HUNGARIANS AND EUROPE IN THE EARLY MIDDLE AGES An Introduction to Early Hungarian History by
a n d r As r o n a -tas
» *» ‘■r C E U <4*
P RESS
Central ^European University Press
First published in Hungarian as A honfoglalo m agyar nep in 1996 by Balassi Kiado, Budapest
English edition 1999 by Central European University Press Oktober 6. utca 12 H-1051 Budapest Hungary 400 West 59th Street N ew York, N Y 10019 USA Translated by Nicholas Bodoczky Maps by DIM AP
© 1996, 1997 by Andras Rona-Tas English translation © 1999 by N icholas B odoczky
Distributed in the UK and Western Europe by Plymbridge Distributors Ltd., Estover Road, Plymouth PL6 7PZ, United Kingdom A ll rights reserved. N o part o f this publication may be reproduced, stored in a retrieval sys tem, or transmitted, in any form or by any means, without the permission o f the Publisher.
ISBN: 978-963-9116-48-1
Library o f Congress Cataloging in Publication Data A CIP catalog record for this book is available upon request
To my parents and grandchildren
CONTENTS
PREFACE.....................................................................................................................................................xiii ACKNOWLEDGEMENTS TRANSLATOR’S NOTES
......................................................................................................................
xix
.........................................................................................................................
xxi
PART ONE
A M E T H O D O L O G IC A L IN TR O D U C TIO N A N D THE SO U R C ES I. INTRODUCTION.............................................................................................................................. 1. Terminology, m ethods
3
...............................................................................................................
3
a) Histoiy, proto-histoiy, ancient h i s t o r y ..............................................................................
3
b) Peoples, ethnic groups, language, ethnic names, culture, n a t i o n ..................................5 2. Chronology and chronological assum ptions a) The geohistorical eras b) The geobotanical zones
..................................................................... 16
...............................................................................................................16 ............................................................................................................17
c) The Stone and Metal A g e s ....................................................................................................... 20 d) Some co-ordinates in world histoiy
...................................................................................22
e) Chronology and c o n tin u a n c e .................................................................................................23 3. The role o f the natural sciences in establishing chronology..........................................
24
a) The determination o f a g e ....................................................................................................... 24 b) The determination o f o r ig in .................................................................................................... 29 c) The determination o f e n v ir o n m e n t...................................................................................... 29 d) The determination of production and te ch n o lo g y ...........................................................30 4. Other ancillary disciplines in establishing c h r o n o lo g y .................................................
30
a) N u m is m a tic s............................................................................................................................... 30 b) Archaeology c) Linguistics
.................................................................................................................................31 ...................................................................................................................................32
N o t e s .....................................................................................................................................................39
viii
Concents
II. THE SOURCES..................................................................................... 43 1. The concept of source material................................................................. 43 2. Source criticism................................................................................ 44 3. The written sources............................................................................... 45 a) Byzantine sources .............................................................................46 b) Latin sources.....................................................................................55 Sources from outside Hungary........................................................... 55 Hungarian sources..........................................................................58 c) Slavonic sources ................................................................................60 The beginnings of Slavonic literacy.................................................... 60 The legends of Cyril and Methodius.................................................... 60 Other early Bolghar accounts........................................................... 61 The Bulghar regal list ..................................................................... 61 Ancient Russian chronicles ......................................................... . 62 d) Middle Iranian sources.......................................................................63 e) Sources by Muslim authors..................................................................64 Sources in Arabic............................................................................ 67 Sources in Persian..........................................................................73 f) Syrian sources.................................................................................... 74 g) Armenian sources..............................................................................76 h) Georgian sources................................................................................79 i) Turkic sources....................................................................................80 j) Tibetan sources.................................................................................. 82 k) Chinese sources ................................................................................85 A brief history of China...................................................................85 Chinese script.................................................................................86 Transcription of Chinese script............................................................88 Chinese historical sources................................................................ 89 I) The Hebrewsources ...........................................................................90 4. The language as a source.........................................................................92 5. The archaeological sources.................................................................. 116 6. Ethnographic sources......................................................................... 140 7. The anthropological sources................................................................ 153 Notes............................................................................................... 164
ix
Concents PART TWO
RELATIVES AND NEIGHBOURS III. THE RELATIVES.................................................................................... 171 1. Linguistic relationship .......................................................................171 2. The Uralic languages and peoples ......................................................... 174 a) The proto-language .......................................................................174 b) The Samoyedic languages and peoples................................................ 177 c) The Ob-Ugrian languages and peoples................................................ 178 d) The Permian languages and peoples .................................................. 179 e) The Volga Finno-Ugrian languages and peoples .....................................181 The Cheremis language and people.................................................. 181 The Mordvin language and people .................................................. 182 f) Extinct Finno-Ugrian languages and peoples......................................... 182 g) The Balto-Finnic languages and peoples ..............................................183 The Finnish language and people.................................................... 183 The Estonian language and people .................................................. 184 Minor Finnic languages and peoples of the Baltic ............................... 184 Notes................................................................................................185 IV. THE NEIGHBOURS ..............................................................................187 1. Early Indo-European languages and peoples ........................................... 187 a) The beginnings of Indo-European languages and the early migrations......... 187 b) The Tocliarian language and people.................................................... 192 c) The Iranian languages and peoples.................................................... 195 d) The Alani.................................................................................... 200 e) The Ostrogoths of the Crimea........................................................... 203 2. The Xiongnu and the Huns.................................................................. 203 3. The early Turkic peoples.......................................................................209 a) The emergence of the Oghur peoples and the Avars................................ 209 b) The Sabirs.................................................................................... 212 c) The Ruanruan and the European Avars................................................213 d) The Turk people in Europe................................................................214 e) The Bulghars................................................................................215 Khuvrat’s Bulghar state................................................................219 The Volga Bulghars.....................................................................220 The Danube Bulghars.................................................................. 227 f) The Khazars.................................................................................. 228 g) The Pechenegs ........... ..................................................................234
X
Concents 4. The Slavonic peoples
............................................................................................................... 239
a) The Southern S la v s ............................................................................................................... 240 b) The Alpine and the Western S la v s................................................................................... 242 c) The Slavs of Pannonia, the Moravians and the Slavonic co n v e rsio n ..............................................................................................243 d) The Eastern Slavs and the Rus o f K iev.............................................................................246 N o t e s ..................................................................................................................................................247 V. EURASIA IN THE 9 th AND 10th CENTURIES......................................................................... 251 1. The end o f the Uighur E m p ir e .............................................................................................. 252 2. The Khitai and C h in a ............................................................................................................... 253 3. The Kharakhanids and the Black Khitai
.............................................................................255
4. The Oghuz and the Seljuk........................................................................................................ 256 5. The Khazars and the R u s ........................................................................................................ 257
6. Byzantium
and the Danube B u lg h a r s ................................................................................ 257
7. Rome and the F r a n k s ...................................................................................................... . 259
8. The Avars and
Slavs in the Carpathian B a sin ......................................................................261
9. The Carpathian Basin on the eve o f the C o n q u est........................................................... 263 N o t e s ..................................................................................................................................................266
PART THREE
FR O M THE URALS TO THE CARPATHIAN BASIN VI. THE NAMES OF THE MAGYARS BEFORE THE FOUNDATION OF STATE ................................................................................................. 271 1. Ethnic names: characteristics and o r i g i n s ......................................................................... 271 2. Turk
.............................................................................................................................................. 275
3. Onoghur, Ungar, Hungarus, H u n g a ria n .............................................................................282 4. Savarti A sfa li................................................................................................................................ 288 5. Bashkir...........................................................................................................................................289
6. M a jg a r ...........................................................................................................................................294 7. M agyar...........................................................................................................................................297
8. Other
names o f the M a g y a rs................................................................................................. 308
a) Scythian....................................................................................................................................308 b) H u n ...........................................................................................................................................309 c)A va r .........................................................................................................................................309 d)
Other n a m e s ................................................................................................................... 309
9. A historical su m m a r y ............................................................................................................... 310 N o t e s ....................................................................................................................................................... 311
Contents VII.
XI
URHEIAMTS AND M IGRATION.................................................................................................315 1. General q u e s tio n s ..................................................................................................................... 315 2. The migrations o f the p ro to -M a g y a r s................................................................................317 3. The migration o f the Magyars from the Urals to the C a r p a th ia n s............................319 N o t e s ................................................................................................................................................. 324
VIII. THE CONQUEST............................................................................................................................ 325 1. The Magyars in the E telk o z.................................................................................................... 325 2. The historical preliminaries o f the C o n q u e s t ..................................................................330 3. The C o n q u e s t ............................................................................................................................ 332 N o t e s ................................................................................................................................................. 338
IX. THE MAGYARS IN THE CARPATHIAN BASIN.....................................................................
339
1. The name o f the Magyar tribal con fed eration ..................................................................340 2. Political o r g a n isa tio n ...............................................................................................................341 3. Social s t r u c tu r e .........................................................................................................................354 4. E conom y.......................................................................................................................................360 5. Religion, lore, c u ltu r e .............................................................................................................. 364 N o t e s ................................................................................................................................................. 370 X. THE INTEGRATION OF THE MAGYARS WITHIN EUROPE.................................................373 1. The integration o f peoples, the types o f ethnic change in the M iddle Ages 2. The third integration o f the Magyars
. . . .
373
................................................................................380
N o t e s .................................................................................................................................................383 XI. SUMMARY OVERVIEW ...............................................................................................................385
PART FOUR
RECENT RESEARCH A N D STU DIES XII. AN OVERVIEW OF THE STUDY OF ANCIENT HUNGARIAN HISTORY..........................................................................................395 N o t e s ................................................................................................................................................. 411
XIII. THE LEVEDI QUESTION AND THE EARLIEST HUNGARIAN CHRONICLE........................................................................................................413 N o t e s ................................................................................................................................................. 421
Contents
xii
XIV. HISTORICAL TRADITIONS, ATTILA AND THE HUNNISH-MAGYAR KINSHIP .................................................. 423 Notes................................................................................................ 427 XV. THE EAST MAGYARS, THE BASHKIRIAN TRIBAL NAMES AND YUGRIA..........................................................................................429 Notes...................................................................................................436 XVI. THE 57ЁКЕЕУ RUNIFORM SCRIPT.................................................................437 Notes.................................................................................................. 444
APPENDICES BIBLIOGRAPHY.............................................................................................. 447 CHART OF RULERS.......................................................................................... 491 CHRONOLOGICAL INDEX .................................................... , .................... 495 INDEX OF PROPER NAMES...............................................................................515 INDEX OF WORDS, ETHNIC AND TRIBAL NAMES........................................... 539 LIST OF MAPS, FIGURES AND PLATES................................................................. 561 SOURCES OF MAPS, FIGURES AND PLATES..................................................... 565
PREFACE
This book is a m odified and extended version o f m y book/i honfoglalo m agyar пёр [The conquering M agyar people] which appeared in 1996. The first Hungarian edition was out o f print within three m onths, the second edition appeared in early 1997, and the third edition is going into print soon. In 1974, som e sixty scholars joined forces in Szeged to lay the foundations o f a new approach to the ancient history o f the Hungarians, or ‘M agyars’, as they call them selves. A five-volum e study appeared as a result, entitled Bevezetes a m agyar ostdrtenet kutatasanak forrasaiba (Introduction to the sources o f Hungarian proto-history), published in Hungarian between 1976 and 1982. The team had very little time to com plete the task, and faced great difficulties. The outcom e had m any shortcom ings, only som e o f which can be m entioned here. The publication lacked overt theoretical foundations, on account o f the fact that the period the study was conceived in was still laden with M arxist ideology. Also, the scholarly standards were uneven, for we were unable to win or find the m ost suitable scholars to write certain chapters. Some o f the chapters were not com pleted in time and had to be covered from second-hand literature. The research group received no financial support whatsoever, and the volumes were eventually published as university text books. Nevertheless, despite all o f its shortcom ings, the w ork was a great success. The focusing o f the volum es o f the Szeged Proto-H istory R esearch Group (M agyar O stdrteneti M unkakozosseg) on source m aterials was not a m atter o f chance. It was justified to broaden the concept o f “source m aterial” and to encom pass all data that would contribute to historical reconstruction. N ot only the w ritten sources, but also all linguistic, archaeological and anthropological data had to be considered— the latter with the same rigour as the written sources. We ourselves were astonished by the fact that in the past decades no new sources had been discovered, and the critical evaluation o f the known sources had long since been a neglected field. Instead o f new critical editions, the literature tended to com bine, over and over again, the existing data, often cited inaccurately. A few o f these com binations were truly ingenious, but m ost
xiv
Preface
o f them were unprofessional. Therefore, it becam e very im portant to collect, select and introduce new sources. It was also clear that the study o f these sources had to regain its position in the curricula o f the universities. For decades no scholars had been trained in this very difficult field o f research. This accounted for the fact, am ong others, that the w ork was published as a university textbook. I m yself began to teach early history and its connected fields in 1974 at the Attila Jozsef University o f Sciences in Szeged. In m any respects this book was a sum m ary o f my lectures. It was, however, addressed to a broader readership, to university students, to teachers o f history and to intellectuals interested in the early history o f the Eurasiatic continent. The first part con centrates on m ethodology and the sources. One o f the m ost difficult issues regarding this type o f history is that the assessm ent o f the different sources requires different competencies. Evidently, no single scholar has equal com petence in all o f the different fields. Therefore, the solution to the problem is collaboration. Collaboration, however, m ust follow a few basic princi ples. Firstly, the collaborators m ust understand the critical m ethods and basic constraints o f each others’ fields o f research. Secondly, the few er hypotheses a scientific claim requires, the greater probability and acceptability it has. In other words, the fewer “asterisked” hypotheses needed, the m ore probable the result is. O f course, the use o f a limited num ber o f w ell-defined hypotheses is perm issible, but one should not confuse hypotheses w ith facts, and one should certainly never take for granted a hypothetical claim o f one related discipline in another related discipline and use that claim to construct another hypothesis. The third principle is that, although one should avoid infertile hyper-criticism , severity o f criticism is m ost needed in those fields where the scantiness o f the sources could prom pt one to take advantage o f obscurity. It is better, perhaps, to confess where present-day knowledge ends, and state openly if one cannot be sure. Open or unsolved questions can help, not hinder, the developm ent o f scholarship. Finally, as a fourth principle, it m ust be accepted that all related disciplines have their own m ethodologies w hich m ust be applied with full severity, w ithout giving in for the other related disciplines’ sake. Besides drawing on the research and results o f the m em bers o f the Szeged Proto-H istory Research Group, this book also relies on the w ork o f other Hungarian and foreign scholars. This is self-evident— the reason w hy I nev ertheless have to point it out here is because, lim ited by its genre, this book cannot give an overview o f the state o f the art, and does not w ish to argue with other views. I have intentionally neglected to m ention authors by nam e in the m ain body o f the text, for the reason that in various scientific articles and essays, i.e. in works o f a different genre, I have already presented our debates, as well as the assessm ent o f their opinions and argum ents— or intend to do so
Preface
XV
in future. I am fully aw are o f opinions differing from my own, and the text should certainly m ake that clear. I believe that non-H ungarian readers would find m ost o f such discussion uninteresting. However, in the reference sec tions after each chapter I have given a b rief overview o f the literature (more elaborate and different to the same parts in the H ungarian version). Those who are interested in the last few decades o f research in H ungary should read Chapter XII. M y approach to the research o f the early history o f the M agyars and Eurasia differs from similar, previous undertakings. Firstly, I avoid the term “prehis tory” where possible. I give m y reasons on p. 4. I have taken a different approach from m y colleagues to the Hungarian chronicles w ritten in Latin. The earliest Hungarian chronicle that has com e down to us is a work written in the last years o f the 12 th century, m ost probably a few years after the Hungarian King Bela I ll’s death in 1196. The anonym ous author o f this work (called ‘A nonym us’), and the authors o f the later chronicles, for instance Simon de K eza or the author o f the Chronicon pictum (Illum inated Chronicle) alm ost certainly drew on earlier works. Nevertheless, these chronicles— the texts o f which have come down to us only in later copies— contain the early history o f the M agyars as it w as conceived 300 years after the Conquest. The critical study o f the Hungarian chronicles has been undertaken in Hungary, with good results. Nevertheless, in discussing the conquering M agyars and the history o f the M agyar Conquest, I have taken a new approach vis-a-vis the accepted— yet hardly regular— practice. I find it inadm issible to accept as genuine that portion o f a given chronicle w hich fits the concept in mind. This does not mean that chronicles are useless as sources. A t a closer look, they can reveal im portant elem ents o f truth. It is the procedure, in m y view, that should be reversed. An attem pt should be made to reconstruct history w ith the help o f contem porary sources, or with the help o f those w hich are nearest to the events. This approach gives a more m eagre history, which then can be collated with the chronicles. The truly authentic portions o f the later chronicles will be those that are identical with the reconstruction. Those which, at best, do not run counter to it will rem ain food for thought. H ence the fact that the reader will not find anything in this book which entered historiography from the chronicles only. It fails to consider m any picturesque parts o f early Hungarian history recounted in some Hungarian and non-H ungarian hand books. It is one o f ancient Hungarian history’s peculiarities that it has m any un w ritten sources, as well as a great m any sources written in a num ber o f different languages. The research o f language history and o f the sources has developed at a great pace in international science, but not, alas, in m any fields in Hungary. Due to this, long-obsolete opinions, provably incorrect data, and
XV I
Preface
results achieved by m ethods unacceptable today are currently in circulation in Hungary. In Hungary m uch o f the investigations o f the M agyars’ early history have been conducted by linguists. This was and is, as we shall see, unavoidable. We cannot fail to consider linguistic facts, such as proper nam es, even in periods for which we do have written sources. However, the role o f linguistics in the reconstruction o f the M agyar C onquest has often been exaggerated in the past. The “furor etym ologicus” has ham pered research heavily. Even if the Turkic etym ology o f one or another nam e o f a tribe turns out to be correct, it does not m ean that the tribesm en necessarily spoke a Turkic language. We m ust not lose sight o f the evidence revealed by the study o f proper nouns— that is, the nam es o f peoples and tribes, personal and place names. Provided their reconstruction is correct, they are linguistic facts. In historical reconstruction, however, they can only be used within limits. This is prim arily because proper nam es do not have a literal m eaning, but only a denotative function, w hich is why their etym ology is rather vague. The fact that some kings o f the House o f Аграф&еге called Andras (Andreas, Andrew) does not im ply that the House o f A rpad had a Greek origin, not even if the m others o f these kings were Greek. However, the Greek origin o f the name can be safely established. Yet it remains to be answered how it infiltrated into the Hungarian language, and how the nam e that originally m eant ‘m an’ be came a personal name also in Hungarian. Even very early on, popular etym ology was at w ork in m any proper names. For instance, Isidorus o f Sevilla (565-636) has hundreds o f these, not to m en tion the Hungarian Anonym us who, like in m any other things, succum bed to the current fashions. People simply cannot accept that proper nam es have a deictic function only, and will, therefore, assign a “proper” m eaning to truly m eaningless proper names. We shall see some clear exam ples later. Naturally, any language is in itself a veritable treasure trove o f the history o f those professing that language. But still, one m ust rem ain strictly critical. The Hungarian words terd ‘knee’, kar ‘arm ’, and gyom or ‘stom ach’ are o f Turkic origin, yet it does not m ean that the M agyars were laden with physical defects before they encountered the Turks. In treating the linguistic sources I have sought to keep am biguous statem ents at bay— as in all o f the chapters. Hence, the reader will miss m any linguistic explanations and rationale cur rently in circulation. I have included in the book a m ethodological introduction to the use o f linguistic data for purposes o f historical reconstruction. This is more detailed than the treatm ent o f the m ethodology for using the other source types. Unfortunately, at times, not only linguistics, but archaeology also trans gresses its field o f authority. In Hungarian cem eteries, from the 1960s, cre mation becam e more com m on than interm ent. This, com bined w ith the oc
Preface
xvii
cupation o f H ungary in 1945 by the Red Army, could easily lead one to the assum ption that the ethnic com position o f the H ungarians changed, and a m ajor new population settled in Hungary. The fact is, that no such change occurred. A rchaeological sources in them selves cannot function as evidence o f ethnohistoric change. Very often, erroneously, archaeological cultures have been identified with ethnic groups. A t the same time, archaeology is a dis cipline which operates with an ever-increasing num ber o f finds and always with new acquisitions, consequently all new archaeological facts require clar ification. For this reason, I have devoted special attention to the m ethods o f establishing chronology in archaeology, as well as to other m ethodological issues related to archaeological finds. However, I have neglected to discuss m any archaeological cultures whose ethnic identification I could not deem guaranteed. In this book, I considered the ethnic group as the subject o f historical changes. In the m ethodological chapter I sought to m ake clear m y concept o f ethnicity and ethnic history. I em phasised that it was never the origin, but rather, the historical change o f an ethnic group that was relevant to history. In this respect I argue with the ethnographic approach taken by the earlier Hungarian schools o f cultural anthropology. Descriptions offered by ethnology or comparative cultural anthropology can be used as analogies, in certain cases. I have m ade good use o f m y earlier field-w ork am ong the nomads o f Central Asia, especially o f the ethnography o f the M ongols. However, the use o f analogies is restricted. Some custom s can change within a very short period o f time, while others m ay remain unchanged for hundreds o f years. I used such analogies w ith due caution and only w hen it was unavoidable. Some o f the Hungarian handbooks on early M agyar history do not give account o f the broader context o f the history o f Eastern and Central Europe in the early M iddle Ages. Yet m any o f the handbooks on the early M iddle Ages were unable to include the Turks and the Hungarians in general history. Although the focus o f my study is the early history o f the M agyars, I have sought to place their history in the context o f their contem porary global history. 1 have dealt with many details o f the topics treated in a m ore general way in this book in several o f my papers. I shall refer to them only in the notes or the general bibliography. M any o f them were published in languages other than Hungarian, and are thus accessible to the non-H ungarian reader. As this book is being published, I am writing a m onograph discussing Turkic loan words in the Hungarian language, the new results o f w hich I have also used in this book.
xviii
Preface
I felt I had to discuss some o f the topics referred to in this book in greater detail. These I placed in Chapters X III-X V I. I also had to introduce som e con ventions. I have distinguished the Turkic-speaking B ulghars from the Sla vonic-speaking Bolghars by the two variants o f the same nam e, although I m ight have done it the other way round. I have to dwell here upon a seemingly term inological question. In H ungar ian historiography the events betw een 885 and 902 are referred to as the honfoglalas, which is a loan-translation o f (but not entirely equivalent to) the G erm an Landnahme. In the English-language literature these sam e events are in m ost cases described as the Conquest— although, o f course, the historical connotations o f the acts o f W illiam the Conqueror in 1066 and those o f the M agyars in the Carpathian Basin are not the same. The H ungarian term honfoglalas literally m eans ‘the settlem ent in the fatherland’, and is a term fully devoid o f connotations suggesting ‘taking by force’. Som e scholars w riting in English have used the term “original settlem ent” , or sim ply “ settle m ent”. However, ‘settlem ent’ has connotations which include the ‘settlem ent o f a nom adic people’, ‘to becom e settled in villages’, etc. which, as w e shall see, was not exactly the case. It has also been suggested to retain the Hungarian word (as, for instance, the Germ an word Urheimat), in m odern English. This w ould have m ade the text very aw kw ard to read. Having weighed up the pros and cons I decided to use the accepted term inology o f the B ritish and A m erican literature— with the caveat that the term inology is never a perfect reflection o f the actual historical events. However, to distinguish it from other conquests, I have spelt it w ith a capital ‘C ’. Finally, I would like to stress that I do not consider this book as a scholar ly m onograph. Its aim is to inform rather than to argue. Som e o f the new claim s in this book are dealt with in greater detail in Chapters X III-X V I. The educated European and A m erican reader m ay be interested who these M agyars are, where they came from, and last but not least, w here our know l edge about the form ation and early history o f the M agyars com es from. I sin cerely hope that this knowledge will find its w ay to textbooks and encyclo paedias w hich presently contain m uch obsolete inform ation about the topics dealt w ith in this book.
ACKNOWLEDGEMENTS
I would like to express how greatly indebted I am to m y professors for in troducing me to this elaborate field o f science. I learnt the basics o f Turkology and Hungarian prehistoric research, as well as the m ethodology, from Julius (Gyula) Nemeth. The fact that hardly any o f his results have w ithstood the test o f time is, paradoxically, a result o f his very own w ork as a scholar and teacher. His studying the form ation o f the conquering M agyars, and not their origins was decisive, as he thus established the basis for future attempts. Louis (Lajos) Ligeti greatly expanded the pool o f sources, while at the same time introducing significantly m ore severe critical methods. From Istvan Talasi, I received the basics in the application o f ethnography in his torical studies. Gyula Ortutay introduced me to the field o f com parative universal folkloristics. O f m y m any colleagues, whose scientific w ork and personal assistance was decisive in the course o f m y work, I shall only m ention those who contributed greatly to this book. Their nam es, in alphabetical order, are: Lorand Benko, Istvan Bona, Ildiko Ecsedy, Gyorgy Gyorffy, Peter Hajdu, Janos Harm atta, Gyula Kristo and G yula Laszlo. The bibliography o f this book attests to the fact that it was not merely courtesy that persuaded me to m ention their names. I w ould like to rem em ber Jeno Sziics who, in m any respects, effected a turn in the research o f ancient Hungarian history, and whose theoretical w ork is considered significant to this day (see pp. 396-397). Form er students o f mine, currently colleagues, also contributed to this work. Eva Csaki, Eszter Lenart and Janos Sipos gave im m ense help in editing and formatting. Being the first educated readers o f the book, they m ade many useful com m ents. Istvan Zim onyi has com piled a new translation o f his se lection o f M uslim authors. I thank him for m aking available his then un published m anuscript. Since then, these texts, translated by Zim onyi and others, have been published in a book edited by Gyula Kristo, containing the w ritten sources o f the conquest period (Kristo, 1995d). Klara A gyagasi helped me find m y bearings in contem porary Slavistic literature.
XX
Acknowledgements
A rpad Berta and Istvan V asary very kindly agreed to read this book from the expert’s viewpoint. I would like to take the opportunity to thank them for their invaluable work and their opinions and com m ents which inspired recon sideration. M any o f these I was able to incorporate in the book, but som e o f them I did not adopt— for lack o f space, or because they had not convinced me. I have learnt m uch from the discussions that follow ed the publication o f m y Hungarian book. Two w orkshops w ere dedicated to this w ork, one at the H ungarian A cadem y o f Sciences and the other at the A ttila Jo zsef University, Szeged. Some o f m y colleagues gave m e their contributions also in a w ritten form, am ong them Istvan Bona and Sam u Szadeczky-K ardoss. I owe sincere thanks to all o f them and I have done m y best to correct som e m istakes and inaccuracies m entioned by them. Ferenc M akk w rote a lengthy book review (М акк 1997) which gave stimulus for rethinking certain issues. M y answ er (Rona-Tas 1998e) and his rejoinder (M akk 1998) started a discussion. The preparation o f the English version was in fact a re-w riting o f the original H ungarian, since the expected readers and their background are different from the Hungarians. The English edition offered a possibility to correct som e o f the m istakes which unavoidably crept into the first two editions. Since 1995 I reached some new results, especially in the study o f the history o f the Khitans and the Bulghars w hich I included in the English version. I added a few new subchapters as well. All this was much helped by the R esearch Prize aw ard ed by the H um boldt Foundation in 1995. I offer m y sincere thanks to the G eorg-A ugust University, Gottingen, to m y host Professor Dr. K laus Rohrborn and to the H um boldt Foundation for their generosity. I w ould like to offer m y sincere thanks to Nicholas Bodoczky, the translator o f this book, with w hom we w orked long m onths together. The help o f Peter Tam asi and his team was indispensable in the extrem ely difficult w ork o f editing the text and the indexes. I was greatly inspired by the fact that the Szeged Proto-H istory Research Group, w hich cam e to life in the process o f creating this work, was supported by Dom okos Kosary, the then President o f the H ungarian A cadem y o f Sci ences, by Laszlo Keviczky, Secretary G eneral o f the H ungarian A cadem y o f Sciences, and by Rezso M eszaros, the R ector o f the A ttila Jozsef U niversity o f Sciences. It is no fault o f theirs, and neither is it a question o f a lack o f funds, that the research team was to set out in early 1996 w ith one m em ber only. M ay the nam es o f those who, during the past decades, m ade considerable efforts to prevent this w ork from com ing into being, be conveniently shrouded in obscurity. Szeged, D ecem ber 1998 The author
TRANSLATOR'S NOTES
All foreign words are transcribed into English, with the exception o f those which were originally w ritten in a Latin alphabet. In alm ost all cases, the usage found in the English literature has been adopted. Sounds usually transcribed in the scholarly literature as c, s, z are transcribed here as ch, sh, zh respectively. The affricate j is rendered by a sim ple j. All Chinese nam es, titles and words are transcribed according to the socalled pinyin system which is the official Latin transcription in China. W here necessary, the transliteration o f the original (non-Chinese) nam es or words are included, in parentheses. The bibliography includes the scientific transliteration o f the nam es written in non-Latin alphabets, and the index includes the scientific transliteration o f the proper nam es where necessary. The Hungarian nam es in the book, such as Alm os, Termecsii, Etelkoz, etc., are spelt according to the m odern Hungarian orthography. F or readers unfa m iliar with Hungarian pronunciation, here is a list o f Hungarian sounds with their transcription according to the International Phonetic A lphabet (IPA):
Consonants The consonants b, d, f, g, h, к, I, m, n, p, t, v and z are pronounced as in English, but unaspirated; r has an Italian-like trill. Hungarian consonant
с cs dz dzs
gy j
ly
Transcription
ts
English (or Germ an) equivalent
tX dz
gets chip lends
d3 dj j j
M duke yes yes
xxii
Translator's notes
Hungarian consonant ny s sz ty zs
Transcription jl J s tj 3
English (or Germ an) equivalent mw she sit /um our decision
Vowels a a e ё i i о o
a a: e e: i i: о o:
6
0
6
0:
u u u u
u u: у y:
hot park bed nam e sit see corn core Germ an Bohm en G erm an schon hook soon G erm an wi/rde G erm an kzVhn
PART ONE
A M E T H O D O L O G IC A L IN T R O D U C T IO N A N D THE SO URCES
1. INTRODUCTION
1. TERMINOLOGY, METHODS
a) History, proto-histoiy, ancient histoiy History is the chronological existence o f human societies. Historical science describes social existence above all— but not exclusively— from that view point. The existence o f human societies and their groups is continuous, which does not imply, however, that the history o f a com m unity lacks im portant or less im portant events, or that certain com m unities do not em erge, prosper and decline, or even disappear. Yet the period nam es used in historical sci ence— such as ‘antiquity’, ‘the M iddle A ges’ or ‘the N eolithic A g e’— are cat egories which serve description, understanding, education and com m uni cation between historians, rather than term s adequately reflecting real changes in history. The prim ary role o f historical science is to reconstruct the stories o f socie ties. Naturally, m yriads o f points o f view and m ethods are available in this reconstruction process which will nevertheless always bring about abstraction and the highlighting o f significant elem ents and relations, or those held to be significant. M oreover, reconstruction is always reduction too, for it does not describe every single event, but m erely some o f them. In the course o f reconstruction, historical science nevertheless seeks to allow for as many conclusions as possible regarding things that it does not and cannot de scribe— deductible from the few things that it can describe. A great m any viewpoints and methods are available in the reconstruction o f history, but because it is im possible to sim ultaneously take contradicting viewpoints into consideration (only com plem enting ones), it is expected o f a historian to m ake his own viewpoints or system o f viewpoints clear in the introduction o f his work. This book focuses on a certain category o f people, to be defined below. Above all, it examines how that people was formed. It wishes to concentrate research on the formation and transform ation o f the people, not on origin. To this end, it draws on all the data contained in the sources, and if it becomes unavoidable, it m akes up for the lack o f sources with close analogies. It regards the form ation o f the people as a lengthy his torical process, and examines it in the very context o f w orld history which the form ation o f that people was itself part of. In the case o f the Hungarians,
4
Methodological introduction and the sources
this m eans, prim arily European, m ore specifically C entral and Eastern E uro pean history; one cannot, however, disregard the Eurasian steppe or the non-European parts o f the M editerranean, owing to the fact that the form er saw m ajor m igrations, and the latter the expansion o f the M uslim world. Conventionally, history is divided into two large periods. The first period covers the tim e from w hich no w ritten sources have com e dow n to us, hence it is only reconstructible with the help o f indirect— non-w ritten— sources. This period is conventionally called proto-history. However, the only argu m ents to support this bipartite division o f history seem to be convention and practicality. The historical processes them selves are unlikely to change m erely because records are m ade o f them , or o f certain elem ents o f them. Further m ore, classifying a phase in the history o f a people as proto-history on the grounds that no written records o f it exist, will not change the events o f the past even if such w ritten sources are later discovered. Naturally, our know l edge about that phase will change, but then, the discovery o f non-w ritten sources can also change our knowledge. The reason w hy this needs to be stressed is because the technical term s “proto-history” (in H ungarian: ostdrtenet) and “Hungarian proto-history” have gained ground in H ungarian histori cal science, and even outside the bounds o f the science. For w ant o f better expressions, I m y self tend to use them , but they are hardly adequate. H ungar ian proto-history is generally held to have ended in the 9th century AD— at a tim e w hen other, neighbouring peoples had long since entered their period o f w ritten history. The use o f technical term s is not, o f course, especially im portant. Still, I agree with those who prefer to say ‘ancient H ungarian h is to ry ’ or ‘early Hungarian history’, rather than ‘H ungarian proto-history’ or ‘Hungarian prehistory’. The tw o questions regarding ancient H ungarian history are, then: w hen did it begin and w hen did it end? One can now give an answer, bearing in m ind that this classification is, in effect, auxiliary rather than rudim entary. The closest answers we can get are by clarifying w hose history w e are talking about, that is, who the subject o f this historical process is. B ut to set some kind o f tim e lim it for our discussions before focusing on the details, we should establish at the outset that the history o f the M agyars began w hen they becam e a people proper. A nything before that will be considered the prehistory o f the M agyars. As the following chapter will point out, that is not a simple question either. Several events can m ark the end o f the ancient history o f the M agyars, one o f them being the conquest o f the C arpathian Basin, the other the founding o f the Hungarian state. Clearly, the elem ents o f statehood date back to pre-C onquest times, and the C onquest w as a m ajor step tow ards the founding o f a state. It is also clear that, as regards the founding o f state, the coronation o f Saint Stephen was, above all, a sym bolical and legal act and an
5
Introduction
outstanding event in setting up the state, but naturally the developm ent o f statehood did not stop there. A fter explicit and m anifold deliberation, one can establish that ancient Hungarian history lasted from the M agyars becom ing a people proper to the founding o f the Hungarian state, the date o f which is associated with the sym bolic and legal event o f the coronation o f Saint Stephen in 1000 A D , in fact on the first o f January, 1001.
b) Peoples, ethnic groups, language, ethnic names, culture, nation The use o f various concepts and technical term s requires great care. We use words to denote concepts and technical terms: and those words have m eaning and usage. Pertaining to the signifier (i.e. the word), m eaning is, in effect, none other than the com plex entirety o f the rules that describe how and to w hat purpose a word can be employed. This is im portant to establish, because the sam e w ord m ay be used differently according to the period and the context. Published in 1972, the D ictionary o f the H ungarian Language provides five meanings o f the word пёр ‘people’, with examples in all cases. The exam ple quoted to illustrate the fourth meaning, ‘a large group o f people’, is hadi пёр ‘arm ed group’, ‘arm y’. This particular use o f the fourth m eaning o f the word is now obsolete in the Hungarian language, yet it lives on in com pounds like nepbetegseg ‘pandem ic disease’. One o f the m ost com m on m istakes in the use o f technical term s is to project the m odem m eaning and the rules o f usage o f a word back to a period when the concept denoted by that word either did not exist yet, or both w ord and concept existed, but the word had a different m eaning and different rules o f usage. Unfortunately, no real distinction is m ade betw een the technical term s people and nation. A ssum ing that the concepts denoted by ‘people’ and ‘n ation’ w ere identical in the 11th century, for instance, can bring about m ajor m isapprehension. Actually, what we call people today w as different then, and w hat w e call nation today did not exist. Social groups, as w ell as the nam e o f those groups, have a history o f their own, yet the two are by no m eans necessarily identical. Next, let us exam ine the various uses o f the word пёр ‘peo p le’ in H ungarian com pounds currently in use. Here are some examples: if a statistician takes a population census (nepszam lalas), he m ust provide him self a definition o f w hat he is about to count. If a literary scholar sets out to study Hungarian populist writers (nepi irok), he is focusing on som ething quite different than the expert o f econom ic geography who will delve into the issues o f depopu lation {elnёptelenedёs). W hen a historian studies the m igration o f peoples
6
Methodological introduction and the sources
(nepvcindorlds), he has a quite different notion o f people than a psychologist who concerns him self with the problem s o f popularity (nepszeruseg). But an ethnographer (neprajzos ) is also looking at som ething different, depending w hether he opts for the traditional approach, or takes up a m odern one. In the form er case he will exam ine the m edia o f peasant culture, while in the latter he will m ake his subject com m unities o f any profession, definable by com m on custom s and culture. Taken from the Hungarian language, the above exam ples attest to the fact that even the various sciences tend to use term s differently. And as one shifts from the m ore rigorous sciences to the realm o f everyday language, the m eaning o f technical expressions really fades, and the bounda ries o f usage rules dissolve. W hat this suggests is that such term s cannot be applied, or they cannot be used to develop theories. The m ajority o f scientific debates are a hopeless drift o f words, due to the lack o f consensus regarding the concepts and the words denoting concepts. There are two ways to resolve the situation. Either som e kind o f definition is created w hich the users o f the expression agree on: here and now, in this book, in this science, or at this m eeting we define, for instance, the word people thus. Or, alternatively, one m ay attem pt to avoid polysem ous words, and introduce a scientific term. It is undoubtedly m ore convenient to introduce the Greek word ethnos, although virtually disregarding w hat the w ord once m eant in A thens, or how the Greeks them selves used it in the tim e o f A lexander the Great. W hat justifies the latter procedure is that our conclusions will be influenced less by the use o f a Greek word than the history, m eaning and usage o f a proposed word o f a present-day language. However, nothing exempts us from defining the m eaning we denote ethnos. And once we have defined it, we are free to affirm that we shall be using the word people the same way in this context. The term ethnos and the word people — as a technical term— will hereafter be considered identical. The first will be used in a theoretical context, the other in the narrative sections. Ethnos, or, in context, people, w ill mean a historically evolved group with a common semiotic system, whose members consciously distinguish them selves from other ethnic entities, and which possesses a permanent self-des ignation. O f all sem iotic systems, language is the m ost im portant. The reason why ethnos is not defined as being a group with a com m on language (am ong others) is not m erely because people and language m ust be distinguished from one another. We shall consider culture, too, as a sem iotic system: a system o f signs and sym bols which the m em bers o f the given group use to com m unicate and to identify each other. Also, there are some exceptional cases when ethnic identity does not presuppose linguistic identity. The sem iotic system may include costum e which, for instance, reveals not only ethnic belonging, but
Introduction
7
also marital and social status. Costum e is m ore than dress/clothing in that it forms part o f such systems. Sem iotic systems also include burial customs, greetings, etc. It is easier to see that identical language does not necessarily m ean identi cal ethnos, although that is preponderantly the case. The A m ericans and the English, the Spanish-speaking South A m erican Indians and the Spaniards o f Spain, or the Germ ans o f Switzerland and o f Germ any are evidently different ethnic entities. W hen several different groups belong to the same ethnos, their different languages notw ithstanding, the situation is more com plicated. There are two types o f such cases. In the first case, a people shifts from one language to another, retaining its ethnic identity. The French, for instance, fall into this category. The Celtic and Germ anic locals were subdued by a governing stratum o f Rom an origin. The m ajority then adopted the language o f the minority. N ot only did the biological com m unity rem ain identical, but also m any elem ents o f custom and culture. The change o f one portion o f a semiotic system does not necessarily entail the change o f its other parts. M aking its way towards the east from the Iberian Peninsula, the Jewish com m unity reached the Rhine region in the 10th century. Here it abandoned its native Hebrew tongue to the German dialect characteristic o f that region, which naturally features Hebrew elem ents, some Old French picked up during the journey, as well as some Slavonic from later on. This Y iddish-speaking com m unity retained its ethnic identity despite the change o f language. If the shift o f language takes place slowly, in three or four generations, for instance, then the ethnic group can “translocate” , through the language shift, all those elements o f its form erly com m on system o f signs which were not strictly bound by language, and can thus preserve its name, ethnic identity, and its isolation from the others. In many cases the language shift is follow ed by ethnic change, at a slower pace. The Cum anians o f H ungary becam e part o f the Hungarian ethnos much slower than they lost their Turkic language. Until very recent times, they have formed an independent ethnic group within the H ungarian ethnos. Inhabitants o f the M ongolian Empire did not attribute significance to the fact that they spoke Turkic and not M ongolian; not only did they call them selves Tatars, but also considered their group to have had M ongolian ancestors in their legends o f origin. As a prerequisite, they were, to a certain extent, bilingual. This type o f bilingualism was m ore com m on in the past than it is today. Political loyalty or political m im icry m ust have brought about such alterations o f consciousness which, despite the use o f a different language, resulted in an identical sem iotic system m anifesting itself in com m on self-
8
Methodological introduction and the sources
confinem ent and self-designation. Such cases are, however, rare and hardly, if ever, lasting. Am ong the im portant characteristics o f ethnos or people is the fact that it has a com m on ‘w e-consciousness’ which sets it apart from the other com m u nities o f a sim ilar order. These w e-consciousnesses are, naturally, interm in gled or constructed on one another, and hence one m ay share the w e-con sciousness o f a ham let, a village, the land, the country or the continent. W ith ethnic groups, however, this always com es with an ethnic nam e, a self-desig nation which is a perm anent characteristic o f that community. Perm anence is very im portant, because groups, m oreover ethnic groups, can change their self-designation. There are three types o f Uighurs living in C hina today. The so-called Yellow Uighurs, the descendants o f the ancient U ighur people, live in the Chinese province o f Gansu. One group o f theirs preserved its old Turkic language, another group becam e M ongolianised and currently speaks m odem M ongolian. The large population o f Uighurs living prim arily in C hina and also, in few er num bers, in Russian Turkestan adopted this ethnic nam e as a result o f a decision m ade by the leading m em bers o f the local literati. The nam e o f the Turkic-speaking south Siberian O yrots was changed to A ltai by a decree. Identical ethnic nam e does not necessarily m ean identical ethnic content. This point, and ethnic nam es in general, will be discussed below. Although the m ost im portant properties o f ethnos are a com m on sem iotic system , com m on w e-consciousness and a perm anent self-designation, there are a great m any other traits which appear together w ith groups definable as ethnos. These include, am ong others, com m on origin, or the consciousness o f com m on descent. It m ust be stressed that the consciousness o f com mon descent, i.e. the belief that the com m unity has a com m on forebear, very rarely reflects actual facts. It originates from the “enlargem ent” o f the fam ily as a key social group. Like a fam ily (the paternal line) is a descendant o f one ancestor, likewise com m on descent plays a role in the identity o f the larger units, in for instance the clans, tribes, or the people as a whole. But the larger and m ore com plex the unit is, the vaguer the identity o f actual biological descent. However, “in return” the consciousness o f descent becom es m ore significant, and this has a legitim ising role, i.e. it contributes to com m unity bonding and the establishm ent o f com m unity rights. The consciousness o f com m on descent is alw ays artificial in the sense that it is propagated by the authoritative m em bers o f the given community. It is fashioned and passed down w ithin the com m unity by certain groups responsible for sustaining and m ediating culture: sham ans, priests, chroniclers, bards, i.e. the “m edium s” o f ancient societies. W hile com m on descent is not, generally speaking, genuine, the existence, operation and role
Introduction
9
o f descent consciousness is very m uch genuine; moreover, it plays an im por tant role in the developm ent o f ethnic identity. The “ forebears” o f com m on descent are real or m ythical persons, but are often anim als or even plants. The term we use in the latter case is totem . In the history o f ethnos, instances o f tracing back to several different com m on ancestors also often occurs, due to the fact that the logic— or, to borrow Karoly M arot’s term , the pre-logical reasoning— that prevails in these processes o f consciousness is rather unique or, one m ight say, m ythological. This accounts for the fact that the leading strata o f the conquering M agyars believed that they were o f A ttila’s progeny, while they concurrently held that the forebear o f the ruling M agyar dynasty was a haw k (see C hapter XIV). As it will be pointed out, the descent from Attila is not a historical fact, but rather, a legitim ising “reference” . Prior to the time o f Chingis Khan, the leading strata o f a num ber o f nom adic steppe peoples believed they were descended from Attila, ju st as Christian peoples in the M iddle Ages were convinced they were the progeny o f the sons or grandsons o f Noah while, at the same time, their very own rulers held they were the offspring o f Caesar or Charlem agne. The consciousness o f com m on descent o f different peoples, however, greatly varied during the ages— partly depending on the characteristics o f their given cultures. And until the end o f the M iddle Ages, the b elief in a com m on ancestor was not a prerequisite o f ethnic identity, but rather, a frequent con comitant. Common land plays an im portant role in the continuity o f an ethnos. Evidently, if one part o f the group m oves away, or the com m unication between certain groups o f the ethnos is hindered by natural obstacles (such as a large river, or insuperable m ountains and forests), the com m on culture and lan guage, that is, the sem iotic system , will shatter. Any innovation conceived in one place will not spread over to the isolated group. This process is not irreversible, because the groups can m aintain connections and “keep in touch” for some time. This m ust have happened to the various segregated groups o f M agyars prior to the Conquest. Com m on land is not then a prerequisite o f a com m on ethnos, but if the com m on land ceases to exist, the splintered groups o f that ethnos will peel o ff to develop as separate ethnic entities, and eventually differences will outnum ber similarities. A common political organisation can also play a role in m aintaining the unity o f an ethnos, w hether it is a tribal confederation, a nom adic em pire, or a prim aeval state. A political organisation’s chief significance can be in form ing the com m on w e-consciousness. Underlying the endless fights is always the “them ” versus “us” opposition. The fights for hunting and grazing grounds, for trade and levying taxes called for the constant reinforcem ent— or often alteration— o f the consciousness o f group identity. At the sam e time, belong
10
M ethodological introduction and the sources
ing to one ethnos is not identical, or is not necessarily identical, with belong ing to the sam e political organisation. N om adic political organisations were powerful forces in fostering ethnic identity: ethnic m im icry can be well observed here, as for instance, in the M ongolian Em pire. The pow er o f these nom adic establishm ents, or O riental em pires, lay in the fact that they would, from time to time, allow ethnic separation. M aking fairly steady headw ay under unchanged circum stances in the long term , the otherw ise extrem ely slow process o f ethnic assim ilation can be achieved, as a m atter o f fact, by the toleration o f separation. China or the Ottom an Em pire are good exam ples here. In the 16th and 17th centuries, the dem arcation o f the Jew ry living under the O ttom an Turkish Em pire was much better tolerated than in m any European Christian states. The Nobel Prize w inning w riter Elias Canetti gives a very vivid account o f this. The status o f the M agyars w ithin the K hazar Em pire m ust have been especially interesting in this respect, because despite the fact that the M agyars m aintained m anifold ties with the culture o f the K hazar Empire, they eventually succeeded in preserving their ethnic identity. Political organisations have yet another role regarding ethnic groups. The “us” versus “them ” opposition dem ands “borders” o f som e description. “That is w here our area ends”, “this is up to where we dw ell” , “this is how far ‘our land’ extends”. These borders are very diverse, and projecting the notion o f m odem frontier lines back to ancient tim es w ould be a great m istake. However, the “m anagem ent” o f geographical space is one o f the functions o f political organisations. Political organisations play a very im portant role in separation. The “m anagem ent” o f geographical space, and the staking out o f the im portant points o f space not only occurred at the perpetually debated points o f the borders, but also in the choice o f the “holy places” o f an ethnos and in reinforcing its cult. The holy m ountain o f the Turks, the Otiiken (in the Hangai m ountain range o f the Republic o f M ongolia) assum ed a very im portant role in the identity o f the Turkic people betw een the 5th and 7th centuries, as at tested by early- 8 th-century ancient Turkic inscriptions. The real or presum ed grave o f the founder o f an em pire or religion can also becom e an im portant com m unity sym bol, an intrinsic part o f identity. The grave o f Attila, Genghis Khan, or M uham m ad can serve concurrently as a sym bol o f religious and com m unity identity. Naturally, religion or religious beliefs can play a part in the identity o f an ethnos. The com m on identity o f people who w orship the sam e divinities or god is also distinguished by the fact that another com m unity adores other gods and perform s different rites. Having ascertained that it was not religious difference that m ade the m em bers o f one ethnos form a separate ethnos, likewise it is easy to see that the m em bers o f different religions can nonetheless belong to one ethnos, on account o f the identicalness o f their other constituent
Introduction
11
sem iotic systems. Protestantism and C atholicism did not bring about ethnic segregation am ong Hungarians, yet the situation with the O rthodox, Catholic and M uslim faiths in the form er Yugoslavia is quite a different matter. Let us now exam ine how contem poraneous sources describe the concept o f ethnos. In the w inter o f 558-559, the Byzantine em peror Justinian tried to persuade the prince o f the Turkic U trigur people to attack the Kutrigur, who posed a threat to Byzantium at that time. The m ost detailed description o f the event was recorded by M enander Protector. The em peror prom ised the ruler o f the Utrigur that, provided he defeated the Kutrigur, he would inherit the money the em peror annually paid the ruler o f the Kutrigur. H ow ever much he would have liked to have m aintained ties w ith Byzantium , the prince o f the Utrigur nevertheless refused, arguing that it would be an execrable and im proper thing to elim inate the Kutrigur, his relatives who belonged to the same people (phyle), for “they not only speak the sam e language as we [homoglossoi], and have dwellings [hom oskenoi], clothing and a w ay o f life [homoia hrontai stole kai diaite] sim ilar to our own, but they also have the same descent [ksyngeneis]. Yet we shall certainly take aw ay the horses o f the Kutrigur for our own use (which Justinian him self had also dem anded), to prevent them from riding them in their forays on the R om ans” . Undoubtedly, what we are witnessing here is a G reek’s interpretation o f the w ay the Turkic nomads o f the East European steppe saw the notion o f identical ethnic group. To the M agyars, who then already lived in a Turkic environm ent, these re lations m ust have been essentially similar. O ver and above the com m on sem iotic system — i.e. language, sim ilar dwellings, clothing and w ay o f life— the consciousness o f com m on descent was also important. And even if they were ruled by different chieftains, and they occasionally stole each o th er’s horses, they nevertheless belonged to one ethnos, and to a certain extent they outwardly acted together. Regino, whose scholarly activities are highly significant in connection with the Conquest-period history o f the Hungarian people, w rote his Synodalibus causis around 906. In it, he establishes that the difference betw een peoples lies in that they differ in origin, m orals, language and laws (se discrepant genere, moribus, lingua, legibus). If, however, we take a look at the nam es o f peoples listed or m entioned by Regino, we find that either he nam es peoples according to the land where they dwell (austrasii, aquitani, carantani, pannonii), or he returns to the old tribal names (e.g. saxones, baioarii). Actually, in the 10 th century the only significant question, w hich m ade a difference, was w hether a people was Christian (populus christianus), or an enem y o f the faith, fierce and disgraceful (ferocissima, nefondissima). The language and m or als o f the latter type were sim ply classified as barbaric (lingua et m oribus barbari), and differentiation between them was not considered w orthy o f dis
12
M ethodological introduction and the sources
cussion. O f course, if those barbarian peoples were to becom e allies, their differences im m ediately gained significance. There m ay have been a differ ence between the 10th-century chroniclers and the 6 th-century Greek authors in attitude, but certainly not in fundamental principles! To summarise: the definition o f the concept o f ethnos used in this book com prises some indispensable elements. These include a com m on sem iotic system , w e-consciousness and a perm anent self-designation. These are the constituent elements. Yet there are various other factors that play an im por tant part in the subsistence or decline— i.e. in the life— o f an ethnos which, however, are not necessarily preconditions to the identity o f a people, and whose absence does not alter the fact that we are talking about an ethnos. These prim arily include the consciousness o f com m on descent, com m on territory, com m on political organisation and com m on religion. These are theformative
elements. Finally, it m ust be stressed that although ethnos is a general category, its varieties, operational structures and the priority o f its elem ents are, neverthe less, highly diverse and variable in space and time. Ethnos is, therefore, a historical category. Consequently, a broader definition can now be given: ethnos— people— is a general category, but in substance historical. Ethnos is that historically formed aggregate o f people which has a com m on sem iotic system; whose m em bers think o f them selves as being separate; which possesses a perm anent self-designation; and in whose form ation the con sciousness o f com m on descent, com m on territory, com m on political organi sation, com m on religion and other factors can play an im portant role. The existence o f the latter can only help— their absence hinder— the existence o f an ethnos. The science concerned with ethnos has a subsidiary discipline w hich is concerned with the genesis o f peoples. For a long time in the past, ethnogenesis gave priority to the origin o f a people, to the detrim ent o f the formation o f a people. Researching into the origins m eans focusing on the forebears, while exam ining the form ation o f a people involves studying the com ponents and processes. The rather illusive idea underlying research o f the forebears is that any person or com m unity has an ancestor from which it is descended in a single, direct line. W hat this suggests is that the ancestor in a patrilineal society is the father, the father’s father, etc., and the ancestor o f the ethnos, too, can be traced back along sim ilar lines. This approach goes back to the early legal system which regulated the inheritance o f property and pow er by registering paternal descent. But w hat actually is the situation? Let us take as an exam ple a currentday Hungarian. He or she has, or had, two parents, four grandparents and eight great-grandparents. Let us suppose, for the sake o f sim plicity, that four
introduction
13
generations succeed one another in every century, i.e. everyone has a child at the age o f 25. This m eans forty generations in a thousand years, that is, a m illennium ago each individual should have had 2 40 ancestors. W hich is the equivalent o f 1,099,512,000,000— roughly 1100 billion individuals. Given that the Earth has 4 billion inhabitants today, the above num ber is 270 times the E arth’s population. In reality, the num ber o f actual ancestors is, naturally, lower— due to the fact that society generally disallows (or, at least, penalises) m arriage betw een second and third cousins, but not between fourth, fifth and rem oter cousins. Consequently, the children bom in such m arriages m ay have had the same ancestor— five generations down, for instance. This m eans that 2 6 (64), the theoretical num ber o f ancestors, m ust be reduced by tw o in the sixth generation, provided the parents w ere fourth cousins. In such cases, then, the num ber o f ancestors is not 64, but only 62. This so-called reduction num ber is not significant in the sixth generation yet, but it increases squarefold, since cousinhood rem oter than the fourth degree is rarely recorded, and m arriage is not forbidden. The calculation o f the exact num ber o f ancestors w ould require social statistics, such as exact childbirth and m arriage rates. Six ancestors will be counted if one confines oneself to the paternal line in tracing back o n e’s forebears six generations down, and if there was no reduction, fifty-eight ancestors will have been neglected. That, naturally, leads to m ajor distortions not only biologically, but also culturally. W hat this m eans in effect, is that not only every individual has significantly less ancestors than is m athem atically possible, but also that all individuals have a large num ber o f com m on ancestors. This is why ethnogenesis theories today are much more concerned, among others, with when and how a new ethnic group emerged, given the many form ative factors. The instance o f ethnic form ation and decay is extrem ely diverse. Reinhard Wenskus, for instance, sought to answer why the Chinese Empire assim ilated the nom adic peoples and why the Rom an Em pire did not. According to Wenskus, the imperial identity in the Rom an provinces was w eaker than the ethnic identity o f the “barbarian” nom adic (e.g. Germ anic) hordes. Thus, the ethnic identity o f the Germ anic tribes changed in substance, but nevertheless rem ained unshaken. As Jeno Szucs pointed out, the “first w ave” o f Germ ans— as well as the Avars, Turks and finally the M agyars who arrived m uch later— lost (or, in the case o f the M agyars, preserved) their ethnic identity in different ways. One o f the m ost exciting issues in the history o f the conquering M agyars is w hy the M agyar people was able to keep its head above water, while other peoples before them in the C arpathian Basin (the Huns, the Gepidae, the Longobardi, the Avars and a great m any sm aller peoples) sim ply disappeared. W alter Pohl established three categories o f Eastern European ethnogenesis: the Avar, the Bolghar and the Slav types. One m ay confidently
14
M ethodological introduction and the sources
add the M agyars to his list. The Avars disappeared, the Bulghars yielded to Slavonic influence, and the Slavs them selves failed to achieve a uniform state. The M agyars, however, survived w ithout changing their language, and also they managed to establish a state. This issue will be discussed in detail in C hapter X (see pp. 373-383). We m ust now touch upon the category o f nation. N ation is a relatively recent political category, although its prelim inaries go back a long way, and its elem ents were present in earlier times. In the M iddle Ages, the crystallisation o f the nation began earlier w ith those peoples whose ethnic and state bounda ries roughly coincided. Looking for a date, one can say that the m odern nation was created in 1791 w hen the French constitution ruled that sovereignty was the nation’s own. True, it is custom ary to talk about the political nation and the cultural nation. C om m on state establishm ent is suggested to bind the nation in the former, while culture in the latter. A lthough the concept o f the cultural nation appeared first in Rousseau and only later in Herder, it prim arily reflected those G erm an, Italian and Polish efforts that sought to establish a com m on state under the pretext o f culture. W hile the political nation is politics m aterialised, the cultural nation is the politics o f m aterialisation. Clearly, political sovereignty, too, is becom ing restricted, and in today’s global world, in the European Union, or even in the Benelux states sovereignty is now here near unlim ited. But even if the substance o f sovereignty is undergoing change, it is unlikely that the concept o f nation will be significantly affected. But it w ould be a m istake to project back to the past an unchanged present-day, past or future concept, and to thus term as “nation” such form ations that did not and could not m eet the criteria. M ention was made o f culture being a system o f signs. We m ust also refer to archaeological culture. A rchaeological culture is not only the collective term for certain types o f objects, but also for burial and— if observable— set tlem ent customs. It is usually nam ed after the m ost im portant or the first excavated site o f a specific type. An archaeological culture has scope in time and space w hich new ly excavated finds m ay modify. The succession o f archaeological cultures in a given place is im portant, for it provides for chronological observations. In many instances, an archaeological culture has, or can have, local varieties, chronological layers o f different ages. W hether archaeological cultures can be associated w ith peoples is a serious problem . Opinions regarding the issue range from full rejection to blind faith. Fallacious reasoning can be extrem ely detrim ental. A person, for w hatever reason, believes that such and such people lived in such and such geographical environm ent at such and such time. He or she finds that archaeologists have identified a given culture o f the same age and locality, w hich leads to the
Introduction
15
assum ption that the people he or she happens to be studying m ay be the vehicle for the archaeological culture in question. In the next round, the archaeologist takes this identity for granted, and claims w ithout any further proof w hatso ever that the archaeological culture in question was created by the m entioned people. Then the next scholar, linguist or historian comes along and refers to the fact that “as X archaeologist has pointed out” the archaeological culture in question confirm s the existence o f the given people. The truth is that, as in many cases, costum e and custom s m ay change w ithout the change o f the people who possess them. Fashion has always had some kind o f a role, even if its history did w ork slightly differently in the past than it does nowadays. It may be true that burial cerem onies and the related beliefs are extrem ely hardy, yet it is equally true that fundam ental beliefs and custom s can change w ithout the transform ation o f the population. In many European countries, but forem ost in Hungary, interm ent was superseded by crem ation in front o f our very eyes, w ithout any m igration activity accounting for it whatsoever. Often immigrants will, for many reasons, adopt the costum e or burial custom s o f the locals, but conversely, the locals, too, often adapt to m inority im migrants who enjoy a high prestige. Identifying archaeological cultures w ith ethnic entities is easiest where the present-day cem eteries o f a group o f people living in one locality can be traced back in time. Naturally even that is no full guarantee. However, the settling in o f a people which introduces a new way o f life, can be very telling. For instance, if an equestrian, nom adic people settles in a local fishing and hunting com m unity, the popula tion will definitely change, but next to nothing can be found out about the extent o f that change. Occasionally, inscriptions inform us about the deceased in a cemetery. For instance, the famous Volga Bulghar epitaphs tell us all about the inhabitants o f the V olga-K am a region in the 13th and 14th centuries. It was am ong these that the ethnic nam e M agyar appeared. However, the large religions like Christianity and Islam on the one hand unify burial custom s, and on the other, they reduce (or often elim inate) the features that im ply ethnic difference. Fortunately, where the great religions cropped up, so did liter acy— which, given an adequate critical approach, we are left to rely on in identifying peoples.
16
M ethodological introduction and the sources
2. CHRONOLOGY A N D CHRO NO LO G ICAL ASSUMPTIONS First and forem ost we m ust establish a chronological setting, and nam e the periods in which the ancient history o f the M agyars took place.
a) The geo historical eras The geohistorical period we shall be concerned w ith is the m ost recent one, the Quaternary. Its first epoch was the Pleistocene w hich began two, or two and h a lf m illion years ago. The reason w hy the Pleistocene is im portant in our context is because it was in this epoch that m ankind evolved. True, it was then H om o sapiens fossilis, or the N eanderthal M an that inhabited one part o f the Earth. This race died out, and was superseded by today’s races. O ften in geological history great parts o f the Earth w ere covered in ice. The m ost recent glacial period, the so-called Wtirm Period lasted until the end o f the P leisto cene, until ca. 10 000 BC. These tw o events delim it the prehistory o f languages and peoples. The N eanderthal M an probably had a sign system o f som e description com prising sounds, w ith w hich he was able to convey his thoughts, but it was different in every respect from later hum an speech. Also, we m ust
Figure 1 The southern frontier o f glaciation in Eurasia
17
Introduction
exclude from the history o f hum anity and o f the individual peoples, those lands which were covered in ice. The glacial zone in the W iirm Period roughly followed the 60th latitude at the Ural M ountains, the 61 st betw een the Ob and the Yenisei head-w aters, and the 64th—65th latitude to the east. The Pleistocene was followed by the Holocene or Recent epoch which is traditionally subdivided into the following periods: Date (BC) 10000-7500 7000-5000 5 0 0 0 - 500 500-
Period Lower Holocene 1 Lower Holocene 2 M iddle Holocene U pper Holocene
Abbreviation
Clim ate
HL HL HL HL
subarctic boreal A tlantic, subboreal subatlantic
1 2 3 4
In the second period o f the Lower H olocene the clim ate o f Eurasia began to thaw, which contributed to the em ergence o f production cultures. As will be pointed out, the form ation o f the M agyars surprisingly coincided w ith the Upper-Holocene period, therefore it happened concurrently w ith a significant change o f climate. Naturally, the clim ates o f the different zones o f the Earth greatly varied, and this determ ined both flora and fauna. We must, therefore, briefly discuss the geobotanical zones o f Central Eurasia.
b) The geobotanical zones M oving from north to south, the northernm ost zone is called the perpetual frost zone. There is no arable land in this area, and it has arctic faunas such as the polar bear, the seal, the walrus, the albatross and the storm petrel. The southern perim eter o f the tundra belt is a line that connects areas with a mean July tem perature o f 10 °C. The zone is characterised by long and harsh winters, and cool and short summers. The subsoil is thin and m eagre, there are no forests, and the area is covered in long-stem m ed vegetation, the tundra. The arctic willow grows here. Its southern regions feature lignified dw arf shrubs and spruce, and other varieties o f fir are not uncom m on either. Its fauna com prises reindeer, alpine sheep, fox and birds o f prey. In sum m er other animals m ove up from the south, and the tundra becom es marshy. The taiga belt is characterised by a very cold w inter and a relatively hot summer. It has arable land which becom es m arshy in summer. Its vegetation features taiga or fir whose varieties include the Siberian spruce, the com m on spruce, the com m on silver fir, the pine, and the deciduous larch. The only other
The geobotanica! zones of Central Eurasia
CN £> u. 3 00
Introduction
19
species that grow, other than firs, are willow and poplar. The northern part o f the taiga is called the dark taiga on account o f the density o f the crowns o f the trees which prevent the developm ent o f undergrowth. To the south, the bright taiga does have undergrowth. The Am ur-Usuri relict is a fascinating area within the taiga. W idespread forest species withdrew here as the glacial period advanced, and spread towards the west as the glacial period ended. The most characteristic anim als o f this area are furry anim als such as the sable, the ermine, the stone marten, the otter, and the beaver, but also the bear, the w olf and the lynx. It is rich in birds, especially water-fowls. The belt o f deciduous forests covers the 4 -500 kilom etre-w ide area that begins with the Carpathians, stretches down to the Southern Ural, gradually narrows towards the R iver Yenisei, and fades away around the Altai M oun tains. Its northern region features a vegetation o f mixed forests (willow, poplar, oak, maple, lime, ash and firs), its central region purely deciduous forests, while the southern parts gradually fade into the steppe. C haracterised by beech in Europe and poplar in Asia, this belt sustains furry anim als, as well as bison and wild boar, and the animals o f the southern steppe tend to m ove north here. The zone has arable areas, and it served as the m ost valuable agricultural land after the clearance o f the forests. The temperate grasslands stretch from the Carpathians through the Ukraine to the Central Volga region, and then follow the borderland o f Siberia up to the River Yenisei. The steppe is grassland with a black subsoil. The now extinct aurochs and the eohippus used to live here. Today’s fauna includes various species o f antelope, wolf, fox, ground squirrel, hamster, field m ouse and the mole-rat. Large birds o f prey are indigenous to all o f its regions. This zone is the m ost suited to nom adic animal husbandry, and it served as the setting for the large migrations. The arid subtropical and tropical zone stretches to the south o f the steppe zone. The mean yearly rainfall is extrem ely low, and the zone features saltwater lakes, waters w ithout outlets, shifting sands, and stone deserts. Very sparse, vegetation here includes various species o f thistles, needlegrass and wormwood. The boundaries o f the geobotanical zones are not clear-cut, and they tend to shift with changes o f climate. The natural environm ent o f peoples living in these borderlands can undergo substantial change as a result o f even the slightest clim atic change.
20
Methodological introduction and the sources
c) The Stone mid M etal Ages The earliest ages were nam ed after the m aterials and the techniques used in im plem ent-m aking. In the rem otest age, m an used unpolished, chipped stone. Scientifically term ed Palaeolithic A ge in those regions which com e into play regarding the prehistory o f the M agyars, this age ended in approx. 20 000 BC. A n interim period ensued, characterised by m ore finely shaped, but still unpolished stone implements: the M esolithic A ge lasted roughly from 20 000 to 5000 BC. It was after 5000 BC that the age o f polished stone im plem ents, the N eolithic A ge began. O f the metals, it was copper that first entered the w orld scene, in w hat is today southern Italy. It becam e widespread throughout the Central Ural region, and in the latter h alf o f the m illennium in Siberia, too. Used together with stone im plem ents, copperware was w idely introduced. Im ported from M eso potam ia through the Caucasus, the first bronze objects cropped up in the first h alf o f the 3rd m illennium BC. Bronze has been shown to have been produced locally as o f the early 2 nd m illennium only, but then it spread apace: it ap peared in the Carpathian Basin betw een 1900 and 1700 B C , reaching the Ural, to day’s Khazakhstan and the regions o f Central Asia, in the m id-m illennium . Superseding the various bronze alloys, iron, too, entered the steppe through the Caucasus, records o f which date back to the 8 th century BC. The new m etal rapidly conquered the world from the Black Sea to China. On the turn o f the 6 th and the 5th centuries BC it m ade its way up the Altai M ountains to Inner Asia, reaching the Transbaikal areas o f Siberia in the 3rd century BC. The ages briefly described above provide a very broad setting w hich further research m ay alter slightly. The N eolithic A ge is o f overriding im portance, for it m arks the beginning o f a production economy. Earlier, gathering, fishing and prim itive forms o f hunting provided m ankind the m eans o f subsistence. This is im portant with respect to ethnic and linguistic history, because the N eolithic Age saw some fundamental changes. In the Palaeolithic and M eso lithic ages, large territories provided the m eans o f subsistence for very small groups o f people. These small groups frequently m oved from one locality to another, consequently a durable com m unication situation could not develop. The developm ent o f a perm anent— from our point o f view, assessable— com m on sem iotic system and culture in a large com m unity postulates cohabitation based on relatively prolonged interaction. Consequently, any fam ily o f lan guages and ethnos could only come into being in the N eolithic Age. Naturally, by the Palaeolithic and M esolithic Ages hum an com m unication substantially differed from anim al com m unication; nonetheless it lacked one im portant elem ent which did not appear until the N eolithic Age, nam ely
Figure 3 Clim atic, geohistorical, archaeological and language historical periods
22
M ethodological introduction and the sources
permanence. Accordingly, there is a sharp line (which lasted for centuries, o f course) betw een the M esolithic and the N eolithic Ages. Even given today’s scientific apparatus, we cannot study the pre-N eolithic languages and peoples, due to the fact that all our know ledge and m ethods pertain to the nascent N eolithic peoples and languages. I f a group w ith a pre-N eolithic culture were to be found som ew here on the globe, this culture would surely be the result o f some secondary developm ent. The Neolithic Age began in the river-valley cultures in the 7th -6 th m illen nia, however, it did not come to the areas we are concerned with before the 6 th m illennium . It is not simply coincidence that the peoples and languages whose ancient history we know m ost about did not emerge earlier than the 6 th or 5th m illennium BC. That date can be assigned to the earliest apprehensible periods and proto-languages o f the Indo-European and Sem itic linguistic com m unities, as well as to the nascent ancient Indo-European and Sem itic peoples w hich shared features o f a perm anent structure— in the m odem sense. W hat this chronological boundary means, in effect, is that those ethnogenetic or linguistic relationship theories which seek to encom pass periods earlier than the Neolithic Age cannot be considered m ore than pure speculation and, consequently, cannot be granted scientific credit. Accordingly, neither can the prehistory o f the M agyars have begun prior to the N eolithic Age, and even then only in those territories o f Eurasia where those nucleus groups are to be found from w hich the M agyars gradually em erged in the 5 th -4 th m illennia BC. As will be pointed out later, it happens to coincide with the putative period o f the Uralic linguistic community. The subsequent m ajor change was brought about by the appearance o f bronze, and it is probably not mere chance that the splintering o f the FinnoUgrian peoples roughly coincides with that period. The scattering o f the Ugrian group which had broken away from the Finno-U grian peoples, and the form ation o f the M agyar people m ust have been concurrent w ith iron gaining currency in the first h alf o f the 1st m illennium BC.
d) Some co-ordinates in world history Let us next attem pt to link the M agyars’ prehistory with som e w ell-know n dates in w orld history. The Uralic com m unity o f peoples m ust have begun to splinter about h a lf a m illennium after the Indo-European com m unity had broken up. At the time when copper first appeared in M esopotam ia, the more prim eval Neolithic forms still existed in the Ural region. T hat attests to the delay o f the periphery o f world history at the time. The form ation o f the Ugrian com m unity, as we have seen, m ust have been concurrent w ith the appearance
23
Introduction
o f copper. That roughly coincides with the age o f the B abylonian Empire (Ham m urapi, the law-giver, lived in the 18th century BC). The C retan Culture prospered then; the m agnificent Palace o f Knossos was erected at that time; and new peoples came into view in Asia Minor, N orthern G reece and N orthern Italy. This period coincides with the beginnings o f Chinese civilisation when the em peror Huangdi ruled the Yellow River region. The M agyars’ independence roughly coincides with the founding o f the Roman Republic (504 B C ), the nascent Athenian democracy, Cleisthenes o f A thens’s constitution (507 B C ), and A eschylus’s dramas. H aving introduced his reforms and com m itted his thoughts to writing in the form o f analecta, the Chinese Kong Fuzi, better known in Europe as Confucius, also lived around then. The dates o f the above sim ultaneous events naturally should not be read verbatim. Our knowledge regarding the dates o f the form ation o f the M agyar people is a lot vaguer than would perm it linking events to years or even decades. The degree o f uncertainty regarding the form ation o f this people is in the order o f centuries.
e) Chronology and continuance Establishing the date, chronology or continuance o f two or m ore events with precise written sources at our disposal is a fairly easy matter. Nevertheless, even in well-recorded cases (the farther back in time we go the scarcer they get) source criticism is extrem ely im portant (m ore about that later). Besides the written sources, however, there are other varieties o f sources that contrib ute to our endeavours in establishing certain particulars o f time. The absolute chronology, i.e. chronology with num erical citation, is natu rally always relative to the calendar system used by the source. We m ust not lose sight o f the fact that accurate observation and registration o f the year, month and day in an astronom ical sense is a very recent achievem ent, although the calendar and year calculations, which rely on them , nevertheless boast traditions going back thousands o f years. Only to a certain degree o f accuracy can calendars keep up with astronom ical events which is why, even today, they occasionally need to be adjusted to astronom ical time, by the system o f leap-years, for instance. H um anity has used a great many different calendar system s, and w ithout accurate knowledge o f these it is im possible to identify any date whatsoever, even in the written sources. However, experts can do the job, given some com petent handbooks. The identification o f date, chronology and continuance w ithout available written sources is significantly harder. There are two ways
24
Methodological introduction and the sources
to establish dates in such cases. One alternative is absolute, w hich involves the conversion o f a date to our own calendar system , expressed in term s o f years. The other alternative is relative. We do not know the date, but we do know a notable event which it came after {post quem dating), or which it preceded (ante quem dating). The actions o f an em peror can only be recorded after the em peror has acted (provided we exclude prophesies), but not in every case do we know how many years later the event was recorded. We do not know when the scribe jotted down the event, but we do know w hen he died. Consequently, his notes m ost have been w ritten before that. In m any instances we only know that an event cannot have occurred before a know n date (ante quem non), or cannot have occurred after it (post quem non). Often, although we are unable to establish the exact date o f an event, the decade or century we can. Historical science has other means, besides written sources, to establish absolute or relative chronology.
3. THE ROLE OF THE NATURAL SCIENCES IN ESTABLISHING CHRONOLOGY The role o f the natural sciences in historical science is increasing apace. In m any cases when examining the genuineness o f a historical source, the same natural scientific m ethods can be applied as in court cases. For instance, paper and ink analysis can be important. There are four areas in which the natural sciences are indispensable for the reconstruction o f past ages and the ancient history o f the M agyars: (a) deter m ination o f age; (b) determ ination o f origin; (c) determ ination o f environ ment; (d) determ ination o f production and technology.
ci) The determination o f age Archaeologists have methods o f their own to determ ine age, but concurrent with these, they increasingly tend to rely on natural scientific m ethods. These m ethods can be grouped twofold. On the one hand, there is a distinction betw een absolute and relative m ethods o f dating, and on the other, betw een the different branches o f natural science whose results history draws on. Consequently, there are, for instance, physical, chemical, botanical, zoologi cal, or geological methods. Naturally, this book cannot attem pt to give an expert presentation o f these, due to the fact that the rapid developm ent o f science and the research work o f entire professions underlies these methods. We can but briefly refer to the m ost im portant ones.
Introduction
25
The prim ary m ethods o f age-dating include those m ethods that are based on radioactive decay. These are called radiom etric m ethods on account o f the fact that radioactive disintegration can be m easured w ith great accuracy. Thus it can be determ ined when certain substances w ere absorbed in an archaeologi cal find. The m ost com m on m ethod is carbon dating, nam ed after the radio active C 14 isotope o f carbon, a product o f cosm ic radiation. The half-life o f carbon-14 is 5730 years. W hat this m eans is that after 5730 years, h alf o f the C 14 atoms will have decayed into nitrogen-14. Living plants and anim als take up the isotope via carbon-dioxide contained in air. H aving penetrated organic tissues, C 14 is no longer exposed to cosmic radiation. If later the plant or the animal, or parts o f it (e.g. leather) are buried, com paring their C 14 content with the С 14 content o f a current-day sample o f the same organic m atter will reveal when the isotope was taken up by that organic matter. And because the age o f organic matter, as well as the time they were buried, can be determ ined more or less accurately, the age o f the artefact can also be established. The method, which experts have been refining for decades now, has its limits, o f course, and its application requires not only expertise— as do all radiom etric m eth ods—-but also a critical approach. Having a relatively fast half-life, carbon-14 is less and less useful as a chronom eter in age-dating finds older than 50,000 years. It was once assum ed that the m agnitude o f cosmic radiation at that time was m ore or less constant. That is unlikely, however, and consequently the error percentage due to the fluctuation o f radiation has to be reduced by other methods. The origin o f the carbon infiltrating organic m atter also varies. Carbon atoms can enter plants from the soil, anim als from other anim als or plants or drinking-water, therefore the ratio o f carbon isotopes is not neces sarily constant. Because the main problem o f the C 14 method is that the isotope has a relatively fast half-life, science has searched for other radioactive elements, with longer half-lives. O ther radiom etric m ethods include the helium m ethod and the lead method. These isotopes feature a considerably longer half-life, therefore, the proportion or ratio o f the daughter isotopes o f lead and helium (the latter is released in the process o f the disintegration o f heavy m etals) to their respective parent isotopes, can inform us about much earlier ages. The potassium to argon and the rubidium to strontium techniques are used to age-date rock. All other geochronom etric m ethods concern us purely for m ethodological reasons. The underlying principles o f the so-called dendrochronological m ethod or tree-ring dating are quite different. The widths and the tones o f colour o f the individual tree-rings reflect the rainfall and tem perature conditions o f the given year, hence there are no identical tree-rings, and adjacent rings form a sharo contrast to one another. However, the tree-ring features o f the same year
26
Methodological introduction and the sources
are very similar. First a tree-ring sequence is established from sam ples taken from living trees. This sequence is extended by the ring sequences o f felled trees— some o f which were alive at the sam e tim e as those living trees from which the first sam ples were taken. Thus, working one’s way backw ards, one can go back several thousands o f years. Sequoias and redw ood— m any o f which are several thousands o f years old— can serve as a control. D endrology is today capable o f establishing the exact age o f a fortunately preserved piece o f wood, in certain areas, from 8000 years ago. M ore accurate results are available by applying the carbon m ethod to the lignin and cellulose content o f the tree-rings. Wood, however, decays relatively quickly, and consequently archaeological finds rarely have wood rem ains that can be tested dendrologically. Also, the clim atic conditions o f each belt are different, hence the tree-ring standards m ust give consideration to tim e and space. The secondary use o f a piece o f wood can also be m isleading: w hen a crypt, for instance, was built from “second-hand” m aterial, taken from an old building, or another know n grave. Notwithstanding these calls for caution, tree-ring dating pro vides archaeologists with a wide range o f chronological data. Natural scientific techniques are not uncom m on in age-dating clay pots or tiles— so frequent among archaeological finds. All objects are continually exposed to ionising radiation. Thermoluminescence dating m easures the em is sion o f light from heated crystals previously exposed to this radiation. At low tem peratures objects will em it this radiation instantly (fluorescence), or prolongedly in the form o f light (phosphorescence). Sensitive devices are capable o f m easuring the unem itted light energy w hich can be released by heating (therm olum inescence). Clay objects are fired at high tem peratures, therefore im m ediately after firing they do not preserve light energy (phospho rescence). Thus the radiation dose is preserved from the very m om ent the object was made. The retained light energy can be m easured and com pared with the results o f laboratory irradiation. This highly refined m ethod has its limits and problems. The discrepancies caused by the location-dependent intensity o f ionising radiation can be elim inated in m any cases w ith com para tive methods. The m ost skilled forgers, however, can produce an artificial radiation value equivalent to that o f a clay object m any thousands o f years old. Also, the ionising effect o f other objects, or possibly the secondary exposure o f an object to fire m ust also be given consideration. N otw ithstanding these problem s, given due caution and circum spection, the m ethod presents the opportunity to age-date clay objects and other archaeological finds from the same site. There is another m ethod for dating clay pots. The Earth has two m agnetic poles— that is general knowledge. The location o f these poles changes at a pace which is m easurable in term s o f decades. Consequently, the direction and
Introduction
27
intensity o f the m agnetic field o f the Earth also changes, and these changes can be traced back relatively accurately. The clay used in the m aking o f pottery contains an array o f different iron-oxides. U nder norm al circum stances the m agnetic poles o f these iron-oxide particles are situated random ly in the clay, hence their outward m agnetism is zero, they neutralise one another. W hen, however, the clay is fired at a high tem perature (approx. 700 °C) the m agnet ism o f m ost o f the iron-oxide particles becom es aligned with the E arth’s m agnetic field. The directional character and intensity o f the m agnetic field existing at the time the pot was last heated up are preserved after it has cooled. Because scientists can describe the past shifts o f the E arth’s m agnetic poles, as well as the change o f the intensity o f magnetism , all that is left is to establish how a given location relates to these. A fired clay pot forgotten in the kiln after it was fired is ideal for the purpose, because not only can the vertical position o f the object at the time o f firing be determ ined (the shape and the firing traditions are the give-away usually), but also its horizontal axis. Good results can be obtained even if only the vertical axis at the time o f firing is known. With clay bricks, however, one can never be sure. Still, archaeom agnetic dating is a fairly accurate m ethod for dating up to approx. 7000 years back. Astronom y can also contribute to absolute dating. O bservation o f the sky at night has always helped Earth dwellers to find their bearings. N ot only the large river-valley cultures, but also the nom adic steppe peoples possessed substantial knowledge about the stationary points in the sky. This enabled them to determ ine directions and arrange their activities according to seasons. These observations have come down to us in w ritten sources, languages and legends. The night sky, however, does not appear the same everyw here and consequently the analysis o f archaic astronom ical nam es can provide infor mation as to where an ethnic group dwelt at the time it gave a vernacular name to one particular star or constellation. The analysis o f the nam es that a people gives its calendar, the years, days and festive days, can shed light on where that people view ed the sky from; accordingly, these calendrical nam es yield a wide range o f inform ation regarding the m igrations, form er hom elands or cultural connections o f the peoples. The tw elve-year calendar o f the steppe peoples— in which each o f the twelve signs o f the zodiac stands for one year— is very ancient indeed. It has been conjectured that the word sarkany ‘dragon’, o f Turkic origin, entered the Hungarian language during the use o f the Turkic tw elve-year calendar. Naturally, the night sky itself is ever-changing, with real events— for instance, solar eclipses, or the appearance o f large m eteors; and seeming ones— for instance, the change o f the Pole Star caused by the m ovem ent o f the globe’s axis. If these were registered in the written sources, we have some im portant chronological points o f reference. Even if they survived in written
28
M ethodological introduction a n d the sources
calendars rooted in plain tradition, they can shed light on the history o f words, nam es and cultures. We shall only m ention some other absolute age-dating techniques. These include the flu o rin e, the ultraviolet ray, the neutron activation, and the derivatographic dating methods. R elative chronology seeks to answer other questions, and consequently it has other working m ethods com m only characterised by the fact that they do not offer an absolute answ er to tim e-related issues. W hat they do inform us about, however, are the geological, geom orphological, clim atological and biological changes— especially the change o f the faunal and floral environ m ent— as well as about the succession o f these changes. A lthough a historian is less concerned with the geophysical strata, in m any cases the dates o f archaeological finds can be identified with full knowledge o f the relative position o f the different geophysical strata. Likewise, one m ust not fail to consider the clim atic, botanical and zoological changes o f an area exam ined from a historical viewpoint. The desertification o f regions, the shifts o f their clim atic and geobotanical zones can contribute im portant inform ation to historically significant events, the lack o f absolute tim e notw ithstanding. Rivers used to play an im portant role in the history o f peoples, in irrigation, transport, or sim ply as sources o f water. However, they did not alw ays flow where they do today, and the hydrographic netw ork o f certain territories m ust have undergone substantial change. Thus, for instance, the R iver Kam a, which played an im portant role in the ancient history o f the M agyars, originally flowed north, and only as a consequence o f the last great glacial period did it turn south and becom e a tributary o f the Volga, and through it, o f the Caspian Lake. The presence o f wild and cultivated plants m ust have been im portant, econom ically speaking. The characteristics o f the flora, however, have under gone m ajor changes in the course o f the ages. Fortunately, am ple am ounts o f pollen have been excellently conserved. Pollen analysis enables us to recon struct the wild and cultivated flora o f an area in a given age. M oreover, it offers som e inferential evidence regarding clim atic changes, too. B ut know ledge about the flora and fauna o f an early period is not only com pelling on account o f the interpretation o f economic issues and m igrations it can afford us, but also because o f the ancient plant and anim al nam es w hich m any m odem languages have preserved. M ore about this issue below (see pp. 34, 93, 111, 192).
29
Introduction
b) The determination o f origin Even in the ancientm ost ages, objects would often travel afar. Objects, knowledge, and even the words designating them, journeyed to and from China and Egypt along the Silk Route. But objects w andered in other im por tant ways, too. We often need to know where exactly the objects buried in a cem etery w ere made. That can shed light on the connections o f a people, and furtherm ore, on the locality they moved from. M etal artefacts are particularly suitable for determ ining where the metal which they are made o f has come from. Spectrum analysis can detect such elem ents in m etals which are char acteristic only o f one particular region, or with luck, one specific mine. This is especially im portant with early bronze items. With cloth, it can be estab lished which com ponents are o f animal or o f plant origin, and w here those raw m aterials cam e from. M any different techniques exist for m anufacturing glass. Given that we know the specific technique, the origin o f a glass object can be determ ined. Also, for instance, the origin o f raw paper m aterial o f excavated written docum ents, can be o f importance.
c) The determination o f environment The historical determ ination o f the natural environm ent seeks to answ er two questions, nam ely w hat kind o f econom y and w ay o f life did the natural conditions o f a given area in a given age provide the m eans for, and w hat kind o f historical, econom ic and population-related changes (decrease or increase, m igrations) did they bring about. Historical clim atology has a key role in the determ ination o f environm ent. Furtherm ore, we m ust distinguish historical climatology, relevant to large regions, from m icro-clim atology which is concerned w ith sm aller areas. With regard to the history o f the M agyars, it is im portant to consider w here they developed their nom adic life style, although the forebears o f the M agyars were hardly involved directly in the global process. We will com e back to the em ergence o f nom adism later (see pp. 142-145 and 320). W hat specific cli matic changes played a role in the M agyars’ switching from an equestrian-hunter type o f economy to being pastoral nomads is quite another matter. Finally, it is o f fundamental significance what the clim ate, phytogeography and zoogeography o f the Carpathian Basin w as like at the time o f the Conquest. We shall discuss these briefly later (see p. 155). The sciences o fpalaeobotany, concerned with the history o f the flora; archaeobotany, researching the history o f dom esti cated plants; palaeozoology, studying the historical changes o f the fauna; and archaeozoology, concerned with the history o f dom esticated anim als have made significant progress in the past decades.
30
Methodological introduction and the sources
d) The determination o f production and technology Historians are generally keen to find out for how m any people a given area was able to provide the m eans o f subsistence. The reconstruction process is facilitated by knowledge about the efficiency o f the m eans o f production under the given set o f circum stances. One can assess carefully, for instance, the depth a specific type o f plough tilled the soil, and the size o f area it was capable o f tilling in a day. The life span o f a harness, or how long a specific w eapon was passed down in a family or was kept in enem y hands can also be o f some interest. The annual metal needs o f a group can be im portant in many respects— provided they used it for m aking tools only— as can the quantities o f jew ellery they m ade or bought, as well as the techniques these were made with. Careful study o f the N agyszentm iklos treasure has enabled us to establish how m any hands used how m any tools in creating the treasure, and w hether certain inscriptions could have been m ade with the same tools as used for the decorative motifs. Thus new knowledge can be obtained about the late Avars— this we shall come back to below (see p. 131).
4. OTHER ANCILLARY DISCIPLINES IN ESTABLISHING CHRONOLOGY
a) Numismatics Coins can provide chronology with much inform ation about the times and places which m oney in circulation reached. Consequently, num ism atics, the study o f coins or medals, is an im portant ancillary discipline to historical science. The issuing o f m oney was always a m onopoly to some degree. The nam e or image o f the issuing ruler or o f a deity w orshipped by the issuer, or some kind o f sym bol, served as a rem inder o f the person who had the coin minted. Coins usually featured inscriptions and tokens. M ore often than not, these enable us to establish the date the coin was issued. M oney was naturally copied, over-struck, used, and forged very early on. Giving these due consid eration, coins can be very helpful in establishing chronology. Coins and m edals generally only allow for p o st quem dating, on account o f the fact that we have no knowledge about how long they w ere in circulation for or were hoarded or used for secondary purposes, e.g. as w om en’s jew ellery. Generally
31
Introduction
speaking, however, coins are in circulation for short periods o f time only, simply because the stocks becom e exhausted. So-called treasure finds are problem atic, since they m ay contain the hoarded coins o f treasuries o f many decades or even centuries earlier. W hen an archaeologist finds such treasure, the oldest and the m ost recent coins can shed light on who hid the m oney and when. The circum stances o f the find m ust be given special attention. It is perfectly feasible that a coin found in a grave is in no way linked to the age the buried person lived in. Possibly we are looking at a grave robber, and a coin robbed in his youth followed him to his grave. H ow ever unlikely it may appear, it has actually happened— possibly due to the workings o f the animals in the soil— that a coin has wandered down from one grave into the grave below, presenting researchers with a seemingly insoluble puzzle! Such cases are rare, o f course, but the critical assessm ent o f finds m ust be open to possibilities o f every nature. Plate III shows the coins unearthed with the famous Conquest-period royal grave o f Karos. Two out o f the three Arabic coins com e from the m int o f Esmail ibn Ahm ad whose attacks forced the Pechenegs to set out on their wanderings. The coins date to 904-905 AD, and were used as funeral obols. The other coins were issued by the Frankish Louis the Infant (900-911). The latter, as well as the third Arabic dirhem are pierced and w ere used to decorate dress hems.
b) Archaeology A rchaeological finds them selves can serve as chronometers. The relative age-dating o f these (in relation to one another), and absolute dating (with a specific num ber o f years given) is one o f archaeology’s m ost im portant fields o f interest. Decoration, form, structure and function can all contribute to establishing dates or periods. The more elem ents an archaeological culture consists of, the more distinctive and isolatable it is, and consequently, the more reliable as regards chronological assumptions. A grave w hich m erely contains a small urn with the ashes o f the deceased is less helpful in this respect than a grave w hich contains the deceased in full m ilitary pomp, w ith accoutrem ents and every object o f his lifetime, or even his dom estic animals. Provided, with the help o f external means, such graves or groups o f graves can be given a chronological setting, researchers will have obtained knowledge about the age o f other, identical or quasi-identical types o f graves. The relative physical position o f graves to one another can also help. Any large cem etery will have an inner chronology o f its own, for the reason that m ost cem eteries are used not ju st for decades, but centuries. However, w hat archaeologists really want
32
Methodological introduction and the sources
to know is the chronology o f archaeological cultures. Besides the above described m ethods, archaeology has many other m eans o f testing its finds for chronology. Naturally, it takes into account w ritten sources. The m ost com m on problem with this is the same as with the ancient period o f M agyar history, nam ely that the sources rarely, if ever, unam biguously state w here exactly a people lives. Consequently, its archaeological identification m ay be doubtful even if the date o f the finds can be established.
c) Linguistics Language can serve as an im portant source in historical reconstruction. We shall consider the relevant parts o f linguistics later (pp. 92-116), and only focus here on some theoretical and m ethodological issues. Language is an ever-changing system whose every part and level has a history o f its own. W hat this m eans is that phonem es, words, the m eaning o f w ords, affixes (inflectional or other), gram m atical structures, phrasal syntax, etc. have their own history. Linguistic change— language history— has general theories o f its own, owing to the fact that m any regularities are characteristic o f every change o f every language, while other regularities pertain to certain languages, or certain parts o f speech only. Such regularities form processes w hich can be traced back. Exam ples o f these will be given w hen the linguistic sources are discussed in Chapter II.4. W hat generates those changes and how they appear on the surface are issues one m ust be fam iliar with. Changes are effected by language usage. Dead languages do not change, unless they are used, as for instance was Latin in the M iddle Ages. Usage can bring about m any different types o f language change. There is a difference between frequently and rarely used words and elem ents. Interest ingly enough, very often it is the m ost frequently used parts that change more often, but the contrary is equally true: the m ost frequently used parts rem ain the m ost stationary ones. There are several reasons for that. One reason for the stationariness o f the m ost frequently used words— verbs, for instance— is that they assum e a m ajor role in com m unication, and consequently in m aintaining linguistic continuity. The ch ief explanation for w hy m ore frequently used w ords change faster is w ear and tear. The m ore frequently a word is used the greater the probability that a new variant emerges and becom es rooted. Language alw ays has a tendency to economise. If a w ord is too long, it will w ear in the course o f usage, but if it is too short, it hinders com m unication. Accordingly, m onosyl labic words stretch out and become longer. One com m on feature o f all languages is that words are preponderantly disyllabic— languages have a
Introduction
33
propensity for that. Sounds, too, keep changing, but when the pronunciation o f two sounds becom es very similar, understanding becom es difficult. Thus, either the two sounds eventually coincide in the course o f their history, or the process o f sound change comes to a halt. A state o f isolation favours change, for if a word has m any derivations, the com m on stem will be preserved in all o f them. Anyone w ho has learned an Indo-European language— Latin, Eng lish, Germ an or Russian— will know that the conjugation o f the m ost fre quently used verbs is irregular. Yet irregular verbs are always m ore scarce than regular ones. R egular verbs form an ever-expanding majority, while the num ber o f irregular verbs remains invariable, or decreases. The present-day Hungarian language has less and less verbs with the -ik ending, and in many cases the less frequently used verbs with this ending no longer follow the conjugation characteristic o f this group o f verbs. Although all languages are com m only characterised by perpetual change, it is debatable w hether the speed o f language change is always identical. With those languages whose history we can trace back a long way with the help o f written sources, such as Indo-European, Sem itic or Chinese, the experience is that the speed o f the change o f the language and its constituent system s (e.g. sound system, vocabulary, affixes, etc.) is not even. Som etim es the pace o f change accelerates, som etim es it greatly slows down. M any scholars have studied the problem. Changes are usually precipitated by the isolation o f groups o f speakers. The splinter group will subsequently retain certain features which in the main, unfragm ented linguistic com m unity have already been changed, and will thus preserve various ancient elem ents, so-called archaisms. The present-day languages o f the Szekely in Transylvania and o f the M ol davian Hungarians, the Csango, are examples here. Conversely, the lan guage o f the isolated group o f speakers will undergo changes which proceed at a faster pace than in the larger, undivided group. A t the same time, the pace o f language change is surprisingly slower am ong those highly m obile groups o f peoples who m eet m ore regularly, like the nomads, for instance. Regional dialects will develop faster in the language o f settled peoples. By processes o f isolation these dialects m ay subsequently becom e independent languages. It is very difficult to draw the line betw een dialect and “m ain” language— m u tual understanding is a good test, however. The language o f nom adic peoples splinters into dialects much more slowly. No doubt, the lack or insignificance o f dialects can have m any other reasons, such as the interior m igrations o f groups o f peoples, or the high prestige o f a ruling group. W hich brings us to another im portant reason that effects language change: prestige. Language learning itself is a process o f imitation; a child acquires the language by im itating his or her parents and peers. But language groups, too, will change their language depending on whose language usage they begin
34
Methodological introduction and the sources
to imitate. Even in the days w hen printed writing, television and radio did not exist, there were groups whose political, cultural or econom ic prestige was high and w hose speech, pronunciation or vem aculism s the m ajority o f society sought to im itate and master. Linguistic change often started earlier in certain groups o f speakers, from where it spread to the other groups. Consequently, it frequently happens that old and new forms coexist in a language. This can com e in very handy, because linguistic changes can be used in the chronologi cal reconstruction o f history, provided one has extensive know ledge o f the linguistic change theories and research methods. The so-called biogeographical m ethodused to be the m ost com m on m ethod in the reconstruction o f ancient history. Naturally, the distribution o f plants and anim als can change. O f the cultivated plants, potato and sw eet com were im ported to Europe after the discovery o f Am erica. The nam es o f these plants shed light on their transit route, and also attest to popular linguistic creativity. Both the fruit and the designation narancs ‘orange’, o f Indian origin, entered H ungary and its language by Persian, A rabic and eventually Italian m ediation. W hen eventually a people m igrates, the plant and anim al nam es o f its form er hom eland w ill travel with it, so long as those plants and anim als can be found in the environm ents the people passes through. W hen, however, the flora and fauna lacks those plants or anim als for a long period o f time, their nam es will be forgotten. If subsequently the people encounters them again in the course o f its wanderings, it will coin new words to designate them. Such new nam es can be borrowed from locals, freshly coined, or form ed from existing plant nam es, by adding on an adjective, for instance. It is interesting in this respect that, as opposed to all the other Finno-U grian languages, the only word in Hungarian collectively designating coniferous trees is fe n y o ‘fir’, w hile the language boasts a vast array o f nam es for deciduous species. This goes to show that, for a long period o f time, the M agyars m ust have lived in a biogeographi cal environm ent which essentially lacked a coniferous flora. A closer look at the other Finno-U grian languages reveals that the various conifers have nam es o f their own. Thus, at least four coniferous species can be distinguished in the Uralic languages, nam ely the spruce fir, the arolla pine, the silver fir, and the larch. However, o f the deciduous trees only the elm existed in the proto-language. Therefore, it is the other way round in the Uralic languages: it has one deciduous tree and several coniferous species. Pollen analysis, m entioned above, as well as various other m ethods have enabled us to reconstruct where and when this mix o f trees was characteristic o f one particular large area. Rapidly spreading afresh after the last great glacial period, the taiga featured this particular flora in the 6 th -5 th m illennia BC in the Ural region. Hence, the Urheimat o f the Uralic peoples and the Uralic languages is thought to have been located there (cf. pp. 93-94).
35
Introduction
Southern boundary of P ic e a { t$ n r ' >
„ Barents
European distribution of Pinus
sibirica (arotla pine; , - ■_
Southern, eastern and northern boundaries of Quercus (oak)
l i i l
Northern bbundary of the spread of fhe hedgehog 4
Lhk .S aint
Petersburg
M osco w
•,A„
В 1 “ С (,
Figure 4 Biogeographical features o f the Uralic Urheimat This m ethod hitherto enjoyed high popularity. Experim ents w ere m ade to locate the Urheimats o f the Indo-European peoples by exam ining the nam es o f fish species, a m ethod which the research o f the Finno-U grian languages adopted. It transpired, however, that fish and other anim al nam es are too readily transferred to similar, yet taxonom ically rem ote species. For a long
36
Methodological introduction and the sources
time it was thought that the Hungarian name for ‘b e e ’, me/?, and the spreading o f the m elliferous bee could afford some historical assum ptions. But, as it transpired, m any peoples had no separate nam e to distinguish ‘wild b e e ’ and ‘honey b ee’, yet the biogeographical distribution o f the two species differs greatly. The reconstruction o f the biogeography o f m any thousands o f years ago has undergone rapid developm ent over the past decades. For instance, today we can draw the northern boundaries o f the distribution o f koris ‘a s h ’, alma ‘apple’ and vadszolo ‘w oodbine’ a lot m ore precisely. All three H ungar ian plant names are o f Turkic origin, and we have exact know ledge about where the M agyars m ust have adopted the words, and w here these plants and their names could not be adopted. Naturally, such linguistic evidence is not enough in itself. A lm a ‘apple’ could easily have entered the language through trade, as did narancs ‘orange’. As a m atter o f fact, throughout the w orld the equivalents o f ‘apple’ originate from the diverse offshoots o f one very ancient name. The Indo-European, Turkic and Hungarian nam es for ‘apple’ are distantly related, and the word was known as early as the Hettite language o f Asia Minor. The Turkic word is o f Indo-European origin (see p. 192). The Turkic equivalents o f the Hungarian szold ‘grape’, ‘v ine’ are used to denote w oodbine everywhere, yet it seems m ore probable that, together with bor ‘w in e’ and m any other Hungarian vinicultural expressions, the w ord was adopted north o f the Caucasus, near the shores o f the Black Sea. Historical and ethnological evidence from Hungarian viniculture attests to this (see pp. I l l and 141). In this case, biogeography helps to determ ine the place rather than the date. However, the Hungarian szold typifies a Turkic sound change which cannot be very old, and certainly cannot have occurred prior to the end o f the 8th or the early 9th century. Given these facts, the m ethod affords con clusions regarding both location and time. A nother theory contends that languages have parts whose speed o f change is invariable. The glottochronological m ethod was propounded by Swadesh and his team. Swadesh held that the core vocabulary o f languages changed at an invariable speed. He collectively term ed “core vocabulary” the m ost im portant and m ost frequently used words o f a language, and picked from this assortm ent a hundred words, or rather m eanings, which he held to be the m ost essential. These included words such as ‘all’, ‘and’, ‘anim al’, ‘ashes’, ‘a t’, ‘b ack ’, ‘bad’, ‘b ark’, ‘because’, ‘b elly ’, ‘b ig ’, ‘b ite’, ‘b lack ’, ‘b lo w ’, ‘bo n e’, ‘breathe’, ‘b u rn ’, and ‘child’. He also set up control lists alongside his basic word lists. He then introduced the rule which he had inferred from the history o f the longest observable languages. According to the rule, 14% o f the “m ost fundam ental” one hundred words dropped out in a thousand years, or were replaced by new words. As the second m illennium passed by, 86% o f the leftover 86 w ords remained. W hat this means, in effect, is that given a com m on
Ъ1
introduction
language group which features a core vocabulary o f one hundred words, and from w hich younger languages peel off, a thousand years after breaking up, those languages will have preserved 8 6 % o f the com m on word-stock. H ow ever, there is no guarantee that all splinter groups will have “replaced” the same 14 words. The probability must, therefore, be exam ined o f the same word dropping out o f both languages, or from one or the other only. Swadesh set up a formula which produced the following result: The com m on “top 100” lists o f two languages will share 74% o f their words after 1000 years o f separation; 55% after 2000 years; 41%) after 3000 years; 30% after 4000 years; 22% after 5000 years; 16% after 6000 years; and 12% after 7000 years. Conversely, given a 30% identity o f the core vocabularies o f two related languages today, the above chart tells us that these language speakers sepa rated 4000 years ago, that is, their com m on original language splintered 4000 years ago. The m ethod provoked much interest and debate. Some claim ed that it was the C 14 m ethod o f linguistics. The m ethod was im proved, and researchers tested it for several languages. Thus, the separation o f the M agyar and Finnic languages, that is, the splintering o f the Finno-Ugrian proto-language, was dated 4000-5000 years ago, which surprisingly coincides with earlier results obtained from other methods. Nonetheless, the glottochronological m ethod did not prove to be successful, for two reasons. One is that it is im possible to draw a “top 100” list valid for each and every language. For instance, Sw adesh’s list included am ong the first twenty words the English at preposition, on the grounds, reasoned Swadesh, that it was the m ost com m on and m ost fundamental expression denoting location. However, the Finno-Ugrian languages, hence Hungarian, do not have prepositions, and express relations o f location with affixes (the Hungarian equivalent o f at is the affix -nal). Yet affixes can hardly be included in a vocabulary. The other reason is that Sw adesh’s definition o f core vocabulary tended to confuse the notion o f “m ost frequent” and “m ost fundam ental” . Usage frequency varies. By now, we have a great m any word occurrence dictionaries which have processed vast bodies o f texts. These tell us that the equivalent o f the word ‘m eat’ is am ong the first five hundred m ost frequently used words in Spanish, whereas, in French, it is m erely am ong the first three
38
M ethodological introduction and the sources
thousand. Also, the category “m ost fundam ental” is problem atic. M any lan guages do not use the conjunctive and for gram m atical co-ordination, there fore these languages either lack the word altogether, or assign it other functions. A nd other unforeseen problem s arose. In old G erm an (Gothic), the w ord denoting ‘m an’ was originally wafr which M ann superseded by process o f secondary change. However, this brought about secondary agreem ent with the English word man. The sam e thing happened with the words equivalent to ‘tru e’ and ‘to w ash’, both o f which are, however, included in the list. Very often the m eanings o f the m ost fundam ental words overlap, as, for instance, in the com pounds o f the Hungarian words kez ‘hand’ and kar ‘arm ’; sem an tically speaking,felkezii andfelka ru indistinguishably m ean ‘one-arm ed’. The w ord a scholar will add to the “top 10 0 ” list in such cases will be the one that is best suited for the specific scientific test. No linguist today w ould use the glottochronological method. Nevertheless, the scientific debate o f four dec ades has produced much im portant knowledge about linguistic change in general, the changes o f specific languages, and about the usability o f change descriptions for chronology. The negative outcom e o f the debate has also proved to be instructive, inasm uch that it confirm ed the speed o f linguistic change is never constant, not even in the seem ingly m ost stationary parts o f the core vocabulary. This is hardly surprising. The circum stances are im por tant. The drastic changes o f history can stim ulate language change, more peaceful conditions tend to slow down the process, w hereas certain types o f life style, such as nom adism , stabilise linguistic change. Although the speed o f language change is not invariable, know ledge o f the changes can nevertheless help in dating historical events. Here is an example. We know that in old Finno-Ugrian the initial k- changed to h- in Hungarian words containing back vowels (for example, in the Hungarian hab ‘foam ’, had ‘arm y’, hagym a ‘onion’, hal ‘fish’, haj ‘hair’, hall ‘to h ea r’, him ‘m ale’, ho ‘snow ’, hollo ‘raven’, husz ‘tw enty’, huz ‘to puli’, and m any other words). Seeing that in the m ajority o f Turkic loan words the w ord-initial k- has rem ained invariable (such as in kantar ‘bridle’, kanyaro ‘m easles’, кари ‘gate’, karam ‘(cattle)pen’, karvaly ‘sparrow -haw k’, kom lo ‘h o p ’ (the plant), кого ‘w eed ’, kos ‘ram ’ (the anim al), etc.) we are inclined to assum e that the M agyars could only have adopted them after the ancient Finno-U grian wordinitial k- changed to h-. But for that matter, the adopted Turkic /к/, too, m ight easily have changed to 1Ы. I f we knew when the shift from k- to h- ended, we w ould also be able to establish after w hat date the M agyars and the Turkic lived together. That, however, is very difficult to ascertain. The closest relatives o f Hungarian, Vogul and Ostyak, feature sim ilar changes, but not in every dialect. This strongly suggests that the к > h change in the Ob-U grian languages is m ore recent, and it is independent o f the same change in the
39
Introduction
Hungarian language. Further difficulties arise from the fact that there are quite a few Turkic words whose w ord-initial k- is equivalent to h- in the Hungarian, such as hajo ‘b o at’, hom ok ‘sand’, hitvdny ‘contem ptible’. For some time the same group o f words was thought to include hattyu ‘sw an’ and hod ‘b eav er’. Even distinguished scholars thought that these items entered the Hungarian prior to k- becom ing h-, on account o f a putative period o f very early M agyar-Turkic interaction which was followed by another sim ilar period later on. This conjecture could have contributed to establishing relative chronology. Unfortunately, however, the Turkic etym ology o f the w ords hattyu ‘sw an’ and h od ‘b eaver’ proved to be mistaken. But our loss on one side proved to be a gain on the other; as it transpired, the other words entered the Hungarian from one Turkic language— nam ely the prevailing language o f the K hazar Em pire— in which the same change had also previously occurred. This we know from the fact, am ong others, that a Khazarian river, for instance, was called H ara siu or H ara siv which in Turkic is K ara sub (or later K ara su), and the title o f the K hazar ruler was haghan or hakhan instead o f kaghan. The em erging к > h sound shift in the K hazar dialects, as attested by many sources, could not have begun earlier than the late 8 th or early 9th century. Thus, in Ibn Fadlan, the Arabic kadzi ‘cadi’ (judge) appears in the K hazar form o f hazi. Hence, this interaction at least can be dated with linguistic methods. Linguistically speaking, it can be very interesting w hen a language adopts the same word twice, and the two variants o f that w ord shed light on the two periods o f the source language. Hungarian is fortunate in this respect, since alongside karvaly ‘sparrow -haw k’, o f Turkic origin, another Hungarian bird name, herjd, w hich disappeared in the 19th century, happens to be a later variant o f the same Turkic word in which the к > h shift had already taken place. Although only a single variant o f the above quoted szolo ‘grape’, ‘vine’ entered the Hungarian, gyiimdlcs ‘fruit’ evolved from the derivation o f the very same Turkic stem, providing a fine parallel for the chronology o f kai-valy and herjd ‘sparrow -haw k’, gyiimdlcs ‘fruit’ and szolo ‘grape’, ‘v in e’.
NOTES 7. Tenninology, methods It is extremely difficult to translate the Hungarian term ostortenet. The word has a historically conditioned connotation. Literally it can be rendered both as ‘the history o f the ancestors (as ‘ancestor’) and as ‘prehistory, the time before history’. In m ost cases it denotes the ‘ori g in s’, the ‘history o f the origins’. The term ostortenet was not in favour for a long time, while beginning with the 1980s it has becem e a fashion. This is behind the fact that in som e cases
40
Methodological introduction a n d the sources
I have used proto-history, in others p rehistory for translating the Hungarian ostdrtenet. The issues o f ethnicity, people, nation and linguistic relationship, as w ell as the earlier literature, are discussed in R6na-Tas (1978a) which was an extended version o f my dissertation for the D Sc degree at the Hungarian Academy o f Sciences. Later I developed my view s and they have been summarised in Rona-Tas (1988a), a paper read before the Academ y o f Rhein-Westphalia; see also Rona-Tas (1989). M y view s were greatly influenced by the works o f Jeno Sziics who first outlined his ideas in Sziics (1966). H is ideas can be follow ed in Sziics (1981) in German, and in Sziics (1986) in French. H is book on the “three historical regions o f Europe” had a great impact on Hungarian historical and political thinking. The English version was published in 1983, the German in 1990. For his other works see the bibliography. The Hungarian K om oroczy discussed the concept o f cthnicity and nation in K om oroczy (1992). Packed with important conclusions, his studies unfortunately disregard the differences between ethnos and nation. See also E. Canetti’s (1977) autobiographical novel. The full Hungarian translation o f Mendander Protector’s text is available in Lukinich (1905, pp. 5 9 -6 0 ), which I adopted with minor m odifications in R6na-Tas (1989, p. 9 = Rona-Tas 1995a, p. 175). The famous place is also cited by Pohl (1988, p. 21). The Greek original is available in D e B oor’s edition E xcerpta (1903, pp. 170-171). Wenskus (1961) is a fundamental work in the field. H. Wolfram continued the work Wenskus had begun, see, for instance, Wolfram (1979b, 2nd edition 1980), as w ell as the publications o f the Vienna scientific circle, hence W olfram -Daim (1980); W olfram - Schwarcz (1985); Friesinger-Daim (1985); W olfram-Pohl (1990); Friesinger-D aim (1990). For W olf ram’s article on the types o f ethnogenesis see Beumann-SchrOder (1985, pp. 9 7 -1 5 2 ). For Pohl’s article on the types o f ethnogenesis see W olfram -Pohl (1990, pp. 113-124). For further research consult E ntstehung (1985); Girtler (1982). The works o f som e German scholars o f this field are published in Studien zu r E thnogenese I—II, 1985, 1988. O f the articles published in these volum es, special mention must be made o f the study o f W.E. MOhlmann (S tu d ien 1 , 1985, pp. 9 -2 8 ); J. Untermann (Studien I, pp. 13 3 -1 6 4 ) discussed the relations o f language and ethnic history. The same publication includes R6na-Tas (1988a, pp. 1 0 7 -1 4 2 = Rona-Tas 1995a, pp. 2 7 5 -3 1 0 ). Although not fully devoid o f ideological constraints, van den Berghe (1987) gives a good overview o f the issues o f ethnicity. Bromlej (1976) is outdated, but interesting. Anderson (1990) represents a modern trend in nation research. He claim s that modern com munication technology has significantly contributed to the em ergence o f nations. See also G ellner (1987). The latest overview available to me on the issues o f nation and nationalism w as Greenfield (1992), the topic is, however, inexhaustible and the bibliography endless.
2. Chronology and chronological assumptions and
3. The role of the natural sciences in determining age I wrote these chapters primarily drawing on the studies contained in H ajdii-K rist6-Rona-Tas (volum e IV), and especially the chapters written by Mrs M agda K om l6s JSrai, L4szl6 Kordos, Sandor Som ogyi, Janos Tardy and Peter Marton. I would hereby like to take the opportunity to thank these authors for their contribution. For a summary see Rona-Tas (1979). I am aware that research has progressed in the last decades. For an up-to-date natural geographical outlook on
Introduction
41
the Carpathian Basin and the EtelkOz, see GyOrfFy-Zolyomi (1994, pp. 3 4 -3 7 ) which also has an ample bibliography. Consult Rona-Tas (1978a) on linguistics as an ancillary discipline used for establishing chronology. For the latest literature in historical linguistics, a review o f research, and a bibliography, see Bynon (1977) and Anttila (1989). Sociolinguistics lends fresh help to historical linguistics and here the works o f Labov are o f eminent importance. There exist a handful o f university textbooks on sociolinguistics and most deal with historical changes as well. See Labov (1994). The spreading o f coniferous and deciduous trees, and the questions o f the U rheim at o f the Uralic peoples have been discussed in detail by Hajdii (1971, 1976), H ajdu-D om okos (1978, pp. 4 5 -5 7 ; 1987, pp. 2 7 3 -2 9 9 ). The Swadesh method was first propounded in Swadesh (1950); the first decade o f the method was reviewed by H ym es (1960). Literature and Hungarian opinions are available in Fodor (1961) and Rona-Tas (1978a, pp. 2 4 3 -2 5 1 ). The questions o f к > h chronology were first discussed by Barczi (1 9 5 2 ,1 9 6 5 ,1 9 7 1 ,1 9 7 2 ). See also Ligeti (1986, pp. 2 7 -2 8 ), and Rona-Tas (1988d) and below (p. 106).
11. THE SOURCES
1. THE CONCEPT OF SOURCE MATERIAL In this work the concept o f “source m aterial” is broader than usual. This expanded concept was first em ployed in 1976 in the volum es o f the Szeged Proto-History Research Group. We consider as sources, all factual material from which inform ation can be obtained, directly or indirectly, regarding the ancient history o f the M agyars. Since written m aterial on the period is relatively scarce, naturally the significance o f other sources is greater. Besides written sources we have used language itself as a historical source, and also archaeological research, physical anthropology, ethnography and num erous other scientific methods. This complex or interdisciplinary approach was always utilised in research. One great problem is that the scope o f some researchers— due to the rapid “expansion” apparent in these sciences— sim ply cannot cover all the different source fields. The use o f incom plete, old or simply out-of-date results from neighbouring branches o f science also poses a threat. Sometim es we experience very specific selections. Am ong the results o f related sciences there is a tendency to withhold those findings w hich either do not agree or could only w ith difficulty be reconciled with individual “pet” theories. At other times we can observe that type o f indirect p ro o f which is m erely m entioned som ewhere as a potential hypothesis but w hich is accepted as p ro o f in another science, and as such, theories are then built on the back o f it. A fter this, the person who put forward the idea as a hypothesis in the first place then quotes the findings o f his colleagues as proving his idea. It is not possible to com pletely elim inate these difficulties as they go hand in hand with any study o f a com plex science such as this. Even so, every effort m ust be made to avoid these pitfalls. This is only possible if one has a good under standing o f the nature o f the sources, and their limitations, and thus in our work we endeavour to speak not only about the evidence provided by the sources but also about the lim itations o f these sources, and in every case we com plete a source critique. Naturally this does not m ean that every person, in every case, has to start everything right from the beginning, but a sober critical approach always provides greater safety.
44
Methodological introduction and the sources
2. SOURCE CRITICISM Every source type and every individual source dem ands a different critical approach, and there are certain general principles w hich are equally valid for large source groups. Let us take a sim ple historical event, for instance the rise to pow er o f a ruler. The event itself occurs in a tem poral frame, and it is by no m eans certain that the entire process will be observed, rather perhaps only highly sym bolic events (election, coronation, etc.). The event is recorded, the record passes into the written source, the written source is com pleted, the author finishes or m aybe edits his work, and then later he m ay even transcribe or enlarge the docum ent. A fter com pletion, and som etim es w ithin the lifetim e o f the author and under his supervision, it is copied, the final w ork is then later still copied again, w hile later writers take extracts or quote from it. These nine stages describe m erely the sim plest o f cases. It is not uncom m on to find corrections, om issions and later additions, etc. m ade to details in the source. The m ajority o f copies are m ade long after com pletion o f the original, and thus it is necessary to reconstruct retrospectively the original text from the copies. Occasionally, to decide what exactly was in the original, it is necessary to com pare the copies and their provenance. And we are yet to m ention forgeries, errors and m istakes, w hether intentional or from sheer ignorance. C learing up all these presum es an understanding not only o f the language o f the sources but o f the writing, the specific orthography and abbreviations as well. Naturally, historians m ust often rely upon those who professionally edited the sources, because they are sim ply unable to com plete source critiques in every case. However, an effort m ust be m ade to use critiques, and if the historian is unable to com plete such, it m ust be clear exactly w hat problem s the source brings with it. Source criticism cannot only enquire into the understanding o f the language and writing, the palaeography o f the source, but the publisher m ust be in possession o f those facts upon w hich unclear, problem atic aspects can be interpreted and explained. W ith the advances being m ade by researchers, science is constantly being enriched by ever new er discoveries, and for this reason source critiques should be continually updated for some tim e to come, even if the original w ork was com pleted in a fully professional m anner in its time. Source interpretations m ust be separated from critiques, and this is true even if the task is not always straightforward. The interpretation o f source infor m ation, o f course, influences (or can have an influence on) the critique. However, it is m ethodologically pertinent to differentiate betw een critique and interpretation. Unfortunately, one has to note that am ong the w ritten sources covering ancient Hungarian history there are only very few where the critical analysis
45
The sources
has been carried out to a satisfactory and m odern standard. A notable exception is the publication o f the D e adm inistrando imperio. Sources can be interpreted in a m ultitude o f w ays. In the sources them selves, the quantity and quality o f inform ation directly related to the M agyars, as well as data which can be linked to this topic, are restricted. For the period prior to the Conquest we have sufficient sources relating to the neighbours o f the M agyars to be able to reconstruct the history o f the area subsequently also settled by the M agyars. In the following we attem pt to sketch out an Eastern European history in w hich the pre-Conquest M agyars undoubtedly had a part. Thus we have only taken into consideration w hat the sources have to say about the area bounded by the Eastern European steppe and forested steppe belt, the Ural m ountains and river, the region defined by the Caucasus, the Black Sea and the eastern side o f the Carpathians, and those details relating to the Balkan region. Those sources relating to the state o f the Carpathian B asin prior to the Conquest, respectively to the Conquest itself, constitute a separate group.
3. THE WRITTEN SOURCES The b rief outline o f written sources is presented according to the language o f the source, and if necessary, within this subdivided according to the type o f writing. In the survey o f sources, we generally do not go back earlier than the 5 th, and in some cases the 6 th century BC. The works o f early writers frequently alluded to such sources w hich possibly could be related to the M agyars or their predecessors, but from our point o f view these can be disregarded. The Yiirka people m entioned by Herodotus (d. 425 B C ) are generally related to the Ugor people and the ancestors o f the M agyars on account o f the sim ilar sound o f the nam es and their location. Behind the nam es o f peoples m entioned in different steppe descriptions, others claim to see the M agyars or peoples related to the M agyars. We will not deal with these m atters here. However, it is im portant to m ention Ptolem y who lived in ancient A lexandria and who put together a collection o f m aps from earlier w ork by M arinos o f Tyros com piled betw een 107 and 114 A D , bringing them together in an eight-volum e cosm og raphy. The collection was com pleted betw een 151 and 178 A D , and although there is no direct M agyar connection, the nam e o f the Finnish people appears here for the first time. The Ptolem y work served as a basis for later geographi cal works, and the Arabs also adopted it. A lthough the structure and the descriptive principles o f the original work were slowly m odified, it still fundam entally defined geographical studies right through to the end o f the M iddle Ages, and as such, also those works which are im portant from our
46
Methodological introduction and the sources
point o f view. The other early sources are only im portant because in many cases quotations taken from them colour, but o f course also corrupt, w orks by later writers.
a) Byzantine sources The Byzantine Em pire was established on the basis o f Rom an law, in a Greek-speaking area and professing C hristianity as a state religion. Dual rule o f the Greek-Rom an Em pire lasted for some time. A fter the downfall o f the West Rom an Em pire (476), Byzantium becam e fully independent. In 740 it halted the advancing Arabs, but it lost Ravenna in 751. In 843 the civil war betw een the iconoclasts and iconodules (the Iconoclastic C ontroversy) was finally term inated, a struggle which had considerably w eakened the B yzantine Empire. The year 867 m arked the break from the Rom an C hurch (for details see pp. 257-260), although the date o f official rupture was 1054. Leo the Wise took to the throne in 8 8 6 , and the B yzantine Em pire underw ent a resurgence. The southern shore o f the Black Sea, today’s Anatolia, and the Balkans to the D anube w ere under the sphere o f influence o f Byzantium . Im portant trading towns were located in the Crim ean Peninsula under B yzantine rule. To the east, the em pire bordered on Sasanian Persia, and betw een these tw o worlds lived such peoples as the Arm enians, Georgians and Syrians. Efficient protec tion o f the Byzantine E m pire’s northern borders and trade routes dem anded inform ation about happenings in the steppe. Frequently, protection m eant form ing an alliance w ith one group or people against another. In order to do this it was necessary to have a good grasp o f the history, pow er relations, location, m ilitary strength and tactics o f these peoples. The m ost im portant genres in B yzantine historical literature are w ritten works covering the history o f the empire, ecclesiastical historical w orks and route itineraries. But we also find historical references in such genres as courtly verses praising the em perors, the panegyrics. Byzantine works were w ritten in M iddle Greek. By this tim e the reading o f characters and character clusters had deviated from that o f the classical Greek period. A t the sam e time, however, w riters with a predilection for the use o f contem porary syntax and gram m ar em ployed classical rules as w ell, and these w ere all m ixed in num erous w orks (see Figure 7 on page 50). To present an exam ple o f ju st the sort o f variable results a single source can provide, it is worth taking a b rief look at the w ork entitled The G overning o f Empire. Am ong those sources that the author, Em peror Constantine VII Porphyrogenitus (913-959), used were original texts, reports, accounts and already com pleted sum m aries w hich were extracted for him. These w ere transcribed
ш<
W <~>
Figure 5 The Roman Empire under Augustus (63 BC - 14 AD)
if ? ! , ! *
D
О Э) <
ТФЗ тз с и
а.
о
.2
ф
££
а
CL
О
Figure 6 The Byzantine Empire under Justinian (527-565)
The sources
49
and copied for the emperor, who generally dictated or som etim es wrote him self using these sources. Thus w ere those parts com pleted w hich w ere then originally brought together in two works with the titles On Peoples and On Provinces. These works, m ay have been com pleted around 945, but im m edi ately afterwards the em peror re-edited the whole work, and the final version was only ready by around 952. The court scribe, or scribes, then m ade copies o f the work. These have been lost, but a new copy was m ade around 979. This was lost too, but some time between 1059 and 1081 yet another copy was made, and luckily this has survived. C urrently it is stored in Paris. Am ong its many readers there were those who added their own rem arks, those who m ade corrections, and those who m ade notes in the margin o f the codex. Occasional unknown or indecipherable words were supplem ented or corrected. Two copies were made o f the Paris document. There are notes and corrections which date from before the copies were m ade, while others date from later. A copy stored in the Vatican was com pleted in 1509, and one in M odena between 1560 and 1568. N ot long after the Vatican m anuscript was com pleted, a new copy o f it was made between 1509 and 1529. The work was first published by M eursius in 1611 on the basis o f the Vatican m anuscript, and then in 1711 by Bandur, who knew the Paris m anuscript and was able to read those parts which are unreadable today. B ekker’s 1840 edition is based on B andur’s text. The critical edition o f the text by M oravcsik and Jenkins is based on an understanding o f this background to the material, and it endeavoured to reconstruct the m ost accurate picture possible o f the original text. A ssistance was provided by the fact that The Governing o f Em pire contains quotations from earlier written works. Am ong these, we know o f some which derive from other textual traditions, and thus it is possible to establish how copies were made during the imperial age. But naturally, some parts o f The Governing o f Empire from around 952 were o f a quite variable quality, there are parts edited precisely and parts negligently, some with a great deal o f attention and others with less care. In approaching such a text, there are two extremes. The excessive critic cannot, in the end, accept any parts o f the w ork as authentic. On the other hand the naive reader considers every letter, every assertion noted down, as the absolute truth. At the same time it is not ju st a m atter o f finding a m iddle way, but rather to back up reasoned criticism with philological argument. We should not blithely correct the text as we please, but then again w e cannot accept it as fact carved in stone and com pletely unalterable. The basic principle rem ains the principle o f the least possible change. Any alteration m ust always be in harmony with the entire work, its spirit and details, and furtherm ore with those facts deriving from other sour&es.
50
Methodological introduction and the sources
Greek character
Greek cluster
a
Transliteration
Transcription accord
according
ing to M iddle Greek
to classical Greek
pronunciation
a
a
ai
ai
e
Ott)
au
au, av
P
b
v
У
g
g
gg g + i,e
ng y+i, e
ук
gk
nk
U
gkh
nh
5
d
d
e
e
e
YY fr \, E
ei
ei
i
EV
eu
eu, ev z
С
z
л
с
ё
6 i к X
th
th
n v vt
S о
71 P
i
i
k
k
1
1
m
m
mp
mb, b
n
n
nt
nd
X
ks
0 01
oi
0 1
ou
ou
u
P r
P r
s
s
a.S T u
t
t
ii
i
ui
i
Ф
ph
f
X V
kh
h
ps
ps
CO
6
о
Figure 7 Transliteration and transcription o f G reek characters and clusters
The sources
51
The appearance o f the Huns was dealt with by several B yzantine writers, the m ost im portant being Priskos the Rhetor born betw een 410 and 420 AD. He accom panied senior am bassadors calling on the court o f Attila in 448^449, com piling in eight books the events o f the period from both his own notes and other sources. The w ork in all likelihood covered the period betw een 411 and 472. He copied up notes on his trip to the Huns im m ediately upon his return home, and slotted in the description to his work, finally com pleted 23 years later. The original w ork is lost, and we only know about it through extracts. Several m ade abridgem ents, for instance Cassiodorus who died around 580, from whose work these parts are only known to us from the quotations o f Jordanes. The main source, sum m aries prepared by C onstantine, dates from around 950. Copies o f the sum m aries that we know about are no earlier than from the 16th century. Follow ing the disintegration o f the Hun Empire, the first references ap peared in Byzantine literature to definite Turkic-speaking peoples, among them the Onogurs. From among the sources covering the period between the appearance o f the Turk peoples and the Avars, it is worth m entioning the cosm ography com piled around 540 by world traveller Cosmas Indicopleustes. The oldest known copies o f his work can be dated to the 8 th or 9th century. The rem aining part o f the 18-volume world chronicle o f John Malalas, in the existing copy, ends at the year 563. The author probably died in 578, and so it is possible that the lost parts contained further chapters. M any have abridged this work. One m anuscript contains the com plete rem aining text, this being an 1 lth-century copy o f contem porary extracts from the original. Q uotations from the work are still with us in the works o f num erous other writers. One o f the m ost significant historians o f the age was Procopius; we know o f three works by him. The eight-volum e H istory detailing the wars o f Justinian could have been com pleted around 553. His second work later becam e know n as the Anecdota, or the Secret H istory, while the third work gave an account o f im perial constructions. The earliest known copy o f the historical work— with the exception o f one part from the 13th century— originates from the 14th century. A gathias continued the work o f Procopios, treating the period be tween 552 and 558, and noting the appearance o f the Avars. The earliest known copy o f this w ork dates from the 10 th to 1 1th centuries. The appearance o f the Turk peoples and the Avars only elicited greater attention in writers o f the follow ing period. A ten-volum e history by Theophanes Bizantius, covering the years between 566 and 581, survived only in a few short extracts. The work by Protector M enander follow ed events until 582. Similarly, his w ork did not survive the passing centuries, but ex tracts (copies from the 16th century) by later B yzantine w riters, Theophilactus Sim ocattes and Constantine, preserved m uch o f the work.
52
Methodological introduction and the sources
M enander’s w ork was continued by Theophylactus Sim ocattes, in whose book events w ere follow ed until 602, and the death o f Em peror M aurice. The oldest extant copy dates from the 11 th—12th centuries. U nder M aurice a w ork on m ilitary tactics (Strategicon) was prepared, including notes on nom adic warfare. A lthough the w ork included m any elem ents taken from classical G reek tradition, such as com m onplaces on the Scythians, the text also included notes on the Turk peoples. C ertain inform ation reflects the position in the last decades o f the 6 th century. Besides extracts, the w ork survived in several copies. The oldest known copy has been dated to the 10th century. During the rule o f H eraclius (610-641) B yzantium developed close ties with the Avars and various Turkic peoples. In the constant cam paigns w aged against Persia, the em pire took advantage o f m ilitary assistance from steppe dwellers. Heraclius replied to a com bined attack m ade up o f an alliance o f Persians, Avars and Slavs by going on the offensive, and in 627, with help from the Khazars, he inflicted a decisive defeat on the Sasanian Persian Empire. A w ork by an unknow n writer, know n as the E aster Chronicle because it also gave details on calculating Easter, dealt w ith the period. The chronicle introduces earlier periods on the basis o f w orks by other authors, although it contains independent inform ation regarding the contem porary period up to 628. The earliest existing m anuscript dates from the 10th century. The Avars are m entioned in a book o f serm ons by Theodorus Sinkellos originating from around 627 (10 t h - 11 th-century copy) and in a history in verse by Georgius Pisides (14th-century copy, including extracts by other authors). The Introduction, com pleted before 715, w ritten for the A gathon alludes to historical events surrounding the Khazars and the B ulghars driving into the Balkans. The work, docum enting the life o f the patron saint o f Thessaloniki, Saint D em eter (m artyred before 305), also contains num erous later historical facts. The part covering 7th-century historical events can be dated to the end o f the 7th century (copy from the 12th century). B yzantine sources list contem porary sees, such as the one o f the C rim ean m etropolitan, several times. We know o f such a list from the 8 th century (betw een 733 and 746, copy from the 14th century) and from the beginning o f the 9 th century (betw een 805 and 815). Theophanes w ho died in 818 wrote his highly influential historical w ork C hronographia betw een 810 and 814, w hich he com piled in part by incorpo rating extracts o f the works o f predecessors, w hile he gathered inform ation partly from his contem porary writers and partly eyew itness accounts for those chapters connected to the 7th and 8 th centuries. The w ork w as translated into Latin som e tim e betw een 873 and 875 (copy from the 10th century). Broadly speaking, Patriarch N icephorus (d. 829) w orked from the sam e sources. His
The sources
53
short history covers the period betw een 602 and 769 (copy from the 12th—13th centuries). W riters o f the following period w ere able to report on the events leading up to the Conquest and provide inform ation related to the Conquest. These works were com pleted under two em perors, Leo the Wise (886-912) and Constantine Porphyrogenitus (913-959). Philotheas, who wrote a handbook on the proto col o f the Byzantine court in 899, and Arethas, author o f a speech praising Leo the Wise in 902, were the first to m ention the M agyars under the nam e Turki (Turkoi), that is Turk. Around this time, the M agyars w ere frequently m en tioned in b rief and by different names. A few w orks on strategy also probably term ed the M agyars as Turk. But Leon D iakonus m entioned the M agyars as Huns and Scythians in his work originating from around 992, and w hich covers events between 959 and 976 (12th-century copy). M eanw hile, in connection with the Italian raids the Chronicle o f the Popes (copy from the 13 th century), docum enting the period between 891 and 929, m entions in the events o f 922 the M agyars under the nam e Ungri, term inology also em ployed in the work M iracles o f Saint George and in an 11 th-century surviving copy o f the L ife o f Saint B asil (Basileios, d. 944). A treatise entitled Tactics by Leo the Wise com pleted around 904 used as its main source the sim ilar work by M aurice, as well as experiences from the em peror’s own time. Leo the Wise m ade additions in at least six places to the M aurice w ork relating to the M agyars. The w ork refers not only to M agyar warfare but also to their way o f living, their role in the B ulghar-B yzantine w ar (894-896), as well as to the alliance with them. The earliest copy dates from the 10 th century. Nicholas M isticus, patriarch o f C onstantinople who died in 925, referred to the M agyars as “Western Turk peoples” in a letter to the B ulghar ruler Simeon ( 10 th-century copy). The m ost im portant Byzantine source regarding the conquering M agyars is a w ork by Constantine Porphyrogenitus. He received the nam e Porphyrogeni tus (B om in Purple) after the purple m arble cham ber where the m le r’s first-born son, heir to the throne, was bom . However, such a nam e w as given to others bom heir to the throne as well. He was only seven w hen his father died (912), and thus first his uncle, and then follow ing his u ncle’s death the admiral o f the fleet, Rom anus Lecapenus, took over the reins o f power. Constantine only exercised true pow er from 944 after R om anus Lecapenus was sent into exile. In the intervening period Constantine retreated from public life and devoted him self to the sciences. He w orked together w ith a large staff who collected and abridged works o f earlier writers. This collection o f extracts preserved the works o f num erous historical chroniclers. C onstantine him self was a prolific writer, and it is*1difficult to say which, from am ong the many
54
Methodological introduction and the sources
works, he actually wrote himself. In any case these works survived under the nam e o f the emperor. As far as we are concerned the m ost im portant work was originally unnam ed; a Greek inscription revealed that the author presented the work to his son, who was then already an emperor. The w ork was given the title D e adm inistrando imperio by M eursius, the first publisher, in 1611. Since then, this w ork in Greek is com m only referred to by its Latin name. The work took its final form in 951-952, and in all likelihood Constantine presented it to his son in 952. Originally each part was written separately and at a different time. Based on internal dating, it is possible to ascertain that it was w ritten starting from 948, furtherm ore that certain parts were com pleted indepen dently o f one another. It appears that originally Constantine wished to write a large encyclopaedic sum m ary o f peoples (peri ethnon), but he later changed his m ind in favour o f putting together a scholarly collection dedicated to his son. The m ajority o f those parts written to m eet the original intent were retained in the work, and it is worth noting that they did not always suit the later structure. Constantine strove for exactness in his use o f sources, and where there has been an opportunity to check back, it is possible to establish that he follow ed earlier works very precisely. He not only em ployed written sources, but reports o f em issaries and eyewitness accounts as well, and it is evident that som e o f these were actually noted down by his own scribes. Those sections concerning the M agyars are derived from several sources. A m ong these were two M agyars, Termecsii and Bulcsu, who, obviously with the help o f an interpreter, gave answers to questions posed about their people. It appears as though these were highly specific questions. At the end o f C hapter 40, six sentences begin with ‘it is know n’ ( ’isteon). These were answ ers given to such questions as “who is who in H ungary” . It is quite possible that there were other records on the M agyars in the im perial court, and not everything came from the same source. The text o f D e adm inistrando im perio— as we have seen above— only survives to this day in an 11 th-century B yzantine copy. Porphyrogenitus’s other project was in reality a collection o f w orks, also generally known by the Latin nam e given by the publisher: D e cerim oniis aulae Byzantinae. The work docum ents court cerem onies, but it also covers num erous other topics as well. A fter thorough exam ination it becam e evident that the available original 12 th-century m anuscript cannot be the work o f Constantine Porphyrogenitus, as his burial place is also m entioned w ithin, and it cannot be a direct copy o f the original work either. On the other hand, other parts o f the work contain the introductory words to the em peror’s son. Georgius M onachus wrote a World Chronicle on the history o f the world known by him until the year 842. This w ork was then continued by an unknown writer. It closes with the death o f Rom anus Lecapenus, and it w ould
The sources
55
appear that it was som ewhat hostile towards Constantine, who was at that time rather ignored. This is im portant because it is likely the w riter was not from among the group surrounding Constantine. Two com pilations o f the Continu ation o f the Chronicle by George the M onk ( Georgius m onachus continuatus) survived. Version В is later and som ewhat m ore extensive than version A. Both have 11th-century copies. There were also continuations o f the original Theophanes w orld chronicle extending to 813. These writers divided the works into six volum es covering events up to 961 (Theophanes continuatus). According to the title it was prepared for Constantine Porphyrogenitus, who also provided sources for the compilation. The m ain source for the part covering the period 886-948 was the above-m entioned Continuation o f the Chronicle by George the M onk version B. The Vatican has a copy dating from the 11th century. Later B yzantine works only supply occasional references to the history o f the conquering M agyars. A few works copied or abridged some earlier sources as well, the m ajority o f which are im portant sources o f the history o f Hungary in the age o f the Arpad dynasty.
b) Latin sources Sources from outside Hungary Ancient Latin sources do not contain any inform ation that we can use. There are occasional references to the names o f Finnish and Estonian peoples. The Yiirka people found by Herodotus were absorbed into m aterial, although in places they appear in the form Turcae, but these sources are interesting only from the point o f view that in later sources containing m aterial interesting from our aspect there are many quotations taken from the ancient classics. The earliest author writing in Latin— and worth citing here— was the Rom an w riter o f Gothic origin Jordanes, who com pleted his w ork (De origine actibusque Getarum, or in b rief Getica) in the m iddle o f the 6 th century. His work is certainly an adaptation o f the history o f the Goths by Cassiodorus. The now lost work o f Cassiodorus covered the history o f the G oths until 533, followed by Jordanes’s own work for the period up until 551. The latter’s Latin did not always m eet the norm s o f classical grammar. This work, w hich is also im portant as regards the relation between the Huns and the Goths, was frequently used by later literary, thus Hungarian-related sources as well. The history o f the sword o f Attila also derives from Jordanes. A near contem porary copy has survived.
56
Methodological introduction and the sources
A 7th-century w ork by an unknown writer, usually quoted as being entitled the Exordia Scythica, is im portant for us because a later copy was one o f the sources em ployed by the Hungarian chronicler Anonym us. The w ork is actually an abridgem ent o f a 2nd-century work, strongly influenced by H ero dotus. The Cosmographia was possibly created around 700, in Ravenna, edited by an unknow n w riter (Geographus Ravennas). A m onk from Pisa took an extract from the w ork in 1118. The work includes references to O nogoria and the Khazars. A t the end o f the 8 th century we find a work w ritten under the nom -deplum e o f A ethicus (Ethicus) Ister. The work, a fantastic description o f the w orld (Cosmographia), m entions the Turk people living in the Black Sea region. A ppended to the w ork is an alphabet, about w hich the first publishers stated that it contained invented characters. In attem pting to refute this, several experim ents were carried out unsuccessfully, which aim ed to show a connec tion with Turkic or Avar runiform script. The end o f the M erovingian dynasty was signalled in 751 with the crowning o f Pepin the Short as “king o f all the Franks” . U nder the rule o f his son, Charles the Great (Charlem agne) (768-814, em peror from 800), the new C arolingian Em pire began to flourish, and ever closer ties were established with the Avar Empire. Thus, sources from this time have im portant inform ation to tell us about the eve o f the Conquest and details related to the C onquest period. On the orders o f Aldwin, archbishop o f Salzburg (859-873), an unknown m onk wrote the Conversion o f the Bavarians and Carantans (Conversio bagoariorum et carantanorum) in 870. The purpose o f the w ork was to support the Salzburg see’s legal claim to Lower Pannonia. The work, w hich analyses the m issionary activities o f the Salzburg see, contains m uch im portant infor m ation about the history o f Pannonia prior to the Conquest. The nine m anu script copies w hich are available to us today date from the 12th century. An early extract was also prepared from the Conversio at the turn o f the 12 t h - 13 th centuries (Excerptum de karentanis) in w hich there are a few w ords which do not originate from the Conversio. Am ong other sources it is im portant to highlight the A nnales iuvavenses m axim i, in the original text o f which there is m ention o f a jo in t attack launched by the M agyars and Khavars around Vienna in 881. Theotmar, archbishop o f Salzburg (873-907), fell in battle fighting the M agyars at Pozsony. A letter (long regarded as a fake) w ritten to Pope John IX in the spring o f 900, contained im portant inform ation not only about argum ents raging w ith the see o f Passau, but also relating to the history o f the Avars, the Slavs and the Magyars.
The sources
57
Between 892 and 899, Regino (d. 915) acted as abbot o f the Priim m onastery situated 100 km w est o f Koblenz. He resided in Trier from 899 until his death. Here he com pleted his World Chronicle, the last year being 906. The chronicle docum ents events following the death o f Charles the G reat on the basis o f verbal statem ents and contem porary sources. The M agyars are m entioned for the first tim e in the year 889, and here the author sum m arised everything he knew about the conquering M agyars. Copies o f this work date from the 10th century. Later writers frequently took extracts. Am ong the m any m onastic annals, that o f the Fulda m onastery is critically important. Notes w ritten year after year were occasionally collected together in “instalm ents” . The last instalm ent o f the annals— which covered events from 680 to 901— and in which we can find details related to the M agyars, was not actually prepared in Fulda but in Bavaria. After this cam e the western Latin sources, authorities on the age o f incur sions. However, a few writers used the surprise caused by the sudden appear ance o f the M agyars to portray these new com ers, and to record earlier events. Am ong the resultant w orks the one by Liudprand who died in 972 as bishop o f Crem ona stands out as particularly significant. As m inister o f B erengar II, king o f northern Italy, Liudprand travelled to Byzantium in 949 and m et C onstantine Porphyrogenitus as well. Three years after the battle o f the Lechfeld he m oved from Italy to the court o f King Otto, where he wrote Antapodosis (Repaym ent) (i.e. in answer to slights he suffered at the court o f Berengar). There are details o f raids by M agyars in Italy, but references are also m ade to earlier periods. In addition, m ore inform ation is given about the raids in two other works (Historia Ottonis and Relatio de legatione Constantinopolitana), 10 th-century copies o f w hich have survived. Between 1077 and 1079, Lam pert o f H ersfeld (1025-1081) w rote about the history o f the Hersfeld m onastery in his annals. He noted that in 1071 King Henry IV visited the monastery, and on his departure one o f his captains, Leopold o f M erseburg, fell aw kwardly from his horse in such a w ay as to die from injuries he sustained from his own sword. The king returned to H ersfeld and granted the m onastery 30 villages so that the m onks w ould pray for the salvation o f the soul o f Leopold. In relation to the incident, Lam pert notes that the sword w hich caused the accident was the very sword w hich Anastasia, wife o f the Hungarian King Andrew I and m other o f Salam on, originally presented as being the sword o f Attila to the Bavarian prince Otto in return for helping her son to the throne in 1063. Lam pert him self travelled to Hungary in 1058, learning at first hand about circum stances there. M any charters originated from the Frankish Em pire and the eastern border region (Ostm ark). Am ong them , several are directly related to the history o f the Carpathian Basin b efo reth e Conquest or at the time o f the Conquest. In a
58
Methodological introduction and the sources
few rare cases the original docum ent itself has survived, for instance the deed o f foundation o f the abbey o f Krems from 777. In this w e can find the first reference to the office o f zhupan (zsupari) in the form jo p a n . The charter o f the king o f the eastern Franks, Louis o f Germany, dated 8 May, 860 in Regensburg, granted land to the m onastery next to M attsee. In the charter one o f the borders is denoted as uuangariorum marcha, “the m arches o f the Vangaros”, w hich we will com e across again in the discussion on the ungri people (see p. 285). Besides the charters, other types o f sources can also be o f assistance. Some tim e before 923 a letter in Latin was despatched to Dado, bishop o f Verdun, in w hich the author w rote o f the origins and nam es o f the M agyars who were then present in the area. We will quote from this letter later (see p. 282). We will also refer to m onastery registers in connection with the ungri p eople’s nam e (see p. 285).
Hungarian sources The Latin-language sources written in H ungary date from after the Conquest. We do not have the opportunity here to provide even a b rief survey o f the two main groups, narrative sources and charters. From our point o f view, we only have to note the following. It is certain that the com pilation o f a chronicle was started in the royal court o f the House o f A rpad in the 11th century, based on w estern models. W hat was im portant in the so-called ancient chronicle was— underlining the acclim atisation o f the M agyars in Europe— the reliance on w estern sources. Authors and writers o f w estern chronicles frequently travelled to H ungary in the 11 th century, in the process transm itting exam ples and sources. C ourt clerics travelled abroad, becom ing w ell versed in the chronicles prepared in the courts o f foreign rulers, docum ents w hich also served to legitim ise claims. This should not be taken as m eaning that early Hungarian chronicles were com pletely o f w estern origin, but ju st that they were prepared on the basis o f w estern patterns and styles. Text-wise, extant or known chronicles can, broadly speaking, be split into two groups. In essence the first group com prises A nonym us’s Gesta. A nonymus used a version o f the ancient gesta. To a certain extent, it was independent o f other chronicle traditions. The latter traditions lead on the one hand to the chronicles o f Simon de К ёга w ritten betw een 1282 and 1285 through a version w hich although lost can in part be reconstructed, and on the other hand to the Illum inated Chronicle w ritten in 1358. Besides the Illum inated Chronicle and continuations, the B uda Chronicle o f 1473 has preserved details w hich are not to be found in the chronicle o f de K eza or the Illum inated Chronicle tradition.
The sources
59
Furtherm ore, this chronicle branch has secondary branches: versions in which we occasionally find specific events which cannot be found elsewhere. At this point we should therefore briefly m ention the w ork o f the notary to King Bela III, nam ed in the chronicle as M aster P., but generally known as Anonymus. Today researchers agree that the author o f the Gesta was notary to King Bela III, and the work was w ritten a few years after the death o f Bela III, that is after 23 April 1196. Thus the Gesta dates from the last years o f the 12th century. In style it is a historical rom ance, allowing m any o f the leading nobility o f the age to play a role. Anonym us was able to com bine on the one hand noble family traditions, and on the other historical legends connected to place nam es or made up by the author. The structure on which he strung together his stories m ust have been that o f a copy o f the lost ancient chronicle. From this he linked Almos, Arpad and his dynasty to Attila. A nonym us lived some three hundred years after the Conquest, and that is why we have learned more about the conquering M agyars from contem porary western sources than from Hungarian. Even so, A nonym us rem ains a prim ary source for inform ation on the language, geography and the history o f names o f his own period. The work can be used only very occasionally for direct inform ation regarding the conquering M agyars, and then only w ith the greatest care. There again, indirectly— because o f the rich linguistic m aterial— it is still an indispensable source for us. Likewise, charters from the age o f Arpad are im portant for their linguistic and geographical inform ation. The last few years have seen enorm ous strides taken in the critical editions o f the Arpad-age charters. The earliest such charters were drawn up in the 11th century, but few o f them are original. O f surviving copies there are a few authenticated docum ents, but because o f obvious interests surrounding ownership rights, m any are fakes or have been corrupted with insertions. Over time, the num ber o f sources increased, but naturally these can only be used indirectly for details regarding the conquering M agyars. By the m iddle o f the 11th century, descendants o f the conquering M agyars w ere already in the sixth-eighth generation. Personal nam es or names that survived in place nam es and which appeared in charters reflect a fundam entally altered historical circum stance. It is conspicuous in this m a terial ju st how m any obviously Turkic-origin personal nam es there are. The question im m ediately arises as to w hether these w ere surviving per sonal nam es dating from before the Conquest, or should we be seeking other linguistic reasons. The continuation o f clan and tribal nam es is m uch more obvious. The nam es o f the conquering tribes, clans and the peoples who joined them can be reconstructed after the critical evaluation o f charters from the age o f the Arpad dynasty.
60
Methodological introduction and the sources
c) Slavonic sources The beginnings of Slavonic literacy Today, scientists fully agree that the creator o f Slavonic w riting was that Constantine w ho prior to his death in 869 in Rom e adopted the m onastic name Cyril. There is also no dispute about the fact that the original alphabet invented by Cyril w as the so-called G lagolitic alphabet, w hich was em ployed to record the first Christian religious texts. The alphabet was probably com piled in M oravia around 863. This was later succeeded by a type o f writing, later called Cyrillic after Cyril, and which is closer to Greek. A few characters from the Glagolitic alphabet were transferred into the Cyrillic alphabet. Early G lagoli tic texts were later rewritten in Cyrillic. The origin o f the Glagolitic alphabet is o f interest to us because according to sources, the Slavs had a runiform script, elem ents o f which G lagolitic writing partly used to record Slavonic phonem es w hich had no equivalent in Greek or in any other w riting system know n to Constantine. This putative runiform alphabet m ay be linked to both the Avar and the Szekely runiform letters (see Chapter XVI). Although the first records o f G lagolitic writing were in M oravia, it was not used there for long. Follow ing the death o f M ethodius in 885, his followers were driven out o f M oravia. They were given refuge by B olghar ruler Boris (852-889) who was baptised in 864, and with the enthronem ent o f Simeon (893-927) the Slavonic old Bolghar language becam e the official language o f Bulgharia in 893. From then on first translations, and later independent works were w ritten in the so-called Old Church Slavic or Old B olghar language.
The legends of Cyril and M ethodius The lives o f the two Slavonic apostles are generally called the Pannon Legends. The L egend o f Constantine, a life o f Constantine who later adopted the m onastic nam e Cyril, was originally w ritten around 870 in Rom e after his death. The legend is probably the w ork o f his brother, M ethodius. In all likelihood it was drafted for the cerem ony o f canonisation, and was w ritten in Greek. The Slavonic-language version prepared by M ethodius cannot be from m uch later. The latter was probably recorded first in Glagolitic script and later transcribed into Cyrillic. O f the 50 known copies, not one is earlier than the 15th century. These copies can be divided into two groups, Russian and Serbian redactions.
61
The sources
The life o f M ethodius was preserved in the Legend o f M ethodius, probably written by a follower, and dating from some time before the expulsion o f the followers from M oravia, that is before the end o f 885. O f the eight known copies in existence the oldest is from the 12 th century.
O ther early Bolghar accounts The lives o f earlier Slavonic saints are contained in a collection dedicated to a soldier who lived during the rule o f Boris and Simeon, B olgharia’s first Christian rulers. The w ork tells o f the soldier’s m iraculous escape from the M agyars (ugri) in relation to events taking place betw een 894 and 896. Naum was am ong several followers o f M ethodius who fled to the Bolghar ruler Boris; Naum continued his activities in Preslav, then capital o f the Bolghars. A record o f his life dates from about 924. In it we can read o f the M agyar Conquest, and o f how the M oravians fled before the M agyars to the Bolghars. The work was originally w ritten in Greek, but it survived only in an Old Church Slavic translation. The oldest extant copy dates from the 15th century. A longer narration relates the foundation o f the Bolghar church and the split between the Latin people and the Greeks. The work was originally penned in Greek, and can be traced to around the 12th century, but certain parts originate from 1Oth—1 lth-century sources. The Greek text has not survived, but it was translated into Old C hurch Slavic, probably some time in the 14th century. The text traces the baptism into the Christian faith o f the different peoples, and it is here that we find reference to the word the M agyars used for them selves, magere. According to the source, two M agyar chieftains travelled to Constantinople to be baptised. Some believe these two w ere Termecsii and Bulcsu.
The Bulghar regal list The w ork contains a list o f the Danube Bulghar rulers. A fter the nam e o f each ruler there is a num ber written in letters indicating for how m any years the prince reigned, and after this the name o f the ruling clan. There then follows a phrase “his year (let yem u) was X X X ” . The text, translated probably from Danube Bulghar into Old Church Slavic, has survived in three m anuscripts, the oldest dating from the end o f the 15 th century and the other two from the 16th century. The texts have been severely corrupted, allowing m uch room for speculation. The first name in the list is Avitohol, m ost probably Attila,
62
M ethodological introduction and the sources
the dynasty’s legitim ising founder; the second Im ik, A ttila’s son; the fourth Khuvrat, ruler o f G reater Bulgharia; and the sixth nam e is A sparukh, who actually founded the Danube Bulghar Empire. The first nine rulers belonged to the Dulo dynasty, the last four to the Vokil dynasty. The clearly Danube B ulghar expressions— which generally consist o f two w ords— com ing after the phrase “his year” have sparked the greatest debate. M ost o f those persons appearing in the list o f princes can be identified in Byzantine sources. O f the Danube Bulghar words, only a few have been deciphered with any certainty.
Ancient Russian chronicles The conversion o f the Kiev state to the Christian faith is com m only dated from the baptism o f V ladim ir in 988, although Christianity had appeared much earlier, at least during the rule o f Igor (913-945). In any case, it was only several decades after 988 that the m etropolitan seat was actually established, and m onks installed here began to chronicle events. R ussian annals are generally considered to have been started betw een 1037 and 1039. The first m aterial was put into chronological order by igumen, that is abbot, N ikon o f the Kiev cave m onastery in the 11th century, supplem ented with fur ther m aterial to 1073. This early work was expanded with inform ation from chronicles from other m onasteries to 1095. It is generally term ed the prim ary com pilation (nachalniy sv o d ). Nestor, a m onk in the Kiev m onastery, used this prim ary com pilation as a base in 1113 for his new version o f the chron icle. This new version w as in turn then revised for V ladim ir M onom akh, an enem y o fN e sto r’s com m issioner, Svatopolk. Follow ing Svatopolk’s death, M onom akh seized the Kiev throne. The revised chronicle was com pleted in 1116 by Silvester, abbot o f the Vidubec monastery. This becam e the chron icle’s second edition. However, the Kiev m onks, seeking the patronage o f M onom akh, began to rework the text once again, subm itting it to the new ruler in 1118. This was the third and last edition o f the early chronicle. Copies found their w ay to different m onasteries where they w ere either copied further or supplem ented with local events, in other words, the chronicle was continued. The oldest surviving copy is from the 14th century. M aterial for the Russian prim ary chronicle relating to the 9th and 10th cen turies was taken partly from local tradition, and partly from B yzantine sources. The chronicle was earlier known as the N estor Chronicle, w hile today it is m ore often called by its Russian title, Povest (Povest vremennyh let). As one can see from the above, inform ation in the old Russian prim ary chronicle related to the pre-Conquest M agyars has to be treated w ith extrem e caution.
63
The sources
d) Middle Iranian sources The M iddle Iranian period ran from the 4 th -3 rd centuries BC until the Arab conquest, that is to the 7th century, although according to the reckoning o f some, until the 9th century. W ithin this, the period at the beginning o f the 3rd century AD is clearly defined by the ruling Sassanids in Persia (Ardashir I 227-241). The rule o f the Sasanian dynasty was brought to an end by the Arabs in 637. At this time, o f the peoples living in Iran the Persians, the Sogdians, the Sakas, the Bactrians and the Hephthalites had their own forms o f writing. These scripts were used to record religious and lay texts, as well as inscriptions on m edals and other objects. Although to date, no inform ation directly related to the M agyars has been discovered in these sources, they do still throw im portant light on the Turkic peoples— am ong them the K haz ars— from the 6 th century. At that time these Turkic peoples w ere neighbours or close allies o f the M agyars. M iddle Persian literature was w ritten in Pahlavi. Pahlavi is one continuation o f the Aram aic branch o f the Semitic family o f writing, containing so-called ideogram s. This m eans that words w ritten in A ram aic were retained in the Persian texts, but they were not read letter for letter. R ather the word was read and understood according to the m eaning o f the Aram aic word, in Persian, as for instance, in an English text the French w ord w ritten roi w ould be read as king. One o f the m ost im portant sources is the Book o f Rulers (H vaday nam ag) which although now lost was drawn on by m any M iddle Persian and new Persian (e.g. Sahnam e) sources. A nother genre was the apocalyptic literature relating the vision o f the end o f the world, through which many peoples were introduced, am ong them the Turks. The Bahm an Yast is one such representative o f the genre. Im portant historical references are to be found in Zoroastrian religious literature, am ong them the D enkard and the Bundahishn. Persian religious leaders played a particularly significant role in the dis sem ination o f M anichaeism throughout Central Asia, an influence w hich also reached the Turkic peoples. Texts were generally w ritten in the so-called M anichaean script, a version o f Palm yrian from the Sem itic fam ily o f alpha bets, further developed by M ani. Texts in M iddle Persian M anichaean speak about, for example, the court titles o f the Uighurs. A few M iddle Persian texts have survived only in Syrian or Arab translation. These also contain inform ation about Turkic peoples, for instance the Kangars. The Sogdian m aterial is particularly illum inating. M aterial w ritten in Sogdian script (which refers back to Aram aic) dating from the 2nd and 3rd cen turies records steppe history, and here, for example, we find the Xiongnus
64
M ethodological introduction and the sources
m entioned with the nam e Hun. Today we know that the court language o f the First Turkic Khaghanate was Sogdian, and therefore it played a role in the m iddle o f the 6 th century sim ilar to that o f Latin in m ediaeval Europe. In scriptions created during the First Turkic K haghanate were also com pleted in the Sogdian language. Am ong these inscriptions, we know o f but one that has survived: the B ughut inscription. There are m ulti-language inscriptions from the tim e o f the Uighur Khaghanate which also contain text in the Sogdian language: the K harabalghasun inscription. Charters and letters in Sogdian reflecting the everyday life o f Turkic society are also important. The Sakas organised their lasting rule from K hotan in Turkestan; the state survived until the M uslim invasion in 1000. They follow ed the B uddhist faith, and their written docum ents can be traced as a Central A sian offshoot o f Indian writing. K hotanese rulers interm arried w ith the Turkic leading strata, and were related to those Turkic peoples who had settled in A fghanistan, but they also had affinities with the Turkic peoples o f the steppe. Thus we find m uch in Khotanese texts about the Turkic peoples, and we know o f a Turkic-K hotanese glossary dating from the 10 th century.
e) Sources by M uslim authors In 622 AD, M uham m ad (m uham m ad ‘praised, glorious’) was forced to flee M ecca for the safety o f M edina where he sought alliances against his own tribe, the Kuraysh. N ot only does Arab history start w ith his exodus (hegira), but am ong the M uslim religious world, dates are also reckoned from this point, and even today in every M uslim country tim e is counted from this focal event. By the tim e o f M uham m ad’s death in 632, he had succeeded in uniting the m ain tribes o f the Arab Peninsula. His first successor, the first caliph (Arabic kalafa ‘to follow ’ > kalifa ‘successor’) Abu Bakr, and his troops reached the borders o f Syria and Persia. The second caliph, Omar, established the foun dations o f the A rabic world empire. O m ar took Dam ascus (635) and Jerusalem (636), and then turned against Persia. He captured Ctesiphon (636), and delivered the final blow to the Sassanid Em pire at N ehavend in 642. His generals then drove the Byzantine troops from Alexandria. In 644, the next caliph continued the conquest o f the known world. He attacked Byzantium , and established a fleet o f Arab warships. In 656, Ali, M uham m ad’s son-in-law, in a bid for power, opened hostilities against the w idow o f M uham m ad, follow ing w hich the M uham m adan world suffered its first and still lasting m a jo r split into those following Ali, the Shiites, and the Sunnis, supporters o f the successor dynasty, the Um ayyads. In 661 M u’aw iya founded the Um ayyad
Figure 8 The spread of Islam until 750
66
Methodological introduction and the sources
I
P r a n k is h
C A U W f t ^ КЙ1<ЙЭОЙ'^, ;
Q P to te d o
'
X
}
\
4
V
/' :J
4 x
/ 4
t
'
r"
\
' y n) T rs\
./ }
- !
; « u k h .,.0 .7 m
!
^ n lg n tj^ p fe
■ S
'4V-' v Г-
«fr ,
% ■;?CLIBYA airo ''
p"-.v.
4 "t", „ „
;
«Mosul
■:
'' т i:
f
< „„i
о г •
i
i
,
• -I ■. ■■
Jam jbiU s ‘ Samara ■ «Jerusalem
,V /
S a i iM t k a n d ‘ ^ 9
'■! ! ’ I R I .
Creb;;s,fc >
-4...
i ‘’t
‘
t ft ,
{,
.....'>•&>&$
^ B e g lid a d
/
У
л ь »M edina
■■>
• Mecca
Figure 9 Islam ic dynasties in the 9th-10th centuries
dynasty and battled successfully against the sons o f Ali. He organised the country’s military, financial and political system. Dam ascus was settled on as the capital. However, Byzantium w ithstood all attacks, and thus atten tion turned eastward: troops o f the caliph captured Khabul, B ukhara and Sam arkand. The form er Carthage fell to the Arabs in 698, and their own currency, the dirham, was introduced. By 711 the Indus valley w as under their control, while in the same year the suprem e com m ander, Tariq, crossed into Spain (Jebel al-Tariq ‘Tariq H ill’ > Gibraltar). In 732, the European powers under the leadership o f Charles M artel halted the M uslim Arab invasion at the gates o f Poitiers. On the eastern front, a coalition o f forces headed by Chinese generals stopped the Arabs at the River Talas east o f the Lake Aral in 751. The Khazars halted the incursion o f Arabs across the Caucasus, and they were unable to take Byzantium . In 750 a new dynasty, the A bbasids, came to rule the Arab state. The seat o f the em pire was shifted to Baghdad, and with Persian assistance it was com pletely reorganised. The direction o f the em pire’s military, political and indeed cultural affairs quickly fell into the hands o f the Persians who in the m eantim e had been com pletely converted to Islam. Harun al-R asid (786-809) ruled during the
67
The sources
second culm inating point in the history o f the Arab Empire. He is know n in the tales o f The Thousand and One Nights. The Arabs adopted m any elements o f Greek culture and translated m any o f their works. Arab culture built on Hellenistic, Persian and Indian elem ents reached its apogee in the 10th century. It is in this environm ent that we m ust seek those sources containing references to the conquering M agyars and neighbouring peoples. In using Arab written sources, attention has to be paid to the fact that Sem itic-origin writing originally com pletely ignored vowels; the m eaning was carried solely by the consonants. Arabic writing has several consonant marks which have no equivalents in non-Sem itic languages, while it does not have separate letters for a few such sounds o f the non-Sem itic languages. Certain letters can only be differentiated by one or several points placed below or above the character, furtherm ore these diacritical points are frequently om itted in scripts. The reading o f Arabic words generally causes no problem , but when foreign words or nam es have to be transcribed, this not only opens up the opportunity for mistakes, but the scribe will often try to m odify the transcription in favour o f a sim ilarly constructed word already existing in Arabic.
Sources in Arabic Geographers. “G reek” and “Persian” approaches to the w orld are m ixed in Arabic geographical literature; some schools further developed the m apping m ethod o f Ptolemy, adding new discoveries, while other writers arranged the world according to the points o f the com pass alone. The efficient governing o f the Arabian empire, which had unexpectedly grown so huge, also dem anded that geographical knowledge be constantly updated and expanded. In this, key roles were played by traders and m inisters w ho diligently reported to Baghdad on their journeys, and if not in person, then to som eone else who was on his way to Baghdad. It is no w onder that the favourite genre and at the same time the m ost recurrent book title o f the age was The B ook o f Routes and K ing doms (Kitab al-masaliq val-mamaliq). The m aps came com plete w ith cap tions which could be extended w ith new inform ation. The Arabs, too, did not recognise au th o r’s rights, and thus they frequently copied each o th er’s work without due recognition. However, in many cases it is possible to pinpoint the author o f a particular piece o f information. From the 12th century prim ary geographical literature ended, but at the same tim e there was an upsurge in grand lexicons and the com pilation o f collected works, and this was im por
68
M ethodological introduction and the sources
tant because in m any cases only these preserved the contents o f m uch earlier works. Al-Khwarezmi, who was bom in the town o f K hw arezm and died around 847, w rote addenda and explanations to a map o f the w orld by Ptolemy. In the explanations he concerned him self with Eastern Europe. Likew ise in the m iddle o f the 9th century al-Jarm i, on his return hom e from im prisonm ent in Byzantium , com pleted a work on the B yzantine Em pire and its neighbours. Although the work was subsequently lost, there are many extracts. Sim ilarly Harun ibn Yahya languished in jail in Byzantium around 900. His work de scribing the B yzantine Em pire and its peoples, as well as his own journeys, only survived in quotations used by other authors, including Ibn Rusta. The caliph’s confidant Ibn Hordadzbeh, founder o f The B ook o f Routes and K ingdom s genre, died in 911. He is usually m entioned as the m inister o f post, but in truth his m ission was probably to obtain intelligence about m ilitary and trading routes. His first known collections were com piled betw een 846 and 847, and he revised the work between 885 and 8 8 6 . He kept num erous route reports, am ong them one on Central Asia, the Report on the R oute o f Salam the Interpreter from before 847. The w ork o f Ibn H ordadzbeh form ed the b a sis for Arab geographical studies at the end o f the 9th century. The earliest surviving m anuscript o f the work dates from before the 12 th century. Al-Yakubi (d. 897), Ibn al-Fakih (9th century), Kudam a (died ca. 958) and others contributed to these foundations with new material. We m ust deal with Ibn Fadlan in slightly m ore detail, since he is particularly im portant for an appreciation o f M uslim sources dealing with the history o f the conquering M agyars. W ithout w ishing to touch here on questions (dis cussed later in m ore detail) o f B ulghar-Turkic history relating to the history o f the conquering M agyars, we m ust mention what is absolutely necessary for an understanding o f the political-historical background to the journey o f Ibn Fadlan. At the beginning o f the 10th century, significant groups o f Bulghars seeking to escape K hazar rule and centralised pressure lived around the area defined by the rivers V olga-K am a-C herem shan. The increasingly inde pendent Bulghar mler, while organising neighbouring tribes, still rem ained nom inally a K hazar subject. The control o f Volga trade as w ell as the key position o f being m ler o f trade routes, prim arily to K hw arezm , east o f the Slavonic territories, significantly contributed to econom ic strengthening. In this situation, the caliph in Baghdad represented a natural ally for the Bulghar m ler in opposition to the Khazars with its leading strata o f the Jewish faith. For this reason, he requested in a letter the support of, and an alliance with, Baghdad. It was proposed that the caliph despatch a person w ho could teach the Bulghar m ler the word o f Allah, Islamic law, and it was also suggested that a m osque and m ihrab be built for him in order to help him in converting
The sources
69
the w hole country. Further assistance was requested by the Bulghars in con structing fortresses and castles to defend them selves against opposing kings. It is clear from extracts from the letter that in exchange for conversion to Islam, m ilitary assistance w ould be provided, also for defence against the Khazars. Interestingly enough, the Bulghar ru ler’s letter was actually taken to the caliph by a Khazar, al-Hazari, obviously a m em ber o f the K hazar rul e r’s opposition who had escaped to the Volga Bulghars. In response, caliph A l-M uqtadir (908-932) com m issioned a delegation to make arrangements. The costs were covered by the incom e from an estate near Khwarezm which had recently been confiscated by the state from a fabulously w ealthy treasurer who had fallen out o f favour. The caliph’s delegation com prised four persons: the leader Sausan al-Rasi, Tegin al-Turki, Bars al-Saqlabi and Ibn Fadlan, that is there were two Turks, o f whom Fadlan referred to one as just “the Turkic” (al-Turki), the other being a Volga Bulghar. In this text the nam e Saqaliba, generally signifying the Slavs, always means the Volga Bulghars. The dele gation left Baghdad on 2 April 921, and travelled only indirectly to the Volga Bulghars. There were two reasons for this. One was, obviously, to avoid the Khazars, the other that they had to pick up m oney in Khwarezm . In all likelihood there was in addition a third reason for this roundabout route. At that time, Khwarezm bordered on the ruling Sam anid dynasty’s centre o f Bukhara, and to a certain extent it was dependent on it. The Sam anids, form ally vassals to the caliph o f Baghdad, had earlier also attem pted to convert heathen Turkic tribespeople living beyond the K hazar Em pire to the Islam ic faith, and thus to extend the Arab alliance. It is likely that Khwarezm and Bukhara thus played an im portant role even then in the conversion o f the Volga Bulghars. On arrival in Bukhara, Ibn Fadlan and his com panions im m ediately w ent to al-Jayhani, m inister o f the local ruler, “Baghdad governor” N asr ibn Ahmad. In his diary, Ibn Fadlan only referred to al-Jayhani as “the respected elder” . Jayhani arranged a servant and accom m odation for them, and gave instruc tions that all their wishes be met. For a variety o f reasons, Ibn Fadlan and his com panions rem ained in Bukhara longer than originally planned, and as he wrote, by the 28th day they were highly im patient, fearing that the onset o f w inter would obstruct their further passage. From our aspect the relevance is that at this time Jayhani was an old man, and the travellers w ere together with him in Bukhara for nearly a month. Because there were several Jayhani’s, the source under the name Jayhani is probably not the w ork o f one single Jayhani, still it is clear that here we are talking about Abu A bdullah M uham m ad ibn Ahm ad al-Jayhani, whose son took over his m inisterial post in 937. This Jayhani died in 941, and his grandchild becam e m inister in 976. It is highly likely that he then com pleted his grandfather’s work. At any rate Ibn Fadlan and his fellow travellers spent long enough in B ukhara to allow Jayhani to
70
Methodological introduction and the sources
question them. The Jayhani report on these m eetings contains parts that Ibn Fadlan could not have known then. Unfortunately we do not know w hat route Ibn Fadlan took on his return to Baghdad, but it is unlikely that he took any other than the one he came on. Even though it is true to say that Ibn F adlan’s travel record ends with a description o f the K hazar Em pire, it w as probably written from inform ation supplied by his com panion al-H azari, and not from personal experience. Naturally it is necessary to note that one o f the m em bers o f the m ission in Bukhara was a Volga Bulghar, and Jayhani could have received a lot o f inform ation from him as well. In any case, one thing is certain and that is that m aterial on Jayhani’s report com prised m any parts, and some o f his sources were contemporary. Jayhani’s other know n close source, which we can actually trace was al-Balhi. Balhi died in 934, leaving a detailed geographical w ork o f w hich nothing rem ains, although the larger part is preserved in extracts by Istahri and Ibn Hauqal. Balhi and Jayhani had such a close personal relationship that Balhi sent slave girls to his friend. A nother writer, A bu D ulaf, recorded in 943 that as an emissary, he travelled through Turkic, Persian and A rm enian territories, and collected inform ation. O f his two route itineraries, the one concerning the Turks is not reliable, being full o f secondary and som etim es inaccurate inform ation. However, he did write that he travelled on the orders o f Jayhani, and thus this Jayhani can only be the son (d. 941), as the first Jayhani— the “respected elder” in the tim e o f Ibn Fadlan— was unlikely to be still living (although it cannot be ruled out). As we have seen above, Ibn Fadlan travelled from B ukhara to K hw arezm , from w here via the land o f the Pechenegs and the Bashkiris he arrived in the land o f the Volga Bulghars. For a long time the work o f Ibn Fadlan was only know n from extracts in the great geographical dictionary by Yakut. Then, in M ashhad in A fghanistan, a m ore com plete m anuscript w as discovered in 1923, and a photocopy was m ade for H erzfeld for the first time. A second photocopy was arranged in 1935 for the Soviet Academy. The text was then published— partly independently o f each other— in the Soviet Union by K ovalevsky (at that time w ithout his nam e, but under the nam e o f the editor Krachkovsky) and in Bonn by Zeki Validi Togan in 1939. A photocopy was also m ade for Ligeti who travelled to A fghanistan in 1936. This docum ent is kept in Budapest. Facsim iles o f this copy (com plete w ith notes) w ere publish ed by K aroly Czegledy in 1950. O f course, the M ashhad docum ent is also not original, but an abbreviated, in m any places corrupted and relatively late, copy. We know this because works by A hm ad Tusi and N ajib H am adani w ho lived in the 12th century, and the w ork o f Am in R azi w ritten in 1593 all quote freely from Ibn F adlan’s work, while they also contain parts w hich cannot be found in the dictionary o f Yakut or the M ashhad m anuscript. Yakut m ade a note that he w orked from several Ibn Fadlan m anuscripts. The M ashhad m anuscript
The sources
71
is not com plete, and it certainly lacks the end, thus it is only possible to reconstruct the original with the help o f the later copies and the Yakut dictionary. As we have seen above, questions concerning the Jayhani source critique have not been fully resolved. The dates o f Ibn Rusta, another im portant Arab writer, are sim ilarly disputed. Only the last volum e o f his seven-volum e work survived. In it there are parts which correspond to the Jayhani work, and for this reason m any date the work later than the second decade o f the 1Oth century. According to others, the work dates from before 913, and was possibly written in 903. The reasons they give are so-called ex silentio, referring to certain events w hich occurred after these dates and are m issing from the work. Countering the opinion that Ibn Rusta could use the Jayhani w ork only after the second decade o f the 10 th century com es the response that the quoted sections originate not from Jayhani, but from an earlier source or sources used by Jayhani. The current state o f the debate favours those who reason that Ibn R usta’s work is later rather than earlier. The Persian w riter Gardizi— we speak o f him in m ore detail below — also used Jayhani’s work, as did the Persian w ork H udiid al-alam w ritten in 982/983, the 1 lth-century al-Bakri, M arvazi from around 1120, the 13th-cen tury Aufi, and Sukrullah writing in Persian in 1456. The latter’s work was translated into the Turkic language, respectively revised, and some details appear in these Turkic versions, the originals o f which have not yet been found, in the text o f later m ediators. The accom plishm ents o f al-Balhi w ere continued by al-Istahri who lived in the 10th century, and who, after spending m any years in Baghdad, eventu ally died in Samarkand. There were two versions o f his work. The first was translated into Persian before 933, w hile the second version dates from around 951. Ibn Hauqal, who joined Istahri in Baghdad, eventually becom ing his follower, travelled extensively throughout the then know n Islam ic w orld and in bordering regions betw een 943 and 973. His w ork was w ritten before 987, com plem enting the B alhi-Istahri w ork with his own findings. Ibn Hauqal also noted that on his travels he took with him the works o f Hordadzbeh, Jayhani and Kudama, and made notes on the spot directly into their m anuscripts. M uqaddasi or M aqdisi (known as “the Jerusalem ite”; the A rabic nam e o f the town is al-Quds, hence al-M uqaddas), the last im portant representative o f the Balhi school, who enriched the geographical docum ents with much new information, probably died some time around the year 1 0 0 0 . W hile the above-m entioned were travellers or inhabitants o f the M uslim east, al-Bakri (d. 1094) never left the M uslim state established on the Iberian Peninsula, Um ayyad Andalusia. It is surprising ju st how well he knew early
72
M ethodological introduction and the sources
geographical literature, or in any case he took extracts from num erous earlier writers. Extracts taken from Jayhani’s work are shorter than those taken by Ibn Rusta, but he selected inform ation about other Eastern European peoples the like o f which cannot be found elsewhere. M ore detail is provided later (see pp. 291-293) about A bu H am id al-G arnati who travelled in H ungary between 1150 and 1153, and died in Dam ascus in 1170. It is necessary to briefly introduce Idrisi, who lived and w rote in Sicily, and who died around 1165. Com m issioned by Roger II, ruler o f Sicily, he w rote a book on the world as it was then perceived. Know n as The B ook o f Roger, this is a particularly im portant work in which contem porary H ungary appeared, although the writer had used the works o f several earlier authors. We have already referred to the geographical dictionary by Yakut (d. 1229). In his work, com pleted in 1224, individual places and peoples appeared under the place nam es in Arabic alphabetic order. W hen talking about places and peoples he quoted the works o f earlier authors, nam ed them, and generally conscientiously copied extracts. But there is also inform ation collected by the author him self, for instance on the Hungarians. Although he gave the correct pronunciation beside the place nam es, that is the author wrote dow n the vowels, this is o f little use because the w ritten forms w ere generally gathered from other books and manuscripts. A fter Yakut, Arabic geographical studies do not, in practice, contain any thing new for us. Certain writers had access to early docum ents since lost, while others influenced European literature because they w ere recognised and translated early on, such as A bul Fida (d. 1331). Historians. Insofar as Arab geographers wrote about the geography o f the Islam ic world, the aim o f Arab historians was to preserve the history o f the Islam ic conquests. As such, much was naturally written about the peoples o f Eastern Europe. One o f the earliest writers was al-Baladhuri w hose work, H istory o f the M uslim Conquests, was com pleted in 892, ju st before his death. Com pleted in 872, the World H istory o f al-Yakubi is alm ost contem porary. This world history describes, am ong others, the known history o f non-A rabic peoples, including the Turkic, covering the period from the Jewish patriarchs to the time o f M uham m ad. Al-Tabari (d. 923) wrote a sum m ary o f world history up until M uham m ad, and from then he changed to an annalistic form o f presentation. M any parts o f the original are lost, the w ork is known through copies, w hile several writers continued Tabari’s labour. The works o f the 10th-century al-D inavari and the
73
The sources
Syrian Agapius writing in Arabic are significant because o f their references to the Khazars. However, the historian al-Masudi, who died in Cairo around 956, is far more im portant than those authors already mentioned. He w rote an abridged history o f Islam com plete with a thorough geographical introduction. The work was originally entitled The B ook o f G old Fields and Precious Stone M ines, and was w ritten between 943 and 947. He also created several other works, am ong them the extant B ook o f Warning and Correction, in which, not long before his death, he endeavoured to bring some order to his historical-geographical understanding. Al-Biruni, bom in Khwarezm in 973, included ancient historical inform a tion in his im portant work which also contained m uch that was new. The works o f Spanish Arab historians contain indirect inform ation relating to the conquering M agyars. We have already seen the broad geographical knowledge o f the Andalusian al-Bakri. A local Arab centre o f historiography had been established in Cordova even before the 10th century, the focus o f which, naturally enough, was events in the Iberian Peninsula, but m any authors also covered other parts o f the Islam ic world. Unfortunately, this rich and well preserved literature remains to this very day w ell-nigh unstudied. W ithout doubt, o f all the writers Ibn Hayyan takes a particularly distinguished place. Some are o f the opinion that he was the author o f a 60-volum e grand work, one part o f which is the M uqtabas (M uqtabis, Extract), the fifth volum e o f which was published in 1979. The publication was based on the only known m anuscript (which is certainly a copy) kept in Rabat, M orocco. Ibn Hayyan also wrote about the spread o f Islam with particular attention to the Iberian Peninsula. The w ork was com pleted ju st before the author died in 1076. He em ployed num erous earlier works, from one o f w hich he took records o f the M agyar raids into Spain in 942. W hen m entioning the M agyars he used the word Turk to explain who they were, inserting some inform ation on the M agyars and the nam es o f the seven M agyar chieftains. These nam es were seriously dam aged in the course o f repeated copying; deciphering w ill require the help o f the M uqtabas source itself or other copies.
Sources in Persian Within the Abbasid C aliphate the larger provinces becam e increasingly inde pendent, while still formally recognising the authority o f Baghdad, the Arab centre. The G aznavid family in Persia was one such sem i-independent dy nasty. With the expansion o f pow er o f the Persian local ruling stratum , the cultivation o f Persian language literature also began.
74
M ethodological introduction and the sources
The Regions o f the World (H udiid al-alam) is a geographical work by an unknown author. It is likely, however, that the author lived in Northern Afghanistan, and according to the dedication in the w ork it could have been com pleted for one o f the local feudal dynasties in 982/983. The only known copy dates from 1258. Gardizi lived in the Gaznavid court in the 11 th century. His w ork, The Book o fth e E m bellishm ent o f the Reports, w ritten between 1050 and 1052, contains both a history o f Persia and the histories o f neighbouring peoples. The sections relating to the Turkic peoples refer in part to a so-called Pseudo Ibn M uqqafa and partly to Jayhani or Hordadzbeh. However, the author used a Jayhani m anuscript which differed from that used by Ibn Rusta. O f the two known m anuscripts, the earlier is a 1 6 th -17th-century copy, while the other dates from the 18th century. The Persian language was em ployed by ever m ore writers from the 12th century, am ong them a great m any historians. They did not have any direct inform ation relating to the M agyars prior to the Conquest, and w hat little they did have on the early Turkic peoples can be traced back to earlier works. One such exam ple is the faithful chronicler o f steppe history, Juzjani, who com pleted his work in 1260. In fact with the M ongol period Juvaini (d. 1283) and R ashidud-D in (d. 1318) represented a new historians’ school; their works are im portant for M ongol-era nam es o f the M agyars.
f ) Syrian sources A group o f Syrian people living on the fringes o f Byzantium and Persia spoke an old language related to the Sem itic language family. The survival o f the com m unity is largely put down to the fact that the western group o f Syrians professed M onophysite Jacobite Christianity, and the eastern com m unity N estorian Christianity. The significance o f this is that although their groups, one under the sway o f Byzantium — centred on Edessa— and the other under the protection o f the Persians— centred on N isibis— engaged in internecine warfare, at the same tim e they played a sort o f m ediatory role betw een the two great cultures. Their significant ecclesiastical historical literature contains a w ealth o f detail. Scripts were recorded in a version o f A ram aic-S em itic, so-called Estrangelo (“rounded”) Syrian writing, and o f the tw o further developed form s o f Estrangelo the western Jacobites used Serto and the eastern Syrians Nestorian. A sm aller sect, the M elkites, used Arabic, allowing them to act as m ediators with the A rabic-language world. Some o f their groups, such as the M aronites, still play a significant role in Lebanon to this very day.
The sources
75
Legends w oven around the eastern m ilitary cam paigns o f A lexander the Great spread throughout the east, even reaching the M ongols. Contem porary geographical and historical elem ents were introduced into local variants o f the legends, w hich were noted down as well. Thus, we find references to the steppes in the A lexander the Great Syrian Legends. In the sam e w ay ethno graphic elem ents creep into so-called doom sday literature, in the chapter on the peoples o f Gog and Magog. The Syrian revision o f the ecclesiastical history by Zakarias, bishop o f M itilen, and originally w ritten in Greek, is an im portant w ork from the aspect o f steppe history. The work deals with the period betw een 450 and 518. A Syrian geographical appendix (not to be found in the original) was added, and thus it is usually referred to as Pseudo Z akarias’ work. Textual evidence indicates that the appendix was com pleted in 555, and it lists the peoples liv ing beyond the Caucasus, that is north o f the Caucasus. Although the only surviving copies date from around 600, and there is some doubt as to the credibility o f the data, as far as m ore recent inform ation goes, like the form o f the nam e o f the K hazar people recently discovered from Turkic sources, it does have some authenticity. But a word o f caution: some parts o f the older data can be traced back to Byzantine sources, e.g. Priskos. The Syrians started writing chronicles in the first h alf o f the 6 th century. These chronicles frequently copied each other, or new insertions were made into old texts. There are references to the Huns in early Syrian chronicles, while later chronicles m ention the appearance o f the Avars and the Turks. It is particularly worth highlighting the chronicle o f M ichael o f Syria com pleted in 1199, a detailed history o f the w estern Syrian (Jacobite) church. Events recorded in this great historical w ork com prising 21 volum es are presented in three parallel columns: w orld history, ecclesiastical history, and natural events. For some time the text was only known in Arab translation, w ritten in A rm enianand Syrian script, but later a Syrian language and Syrian script version was found in the m onastery at Edessa. Even though the w orld history section is m ostly a com pilation, it still preserves m uch inform ation on the history o f the steppes which cannot be found in any other surviving material. B ar Hebraeus (d. 1286), son o f a doctor born into a Jew ish fam ily but converted to the Christian faith, becam e leader o f one o f the groups o f Jacobites. The earlier sections o f his great historical work are a slightly revised version o f the M ichael o f Syria World H istory Column. H ebraeus him self wrote the section on events after 1199. Its significance lies in the Arabic translation which he com pleted. This Arabic translation has m uch to offer in resolving certain problem s o f source criticism and reading. He also used other
76
Methodological introduction and the sources
sources in preparing the Arabic translation. In addition he w rote an ecclesias tical history. The H istory o f Saints a n d M artyrs contains the lives o f several saints.
g) Armenian sources The beginning o f the history o f the people speaking the A rm enian lan guage— an independent Indo-European language branch— is traceable to the first m illennium BC. The state founded by tribes who had m igrated into the site o f the ancient U rartu culture soon came up against the Greeks and Iranian peoples, but it always m anaged to preserve its distinct cultural, and in m any cases political, independence. C hristianity becam e a state religion in A rm enia in 301 AD. A round 400, the A rm enians established their ow n alphabet using elem ents o f Sem itic writing strongly influenced by the system o f G reek literacy, and began translating the Bible into Arm enian. The m ajority o f the great w orks o f G reek culture w ere translated into A rm enian from the 5th century. Parallel with this developm ent, independent A rm enian literature was born as well. Scribes in the royal court, in m onasteries and in the palaces o f the aristocracy began noting events and glorifying the past, leading to the creation o f A rm enian chronicles. This heroic chronicle tradition recorded for posterity the wars against the Persians and the barbarian peoples, and such historical inform ation was also included in w orks recording the lives o f the saints o f the A rm enian church. A round 640 A rm enia was invaded by the Arabs, and although the state then becam e a suzerain it m anaged to preserve its religious and cultural identity. A rm enia suffered greatly during the A rab-K hazar wars w hich spilled over into its territory, and thus we find a great deal o f inform ation relating to this period in the chronicles. The Bagratid dynasty, w hich had been helped to pow er by the Arabs, consolidated its position in 885, and in recognition o f this, the B yzantine em peror sent a crown in 8 8 6 . A rm enian historical literature w ent through a golden age in the 10th century. The Bagratid Kingdom collapsed in 1045, having becom e victim o f the B yzantine-Seljuk war. So started the m ass m igration o f Arm enians. A new A rm enian Kingdom was founded in 1080 around the southeastern shore o f today’s Turkey, in the form er province o f Cilicia. The dom inant dialects in this new state were w est Armenian. The Arm enian alphabet and orthography are relatively conservative. The characters have eastern and w estern A rm enian readings. The eastern is closer to old Arm enian. The overwhelm ing m ajority o f A rm enian literature today is available only in not older than 12 th-century copy form.
Figure 10 The Armenian Em pire
78
Methodological introduction and the sources
Early Arm enian chronicles survived only in copied form ats transcribed many times, but we do find some references to the Huns and the Ephthalites. The great historical w ork H istory o f Arm enia, conventionally attributed to M ovses H orenaci (Moses o f Horene), was com pleted in the 440s, leading many to presum e that the author lived in the 5th century. The oldest extant m anuscript fragm ent dates from not earlier than the 10 th century, and it is highly likely that it was copied m any times, supplem ented, and had anachro nistic interpolations. It is particularly disturbing to find references to later northern peoples in legendary histories, e.g. the Vlendur Bulkars, who are placed into the A rsak I era, that is the 2nd century BC. Arm enian additions to the basic geographical work o f Ptolem y w ere also started in the 7th century. Because earlier it was believed to be the work o f Horenaci, it is com m only term ed the Geography by Pseudo M ovses Horenaci. More recently, Ananias o fS h ira k has been suggested as the author. The work was revised m any times, but the basic editing dates from before the Arab invasion, perhaps around 640. The work includes a later intrapolation on the peoples o f the north— that is those living north o f the Caucasus— m aking it one o f the m ost im portant sources on the history o f Eastern Europe, and w ithin this the history o f the Khazars and M agyars from the 6 th—7 th centuries. There are m ore than fifty known m anuscripts o f the w ork in existence, but o f the extended versions it appears that the m ost useful is a m anuscript dating from the 15 th century. The identity o f an author who wrote a chronicle around 687, patterned on earlier such forms, has rem ained unknown, but the last part m entions a few events from the contem porary period. The earliest known m anuscript dates from the 10 th century. A w ork com m only referred to as the Sebeos was com piled from m any separate parts. The earliest surviving m anuscript dates from 1672, but unfor tunately this preserved at best mere fragm ents o f the com plete w ork o f the author, who probably lived in the 7th century. At the same time, other works were also copied into early manuscripts, such as an extract o f a 1Oth-century w ork prepared in the 15th—16th centuries. Despite this, im portant 6th-7th-century steppe history has been preserved, and it rem ains an im portant source for the history o f the Khazars. L evond (died ca. 790) probably finished his work shortly before his death. The work deals with the history o f A rm enia betw een 732 and 789. A signifi cant part (a theological dispute between the Byzantine em peror and the caliph) is a later interpolation. The only m anuscript o f the w ork was prepared in the 13th century. It ranks as an im portant source on the A rab-K hazar wars which devastated Armenia.
79
The sources
The so-called H istory o f Albania, probably com pleted in the 10th century but with 1 1t h - 12 th-century interpolations, should be seen as a history o f Caucasian Albania. The nam e o f the writer is uncertain, some refer to him under the name M ovses K alankatvaci (originating from Kalankatuyk), and others as D ashuranci (originating from Dashuran). The work can also be considered an im portant source for steppe and K hazar history. Later A rm enian ecclesiastical and historical literature is only w orth exam ining for its extracts and quotes preserved from earlier works.
h) Georgian sources There is no affinity between the several different languages found in the Caucasus today. However, the so-called Kartvelian language group is particu larly noticeable in this diverse group, o f which Georgian is the m ost recog nised, although M ingrelian-Laz and Svan also belong here. The first G eorgian state grouping the ancient inhabitants o f the Caucasus was established at the end o f the 3rd century (although there had been earlier moves in this direction). Then at the beginning o f the 4th century, parallel with the Arm enians, the state becam e Christian. During the 7th century it also fell under Arab rule. The Georgian branch o f the Bagratid dynasty proclaim ed independence in 930. In the 12th century this small kingdom experienced its first golden age, a period w hich was brought to an abrupt end by the M ongols. Georgian writing also originates from a branch o f A ram aic o f the Semitic family o f alphabets, although the Greek alphabet also played a role in its formation. The oldest m aterial dates from the 5th century, w hen the Bible was translated into Georgian and the first inscriptions were recorded. Translations aside, the first original Georgian m anuscripts recorded the lives o f Georgian martyrs and saints. These docum ents were later copied and supplem ented with subsequent events relating to the saints or places connected with them. In the course o f writing these hagiographies, a great deal o f inform ation on the neighbouring peoples was included, for instance on the K hazars, and the A rab-K hazar wars were also mentioned. The work The M artyrdom o f Abo, dating from around the end o f the 8 th century provides us with im portant historical source inform ation on contem porary Georgia and its neighbouring peoples. The Georgian Chronicle, the Kartlis chovreba (the с and the h are pro nounced separately), a history o f Kartlis, that is, Georgia, was com piled from numerous earlier historical works. The first ten volum es cover the history o f Georgia from its earliest times to the 14th century. The first chapters were written with the help o f early chronicles, thus the 11 th-century works by Leonti
80
M ethodological introduction and the sources
M roveli and Juanser cover history from the first stages to the 5th century, and from the 5th to the 8 th century respectively. The section on 8 th- to 9th-century history was w ritten by an unknow n author, and it is here that we can find much on those peoples w ho lived to the north o f the Caucasus and in the Caucasian mountains, as well as on the A rab-K hazar wars. The earliest known and surviving version o f the chronicle was edited in the 15th century. There is also an Arm enian translation o f the chronicle dating from betw een 1279 and 1311.
i) Turkic sources A fter defeating the Ruanruans around 552, the Turks founded the First Turk Khaghanate. W ithin ju st a few years the em pire was m aster o f the entire steppe. A fter 556, the Turkic peoples o f the Eastern European steppe w ere also under the em pire’s dom ination, and it then opened contacts with Persia and the B yzantine Empire. Around 630, internal and external w arfare gradually underm ined the tribal confederations, leading to a split into the w estern and eastern Turk empires. A fter 680 the eastern Turks, subjugated by the Chinese, revolted and established the Second Turk K haghanate. This w as brought to an end in the 740s by a tribal confederation led by the Uighurs. The Uighurs, headed by an alliance o f the N ine O ghuz ( Tokhuz Oghuz) tribal union, quickly rose to becom e rulers o f today’s M ongolia and the eastern steppes. Uighur rule was in its turn broken in 840 by the Khirghiz. The focal point o f the Uighurs was shifted to later Turkestan. Three sm aller states w ere created, o f w hich Turfan and Gansu rose to prom inence. Khirghiz hegem ony on the steppe was shattered by the arrival o f the Khara-K hitai around 1112. The chancellery o f the eastern territories o f the First Turk K haghanate was in Sogdian hands, and as an inscription from B ughut (betw een 572 and 580) attests, Sogdian language and writing w ere em ployed. A section o f 6 th-century Sogdian coins and objects can be directly linked to the Turks. During the second h alf o f the 7th century a new form o f script using carving techniques spread across the entire steppe region. We will com e back to the western, that is the Eastern European, versions later. The Turkic language replaced Sogdian during the Second Turk K haghanate, and a runiform script created to m eet the dem ands o f the Turkic language was introduced in place o f Sogdian writing. This w riting quickly spread and becam e dom inant throughout the eastern h a lf o f the steppe, although Sogdian w riting and language were reverted to on occasions. A t the end o f the 10th century other form s o f w riting overshadowed the Turkic alphabet. The m ajority o f Turkic runiform script is to be found in inscriptions carved into stone, but we also
The sources
81
know o f texts being written on paper, scratched onto objects or engraved onto coins. We now know o f nearly 300 objects displaying Turkic runiform inscrip tions. These are usually divided into seven geographical groups: 1. the area around Lake Baikal and the R iver Lena; 2. the area around the R iver Yenisei; 3. the area o f the Republic o f M ongolia; 4. the area around the Altai M ountains; 5. Eastern Turkestan and the D unhuang caves; 6 . Khirghizistan and Khazakhstan; 7. Ferghana, the Altai M ountains and North Tokharistan. Contrary to earlier opinions w hich believed that the inscriptions o f groups 2 and 4 w ere the earliest, today we are certain that the oldest inscriptions come from group 3. The m onum ental inscriptions o f the Second Turk Khaghanate and the U ighur Khaghanate are to be found here. Some thirty inscriptions have been found on the territory o f the Republic o f M ongolia. Am ong them the most im portant (which can be dated) are: Inscriptions o f the First Turk Khaghanate: 1. Koli chor or Ihe Hoshiitii inscription 720-725; 2. Ongi inscription, 720 or 732; 3. Ihe Ashete inscription, 724; 4. M inister Tonyukhukh inscription, 726; 5. Kol Tegin (prince) inscription, 732; 6 . Bilge Khaghan inscription, 735. Inscriptions o f the Uighur Khaghanate: 1. Taihir or Hoitu Tam ir ten inscriptions, the seventh 735, the others between 744 and 756; 2. Tez II inscription, 750; 3. Terh inscription, 753-754; 4. M oyin chor or Shine usu inscription, 759-760; 5. Sevrei sum un inscription, 763; 6 . K harabalghasun I inscription, 810 or 821; 7. Suj inscription, 840. Some o f the other M ongolian inscriptions can only be dated far more generally, and in some cases they cannot yet be ascribed any date at all, but none can be later than the m iddle o f the 9th century. The Kol Tegin and Bilge
82
Methodological introduction and the sources
K haghan inscriptions are twin inscriptions containing partly identical texts, and are to be found standing by the R iver O rkhon, which is w hy these inscriptions, and indeed som etim es the w hole group, are com m only called the Orkhon inscriptions. These inscriptions date from some tw o hundred years before the M agyar Conquest. They stand in the eastern part o f the steppe where to the w est the M agyars lived in close contact w ith the Turkic peoples. The m ajority o f the inscriptions serve as m em orials to the dead. The texts recording the lives o f dead people, and praising their heroic deeds, are contem porary and for the m ost part authentic. The texts are not copies, but having defied the elem ents for centuries they survive to this day as originals w here they were first erected. As the U ighur Khaghanate flourished so it attracted the proselytising religions. Soon the M anichaean, the Nestorian C hristian and the Buddhist faiths and m issionaries— who w ere not above rivalling each other for souls— converted large num bers o f Turkic groups to their faiths. In 762, the U ighur K haghan converted to M anichaeism. The three m ajor religions brought with them different scripts. B uddhist texts w ere w ritten in an U ighur form refined from Sogdian and in the so-called Brahm i script a Central A sian version o f Indian writing. M anichaean texts were recorded in the w riting o f the Iranian M anichaeans and later in Uighur writing. The N estorians used a version o f Syrian writing. First religious literature translated from foreign languages, and then religious texts originally in the Turkic language appeared in these different forms, although those texts which date from the 9th century also contain references to contem porary history. The beginning o f the 1Oth century saw the birth o f geographical and historical records. A significant proportion o f the literature from the period im m ediately follow ing the downfall o f the Uighur Em pire and the establishm ent o f the southern Uighur states has com e to light from excavations carried out around Turfan and in the famous Dunhuang caves. Am ong this m aterial are originals and m any early, p re - 12 th-century, copies.
j ) Tibetan sources A t the beginning o f the 7th century a new em pire arose on the Tibetan plateau, bounded by the Him alayas, the Turkestan Basin, and China. B y the m iddle o f the 7th century the Old Tibetan Em pire had strengthened to such a degree that it rated as an im portant historical factor in Eurasia. Troops o f the Tibetan king m arched to the Chinese capital in 763, w hile Turkestan and the border regions o f China cam e under Tibetan rule. In the battle o f the R iver Talas o f 751, the Tibetans took the side o f the Arabs against the Chinese. Reports that the
83
The sources
Tibetans enjoyed close ties w ith the Turkic w orld com e not only from Tibetan but from old Turkic sources as well. A Tibetan m inister was present at the burial o f Turk khaghan Kol Tegin in 732. Tibetan rulers adopted a northern variation o f Buddhism , and at the beginning o f the 7th century a Central Asian version o f Indian writing was adapted to the requirem ents o f the Tibetan language. Thus was established the astonishingly rich Tibetan w ritten litera ture. The very first texts were religious texts translated from Sanskrit and Chinese, but from the m iddle o f the 7th century independent Tibetan literature also appeared. The golden age o f the Old Tibetan Em pire w as brought to an end in 842 by an anti-B uddhist uprising. The m ajority o f m onasteries in central Tibet along with their libraries were destroyed, although inscriptions survived. However, Tibetan literature located in m onasteries on the fringes o f the Old Tibetan Empire passed unscathed through these difficult times, and num erous copies o f central Tibetan w ritten literature were also preserved here. The fringes o f the empire, and particularly the Chinese border region, also m aintained their traditions o f independent literature; 9th- to 1lth-century texts were stored in the walled-in Dunhuang cave library, and have com e to light in excavations in Turkestan.
Figure 11 The Old Tibetan Empire and its contacts
84
Methodological introduction am i the sources
Buddhism received fresh im petus at the end o f the 11th century, followed alm ost im m ediately by a resurgence in Tibet and in m onastic life in the country. The rich religious and historical literature o f Tibet has preserved, in copies or in extracts, m any early sources which in their original form have long been lost, or if they did survive are barely legible, such as som e central Tibetan inscriptions. It is possible to divide Old Tibetan sources into four groups according to content and character: 1. Charters, for instance royal edicts, m ilitary and other adm inistrative documents; 2. Inscriptions, am ong them epitaphs, privileges and inscriptions recording the transfer o f gifts; 3. Statutes; 4. Historical works, thus chronicles, annals, works o f church history. The rich collection o f adm inistrative m aterial includes a great deal o f local inform ation related to the Turkic peoples. Am ong the best known are reports o f U ighur envoys, the extracts o f which were translated into Tibetan, and which contain im portant inform ation on the Turkic peoples, including the Pechenegs. Currently, we know o f 18 central Tibetan inscriptions, the oldest dated one being the Zhol inscription from 764, the last the text o f a peace treaty betw een the Tibetans and Chinese inscribed in two languages and dated 821/822. The O ld Tibetan Annals record events on an annual basis starting with the year 650. Several chronicle fragm ents have also survived. The great Old Tibetan Chronicle is the m ost com prehensive, even though som e om issions in the copy have caused serious problem s in textual criticism . It gives an account o f the legendary forebears o f Tibetan royalty through to events from the first decades o f the actual foundation o f the m onarchy, w riting about the adoption o f Buddhism but, im portantly, w ithout the characteristic distortion o f later Buddhist historical works. The text includes not only reference to the Turks under the nam e Dru-gu, but num erous other Turkic peoples, for instance the Uighurs, Turgesh, Kharlukhs, Basm ils, Khirghiz, Pechenegs, the different Iranian peoples and the Arabs. N aturally enough, the focal point o f history is the relationship with the Chinese. This was not only a m ilitary-political struggle, but also one betw een the different trends o f Buddhism , one prom oted by China and one opposed to China. The reason this is im portant from the aspect o f historical sources is because in the historical events and interests behind these apparently purely B uddhist religious debates are concealed the peoples o f Central Asia, am ong them figures from steppe history, and the events surrounding these peoples.
85
The sources
к) Chinese sources A brief history of China The core o f ancient C hina was established at the beginning o f the 2nd m il lennium BC on the central region o f the Yellow River (Huang-he), at the point where the north-south flowing river suddenly breaks to the east, at the confluence with the R iver Wei. Before long, the culture spread towards the coast, and in the Bronze Age, some time around 1600 BC, the C entral Empire, the very core o f later China was established partly under the legendary Shang or Yin dynasty. The Zhou dynasty centred on Loyang was form ed in the 11 th century BC, rapidly extending its authority over a large new territory. The rule o f the Zhou dynasty lasted— with some m inor breaks— until the 3rd century BC. The end o f this period was m arked by constant w arfare betw een individual vassal states across China. The succession o f dynasties ruling a unified China was ushered in by the rise to power o f the Qin dynasty. In the 3rd century BC (206 BC), the Han dynasty took over. The Great Wall o f C hina was constructed to keep out the X iongnu or Asian Huns who launched attacks from the north. The Early or Western Han dynasty ruled until 9 AD, to be followed by the Later or Eastern Han dynasty (25-221 AD). With the collapse o f the Han dynasty, China split into three parts, a situation which lasted until 264. A fter this, a bew ildering succession o f short-lived dynasties ruled the north or south. Only one o f these dynasties is o f interest to us: the Tuoba Wei dynasty which ruled the north from 386 to 532, because they were o f M ongolian descent. China was reunified around 580 AD by the Sui dynasty, to be succeeded by the Tang dynasty which retained pow er until 907, when China once again fragmented. Five dynasties ruled alongside and fought each other. In 960, the Song dynasty once again unified China. In the north the K hitai Liao dynasty rose to pow er in 907, and the Jurchen Jin in 1115. This latter dynasty together with the Song dynasty were overthrown by the M ongols, who ruled under the nam e o f the Yuan dynasty until 1368. The Chinese M ing dynasty, successors to the Yuan, were in turn suppressed by the M anchu Qing dynasty in 1644. The dynasty o f the M anchus was brought to an end in 1912 by the first Chinese Revolution. It is thus clear that the sim plest way to divide up Chinese history is by dynasties, with subdivisions by rulers. Rulers had one or m ore so-called ruling periods, m arked by im perial slogans (nienhao). The year o f notable events is given by the year o f the im perial period.
Metnoiiological introduction and the sources
. % > . Lake И r . U L ake B alkhash
E
Tim qhuz КОШ1Л:
a
s :ь г
-'l
и
г
к
л~
, - d m 700 J f
J
K harhikhs JAPAN
K hasghar / "' M
Logins 'Changan . i
, .
1langchou* С
^ H ^ lV
NM
, ,«*/
1
^
■ipulrtl
X
j’
•
v
( £«.vr
A
ch in a /
«
Sea
-i I fO / /I
4 1 .^ 4 г % Mrihes..
I N D I A
n
.У
Va
B ay o f B engal
f >
4'-J
CY
i r ^
\
c,o^ 1
\ \ K1IMKR ’ % \ .
Figure 12 Central Asia and China under the Tang dynasty (618-907)
Chinese script C hinese script, as with the written forms o f m ost o f the w orld’s languages, evolved from icons. The earliest extant exam ples date from the Shang dynasty. The writing system , whose fundamental principles have rem ained unchanged to this day, took shape relatively quickly. Barring a few exceptions, Chinese characters consist o f two elements. The conceptual elem ent carries the ‘m ean ing’, while the phonetic elem ent indicates the pronunciation. We can get a general idea o f this if we take as an exam ple the w ord bank, m eaning the bank o f a lake. The Chinese would represent this word by a small drawing. Now, if we had to write down the words bank ‘bench’, ‘ro w ’ and ‘bank’ as an institution, in the first case the Chinese w ould write, or to be m ore precise, draw, the w ord bank for ‘shore’ and add a small draw ing o f a bench; in the
The sources
87
Ж ,' u \ o
/ (At I’xilkha^h
^EMPIREKhl/yhU JAPAN KOREA -Х1.Л
NORTH
Khasghar ;
SO NG
DYNASTY
* K h o t< m
( »iii> t ■'
J a n g liu
1030
TIBET Lhasa
v
*
Brahmaputra
C-\
'V'.. Hangchou
Llma
.It.
Se>i
■■ ■ /
SOUTH SONG DYNASTY
'‘T&1K
1 M Dl I A ’agatji
\
>tribesi АШАМК j,
A' В a ¥ a f В e t:;; a I
KHMER
Figure 13 Central A sia and China under the Khitai Liao (907-1119) and the Chinese Song dynasties (960-1278)
latter case it would add a small sym bol for money. In the first case the m eaning would be bank ‘bench’, in the second bank ‘institute which deals with m oney’. If w e then had to write bank counter, tw o signs w ould be needed. The first would consist o f the phonetic elem ent for bank ‘shore’ and the sem antic elem ent ‘m oney’; the second sign o f the phonetic elem ent for ‘shore’ and the sem antic for ‘b ench’. It w ould be pronounced ‘bank bank’ but actually m ean ‘bank counter’. O f course these are non-existing exam ples, and they serve to give an idea o f how Chinese script developed. This peculiar form o f pictographic writing has one serious difficulty. Given that both the sem antic and the phonetic part o f the word are represented by pictographs, it does not follow changes in the (spoken) language. The stock o f sym bols has in principle remained unchanged for thousands o f years, and neither m inor nor m ajor sim plifications have altered the essence. The contem porary pronunciation o f
88
Methodological introduction and the sources
a given Chinese character m ust thus be separately determ ined according to the age in question. Scholars have carried out reconstructions with the aid o f contem porary rhyme dictionaries and other sources, and today it can be reconstructed with a m ore or less reliable degree o f accuracy how a given character was pronounced in a given period o f history. Each character, then, has a current pronunciation and num erous earlier pronunciations. The current pronunciation is dependent upon the region o f China w here the word is spoken, since the various dialects o f contem porary China are so far rem oved from each other as to be m utually incom prehensible to their respective speakers. N ever theless, they are connected in their use o f a com m on w ritten form, albeit pronounced in a variety o f ways. For this reason, it is extrem ely difficult to transcribe Chinese characters into the Rom an alphabet. The official pronun ciation o f the language today is that o f the capital Peking (Beijing). In sim pler Chinese dictionaries, the pronunciation o f Chinese characters is rendered into Rom an script according to the spoken dialect o f Peking, and this is also the version used by international scholarship. If, however, a m ore ancient pronun ciation o f a given character m ust be reconstructed, then only a scholarly phonetic transcription will suffice. Sinology m akes a distinction betw een A ncient Chinese, M iddle Chinese, Early M odem Chinese and C ontem porary Chinese pronunciation. A ncient Chinese, form erly know n as A rchaic Chinese, derives from the earliest period from which tangible evidence rem ains, the Han period. M iddle Chinese pronunciation has an earlier variant, roughly reflecting the state o f the language as it evolved around 400 AD and was set down around 600 AD, as well as a later variant, reconstructed based m ostly on sources from the Tang and Song dynasties. This latter is not only a later variant, but represents a different dialect. Early M odem Chinese gives us a phonetic picture o f the so-called M andarin, or clerical language current in the Yuan dynasty, w hich is closely related to M odem Chinese.
Transcription of Chinese script There are m any know n m ethods o f transcribing Chinese script into the Rom an alphabet. Recently, however, use o f the so-called pinyin system o f transcrip tion into Latin script, as officially adopted by the Chinese them selves, has been gaining ever increasing ground. The characters it uses m ust generally be pronounced as they would be by an English speaker. H ow ever “q ” is like ch, “r” like zh, “x” like hs. In this book all Chinese proper nouns and other w ords occur in p inyin transcription.
89
The sources
Chinese historical sources Am ong surviving Chinese inscriptions, the greatest am ount o f historical data is contained in grave epitaphs and titles o f privilege. These latter relate historical events where, by virtue o f their participation, the person concerned gained certain privileges. During the reign o f each successive dynasty, official court records were continuously kept, arranged according to years. W hen an im portant individual passed away, a biography would be prepared. O f great im portance as a literary genre in the Chinese state was the petition, in which zealous functionaries or subjects w ould call the attention o f the Em peror and his court to certain events and tasks at hand. The court itself w ould publish a variety o f docum ents, copies o f which w ould be stored in the im perial archives. These contem porary w rit ings would also be consulted after the fall o f each dynasty, when the succeed ing dynasty would have its historians set down the history o f its predecessor. Chinese historical writing em ployed a rich schematic system. Particularly when describing the history o f foreign, barbarian peoples, it showed a special predilection for formulas adopted from earlier times. Thus it was not always o f vital im portance to record the exact provenance o f a new ly em erged people, nor to reveal from which earlier people they m ight be descended, nor even what might be the real nature o f their custom s, but rather to adapt one o f the already recognised m odels to their description o f the newcom ers. Chinese historical literature proper begins with the work o f Sim a Qian (145-86 B C ) known as the Shiqi, prepared by order o f the Han dynasty in around 90 BC. In this the author sets down the history o f China from its m ythical beginnings up to his own time. The basic structure o f the work breaks down into five chapters: (1) M ain chronicles; (2) Chronological charts; (3) M onographs (on cerem onies, m usic, the calendar, astronomy, etc.); (4) He reditary families; (5) Biographies. In essence this structure becam e the model for subsequent works, while the Shiqi’s depiction o f barbarian peoples sim i larly served as a m odel for later authors to follow. The history o f the Early Han dynasty was com piled prior to 92 B C , and that o f the L ate H an dynasty in around 300 A D . The history o f the Three K ingdom s was prepared in around 289, and that o f the Yin dynasty ( Western Yin 265-419) only as late as between 644 and 646. The chronicle o f the Song dynasty was com pleted in 488, that o f the Southern Chi in around 530, and those o f the L iang and Chen dynasties in 636. The annals o f the M ongol-descended Tuoba Wei dynasty w ere finished in about 554, and those o f the Northern Chi and Zhou dynasties in 636. A com prehensive history o f the northern and southern dynasties was also put together in 659. An older chronicle o f the Tang dynasty was com pleted in 945, and a newer, revised and updated version, the new Tang chronicles, in 1060.
90
Methodological introduction and the sources
Besides these dynastic histories, m ediaeval China also saw the com pila tion o f several encyclopaedias, sim ilarly rich in historical m aterial arranged according to topics or peoples. As early as the 11th century, collections o f historical docum ents were assembled, wherein copies o f earlier records were stored in them atic order. These dynastic chronicles and other historical sources have not survived in their original form, and m odem critical editions have been prepared based on a variety o f m ultiple copies, block prints and revised versions. R ecent archae ological excavations, however, have turned up a num ber o f older editions or fragm ents thereof. The dynastic chronicles, w hich w ere always put together by the succeeding dynasty, partly to seal the legitim acy o f its own rule, are generally reliable with regard to the area w hich concerns us, the history o f the steppes. A lthough the Chinese court was not always consistently ‘far-sighted’, there w ere periods when it was able to follow events as far as E urope’s border, all the way to the Caucasus or the River Volga. The chronicles o f the M ongol dynasty even give an account o f the Tatar invasion o f Hungary in 1241-1242. Besides m ilitary expeditions, a succession o f Chinese travellers, envoys and B uddhist m ission aries, who roved far and wide and then returned hom e with tales o f distant lands, greatly contributed to C hina’s knowledge o f the outside world. M any o f these reports have survived, as they were filed aw ay am ong official records. The nam es o f peoples and places featured in Chinese sources are exceedingly difficult to m ake consistent with the names in w estern sources, having been passed down through a variety o f different speakers. The Chinese form often does not provide the local pronunciation o f a place nam e or people, but rather reflects the form used in the language o f some interm ediary m erchant, people or neighbour.
I) The Hebrew sources The Jew ish people scattered far and wide in large num bers in the 3rd century, into the various provinces o f the Rom an Empire. Their situation was particu larly favourable in the Caliphate o f Cordova. Although by the 10th century Jews had established settlem ents in alm ost every country in Europe, the m ajor m igration took them up from the Iberian Peninsula into the regions o f the Rhine, where a section o f the Jewry adopted a local G erm an dialect which later evolved into Yiddish. The m ore significant Jewish centres in the C aliph ate o f Cordova, France and Italy, and subsequently in the Rhine provinces, kept in close contact with each other, building up a fully-developed trading netw ork organised by the Radanite m erchant house from its seat in M arseilles.
The sources
91
The C ordovan ruler Abdul Rahm an III (912-961), under whose reign the Cordovan state blossom ed, m aintained contacts with every significant country in the world, and even m anaged to take over the institution o f the caliphate from the failing Abbassid dynasty. His m inister o f finance, the Jew H asday ibn Shaprut, learned that there existed a country in the east, the K hazar Empire, whose ruler was o f the Jewish faith. First through the m ediation o f Byzantium , and when this failed, via Europe and hence through Hungary, he dispatched a letter to Joseph, the sovereign o f the Khazars. The khaghan received the letter and replied to the greater part o f the queries contained within, although evading those o f a m ilitary nature. A lengthy debate has evolved around the letter and the question o f its authenticity. The original m ust have been w ritten before 962. The reply o f the khaghan Joseph, m eanw hile, is extant in both a shorter and a longer version. Although the m anuscripts currently known are certainly copies, the texts are quoted in detail as early as the first decade o f the 12th century by the Catalonian Judah ben Barzillai, and so even if they were to prove to be forgeries, they are still docum ents o f considerable age. A num ber o f errors in the text point to an interm ediate Arab source. The St. Petersburg m anuscript o f the longer version m ay originate from the 13th century, while the shorter redaction o f H asday’s letter and khaghan Joseph’s reply survives only in the form o f a 16th-century m anuscript kept in Oxford. The St. Petersburg m anuscript was brought from Cairo in the 1860s. The entire correspondence was published as early as 1577, together with the other two texts also based on the Cairo copy. Several other H ebrew docum ents are also known from the 10th century. Im portant am ong these is the letter from Kiev known as the Schechter text, m entioned in regard to the alleged Savarti Asfali name o f the Hungarians (see p. 288). This is a letter to H asday from an unidentified Jew concerning the Khazars. The Sepher Yosippon was w ritten in Italy around 940. In the light o f these w orks, the question o f the authenticity o f other contem porary correspondence regarding the K hazars also becam e clearer. The Jews were driven out of England in 1290, and from France in 1394. In 1492, the year o f the discovery o f the N ew World, the arm ies o f Ferdinand and Isabella put an end to Arab rule in Spain. The great m ajority o f the Jewry, which had until then lived under the shelter o f Arab rule, fled eastw ard to escape the new inquisition. Passing through the O ttom an Turkish Em pire, they reached the territory o f today’s Ukraine, Poland and Lithuania, where they met with Jews who had been continuously m igrating there through Germ any since the 12th century. East European Jews thus m igrated from the w est to the east o f the continent, and w ere not descended from the inhabitants o f the Khazar Empire.
Methodological introduction and the sources
92
&
I Am erica r .^ : :AmfASaC-
...»j
j . I „ £_Л
« = '/
I ^
Y*
' ^
/J.
^ ‘
, ,
И З 6th -
V
x' I ft
i ij»n "P
.
LU
• 4
e m p ir e
IftjNHAND; *>,•; *•••;•.\
Gcnnv.i
- 1Й ^
11th centuries
L....J 11th - 13th centuries E H 14th century
S Л
..... t.l HMAN' V " ' B O M A N -P ^
M .W C E
1?
<
^XX,;
'
■
,
I.
•x? / ^ 1-'} I 'V.-y/' -M
7
«''t
The spread of Jews
\
s%r S
Kiev
15th - 1 6th centuries
'V
..
f дю ,
„
о ш ди л
Venice., ‘
,
ARAGONI
>DjSttKW us Ashkenazi migration < = 18th - 20th centuries 19th - 20th centuries Sephardi migration > 15th - 18th centuries •*“ — 17th - 18th centuries ---------State boundaries in the 15th century
Jerusalem
Figure 14 W anderings o f the Jews expelled from Spain A large portion o f works w hich described the history o f the Jews in the 9th and 10th centuries perished or w ere lost, but fragm ents remain. Am ong these, particular value m ay be attached to the observations o f the R adanite m erchants who m aintained contact w ith the conquering M agyars.
4. THE LANGUAGE AS A SOURCE The Hungarian language itself represents a source for H ungarian history. The language retains the circum stances o f its own genesis, and preserves each and every contact through history w hich involved linguistic com m unication in one form or another. For this reason, the Hungarian language can be seen in a certain regard as the im print o f Hungarian history.
The sources
93
Words change in a given language according to the rules o f that language. However, if a w ord is adopted by another language, then from that point on it no longer follows changes in the original language, but rather those o f its adoptive language. Such adoption, or borrowing from another language, always m eans assim ilation into the system o f the adoptive language. Let us take an example. In the Proto-Indo-European language, the num ber five could be heard as penkve. Through the m illennia this steadily evolved in each o f the various Indo-European languages, w ith the ancient penkve becom ing the Germ an f u n f ( by way o f the transitional form fin ke), the English fiv e , the Latin quinque (by w ay o f pinkve), the ancient Slavic penti (by way o f p en k i), the G reekpenta, and the Indiepancha. The Latin quinque then becam e the Italian cinque and the French cinq, while the ancient Slavic penti becam e the Russian pyaty, the Bulgharian, Serbian and Croatian pet, the Slovak p a t, the Polish pyench, and so on. This Indo-European word is also recognisable in at least three forms in the Hungarian language. The Hungarian w o rd p en tek ‘Friday’ is boiTOwed from the Old S lavicpen tik ‘fifth (day)’, w here the IvJ phonem e is still preserved, as it also is in sim ilarly Slavic-derived Hungarian words szent ‘saint’ or ‘sacred’, and donga ‘stave’. M eanwhile, our w ordp eta k (which survives in the saying egy petdkot sem i r ‘it’s not w orth a penny’) comes from the Croatian form o f this same w ord, p etek ‘five (-penny piece)’, where the phonem e /x\I is no longer present. Thirdly, the Hungarian w ord pun ko sd ‘Pentecost’, although not taken directly from the Greek, conceals the Greek penta-kosta ‘fiftieth (day after E aster)’. For a time there were no words in Hungarian that began with sk-, sp-, st- or szk-, and thus the Latin schola becam e iskola ‘school’, the Slavic-Avar span becam e ispan ‘a title’, the Germ an Strang becam e istrang ‘traces (o f a harness) ’, and the Latin Stephcmus becam e Istvan ‘S tephen’, in order to conform to the rules o f the Hungarian language. A fter a while, however, this rule ceased to apply, and there is no longer need to prefix consonant clusters at the beginning o f new ly adopted words, such as, for example, the H ungarian studio ‘studio’, skala ‘seale ', start ‘start’, sparga ‘asparagus’, and spendt ‘spinach’. A lthough the latter word has been m entioned in this form in various sources since 1500, to this day certain dialects are nevertheless still fam iliar with the form ispindt. These regularities m ake it possible to attach a date to the various changes and adoptions o f words, as well as to the evolution o f these regular forms. At the same time, they help us to draw conclusions about the way in which contact betw een speakers o f the languages concerned m ay have developed. The com m on Urheimat o f the Uralic peoples m ust have been located where the four different species o f coniferous tree (‘spruce’, ‘arolla p in e’, ‘silver fir’, and ‘larch’), as well as the deciduous ‘elm ’, could be found together in one region. This is to say that the nam es o f these trees were com m on to all the
94
M ethodological introduction and the sources
U ralic or Finno-Ugrian languages. Even w hen the w ords w ere lost from one language or the other, as for exam ple happened with Hungarian, it is still possible to reconstruct their earlier existence. The period concerned here was following the last Ice Age, at the flowering o f the Neolithic era. The com m on presence o f these trees can be proven through pollen analysis and other scientific methods. The region itself probably covered roughly the territories o f the Volga and Kam a rivers and the m id-ranges o f the Ural M ountains, but m ay also have extended som ew hat beyond the Urals, w hich never constituted a geographical dividing line. The shared territory o f the various peoples speaking the Proto-U ralic tongue probably began to break up around 4000 в с (see also pp. 34-35). Attem pts both past and present have been m ade to go yet one step further back in time, to postulate the theory that the Uralic languages once belonged to an even larger fam ily o f languages. Views o f this kind include tw o w orthy o f special m ention here. The so-called N ostratic theory (from the Latin word nosier ‘ou rs’) endeavours to prove that the Uralic, Altaic, Indo-European, Sem itic-H am itic, Dravidian and K artvelian language groups (the latter includ ing G eorgian and its close relatives) can be traced back to a com m on ancient language, to ‘o u r’ language. A lthough this theory, propounded by H. Pedersen, and elaborated by a research team in M oscow, is today enjoying som ething o f a second flowering in the United States, it is im probable and uses com plex suppositions and unprovable assertions to bolster its argument. From our view point, moreover, this hypothesised ancient language w ould be so far in the distant past— around 7-8000 BC, or some 10,000 years ago— that w e could still happily disregard it even if fresh data were to back up the supposition. The prospects o f such data em erging, however, are non-existent. The other theory to be m entioned traces back the U ralic and A ltaic lan guages to a com m on, ancient parent. No notew orthy exam ple o f the UralicA ltaic linguistic relationship rem ains today. However, interesting structural parallels do exist betw een the tw o language groups, and as w e shall see, early interlinguistic contact did occur. Still, this is insufficient for the reconstruction o f a com m on ancient language. Attem pts have been m ade using linguistic m ethods to determ ine w ho m ight have lived in the vicinity o f this, then still scarcely heterogeneous, Uralic community. Two linguistic groups have been taken into serious consideration. The Altaic languages w ithout doubt came into contact w ith other languages at a very early stage. Such contacts, however, can be divided into two dis tinctive groups. Although a num ber o f early analogies betw een the U ralic and Turkic languages do exist, for the tim e being these are very few in number, including, for example, the nam e o f a species o f w ild duck, the angga. M eanw hile, a surprising, albeit not very great, num ber o f analogies can be
The sources
95
M ANCHU -TUNGU ZIAN U RALIC
M ONG OLIAN TURKIC
IND O-EUROPEAN
Figure 15 Relative locations o f the Uralic and the neighbouring peoples
found betw een the U ralic and M anchu-Tunguzian languages. For example, the ancient Uralic form o f the Hungarian eger ‘m ouse’ can be found in every language o f the M anchu-Tunguzian family. A t the same time, no early contact whatsoever can be proven with the third Altaic subfamily, the M ongolian. As we shall see later, the relationship betw een the various subfam ilies o f the Altaic language group is likewise hotly disputed. For this reason, we are able to say that while the speakers o f the Uralic languages lived in a region where they cam e into loose contact with the ancestors o f speakers o f the Turkic languages on the one side, and o f M anchu-Tunguzian languages on the other, they did not live together with a people w ho spoke a com m on, ancient Altaic language. Early connections between the Uralic and Indo-European languages have also been exhaustively investigated. This is o f great im portance because the Indo-European languages have bequeathed us a very ancient and diverse body o f historical source material, and consequently the early evolution o f the Indo-European languages can be far m ore reliably reconstructed than that o f languages o f which only com paratively recent historical evidence remains. Here, too, genuinely early and indisputable analogies are very few and far between. Nevertheless, the Hungarian words in ‘sinew ’, nev ‘nam e’ and vas ‘iron’, ancient words which can be dated back to the Uralic period, ultim ately are o f Indo-European origin. Thus the word in is o f identical origin to the Germ an Sehne ‘sinew ’ or ‘string’, w hile the Hungarian nev shares the deriva tion o f the English name. M eanwhile, the sim ilarly Indo-European-derived Hungarian w ord vas ‘iron’, probably with the original m eaning ‘o re’, was a word denoting a naturally occurring m etal w ith which the U ralic com m unities may well have become acquainted even before the Iron Age.
96
Methodological introduction and the sources
The fact that early analogies scarcely exist (but do) betw een the Uralic and Turkic, Uralic and M anchu-Tunguzian, and U ralic and Indo-European lan guage groups suggests that contact could not have been very intensive, and that neighbouring peoples w ere not on a significantly dissim ilar, m ore ad vanced level o f culture; in other words, contacts such as they w ere m ust have been very early indeed. All things considered, linguistic data indicates that the Uralic Urheimat was situated som ewhere between the Turkic, M anchu-Tun guzian and Indo-European peoples, or at least in a territory w here contact betw een these peoples was possible. This territory m ost likely com prised the central and the southern ranges o f the Ural M ountains. The M anchu-Tunguz w ould have been to the northeast o f this, the Turkic peoples to the southeast, and the Indo-Europeans to the south and southwest. Based on the available linguistic data, therefore, the oldest area o f habitation that we can reconstruct for the ancestors o f the Uralic peoples, am ong them groups speaking the Hungarian language, is represented by the central and southern regions o f the Urals. In around 4000 BC, the ancestors o f the Finno-U grian peoples separated from the larger Uralic community. The Indo-European peoples also scattered abroad, by w hich time the com m on ancestor o f the later Indie and Iranian languages— known as Indo-Iranian— had already becom e detached. From this point on, evidence can be dem onstrated in the Finno-U grian languages solely o f contact with the Indo-Iranian, and subsequently Iranian and Tocharian languages. The lands inhabited by speakers o f the Finno-U grian languages expanded significantly on the western side o f the Urals, chiefly tow ard the w est and northwest. At the same time, to the south and southeast, the A ncient Iranian peoples had begun to advance northward. This can principally be ascertained from the large num ber o f words borrow ed from A ncient Iranian to be found in the Finno-Ugrian languages. The Ancient Iranian period lasted from around 4000 BC until about 800 BC. Hungarian words such as dr (in its sense o f ‘value’), dr (in its sense o f ‘tool’), haz ‘house’, meh ‘b e e ’, mez ‘honey’, szaz ‘hundred’, szarv ‘horn’, agyar ‘tusk’, ostor ‘w h ip ’, tehen ‘co w ’, tej ‘m ilk \ f e j n i ‘to m ilk (a cow )’, het ‘seven’, and m any others besides all derive from Iranian languages o f various periods. We can also see from these words that the Finno-U grian peoples came into contact w ith a people w ho was fam iliar with agriculture and kept animals. The beginning o f these contacts can be established by exam ining Iranian words w hose date o f origin is easier than m ost to pinpoint accurately. The oldest forms o f these, as they appear in the Finno-U grian languages, can be dated to around the 3rd m illennium BC. As early as around 2000 BC, the com m on ancestors o f the peoples who speak the Voghul, Ostyak and Hungarian languages, the Ugrians, separated from the
The sources
97
Finno-Ugrian linguistic community. N evertheless, the contact with the Iranian languages endured. W hile a significant num ber o f words borrowed from the A ncient Iranian period can be found in the Ugrian languages, the situation changed m oving into the early part o f the O ld Iranian period. A few words such as arany ‘gold’ (which in the other Ugrian languages m eant ‘copper’), or tal ‘d ish ’ (a w an dering word o f Indian origin), tend to suggest contacts o f a com m ercial nature. Attem pts have been made to dem onstrate that words w ere also borrow ed from the Turkic languages during the Ugrian period. H attyu ‘sw an’, szd ‘w o rd ’, Id ‘h orse’, ho d ‘beaver’, ir ‘w rite’, hajo ‘boat’, hom ok ‘sand’, nyak ‘neck’, nyereg ‘saddle’ and nyar ‘sum m er’ have been suggested. However, these suppositions have not weathered critical examination. Only the argum ent for szd ‘w ord’ holds w ater in every respect, while the word ir ‘w rite’ has been independently borrowed in both Hungarian and the O b-U grian languages. The remaining words present unsolvable phonetic or semantic problem s either from the Turkic or the Ugrian side, or both; or are adoptions from the Turkic not in the Ugrian period (hajo, homok)\ or are quite clearly o f non-Turkic origin (nyar). These findings together indicate that the population w hich spoke the Ugrian language m ay have lived in the first h alf o f the 1st m illennium BC in a region where they did not come into close contact with either the Iranian or the Turkic world. At the same time, a good num ber o f H ungarian words from the Ugrian period (lo ‘horse’, nyereg ‘saddle’, f e k ‘bridle’, szeker ‘wagon ’) suggest that the Ugrians were already an equestrian people, although they probably used saddle-anim als prim arily for transport and hunting pur poses, rather than for more significant nom adic wanderings. The separation o f the Hungarian people from the larger com m unity o f Ugrian peoples cannot be fixed to a precise point in tim e using the tools o f linguistic science. Nevertheless, it is in the Hungarian from this indeterm inate time onward that we find a significant num ber o f words borrow ed from M iddle Iranian, and within this specifically eastern M iddle Iranian, no sign o f which can be found in any other Ugrian language. W hile perhaps no m ore than six words entered H ungarian’s Ugrian-age predecessor from Old Iranian (8th-2nd centuries B C ), certain scholars believe that Hungarian borrow ed around 45 words from the M iddle Iranian languages (from the 2nd century B C , until the 7th— and in some opinions— the 9th century A D ). Although it is indisputable that both older and m ore recent adoptions can be found am ong these M iddle Iranian borrowed words, not one o f them possesses a phonetic m arker which would enable us to date these words to the first half o f the Ancient Hungarian period. On the contrary, these adopted words show the kind o f m utations which occurred in the language during the second half o f the A ncient Hungarian period. Thus, for example, the Hungarian word kincs ‘treasure’ was taken up
98
Methodological introduction and the sources
at a time when there was as yet no Hungarian word beginning w ith /g/, and consequently the Hungarian language substituted а /к / for the /g/ phonem e in the Iranian genj. A t the same time, the phonetic cluster /nch/ already becam e /j/ (as in m ancha > maj), given that the /nch/ in kincs has been preserved. M eanw hile, the word gazdag ‘rich ’ was adopted into Hungarian from an Iranian language, nam ely the ancestor o f O ssetic (com pare the O ssetic gazdig), at a point when the status o f /g/ as an initial had becom e established. The borrow ing o f Turkic words also began at this time. These, however, can be dem onstrably shown to have passed into the Hungarian language during the second h alf o f the A ncient Hungarian period. It w ould appear, then, that the speakers o f the Hungarian language, follow ing their split from the larger Ugrian com m unity, lived for a time in a region where they had very little or no contact with Iranian or Turkic peoples, but then all o f a sudden saw their language inundated with an increasing num ber o f Iranian, and still greater num ber o f Turkic words. The question is: does any trace o f this period rem ain in the language? The H ungarian language preserves not only the traces o f contact w ith other, different families o f languages, but also o f relations with kindred peoples. As far as earlier periods are concerned, such evidence is difficult to reconstruct, ju st as it is difficult to dem onstrate the existence o f Dutch loan w ords in G erm an, or Ukrainian words in Russian, but the task is not im possible. It seems that the Hungarian language holds on to traces o f its contact w ith the Perm ian languages, i.e. the com m on ancestor o f the Zyryan and Votyak languages, two such types o f traces being distinguishable. On the one hand, Hungarian has a num ber o f borrow ed words such as eziist ‘silver’ and kenyer ‘bread ’. Even m ore interesting is the adoption o f certain elem ents o f a w ord’s construction. Thus the latter phonetic elem ent o f the Hungarian num erals kilenc ‘nine’ and harm inc ‘thirty’ can be attributed to a Perm ian word m eaning ‘ten ’ (in sim plified terms: [-mis] > [-ns] > [-nc]). K ilenc ‘nine’ is the num ber ‘outside’ ten, (the num ber before ten), while harm inc ‘thirty’ is equal to harom tiz ‘three ten s’. M eanwhile, the num eral nyolc ‘eight’ has m erely taken an analogical phonetic ending. It seems likely that the order o f num erals in today’s H ungar ian m ay have taken its final shape at this time. At the sam e time, the em erging o f voiced plosives (/b/, /g/ and /d/) in word initial position occurred solely in the Perm ian languages and Hungarian, while nasal consonants preceding other consonants disappeared, thus effecting the follow ing changes in the Perm ian languages and Hungarian: /nt/ > /d/, / т р / > /Ь/, /пк/ > /g/, and /nch/ > /j/. For exam ple, the w ord agyar ‘tusk’ harks back to the form *onychara o f the Finno-U gric age. From this cam e the Ancient Perm ian *vojer (in today’s Zyrian *vodzer, in Votyak vajer), and so too in Hungarian ojar, and thence to the m odern form agyar, while the closest linguistic relatives, the Voguls,
The sources
99
have anjer, and the Ostyaks anzhar, thus preserving the nasal consonant. These m utations m ay o f course have occurred entirely independently o f each other, and also took place in other linguistic families. All the same, it does appear that, at least in this case, we are dealing with w hat am ounts to a regional linguistic phenom enon, in other words a phenom enon which occurred sim ul taneously over a large area and am ong peoples speaking a variety o f languages. This is known as an areal phenom enon. As we have indicated in the above, while in the Old Iranian period— i.e. between 800 BC and 300-200 BC — scarcely any Iranian loan w ords can be found in Hungarian, at the same time some surprising signs point to increased Iranian contact with the speakers o f the A ncient Perm ian languages. Indeed, the stock o f Iranian loan words in the Perm ian languages begins to swell considerably at this time. Speakers o f the shared ancestral language o f the Zyrians and Votyaks adopted about five times as m any Old Iranian words as the ancestors o f Ugrian language speakers did. This indicates that the Permians came som ewhere between the H ungarians’ ancestors and the Iranians, or in other words, we m ust suppose an advance o f the Iranian peoples which brought significant groups o f Iranians into the vicinity o f Perm ian language speakers, but w hich did not take them into the territories o f H ungarians’ ancestors. This m uch the language has to say as a historical source o f inform ation regarding the period at hand. The second h alf o f the A ncient Hungarian period again saw increasingly significant contacts with the Iranian peoples. This time also saw the beginning o f Turkic-H ungarian com m unication. To m ake clear how this process can be conceived in term s o f the history o f the Hungarian language, we will use here a schem atic model. In the diagram , the letter С represents every possible consonant in the given language, while the letter V represents every possible vowel. The G reek letter gam m a [y] represents a voiced spirant consonant, such as the Hungarian /h/ pronounced as voiced. The Greek letter eta [r\] sym bolises the sound w hich we find on the end o f the Hungarian words harang ‘b ell’, borong ‘be sorrow ful/cloud o v er’, and lang ‘flam e’, where we pronounce a sound som ewhere betw een M and /n + g/, as opposed, say, to our word langos ‘a large, savoury fried doughnut’, w here the /п/ and the /g/ are pronounced separately. The raised v signifies that a vowel is m ore closed, i.e. an /о/ becom ing Ini, or an / 6 / becom ing /ii/, etc., at the same time being spoken shorter. This short, closed vowel has now disappeared from the end o f words in Hungarian (for instance: hodu > had ‘arm y/w ar’).
100
Methodological introduction and the sources
Form ation o f guttural word finals in Hungarian Period
Finno-Ugrian 0
Ancient Hungarian 2 3
1
Old Hungarian 4
M iddle H ungarian 5
Category I CVKVKV CVKVCV
>
CVKKV
>
CVKKV
>
CVKv
>
CVKv
>
CVK
CVnKV
>
CVtiGV
>
CVGV
>
CVGV
>
CVG
CVKVKV C V t|V C V
C ategory II CVKV
>
CVKV
>
CVGV
>
CVyV
>
CVhv
CVt) V
>
CVGV
>
CVyV
>
CVyV
>
CVhV
cvv >
CVV
Examples: 1. PFU *pakke > fa k k e >fa k f > fe : k ‘ >f e k ‘h alter’ 2. PFU *pw]ke > furjge > fo g f > fo g ' > fo g - ‘to grasp’ 3. PFU *yoka> yo g o > y o g u > y o h u > jo (in its form er sense o f ‘riv er’, e.g. in the latter part o f the river nam es Sajo, Hejo) 4. PFU *pat]a > fa g e > fa g f >fa h ' >fe j/fo ‘h ead’/ ‘m ain’ These exam ples o f course stand in for other words as well. Thus as far as the configuration o f phonetic sounds within the word is concerned, the word mancha, which in the late A ncient Hungarian period becam e m aja (and later m agy o f m agyar ‘H ungarian’), belongs under the second example. We can see, then, that two subperiods can be distinguished w ithin A ncient Hungarian (see colum ns 2 and 3 o f the above diagram ), and one o f these, the period num bered 2 , we shall classify as early, while the other period, num bered 3, we shall classify as late Ancient Hungarian. We have observed that words adopted from both M iddle Iranian and Old Turkic entered the language in the third period in the diagram , i.e. in the late Ancient H ungarian period. The fact that the A ncient Hungarian, O ld Turkic and M iddle Iranian periods are here revealed as contem poraneous should not bother anyone, since these chrono logical designations are connected to different periods w ithin each language’s history, and there w ould be no point, m erely for the sake o f consistency, in
The sources
101
altering designations w hich are already accepted in Hungarian, Turkic and Iranian linguistic science respectively. However, for the sake o f easy com pa rability, we have presented these periods in a table (see Figure 3 on p. 21). As the Iranian-derived word kincs ‘treasure’ and Turkic-derived kantar ‘b rid le’ or kender ‘hem p’ dem onstrate, these words entered the Hungarian language after the cluster /nch/ becam e /j/, and /nt/ becam e /d/. That is, if they had entered earlier they would have taken part in the above linguistic proc esses, and would be spelt in today’s Hungarian as *kij, *kadar and *keder, respectively. We have arrived, therefore, at the second, later phase o f the A ncient Hungarian period (see the column m arked 3 in the above diagram ). We do not know precisely when it m ay have begun, but we have found that it was after H ungarian-Perm ian contacts began. From the Iranian point o f view, given that the M iddle Iranian period began in the 2nd century B C , we can locate this period to after the 2nd century BC. From the Turkic point o f view, m eanwhile, the chronology o f this time cannot yet be supported by w ritten evidence, and thus we m ust briefly present here the set o f linguistic tools which can help us to establish a Turkic linguistic chronology. We will return afterw ards to the question o f precisely when Old Turkic borrow ed words m ay have entered Hungarian. The question o f an A ltaic linguistic affinity at this point is unavoidable. Certain scholars believe that the Turkic languages m ake up one branch o f the Altaic family o f languages. According to this view, the Turkic language began an independent life o f its own having split from the other A ltaic languages. The A ltaic family also contains, besides the Turkic languages, the M ongolian and M anchu-Tunguzian linguistic subfamilies. Renew ed attem pts have been made to include Korean and Japanese in the sphere o f languages w ith an Altaic linguistic affinity, but such theories, although proclaim ed loudly, are still only at a very elem entary stage. Various arguments supporting relationships within the Altaic linguistic family tend to collide with hard historical facts. It is indisputable that there is a great deal o f early regular correspondence between the Turkic, the Mongolian and the Manchu-Tunguzian languages. The question, however, is to w hat can these correspondences be attributed? A fter all, the fact o f their antiquity and regular correspondence rem ains valid w hether these languages are actually related, or w hether the correspondences are m erely the consequence o f early contacts am ong the respective peoples. Let us exam ine two examples. In the Turkic family, the word m eaning ‘o x ’ is spoken in most languages as okiiz, or some standard linguistic variant o f this. However, in Chuvash, the m inor language spoken in the vicinity o f the River Volga, we find a form o f the word derived from the earlier form o f okiir. M eanw hile, am ong names given to various peoples, the name Oghuz was used sim ultaneously
102
Methodological introduction and the sources
with the nam e Oghur, in other words in one group o f the Turkic languages the word ended in Ы , while in the other it ended in /г/. On this evidence, it has becom e the custom to refer to the Turkic languages either as z-Turkic, Oghuz or common Turkic, or r-Turkic, Oghur, or som etim es Chuvash or BulgharTurkic. In addition, we find the form iiker in M ongolian (or hiiker in M iddle M ongolian). Two explanations are possible. On the one hand, we begin with the view that the /г/ is the original final o f these words, which opens up two possible chains o f developm ent. The first is that the form okiir is an ancient Altaic form, which becam e iiker in M ongolian and okiir in Turkic, the latter then changing from okiir to okiiz in the m ajority o f Turkic languages. It is also conceivable that okiir is the ancient Turkic form, which was then borrowed by M ongolian. The second explanation is that the okiiz form is the elder o f the two. In this case, however, only one chain o f developm ent is possible, nam ely that M ongolian (and Hungarian) borrowed this word from a Turkic language in which the Izl > /г/ transform ation had already taken place. Thus, there are in fact a total o f three possible explanations: I
II
Altaic/- Turkic r —» M ongolian r
Turkic r
T urkics
M ongolian r
Turkic z
Turkic r
III Turkic z > Turkic r —> M ongolian r
Turkic 2
Turkic r
Turkic r
Debate over this question has been raging for alm ost one hundred years, during which time a great deal o f m inor details have been clarified. It would appear that in recent times explanation III has finally proven m ost able to w ithstand the criticism s o f general linguistics and language history. Debate only seems to revolve around w hether the /z/ or the /г/ is the original. This is to say that in Turkic there exists another /г/, found, for exam ple, in such com m on Turkic words as kara ‘black’ or er ‘m an’. These words contain an /г/, and nowhere a Izl, in all Turkic languages. The argum ent is thus based around w hat the original state o f the language was: w hether it was w hen the /z/:/r/ phonem ic opposition existed; or if it was w hen there was only an /г/, which in certain words subsequently becam e a Izl, and in others rem ained an /г/. Although researchers have attem pted to avoid answ ering this question by postulating the existence o f two kinds o f /г/, there is no evidence or likelihood o f this, nor any parallel in other languages. At the same time, from a m eth odological point o f view this is irrelevant, since if the original oppositions o f two types o f /г/ had disappeared in the O ghur or C huvash languages, this would
The sources
103
still m ean that this was a secondary, or later change, since an originally existing opposition would have ceased to exist. According to this view, then, there originally existed a /z/:/r/ phonem ic opposition in the Turkic languages, which disappeared from the O ghur sub group o f Turkic languages. It was subsequent to this process that the M on golians, and later the Hungarians began to borrow words from this language. We m ust therefore attem pt to determ ine exactly when this /z/:/r/ dichotom y disappeared from the O ghur languages. A suitable exam ple to help us here may be the Turkic w ord for ‘stirrup’. The stirrup is not a particularly ancient innovation, and how ever inconceivable it m ay seem, neither the Rom ans nor the Huns were fam iliar with them. The stirrup was originally a sim ple loop o f leather, which merely helped the rider climb onto the horse’s back. However, once horsemen began to use such loops on both flanks o f the horse, cavalry warfare was revolutionised, as it thus enabled the rider to rise up in the saddle and turn around to fire arrows at his pursuers. Leather loops w ere soon replaced by metal stirrups which at first im itated the shape o f the leather loops. This made com bat still easier for the horsem an, facilitating the use o f the sword and lance. The appearance o f stirrups can be dated relatively accurately. A num ber o f exam ples o f metal stirrups have been found in Eastern Asia dating from the 3rd century A D , and they can be seen both in Chinese il lustrations and graves from this time onward. They were introduced into Europe by the Avars. We must, however, take into consideration that the antecedent o f the metal stirrup, the leather stirrup, m ay have preceded the metal version by several hundred years. W eighing up the entire body o f data, we cannot date the appearance o f the stirrup to earlier than a few centuries before the birth o f Christ. The stirrup has the same nam e in every language o f the Turkic family, but with an /г/ in accordance with the rules o f r-Turkic, and with a IzJ in z-Turkic (taking the respective forms o f irenge, and izertge or iizengu). On the one hand, this m eans that the branching o ff o f the various Turkic languages occurred after the appearance o f the stirrup, and on the other hand it indicates that the m uch-debated phonetic change m ay also have occurred only after the appear ance o f the stirrup, i.e. in the last: centuries BC, since the w ord m eaning stirrup was involved in this process o f change. We can o f course no longer posit the existence o f a still uniform Altaic proto-language at this time. Consequently, explanation I can be ruled out for historical reasons. From a historical aspect, explanations II and III w ould also m ean that the C huvash- or O ghur-type languages evolved in the centuries before Christ. At the sam e time, as we have seen above, out o f general linguistic considerations, we m ust assum e that the /z/:/r/ dichotom y disappeared from the Oghur languages. This chronology is
104
Methodological introduction and the sources
supported by other data. Thus the earliest Turkic loan words in the Sam oyed language m ust have entered the then still uniform com m on Sam oyedic before the separation o f the Sam oyedic languages, in other words in the centuries around the birth o f Christ. Am ong these there are words w hich already bear the r-Turkic form. It w as at this time that the bulk o f r-Turkic forms m ay also have entered the M ongolian language. This view is also supported by the existence o f early Tocharian loan w ords in the Turkic language, around w hich a lively debate has flared up in recent times. All this m ust be clarified because the overw helm ing m ajority o f loan w ords w hich entered the Hungarian language in the A ncient Hungarian period were o f r-Turkic or Chuvash-type Turkic origin. M oreover, this could only have happened after the Oghur-type languages had separated from the other Turkic languages, around or shortly after the birth o f Christ. This thus serves to confirm two independent linguistic sources, nam ely the words adopted from the M iddle Iranian and Chuvash-type Old Turkic, which m ay both have entered the Hungarian language only in the early centuries after C hrist’s birth, or in term s o f the history o f the Hungarian language, during the later A ncient Hungarian period. In this way, we have o f course m erely obtained a date after w hich (post quem ) contacts m ay have been established. We m ust further look into the question o f w hether linguistic data can be isolated w hich m ight indicate when and w here the Hungarian language adopted these words. The current convention is to differentiate betw een the follow ing M iddle Iranian languages: (1) Western M iddle Iranian or M iddle Persian (southw est ern) and Parthian (northwestern); (2) Eastern M iddle Iranian, including K hotanese Saka, Bactrian, Sogdian, K hwarezmian and Alanian. With the exception o f A lanian, we know o f self-penned texts in each o f these languages. The last three have significance as far as the history o f the Hungarian language is concerned. O f these, a particularly im portant role is played by the A lanian language, o f which several dialects m ay have existed. The Jasz (Jazygian) people who m igrated to Hungary spoke one dialect o f the A lanian language, and continued to speak it even as late as the 15th century, as w e can see from a Hungarian Jasz glossary. At the same time, a very close relative is the Caucasian language o f the Ossets, o f which the two m ain dialects are Iron and Digor. A very helpful contribution to the historical investigation o f the conquering H ungarians is provided by the words borrow ed from Turkic, which, as has long been recognised by Hungarian scholarship, have done m uch to assist the exploration o f details and m ajor interrelationships. With regard to the Turkic loan words in the Hungarian language, the question to ask is how m uch can
The sources
105
these help to determ ine where and for how long the H ungarians dw elt prior to the Conquest, w hat was the nature o f their social and political structure and culture, with which Turkic peoples did they com e into contact, and w hat kind o f conditions did they find in the Carpathian Basin. The Turkic words in the Hungarian language are usually divided into three main groups: words adopted before the Conquest, those borrow ed from im m igrants follow ing the Conquest, and those taken from O ttom an Turkish. Words which entered the language during the Turkish occupation (16th—17th centuries) w e m ay set aside here, as they do not help to clarify questions relating to the conquering Hungarians. The languages o f the Cum an, Pecheneg and Uzian groups which m igrated to H ungary are o f greater interest. On the one hand, they are im portant sources regarding the language o f the Turkic peoples who lived east o f the Carpathians, while on the other hand knowledge o f them m ay be essential in singling out the borrowed words o f the pre-C on quest period. In m any cases, unfortunately, there are no criteria that reveal w hether a Turkic loan word belongs to the first or the second period. In the case o f the words borrowed from Turkic before the Conquest, chronological queries also arise: such as w hen did Turkic-H ungarian contacts begin, what periods can we divide this era into, where the adoption o f words m ight have occurred, w hether we can isolate the language o f the Turkic groups which attached them selves to the Hungarians or not, and if it can be successfully determ ined w hether a Turkic language was spoken in the C arpathian Basin at the tim e o f the Conquest. Linguistic criteria can be divided into three principal groups. T he first group contains criteria w hich unequivocally point to a Chuvash-type language. In the second category, we m ay include criteria which point only indirectly to Chuvash origin. There are characteristic features which occur in some words alongside unequivocal Chuvash criteria, thus indirectly indicating a Chuvash type, but which o f them selves are typical not only o f the C huvash language, but o f other Turkic languages as well. Finally, the third group includes those words which, while displaying none o f the above criteria, nevertheless possess chronological criteria which unequivocally show that they cannot be Turkic loan words adopted after the Conquest, or w hich in term s o f linguistic geography are so widespread am ong the Turkic languages as to m ake it probable that they w ere adopted from a Turkic language o f C huvash type. There are currently around 450 borrowed words belonging to the pre-C on quest or pre-O ttom an categories. These 450 words are o f course not o f equal value as far as the reliability o f their etym ology is concerned. We shall mention here only those words o f a certain or very probable Turkic origin, which also assist us in our historical reconstruction.
106
Methodological introduction and the sources
Unequivocal Chuvash criteria can be found in the follow ing types o f words: 1. W here in com m on Turkic there is a /z/, we find an /г/ in Hungarian: borju ‘c a lf’
The sources
107
In Hungarian, the phonetic cluster short vow el + / / / often changes to a long vowel, in such a way that the / 1/ is dropped; or vice versa, the long vowel becom ing shorter and a so-called inorganic /1/ being inserted. This even occurs in Turkic borrowed words, thus the Turkic agachchi becam e acs ‘carpenter’ in Hungarian, while there is also an older H ungarian form: ales. Likewise, the Turkic biigiichi becam e bdes in Hungarian, and later boles ‘sag e’. The diffi culty is caused by the fact that in Hungarian the /lch/ also appears where there was an /lch / in the original Turkic, thus the Hungarian kolcson ‘loan’ is derived from the Turkic form o f kilchen. If now and then an original /sh/ phonem e in Turkic becam e /ch/, then /lch/, and later /1/, as happened in Chuvash, then it cannot be said that the Hungarian adopted the /ch/, as in the word biicsii ‘farew ell’, or the /lch/, as in the word bolcso ‘cradle’; but (perhaps) did after all take the /ch/, as in the form *becsit, and added an /1/ in Hungarian. Such, for example, is the Hungarian word gyiim olcs ‘fruit’, which m ay have been borrowed from the Chuvash form jem ich or jem ilch. W hatever the explana tion, if the equivalent o f the com m on Turkic /sh/ phonem e is the Hungarian /ch/ or /lch/, this is a Chuvash criterion. It would be also conceivable that in this case we are dealing with an early change in Hungarian, i.e. that Hungarian adopted the Turkic word with an /sh/, which then becam e /ch/ in Hungarian. In this event, the Hungarian /ch/ and /lch/ would not be C huvash criteria. In numerous instances, the original Turkic /ch/ p h o n d n e appears in the H ungar ian as an /sh/: e.g.: the Hungarian kos [kosh] ‘ram ’ <— *kocs, Hungarian kes-ik [kesh-] ‘to be late’ <— *kecs-, Hungarian kis [kish] ‘sm all’ <— *kichi —» kicsi [kichi]. We may also hypothesise that we are dealing w ith an early Hungarian /sh/ ~ /ch/ dialectal change. Which is to say that the Hungarian language may have adopted these words at any stage o f the Turkic /sh/ > /ch/ > /lch/ > /1/ chain o f development. In other instances the word appears only in Chuvash, such as, for example, the Turkic original o f diszno ‘p ig ’; or its m eaning and form are present only in Chuvash, such as the source form o f the Hungarian w ord eke ‘plough’. The root verb ek- ‘to throw /sow ’ is also present in other Turkic languages, but the derivation eke ‘plough’ formed from this only in Chuvash. The original o f the Hungarian word tiikor ‘m irror’ can be found in several Turkic languages, always with the m eaning ‘round (object)’, but with the m eaning ‘m irro r’ only in the Chuvash language and Old Church Slavic. The latter adopted the word from the Danube Bulghar language. Unm istakable C huvash phonetic criteria can be found in the Hungarian word szolo ‘vine/grape’. This word exists in Chuvash, but means only ‘berry’ or ‘wild vine’. The Hungarian word arok ‘ditch’ contains no phonetic criteria whatsoever, but occurs in the text o f the
108
Methodological introduction and the sources
D eed o f Foundation o fT ih a n y Abbey from 1055, and is m ost likely a word which precedes the Hungarian Conquest. Based on the evidence o f all these linguistic sources, we are able to state that the Hungarian people lived together with a people or peoples who spoke a type o f the Chuvash Turkic language. W ithin this large num ber o f loan words, we are able to distinguish at least three Old C huvash (OCh) chrono logical phases. In order to dem onstrate this process, we m ust m ention the fact that in Late Ancient Hungarian (LAH) only a short vowel could have appeared at the end o f a word, and consequently when the Hungarian language adopted a word-final consonant, it w ould generally adapt the new word to the system by adding on a vowel. This short vowel was subsequently dropped in Old Hungarian (OH). Let us see what happened with the Proto-Turkic (PT) word-final /-k/: Adaptation o f Turkic guttural w ord endings to the H ungarian w ord struc ture. Third period, LAH III PT * kum uki > OCh *humuki —> LAH *humuki > OH hom ok ‘san d ’ PT * kuruk > OCh *hurtik —» LAH *huruku > OH hurok ‘loop’ PT * kosheke > OCh *kdleke —> LAH *kdldke > OH kolyok ‘k id ’, ‘young anim al’ Second period, LAH II PT *sivrik> OCh *shiirik> *shiirig —» LAH *shiirege > OH *soreg ‘a type o f fish ’ PT *ongmek > OCh *omek > *omeg —> LAH *iimege > OH *iimeg > imeg > ing ‘shirt’ Iranian je le k —> Old Turkic *jelek > OCh *szjelek > *szjeleg —» LAH *silege > OH siw eg > siiveg ‘cap’ First period, LAH I PT jtizu k > OCh *jiiruk > *jiiriig —> LAH *jiiriigu > OH gyuru ‘rin g ’ PT kajitk > OCh *hajuk > *hajug -» LAH *hajugu > OH hajo ‘sh ip ’ PT kum lak > OCh *kumlak > *kum lag —> LAH *kumlagu > OH komlo ‘h op s’ PT tisek > OCh *silek > *sileg —> LAH *silege > OH *sileu > siillo ‘zander (a fish)’
The sources
109
The following exam ple belongs to the second period, but cannot be derived from a Chuvash-type language, since we would expect an /sh/ phonem e at the beginning o f the word: PT singek > Old Turkic sinak > sinag —> LAH sinagu > sunagu > OH sunog, > szim yog ‘m osquito’ We can see that the /-к/ at the end o f a word was preserved in the m ost recently adopted words, while in the middle period it becam e a /g/, and in the oldest words disappeared through a diphthong. In the second period at least, Hungarians had already come into contact with speakers o f non-C huvash type Turkic languages as well. Thus it is that the Chuvash-type words adopted during the Late Ancient Hungarian period flowed into the Hungarian language over such a long time that, using the tools o f linguistics, we can distinguish three subperiods within this period. A strikingly large num ber o f the adopted words indicate that the Hungarian people m ust have lived for a com paratively long time with these Turkic speakers o f a Chuvash-type language. All this occurred sim ultaneously with a significant transform ation in the way o f life o f the Hungarians. The adoption o f a word signifying an object does not necessarily m ean that the object was previously unknown to speakers o f the recipient language. As we have already pointed out, the Hungarian words terd ‘knee’, koldok ‘navel’ and gyom or ‘stom ach’ are o f Turkic origin, which o f course does not mean that before their encounter with the Turks the Hungarian people had no knowledge o f their own bodies. Nevertheless, these words did not enter the Hungarian language as nam es for parts o f the human body but, concurrent with the increased and intensified practice o f anim al husbandry, were adopted as the nam es for animal parts, and were only later used m ore generally to specify parts o f the human body. The Hungarian word birka ‘sheep’ is o f M oravian origin, and was introduced into H ungary by C zech-M oravian shep herds together with a new form o f sheep farming, which o f course is not to say that the keeping o f sheep had not for a long time been a fam iliar enough activity am ong Hungarians. The adoption o f the word eke ‘plough’, m ean while, does not m ean that Hungarians had no previous know ledge o f agricul ture or the plough, but means rather that they learned to use a new, m uch more efficient type o f plough. However, before we get down to the business o f review ing the various sem antic groups o f borrow ed Turkic words, we m ust also note that conspicu ous gaps exist. For example, there are no technical term s connected to the names or processing o f metals among the Turkic loan words in the Hungarian language. This is no wonder, given that the Iron Age may have reached the Hungarians in the m iddle o f the 1st m illennium BC, and as we have seen,
110
Methodological introduction and the sources
contact betw een the Hungarians and Turks occurred later than this. Am ong H ungarian words o f Turkic origin many terms can be found belonging to the area o f anim al husbandry, but there are no term s related to the keeping o f horses (except perhaps csodor ‘stallion’). A t the same time, there are many nam es for parts o f the harness (arkany ‘pole lasso’, beklyo ‘hobble’, gyeplo ‘rein s’ and kantar ‘bridle’). Significantly, however, the word kengyel ‘stirrup’ is not o f Turkic origin. Given that in Eastern Europe the stirrup entered w idespread use after the 6 th century (see p. 103), the w ord kengyel m ay well have been a Hungarian innovation, ju st like the Germ an word for stirrup (Stegreif). This m ight reinforce the view that the Hungarians w ere already an equestrian people when they came into contact with the Turks. The nature o f the h o rse’s use, however, changed fundamentally. In any event, it is certain that the Hungarians were introduced to com pletely new forms o f animal husbandry and agriculture. This assum ption is backed up by the store o f Hungarian words o f Turkic origin connected to the practice o f anim al husbandry, such as alacs ‘spotted (o f an anim al)’ arkany ‘pole lasso’, artany ‘(castrated) hog’, barom ‘cattle’, belyeg ‘brandm ark’, bika ‘b u ll’, borju ‘ca lf’, diszno ‘p ig ’, gyapjii ‘w ool’, h o ‘butterm ilk’, karam ‘pen (for anim als)’, kecske ‘goat’, kom ondor ‘(one type of) sheepdog’, кори ‘churn’, kos ‘ram ’, kuvasz ‘(one type of) sheepdog’, okdr ‘o x ’, ol ‘sty ’, olio ‘kid (a young goat)’, serte ‘bristle’, tarka ‘piebald’, teve ‘cam el’, tino ‘young o x ’, toklyo ‘year-old lam b’, turd ‘cottage cheese’, tyiik ‘h en ’, tino ‘h eifer’, iirii ‘sheep’, tivecs ‘ew e’, and valyit ‘trough’, as well as words related to agricul ture, such as alma ‘apple’, arat ‘to harvest’, arok ‘ditch’, drpa ‘barley’, aszok ‘gantry’, bor ‘w ine’, borso ‘p ea’, buza ‘w heat’, csepii ‘thresher’, csiger ‘a bad w in e’, dara ‘groats’, did ‘w alnut’, eke ‘plough’, gyom ‘w eed’, gyiim dlcs ‘fruit’, kender ‘h em p’, kepe ‘stook’, kerodzik ‘to rum inate’, kert ‘garden’, kom lo ‘hop (the p lan t)’, кдкёпу ‘sloe’, kolyu ‘pounder’, korte ‘p e a r’, ocsii ‘tailings’, orso ‘reel’, orol ‘grind’, tarlo ‘fallow land’, tilo ‘sw ingle’, tonna ‘horseradish’, sarlo ‘sickle’, szor ‘to throw ’ or ‘to separate tailings from w heat’, som ‘dogberry’, szold ‘grape’, ‘v ine’, and szur ‘to filter (w ine)’. We have observed that dating further back than these borrow ed Turkic words connected to the keeping o f livestock are Hungarian words o f Ancient Iranian origin such as tehen ‘cow ’, tej ‘m ilk ’ and fe jn i ‘to m ilk’, while the basic term s o f equestrianism also date from the Ugrian period. Still, com pared to the circum stances o f these earlier adoptions, the transform ation in animal husbandry was o f a fundam ental nature. We can m easure its significance if w e com pare this store o f words with the Hungarian words borrow ed from the Slavic languages. W ithin the Hungarian language’s vocabulary relating to anim al husbandry a num ber o f words o f Slavic origin can be found, including abrak ‘fodder’,
The sources
111
akol ‘sheep-pen’, barany ‘lam b’, birka ‘sheep’, bivaly ‘b iso n ’, csorda ‘herd o f cattle’, gacser ‘drake’, iga ‘yoke \ j e r k e ‘ew e’, kacsa ‘duck’, kakas ‘cock erel’, kanca ‘m are’, malac ‘piglet’, mangalica (a type o f pig), m iskarolni ‘to geld’, dsztoke ‘p ro d ’,pasztor ‘herdsm an’,pata ‘h o o f’,patko ‘horseshoe’, suta ‘d o e’ and tarho ‘curds’. Although these words reflect a m ethod o f keeping anim als based around the stable, several levels o f this practice existed in Hungary. The situation is different as far as the Hungarian words o f Slavic origin, relating to agriculture, are concerned, such as asztag ‘stack (o f corn)’, bab ‘bean’, barack ‘peach/apricot’, barazda ‘furrow ’, borona ‘harrow ’, cekla ‘beetroot’, cirok ‘broom com ’, csep ‘thresher’, cseresznye ‘cherry’, csoroszlya ‘coulter’, dinnye ‘m elon’, gabona ‘grains’, garat ‘hopper’, hajdina ‘buck w heat’, kakat-szeg ‘a nail in the plough’, kalangya ‘haystack’, kalasz ‘ear (of w heat)’, kapdl ‘to h o e’, kasza ‘scythe’, kazal ‘haystack’, konkoly ‘corn cockle’, korpa ‘b ran’, lapat ‘spade’, len ‘flax’, molnar ‘m iller',p a jta ‘stable’, p a rla g ‘fallow land\p a s z u ly ‘bean’,pohanka ‘buckw heat’, repa ‘b eet’, retek ‘radish’, rozs ‘ry e’,szalm a ‘hay’, szam orodni (a kind o f vine), szecska ‘ch a ff’, szena ‘hay’, szuszek ‘w heat container’, tarack ‘stolon’, uborka ‘cucum ber’, ugar ‘fallow land’, villa ‘pitch-fork’ and zab ‘o at’. These reflect significant changes in agricultural technology. If we consider the Hungarian words o f Turkic origin relating to agriculture, it becom es abundantly clear that the basic vocabulary relating to the cultiva tion o f wine is o f Turkic origin. Such words include: szolo ‘vine/grape’, bor ‘w in e’, csiger ‘a bad w ine’, sepro ‘dregs’, aszok ‘gantry’ (for supporting barrels) and sziir ‘to filter (w ine)’. These words can only have been adopted in the vicinity o f the River Kuban or on the northern shores o f the Black Sea. There was no cultivation o f vines or fruit in the territories o f the M iddle Volga and K am a rivers at this time. As regards the words korte ‘p ea r’, alm a ‘apple’, did ‘n u t’ and sow ‘dogberry’, it can o f course be postulated that the Hungarians encountered these on com m ercial travels northward (just as the Hungarians did not com e to know the orange in the place o f its cultivation), but these words together nevertheless indicate a horticultural or fruit-growing culture, or at least a phytogeographical environm ent to which these plants w ere indigenous. This m ust again have been the Kuban river region and the northern environs o f the Black Sea. The data relating to viticulture and fruit are m utually supportive. In point o f fact, the Hungarian w ord kert ‘garden’ is probably of Indo-European origin, perhaps Iranian, although com ing through the m edia tion o f the Turks. O f the Hungarian w ords o f Turkic origin, the num ber o f term s belonging to the term inology o f the state, public life and religion is strikingly large. Such words include: a l ‘false’, bator ‘brave’, ber ‘paym ent’, betu ‘letter (charac-
112
Methodological introduction and the sources
ter)’, bilincs ‘handcuffs’, bird ‘ju d g e ’, bocsat ‘to pardon’, boszorkany ‘w itch ’, bun ‘sin ’, bubaj ‘charm ’, bo ‘wealthy, noble’, bojt ‘lent’, boles ‘w ise’, borton ‘prison’, bucsu ‘farew ell’, csata ‘b attle’, csosz ‘field-guard’, egyhaz ‘church’, erdem ‘m erit’, erkolcs ‘m orals’, eskti ‘oath’, gyalaz ‘to slander’, gyanu ‘sus picion’, gyon ‘to confess’, gyiilol ‘to h ate’, im dd ‘to adore’, ir ‘to w rite’, kin ‘p ain ’, koldul ‘to b eg ’, koporso ‘coffin’, kolcson ‘loan’, orvos ‘d o cto r’, sereg ‘arm y ’, sir ‘tom b’, tanacs ‘counsel’, tanu ‘w itness’, telek ‘p lo t’, terem ‘h all’, tolmacs ‘interpreter’, tor ‘feast’, tomeny ‘abundant’ (originally ‘ten thousand’, a m ilitary unit), torveny ‘law ’, ur ‘lord’, iidiil ‘to rest’, unnep ‘festival’. For such a large quantity o f this type o f term to have entered the Hungarian language, the Hungarian people would have to have been full participants in some kind o f Turkic political system , tribal confederation, em pire or state. In this regard, the Sabirs, the Turks, the O noghur-Bulghars and the K hazars may, in principle, each come into consideration, being Turkic-language ethnic form ations o f a kind which headed a political structure for shorter or longer periods on the territory here considered feasible based on other considerations o f linguistic geography. The Turks m ay be ruled out on phonetic grounds, as we are well fam iliar with their language from Turkic inscriptions. Their significance was anyw ay only transitional in Eastern Europe. We are unable to take the Sabirs into consideration as we know nothing o f their language. The few “Sabir” personal nam es which have survived in B yzantine sources are insufficient for the identification o f it. Anyway, we m ay set aside the question o f the Sabirs since, although they played an im portant role in Eastern Europe in the first h alf o f the 6 th century, their significance sw iftly declined in 558, follow ing their defeat at the hands o f the Avars, and by the seventh decade o f the 6 th century they had disappeared from the scope o f w ritten sources. They could thus have scarcely played a significant role in the form ation o f the vocabulary o f the Hungarian language. In effect, only the Bulghar-Turkic and K hazar Em pires are w orthy o f consideration. N aturally we cannot allow ourselves to forget that these nom adic em pires em braced m any languages, and in each Turkic empire there dwelt both foreigners and speakers o f a variety o f Turkic languages. Nevertheless, w hen foreigners such as the Hungarians, for example, succeeded in mastering one o f the languages o f an empire, then this would most assuredly be the language o f the ruling group. The O noghur-Bulghars w ere in fact participants in Eastern European his tory from the second h alf o f the 5th century. However, no direct evidence o f their language has survived. Two branches o f this people played a role follow ing the dissolution o f the Bulghar Em pire o f Khuvrat. One group were the people o f the Danube Bulghar Empire. We have a handful o f sources relating to this language. These are:
The sources
113
1. Inscriptions in Greek script and in Greek language from the territory o f the Danube Bulghar Empire. These inscriptions contain a few titles and names. There are close to one hundred inscriptions, or fragm ents o f inscriptions, surviving, and a few dozen com m on words can be found am ong them. However, a portion o f these are official titles. 2. Fragm ents o f Danube Bulghar texts in Greek script. These contain the nam es o f weapons. There are scarcely any Turkic words am ong them. 3. The so-called D anube Bulghar regal list. This enum erates the Danube Bulghar princes, giving the dates o f their rule after each name. The problem with this list, however, is that the 1Oth-century original has survived in copies from the end o f the 15th century and the beginning o f the 16th century in Cyrillic script. The copiers so distorted the words, which were unknow n to them, as to m ake them unsuitable for a scholarly reconstruction o f the text (see pp. 61-62). 4. Runiform fragments. These are practically undeciphered, and because o f their brevity there is little prospect o f them being deciphered in the near future. As far as the history o f writing in Eastern Europe is concerned, however, they are very interesting. 5. Loan words which entered Old Church Slavic or Bolghar Slavic from the Danube Bulghar language. W ithout any doubt, the language o f the Bulghars was Turkic, nevertheless the question o f which Turkic language they spoke cannot be answ ered for the time being. The word for an official title appears am ong the inscriptions in the form o f kolovros. If this is taken to be the Grecised form o f kolovur, and can be identified with the com m on Turkic title o f kulavuz (which is actually probably o f Iranian origin), then we have an exam ple o f the Danube Bulghar /г/ equivalent o f the com m on Turkic lz/. Still, this one piece o f data— even if it proves correct— is not enough to construct a theory around. The earliest evidence o f the Volga Bulghar language, aside from a few personal and place names, are the names o f rivers and other w ords recorded by Ibn Fadlan, followed by the so-called Volga Bulghar epitaphs, which we shall discuss shortly in relation to the nam es given to the Hungarians, and thirdly, the loan words adopted by the neighbouring Finno-U grian and Turkic languages. In the A rabic-language grave epitaphs, both the date and certain parts o f the name, and som etimes one or two other phrases, are in Turkic. We know o f two types o f inscription, one o f which uses a Turkic language o f the Chuvash type. The earliest dated relic is from 1281, while the Chuvash-type inscriptions disappear from around 1360. The Chuvash elem ents o f the Volga Bulghar grave epitaphs reflect not the direct antecedents o f today’s C huvash language, but rather a separate, now
114
Methodological introduction and the sources
extinct, Chuvash dialect. The two dialects, the Volga Bulghar and the prede cessor o f today’s Chuvash, w ere nevertheless very closely related. For this reason, the Chuvash o f today can be used in reconstructions, albeit with a certain am ount o f reservation and precaution. The view is held by some that the Chuvash-type inscriptions are not those o f the Volga Bulghars. The Volga Bulghars left behind the other type o f inscriptions in the com m on Turkic language. This suggests that the Volga Bulghar r-Turkic inscriptions perhaps reflect the language o f surviving m em bers o f the Khazar people who fled northward, and would m ean that the Chuvash are also the descendants o f the Khazars. Unfortunately, reconstruction o f the Khazar language is also m aking only slow progress, since we have alm ost no data at our disposal. The recognised titles o f officials and personal names are scarcely adequate for the determ ina tion o f linguistic affiliation. The onom astic m aterial consists o f the nam es o f four tribes, and some 40 personal nam es and titles (the latter two not always distinguishable). These throw a certain am ount o f light on the state o f the K hazar language, but not one is suitable for more precisely determ ining the language o f the Khazars. The situation is som ewhat m ore prom ising with regard to the place nam es and geographical names. Am ong these, there are one or two o f an unm istakably Chuvash hue. Such, for exam ple, are the nam es for the Fortress at Sharkel, built in the year 838 beside the R iver Don with B yzantine help, and the Khazar name o f a river, the H ara sin. For a long time researchers believed that the nam e o f the Khazars as a people ruled out the theory that they spoke a language o f the Chuvash type, since the w ord K hazar was coined by the Khazars them selves, and contains a l-z-l. However, recently discovered Old Turkic runiform inscriptions and other data have unequivocally proven that the Turkic nam e for the K hazars was kasar. This m eans that Khazar cannot be ruled out as belonging to the C huvash group o f languages, altering the previous assertion that m eant only the BulgharTurks were taken into consideration. In the final analysis, then, both sources are feasible. Using the m ethods o f linguistics we are not even able to exclude the possibility that loan words which entered the Hungarian language prior to the C onquest period may include those o f both Bulghar-Turkic and K hazar origin. Yet m ore complex is the question o f w hether Turkic-speaking peoples lived in the Carpathian Basin at the time o f the Hungarian Conquest. It appears that research today is able to produce fresh results in this area as well. M uch has been said o f the language o f the Avars and the peoples who lived under Avar rule, but the surviving m aterial again com prises m ostly personal and official nam es, w hich are not really suitable for determ ining the nature o f a language. W hile it is certain that Slavic peoples lived in the C arpathian Basin at the time
The sources
115
o f the Conquest— som ething we shall discuss later— there are a num ber of phenom ena which cannot be explained m erely by a Slavic presence. Am ong the official ranks at the time o f Arpad, an im portant role was played by the ispan (a type o f local steward or governor). The older form o f this word, span [shpan], can be easily traced, and in Hungarian evolved into ispan in m uch the same way as the Latin scola becam e iskola [ishkola] ‘school’, and Stephan becam e Istvan [ishtvan]. However, this span form cannot be attrib uted to the Slavic. It is clearly identical w ith the official title o f zhupan, which can be shown to have been an im portant rank in the C arpathian Basin from 111 onward. The dropping o f the vowel -u- cannot, however, be explained by the Slavic, and Hungarian could not have developed ispan from the form o f zhupan. The key to the m ystery can clearly only be held by a language in which the stress in the word was placed on the final syllable, and in which the vowel -u- was reduced in an unstressed position, and finally disappeared. Zhupan cannot be a Turkic word, because in Turkic a word could not begin w ith /zh/. N either could the Slavic zhupan form be explained by a Turkic choban form. If, however, a Turkic people had borrowed the form zhupan from a Slavic language, then it may have quite properly become shupan, and later perhaps shpan. Several unsolved problem s also revolve around the Hungarian official title nadorispan ‘palatine’. Am ong the early forms o f the word can be found nadorispan, nandorispan, and landorispan. If, from am ong these, the nandor prefix is taken as the original, then this m ight be the Hungarian nam e for the Danube Bulghar-Turks or the Onoghundurs: just as the city o f B elgrade was known in Early Hungarian as Nandorfejervar (literally, ‘W hite City o f the N andor’, i.e. the Bulghars). It is a rare, but not unique, occurrence for a title to be derived from the name o f a people. The eastern Slavic Boyars is the name for the ancient feudal noble class o f the boyarin (as opposed to the court nobility, the dvoryanin), the old form being bolyarin. This is the ethnic nam e Bolghar together with the Slavic affix [-in]. The [-lg-] > [-ly-] change occurred in the region o f the Volga as well, where the old place nam e Bulghar becam e Bulyar, then B ilyar or Bular (according to the Hungarian Anonymus: Bular). The title o f nandorispan was therefore originally the ‘O noghundur (Bulghar) zhupan’, the ch ief o f the Turkic population, just as the szekelyispan was the head o f the Szekely people. Just as the Latin name Caesar gave us the titles K aiser, Czar and the Hungarian csaszar (while the original Turkic form o f the K hazar nam e was Kasar), thus the nam e o f khaghan Bayan, the founder o f the Avar Empire, gave the title o f ban ‘B an’. The name o f khaghan Bayan is Turkic, not M ongolian. Although in the Turkic languages, the w ord m eaning ‘rich ’ always takes the form bay, in C huvash it is puyan, which can be traced back to an
116
M ethodological introduction and the sources
earlier bayan form. This m ay suggest that at least the leading stratum o f the Avars spoke a Chuvash-type Turkic language. We shall return to this when discussing the Avars in more detail. In Turkic, however, the [-y-] is not dropped from between the two vowels. It m ay in fact have been dropped only in certain Slavic languages, for exam ple in Croatian, while the Slavic o f Pannonia m ay also have followed this change. We are therefore dealing here with the Slavic form o f a Turkic word. The Conversio bagoariorum et carantanorum com piled in 870, w ith regard to a Church consecrated in 8 6 6 in the region o f Lake Balaton, m entions the place nam e o f Termperhc. This sam e place nam e features in a docum ent drawn up in 860. The second elem ent o f the place name is the G erm an w ord -berg ‘h ill’. The nam e also appears in Baranya county in a docum ent from 1332 in the form o f Teremhegy. The w ord terem, here m eaning ‘p alace’, is m ost likely o f G reek origin, but was also current am ong the Turks and the Slavs. In this event, it entered the Slavic through the m ediation o f the Turkic. The title o f palatinus ‘palatine’, o rp a lo ta g ro f was termecsii, w hich becam e the personal name Termecsii, borne by the grandson o f A rpad who visited C onstantinople in around 950. The title o f termecsii is only indirectly related to the K hazar title o f termecs ~ tarmacs recorded from 730. The Hungarian w ord belyeg ‘stam p’ m eant a m ark branded onto an anim al with a red-hot iron. In Turkic we m ust begin from the form belek. The Slavic word o f identical m eaning, beleg, is also derived from the Turkic. By linguistic m eans we are able to dem onstrate that the w ord entered Hungarian either from the Turkic, or from the language o f a Turkic people w ho were becom ing Slavicised. This Slavic-oriented Avar-Turkic language also lent words to Hungarian, although not in very great numbers.
5. THE ARCHAEOLOGICAL SOURCES W hile the discovery o f fresh linguistic sources o f data is a very rare occur rence, advancem ent o f research in this area being confined m ainly to the refinem ent o f methodology, the situation is entirely different in the case o f archaeology. New m aterial is constantly em erging from the bow els o f the earth, while other relics gathering dust in the depositories o f m useum s becom e ever more readily accessible. The difficulty with archaeological m aterial lies not m erely in the problem s o f determ ining age, as we have seen in the above, but also in attaching each archaeological culture, group o f finds or objects to a given people. Researchers do not always succeed in finding the sober m iddle road betw een too rapid conclusions and total scepticism .
The sources
117
The treasure o f archaeological findings relating to the conquering M agyars is w ell known to science, and even if certain m atters o f detail rem ain unre solved, this is chiefly the result o f advancem ents in m ethodology, which although producing more results, also generate more problem s. For a long time Hungarian archaeologists were sustained by the hope that the com paratively easily identifiable store o f archaeological finds from the Conquest period could be traced back in time, that the place from which the M agyar tribal confederation launched its Conquest o f the future hom eland would be easy to locate. A fter all, we are fam iliar with relics from even the earliest generations following the M agyar Conquest, and by no m eans can it be assum ed that the possessions or burial custom s o f the first-generation conquering M agyars buried in the C arpathian Basin w ould have been signifi cantly different from those o f their fathers’ or grandfathers’, laid to rest in the pre-Conquest Etelkoz region. Before the First World War, a num ber o f attempts w ere m ade to unearth this m aterial, but not only was it a case o f trying to find a needle in a haystack, but moreover, the ‘haystack’ at this time, nam ely the archaeological findings from southern Russia, were not system atically categorised, and thus it was im possible to know the local context o f each potentially interesting find. Burial sites from the pre-C onquest period were not found (or if rum our o f such a site began to circulate, it w ould later prove to be false), and for a long time it was believed that the reason for this was the closing down o f frontiers w ith the em erging Soviet Union in the 1920s, which prevented Hungarian archaeologists from carrying out their field work. The com plaint was ever m ore frequently heard, albeit not through official chan nels, that Soviet archaeologists, partly with regard to their professional stand ards, and partly because o f their particular prejudice against nom adic peoples, scarcely carried out any excavations in these regions, and if they did actually find anything, would fail to recognise or be unw illing to identify it as H un garian. For this reason it was im possible to identify M agyar m aterial from before the Conquest period. Later, in the 1950s and 1960s, H ungarian archae ologists were given the opportunity to em bark on research expeditions, and from the 1970s jo in t excavations took place. An incredibly rich treasure o f archaeological m aterial began to emerge from southern Russia, the Volga and northern Caucasus regions, and while the publication o f finds was not always o f a desirable standard, after a time both the sheer abundance o f m aterial and the sporadic nature o f publications began to hinder research into M agyar parallels. From the end o f the 1970s, however, the surprising conclusion began to emerge that despite all endeavours, and even with possession o f a great store o f com parative m aterial, it was still im possible to trace back, either in time or space, the archaeological relics o f the pre-C onquest M agyars.
118
/\A ethodobpctd introduction and the sources
C onquest-period sites beyond the Carpathians: (1) Przem ysl; (2) Sudova Vishnya; (3) Krilos; (4) Frumu§ica; (5) Probota; ( 6 ) Groze§ti; (7) Bucharest-Tei; ( 8 ) Subbotici
This state o f affairs then becam e m ore precisely defined, and in three directions. In the eastern lower reaches o f the Carpathians, burial sites did nevertheless begin to em erge which yielded archaeological m aterial displaying wellfounded sim ilarities w ith relics o f the conquering M agyars. This strip o f territory em bracing the Carpathians from the east, from Przem ysl in the north to Bucharest in the south, has thus far provided us w ith six archaeological sites (Przem ysl, Sudova, Krilos, Probota, Groze?ti and Bucharest-Tei). There can scarcely be any doubt that we are dealing here w ith lookout garrisons posted from the Carpathian Basin, i.e. after the C onquest o f the hom eland, and that here w e find evidence o f the inhabitants o f sentry posts set up to w atch over the outerm ost reaches o f the new ly founded settlem ent, not the graves o f those who w ere m aking their way toward the Carpathian Basin. It is som ew hat more difficult to determ ine the nature o f two other sites located further to the east,
The sources
119
one in M oldavia (Frumu§ica), the other beside the River Ingul (Subbotici). The chronology and other circum stances o f the finds are yet to be adequately ascertained, and these m ay be the remains o f groups which w ere either left behind, or chose to settle, or perhaps even returned from w anderings else where. In all events, these sites do not help solve the problem o f finding a location for the M agyar people prior to the Conquest o f their hom eland. The second direction in which renewed waves o f em erging data point is concerned with the individual types o f archaeological relics o f the conquering M agyars, the various ornam entation and techniques applied, one or other o f which can be found on each archaeological object found outside o f the Carpathian Basin. Here we are looking not for the antecedents o f individual stylistic elem ents, but for instances o f exact likeness. Recently, for the occa sion o f an exhibition in Hungary, the celebrated Sword o f Kiev, often used in com parison with H ungarian archaeological finds, underw ent a successful restoration in Budapest. Thus, much has been determ ined o f this, by now scrutinised, m asterpiece that “its blade m ay have been fashioned som ewhere in the region o f the Rhine, and its hilt end and cross-piece ornam ented by Viking m aster craftsm en, while its plate o f silver— decorated in very close conform ity to Hungarian taste— was in all probability attached to the hilt in Kiev” . In any event, it is a good example o f ju st how com plex a historical background can be revealed when an object is subjected to thorough exam i nation, and suggests that we can take for granted a sim ilar result even where such a thorough exam ination is not possible. In the case o f the Hungarianstyled silver plate on the Sword o f Kiev (see Plates 1/1 and 1/2), we still do not know where the craftsm an lived, for whom he was working, and to which workshop he belonged. Caches o f coins have already proved that as early as the decade following the Conquest, m erchants arrived from the region o f the Caspian Lake, and that such contacts were clearly reciprocal. Each object, as a piece o f m erchandise, m ay have travelled quite considerably, even as far as Sweden, while either voluntarily or o f necessity, craftsm en were also capable o f journeying very great distances. To what degree they would have then retained the M agyar style in their new location obviously depended partly on their new masters. For this reason, from the Caucasus to Scandinavia, certain individual objects displaying unm istakable Hungarian parallels, or the orna m entation to be found on such objects, can be regarded exclusively as pro viding clues to the history o f the given object itself, and not as sources o f information regarding a people’s history— although they m ay serve well to indicate the trading routes, connections and econom ic contacts o f the age. While we are unable to take a firm stand regarding the ethnic provenance o f these craftsm en, it is conceivable that am ong them there w ere not only M agyars, but also wandering craftsm en belonging to other peoples. This
120
Methodological introduction and the sources
question m ay be further com plicated by the possibility that associates o f a craftsm an arriving in Hungary from a foreign workshop m ight have continued to m anufacture the same type o f object in their original w orkplace which their colleague who had em igrated to Hungary had introduced, and perhaps brought into general use in the Carpathian Basin. Under these circum stances, we can again obtain im portant inform ation regarding economic channels and trading contacts, but nothing about the m ovem ent o f peoples. The third group o f evidence is the m uch-debated body o f archaeological m aterial which has come to light at Bolshie Tigani, near to the confluence o f the Volga and Kama rivers. The finds uncovered here bear a very close resem blance indeed to the Hungarian finds from the Carpathian Basin, and experts are in little doubt that the Bolshie Tigani burial site contains the rem ains o f Magyars. The sim ilarities are unquestionably striking, particularly as regards the partial burial o f horses. Unfortunately, however, the determ ination o f exact or partial correspondences has not been followed by a thorough analysis o f the dissim ilarities. There are no metal sabretache plates, for exam ple, while the three-leaved belt-end decoration is far rem oved from palm -leaf ornam en tation, and so on. There has scarcely been any debate with regard to the dating o f the cemetery. The m ost recent coin found in the graves was m inted in 900 A D , and evidently ended up in the earth a few years or decades after. Debate is centred around how we may reconstruct the history o f this sm all ethnic group which is regarded as Magyar. There is agreem ent that these people may be identical with those whose descendants were m et by Friar Julianus not far from the capital o f the Volga Bulghars in the year 1235, and whose name appears in an inscription from 1311 which we shall discuss later (see pp. 301-302). From this point on, however, there are two divergent opinions which have two diam etrically opposed sets o f precedents. There is some argum ent as to when this group o f M agyars arrived in Bolshie Tigani and the environs o f the River Kama. The first opinion states that it was from here that the M agyar people set out on their final peregrination ending in the occupation o f the ultim ate hom eland, while the group based around B olshie Tigani rem ained in the form er hom eland. The second opinion, m eanw hile, has it that this M agyar people drew back northward from the south, having gone their separate ways further south, with a section o f the M agyar people proceeding westw ard, and the other section, joining up with the Volga Bulghars as they withdrew northward, m igrating into the region o f the Volga and K am a confluence. It is clear that w hichever point o f view we choose to stand by, as far as the early history o f the M agyars and o f Eastern Europe in general is concerned, we have hit upon a question o f vital importance. Here it is only archaeological argu m ents that we are holding up for com parison. The argum ent w hich takes the
The sources
121
view that the M agyars in the K ama region “rem ained in place”, in other words in the one-tim e Urheimat, thus reasons that the Bolshie Tigani culture was not isolated. Strikingly sim ilar burial practices can be inferred from earlier archae ological m aterial unearthed in the vicinity, nam ely that belonging to the K ushnarenkovo culture. This theory dates the beginning o f the Kushnarenkovo culture to the 6 th century, thus allowing the local antecedents o f the Bolshie Tigani culture to be traced backward. The burial sites at Kushnarenkovo, 50 km to the northw est o f Ufa, and Sterlitamak, some 130 km to the south o f Ufa, together with the archaeological culture which has come to be connected with them, have long occupied the m inds o f experts in the field. The type o f ceram ic objects unearthed at these sites can also be found on the far side o f the Urals, and for this reason some have traced the origin o f this culture to w estern Siberia. However, the uni formity and chronology o f this culture are still the subject o f dispute, although it can at least be said with certainty that it still existed in the 9th century. This is important because the territory it occupied is by and large identical to the territory where Ibn Fadlan encountered what he named the Bashkir people in the year 922. Bashkirian researchers wasted no time in identifying the Kushnarenkovo-Sterlitamak culture as Old Bashkirian. The remains o f the people who populate the 150 or so graves o f the Bolshie Tigani cemetery m ay thus actually reveal a connection with this territory. While the archaeological relics o f Bolshie Tigani reveal a link with the Kushnarenkovo finds, the testimony provided by the Bolshie Tarhani and Tankeyevka burial sites cannot be disregarded; these latter displaying an unequivocal connection to the Volga Bulghars. Bolshie Tarhani is located at the Kuybishev bend, and is the earliest known burial site o f the northbound Volga Bulghars, w hile the cem etery at Tankeyev ka is one o f those very few Volga Bulghar burial grounds lying on the right, or northern side o f the River Kama, and belonged to the garrison posted to the far side o f the Kama, the river which roughly constituted the Volga Bulghars’ line o f defence. Accurate dating o f the Bolshie Tarhani burial site is as sisted by caches o f coins, the m ost recent o f which m ay date from 787-789, and which can hardly have ended up in the ground m uch earlier than 800. M eanwhile, two dirhem coins found in the Tankeyevka cem etery can be respectively dated to 846 and 892-902. This m eans that the Volga Bulghars m ay have crossed the line o f the Kama in around the year 900. The Bolshie Tigani cem etery can be dated to betw een 850 and 920. The question is: where and when did those buried here come into contact with the Volga Bulghars? Given that the territory o f Bolshie Tigani was reached by the Volga Bulghars only after 850, then either contacts began at around this time, or the two groups arrived simultaneously. The first alternative can be ruled out on archaeological grounds, lin g u is tic arguments we will deal with sepa-
122
Methodological introduction and the sources
rately later. The second alternative is, however, feasible. Throughout the territory bordered by the Volga, Kama, Cherem shan and M ayna, there are m ore than one thousand archaeological sites that can be dated to betw een 800 and 1200 AD , and these are all Volga Bulghar sites. This region is the m ost thoroughly explored territory in Russia in archaeological terms. W hile it is no sm all feat that archaeologists have succeeded in finding traces o f the M agyars in this ocean o f evidence, nowhere can there be found any archaeological relic which w ould prove to be not only connected to the finds o f Bolshie Tigani— as is the m aterial unearthed in Kushnarenkovo, B olshie Tarhani or Tankeyevka (entirely understandably if such links came about through m arriage)— but which m ight also be revealed as an antecedent o f Bolshie Tigani. The highly specialised horse burial revealed in the grave at Sterlitam ak is o f course worthy o f attention, but only because it m ay shed light on the origins o f this charac teristic form o f burial. The m atter o f funereal shrouds is also not w ithout interest. A great num ber o f such funereal shrouds or face coverings have been found in C onquest-pe riod M agyar cemeteries, where the position o f the eyes and the m outh is indi cated by pieces o f m etal sewn onto canvas, or som etim es leather (see Plate 11/2). In exceptional cases, as in R akam az-Strazsadom b (NE Hungary), the face m ask is m ade o f metal, with openings indicating the position o f the eyes and the m outh (see Plate II/l). Two types are known in the province o f the Rivers Volga and Kama. One is a gilded silver or plain silver cerem ent covering the entire face, and leaving openings for the eyes and m outh (such as that found in Bolshie Tigani), while another m erely places plates o f a precious metal sewn onto leather or canvas over the eyes and mouth. The variant leaving openings and the type where the eyes are covered reveal two differing spiritual beliefs. One “protects” the deceased against evil spirits w hich m ight break through openings, while the other “eternalises” the deceased, setting the face in perpetuity. The theory that these represent ancient Finno-U grian or Ugrian burial rites, which the M agyars brought with them from the region o f the Kama, scarcely w ithstands close scrutiny. Such funereal shrouds and masks are far m ore widespread, and cannot be attached to any particular ethnic group. Even if plates such as those designed to be placed over the eyes and the m outh were taken separately in an attem pt to prove such an ancient connection, the problem w ould still rem ain that this type appears in the territory o f the Urals only in the 8 th century, in the later Lom ovato culture. It is quite im possible, therefore, that we are dealing here w ith some kind o f “ancient” Finno-U grian phenom enon, as was claim ed by some scholars. A rchaeological arguments, therefore, do not support the theory that the M agyars em erged from this region. We shall, however, return to the question in connection with M agyar groups left behind in the east (see C hapter XV).
The sources
123
The burial site o f the M agyars o f Bolshie Tigani does nevertheless provide us with one other piece o f testimony. If taken to be the resting place o f M agyars who m igrated northw ard from the south or southw est together with Volga Bulghars in the 8 th century, then this w ould suggest that the burial custom s o f the M agyar people had already become fairly uniform by the tim e the M agyar groups separated. A parallel phenom enon was the preservation o f self-designation in both groups (see pp. 297-308). If, however, this can be established, as it m ost certainly can, then once again we find ourselves facing the problem o f why no M agyar archaeological evidence can be found in the one-tim e Etelkoz and Levedia regions, or indeed anyw here the M agyar people dwelt in earlier times. At the same time, the theory that the archaeological culture o f the M agyars took shape in the Carpathian Basin itself also falls by the wayside. In order to exam ine the archaeological culture o f the M agyar people, we m ust briefly survey the archaeological antecedents in the Carpathian Basin. Although archaeological evidence, and even inscriptions, o f Celts, Romans, Huns, Gepidae, Longobardi and other peoples can be found in great abundance in the C arpathian Basin, from the point o f view o f the M agyar people only the Avars who im m ediately proceeded them in the Carpathian Basin have any direct significance. Rather than the archaeological relics o f the Avars them selves, it is more appropriate to speak here o f the archaeological relics o f the Avar period, since in this way we use term inology w hich does not attach such relics to a specific people. As a result o f the m ost recent research we can divide the period in question into four subperiods: (1) Early Avar period; (2) transitional period; (3) Late Avar period; (4) closing phase o f the Avar period. In the year 567, the king o f the Longobardi form ed an alliance with the Avars as they arrived from the east, inflicting an annihilating defeat on the Gepidae who then lived in the eastern half o f the Carpathian Basin. Sub sequently recognising their strategic position, the Longobardi w ithdrew to northern Italy after Easter in 568 to the region o f Spoleto, Benvenuto and to Lom bardy which bears their name to this day. The Avars follow ed in their footsteps into the Carpathian Basin, and with them arrived a new archaeologi cal culture. This lasted until the 640s, when the Bulghars o f K huvrat from the east, and the Slavic tribes o f D alm atia from the southwest, joining forces with the subject peoples o f the em pire, dealt a decisive defeat on the Avar Empire. It was at this tim e that the characteristic archaeological culture o f the early Avar period came to a close, and the relics o f several subsequent decades reveal transitional forms. A round 670, the Khazars overthrew the O noghur-Bulghar em pire o f Khu vrat, after which one group oftthe Onoghurs m igrated to the Balkans, while
124
Methodological introduction and the sources
another departed for the territory o f the Avar Empire. The O noghur chieftain was elevated to a high rank in the Avar Em pire, and archaeological rem ains o f his entourage can be found separately along the banks o f the R iver Sio and in the M ezoseg region. A t the end o f this m ajor internal transform ation, i.e. around the year 700, pow er was once again consolidated in the Avar Em pire, leading to the flow ering o f an archaeological culture w hich we shall call the Late Avar p e riod. C harles the Great (Charlem agne) and the Frankish Em pire launched
JCANNAB1ACA /
/
•v#
/
(£
ds
AUV N O VA *
X W
Риги od* , Ж
^
О*5
/ A D ^ E X U M i^ H 'A R B A M I A # ^
-
4 № г% гл г л с г о а VXRRABONA URIGETIO ®JJL.CIbIA CASTRA •• ' . „ . m c r i , "•* ' \% % ® A Q U 1N C U M >Ч.:сКЬШ :А®^ -• A - y f Albertfalva
4A V A R IA ^ .
V ,> ,,
•
.-
-M C JttN T lA N A '' ' ' ' ' ' '
/ vn
.S iih i
SAL!
{
•
Loeica
VS "
CF1 F b ? l;' r^ ^ 4 ,
*. * * a x d a v c
tMONA ATRANS ,
®
4- -
4
| PRAHTO RfUM * ^ LATOBICORUM
/
*
I
10VIA
\1
j
QRPI AszSr
i
Х .ТГС(,~1СЛ
>
/”
giUSSO NiUM D
* 'A e|
s TATUAS
i
'
’*
'
2 ~
-
FLORENT1AM °
'"S., A L H N U M * l ^
.
NOVAS
®
M U R S E L L A * >m J E jI'O B U R G IU M .
И; ■ ~ ® .
fe-.TUS SAUNA "pNThRClbA
A U S < £ *^L T A R1PA
S O ^ lA N ^ l ® L l
"^VNpA UTONIA
.)
;
S, ,!)•?
1_ Motel ice
\
-
«•
H rte n y jw ta *
AQ UAE ! A | a 6 ” ' %
Я.ГЛ'Пч! VKJDUNUM
^ Л М Р О Ш
I^ IA N A t NKAMAT,A « С
•' \-0 ii>kpmzt
м
^ Щ
Le gionary camp
ra m „
I2
Cotonia, municipium Ottifir settlement
СВАЬЧЕ® д а ™ , * ■ (3QNONiA ^ U D A L I A * : BASS1/\NA ^ 4 SiRfyUOMis *' 4 Ш Й В Д » ,
Roman names of the settlements Present nam es of the settlements
5 FOSSAE ) 6 Hrtkcwci j
Figure 17 Rom an Pannonia
■wg(}§[
2с О)
0 -£ \о 1 Л
Е *5 л & Л и О) л л о Л -о О)
00
126
Methodological introduction and the sources
attacks against the Avar Empire in around 791. Internal strife, the conversion to C hristianity o f a section o f the Avar leadership and fresh C arolingian attacks ultim ately brought the end to the em pire (as described in greater detail on pp. 263-266). However, in the last years o f the 8 th century— contrary to m any opinions until now— life did not com e to an end, and the land o f Pannonia was not reduced to devastation. On the one hand, a Slavic centre was created in Pannonia with Zalavar (M osaburg) as its seat, while on the other the Frankish influence began to increase, an influence subsequently reflected in the archae ological material. Besides this, a group o f archaeological finds can be distin guished which is contem porary with the ceram ic finds o f Carinthia, or with the culture know n to some scholars as the N agyszentm iklos-B IatnicaM ikulcice parallel, which was, m ore correctly, proposed the designation o f “closing phase o f the Avar period” . Evidence o f this culture can be found in Transdanubia, and in all likelihood it was also present in the region betw een the Danube and Tisza rivers. Later, around 830, the Danube Bulghars, who had in no small part contrib uted to the downfall o f the Avar Empire, seized possession o f the Maros (Mure§) valley, m ainly in order to secure them selves the salt-m ines, and in all probability gold-m ines in the vicinity. Here, a characteristic body o f Danube Bulghar archaeological m aterial can be found. A m ong the clearly distinguishable relics o f the Early Avar period, we find a num ber o f press-forged belt hoops (see Plate IV). It was from this period that we find long and loop-sided stirrups, straight-bladed swords with P-shaped handles, scaled arm our fashioned from w rought iron plates linked together, and quivers with edges trim m ed with carved battens o f bone. In the female graves, m eanw hile, were discovered plain bone needle-cases and spherical or prism -shaped hanging earrings. In the middle, transitional period, a num ber o f new objects appeared, including cellular, circular brooches inlaid with glass, and cylindrical, latheturned needle-cases, while am ong w eapons, the sword gave w ay to a straighthilted, single-edged sabre. In the late Avar period, we find that types o f objects from the transitional period continued to be made, the lathe-decorated needle-cases am ong them, and runiform inscriptions also begin to appear on them. The m ost striking finds, however, are the belt m ounts cast o f bronze, and often gilded or plated with silver (see Plate V). On the ends o f larger belts can be seen scenes o f anim al struggles, griffins, hunting scenes, the Tree o f Life, and m an fighting lion, while on other belt-ends, or som etim es on the reverse side o f the larger belts, we find creepers, vine leaf clusters and again the Tree o f Life. Belt hoops are rectangular and decorated with creepers, plant elem ents being the preva-
The sources
127
lent feature. For this reason these types o f finds came to be know n in b rief as the griffm -creeper archaeological culture. A new type o f sabre also appeared, as did w om en’s earrings augm ented with pendants o f fake pearl or glass, and bracelets decorated with ribbons and precious stones. At the cem etery o f Szarvas (Bekes county, eastern Hungary) belonging to this archaeological culture, dating to the second h alf o f the 8 th century, was found a prism atic needle-case o f sheep bone, with a line o f inscription on each o f its four sides. The whole inscription consists o f some 59 characters scratched into the surface (or at least this m any can be distinguished for certain, while in worn areas o f the surface there were, or m ay once have been one or two additional letters). The alphabet o f the inscription is in essence identical
F igure 19 T he B u yla in scrip tion on the N a g y sze n tm ik lo s plate
Merliodoloyicul introduction and the sources
128
4
-b M-И “V A
Щ
2А
$
й
l/i
ш
т
2/5
* fto m Figure 20 Inscriptions o f the Nagyszentm iklos treasure I (1 -6 )
ь
The sources
129
y% \
?/t
?/l
№ a/2
C O
m
M Ло
9
81 'Й .
f
t лъ
\
Figure 21 Inscriptions o f the Nagyszentm iklos treasure II (7-14) and the tent (15)
130
M ethodological introduction and the sou red
The sources
131
to that o f the runiform inscription revealed in the other im portant find o f the period, the treasure o f Nagyszentm iklos. D uring a fresh exam ination o f the earlier find m aterial, it em erged that runiform inscriptions can be found on other needle-cases and various other objects. The late Avar period cem etery o f Szarvas w hich gave us the bone case also produced another inscription on a belt-end. The treasure o f Nagyszentm iklos is a gold hoard o f 23 buried objects, whose age and origin have rem ained uncertain since its discovery in 1799. The objects originate from a variety o f periods, and several objects have cop ies in the hoard. The treasure includes a Greek-language inscription in Greek characters, as well as a non-G reek inscription using G reek letters, and a num ber o f runiform inscriptions. Am ong the runiform inscriptions, there are a num ber which occur several times. A total o f 14 types o f inscription can be distinguished, together with a sketch depicting nom adic tents. One particular inscription appears five times, while two others each feature twice. Based on analogies with the above-m entioned bone needle-case inscription o f Szarvas, whose date o f origin can be established com paratively accurately, w e are now able to date the inscriptions from the Nagyszentm iklos hoard to the second h alf o f the 8 th century. The m ost recent survey o f the treasure has succeeded in establishing not only the chronological relationship betw een the individual objects, but also the phases o f alteration carried out on these sam e objects. Perhaps the most im portant finding is that the inscriptions were scratched onto the objects at a com paratively late date, while the clasps to be found on a num ber o f the objects, which would have enabled them to be suspended from a belt, are also o f later origin, as indeed are the handles o f several drinking vessels. D istinc tions have been m ade between the earlier and later attachm ent o f such clasps. The craftsm an who carried out the attachm ent o f clasps paid no regard to the sole engraved inscription to be found am ong the objects. A cross also appears on one o f the objects, and in this we have been able to distinguish the various phases o f the treasure’s transform ation. However, in some cases it can be proved that the object and its inscription are o f the sam e age. Those inscrip tions which are repeated on several occasions are also contem poraneous. In any event, the history o f the treasure includes a “C hristian” phase, a “runi form ” phase, and a third phase during which alterations w ere m ade to the objects so that they could be hung from belts. From the non-Greek-language inscription which uses Greek letters, two names and titles can clearly be read: Buy/a zoapan and B utaul zoapan. The title o f zoapan is written unusually, on one occasion with an Om ega, and on the other with an Omikron. The two personal nam es are w ithout doubt Turkic, but the language o f the inscription, as with the language o f the runiform
132
M ethodological introduction and the sources
О
1cm
Figure 23 H air pin, Przem ysl
inscriptions, has so far not been satisfactorily deciphered. It is very probable that they are in a Turkic language, but the inscriptions are too short, and we have no bilingual inscriptions at our disposal. For this reason, attem pts to decipher the inscriptions have so far proved inconclusive. M uch is certain, however, that both the Szarvas and the N agyszentm iklos inscriptions were w ritten in the language o f one o f the peoples o f the late Avar period, and that this people em ployed a form o f w riting that did not use Rom an characters. It was a w riting used by a people w hose local leaders had Turkic nam es and bore Slavic titles. This constitutes a substantial step forward as regards the identi fication o f the peoples o f the late Avar period. The runiform inscriptions, in respect o f type, belong to the fam ily o f Eastern European runiform writing. A rchaeological finds containing Eastern European runiform w ritings have m ultiplied considerably over recent decades. These finds can be divided into the follow ing geographical groups: (1) the N orthern Caucasus and R iver K u ban region; (2) the territory o f the Volga and Don rivers; (3) the region o f Dobruja; (4) Eastern Bulgharia; and (5) the C arpathian Basin. Unfortunately, the overw helm ing m ajority o f over one hundred inscriptions have not been reliably published. All are very short, and we know o f only two inscriptions longer than that on the Szarvas needle-case, and even these contain only a few characters more (Elista, M ayak), while there exist no inscriptions in two lan guages. It is certain that we are dealing here with several local variants o f one script, and it is a striking fact that this alphabet appeared in the land o f the Khazars and the Onoghur-Bulghars. N ot long ago in Siberia, where Eastern Turkic inscriptions can be found, three inscriptions were discovered w hich are not a part o f the already deciphered and w ell-know n body o f Eastern Turkic inscriptions, but which belong to the above-m entioned Eastern European type o f writing. These w ritings are im portant from several aspects with regard to the conquering M agyars. On the one hand, the greater part o f finds have com e to
The sources
133
light in territories which had, or m ay have had, some connection w ith the conquering M agyars. On the other hand, to this day the origin o f runiform inscriptions found in the 8гёке1у region (o f eastern Transylvania) has re m ained unresolved (see Chapter XVI for details). A significant breakthrough was the discovery o f a runiform inscription from the C onquest period on a bone-covered quiver opening that em erged from excavations at H om okm 6 g y Halom near K alocsa (southern Hungary), w hich added a new link to the chain o f inquiry. In Hungarian, both the verb ir ‘w rite’ and the noun betu Tetter (of the alphabet)’ are o f Turkic origin, and it is ju st possible that the M agyars adopted this version o f runiform writing in Turkic, probably Onoghur-Khazar, surroundings, ju st as they did the m ajority o f the borrowed words. The Sz 6 kely runiform script, however, contains characters o f Greek origin as com m uni cated through Slavic, and we are now better able to explain the problem o f its origin in the light o f new archaeological finds (see C hapter XVI). The rich supply o f archaeological m aterial from the late Avar period gradually began to dry up in the 9th century as it becam e m ixed with Slavic and Frankish m aterial. It appears, however, that at least in certain regions, a localised archaeological culture o f the closing part o f the Avar period becam e established for a short time, and lived to see in the M agyar conquerors. One typical feature o f this culture was the secondary utilisation o f objects from the late Avar period, as can also be observed in a num ber o f graves from the Conquest period. We m ust probably reckon on the presence o f Slavic peoples as early as the beginnings o f the Avar Empire. Still, the task o f isolating archaeological finds pertaining to the Slavs meets with not inconsiderable difficulties. The gener ally held view is that, as opposed to the A vars’ skeletal form o f burial, the crem ation o f the dead and subsequent burial o f ashes in urns is indicative o f a Slavic-type settlement. This form o f burial, however, by its nature leaves little trace, and thus the sm aller num ber o f finds and paucity o f grave goods may actually be a consequence o f this particular burial practice. A separate chapter should be devoted to the excavations carried out in Zalavar (at the southwest end o f Lake Balaton in w estern H ungary) and its environs. The castle com plex found here survived to see the M agyar Conquest, and saw further extensions during the era o f Arpad. Its foundations w ere in all probability laid for Pribina between 840 and the consecration o f the church in 850. Debate is constantly raging regarding the attachm ent to a particular ethnic group o f those who lived around Zalavar and the objects w hich they left behind. H istorically speaking, the presence o f both southern and northern (M oravian) elem ents can be expected. In any event, the m ajority o f this population were among those who settled in around the year 840, and their range o f influence certainly cannot have been as great as that o f other Slavic
134
Methodological introduction and the sources
groups which by then were undoubtedly com m on elsew here in the C arpathian Basin. M ore or less sim ultaneously with the C onquest came the appearance o f a culture w hich came to be known as the culture o f Biyelo-Brdo, after a settle m ent in the old county o f Veroce beside the R iver Danube. C om m on am ong the finds from this location are S-ended hair bands, while the burial custom s are o f several types. Grave goods, however, are very m eagre indeed. P re viously the entire culture had been regarded as a legacy o f the Slavs, but is now thought o f as com prising the burial sites o f com m on M agyars. Follow ing all these archaeological antecedents, the M agyars appeared, leaving behind them an entirely characteristic, easily distinguishable archae ological legacy. Archaeological evidence o f the conquering M agyars does not cease im m ediately upon the founding o f the state o f Saint Stephen, and graves from as late as the 6 th and 7th decades o f the 11 th century have been excavated w hich, although dated by coins from the age o f Arpad, preserve in one way or another the characteristics o f archaeological finds o f the C onquest period. From this tim e on, however, this type o f burial site rapidly disappeared, giving way everyw here to the practice o f Christian burial w ithout grave goods. These graves thus contain m aterial evidence from a period o f some 180 years in length. It is worthw hile m aking special m ention o f the graves o f m arauding M agyars, or at least those with coins looted, acquired or received during the period o f the M agyar raids on Western Europe. At the same time, it is sur prising that practically no foreign objects have been found in these graves. M etal articles were in all probability m elted down. W hat, then, are the historical questions which are answered, or m ight be answ ered by this rich and thoroughly processed body o f archaeological data from the C onquest period? First o f all, let us em phasise that the archaeological m aterial reflects the situation at a given time. Although this situation carries traces o f its historical antecedents, we cannot expect too m uch from this. W hile it is true that burial custom s and concepts o f the afterlife are very durable, and m ay rem ain unchanged for a long time, they nevertheless do go through surprisingly rapid change on occasions, and such changes are not always m otivated by historical or ethnic transform ations. For instance, as was stressed above, it w ould be a m istake to assum e that the ethnic origin o f the population o f B udapest could be determ ined from the proliferation o f burials by crem ation here, as coincid ing with the presence o f Soviet troops. Similarly, it w ould be a m istake to conclude from a rapid increase in the wearing o f jeans that a significant influx o f A m ericans has occurred. Custom s and dress are characteristics o f an ethnic group, but only in the context o f the prevalent environm ent. We can see the
The sources
135
signs o f a surprisingly uniform culture on the steppes o f the 9th and 10th centuries, w hich nevertheless has rapidly changing elem ents at the local level. At the sam e time, m ilitary expeditions, m erchants and prisoners o f w ar both im ported and exported objects, ideas, fashions and technical developm ents. The archaeological excavator, no m atter how rich the m aterial, is at the m ercy o f chance. For this reason, periodic efforts am ong archaeologists to m ake ethnic-based differentiations betw een each object, phenom enon or even group o f phenom ena from the Conquest period (e.g. the Khavars) have so far proved unsuccessful. The Conquest period graves o f course speak volum es to an expert in the field. They reveal to w hat degree the leading strata o f society w ere isolated from the sim ple, com m on people. The exceptionally rich graves (such as at Кепёг1б, G esztered, Rakam az and Karos), which are usually known as princely, chieftains’ or tribal leaders’ graves, can in some cases be easily dated precisely because o f their opulence. W herever dating was possible, archae ologists have been able to establish that the graves originate from the first h alf o f the 10th century. From these it is possible, to a certain extent, to isolate the archaeological m aterial o f typical Conquest period m ilitary escorts, and that o f independent clan leaders and m em bers. Finally, the rem ains o f common people, who lived a life o f servitude and bore the hardships o f everyday work, can also be found am ong the graves. Social standing was probably indicated by objects other than those which designated m ilitary rank. Objects distinguishing rank m ainly included belts, harnesses, sabretaches, sabres, bow holders and quivers. Belts had a special role for men. In the Turkic language, the word m eaning ‘b elt’ and the word m eaning ‘ran k ’ or ‘social status’ are identical (kur). The Turkic polym ath and lexicographer Khashghari, who was alm ost a contem porary o f the M agyar Conquest, wrote around 1074 that the sentence m eaning “m y belt is large” (m ening kurim ulug) also m eans “I have a high rank”. In the sim ilarly Old Turkic Khutadghu bilig (W isdom o f Royal Glory) written in 1069 in the m irror-for-princes style, we read the phrase: “do not allow the sm all-belted ones to mix with the large” , in other words, do not let those o f low rank m ingle w ith those o f high rank; while the expression “he has the belt o f a beg” (bu beglik kuri) m eans that the person in question holds the rank o f a beg. The O ld Turkic word kurdash— literally ‘belt-m ate’— m eans ‘one o f identical rank, belonging to the same social stratum ’. In M ongolian, one way o f describing a wom an is ‘beltless’, as opposed to the belted men. The conquering M agyars wore em bossed belts cast o f silver (see Plates V I-V II), which in the case o f those o f com m on rank were unom am ented and held together with an iron clasp. Each belt had a clasp, and a num ber o f ob jects w ould have been suspended from the strap, including a sabre, bow
136
Methodological introduction a n d the sources
holder, quiver and sabretache (see Plates V III/1-4, IX). The sabretache was closed with a small slat o f bone and attached to the belt w ith an em bossed hook. Besides belt accessories, the sabretaches are the m ost characteristic finds from the graves o f the C onquest period. They w ere strengthened w ith m etal plates, generally o f silver. A t the side o f each bag, a strap w as threaded through, and both this strap and that w hich attached the bag to the belt w ere decorated w ith m ountings. The sabretache, w hich fulfilled the function o f a pocket, would have held fire-m aking tools: flint, steel and tinderbox (see Plates X /l-2 , X I/1-2, XII, XIII, XIV and XV). A slightly curved-bladed sabre is typical o f the graves o f the early conquer ors, w hile later graves see the appearance o f double-edged straight swords, interestingly enough often found in com m oners’ cemeteries. The thoroughlyfinished reflex bow s w ere reinforced w ith plates o f horn, and often had braces fashioned from stag antler. Battle hatchets and lances, m eanw hile, are surpris ingly rare. Harnesses can be w ell reconstructed from the finds o f the C onquest period graves, as can saddles (see Plate XVI) and stirrups. The typical, but not ex clusive, form o f the latter was a pear shape. Several m ethods o f suspending the stirrups have also been dem onstrated. Based on the objects unearthed from fem ale graves, a good picture o f the clothing and jew ellery o f the conquering M agyars’ w om en can be obtained. They certainly wore head-dresses and hair adornm ents, m ostly decorations braided into the hair, and earrings. Their clothes w ere ornam ented w ith m etal pendants, and they w ore necklaces and bracelets. Their dresses opened in the m iddle and reached to the knee, and w ere adorned with collar decorations, and various other em bossed em bellishm ents and buttons sew n on. The ornam entation o f objects o f the C onquest period is very varied indeed, but the m ain characteristic feature is a stylised flow er consisting o f 3 -5 petals, w here the edges o f the leaves are shaded with grooves. The leaves, leaf-stem s and creepers w hich jo in them together, often intertw ining in a com plicated pattern, cover the w hole surface o f a given object. This style o f ornam entation is typical o f the archaeological relics o f the M agyars. Am ong the depictions o f anim als is often found a creature w ith the body o f a horse, and the head and tail o f a griffin, and som etim es heavily stylised portrayals o f predatory birds or other animals. Clearly, in the years follow ing the C onquest, great value was attached to the various craftsm en, some o f w hom m ay have com e in w ith the conquering M agyars, or m ight have m oved in at the M agyars’ invitation, but o f w hom the m ajority w ere obviously locals, joining up w ith the M agyar people in the hope o f protection and accom m odation, and striving hard to fulfil their m asters’
The sources
Figure 24 Drawing o f a Conquest-period grave. M indszent (Csongrad county, Hun gary)— Koszorudtilo, Grave 2
The reconstructed objects and grave goods o f a m ounted archer: (A ) saddle under his head: (1) pair o f stirrups, (2) snaffle, (3) twenty-one silver plates decorating the har ness or the saddle cover; (B ) relaxed bow: (4) bone plates covering the nocks and (5) the grip; (C) arrow quiver: ( 6 ) two arrowheads, (7) bone plates enframing the quiver opening, ( 8) remains o f the iron straps o f the quiver, (9) iron braces supporting the quiver, (10) quiver’s locket; other: ( 11 ) belt buckle, ( 12 ) h orse sk u ll and leg b o n es. (B a se d on Csallany.)
137
Drawing o f a C onquest-period grave. K en ezlo (B o rso d -A b au j-Z em p l 6 n county, Hungary)— Farkaszug, C em etery I, Grave 11 Reconstructed objects and grave goods o f a mounted archer with a harness decorated with coins: (A ) arrow-quiver placed over the left thigh: ( 1) locket, ( 2 ) cover plates o f quiver mouth with 8 arrowheads, (3) lockets o f quiver, (4) nails o f the quiver bottom; (B ) saddle placed on the thighs: (5 ) snaffle for a foal, ( 6 ) pair o f stirrups, (7) strap buckle, ( 8 ) sixteen perforated Italian silver coins from the rule o f Berengar I (8 8 8 -9 1 5 ) and the joint rule o f H ugo o f Provence and Lothar II (9 3 1 -9 4 5 ) fixed to the straps o f the harness, (9) the dislocated seventeenth coin; (C) bow quiver: ( 10 ) cover plate o f quiver mouth. ( 11 ) plait ring, ( 12 ) mounts o f the belt, (13) dislo cated buckle and locket, (14) horse skull and leg bones. (Based on A. Josa.)
138
Methodological introduction and the sources
demands. M etallurgy m ust have been continuously practised even follow ing the Conquest, while potters m anufacturing earthenw are vessels were also im portant craftsm en. Evidence o f these is well docum ented by archaeological excavations. Am ong the body o f ceram ic objects o f a rich variety o f forms, the suspendable clay pots are particularly conspicuous. The grave pits were o f w est-east orientation and o f a rounded rectangular shape, the deceased placed w ithin with face and legs pointing tow ard the east. With regard to the orientation o f both graves and tents, the steppe can be divided into two large regions. Southern orientation is typical o f the eastern h alf o f the steppe, and is observed at the sites o f the M ongolians and their neighbours. Eastern orientation, on the other hand, w hile know n at the sites o f every nom adic Turkic people, is also found am ong the burial custom s o f other peoples. The bodies were placed on their backs and in an elongated position in these graves. Am ong the graves o f those o f higher rank, tw o forms o f horse burial were discovered. The m ore com m on form saw the butchered horse flayed, with the skin and still intact head and limbs placed in the grave, very often alongside the legs o f the deceased. M ore rare were burials o f stuffed horse skin. From the bones it can be ascertained that the deceased person’s favourite steed was buried (and not an old nag or— perhaps because o f its m eat— young horse), on rare occa sions alongside wom en and children. We have already discussed the shrouds. It is not a characteristic o f the graves o f the M agyar conquerors that items o f food or drink were given for the long journey into the next world. Besides independently-lying, single-occupant graves, there also exist burial sites o f sm all groups (clans or extended fam ilies), or o f larger com m unities. There is some doubt as to how m uch the layout o f cem eteries reflected the order o f rank in society. It is possible that once the head o f a clan had been buried, the other m em bers o f the clan w ould then be buried one after the other in a row beside him, but it is also conceivable that in the centre o f such a row-like cem etery arrangem ent, a place was left free for the senior m em bers o f the clan. However, besides exam ples w hich suggest this schem e, there are num erous other cem eteries where no burial according to an “order o f rank” is discernible. The archaeological finds o f the C onquest period, as we have seen, are prim ary sources o f data regarding the m aterial culture o f the conquering M agyars, while also indirectly helping us reconstruct their society and system o f beliefs. The fact that this archaeological culture came to an end in the 11th century, giving way to an entirely different system o f burial, does not o f course mean that a new people m oved into the Carpathian B asin at this time. The new -style cem eteries which grew up around the churches were the reflection o f a fundam ental transform ation in society.
The sources
139
We have observed the failure o f an attem pt to trace back the body o f Hungarian archaeological finds, them selves well defined, in a direction op posed to the w estward advance o f the M agyar people. Given that from the point o f view o f ethnic identity the archaeological findings are o f no use prior to the Conquest (the m aterial from the region o f the Kama proving an exception), we m ay m ake use o f this archaeological m aterial as a source o f inform ation with regard to the early history o f the M agyars in two other ways. As we shall see from an exam ination o f the various nam es given to the M agyars, it is possible to determ ine the identity o f the M agyars’ neighbours before the Conquest o f the final hom eland. For this reason, the archaeology o f the Eastern Europe o f the 6 th—9th centuries is also significant, even if we are unable here to directly pinpoint M agyar groups. On the one hand, the archaeological relics o f the Onoghur-Bulghars, Khazars and Slavs assist in the reconstruction o f the history o f these peoples, and thus indirectly the history o f the M agyars. On the other, they m ay serve to indicate the generally prevailing level o f econom ic activity, technological developm ent and culture, which may be expanded m ore broadly to apply also to the M agyars. An im portant place am ong these archaeological cultures is occupied by the so-called Saltovo-M ayak culture, whose beginning can be traced to the m iddle o f the 8 th century, and conclusion to the end o f the 9th and early 10th centuries. By now three variants o f this culture have been distinguished, a northern, a southern and an eastern territory, o f which the southern territory is usually attached to the Onoghurs, and the eastern to the Khazars. It cannot yet be proven that the northern territory is connected to the Alanians (see pp. 200-203). The Saltovo-M ayak culture, or cultures, indicate(s) an intensely agricul tural, livestock-breeding, sem i-nom adic econom ic system. In other w ords, a system such as that reflected in the Turkic loan words to be found in the Hungarian language. Very intensive excavations are in progress over the territories o f today’s Finno-Ugrian peoples, and in many cases these have helped to successfully trace back in archaeological term s the continuous history o f peoples living in those regions today. Unfortunately, however, local archaeologists in this region are not entirely w ithout certain prejudices, m aking the evaluation o f local data som ewhat more difficult. The identification o f one archaeological culture or another with the “U grians” or with the “Perm ians” is generally backed up by linguistic arguments and local, autochthonous theories, but these have little real archaeological basis. Despite all this, it is worthwhile paying attention to those archaeological cultures which in both space and time are genuinely close to locations and
140
M ethodological introduction and the sources
Variants of the Saitovo-Mayak culture A. Upper Don region B. Lower Don region C. North Caucasian variant D. Dagestan variant E. Crimean variant F. Unexplored steppe variant G. Danube-Balkan variant
East Slav archaeological cultures ------ Romni-Borshevo culture ......
— •-
Kurghan grave sites of the Polyans in the 10th—12th centuries Luka culture
1. 2. 3. 4. 5. 6. 7. 8. 9.
Semender 3 • Balanjar Derbent Etil Sharkel Kerch (Tamatarhan) Kherson Kiev Chernigov
Figure 26 The Saltovo-M ayak culture, the neighbouring archaeological cultures and the possible location o f m ajor towns
periods deduced through other sources. Here, however, archaeology is not a source o f inform ation about ethnic history, but at the very best leaves behind an im print o f the various econom ic, technological and social transform ations in the region.
6. ETHNOGRAPHIC SOURCES Prior to the m iddle o f the 20th century in Hungary, the focus o f cultural anthropology or ethnography was peasant culture. However, from the second h alf o f the century onwards the outlook broadened and began to look at the hum an cultural aspects and questions regarding cultural consolidation. It was
The sources
141
long believed that archaic cultural aspects could be o f aid in creating a re construction o f the conquering M agyars’ culture. This m ethod was used to exam ine the so-called double-yard settlem ent system w hich was in use on the Great H ungarian Plain and its peripheries, w hereby the stock-farm er peasant had an inner piece o f land, where he lived, and an outer one, w here he kept his anim als. However, the earlier theory that this settlem ent system carried traces o f a pre-C onquest nom adic settlem ent form has been dism issed. It has also been suggested that o f the two corn-w inning techniques, treading-out was used prior to the Conquest, while the knowledge o f threshing was acquired in the C arpathian Basin. This was also refuted later. O ther theories have conjec tured that the dom estic fam ily arrangem ent, whereby several generations live together in ‘extendedfam ilies ’or clans, a form that has survived— particularly in N orthern Hungary— to the 20th century, was derived from the conquering M agyars. But even this was found to be incorrect, as it transpired that it originated from after the Turkish occupation o f Hungary, when it developed as a tax avoidance measure. N either did the opinion that the M agyars’ belief in wizardry and witchcraft contained traces o f Sham anism stand up to scien tific scrutiny. It can, therefore, generally be said that these ethnographic char acteristics, which were indeed archaic by the 19th century, developed after the Conquest, and neither the com parative studies nor the careful exam ination o f historical data could yield any evidence that they had existed prior to that. It is possible though that the unique duality o f Hungarian viniculture preceded the Conquest. In the 19th century, viniculture in the northeastern wine district and the Transdanubian region was still considerably different. C ertain aspects o f Transdanubian vine-growing, such as the tools used, can be traced back to Rom an times. It was this, as w ell as certain techniques used in the northern region o f the B lack Sea, such as the use o f hum an w ine-pressers to obtain the pressure (as opposed to the use o f mechanical pressers), that gave rise to the theory that the Tokaj region’s viniculture, in N ortheastern H ungary was o f Oriental origin and came with the Hungarians into the Carpathian Basin. The key reasoning behind this was based on the vinicultural loan w ords in use, and the fact that the nam e Tokaj is o f Turkic origin. The fact is, however, that the original nam e o f the hill was K opaszhegy (B ald Hill), and was only later nam ed Tokaj after the nam e o f the nearby village. However, it is disputed where the nam e Tokaj, first m entioned in 1333, had com e from. It is possible that we are looking at the derivation o f a Turkic word referring to the shrubs that grow on the shallows o f river bends. B ut even if that is the case, it has very little to do with viniculture. Since the publication o f the research w ork o f Вё1а Bartok, ethnographers have been exam ining the oldest form s o f Hungarian folk music. One such form is characterised by the use o f the /д-pentatonic scale w hich lacks sem itones,
142
Methodological introduction and the sources
and o f a m elodic scheme in which the second h a lf o f the m elody repeats the first h a lf at a fifth below. It was believed for som e tim e that the sim ilarities betw een Cherem is m elody patterns and th ispentatonic ‘fifth -sh ifte d ’ m elodic structure w ere the key to the origins o f Hungarian folk m usic. However, detailed studies and further local field research have shown that the C herem is m elodies in question are not necessarily as old as was earlier suggested and that they developed at the confluence o f the Cherem is and C huvash language territories. A lthough this claim refutes that com m on features survived in two distant regions— w hich w ould suggest a cultural link in the genetic sense— it nevertheless indicates an interesting typological parallel, nam ely that a unique m elodic form like this could develop on the border o f Turkic and Finno-U grian folk m usic regions. It is with the exam ination o f this typological parallel that ethnography can contribute m ost to the research o f the origins o f the conquering M agyars. But before w e exam ine this in depth, there is one other connection to be m entioned. The folklore, beliefs and m yths o f the O b-U grian people have been well docum ented since the 19th century. It becam e clear that the horse and the horsem an play an im portant role in the life o f the O b-U grian people. Yet the horse has no im portance in contem porary dom estic arrangem ents in these areas, and is equally insignificant in the folklore o f the peoples o f the taiga belt. This w ould indicate that one com ponent o f the people m oved from further south, from the greener steppe where an equestrian life style was predom inant. Ugrian horse-breeding, however, has a bearing on the form ation o f the M agyar people. Therefore, this connection can be used, albeit with m uch caution, to reconstruct the Ob-U grian and indirectly the Ugrian people’s history. Ethnography can contribute considerably to the understanding o f the char acteristics and developm ent o f the nom adic life style. A lthough these analo gies cannot be used in their own right, nevertheless, they serve to elucidate. In som e cases they m ay pose questions w hich thereafter have to be answ ered by detailed research. But taking into consideration that the nom adic life style is in decline, the ethnography o f the 19th- and 2 0 th-century nom adic peoples can provide invaluable information. In 1 9 5 8 ,1 spent som e time studying the life o f the huntsm en o f northw est M ongolia. The study o f the solitary trapper roam ing the forests, even using a bow and arrows, was in itself an interesting life style to observe. B ut more im portant was the relationship betw een these huntsm en and the nom adic stock-breeders living further to the south. A connection em erged betw een the two. I f one o f the nom adic fam ilies had becom e im poverished as a result o f losing its anim als due to some m isfortune, it could m ove into the forests and turn to hunting and trapping. B ut this “m igration” worked both ways. Some o f the hunters becam e rich enough to buy anim als, m ove to the greener areas
The sources
143
o f the steppe and pursue a nom adic life style. It is clear that such changes in life style also occurred on a larger, historical scale to whole groups o f peoples. Some written evidence exists o f this, albeit from the 12th and 13th centuries. It has been debated for the past half-century w hether the conquering M agyars were nom adic or only partially nom adic, and if the latter was the case, what exactly “partially” nom adic meant. In M ongolia, three forms o f agriculture w ere present. One w as the Chinesetype irrigation system , another was the European-style large estate method, and the third was a m ethod used by the nom adic farmers. It is im portant to note that the Chinese irrigation m ethod was practised by Chinese im m igrants, with Chinese tools and a Chinese crop-m anagem ent system. The nomadic lords o f these regions were tolerant and even supportive tow ards these activities, because they had a dem and for the goods. However, once the Chinese reached a certain level o f wealth, they w ere hounded from the region, as they had becom e a threat to the political establishm ent. A few decades later a new batch o f Chinese im migrants would start the sam e practice all over again in the same region. This was a particular form o f intercourse betw een different ethnic groups. For our purposes, however, the nom adic fa n n e r’s method is the most interesting. The w inter lodgings were in the river valleys which were m ore protected from the harsh winters. In the spring, the nom ads sowed the grain in the virgin earth, ploughed by simple w ooden ploughs. No harrows w ere used: the larger pieces o f earth were broken up by hand. The seeds w ere also sown by hand, although there is evidence o f some m achinery used in Asia. Follow ing this, the nom ads m oved on to the sum m er pastures. The distance betw een the farthest points o f the w inter lodgings and sum m er grazing grounds depended on the clim ate as well as on the geobotanical environm ent. The closer a tribe lived to the arid, desert steppe, the w ider the area they roam ed. They arrived back for the harvest in the early autumn. They used hands, occasionally sickles, to gather in the crops. These crops were placed on the ground in the barnyard and the seeds were extracted by leading the horses, occasionally the oxen, over them. One o f the tribesm en w ould stand in the m iddle with the bridles o f six to ten unshod horses in one hand and a whip in the other driving the anim als round in circles over the gathered crops. The C hinese, however, used threshing cylinders. Once the grains were extracted, w ooden spades were used to throw the seeds into the w ind to separate the ch aff from the wheat, barley or rye. The grain was then put into leather sacks and som e o f the straw was used to line pits, dug into the ground which provided storage for the sacks o f grain, keeping them from the w inter frost. Thereafter, the grains w ould be turned into flour either by hand, using two grinding-stones, or w ith larger grinding-stones driven by horses. Follow ing the harvest the nom ads w ould
144
Methodological introduction and the sources
Figure 27 M ongolian tent roof-hoop
Figure 28 Wooden fram ework o f a K hazakh tent
145
The sources
A Шжт*тШ т§
т
^
гШ
*
f f ';
mm
«Я И ^Рч- ^
"■'< . „*>*"'
4
-' '
V
Щ , i f :4 Ж ;М
Figure 29 W ooden fram ew ork o f a M ongolian tent
move on to their w inter lodgings w hich would not necessarily be in the same place as in the previous year. The w ooden plough was light enough to carry with them, along with tents and other accessories. Thus, as the M ongols put it: “the land also m igrates”. It is therefore likely that they ploughed previously untouched land every year. This is one form o f nom adic agriculture for which evidence can be found even in the 20th century. Nevertheless, it is worth com paring this with Jayhani’s description o f the M ajgars’ life style: “W hen the w inter m onths are approaching they all move down to the river valley which is nearest to them. They stay there for the w inter and fish [...] The M ajgars’ land has m any trees and rivers. Its soil is damp. They have a lot o f arable land.” Scientific research could not, as yet, find correlation betw een the nom adic life style and arable land, and therefore concludes that these state ments refer to different eras. This M ongol ethnographic analogy m ay provide useful evidence. A rchaeologists excavating Conquest-period settlem ents have observed circular-shaped ditches which, given their radius, would suggest that som e form o f a building— m ost likely to be a tent— once stood there. The question is, what type o f tent?
146
Methodological introduction and the sources
Figure 30 D epiction o f a tent on a Chinese Tang dynasty lacquered bowl
Ethnographic research has shown that in Eurasia and N orth A frica two types o f tents w ere in use. The so-called “black tent” was com m only in use from M orocco to A fghanistan. This had a top cover, vertical sides and was norm ally oblong-shaped. The inside-planes could be m ade o f different m aterials but the outside covers were black canvas or som e other black m aterial, hence the name. This type had m any derivations. The other m ain type was the circularbased yurt. The walls o f the yurt were constructed from w ood into a lath mesh. A circle was then formed from this lath m esh, onto w hich w ooden roof-poles were placed to form the ro o f which at the top supported the ‘ro o f h o o p ’ which served both as a ventilation outlet as well as an opening to let in sunlight. A w ooden-fram ed door was hung onto this structure. This yurt-type, too, had m any varieties, am ong them the Turkic and the M ongolian ones w hich had different roof-pole arrangem ents. The M ongolian roof-poles w ere straight, while the Turkic ones were curved at the low er end w here they w ere connected to the lath m esh structure. This lent a m ore dom e-like appearance to the roof. Arab sources refer to this type o f yurt as being used by the M ajgars. The yurts could be dism antled in two hours and packed onto the carts that follow ed the nomads. Some yurts, however, w ere built onto the carts and did
The sources
147
not require disassem bling at all. These “high carts” are m entioned by the Chinese sources in descriptions o f the Turk people. Some yurts w ere sim pler and did not have a wooden ‘ro o f hoop’ at the top. It is a depiction o f such a yurt that can be seen engraved on one o f the item s o f the N agyszentm iklos treasure. The interior arrangem ent o f the yurt also follows a strict order. The M ongols have their doors facing south, while the Turkic always have theirs facing east. This corresponds with their burial directions. Entering through the doorway the left side o f the yurt is to the right, as the following directions are given from a position facing the door from the inside. The left side o f the yurt is the fam ily’s and the w om en’s side, while the right side is for the m en and the guests. Opposite the door is the shrine where, in earlier days, the M ongols used to place statues o f Buddha and before that, other cult objects. During the socialist regim e the party secretary’s picture was hung here. In the centre o f the yurt stands the fireplace which used to be occupied by a three-legged cauldron, and dried dung was used for fuel. The saddles and the tools used by the m en are hung on the right side o f the yurt while the kitchen utensils are on the left. They sleep on felt carpets and use fur blankets. M any o f the M ongols who settled in towns continued to live in yurts for some time. These yurts were constructed on a clay foundation, and some o f the yurts we came across were actually built from clay. It is, therefore, possible that the circular ditches found by the archaeologists w ere in fact foundations o f clay, not felt or canvas, buildings. All this inform ation, however, provides little m ore than an analogy in trying to reconstruct the tent used by the conquering M agyars. Little is know n about the conquering M agyars’ religion. Arab sources sug gest that the M agyars were fire-worshippers. But there is no concrete evidence to suggest that this was the case. The w orld religions on the Eurasian steppe, such as C hristianity (particu larly its Nestorian variety), M anichaeism , Buddhism and Islam m ade great efforts to convert the nomads. In relation to the M agyar conquerors, we are m ost interested in w hat their body o f beliefs were before their first contact with the w orld religions, and when they were first subjected to the influence o f these religions. The religion o f the nom adic people o f the steppe has been revealed to some extent from early docum entation. Ibn Fadlan reports on the phallic cult o f the B ashkir people. He describes how he observed a penis-sized piece o f wood being worshipped, following which, he got his interpreter to ask why this piece o f wood had becom e a deity. The B ashkir gave the following answer: “Because I m yself was begotten by som ething similar, and I know o f no other creator.” Ibn Fadlan then lists the twelve gods w orshipped by the Bashkir people. These
148
Methodological introduction and the sources
are: winter, summer, rain, wind, the tree, the hum an being, the horse, water, the sun, night, death and the earth gods. B ut the God o f the H eavens is the greatest o f all. In Yakut’s listing o f these gods, the god o f life is added alongside the god o f death, but his list adds up to fourteen gods. In both cases, in the centre o f all was the sky in the role o f a suprem e deity. The Turkic nam e for this god, quoted by Ibn Fadlan from the Oguz, was Tengri. In the history o f religions the nom adic form o f religion, with a suprem e deity representing the Heavens, is known as Tengrism. This is typical in large nom adic em pires where all the pow er is in the hand o f one ruler, reflected in the religion o f these tribes. According to Turk inscriptions, the T urk ruler was descended from the God o f the Heavens and had sim ilar pow ers. It is im portant to understand the m ythological function o f this god. A ncient Turkic w ritings suggest that he created the world. Linguistic research, how ever, shows that this did not m ean a creation from nothing but a creation from an uncultivated mass. We have som e evidence to suggest that the conquering M agyars had a sim ilar religion. The Turk inscriptions use the verb ya ra t in describing the process o f creation by the G od o f the Heavens. D ated 732, the K ol Tegin inscription reads: “ I, the H eaven-like, created by the Heaven, the Turk Bilge K haghan” (tengri teg tengri yaratm ish ttirk B ilge khaghan). The m eaning o f this Turkic verb is “to create som ething new from existing parts” (e.g. sentences from words, buildings from bricks), as opposed to the concept o f creation by “division” (e.g. o f light and darkness) and creation from nihility. The conquering M agyars adopted this word from the Turkic language, keeping the m eaning o f it which lives on today in the Hungarian word gyart ‘m anu facture’, ‘produce’, ‘fabricate’. M uch is now know n about the characteristics o f Tengrism . It had many variants, but its m ost advanced form was reached am ong the Khazars. It is know n that Sham anism coexisted with this, w hich often m akes it difficult to differentiate betw een notions o f the tw o types o f religious beliefs w hich were, in any case, likely to be closely linked. One o f T engrism ’s typical features was the Sacral Kingship. Ibn Fadlan writes about the K hazar Sacral King who is entrusted with the people’s w ell-being and w ho is ritually exe cuted once he has served a specific term. This institution o f Sacral Kingship is also apparent in the history o f the Central A sian Turk people. Al-Istahri o f the 10th century wrote the following about the K hazar people: “W hen they have found a candidate to becom e the khaghan, they take a silk string and start strangling him until he loses his breath. They then ask him and he w ill tell them for how m any years, he w ishes to rule. If he does not becom e deceased w ithin these years, he will be m urdered when the num ber o f years he has nam ed are up.” The chronicles o f the Chinese Zhou dynasty say the follow ing
The sources
149
about the Turk people: “If a new leader is chosen, those o f a high rank around him will wrap him round with felt and turn him round nine times. They then sit him onto a horse and am idst an array o f bow ing they m ake him ride around. They then start to strangle him w ith a silk string until he is barely alive. They release the string and quickly ask him: ‘how long do you w ant to rule fo r’? The khaghan utters and m urm urs som e incom prehensible w ords in a sem icon scious state. They w ork out from these sounds how many years he has to rule.” Some recently found Uighur runiform scripts, dated from 750, also refer to this tradition. The holiness o f the king has, naturally, no connection to Christianity. These sources clearly state that the king was holy because he had khut. On the so-called “sm aller” Uighur inscription found at K harabalghasun, the deceased says the following words about himself: “ Having bought the East under the range o f my bows and arrows, and driven the West under m y power, my khut had shrunk in the blue sky, m y road had shortened on the brown earth, and so I m et m y dem ise, Buqa ogul, I am now dead.” The people begin to have bad fortune when the leader loses his khut and he finally dies w hen his khut ceases. During his lifetim e the leader has khut or holy khut (khutlug, idukh khut). The khut is a substance that brings luck, and com parative research shows that this was very m uch a m aterial notion. Here is the Khirghiz description o f khut: (1) A dark red gelatinous m aterial which drops through the ro o f outlet o f the yurt. Touching it brings luck but only in the hands o f good and honest people. In the hands o f evil people it will turn into a piece o f dung. (2) A talism an that protects people and anim als. Little hum an figures m ade o f canvas or lead, dressed in red or blue rags and decorated with m other-of-pearl buttons which are attached by seven fine wires. These are washed with w ater which is then sprinkled onto the anim als. (3) Idols (particularly in the m ythology o f neighbouring peoples such as the Kalmtik). (4) Vitality and the spirit. (5) Luck and happiness. It is, therefore, clear that the presence o f vitality im agined in some material form will ensure the leader’s holiness. M any Siberian am ulets have been found and a num ber o f objects attributed to the conquerors have been considered to be am ulets which w ere also know n as ongon. This word is o f M ongolian origin and becam e part o f scientific term inology from its use by the B uryat people. Their function w ith the Turkic nomads was linked to the khut. The consequences o f losing the khut are described in detail by M asudi. He writes that one o f the kings o f the K hazar people exercises all actual pow er over the people, including the other king, the hakhan. The hakhan lives in his palace, he cannot ride out from there, m eet any o f the noblem an or the ordinary people. He is not even allow ed to leave his quarters w here he lives with his wives. He does not com m and or forbid, and he has no say in the state’s business. Nevertheless, the kin g ’s sovereignty is not guaranteed unless the
150
Methodological introduction and the sources
hakhan is known to be present with him in the capital, in his palace. If K hazars are in distress, starving, or if w ar breaks out, the people and the nobility will hurry to the king and claim that the hakhan is a bad om en and good for nothing. They will ask the king to “kill him or allow us to do it. Som etim es the king will take this step him self, at other times he will plead for the hakhan's life, claim ing that he cannot find him responsible for any o f the m isfortunes. That is the current-day K hazar practice”. This description o f the K hazar tradition is from before 956 and refers to the final phase o f this sacral dual kingship. M ore than thirty years earlier, Ibn Fadlan writes the follow ing about the K hazar leader w ho is known as the khaghan. He claim s that the khaghan only appears in public every four months, and otherw ise lives in com plete solitary confinem ent. He him self is referred to as the great khaghan and his deputy is known as the khaghan-beh: “The khaghan-beh appears before the king every day and bows down in front o f him in respect. He goes bare feet carrying a piece o f wood. W hen he greets the king he lights the piece o f wood and then sits down next to him on the throne, on the kin g ’s right.” It is worth noting that the stick the second king lights is a torch w hich was called ju la in the West Turkic language {yula in m ost Turkic languages and ju la in M ongolian). This is where the second M agyar k in g ’s title derives from: ju la or jila (see pp. 343-347). Ibn Fadlan continues his report by describing the great khaghan’s harem and burial traditions, and then goes on to say that w hen the king rides out o f town he is follow ed by the w hole o f the army. The soldiers, however, will stay one mile behind him. I f any o f his people catch a glim pse o f him they w ill fall to the ground and keep their heads on the ground until he has passed. The great king will reign for forty years. If he lives a day beyond that, he will be killed by his people who will claim “he has lost his sense and his judgem ent has weakened” . This was, then, a som ewhat earlier phase o f the sacralisation o f the holy kings. This sam e approach could be seen in the Jayhani tradition. The presence o f the holy king, the H azar hakhan, and the suprem e com m ander, the ishad, is also apparent in this tradition. The K hazar people obeyed both o f these figures and took orders from the latter. According to this tradition, w hen the m ilitary com m ander rode out o f town for any reason, a sun-disk shaped object was carried in front o f him which was suspended in a drum -like fashion. A horsem an would carry this object and the arm y would follow the com m ander w atching the light from the object. In G ardizi’s account, this object is replaced by a flam e from a torch covered in wax. This ritual w ould suggest that, despite the fact that both the “great” king and the com m ander were o f Jewish religion, they w ere also sun w orshippers.
The sources
151
M ovses Dashuranci writes about the coexistence o f the Sun God and the God o f the Heavens. He says that the Huns o f the N orthern Caucasus believed in two gods. One is called Tangri han, that is Tengri khan, who is thought to be identical to the Persian Aspandiat, and for whom horses w ere sacrificed, and another who is called Kuar, whose victim s are struck dow n by lightening. K uar is a word o f Persian origin and has the sam e m eaning as the M iddle Persian hvar which means sun. However, the Jayhani tradition does not include the ritual execution o f the king. It is, therefore, clear that the sacral dual kingship was a changing tradition among the K hazar people, which evolved in several stages and reached its most advanced form in the m id - 10 th century. There is no doubt that between the 6 th and 9th centuries Tengrism was the religion am ong the nom ads oi'the steppe and, therefore, we have to suppose that it was also the religion o f the M agyars. Tengrism represented a transitional stage prior to the shift towards the world religions. How this set o f beliefs transform ed to becom e part o f the world religions and how the C entral Asian syncretistic forms o f the world religions came to be m ay only concern us in terms o f distant analogies. Nevertheless, it is im portant to stress that there is no trace o f a sacral kinship am ong the M agyars— as some scholars have propounded. All this, however, does not m ean that beside the Tengrism o f the ёШе there was no presence o f Shamanism. Tengrism and Sham anism together provided the w ider fram ew ork into which the conquering M agyars’ beliefs can be placed. Sham anism is a unique system o f beliefs which can be linked to m any other religions, but one which, nevertheless, is different from all o f them in a few basic aspects. In the centre o f the religion stands the Shaman. The Sham an is different from all other active religious characters, such as the soothsayer, the m edicine man or the witch. The Sham an is a m an— or a wom an— who has inherited the pow er from his ancestors and further developed the skills to create a link between his people and the next world, and is capable o f carrying him self into a trance. This trance places him into a state o f consciousness in which he becom es alm ost totally detached from the visual and sound effects o f the real world. The Sham an is capable o f living with the psychological pressure this puts on him. He has the ability to create contact with the spirits, more often than not the guardian ghosts o f the other world. This occurs in a deeper trance, in the course o f which the Sham an’s soul flies. This is when he is capable o f w andering in the other world and contacting the dead. It is said that when the souls are near ing— som etimes from very far places— their approach is signalled by the
152
Methodological introduction and the sources
beginning o f the trance. These beings can be varied but are determ ined by the regions o f the other world the Sham an is m oving in. His main task is to provide inform ation and help. He establishes contact with the other world in order to serve his people. The Sham an becom es the m essenger and interpreter o f m essages received from the other w orld and thus protects his people from it. The Sham an’s ritual dress, his body-paints and his m ovem ents were determ ined by his com m unity’s traditions. For his journeys to the other w orld he w ould use an object, in m ost cases a drum, on which he “rode” . The m odem research on Sham anism had to be collated, sim ply because in earlier research m any unrelated phenom ena w ere linked to it. The inform ation provided by com parative ethnography indicates that the conquering M agyars had links with Sham anism , although there is no concrete evidence to prove this. The various depictions o f the soul in H ungarian folklore and on archae ological finds are rem nants o f the old beliefs, but in their own right they still provide no evidence o f the presence o f Sham anism. It is also certain that the M agyars came across Judaism , C hristianity and the M uslim faith prior to the Conquest. It can be proven that Cyril m et the M agyars before the Conquest. Kiev in K hazar is Sam bat, the same as the Hungarian word szom bat ‘S aturday’, which is likely to have been derived from the K hazar Jews living in Kiev. The Hungarian word szom bat is also said to have derived from the Slavic w ord sonbota. This claim , however— on account o f the final a o f the Slavic w ord— can be dism issed. The Hungarian w ord szom bat clearly refers to the time o f w eek w hen the K hazar people had an officially accepted day o f worship. The K hazar word is o f G reek origin. The H ungarian w ord vasar ‘m arket’, o f East-Iranian origin, in all probability com es from the Alanian word vachar. The day o f the m arket is in Hungarian vasarnap ‘S unday’ (literally, ‘m arket d ay’), indicating early A lanian-H ungarian contacts. The Slavic w ords sereda, m arking the m iddle o f the w eek (which becam e the Hungarian szerda ‘W ednesday’); chetvertek, the fourth day o f the w eek (which is the origin o f the Hungarian w ord csiitortok ‘T hursday’); and pentek, the fifth day o f the w eek (hence the Hungarian p entek ‘F riday’) are indications o f S lav-M agyar cohabitation. The words hetfo ‘M onday’ and kedd ‘T uesday’ are o f H ungarian origin. The first day o f the week, hetfo, literally m eaning ‘head o f the w eek ’, is described in the same way in the Turkic languages (and hence in Chuvash), using their own vernacular word o f ‘h ead’, ‘to p ’. This is how the B ulghars m ight also have described the first day o f the w eek and the M agyars m ight have sim ply translated that. The word ‘second’ in H ungarian, ketted, described the second day o f the week, hence the H ungarian kedd ‘T uesday’. These are ju st a few o f m any exam ples that show how ethnography
153
The sources
and linguistics play an essential part in reconstructing the history o f the conquering M agyars. As seen here, the Hungarian vocabulary carries m any traces o f history. However, there are no ethnographic sources which can yield inform ation to us, regarding the effect o f the world religions on the M agyars prior to the Conquest.
7. THE ANTHROPOLOGICAL SOURCES The conquering M agyars buried their dead in graves deep in the ground where their bones w ere preserved. The question, therefore, clearly arises: w hat can these bones tell us? So-called physical anthropology has two m ain schools. The first one, the classical approach, takes the size and proportion o f the bones as its starting point. It categorises these, and using these categories attem pts to establish different types. It m easures, for instance, the length, the base and the width o f the skull, the forehead, the brain-case, the cheekbone, the orbital and the nasal bone. N orm ally fifteen m easuring points are set on the brain-case and sixteen on the face. A further m easurem ent is used w hich is a horizontal line linking a point below each ear to a point below the left orbital. These lengths and widths give the unique nature o f each skull. The figures are used to categorise the population o f each grave. The m easurem ents o f the early skulls found in Eurasia and those o f m odern m an serve to establish larger groups and sub groups as well as provide inform ation on the transition betw een each category. Physical anthropology considers the whole o f hum anity to be one species, as each group o f people is capable o f crossbreeding w ith any o f the other groups. A nthropology divides the hum an population o f the w orld into three or five groups: (1) Europid, (2) M ongoloid, (3) Negroid, (4) Veddo-Australoid, (5) Am ericanoid. The last two are thought to have derived from the first three. These are the m ain groups from each continent. The first three are also known as subspecies, or greater races. The word ‘race’ indicates the physical and bio logical differences within the species, and it is these differences w hich have been exploited and m isused by racism. The Europid and M ongoloid subspe cies are further divided into m icro-subspecies. It is on the dem arcation line o f the two subspecies that the Europo-M ongoloid group developed. This has Uralic, or Europo-Sibirid, Turano-South Siberian and Yenisei subgroups. With a lot o f crossbreeding betw een the groups the m igratory people, if set tled long enough in one area, could also develop local subgroups. The Euro-
154
Methodological introduction and the sources
pid subspecies includes, am ong others, the Pam irian, the D inaric and the A rm enid-N ear Eastern subgroups. The Asian subspecies is called M ongoloid on account o f the fact that w hen these groups w ere first recorded in the 13th-14th centuries these people were all under the rule o f the M ongols. The ethnic M ongols, however, are not a dom inant group in this subspecies. Unfortunately, in H ungarian anthropology there is a lot o f confusion regarding the nam es o f groups. The words describing the subspecies have the ending -id, while groups sim ilar to the subspecies, but not necessarily genetically related, attract the suffix -oid. This latter one, however, cannot be anthropologically defined. I shall use in accordance with international practice the suffix -oid, and the suffix -id I em ploy to describe other m icro-species. In any case, it cannot be claim ed that the M ongoloid subspecies is identical with the M ongol ethnicity. Provided that the historical im plications are clear, the term m ay be used, bearing in m ind that the m ajority o f M ongols are M ongoloids but the m ajority o f M ongoloids are not M ongols. Today, m ore than one billion Chinese inhabit this planet and the vast m ajority o f them are M ongoloid. From the exam ination o f the bones, the sex o f the person and the age at which he or she died can be ascertained. The latest research o f this kind exam ined the intact skulls o f the conquering M agyars from three generations. From seventy-one archaeological sites, three hundred and fifty-three adult skulls were com pared. The findings w ere divided into five geographical regions: (A) Pest county and northern B acs-K iskun county; (B) Szabolcs-Szatm ar-Bereg county and the south-w est parts o f Borsod-A bauj-Zem plen county; (C) Fejer and Heves counties, B acs-K iskun county’s tw o Danube areas, and som e finds from the northw est Hungarian plain (Kisalfold); (D) Bekes county; and (E) areas around G alanta and Dunaszerdahely (see Figure 31). Thorough research came up with the follow ing results: the exam ination o f the skulls determ ined that the conquering M agyars can be divided into two groups. The first group (I) includes the skulls from the aforem entioned groups A, B, and E. These are typically wide (when looking at the w idth in proportion to the length o f the skull) and have a high percentage, at least forty, o f the Europo-M ongoloid subtype. The second group (II), С and D, has narrow skulls and the Europo-M ongoloid type is no m ore than six per cent present in it. There is a clear geographical divide betw een the two groups. The border line between them runs along the left bank o f the Danube tow ards the north, then turns eastw ards at northern B udapest and runs till N agytarcsa and finally ends abeam o f A ldebro, Kal and Eger. The conquerors o f group С occupied the rolling country o f Transdanubia and the N orthern M ountains. G roup A occu pied the area betw een the Rivers Danube and Tisza, w hile group В the northern
The sources
155
Figure 31 Physical anthropological groups o f the conquering M agyars
Tisza areas, and group E occupied the green steppe region betw een the Rivers Vag and Nitra. If these areas o f occupancy are com pared to the Carpathian B asin’s geobotanical m aps at the age o f Conquest, it becom es apparent that the territories o f group I are those o f the (a) sand dunes and oak forests o f the lowlands, the juniper forests, the Pontus-sub-M editerranean-type deserts, the alkaline areas among the sand dunes; (b) the willow and poplar groves o f the river banks, the m ixed elm, oak and ash forests, the reedy swam ps, the alkaline oak forests and the moors. The territories o f group II typically include the (c) foresty steppe areas, the partially sub-M editerranean-type Tatar m aple-tom entose oak forest o f the Great Hungarian Plain (Alfold) with areas o f loess clearings; and (d) the closed m ature oak forests o f the Pannonian downland. Researchers claim that group II had sim ilar characteristics to the Iron Age people living to the w est o f the R iver Don, w hile group I could be linked to those living to the east o f the river. This becom es even m ore interesting when the geobotanical features o f these areas are compared, because w est o f the River Don was the area o f the (e) peduncular broad-leaved oak forests with a
156
M ethodological introduction and the sources
Figure 32 Vegetation o f the Carpathian B asin and the Etelkoz (reconstructed by B. Zolyom
The sources
157
L egend on p a g e 158
158
Methodological introduction and the sources
scattering o f hornbeam ; (f) the sparse maple and oak forests o f the M oldovaPodola loess ridge; and (g) the rich, green Eastern European steppe. It appears that group II w as socially inferior to group I because in two o f the burial areas classified as group II, skulls resem bling those o f group I w ere found in rich and lavish graves. There are no significant differences in the num ber o f m en and wom en, and there were no apparent shortages o f females in the first three generations which also gives an indication o f family life. The average m ale height was 168.83 centim etres (5'6Vi"), the female average 157.93 centim etres (5'2") w hich can be considered reasonably high. In comparison, in 1973 the average height o f
L e g e n d to Figure 32 1. W illow and poplar groves on river banks, mixed elm, oak and ash forests, reedy swamps, alkaline oak forests and moors = (b) 2. Sand dunes and oak forests o f the low landsjuniper forests, Pontus-sub-Mediterranean-type deserts, alkaline areas among sand dunes = (a) 3. Grassy loess ridges o f the plain with areas o f Tatar maple and oak forests and heather 4. Foresty steppe areas, partially sub-Mediterranean-type Tatar m aple-tomentose oak forest o f the Great Hungarian Plain with areas o f loess clearings = (c) 5. The Wallachian forest-steppe’s Tatar maple loess-oak 6. The m eadows o f the edge o f the plain with scatterings o f maple and oak trees with a cool continental climate 7. Sparse maple and oak forests o f the M oldova-Podolan loess ridge = (f) 8. The Eastern European zone o f peduncular broad-leaved oak forests with scattering o f hornbeam = (e) 9. Rich and green steppe = (g) 10. Low-grass steppe with Artemisia 11. The closed mature oak forests o f the Pannonian downland = (d) 12. Central European dry oak forests 13. Sub-Mediterranean tomentose oak forests and open ash forests with rich undergrowth 14. Turkey-oak and farnetto-oak matured forests o f the hilly Dacia and M oesia regions 15. The silver linden-peduncular oak-forests o f the Balkan Mountains 16. Sub-Mediterranean-Illyrian tomentose oak shrubberies, ash and yoke-elm m ixed forests, eastern hornbeam and its overgrown variants 17. Pontian-sub-Mediterranean mixed forests 18. Central European sheltered mature nonpeduncular hornbeam and oak forests with areas o f beech trees and peduncular hornbeam oak forests on the peripheries o f flood area 19. Central and South European mountainous sheltered beech forests, beech forests with fir trees o f an Illyrian variety in the south-west and a Carpathian Transylvanian variety in the north-east 20. European high-mountainous spruce forests, sub-Alpine yellow -pine forests and Alpine and sub-Alpine grassy areas 21. Central-European acidic-soil pine forests with m ixed oak forests 22. Northeast European fir forests with an increasing population o f beech trees and many oak trees
The sources
159
Hungarian conscripts was 171.15 centim etres {5'lVi"). Some o f the skulls had signs o f either real or sym bolic trepanation as an indication o f healing practices. It is im portant to note that a section o f the people o f the Avar era in Trans danubia, part o f southern Slovakia today, and in areas east o f the R iver Tisza survived to see the Conquest and assim ilated with the people o f the A rpad era. There are no signs o f significant num bers o f M ongoloid people surviving the Avar era. Due to the volum e o f m aterial and the lim itations o f the m ethodology, these pieces o f research can only be used in a cautious m anner to reconstruct the whole picture. Regarding the num bers o f the conquering M agyars the sources are silent or based on topoi. Some scholars claim that the population o f the conquering M agyars had to be larger than the population o f those then living in the Carpathian Basin, since otherwise the M agyars would have been unable to retain their language. Although this claim cannot be accepted in term s o f methodology, nevertheless, it m ay still be true, since population ratios are not considered in global terms. At the time o f the Conquest the Carpathian Basin was politically and ethnically heterogeneous. The sm aller groups fighting each other would have sought protection with the M agyars, and in doing so, joined them. The toponym s o f the 12th and 13 th centuries show that place nam es were rapidly Hungaricised. This was prim arily due to political and religious reasons, and was not related to ethnicity, nor the size o f the popula tion. N ot so long ago, in a workshop, the question o f the population size o f the conquering M agyars arose. The opinions were varied, but the figures were all around several hundred thousands. In relation to this, the 199 m ale and 154 female skeletons found constitute a very small num ber and cannot be used as a representative sample. Therefore, the conclusions drawn from these findings m ust be treated with great caution. D iscovered during the past few years, m any new findings are likely to change this picture. Nevertheless, we com pleted statistical studies with the sam ples available to us, and these provided satisfactory results. Figure 33 gives a good indication o f the ratio o f the different skull-types in each group. The average deviation o f the figures in each group do not show fundam ental differences. It is notable that the typology o f the male and female skulls are identical w ithin each group which suggests that there was no significant crossbreeding betw een them. A longside the classical anthropological research o f m easuring bone proportions, attempts have been made, as early as the beginning o f the 20th century, to introduce biological considerations into the research. At the beginning, the ratio o f people with blood groups А, В, AB and 0 were ob served, and later groups A1 and A2 were brought into the research. The results showed that the proportion o f people o f each blood group were sig-
Methodological introduction and the sources
160
30т-------
20 -
10-
x 64.0
68.0 66.0
72.0 70.0
76.0 74.0
80.0 78.0
84.0 82.0
88.0 86.0
92.0 90.0
94.0
С and D groups (N=194). Average 77.9. Standard deviation 4.73.
40 -
30 -
20
10
-
-
64.0
68.0 66.0
72.0 70.0
76.0 74.0
80.0 78.0
84.0 82.0
88.0 86.0
92.0 90.0
94.0
A, В and E groups (N=121). Average 81.7. Standard deviation 4.78. Figure 33 The skull m easurem ents o f the two groups o f the conquering M agyars (w idest s k u ll: longest skull)
nificantly different on each continent. For instance, the num ber o f people in Europe with blood group A decreased towards the east, w hile the num ber o f those with group В increased. N ew factors w ere introduced into this research, such as the Rh factor, and taking into consideration several different aspects, so-called im m unological races w ere constructed. This, how ever, did not produce results that could help in constructing a historical picture, as for instance, it becam e clear that epidem ics claim ed m ore victim s from certain
The sources
161
im m unological groups than from others, therefore the decline o f a particular im munological group o f people did not necessarily entail a change in the general population. It was also frustrating that the m utations o f the serological factors did not provide any historical inform ation to build on. Finally, the fact that the hypothesis could only be tested on the living population m eant that the figures could not be applied to a historical study. The ratio o f the different im munological groups could have evolved the way it did for m any different reasons. M odem research cannot differentiate betw een the groups which expanded quickly and those which took considerable tim e to build up. Im m une-biology and, within this field, the rapidly expanding research o f im muno-globulin have been used in attem pts to answ er historical questions. From the five existing im m uno-globulins in the hum an body, one, the so-called gam m a-globulin (Gm) was examined. Every such im m uno-globulin consists o f a light and a heavy chain, and one part o f these protein chains is constant. This constant section contains chem ically very different constituents which determ ine the chain. These are known as the determ inants. The chem i cal characteristics differentiate the particles and are know n as m arkers. M any o f these m arkers have been identified and their hereditary nature determ ined. It has been found that m arkers are not inherited independently o f each other, but in groups. The genes w hich determ ine each block are inherited on the 14th chromosom e in a co-dom inant way, that is to say that the m other’s or the father’s genes are not dom inant but every inherited parental gene will appear in the offspring. The Gm -m arkers are directly inherited and do not change over a life span. However, there are genetic m utations w hich m ay lead to new markers evolving; and in the build-up o f the chrom osom es there can also be changes which m ay cause irregularities in the inheritance o f the Gm -system . Scientists claim that the probability for this is very low ( K H -1 0 -6) The new m arker system is indicated with num bers, and those m arkers which are regularly inherited together (haplo-types) are placed in brackets. Based on this system , the European race is determ ined by the following groups: (1, 17, 21, 26), (1, 2, 17, 21, 26), (3, 5, 10, 11, 13, 14, 26), while the M ongoloid subspecies by (1, 1 7 ,2 1 ,2 6 ), (1 ,2 ,1 7 ,2 1 ,2 6 ) , (1 ,1 0 ,1 1 , 1 3 ,1 7 ) , ( 1 , 3 ,5 , 10, 11, 13, 14, 26). It is apparent that the num ber ‘ 1 ’ m arker does not appear in one o f the haplo-types in the European group, w hile it does appear in every M ongoloid group. In any European group that contains 1, 21 will also be present, while the same cannot be said for all M ongoloid groups. These combinations indicate a different hereditary order. Using these orders and the num ber o f recurrences, the frequency could be graphically illustrated. Studies conducted in Hungary show that 95% o f the Hungarian people belong to the European group, and a com parative study o f 28 other European populations found that the deviation rem ained under 5%. This 5% is M ongoloid but
16 2
M ethodological introduction and the sources
one h a lf is from the northern, the other from the southern subgroup o f the M ongoloid subspecies. To get an idea o f the recurrence, it is worth looking at the frequency o f the northern M ongoloid group (1, 13, 15, 16). The frequency o f this in the Hungarian study is 0.0169, while am ong the M ongolian northern Buryat people it is 0.307 and am ong the southern B uryat people 0.272, and in the M ongolian Republic it is 0.140. This com ponent, even in South Asia is twice as high as the Hungarian; and the low com ponent am ong the M agyars is only m atched by the Peruvian Kechua people (0.016) and figures from Thailand (0.015). It is not surprising that the M agyars are biologically identical to the other European peoples, and neither is it unexpected that there are two com ponents which have Asian traces. These figures, however, show little m ore than this, and it rem ains uncertain how this ratio developed. It is, for instance, quite possible that the two M ongoloid com ponents originated from different periods (e.g. from the Tatar and the Turkish occupation o f H ungary in the 13th and the 16th-17th centuries respectively); or they m ay have originated from two different groups altogether (such as the A natolian and Crim ean population during the age o f Ottom an occupation). However, it is equally possible that a section o f H ungary’s Gypsy population brought these com ponents along with them. We cannot tell w hether these com ponents built-up gradually or whether they decreased to the levels observed today. No m ethod currently offers an answ er to these questions and, therefore, the results o f these pieces o f research cannot be used for the purpose o f historical reconstruction. N evertheless, it is im portant to m ention these studies, as hopefully the tim e will soon com e when certain biochem ical m aterials can be extracted from bones w hich will provide inform ation for historical analysis, some substance that will reveal traces o f history. W hile the research o f the gam m a-globulin m arkers was based on hereditary characteristics, population genetics examines the different genetically inher ited features am ong larger populations. One advantage o f w ell-organised and conducted m ulti-factor research is that the percentage deviation o f the con stituent tests will balance each other out. 15-20 genetically inherited char acteristics m ay give a correct ratio for each group o f people. The ratio of occurrence, in for instance a group o f 1 00,0 0 0 people, can be linked to indi vidual groups o f people, and com parisons can then be m ade betw een these groups. The inform ation can be arranged in such an order that the area between identity and the largest difference is divided into four. The term s “identical”, “short”, “m edium ”, “long” and “very long” “genetic distances” are derived from this. This does not refer to a general genetic distance but to the num erical
The sources
163
difference between the exam ined frequency o f occurrences o f genetically inherited characteristics. The genetic ratios o f the Hungarian ethnic groups do not show any distinc tive differences. Com parisons with foreign samples, however, do. For in stance, the difference betw een the Hungarian figures and those o f the M ongoloid subspecies, for certain genetic characteristics, are as high as 14, while the difference between the M agyar and Slav people, and the M agyars and Germ ans was 1.1 and 1.2 respectively. This shows that the genetic stock o f the M agyars is certainly identical with other European people. Looking at even closer correlations, however, cannot be considered as a reliable source o f evidence. For instance, com parison o f the Hungarian, Turkic, Finnish or Iranian figures reveals their relation to each other (w hether they are close or not) is not one that can be evaluated in a historical sense. This is simply because the research based on the frequency and ratio o f genetic groups o f the living population cannot give any indication o f the reasons they developed the way they did. It has to be understood that behind the sim ilar figures an identical history may stand, but equally, there may be num erous other factors such as: 1. Two different groups m ay have com e under the influence o f a third one, and thus their originally different ratios m ay have evened out. 2. Two originally com pletely different groups m ay have come into contact with each other with an effect o f the different ratios balancing out. 3. Two originally com pletely different groups m ay have becom e sim ilar as a result o f epidem ics, biological effects or the need to adjust to a certain environment. For the same reasons, two originally sim ilar groups my have becom e quite different in their genetic ratio. Finally, the two processes, o f convergence and divergence, could have had num erous other complex variations. It is also im portant to m ention the difficulties behind the sam pling process. For instance, in the case o f the Hungarian ethnic groups, it is often hard to define who belongs to which group and where the dem ographic boundaries o f these groups are. It is also o f some concern that sim ilar genetic research has not been carried out am ong the different people o f H ungary’s neighbouring countries (or else, the m ethodology o f sam pling was different). The im portance o f population genetic research is in its ability to pose new questions w hich can m ark out new areas o f research. It reveals facts for which population genetics, in its own right, cannot provide adequate interpretation. It cannot shed light on historical processes, itself being built on historical hypotheses.
M ethodological introduction and the sources
164
NOTES For the origins o f the De adm inistrando im perio, which are a lot more com plex than described above, see M oravcsik (1983, v o l.l, pp. 361-367); Jenkins (1962, pp. 1-8). The description o f the Byzantine sources is based on the relevant chapters o f HajdiiKrist6 - R 6 na-Tas (1976, vol. 1/2),. which are the work o f S. Szadeczky-Kardoss, F. Makk and T. Olajos (1976). There are further references made in their work. Each Greek author is also m entioned in the first volum e o f M oravcsik (1983). Sources on Leo the W ise and subsequent ones referring to the Magyars can be found in the preface o f M oravcsik (1984). The Hungarian translations o f the Byzantine sources relating to the Avars and the commentaries on them were first published by Szideczky-K ardoss in A rchaeoldgiai E rtesito between the years 1978 and 1986, and later with a few additions they were com piled as a book: Szadeczky-Kardoss (1992). It can only be hoped that further volum es will follow. Further information on the developm ent o f the research o f these authors can also be found in the continually updated P auly..IVissowa E ncyclopaedia. Another useful source o f informa tion is the Z ie g le r-S o n th e im er E ncyclopaedia or the “short Pauly” which was originally published in five volum es between 1964 and 1975. The m ost accessible version o f it being the paperback edition published in 1979, in Munich. N ow the basic work to be consulted is The O xford D ictionary o f B yzantium o f Kazhdan-Talbot (1991), the reference to which I thank Szideczky-K ardoss. Mar6 ti and Lakatos wrote the chapters on the Latin sources in H ajdii-K risto-Rona-Tas (1976, vol. 1/2). The western mediaeval Latin sources were presented by K aricsonyi and Szegfii (1976) and the Latin sources in Hungary by Kristo (1976). The Kristo (1994c) L atin sources section was written by Laszlo Szegfii. The western Latin sources were selected by Gombos (1 9 3 7 -1 9 4 3 ), based on the then available German publications. The G om bos collection is very useful in locating Latin quotes, however, it does not satisfy the requirements o f a critical edition— which it does not intend to be. Where I refer to a Hungarian translation, I do that based on GyCrffy (1975c). The Latin sources in G yorffy (1975c) were translated by J&nos Horvath. The prefaces for each translation are useful references. In many cases I used these references as source material. I also discussed the matter o f “Attila’s sword” which was noted by Lampert o f Hersfeld in Rona-Tas (1988a, pp. 112-114). In this case I used the German critical edition (Holder-Egger 1894). That is also where I discussed the related earlier Hungarian literature. The deed o f foundation o f the Krems Abbey was published by Hagn (1852). Louis o f Germ any’s deed was dealt with by Olajos (1969). For the letter written to the Bishop o f Verdun, see pp. 2 8 2 -2 8 3 . The deeds o f the Arp^d period were published by GyOrffy (1997). The Latin sources written in Hungary were published by Szentpdtery (1 9 3 7 -1 9 3 8 ). The analysis o f the Hungarian chronicles was published by Krist6 in Hungarian in 1980 and 1994b, and in English in 1996. For GyOrffy’s latest view s see G yorffy (1993b) which also includes the reprint o f his work first published in 1948. The 13th-century Hungarian chronicler Anonym us has a very useful facsim ile edition, with
a translation
by Pais, including P ais’s and G yorffy’s notes (see Pais
1975b). The Anonym us research was summarised in Csapodi (1978), G yorffy’s view s in G yorffy (1988). The Anonym us article in Kristo (1994c) was written by Zoltan Korde. For a more recent summary on the Hungarian A nonym us research see also Thoroczkay (1 9 9 4 -1 9 9 5 ). For other and earlier bibliographies, see Kosary (1951). The introduction to the Slavic sources in H ajdu-K risto-Rona-Tas (1976, vol. 1/2) was written by Istvan Ferincz. The Hungarian problems in connection with the Cyril and Methodius legend were discussed in Kiraly (1974) and in H. Toth (1981). The most recent study is Eggers (1996). The first Hungarian translation o f the primary Russian chronicle is by Hodinka (1916).
The sources
165
Today, the m ulti-volum e Russian chronicles are available, known by their abbreviation PVL (P o vest’ vrem ennyh let). Many have taken part in its edition since 1951, led by Lihachov. The article about it in Kristo (1994c) was written by Marta Font, The critical edition and English translation o f the Laurentian version is by Cross-Sherbowitz-W etzar (1953). The M iddle Persian sources are absent from Hajdii-K risto-Rona-Tas (1976, vol. 1/2) and Kristo (1994c). Janos Harmatta’s summary was published only later, in 1993. Several aspects o f the sources in M iddle Persian were discussed in studies written by C zegledy (see his reprint volume, Czegledy 1985) and the English translation o f his book C zegledy (1983a). The M uslim sources were summarised in Hajdu-K risto-Rona-Tas (1976, vol. 1/2) by Tamas Ivanyi, integrating Mihaly K m osko’s unpublished manuscripts. See now also Zimonyi (1991) and K mosk 6 (1997). The summary to Kristo (1994c) and the articles on each author ( Ibn Rusta, Ibn Fadlan, Gardizi, H u d u d a l-alam , al-Bakri, Marvazi, Jayhani, Idrisi, Yakut, Ibn Hayyan, Abu-Hamid al-Gamati) were written by Istvan Zimonyi. G ood summaries are: Sezgin (1967) and Miquel (1973). The 5th volum e o f Ibn H ayyan’s book M uqtabas was published by Chalmeta (1979), see also Chalmeta (1976). The Arab historical literature o f Spain is examined in Bojko (1977). Individual authors are also discussed in the editions o f Gibb et al. (1 9 6 0 -1 9 9 3 ) published in English. Important com ments on the M uslim sources o f ancient Hungarian history, including the much discussed Jayhani question, can be found in Zim onyi (1990). For Ibn Fadlan’s tw o important editions see Togan (1939), Kovalevskij (1956)— the latter is the second publication o f the author’s work. (Published in 1939 by Krachkovskij, the first publication appeared without mention o f the authors name, for political reasons.) The photograph taken by Ligeti o f the Mashad manuscript was published by Czegledy (1 9 5 0 -1 9 5 1 ), see also K m ietovicz et al. (1985). The Syrian sources were introduced by Tamas Ivanyi, based on Mihaly K m osko’s unpub lished works. K m osk 6 ’s original research area was Old Syriac. Further information on him can be found in Kristo (1994c, p. 645). The Armenian sources were discussed in Hajdii-K risto-Rona-Tas (1976, vol. 1/2) by Odon Schutz, see Schiitz (1995). A bibliography o f the works o f Schiitz can be found in volum e 50 o f A cta O rientalia H ungarica (1997), pp. V II-X . The Syrian and Armenian sources referring to the Khazar people are quoted by Marquart (1903), Dunlop (1954/1967), Golden (1980) and Ludwig (1982). The Georgian sources in H ajdii-K risto-Rona-Tas (1976, vol. 1/2) were summarised by Margit Bir 6 . I wrote the com ments on the Turkic and Tibetan sources in Hajdii-K risto-Rona-Tas (1976, vol. 1/2). I discuss the two groups o f sources in more detail in two studies (R 6 na-Tas 1996d, 1996e) which were published in the series H onfoglalasrol s o k sz e m m e l [Different view s on the Conquest] in the volum e about the sources. The summary o fth e C hinese sources in H ajdii-K risto-Rona-Tas (1976, vol. 1/2) was written by Ildiko Ecsedy. A very useful multi-lingual reference book on several transcriptions o f Chinese words, including the p in yin transcription, is C songor-Ferenczy (1993). Questions regarding Chinese transcriptions can be found in Rona-Tas (1985b, pp. 3 0 5 -3 4 7 ). The two basic works for reconstruction o f Chinese are Karlgren (1957) and the Pulleyblank (1991) lexicon. H ajdii-K risto-Rona-Tas (1976, vol. 1/2) did not include the Hebrew sources. Follow ing the completion o f the Hungarian original o f this book, two significant works were published which need to be mentioned here. The first is titled A honfoglalas koranak irott fo rra s a i [The written sources o f the Conquest period], edited by Gyula Kristo (1995d), the second book, edited by Laszlo KovacS and Laszlo Veszprdmy (1996) is a compilation o f works
166
M ethodological introduction and the sources
entitled A h onfoglalaskor irott fo rra s a i [Written sources from the age o f Conquest], It is likely that the two works, which are very different in their contents and style, will nevertheless cause much confusion with their near-identical titles. Kristo (1995d) contains the translations o f the sources referring to the Magyars. The editor split the subject-matter into five chapters: M uslim sources; sources written in Greek; Slavic sources, sources written in Latin and in Hungarian. The sources in Hungarian are represented by the Song on the Conquest o f Pannonia, presumably written by Demeter Cs&ti in the 16th century, and another fragment. The translations o f the individual sources are each prefaced. The anthology is considerably more detailed than Gyorffy (1 975c), from which som e translations were taken. However, in m ost cases the translations are new or reflect new research results. The notes are very much different in character and volum e and occasionally require substantive evidence, in other cases they contain philological com ments. The compilation o f the Hungarian translation o f the sources referring to the Magyars is useful as a genre and can serve as a quick reference, but has its dow nsides. B y rem oving a text from the greater context, without the know ledge o f the author’s regular vocabulary and the structure o f the text, the university student and the lay person, for whom the book was w ritten, may com e to incorrect conclusions. The K ovacs-V eszprcm y (1996) volum e contains the lecture transcripts, with additional literature, given by researchers at a conference in D ecem ber 1993, organised by the Hungarian Academ y o f Scien ce’s Com mittee for Magyar Proto-History. Practically every single article o f this book contains new material on the chosen source subject. It is unfortunate that this collection contains the older translation o f the K 6 I Tegin inscription without taking into account the results o f the latest research, as in the light o f these works many sections o f the old translation are obsolete. The follow ing sentence in the preface which claim s that “Turkic writing was introduced into western A sia (Yenisei and Talas Valley inscriptions) and into the Khazar Empire by the West Turkic people, which was how the Magyars got to know it, and among whom it remained to be in use until the 16th century, known as ‘Szekely runiform script” ', contains geographical, substantive and chronological errors. However, the studies in that book make invaluable new observations accessible in Hungarian and several o f the articles have bibliographies referring to new Hungarian translations o f source materials. I have discussed the question o f language as source material in many o f my previous works. An overall informative guide is Rona-Tas (1978a) on linguistic relationships. The revision o f Gyula N em eth’s work on the ancient relationship between the Uralic languages and the Turkic languages is published in Rona-Tas (1 983b). The relationships between the Uralic, Finno-Ugric and Turkic languages are discussed in Rona-Tas (1988d). The Irano-Finno-U grian and the Irano-Hungarian linguistic relations are discussed in two monographs, Joki (1973) and Rddei (1986), and are also discussed in a chapter o f Redei (1988). A good, although som ewhat inconsistent, account o f the Irano-Hungarian linguistic relations is Harmatta (1977). This article has no bibliography. The Ob-Ugrian languages’ Iranian loan words are discussed by Korenchy ( 1976). This work also looks at the words that made their way into Hungarian and has a good, albeit by now slightly outdated, bibliography. The relevant chapter on the Iranian languages o f Hajdu-K risto-Rona-Tas (1976, vol. 1/2) was also written by Eva Korenchy (1976). Pages 101-103 o f that book include an overview o f the history o f research, earlier literature, and a bibliography can also be found there. The earlier H ungaro-Perm ic connection has been overem phasised by Elemer Moor, who suggests that even the name M agyar com es from this connection. He claim s that the m a n s word had a second component, se r or seri, added to it and, thus, by losing the sound Is/ it becam e er or eri. The word ser, according to M o 6 r (1953), lived on in the ethnonym Zyryan. Pais (1953) claim s that the Nydk tribe represented the Permic connection. The critical analysis o f the previous literature was done by Redei (1964, 1969), which closed the debate for som e time.
167
The sources
His basic claim was that, although a loose connection could be detected, there is no result o f major mutual influence handed down from this relationship. The question has now arisen again and warrants a new examination. The Hungarian loan words taken from Turkic were treated by Zsuzsa Kakuk in Benko (1 9 6 7 -1 9 8 4 ) and revised by Ligeti. 1 was in charge o f the words o f Turkic origin in Benko (1 9 9 2 -1 9 9 5 ). Ligeti’s earlier articles are also o f great value as is Ligeti (1986). I discuss the Turkic etymology, found in this book, in a monograph about the Turkic loan words in Hungarian (to be published). I wrote on the guttural word endings in R 6 na-Tas (1997d). I dealt with the problems o f rotacism and lambdacism in depth, with the inclusion o f all available material, in my D Sc dissertation (Rona-Tas 1970), and in several articles (R 6 na-Tas 1973, 1974a, 1980b, 1981, 1982d, 1988d) and in the introduction to my Chuvash text book (Rona-Tas 1978b). On the criteria o f the Chuvash loan words see Rona-Tas (1978b) and Ligeti (1986). The monograph o f written Greek material on the Danube Bulghar linguistic history is by BeSevliev (1963, 1979). The monograph o f the list o f the Danube Bulghar rulers, with conclusions difficult to accept, is by Pritsak (1 9 5 3 -1 9 5 4 ). The Volga Bulghar inscriptions were published by Jusupov (1960), R 6 na-Tas-Fodor (1973), Hakimzyanov (1978), Tekin (1988), Erdal (1993b); the historic sources o f the Chuvash and the Chuvash-type languages are discussed in R 6 na-Tas (1982d), the word terem and the name Term ecsu are discussed in Rona-Tas (1 996g). In a recent publication Benko (1998) has expressed doubts on the etymology. His main argument is built on the fact that in recent Hungarian the second vow el o f terem is a closed and not an open e, and closed e vow els o f Old Hungarian are not transcribed by Greek alpha. Since in the original, be it Turkic or Slavic, there was no closed e, this argument is not conclusive. The present Hungarian form is due to a longer adaptation to the structure o f the Hungarian language. The references on archaeology in this book are largely based on the archaeological chapters o f H ajdii-K risto-Rona-Tas (1976, vol. 1/2), in particular the work o f Csanad Balint. The archaeological sources up to the period o f the 4th century
ad
were summarised by Bela Kurti.
Csanad Balint summarised the eastern relations, the archaeology o f the Carpathian Basin and that o f the conquering Magyars. A separate volum e had been published with the pictures o f the objects found (Balint 1989b). The most successful discussion on the early archaeological findings o f the ancient history o f the Finno-Ugrian people is by Fodor (1973) and the best known one Fodor (1975), in English Fodor (1 982b), in German Fodor (1 982a). Archaeological findings linked to the Magyars outside the Carpathian Basin are discussed in Fodor (1994). I discuss the question o f the Volga Magyars and the Volga Bulghar archaeology in detail in my assessm ent o f Istvan Fodor’s dissertation (Rona-Tas 1978-1980, published in 1988, see 1988e). Halikova published several studies on the B olshie Tigani finds (H alikova 1976, H alikova-K azakov 1977, Halikova 1978, H alikova-H alikov 1981). See also Fodor (1994). I also wrote about the funeral shroud in relation to the Fodor-dissertation (Rona-Tas 1988e, see also M. Benko 1992/1993). Important studies written on Avar and Slav archaeology and the history o f the Pannonian Slavs are by Istvan Bona. Two overview s can be found in Bona (1984a, 1994a, see also Bona 197 1 ,1 990e). On the Avar archaeology’s eastern links see Erdelyi (1982). The best monograph on the Avars is Pohl (1988; see also my review in Rona-Tas 1991c). On the archaeological excavations at Zalavar and the Pannonian-Slav settlements o f Pribina-Kotsel see Cs. Sos (1973). The needle-casc o f Szarvas was found by Iren Juhasz and published in Juhasz (1983, 1985). The latest publication on the inscriptions on the N agyszentm ikl 6 s Treasure and on the questions o f Eastern European runiform writing is by GObl-Rona-Tas ( 1995). The thorough bibliography
168
M ethodological introduction and the sources
is the work o f Maria Ivanics. On the Hungarian words ir 4vrite’ and betii ‘letter (=character)’ see Rona-Tas (1992a), in the volum e on the Hungarian runiform writing, see Sandor (1992b). The chapters on Magyar ethnography in Hajdu-K risto-Rona-Tas (1976, vol. 1/2) were written by Laszlo K osa and Imre Katona. Questions on agriculture, settlement types, livestockfarming, threshing and sow ing are summarised by Kosa. Social ethnography was discussed by Katona. M y references on the extended family units are from my unpublished M A thesis on the extended fam ilies o f Orhalom (Northeast Hungary). For M ongolian ethnographic material see Rona-Tas (1959, 1961a, 1961b, 1963, 1972, 1980a). On Tengrism see R o u x (1 9 5 6 , 1958, 1962) and Rona-Tas (1987a). The data on the Khirghiz kut is by Judahin (1965, p. 452). On the smaller Kharabalghasun inscription, see Sinehiiij (1980, pp. 42^13). TheNorthern Zhou-dynasty’s annals were reported by Liu (1958, p. 8 ). M asudi’s report quoted here does not appear in the M asudi-translation o f Gyorffy (1975c), nor in Kristo (1995d). It is discussed, however, by Ludwig (1982). Ibn Fadlan’s report was published in Togan (1939) and K ovalevskij (1956), see p. 165 above. M ovses Dashuranci’s description can be found in Ludwig ( 1982, p. 310). Shamanism is discussed in Johansen (1987) and Hultkrantz (1993). Both articles contain further bibliographies. The chapter on anthropology in Hajdii-K risto-Rona-Tas (1976, vol. 1/2) was written by Pal Liptak. It can be found in more detail in Liptak (1971). The categorising o f the latest results into A -E groups and the maps are by Ery (1994, see also Ery 1983). The data were statistically verified and the histogram was created by A kos Rona-Tas. The source for the gamma-globulin markers and the latest results on the genetic research o f the Magyars can be found in Tauszik (1990a, 1990b), Czeizel (1990) and the review by Rona-Tas (1990b).
PART T W O
RELATIVES A N D N E I G H B O U R S
111. THE RELATIVES
1. LINGUISTIC RELATIONSHIP Unlike our neighbours and friends, we cannot choose our relatives. Our approach to the ancient affinities o f the Hungarian language m ust therefore be unclouded by em otional judgem ents or prejudice. We m ust be clear, however, w hat linguistic kinship is, or rather what is m eant by the term in this book. Am ong people, there are two kinds o f relatives: those linked by descent, and those acquired by m arriage, where the bonds are not o f blood. Applied to language, the term will herein have the first sense. The concept o f linguistic relationship goes back a long way; it is the real substance o f the Tower o f Babel story. The first use o f the m odem scientific concept o f linguistic relationship is norm ally attributed to a lecture given to the Royal Asian Society on 2nd February 1786 by Sir W illiam Jones, in which he propounded the m odem theory o f the relationships am ong Indo-European languages. In his Origin o f Species, published in 1859, Darwin used the relationships o f languages to illustrate the developm ent o f species. Extinct languages assum e the same position as extinct species. And although we do indeed talk o f genetic relationships betw een languages, and therefore apply a biological parallel, the theory o f linguistic relationship is not, however, the product o f Darwinism. The m ultitude o f the w orld’s languages can be ordered in three different ways. First, languages can be grouped geographically. This is som etim es un avoidable, and in the Caucasus, for example, where there are so m any different languages, it is the only way. However, it can never be the basis for an unam biguous classification, because taking the same example, nobody can say where the borders o f the Caucasus are, nor can the R ussian spoken in the Caucasus be easily accounted for. The second is typological classification. In contrast to the languages o f its neighbours, m ost o f the gram m ar o f Hungarian is built from constructions— “agglutination”— o f affixes, suffixes, and word stems. Languages can be classified typologically in different ways. One scheme groups them into “agglutinative” , “synthetic” (inflected, like the Indo-European languages) and “isolating” (analytic, like old Chinese) types. Other divisions m ay be made according to the com pulsory order o f subject,
172
Relatives an d neighbours
predicate and object, or w hether the logical subject o f the sentence consists o f a subject case or has an instrum ental com plem ent. These typological com pari sons shed m uch light on the w orkings o f a language. They are also com plete, in that all languages can be assigned to categories in any o f the schem es, but they are not unam biguous, because one language can belong to tw o or m ore different typological groups. This typological sim ilarity is often m istaken for kinship. Sim ilar languages are now adays considered to be related from the m istaken inference that “since relatives look sim ilar, sim ilarity im plies relat edness” . But since they are always changing, languages also change their type. English is losing its last gram m atical inflections, and the role o f its gram m ati cal w ord order is getting closer to that o f Chinese. In the follow ing two sentences, it is clear only from the word order w ho kills whom: The lion kills the snake. The snake kills the lion. In Hungarian, the subject and object are distinguished by suffixes, and in the equivalent o f the first sentence, the places o f the lion and the snake w ithin the sentence can be reversed w ithout any benefit to the snake. This m eans that the order o f the parts o f speech can be freely changed w ithout changing the m eaning. Genetic relationships o f languages, i.e. grouping by linguistic family, is com plete, because every language belongs to som e family, although there are nowadays som e languages w ithout relatives. It is unam biguous, because one language cannot belong to m ore than one family, and not arbitrary, because it can be determ ined uniquely from each language w hich fam ily it belongs to. O f course linguistic relationship is m ore than ju st a m eans o f grouping languages. Two or m ore languages are said to be related if their principal subsystem s can be historically traced to a com m on ancestor language. Sub system s include basic vocabulary, phonem ic structure, m orphology, and gram m atical-functional relationships. Tracing im plies the ability to state the rules describing how the proto-language diverged and how the related languages em erged from it. Related languages m ust be linked not by individual words w hich look and sound the same, but by consistent correspondences. There are m any different kinds o f consistent correspondences, and only one o f these indicates that the languages are actually related. Loan w ords also have a system , and have regular correspondences w ith the language they are bor row ed from, but these are not evidence o f a basic relationship. Sim ilarities betw een languages can also have several other causes. Very com m on are “elem entary” w ords, such as onom atopoeic words. Birds are nam ed onom atopoeically in m ost languages, so that the cuckoo has been given a sim ilar or identical nam e in several languages, com pletely inde pendently. Such elem entary sim ilarity is typical o f infant w ords. In nearly all
The relatives
173
the languages o f the world, children’s first verbal interchanges w ith their parents are through phonetic forms like papa, m ama, dada, etc., w ithout these languages being related. In m any cases, the reason for sim ilarity is sim ply coincidence. There is no relationship betw een the Chinese and Hungarian words for wom an, which are ‘ям’ and ‘n o ’ respectively, nor betw een the English w ord ‘cu t’ and the sim ilar-sounding word o f the sam e m eaning in Arabic (see also p. 298). Because o f the sm all num ber o f sounds and the restricted possible connections am ong them, there will always be such coincidental correspondences between languages. It is surprising there are not more. However, these never fit into a larger system. Sim ilarity can also be due to convergent changes. The w ord m eaning ‘b o y ’ in Rom anian sounds alm ost the same as in Hungarian (/гм), but there is nothing in com m on betw een their respective origins. One derives from the Latin filiu s and the other from the Finno-Ugrian *poj, w hich w ent through changes re sulting in accidental similarity. W here languages coexist for extended periods, secondary correspondences arise which are not inherited from a shared origin. There are several such phenom ena am ong B alkan languages, and some in the Volga region. For example, in all o f the Volga region’s languages, the illabial /а/ has transform ed into the labial /а/, form ed with rounded lips, identical to the M agyar short /а/, but at varying times. Although such regional correspondences can becom e very influential after long times o f proxim ity, they do not result in a new language. The significance o f this to the history o f Hungarian is that, as we have seen, the language m ust have had intensive exposure first to Iranian, and then later to Turkic languages. These w ould have resulted in both borrowings and m utual changes, some o f which can actually be dem onstrated, but did not lead to the evolution o f a new language. The language o f the speakers was not replaced, it only changed. The system o f sim ilarities, correspondences and deviations betw een the languages can be analysed using increasingly refined scientific techniques. The U ralic origin o f the Hungarian language can be unam biguously proven from the history o f the most basic vocabulary elem ents, gram m atical signs and the m ajor points o f grammar. But it is not the origin itself w hich is in teresting so m uch as the question o f what linguistic features have evolved as Hungarian has passed through history as a separate language. A ncient inherited forms and external elem ents can quite easily be distinguished, but this is not so in the case o f internal transform ation o f w hat has been inherited. There are several Hungarian suffixes w hich are not part o f the m odem language, but whose existence in older versions o f the language can be dem onstrated. Some o f these definitely em erged during the language’s sepa-
174
Relatives and neighbours
rate existence. In m any cases their function is not even clear, there being no apparent difference betw een stems and derivatives. This internal course o f change w ithin the Hungarian language m ust have been particularly influential in the earliest period o f A ncient Hungarian. How is this m ajor change to be assessed? Did it depend on the length o f the period, on the effect o f— or isolation from— external influences, or a com bination o f these? The answer to these questions can be sought through reconstruction o f the history o f speakers o f the Hungarian language. The links between linguistic and ethnic relationships has already been discussed (see pp. 5 -12), and so we can proceed to the links betw een the Uralic languages.
2. THE URALIC LANGUAGES A N D PEOPLES
a) The pro co-language The term Uralic indicates the com bination o f the Finno-U grian and the Sam oyedic language family, and proto-U ralic their com m on reconstructed ancestor. In some cases the term Finno-U grian is used instead o f Uralic, following the older custom. Thus the five-yearly International Finno-U grian Congresses are really concerned with the study o f the Uralic peoples, includ ing Samoyedic. Speakers o f Uralic languages do not live very far apart from each other. The greatest separation is betw een the speakers o f the Hungarian and Finnish languages. It is thus to be expected that Uralic languages in close proxim ity to each other will contain many typological correspondences arising from areal or secondary effects, since these can occur ju st as well betw een related languages as betw een those that are unrelated. They are im portant inasm uch that secondary correspondences have to be elim inated in order to enable reconstruction o f linguistic affinities. For example, m any languages— includ ing old Indo-European— have a “dual”, i.e. a third category additional to singular and plural. This dual exists in the Ob-U grian languages, but not in H ungarian. Did it drop out o f Hungarian, or was it introduced secondarily into O b-U grian languages? The fact that the dual can also be reconstructed for the Sam oyedic languages suggests that this is an old areal phenom enon, and although there are also traces o f it in Lapp, it cannot be assum ed that dual existed in the ancient Uralic languages, because the Lapp speakers m igrated from the region where this areal dual phenom enon existed.
175
The relatives
Вa r en l s $ ea 4 ‘! I t i ",■
5. f c ! » \
c4
V'irttfa-
,л. W/ I
/
' Z«*efWv i -prqlsi
I"
Ч , ' I 'I
*1 'I Ifi «. f ’ ’ 1
Л'Г
J /'
LakeBalkhash
<>x ,
Ural neolithic culture Kelteminar neolithic culture
Figure 34 The Ural and the K eltem inar cultures in the 4 th -2 n d m illennia BC
The proto-U ralic, sim ilarly to other proto-languages, was certainly not com pletely hom ogeneous. Som e differences betw een later separate languages could have already existed in basic-language dialects and carried over into languages as they split off. Geographical areas containing identical linguistic phenom ena are delineated by lines called isoglosses. Some o f these isoglosses encompass dialects that later becam e separate languages, but there are some
176
Relatives and neighbours
B arents
$ea
Л 'К а га Г )М *
Ф
Ч 1 ,
'%
Ч-ч
l,
"" й ^ м Г " - - '. . x
_
Vogul j
'
‘
IJ R A L U IIH E IM A T
^
Chezenjis
v
~
^ъ
-■
^
H u n g a r ia n
<1
’
B lack Sea
-ч
л' iП’
Baltic Finnic (Finnish . . . and. Estonian) 2 [ _ j Lapp |— I Volga Finnic J 1— I (Mordvin, Cheremis) л (TJTTJ Permian (Zyryan, И1Ш Votyak) с I I Ugrian (Vogul, Ostyak, 1— I Magyar)
D on/
Cfaspiytt^ q.Atke-Sf
Uralic languages Finno-Ugrian
д
/fjral
j
'
U
MllQIi o r d v in
**»/ /""^4. \
i \
LUKCN \n }f
;y I__ j Sam oyedic
Figure 35 The Ural Urheimat and the distribution o f Uralic languages today
phenom ena in the proto-language whose continuations can be detected in m ore than one derivative language. There are two possible reasons for the absence o f certain linguistic phenom ena in some U ralic languages and their existence in others. The absences m ay be due to their disappearance in the m eantim e, but it m ay also be that they w ere m issing from the dialects o f the original proto-language. In Ob-U grian languages there is a passive form o f the verb, and H ungarian also recognises passive forms. It seem s that these appeared independently in both languages by a secondary route, ju st as the definite verb ending did. However, it is probable that the fundam ental linguis tic conditions for the definite conjugation evolved in the com m on Ugrian language. The value o f these two exam ples is only in show ing that reconstruc tion o f linguistic affinities can only be accom plished w ith highly refined m ethodology and by taking into account several historical questions.
177
The relatives
O u a n tity Dual M u ltifu n c tio n a l ro ot i Interna l/external local suffixes
- Preterit ш т т т ш т тш я
In fin itiv e -jji
• • • • • • • • • Conversion SVO -tendency
Figure 36 Typological classification o f the U ralic languages, after P 6 ter Hajdu
b) The Samoyedic languages and peoples The ethnic name Sam oyed m eans ‘self-eater’ in m odem Russian, but this is the folk-etym ological corruption o f some externally designated name. The Sam oyedic languages can nowadays be divided into three groups: northern, southeastern and southwestern groups. The date when the shared Sam oyedic language divided can be placed at around the time o f Christ. The shared Sam oyedic proto-language was probably spoken in the region o f the Sayan M ountains.
178
Relatives and neighbours
The northern Sam oyedic languages include the N enets, the Enets and the N ganasan languages. The nam es o f all three ethnic groups are self-given, and all derive from a shared word m eaning ‘m an’. The N enets people (also Yurak, or Yurak-Sam oyedic), num bering about 35,000, live thinly scattered on the wide expanse o f tundra betw een the northern course o f the Dvina and the northern estuary o f the R iver Yenisei. At the south they com e into contact w ith the Ob-Ugrians. They speak several highly differentiated dialects. The nam e Yurak originates from the nam e o f the N enets in Ob-Ugrian. A long the lower, i.e. northern, course o f the Yenisei live around 400 Enets (Yenisei Sam oyed) who preserve their own language. The N ganasan language (Tavgi Sam oyed) is spoken by about 1300 people on the Taim yr Peninsula. The Selkup people live to the southw est beside the N arym , Tim, K et and Vasyugan tributaries o f the m iddle Ob. The northern group lives beside the R iver Taz, and beside the left two branches o f the Yenisei, the Turuhan and the Yelogui. The northern group only m oved into their present location from the m iddle O b region. The first syllable o f the nam e m eans ‘land’ and the -kup means ‘m an ’, and is etym ologically related to the Hungarian w ord him ‘m ale’. In the southeast, the Sayan Sam oyedic peoples lived in the Sayan M oun tains. T heir languages are now extinct, but valuable linguistic records were m ade in the 18th and 19th centuries. The best-know n Sayan Sam oyedic lan guage is the Kamas. Its last two speakers revived their com m on language, and after one o f them died, the last Kam as w om an m ade a personal appear ance and contribution, at the 1970 International Finno-U grian Congress. Her m other tongue and her use o f it gave an interesting exam ple o f linguistic extinction and preservation. M ost Sayan Sam oyedic language speakers were Turkified in the first h a lf o f the last century, with the loss o f the Karagas, K oybal, M otor, Soyot and Taigi languages. A t the turn o f the century they were already nam es o f Siberian Turk dialects or languages. The nam e o f the M otor language probably derives from the Turk word bagatur, w hich is the root o f the Hungarian word bator (brave), and whose final origin is obscure, but probably Iranian.
c) The Ob-Ugrian languages and peoples The Voguls, or as they call them selves, the Mcmshi, live betw een the Urals and the low er course o f the Ob. O nly about a h a lf o f the 8500 Voguls speak M anshi. Their language is divided into several dialects, generally nam ed after the rivers around w hich their speakers live. The area is now inhabited only patchily. Speakers o f dialects w ho live far apart no longer understand each other. Their
179
The relatives
literary language is based on the northern dialect. Some dialects have now becom e extinct, and the people have becom e Turkified or Russified. The Ostyaks, or as they call them selves, the H antis, live in the vast area between the Yenisei and the m iddle and low er reaches o f the Ob. Around 70% o f the 22,500 Ostyaks speak their m other tongue. Their dialects are clas sified into northern, eastern and southern groups. Differences betw een them are so deep that there have been attem pts to create several different literary languages. The name o f the R ussian Ostyak people is probably o f Turk origin, derived from one o f the old names for the Bashkirs, Ishtek. The nam e also used to denote the palaeo-Siberian Kets ( Yenisei Ostyaks) and the Selkups. The Ob-U grian languages contain m any words borrowed from Turkic and Zyryan, as well as from old and m odem Russian. In the old Russian chronicles, the people were called the Yugrians, and their country Yugria. In the 12th century, their principal groups lived on the western side o f the Urals. A fter the 13th century, they m igrated east to escape the Zyryans and the Russians. In the 14th century, their centre was still in the Ob region. After their subjugation, Ivan III assum ed the title o f the G reat Duke o f Yugria, which appears in a letter w ritten to the “king o f the U grians”, that is to King M atthias o f Hungary on 29 July 1488. The Vogul nam e first appears in 14th-century sources, and the Ostyak in 16th-century sources. Russian sources ceased to refer to the Yugri people in the 17th century. However, there are reports o f Vogul populations in the K ama and Chusovaya regions up to the 17th and 18th centuries. The last report is o f 69 Vogul inhabitants in the Cherdyn district on the w estern side o f the Urals in 1860. In the 14th century, a Khanate was established under the leadership o f Turkic groups o f Kipchak origin. This was gradually broken up by Russian expeditionary forces from the 16th century onwards.
d) The Pennian languages and peoples The Zyryans, or as they call them selves, the K omis, live along the banks o f the right tributary o f the North Dvina, the Vychegda, the M ezen, w hich flow into the W hite Sea, and the Pechora and its tributaries, which flows into the Barents Sea. The 1990 census gives the num ber o f people declaring them selves to be Komi as 496,000, and 83% o f these give one o f the two li terary languages as their m other tongue. The effective centre o f the Komi population is Siktivkar, the capital o f the republic, at the confluence o f the Sisola and the Vychegda. South from here, centred around Kudim kar, in the Perm area, they speak Permyak. A small Kom i group also lives to the east, in the Yazva valley. The two main groups produced two separate literary lan-
180
Relatives and neighbours
guages. The Zyryan language is divided am ong four larger dialect groups, prim arily according to the representation o f the original /V sound. The two literary languages belong to the same dialect type. In the 14th century, the subsequently sanctified Bishop Stephen o f Perm devised an alphabet using G reek and Cyrillic letters, for the purpose o f spreading the faith. Several ecclesiastical texts from Perm w ritten in Zyryan have survived from between the 14th and 16th centuries. These, however, did not give rise to a widespread use o f the written language, far less to a literary language. The Votyaks, or as they call them selves, the Udmurts, live in the area bounded by the River Vyatka and the low er course o f the Kama. In 1990, 83% o f the approxim ately 746,000 Votyaks declared Votyak to be their m oth er tongue. The capital o f their republic is Izhevsk, on the bank o f the R iver Izh. The first syllable o f the Udm urt ethnic nam e can be traced to the nam e used by the Cheremis, Odo, or to a word w hich m eans ‘field’. The second syl lable is o f Iranian origin and m eans ‘m an’. The first part o f the nam e Votyak is identical with the u d or odo element, and the second is the R ussian ‘peo p le’ suffix. The level o f correspondence in vocabulary and form s betw een the Zyryan and Votyak languages is 80%, but the pronunciation deviates to such an extent that speakers o f the two languages cannot understand each other. In the 10th century, when the Perm ian language o f the K am a and Vyatka areas was still shared, its speakers came into contact w ith the B ulghar-Turkic population. This shared Perm ian language has C huvash loan words, which have significance for the chronology o f sim ilar C huvash loan w ords in the Hungarian language. The split o f the two Perm ian languages m ust be dated to after this period. Between the 11th and 13th centuries, the Zyryans w ere in contact w ith the N ovgorod R ussian and Volga Bulghar peoples, and later w ith the K ipchak people. The Votyaks came under the influence o f the Volga Bulghars, and from the 13th century, very strongly, o f the Tatars o f Khazan. The Volga region fell victim to the same M ongol attack o f 1235-36 w hich destroyed the M agna H ungaria and M agna Bulgharia in the K am a region, and fundam entally altered the Votyak pattern o f settlem ent. They have m ore or less been living in their present areas since that time.
181
The relatives
e) The Volga Finno-Ugrian languages and peoples Earlier there was little doubt that the Cheremis and M ordvin languages were derived from a shared precursor. It is becom ing an increasingly com m on view, however, that there was never such a shared Volga Finnish era, or if there was, it was very short.
The Cheremis language and people The Cheremis, or as they call themselves, the M ari, people num ber some 670,800. M ore than 90% o f them speak their m other tongue. The origin o f the Cheremis name remains unclear. Jordanes m entions am ong the peoples subjugated by the Gothic King Ermanrich, along with the M erens and M ordens peoples, a people called the Sremniscans. (In the m anuscripts, Rem niscans and Im niscaris can also be found.) Researchers identify the first two as references to the M erya and the M ordvin, naturally not around 350 when Erm anrich ruled, but at the tim e w hen Jordanes did his work, in the m iddle o f the 6 th century. The third nam e is linked by some to the nam e o f the Cheremis people, and by others to the nam e o f the R iver Cheremshan. For a long time, Cheremis was also the name used for Chuvash in travellers’ accounts. The Tatars now refer to the Cheremis as Chirmesh, a R ussian form, the origin o f which is not clear. In Chuvash, their nam e is Syarmas, w hich is the regular continuation o f the earlier Chermish. The nam e o f the M ari people also m eans ‘m an’ (see p. 303). There is a m ountain dialect o f Cherem is spoken on the right bank o f the Volga, close to the Chuvash. The other dialects, collectively known as m eadow Cheremis, are extrem ely diverse. The eastern Cheremis, who live in Bashkiria, split off in the 18th century. M ountain and M eadow Cheremis have separate literary languages. Cherem is has very old Chuvash-type loan words. There is a debate over when these loan words were acquired. It is decided by M ongol loan-w ords which passed into Volga Bulghar or M iddle Chuvash, and from there to Cheremis. These cannot be older than the 13th century, and display the same features as the other Chuvash-type loan words. The start o f contact betw een the Cheremis and the Chuvash m ust therefore be put to the 13th century. Some Hungarian word equivalents or phonetic forms originating in the Chuvash language have only been preserved in Cherem is, and so Cherem is inform ation is very important. The Cherem is loan words from Chuvash, which num ber over 1 0 0 0 , also have the power to settle the issue o f the historical relationships betw een the Cherem is dialects.
182
Relatives a n d neighbours
The sole reason that can be adduced for linguistic relations between Chuvash and Cherem is not predating the 13th century is that the Cherem is moved into the region bordering the C huvash only shortly earlier. M igrating from the west, probably from the Oka region, the Cherem is m ay have m et the w estw ard-m oving Chuvash w est o f the Volga bend. The Chuvash influence was followed by a strong Kazan Tatar influence. This continued even after the capture o f Kazan by the Russians in 1551. The Russian influence, w hich is apparent in the m iddle Volga region from the 9th century, becam e m uch more intensive from the 16th century.
The M ordvin language and people There are 1.2 m illion M ordvins, but they are thinly spread out betw een the Oka and the Belaya. As noted above, their nam e was first recorded by Jor danes. Byzantine, Arab and later Russian sources m ention them, and Julianus, a Hungarian traveller o f the 13 th century also recorded their name. The M ordvin language is divided into two, near-independent branches. The western group living in the R iver M oksha area and its surroundings are known as M oksha Mordvin or sim ply M oksha, the eastern group beside the R iver Sura are know n as E rzya M ordvin or Erzya. The linguistic boundary betw een the two groups falls along a tributary o f the R iver Alatir, the River Insar. Speakers o f the two languages can understand each other only w ith great difficulty, and two literary languages have been created. It is interesting that, w hereas there are a great num ber o f Chuvash-type Turk loan words in C herem is and many in the Perm ian languages, there are practically none in the M ordvin dialects. This can hardly be due to these peoples not com ing into contact with each other, because it is difficult to conceive that the M ordvins could have been isolated from the Bulghars and the Khazars. Presum ably their cultural and econom ic life was such that living with Turk neighbours had no effect. This could only have been possible if they lived on equal social and political terms with them.
f ) Extinct Finno-Ugrian languages and peoples There is a view, held by some, that the whole o f Russia really consists o f Finno-U grian peoples assim ilated one after another by the incom ing Slavs. The river nam es produced in evidence are not unam biguous. W hile it may not yet be possible to prove-—or indeed to disprove— the assertion o f a Finno-U grian population stretching across Russia, there is a great deal o f evi dence that Finno-U grian peoples w ere absorbed by Slavonic peoples and dis
183
The relatives
appeared. This was the fate o f the M eiya, M eshcher and M uroma peoples m entioned in early R ussian sources. The M erya or Meri lived in the vicinity o f Rostov betw een the Volga and the Suzdal, and their Slavisation m ay have started in the 10 th century, becom ing com plete on the establishm ent o f the R ostov-Suzdal Principality. The M eshchers lived on the right bank o f the Volga, and the M urom s between the lower Oka and the Klyazma. The course o f their Slavisation probably came to a conclusion in the 11th century. The Teryuhan, living between the Volga and the Oka, are Erzya M ordvin who Russified in the 18th century, but preserved their origins in their culture and consciousness. The Karatai, living in the Tatar Republic, are Tatarised M ordvins, who have retained their ethnic distinction within the Tatars up to this century.
g) The Balto-Finnic languages and peoples The Finnish language and people O f the 5 m illion inhabitants o f Finland, around 7% speak Sw edish as their first language and the rest Finnish. The Finns are m entioned by authors o f ancient times. The name o f the Finn people is o f Scandinavian origin, and m ay have first been applied to the incom ing Lapps. The Finns refer to them selves by the nam e Suomi, the origin o f w hich is uncertain, but was earlier attached to a Baltic Finnic tribe. Early Germ anic, Baltic and Slavonic loan words were o f great assistance in tracing the history o f the Finnish language. The earliest records o f the Finnish language are m arginal notes w ritten into Latin texts o f the 13th century. The first continuous text in the language dates from the first h alf o f the 16th century. Dialects o f Finnish differ considerably from each other. Seven m ajor dialects are norm ally identified, and these are distributed among two larger groups, the Western and the Eastern. There is considerable dispute over when the Finns arrived in their present hom eland. The m ost w idely-held view is that some tribes from the ancient hom e o f the Finns south o f the Finnish G ulf started to inhabit the northern shore o f the Finnish G ulf and the interior o f the country in the first centuries A D . Others are o f the opinion that Finns, or a related people, have lived in the area o f Finland since the 3rd m illennium BC. In any case, there are no Turkic loan words to be found in any o f the Baltic Finnic languages. Finnish and M agyar are the two m ost distant m em bers o f the Finno-U grian language group, the distance between them being sim ilar to that betw een Latin and Persian. Yet there are m any shared basic words, such as the Hungarian had ‘arm y’ or ‘w a r’, the Finnish kunta, the Hungarian hagy 'leave’, the Finnish
184
Relatives and neighbours
kadota, the Hungarian haj ‘fat’, the Finnish kuu, the Hungarian hal ‘fish’, the Finnish kala, the Hungarian hal ‘die’, the Finnish kuolla, the H ungarian hall ‘h e a r’, the Finnish kuula, etc. There are suffixes w ith com m on origins, and some old shared derivatives.
The Estonian language and people There are som e one m illion Estonians alive today, m ost o f them in the Estonian Republic. The name Estonian has only been used by Estonians since the m iddle o f the 19th century to refer to them selves. It first appeared in T acitus’s w ork Germania (98 BC) in the form aestii, and has since appeared in num erous other sources, but it was probably coined by the non-Finno-U grian Baltic people (Latvians, Lithuanians, O ld Prussians). The w ord itself m ay be o f G erm anic origin, and only taken over by the Finno-U grian Estonians at the end o f the 18th century. The Estonians form erly referred to them selves as M aam ees, which is m ade up o f the words m aa ‘land’ and m ees ‘person, m an’. Estonian is divided into two dialects, one spoken by north and northeastern Estonians, and the other by those in the south. The northern dialect is closer to Finnish. M ost o f the Estonians are Lutheran, and a m inority G reek Catholic. This group preserves m any ancient features. The oldest recorded fragm ents o f the Estonian language date from the beginning o f the 13th century, and the first continuous Estonian texts appeared in the 16th century. There are m any Baltic, Germ anic and Slavic words in Estonian inherited from m iddle Finnish, with m any elem ents taken over later from Germ an, R ussian and Lithuanian. W hen the Finnish tribes m igrated, the Estonians probably m ore or less stayed in place, and evolved out o f two m iddle Finnic groups, the local Baltic people being assim ilated. Estonia was invaded by G erm an and D anish forces in the 13th century, the Danes later selling their part to the Teutonic K nights in 1347. In the 17th century, the country fell under Sw edish dom inion. The Tartu University was founded in 1632, and the Tallin Press in 1634. Two literary languages evolved, based on the Tallin (northern) and Tartu (southern) dialects. In 1940 the Soviets invaded the country, follow ed by the G erm ans in 1941, and the Soviets again in 1944. Estonia regained its independence in 1991.
M inor Finnic languages and peoples of the Baltic The Karelian language was once treated as a separate group, but it is actually a dialect o f Finnish, some o f w hose speakers lived in the form er Soviet Union. The num ber o f the K arelians in Russia was recently som e 131,000, 63% o f
185
The relatives
whom spoke their m other tongue. T heir nam e has appeared in the records since the 9th century. M ost o f them live in the vicinity o f Petrozavodsk and Tver. There are substantially greater differences betw een Finnish and the sm aller Finnic languages which are extinct or dying out. These are Izhor (Inkeri), Vepsian, Votic, and Livian. In 1970, some 800 people in the G u lf o f Finland declared them selves to be Izhor, but only one-third o f them spoke the Izhor language. They live between the lakes Ladoga and O nega and the W hite Lake. Four dialects can be distin guished, and they have been very rapidly becom ing Russified. The Vodians live in a few villages in the Estonian R epublic and around St. Petersburg. Very recently, researchers found a few old people who spoke the Vodian language. M ost o f them have becom e assim ilated into their Estonian or Russian surroundings. At the end o f the last century there were still about 3000 thousand Livian people in the north o f Latvia, on the K urland Peninsula. They now num ber a mere 500. The sm aller Baltic Finnic languages have retained m any old features, which makes them im portant for research into the history o f the Baltic Finnic languages.
NOTES For earlier research and theoretical issues o f linguistic relationship, see my monograph. R6na-Tas (1978a). For the Uralic languages and peoples, with an introduction into Uralic studies, see Hajdu (1971, 1976, 1981), H ajdii-D om okos (1987). The sources o f the latest statistical data are Janhunen (1991) and a communication by Hajdu for which I offer my sincere thanks.
IV. THE NEIGHBOURS 1. EARLY INDO-EUROPEAN LANGUAGES A N D PEOPLES
a) The beginnings o f Indo-European languages and the early migrations We have seen that in order to reconstruct the past history o f the conquering M agyar people it is im portant to know the main points o f the Indo-European peoples ’ history. This is because the dom inant early contacts w ere with Iranian groups. The oldest written exam ples o f the Indo-European languages date from around 2000 BC, and were found in Eastern A natolia in m odern Turkey. These are non-Indo-European texts in which the nam es o f Indo-European language speakers occur. Records w ritten in the languages them selves exist from shortly afterwards, such as those from about 2000 BC found in the m iddle o f m odem Turkey, in the area occupied by the Hatti Empire. The language is that o f the Hittites, who them selves set up an em pire there around 1640 BC. These relics show that the “A natolian” group o f Indo-European languages (H ittite, Luwian and Palaic) had already consisted o f separate languages for som e time. This is upheld by earlier river and lake nam es dating from the 2nd m illennium BC, that have elem ents which can be deciphered from Indo-European. This puts the date o f the proto-language o f the ancient Indo-European A natolian lan guages at som e time in the 4th m illennium BC. The forebears o f other large language groups can also be placed in the 3rd m illennium . W eighing up all o f these facts puts the break-up o f the Proto-Indo-European, still unified but made up o f different dialects, at the end o f the 5th, or the start o f the 4th m il lennium BC. The question rem ains o f where this language was spoken, or in other words, what was the Urheimat 0 f the Indo-European languages? There is a very heated debate over the issue; the different regions suggested are shown in Figure 37, along with the nam es o f their proponents. The question has a M agyar aspect too, albeit indirectly. Featuring prom i nently in the arguments for and against the various Urheimat theories are the non-Indo-European words w hich passed into the Indo-European language. Some o f these ancient loan words also found their way from Indo-European into Hungarian. One o f the m ain weapons and im plem ents o f the B ronze Age was the axe, which carried cult significance in some B ronze-A ge cultures. The
188
Relatives and neighbour.
Figure 37 The Indo-European peoples’ Urheimat. Views since 1960
Figure 38 A ncient Indo-European languages, Hittite, Luwian and Palaic, am ong Anatolian and M iddle-Eastern non-Indo-European languages
The neighbours
189
Figure 39 The evolution o f Indo-European languages, after G am krelidze and Ivanov
Semitic nam e o f the axe is the noun derived from the verb m ade up o f the roots p-l-k, and m eaning ‘to split’, i.e. that w hich splits, or w hat we split with. This existed as the forms p ilakka in Akkadian, p elka in Old Syrian and pilka in M andean. The Indo-European languages adopted this as *peleku and *phelekhu. The word is certainly very old, because it already existed in the M ycenaean and Cretan culture, and also occurs in H om er in the fo rm pelekiis. The w ord also passed into the Turk languages, appearing in the form balqa in Old Turkic. Through another route it arrived in M ongol in the form *paluka, from which haluka and the m odem M ongol form aluka were derived, every where with the m eaning ‘ax e’ or ‘hatchet’. The word was passed down from the basic Indo-European language into Indo-Iranian, and here it started with the form peleku, from w hich it progressed regularly to paraku > parathu > partha. It was borrow ed from Old Persian by Ossetic and Tocharian. This form also probably passed through Finno-Ugrian to C huvash w here it now appears regularly in the form purta. There w as a tim e and an Iranian dialect, however, in which the /г/ sound becam e /1/. It was from this language that the p alta form entered Turk, and the Turk balta into M agyar where it rem ains as balta.
190
Relatives and neighbours
This loan word does not standalone. The H ungarian w ord szekerce ‘hatchet’ derives from Southern Slavonic. Its Slavic basic word, sekira, goes back through the Latin securis ‘w eapon, chariot’ to an Indo-European w ord *sekur which m ay be o f Sem itic origin; see the A kkadian shukurru ‘lance, piercing w eapon’, and the Hebrew segor ‘axe’, influenced by Indo-European sek- ‘to cu t’. This all needed saying because these are im portant, typical loan w ords for tools carrying cult associations. The axe turns up frequently in Volga BronzeAge figuration. The Hungarian word for ‘seven’, het, was originally o f the form *et, w ith the h attached by analogy with the w ord for ‘six ’, hat. The word also exists in the other Ugrian languages, and it is possible to reconstruct the
Figure 40 Isoglosses o f the Indo-European languages, after A nttila (1 -2 4 phonetic and m orphological isoglosses)
191
The neighbours
f
„ \° д и » \ _
C ?/,-/r
к
.. i re ^
U
'
B a ltic
'
г e r т ° П' C ^ Slavonic V C e l t i c £ f~
,v
!
0o v ч-<> г
.
(. -
% ^ ^ 0 с Ь а г / 0/?
ft* < Чч y
и
а n i q h
\
\> ш & а” ППеп1'ъ Г '
^ /- 'h A n a to Itan <" .
//?
Ira ,, /
о
,~x^
4 ' J
4
w
\
s '\ 4„ \ \ ""\
^
'r
>\
^
.....<
s С
J / , !p r / ' ,y \
Figure 41 The main Indo-European languages and their early locations
Ugrian proto-language form *‘septe. This is a loan word from the Indo-Euro pean form *septem. But the Indo-European word itself is probably o f Semitic origin, and was brought over into Indo-European with the cultic num ber seven. The ancient Semitic sabc ‘seven’ is also the root o f the Hebrew sabbat, the name o f the seventh day. So the Hungarian words h it ‘seven’, szeptem ber ‘Septem ber’, and szom bat ‘Saturday’ can all be traced to the same ancient origin. Locating the Urheimat o f Indo-European involves studying languages other than Semitic. One o f these is Hattie, a non-Indo-European language whose relics are know n from the period before the Hittites. The Hittites took over the worship o f the sun god from the Haitians, and w ith it the Hattie word for the sun god. The Hattie name Eshtan transform ed into Ishtanu ‘sun, sun god’, from which, m ost likely through Caucasian and K hazar interm ediaries, it passed into Hungarian as the word for God, lsten. However, such cultural loan words cannot prove relationships between ancient peoples, nor determine where the Urheimat o f Indo-European was located.
192
Relatives and neighbours
Also unsuitable for such a purpose are the affinities o f the Hungarian word for apple, alma. The word exists in every Indo-European language, and the Germ an Apfel, the English apple and the Russian yabloko are all m em bers o f this word-fam ily. The form *ablu~aplu can be reconstructed for the IndoEuropean languages. Hittite has the word sham lu, however, and this is old er than ablu. The Hittite /sh/ regularly disappeared in the other Indo-Euro pean languages, and the ml passed through the juncture m pl to becom e -bl/pl-. From the Hittite sham lu form came *amlu, which, perhaps through Tocharian m ediation appeared in the Turkic languages as *amla and subsequently alma. The Hungarian alma ‘apple’ is o f Turkic origin. A lthough in m ost o f the Indo-European languages the word has recurrent correspondences, some Indo-European languages borrowed it from others. The ancient Indo-Euro pean word for ‘apple’ only m eans that the apple was indigenous in the Indo-European Urheimat. For the Turkic languages, however, it is im portant that the w ord retained the form o f an old Indo-European version. W herever the Urheimat o f Indo-European was, for the history o f the M agyars it is only the history o f the Tocharian and Iranian languages which are im portant. This requires us to take a b rief look at the historical distribution o f Indo-European languages, as shown in Figure 39.
b) The Tocharian language and people Excavations in the Central Asian region o f Turfan have produced texts written in a language which was not Iranian, but was Indo-European. Two dialects have been found quite far from each other, one used in the K ucha and K arashar regions, and the other around Turfan. Buddhist texts w ere w ritten in this language, o f which several were also translated into Turkic. One o f these texts m entions that the Turkic text was translated from Tocharian. The Tocharians had a long history behind them by this time. They had m igrated a long way east o f the Indo-Europeans, right up to the Chinese border. It m ay be that the Chinese learned bee-keeping from them, because the Chinese word for honey is an ancient Tocharian loan word. The people appearing in Chinese sources under the nam e Yuezhi were defeated by the X iongnu around 175 BC. Some o f them drifted w est in the 2nd century BC. Their fate thereafter is o f no interest to us, but w hat is im portant is that a people speaking a language o f the so-called C entum group o f the Indo-European language group, close to Italo-Celtic, m igrated from the Indo-European Urheimat into East Asia. It is not known w hen or how quickly this m ovem ent took place. M aybe they follow ed the Silk Route, but they certainly spent extended periods at one or two places on the
The neighbours
&
Amtnmom Ат Щ т
L,de l*A* ]■_•*}.
Aial ,
Bitlklknh ' A-
•
^-0 c h a r ia n s
j ls. Д8
-ji Lop-nor
% ll*W
wv.
K » n lu n
M ountain*
Figure 42 The hypothesised emergence o f the Tocharians, the Andronovo and the Afanasyevo cultures
Figure 43 The Tocharians and their neighbours in the Tarim Basin
194
Relatives and neighbours
way. In some opinions Tocharian words, and not ju st Iranian elem ents, passed into the Finno-Ugrian languages. However, the evidence brought forward to prove this does not stand up to scrutiny. So we can still only assum e that the Tocharians also played a part in passing on some m igrating culture words. The issue is clouded by recent p roof that several Tocharian words are not ancient Indo-European words, but loan words from Iranian languages, which m ight also have been the route by w hich these words w ere incorporated into the Finno-U grian languages. It is m uch m ore likely that words passed from Tocharian into ancient Turkic. Retreating w est from the Chinese border in the 1st century AD, and settling down divided into two groups in the Turfan region, the Tocharians once again m et with the Turks, who were by then a m ajor political organisation, and it was from these later contacts that the Tocharian words entered Turkic. These spread through Buddhist translations, and from our point o f view can be ne glected. The earlier Tocharian-Turkic contact can only be put som ew here in the 1st m illennium BC, ju st at the time w hen the Tocharians were m oving east. The earlier T ocharian-Turkic contact could be im portant to the question o f the Hungarian language’s C huvash loan words. The H ungarian w ord okor ‘o x ’ com es from the Chuvash *okiir, and not from the C om m on Turk okiiz. A very long tim e ago, there was an attem pt to link this word with Indo-European words. Those that considered /г/ as the original sound associated the Turk word with the L atin pecus, pecoris ‘cattle’. However, the Latin /s/ ending and the -r- in the declined derivative are later, Italic changes, the phenom enon which is know n in Latin linguistic history as rotacism ( Venus > *venezis > veneris), i.e. w hen the /s/ betw een vowels becom es Izl and then /г/. So the Indo-Euro pean root was actually *peku. It was also thought that the /р/ disappeared via an /h/, and that the end o f the word is a Turkic suffix. In this case, the history o f the w ord cannot be used to throw light on w hether the /г/ w as form erly Izl. If Turkic borrow ed the form *peku, it could only have done so from Tocharian, because in the Iranian languages the phonem e /к/ becam e Isl (see Sogdian pasu). Such a word has not yet been found in Tocharian, however. Others considered that the Turk word okiiz was related to the form okso w hich occurs in Tocharian, Ochs in G erm an and ox in English. The recon structed Indo-European proto-language form o f this is *ukos. This gave rise to okiiz and okuz in Turkic (where both exist), and then cam e the change z > r, and okiir passed into Hungarian and M ongol. The w eak point o f these views is that /h/ does exist in M iddle M ongol, and the form hiikiir has been found in early language records. The initial consonant /h/ usually derives from an earlier w ord initial p -, and as such, the form *рдкйг < *pdktis should be reconstructed in Turkic. This, however, cannot be traced to the Indo-European *ukos. This difficulty can perhaps be explained by a so-called cockney Ihl,
195
The neighbours
that is, the ability o f an /h/ to emerge by a secondary process before vowel initials. An exam ple o f this exists for this very word in two Turk languages: the form htikiz exists in U zbek and hokiiz in New-Uighur. And to dispel all doubt as far as the secondary /h/ is concerned, the word for car in U zbek is havtomobil, where it is beyond dispute that the Uzbek language has appended an /h/ in front o f the w ord borrowed from Russian. Such exam ples also appear in the M agyar linguistic record. A lake is nam ed in a docum ent w ritten in 1326 and surviving in a copy dating from 1428, as H w kurithou, which, when rendered into m odem Hungarian, w ould be written as Hukiirito. There are many 13th-century references to the form w ithout/h/, showing that this really is an occurrence o f the secondary /h/ within Hungarian. The value o f the above is that it dem onstrates the assistance that can be drawn from Tocharian, as the earliest eastw ard-m igrating Indo-European language, in reconstructing the background o f ancient M agyar history.
c) The Iranian languages and peoples Since the Indie people who broke o ff from the Indo-European com m unity arrived in the Indian subcontinent by 1500 BC at the latest, but probably well before that, the break-up o f com m on Indo-Iranian m ust be put before this time. The first peoples o f the w estern steppe m entioned by Greek sources were the Cimmerians. Their earliest traces can be dated at around 1000-800 BC. The C im m erians’ linguistic and ethnic affiliations are keenly disputed, some scholars claim ing they spoke an Iranian language, others that their language was unknown. In the 8th-7th centuries BC, the Scythians arrived and took over the dom inion o f the steppe. Their nam e is norm ally explained as originating from a w ord m eaning ‘shoot (an arrow )’. Some are o f the opinion that the name o f the Sogdian people also stems from the sam e proto-form . The Greek sources m ention the eastern branch o f the Scythians by the nam e Saka. The Assyrian sources m ention the M edes for the first time in 835 BC. M ede em perors founded an em pire south o f the Caspian Lake, and, with assistance o f the M esopotam ian people, struggled against the Cim m erians and Scythians. In the m iddle o f the 6 th century B C , the A chaem enid em perors founded the Persian Empire. Darius consolidated im perial pow er around 520 BC and made Persia into a great empire. The Persians occupied the area o f m odem Turkey, M esopotam ia and Egypt, and at the other extrem e spread their dom inion northward and eastward. This em pire was brought to an end in 330 BC by A lexander the Great. A lexander also occupied the lands o f the Parthians, the Bactrians and the Sogdians, and in 328 BC took as his wife Roxana, the
Relatives and neighbours
196
daughter o f the Sogdian king. This was the beginning o f the H ellenisation o f Bactria. W hile the M edes and the Persians were establishing em pires in the south, the lands to the north o f them started to receive settlem ents o f eastern Irani ans. In the m iddle o f the 1st m illennium BC, the settlers o f K hw arezm and Sogdiana, the Khwarezm ians and the Sogdians, and the Sakas w ho settled the East Turkestan basin, centred on Khotan, w ere all Iranian people, and had developed w ritten languages o f their own. A lthough they w ere subjected to Persian control as long as the Persian Em pire flourished, this w as m ore politi cal dependence than a close linkage betw een the peoples. A round 247 BC , the Parthians took over the eastern part o f Persia held by the Seleucid dynasty which had follow ed A lexander the Great. In the 1st century BC, the Parthians ruled to the south, and the M edes southwest, o f the Caspian Lake. On the steppe, the Scythians w ere ousted by the Sarm atians around the 3rd century BC. This people m ay have been the descendants o f the Saurom atas m entioned by Herodotus as being Scythians who lived in the vicinity o f the Don and spoke the Scythian language badly. The Sarm atians gave rise to what later becam e independent peoples: the Roxolani, the A o rs and the Alani. These peoples were ruling the western ranges o f the steppe by the tim e o f Christ. / (
У
• ..w,*
A }
X
>
■ .
S
i
.S c y t h k m s
V 4-- A, - - ••
b ~
AiV/rKiliarts
-/jf/W 1 * a4 \
*
■<-
•
-
\
\ и
/Mans
, 4:' .
{ .u f * ' ' iV a t,
\
S o g d ia n s
с? V' ■ V\
"v./ % $
V\
^ V i; \ \
■’£ .
i
-> '.
K ln u a r te in ic m ti
,
i
T i u i ’ >!
S a k .h
/
P e r s ia n
i
1
P.mhtii
I ‘ j Eastern Iranian lan^yagt area Г~1 Vijestern Iranian lang'ttlge ar«p
^ £ ,j
•
Figure 44. The early distribution o f the Iranian peoples
I
197
The neighbours
г у t h Ji а л
.
Mas Si
Lake' Aiaii, O k, ■Vi.
Bl n, «5s; >'*| S&tfefe»' *M itat«: C y p r u ss
^SOGDIANA
^ А Р Р А ° ° ^ . -v, \
.. ;
«*?>* *»*«•*«» ;
д
с
ARMENI A
« t\IJi ! -Л ;
r* Bjsctra' B A C T R IA ^
*
NhUitana
£
Й Г .,# J"гныЬ'т * / > o 'Г '• -'О
Memphis •!*"'^52 5 / / Petusnrn 1 c v FI
Pcrsepolts ■> PERSIS
Tlu'bo ,
у
%
.
GEDROSiA
Figure 45 Developm ent o f the Persian Em pire ( 6 th—5th centuries
/
,'v
i '/
BC)
By the time the Turkic tribes arrived in the w estern region o f the steppe, the only old Iranian nom ads that they found were the Alani. The other three Iranian peoples, the Sogdians, the K hwarezm ians and the Sakas had settled at the main trading stations. In 227 A D , a new dynasty took up pow er in Persia, the Sasanians, who held them selves to be the successors o f the Achaem enids, and revived the Zoroastrian religion. For a short while (217-277), state support was also received by the religion o f M ani, M anichaeism , w hich proclaim s the dualism o f the world. Around 350 A D , a new m igration took place from the north into the old land o f the Sogdians between the two rivers that flow into Lake Aral, the Oxus and the Yaxartes (also known as the Am u Darya and the Syr Darya). The peoples involved are m entioned under the nam es o f Far, Hyon, and H ephthalite or Ephthalite. The Hephthal or Ephthal dynasty had the upper hand from the middle o f the 5th century. The latter are som etim es referred to in the literature as the White Huns. This nom ad em pire spread its dom ination to the Caspian Lake and besieged the borders o f Persia. The Sasanian Persian Emperor, Shapur II (309-379) fought them hard and with varying degrees o f success.
198
Relatives and neighbours
\
V o*-
7
x
l r
*i
' "
S £
■Tv r
■♦
:
•/ ./
4 \ ■}. -
/
X) f
\ ¥
\
H/N-4 \ r ' K**g
W
r Ли X
" ш^ fc € с ® га ш J S v* T3 J~ о
Ш-О a
Ш 6
с
ф < -a -c g *i_- ^о7C»3 ’Oяз <* Ш ®D
г
В .2 Ш о
*. < Г С P »# x ш Ф <
JZ
Figure 46 Alexander the G reat’s empire (336-323 BC) and the Sogdians
•-
и
The neighbours
199
Figure 47 The Parthian and the M ede Empire (3 rd -ls t centuries BC)
The Var-Hyon migration is related to the m igrations o f other peoples which have a m ore direct bearing on the history o f the M agyars. Peoples arrived in the Khazakh steppe at this time who were very probably already Turkic speakers, and who were subsequently involved in the Onoghur m igration. The later Avar m igration also has a connection to the break-up o f the Hephthalite Empire, because the Var and Hyon groups were am ong the peoples o f this empire too. We will return in due course to this im portant event for Hungarian history.
200
Relatives and neighbours
d) The Alani The Alani were a people who survived the Iranian era on the steppe, and w ent on to assum e a position as a m ajor force there. The Alani probably assim ilated the Ao/'s and A s peoples, some scholars believing that the two latter, or at least their nam es, were originally the same, and the Alani w ere one o f their leading tribes. The 1Oth-century A rabic sources (Ibn Rusta) distinguish the Alani (al-Lan) and the A sians, but the A sians only start to appear more frequently in the sources after the M ongol period. It is uncertain w hether the people appearing as A z in Turk inscriptions are the same as the Asians. The general view is that the As form becam e Yas in Slavonic, although in Chuvash-type Turkic languages the As form also regularly b e com es Yas. The Turkic w ord m eaning ‘erm ine’, as, is in m odern C h u v a s h e s , which can be traced directly to the form К к. The O ssetian people and language are clearly derived from those o f the A lani-A sian people, and the nam e Ossetian is derived from the Georgian version o f As. W estern and C hinese sources m ention the A sian people am ong the Scythians relatively early, with the form A siani also occurring occasionally. Strabon (64 B C - 1 9 A D ) m entions the A n / people am ong those Scythians who conquered Bactria from the Greeks. Perhaps they fell under Alani leader ship in the 1st century along with the Aorsian alliance. If we discount some contem porary references to the Asian people placing them in the regions o f Sogdiana and Bactria, the Alani first appear in the sources from the 6 th century. The Byzantine sources follow their history in particular detail, because the Alani, or a part o f them, becam e C hristians very early. Procopius first m entions them before 552 as people living north o f the C aucasus, along with the A bhazians and the Zikh people, and as friends o f the Rom ans. Behind them, i.e. to the north, lived the Sabirs. Chinese sources, especially the Tangshu, recall the Alani in several places. They are m entioned as belonging to the w estern Turk people in descriptions o f the latter, and in the geographical chapter o f the Tangshu it is w ritten that they lived east o f the Byzantine Em pire (Fulin), together with other tribes w hich are referred to by nam es o f disputed identity, but include what is quite probably a transcription o f *Ongur or *Ungur. These findings relate to the 7th century. M ichael the Syrian, in the work already cited (see pp. 75 and also 229 be low), writes that during the reign o f Em peror M aurice, Bulgharios, one o f three brothers from the east, m igrated to the Danube and there founded the Danube B ulghar Empire. The other two brothers w ent to the land o f the Alani, w hich was called Barsilia. Its cities were built by the Rom ans, and were close to the Caspian Lake and the Gate o f Toraye. This Gate o f Toraye w as probably
The neighbours
201
the Darial Pass, which got its name, D ar-iA lan, the “Gate o f the A lani” , from the Persians. The Darial Pass lay north o f m odem Tiflis, and stretched to Vladikavkaz, i.e. beside the road to Ossetia. According to M ichael the Syrian, or rather his sources, the B ulghars and the Vongurs (Vogurs) had been Christians earlier, but after a foreign people had taken control o f them they w ere called Khazars, after the nam e o f the older brother, Khazarig. This story m ust refer to the period around 678, and recalls the K hazars’ occupation o f Alani lands at an earlier date. Arab sources frequently m ention the Alani from the 8 th century onwards, in reports o f m ilitary expeditions through the Caucasus. In 725, the Arab warlord Jarrah led a cam paign against the Khazars, in the course o f which he attacked the Alani at the Darial Pass and exacted tribute from them. In 737, M arvan tricked the Khazars into thinking that the m ilitary expedition he was preparing was to be against the Alani, thus causing them to drop their vigilance. He attacked sim ultaneously across the D erbend and Darial Passes, but it seems that the m ain arm y attacked the Khazars through the Alani Empire, i.e. over the Darial Pass and across Alania. Around 760, under the leadership o f Ras Tarkhan, the Khazars drove south and attacked the Alani. It is m entioned that in 854, the M uslim governor o f Arm enia and A zerbaijan perm itted some K hazar families to m ove with their leader, Buga the Elder, into an area ruled by Islam; 3000 Alani families went with them into the Darial Pass region. Porphyrogenitus wrote o f the Alani in detail around 945-950. He pointed out to his successors that the Alani could still attack the Khazars, w hose nine richest provinces, from which they drew all o f their wealth, lay adjacent to the Alani lands, and an attack would therefore w reak great dam age on the Khazars. Additionally, if the Khazars, com ing from Sharkel, attacked K herson and the “R egions”, i.e. the Byzantine possessions in the Crim ea, then it would be expedient if the Alani attacked the Khazars, because being unable to fight on two fronts, they would leave Kherson in peace. In descriptions o f journeys, it is w ritten that beyond Sharkel and Tamatarhan is the R iver Ukruh, and then the R iver N ikopsis, which flows into the Zikhia. Then follows Papagaia, K asakhia and the Caucasian M ountains, behind which lie the Alani. Then there are islands in the Caspian Lake w here the Zikhians and others escaped from Alani attacks. Descriptions o f the Pecheneg lands m ention that the Pechenegs living beyond the D nieper w ere confronted by Uzians, Khazars and Alani from the east and north. According to M asudi, who died in 956, the Turkic Bajna, Bajnak, Bajghard and Nokarda lived w est o f the Khazars and Alani. W hatever the status o f this
202
Relatives and neighbours
clearly degraded source is, the Khazars and the Alani appear together here too. M asudi describes the Alani Kingdom in detail, and gives the title o f its em perors, w hich is Karkundaj. T heir capital was called M agas. M asudi claims that the nam e o f the capital has the m eaning ‘religion’, but this is obviously an explanation from popular etymology. M asudi claim s they w ere C hristian until 932, Ibn Rusta holds that only the rulers w ere C hristian, and at all events their bishops and priests were banished in 932. In prehistoric tim es, a bridge and a castle were built betw een the Caucasus and the land o f the A lani to stop them intruding into Persia. This is a reference to the D arial Pass. According to Biruni, bom in 973, the Alani and the Asians lived betw een the Caspian Lake and Lake Aral. The Alani also appear in the records o f the Jewish traveller, Benjam in o f Tudela (around 1170), as living under the rule o f the Prince o f Captivity, w hich could be a reference to their earlier allegiance to the K hazar ruler. The H udiid al-alam, written around 982, precisely describes the country o f the Alani. Its king and some o f its people w ere C hristian, and the rest idolworshippers. One o f their lands was Khasakh, identical to the K asakhia w hich occurs in Porphyrogenitus’s work. They are the Circassians, know n in R ussian as K asog, who spoke a Caucasian language. It m entions the Alanian Gate (Dar-i Alan), which was a city built on the sum m it o f a hill as a fortress, and where protection was provided by a guard o f 1 000 soldiers, changed every day. There are also records in the correspondence betw een K haghan Joseph and Hasdai ibn Shaprut, and in the Hebrew “C am bridge D ocum ent”, o f the Alani and the A lanian or Darial Pass. The capital o f the Alani was destroyed by the M ongols in 1239. A fter being m ade hom eless, some o f the A lani or Asians stayed in place, but a section o f them joined w ith the Cum ans, and around 1245 cam e to H ungary with the second m igration o f the Cum ans, and were settled by Bela IV in w hat becam e the Yas province (Jaszsag). The Latin sources o f the M iddle Ages called this province Yazygia by an erroneous equation with the Yazig people in Strabon (VII, 2.4). Some archaeologists think that after M arvan’s attack in 737, a large section o f A lani escaped to the north and settled down. It is true that in the area o f the upper R iver Donets there are finds from the Saltovo-M ayak culture, such as catacom b-like graves, o f which parallels have been found in areas inhabited by the C aucasian Alani. It is difficult, however, to use these and the anthro pological features usually m entioned in association with them , as a basis for proposing the existence o f a northern Alani group. It cannot, o f course, be excluded that as a result o f the M arvan attack, w hich certainly induced large m ovem ents o f people, Alani groups really did com e to the D onets-D on area, but this cannot be dem onstrated from any other sources. Some have proposed that these northerly-shifting Alani groups w ere Turkified at an early stage. All
203
The neighbours
this is unverifiable speculation, however. In any case, different ethnic groups m oved and m igrated m ore freely w ithin the K hazar Em pire ju st at the tim e o f the Pax K hazarica, i.e. in the 9th century.
e) The Ostrogoths o f the Crimea The Ostrogoths did not speak an Iranian, but a Germ anic language. The first groups appeared around 150 A D on the north shore o f the Black Sea. Their settlement along the lower course o f the D nieper was com plete by about 230. They had already clashed w ith the Rom an Em pire in the 3rd century, but their intrusions into the Balkans stopped at the end o f the 3rd century. A t that time, the Visigoths and the Ostrogoths m ade up a great tribal alliance. One group o f Visigoths w ere converted to C hristianity under W ulfila (d. 383 A D ) and settled in the Roman Empire in 348. W ulfila translated the Bible into the Gothic language. The m ain body o f Visigoths traversed Europe and settled down in Spain, where they were dispersed by the Arabs in the 8 th century. Around 360, the Ostrogoths, under Erm anrich, established a m ajor empire between the D niester and the Don, but it was overthrown by the Huns in alliance with the Alani after 370. M any o f them m igrated w est and fought in alliance with Attila, and later appeared in Pannonia and Italy. One group was converted to Christianity, however, and rem ained in the Crim ean Peninsula. There are records o f the Crim ean Ostrogoths dating from 300. They lived in the vicinity o f K herson (today Sevastopol), and their language survived right up to the 16th century, when Ogier Ghislain de Busbecq (1560-1562) set down a small vocabulary and a short linguistic record. The O strogoths o f the Crim ea were subordinated to K hazar rule in 787, w hich was displaced by Byzantine power before 795. They featured prom inently in K hazar-B yzantine wars, in Crimean trade, and in the spreading o f Christianity.
2. THE XIONGNU AND THE HUNS Chinese history is deeply coloured by struggles with barbarians from the north and west. These barbarians swept into China from the steppe, som e o f them gaining control o f the northern provinces o f China, and even form ing dynas ties, before becom ing Sinified and assim ilated into the Chinese m asses. Other groups retreated on to the steppe, where they either disappeared or drove westward. This historical scenario was successively repeated for a thousand years. The people o f interest to us here is that w hich cropped up in the Chinese sources around 318 BC under the name Xiongnu. They w ere classical nomads, to the extent that in later Chinese sources all nomads were “X iongnu” , ju st as
204
Relatives and neighbours
for the Greeks they were all “Scythians” . D escriptions o f nom adic life repeatedly use the parlance applied to the Xiongnu. The Turks are also ascribed features copied from the X iongnu stories, em bellished w ith new observations. The first great X iongnu chief, M aodun (ca. 2 0 9 -1 7 4 B C ) organised his em pire into a great pow er with cruelty which shocked even the Chinese. The em pire stretched from Korea in the east to Lake Balkhash in the west. The title o f the chieftain was Shanyu. The following century was passed fighting, with varying levels o f success, against the Chinese, during w hich the X iongnu took m uch from Chinese culture, and the Chinese adopted the X iongnu’s cavalry tactics. In the m iddle o f the 1st century AD, the Chinese took advantage o f internal pow er struggles betw een X iongnu leaders and launched counter attacks. The leader o f one o f the X iongnu groups, Zhizhi, brother o f the chief, fled w est after his defeat. The rest o f the X iongnu’s history follows the usual pattern: som e o f them retreated north, but the greater part were assim ilated into the northern population o f China. A western X iongnu Empire was set up covering h alf o f the O ld E m pire’s lands, and the X iongnu flourished briefly for a second time. This time, the people overrun by the w estern X iongnu were certainly Turkic, although the nam es o f the peoples referred to in Chinese sources cannot always be identi fied w ith people known from elsewhere. The w estern X iongnu Em pire spread as far as the Rivers Talas and Chu. The Chinese once again w ent on the attack, dealing the West Xiongnu Em pire a decisive defeat in 36 BC and capturing its capital. Zhizhi lost his life in the battle. Some o f the Xiongnu, who had fled north and regrouped, attacked China again in the early years AD, but suffered a defeat in 91 AD and turned westward. This second X iongnu wave also found a hom e in the w estern part o f Central Asia, in the Tianshan region. Hardly anything is known o f the X iongnu’s language. As in m ost nomad em pires, there were certainly m any peoples o f varying languages and ethnic origins. A ttem pts have been m ade to link the few nam es, com m on words and one short, ten-syllable Xiongnu verse that crop up in the sources with nearly every early language in the world. Assessing the scant and contradictory records is made even m ore difficult by the unsuitability o f Chinese script for rendering foreign words, and by the fact that the words recorded are m ainly culture and loan words. The X iongnu issue is im portant for us because o f the long, if unproved, suspicion that there are historical links o f som e kind betw een the X iongnu and the European Huns. Recent research has put it beyond doubt that the Sogdians, w ho controlled trade throughout the entire steppe, knew the X iongnu as Huns. For this reason it has becom e custom ary to refer to the X iongnu as Asian Huns,
205
The neighbours
Figure 48 China and the X iongnu
but there is little m ore to identify the X iongnu as Huns. In the waves o f Hun m igration which later swept Europe, the X iongnu aristocracy w hich controlled the steppe peoples w ould certainly have been prom inently involved, and al though they m ay have held high influence within the H unnish Em pire, they did not necessarily control it.
Relatives and neighbours
206
7 B%ANmX \ ^ « 7i<0 > '■ & /I' '--'A *f~j 4 ] ^ Т гл,,Г ^A rge- t>vamn; SArgen^;»^,» >»s^« « >
’p"'ntll":,d' 11 r
“ * АчшЙ^Лф, а Л 4 *>
4 Meditifinuni» ~ c£*A X 'к У Ь : гга И а Т % ^
4-PRbUINCl/ .„A :
i^N O N A R lA
I
•
Ш ( > \l M SASSAM-
M ^»
D/\Rl
■ T H R A C IA
. \l^ A C 1Л A ^ C^ ^ t o ■ r i t ]■ ! . jjiofb P 0 N n C'
ПгХы$
SUBt JRBiCARlA
l-mdjypUlA (дЪ| \f>д ‘» Ip! N8
\ i ,h,
■,r (t,
„ьЗ
'‘ACHAfvV. •
?
A FRi СA \ •
I
\btcilta
The border of the Western and Eastern Empire з (395/ADj
Alexandria ■ V ’V v. A i . G Y PT1US V:> v’\
Areas where the Franks, the Ostrogoths and ths Visigoths settled
I
V
Figure 49 The split o f the Roman Empire in the 4th century, the Hun attack, and the Goths
W hen the Huns crossed the Volga in 375 A D , they were not com pletely unknow n in Europe. A few years earlier, Byzantium and Sasanian Persia had suspended their, otherw ise life-or-death, struggle with each other to defend the C aucasian passes against the Huns. Sw eeping across the Eastern European steppe, the Huns enslaved the A lani, who had attacked the Goths, who in turn had borne dow n on other Germ anic tribes. Earlier, the Ostrogoths, under Erm anrich, had built up a form idable empire. A fter the Alani had been overrun, the Huns clashed with the Ostrogoths, who at this time w ere living on the Crim ean Peninsula. They enslaved or enlisted the m ajority o f them, although one group asked to be handed over to the Rom an Em pire. The Alani and O strogoth populations thus joined those enslaved and enlisted people who had com e with the Huns from Asia, beyond the Volga. In 395 AD, the m ain stronghold o f the Huns m ay have lain north o f the C aucasus, because it was then that they crossed these m ountains and took their forces into battle on the lands o f Persia and Byzantium . From the 400s, the
207
The neighbours
Figure 50 The Huns in Europe
centre o f the H unnish Empire gradually shifted to what were later Hungarian lands, between the Danube and the Tisza, and Pannonia. Bleda and Attila, who came to power in 434, had their base betw een the Danube and the Tisza. A fter consolidating their power, the Huns spread west to the Rhine, north to the Baltic Sea, and east to the Don. In a great battle in 436, they defeated the Germ anic Burgundians who lived around Worms. The m em ory o f this battle is preserved in the m uch later epos Das N ibelungenlied (The Song o f the Nibelungs). In 441, the Huns turned against the Eastern Roman Empire, i.e. Byzantium. Continual fighting was occasionally inter rupted by an exchange o f envoys and a peace treaty. A m em ber o f one o f these delegations was Priskos, whose writings remain the best surviving description o f the Hunnish court and people. In 445 A ttila becam e the Great King o f the Huns. In spring 451, Attila and his Gothic allies crossed the Rhine and attacked the Visigoth king, whose seat was in what is now Toulouse. A great battle took place on the Cam pus M auriacus near Troyes, where the Romans under General Aetius fought alongside the Visigoths. There were heavy casualties on both sides, and the battle ended with the defeat, but not destruction, o f the Huns. It did not, at this time, decide the fate o f the Hunnish Em pire and Europe. In 452 Attila attacked Italy, taking and sacking A quileia, but proved unable to follow this up with further significant successes. Roman counter-attacks and epidem ics decim ated his army. The Rom an Em peror M arcianus (450-457) resolved to send em issaries to Attila. The chief em issary was Pope Leo I (440-461),
208
Relatives and neighbours
and after the talks, Attila led his arm y home. One year later, in 453, A ttila died unexpectedly, and the em pire w hich he had held together disintegrated w ithin months. The rem ainder o f the H unnish armies w ithdrew to the area below the Caucasus. Little is known o f w hat becam e o f them ; m ost o f the tribes who appeared later under the nam e Hun w ere not A ttila’s Huns, but new peoples. A ttila’s em pire was the greatest in the history o f the steppe nom ads until Chingis. M any have attem pted to reconstruct the H uns’ language. It is inter esting that the very scant sources o f inform ation are often m utually contradic tory. Three H unnish com m on w ords have been recorded: strava ‘funeral feast’, medos ‘honey b eer’ and kam on ‘m illet b e e r’. O f these, the first is certainly Slavonic, the second probably so too, and the origin o f the third is com pletely obscure. But these are cultural w ords and it is not possible to m ake inferences from them concerning the language o f the Huns, only that drinks with Slavonic nam es were drunk in A ttila’s court. It is unlikely that Priskos would have had Slav interpreters. H unnish personal nam es hardly yield any inform ation. We do not know w hat they m ean, and their phonetic transcription is dubious. A ttila’s nam e was certainly Gothic, because the -Ha is a w ellknow n Gothic ending which occurs in m any G othic nam es (such as the Goth bishop W ulfila, and the G othic personal nam es A nsila, Hunila and Totila). But as has already been pointed out, the fact that the H ungarian K ing A ndrew had a nam e o f G reek origin does not m ake the H ungarians Greeks. Thus the m ost appropriate attitude to the nam es o f European Huns is that there w ere certainly Slavs, Goths and Iranians am ong the people, but it is not know n for sure w hich language was spoken am ong the rulers. It is possible, w ithout there being serious evidence for it, that a Turkic language was am ong those spoken in the H unnish Empire. The Huns determ ined the history o f the Eastern European steppe, and subsequently that o f the whole o f Europe, for a hundred years or m ore, and caused m ajor changes in the areas lying north o f the steppe. The best-know n X iongnu find is the set o f royal graves uncovered near N oyon ula in the forest region o f North-M ongolia. W hat is im portant for us is that the nature o f finds in the V olga-K am a region and in the Urals go through a change in the 4th century A D , w ith features appearing w hich are indicative o f large-anim al husbandry and a nom adic w ay o f life. K hazan archaeologists link this era with the first settlem ents o f Turkic people. It is difficult to prove this relation betw een changes and ethnicity. The effect o f the Huns, or rather the changes occurring sim ultaneously w ith the H uns’ appearance m ust, how ever, be ac cepted as a fact. This strongly suggests that w herever the M agyars w ere in the 4th and 5th centuries, they m ay have been affected by the w aves o f H un nish m igration.
209
The neighbours
3. THE EARLY TURKIC PEOPLES
a) The emergence o f the Oghur peoples and the Avars The fam ous and m uch-cited report o f Priskos the R hetor on the events following the death o f A ttila states that the Saraghurs, the Ughors and the Onoghurs sent em issaries to Byzantium , from w hom it w as learned that great changes had taken place on the steppe. The peoples living on the sea coast had driven out the Avars, the Avars had sent the Sabirs into flight, and the Sabirs had attacked the Saraghurs and Ughors who had therefore left their dwelling places and appeared at the entrance to the Caucasus. This all occurred around 463 AD. The new peoples em erged ju st at the place where the Huns had re treated after A ttila’s death. The new arrivals clearly assim ilated the rest o f the Hunnish population. The new people had com e from the east, like other nom adic peoples, and their arrival heralded a new era on the steppe. The “Iranian” era lasted more or less up to the em ergence o f the Huns. This does not m ean that the Alani, for instance, com pletely disappeared with the em ergence o f the Huns. All that can be said is that hegem ony on the steppe after the Huns up to the M ongol invasion was exercised by Turkic peoples o f whom the first w ere the Oghurs or Ughurs or Ughors. W hich o f the three forms o f the nam e is the earliest is a question to which we will return later. The nam es o f the O ghur tribes, Saraghur or Sharaghur (Sara O ghur or Shara Oghur), O noghur (On Oghur), Khuturghur (K hutur Oghur), Uturghur (U tur Oghur) indicate for certain that the nam es o f the peoples w ere r-Turkic, and therefore C huvash-type, with the place o f Oghuz being taken by Oghur. It can only be supposed that khutur is actually a version o f the r-Turkic form tokhur o f the Turkic tokhuz ‘nine’, or that utur is the r-Turkic version utur o f the Turkic otuz ‘thirty’; i.e. that as well as a “ten” O ghur there was a “nine” O ghur and a “thirty” Oghur. Such names o f tribal confederations exist elsewhere: in the Turk inscriptions there is Tokhuz Oghuz, which is one o f the nam es o f the U ighur tribal confederation, and there is Otuz Tatar, the thirty Tatar tribes’ confederation. M ore im portant for us is the question o f where these O ghur Turks cam e from. Around 463, the Oghurs occupied an area which spread from the land be tween the D on and the Volga w est to the Sea o f Azov, i.e. to the M aeotis, and south to the Caucasus. They had come from the east, and w ere followed by the Sabirs, w ho settled by the lower course o f the Volga, and were driven out by the Avars. The location o f the Avars in the 460s should therefore be examined.
210
Relatives and neighbours
The Turks in the 4th century •C .I-’ 1st Turk Khaeharmte 2nd Turk Khgsj’hariate
,..•///
■j
/■
/
<
/'
^ <>~х r-/ \ 4
Ld<-
K h a z a k ste p pe
' Oghur migration
.:‘4 „
l ^ M k im s h
; ?■
Turks against China
V * =: ■~~ ~-c „
11?:
& .“ v------- " i:#f«r
... ....
г"Щ:-yf 'VШ tf' ' ' ■ ? i/ 'f J ‘
У'
;
. is "
/
А « ы .М
^
* * ‘
, l^ s k i
■'■Ч'л
< "v // ^
** " -■*
'
f ...' '
/'
% „ Vunert**'
^ .
1Chinese attack (around 630)
:
’* '
I
■ 4
Figure 51 Early m igrations o f the Turkic peoples
A round 350, the following events occurred on the K hazakh steppe: the Huns set o ff towards Europe, and the Var and Hyonite tribes, driving south, attacked Persia. A new people m igrated northeast from the strategically im portant area, whose nam e has been passed down in the Chinese sources, but has been the subject o f much dispute. Before we address this problem , some attention m ust be given to the fact, discussed in detail later, that the Avars joined forces with the Varhons and the occupants o f m odern M ongolia, the Ruanruan, before they turned up in Europe. The 6 th- and 7th-century Chinese sources give three Chinese characters for the nam es o f the leading Ruanruan tribes, rendered as Yujiulu. R econstruction o f the early Chinese pronunciation o f this yields *ugur(i). The i is here put in brackets because the Chinese transcribed it thus in both cases if there w as only an -r (there may have been another way too) or an -ri in the original. The story o f this tribal nam e can be followed for a long tim e in the Chinese sources, because the tribe, or part o f it, stayed in the east for a long time, and later j oined the Khitai dynasty which came to pow er in the 10th century.
The neighbours
211
Theophilaktus Sim okattes’s History o f the World, com pleted in 638, de scribes in detail the break-up o f the Avar Em pire in the m iddle o f the 6 th century. We will consider only a part o f it here: “Thereafter the (Turk) Khaghan turned to another enterprise, and enslaved the entire Oghor tribe, w hich was one o f the m ost powerful, since it had a great m any m em bers, and was accom plished in the art o f war. In consequence they occupied their lands to the east, beside the River Til, w hich the Turks call M elas.” M elas is a G reek word m eaning “black”, and appears in Herodotus and Strabon as the nam e o f several rivers. So all that we can take from this is that the Turk called the Til river ‘b lack ’. The Chuvash call the R iver Kama the Black Adil (this is the Chuvash form o f the Etil). The name Oghor appears in one ofT heophilaktus’s sources, M enander (who m ust have written his works betw een 582 and 602), in the form Ughur, i.e. in exactly the form which can be reconstructed from the Chinese sources. However, it is highly probable that the Ughur form is secondary, and that the original form o f the name was Oghur, which soon became Ughur, as written in several sources, and, more rarely, Oghor. It is very likely that the name o f the ‘ten tribes’ people also evolved from the form on ‘ten ’ Ughur and becam e Onughur. Theophilaktus Sim okattes continues: “The leaders o f this people [i.e. the Oghors] were called Var and Hun, and the different parts o f the people were accordingly called Var and Hunni. Later, when Justinian took possession o f imperial power [527-565], the Vars and the Hunnis, breaking away from the ancient tribe, fled to Europe and settled there. These people called them selves Avars, and their ruler was honoured with the title khaghan.” There follows the often-quoted part which tells o f how the Barselt, the Onoghur, the Sabir and other Hunnish tribes took such fright from the Vars and the Hunnis, believing them to be the Avars, that they attem pted to win their m ercy with lavish gifts. The Vars and the Hunnis saw how fortunate this name was, and started to call them selves Avar s. But right up to the time o f Theophilaktus Simokattes, one part o f these pseudo-Avars bore the ancient name Var and the other part the name Hunni. Although 5th-6th-century Chinese hold the Ughur to be the leading tribe o f the Ruanruans, and 5th-6th-century Byzantine sources state that the Turk Khaghan conquered the Ughurs, led by the Vars and the Hunnis, they are referring to the same people and partly to the same events. These events started in the 460s. Around 450, the Ruanruan, or Avar, khaghan defeated the Ughurs, who were living in the lands vacated by the westward-fleeing Huns in 350. He placed Avar leaders at their head, called the Vars. The form Var can also be found in Chinese sources as a sign which is nowadays read hua. This highly-cultured Ughur people then started to play a significant part in the Ruanruan Em pire, however, and becam e the leading tribe o f the Empire. Later, a part o f the Ughurs m ounted a rebellion, and the
212
Relatives a n d neighbours
Avars formed an alliance with the Sabirs to put them down. The Sabirs attacked the rebellious Ughurs from the north, and the Avars from the east and perhaps from the southeast. The Ughurs fled w est to the V olga-D on region, below the Caucasus, and in 463 their envoys m ade their appearance in the court o f Byzantium . W hen the Turk ruler turned against the Ughurs after defeating the Ruanruan, they were already living north o f the Caucasus. This is im portant for the story o f the Ungri nam e o f the M agyar people, but also because the O ghur or U ghur people, speaking an r-Turkic, i.e. C huvashtype language, could hitherto only be proven from the sources as being present w est o f the Urals from the 460s. It is also im portant because it sheds more light on the role that the Oghurs played in the history o f the Avars. This group came from the east, from the direction o f the K hazakh steppe, and so it w ould be useful to establish where the Oghurs originally cam e from. Theophilaktus Sim okattes wrote that a city they had established, B akath, had been levelled by an earthquake, and adds that Sogdiana w as also destroyed by earthquake and plague. M any thought that this Bakath m ust have been identical with Bactria, the colony founded by A lexander in the area o f m odern Afghanistan and which flourished until the M iddle Ages. However, Bactria received its nam e from the R iver Bactros. Recently, it was pointed out that the ending -kath is Sogdian, and m eans ‘tow n’. The city has been identified from Arab and Persian sources as being in the region o f the Yaxartes, i.e. the Syr Darya. It is inconceivable that T urkic-M agyar contacts could have begun here, in the earlier land o f the Oghurs. Also, this w ould be in contradiction with historical linguistic considerations.
b) The Sabirs As we have seen, the O ghur tribes fled from the Sabirs, w ho caught up with them in the Volga region. A fter a long struggle, the Sabirs gained hegem ony around 506. We know very little about the Sabirs, or as they are know n in som ew hat later sources, the Savirs. Their nam e is usually linked with the nam e o f Siberia, and it was recently suggested that the original form o f the ethnic nam e w as Syipir or Syepir, linked by others to the nam e o f the Volga B ulghar city and tribe, Suvar or Sovar. From this was derived the first part o f the nam e Savarti Asfali, used for the M agyars, in the w ork o f Constantine Porphyro genitus. The nam e Sapiri, referring to a people, occurs in a Turfan-region Sogdian source. The relationship between the evidence quoted is, at least, ex trem ely uncertain. N othing is known o f the Sabirian language, w hich was probably Turkic, but the few nam es which have been recorded as Sabirian are
213
The neighbours
not sufficient to prove this. In any case, Sabir dom ination did not last long. They were involved in the B yzantine-Persian wars, but the Avars, who ap peared around 555, overthrew their rule and they did not subsequently play any im portant role.
c) The Ruanruan and the European Avars The people w hich appeared in the land o f m odem M ongolia is first m entioned in the chronicles o f 385. Their name is written in several ways in Chinese sources, but the form m ost com m only used is Ruanruan. In the 5th century they established an enormous empire, part o f which was the H ua tribe m entioned above, whose name in the Chinese pronunciation o f the tim e was var. The title khaghan is m entioned in the Chinese sources from the 3rd century onward, and also crops up am ong the titles o f the X ianbei tribes. In 546 A D the Tiele people rose up against the Ruanruan. This ethnic group was very probably the sam e as the Ughur who were involved in the events occurring around 460 recounted above. Prom inent in putting down this uprising was a people which very soon becam e known throughout the steppe as the Turk people. Some sources claim that these Turks were previously the R uanruan’s smiths. The leader o f the Turk tribal alliance asked for the hand o f a Ruanruan princess as reward for victory. W hen this was refused, he attacked the Ruanruan ruler, with Chinese support, in 552. The Turks were victorious, and the Ruanruan ruler com m itted suicide. Som e o f the Ruanruans fled to the Chinese, others withdrew eastward, but m ost o f them took the road west. A Syrian source appended to the work o f Zakarias the Rhetor, o f 555 AD, states that m any nom adic people live north o f the Caucasus, am ong w hich are listed the Avars (Ongur, Oghur, Burghar [Bulghar], Khurturghur, Abar, Khasar, Sarurghur, Dirmir, Bagirsik, Kvaliz, Abdel, Eftaliyt). This im plies that three years after their defeat by the Turk Em pire, the Ruanruan were already in Europe. By 561 the Frankish rulers were struggling with them. The argum ent as to w hether the Ruanruan are identical to the European Avars has now essentially been settled, inasm uch as the question itself is clearly imprecise. Just as the Huns are not the same as the X iongnu, the Avars do not exactly coincide with the Ruanruan. However, it was m entioned above that one branch o f the Ruanruan was called Var, and they were placed at the head o f the Ughurs after subjugating them. The Vars em erged in Europe together with the Ephthalite Huns or Chionites. The nam e o f the Varkhonni tribal alliance— Varkhon (Ouarkhonitai) in M enander— m ay be preserved in the Varkony place names in Hungary. Varkhon or Avar only refers to the ruling
214
Relatives and neighbours
circles o f the tribal alliance, perhaps only to the original holders o f the names. Nevertheless, it is normal practice in the literature to refer to the Ruanruan as “Asian Avars”, which only means that part o f the leading circles o f the two empires was the same. What is new is that the Ughurs had an important part in the history o f the Asian and the formation o f the European Avars (see pp. 261-266).
d) The Turk people in Europe The Turk Em pire was founded by the Turk tribes who rose under the leadership o f Bum in against the Ruanruan. The founder o f the em pire, w ho proclaim ed him self khaghan, died in 553, and was followed for a short w hile by his elder son Kolo, and then by his younger son, M ugan khaghan (553-572), w ho laid the foundations o f the great empire. He subjugated the K hitai, the Ephthalites, and the Khirghiz. The E m peror’s uncle, Ishtem i, led the w estern cam paigns, m ost im portantly the fight for control o f the Silk Route. In 557, an alliance was forged between Ishtemi and the Persian Em peror Husrau A noshirvan against the Ephthalites. A fter jo in t victory, the Turks occupied the area north o f the Am u Darya and no doubt took over the form er lands o f the Oghurs. They also took into their service the Sogdians, who w ere the real controllers o f the silk trade. The Sogdian traders not only received protection from the Turks, but in return contributed to the organisation o f the em pire and the form ation o f its chancellery. Sogdian becam e the official language o f the First Turk Khaghanate, and the language o f docum ents and inscriptions. A con tributory factor m ay have been that Sogdian and Turk populations had already lived together in m any places. The gradual Turkification o f a large section o f the Sogdians had begun. In 568, Sogd em issaries, led by M aniakh, crossed the land o f the A lani and arrived in Byzantium . They held talks in the nam e o f Silzibulos, who must have been khaghan Ishtemi. The Turks offered an alliance against Persia and the “absconded slaves”, the Avars. Already struggling against the continually raiding Avars, and the old enemy, Persia, the Byzantines w ere pleased to receive the offer but did not hurry to reply. They also sent a delegation to the Turks, led by Zem arkhos, partly to conduct talks, but m ore im portantly to gain intelligence. Although the real issue was the security o f the Silk Route, Byzantium had ju st started producing silk out o f cocoons stolen from China. Appraising the diplom atic reports, M enander wrote that the Ughurs lived w est o f the Volga and accepted the suprem acy o f the Turks, who had also gained the subm ission o f the Alani. The local pow er o f the U ghurs may therefore have been retained. The centre o f the W estern Turkic Em pire was alw ays in Western Asia, perhaps in the region o f the Altai, i.e. the Golden
215
The neighbours
M ountain. The title o f the ruler was Yabghu. The W estern Turkic form Jebu or Jevu was the title borne by Ishtemi, and gave rise to the nam e o f the M agyar chieftain Geza, Saint Stephen’s father, whose actual nam e was Jeucha or Jevicha. The current form G iza is due to a later m isinterpretation. W estern Turk authority was o f variable intensity, but during their rule the ethnic com position o f Eastern Europe did not change substantially. This is im portant in view o f archaeological finds in the V olga-K am a region, which show up further changes in the 5th and 6 th centuries. M uch silver was brought here, m ostly from Sasanid Persia, but also from nearly every region from Byzantium to Khwarezm. Some researchers have attem pted to prove the emergence o f new Turkic peoples behind this, linking the m igration with the western expansion o f the Turk Empire and the ethnogenesis o f the Kipchak speakers o f the Volga (Tatars and Bashkirs). This is not supported by evidence, however.
e) The Bulghars At the beginning o f the 600s, the O ghur population regained independence under the leadership o f the Bulghar tribes. The Bulghars are m entioned nearly everywhere, together with, or as identical to, the Onoghimdurs. Theophanes simply calls the Bulghars Unnoghundur-Bulghar, and Porphyrogenitus also writes that the Bulghars “used to” call them selves Onoghundur. Arm enian sources m ention them under the name Olhontor B ulkar and later Vlendur Bulkar. The nam e is w ritten in the H udiidal-alam as Vunundur, by M asudi as Wulundur, in a letter by the K hazar khaghan Joseph as Vununtur. These forms are the only basis for explaining the Hungarian nam e for the Balkan Bulghars, Nandur, and later Nandor. The appearance o f the Bulghars is also noted by the A rm enian Geography, formerly attributed to M ovses Horenaci, and in m ore recent tim es to Ananias o f Shirak. The greater part o f this work was alm ost certainly w ritten before the conquest o f the Arabs, but some parts o f it are end-of-7th-century inser tions. M ention is m ade o f Asparukh, who fled to the w est and settled in the Balkans. This happened around 679 A D . For a long tim e now, the original homeland o f the Bulghars was said to be the region o f the R iver Kuban. This claim, however, needs to be re-examined. The text is not only an Arm enian translation o f Ptolem y’s work, but it has also been am ended w ith data contem porary to the authors. Geography m entions, in the chapter on Thrace, the Danube, which has six tributaries “and which forms a lake, and an island called Piwki (Ptolemy: P e u k i lim ni). On this island lives Asparukh, son o f Khuvrat, a fugitive from the Khazars from the m ountains o f the Bulghars, w ho expelled
216
Relatives and neighbours
the Avar nation and settled there” (H ew sen’s translation). W here should one look for these “m ountains o f the B ulghars”, then? The description o f Sarm atia reads: “The son o f Kuvrat fled from the Hippie (Jiakanri) m ountains.” In Ptolem y the name o f the m ountain is Hippikos. The A rm enian author alludes to a certain K r a k ’s learn. The latter is ‘R aven m ountain’, or K orax in Greek (Pliny Coraxici; the Korax o f Strabon is in Greece). Some authors thought that k r a k ’s is an Arm enian rendering o f the M iddle Persian Pahlavi kurrag ‘fo al’, ‘co lt’; hence the Hippikos and the Korax are the sam e m ountain at the western part o f the Caucasus. This conjecture has for a long time m isled scholars. According to Ptolemy, six rivers run from the Hippie M ountains— the A rm enian version, however, counted only five, w ithout nam ing them. They are tributaries to the M aeotis, that is to the A zovian Sea. The A rm enian G eography next m entions two rivers. One o f them is the Vardanes, later called the Kuban; the other runs into the M aeotis at Anakopia. The Turks and the Bulghars reside to the north o f this river. This also confirms the assum ption that the ‘Hippie M ountains’ could not be the western slopes o f the C au casus— and had to be north o f the Kuban. This is further corroborated by a third m ention o f the ‘Hippie M ountains’. The author o f this part o f the A rm enian Geography writes that the Volga (/ra; in Ptolem y Rha) has two headw aters in the north o f the U nknown Land. These unite near the Hippie M ountains from which a branch o f the Don (Tanais) flows, w hich runs into the G u lf o f M aeotis. The Bulghars are nam ed, according to the A rm enian author, after the rivers: K up’i Bulghars, D uch’i Bulghars, O lhontor B ulghars and C h ’adar Bulghars. These nam es were unknow n to Ptolemy. This Kuphi m ust be identical with the one quoted below, and with the R iver Bug. The Arm enian author described the land from east to west, as did other contem porary authors, thus at the Volga we find the Khazars, w est o f them the M agyars under the name Turk. Further to the w est lived the B ulghars am ong the Don and the Bug. Asparukh fled from the m ountains north o f the Don and the Bug to the west. These m ountains were also called B ulghar M ountains and this region was later called B lack B ulgharia (maure Boulgaria) by C onstan tine. The com m on source used by the 9th-century Theophanes and N icephorus could hardly have been earlier than the 8 th century. A ccording to them Bulgharia lay betw een the Kuphis (in Theophanes, and Kophis in Nicephorus) and the M aeotis. This used to be identified as Kuban, and so K huvrat’s B ulgharia was also placed in the Kuban area. B ut in Porphyrogenitus, C hap ter 42, in the description o f the road from the Danube to Sharkel, we read that Sharkel is 60 days’ journey from the Danube, and the largest rivers in the intervening area are the D niester (Danastris) and the D nieper (Danapris), but there were also sm aller rivers in the region, such as the Singul (or
217
The neighbours
more properly Ingul), the Hibil, the Alm atai, the Kuphis and (or i.e.) the Bug (Bogou). The description o f Etelkoz in C hapter 38 m entions the Kubu (Koubou), w hich is w ithout doubt the nam e o f the Bug there. W hether or not the Kuphis is identical to the Kubu and the Bug, or another river betw een the D niester and the Dnieper, such as the Ingulets, the Kuphis in Theophanes and Nicephorus cannot be the Kuban, but only a river w est o f the Dnieper. This m eans that the Byzantine sources also locate K huvrat’s B ulgharia in the place where K huvrat’s grave has actually been found (see below). This area also appears under the nam e Black Bulgharia. Porphyrogenitus writes, in the same Chapter 42, that the Rus came along the R iver Dnieper, crossed Black B ul gharia, and then w ent to Khazaria and Syria. If there was a B ulgharia prior to the 6 th century, and it was in the Kuban region adjoining the Caucasus, then the Bulghars vacated it at the end o f the 6 th century and were living in the Dnieper area from the beginning o f the 7th century. Recently, the grave o f Khuvrat, the last Bulghar emperor, was successfully identified. The identification was m ade certain by the discovery in the grave o f the E m peror’s signet ring, on which his nam e appears in the form Hovratu (hovratou). The initial h- rendered as a back-sounding k- in the personal name Khuvrat (K ovrat) is very im portant as regards loan words in the Hungarian language (see Figures 52, 53 and 54). The royal grave is in the Ukraine, near M ala Pereshchepino (or in Russian M alaya Pereshchepina), which is 13 kilometres southeast o f Poltava, not far from the River Vorskla, which is a tributary to the left o f the Dnieper.
Figure 52 K huvrat’s grave
218
Relatives and neighbours
\
Harkov
r~* „
/
_ _
* 4
..............................
Senzhari
>4 i
gmepchug
. _ ° r ^ 'n r a ^
„.Г45
wi
L ^ te a e ^ s yAVozrtesenki Zaporozhe
J
'•г
ж г^-^
Khersc &Уг A Kdegey
Figure 53 Royal grave sites in the Ukraine steppe zone (7th century)
la
.в x
x
x
X
2a
у
xov
Xой
Cs p м
Jl
x ov/®6
*
^°u/5
n
P
U JL j
xov/5f WT
P j s
P i b
P § fe
Л
Л
w
X0VI^S n
U
X0U^ WT XovA?wtou
P ., , П P .
J l ii
U Ju ^
ft F
xov/5eI4,Tte,T xou/5f t4T on 7Tt«lTgW X W JTgttJK
J rp » n £ t4 f
Figure 54 Decipherm ent o f the inscription on K huvrat’s signet ring
219
The neighbours
Khuvrat's Bulghar state According to the Byzantine sources, K huvrat rose up against the Avar chief tain in 635, and founded an independent empire. This statem ent is interesting because it is hardly conceivable that the rule o f the Pannonian Avars could have stretched as far as the Kuban. The grave o f K huvrat displays m any p ar allels w ith early age Avar royal graves (Bocsa and Kunbabony). The location hitherto proposed for K huvrat’s short-lived Bulghar Em pire m ust therefore be reviewed. The geographical circum stances o f K huvrat’s D nieper-region em pire are very highly rem iniscent o f the Etelkoz area. N um ism atic data shows that K huvrat m ust have died not long after 650. The sources w rite that he died shortly after the death o f B yzantine Em peror Heraclius in 641, leaving five sons. The em pire was only brought down by the attack o f the K hazars around 670. This set o ff the m igration which culm i nated in the Danube Bulghars under Asparukh conquering their homeland.
Figure 55 K huvrat’s Bulghar Em pire (7th century)
220
Relatives and neighbours
O noghundur Bulghars also em erged in the Avar Em pire, on the shore o f the A driatic and around Ravenna. Two hundred and tw enty years later, the M agyars w ere party to events which w ere very sim ilar in m any respects.
The Volga Bulghars W hereas one part o f the Bulghars m igrated westw ard, and took part with varying success in the history o f the Balkans, the C arpathian Basin and Italy, others, under (Bat) Bayan subm itted to the Khazars. In return, they clearly received some reduced degree o f autonomy. Although it is not know n p re cisely where (Bat) Bayan and his Kotragos peoples, who still featured in the story, w ere located, they m ust originally have been in the eastern h a lf o f the Bulghar Empire. The area inhabited by the Bulghars w ho rem ained in place m ust therefore have been som ew here betw een the D nieper and the Don. N icephorus writes that Kotragos crossed to the far side (clearly the eastern, the left bank) o f the Don. As we shall see, the lands east o f the Don along the B lack Sea coast and both sides o f the Volga were controlled by the Khazars. A t that time, i.e. after 670, the Bulghars thus had no opportunity, nor reason, to m igrate elsewhere. The m om ent for that only came a good sixty years later. During the K hazar-A rab wars, in 723, the Arabs, attacking from the south, succeeded in breaking through the K hazars’ defensive lines on the Caucasus and occupied the form er K hazar capital, Sam andar and Balanjar. The K hazars hurriedly m oved their centre to the w ell-defended Volga delta, and thenceforth the capital city nam ed after the R iver Etil becam e the C ity o f the K han (Hanbalik). The later Caliph M arvan, however, launched a renew ed attack against the Khazars in 737, and the K hazar chief was forced to flee. His route led north along the eastern bank o f the Volga. The Arab arm y follow ed them on the w estern bank o f the Volga, and in a night am bush captured the K hazar Khaghan, who only escaped a worse fate by converting to the Islam ic faith and m aking an open oath o f allegiance to the Caliph. The Arab forces withdrew, and after spending a short tim e supervising the areas north o f the C aucasus, returned home. There is no concrete evidence as to w hether the B ulghars offered any kind o f support to the A rabs against their m asters, the Khazars, during M arvan’s 737 attack, but it was m ost probably at this time that very good relations began to develop betw een the B ulghars and the Arabs, w hich reached their peak in the diplom atic m ission o f 921, described by Ibn Fadlan. A fter the K hazars’ hum iliating defeat by the Arabs and the return o f their khaghan to Etil, a tem porary vacuum em erged on the w estern bank o f the Volga. This encouraged the B ulghars living under the K hazar yoke to cross
221
The neighbours
Figure 56 Bulghar migrati ons at the end o f the 7th century
the narrow D on-V olga elbow and set o ff north along the Volga, with the aim o f gradually freeing them selves from central control. A round 750, they reached the low er course o f the Volga at the bend w here it is joined by the Rivers Sam ara, Kinel and Sok, known as the Sam ara bend. This alm ost peninsula-like area (nowadays actually an island on which the Zhiguli hills are situated) surrounded by the Volga bend, and providing excellent defence, was the central stronghold o f the Volga Bulghars. Because o f the strategic nature o f the area, a city was also built here w hich stood until the 10 th century. Its ruins lie beside the m odern village o f Vali and are known as “M urom Castle” (M urom skoye Gorodishche) in the archaeological literature. Avoiding the reviving Khazar pressure, the Bulghars first proceeded north along the right bank o f the Volga, and set up their next stronghold in the area bounded by the Rivers Kilyna, Sviyaga and Volga, arriving a generation later. The Kilyna has a left tributary called the Bidenga, which is joined on its left side
222
Relatives and neighbours
Figure 57 Bulghars, M agyars and Alani in the K hazar Em pire (7th—9th centuries). The northerly m igration o f the Volga Bulghars
by the Tarhanka, by whose bank the Bolshie Tarhani excavations discovered a Bulghar cem etery identified as dating from ju st this time, the end o f the 8 th century. D ating has been verified by coins found there. One was an Arab dirham from betw een 775 and 809, and the other a K hw arezm Arab coin from between 762 and 787. The Volga Bulghars m ust have crossed the Volga at the level o f the River M ayna, where they reinforced the southern frontier o f their empire, roughly following the valley o f the M ayna and the Cheremshan, and expanded in the east up to the River Shishme. In the north, the Bulghars reached the Kama in the 900s, but only set up expeditionary settlements on its northern bank. These have been dated by coin finds. The two latest, in the Tankeyevka graveyard on the northern bank o f the Kama can be dated to 846 and 892-902 respectively. The Volga Bulghar king had strengthened his relations with K hw arezm by around 900, and becam e a m ajor partner o f the Arab C aliphate which was not
223
The neighbours
prevailing against the Khazars. This resulted in the journey o f Ibn Fadlan which started in 921 (see pp. 68-71). It is difficult to assess w hether there had already been attem pts to convert the Volga Bulghars to Islam, but such missions could have occurred in 737 and thereafter. The sources are contra dictory on this point. However, Ibn Fadlan, who arrived in the court o f the
> ^ < 1 b ars
*........... //
S.. \
M
vfli
Sgl
0 Vv
A \> ЫШ
1
StUi:
"T t A
/fc?
4 ■y
V. 4, >*ч»< О Figure 58 Ibn F adlan’s journey
224
Relatives and neighbours
Bulghar ruler on 12 M ay 922, described in detail that there w ere already people from Baghdad in the royal court (for exam ple a tailor), that there were m uezzins, and that Friday prayers included the title o f “Yiltever, the Bulghar K ing” . A t the talks, the Bulghar King enquired as to how Friday prayers should properly be said. Ibn Fadlan answered, “ in the nam e o f you and your father” . To this the King responded: “M y father was a heathen, and so I do not wish to utter his nam e in the prayer booth.” Thereafter the King, w hose nam e was Alm ish ibn Shelkey, took up the name o f the Caliph, Jafar, and as his father’s nam e, as custom ary for converts to Islam, he chose Abdallah. From then on, the serm on incorporated the prayer: “Oh, m y God, endow w ith well-being your servant Jafar, son o f Abdallah, ruler o f the Bulghars, client o f the lord o f the believers!” O n the basis o f this and other details o f Ibn F adlan’s report, it seem s that Islam is unlikely to have been adopted am ong Volga Bulghars before 900. C ertainly it was only after founding their state, or during it, and partly to counterbalance the Jewish-faith Khazars, that they started to forge links with Islam. This does not m ean that there w ere no earlier attem pts to convert the Volga Bulghars. Ibn al-Nadim , w riting o f Caliph al-M a 'm un (813-833) and his library in 988, m entions a book with the title: R esponses to questions p u t to al-M am un by the B urghar king concerning Islam a n d the unity (in theology). The book, w hich is lost, caused a controversy w hich is probably exaggerated, since it is quite conceivable that, breaking aw ay from the Khazars at the beginning o f the 9th century, the Bulghar K ing had already started to put out feelers in Baghdad. Several tribes can be distinguished within the Volga B ulghar Empire: the Suvars, the Barandjars, the Bersuls and the Esegels. These nam es were recorded in Arab w ritten sources, and so m ay be read several w ays, but certain readings can be excluded and some o f the possible readings are also supported by B yzantine sources. The erroneous Suvaz form o f the ethnic nam e Suvar gave rise to the theory that it was the origin o f the ethnic nam e Chuvash. This would be im possible even if the letter wrongly given a dot in later copies o f Ibn Fadlan (the Arab letter z differs from the Arab letter r by a superposed dot) reflected the proper form. N either can the name Suvar be linked to the ethnic nam e Sabir, even if it is read as Sovar, which is theoretically possible. The B aranjar form, however, can confidently be identified with the city o f B alanjar below the Caucasus w hich was captured by the Arabs in 723. It can only be im agined that the city got its name from this people, who fled from the A rab attack and jo ined up with the Bulghars. The front vow els o f Bersiile, w hich cannot be rendered into Arabic script, are provided by the form bersil from the Turkic inscription o f Terh in Central Asia, and the form Bersilia from B yzantine sources. In contrast, the Tibetan transcriptionp a r-sil has no value as evidence.
The neighbours
225
In an A rm enian source (a later extract from Pseudo M ovses Horenaci), it is w ritten that Khatun, the wife o f the K hazar khaghan, cam e from the Bersil (Barsilk) people. The reading asagal, appearing also in the H udiidal-alam , is written in tw o different ways by Ibn Rusta: in one place the w ord contains the Arabic letter gain, and in the other the A rabic letter k e f After the 12th century, a custom em erged in Arab rendering o f foreign, e.g. Turkic, words w hich used gain in back w ords and k e f in front words. However, this was not the case in the 11th century. In the great Arab dictionary com pleted around 1072 by Khashghari, the Arab letters are not used in this w ay either, and so later spelling rules should not be applied retrospectively to this period. For Persian authors, the k e f could m ean /к/ or /g/, but Arab authors did not recognise the fro n t/g / (it exists only in some A rabic dialects in place o f the A rabic /j/). Early Arab authors substituted the foreign g with k e f or gain, the sym bol for the back-sounding g, and so from Ibn R usta’s k e f and gain it is possible to suppose the pronunciation g if the transcription is to be taken seriously. The transcrip tion by the Persian Gardizi and the k e f in the H udud al-alam do not decide whether к or g was involved or w hether the sound o f the Turkic nam e was back or front. The two possible readings are therefore asagal and esegel. We have decided on the reading esegel, because the chronicles o f the Chinese Tang dynasty list the Turk tribes in relation to the events o f 651, and there is a name appearing twice, o f which the m ost likely reconstruction is *asegel. The reason for this slightly extended treatm ent o f a Volga B ulghar tribal name is that it is frequently linked with the nam e o f the H ungarian-speaking Szekely people living in Transylvania (Rom ania). This runs into serious phonetic difficulties, however, and is therefore inadmissible. The nam e o f the Bulghar king is given by Ibn Fadlan as Alm .sh ibn Sh.lky. The last vowel o f the personal nam e is not written, and so can be read as either Almish or Alm ush, o f which the latter becam e popular because o f the nam e o f the M agyar chieftain Almos. In Ibn R usta’s m anuscript it consists о f Alm .sh. The A lm ush form w hich circulated at large was the extension by the publisher o f the Arab text. C om parison with Turkic data effectively narrows the reading down to Alm ish. Its literal m eaning is ‘the taken’ from al- and ‘to take’. The expression el almish ‘to take or acquire an em pire’ also occurs as a Turkic name (read as II almish). The only rem aining problem regarding the Bulghar nam e has since been solved, since in a Chuvash-type language the Com m on Turkic /sh/ phonem e should regularly correspond to /1/. Hence, the *Almil form should be expected. But there is no III in the m odern C huvash form o f the Turkic w ordjem ish ‘fruit’. Instead there is /sy/, and the w ord has the form syimesy. The Chuvash form and the Hungarian gyiimolcs ‘fruit’ derive via an interm ediate form *jemich from the form *jemish. W hether the nam e con tained A lm ich or Alm ish, the Arab author wrote it as Almish.
226
Relatives and neighbours
The nam e o f Alm os, father o f Arpad, occurs in two m utually independent sources: as Alm uch (Almouts) in Porphyrogenitus and Alm ush in the H ungar ian chronicler Anonym us. The nam e later occurs with Prince A lm os, broth er o f K ing Colom an (died in 1127), and so it was certainly a living nam e in the house o f Arpad. It is, o f course, quite im possible to conceive o f a rela tion betw een the Volga B ulghar ruler, even via legend, and the ancestor o f the M agyar house o f Arpad. The nam e A lm ish was know n on the Eastern European steppe, however. A nonym us’s explanation that the nam e o f Alm os derives from the M agyar word alom, ‘dream ’, is a folk etym ological invention typical o f the Nam eless One. N ot a single vowel o f the nam e o f the Volga B ulghar ru ler’s father is given. The initial sh- is highly suggestive, however. W hatever the reading o f the word is (and the first syllable can only be i or long a ), the w ord is C huvash-type. Also typically C huvash is the ru ler’s title, yiltever, also read in the form yiltaw ar, since it is w ritten with the “em phatic” A rabic t-, w hich was later actually used in transcribing back-sounding foreign words. This occurs in Turkic languages in the form elteber, and also occurs in K hazar titles. Such was the title o f Elteber Alp, who m arried the daughter o f the K hazar K haghan around 680. This title was bestow ed on office-bearers w ho exercised pow er over foreign subjects, and contrary to appearances was probably not originally a Turkic word. The Bulghar w ord for ‘birch’ is given in Ibn Fadlan as hazing (with the z pronounced interdentally). In O ld Turkic the equivalent is kading w hich subsequently changed to kaying, and is huran in m odern Chuvash. This m eans that the changes к > h and d > z w ere com pleted in Volga B ulghar by the beginning o f the 10th century, and as w e saw in K huvrat’s nam e, the к > h change had already begun in the 7th century. The nam e o f a typical drink o f the V olga-K am a region, honey beer, was noted by Ibn Fadlan as suchii. This occurs in the other Turkic languages in the form siichig, w hich im plies that the ending -g had already disappeared by the beginning o f the 10 th century. The goblet from w hich it was drunk was called sahrah. This is the C huvash form o f the Turkic sagrak. So it is evident that the term inal gutturals had already becom e fricatives. This evidence proves beyond doubt that the Volga B ulghars spoke a C huvash-type language at the beginning o f the 10th century. Their language is attested to by grave inscriptions w hich have been found in the area o f Volga B ulgharia dating from the years 1281 to the 1350s. It is not certain that this language was the ancestor o f Chuvash, but it was very close to it. It cannot, o f course, be denied for certain that the Suvars, the B aranjars and the Esegels, or certain sections o f them, spoke languages other than C huvash Turkic. Various Turkic groups, traders, refugees and accom panying peoples
221
The neighbours
m ay also have lived am ong the Bulghars. The Rus o f K iev and Slav traders had already m ade their appearances by this time. The Volga B ulghars had links to the great trading operations o f Eurasia.
The Danube Bulghars One o f the figures in the story o f K huvrat’s five sons, A sparukh, crossed the lower course o f the Danube around 679, and occupied the right bank o f the Danube and the coast. He established a new em pire centred on present Pliska, in which a large upper Turkic class ruled over the local Slav population. Asparukh died around 701 and his successor, Tervel, com pleted the organisa tion o f the empire. They soon interfered in internal Byzantine pow er struggles, and then joined with the B yzantine side in beating o ff A rab attacks on Constantinople in 717-718. In 739, the last o f the Dulo dynasty, w hich had founded the state, died. It is possible that this clan nam e is linked to the nam e o f the chieftain Dula m entioned in the Hungarian Chronicle, but cannot have anything to do w ith the title Gyula and the nam e that derived from it. The new rulers were drawn from the Vokil dynasty. Various rulers follow ed each other with rapid succession on the Bulghar throne betw een 756 and 772, no doubt as a result o f bloody internal wars. In 777, a king by the nam e o f Telerig lost out in the internal struggle and fled to Byzantium w here he converted to Christianity. However, Telerig was unable to settle the internal affairs o f the Bulghars even w ith Byzantine support. This was only achieved by Krum , who belonged to the Pannonia branch o f the O noghundur Bulghars, and acceded to the throne with their assistance. Krum quickly turned against the Avars, dealing them a heavy defeat, and then after securing his northern borders turned against Byzantium , conducting long fights against Em peror N icepho rus. Although the fortunes o f w ar initially favoured the Byzantines, who took Pliska on several occasions up to 8 11, ultim ate victory was reaped by Krum, who captured Nicephorus, executed him, and follow ing nom ad custom had his skull m ade into a drinking cup. The advance o f the B ulghars was halted by a reorganised Byzantine em pire under M ichael I. K rum ’s legacy upon his death in 814 was a Bulgharia in full strength. A fter b rief internal struggles and a few interm ediate rulers, K rum ’s son O m urtagh (814-831) consolidated his grip on power. He m ade peace w ith Byzantium , and set about building up his country. The conversion o f the Bulghars gradually com m enced, although resisted by O m urtagh him self, but some o f his successors took up the faith, clearly in search o f support in the pow er struggle. For a while, however, vic tory was won by resisters to conversion, M alam ir and his successor Persyan (or Presyam).
228
Relatives and neighbours
Boris came to the throne in 852, and until his death in 889 was a party to the events which directly preceded the Hungarian Conquest, and so the story o f Bulgharia under the rule o f Boris and Sim eon (893-927) will be discussed along with the events o f the Conquest. The Danube Bulghars used the Greek language and script in their records, and those which have survived also exist in the form o f inscriptions. The rest have been lost or passed into larger collections, but som e w ere subsequently translated into Old Church Slavic after C hristianity had been adopted and the Slavonic language had becom e prevalent. The surviving later copies o f these early texts, one o f which is the list o f Bulghar kings already discussed (see pp. 61-62), are the culm ination o f a com plicated heritage. F or a long tim e, the Bulgharian state retained a Turkic organisation, which gradually adopted elem ents borrow ed from Byzantium . Some o f the titles have still not been deciphered, but kavhan, boyla, khan, kolovur, bagatur, tarkhan and sam psi are Turkic in origin, or were at least used by Turkic peoples. In the m eantim e, the leading Turkic class gradually dim inished, and becam e assim ilated in to the Slav population. By the end o f the 7th century, bilinguality was w ide spread. O m urtag’s three sons (also) had Slavonic nam es (Voyin, Zvinitsa and M alamir). The linguistic effect o f T urkic-Slav coexistence is also reflected in inscriptions. Am ong the titles o f court office bearers was, for exam ple, the ichirgu boyla, the ‘internal boyla’ o f which parallels are attested in Turkic sources o f Central Asia. From a Cyrillic inscription o f around 969 it is known that at the end o f the era this would be approxim ately uttered as chregubilya. The second elem ent o f the title is the origin o f the Hungarian nam e Bela, used by several kings o f the Arpad dynasty.
f ) The Khazars Researchers into the K hazar issue were for a long time inhibited by the assum ption that the form Khazar was original. The discovery and publication o f Terh and Tez U ighur inscriptions revealed that the original Turkic nam e of the Khazars was Khasar. This m ade other evidence w hich had hitherto been difficult or im possible to interpret accessible, and m ore could be put in correct chronological order. The K hasar ethnic nam e ultim ately stem s from the title Caesar, which reached the M iddle Persian sources, am ong others. In 739, a ruler o f the Turkic dynasty in Gandhara abdicated in favour o f his son. The Chinese sources w hich record this and the coins issued by the new king both give the kin g ’s nam e as From Kesar. The From is the Iranian form o f Rome, and the K esar is the title Caesar. This nam e passed over to the Tibetans, who knew it in the form Phrom Ge-sar. This “Rom an” Ge-sar also appeared in the
The neighbours
229
Tibetan sources as “Turk”, i.e. Dru-gu Ge-sar, and after a w hile becam e the hero o f a Tibetan epic, w hich now has a long and w idespread heritage in Tibet and its surroundings. This Tibetan epic was borrow ed by the M ongols and survived am ong them as the Geser Epic. The K esar title also passed over to the Turkic-speaking population. As a result o f the stressed second syllable in Turkic, K esar becam e Khasar, and the title survived as both an ethnic and a personal name. The m ost prom inent o f the personages w ho bore the nam e was Chingis K han’s brother. There are not m any examples o f ethnic groups’ nam es deriving from a title, but am ong them are the nam e K erel for the M agyars and the Turkic Yabghu, which derives from the title Yabghu. This type o f ethnic name evolved from the expressions “K hasar’s people” and “Y abghu’s people” . Another barrier to research into K hazar origins has also been lifted. There was a view that if the K hazar people were self-designated, the language im plied by the m id-word -z- could not be Chuvash type. Quite apart from the fact that ethnic names say nothing about the peoples them selves, it is now known that the -z- in the name o f the Khazars is secondary. Chinese sources also list a tribe by the name o f K hasar am ong the Central Asian U ighur tribes. There is some debate as to how they are related. This Uighur tribe only appears in the Chinese sources in the period following the middle o f the 8 th century, and so it cannot be ruled out that a group o f European Khazars m igrated east— perhaps at the time o f M arvan’s attack in 737— and joined forces with the Uighurs who took over pow er from the Turks after 750. If this was the case, then the Uighurs could not have been the ancestors o f the Khazars, as some thought. It also conflicts with the view that links the Khazars with the Akatirs. The original form o f the ethnic nam e that appears in Priskos is akatir, but after the Latin phonem ic change ti > chi, copyists wrote it as akatziroi. The m ore reliable m anuscripts give it as akatir. This only later became akatsir in Jordanes and Cassiodorus. The popularity o f the form akatsir is due to the Turkic etym ology produced for it (agach eri, ‘people of the w ood’). The m ost im portant evidence, hitherto ignored precisely because o f its reading, is an appendix already quoted: the Syrian geographical description written for the ecclesiastical history o f Zakarius the R hetor also lists the K hasar people am ong the peoples living north o f the Caucasus. As we have seen, this dates from 555. M ichael the Syrian wrote the history o f three peoples in his chronicle. The ancestors o f the three peoples are three brothers, called Hazarig, Bulgharios and Pugur. The form o f the third nam e is a produce o f text corruption, but can be read correctly as Wugur, one form o f the ethnic nam e Oghur. Bulgharios and his people m igrated to the Danube, and the other tw o peoples w ent to the land o f the Alani, which they called Bersilia. M ichael the Syrian wrote his
230
Relatives and neighbours
chronicle before 1199, but adapted old B yzantine and M iddle Persian sources. W hat is im portant for us is that it refers to events around 670, in w hich the K hazars and the Ughurs appear together in the foundation o f an empire. The K hazars’ first prom inent historical appearance can be put at around 620. The Byzantine Em peror H eraclius (610-641) secured their alliance for a cam paign against the Persians. The Khazars assem bled 40,000 soldiers for the siege o f Tiflis in 627. The ruler o f the Khazars at that tim e bore the title je b g u or je b u , versions o f jabgu. Their m ajor cities were B alanjar and Sam andar, below the Caucasus. A fter the collapse o f Western Turk pow er (about 630) the K hazars finally secured their independence and the K hazar ruler adopted the title Khaghan. The first known m ention o f this is in 652-653. This m ore or less coincided with the death o f Khuvrat. B etw een 650 and 670, the K hazars abolished the Bulghar Empire. As was discussed earlier, one part o f the Bulghars w ent west and the other subm itted. The w rit o f the Khazars at that time ran up to the Crim ean Peninsula, w here the Byzantines m anaged to retain the city o f Kherson in the m iddle o f the 7th century. In 695, the B yzantine Em peror Justinian II was banished to Kherson, w here he m ar ried the K hazar K haghan’s elder sister, who was baptised as Theodora. The K hazars did not offer Justinian support, however, although he ultim ately regained his throne with Bulghar assistance (705-711). The leader o f an uprising against him in Kherson, Philippikos Bardanes, was assisted by K hazar forces, and in 711 put an end to Justinian’s second reign. A fter their conquest o f Persia, the Arabs advanced further to the east and also landed at Gibraltar to invade the Iberian Peninsula, which they m arched right through, only to be halted by Charles M artel near Poitiers in 732, as we saw (p. 6 6 ). In 737, they defeated the Ttirgesh in C entral Asia, and in 751, in the battle o f Talas Valley the Chinese and their allies, although here their advance was finally halted. The Arabs soon grouped below C onstantinople, and attacked the Khazars via the Caucasus. The Khazars played an im portant part in averting this threat to the whole o f Europe. They won a m ajor battle against the Arabs near A rdabil in 730 and forged an alliance with Byzantium . In 732, the son o f Em peror Leo III o f Byzantium , Constantine, m arried the daughter o f the K hazar K haghan, who was called by the Turkic name Chichek, m eaning ‘flow er ’, but in Christendom she received the nam e Irene. Their son Leo IV (775-780) was also know n as “Leo the K hazar” for this reason. The B yzantine-K hazar alliance did not endure, however. We saw above that the A rabs m anaged to force the K hazar K haghan to take up the Islam ic faith for a short tim e in 737. In 760, the Arabs m ade peace with the Khazars, and this tim e they sealed the alliance with a m arriage. B ut it seems that the Arabs only wanted a tem porary peace, because after the death o f the K hazar princess, K hazar-A rab fighting broke out anew
The neighbours
231
and lasted until 764. The last K hazar attack against Arab Transcaucasia was in 798. In the m eantim e, the Khazars extended their pow er to em brace the Ostrogoths o f the Crim ea, but ceded them considerable independence under their own toparch. The sources are very poor regarding the subsequent period. It is known that at the request o f the Khazars, under the direction o f Petronas, the Fortress at Sharkel was built on the low er course o f the Don, in the area o f m odern Tsimlyansk. The Russians, under Kiev leadership, launched m ilitary expeditions in the middle o f the 9th century to police the trade routes to the Caspian Lake. These proceeded along the bank o f the Volga but probably only m et the Volga south o f the Sam ara Bend. Such m ilitary expeditions are know n to have taken place in 860, 8 8 0 ,9 0 9 -9 1 0 and 911-912. Some o f the expeditions w ere m ounted in alliance with the Khazars, or at least with their knowledge. A fter large quantities o f booty had been obtained, however, relations betw een the Princi pality o f K iev and the Khazars deteriorated. D espite the K hazars’ em ploym ent o f K hw arezm ian m ercenaries, they were only able to hold out for a short time against a decisive attack by Svyatoslav in 965. By the end o f the century, the K hazar Em pire had gradually disintegrated. The religion o f the Khazars in the 7th century was Tengrism. This was described quite thoroughly by the bishop o f the Caucasian Albanians, Israel, who visited the “North Caucasian H uns” in 681. The description has survived in the text o f M oses o f Dashuranci. The K hazars soon converted to the Jewish faith, however. The political reasons for this were clear: choosing a third religion betw een the Islam o f the A rabs and the C hristianity o f Byzantium enabled them to avoid becom ing dependent on either. M any details o f the conversion and the nature o f the K hazars’ Judaism were until recently dis puted, however. In contrast with the widespread view that Judaism was not a proselytising faith, there are exam ples o f it being adopted by various ethnic groups in certain historical periods. In the period around the birth o f Christ, the royal house and people o f the small Syrian kingdom A diabene (or Edayab in the Syrian sources), in the vicinity o f m odem M osul, converted to Judaism to secure their independence from both the Romans and the Parthians. In 6 th-century Yemen, the royal house converted to Judaism for political considerations, in order to distance them selves from Christian Ethiopia. A t this time, on the larger scale, this im plied opposition to B yzantium and friendliness in relation to Persia. Towards the end o f the 1st m illennium , two m ovem ents em erged w ithin Jewry. One, the “ Spanish line”, rigidly opposed conversion, w hereas the other, which spread through France and Germany, and as far as Kiev, was m ore favourably disposed to it. This dispute also influenced the decision o f the Khazars.
232
Relatives and neighbours
The K hazars’ conversion proceeded in several stages. The earliest source (M asudi) claim s that the first conversion took place during the tim e o f Caliph Harun al-Rashid (786-809). According to the K hazars’ Jew ish tradition, the first converted king was Bulan, but the faith only started to spread after his third successor, Obadiya, had “revived the em pire and strengthened the true faith” around 800. Bulan ruled around 730, and it is highly im probable that any kind o f official conversion took place at this time, or if it did, then only for a very short period. In his w ork w ritten in 864 in W estphalia, D ruthm ar o f A quitain noted that “all o f the Gazars follow Judaism .” This is, in fact, one o f the m ost disputed issues. In the initial period, Judaism started to spread am ong the upper circles o f the Khazars. The view o f the Arab sources which discuss the issue in detail is quite straightforw ard. If very short reports on the K hazars are discounted, then it clearly em erges that the K hazar Khaghan, nobles and high officials follow ed the Jew ish faith. “The religion o f the rem ainder is sim ilar to that o f the Turks” , wrote Ibn Rusta. The Jew ish nobility w as joined by im m igrants and refugees. A t this time, there was probably a Jew ish com m unity living in the Crim ea, in Phanagoria (according to Theophanes, or in Sam kars according to Ibn al Fakih), and a sm aller diaspora in the North Caucasus. Jews arrived from M uslim cities, and there was a continuous stream fleeing sporadic eruptions o f Jew ish oppression in Byzantium and Persia. Ibn Fadlan w rote that the M uslim s dem olished a synagogue in D ar al-B abunaj, and in return the K hazar K haghan destroyed the capital city’s m osque and had the m uezzin killed. The flow o f refugees stepped up in 943, when R om anus I Lacapenus, the de fa c to B yzantine ruler (920-944) started to forcibly convert the Jews. A fter official conversion in the 9th century, then, the K hazar Jew ish population had a diverse com position but was still very sm all in num ber com pared to the population as a whole. There were two basic strands within the Jewish religion o f the time. There was the K araim (“the literate”, plural, singular karai, giving the nam e K araite) w hich only recognised the five books o f M oses, and R abbinic Judaism , w hich also follow ed later holy scriptures, the Talmud. A lthough it is possible that there were sm all K araim com m unities in K hazaria, the religion adopted by the ruling K hazar circles was R abbinic Judaism . From historical and linguistic evidence, it is unlikely that the sm all Turkic-speaking K araim Jew ish ethnic group living in m odem Poland and Lithuania, o f w hich one branch also lived in the Crim ea, are the descendants o f the Khazars. A t m ost, it is conceivable that the sm aller Karaite com m unity w hich lived in K hazaria gained the K ipchak type Turkic language, that they speak today, through an exchange o f language.
The neighbours
233
From M asudi, it is known that M uslim s, Christians, Jews and pagans lived together in the K hazar capital. The sources give a detailed description o f the K hazars’ “double kingship” . The king o f the Khazars was held to be holy and inaccessible, and responsible for the flourishing o f the em pire through his m ystical power, i.e. his khut (see pp. 149 and 150 for m ore on this). Real affairs w ere directed by the m ili tary leader, whose title was K haghan-beg, beg, shad, or yilig. A ccording to Ibn Fadlan, the H akhan-beg was followed by the Kiindii (m isw ritten as Kundiir) H akhan and the Jaushigir. The titles Kiindii, Beg and Yilig are also found am ong the M agyars o f the Conquest, partly as titles and partly as names. The intricacies o f double kingship and its custom s as they stood in the second h alf o f the 9th and the beginning o f the 10th centuries are described by M uslim authors such as Jayhani, Ibn Rusta, Gardezi and Ibn Fadlan. How the double kingship, which m ust have had a bearing on the spiritual world as a whole, and the K hazars’ Jewish faith regarded each other is a question which prompts some speculation. A review o f the evidence shows that the official adoption o f the Jew ish faith around 800 did not affect in the slightest the double kingship and its associated rituals. A different view is obtained, naturally enough, from Jewish sources, particularly a letter from K haghan Joseph. This was w ritten around 960, however, and clearly represents the attitude o f Jewish parties in the court. The Khazars certainly received a pilgrim age from Cyril and M ethodius in 861, obviously with a proselytising aim. According to the legend o f Cyril, the chief adviser to the K hazar king was a Jew (w ho according to the m issionaries was “m inded to adopt Christianity”) but he also had a M uslim adviser. It may be concluded from the Cyril m ission’s lack o f any great success that the Judaisation o f the ruling circles was com plete by around 860. W hereas there is considerable literature on the K hazars’ Jewish faith, less attention has been paid to the spread o f Christianity. C hristianity had consid erable influence in the K hazar Empire from the 7th century onwards. This was partly, but not exclusively, due to the proselytising activities o f the Christian Albanians o f Caucasia. There lived Christian com m unities in the Crim ea, the home o f the “Onoghur bishopric” , am ong them Ostrogoths. It was at this time that the iconoclasm dispute was raging, and the m any people from the factions who lost out, especially from am ong the iconodules who had opposed the iconoclasts, fled to the Crim ean Peninsula and to the safety o f K hazaria, which controlled the C rim ea. As m entioned above, C onstantinus (C yril) and Methodius also attem pted to spread the faith am ong the Khazars. According to legend, two hundred K hazars w ere baptised on hearing the brothers preach. It is not im possible, however, that C onstantinus m ade translations into K hazar during his K hazar studies.
234
Relatives and neighbours
The K hw arezm ians, who assum ed an increasing role in K hazar adm inistra tion, retained their Islamic faith. Ibn Fadlan gives the title o f their leading figures as hazi (h.z. in the m anuscript), w hich is the K hazar version, w ith initial /?-, o f the Arabic kadzi ‘cadi (judge)’. These K hw arezm ian M uslim s later em erged in Hungary under the nam e Kaliz.
g) The Pechenegs The origins o f the Pechenegs and their relationship w ith the K angars need not concern us for the m om ent (see pp. 416-421). W hen considering the conquer ing M agyars, only the situation at the beginning o f the 9th century is important. At this time the Oghuz lived in the vicinity o f Lake Aral, the K im eks beside the R iver Irtysh and the Kharlukhs som ewhere in the area o f the Rivers Talas and Chu. The K harlukhs m oved w est after their defeat by the Uighurs near the source o f the Selenga in 755. Alliances had been m ade, and wars fought, betw een these four peoples in the past, but at this time there was a K im ek-K harlukh-O ghuz coalition against the Pechenegs. The fight was clearly for the strategic area betw een Lake Aral and the Urals through which the nom adic peoples proceeded westward. Quite substantial records have survived o f the Kim eks. They were, like all peoples, a mixture. The tribal division described by Gardizi m entions seven tribes: the Im i, the Im ek, the Tatar, the Blandr, the K ifchak, the Lanikaz, and \hzA jla d . According to the H ududal-alam , eleven chieftains inherited control over the regions from their fathers. The origin o f the Im ek or Yimek tribal name is identical to that o f the K im ek tribal name: disappearance o f the initial k- is a peculiarity o f the K ipchak language. In reality, the K ipchak tribe was also a m em ber o f the alliance. The Lanikaz was probably the Alan-i-kaz(ar), i.e. the “K hazar” Alani, and blandr was interpreted as Bayandur, but it could also have been blandur, i.e. the Persian version o f vlandur, derived from the name o f the Onoghundurs. There are sim ilar descriptions o f the K harlukhs and the Oghuz. The H udud al-alam also writes o f the Kim eks that, w hen they were at peace with the Oghuz, they visited them in w inter (in the south). Gardizi also describes the route to the Pechenegs. He set o ff from Giirgench (now Urgench), passed the Khwarezm m ountains and Lake Aral, and crossed a dry desert where for nine days he came upon a well daily or tw ice daily. Here, he had to climb down a rope to draw w ater for the animals. On the tenth day the traveller reached springs, and then the river. Here lived all kinds of birds and wild animals, but little grass. This country took sixteen days to cross, and the traveller arrived at the Pechenegs on the seventeenth day. The route was alm ost identical to that taken by Ibn Fadlan.
The neighbours
235
The desert m ust have been the Kharakhum , and the river the Irgiz. The Sakaliba (Saklab) m entioned by Gardizi could only have been the Volga Bulghars, as the nam e o f the people was also used by Ibn Fadlan. This is consistent with w hat was written by Porphyrogenitus on the hom e o f the Pechenegs which was originally beside the R iver Etil and beside the River Yeyik [Ural], and shared borders with the Khazars and the so-called Uz [Oghuz] (see p. 237). In spring 922, after leaving the Oghuz, Ibn Fadlan visited a Pecheneg community. He describes them as very poor in com parison with the Oghuz. He only stayed with them for one day, and his next stop was at the R iver Yeyik (Ural), which he crossed to reach the Bashjird. The Pechenegs are also located here by an U ighur em issary’s report, o f which an extract survives in a Tibetan translation, giving the nam e o f the Pechenegs in the form be-cha-nag. These four independent sources indicate where the Pechenegs lived, the place that m ight be term ed their Urheimat. The date was naturally not 922, but earlier; however, but some Pecheneg tribes rem ained in the area— and there are sources attesting to this— after the m ajority m igrated west. Such major operations led to factions dissatisfied with the leadership breaking away, just as happened am ong the Avars and the M agyars. This is w hat caused Ibn Fadlan in 922 to find a very poor “rem nant” group, and indeed the local Pechenegs succum bed to Oghuz control around that time, or a little later. Porphyrogenitus also writes that right up until his own tim e there were some Pechenegs living under Oghuz rule. He writes that when the Pechenegs were banished from their original hom eland, “some o f them o f their own wish and by personal decision stayed behind, joining with the so-called Uz, and live among them to this day”. There follows a description o f their clothing, w hich was cut short as a rem inder that they had cut them selves o ff from their own people. These were the Pechenegs who rem ained in their old lands. Khashghari also talks o f two Pecheneg tribes. One lived near Byzantium , and the other was an Oghuz tribe. The Huditd al-alam also distinguishes two kinds o f Pechenegs, calling one o f them ‘T urk’, and the other ‘K hazarian’. Description o f the Pechenegs continued with Gardizi. The length o f the Pechenegs’ country was a long thirty-day journey, and was bounded by the Kipchak to the north, the Khazars to the southwest, the Ghuzz (Oghuz) to the east and the Sakaliba people to the (north)west. These neighbours had been manipulated by the Pechenegs into fighting each other and taking prisoners who w ere subsequently sold into slavery by the Pechenegs. These Pechenegs were very rich, and had m any horses and sheep. They also had a great quantity of gold and silver vessels and m any weapons; they wore m uch silver and had
Figure 59 Central Asia in the 9th-10th centuries
236 Relatives and neighbours
The neighbours
237
flags and lance-pennants which they rose as they charged and horns m ade from ox-hom s w hich they sounded in battle. As has been proven by several authors, this description is o f the Pechenegs in their new country and in full power. The slaves could only have been sold at one o f the ports o f the Black Sea. Porphyrogenitus on several occasions tells o f where the Pechenegs lived. As we m entioned above (p. 235), C hapter 37 o f D e adm inistrando imperio, covering the Pechenegs, starts thus: “Originally, the Pechenegs had their dwelling on the R iver Atil, and likewise on the R iver Gei'ch [Yeyik, i. e. Ural], having com m on frontiers with the Khazars and the so-called U zes” , and continues: “B ut fifty years ago the so-called Uzes m ade com m on cause with the Khazars and joined battle with the Pechenegs and prevailed over them and expelled them from their country, which the so-called Uzes have occupied till this day. The Pechenegs fled and wandered around, casting about for a place for their settlem ent; and when they reached the land which they now possess and found the Turks living in it, they defeated them in battle and expelled and cast them out, and settled in it, and have been m asters o f this country, as has been said, for forty-five years to this day” (translated by Jenkins). There follows a description o f the provinces (thema) inhabited by the Pecheneg tribes. Four tribes lived across the R iver D nieper towards the east and north, looking towards Uzia, Khazaria, Alania, K herson and the rest o f the regions. The other four tribes lived on the D nieper to the w est and north. One o f these provinces was Yazikapan, close to Bulgharia, another, Low er Yula, lay close to Turkia, and a third, Yavdiertim, adjoined the lands o f peoples who paid tribute to Russia, the Ultin, the Dervlenin, the Lenzenin and the rest o f the Slavs. The description devotes less detail to the lands o f the w estern Pecheneg tribes than to those in the east. He then states that the country o f the Pechenegs (Patzinakia) lies a distance o f five days’ journey from Uzia and K hazaria, six days from A lania, ten days from M ordia, one day from Russia, four days from Turkia, h a lf a day from Bulgharia, is very close to Kherson, and even closer to the Bosphorus. The Em peror defines the location o f the Pechenegs som e what differently in C hapter 38, w hich deals w ith the M agyars. Here he writes: “The place o f the Pechenegs, in which at that time the Turks lived, is nam ed after the local rivers. The rivers are these: the first is that called Baruh, the second that called Kuvu (Koubou), the third river that called Trullos, the fourth that called Brutos, and the fifth river that called Seretos” (translated by Jenkins). I f the five rivers have been correctly identified, they are the Dnieper, Bug, Dniester, Prut and Seret. This is therefore the area inhabited by the second four Pecheneg tribes, only defined in term s o f the rivers. The difference between the two passages is that in the first the Em peror gives the entire Pecheneg area, and in the second the hom e o f the Pecheneg tribes w hich were
238
Relatives and neighbours
earlier held by the M agyars. This was m ade clear by the Em peror him self w hen he wrote after the earlier “K angar-Pecheneg attack” that the Turk arm y split into two follow ing its defeat. “One part w ent eastw ards and settled in the region o f Persia, they to this day are called by the ancient denom ination o f the Turks ‘Savarti A sfali’, but the other part together with their voivode and ch ief Levedias, settled in the western region [see below p. 421 ], in the places called Etelkiizii (Atelkouzou), in which places the nation o f the Pechenegs now live” (translated by Jenkins). In C hapter 40, w hich returns to the history o f the M agyars after the K havar story, the Em peror states that “returning from the attack on the B olghars, the Turks found that the Pechenegs had destroyed their fam ilies and banished the remainder, and so they settled on the land w here they lived in the tim e o f the Em peror” . Later he remarks: “The place w here the Turks had dw elt earlier is called after the two rivers which cross it, the Etel and the Ktizii (Etel kai K ouzou).” This is an insertion w hich does not fit the text. In Porphyrogenitus’s work, the dw elling place o f the Pechenegs and its precise borders are given on several occasions, in different parts o f the work, and there can be no un certainty as to where they lived around 945, nor can it be doubted that the w estern h a lf o f this was the sam e Etelkdz which was the hom e o f the M agyars before the Conquest. This area lies betw een the low er course o f the Danube and the Dnieper. The area betw een the D nieper and the Don, w hich was freed by the B ulghars ranging north along the Volga, was also subject to Pecheneg control. The Pechenegs were therefore living in the area betw een the D anube and the Don after their victory over the M agyars. The Pechenegs are also m en tioned as being beside the Don (which is considered to flow into the Volga) in the HudUd al-alam. The question is sim ply whether, betw een the two Pecheneg lands, the old one betw een the Urals and Khazaria, and their new hom e around 950, they tarried at an interm ediate location. The Em peror writes in the chapter on the Pechenegs that they had lived for fifty-five years in the place w here they were living at the time. However, the introductory section states that the m ilitary expedition started out fifty years before. The five-year difference has been explained in various ways, by text degradation in the first or the second place, or by proposing that there were in fact two Pecheneg attacks. B ut it is inconceivable that the first Pecheneg attack took place later (only fifty years previously) than the second (fifty-five years previously), and on the principle o f the sm allest correction, the first place should be understood as fifty-five years, but not necessarily read as such, as m ost have, because this part may have been w ritten five years earlier. That is, we have to im agine that the date o f arrival in the new hom e is only given roughly “som e fifty years ago” in the
239
The neighbours
first place. The fifty-five was entered as the exact figure when the text was edited. Inferences regarding the language o f the Pechenegs can be m ade from the names o f the Pecheneg provinces. Unfortunately, the forms o f these provinces’ names that have achieved currency am ong researchers deviate from those written in the original m anuscript. For instance, the second o f the eastern provinces, Siirukalpei (suroukalpee) was read as Surukiilbey. The third vowel’s alpha was em ended to omicron ypsilon, and then read as ii, because this better fitted the supposed Turkic etymology. The word w ritten as chopon is regularly rendered in the form chaban, and the word ertim in the form erdim, etc. No great significance should be ascribed to the m entioning o f the Pech enegs am ong the Oghuz tribes, since the old tribal alliances were not assem bled on the basis o f linguistic affinity. It is certain from the nam es that the language was Turkic, but not o f the Chuvash-type Turkic. Later-recorded “Kipchak” features cannot be detected am ong them. This could o f course in theory be ascribed to their not yet having evolved. W hatever the case, a new analysis o f the Pecheneg province names will yield interesting results. Overall, it can be said o f the Pecheneg story that the contents o f the Pech eneg chapter in the work D e adm inistrando imperio m ust be regarded as authentic, and an explanation m ust be found as to why there are conflicting facts in the chapter on the M agyars. W hat can be definitely stated here is that the Pechenegs m igrated w est from their eastern hom e straddling the Urals, and that this m igration ended with the expulsion o f the M agyars from Etelkoz in 895.
4. THE SLAVONIC PEOPLES The long prehistory o f the proto-Slavonic speakers who broke aw ay from the Baltic-Slav population can for our present purposes be ignored. The individual features o f the three great Slavonic language groups, the W estern (Czech, Slovak, Polish, Slovenian, Sorb), the Southern (Serbian, Croatian, Bolgharian) and Eastern (Ukrainian, Belorussian, Russian) evolved relatively late. A l though their diverging tendencies can be detected in the 8 th and 9th centuries, it is difficult to determ ine the origin o f individual elem ents o f the languages. The dispersal o f the Old Slavonic people was slow and continuous. The Iranian loan words o f the proto-Slavonic language roughly determ ine that they must have lived in the land o f the modern Ukraine before dispersing. The origin o f the name Slav is uncertain, but as for a long tim e they w ere the source of a large num ber o f slaves, their nam e gave rise to sclavus ‘slave, servant’, or in the Arabic sources saklabi (singular) and sakaliba (plural).
240
■ #
?
| p :;i
North $ 1
Relatives an d neighbours
Sea
,,,
\
% ^
f
С ,
Onega
^
ж-%%. ,: "'V%‘p 4- v-^ ;:a.0 • % /... и i «■ > I у ft a r"\ ; „ Г 'ы '1 • & W / ^ ' ГЛ*К {W‘Lc 4
...'<%X:,<<:'V£v:.,>:'.•>:
:'■■■■£
.
O :?}-1
1
.n s 0
f-
=Л • O' /<,4 / • "i 1 as; % . *' c? e rVm / n n ,, . ■ ic * '> 1 *
S l a v s
■
-4 ^ “V
'•
£ j r ,
, ■0'-'
•
^
*
Aiani
x "v,. ,
:*\,
<л" ^ ;
■я v :\Ъi £• .
, ,d ,
«*■• -.««JHbractear
%
f
Щ аск
,1 ..~ ( V^nW^ntmnn!^
'ч О ^
с,,
\ч
,# с а
Ph.s-.x •'«'j,
'. J W / v 0 is ^ ' , Л’ « о да м jr-j V - : ^ *Ь ; ° ^М'AJV '■" '' л r- SlPl }t E i ~4
I
*
A la n i .г .
Ч. “ >
PERSIAN
.. •£ ^
-S'
ЕМ f I R E
«га Figure 60 The distribution o f the Slavs betw een 450 and 550
a) 7~/z Southern Slavs The nam e o f the Slavs is first m entioned in the m iddle o f the 6 th century by Procopius who writes o f the year 527 that the A nts and the Sklavens attacked B yzantium ’s Baltic provinces. In the second h a lf o f the 6 th century, there may already have been some scattered Slav settlem ents south o f the Danube, but the larger Slav groups gradually pressed w est and north. A m ajor change was brought by the com ing o f the Avars. The Avars took Slavs w ith them as allies under their organisation, and their m ilitary expeditions opened the w ay for autonom ous m ovem ents o f Slav groups. In part, they caused Slavs to flee from their attacks and move to new areas. A contributory factor was B yzantium ’s preoccupation with fighting the Persians (and later the Arabs). In 582, the Avars captured Sirmium, opening up a gap through w hich the Slavs could drive their way into the Balkans. By the end o f the 6 th century there were m ajor Slav settlem ents in the northeastern regions o f the Balkans. The Byzan tines recaptured Sirm ium in 600, but they could not halt the advance o f the Slavs. With H eraclius (610-641) held down by the Persians, new Slav popu lations were able to settle in the Balkans. B y the beginning o f the 7th century,
241
The neighbours
у
|
'
1
\ I
T „
' Д.ч'-
4 **«
3
/ “• in
•
<•
‘ i
l
pAjsW SH
**•
i A
4 v -
*
•
' V,
1/
-
\ ,/ " '"""’ '4
-•
:
s....
ъ »'
,
_
,
“5.
""■)
X :
/'•■' \
Й Й ^
" " 4^ Turks
4
£
л
; ^f ’ ijjH f o ' V
..
Л
i
^ i
V A l a n i f*
\
•!
X
* % ■
j
Ls.uO
t
. ,.f
-•
,vf"
-
'4
4
s'
*
Slant,
'■** s
..,
‘
.
s , f i
Д Ч 'Л н Ч ; E s v ilfm i; \
C l , ,
1
— s
'.
'■ .■ ■\ C1 '
* • « ,./% ,
em w re
C V *
‘^ВеЫ?0У}‘О / ' :.iX
~ p'^' :H;!
litls-r’ <
,,
.p L a k e O nega
(
% ,, <Л П ,
*’i* л Г ' \ L
}
i * /
4
i,
f,
I
I
f
':' ф
|
4’„ .
1.
JvzaM m rn,-'"'
e A s TERN- ROMAN
CV
EMf4»'E
’•-
.. "**:
.SV„
.
;>
1 '
!,; ■
!'
"” 4
" V:
.. -v-''
..: '
/
Я * 'г \ 'л <■ ' >' a
4
\
\
\
PERSIAN
EM P /R E (;s,: /
Figure 6 1 The distribution o f the Slavs betw een 600 and 700
large populations had reached the A driatic coast. The incoming Slavs gradu ally assim ilated the native Balkan population, and so it is difficult to follow the ethnic processes archaeologically. In any event, the Slavs were operating autonomously next to the Avars in the 7th century, and on occasion actually clashed with them. The western sources note the rising independence o f the Slavs, and record their m ilitary expeditions, led by the Frankish trader Samo, against the Avars in the 620s. In 626, the Persians allied with the Avars and the Slavs against Byzantium . There are records o f a Slav-A var w ar around 629, which was rapidly followed by an attack on the Avars by the then Croatian-led Slavs in conjunction with the Turkic O noghundur B ulghars from the east. Local Serb and Croatian princes em erged as independent forces in the 620s. Their origin is disputed, some opinions claim ing that they came from the areas o f “the white Serbs and the white C roatians” in the north, but in any event they organised the Balkan Slavs. The account o f events related by Saint D em eter's M iracles offers inform a tion on the Slavs w hich would take a separate chapter to analyse. It follows the affairs o f Kuver, a leader whose origins were in the Avar Em pire, but was
242
Relatives and neighbours
probably Onoghundur-Bulghar, and the Slavs who m ade their appearance with him. In the 670s, the Bulghar Turks driving into the Balkans settled am ong the local Slav population, and in general followed the same policy as the Avars, i.e. they relied on the leaders o f local com m unities. Slavs w ere involved in the b rief period o f anarchy w hich follow ed the dying out o f the Bulghar ruling line in 739. Although Krum and O m urtag (814-831) m anaged to reim pose central control, and a leading role was still played by the Bulghars, the Slavs gained increasing influence in the Danube Bulghar Empire. The Slavs living next to the R iver Tim ok (the Timochans) cam e into conflict with the Serbs, who had built up their strength at the beginning o f the 9th century. The Bulghars also struggled against the new Serb power. They led m ilitary expeditions against the Serb chieftain V lastim ir in the period between 839 and 850. In 827, Bulgharian forces attacked and captured the T im ochans’ lands around Sirm ium and eastern Slavonia. Under Bulghar pressure, the Slav tribe A bodrit crossed the Danube, and drove into the Tem eskoz, but around 830, O m urtag dealt them a defeat and spread his control up to the line o f the River M aros. There are no reliable sources concerning the following period. It is certain that the Bulghar rulers exercised control over the M aros valley salt mines in 892, and thereby held sway over the local Slav population.
b) The Alpine and the Western Slavs The W estern Slavs, taking advantage o f the tem porary w eaknesses in the centres o f power, fought successfully for their independence. They w ere at the w estern side o f a Slav world w hich was gradually taking shape to the east of the E lba-S alzburg-A quileia line, and being penetrated from the w est by the Franks, and from the east first by the Avars and Bulghars, and then by the M agyars o f the Conquest. In the 6 th and 7th centuries, the Slavs were not yet especially prom inent in the C arpathian Basin. Their stature grew first in Transylvania, at the beginning o f the 7th century. A t the sam e time, at the eastern perim eter o f the Alps, in Dalm atia and in the southw estern area o f the m odem Czech Republic, and the w est o f Slovakia, unified Slav blocks took shape from the 620s and 630s w hich soon started to organise them selves politically. Porphyrogenitus m entions around 950 the Lenzen tribe (p. 237), w hich was tributary to Kiev. This nam e is the origin o f the Hungarian name for the Poles, Lengyel, which they learned in the Etelkoz.
243
The neighbours
c) The Slavs o f Pannonia, the Moravians and the Slavonic conversion It was first recorded in 822 that the Slavs living in the M oravian region sent an em issary to the Frankish court. The successor o f the first autonom ous ruler M oym ir (830-846) w as his nephew Rastislav (846-870), w ho had been raised in the Frankish court, and after acceding to the throne soon attem pted to secure independence from the Frankish king. By around 860, he had built up m ilitary strength and it becam e clear that he could count on m ajor allies if he adopted Christianity. However, he also w anted to prevent the influence o f Frankish m issionaries from rising, so in 862 he sent to Constantinople for proselytising priests. This he dared to do because a favourable situation had been created: the son o f Louis the Germ an, Carlom an, rose up against his father, and R as tislav called on the alliance o f the M agyars to free him self from Frankish dependence. To understand w hat followed, we m ust take a step backwards in time. In the Nitra area, Prince Pribina, in pursuit o f more independent power, clashed with M oymir, who had absolutist aspirations, and was forced to flee. He was backed by Ratbod, the baron o f the eastern fringe o f the Frankish Empire. In 833, Ratbod introduced Pribina as a Frankophilic Slav tribal chief to Louis the Germ an, grandson o f Charlemagne. H ow ever relations between Ratbod and Pribina rapidly deteriorated, and Pribina had to flee once more. This time he turned to the Bolghar king, M alamir, who had fought hard against the Franks in the past, but had m ade peace with them in 832. Pribina was unable to persuade M alam ir to act against the Franks, and continued his flight. At the western end o f the Drava there was another small Slav possession. Its ruling noble, Salaho, had great influence in the Frankish court. When Pribina finally cam e to him, he was accepted and Salaho persuaded Louis the German to order Ratbod to come to a settlem ent w ith Pribina. The Frankish court very probably wished to curtail R atbod’s power, and in addition was seeking an ally against M oymir, and so in 840 Pribina was granted extensive estates beside the R iver Zala (in present Western Hungary). Shortly afterwards, a battle was started against the M oravian chieftain Moymir. Pribina obviously played a prom inent role in the fight, and in 846, after victory over M oymir, gained further estates near W ulka in modern Burgenland (Austria). He was also granted a title to go with his lands, becoming the border guard baron o f the Drava region. His duty was to gather, and keep loyal to the Franks, groups o f Slavs fleeing from various directions and causes. To carry this out he built a new stronghold in the sw am py land next to the R iver Zala, which was called M osaburg (present Zalavar). In 850, a church was built in hom age to Mary. O f the prom inent personages who
244
Relatives and neighbours
arrived for the consecration, fifteen had Slav nam es, sixteen G erm an, and one Latin. U ncertainty surrounds P ribina’s death. Possibly the Frankish prince Carlom an, who rose against his father, captured him and handed him over to the M oravians. P ribina’s son K otsel fled to Louis the G erm an in Regensburg. Louis the G erm an conducted a bitter struggle against his son C arlom an, and only after victory over him was he able to instate K otsel in his fath er’s place. K otsel received both his father’s lands and titles, and exercised control over Transdanubia. He first o f all had to settle the m atter o f church taxes. These w ere controlled by the Bishop o f Salzburg, who was not regarded favourably in the circles surrounding Kotsel, and the tim e seem ed right for establishing a bishopric in Pannonia. Kotsel pursued this by petitioning Rom e directly. The Pope, N icholas I (858-867), im m ediately recognised the significance o f establishing a bishopric loyal to Rom e in Slav lands, since the B olghars had for a long tim e been negotiating with both Rom e and C onstantinople. The Pope had earlier succeeded in bringing D alm atia into his sphere o f influence. K otsel’s loyal follower, Oswald, was actually appointed as B ishop o f Pannonia by the Pope, but the Bishop o f Salzburg, Aldwin, w asted no tim e in having O swald captured and locked up in a monastery. It was then that M oym ir’s successor Rastislav, Prince o f M oravia, sent to the Byzantine Em peror for proselytising priests. The Em peror did not at that m om ent w ish to provoke the ire o f the Franks, and so instead o f a proselytising bishop, only sent the M oravian chieftain a pair o f brothers from Thessaloniki, C onstantinus and M ethodius. The two brothers arrived in the P rince’s capital in 863. Both were bilingual because— although there is som e dispute as to their origins— they received a bilingual Greek and Slav education in the town o f their birth. However, Louis the Germ an did attack, and in 864 defeated the independence-building Rastislav near Оёуёпу (Devin). Rastislav had to renew his vow o f loyalty, and as security sent his nephew, Svatopluk, Prince o f Nitra (who appears in Frankish sources as Zventibald). In the m eantim e, the Bolghars were subjected to pressure from Byzantium , from w hich they w ished to escape by turning to Rom e w ith a plan for establishing an independent Bolghar church. They w on support from the Bishop o f Passau, who had had prior relations w ith the Bolghars. There was bitter rivalry betw een the bishops o f Passau and Salzburg, how ever, and the Salzburgers persuaded Rome to restrain Passau. With the even greater m ilitary m ight that follow ed victory over the Arabs, Byzantium turned against the Bolghars, dealing them a m ortal blow, and stipulated in the peace treaty o f 8 68 that the B olghar church should be set up under their supervision. The Bolghar K han Boris again sent to Rom e in secret, but it was decided there that the situation did not perm it intervention.
The neighbours
245
It was in this confused state o f affairs that Constantinus and M ethodius had to make w hat headway they could in the M oravian court. They decided that they could m ost clearly settle their affairs if they travelled to Rom e, which they did in 867. The central aim o f their petition was the granting o f perm ission for a Slavonic liturgy, and to create a w ritten language involving the alphabet they had invented (this was the Glagolitic and not the C yrillic alphabet— see pp. 60-61) into w hich they could translate the holy texts. A t stake was the conversion o f the Slavs. Consequently, Pope Hadrian II (867-872) reached an accord with the brothers on a Latin-Slav liturgy, and gave them his perm ission for its use. The two brothers, on their way to Rome from the M oravian court, visited Kotsel in M osaburg. Kotsel received them with great respect, because with his Frankophile Slav past he clearly understood the significance o f the proselytising brothers’ journey. He ordered pupils from his circle to study under them, and allegedly learned to read in G lagolitic script him self. Some have claim ed that it was he who persuaded the brothers to first seek support from Rome, in which he was clearly influenced by his previous attem pt to establish an independent Pannonian bishopric. Constantinus, who in the m eantim e had adopted the religious nam e o f Cyril, died in Rome in 869, and M ethodius returned to Kotsel. But he was not appointed as Bishop o f Pannonia, because com plex scenarios w ere being played out in Rome and Constantinople. M ethodius was once m ore sent by Kotsel to Rom e, whereupon he was consecrated as B ishop o f Sirm ium (Mitrovica). In the m eantim e the Bolghar church affair had been settled, and Boris, who inclined to Rome, had to concede: Bulgharia, in 870, subjected itself in ecclesiastical affairs to the authority o f Constantinople. The Bolghars captured Sirm ium and established it as a see. In the same year, the M oravian R.astislav once again confronted the Frankish king. However, his nephew Svatopluk captured him and turned him over to the Franks, who blinded him and locked him in a m onastery in 870. To definitely prevent a com eback, they also took M ethodius prisoner, brought him before the ecclesiastical court in Regensburg on trum ped-up charges, and sentenced him. Although M ethodius was released from im prisonm ent a few years later, it was the end o f the dream o f a Slavonic-liturgy church loyal to Rome, since Pope John VIII (872-882) prohibited the use o f the Slavonic liturgy. The Salzburg and Passau church leaders’ rivalry had succeeded in strangling the experim ent at birth. W hen M ethodius was released, K otsel was already dead, and his successor, Gozwin, did not support him. The M oravian nobles did not put up w ith Frankish supremacy, however, and Svatopluk supported them secretly at first, and then openly. The w ar ended with the peace o f Forchheim in 874. Svatopluk invited the B ishop o f Passau to reorganise the M oravian church, and brought in M ethodius at the sam e time.
246
Relatives and neighbours
It was this situation that sparked off the conflict. Rome sum m oned M ethodius to face accusations, and although he was absolved in 880, his opponent, W iching, was m ade bishop o f Nitra. M ethodius lived for another five years and was active in S vatopluk’s court, but after his death in 885 his pupils were banished, m ost o f them fleeing to the Bolghars, m arking a new beginning for the Old Church Slavic tradition, this time in Bulgharia. Louis the Germ an died in 876. B avaria and the eastern flanks gave their allegiance to Carlom an, who died in 880. His successor was C harles III, who reunited the em pire after the death o f his brothers Carlom an and Louis III, but was forced to abdicate in 887. His successor A m ulf, son o f Carlom an, was crow ned em peror by the Pope in 896. Exploiting the warring between Louis the G erm an’s sons, Svatopluk laid waste to Transdanubia, but made peace with the heir to the throne, A m ulf, in 885. This did not last long. Arnulf, rem em bering that Rastislav had achieved success by calling in the M agyars, him self com m issioned M agyar support in attacking Svatopluk in 892. Svatopluk died in 894. His country was divided am ong his sons, am ong them M oym ir II and the Frankophile Svatopluk II. The days o f the M oravian Em pire were numbered.
d) The Eastern Slavs and the Rus o f Kiev The land o f the m odern Ukraine, Belorussia and European Russia was occu pied by Finno-Ugrian and Slav peoples. To the east, the border w ith Slav tribes around 900 was the N izhniy N ovgorod-R yazan line, i.e. the valley o f the River Oka. To the south, during the age o f m igrations, the Slavs w ere not present on the steppe o f their own accord, but as slaves or conscripted soldiers. The forest region and river valleys, however, provided the security which enabled tribes known as th eP o lya n e (people o f the fields), Severyane (northerners), Krvichi, Vyatichi or Slovene to survive. The Polyans, the Severyans and the Vyatichs for a while paid tribute to the Khazars. At other times tributes w ere extracted by the rulers o f the steppe at the time, and slaves were taken from them for sale. In the 9th century, the flourishing o f the Arab Em pire and Byzantium stim ulated trade, which was controlled by the trading houses and shipping com panies. W ell-organised free bands incorporating northern (Norman) G erm anic warriors acquired great wealth, and it also becam e im portant to them that the north-south routes, and the transport o f goods along the rivers, be secure. Certain sources m ention a Rus Khaghanate in the 830s and 840s w hich clearly had strong links to the Khazars, but whose precise location cannot be determined. The title o f later K iev rulers also contained that of
247
The neighbours
khaghan. As towns becam e established, the northern Germ anic Varangians set up trading depots and centres. In Novgorod, the “new tow n” , one o f the leaders o f the Varangians, Rurik, was proclaim ed emperor. Rurik and his two accomplices, A skold and Dir, captured Kiev. A fter R urik’s death (around 879), power passed to his relative Oleg, who occupied Kiev in 880-882, executed Askold and Dir, united the surrounding tribes, and by 907 attacked Byzantium . Rurik’s work was continued by his son Igor, who succeeded him to the throne in 913. Igor also started by attacking Byzantium , but subsequently m ade a trading agreem ent, and thus the road to the conversion o f K iev was opened. Igor died in 945, and his widow took up the Christian faith in 957. The Kiev Prince Svyatoslav (961-972) defeated the Khazars and the Danube Bulghars, but on his way hom e in 972 was killed by the Pechenegs at the instigation o f Byzantium. His successor V ladim ir consolidated his pow er with the assistance of the Varangians, and after taking a B yzantine wife, was converted to Christianity in 988.
NOTES Two excellent accounts o f research into Indo-European peoples and languages are Gamkrelidze-I vanov (1995) (the Russian original 1984) and M allory (1989). The G am krelidze-Ivanov book and its new results have been covered in several dozen articles and studies. For its Altaic aspects, see Rona-Tas (1988c). Good further bibliography is given in M allory’s book, see also Harmatta (1971). For Indo-European words, e.g. those meaning ‘axe’ and ‘hatchet’, see Gamkrelidze-Ivanov (1984, p. 716), Poppe (1954), Nem eth (1934, p. 85; 1942-1947, p. 87) and Benko (1 9 6 7 -1 9 8 4 ). For the Hittite form Istanu, see G am krelidze-Ivanov (1984, pp. 684 and 897). For the Iranian names for the Sun see Jeremias-Maroth (1977). The form and spread o f the Hittite form for the word ‘apple’, see Gam krelidze-Ivanov (1984, pp. 6 3 7 -6 3 9 ). The most recent works on the word for ‘apple’ are Joki (1963) and Erdal (1993a). For the Indo-European equivalents o f the Turkic word for ‘o x ’ (Old Turkic okiiz, Ancient Oghur okur) see R am stedt(1947, p. 25; 1957, p. 104), D oerfer( 1963-1975, pp. 5 3 8 -5 4 0 ). The Turkic word okiiz has been linked to the Tocharian okso, follow ing Munkacsi, by Nem eth (1934, p. 168), in German (1 9 4 2 -1 9 4 7 , p. 94). I suggested the re-examination o f Tocharian-Turkic links in Rona-Tas (1974a). The corre spondences discussed there have recently been re-examined by Reinhart (1990). The latest results o f Tocharian research are presented by W. Thomas (1985). see also van W indekens (1976). The etym ological dictionary o f the Ossetian language (Abaev 19 5 8 -1 9 8 9 ), is now complete. It is striking that although there are Turkic loan words at several levels in the Ossetian language, no words o f certain Chuvash-type origin have yet been found. The ancient Ossetian Nart epic refers to the Agyr/Agur people, which is in all probability an equivalent to the ethnonym Oghur (Abaev 1958, pp. 3 6 -3 7 ).
248
Relatives and neighbours
For the Chinese sources relating to the Alani, see Chavannes (1900) and Liu (1958). Arabic and Byzantine sources are quoted by Dunlop (1954/1967) and Marquart (1903). Concerning M ichael o f Syria, see Rona-Tas (1982a). M inorsky (1937) discusses the Alani in detail in the notes to the H udiid al-alam . For the language o f the Goths o f Crimea, see Stearns (1978). For the European Huns and the early history o f the steppe in general see G olden (1992). The best monographs on the European Huns are M aenchen-H elfen (1978 which is the expanded German edition o f the English-language original o f 1973) and Bona (1 990a, 1991). The issue o f the name o f the leading tribe o f the Ruanruan is touched on in H. Franke’s (1969) article on the Khitai language. The identity o f the tribal names was dealt with earlier by Chavannes and Otto Franke. 1 discuss the Turkic elem ents o f the Khitai language and the Turkic-origin Khitai tribal names in an as yet unpublished study. Information from Chinese sources on the Ruanruan has been collected and made available in Hungarian by Csongor (1993). For the appearance o f the Khitai in the history o f the Ruanruan in 553, see Csongor (1993, p. 46). The issues o f the Priskos literary heritage are covered by M oravcsik (1983, pp. 485^186). For information on the tribal names o f the Oghur, Ughur, Onoghur, etc. peoples and tribal names see M oravcsik (1983), under the corresponding headings. The events o f 350 on the Khazakh steppe are discussed by Czegledy (1969, pp. 9 2 -9 3 ). Nem eth has also expressed his opinion in several places that the ethnic name U trighur can be traced to *utur (see Chuvash vatar, Turkic otuz), and kutrigur can be linked to the numerals tokhur ‘nine’ (Chuvash tahhar, Turkic tokuz), see recently N em eth (1991, p. 132), I used the Greek text o f Theophilaktus Simokattes in the 1972 Boor-W irth edition. For the part on the Oghurs, see page 258 o f that edition. For a translation, see Schreiner (1985, p. 187). The former pronunciation o f the Chinese name hua as var is discussed by Czegledy (1969, p. 141). The H ua country has been covered most recently by Csongor (1993, pp. 6 7 -6 8 ). The Chinese pronunciation is more precisely hvar. The identification o f the second syllable o f the town name o f Bakath is covered by Czeglddy (1969, pp. 92 and 138), and Harmatta (1992, p. 257). For the latest view s on the name o f the Sabir people see Harmatta (1992a, pp. 257 -2 6 2 ). Harmatta considers the tribal or ethnic name Sabir to be o f Iranian, or more exactly Saka, origin. The question and literature o f the “Eastern Silver” is covered by Balint in H ajdii-K ristoRona-Tas (1976, vol. I / l , p . 9 1 ) . For the identification o f Khuvrat’s grave, see Werner (1984) and Balint (1984), but both are o f the view that Khuvrat only moved to the place o f his subsequent burial shortly before his death. The location o f Khuvrat’s Bulgharia was identified essentially correctly from text analysis by Lauterbach (1967). The history o f the Bulghars is given in Zimonyi (1990). An earlier treatment is Smirnov (1951). For a review o f research history', with particular regard to the Volga Bulghar inscriptions, see Rona-Tas-Fodor (1973). The more recent publications are Hakimzjanov (1978), Tekin (1988) and Erdal (1 993b). For the B olshie Tarhani excavations see G ening-H alikov (1964). The chronological classification and com plete discussion o f the Volga Bulghar archaeological finds is given by Fahrutdinov (1975), with excellent maps. I analysed Ibn Fadlan’s Volga Bulghar language evidence in Rona-Tas (1982c, pp. 2 1 0 -2 1 4 ). The inscriptions and language o f the Danube Bulghars have already been m entioned above (see pp. 113 and 167). Fine (1983) and Golden (1992) give treatments o f the history o f the Danube Bulghars and the rich historical literature relating to them. I considered the ethnonym Khazar in Rona-Tas (1982a and 1983a) (with English summa ries). See also Bazin (1 9 8 1 -1 9 8 2 ). For the Khazars, see Dunlop (1954), G olden (1980) and
The neighbours
249
Ludwig (1982). Conversion o f the Khazars to Judaism is covered in Dunlop (1954), Pritsak (1978) and Golden (1983). The history o f research into Pecheneg-M agyar contacts is covered by Kristo (1980, pp. 97-115). For GyOrffy’s opinion see G yorffy (1971). I used Gardizi in M artinez’s 1982 edition. The best edition o f Khashghari is D ankoff-K elly (19 8 2 -1 9 8 5 ). The H udiid al-alam edition used by me is M inorsky (1937), I did not access more recent editions. For the history o f the Balkan Slavs, see Fine (1983). An account o f the Slav-Sk lavin -Sakaliba issue and the lessons o f the M uslim sources is given by Golden (1993b, pp. 8 7 2 -8 7 8 ). The Rus issue and the beginnings o f the Principality o f Kiev are also covered by G olden (1993a, pp. 61 7 -6 3 0 ). Several aspects o f Slav history in connection with the Magyars are discusscd by Boba, among others, as in Boba (1 9 6 7 ,1 9 8 2 -1 9 8 3 , 1984, 1993). For the Slavs o f Pannonia see Cs. S6s (1973) and the literature given there, and Bona (1984a, 1990d). A s regards Constantinus-Cyril and M ethodius, see above (pp. 60 -6 1 ); on their assumed Magyar connections consult H. Toth (1981) and Kiraly (1974). Recently a dispute has developed over the “Great M oravia” issue. Som e were o f the opinion that there also existed a “Southern” Moravia. Since the arguments in favour o f the “ Southern” Moravia did not convince me, I considered it unnecessary to deal with the question in relation to the Conquest. The issue is, however, far from being solved. This has been recently done by Martin Eggers in two excellent works (Eggers 1995, 1996). I got the two books only after having closed the work on the English version o f this book. Mis analysis gives a detailed background o f the ethnic history o f the Carpathian Basin, o f the history o f Moravia and its rulers. Eggers refutes the argumentation o f Imre Boba who was, if not the first, but the m ost ardent (and learned) protagonist o f the idea o f the two M oravias and especially o f “Southern” Moravia. Having read the two monographs o f Eggers, I see no urgent need to change anything in my wording. 1 hope, however, that 1 can use his data and reasoning in the preparation o f a second edition and/or a German publication. The chapter on the ethnic diversity in the Carpathian Basin on the dawn o f the Hungarian Conquest will profit much from his works.
V. EURASIA IN THE 9th A N D
10th CENTURIES
This book cannot give m ore than a b ird ’s eye view o f the historical situation o f the two continents, and in doing so, only o f the events that are directly or indirectly related to the history o f the M agyars. Such a superficial overview seems to be nevertheless im portant in order to understand the historical framework within which the fundam ental changes o f the history o f the Magyars occurred. The 9th and 10th centuries saw m ajor changes in the relations o f the nom adic and settled peoples. The em ergence o f a new m ilitary force, the light cavalry o f the nom ads, as well as the special system o f incursions, can be connected to the use o f a new invention, the stirrup. The stirrup was invented in the first centuries around the birth o f Christ. For a long time it was used alm ost exclusively by sick or elderly people and pregnant women, prim arily for getting into the saddle and to prevent them from falling off. It was first made o f a leather strap, occasionally o f wood, and only later, from the 3rd century A D , o f metal. Before the Avars appeared it was not part o f the regular m ilitary equipm ent and was not used in battle at all, but from the 6th-7th centuries it began to play a decisive role in nom adic warfare. The stirrup made firing an arrow backwards easy, and this led to the developm ent o f the unique tactics hailed or disparaged by m any contem porary authors. Following an initial attack, the nom ads w ould turn back and pretend to flee. After a time, however, the warriors would turn round in their saddles and unleash a hail o f arrows at the pursuing enemy. The technique o f shooting backwards from the saddle is a lot older than the stirrup itself; however, standing up on the stirrup m ade the turn backwards and shooting more effective. The settled people m anaged to m ake the m ost o f the new invention, and consequently at the end o f the 8 th century the pow er o f the Turks and the Avars dom inating the central parts o f Eurasia began to decay. O f course, the stirrup was not the only factor that influenced the changes. The astonishing expansion o f the Arab-M uslim world, w hich reached its first peak in the 8 th century, changed the situation both in A sia and in Europe. The birth o f a new European order under Charles the Great, the p a x khazarica in Eastern Europe
252
Relatives and neighbours
also had an im pact on the history o f the 9th and 1Oth centuries. Eurasian trade underw ent som e significant changes. We shall now give a m ore detailed description o f this procedure.
1. THE END O FTHE UIGHUR EMPIRE In alliance with the K harlukhs and the Basm ils, the Uighurs defeated the second Turk Khaghanate in about 740, but then turned against their allies and ousted the Basm ils and Kharlukhs from power. The Kharlukhs m oved w est wards. A fter 744, the Uighur K haghanate flourished for about hundred years. Its high culture, rich towns and role in the trade o f Central A sia w ere praised both by the Arabs and the Chinese. The Uighur Khaghans were involved in dom estic Chinese affairs, and during the course o f a visit to the Chinese capital in 762, Bogu K haghan converted to M anichaeism w hich was then one o f the strongest religions in China. At the same tim e Buddhist m onks also found their way into the prosperous U ighur K haghanate. In 832, the U ighur khaghan was m urdered by some o f his m inisters, and dom estic discord was augm ented by attacks o f the Khirghiz and the Tibetan armies. In 840, the U ighur tribal confederation broke up, and som e groups fled to the west, som e to the south. Those o f them who w andered to the south founded several small principalities. M ention m ust be made o f the Uighurs o f the Five Cities (B eshbalik, Penjikent), the Khocho Uighurs, and the Uighurs o f the Gansu province in China. During the following centuries the Uighurs played an im portant role in the cultural life o f Central Asia— a role which was m ore prom inent than their political significance. It could be o f im portance in connection w ith the history o f Eastern Europe that one Uighur tribe had the nam e K hasar, w hich rings a bell, being quasi-identical with the Turkic nam e o f the Khazars. A nother tribe had the nam e Uturkar or H uturkar which m ay be connected w ith either the Utrigurs or the Kutrigurs o f the O ghur tribal confederation. Such facts, however, are insufficient in them selves to establish a closer connection betw een the Uighurs and the Khazars; w hat they do show is that tribes often segregated, and the various splinter groups could surface in places very far from each other. It is possible that the K hasar tribe o f the Uighurs w as an older unit and rem ained in Central Asia, while other parts m igrated to Eastern Europe; but equally, it is possible that a group o f K hazars m igrated to Central A sia and joined the Uighurs, like the Khavars did the M agyars. The debate continues about this enigm atic issue and w hatever the outcom e will be, there is one thing for sure: the Oghur groups played a role o f some description in the U ighur tribal confederation. However, we have face up to the fact that although source m aterial regarding this issue is expanding, no linguistic
Eurasia in [he 9th and 10th centuries
253
evidence has yet been singled out from the otherw ise rich U ighur material. What could account for this is that the w ritten docum ents were, perhaps, based on a non-O ghur dialect. The Khirghiz, who played an im portant role in the collapse o f the U ighur Khaghanate in present M ongolia, did not build up a new power. They lived in South Siberia w here they left runiform inscriptions behind. Later, a part o f them m oved to the region w hich is now called the Khirghiz Republic. Interestingly enough, traces o f their Siberian hom eland can be found even in the m odem Khirghiz language. The tribal confederation w hich followed the Khirghiz spoke a M ongolian language and becam e known as the Khitai or Khitan.
2. THE KHITAI A N D CH IN A It has been established that the M ongolian language was indubitably spoken even before the appearance o f the Khitai. The tribal confederation called Xianbei by the Chinese sources first cropped up in the 3rd century BC. Am ong them we find tribes whose names bear com parison with the nam e o f the Mongols. M uch is recorded about the groups which occupied parts o f China and founded dynasties. Among them we have to m ention the Tuoba who founded the Wei dynasty in 386 A D . The Tuyiihun, one branch o f the Xianbei, moved to the south and formed an em pire in the vicinity o f the northern borders of Tibet. This small state flourished betw een 541 and 700. In the Tibetan sources this group was called Azha. The title o f khaghan appeared alm ost at the same time among the Ruanruan and the Tuyiihun. The Tuyiihun had another title, khapagan, which is noted by the Chinese sources at the end o f the 6 th century. The latter is the same title we com e across am ong the Avars of the Carpathian Basin. The Khitai reached the peak of their pow er at the beginning o f the 10th century. Their nam e figures in sources since the 5th century. We can read about them not only in the Chinese sources, but also on the stelae, w ritten by the Turks in Turkic runiform script. Their name features as K hitany on the stelae. In m ost Turkic languages the final -ny (palatal Inf) becam e -y, but also -n in some o f them, as well as in M ongolian. This accounts for the tw o variants Khitai and Khitan. The R ussian nam e o f China, K it ay, reflects the first form, and the Cathay o f the travellers o f the M iddle Ages, w ho visited M ongolia or China after the M ongols o f Chingis Khan came to power, is a variant o f the same name. The head o f the K hitai tribal confederation founded the Liao dynasty in 916. The conquest o f the north o f China happened at alm ost the sam e tim e as the
254
Relatives and neighbours
M agyars’ Conquest o f the C arpathian Basin. A lthough both o f these conquests w ere a case o f peoples w ith a nom adic background encountering settled civilisations, the set o f circum stances, the underlying processes and the outcom es w ere different. N evertheless, the com parison o f these two histories can afford some inter esting results. Firstly, it m ust be stressed that Khitai society featured a significant Turkic element, m ost probably o f U ighur origin. The clan from w hich the Khitai em perors took their w ives was Turkic, and the Turks also played an im portant role in the political life o f the Liao Em pire. Fortunately, copious evidence about them has come down to us. The assessm ent o f the Khitai Liao dynasty wholly depended on the succeeding dynasties. The Chinese Song dynasty in the south reigned sim ultaneously with the Liao dynasty. The “national” historical tradition claim s that the Liao w ere usurpers, and consequently their history can only be reconstructed with the help o f the Song Annals. However, the Jurchen Jin dynasty, follow ers o f the Khitai, and later the M ongolian Yuan dynasty, who ruled over the whole o f China after 1280, considered the Liao as their righteous predecessors. This caused prob lems in the com pilation o f the dynastic history o f the Liao, the L iao shi. It was rew ritten, am ended and censured m any tim es over. The Jurchen scholars added to the confusion, by replacing the Old M ongolian (K hitai) term s and titles with Jurchen ones (even in retrospective), with not so m uch as alluding to the “corrections” . Giving full consideration to these textual problem s will leave us with a highly detailed description o f the society, culture and political life o f the Khitais. The description o f Liao society, sum m arised in a m onograph by W ittfogel and Feng Chia-Sheng, is very inform ative. The sm allest unit o f K hitai society was w hat the M ongols called ayil, and the Turks aul. Its K hitai equivalent has not com e dow n to us, all we know is that it was replaced by the Jurchen name ghasha. The ghashas o f related kinsm en form ed a branch, and related branches were m em bers o f a clan. Clans were grouped into subtribes and several subtribal com ponents form ed a tribe. The m em bers o f a clan bore the same clan nam e, they venerated the same ancestor and m arriage was forbidden am ongst the m em bers o f the same clan. The subtribes had special rights and som etim es they or their head were accorded tribal rights. The title o f the second leader o f the subtribe was darkan. This w ell-know n title, w hich existed am ong m any Turkic peoples as well, gave the nam e to the M agyar tribe Tarjan (read as Taryan). The history o f the tribal system o f the K hitai is particularly enlightening. A round 600 A D , the num ber o f the Khitai tribes increased from eight to ten. In the 7th century the ten tribes were reduced to eight, but two new tribes joined, leaving the previous total o f ten unchanged. A fter the K hitai conquest
Eurasia in the 9th and 10th centuries
255
o f N orth China the tribal confederation was reorganised on m any occasions. It was divided into “Inner Tribes” and “Outer Tribes”. The “Inner Tribes” were the original m embers o f the Khitai tribal confederation, while the “ Outer Tribes” were those who joined the confederation later. Also, there were various types o f dependent tribes or peoples. The rulers always belonged to the Khitai Yelyu clan, and their consorts to the Xiao clan. The Xiao clan was o f Uighur origin. The “Inner Tribes” were divided in two subgroups, those associated with the ruler and those w ith the queen. The Yelyu subgroup was further divided into the ruling tribe o f the Emperor, and the groups o f his paternal uncles. R eorganisation m eant dividing or m erging existing tribes, or creating new ones from subtribes. Tribes were organised from prisoners o f w ar or other foreign groups who joined the tribal confederation for w hatever reason. This tribal system naturally had a hierarchy. O f the ten tribes which existed at the time o f the foundation o f the Liao dynasty, the first was the tribe led by the Yelyu clan (which gave the Em peror), w hile the nine others w ere tribes o f the Yaolien clan, who had fallen from pow er in a struggle against the Yelyu. Consequently, the nine Yaolien tribes were lower in rank, had lim ited rights, and even as a whole they had less pow er than the sole tribe o f the Emperor. This m ay have been the case with the M agyar tribes as well, w here the tribe o f the later Arpad dynasty, the M egyer, had greater pow er and rank than all the others together. There is an interesting tribe in the Khitai confederation w hich appears, in Chinese transcription, under the nam es Wuguli, Yuguli, Yujielyu, and Yujueli. These are transcriptions o f the original tribal nam e w hich sounded ‘U g h u r’ or ‘Yughur’. We can trace this nam e back to the Ruanruan era in which the same tribe assum ed a leading role (see p. 210). As a m atter o f fact, this happens to be a late form o f the nam e Oghur, and as w e have seen (pp. 211-212), the Ten Oghur or Onoghur were an im portant tribal confederation in Eastern Europe. Their nam e was transferred to the M agyars, from which— through Slavic mediation— the nam e Ungri and Hungarian finally cam e into existence. Extremely im portant is the description o f how the Khitai shifted from their original “nom adic” agricultural econom y to the one prevailing am ong the Chinese. Their agricultural roots m ust have strongly resem bled the M agyars’, but their traditions took a different direction, and so the outcom e was quite different.
3. THE KHARAKHAN1DS A N D THE BLACK KHITAI As seen above (p. 252), following the victory o f the Uighurs over the Turk Khaghanate, the K harlukhs (form er allies o f the Uighurs) were also over thrown and driven westwards. A fter sporadic fights with the neighbouring
256
Relatives a n d neighbours
peoples, they settled around Balasaghun and Khashghar. The founder o f the new em pire, Satuk Bughra khan, died in 955. He was the first significant Turkic ruler to convert to the Islam ic faith. His son extended the realm , and hence the religion o f M uham m ad. The Islam isation process brought into existence a new culture and literacy in Arabic, and shortly, in the 11 th century, im portant works w ere conceived about the language and custom s o f the Turks. M ahm ud al-K hashghari wrote a great lexicon w hich contained a glossary o f m any hundreds o f w ords translated into Arabic, giving am ple exam ples o f w ord usage; also a vivid description o f the everyday life o f the Turks o f the 11th century w as included. K hashghari’s w ork is a truly invaluable source, as is Khass H ajib’s Wisdom o f R oyal Glory w hich he wrote in 1069. Although both works w ere w ritten by followers o f the Islam ic faith, they conjure up the archaic w orld o f the Turks prior to Islam isation, which w as, im portantly, contem porary to the M agyars’ Conquest o f the C arpathian Basin. In the lexicon we find the title yugrush, a position w hich (sim ilar to the Kharakhanids) cam e next after the khaghan o f the Avars; and the tudun, another im portant title am ong the Avars. Khashghari referred to m any other titles, some o f w hich occurred am ong the M agyars, such as ya b g u and tarhan. K hashghari described the tribal and kinship organisation o f the Turks and their “pagan” system o f beliefs. From the works o f K hass H ajib and Khashghari we can reconstruct the political organisation and the everyday life o f the Turks in that region. The K harakhanid em pire was overthrow n in the 12th century by the B lack Khitai, a branch o f the Khitai tribal confederation. The Black K hitai did not subm erge in Chinese culture, but m igrated w est when the Khitai Liao was overpow ered by another nom adic people, the Jurchen, the language o f whom was related to the M anchus’ language.
4. THE OGHUZ A N D THE SEUUK In the 9th century the Oghuz people had their dw ellings on the shores o f Lake Aral. Their centre was located in the north A ral area. W hen the Sam anid ruler crushed the K harlukhs in 893, the O ghuz tribes gained greater im portance. Their nam e features in the sources also as Uz or Ghuzz. O ghuz is identical to the tribal nam e Oghur, with the difference that the form er w as used am ong those Turks who spoke the so-called z-Turkic language. One group o f the Oghuz, under the nam e Seljuk, set out to build a new em pire w hich flourished under Togril khan, who died in 1063. The nam e Toghril has the m eaning o f ‘predator b ird ’ which happened to be the totem ic anim al o f the A rpadian M agyars, and w hose Hungarian nam e, turul, was borrow ed from Turkic. The
Eurasia in [lie 9[h and Writ centuries
257
Oghuz people played an im portant role in the history o f the M agyars, in so far as they defeated, in 893, those Pechenegs who subsequently invaded the M agyars in the Etelkoz (see pp. 330-331).
5. THE KHAZARS A N D THE RUS In the second h a lf o f the 9th century, the history o f the K hazar Em pire saw some im portant changes. The upper social stratum , w hich had converted to Judaism, becam e isolated from the rest o f the people. Their arm y was only m oderately supported by the w ealthy nobles, so they had to engage m ercenar ies. M ost o f these soldiers came from K hw arezm , and follow ed the Islam ic faith. Although the great Arab threat had been brought to an end, new enemies rose in the east. O f these, the Oghuz or Uz and the Pechenegs w ere unable to overthrow the Khazars, but nevertheless engaged considerable forces. The Khazars’ erstw hile allies, the Byzantine em perors, now sided w ith the nom ads, with the help o f whom they tried to bring down the Khazars. Their relationship with the w arrior-trader groups o f the Rus was more complicated. From the m iddle o f the 9th century they m aintained friendly relations, and organised jo in t raids in the region o f the C aspian Lake. Later, however, the Rus gained the upper hand in these raids, and the K hazars were left with only a share o f the booty. They then tried to tax the Rus m ilitary adventurers. As the Persian source, the H udud al-alam, rem arked “the wealth and the well-being o f the king o f the Khazars are m ostly from the m aritim e customs” . The trade relations changed to a substantial degree. The K hazars were weakened not only by the secession o f the K havars (which, in turn, strengthened the M agyars), but also by the dissidence o f other groups who joined the Arabs or w ent to seek their fortune in the B yzantine court or army. As viewed from the outside, the K hazars’ Em pire prospered in the last decade o f the 9th century and both the econom y and culture had reached their zenith. However, due to the workings o f the pow ers behind the scenes, decay had already began. In this procedure, the M agyars and later the Volga B ulghars played an im portant role. Their cautious but persistent m ove tow ards inde pendence took advantage o f the K hazar central po w er’s loosening grasp.
6. BYZANTIUM A N D THE DANUBE BULGHARS From the beginning o f the 7th century Slavic groups m igrated to the Balkans and Greece. It took the Byzantine em perors two hundred years to m oderately keep them at bay, or to settle them down. Efforts to “reoccupy” Peloponnisos and Hellas from the Slavs were only rewarded with success in the early 9th
258
Relatives and neighbours
century. Cities such as Thessaloniki rem ained bilingual. The aboriginal in habitants o f Thrace and M acedonia were fully Slavicised, and the Turkic rulers o f B ulgharia faced the same fate. The leader o f the so-called Thom as-rebellion (820-823) was a Slav soldier in the Byzantine army. In the 9th century, the struggle against the Arabs continued, but Crete (826) and M alta (870) fell into the hands o f the Arabs. However, in 843 the Em press Theodora, acting on b eh alf o f her son M ichael III, finally put an end to the dom estic struggles surrounding the icons which had begun in 726. The Iconoclast fraction was defeated, and the people w ere satisfied with the outcome. N evertheless, social problem s could not be solved straight away. Rooted in social confusion, the m ovem ent o f the Paulikians was put down only in 872. It had been one o f the first tasks o f the new M acedonian dynasty in Byzantium , whose first ruler was Basil I, father o f Leo the Wise (although certain sources claim that L eo’s actual father was M ichael III). Leo ascended to the throne in 8 8 6 . He was not only wise and learned, but was also an excellent organiser. He defended success fully the em pire against the Arabs in Syria, he beat back the Arab pirates in the high seas where they had attacked the Greek navy, and he protected the sm all islands, as well as cities like Thessaloniki. Leo was no warrior, he did not take part in m ilitary actions personally. As a pupil o f the Patriarch Photius, he took great interest in literature and scholarship, hence his sobriquet “the W ise” . His greatest challenge came, however, from the north. In 893, Simeon becam e ruler o f the Danube Bulghars. This brings us to the events which will be discussed in connection with the M agyar C onquest (see pp. 330-335). The details will be dealt with below, but it m ust be stressed that the tension betw een the Bulghars and Byzantium was o f an econom ic nature. The Bulghar ruler Krum proposed in 812 to regulate trade betw een B ulgharia and B yzan tium, and the m easures taken by Leo the W ise included trade restrictions to the disadvantage o f the Bulghar merchants. The return o f im age-worship did not ease the problem s with the papal Rome. Behind dogm atic discussions, claims o f suprem acy and the calcula tions o f Eastern and Western Em perors prevented reconciliation. The fate o f the Cyril and M ethodius legacy (see pp. 244-246) gives an idea o f the everyday aspects o f the problem . Things were dealt with far from Rome and C onstantinople. The very shaky equilibrium betw een Byzantium and the Danube Bulghars on the one side, and with Constantinople and Rome on the other, fram ed the events o f the M agyar Conquest.
Eurasia in the 9th and W di centuries
259
7. ROME A N D THE FRANKS To understand the situation in North Italy and the eastern part o f the Frank Empire, we have to go back a few steps in history. As we have seen, in Byzantium Basil I consolidated the central power. He was also successful in his struggle with the Arabs. One o f the first acts o f Basil was the deposition o f Patriarch Photius (867) and the reinstatem ent o f the form er Patriarch o f Constatinople, Ignatius (847-858, 867-877). As one o f their first acts Basil and Ignatius acknowledged the authority o f Rome. Pope N icholas I, who died on the 13th o f Novem ber 867, could not witness this gesture and the fruit o f his energetic activity, because the letter with the great m essage arrived after his death. One o f the m ost im portant reasons for the clashes betw een Rome and C onstantinople was the situation o f the Bulgharian Church. On request o f the Bulgharian ruler Boris I, Pope Nicholas I sent m issionary bishops to Bulgharia and am ong them was the bishop o f Porto, Form osus, the nam e o f whom we shall m eet later. This was unacceptable for Photius who convoked a synod to C onstantinople in August 867, which excom m unicated the bishop o f Rome. In fact this was an answer to the synod o f Lateran (863), which excomm unicated Photius. The successor o f N icholas I was Hadrian II who had other problem s. The prince o f Spoleto, Lambert, devastated Rom e, while King Louis II fought against the Arabs in South Italy. Basil tried to form an alliance with Louis II, but this failed. The Byzantine occupation o f Bari and Tarentum and the successful operations o f General N icephoros Phocas against the Arabs in South Italy naturally increased the Byzantine influence around the end o f the reign o f Basil, and with this, the im portance o f the western alliance decreased. A t the Fourth Council o f C onstantinople (869) Photius was officially deposed and anathem atised in the presence o f the legates o f the Pope. But then an unexpected turn followed. Three days after the end o f the Council, the delegates o f the Bulgharian ruler Boris I arrived. They asked for a decision, where their diocese should belong, to Rome or to Constantinople. This was a good strategy o f the Bulgharian ruler who wanted an independent B ulghar Church. The Council reassem bled and under the influence o f the eastern patriarchs decided in favour o f Constantinople. Boris also had problem s with Rome, because Rome refused two o f his candidates for the new archbishopric in Bulgharia. One o f them was Formosus. The Popes in Rome w ere afraid o f the growing influence o f Form osus. This m ade B asil’s hand free, and he succeeded in achieving the rem oval o f the Latin priests from Bulgharia. This sealed the fate o f B ulgharia and the Bulgharian Church, which has rem ained in the sphere o f the orthodox Church until the present time. On the other hand, the fate o f the Bulgharian Church was a negative exam ple for the Hungarians in the 10th century. In com pensation for the loss o f the Bulgharian Church
260
Relatives and neighbours
Hadrian supported the m issionary activities o f Cyril and M ethodius (see pp. 244-246). Hadrian died in 872 and his successor John VIII (872-882) took part in the political life o f the Frank Empire. In 875 he crow ned C harles the Bald, and not Louis the Germ an, as Em peror in Rome. In the next year he excom m unicated Form osus, the Bishop o f Porto as a potential rival. The son o f Louis the Germ an, Carlom an, invaded Italy and Charles the B ald had to flee. He died the sam e year (877). C arlom an asked for the crown. John VIII tried to gain time. The allies o f Carlom an, am ong them the prince o f Spoleto, however, im prisoned the Pope, who nevertheless refused to offer the crown to C arlom an and fled to Provence. In 881 John VIII crow ned the brother o f C arlom an, Charles the Fat. In Byzantium Em peror Basil recognised that it was a m istake to dism iss Photius, and recalled him. First he was entrusted w ith the education o f the E m peror’s children, am ong them the later Leo the Wise. A fter the death o f Ignatius, the Em peror offered Photius the patriarchal throne, who accepted it and convoked a council in Constantinople. The relationship betw een Constantinople and Rom e w orsened step by step. The C ouncil o f Constantinople nom inally accepted the rights o f Rom e in B ulgharia, however, now w ith the im portant reservation that the m issionary w ork o f the Greek priests should not be limited. In a counter m ove John VIII allow ed the use o f the Slavic language in the Church liturgy for M ethodius (880). The next Pope, M arinus I (also called in some sources M artinus II), w ith drew the excom m unication o f Form osus and his followers in order to get help from Charles the Fat. He could also persuade C harles to rem ove Prince G uido III from Spoleto, who tried to exert his pow er in Rome. M arinus died in 885 on his w ay to Worms, w here he was expected to arrive and help the ille gitim ate child o f Charles in the struggle for power. The follow ing Pope, Stephen V (VI), after the abdication and death o f C harles ( 8 8 8 ) asked A m u lf to protect him against the inner enem ies and the Arabs. A m u lf had, however, first to secure his pow er and could not send troops at once. T herefore Stephen changed his alliance and asked for the help o f G uido III o f Spoleto. The price was the coronation o f Guido as Holy Rom an Em peror in 891. In 8 8 6 Leo the W ise follow ed Basil and being afraid o f the great popularity o f Photius, he deposed the patriarch and raised his own brother Stephen to the patriarchal throne. Stephen o f Constantinople was w arm ly accepted by Stephen o f Rome. On the other hand, Pope Stephen ended the Slavic liturgy and backed W iching, the great enem y o f M ethodius and his followers (see p. 246). Pope Stephen died in 891. The Bishop o f Porto, Form osus, was over 75 w hen he was elected in 891 as the next Pope. The House o f Spoleto exerted its patronage and Form osus had again to crown Guido, and his son Lam bert as со-regent in 892. The tyranny o f the dynasty o f Spoleto was so intolerable that in the autum n o f 893
Eurasia in the 9th and 10th centuries
261
Form osus turned to Arnulf, who one year earlier asked for the help o f the Hungarians (see pp. 246,331). A rnulf first had to secure his kingship (894) and then m arched to Rom e in 896. In February 896 A rnulf was crow ned in Rome, but soon he had to leave, and Form osus died. Bonifatius, the next Pope, ruled only 15 days. The next Pope Stephen VI (VII) was first on the side o f Arnulf, but then changed to the House o f Spoleto and becam e an ally o f Lam bert o f Spoleto. It was he who organised the m acabre “Synod o f the C orpse” . They exhum ed the corpse o f Form osus and sat him on the throne. The Synod made a legal process against the dead. He was found guilty and his corpse was thrown into the Tiber. A bout the next two Popes, Rom anus and Theodorus II, we know practically nothing. Pope John IX (898-900) was a prot 6 ge o f Lam bert o f Spoleto. A Synod o f Rome withdrew the sanctions against Form osus, but also questioned the legitim ity o f the coronation o f Arnulf. A t the same time it was ruled that in the future the Pope should be consecrated in the presence o f the em issaries o f the Emperor. A second Synod in Ravenna regulated the relations between the House o f Spoleto and the Holy See. John could not calm down the fights in Rome betw een the parties o f the late Form osus and his enemies. Pope John died in January 900. The situation became so confused that it is unclear when the next pope, B enedict IV, was elected. The date can be put to either M ay or June. The political situation was also very shaky. Lam bert died on the 15th o f O ctober 898 and left no male descendants. The pretender to the Holy Crown o f the Rom an Em peror was King Berengar I o f Friaul, from 8 88 King o f Italy. In the w inter o f 899/900 the Hungarians as allies o f Arnulf, however, defeated Berengar. As we can see, in the years around the Hungarian Conquest, Rom e, North Italy and the eastern part o f the Frank Empire were w eak and deeply involved in internal troubles.
8. THE AVARS A N D SIAVS IN THE CARPATHIAN BASIN As referred to above (p. 213), the Avars had advanced up to the Caucasus area by 555. Around 560, the ruler o f the Avars was called Bayan whose nam e features Chuvash-type linguistic traits. We also know that the O ghur or U ghur people, who spoke a Chuvash-type language, assum ed a m ajor role in the formation o f the European Avars (see pp. 213-214). In 562, one group o f Avars fought the Frankish ruler Sigibert I at the River Elba. Its attacks were repelled, but in 566 the Avars defeated the Frankish armies, captured Sigibert who paid a ransom o f food supplies, and finally concluded an alliance with the Avars. In
262
Relatives and neighbours
562, the Avars reached the Low er Danube, and sought to contact Byzantium . A scending to his throne in 565, the Byzantine em peror Justine II refused to com ply with the A vars’ dem ands. C oncurrently with this the G epidae fought the Longobardi in the C arpathian Basin. The G epidae asked for and received support from Byzantium against their foes, on condition that they ceded Sir mium. However, having defeated the Longobardi, the G epidae did not live up to their prom ise m ade to Constantinople. In 566, in alliance with the Avars, the Longobardi crushed the Gepidae kingdom which the Avars subsequently occupied. However, the Longobardi were wary o f the proxim ity o f the Avars, so they solem nly agreed to withdraw, an event which took place on Easter Day in 568. The Avars, then, took possession o f the entire Carpathian Basin. In the 570s the Avars made frequent raids on Byzantium , at that tim e involved in its cam paign against Persia. In 574, they trium phed over Tiberius who had been elected em peror in the m eantim e, and whose im m ense annual taxes only could buy peace from the Avars. In the following years, however, B yzantium ’s Balkan estates suffered a series o f Slavic incursions, and in 582 the Avars took Sirmium. Bayan, who died around 583, was succeeded on the throne by one o f his sons. A round 583, various groups fleeing from the West Turks sought refuge with the Avars. There was a growing Avar and Slav pressure on Byzantium until 591 when the Byzantine em peror concluded peace with the Persian ruler, and was consequently able to disengage forces and launch them against the Avars. Follow ing fights o f varying success, Byzantium advanced up to the line o f the River Tisza, so the Avars joined forces with the Longobardi in N orth Italy and w ith the Franks. M eanw hile, having taken up the fights w ith Persia again, Byzantium sought to buy peace from the Avars by increasing its annual taxes. Until 620, Byzantium suffered a new series o f Avar and Slavic campaigns, to prevent which, even higher taxes were agreed on. In 626 Avar and Slavic arm ies besieged C onstantinople, w ithout success, however. In 627, the B yzantine em peror Heraclius trium phed over Persia with K hazar support. In 630, B yzantine-Frankish diplom atic contacts w ere established w hich contributed to the reinforcem ent o f B yzantium ’s position. Soon, inter nal fights broke out within the Avar Em pire. The conflict o f Avar and B ulghar pretenders ended w ith the victory o f the Avars, and one group o f the defeated B ulghars fled to Bavaria in 631; however, the Bavarians m assacred them. It has been suggested that the m ediaeval m ass-grave at Sankt Florian, near Linz in Austria, contains the bones o f these slaughtered Bulghars. Following K huvrat’s death around 670, one o f the O noghundur-B ulghar m igratory waves reached the Avar Empire. In 679, an Avar group o f envoys
Eurasia in die 9th and 10th centuries
263
participated in the festive events in Constantinople, celebrating victory over the Arabs. In 682, K uver and his group decided to m ove from the Avar Empire, and joined the Bulghars. In 768, an Avar delegation visited the court o f the freshly crowned Charles the Great (Charlem agne), king o f the Franks.
9. THE CARPATHIAN BASIN O N THE EVE OF THE CONQUEST In 788, the Franks crossed the R iver Enns from where they launched all o f their anti-Avar m ilitary operations. The Avars, who at that tim e possessed significant advanced posts in the Vienna Basin, m ade an abortive attem pt in 790 to com e to an agreem ent with the Franks regarding frontier m atters. In 791, the armies o f Charles the Great launched an attack against the Avars. One episode o f the w ar featured Erik, prince o f Friaul, who, taking advantage o f the A var’s fragile position, made a surprise attack on the unprotected seat o f the khaghan, robbing the town o f m any treasures w hich the prince transported to Aachen, to the court o f C harles the Great. In that year the khaghan died, as well as a person bearing the title o fyugurush. It seems that concurrently with the Frankish and Bulghar attacks, the Avar Em pire was being w eakened by an increasingly violent dom estic feud, too. N otw ithstanding the fortunes o f war, until 802 the Avars m anaged to hold onto their advanced positions as far as the Vienna Forest. The dom estic fights ended w ith the victory o f the faction that supported peace with the Franks, and in 796 the Avar E m pire’s second man, the tudim, visited the Frankish court where he was christened. In the following year the role o f the Salzburg archdiocese increased within the Avar m ission. It seems that, seeing the weakening o f the Avar Em pire, the Bulghars did not rem ain idle either. Leaping into action, the B ulghar ruler Krum launched significant attacks upon the Avars. Probably not wholly independently o f this, in 804 the Avar ruler, K hapkhan Theodorus, also journeyed to the Frankish court . His nam e, w hich has an East C hristian ring, attests to the fact that the Avars sought allies in B yzantium and its associates. However, Theodorus soon died. His successor, Abraham , was christened in the River Fischa in 805. Some sources refer to his son as Isaac. Following a b rief period o f respite, the Avar dom estic struggles broke out again. In 811, several Avar officials visited the Frankish court. Probably as a result, Charles the Great sent troops to support the Avar faction loyal to him. Eleven years later, in 822, the Avar khaghan and his train arrived at the imperial assembly in Frankfurt to vouch for his loyalty to the Franks. The Christian mission becam e a m uch-debated issue on the agenda o f A var-Frankish nego tiations. In 829 the F ranks’ outposts in the east w ere shared betw een the
264
Relatives and neighbours
Salzburg archdiocese and the Passau bishopric. The territories southeast o f the R iver Raba were ceded to Salzburg, the areas up to the Raba to Passau. As from the 820s the population o f Transdanubia underw ent signifi cant changes. South Slav refugees settled from the Drava region, especially the so-called Tim ochani, from the River Tim ok area. As seen above (see pp. 243-244), local pow er was seized by the Slavs after 840, and it was at that tim e that Pribina acquired his estates at Zalavar (M osaburg) where a church was consecrated in honour o f the Virgin Mary. It was no m ere chance that in the 860s Transdanubia often featured as Sclavinia in the sources. The name Karinthia was extended at that tim e to Transdanubia due to C hurch political reasons. The Carantans m entioned in the Conversio (described above, see p. 56) were, in effect, the people that populated Transdanubia. C reated in 870, the Conversio is probably the last source in w hich the Avars occur as a separate entity. It defined the inhabitants o f the territories under the suprem acy o f the Salzburg archbishop as being “the people that rem ained o f the Huns and the Slavs in those parts” (populum qui remansit de H unis et Sclavis in illis partibus). Naturally, the designation ‘H uns’, as used by the author o f the Conversio, is here to be understood as m eaning the Avars. In the m iddle o f the 9th century the political organisation o f the Avars m ust certainly have collapsed. This, however, did not entail the im m ediate disap pearance o f the Avars from the Central European scene. We have linguistic evidence in support o f the fact that the gradual assim ilation o f the Avars into the Slavic com m unity lasted at least three generations. A longside the Slavs, it was this near-bilingual (i.e. speaking a Turkic and Slavic language) popu lation undergoing Slavisation that the conquering M agyars encountered— in Pannonia at least. The Slavs were divided into groups o f South Slavs and West Slavs. Their centres were inhabited by Frankish, Bavarian and other Germ anic settlers. W hile m uch inform ation has com e down to us in the w ritten sources regarding the Slavic-Avar population o f Transdanubia, evidence o f these peoples in the Alfold (G reat Hungarian Plain) region is scanty. It is alm ost certain, however, that yielding to C hristian-Frankish rule, the Avars held groups w hich sought to shirk unpleasant duties. These groups disappeared in the Alfold. The Avar w ilderness (solitudo avarorum) denotes these barren Alfold lands occupied by the Avars. The treasure find known today as the N agyszentm iklos treasure m ay have been the w ealth o f a fleeing Avar high dignitary. We do not know w hen it was hidden, but parallels with the Szarvas finds bearing sim ilar runiform inscrip tions, seem to indicate that the treasure was last “used” in the latter h alf o f the 8 th century or possibly the first h alf o f the 9th. It has been established that the treasure underw ent a “Christian phase” , but later the jugs were converted so
Figure 62 Peoples in the Carpathian Biasin in the 9th century. Avar and Avar-influenced Slavic archaeological sites in Transdanubia and A ustria
266
Relatives and neighbours
as to enable them to be worn on belts when riding. This process neglected the Christian motifs (see p. 131). The m iddle o f the 9th century saw an influx o f Slavic settlers to the Upper Tisza and the River Szamos regions, as well as to the M aros valley. At the time o f the M agyar Conquest the Maros valley was under B ulghar supremacy. The fortress o f C songrad (Chorniy Grad ‘black castle’ at the m eeting o f the Rivers Koros and Tisza), too, served as a B ulghar or B olghar-S lavic guard outpost (see p. 242). As a whole, it can be said that at the time o f the C onquest, there was no significant political organisation in the Carpathian Basin. Transdanubia fell under the influence o f the Franks, the K isalfold (N orthw est Plain) and the northeast areas under the M oravian Principality o f Nitra, while the M aros valley and Sirm ium w ere under the Danube Bulghars. The population o f the Carpathian Basin, however, was highly varied. It featured West, South and East Slavic groups, Slavising Avars, Franks and Bavarians could be found in the tow ns and in the vicinity o f the churches, as well as a great m any other peoples— all w ithout form ing a truly hom ogeneous population. This factor, am ong others, accounts for many as yet unclear points in the history o f the M agyar C onquest and the M agyars in general.
NOTES The most recent overview o f the history o f the Uighurs is Sinor (1997). See further Hamilton (1955, 1962), Mackerras (1972), Senga (1990), Golden (1992), and for the late history o f the Uighurs Maljavkin (1983) and Pinks (1986). For the history o f the Khitai, the basic work remains W ittfogel-Fcng (1949), the linguistic parts are, however, full o f unfounded claims. M ost o f them are due to a misunderstanding o f the source, m ixing later Jurchen replacements with their genuine Khitan originals. The consultant for Altaic linguistics was M enges (1968). The remarks o f H. Franke (1969) are important; Sinor (1995) is interesting. For this section, as for other Khitai matters discussed in this book, I have used the text o f my lectures on the language o f the Khitans given in 1995/96 and 1996/97 at the University in Szeged. On the Kharakhanids see Pritsak (1 9 5 3 -1 9 5 4 ), G olden (1992, pp. 2 1 4 -2 1 6 ), on the Kharakhitais see W ittfogel-Feng (1949, pp. 6 1 9 -6 7 4 ). The Oghuz and the Seljuk are dealt with in G olden (1992, pp. 205-211 and 2 1 9 -2 2 5 respectively). For the history o f the Khazars see Dunlop (1954, reprinted 1967), Artanomov (1962), Pritsak (1978), Ludwig (1982, 1983); G olden (1983, 1992, pp. 2 3 3 -2 4 4 ); M agom edov (1983). My papers on the Khazars (1982a, 1983a) were published in Hungarian with English summaries. For an overview o f Avar history see Bona (1984a, 1984b), Pohl (1988) and Bona (1994a). The sources were published in Hungarian in Szideczky-K ardoss (1 992), and more recently in Olajos (1996). I drew on Wolfram (1979a) for the chronicle C onversio bagoariorum et carantanorum jxnA recently used the new edition o f LoSck (1997). On the treasure o f N agy-
267
Eurasia in the 9th and 10th centuries szentmiklos see L aszlo-R acz (1983a, English version 1984),
and G obl-Rona-Tas (1995,
especially the bibliography o f the latter, pp. 6 4 -7 7 , compiled by Maria Ivanics). Two basic works can be recommended on the history o f Byzantium; both have been many times re-edited, amended and revised: G. Ostrogorsky (first edition 1 9 4 0 ,1 have used the 1993 reprint o f the second English edition) and A .A . Vasilev (first published in 1917, first English edition 19 2 8 /2 9 ,1 have used the 1984 reprint o f the edition o f 1958). A good handbook on the history o f the Popes is by J.N.D. K elly (1988a, 1988b).
PART THREE
FROM THE URALS TO THE CARPATHIAN BASIN
VI. THE NAMES OF THE MAGYARS BEFORE THE FOUNDATION OF STATE
1. ETHNIC NAMES: CHARACTERISTICS A N D ORIGINS We have seen the im portance o f an ethnic nam e in defining a people or an ethnos. Its im portance is o f sim ilar stature in the sources, which are often responsible for attaching nam es to ethnic groups. So before discussing the various ethnic nam es o f the M agyars, let us briefly exam ine the general issues surrounding ethnic names. Ethnic names fall into two basic categories according to their use. The first category is made up o f self-designations, the names by which peoples call them selves, and the second is that o f external designations, the nam es used by outsiders. A lthough these two categories are quite sharply delineated, there can still be historical transitions between them. An ethnic group can even learn and call itself by a name which was originally applied by outsiders. This is the exception, however. M ore com m on is for the name o f one people to be transm itted to another and then be adopted as its own. This has happened to the French and the Russians. Frangais derives from the Franks, w ho were originally a Germ anic tribe. The name Rus, by which the Russians call them selves, is not Slavonic, but one o f the nam es o f the northern Germ anic Vikings, although it could hardly have been the nam e that a Viking ethnic group applied to itself, and is much more likely to have been assigned to ‘red-hairs’ or, as others think, to ‘row ers’. Such considerations lead to the distinction betw een prim ary and secondary self-designation. A self-designation is prim ary if it has been assigned to the ethnic group from its own language. These designations generally have the m eaning ‘m an’, ‘hum an b eing’ or ‘person’. The group calls itself person, the rest strangers, barbarians, speakers o f unintelligible languages. For exam ple, the self-desig nation o f the Zyryans, Komi, m eans ‘m an’, and is historically traceable to the same ancient word as the Hungarian word him ‘m ale’. The nam e o f one o f the Sam oyed peoples, Selkup, m eans ‘forest m an’ and the -kup is historically equivalent to the words kom i and him. The self-designation o f the Cheremis, Mari, is o f Indo-European origin, where it m eant ‘person, young person, young m an ’. The second com ponent o f the Votyak’s self-designation, the murt part o f the word Udmurt, is also an Indo-European word, w hich has the
272
From the U ra b to the Carpathian Basin
m eaning ‘person, m ortal’ in Indo-European languages, and is ultim ately related to the Latin w ord m ors ‘death’. We shall see that this kind o f self-designation, involving a word m eaning ‘m an’ in a different language, will also be m anifest in the ethnic nam e o f the M agyars. Secondary self-designations arise from the nam es o f other peoples, and ultim ately derive from prim ary self-designations or external designations. The vast m ajority o f ethnic nam es are external designations, and there are com m on types o f these. One group o f ethnic nam es involves geographical designations. Exam ples o f these are Latin (< Latium ), R om anian (< Rom e), and Am erican (< America). In very rare cases, an ethnic nam e derives from an ancestor, or founder o f an empire. Such is the nam e Osman or Ottoman w hich designates the people o f the Turkish Em pire. It derives from the origi nally A rabic nam e o f its founder Osm an (1259-1326). The tw o form s o f the nam e arise from two non-A rabic pronunciations o f the original. The Kipchak-Turkish ethnic nam e N oghay also derives from a personal nam e, w hich in turn com es from a M ongolian w ord, nokhai, m eaning ‘dog’. It is know n o f the Uzbeks that some hundred years after the death o f the historical, but subsequently legendary, U zbek Khan, an ethnic group started to call itself Uzbek. N o such ethnic nam es have been found w hich em erged before the M ongol Em pire, i.e. before the 13th century. This type should not be confused with the cases in w hich a personal name is the sam e as that o f a tribe or people. I f there is a tribe called Yaghlakhar, and sources m ention a person called Yaghlakhar, this only m eans that he took his nam e from the tribe, and not that the nam e o f the tribe derived from a person o f that name. Chingis K han’s younger brother was called Khasar, which is the original form K hasar o f the ethnic name Khazar. The reverse is very common. Personal nam es are fashioned out o f existing ethnic nam es, and the origin o f the people is traced to a person who never existed in history. Thus the hero o f a legend related in a H ungarian chronicle, M agor, got his nam e from the ethnic nam e M agyar, along w ith his partner H unor who was invented from the ethnic nam e Hun. A sim ilar history is know n from the report by M arco Polo. A ccording to him , the descendants o f Gog and M agog w ere the U ncs and the M ongols. Unc was originally the nam e o f the Nestorian K ereit king O ng-khan, but later his nam e was confused with the name o f the Onggiit, likewise a N estorian group in C entral Asia. The nam e o f the M ongols was changed to M oghul in W estern Asia fairly early on, hence the nam e o f the M oghuls o f India. The resem blance o f Ong and M ogh-ul to Gog and M agog w as enough to build a w hole story in the 13 th century, and claim that the M ongols w ere the descendants o f M agog. There is a very rare type o f nam e whose origin really does incorporate a personal nam e. This is the designation o f a people by its leader’s title. O ne o f
The names o f the Magyars before the found ation o f state
273
these is the nam e kereI or keral, which is found in the 13th-century The Secret History o f the M ongols and in som e other sources. It designates a people with a king, and refers to the M agyars. The Hungarian title kiraly ‘kin g ’ derives through Slavonic m ediation (krai ‘kin g ’) from Charles the Great: Karl. We saw earlier that the nam e o f the Khazars is ultim ately derived from the Latin personal nam e Caesar, but this firstly gave rise to a title (cf. K aiser in German, csaszar ‘em peror’ in Hungarian via Slavonic, and czar in som e Slavonic languages), appearing am ong the Iranian peoples in the form kesar, and the title subsequently becam e khasar am ong the Turks. The ethnic nam e “A m eri can” originated from the nam e o f a continent nam ed after A m erigo Vespucci. The Hungarian nam e Em ericus (like the nam e o f the son o f K ing Saint Stephen) is a version o f the G erm an Henrik, and from Em ericus evolved Am erigo, w hich was fashionable for a while in northern Italy alongside Enrico, which com es directly from German. So the M agyars also had a historical part in the nam ing o f America. Exam ples can also be found o f an ethnic name deriving from a title which designates a leader or an honoured figure, but w hich cannot be traced to any personal name. Such is the em ergence o f the ethnic nam e y a b g h u from a title o f unknow n origin. A title like this can often becom e a personal nam e itself. M any A rpadian M agyar nam es o f Turkic origin derive from titles. The original form o f the Hungarian nam e Geza is Gyeiicsa (Gyeucha), based on the word gyeu, which can be traced to the tM eyabgu. Thus tribal or ethnic nam es often m anifest a secondary correspondence with personal names. In m any cases, the ethnic nam e refers to the organisation o f the people. Such is O noghur ‘ten O ghurs’, o f w hich m ore will be said later, and such was the designation o f the M agyars as hetm agyar ‘seven M agyars’. Ethnic nam es com pounded with num bers are particularly com m on am ong Turkic ethnic groups, such as N ine Oghuz (Tokhuz Oghuz), Three K harlukh (Uch Kharlukh), Thirty Tatar (Otuz Tatar). Although these are actually nam es o f tribal alli ances, they often becam e ethnic nam es later. Particularly com m on is an ethnic name m ade up o f a leading people or tribe, such as K hazar or Tatar. Ethnic nam es can be based on the nam e o f the land, for exam ple M ongol H oi-yin lrgen ‘people o f the forest’, although there has been m istaken classification o f the ethnic nam es A katir and A katsir as this type. One type o f external designation is a reference to the unintelligibility o f a people’s language. The Hungarian w ord for G erm an, nemet, is o f Slavonic origin, nemets, and com es from the same root as the Slavonic word nem oy ‘dum b’, i.e. the Germ ans are the people which to their Slav neighbours were dumb, speaking an unintelli gible language. In m any cases a typical feature o f an ethnic group is w here its name stems from. The Tirpakpeop\e who live around N yiregyhaza in H ungary speak a dialect o f Slovak in which the very com m only-used Slovak w ord teprv
274
From rite Urals со the Carpathian Basin
‘only, ju s t’ is pronounced trpov. Old R ussian sources called the Cum ans o f South Russia Polovets, the root o f the R ussian word originally m eaning 'p a le ’. This probably referred to their unusual anthropological features o f fair skin and hair, and is unlikely to carry any m eaning involving colour symbology. In any case, the m ediaeval G erm an nam e for the Cum ans (Falben ‘the p ale’) contains the sam e reference, and it is possible that underlying their H ungarian ethnic nam e К ип is a Turkic word o f sim ilar m eaning. The Turkic-speaking Kharakhalpakh are nam ed after their unique dress involving a black hat (khara ‘black’ and khalpakh ‘h at’). There are rare instances o f a people being nam ed after its leading tribe. This m ay be the origin o f the Turk people’s name, but it is not as com m on as notable researchers o f the subject used to think. They took the m ethodologically inadm issible approach o f not distinguishing ethnic nam es from the nam es o f tribes and clans. It is usual to m ention am ong clan nam es the so-called totem istic nam es deriving from descendancy from sacred anim als, or m ore rarely plants. M ost o f the known cases are not ethnic, but clan nam es, and m any are derived from personified anim al ancestors or ancestors with nam es identical to anim als. It is rare for a clan nam e to becom e a tribal name, and even rarer for a tribal nam e to becom e an ethnic nam e, and even am ong the enorm ous am ount o f Turkish data there is not a single exam ple to be found o f an ethnic nam e with a totem istic origin. This is surely not a coincidence. Ethnic groups’ self-designations are never based on outstanding faculties or historical events. Ethnic nam es ascribed to words m eaning heroic, brave, strong, outstanding, etc., or surviving, rebelling, or “a fragm ent”, do not stand up to etym ological criticism. N either is it conceivable that people could col lectively decide to express consciousness o f them selves in such a way. T he sam e people is often known under different nam es by its different neighbours, usually after a section o f the people or a tribe w hich lives close by or has an intensive relationship with them. The French call the Germ ans Allem ande (after the A lem an tribe), and the Finns call them Saksa (after the Saxon people), the English use the Latin-based German, and the Russians nem ets, from the sam e root as the corresponding H ungarian w ord nem et, from nem oy ‘du m b ’. Som etim es the nam e o f the people is the sam e as that o f the country, like Germans and G erm any, and som etim es not, like Russian nem ets for ‘G erm ans’ and Germ aniya for ‘G erm any’. Although, as will be shown later, som e ethnic nam es can be relatively easily linked to a com m on word in the language, som e originally only designated the people, and were artificially related with com m on w ords w hich sound the sam e or very similar. The m odern self-designation o f the B ashkirs is Bashkort. The B ashkirs now adays trace this to two Turkic words, bash m eaning ‘h ead’
The names o f the M agyars before the foundation o f state
275
and khort m eaning ‘w o lf’, ascribing to their own nam e the m eaning ‘head wolf, w o lf leader’. This explanation, which is unlikely even on typological grounds, fails, apart from anything else, with knowledge o f the old forms o f the B ashkir ethnic nam e (discussed below) which rule out this interpretation. There are hundreds o f these kinds o f secondary folk etym ologies. From Isidorus o f Seville to the Turkom an Abul Gazi, sources contain an incredible num ber o f false popular etym ologies. This also applies to nearly every explanation that the Hungarian Anonym us gave for ethnic nam es, but in so doing he was m erely following the scholarly custom o f his time. We should also draw lessons from cases where we can check the older form o f an ethnic name by looking at the people’s history, noting that the history o f most ethnic names is unknown. Put it another way, we only have any chance o f determ ining the origin and etym ology o f an ethnic nam e i f we know its history, its earlier forms and their use. I f we only know the m odern form, then it is only very rarely that there is any worth in m aking attem pts at its etym ol ogy. The origins o f m ost ethnic nam es are obscure, because they are passed from one language to the next. The designation o f m any extinct peoples and languages persists. There is no longer any trace o f the Alem ans, after whom the French nam ed the Germans. W ell-known ethnic names include m any based on m istaken designation or misunderstanding. The Am erican Indians have no relation with India, except to the extent that Colum bus thought he had reached India. There are several kinds o f m istaken or m isleading ethnic nam e designations. The m ost com m on type is found in early w ritten sources. A ncient Greek, C hi nese or Persian authors were not interested in precisely identifying peoples. It was m ore im portant for them to assess their m ilitary strengths, the extent o f their organisation and their way o f life. So if a nom adic people was replaced by another in the same place, the old nam e was applied to the new people. The Greek sources called a whole series o f peoples Scyths, and the sam e has been done with the nam es Hun, Avar and— significantly for the M agyars— Turk. Ethnic nam es are only helpful to the historian if it can be precisely determined, for an external designation, when and which people used the name. Uncovering the story o f the name can also help. For a self-designation, the task is to determ ine the time when the people started calling them selves by the name, and to find the origin o f the name.
2. TURK The M agyars are called Turks in several sources. Som e tim e after 904, the Byzantine Em peror Leo the Wise wrote a book containing descriptions o f m ilitary m easures taken as protection against the nom adic peoples. This work
276
From the llra ls to the Carpathian Basin
was an adaptation o f a m anual w ritten by M aurice at the end o f the 6 th or the beginning o f the 7th century. M uch that he wrote, therefore, concerned the Turks o f the 6 th century, but in the parts involving new data there are m any references to the M agyars. In the Byzantine G reek text, the nam e o f the Turks had the form Turkoi. The G reek ending -oi— w hich w as probably pronounced /i/ at that tim e— was appended to the ethnic nam e to m ake it fit into the Greek word-final scheme. One o f the reasons it is know n that Leo used the nam e Turks for the M agyars is that in 902 the court poet, Arethas, w rote the Em peror a table-hym n o f praise in w hich the ethnic nam e Turkoi is w ithout doubt applied to the M agyars. However, it can be shown that Turkoi appears in m any places in L eo’s w ork referring to peoples other than the M agyars, and not always Turkic peoples, whose m ilitary practices and w ay o f life w ere the sam e as those o f the Turks. For Leo, the Turks were part o f the Scyths, sim ply by virtue o f their way o f fighting, as it resem bled the descriptions o f Scyths in the much older Greek sources. Exactly where the nam e identifies the M agyars can only be determ ined by very careful analysis o f the text. Constantine Porphyrogenitus’s work De adm inistrando imperio, discussed above (see pp. 46-49), includes several detailed accounts o f the M agyars, and he called them Turk on each occasion. He called the new lands occupied by the M agyars Turkia. But he also carefully distinguishes them from Turkic peoples such as the Khazars, the Pechenegs and the Bulghars. Since Constantine Porphyrogenitus received much o f his inform ation from the M agyars them selves, the question arises as to w hether the M agyars also called them selves Turks at that time. There is no direct evidence for this, though, w hich m akes it unlikely. W hen I m y self go abroad, I say to an English-speaker that I am a H ungarian, not that I am a M agyar. So when A rpad’s great-grandsons Termecsii and Bulcsu, visited the B yzantine court and spoke in a foreign language, then they would also have spoken the nam e o f the M agyars in the foreign tongue. It is likely, however, that they were assisted by a Slav interpreter who translated into Greek. W hat we do not know is w hether Termecsii and Bulcsu spoke M agyar to a Slav interpreter w ho also knew their language, or, and this is m ost likely, Slavonic, which the interpreter then rendered in Greek. The Slav interpreter w ould quite definitely have know n that the M agyars were know n in G reek as Turkoi, and w ould have used this name. The visiting M agyars’ statem ents were probably not delivered directly to the Emperor, but carefully noted down in the court. The scribe w ould probably have w ritten Turkoi even if the interpreter or the M agyars got their languages confused and accidentally used the M a gyar or Slav name. In any case, whereas every use o f Turkoi in Leo the W ise’s work m ust be analysed very carefully to determ ine w hether or not it refers to the M agyars, in Constantine Porphyrogenitus’s w riting it practically always does.
The names o f the M agyars before the foundation o f state
277
The Holy Crown o f Hungary, although norm ally referred to as Saint Stephen’s Crown, was not actually received by Saint Stephen. The crown is m ade up o f two parts. The low er part, which the B yzantine em peror sent to King G eza I (1074—1077) is called the Greek Crown. The upper part is the Latin Crown. Although there is a m inor argum ent am ong experts as to the exact date that the crown was made and sent, it w as definitely betw een 1067 and 1075. W hat is interesting for us is the inscription on the G reek Crown: Gyeovicha, the loyal king o f the Turks (Geovitsas pistos krales tourkias). The m odern Hungarian nam e G 6 za is based on the erroneous reading o f an old text, but was so widespread by the time the correct reading was established that it has rem ained in use. It was pronounced in the H ungarian language o f the time as Gyeiicha, Gyeicha or Gyevicha. This is confirm ed by the inscrip tion on coins m inted by Geza I, where the kin g ’s nam e is w ritten in the form GEVCA, from which the form Gyeiicha or G yeucha can clearly be inferred. Even the inscription on the Holy Crown, the m ost im portant relic o f Hungarian history, tells us that its w earer is “King o f the Turks”. The nam e o f the king, Geza, is based on the Western Turkic equivalent (jevu) o f the eastern Turkic title yabghu. This title had long been adapted into the H ungarian language, because the word ending is Hungarian, probably with a dim inutive function. The title krai > kiraly ‘king’, how ever,'as m entioned earlier, passed to the M agyars from the Slavs. We will later consider in more detail this peculiar coexistence o f Turkic, M agyar and Slav elements. Later Byzantine sources also refer to the M agyars as Turks, but this is ju st the continuation o f the practice started at the time o f the Conquest, and so this data can be set aside. The M agyars are also called Turks by certain authors w riting in Arabic. The Arab author Ibn Hayyan wrote his w orks in Spain around 1075, but used 10thcentury sources. His description o f the adventuring detachm ents o f M agyars who reached the Lerida area o f Spain refers to them as Turks (al-tiirk). A lm ost contem porary to him, also in Spain, was al-Bakri, who also uses the nam e Turk for the M agyars. The Um ayyad dynasty which ruled Spain at the time had good connections w ith Byzantium , and so it is probable that the Arab authors also adopted the Greek designation. B efore exam ining w hy the M agyars were called Turks, and the origin o f the nam e, we m ust first attem pt to establish w hether the M agyars called them selves Turks at the tim e o f the Conquest, or w hether they knew or used the nam e at all. In the oldest surviving text containing continuous Hungarian sentences, The D eed o f Foundation o f Tihany Abbey (1055), there is a description o f the estate boundary beside the Balaton which includes the nam e o f a sm all lake. The boundary first led to a low hill, and from there to Lake Turku (adlacum Turku).
278
From the Urals to the Carpathian Basin
This word should definitely be read Tiirkii, but the Hungarian vowel ending -ii had the same purpose as the Greek -oi w hen applied to foreign names: it served to adapt it to the structure o f the Hungarian language. This is held to be likely even if—as we shall see— there m ight have been an -ii at the end o f the original ethnic name. In The D eed o f Foundation o f Tihany A bbey nearly every word which later becam e m onosyllabic ended in a vowel (varu, hodu, utu, babu, hurhu, harmu, liku, szeku, szilu, which w ithout the -u are nearly the same as the m odern Hungarian words for ‘castle’, ‘w a r’ or ‘m ilitary’, ‘ro ad ’, ‘b ean ’, ‘dried-up route o f stream ’, ‘three’, ‘hole’, ‘ch air’ and ‘e lm ’). Some also occur as m onosyllabic words, which shows that the closed word-final vow els at the end o f later m onosyllabic w ords had already begun to disappear, as in ohut = o u t ‘old road’ (but: utu ‘road’). The Turk ethnic nam e also occurs in som e A rpad-era docum ents as a personal and place nam e, in the form Thurk, Turk, Turuk, Turcu, and later Thewrewk, Twrek, etc. Rivers can also be designated by ethnic nam es, and although this is a rare category, there is a village called O roszpatak, ‘Russian riv er’ in eastern Slovakia, and one o f the nam es o f the Black Sea ju st at the tim e o f the Conquest was the K hazar Sea. This is a m ore likely explanation than that the lake was called after som ebody called Turk, or that the lake’s nam e actually involved the Hungarian word torek, ‘c h a ff. In any case, it is certain that although the Byzantines called the M agyars Turks, contem porary sources rule out the likelihood that the M agyars, who knew the nam e, used it for them selves. The M agyars probably used the term Turk for one or all o f the Turkic-speaking ethnic groups who lived am ong them. The old w ord tiirk in the H ungarian language becam e first tiirkii, then tiiriik, and finally torok. After the O ttom an Turkish conquest, use o f the term revived and subsequently spread to encom pass every Turkic-speaking people. In order to take a closer look at where and w hy the M agyars w ere called Turks, and by whom , we m ust first consider the history o f the ethnic name Turk. Since the original form Tiirk is the origin o f both Turk and Turkic, we m ust first m ake clear the conventions o f using Turk for ethnic groups and Early Turkic political groupings, and Turkic for peoples who speak Turkic languages and for the languages themselves. A round 551 A D , a new ethnic grouping established an em pire in Inner Asia. This group called itself Turk. It had its own runiform system o f w riting, and referred to itself in inscriptions from the beginning o f the 8 th century as the Turk people (Tiirk Bodun). In the oldest inscriptions the w ord is w ritten with a kind o f letter [k], which according to the reading rules, should be read with a preceding or succeeding vowel, either /й/ or / 6 /. The possible readings are therefore Tiirkii or Tiiriik, or less probably Tiirko or Tiirok. The reading w ould also allow the first vowel to be / 6 /, but because o f other transcriptions and the
Thf names o f the M agyars before the foundation o f state
279
history o f the w ord this m ust be excluded. In later records, this particular spelling disappears, and only Turk can be read, so that the readings Tiiriik and T iro k can also be excluded. It is puzzling, however, that the /и/ vowel at the end o f the w ord is not written with the letter [ti] that the writing system included, and according to the spelling custom, it was m ore com m only om itted within the word, but always written at the end. The only reason for this could be that the /и/ vowel was atrophying, had a reduced pronunciation, so that it was very short, subsequently dim inished, and soon disappeared. It is known that such vowels in other words were not included explicitly in Early Turk runiform writings. The vowel was sim ilar to the closed vowels in final position in Old Hungarian, /и/, /и/, /i/ and /е/. The ethnic nam e Turk was w ritten down at a very early stage by literate neighbours. The Chinese recorded the nam e in their hieroglyphic system as tujue. Reconstruction gives the Chinese transcription as being the rendering o f a foreign form tiirkiit. Although some thought that this form was a form o f the ethnic nam e Turk with a Turkic or M ongol plural ending, it can nowadays be stated with confidence that this form did not pass directly into Chinese from the Turks, but via the Sogdians. The Sogdians (see pp. 196 and 214), a trading people who spoke an Iranian language, were very prom inent in the Turk Empire and Asian trade, to the extent that the earliest Turk inscriptions were written in the Sogdian language and writing system. On the recently discov ered Sogdian-language B ughut inscription, which refers to a Turk ruler and dates from around 572, the nam e o f the Turks is w ritten as tiirkiit (trkwt). We are dealing here with the Sogdian plural ending or perhaps with the Sogdian -kut adjective ending, and the original form turkkut sim plified so as to becom e turkut or tiirkiit. A nother form which can help in revealing the origin, apart from the Chinese and Sogdian, is the Khotan Saka nam e for the Turks. Several hundred years prior to the em ergence o f the Turks, an Iranian-speaking people, the Saka, settled in a Turkestan oasis town called Khotan (see p. 196). This people took up the Buddhist religion and used a version o f Indian script for their own language. The ethnic nam e ttrukd appears in Khotanese Saka sources. The final letter [a] at the tim e o f writing no longer indicated a vow el at the end o f the word, and the initial tt is a device to ensure that the word is read with a t. No distinction is made in the language, or in its script, betw een the /и/ and the /й/. Thus the Khotanese form m ust be read with the pronunciation truk or triik, and this derives from an older form, truka or triikii. If we consider w hat relation there could be betw een the tiirkii in the Turkish inscriptions and the triikii o f the Khotanese Saka sources, then the only conclusion we can arrive at, how ever surprising, is that the Khotanese data are older and m ore original. The tiirk or tiirkii form could, w ithout any problem , have been rendered into
280
From the llr a b со the Carpathian Basin
Khotanese script, but it was not. In contrast, the phonetic cluster tr is not possible in the Turkic languages, and m ust certainly have been assim ilated. The Turks had several ways o f doing this, one o f w hich was m etathesis. A fter some tim e it becam e possible in the Old Turkic languages to write the -rk juncture (see Old Turkic erk, ‘strength’, berk ‘h ard ’, irk, ‘divination rune’, ark, ‘chattels’, ‘m ovable property’, etc.). The initial juncture tr in foreign words w as also for a long tim e assim ilated in the H ungarian language too. This is how the Slovakian trpak becam e tirpak (a H ungarian Slovak ethnic group) in Hungarian, and the Slavonic troszk becam e tarack ‘couch grass’. Thus, the Turkic language fitted the foreign w ord truku or triikii into its own system as tiirkii. This could not have occurred the other w ay round, because the tiirk or tiirkii form could have been transcribed into K hotanese quite straightforwardly. The question is w hether this first, apparently surprising, result that the ethnic nam e Turk is o f external origin is supported by other evidence. Judgem ent is assisted by Tibetan-language sources. The Tibetans estab lished their first great em pire in the 7th century, w hich was soon to play a m ajor part in Asian history. They adopted B uddhism , and Indian script reached th e n f via in n e r Asia. M any w ritten sources, chronicles and annals are know n from the m iddle o f the 7th century: hundreds and thousands o f docum ents o f Tibetan central power. They have m ostly been discovered from excavations, but have also survived in walled-in libraries and in the form o f preserved original inscriptions. Tibetans called the Turks by the nam e D ru-gu. This form preserved the original Khotanese Saka form Triikii. The Tibetans could also have transcribed the name Tiirk. The Turkic tribal nam e w hich later becam e known directly as Tiirgesh or Tiirgish is found in Tibetan sources in the form Dur-gyis. The 7 t h - 10th-century Tibetan sources therefore provide im portant evidence. It w ould not be the only case o f a people’s self-designation being originally external: it is enough to think o f the self-designations o f the French and the Russians. However, the historical question o f how it occurred rem ains, and w hether there is historical evidence to support it. It has long been noted by researchers that the clan nam e associated w ith the first rulers o f the Turk Em pire founded in the 6 th century was A shina, and that this and the nam es o f the first rulers and their know n relatives w ere clearly not Turkic. It is highly probable that we are dealing w ith a royal fam ily and clan o f Iranian origin, alm ost certainly Saka, w hich after gaining dom inion over the Turks becam e Turkified itself. This is supported by evidence from the latest research that the clan nam e A shina was o f Saka origin. The m eaning o f the Saka w ord asseina is ‘b lu e’, and the Turkic dynasty calls itself in m any inscriptions by the nam e kok tiirk, i.e. the Blue Turks.
The names o f d ie M agyars before die fo u nda tio n o f state
281
A strange story is found in the Chinese annals. The annals o f the Sui dynasty (betw een 636 and 656) state that the ancestors o f the Turks settled in the Altai Mountains, where they specialised in the occupation o f metal forging. Since the Altai M ountains look like a helmet, and this “helmet” in their language sounds like tiirk, this is why they called them selves by this name. So far no Turkish language has been found which has a w ord m eaning ‘helm et’ or any- thing sim ilar w hich sounds rem otely like tiirk. By contrast, the w ord for ‘lid ’ in Saka is tturaka. This at least m eans that Khotanese Saka speakers or those with a Saka tradition could have contributed to the form ation o f a popular etymology. O f the m any etym ologies o f the T urks’ nam e, the m ost popular to date, and for a while m ost acceptable, was that it derived from a Turkic word m eaning ‘strong’. But it has turned out that the com m on w ord which occurs in O ld and M iddle Turkic languages and w hich is w ritten in the form tiirk, does not m ean ‘strong’. There is a w ord o f this form, but it means ‘the m iddle o f the time o f the ripening o f the fruit, the resting time o f the day, or young people’s first period o f m aturity’. The ‘strong’ m eaning o f the word lies only in one com ponent o f erk tiirk. The first part o f this m eans ‘strong’, and in fact the H un garian word его ‘strength’ is also derived from this Turkic word. The m eaning o f the second part is approxim ately ‘flourishing, being in full strength’. The expression only m eans ‘strong, flourishing’ with both o f the w ords together. We can now reconstruct the history o f the ethnic nam e Turk as follows. The word is o f East Iranian, m ost probably Saka, origin, and is the nam e o f a ruling tribe whose leading clan Ashina conquered the Turks, reorganised them, but itself rapidly becam e Turkified. In the process, they changed their original name triikii to fit the Turkish pronunciation, so that it becam e tiirkii and then tiirk. The Turkified ruling class was prepared to accept this change because the new nam e sounded like the Turkish com m on w ord tiirk m eaning ‘flour ishing, being in full strength’. Behind this designation by the leading class would, no doubt, have been some kind o f word m agic w hich hoped that applying the nam e w ould c a n y this m eaning with it. Such a b elief is also found among other A siatic peoples. A fter the foundation o f the Turk Em pire the nam e o f the leading tribe passed to the tribal federation and all o f its peoples. The only m odern Turkic-speaking peoples which do not recognise the old nam e Tiirk are the Yakut and the Chuvash, the latter only learning it later from the Turkic-speaking K hazan Tatars. This shows that the ancestors o f these peoples had split o ff from the Turkic-speaking com m unity m uch earlier. The Turk Khaghanate subjugated the Turkic peoples who had arrived before them on the Eastern European steppe, and Turk em issaries appeared at the Byzantine court by 568. From then on the B yzantine sources refer to the Eastern European Turkic peoples as Turkoi. The nam e first appears in the
282
From the Urals Co the Carpathian Basin
works o f Agathius, who died in 582, telling the history o f Byzantine em perors up to 558. They crop up in Theophanes B yzantios’s history o f B yzantium up to 581, and are m entioned m any times in the w ork written by M enander between 584 and 602. A ccording to Tabari, al-A thir and Baladhuri, the Arabs attacked the Khazars in 652-653. H aving occupied Derbend, they advanced towards Balanjar. The Khazars defending Balanjar w ere rescued by the Turks. The events are also recorded by the A rm enian Sebeos. The sources m ake it clear that the nam e Turk here designates the western part o f the Turkic K haghanate. The Georgian Chronicle refers to the K hazars in 626-628 as the ‘West T urks’ w ho w ere then opposed to the East Turks o f Central Asia. Short after 679 the Arm enian G eography m entions the Turks together w ith the Khazars; this m ay be the first record o f the M agyars. A round 813, Theophanes uses— alongside the generic name Turk— ‘East T urk’ for the designation o f the Khazars, and in the con text, the ‘West T urks’ may actually have m eant the M agyars. We know that Nicholas M isticus refers to the M agyars as ‘West T urks’ in 924-925. In the 9th century the nam e Turk was m ainly used to designate the Khazars. It cannot be a coincidence that the M agyars, splintering o ff the K hazar Em pire, were designated by this name.
3. O NO G H UR, UNGAR, HUNGARUS, HUNG ARIAN In a letter to Dado, Bishop o f Verdun, w hose death in 923 proves that the letter refers to the period prior to that date, the unknow n correspondent poses a theological question: does the evil m anifest in history com e from God? In regard, the scholar cites the famous part o f the Old Testam ent book o f Ezekiel, which relates the story o f the people o f Gog and M agog (actually M agog in the country o f Gog). Aware o f the com ing turn o f the m illennium , he assesses the likelihood o f an apocalyptic end to the world. The author is optim istic with regard to the end o f the world, but could not avoid the question that preoccu pied the w hole o f literate C hristendom at that time: where did they com e from, this people who had swept in from the east to threaten the entire w estern world, the M agyars? The letter writer, at the turn o f the 9th and 10th centuries, relates to his friend the theories in circulation accounting for the origin o f the M agyars. The first theory, w hich the w riter discounts as claptrap, is w ide spread in the countries o f both the writer and the addressee. This has it that the accursed M agyars are none other than the people o f Gog and M agog. There is another opinion which identifies the people o f Gog and M agog with the wild and teem ing Scythians who lived beyond the Caucasus, in the swamps o f the M aeotis, beside the Caspian Lake and all the way to India. Our author
The names o f the M agyars before the foundation o f state
283
is o f the view that this is also mistaken, since the people o f Gog and M agog are not to be understood as an actual people, but as the heretics: it is they who endanger the true Church. But who, then, are the M agyars? Our w ell-inform ed correspondent states that their nam e is unknown in the sources. How can this be? Unlike Regino, who also tackled the question, but could find no answer, our author has a sim ple solution. Just as the River Tiber was once called the Albula, and Italy was once known as Saturnia, the M agyars are not m entioned by this nam e in the old books because they once had another name, and “b e cause their name has changed they cannot be identified in the texts” . He then, after expressing some reservations, recounts the third theory on the origin of the M agyars, which he heard from his elders, “ ...be it historical truth or fable” . The story runs thus. At one time famine struck the people o f Pannonia, Istria and Illyria. All o f the rulers o f the area gathered in council, and then selected those who were capable o f saving them selves from famine. The others were banished to the desert and unknown regions, where they w ould either die or be capable o f surviving. The expelled peoples reached the m arshes o f the Maeotis, where they survived by hunting birds and game. Their num bers burgeoned. Then comes the explanation: “and after the fam ine that they had withstood, they w ere called Hungri” (et a fa m e quam sustineant H ungri vocati sunt). The author thereby reveals his native tongue, because it is only in Old German, m ore particularly w hat is known as Old High Germ an, that the word meaning ‘hunger’ (H ungar, the m odem Germ an H unger) coincides with the name o f the M agyars. One group o f sources contem poraneous with the C onquest use the forms Ungar or H ungar for the nam e o f the M agyars, to which are added the Latinising ending -us, to produce Ungarus and Hungaerus, and the w ord also occurs in the plural form Ungri and Hungri. This is the nam e applied to the M agyars in the earliest Latin sources o f the Conquest era. From the first h alf o f the 1Oth century, the form Ungri (given in G reek in the form Ouggroi) also appeared for the M agyars in the Byzantine sources. First o f all, the m atter o f the forms with and w ithout the initial h- m ust be cleared up. In the French language, it was about that time that the initial hbegan to disappear, which French spelling still indicates for older Latin-root words, but is not pronounced. In foreign, non-Latin origin words, the w riter or scribe was uncertain w hether to write the h- or not. For this reason he sometimes wrote it even when it had no reason to be there. The nam e o f the M agyars under discussion was new to m onks who had been brought up in the French school, or w ere them selves French, and so som etimes they wrote the h-, and som etim es they did not. As the identification with the Huns took hold, however, the h- form becam e prevalent am ong the French. It was from them that it passed to the English, to gain ground as Hungarian. A m ong the
284
From the Urals to the Carpathian Basin
Germ ans, the two forms survived together for longer, but ultim ately the form without h- ( Ungar) becam e prevalent. The Ungri form is o f Slavic origin. In Old Church Slavic, if there was a nasal consonant (n, m, ng) after an о vowel, then the о itself becam e nasal (as at the end o f words in m odem French), and later the nasal on becam e un, and the nasal elem ent ultim ately disappeared from the un. Thus from the plural form Ongre there was a system atic shift via Ungri to Ugri. In old R ussian texts, until Polish influence in the Slav world led to the spread o f the Polish nam e for the M agyar people, Vengri, M agyars w ere referred to by the nam e Ugri. The form Ongre is the Slavic plural o f Onghur or O noghur or m ore likely Onughur. The Slavic form o f the nam e, w hich is found everyw here in the earliest Slav sources, thus derives from the form O noghur/O nughur or, with the elim ination o f the m iddle vowel, Onghur. The nam e O noghur is very com m on in the various sources. In order to determ ine who first used this name for the M agyars and when, we m ust first look at the origin o f the ethnic nam e O noghur and the history o f the Onoghurs. The first part o f the nam e, on, is the Turkic num ber ‘te n ’. The second part is the ethnic nam e Oghur. The ethnic nam e O ghuz also occurs frequently in C entral A sian Turkic inscriptions and later in other sources. The N ine Oghuz tribe, the Tokhuz Oghuz was the core o f the U ighur tribal alliance. Later m i grating west, the Oghuz are found in the sources under the nam e Uz. The rela tion o f the w ords Oghur and Oghuz is equivalent to that o f okiir and okiiz ‘ox ’. T he first belongs to a Chuvash-type language, and the second, containing a z, is a C om m on Turkic word. It turns out that the form ending in z is older, but a discussion o f the evidence for this w ould not be relevant to the issue o f the M agyars’ ethnic name. Incidentally, the persistent m isconception to the effect that the ethnic nam es Oghuz and O ghur are in som e w ay related to the Turkic w ord ok ‘arrow ’ is com pletely w ithout foundation. It is true that the Ten Arrow People (On Okh Boduri) does appear in Turkic inscriptions, bu t the form ok in the Old Turkic language could not have becom e O ghuz or Oghur. The m ain events o f Onoghur history have already been covered (pp. 209 228). It is interesting that the groups that occupied the land next to the Danube retained both the nam es O noghur and Bulghar. As well as the ethnic name O noghur, or in place o f it, we often com e across O noghundur and its later versions Vlendur and N andor. The nam e used by the M agyars for the Balkan O noghurs was N andor, and m ediaeval Hungarian sources refer to Belgrade as N andorfejervar, i.e. as the W hite Castle o f the O noghur-Bulghars. There is now a w ealth o f evidence to support the opinion that the O noghurs lived in the C arpathian Basin after 642. In 791, Charles the G reat attacked the Avar Empire. This attack is m entioned in m any annals. One o f these is the
The names o f the M agyars before the foundation o f state
285
Annales gem m eticenses, w hich states that Charles the Great destroyed the Hungarian em pire (Karolus Hungarorum regnum vastat). O ther sources con tain the forms Hungarorum and Vungrorum. Since the chronicles were copied several tim es, it cannot be ruled out that these nam es found their way into the text at a later date, ju st as it is nowadays custom ary to say that the M agyars traversed the “southw est Russian steppe”, m eaning the southern part o f m odern Russia, and not the southern part o f R ussian lands at that time. But it is also quite possible that the ethnic nam es H ungar or Vungar w ere included in the original text. This highlights the im portance o f a contem porary deed describing a donation o f lands made on 8 M ay 860 to the M attsee M onastery by the Eastern Frankish Emperor, Louis o f Germany. It m entions the border lands or the m arshlands o f the Wangars (uuangariorum marcha). Here, there can be no talk o f later substitution. The w estern m onasteries record the adm ission, visit, or death o f m onks who may have had some relation with these Pannonia or B alkan Onoghurs. Unfortunately the first careful overall study o f this extensive source has not been follow ed up by further research. I will only quote the earliest o f the data: Hungarius (797-809), H ungaer(us), Hunger(us), Hounger(us) (761), Ungarus (731-736), Ungerus (9th century), Onger (9th century), Wanger (812). Some o f these could hardly be related to the nam e o f the Onoghurs. As is also seen from the structure o f the nam e o f Hungaer, noted in St. G allen Abbey in 761, his father’s nam e was Teutgaer (Ego Hungaer, filiu s Teutgaeri). Teut is the nam e o f the Germ ans (the word D eutsch derives from the adjective form Teutisch), and so a Germ an personal name has been form ed from the very com mon ethnic nam e H un, also used for the Avars, along with the G erm anic ending -gaer. In other cases there is a suspicion o f later, i.e. 9 t h - 10th-century, transcription or falsification. However, m ost o f the data are from after 796, and a section o f the Avars had becom e Christianised by this tim e (see pp. 263-266). This group o f Avars m ight conceivably have been the origin o f the m onks w ho used some form o f the ethnic name Onoghur as their identifying name. O f course, the U ngar who is recorded in 926 by Ekkehard as being baptised and m arried, and whose wife bore him a child, m ay actually have been Magyar. The story o f the ethnic name O noghur in Europe can thus be traced from the m iddle o f the 5th century up to the tim e o f the Conquest. Until the 7th century, the ethnic nam e Onoghur denoted the people w ho lived on the Pontus steppe. A fter the fall o f K huvrat’s em pire, the part o f the Turkic people who m oved to the Danube area w ere called O noghur and Onoghundur, and the subsequent versions o f these, and Bulghar. It is also clear that the ethnic names Onoghur and Onughur were used for the M agyars by the Slavs. All that remains to determ ine is the time and the place.
286
From the llra ls to the Carpathian Basin
The Slavs em erged in the Balkans over the period from the m iddle o f the 6 th century up to the 630s. A round 670, the O noghur-Bulghars w ho arrived in the Danube area were certainly know n by these two nam es. The ethnic nam e O noghur persisted from that tim e on am ong the Slavs as the nam e o f the peoples who followed a nom adic w ay o f life and spoke a Turkic language or had a Turkic culture. It is highly probable, but difficult to prove, that the Slavic version o f the ethnic nam e O noghur was extended to the M agyars living in the Etelkoz. The ancient Russian chronicles note that the M agyars passed below Kiev, and one place near Kiev was still know n at the tim e o f writing the chronicles as the Hill o f the Ugrians. The R ussian chronicles m ention the W hite Ugrians, who, following the Bulghars, settled on the lands o f the Slovenes by the banks o f the Danube. These Ugrians em erged, so the prim ary chronicle claim s, at the time o f Em peror H eraclius (610-641). Since the Russian prim ary chronicle was w ritten in the 11 th century, no great im portance is attached to whom the Slavs actually called W hite Ugrians, i.e. M agyars, in the 7th century. In any case, the Slavs had relations w ith the W hite Ugrians from the beginning o f the 7th century. W hat is certain is that the neighbours o f the Slavs adopted the Slav form Ungri (< ongre) to designate the M agyars, and that this could not have occurred later than the 9th century, because subsequently the nasal elem ent disappeared from the word. W eighing up the arguments from history and the process o f phonetic change, the Slavic form m ust have passed to the Slavs’ neighbours in the late 8 th or the 9th century. The legends o f Cyril (or Constantinus), and o f M ethodius, also include Ugri as the nam e o f the M agyars. Constantinus, w ho adopted the nam e Cyril shortly before his death, died in 869. M ethodius’s death can be put at around 885. The texts recording their lives originate from the 9th century, but, as w e have seen, no available m anuscript o f the Constantinus legend is earlier than a 15th-century copy, and the earliest M ethodius legend m anuscript is from the 12th century. Constantinus and M ethodius m ay have m et w ith M agyars on the Crim ean Peninsula in 861, and M ethodius with the “ Ugrian K ing” on the bank o f the Danube in 881-882, som ewhere on the low er course o f the river, on his way to Constantinople. However, the latter m ay only be a late interpolation (see pp. 60 and 2 4 5-246) replacing the description o f a m eeting w ith Charles the Fat. In the original m anuscripts the form was quite certainly Ungri, w hich m eans that the Slavs used this nam e for the M agyars after the m iddle o f the 9th century. It was from them that the S lavs’ neighbours took the word and, as evidence from sources attests, only used it for the M agyars from the year 882 onwards. A fter the Conquest this becam e the m ost frequent designation o f the M agyars in Latin sources. In the H ebrew-language work Sepher Yosippon, w hich em erged in Italy around 940 (see p. 91), Togorm a’s ten sons and their peoples are fitted into
The names o f the M agyars before the found ation o f state
287
the Biblical genealogy. Ten tribal nam es can be identified in the H ebrew text, the K hazars, the Pechenegs, the Bulghars, the K nbyna (?), the Torks, the Guzz, the Tulmach, the Alani and the Zikh. The ninth tribe on the list was the Ugr. The ethnic nam e Tork is well known from Slavic sources, and its origin is identical to that o f Turk, and the Zikh was a Caucasian people, probably the Georgians. The Guzz bear a version o f the ethnic name Oghuz. Tulm ach is the name o f a Pecheneg tribe and identical in origin to the Hungarian w ord tolmacs ‘interpreter’. The form Ugr (awgr) is im portant because this is a definite reference to the M agyars. The text states that all these peoples live east o f the River Atil, except the Ugrians, the Bulghars and the Pechenegs, who live on the Danube. The data o f the Sepher Yosippon cannot be o f G erm an or Latin origin, and definitely derive from a Slavic source. Such sources m ust date from the tim e when the nasal sound had already disappeared from m ost Slavic languages. However, it does not contain the Slavic plural. Hasday ibn Shaprut, the Jewish m inister o f finance to the court o f the Arab U m ayyad in Spain, wrote a letter to the K hazar K haghan Joseph between 955 and 962 (see p. 91) in which the M agyars appear under the nam e Hungrin. Here the ending is clearly Slavic, the nasal sound is still retained, and there is an [h] at the beginning o f the word. It can only be im agined that this is the Slavic version o f the Latin form Hungar. Since the letter’s transm ission route is roughly known, it is very probable that the Jewish com m unity living in the Carpathian Basin in the 10th century also played some part in carrying it. A H ebrew-language A ustrian grave inscription, m arked w ith the date 1130, records that the deceased, whose name was Sabatai, was killed in Hungary. The nam e o f Hungary on the inscription is Ungriya. In m ediaeval Hungarian H ebrew sources, the M agyars are referred to by the nam e Agarenus, probably derived from the Biblical nam e Hagar. How ever, its existence is due to a popular-etym ological “correction”. O m itting the second letter o f ugr (written in Hebrew as awgr) enables the form agar to be read. The Italian Sepher Yosippon data and the Hungarian Hebrew data therefore derive from the same Slavic form. In Greek sources the ethnic nam e Ungroi (pronounced ungri) only appears in the second h alf o f the 10t.h century, and only refers to the M agyars. It is certain that the Byzantine authors obtained the designation from the Slavs. As seen above, the Byzantines norm ally called the M agyars Turks. In the final analysis, then, the M agyars had a nam e given by the Slavs and adopted at various times and in various forms by contem porary European peoples.
288
Front tlw Urals to the Carpathian Basin
4. SAVART1 ASFALI Constantine Porphyrogenitus writes about the Turks, i.e. the M agyars, in Chapter 38 o f his great work. A fter describing where they had form erly lived under the leadership o f Levedi, he continues: “Now, the Pechenegs who were previously called ‘K angar’ (for this ‘K angar’ was a nam e signifying nobility and valour am ong them), these, then, stirred up w ar against the K hazars and, being defeated, were forced to quit their own land and to settle in that o f the Turks. A nd when battle w as joined betw een the Turks and the Pechenegs who were at that time called ‘K angar’, the arm y o f the Turks was defeated and split into two parts. One part w ent eastw ards and settled in the region o f Persia, and they are to this day called by the ancient denom ination o f the Turks ‘Savarti A sfali’, but the other part, together with their voivode and chief Levedias, settled in the western region, in places called Atelkouzou, in w hich places the nation o f the Pechenegs now lives.” (Translated by Jenkins.) Both the Pecheneg-K angar and the M agyar-S avarti Asfali issues, like m ost unsolved problem s, have enormous literature. The details will be discussed below (see C hapter XIII). Close attention to the text produces a strong im pression o f a parallel between the “former nam e” o f the Pechenegs and the “old nam e” o f the M agyars. This seems to suggest that one o f Porphyrogeni tu s ’s sources m entions a w ar between the Kangars and the Savartis, and for some reason the Em peror has connected it to the wars between the Pechenegs and the M agyars. The com m on feature o f the two ethnic nam es is that both occur in some form in the Caucasus area. The ethnic nam e K angar occurs in Syrian, M iddle Persian and, m ost probably, in A rm enian sources. The name Savarti sounds very sim ilar to an ethnic nam e w hich is m entioned in another work by Porphyrogenitus as m eaning in A rm enian popular etym ology “black boys” . That the Kangars w ere noble and heroic, and the Savartis “steadfast” (the Greek word used is asphales, “unwavering, strong, firm, reliable”) are w riters’ cliches which refer to some longer story which the abridgers short ened. The fact that the second word o f the ethnic nam e Savarti Asfali is a com m on Greek word, w hich appears here as a perm anent adjective, suggests an artificial designation. For this reason there is little worth in striving, on the basis o f the n am e’s structure, to connect the ethnic nam e Savarti to other known nam es, such as the nam e o f the people that gave Siberia its nam e, Sabir, or to Suvar, the name o f one o f the Volga B ulghar tribes. There is a view that the ethnic nam e Savarti Asfali, w hich hitherto existed only in Porphyrogenitus’s work, and even there only in one place, is some invented term which has no historical credence. But a recent discovery in another source has endowed the name with new interest.
The names o f the M agyars before the found ation o f state
289
The text in question is the “Kievan Letter”, w ritten in Hebrew, w ith H ebrew script, w hich was discovered in 1962 and published in 1982. It was probably written in the 1Oth century, and the copy found in Cairo is a copy from very soon after. The letter tells o f a w ealthy K ievan personage, one M ar Jakob ben R. H annukah, who has fallen into serious trouble by virtue o f standing surety for a large loan taken out by his brother. The brother has been robbed and m urdered, and M ar Jakob imprisoned. The Kiev Jewish com m unity redeem ed him and gave him the letter in question so that he m ay approach Jewish com m unities in order to collect the m issing money. The letter testifies to this and is signed by ten persons. One o f the ten is Simson Jehudah, know n as the Savarti (Simson Jehuda ha m ekhunes Savarte). It is clear from the structure o f the nam e that Savarti refers to a distinguishing feature o f the person. O f course we do not know w hat this feature was. It m ay have been that Simson Jehudah was one o f the Savartis who converted to the Jew ish faith. Rare cases o f non-Jews converting to Judaism are know n in the K hazar Em pire, but such appellations can have m any different reasons. Sim son m ay have lived in a Kievan Savarti house, or there m ay have been som e story w hich linked him to the Savartis, and because o f which Savarti was appended to his name. W hatever the reason for this instance o f a given name, it is sure that the name Savarti has reappeared in a second, independent source. It m ust be noted that both the Greek and the Hebrew form could also be read Severti. Beyond the data o f the sources, we are reduced to guessing w hat could have led Porphyrogenitus to link the M agyars with the Savarti people. There are no other facts to support the assertion that it was the old nam e o f the “Turks”, i.e. the M agyars. The Magyars;’ alleged Savarti nam e could at m ost be said to contain the historical inform ation that one group o f M agyars m igrated east wards (see Chapters XIII and XV).
5. BASHKIR The M agyars are called Bashkir in several sources. Bashkir is nowadays the name o f a people who live south o f the Kam a and w est o f the Urals, w ith an adm inistrative centre in Ufa (Ofo in the B ashkir language). The Bashkirs speak a Kipchak-type Turkic language. Here we m ust first determ ine the original, or the oldest reconstructible form o f the ethnic nam e Bashkir. Tenth-century sources can be divided into three categories. The first con tains the works o f al-Balhi, who died in 934, and those by al-Istrahri and Ibn Haukal in which his work was used and abridged. In this category the ethnic name occurs in the form Bashjirt. The second category is that o f Ibn F adlan’s work. In the description o f his journey made in 921-922, the ethnic nam e takes
290
From the Urals to the Carpathian Basin
the form Bashghird. This form also appears in the work o f Abu Hamid al-Garnati, which is discussed below. Khashghari, writing around 1074, uses the form Bashghirt, and the form Bashgird appears in the great geographical dictionary o f Yakut, who died in 1229. The third form is Bajghird, and is found in the work o f M asudi, who died around 956. The ethnic name Bashkir is also known from later sources. The Secret H istory o fth e M ongols, w ritten around the m iddle o f the 13 th century, contains the ethnic name in the form Bajigit. The Persian historians o f the M ongol era (Juvaini, Rashid ud-Din) and the Turkish authors who follow ed them (such as A bul Gazi) used the form Bashgird. The European travellers o f the M ongol era, Plano Carpini, Rubruk and Benedictus Polonus used forms such as Baskar and Baskatur, which can also be read as B ashkar and B ashkatur (spelling variants: Baschart, Biscart, Bascart, Bashart, Bistart, Bastard, Bascard, Pascatur, Pascatir, P ascatu). The Russians adopted the nam e in the form Bashkir, the ending -t disap peared in Russian, and the word thereafter passed into European languages from the Russian. The self-designation o f B ashkirs now adays is Bashkort. The conclusion drawn from surveying the three 10th-century versions o f the nam e is that the earliest form that can be reconstructed for the B ashkirs’ ethnic nam e is *Bachgird. This is the proto-form which allow s all the others to be derived. The /j/ sound in the B ashjirt forms is an Arabic substitution for the Turkic /g/, but the change ch > sh in the ethnic nam e is Turkic. Here we have to face the difficulty that the /ch/ sound could only be distinguished by the Persian authors. Arabic authors wrote a foreign /ch/ sound as [j] or [sh]. If the original had been /j/, however, the /sh/ could not be accounted for either in Arabic, or in Turkic. After the /sh/ had evolved, the follow ing /g/ lost its voicing, and resem bled а /к/. It cannot be determ ined w hether the w ord ending was originally /d/ or Л/, but this is not particularly significant in our context. There are no good sources for the B ashkirs’ history. M odem B ashkirs have a rich historical tradition, fam ily trees and national legends, which are now easily accessible through the publications o f B ashkir scholars. These histories only extend, at m ost, to the M ongol era, however. From the 14th-15th cen turies they faithfully record their m igrations and relations, and the im m igra tion o f other tribes into their lands. There are some sources which m ention the Bashkirs, but surprisingly, they confuse the Bashkirs and the M agyars. In the group o f sources based on al-Balhi, this is expressed by the statem ent that there are tw o kinds o f Bashkirs. One live at the end o f the Guz, behind the Bulghars, and are said to num ber 2000 people. They live under the protection o f forests, and are therefore difficult to attack. They are subjects o f the Bulghars. The other Bashkirs, the
The names o f the M agyars before the fo u nd atio n o f stare
291
larger, m ore populous group, live in lands bordering on the Pechenegs. Both they and the Pechenegs are Turkic, and share borders with the Byzantines. M asudi produced his w ork from earlier geographical descriptions and con tem porary reports. His original text has not survived, but he h im self m ade an extract o f w hich a copy has survived (see p. 73). One o f M asudi’s works, the M urnj adz-D zahab states that the Khazars and Alani have borders to the west with four Turkic peoples which can trace their fam ily trees to a com m on ancestor. Each has a separate ruler, and the extent o f their land is equal to several days’ journey. Their country stretches to the M aeotis. Their raids have taken them as far as Byzantium and Andalusia, where they prevailed over the peoples there. They live at peace with both the K hazar king and the Alani prince. Their lands border on the country o f the Khazars. Their first tribe is the b.j.n.y, and the second, neighbouring tribe is the b.jg.rd. Beside the latter lives a tribe called b.j.nak, which is the m ost w arlike o f all o f these peoples. Finally, across the border is the fourth tribe, w hich is called nwk.rda. This list is repeated in another work, the Kitab at-Tanbih. Here the four tribes are part o f the nom adic Turkic, who are collectively called al-w and.riyyat after a city called Wand.r. They, together with the Burghars, attacked B yzantium in 932, and closed o ff the land route between Byzantium and Rome. We will have reason later to discuss the questions thrown up by this text, but it is certain that here Bashkir occurs as one o f the names for the M agyars. In relating his journeys, Ibn Fadlan sets down precisely where he crossed the land o f the Bashkirs, nam ing the rivers crossed by his caravan on the road to the capital o f the Volga Bulghars. Crossing the R iver Ural he noted another six rivers. Several o f these are easily identified, such as the Sam ara, the Kinel and the Sok. Ibn Fadlan describes the land and custom s o f the Bashkirs in detail, and then notes the rivers he crossed after he had left the B ashkirs’ lands. Some o f these can also be identified, and so the area inhabited by the Bashkirs in 922 can be firm ly delineated. It lay betw een the River Ural and the River Kundurcha. Ibn Fadlan states that the Bashkirs are a Turkic people which com m ands awe by virtue o f being the m ost fearsome and bravest o f the Turkic. The Arab Ibn Fadlan questioned them via a B ashkir interpreter who had converted to the Islam ic faith. However, he does not m ention w hether they spoke a language which was som ehow different from the other Turkic lan guages (see Figure 58 on p. 223). The key to understanding who called the M agyars Bashkirs, and why, is found in the work o f Abu Ham id al-G am ati. The author was bom in Granada in 1080. He first studied in Alexandria, and arrived in Baghdad in 1122. After various journeys, he settled in Saksin, on the Volga estuary, and then travelled among the Bulghars. He left the Volga B ulghars’ capital in 1150, and crossing Russian lands, through Kiev, visited the Pechenegs. From there he progressed
292
From the Urals to the Carpathian Basin
to Hungary. He spent three years in Hungary, and left an interesting description o f the contem porary state o f affairs there. Leaving his son in H ungary in 1153, he travelled to Saksin, and then m ade a pilgrim age to M ecca. Reaching Baghdad in 1155, he wrote his w ork on the astonishing things he had seen on his journeys. Encouraged by his success, he w rote a second w ork in 1162. He died in 1170 (see Figure 63). D escribing his arrival at the country he alternately calls Bashghird and Unguriya, he writes: “A fter I arrived in U nguriya, where
7ТГ
J
} 7?
Lake Ladogi ^К^аке Onega
Ье1<ч> Ozero X
■ -
|ч / , Bulghnr, v
I
oY-11133/1134) -у;: • (1150)
7 T iA •фч6 \ '\ .
Л’а!>
\ Kieve'i ймке) у?
K1NGDOMOF .. HUNGARY
V1
- 7 .vn<1156-1153)^
I
S a k s in i'
Bukhara
(1131| ‘i
(1153)ч
ВiUu'k Sen Constarjtjnopte'’
V.
D e r b e n f f ’' i "
■Ainu!
(1430-1150)*
M erv*
Г--'1..
У
Ntshapur
% v .
M osul
\,RaU
^M^Abhajy. ' - u *
Baalbek
Damascus
/ *
BasrlT^
I
f 18 -1 t-йt h
4 j
}
\ ^ 4* 1
if Isfahan
-1 127)
(1122)
r (1169/ (1169/1170)
A lexandra
(1117/1118)1
'”"4*%«hdad
&
/ Figure 63 Abu H am id al-G am ati’s journey
Г / / I \fs /
The names o f [lie M agyars before the foundation o f state
293
there lives the Bashgird people.” He calls the Hungarian king a B ashgird king, who is called Krali. The only reason why som ebody w ould have called the C arpathian-B asin M agyars B ashkirs in the 12th century is that he had com e from a place where M agyars were called Bashkirs. A l-G am ati had com e from the Volga Bulghars, and used their designation. Some argum ents can be adduced to reinforce this. A l-G am ati called Kiev M on Kermen. This has been m isread by publishers o f the text to date as G ur Kerm en or Gurkum an. O nly in the languages o f the Volga Bulghars and the m odern Chuvash does there exist a word m eaning ‘b ig ’ which has an old form o f mon. In the K ipchak languages K iev is called Man Kerm en, and this eventually reached the M ongols. In The Secret H istory o f the M ongols we find the reading M en Kerm en, but this is a m isreading o f Man Kerm en, which has the same spelling. The M ongols learned o f Kiev via the Kipchaks. M eaning ‘great city’, Mon or M an K erm en is the translation o f the Russian Velikiy Gorod, ‘the great city ’, the perm anent nam e o f Kiev. Al-Gam ati, however, did not learn the Russian, Kipchak, Pecheneg or Cum an name o f Kiev, but the Volga Etulghar. He proceeded in the sam e way with the M agyars. The Volga M agyars and the Bashkirs were also identified as one by a traveller o f the M ongol era, Plano Carpini. He writes in his work, pro duced betw een 1247 and 1252, that after the destruction o f the Volga Bul ghar Empire, the Tatars turned “further north” against Bashkiria, i.e. against Greater Hungary (...baschart id est Hungariam magnam...), and crushed it, too. Could the Volga B ulghars have had any idea in 1150 that the M agyars lived far aw ay? This relationship certainly existed, and as well as M ongol-era travellers, there are tw o good pieces o f evidence for it. In the last chapter o f his work, A nonym us m entions Billa and Baksh, and later Heteny, who arrive from the land o f Bular. This B ular is the regular variant o f Bulghar. There are several coins m inted in the seat o f the Volga Bulghar kings, Bulghari, on which can be read the nam e Bular. It now exists in the placenam e Bilyar. The ethnic name Bular also lies behind the nam e B elar in Hungarian chronicles. The other evidence is the journey o f the Hungarian Friar Julianus, who journeyed afar to visit the early relatives o f the Hungarians (see p. 429) Finally, we have to find the reason why the Volga B ulghars called the Danube M agyars Bashkirs. There can only be one explanation: they also called the Volga M agyars by this nam e and they knew that it was the sam e people. This m eans that al-B alhi’s remark, cited above, that there are two kinds o f Bashkirs, one o f which are subjects o f the Bulghars, and the other inhabit lands bordering on the Pechenegs and the Byzantines, refers in both cases to Magyars, nam ely the Volga and the Danube M agyars.
294
From the Urals to the Carpathian Basin
The designation o f the M agyars as Bashkirs has for a long tim e preoccupied scholars. We return to related issues in C hapter XV. Here it will only be stated that the ethnic name Bashkir was the designation o f the M agyars used by the Volga Bulghars, and that the land they inhabited in 922 is given exactly from Ibn F adlan’s account o f his journey. This is also the earliest authentic mention o f the ethnic name Bashkir. There rem ains the question o f when the Volga B ulghars started calling the M agyars Bashkir. As was discussed earlier (see pp. 220-227), this was the 8 th century at the latest. A m ore exact answ er can only be given w hen we discuss the ancient history o f the M agyars. In principle, it could be when the Volga Bulghars first cam e into contact with the M agyars, or becam e neighbours with them.
6. MAJGAR One group o f Arabian and Persian authors m ention the M agyars by the name Majgar. The vowels were not written originally in Arabic script, but later the long vowels were written with the full letters (A, I, W), and short or naturallength vowels with auxiliary symbols. The latter were frequently om itted from the m anuscript altogether, or som etim es erroneously inserted by copyists. In some m anuscripts there is a sym bol above the letter m which can be read as о, и, o or ii, and so the data have been transcribed as Mojgar, This would fit well with the M ogyer written by Anonym us. It is now know n that this extension has nothing to do with the name, but is the interpretative insertion by an Arab copyist. In Arabic, the letter g (gain) is only distinguished from the letter cain by a superposed dot, which indicates an unusual voiced fricative made in the throat using the pharynx. Latin-script transcription represents this sound using a raised с symbol: c. If we start from the prem ise that vowels were not m arked in the original transcription o f the word, then the data m ay be read m.j.g.r, where the dots can be substituted by any vowel, or none. In practice, however, a cluster made up o f mj or g r is inconceivable in Arabic, Turkic, or Hungarian, and so the question o f presence or absence o f a vow el can only arise in the case o f the second point. If the vowel was a in each case, then the A rabic form o f the word could only be read as m ajagar or majgar. O f course, there is no vowel that can be com pletely ruled out as belonging in the place o f the dot, but we w ould have to adduce some external reason for its being present. If we also regard the dot placed above the g as superfluous, then according to Arabic script the readings majacar or m ajcar are also possible. For the sake o f com pleteness, it should be noted that if a superposed dot is placed or om itted elsewhere, then two other sounds (two kinds o f Arabic /h/) could conceivably occupy the place o f /j/, but these can also be ignored.
The names o f the M agyars before the foundation o f state
295
The name M ajgar and reports o f the people it designates also occur in many other sources. Am ong the m ost im portant o f these are the works o f the Arab Ibn Rusta (w ritten around 930) and the Afghanistani Gardizi (written in Persian betw een 1050 and 1053), and the work called Hudiid al-alam written in 982-983 by an unknown author. We also find these reports in works by al-Bakri, w ho died in 1094, and later authors. Although the ethnic nam e is in many cases distorted, for instance in the form m.h.f.r in Bakri (the [g] and the [f] are in certain cases easily confused), two things are clear from the m anuscript record. One is that the entire m anuscript record is derived from a single early source, although later texts contain greater or lesser deviations. It is also apparent that the later sources have no value in their own right, and the various forms o f the ethnic name can only be accounted to the copyist, and do not indicate new knowledge. The ancestor o f all o f these sources is the lost or inaccessible work written by a m inister o f the Sam anide dynasty based in Bukhara, al-Jayhani. Although, as was m entioned earlier (pp. 69-71), not every detail surrounding the person o f the author has been settled, we will, for the sake o f simplicity, henceforward refer to the Jayhani heritage. According to this, one o f the M ajgars’ borders was with the land o f the Pechenegs and the Esegels, who w ere affiliated to the Bulghars. The M ajgars were a kind o f Turk, however, and their chieftain m arched with 20,000 horsemen. Later it is written: “One o f their borders touches the Sea o f Rum into which two rivers flow. One o f these is larger than the Jayhun (the Amu Darya). They inhabit the land between these two rivers.” And later: “Left (west) from them, on the river on the side o f the Slavs, lives a people which belongs to the Byzantines. All o f these are Christian and are called w.n.n.d.r. [...] One o f the two rivers is called the Atil, and the other the Danube [...] W hen the M ajgars are on the bank o f the river, they can see the w.n.n.d.r. people [...] The river which is to the right (east) o f the M ajgars flows towards the Slavs, and from there to the lands o f the Khazars. This is the larger o f the tw o...” And even later: “The M ajgars are fire-worshippers. They swoop on the Slavs (Sakaliba) and proceed along the bank until they reach one o f the Byzantine countries’ ports, which is called K.r.h.” (Read as Kerj, i.e. K erch on the Crimean Peninsula, but perhaps m ore likely Kers, i.e. Kherson.) W hatever position w e take in the debate over the Jayhani report, tw o things are clear and unam biguous. One is that Jayhani also alludes to two kinds o f Majgars. One lived between the Bulghar-Esegels and the Pechenegs, and the other led a nomadic existence betw een the Atil and the Danube. Several other river names could be assigned to the Atil, but a river which flows first through Slav lands and then into the country o f the Khazars, and then flows into the Black Sea, could only be the Dnieper or the Don. Their southern border was
296
From die Urals со die Carpadiian Basin
the B lack Sea, into w hich flowed the two rivers that delineated the lands where they lived. W hen they were on the bank o f the Danube, they could see the Danube Onoghundur-Bulghars, i.e. the W anandars, on the far bank. This report presents the situation prior to the Conquest, and the date can also be identified to some degree o f certainty. It is m ore or less contem porary with the visits by Constantinus (Cyril) and M ethodius. They also m et the M agyars in the Crimea, and so the report concerns the period betw een 860 and 880. Since the Danube Bulghar Christian church finally affiliated to Byzantium in 869, even this interval can be narrowed. The question naturally arises o f why Jayhani, with his wide scope o f attention, adopted data from an earlier period. Since we are dealing with a w ork w hich is only know n through extracts and quotations, we are not in a position to answ er this. However, it is alm ost certain that Jayhani’s basic source concerning peoples far aw ay from him was the w ork o f Ibn Hordadzbeh, who first set dow n his know ledge in 846-847 (see p. 6 8 ), and com pleted his thoroughly revised w ork in 885-886. Ibn H ordadzbeh died in 911. His work, The B ook o f Routes a n d K ingdom s, served as the source and m odel o f later Arab geographers. Jayhani m ay have used a version o f H ordadzbeh’s work which has not survived. However, the m inister in B ukhara expanded on this in his w ork w hich was clearly started after acceding to the post o f m ilitary chief in 913, and was continued by his son and grandson, who bore the sam e name, after 937 and 976 respectively. They retained the basic historical-geographical situation o f 860-880. A nother question has rem ained open: who called the M agyars M ajgar betw een 860 and 880? Only one o f the neighbours who could be candidates remained-—the Khazars. The Jayhani heritage inform s us that the M agyars, at this tim e, had a border with the Khazars, and the source also w rites that the K hazars surrounded them selves w ith a ditch to protect them selves against the M ajgars and other neighbouring peoples. Form erly m istranslated into Hungarian as “surrounded them selves with a w a ir , the ditch is a definite reference to w hat is well know n in Arabian literature as the W ar o f the Ditches. M uham m ad protected him self from the attacks o f hostile Arab equestrian forces in M edina in 627 by surrounding his positions with a ditch, at the advice o f a Persian w ho was experienced in agricultural irrigation. This w ar betw een farm ers and nom ads becam e som e thing o f a sym bol, and the source also refers to this. C onsequently, the construction o f the Fortress at Sharkel cannot be directly inferred from these source data. This passage does m ention that the M ajgar and the K hazars were neighbours, and at that tim e the M ajgars w ere no longer the subjects o f the Khazars, but an independent ethnic group w ho were capable o f threatening even the K hazar Empire. This is, incidentally, also apparent from the reported alliance o f the Khavars, recorded by Porphyrogenitus. One rebellious group
The names o f cite M agyars before the fo u nda tio n o f state
297
only jo ins w ith another people if the second is not subject to the pow er against which the rebellion has been mounted. Even if p ro o f cannot be produced, the circum stances m ake it highly prob able that the M agyars’ M ajgar nam e originates from the Khazars. Two opinions can be adduced in opposition to this. One is that in Turkic languages, there only used to be an /m / sound in the initial position if it was follow ed by a nasal consonant. The other counter-argum ent is that it is not possible to presum e a /j/ sound at the end o f syllables in known old Turkic languages. This can only m ean that this form o f the nam e cannot be o f Turkic origin. In seeking a solution, it m ust be borne firm ly in m ind that M ajgar (M ajagar, M ajacar, etc.) was not sim ply the K hazar nam e for the M agyars, but was the Arabic recording o f the ethnic nam e used by the Khazars to designate the M agyars. This only increases the num ber o f possible original forms to the extent that if the K hazar designation was, for example, M achgar, M achagar or som ething similar, then Arabic would still have rendered it as Majgar. In Persian there is a separate letter designating the consonant /ch/, but A rabic did not recognise such a sound and did not have such a letter, and as was seen in the case o f the ethnic nam e Bashkir, the foreign consonant/ch/ was transcribed with the letters j or sh. This w ould in principle solve the question o f the M ajgar’s /j/ consonant, because in the old Turkic languages there was a syllable-final /ch/, and it is also know n that /j/ was possible betw een vowels (one general’s nam e was Kundajik, for instance). From the K hazar point o f view, then, only the form s Machgar, M achagar or M ajagar can be hypothe sised. However, there is no “Turkic” solution to the initial m-. The old Turkic languages did use the initial m- other than before non-nasal consonants in foreign words, for exam ple in the Kol Tegin inscription near the R iver Orkhon there is a person called M akarach. This is a nam e derived from the Indian M aharaja, but the initial m could not have occurred in a word o f Turkic origin. There are consequently weighty arguments for the view that the nam e M ajgar for the M agyars is o f K hazar origin, and the arguments against only result in the conclusion that the nam e is not originally Turkic. At present no m ore can be said with certainty, and groundless speculation should be avoided. We m ust seek an answer for a logical question, how ever: is the K hazar designation o f the M agyars, M ajgar, independent o f the M agyars’ self-designation?
7. MAGYAR The ethnic nam e M agyar appears relatively late in Hungarian sources. Early data w hich has been connected w ith it, such as a person nam ed Muayeris (written in Greek letters as M ouageris) o f the 6 th century, or a fort called
298
From the Urals to the Carpathian Basin
Matzaron allegedly m entioned in 587, cannot be linked w ith the ethnic nam e Magyar. It has been proved that the ethnic nam e Mazarous, w hich ap pears once in Porphyrogenitus’s work, is a copyist’s m istaken rendering o f Kazarous. In the A rabic language, the verb ja r a has the m eaning ‘to flow ’. The ad jectival form o f this gives m ajaran ‘riv er’, plural m ajarin ‘rivers’. This also occurs in several Arabic geographical nam es, and enthusiastic lay people have often connected it with the name Magyar. It is im probable that the ethnic nam e M agyar is to be sought in the personal name M ogurdi which appears in a copy, m ade betw een 1130 and 1140, o f a 1086 Latin docum ent w ritten in Hungary. There are definite data from the 12th century, however. A Slavonic source w hich is derived from a 12th-cen tury G reek original, states that “the Ugors (Ugri) w hich are called Peons call them selves M ager" (plural Magere). Anonym us writes: “They are called Hungari in foreign languages and M ogers in their own language” (... p e r ydiom a alienigarum hungarii et in sua lingua propria m ogerii vocantur). After the 12th century there is a series o f data in sources from Hungary: M ogorsciget (1225), M agari (1399), M agery (1510), etc. Data on the ethnic nam e M agyar, and the forms w hich have passed into other languages and developed there, occur widely. This data can be divided into two large groups. One can be linked to the M agyars w ho rem ained in the east, and the other involves references to groups who w ere expelled or m igrated from Hungary, or to events within Hungary. The nam es o f M agyars who rem ained in the east have also survived in village nam es. These village names all lie w est o f the Volga. They reflect the era w hen refugees from the M ongol invasion, who reached the Volga M agyars in 1235, crossed the Volga and dispersed on the right bank. Such nam es are M ozhar, M ozharka, M ozharovo, M ochar, M uchar, etc. In these nam e-form s w ritten in Russian, the о or и letter o f the first syllable recorded an о or sim ilar vowel. This is derived from an earlier a vowel. The region contained speakers o f Chuvash, Tatar, Cherem is and other languages, if not sim ultaneously, but by the time the data w ere recorded the phonetic change a > о had been com pleted via a rounded-lips /а/ vowel sim ilar to that in H ungarian. The original /j/ consonant developed differently depending on the interm ediate languages and the dialect o f Russian. Ultimately, however, all o f the data derive from a form o f Majar. In R ussian sources relating to the area, the ethnic nam e M ajar also occurs in docum ents applying to the Volga M agyars. The oldest such datum appears in a contract o f 1483: after the B eserm yan and M ordvin people, the M ocharin people are mentioned.
Figure 64 The distribution of the splintered Volga Magyars after the Mongolian invasion
300
From the Urals to the Carpathian Basin
The Volga M agyars also appear in other w ritten sources. All these are o f the M ongol era, or draw on original sources from that time. One group o f these relate the story o f the M ongol invasions. In M ongol, Chinese and Persian sources the form m ost com m only found is M ajar. The great m ajority o f their data concern Carpathian Basin M agyars, but a small group o f sources, such as the Secret H istory o f the M ongols, on describing the 1221-1222 cam paign o f the M ongol warlord Siibotei, m entions the Volga M agyars, together with the Bashkirs: Kangli, Kibchaut, Bajigit, Orosut, M ajarat, Asut, Sasut, Serkesiit, Kesim ir, Bolar, R aral (= Kerel) (section 262). The list appears on another occasion in the Secret H istory o f the M ongols, in section 270, involv ing the form s M ajar and Kerel. The list o f peoples is not unproblem atic: the form M ajarat which occurs in section 262 could be the M ongol plural o f the forms Majaran or Majarar, or, deriving from Russian, M ajarin, and also because the name Kerel for the M agyars o f Hungary, which exists in a m isw ritten form in the first list, occurs even though the M ongols only visited the C arpathian Basin some tw enty years later, at the tim e o f the M ongol invasion. Although the work was com pleted around 1240, both lists should probably be regarded as being inserted around 1245. The Persian historian Juwayni, o f w hom it is known that he w as still w ork ing in 1260, uses the ethnic nam es Bashgird and K elar in describing the M ongol invasion o f Hungary, and does not m ention the nam e M ajar. The greatest historian o f the M ongol era, Rashid ud-Din, w ho deals with the history o f the period in m uch greater detail, uses the nam e M ajar as w ell as Bashghird and K elar in the work that he com pleted in 1311. Chinese sources, including the annals o f the M ongol dynasty, also mention the M agyars’ ethnic name. In Yuanshi, recording, am ong others, the history o f the M ongol conquests, the form M achar can be reconstructed for the M agyars, although this m ay be a transcription o f Majar. A source o f particular significance is a letter w ritten in Persian which was carried by Plano Carini in 1247. The letter contains a m essage from Giiyuk Khan to Pope Innocent IV. The original was w ritten in M ongolian in 1246, and there was a Persian and a Latin translation o f it. The latter w as produced by Plano Carpini with the help o f local interpreters. In an earlier letter, the Pope adm onished the M ongols for their attack on the M agyars. This adm on ishing text is quoted back in the reply o f the M ongol Khan. The original and the Persian versions contain M ajar, but in the Latin version Plano Carpini translated it using Ungari. M any persons are known from the M ongol era whose nam es are related to the ethnic nam e o f the M agyars. There are princes o f the M ongol dynasty and descendants o f M ongol-era w arlords who acquired their nam es on the occa sion o f victory over the M agyars or in m em ory o f it. These nam es later found
The names o f the M agyars before the foundation o f state
301
life am ong the M ongols and their subjects, and were provided with various personal name endings. Nam es like Majar, M ajartai, M ajarkai becam e quite common. These persons often provided the nam es o f the places w here they were based, as in the often-cited town o f M ajar on the Kum a, the earliest report o f which is by Abul Fida in 1321. The popular form o f carriage w hich was produced in the town also came to be known as a m ajar coach. Contem porary with the advance o f the Ottom an Turks from the 15 th century onwards, the M agyars’ ethnic name appears with increasing regularity in both the A rabic-script Turkic and the Greek-script Late Byzantine sources, so that it was very com m on by the 16th century. By then, however, it was being m entioned in relation to the history o f the Ottom an subjugation o f the M agyars o f the C arpathian Basin. The next data have a different origin. The name o f the Volga M agyars was recently uncovered in a cem etery in Chistopol, beside the River Kama, in an inscription on a gravestone dated to 1311. The significance o f grave inscrip tions in the Volga area was discovered at the time o f Peter the G reat’s travels in 1722. In 1863, records o f two Turkic languages were found to have survived among Arabic-language grave inscriptions. One, which has a single live ver sion, is Chuvash, and is thus very im portant regarding M agyar-T urkic links. This is custom arily attributed to the Volga Bulghars. The other is a M iddle Turkic literary language, adopted from Khwarezm , and used by the incom ing Kipchaks. Several o f the known inscriptions have been published recently. Analysis o f the Volga Bulghar inscriptions was instrum ental in developing the famous Bulghar-Turkic theory (p. 226). The inscriptions are generally som ewhat brief. A part from the nam e o f the deceased, the details and possibly circum stances o f his death, and a quotation from the Quran, there is always the date (in the Arabian system ), and very often some m ention o f the ancestry o f the deceased. The inscription for the interred person who is o f interest to us now gives his date o f death as 9 Decem ber 1311. His nam e was Ismail, “who educated our scholars, loved those who fear God, built m osques, and carried out a m ultitude o f pious acts” . His father was called M ajar Rejep. The nam e Rejep originates from the Arabian m onth o f Rajab, and reflects the custom o f nam ing the child after the month in which he was bom . This also m eans that Ism ail’s grandfather was already M uslim . As well as the com m on nam e Rejep, he was given the distinguishing nam e o f Majar. Such distinguishing nam es based on ethnic names are also found in other inscriptions (e.g. Bulghar, Tatar, Turkm en, and, as discussed below, Ar). If Ismail was, say, sixty years old in 1311, this puts his birth at around 1250. If his father was then about 20 years old, then he m ust have been bom about 1230. This is o f course speculation, but it is certain that this M ajar Rejep was bom roughly at the sam e time as the H ungarian Friar
302
From the Urals to the Carpathian Basin
Julianus’s visit to the Volga M agyars in 1235-1236. He could hardly have received his nam e in m em ory o f the M ongol invasion o f Hungary, because at that tim e the M uslim s beside the Volga were not yet integrated into the M ongol Empire. The 1311 Chistopol inscription m ust be treated as an inde pendent datum, and the name M ajar refers to the self-designation o f the Volga M agyars. We also find M agyar nam es am ong slaves sold in the Crim ea. These are listed on sales contracts originating from the second h a lf o f the 13th century. In m ost cases it is im possible to decide w hether they are Danube or Volga M agyars. There is also a later group o f place nam es derived from the M agyar ethnic name. These are the nam es o f villages and areas inhabited by M agyars exiled from their country during the Ottom an conquest. There are several o f these in Egypt and Turkey. Finally, the M agyar ethnic name also crops up in academ ic literature from the 19th century onward, where we find instances o f the nam e rendered into several languages, such as die Magyaren, the M agyars, les m agyars, etc. M agyar is therefore evidently the self-designation o f the M agyars, but other peoples also used it. The spread o f M ajar as an external designation is as sociated w ith the M ongol era, and so we are left with the questions o f how it spread, why ju st then, and who was responsible for it. This obvious question has not, as yet, been addressed by anyone. The question can be put more precisely as how the nam e M ajar reached the M ongols as an external designation, since from there its course into Persian, then Chinese, and subsequently other sources can be traced. The M ongols m ade contact w ith the Volga M agyars earlier than w ith those who lived in the Carpathian Basin. We have seen that the Volga M agyars were also called M ajar and so it can be inferred, if not proved, that the M ongols em ployed Volga M agyar interpreters prior to their conquests, in preparation for them. It is also very likely that, before the M ongols, the K ipchak-Turks living in the Volga country, and their relatives to the south, the Cum ans, also called the M agyars by the nam e Majar. Finally we m ust address two questions. W hen did the M agyars start using their own nam e, and w hat is its origin? In analysing the origin o f the ethnic nam e M agyar, we m ust first o f all consider the history o f the Hungarian language. The H ungarian consonant gy (which sounds approxim ately equivalent to the first consonant o f the British English duke) is a relatively recent innovation in the Hungarian language. It derives in nearly every case directly from an earlier consonant /j/. Exceptions are rare, such as gyem ant ‘diam ond’ which com es from the Germ an D iamant or the verb gydgyul ‘heal’, which is identical in origin to the H ungarian j o
The names o f the M agyars before the foundation o f state
303
‘good’. The consonant /j/ is itself relatively new. O f the m any origins which gave rise to it, the category which is o f interest here is the shift o f an earlier consonant cluster/nch/ to /j/, during which the Ы disappeared, ju st as the nasal sound has disappeared from before other consonants. An exam ple is in the word agyar ‘tu sk ’, which derived from a proto-form *onchara, or agyek ‘groin’, the root o f which is the proto-form *ancha. Before going any further, it m ust be explained why there is a vowel /е/ in the second syllable in the earliest data. As was m entioned before, the M agyer form existed in the 12th century and appeared also later sporadically. (In A nonym us’s spelling, the letter о denotes a Hungarian rounded-lipped, labial a, and the practice o f writing the consonant g y had not yet developed, which is why he writes the nam e as M oger.) In accordance with the rules o f H un garian vow el harmony, the front second vowel later becam e a back vowel, but the m ixed vowels o f the earlier data still have to be explained. The reason for this strange phenom enon is that the M agyars’ ethnic nam e originally consisted o f two words. The first part, magy-, may derive from the form manch. Evidence for this is provided from the finding o f this designation am ong the M agyars’ closest linguistic relatives. The m odem self-designation o f the Voguls is Manshi, which also derives from a form *manch, as does the nam e o f another close relative, one o f the main groups o f the Ostyaks, the M osh phratria (an exogamic group which interm arries only with other groups). This word orig inally m eant ‘m an’. It was borrowed from an Indo-European language, the proto-w ord o f which is *manu ‘m an’. This is the root o f the English man and the Germ an Mann. To this was appended an ending, for instance in Old Indie, as well as the form manu there is m anush(ya) and in G erm an as well as Mann there is M ensch ‘m an’. This ending is also found in the R ussian word muzh (< monzh < manush), and in the Old Persian word manush. The borrow ing o f a word m eaning ‘m an’ is a phenom enon which can also be seen in other languages, e.g. in Zyryan, where it is mort, and in Udmurt, where it is murt. These, and their relatives in the other Finno-U grian lan guages, are derived from the Indo-European verb mert- ‘to d ie’, and have the m eaning ‘m ortal’, ‘person’, such as the Old Indie marta, the Old Greek mortos, the Arm enian mard, etc. ‘person’ (cf. Latin m orior ‘to d ie’). And as was m entioned earlier (p. 180) this word forms part o f the ethnic name Udmurt. The situation is sim ilar with the Cheremis. The self-designation o f the Cherem is Mari, M ariy is ‘person, husband, C herem is’. The w ord is known in Indo-Iranian languages. In Old Indie it is maryah, in the Avestan language mairya ‘young man, young person’. It is also conceivable that the Indo-European w ord m anush gave rise to an Indo-European ethnic nam e, and that was adopted as the self-designation o f
304
From the Urals to the Carpathian Basin
the M agyars, ju st as F rank was by the French and Rus by the Russians. The process m ay have been encouraged by a word *mancha, *macha, m eaning ‘tale, fable’, and very sim ilar to the word o f sim ilar m eaning in proto-H un garian, w ould also have been associated with the nam e, reinforcing the im pression that the people are ‘speakers’ and the others dum b, i.e. speak in an unintelligible tongue. It was seen above that a sim ilar association w as m ade with the externally-originating nam e Turk. However, the actual root o f the Hungarian w ord m ese ‘tale’ cannot be traced directly to the first part o f the ethnic name. The first part o f the ethnic nam e M agyar seems to be w ell enough served by the prevailing explanation o f its origin. However, the explanations put forw ard for the origins o f the second elem ent, and for the connection o f the two elem ents, seem to run into difficulties. The elem ent er is also a word m eaning ‘m an ’. This is the Finno-U grian w ord w hich lies w ithin the H ungar ian w ords em ber ‘person’ andfe r j ‘husband’. F erj com prises the w ordsf t and erj, and originally m eant ‘m an’. The first part o f the word em ber, the em, m eans ‘w om an’, and the second part ber is identical to the w ord fe rj, so that in Hungarian the ‘person’ is the com bination o f w om an and man. The m ost w idely-held current opinion is that the ethnic nam e M agyar m ust originally have m eant ‘M anshi person’. This is hardly credible. The ethnic nam e always distinguishes the group from others. It is not really possible that a distinction can be m ade from other M anshis on the basis that ‘w e are M anshi people, you are M anshi non-people. ’ There is no such ethnic nam e known. Structures like E nglishm an, deutscher M ensch, russkiy muzh, or, from Turkic, Rum eri ‘G reek p erson’ (actually Rom an person), or even mansh kum ‘Vogul person’, do not becom e ethnic names. Even less acceptable is that the er elem ent is o f Turkic origin. There is such a Turkic word, and it does m ean ‘p erson’ (cf. the H ungarian w ord erdem ‘m erit’ w hich is o f Turkic origin), but it is virtually inconceivable that a Finno-U grian speaking people w ould take one part o f its self-designation from its m other tongue and the other from a foreign tongue, even if the original m eaning o f the nam e was no longer know n to them. One ethnic nam e that has not hitherto been brought into the argum ent is that o f the Er, who were known to live betw een the Volga and the Urals. The m odem form o f the ethnic nam e is Ar, w hich is the nam e under w hich the Udm urts are known to the Tatars, the C huvash and the Cherem is. The nam e also occurs in placenam es: the largest town nam ed after the people is Arsk, w hich lies 65 km northw est o f Kazan. In the C huvash language, A r derives from an older form Ar. This people is also m entioned in old sources. In assessing these, it has to be bom e in m ind that m ost o f the sources are unable to distinguish betw een the vowels /а/ and /а/, because they have the same w ritten form. According to the 10th-century w riter al-Istahri, the land o f the
The names o f the M agyars before the foundation o f state
305
Ar lay beside uninhabited wilderness. It was from here that the best furs and lead w ere brought. This people is also m entioned by A bu H am id al-G am ati, who states that the country o f the Aru or A ru is tributary to the Bulghars. Here they hunt beaver, erm ine and grey squirrel, and in sum m er there is daylight for twenty-tw o hours. It is from them that the best beaver furs w ere obtained. The name A r hoja is legible on an undated Volga B ulghar inscription, found in a village called Tatar K alm ayur (Cherdaklin county, Ulyanov region), which probably dates from the 1340s. The Russian chronicles first mention the A r land (Arskaya Zemlya) under the year 1379. One o f the A r people’s towns and A r princes appear in connection with events o f 1469 and 1489. The area currently inhabited by the Udm urts was form erly the land o f U grian-speaking groups. As late as the 17th century, there w ere Vogul com munities stretching as far as the Kama. A ccording to W itsen (1692), the Voguls lived around Perm. A m anuscript o f 1736 m entions Voguls on the bank o f the Chusovaya. This m akes it probable that the A r people spoke an Ugrian-type Finno-Ugrian language, and their self-designation originates from the word er m eaning ‘p erson’. Magyer m ay be conceived as being com posed o f two ethnic nam es. A s such it could be one o f two types. It either reflects the fusion o f two peoples, i.e. the people o f the Manshi and the Er, or M anshi m ay distinguish a group from other Er people. There are ex am ples o f both. Sources describing the Avars m ention the Varhon people, w hich is a com bination o f Avar and Hun. The Kiirtgyarmat found in Porphyrogenitus only indicates a tem porary m erger o f two tribes, because there are no traces o f related placenam es. The distinguishing-m ark structure is exem plified by one nam e for the Votyak people, Odo Mariy, “Votyak C herem is”, which is used by some Cheremis groups for the Votyaks. There is a long-standing view that the origin o f the nam e o f the leading tribe o f the M agyar Conquest, M egyer, is o f the sam e origin as the ethnic self-des ignation Magyar. Both can be traced back to the form Majer. Interpretations have hitherto spoken o f two kinds o f assim ilation only. There are two unavoid able questions, however: when and how did the two forms evolve? Particular causes for the shift Majer to Majar and then to Magyar only need to be sought in the word stress, which in the Hungarian language alw ays falls on the first syllable. I f a foreign w ord is incorporated into H ungarian, then the stress pattern o f the original foreign word m ay be influential. The Hungarian cseled ‘m aidservant’ and csalad ‘fam ily’ both derive from a Slavonic form chelyad, but the w ord csalad im plies a Slavonic language in w hich the stress was on the last syllable, whereas cseled m ay have been adopted from a Slavonic form where the stress was on the first syllable, but it m ay also have
306
From che llm ls to the Carpathian Basin
been a Hungarian development. In the case o f Majar, however, there is no foreign origin. The developm ent Majer > Majar > Magyar m ust have been purely Hungarian. This is especially im portant because the change took place also in the language o f the Volga M agyars. As we have seen, the placenam es M uchar, M ochar and M ozhar derive from a form Majar. In Turkic languages, the stress is on the last syllable o f the word. Thus in Turkic, where the vowel harm ony is even stricter than in Hungarian, the regular adaptation o f the form Majer would be mejer. This is natural if we bear in mind that the tribal names o f the M agyar people were Turkic, or appeared in Turkic form. The Turkic form o f the name occurs as early as Porphyrogenitus’s work, so that there is an earlier source for the Megyer form than the Magyar. The Greek form is Megeres, which should be read as meyeris. In the Greek orthography the clusters g + e and g + i, and the letter g (gam m a) in B yzantine records should always be read as /у/. This /у/ also often stands for /j/, however, as in the transcription o f another M agyar tribal nam e Gyarm at, w ritten as Germat (pronounced ja r mat) or the nam e o f Gyeiicsa to Geobitsa (pronounced je v i-
cha). Two ethnic nam es in the Volga area have also been connected to this Mejer form: Meshcher and Misher. The ethnic nam e Meshcher, which occurs in Russian sources and m ay have been the nam e o f an extinct Finno-U grian people, m ust be regarded as unrelated to the tribal nam e Mejer, and m ost probable Misher, nowadays the name o f a large Tatar group. The /j/ in the latter can only w ith the greatest difficulty be im agined to have changed to /sh/. It is true that in Khazakh, Nogay and Karakalpak, the original Turkic con sonant /ch/ becam e /sh/, in Chuvash a strongly palatised /sy/, in B ashkir /s/, and in some Tatar dialects /с/, and in others fricativisation had started, som e times nearly only the /sh/ sound could be heard. Assum ing the sequence of changes Mejer > Mecher > Mesher > Misher, however, throws up a handful o f chronological questions. The form meshcher also becom es mesh-sher or mesher in several Russian dialects, from which it is easier to deduce the form misher. The /i/ o f the first syllable is definitely a Tatar developm ent. We will return to this question in connection with the putative M agyar Urheimat in Bashkiria (see Chapter XV). Several im portant conclusions can be made from analysis o f the M agyar ethnic name. It was the self-designation o f the M agyars, m ade out o f two ethnic names, and suggests that the M agyar people em erged from the merger o f two Finno-Ugrian, or m ore precisely Ugrian, groups. Since the Volga M agyars also called them selves M agyars, the self-designation m ust have existed before the two groups splintered. The exact date o f the division must be determ ined later, but it could not have been later than the m iddle o f the 8 th
The mimes o f the M agyars before the foundation o f state
307
century. The fact that the M agyar’s self-designation already existed in the 8 th century does not, o f course, say anything about when the M agyars started to use the name. The stages in evolution o f the ethnic nam e can be reconstructed as: m ancha > m aj + er> M ajer (> M ejer > M egyer) > M agyer > Magyar. The question that m ust be addressed now is w hether there is a connection between the self-designation o f the M agyars and their designation by the Khazars. The problem surrounds the letter g (Arabic gain) in the w ord M ajgar. Some scholars have clearly seen the problem, but thought that the name represented here was, in fact, that o f the Bashkirs. The form B ajgird emerged in the A ncient Turkic language in which there was no initial m-, and it was replaced by a b. There is definitely an example o f this. The Hungarian word bors ‘pepper’ derives from a Turkic form burch, which was in turn taken from the Iranian murj. The Iranian word can be followed along the eastern spice route, but for now it only serves as an example o f the process in certain Early Turkic languages where, if an initial m- in a foreign word does not precede a nasal consonant, it is replaced by a b-. Unfortunately, this suggestion runs into two insurm ountable difficulties. One is that it cannot deal w ith the d at the end o f the name Bajgird, the other is the idea that the -gir elem ent is the product o f some ethnic-nam e form ative suffix cannot be upheld. Sim ilar syllables can be found in the very far east (Yukagir, M anagir, etc.), and even there it is not sure that the -gir at the end o f the ethnic nam e is the same as the ending in the Bachgird/Bajgird form. If we accept that the form M ajgar in every source derives ultim ately from Jayhani or his source, Ibn H ordadzbeh’s work, then it is conceivable that the answer m ust be sought in Arabic script. Until now we had no reason to do so, but now we know that the M agyars’ ethnic name had the form M ajer, and that it was made out o f two ethnic names. It is thus conceivable that in A rabic script the letter originally used was not the Arabic gain, but cain, w hich is written in the same way, distinguished only by a dot. There is a dot on the gain, but not on the cain. This possibility was pointed out earlier, where it was noted that a form Majacar could be hypothesised. The syllable car w ould appear natural to Arabs because the ethnic name o f the Arabs has the same initial letter. If the original used the form Maja er, then the front sound o f the er would be difficult for a foreign ear to hear, but the closure o f the pharynx needed between the two vowels would have a sim ilar phonetic effect as that transcribed by cain. If this hypothesis was to be proved, then it w ould be possible to hypothesise further a M ajaer form underlying the K hazar desig nation o f the M agyars, which implies that the form m ancha > m aja er becam e Majer in the M agyar language. The placing by a copyist o f a dot on the cain could have played a part in m aking the ethnic nam e more sim ilar to the name Bajgird. Another possibility is that the dot that deprived the letter o f its
308
From the llra ls со the Carpathian Basin
function w as placed above the letter under the influence of, or by analogy with, the ethnic nam e Bajgir. In the event o f this som ewhat extended hypothesis being borne out, then it w ould be possible to follow the M agyars’ self-designation m uch further back, at least to the start o f the contact betw een K hazars and M agyars. The suppo sition naturally dem ands further proof, and so cannot yet be em ployed in historical reconstruction. We have also seen that the M agyar nam e also becam e one o f the external designations o f the M agyars during the M ongol era, and that the M ongols were closely involved in this. W hether it reached them directly from interpreters or via Cum an m ediation, it was ultim ately obtained from the Volga M agyars. The M ongols and their auxiliary forces started to use it for the M agyars o f H ungary at the sam e time, however. B efore setting out the conclusions that can be draw n from the history o f the M agyars’ self-designation in respect to the M agyars o f the Conquest, let us consider some other ethnic nam es which have been applied to the M agyars.
8. OTHER NAMES OF THE MAGYARS W hereas the foregoing nam es and their history allow im portant conclusions to be drawn concerning the C onquest-period M agyars, the follow ing nam es have no direct historical value. A t m ost they reflect assessm ents o f the M agyars.
a) Scythian The Scythians em erged in the 8th-7th centuries B C , and after overcom ing the Kim m ers, becam e lords o f the steppe. In all probability they spoke an Iranian language. The Greek sources devoted m uch attention to them, describing their social organisation and their nom adic w ay o f life. In the 3rd century BC, the Sarm atians dealt them a heavy blow, and in the second h a lf o f the 3rd cen tury A D , the advance o f the Goths spelt the end o f their em pire (see p. 203). Subsequently, the nam e Scythian was applied to every nom adic people, so that in the B yzantine sources, the Huns, Onoghurs, Avars, K hazars, Bulghars, Pechenegs, Uzians, Cum ans, Seljuks, M ongols and O ttom an Turks w ere all referred to as Scythians. In som e B yzantine sources, the M agyars are also nam ed thus. Later this nam e crept into other sources, im puting a certain way o f life and an im pression o f antiquity. The Hungarian form szittya is o f Latin origin and has been used since the 18th century.
The names o f the M agyars before the foundation o f state
309
b) Hun The ethnic nam e Hun was used from the 4th century onwards in sources o f neighbouring literate peoples, am ong them Byzantine sources. The nam e Hun later becam e the nam e o f various m ajor nom adic peoples com ing from the east. It was applied to Ephthalites, Akatirs, Sabirs, Onoghurs, Bulghars, Avars, Turks, Uzians, Cum ans, Seljuks, Ottom an Turks and m any sm aller peoples. From the 10th century, the M agyars were also known as Huns. D esignation o f the M agyars as Huns also appeared in m any other sources, several written in Latin script. This is related to the use o f the ethnic nam e H un in Hungarian chronicles. There are also m any data to the effect that after the death o f Attila several steppe peoples or their ruling families claim ed descent from him. Tracing back one’s fam ily tree to Attila was a m eans o f legitim ating the right to power, and so the story o f the Huns becam e very popular am ong many peoples o f the steppe. This reinforced the practice o f calling a people Hun, regardless o f their actual historical relations (see Chapter XIV).
c) Avar Some western sources m ention the M agyars o f the Conquest under the name Avar, because they becam e the inhabitants o f the land o f the Avars, the area which at one time had been the Avar Empire. The value o f such an ethnic nam e is well illustrated by the following quotation from Leo the Wise: “ ...the Scythians, i.e. the Avars and the Turks, and the other Hun people who follow a sim ilar way o f life.”
d) Other names The M agyars are also m entioned by other nam es in various sources. M ost of these are “literary nam es” . The authors wished to show o ff their knowledge o f antiquity by referring to the M agyars by the name o f som e ancient people. One such is the Dutch Latin chronicle o f the Saxon Wars, in w hich the M agyars appear under the nam e Parthian. In m ost cases these nam es are accom panied by som e conceited explanation, as in a Late Byzantine source: “the Paions, i.e. the Ungors or H uns” . The Paions lived in the area o f A ncient M acedonia, and w ere the allies o f the Troyans in Homer, who described their custom s in detail. The nam e m ay have been linked to the M agyars partly because o f their way o f life, and partly because o f the sim ilarity o f the nam e to Pannonia. The names Sarmatian and Yazig were also applied to the M agyars. In the history
310
From the Urals to the Carpathian Basin
o f the steppe, the S annatian tribes fought against the Scythians in the 3rd-2nd centuries BC (see p. 196). The Yazigs lived in the plains o f the Carpathian B asin in the 1st century AD. The nam e o f nearly every people o f antiquity which had some connection with the Carpathian B asin has been applied at some tim e to the M agyars. Thus there are sources w hich refer to M agyars as G epidae, D acians (earlier than to the Rom anians, incidentally), and as Getas.
9. A HISTORICAL SUM M ARY Ethnic nam es designating the M agyars have been seen to fall into two large categories. The first contains those nam es which provide inform ation concern ing the history o f the M agyar people, the nam es in the second category are only o f literary value, but their use m ay be significant for the judgem ent o f the M agyars or for criticism o f sources. N am es that are historically im portant can be divided into tw o categories: those that outsiders called the M agyars and those that they called them selves. An exam ination o f the external designations o f the M agyars enables a recon struction o f who called them w hat and when, from w hich em erges a very interesting picture o f the M agyars’ neighbours prior to the C onquest. The various nam es o f the M agyars w ere given by the Khazars (M ajgar, or possibly M ajaer), the Volga Bulghars (Bashkir), the B yzantine Em pire (Turk) and the Slavs (Ungri). It would be useful to know w hat the Pechenegs called the M agyars, but no sources have been found on this. It is conceivable that the Cum ans called the M agyars M ajar, and the M ongols obtained this nam e by their m ediation. The Pechenegs could hardly have called the M agyars by a name different to that later used by the Cumans. The latter tw o cases are only hypotheses, however. We can say, therefore, that the M agyars’ nam es, i.e. M ajgar, Bashkir, Ungri and Turk, were truly the nam es used for them prior to the C onquest, w hich the conquering M agyars “brought with them ” . B ut it also m eans that the M agyars m ust have been a people o f some significance before the C onquest for their neighbours to have distinguished them by such diverse names. As regards the date, these four designations m ade by different neigh bours are not necessarily sim ultaneous, but they cannot be separated by great intervals, and they certainly sounded like these and w ere used in the century prior to the Conquest. Unfortunately we cannot gain any insight earlier than the 8 th or 9th century via these nam es. In contrast w ith the practice w hich has been com m on hitherto, that the M agyars’ external designations have been regarded as consecutive, it now seems that the external designations are the
The names o f the M agyars before the foundation o f state
311
Figure 65 The M agyars’ self-designations and external designations prior to the Conquest nam es o f the conquering Magy ar people from the 9th century, although it may be that one or two em erged earlier than that. Study o f the M agyar people’s self-designation takes us som ew hat further. It has been shown that self-designation is one o f the im portant elem ents o f the M agyar people’s identification o f them selves as such. The evolution o f the name M agyar can be followed as far back as the form ation o f the people: the alliance o f the M anshi and the E r peoples. This was the nam e o f the M agyar people when they set o ff w estw ard and before the Volga M agyars split off. This is all that can be known o f the history o f the M agyars prior to the Conquest from the history o f the nam e and the analysis o f its sources. If we involve other sources in the reconstruction o f the history o f the M agyars o f the Conquest, then we can proceed further.
NOTES A new stage in the study o f Hungarian names, and particularly ethnic names was opened by the work o f Jinos M elich. A summary o f his view s is M elich (1 9 2 5 -1 9 2 9 ). The early Magyar personal names o f Turkic origin have been discussed by G om bocz (1915). This formed the
312
From the Urals to the Carpathian Basin
foundation for further work by Julius (Gyula) Nem eth. He set out his discourse on Magyar proto-history in numerous articles (collected in Nemeth 1990) and in his central work (Nem eth 1930), basing it almost entirely on the analysis o f ethnic and tribal names. A short overview in French is Nem eth (1949). In his last theoretical review he discussed the methods and the earlier studies o f the ethnic and tribal names, and tried to give a typology o f their semantics and morphological categories (see Nem eth 1991, pp. 5 0 -1 0 8 ). S e e a lso N e m e th (1 9 3 8 ,1 9 4 0 b , 1964, 1965, 1966c, 1969, 1971a, 1971b). A more critical approach is Nem eth (1975). Nem eth col lected a very large mass o f material about modern Turkish tribal names. He analysed the entire accessible Russian and Soviet literature, with its rich data. For this purpose he assem bled, and had it assembled, the bibliography which was finally published under the editorship o f Gyorgy Hazai (I9 6 0 ). He regarded as particularly important the work o f the tw o Bashkir researchers, Garipov and Kuzeev, who published several hundred Bashkir family, lineage, clan and tribal names. The research o fN em eth ’s pupil, Laszlo Rasonyi Nagy, was primarily directed at Turkic personal names. These studies resulted in a rich index o f Turkic personal names, and although it is unpublished, a guide has been produced for it (Baski 1986). Nem eth rightly criticised Rasonyi for his view that an ethnic name might have evolved from a personal name even before the M ongol era. R asonyi’s view is best presented in Rasonyi (1961 and 1976a). The Early-Turkic-origin personal names are the subject o f tw o articles by Ligeti (1978, 1979), but he covered tribal and ethnic names only in connection with two tribal names, Jeno and Gyarmat (Ligeti 1964b). Arpad Berta has suggested new tribal name etym ologies in several articles. He sees the Magyar tribal names as being o f Turkic origin, m ost o f them having started out as military terms (Berta 1989, 1990a, 1990b, 1991, 1992a, 1992b). For the N o g a y people and their ethnic name, see Vasary (1 993b). Nem eth was one o f my professors and w e had long discussions on questions o f m ethodology and historical relevance. I have learned much from him, but I was always more sceptical than him. The designations o f Magyars in foreign sources have long been discussed by Hungarian scholars. The basic work on Latin sources, which is still useful, is the work by Homan first published in 1917 and reissued in 1938 (Homan 1917/1938). A review o f the Greek data was made by M oravcsik (1958/1983). For archaised designations o f the Magyars, see M oravcsik (1 9 2 6 -1 9 3 2 /1 9 7 6 , pp. 3 8 3 -3 8 5 , 1929-1930/1967, pp. 32 0 -3 2 5 ). The substance o f this chapter was first published in the printed text o f my lecture to the Academ y o f Rhine-Westphalia (Rona-Tas 1988a = Rona-Tas 1995a, pp. 2 7 5 -3 1 0 ), where I dealt with earlier view points and gave a bibliography. I first wrote about the Magyar ethnic name in connection with the Chistopol inscription (Rona-Tas 1986 = Rona-Tas 1995a, pp. 135-138); for the names o f the Magyar people, see my inaugural address to the Hungarian Academ y o f Sciences (Rona-Tas 1993 = Rona-Tas 1995a, pp. 2 2 7 -2 5 3 ). For the sources quoted in the text see the chapter on the Sources. The two main viewpoints regarding the origin o f the ethnic name Turk are represented by Pelliot (1915) and Nem eth (1927) on the one side, and Clauson (1962, p. 87) on the other; see also Doerfer’s review (1 9 6 3 -1 9 7 5 , vol. II, pp. 483^195). L igeti’s view is presented in Ligeti (1985, p. 322). For my own research, see Rona-Tas (1991a, pp. 9 -1 3 ). The Bugut inscription was published by KlaStornyj-LivSic (1972). The research into Khotanese Saka language and literature has been reviewed by Emmerick (1979 and 1984), and the words are presented in dictionary form in Bailey (1979). This must be used in conjunction with the corrections presented by Em m erick-Skjaervo (1982, 1985, a third volum e is in print). The data and prob lem s o f Chinese transcriptions o f the ethnic name Turk are discussed by Pelliot (1915), and the Sogdian aspects by Harmatta (1962). The Tibetan data concerning the ethnic name Turk were first gathered by F. W. Thomas (1951, pp. 2 6 7 -3 0 6 ). N ew data have been discovered since
The names o f the M agyars before the foundation o f state
313
then. The ethnic name Turk on the Hungarian royal crown is discussed using data from the latest literature in Rona-Tas (1988a, pp. 119-120 = Rona-Tas 1995a, pp. 2 8 7 -2 8 8 ). For the Deed o f Foundation o f Tihany Abbey and Lake Turku, see Barczi (1951, pp. 1 6 -17). I quoted the letter to Dado from H eilig (1933). A good and still useful review o f the history o f the O noghur people and its name is given in M oravcsik (1930b, in German: 1930a). For more recent material, see Golden (1992, pp. 9 2 -1 0 8 ). The Greek data can be found in Moravcsik (1983). The latest review was by Szadeczky-Kardoss (1970). The Slavonic aspects o f the ethnic name, with particular regard to the biographies o f Cyril and M ethodius, w as discussed by Kiraly (1974). For the O nughur form o f the ethnic name see p. 209. The data on the U uangariorum m archa has been discussed by Olajos (1969). For the H ungarus names in pre-Conquest Latin monastery records see Kiraly (1987). The Sepher Yosippon was published by Flusser (1980, 1981); for its dating, cf. G olb’s comments in Golb-Pritsak (1982), for the ten tribes see vol. I, p. 4. The best edition o f the correspondence between Hasday ibn Shaprut and the Khazar King Joseph is Kokovtsov (1932), on which the “Schechter Text” sheds new light (Golb-Pritsak 1982). The research history was reviewed by Fejgina (1972). For the Austrian Jewish inscrip tion, see Scheiber (1983). Kohn (1881) also deals with the names o f the Magyars in Hebrew sources. M oravcsik (1984) helps greatly in collecting the Byzantine data on the Magyars. Studies on the Savard question have been reviewed by Czegledy (1959), and later by Kristo (1980, pp. 100-115; see also Kristo 1996, pp. 141-142). For new data see Rona-Tas (1984). The designation o f the Magyars as B ashkir in M ongol-era sources has been discussed in detail by Ligeti (1 964a). I reviewed earlier data in Rona-Tas (1 988a, p. 122 = Rona-Tas 1995a, p. 290). For Ibn Fadlan, see Zim onyi (in Kristo 1994c, p. 277) and above, p. 165. For Abu Hamid al-Gamati and the related literature, see Zimonyi (in Kristo 1994c, p. 28). K iev’s name for the Volga Bulghars is discussed in R 6 na-Tas (1992b, in English: 1994b). For the Magyars o f Bashkiria see Chapter X V in the present book. The alleged early occurrences o f the ethnic name M a g ya r and the M ajgariya data in Arabic sources have been discussed in detail by Czegledy (1956), who also quotes K m osko’s remarks (p. 274), that the -u (i.e. the dam m a) is a secondary Arabic element, and the original correct reading is with an -a-, so that it is not M ujgar or M ojgar, but M ajgar. Another suggestion, but one which carries serious problems, concerning the ethnic name M ajgar was made by Ligeti (1986, p. 400). He considers the element -gar, or more precisely according to him -gir, a suffix. The Volga-region placenames which refer to the Magyars, the earlier research and data, and the ethnic names M eshcher and M isher are also discussed by Vasary (1976, 1977). For the Chistopol inscription see Hakimzjanov (1986a, in English: 1986b; and Rona-Tas 1986 = Rona-Tas 1995a, pp. 134-138). Hakimzjanov read in the inscription: M ajar kadi. We had the opportunity to visit the inscription with Hakimzyanov in 1986 and recognised on the spot that the correct reading is M ajar R ejep. For the Eastern European and Asian placenames which have been connected with the ethnic name M agyar, see Nem eth (1991, pp. 3 0 5 -3 1 2 ). For the ethnic name M anshi, and the Finno-Ugrian aspects o f the ethnic name Magyar, see Lako-R edei (1 9 6 7 -1 978/1971, pp. 415^117; Rddei 1979, p. 353; Rddei 1988-1994, vol. Ill, pp. 866 - 868 ). The forms M eshcher, M isher and Mochar, M ozhar have been reviewed using earlier lit erature by V&sary (1976).
VI1.
URHE1M ATS
A N D MIGRATION
1. GENERAL QUESTIONS The literal translation o f the Germ an word Urheimat is ‘original hom eland’, which is based on the obsolete assum ption that peoples have one ancient hom eland from which they emerged. The word can no longer be applied in this sense, bearing in m ind that the developm ent o f peoples is a continuous process whose direction can, at most, be changed by som e significant event only. Accordingly, the use o f the term in the plural, as in the title o f this chapter, is more correct. ‘Urheimats ’, then, should denote those m ajor stages in the formation o f a people which brought about significant change to the life o f the m em bers o f the group (which is prim arily a consideration o f m odern-tim e historians). Form ing part o f the history o f a people, such changes m ay include a splinter group peeling o ff from the m ain community, the beginning o f interaction with another people, the change o f com m unity life style, or a m ajor migration. O f these historical processes, m igration deserves special attention, due to the fact that ethnic m ovem ent has m yriads o f forms. M igration can be a daily event, or related to the seasons. It can be triggered by econom ic considerations, or some violent event; it can be organised or spontaneous, etc. The whole com m unity can participate, or often the m en only, or those directly involved in animal husbandry, or the warriors or the m erchants only, leaving behind the others, who m ay follow them at a different pace. M ovem ent is not always predeterm ined, and very often m igrants are oblivious to the fact that they are on the move, sim ply because m ovem ent is so natural to them. M igration was far m ore integral to ancient societies than it is today. N a turally there were groups that possessed perm anent, stationary dwellings. Thus, for instance, farm ers irrigating their lands by m eans o f canals were evidently m ore bound than those w ho obtained arable land by clearing forests. Certain groups o f townspeople m ost probably held onto their privileged habitation for m any a generation, while others, by exploiting— or forced by— the changes o f trade or political circum stances, w andered from town to town.
316
From the llrals to the Carpathian Basin
The m igration o f ethnic groups, especially in the Eurasian region, was a very m uch m ore natural and perm anent process than one w ould be inclined to think today. A great m any accidental phenom ena, natural catastrophes or even a good crop, a favourable clim atic turn or a rapid increase o f livestock could equally well launch m igration, as could fraternal feuds, the arrival o f groups fleeing from revenge, or the clash o f am bitions for power. This was prepon derantly the case even before the shift to equestrian nom adism . M ovem ent was the m ost intrinsic elem ent o f the life o f equestrian nom adic groups, clans and tribes. It is practically im possible to dem arcate individual com m unities in those areas o f the steppe about w hose history w e have rela tively good sources. Clans or tribes that originated from one ancestor could splinter in a couple o f weeks, unite w ith others, create or break up alliances. The rapid m ovem ent o f peoples organising em pires creates the illusion that the w hole world, and all o f their subjects, w ere in perpetual m ovem ent and m igration. In reality, it is as if we were w itnessing an optical illusion created by the ripples on the surface o f water. The leading stratum gallops across the steppe, subjecting on its way all other peoples it encounters, sw eeping along with it som e o f the strata o f its new subject groups, but the subjected peoples them selves rem ain m ore or less in the place they were. As discussed earlier (pp. 213-214), having been defeated by the Turks in 552, the Avars conjoined the Ephthalite and Chyon groups and arrived in the C aucasus in 555, and the L ow er Danube region by 562. It w ould be a m istake to assum e that this am azing speed— notw ithstanding the “ invention” o f the stirrup w hich aided riding— involved the displacem ent o f a larger group o f people. It certainly happened that a large group set o ff on a long journey, but the larger the group, the slow er its m ovem ents were. Naturally, the m ovem ents o f the w andering hunter on foot, the m ounted huntsm an, the sem i-settled farm er pursuing an arable and anim al-rearing economy, and the cattle-rearing nom ad w ho also grows crops, differ greatly. M ovem ent o f any description can be recurrently cyclic or one-way, in that the annual m ovem ent scene changes in the course o f migration. In this m igration-packed world the w e-consciousness o f the individual groups w as very im portant, for it m eant solidarity and help. Yet these w e-consciousnesses were by no m eans absolute. The secure and established com m u nity o f a clan, or a group having a com m on cult, could fragm ent as a result o f individual, personal reasons; m arriages, friendships, alliances, interests or com m on enemies could contribute to the form ation o f groups having new w e-consciousnesses. In this highly pliable and coloured w orld vibrating from diverse m igrations, perm anence m eant belonging to a com m on people or even to a short-lived, large political establishm ent.
Urheimats and m igration
3 17
We m ust not lose sight o f the aforem entioned in describing and discussing the m igrations o f the conquering M agyars. W hat we are looking at is not the thousand-year ancient history o f a closed and hom ogenous group. We can but locate those areas w hich were positively related to their ancient history. It is in view o f this that we consider the places o f the ancient “Urheimats" .
2. THE M IGRATIONS OF THE PROTO-MAGYARS Taking into consideration all o f the sources at our disposal w e m ust m ake an attem pt to locate the stages o f the em erging M agyar people. N aturally each period has its different types o f source material. Establishing the very beginning is a hard task. W hich was day one in the history o f the M agyars? No m atter w hat we pinpoint as being the outset, it must be established who was where before that. We m ust go back, then, to prehistory, to the Urheimat o f the Uralic peoples, but o f course, not as far as the form ation o f this group o f peoples. The only graspable instance in this ancient history is the period im m ediately preceding the dissolution o f this group. This, considering all available evidence, happened in the central and southern regions o f the Ural M ountains, m ore or less abeam o f the R iver Kam a and its tributaries, but also, naturally, the broad area on either side o f the mountain range, i.e. to the east o f the Urals up to the central section o f the Ob and its headwaters. The gentle and passable slopes o f the Ural M ountains tend to connect peoples, not separate them. Small river and brook valleys, large clearings and the fauna and flora here provided Neolithic m an w ith an ideal domicile. The m eeting point o f the taiga belt and the deciduous forests was a perfectly diverse natural environm ent (see pp. 34-35 and 93-94). It w ould be a m istake, however, to believe that every people that ever emerged on Earth actually survived. One can safely assert that the Early Neolithic age in particular saw the form ation and decay o f a great m any peoples. A m ong them , m any peoples w hose languages w e are unable to reconstruct. Siberia and the A rctic circle are inhabited by a m ultitude o f peoples collectively term ed Palaeosiberian or Palaeoasian peoples. These are in no way related to each other. This group includes the K et people that lives in the Yenisei region and w hich is near extinction (they are also know n as the Yenisei O s ty a h , although their language is not o f Finno-U grian origin); the Yukagir o f the A rctic regions, the K ham chadal and several groups o f the Innuits. In a couple o f decades’ tim e their languages and m em ories will only have been preserved in 18th-20th-century scientific records. B ut then so m any people disappeared w ithout any record taken about them whatsoever! Yet these peoples m ight have left traces in languages that still exist. This is
318
From the Urals to the Carpathian Basin
im portant to acknowledge, because despite every attem pt to reconstruct from present-day languages the language and w orld o f the Finno-U grian peoples, there will always rem ain insoluble parts, on account o f the fact that the Finno-U grian languages preserve the im prints o f extinct languages as well. The peoples o f the Uralic Urheimat included the ancestors o f the future Finno-U grians and Samoyeds, as well as other peoples. The neighbours, too, m ight well have included peoples that disappeared w ithout a trace, but for tunately, we are able to reconstruct the language o f those neighbours whose m odern-tim e descendants we know. A djacent to the greater U rheim at, to the southw est lived the Indo-European peoples, to the southeast the Turkic peo ples, and to the northeast the M anchu-Tunguz peoples. Early interaction with them was sparse, but the paucity o f evidence greatly delim its our know ledge about the period. However, we have no reason to believe that the Uralic peoples splintered earlier than the 4 th -3 rd m illennia BC. The Sam oyedic peoples subsequently m igrated from the Ural region to their Siberian Urheimat where the very slowly changing conditions kept them together for m any thousands o f years. The dissolution o f this group took place roughly at the time o f Christ, and was located to the north o f the Sayan M ountains. The Finno-U grian peoples dispersed prim arily, but not exclusively, in the vast region on the w est side o f the Urals. The spread o f bronze brought about social differentiation, as well as dissim ilarities betw een the individual groups. Even in the Early Bronze Age, sim ple forms o f arable or anim al-rearing econ omies cropped up. B y the 2nd m illennium the area inhabited by the FinnoUgrians was so large that the later language groups em erged, including the forebears o f the Ugrian group. This was the period o f the Late Bronze Age, in which the role o f animal rearing increased alongside an arable economy, established on cleared forests. By the turn o f the 2nd and 1st m illennia the arable and anim al-rearing econom y was predom inant. We have great difficulty in locating the Ugrian Urheimat. As has been discussed earlier (see p. 98), lin guistic evidence indicates that— as opposed to the earlier period w hich fea tured rich Finno-U grian and Early U grian-A ncient-lranian ties— the Old Iranian period (approx. 8th-2nd centuries BC) bears no traces o f sim ilar interaction. An equestrian-huntsm an culture m ust have played an im portant role in the latter h alf o f the Ugrian period alongside the prim eval forms of agriculture and anim al rearing, and when it com es to assigning the locality, these facts unquestionably attest to a deciduous forest region. Also, undoubt edly, the ancient proto-M agyar period o f the freshly separated M agyar group does not feature traces o f close Iranian connections, w hereas interaction with the Perm ic languages is m anifest (see pp. 98-99).
llrheitnacs and m igration
319
W eighing up the vast stock o f evidence, we now think that the U grian period was relatively short. Subsequently the Ob-U grian and M agyar-speaking groups evolved, despite the fact that their splintering was a slow process, in the course o f w hich these Ugrian peoples lived together in one related area for a long time.
3. THE M IG RATIO N OF THE MAGYARS FROM THE URALS TO THE CARPATHIANS The northern and eastern areas o f the Ugrian Urheimat were inhabited by the Ob-Ugrians, the western and southern parts by the M agyars and the Er people. The Ugrian Urheimat was located in the Ural region, prim arily on the western side. However, Ugrian splinter groups are known to have resided to the east o f the Urals, too, by the time which the M agyars m ust have dw elt in the V olga-Kam a region where, in the neighbourhood o f the Perm ic peoples, two groups o f the M agyars, the M anycha and the Er, m aintained very close contacts. The E r group o f peoples is supposed to have spoken a Ugrian language, or to be more precise, a Ugrian dialect which was m ore closely related to the M agyar language than to the Ob-Ugrian. Naturally, it is perfectly feasible that this group o f peoples featured quite different anthropological characteristics altogether; however, there is not a shred o f evidence in support o f that. The locality in which the M agyars, i.e. the M anycha-Er group, em erged was betw een the Volga and the Ural M ountains. At that time, roughly between the 8 th and 5th centuries BC, the M agyars em barked upon their independent existence, and hence the early period o f the proto-M agyar language began. This area and its im m ediate vicinity w ere naturally inhabited by other peoples, too. Perm ic peoples dwelt to the w est and the northwest, and at the far side o f the Perm ians Iranian peoples to the south and southw est which, however, gradually assim ilated into the local com m unities o f Finno-Ugrians. It has been contended that the U grian Urheimat was located in W estern Siberia. This hypothesis started from the prem ise o f a putative early U g rian Turkic connection w hich could only have been established in the vicinity o f the Turkic Urheimat in Siberia. As has been discussed above (see p. 97), the etym ologies o f loan words such as Hungarian hattyu ‘sw an’, hod ‘beav er’, homok ‘sand’, nyereg ‘saddle’, etc. (which are claim ed to have entered the common U grian language at that time) cannot be upheld for various reasons, or else they did not enter the language in the Ugrian period. The O b-U grian’s move across the Urals m ust have occurred fairly early on, but their eastw ard migration was only substantially precipitated by the eastward advance o f the
320
From the Urals to the Carpathian Basin
Slavs in the 8th -9 th centuries. This process w as finally com pleted in the 18th century only. The Ob-U grian peoples slowly traversed the Ural M ountains to the region o f the Ob headwaters. Later, other peoples forced them to m ove north where the Ob-U grians subdued the local inhabitants. The M agyars who lived to the south o f the Urals in the region o f the R iver Ural (or possibly, slightly to the east) gradually shifted to a nom adic life style. The Iranians still held sw ay over the steppe at that time, and it is probably no m atter o f chance that as o f the 2nd century BC , a profusion o f M iddle Iranian w ords m ade their w ay into the Hungarian language (and likewise, into the Turkic). It w ould be a m istake to believe that the M agyars’ shift to a nom adic life style happened in a couple o f decades or m erely a few generations. Q uite to the contrary, the m ajority o f the M agyars continued to pursue a hunter and arable and anim al-rearing econom y in the shelter o f the forests. Som e groups, however, decided that setting out on large-scale equestrian raids w ould pay o ff better. They participated in num erous nom adic “projects”, and as auxiliary troops they acquired a fair am ount o f the loot. Frequently, how ever (especially w hen defeated), they w ould return to their forest-dw elling com m unities. This process w as a long one, and m ight even have lasted over a century. The forest regions o f South Siberia provided the setting for the Turkic, M ongolian and M anchu nom ads, and the Ural M ountains for the M agyars. A clim atic change m ay also have played a part in the change o f life style: the grovy steppe zone gradually shifted north around the early 1st m illennium BC. However, this m erely contributed to the process, and w as hardly decisive. In general, clim atic changes are slow er than w ould propel a group to “follow” the shift o f the m ost advantageous clim atic zone (unless, o f course, the econom y o f the group also begins to change). The centuries prior to and following the birth o f Christ saw m ajor changes on the steppe. W estern groups o f the X iongnu and the Iranian nom ads— hence the Sakas and the Alani— clashed for rule over the steppe. In the west, groups o f G erm anic and Gothic tribes advanced down to the Black Sea. T he next wave o f m igration, however, had not reached the region o f the U ralic passes. The arrival o f the Huns brought a breath o f fresh air to the steppe. The processes resulting in their m ove towards Europe ended in the 3rd-4th centuries. The Hunnish attacks precipitated the developm ent o f steppe trade, and the archaeological finds o f the tem perate forest zone began to feature elem ents o f steppe origin. The M agyars’ change o f life style could hardly have been triggered by the influx o f the Huns, who rapidly advanced across the passes o f the Ural M ountains.
Urheimats and m ig ra tion
321
After the Huns had passed, the first Turkic peoples appeared on the K hazakh steppe, and later also farther west. They preponderantly spoke an Oghur-type Turkic language. This O ghur group had appeared earlier in the C entral Asian steppe, and played a m ajor role in the shift o f life style o f the M ongols. Thorough analysis o f the Turkic loan words, w hich both the M ongolian and Hungarian languages feature, reveals some surprising facts. Thoroughness in this instance concerns the details. Consequently, certain later or secondary correspondences, or m orphologically obscure accords m ust be disregarded. But taking into consideration only the m ost elem entary and the m ost apparent correspondences reveals that both the Hungarian and the M ongolian lan guages feature essentially the same anim al-rearing w ord-stock w hich is o f Chuvash origin: for instance, the Hungarian borju ‘c a lf , bika ‘bull’, okor ‘o x ’, uno ‘co w ’, iirti ‘w ether’, teve ‘cam el’, beklyo ‘fetter’, etc. There are also many com m on agricultural terms, such as arpa ‘barley’, kender ‘hem p ’, bor ‘w ine’, bitza ‘w heat’, did ‘w alnut’, dara ‘groats’, tarld ‘stubble-field’, gyekeny ‘bul rush’, orol ‘to grind’, tild ‘hem p-cutter’, etc. Words like iirom ‘w orm w ood’, tiizok ‘bustard’, goreny ‘polecat’, karvaly ‘sparrow -haw k’, torontal ‘a preda tor b ird ’, etc., w hich exist in M ongolian also, attest to an environm ent charac teristic o f regions south o f the forest zone, o f the steppe and the sub-steppe belt, also indicating w here this Turkic-M ongol interaction took place. The M ongolian equivalents o f the follow ing w ords do not have a parallel H ungar ian form, but belong to the same w ord-stock o f Turkic origin: ‘m olar’, ‘to ch u m ’, ‘donkey’, ‘koum iss’, ia m b ’, ‘blaze’ (white m ark), ‘belly fa t’, ‘prog eny’, ‘w ind-shelter’, as well as a few nam es o f typical anim al colours. This highly intensive T urkic-M ongol connection m ust have ended in the centuries before Christ— this the conclusions regarding the Turkic word for ‘stirrup’ (see pp. 103-104) also confirm. We know by now that not every O ghur Turkic group set off west. Traces exist o f small South Siberian groups holding onto their independence right up until the 7 th -8 th centuries before assim ilating into the C om m on Turkic and M ongolian population, or conjoining the East Asian Khitais. W hat is certain is that they arrived on the Khazakh steppe in the place o f the Huns, or possibly it was they, before going w est in the m id-5th century, who actually triggered the m igration o f the Huns. O ghur-M agyar connections cannot have been very intensive at that time, in the early 5th century, w hen fleeing from the Sabirs in a new m igratory wave the Oghurs arrived at the pre-Caucasus region, and hence within the scope o f Byzantine sources. The question arises w hether it was not then that the M agyars conjoined the O ghur groups. This is unlikely, because it would be difficult to account for the M agyars’ preserving their Finno-U grian tongue in the com pletely Turkifying steppe. Evidently, a m ajor portion o f the M agyar
322
From the Urals to the Carpathian Basin
population did not com e under direct Turkic influence. A lthough m ore and m ore com m unities adopted the nom adic life style, m any groups continued to pursue a m ore peaceful and sim pler arable and anim al-rearing econom y in the protection o f the forests and river valleys. As did, for instance, the M ordvins who dw elt to the w est o f the Magyars. Settling in the pre-Caucasus region and populating the Eastern European steppe, the O ghur tribes consolidated their position under the rule o f the Sabirs and later the West Turks. However, follow ing the storm y passage o f the Avars, the northern region o f the steppe was left deserted. A lthough the W est Turks tem porarily extended their pow er over the peoples o f the Eastern European steppe, the peoples under the suzerainty o f the K hazars— then still under West Turk authority— needed protection against anticipated attacks from the east. The weakening and disappearance o f the West Turk leading stratum at the end o f the 6 th century presented the opportunity for change. W hat exactly these changes w ere can m erely be conjectured. One O ghur group, under Bulghar leadership, m oved from the region o f R iver K uban to the area betw een the Dniester and the Donets w hich was w here K huvrat’s B ulgharia developed. M oving southw est from the Ural region, the M agyars then occupied the B ulghars’ old place around the end o f the 6 th century. Interestingly, this phase o f the M agyars’ m igration happened concurrently w ith the spread o f the stirrup. W hile the B ulghars’ country becam e an Avar protectorate, the M agyars settled into the organisation o f the K hazar Empire. Serving as border guards in the north, they established connections w ith the Alani in the southwest, while also m aintaining ties w ith the Bulghars. The M agyars proceeded with their shift to a m ore intensive arable and anim al-rear ing econom y in the pre-Caucasus region. In 635, K huvrat gained independence from Avar authority, and for a short period the “who rules the Eastern European steppe?” becam e a hotly debated issue. The M agyars sided with the Khazars in the B ulghar-K hazar fights w hich ended w ith K hazar victory around 670 (see p. 219). The w estern groups o f the B ulghars (four o f “the five brothers”) w andered to the B alkans and the C arpathian Basin, and then other splinter groups m igrated even farther. Their place to the w est o f the River D nieper was occupied by the M agyars. A larger Bulghar group rem ained under the suzerainty o f the K hazar Em pire and gradually m oved northeast. B ulghar-M agyar connections m ust have flour ished under K hazar rule, because when in 737 an A rab attack forced the Bulghars under K hazar authority to w ithdraw to north o f the R iver Volga, they w ere jo in ed by a group o f M agyars. T he region nam ed by the M agyars Etelk 6 z lay to the w est o f the River Dnieper. It was to here that those K hazar groups (know n as K havars) fled which, having becom e opposed to central K hazar rule in the internal fights
Figure 66 Migrations of the ancient Magyars
llrheitnacs and m igration
323
324
From the Urals to the Carpathian Basin
triggered by forcible conversions to Jewish faith, rebelled, but w ere sup pressed and forced to flee. In 838, the K hazars constnicted, or rather, rein forced Sharkel. As we have seen above (pp. 234-239), in 893 the Pechenegs left their form er dwellings in the area o f the Ural pass, and crossed the Volga in 894. In turn, the Volga Bulghars m igrated even farther north, and occupied their final hom eland. Concurrent with this, pursued by Pecheneg attacks, the M agyars, too, left the Etelkoz and conquered the Carpathian B asin in 895.
NOTES The literature on the possible dw ellings o f the Uralic and Finno-Ugrian peoples is vast. It would be im possible to discussal! o f these works even briefly. Peter Hajdu and Istvan Fodor are highly recommended. For their works, consult the bibliography section o f this book (see pp. 4 4 7 -4 8 9 ). The Iranian effect on the Permic language is well known. However, the most useable monographs are not directly concerned with this issue. See Lytkin-Guljaev (1970), Joki (1973); there are surprisingly numerous words corresponding in the Permic languages and in the O ssetic, see Abaev (1 9 5 8-1989). Unfortunately, the Iranian elem ents o f the Hungarian lan guage are not discussed by any modern monograph. Munkacsi (1901) is, for understandable reasons, outdated, and even his acceptable or remarkable etym ologies need to be reconsidered. Skoid (1925) is also largely outdated. BSrczi’s (1958) com ment “the w hole issue o f Iranian loan words in the Hungarian language cannot be regarded as clarified” (p. 51) holds true even if minor steps have been made to sort out the matter. Harmatta’s works can be recommended, as for instance Harmatta (1971, 1977), as well as Ligeti (1976, pp. 19-27; 1986, pp. 162-174). The Magyar and M ongol agreements o f Chuvash-type words are discussed in Rona-Tas (1981). The debated issues o f the M agyars’ migrations, their dw ellings in “Levedia” and in the Etelkoz are summarised in Krist6 (1980 and 1996). For G yorffy’s view s see G yorffy (1975c, pp. 5—46). G yorffy wrote about the Conquest, its preliminaries and the settling o f the Magyars in 1984 in the ten-volum e history o f Hungary (see Gyorffy 1984). The questions o f Levedia and the Etelkoz were discussed at a debate on 28th April 1983. The talks, com ments and L igeti’s latterly submitted remarks were published in volum es 80 and 81 o f M agyar N yelv, and in the form o f an independent booklet as volum e 172 o f the series A M a g ya r N yelvtudom anyi T arsasag K iadvanyai [Publications o f the Hungarian Linguistic Society] (1985). Here they are cited as Gyorffy (1985), Benko (1985), Kirdly (1985), Harmatta (1985c), and Ligeti (1985). Vekony (1986) was not included in the volume. These issues are further discussed in Chapters VIII.I (pp. 3 2 5 -3 3 0 ) and XIII (pp. 413—422).
Vlll. THE CONQUEST
1. THE MAGYARS IN THE ETELKOZ The question o f the Etelkoz was recently discussed by renow ned Hungarian scholars. The exposition below relies largely on the m anifold views expressed at this discussion. However, I have questioned some claims taken for granted in recent debates— fully aware o f the fact that m y doubts, or certain elements of them, have already been raised by others, and that I m yself am unable to approach full certainty. Still, I am o f the opinion that placing m ore strictures on the sources at our disposal can help us m ake greater progress. O ne part o f the related issues has been discussed in the chapter on the history o f the Pechenegs (see pp. 234-239), as well as in the chapter on the earlier m igrations of the M agyars (see pp. 315 -324). I give full explanation o f my views in Chapter XIII apropos o f the Levedi question. The main question is: when did the M agyars m ove to the Etelkoz? Two other questions are less debated: where was Levedia and where was the eastern borderline o f Etelkoz? The M agyars at the end o f the 6 th century had long m aintained connec tions w ith various advancing Turkic groups, but they had not yet left their hom eland in the South Urals and the region o f the R iver Ural. A dopt ing the nom adic w ay o f life alongside an anim al-rearing econom y (with the em phasis gradually shifting to the former), the M agyars here estab lished their first contacts with the Khazars who were then organising the West Turkic peoples. The m ain current o f rapid successive changes barely affected the M agyars at that time. A fter the Huns in the m iddle o f the 4th century, the Oghurs around the year 463, and the hegem ony o f the Sabirs, the East Turks extended their pow er over the land. Released from the grip o f the weakening East Turk rule, the m ost powerful group am ong the West Turkic groups were the Khazars. O riginally in alliance with the B ulghars, they established the West Turk Empire. The historical legend recorded by M ichael o f Syria contains some fine lines about this pow er (see pp. 200-201). The alliance o f the Khazars with the Byzantine Em pire greatly contributed to the consolida tion o f their power. Their large-scale, joint m ilitary cam paign against Persia in 627 assum ed an im portant place in the historical records o f all peoples, near and far.
326
From the Urals tv the Carpathian Basin
The insurrection o f the Bulghars, however, brought about decisive changes. In the early 7th century the Bulghars gained independence and set up an independent Bulghar Em pire under K huvrat— not, as it was assum ed earlier (see p. 219), in the Kuban region, but on the two banks o f the R iver Dnieper. The B ulghars were backed by the Avars in their w ar o f independence against the Khazars. Independence, however, turned them against their Avar allies. In the resulting situation, the Khazars were able to crush the Bulghars, forcing large groups o f them to flee to the west, to the area o f today’s Bulgaria, to the Carpathian Basin, to the w est and east coasts o f the A driatic Sea, and hence to the vicinity o f Ravenna. The place o f the B ulghars was subsequently occupied by the M agyars. A round 670, following the death o f Khuvrat, the Bulghar Em pire collapsed— largely due to the attacks o f the jo in t K hazar-M agyar armies. It w ould be very convenient to assum e that Constantine Porphyrogenitus’s conjuring up o f the period o f the three years when the “Turks” (i.e. the M agyars) lived together with the Khazars “and fought in alliance with the Khazars in all their w ars” (translated by Jenkins) was an allusion to this war. Likewise, it m ight be possible that the K hazar khaghan, to seal their alliance, “gave in m arriage to the first voivode o f the Turks, called Levedias, a noble K hazar lady” (trans lated by Jenkins). I f this was the case, Porphyrogenitus’s reference to Levedi w ould be a later insertion. Although this possibility cannot be dropped, we know o f no source in support o f it, so in any case we are left in the dark. Neither does any source positively confirm that there actually were two Pecheneg attacks, as suggested below (pp. 420-421). A round 670, then, the M agyar tribal confederation— or m ore precisely the M agyar tribal confederation o f that time— occupied the region left by the B ulghars betw een the D nieper and the Danube, and m ade arrangem ents to settle down. As exposed below (pp. 418-419), the area form erly referred to as “Levedia” did not exist as a separate Urheimat, and L evedi’s dwellings were, in fact, in the Etelkoz. It is highly probable that the individual M agyar tribes pursued a nom adic w ay o f life betw een the D nieper and the D anube, as did the Pechenegs after them in the very sam e region. The dual kingship m ust have em erged around that time, under K hazar influence. As previously discussed (see pp. 148-150), the em ergence o f the dual kingship involved a long period in history, probably several generations. The ru ler’s legitim acy was granted by the sky, as w ell as by the K hazar ruler, o f course. D etails o f this twofold situation strongly resem ble the situation o f the Volga Bulghars before their conversion to the Islam ic faith. The focal points o f the econom y in the region m ust have been on the eastern side o f Etelkoz for a long time, on account o f the fact that the im portant trade routes converged there and the international centres o f trade, the D nieper and
The Conquest
327
the Crim ea, were located on the eastern side. Late 7th- and early 8 th-century sources report that the Crim ean Peninsula still had B yzantine colonies— in the region o f K herson at least, and around the D nieper delta, w hich was also where m ost o f the northern trade routes converged. This was the period when the K hazar Khaghanate interfered in the dom estic struggles o f the B yzantine Em pire, and also when the new, great foe, the Arab Caliphate emerged. The protagonists o f Eastern European history, from the early 8 th century until the end o f the 9th, w ere the Arabs, the Khazars, the B yzantines, the everadvancing Turkic nom ads, the Slavs advancing east in the forest regions, and the Norm ans storm ing the river valleys from the north. Their story is a history o f wars, alliances, o f trade and plunder. Gardizi writes that in the 1Oth century, at the instigation o f the Pechenegs, the neighbouring peoples m ade forays into each o th er’s countries to take slaves which they later sold to the Pechenegs (a situation which m ust have been sim ilar back in the 8 th—9th centuries). The Jayhani tradition reports, in connection with the M agyars, that w hen ever they m ade attacks on the Slavs, they proceeded along the banks o f the river until they arrived at a harbour belonging to the country o f the Byzantines. According to new and reasonable arguments, this was not Kerch but Kherson. In the 8 th century the balance o f external forces favoured the strengthen ing o f the M agyars in Etelkoz. Two m ajor em pires neighboured their country to the west: the Avar and the Bulghar. A t that time, the Avars w ere preoccupied with dom estic problem s and with Byzantium , and in the second h alf o f the 8 th century their prim e concern was the Franks who had switched to a rather forceful policy in Easten Europe. In the m iddle o f the 8 th century, following a b rief interlude, the new Abbasid Caliphate, which superseded the Um ayyads, channelled its forces against the Khazars. The K hazars m anoeuvred with keen diplom atic sense and a form idable army betw een two great foes, Byzantium and the Arabs, but it w as in their interest to secure them selves from the west, too. The M agyars did the job, in return for w hich they received a share o f the profits o f trade, and also their prestige rose within the K hazar Empire. At this time the M agyars m aintained close ties with the B ulghars and the Alani, too. Both peoples lived under the suzerainty o f the Khazars. As was pointed out above (pp. 220-227), the Bulghars, following the K hazars’ defeat by the Arabs in 737, took advantage o f the transitory weakening o f K hazar central power, and withdrew to the north along the banks o f the Don and the Volga. It seems that around the sam e time there was dom estic discord among the M agyars, as one— evidently significant— M agyar group joined the B ul ghars, and accom panied them on their new Volga Bulghar conquest, in the same fashion as the Khavars escorted the M agyars to the Carpathian Basin.
328
From the Urals to the Carpathian Basin
The M agyars’ connections with the Alani are only indirectly im plied from the sources. The pre-C onquest A lanian loan words entered the Hungarian language on the shores o f the Sea o f Azov. The m ost significant ones include asszony i a d y ’, ‘married w om an’ which actually m eant ‘distinguished, royal lady’ in Old Hungarian. The word was later also used to denote the Virgin M ary and Saint M argaret, daughter o f King Bela IV, and the still existing Hungarian place names A ssz o n y ^ p e , Asszonyfolde (literally, ‘L ady’s K in ’, ‘L ady’s L and’), etc. bear reference to their form erly belonging to the queen’s estates. The m odem Ossetic w ord hsin or ehsine, o f com m on origin with asszony, has the m eaning o f ‘lady’. A 12th-century B yzantine source trans lates the w ord as ‘princess’ (arhontissa). It is highly probable that the Hun garian asszony preserves m em ories o f the royal A lanian-M agyar m atrimony. The story related in the Hungarian chronicles about the M agyars w ho, while hunting in the swam ps o f the M aeotis, i.e. the Sea o f Azov, abducted the Alanian Prince D u la’s daughters (who later w ere m arried to H unor and M agor), was not, perhaps, a mere flight o f fancy. The language spoken in the region between the Donets and Don regions by the so-called A lanian groups o f the Saltovo-M ayak archaeological culture cannot be regarded as a source o f A lanian loan words in the Hungarian language (see p. 202). As a m atter o f fact, no direct source about this period exists. The picture that indirect sources reveal is that, although at the end o f the 8 th century the M agyars nom inally belonged under K hazar suzerainty, they w ere actually free to m anage their dom estic affairs them selves. The K hazar Em pire had not collapsed after the wars with the Arabs, quite the contrary, it had grow n and transform ed. In the first instance it em ployed a form idable arm y o f m ercenar ies, and successfully concluded its struggle w ith the A rabs in 798. The resulting p a x khazarica, i.e. the Eastern European peace established by the Khazars, created a century o f peace and prosperity in the region. The antago nism w ithin the K hazar Em pire at the end o f the 8 th century was largely created by the forcible conversion o f the leading K hazar strata to Judaism . The losers o f the K hazar internal fights, the K havar group, joined the M agyar people. This could not have happened if the M agyars had not already been independent by then. Evidently, the Khavars com prised the peoples o f the w estern areas o f the K hazar Em pire, i.e. the groups that dw elt betw een the Don and the Dnieper. Perhaps it was due to the earlier dam age suffered that in 838 the K hazars erected the Fortress at Sharkel in the low er Don area. The Jayhani tradition holds that the M agyars dw elt betw een tw o rivers, one being the Danube, and the other the Etil, in this case m eaning the Dnieper. A people affiliated w ith Byzantium lived by the Danube on the side o f the Slavs. All o f them w ere C hristians and were called Vanandars. The M agyars saw these Vanandars from the other side o f the Danube (Gardizi goes further, and
The Conquest
329
claims that they actually captured theseN andurs). The D anube B ulghars were baptised betw een 865 and 8 6 8 , and it was not before 870 that they finally succumbed to the authority o f the C onstantinople patriarch. Consequently, this report can only have been noted down after 870. Evidently, the M agyar groups o f the Etelkoz m ust have struggled for pow er among each other all along. It is probable that the dwellings o f the M agyar ruler loyal to the K hazars, the kende, w ere in the eastern areas, while the arm y commander, the jila had his m ain dwellings in the west. This arrangem ent was natural, since the M agyar incursions w hich served conquering and trade purposes, were directed westwards at this time. As early as 839, the M agyars were in the Lower Danube region; Cyril m et a group o f them in 861 in the Crimea; but, as we have seen, in alliance with the M oravian ruler Rastislav, they encountered the Franks in 862, and were near Vienna, together w ith the Khavars, by 881. All o f which leads to the assum ption that, especially after the collapse o f the Avar Empire, the focus o f involvem ent in M agyar m ilitary campaigns shifted to the western tribes o f the Etelkoz tribal confederation. This, to some extent, bears com parison with the situation in the Carpathian Basin a century later, w ith the one great difference that w hile the 1Oth-century struggles ended with the victory o f the H ouse o f Arpad, the dom estic conflicts in the 9th century led to the downfall o f the House o f Levedi, the Kende, and brought the trium ph o f the House o f Arpad, i.e. the jila . We know from the Jayhani tradition that the Etelkoz M agyars lived in dome-shaped tents, they followed the sprouting grass and vegetation, that is they m oved to and fro between their sum m er and w inter dw ellings. W hen winter approached, all tribes m oved to the river nearest to them , and rem ained there throughout the w inter and hunted. The w inter dw ellings, then, were in the river valleys, w hile the sum m er dwellings w ere further up in the hills. The author o f the Jayhani tradition notes that the M agyars’ country abounded in woods and waters, the soil was moist, and they had plenty o f arable land. They always defeated the Slavs that dwelt close by. They levied severe sustenance taxes on them, and treated them as w ar prisoners. W henever the M agyars would go to the port o f the Black Sea, they w ould hold a m arket w ith the Byzantines. Here they w ould sell their Slavic prisoners and buy brocade, woollen carpets and other Byzantine goods. W hile the western sources tell us about the M agyars’ incursions on the west, the M uslim sources speak o f their m ilitary actions in the north, o f the raids on the Slavs for taxes and slaves. It is very interesting and typical that the M agyars had their own arable lands, w hile at the sam e tim e they subjected the Slavs to severe sustenance taxes. Although agriculture played an im portant role in the Etelkoz, the foodstuffs produced w ere nowhere near sufficient. The Jayhani tradition describes the agriculture o f the Slavs, largely dependent on
330
From the Urals to the Carpathian Basin
the production o f honey and pig-keeping. They reared pigs— so Jayhani’s story runs— like other peoples kept sheep. They lived in a forest area, and had neither vineyards, nor arable land. This leads to the assum ption that the sustenance taxes levied by the M agyars m ust have been paid in the form of pork, honey and vegetables, and not wheat. The residence o f the M agyars in the Etelkoz was brought to an end by the weakening o f the House o f Levedi triggered by Pecheneg incursions in the east. Backed by the Khavars (who had recently sided with the M agyars), Alm os, who held the office o f jila at that time, ousted Levedi from his fragile position, and had his own son Arpad elected as ruler. The K hazars acknow l edged the change o f power. The M agyar chieftains made A rpad prince “according to the custom, or ‘zakanon’, o f the K hazars” by lifting him upon a shield (see C hapter XIII). A rpad’s rule in the Etelkoz, however, did not last very long. No available sources report on the time elapsed betw een A rpad’s election as prince and the events o f 894-895. It is highly unlikely that this was a long period o f time, however. The fact that A rpad’s son Liuntika headed the m ilitary campaigns against the Bulghars in 894 suggests that Arpad m ust have been past his prime by then, that is, he was at least forty years old. Also, the general Pecheneg attack on the M agyars m ust have followed the P echenegs’ incursions fairly quickly. Porphyrogenitus wrote that a mere “some years” had passed between A rpad’s election, the Pechenegs’ attack, and the Conquest. “Som e years” m eans four or five at most, but m ore probably two or three— that is, the years around 889 and 890. The question is why Regino referred to the M agyars only in 889. One possible answ er is that news o f the change o f dynasty (or news regarding that year) had not reached Priim in Lotharingia earlier.
2. THE HISTORICAL PRELIMINARIES OF THE CONQUEST We m ust discuss the series o f Eurasian events related to the M agyar Conquest in greater detail. The following account o f events follows a geographical and chronological order. C ontem porary to the M agyar Conquest, the Sam anid dynasty ruled today’s Iran, A fghanistan, and Turkm enia and the related territories. In early 893 the Sam anid ruler N asr ibn A hm ad I died, and was succeeded on the throne by his younger brother Esm ail ibn Ahm ad who im m ediately set about organising m ilitary cam paigns to extend his pow er (see Plate III for the m oney minted by Esmail). In the first instance he besieged the town o f Taraz in the valley o f the R iver Taraz or Talas, which was the stronghold o f the Kharlukhs. The K harlukhs suffered irreparable losses in the attack, and consequently they
The Conquest
331
were unable to sustain the pressure they had been exerting on the neighbouring Oghuz and Kimeks. In turn, the Oghuz deemed that the tim e had com e to build their own empire, and to that end they would first have to drive out the Pechenegs who then ruled the western areas o f the steppe. From G ardizi, we know exactly where the Oghuz dwelt (see above p. 235). In all probability they could count on the co-operation o f the Kim eks, too. In 893 the Pechenegs suffered heavy losses in the fights with the Oghuz, and their tribes split into two groups. One group, as it was pointed out above (see pp. 235-238), was still living close to its form er dwellings in 922. The defeat forced the m ajority o f the Pechenegs, however, to set o ff westward. They m ade an abortive attem pt to break through the K hazar Em pire, following which, in 894, they crossed the River Volga. In the meantim e, the Balkans, too, were bustling with events. In the wake o f various disputes regarding trade, the Bulghar ruler Simeon attacked B yzan tium. The em peror Leo VI the Wise thereupon sent em issaries to the M agyars to win them for an anti-Bulghar alliance. Nicetas Scleros conferred with the M agyar chieftains at the Lower Danube in 882. The im perial arm ada then rowed up the Danube to transport the M agyar troops to the rear o f the Bulghars. The M agyars levied a shattering attack upon the Bulghars, follow ing which their armies devastated Bulgharia and even reached Preslav and M adara. The exact date o f these attacks can be readily established from the sources: they occurred after the solar eclipse in 891 and the death o f the patriarch Stephen in 893— that is, in 894. In order to secure his position against Byzantium , Simeon launched a counter-attack against the M agyars during the follow ing year, to which end he needed allies, and the Pechenegs, who had then arrived at the M agyar frontiers, were ideal for the purpose. It is not certain w hether the Pechenegs attacked the M agyars at the instigation o f the B ulghar ruler, or whether Simeon m erely took advantage o f the Pecheneg attacks and also marched against the M agyars; however, there is general agreem ent on the fact that the M agyar’s defeat in 895 in the Etelkoz was a result o f P echeneg-B ulghar m ilitary operations. The M agyars living in the Etelkoz from the end o f the 7th century had long known the situation in the Carpathian Basin. As has been pointed out earlier (pp. 244-246), in 862 Carlom an turned against his father, the Frankish King Louis o f Germany. Seeking to exploit the feud, and in support o f his efforts o f independence, the M oravian ruling prince Rastislav called in M agyar auxiliary troops, and won a small victory over the Franks, In 881 (by then Svatopluk ruled M oravia), the M agyars again interfered in the M oravian-Frankish struggle, and fought outside Vienna, together with the Khavars. Eventually, in 892, C arlom an’s son Arnulf, the new Frankish ruler, brought in the M agyars, now against Svatopluk. A rnulf concluded an alliance
332
From the Urals со the Carpathian Basin
w ith the Bulghars, too, against the growing influence o f the M oravians. Frankish envoys sailed down the Sava and persuaded the B ulghars to obstruct salt consignm ents from being delivered to M oravia. This m arked the begin ning o f the Frankish-B ulghar coalition. Given these circum stances, it is understandable why Byzantium appealed to the M agyars for an alliance against the F rankish-B ulghar coalition. The Bulghars, however, looked for association with the Pechenegs. In 894, Svatopluk sent envoys to the M agyars for help against the Frank ish -B ulghar confederation. The M agyars arrived along routes they had com e to know in earlier cam paigns, and in alliance w ith Svatopluk they attacked the Franks and the local, affiliated Slavs in Pannonia. Following the death o f Svatopluk (which was unrelated to these fights), his sons shared the M oravian Empire. The M agyar troops ravaging Pannonia withdrew, and it seem s likely that they, or a group o f them , m ay have spent the w inter m onths in the Upper Tisza region. Thus, it indisputably follows that in 894 the M agyars w ere involved in two m ilitary campaigns. One against the Danube B ulghars in alliance w ith B yzan tium, and the other against the Franks in alliance with the M oravians. O f the two cam paigns the Bulghar one was o f greater im portance to them , for they channelled their main forces here. By then, the M agyar armies that withdrew to the region o f the R iver Tisza had obtained extensive and accurate intelligence about the status quo o f the C arpathian Basin, the decline o f M oravian pow er caused by internal conflicts, the dom estic affairs o f the Frankish Em pire, and forem ost, about the internal relations o f the Carpathian Basin. Thereupon they decided to interfere in these dom estic fights with added intensity. In 895, the m ain body o f the M agyar army crossed the C arpathians at the Verecke pass, and launched the m ilitary operation which entered historiography as the ‘C onquest’.
3. THE CONQUEST Accounts o f the C onquest in Hungarian historiography were, for a long time, founded on the stories related by the late Hungarian chronicles. This was follow ed by a period during w hich a m ore critical approach was taken, and it w as then established that the Hungarian chroniclers had draw n on three groups o f sources. Firstly, they m ust have know n a few o f the w estern chronicles (R egino’s, for instance); secondly, they probably used a contem porary copy o f the Hungarian prim ary chronicle; and thirdly, they incorporated the clan traditions o f nobiliary families o f their own time. These they further elaborated in the narrative style in vogue at the time. This type o f source criticism for-
333
The Conquest
Figure 67 The M agyar Conquest warded research in so far as it tracked down the western sources. Trying to single out the elem ents adapted from the lost prim ary chronicle or originating from family traditions (o f equivocal authenticity, but otherw ise interesting), or those purely added by the chroniclers to spice up their account, depended on the com petence, experience and judgem ent o f the historian. This was perfectly natural. Even the m ost up-to-par historian could m erely claim that an event quoted from a chronicle reflected the actual facts o f the C onquest “in all likelihood” or “with great probability” or “m ore likely than not” . Historians o f the ensuing period o f historiography included the contem porary sources in their discussions about the Conquest, but in reconstructing the chain o f events,
334
From the Urals to the Carpathian Basin
they still tended to use those parts o f the chronicles which best fitted the concept they had in mind. Below, we shall attem pt to reverse this process. We shall seek, therefore, to reconstruct the events o f the Conquest from contem porary or near-contem porary— hence extraneous— sources; disregarding, in the first instance, the H un garian traditions, that is, the accounts o f the posterior chronicles. We shall subsequently stack up the resulting (evidently less colourful) picture against the version in the chronicles. Based on the evidence, we can divide the C onquest into three m ain phases. The first began in 895 and ended in 898, and the second lasted from 899 to 900. The third phase, between 900 and 902, concluded the events. The paucity o f evidence regarding the first phase leaves us to conjecture as to whom the M agyars w ere fighting at this time. The sources— Porphyrogeni tu s’s work— tell us that when the M agyars w ent to war, the Pechenegs attacked them, together with Simeon. The M agyars cannot have been involved in the M oravian-Frankish conflict, for they had w ithdraw n from there in 894. There is no indication in the sources w hether they w ere in the U pper Tisza area, or had returned to the Etelkoz. Nevertheless, they launched a new m ilitary cam paign in spring 895— the only conceivable objective being an attack against the Bulghars. The known events point to the fact that the M agyars m ust have wanted to storm the B ulghars’ Transylvanian estates to thus ap proach the Bulghars from the rear. This, however, entailed occupying the Transylvanian salt m ines which earned the M agyars a strategically crucial position in the following years. A t this point w e m ust be rem inded o f the fact that the purpose o f the F rankish-B ulghar alliance, in the first place, was to isolate the M oravians from those Transylvanian salt mines. The M agyars m ust have left a garrison behind to guard their dw ellings in the Etelkoz. The Pechenegs and the Bulghars caught the M agyars o f the Etelkoz in a pincer movement. The Pechenegs launched attacks from the east, the Bulghars from the southwest, that is, the Lower Danube area. The Pechenegs beat the Etelkoz M agyar garrison which fled towards the B ulghars w ho had set them a trap. Porphyrogenitus’s account tells us that the survivors fled back to the Etelkoz. The battle itself is described by two sources. In the Continuation o f the Chronicle o f George the M onk we can read that Sim eon sought peace from the Byzantine em peror who, in turn, sent his emissary, Leon Hoirosphactus, to conclude the pact, whilst sum m oning back his land forces and navy. Sim eon, however, w ithout even deigning to listen to him , had the Byzantine em issary locked up, follow ing which he launched a m ilitary cam paign against the Turks, and “because the Rom ans had failed to grant them support and unw isely left them to fend for them selves, they m assacred every one o f them , further
The Conquest
335
heightening their hubris”. The battle is recounted by the Fulda Annals, too. Relating the events o f the year 895 they m ention that the M agyars (referred to as “Avars” in the text) storm ed the borders o f the Bulghars, but the latter were on their guard, and they killed the greater part o f the M agyar army. In the year 896 the annals give the full particulars o f the battle, indicating that the Bulghars had lost tw enty thousand warriors. Now, if the battle took such a heavy toll on the v icto r’s men, w hat were the M agyars’ losses like?— asks the author. The detachm ents beaten by the Pechenegs and the people o f the base dwellings were forced to flee. Together with the armies returning from the Low er Danube, there was only one conceivable place they could go: to where the M agyars’ undim inished armies resided, that is, to Transylvania and the Upper Tisza region. So there they went, via the passes o f the Carpathian Mountains. There is no indication in the sources as to w ho the com m ander o f the spring 895 M agyar m ilitary operation was, but because no source im plies that A rpad had actually participated or died in the course o f the fights with the Pechenegs and the Bulghars, we have reason to believe that he alone, or jointly with Kursan, led the operation. This does not contradict Porphyrogenitus who claimed that the Pechenegs am bushed the M agyars, and drove them away, together with their prince Arpad. Porphyrogenitus highlighted the gist o f the matter: A rpad’s dwellings had been dem olished, consequently he did not return to there, for he had been driven away. We m ust here consider the instance o f the M agyars’ crossing o f the Danube with B yzantine help in the previous year, 894, following which they defeated the Bulghars. We learn from Porphyrogenitus that A rpad’s son, Liuntika, was the arhon at that time. The em peror had earlier referred to all seven chieftains, including Arpad, as being arhons— w hich m akes it unlikely, as has been suggested, that A rpad was removed from his position for a b rief period in 894. It seems m ore likely, rather, that Liuntika headed the tribe or tribes responsible for the Bulghar military operations. A different interpretation o f the same sources contends that Sim eon pro voked an attack from the M agyars to engage their entire army, follow ing which the Pechenegs storm ed the unguarded M agyar dwellings from the rear. We do not think this is likely, however, because Porphyrogenitus writes that “ [...] when the Turks [M agyars] had gone o ff on a m ilitary expedition, the Pech enegs w ith Symeon came against the Turks [...]” (translated by Jenkins)— that is, the source clearly speaks o f two different m ilitary actions. The conquest o f South Transylvania, form erly under Bulghar rule, and the protective shield o f the Carpathian M ountains provided the arriving M agyars with a m om ent o f respite.
J JO
From the Urals to die Carpadiian Basin
O ur sources do not allow us to say for certain w hether the 895 military cam paign was a conscious Conquest, or w hether it was the outcom e o f an under-pressure situation which the M agyars w isely tried to m ake the m ost of. That the Conquest m ight have been planned is suggested by the fact that, many decades prior to the Conquest, the M agyars had traversed and explored the Carpathian Basin. Yet if they had proposed a m igration (a conquest) for that year, they w ould hardly have left behind vulnerable groups o f their people and their base dwellings. However, the fact that, despite a series o f unfortunate events, the M agyars m anaged to keep their heads above w ater goes to show that they w ere indeed ready to m ove on. The sources rem ain silent about the M agyars over the next four years, which is perfectly understandable. The M agyars m ust have crossed the R iver Tisza, but definitely not the Danube. During these four years they secured their frontiers and recovered and reinforced their authority over the occupied territories. Understanding the second phase o f the C onquest requires know ledge o f the events and m ain characters involved. A m u lf’s delegate, Braslav (who had participated in the wars against the M oravians) ruled Transdanubia. M oravia was ruled by one o f Svatopluk I ’s sons, M oym ir II. Lam bert o f Spoleto (892-898), claim ing to be related to the Longobardi, and B erengar o f Friaul (888-924), a descendant o f C harles the Great, struggled for pow er in North Italy. A rn u lf was crow ned king in 894, and in 896 he m arched to Rom e with his army, to have him self crowned emperor. It seems that the M agyars strove to m aintain peaceful relations with most o f the above rulers. Their endeavours were successful in so far as M oym ir II was concerned, w ith whom they concluded an alliance. We also know that m any o f the M oravians shaved their head in the sam e fashion as the M agyars did, which was no obstacle to ordaining an archbishop and three bishops in M oravia in 898-899, with the perm ission o f the Pope. In the m eantim e, however, the Czech nobiliary factions seeking independence w ere giving M oym ir II a very hard time. They asked A rn u lf for help, and soon the Frankish foreign policy turned in support o f Svatopluk I ’s other son, Svatopluk II. M aking the m ost o f this fraternal strife, the Frankish forces m arched into M oravia and, in league with the Czechs, they severely w eakened the M oravian Empire. In 898 Lam pert died, and A rnulf reckoned that the tim e had com e to extend his pow er over the northern part o f Italy. To achieve this he had to underm ine King B erengar’s position. A rnulf had had good experience with the M agyars earlier and knew he could count on them once more. So he offered the M agyars a deal which m ust have included substantial financial benefits, in exchange
The Conquest
337
for which they were to march on Berengar and shatter his power. The sources claim that A m u lf granted the M agyars m oney and clothing. Their deal was concluded according to both Christian and pagan rituals, swearing by G od and Dog respectively. The M agyars sent troops to attack N orth Italy in spring 899. They spent the whole autum n and winter there, and practising their unusual tactics they beat B erengar’s forces. Even in the following year they w ere pillaging the rich Italian towns, advancing as far as Bologna, and on 29th June 900 they even made an (abortive) attempt to sack Venice. Finally, B erengar bought the withdrawal o f the M agyars by offering food and hostages. In the m eantim e, A m u lf him self leaped into action, and interfered in the M oravians’ dom estic struggle in 899. This was not wholly successful. Fol lowing a b rief period o f retirem ent, M oym ir returned and launched a counter attack. He crossed the Danube, devastated A m u lf’s hinterland, Transdanubia, and beat Braslav, who fled. In the m eantim e, in D ecem ber 899, A m u lf died. The forces returning from Italy and the main M agyar arm y crossing the Danube caught the M oravians in a pincer m ovem ent, who were busy devas tating Transdanubia, and occupied that region. Chronicles relating the events o f the year 900 report that the Bavarians were building defensive lines on the River Enns. Naturally this does not m ean that the Enns m arked the border o f the occu pied territories, but it certainly fell within the sphere o f m ilitary operations. This was neither the first nor the last time that it becam e a frontier river in this sense. In effect, this m arked the end o f the M agyar Conquest. The M agyars spent the next three or four years consolidating their control over the Carpathian Basin. There is one other interesting point regarding the m ilitary operations in Transdanubia, nam ely an analogy applicable— with caution— to the events o f 894-895. The year 894 also saw a successful m ilitary cam paign against the Bulghars. This victory could have led the M agyar supreme com m anders to the assum ption that the returning trium phant arm y and the m ain M agyar forces could catch the Transylvanian Bulghars in a pincer m ovem ent. However, the Bulghars, in league with the Pechenegs, decided on a m anoeuvre along sim ilar lines. Good tim ing on the part o f the Pechenegs earned the M agyars’ foes victory. That was not the case in 900. Following the Conquest, the M agyars were left to secure the gains they had made. They first opted for the pacific alternative and sent peace delegations to the court o f the new Frankish ruler, Louis the Infant. A fter their offer was rejected they took severe measures. The M agyar armies launched an attack against the area under Bavarian control (today’s Austria). Pressing westwards
338
From the Urals to the Carpathian Basin
along both banks o f the Danube, they occupied the Kisalfold (N orthw est P lain) and the territories to the north, possibly even beyond the Enns. The advance o f the sm aller M agyar detachm ent on the north bank o f the Danube was halted by the arm ies o f the M arquis Luitpold and the Bishop o f Passau in Novem ber that year. Having w itnessed the events, the other detachm ents, on the south bank, also decided to turn back. Following a series o f m inor m ilitary opera tions, in 902 the M agyars finally occupied M oravia. One part o f the country’s population fled to the Bulghars, to the Croats, and to the Franks. This marked the end o f the actions aim ed at securing the M agyars’ gains, and o f the process o f the C onquest o f the Carpathian Basin. This is m ore or less the portrait rendered by the sources contem porary or near-contem porary to the Conquest. We are fully aw are that this picture is highly deficient; however, we did not draw on the Hungarian chronicles to colour it. The Hungarian prim ary chronicle is thought to have been first com m itted to writing in the 1070s. T he first draft was rew ritten under K ing Colom an 1 Beauclerc (1095-1116); it is obvious, however, that the text created around 1070 had had its sources and editorial schem es (see pp. 414-415). The his torical traditions then evolved for 220 years before they w ere noted down. We have com e to the point, therefore, where other scholars m ust take over.
NOTES For the literature on the questions o f Levedia and Etelkoz see pp. 324 and 421. The Hungarian word a sszo n y ‘lady, married w om an’, in early Hungarian ‘queen’ was discussed by Ligeti (1986, p. 164), where the word is considered to be an 8th-century loan word o f Alanian origin (see also Benko 1992-1995, vol. I, p. 55). Many scholars have written on the localisation o f Krh (see Nem eth 1930, Benko 1985). Recently Szabolcs PolgSr in a yet unpublished lecture on the 18th September, 1998 and Istvan Bona in his inaugural lecture read at the Academ y o f Sciences (18th March, 1999) have independently dealt with the question. Bona suggested to identify the Krh o f Ibn Rusta with Kherson. The arguments o f Polgar and B6na have convinced me that Kerch can be excluded, and the place where the Magyars sold the slaves was either in Kherson or som ew here near. For the events o f the Conquest and the historical setting, see Gyorffy (1984) and Kristo (1980, pp. 151-228; 1996, pp. 175-203). Both authors build on earlier Hungarian scholarship. In these works there are som e facts, the interpretation o f which 1 disagree with, and som e o f their reconstructions are difficult to accept. The two authors frequently disagree and there exists a growing secondary literature which argues on one or the other side. It would be templing to go into the details here, but I wish to avoid scholarly discussions in this book.
IX.
THE MAGYARS
IN THE CARPATHIAN BASIN
The M agyars o f the Conquest changed their way o f life in the Carpathian Basin. Given the conditions o f the age, this transition required a m ere two hundred years w hich, historically speaking, counts as a brief period. N aturally the transition continued under the rule o f Colom an Beauclerc (1095-1116). The ensuing period was characterised by keeping abreast o f—and not sim ply catching up with— the European trends. A turning point in the process o f change was the establishm ent o f Hungarian statehood, laying the foundations o f political and econom ic organisation, i.e. the procedure which the H ungari ans sym bolically refer to as “ Saint Stephen’s oeuvre” . The foundations of Saint Stephen’s state organisation had been laid by that branch o f the House of Arpad which assum ed a m ajor role in the struggle for pow er back in the 10th century. The new style o f state organisation brought to an end a very controversial and com plex period, known in Hungarian historiography as the period o f “adventure cam paigns” (kalandozasok), in effect, o f incursions. We shall here m erely attem pt to outline the circum stances the conquering M agyars lived under at the time. These conditions did not essentially differ from the conditions they had been used to in the Etelkoz. The only new elem ents were really the tasks they were required to perform as a result o f the Conquest, and relating to the cohabitation with the population they encountered in the Carpathian Basin. Practically no written sources exist regarding the life o f the people in this period. The western sources conceived in the period o f “adventure cam paigns” rarely gave rigorously objective accounts o f the events, and even these were usually spiced up by repeating the old truism s about barbaric incursions and the usual negative popular cliches o f m ediaeval historiographers. O cca sionally, however, for w hatever reason, they actually recorded some fascinat ing events o f reality. The Hungarian chronicles w ritten down two centuries after the C onquest tend to relate the glorious traditions o f the nobiliary families o f the period. Consequently they abound in highly colourful stories o f the past, and give scornful accounts o f the history o f rival fam ilies, or rem ain silent about the often significant past o f these rivals. All this was natural to the
340
From the Urals to the Carpathian Basin
European chronicles o f the age. Still, m any elem ents o f these stories preserve m em ories which, subjected to expert historical analysis, can be “converted” to actual historical facts. Accordingly, the follow ing reconstruction— by com parison w ith the chro nological history o f the C onquest where w e disregarded the posterior chron icles— draws on the relatively scanty evidence contained in the Conquestrelated contem porary, non-H ungarian sources, as well as in the w orks o f the Hungarian chroniclers.
1. THE N A M E OF THE MAGYAR TRIBAL CONFEDERATION The M agyar conquerors’ tribal confederation was called (in m odem H ungar ian) H etm agyar ‘Seven M agyars’. The reconstruction o f the original pronun ciation w ould be roughly hetiimajer or hetiimajeri. The w ay A nonym us spells the designation is hetumoger. The uncertainties o f L atin spelling at the time m ade it difficult to render the labial, short H ungarian /а/, and so it was often spelt as o. Spelling the Hungarian consonant cluster dzs (pronounced as the first letter o f the English ‘ju m p ’ and transliterated as j in this book) in Latin w as particularly troublesom e. To m ake things even w orse, dzs had begun to m ake its shift towards gy (as in the B ritish English ‘duke’), that is, /dy/. Certain records tagged on a final -i. This vocalic final m ust have been added to the designation o f the M agyars at a time w hen the language w ould not allow a word-final consonant. This close and short vowel disappeared in the Old Hungarian period, as did the final -ii from the word hetti ‘seven’. Two independent foreign sources vouch for the above designation. One o f them is Porphyrogenitus’s The Governing o f Em pire, in w hich he m entions the seven chieftains heading the seven tribes; and the other Ibn H ayyan’s 1 lth-century work which, however, preserves an earlier, 10 th-century record with an inserted note. The latter, too, claims that the M agyars had seven chieftains. Naturally, Ibn H ayyan’s inform ation does not pertain to the M a gyars m aking raids on Spain. W hen he got to this point, the author took a b rief interval in his account o f the events, and related the know ledge avail able about the m arauding M agyars in Iberia. As will be pointed out, the fact that the M agyars’ designation at the tim e o f the C onquest stood for ‘Seven M agyars’ does not m ean that they w ere actually constituted o f seven tribes. It cannot even be ascertained w hether the tribal confederation consisted o f seven M agyar tribes. The nam e is m erely one o f m any Turkic designations for tribal confederations, w hich tagged on a num ber preceding the tribal nam e, as in the Three K harlukh (Uch K harlukh), the Three Khurikhan (Uch Khurikhan), the N ine Oghuz (Tokhuz Oghuz), the Ten Oghur
The M agyars in the Carpathian Basin
34 1
(On Oghur), the Thirty Tatar (Otuz Tatar), or the Thirty Oghur (Utur Oghur > Uturghur), etc. I f the M agyars’ nam e had a Turkic equivalent, it w ould have sounded Jetimajer or perhaps Jetimejer. There is no such nam e in the sources, unless, o f course, A nonym us’s Dentumoger was a distorted derivation o f it.
2. POLITICAL ORGANISATION The M agyars’ political organisation at the tim e o f the C onquest was a tribal confederation. Tribal confederations took m any forms. We have relatively good knowledge about the Turkic-type tribal confederation system ; presently we know m ofe about the system o f the East Turks than o f the West Turks. The differences are not insignificant, but where w e are lacking inform ation re garding the West Turks we are com pelled to draw on our know ledge o f the East Turks— with due reservations, o f course. N aturally w e cannot get in volved in discussing the details in this book. Tribal confederations had a leading tribe around w hich the m ulti-faceted system o f the confederation was grouped. It w ould be a m istake to assum e that the M agyar tribes held a great assembly, and by m eans o f a dem ocratic vote they elected one o f the chieftains as prince, whose old tribe then elected a new chieftain, resulting 7 + 1 leaders in the confederation o f the seven tribes. The chieftain o f the leading tribe was the head o f the entire confederation, and at the same tim e he rem ained the leader o f his own tribe, while the other chieftains assum ed a subordinate role. This basic set-up becam e m ore and m ore elaborate as new tribes— w ilfully or by force—joined the confederation. The chieftains o f the old tribes collec tively ruled over the new tribes, under the prince and the leadership o f the confederation. The chieftains o f the new tribes had significantly less power; they were delegated special tasks on the perform ance o f w hich their position (and indeed often their life) depended. Recorded by the Chinese sources at one particular m om ent o f its history, the structure o f the N ine O ghuz (Tokhuz Oghuz ) tribal confederation is especially interesting. Viewed from one perspective it consisted o f 18 tribes, 17 from another, and 9 from a third. The answ er to the riddle is this: the ninth o f the nine tribes itself com prised nine tribes. The sources w ould start by listing the first eight tribes and when they got to the ninth they w ould continue with the first tribe o f the sm aller unit o f nine, and proceeded until the ninth. That made 8 + 9 = 17, but from another perspective there was a confederation o f nine tribes, the ninth unit o f w hich was adjoined by another unit o f nine; that is, the larger unit was the one-ninth entity in the first instance, and in the second, one o f the small tribes, which am ounts to 9 + 9 = 18. But w hen it came
342
From the Urals to the Carpathian Basin
LEADING TRIBE
Figure 68 Structure o f the Nine Oghuz tribal confederation
to negotiations, the confederation acted as a single unit, and the leading tribe— the Uighur— was in control o f the Nine Oghuz (see Figure 6 8 ). The pow er o f the prince was secured, to a great extent, by his m ilitary train or retinue. The p rince’s retinue consisted o f his own clansm en, rather than other chieftains; also, o f m em bers o f other tribes who very often assum ed a key role, partly as hostages and partly because o f their vested interests which bound them to the prince. It so happened that if a tribe rebelled, the kinsman o f that tribe in the escort o f the prince could be used to hold in check the rebelling tribe. Outsiders, too, could be granted m em bership in the escort, and the actual power o f the prince depended on the extent to w hich he was able to m anage and keep control over his escort. This system m ust have achieved, to a certain extent, a balance o f power. The personal authority o f the prince depended partly on the pow er o f his own tribe, partly on the system o f vested interests o f a m ultiply differentiated tribal confederation, and partly on the pow er o f his escort. A dual system o f pow er operated w ithin this basic political fram ew ork. The terms dual kingship and the holy, sacral kingship are, strictly speaking, two aspects o f the sam e concept, even though in Hungarian historiography they have different meanings. This differentiation is m isleading, because sacral kingship intrinsically postulates a dual kingship; conversely, however, a dual kingship is not necessarily sacral. Underlying the dual system is a principle
343
The M agyars in die Carpadiian Basin
concerning the ru ler’s legitim acy o f power. In the dual system the ru ler’s power originates from the heavens, notw ithstanding the circum stances which earned him his throne. As we have seen, everything depended on the ru ler’s kut or holiness or charisma, which he received from the heavens and whose power he used to govern (see pp. 149-150). The longer a dynasty possessed this celestial gift, the less it participated in m anaging the everyday m atters o f the realm. This was characteristic o f all m ediaeval powers, not only in the steppe. Alongside the retired king, the chancellor, the m ajor-dom o, or the com m ander came to the fore in the royal courts. The older the dynasty, the greater the role o f the person actually w ielding authority was. This second position would often be controlled by a clan whose m em bers (passing down the title not from father to son, but to the second eldest clansm an) m ight grow so strong that they would often become the real pow er behind the throne. In this system, then, the duality o f the king and the suprem e com m ander was natural, and their relationship largely depended— beyond their person and the given set o f circum stances— on the ritualisation o f the ru le r’s legitim acy o f power. Naturally this rite only warranted the position o f the sacral king so long as the circum stances o f the com m unity and the tribal confederation were favourable. To outsiders this system m ust have appeared as a dual kingship. However, the only thing a sacral kingship had in com m on with a dual kingship was that, o f the legitim ate ruler and the supreme com m ander w ielding power, the former became the object o f such rites which associated the life o f the king with the fortune o f the tribal confederation. The ru ler’s kut and the peo p le’s— i.e. the tribal confederation’s— kut thus becam e intertw ined. The sources m ust be interpreted in the light o f this. The Jayhani tradition writes thus: Ibn Rusla
Gardizi
The Majgars are a race o f Turks and their leader rides out with horsem en to the number o f 20,000. Their leader is called kilnde [or kende, the first vow el is omitted] but this is only a nominal title, for the name o f the man who is actually king over them is called j ila [or ju la ] and all the Majgars accept the orders o f their j ila [jula] in the matter o f war and defence and the like.
These Majgars are a type o f the Turks. Their leader (sd la r) rides out with 20,000 warriors. This leader they call ktinde [kende]. This is the title o f the greater o f their kings (w a in nam -e m alek-e bozorgtar-e ishan ast). That leader (sa la r) who appoints the functionaries they call j u la [jole]. What the ju la commands, the Majgars do.
The only thing the above sources tell us is that the gyula was in charge o f the m ilitary m atters o f the tribal confederation; w hereas there existed a legitimate ruler who had little influence on arm y-related issues.
344
From the llrals to the Carpathian Basin
A closer look at the Jayhani tradition reveals that this set-up reflects conditions prior to the Conquest. The question arises o f when the dual kinship existed am ong the M agyars, and w hether any other sources referring to it exist? The Levedi story could be regarded as another source. As discussed in C hapter XIII, it relates the ousting o f the form er legitim ate dynasty, when the tribal confederation’s second clan— the suprem e com m ander— took over control. Contem porary or near-contem porary sources are in agreem ent regard ing one question. They talk o f neither L evedi’s nor A lm os’s progeny, but only o f A rpad’s. This m eans that according to the 9 t h - 10th-century Byzantine sources both royal and m ilitary powers tem porarily converged in A rpad’s hands. We know for sure that this was the case in 894, the year w hen N icetas Scleros m et Arpad and Kursan in the Lower Danube area to talk o f w ar against the Bulghars. I f one o f them had been the sacral king, he w ould not have taken part in the negotiations. Two questions rem ain to be answered. One o f them is an event traditionally recounted by the Hungarian chronicles: the m urder o f Almos. One tradition says that “A rpad’s father, A lm os, was killed in Transylvania, for he was not to enter Pannonia” . This has been hitherto explained aw ay as being ritual regicide, a rem nant o f the sacral kingship. This is very unlikely, however. It is hardly conceivable that following the Conquest, A rpad, de facto suprem e com m ander, would m urder his father for ritual or sacral reasons. Possibly Arpad and his father were on bad term s, as w ere Carlom an and his father, Louis o f Germany. If these reports reflect true events, the only possible explanation w ould be that Arpad or som eone in his entourage— perhaps som eone seeking A rpad’s favour m urdered Almos. The other question that rem ains to be answ ered is w ho was the other person Nicetas Scleros met. Historians refer to him as K ursan, a reconstruction o f the nam e Curzan w hich features in the w ork o f the H ungarian Anonym us. The nam e appears in two traditions, independent o f one another. The Byzantine tradition is represented by the anonym ous author o f the sequel to The Chron icle o f George the M onk in w hich the nam e K usan (kousanes) occurs. The variant Kursan (koursane), in a later m anuscript, has been shown to be a copyist’s error. The w estern annals (o f the years 902 and 904) feature the variants Kussal (Chussal: A nnales herem i 904 A nnales alam annici, in w hich it is erroneously spelt Chusdat) and K ussol (Chussol: A nnales sangallenses m aiores 902 <— A nnales alam annici), while the late form o f Kussala (Cussala: Aventinus) originates from Kussal. There is no doubt about the form s Kusan and K usal being related, but it is not clear how (a possible shift o f the word-final consonant from -n to -I is out o f the question). All evidence
The M agyars in the Carpathian Basin
345
points to the fact that word-internal consonant is an Ы and not an /sh/. It is possible, but not certain, that the original word-final consonant was an -/, and that the -n in the Byzantine source is a secondary formation. The link with the form Curzan is not in the least bit straightforw ard, because the regular insertion o f an -r- into com m on Hungarian words (as in hars ‘linden’, nyars ‘sp it’) was a change that occurred later. Consequently, the nam e Kursan is historical convention, as is the nam e form Geza. B ut who was K ursan, or more probably Kusal (Kiisel) or Kusan (Kiisen)? The sources unanim ously vouch for Kusan being the protagonist o f the events o f the second phase o f the Conquest, after 899. There is no reference, however, to w hether his successor or anyone from his clan played a part in the events contem porary with K ursan’s lifetim e or later. Our knowledge o f the order o f succession o f the A rpad dynasty from som ewhat later indicates that Kursan, too, m ust have been a dynasty member, and m ight even have been A rpad’s brother. Visiting the Byzantine court around 950, Termecsii and Bulcsii enum erated A rpad’s four sons who w ere called Tarkacsu, Yeleh, Yutocsa and Zalta. The very same chapter relating their visit claim s earlier on that Liuntika, who probably led the cam paign against the Bulghars, was also A rpad’s son. This goes to show that Termecsii and his com panion neglected to m ention the branches they considered insignificant regarding pow er and authority; and accordingly they om itted Liuntika from their list. The editors o f the text either failed to notice this inconsistency, or sim ply regarded the m atter as too trifling to be o f concern. The claim that one o f the four sons had two nam es (and that one source recorded the one, and other sources the other), is w holly unfounded. It is hardly im aginable that Arpad was A lm os’s only son. As yet, we have no evidence to prove that Kursan was A rpad’s brother. In any case, it was he who led the second phase o f the Conquest and the incursions on the west, and who was tricked by the Bavarians into attending peace talks, and was m urdered on the bank o f the R iver Fischa in 904. I do not consider justified the views w hich contend that the A rpad-K ursan twosom e and the m urder o f Alm os im ply the existence o f a K hazar-type sacral kingship am ong the M agyars o f the Conquest. There was no tim e for that institution to develop after the A rpad dynasty had come to power, and neither were the historical and social conditions favourable. It is quite another m atter that the rulers o f the A rpad dynasty considered them selves sovereigns chosen by the heavens. The problem o f why Kursan alone features in the w estern sources requires further clarification. In C hapter 40 o f his De administrando imperio Constantine Porphyrogeni tus describes the fam ous landm arks and sights o f Turkia, that is, Hungary. He writes o f T rajan’s bridge over the Lower Danube, Em peror C onstantine’s tower
346
From the Urals to the Carpathian Basin
in Belgrade, and o f Sirmium. He continues thus: “ Such are the landm arks and nam es along the River Danube; but the regions above these, w hich com pre hend the w hole settlem ent o f Turkey [Hungary], they now call after the nam es o f the rivers that flow there. The rivers are these: the first river is the Timisis [Temes], the second river the Toutis [Bega], the third river the M orisis [M aros], the fourth river the K risos [Koros], and again another river, the Titza [Tisza]. N eighbours o f the Turks are, on the eastern side the B ulgharians, where the R iver Istros, also called Danube, runs betw een them ; on the northern, the Pechenegs; on the western, the Franks; and on the southern, the C roats.” (Translated by Jenkins.) This description is odd because it suggests that H ungary during the period betw een 940 and 952 consisted m erely o f the Trans-Tisza area. Porphyrogenitus m ight have, perhaps, draw n on the cleric G abriel’s report. Gabriel had been sent by Byzantium as an em issary to the M agyars around 927. Nevertheless, this report describes the actual areas in the possession o f A rpad’s successors, that is, the A rpadian dw ellings. Thus East H ungary was ruled by A rpad’s offspring, and West Hungary, perhaps, by K ursan’s. H ow ever, because the report m entions that Franks occupied the areas further west, A rpad’s successors m ust have considered them selves in charge o f the entire land o f Hungary, but only spoke about their own dw elling in detail. Provided this conjecture corresponds to historical reality, it is very likely that K ursan was a brother to Arpad. Kursan and his offspring headed the raids on the west, and accordingly, their kinsfolk suffered the great losses, too. We know next to nothing about the fam ily background o f Tarhos, Bogat and Salard, the leaders o f the 921 and 924 raids. In 947, a chieftain called Taksony also headed an incursion on Italy. C ertain sources (Aventinus) claim that Taksony died in battle on the Lech plain in 955, together w ith Bulcsu, Csaba, Lei and Sur. Other sources recount, however, that Taksony becam e the head o f the leading tribe around the sam e date— w hich m eans that either the reports o f his death on the Lech plain are m istaken, or that they w ere referring to another chieftain Taksony (Toxus). It is hardly conceivable— although it cannot be ruled out— that after his defeat, on returning hom e he took over pow er in the realm. We know for certain that Bulcsu was not a m em ber o f A rpad’s family. Porphyrogenitus refers to him as being “T urkey’s third prince” (arhon). His title was kharha w hich came after the gyla in rank, according to the emperor. He confirm s elsew here this basic set-up: the first leader o f the M agyar court (khephale prote), the prince, was assisted by the tw o im portant officers, the gyla and the kharha; also, each tribe had a chieftain. W hich m eans that the prince, the gyla and the kharha m ust have constituted the leadership o f the tribal confederation.
The M agyars in the Carpathian Basin
347
N either the title nam e gyla, nor kharha have convincing etym ologies. ( Gyla or gtila, pronounced as yila, m ight reflect the form gyila or jila , but not the form gyula which is a M agyar developm ent; and the Pechenegs had no tribe called jula.) As has been m entioned above, the Jayhani tradition claim ed that the M ajgars had two leaders, a kende or kiinde and a jila. The description o f the system inform s us that, o f the two, the jila was the m ilitary commander. It is hardly doubtful that this jila and the B yzantine version gyla are identical title nam es, but due to the fact that a century separates them w e cannot be so sure that the functions o f the holder o f this title w ere identical at both times. The Byzantine account does not refer to the gyula by name; however, it does tell us that T urkey’s (i.e. H ungary’s) prince (arhon ) in those tim es was Falicsi, A rpad’s grandson. The West Turk political organisation rested upon tw o im portant titles. One o f them, after the king, was the suprem e com m ander, and the other the governor or governors in charge o f joined or subjugated peoples. As has been m entioned above (see p. 226), in the K hazar Em pire the latter was called yiltever w hich corresponds to the East Turk elteber. If w e are looking at a partial adoption o f the K hazar system, then the M agyar kharha m ust have been the governor in charge o f joined or subjugated peoples, w hich explains why he was lower in rank than the gyula. Kursan m ust have originally occupied this position, but it is uncertain w hether his successors w ere his offspring. The title was held by Bulcsu around 950, which is inform ative, because if kharha Bulcsu ruled over the joined peoples, he cannot have been related to A rpad’s clan, for the kharhas w ere never related to it. B ulcsu’s father Kali also held the office o f the kharha. The ru ler probably appointed him after the m urder o f Kursan in 904. Weighing up all the evidence, the leading stratum o f the M agyar’s political organisation can be reconstructed thus: the tribal confederation was headed by the prince, the suprem e com m ander (the gyula), and the kharha who gov erned the jo ined peoples. On the next level cam e the chieftains o f the tribes. Theoretically the conquerors’ tribal confederation consisted o f 7 + 1 tribes, the eighth tribe being the Khavars which originally consisted o f three tribes. This, in the end, m eans ten tribes. The designation Khabar denoting the Khavars, like so m any other names, is Hungarian historical convention. The G reek text reads as Khavar. Although in those days Khabar m ight also have been transliterated by the Greeks in the same way, w e learn from the Latin w ritten sources that the correct reading was indeed Khavar. The variant cowari, featuring in the Annals o f Salzburg, and the G reek khabaroi are to be read as Khavari. The K havars rebelled against the Khazars, but the latter got the better o f them , and “ [...] som e o f them were slain, but others escaped and came and settled w ith the Turks in the land
348
From the Urals to the Carpathian Basin
o f the Pechenegs, and they m ade friends with one another, and w ere called ‘K havars’,” Porphyrogenitus writes. The question is who called them Khavars. The quoted text does not say that one o f the K hazar tribes, the Khavars, rebelled, but rather, that those that joined the M agyars were called ‘K havars’. The ethnic nam e cannot have derived from the Turkic verb kava- ‘to swell (o f a w ound)’ as it was claim ed; sem antic and onom astic problem s make that im possible to establish. I f we w ere to find a related Turkic word, a possibility would be kavir- ‘to collect, assem ble’ which interestingly tallies with kuvra- ‘to be assem bled’, kuvrat- ‘to assem b le ’. (The latter survived in the nam e o f the B ulghar khaghan Khuvrat). M orphologically, however, the derivation is not clear, and m ore likely than not we are again looking at a popular etymology. One o f the leading clans o f the K havars m ust have been called Khavar, a nam e w hich the new group consisting o f three tribes com m only adopted. Popular etym ological explana tions came later, like for exam ple, bodun kavradi ‘brought the people to g eth er’. Sim ilar secondary explanations occurred w ith the Turk and Manys ethnic nam es, and sim ilar stories were told by Anonym us, too. The three K havar tribes— so the em peror tells us— had one chieftain (arhon), a set-up w hich rem ained much the sam e until the times o f the account. O f the eight M agyar tribes the Khavars, who w ere always sent to the front lines in battle, proved to be the strongest and the m ost courageous in war, in reward for w hich they were raised to the level o f the first tribes— we read. Hungarian historiography has long revealed that we are w itnessing here an established m ilitary system w hereby the m ost recently joined people or tribe was sent to the front lines. The above ‘first tribe’ refers to m ilitary position, not to a political rank. The system o f three tribes having one chieftain corre sponds perfectly to the organisational sam ples o f Turkic tribes. This is im portant to acknowledge because it often happened that the m erger o f two or m ore tribes entailed having one chieftain for the new unit. It followed from the K havars’ unique position in the M agyar military organisation that they played a significant part in the incursions on the west. We know that it was together w ith the Khavars that the M agyars w ent to war and advanced to Vienna around 881. This probably m eant that the losses o f the Khavars were greater than o f those that stayed behind in the Carpathian Basin. Two questions require clarification in connection with the Khavars. Is it possible that the leading stratum o f the K havars was converted to Judaism ? The possibility cannot be ruled out. However, w eighing up chronological evidence we can say that the K havars’ joining the M agyars was not wholly independent from the conversion o f the leading K hazar strata to Jew ish faith. Yet we have a relatively good notion about contem porary K hazar conditions,
The M agyars in the Carpathian Basin
349
where the Jewish leading strata lived together in peace with M uslim , Christian and “pagan” com m unities, while the Khazar system o f beliefs was com plex and syncretistic (more about that later). The Central A sian U ighur Em pire can offer some interesting parallels. Here Buddhism and N estorianism flourished under a new ly converted M anichaean ruler. Both the K hazar and the U ighur examples attest to the fact that the K havars’ splintering o ff could only indirectly have been provoked by the conversion to the Jewish faith. The turn evidently interfered with imperial and political interests. The other question is related to the problem o f bilingualism . The relevant Greek text has m any interpretations. “And so (1) to these Turks they taught also the tongue o f the Khazars, and (2) to this day they have this same language, but (3) they have also the other tongue o f the Turks.” (Translated by Jenkins.) The subject o f the clauses o f the sentence is unclear. Either ‘K havars’ is the subject o f all three subordinate clauses, or ‘K havars’ is the subject o f the first clause and ‘Turks’ o f the second and third; or ‘K havars’ is the subject o f the first and third clauses, and ‘T urks’ o f the second. W hat this boils down to is the question “who was bilingual, then?” In my opinion the first version is the correct one, and the report should be read as follows: the Khavars taught the M agyars the tongue o f the Khazars, but the Khavars “to this day” speak this K hazar language, and also learnt the language o f the M agyars— that is, the Khavars were bilingual. The Khavars were under M agyar rule, and the m ilitary lingo o f the tribal confederation m ust have been Hungarian, so they had to learn the language. It is sure that the M agyars first encountered the language o f the K hazars before the Khavars had joined them. However, after the Khavars had joined, the role o f the K hazar language m ust have increased in the M agyar court. The process o f joining was very often confirmed by m atrim ony between the leading clans— as in L evedi’s case, for instance. We know that Levedi had a K hazar wife, consequently the K havar chieftain m ight have taken a M agyar wife. However, having a M agyar wife was not considered to be binding enough, so the K havar chieftain was always supervised by a M agyar governor, the kharha. There are several examples o f this in the K hazar Empire. For instance, the Khazar titleyiltever denoted such a function am ong the Caucasus “H uns” and the Volga Bulghars. If the K havars ’ j oining the M agyars happened shortly after 800, their eighth generation would have lived at the turn o f the m illennium . By that tim e— which roughly coincides with the appearance o f the first traces o f Hungarian literacy— the Khavars had long since lost their original Turkic language, unless o f course in the m eantim e they encountered or m aintained connections with new Turkic im migrants, or with the Turks outside the C arpathian Basin. Around the year 1000, the Khavars belonged to the m ajority o f native speaker
350
From the Urals to the Carpathian Basin
Hungarians. It is perfectly feasible, however, that knowledge o f the K hazar language lived on in a few clans. Porphyrogenitus enum erates the M agyar tribes by nam e, thus: “The first is this aforesaid clan o f the Kabaroi [Khavars] which split o ff from the Khazars; the second, o f Nekis [Nyek]: the third, o f M egeris [M egyer]: the fourth, o f Kourtougerm atos [Kurt and Gyarm at]: the fifth, o f Tarianos [Taijan]: the sixth, Genah [Jeno]: the seventh, Kari [Кёг]: the eighth, Kasi [K eszi].” The sentence has provoked m uch debate on account o f the inconsistency o f its gram m atical structure. Inconsistency begins by the E m peror’s w riting the second to the fifth tribe nam es which com e after the Khavars in the singular genitive, as if he were talking about tribes o f (or belonging to) persons. I do not think that this usage is reflecting a Hungarian gram m atical structure— as it was claimed. It is continuing the previous unit, w hich records the tribe o f the Khavars. The tribal nam es o f Kurt and G yarm at are com bined in the E m peror’s ac count. We have plenty o f examples o f the M agyar tribe nam es being pre served to this day in the nam es o f early M agyar settlem ents. There is no place in Hungary called Ktirtgyarmat; however, Ktirt and G yarm at indi vidually occur frequently as place names. The two tribes m ight have been com bined tem porarily for m ilitary reasons. This rounds out our picture o f the H etm agyar tribal confederation which consisted not o f 7 + 1 tribes, but of 6 + 1 (+ 1) + 1 (+ 3). The Hёtm agyar— literally, ‘Seven M agyars’— tribal con federation was constituted o f eight tribes. This sounds like a contradiction in terms, but in fact the leading tribe— the head o f the tribal confederation— played an extrem ely im portant part. Six tribes joined the original M ajer tribe, form ing a system o f 1 + 6 (while all new tribes that joined w ere fitted into this system ). This is im portant to acknowledge, because it shows that the M ajer tribe substantially outrivalled the other tribes in numbers. M uch has been w ritten on the etym ologies o f the tribal nam es. It m ust be stressed, however, that linguistic historical evidence has no, or not much, bearing on the ethnicity o f the tribe members. We have seen m any examples (e.g. French, Russian, etc.). However, solving the riddle o f nam es offers some interesting historical knowledge. The literature on the etym ology o f names m ay fill w hole libraries, yet probably no etym ology can lay claim to being sound in every respect. The M agyars have two tribe nam es w hose etym ology does not require too many prem ises. These are the nam es M egyer and Tarjan. M egyer (pronounced m ejer in Ancient Hungarian) is identical w ith the ethnic nam e M agyar w hich derives from the form majer. The shift from M ajer to M egyer can be explained by the fact that the w ord stress was on the last syllable, a rem nant o f Turkic word stress patterns (see above, p. 306). In effect, then, M egyer is a variant o f m ajer— as pronounced in Turkic. The etym ology
The M agyars in the Carpathian Basin
351
o f Tarjan is tarkhan > tarhan > tarjan, a Hungarian change w hich was fairly systematic, and m any examples o f it have come down to us. Yet the form tarkhan lived on, as attested to by the w estern sources and the Hungarian place names with the form Tarkany. There are m any instances when two variants o f the same title occur. We have seen examples o f titles becom ing ethnic nam es above (see p. 273). The tribe name Jeno features as Genah in the above quotation. It w as pronounced as Yeneh. This transform ed to the Hungarian Jeno by process o f a regular change. The putative root, the Turkic verb ina‘to trust’, only occurs in the Turkic languages in its back vocalic variant. The so-called prothetic word-initial y - features in a num ber o f other Hungarian words o f Turkic origin. An interesting exam ple is A rpad’s second son’s, Yeleh’s nam e which derived from the Turkic elig, ilig ‘ru le r’, ‘kin g ’. The Hungarian place name Ullo goes back to this ancient word. Thus, if the Turkic word inag ‘friend’, ‘com panion’, ‘m ate’ had a front vowel variant, that is *ineg, the correspondence with the Hungarian Jeneh (read Yeneh) w ould be regular. The Turkic word inag also had the m eaning o f ‘friend o f the ru ler’, ‘m inister’, ‘official’ in m any old Turkic languages. No such front vocalic variant ( *ineg), necessary to establish the etym ology for Jeneh, has yet been found in the Turkic languages, but the phenom enon is not uncom m on. Only the back vowel variants o f the Hungarian word boleny ‘bison’ (and probably also the Hungarian tor ‘trap ’) have been identified in the Turkic languages so far. The Turkic title inanch, deriving from the same root, has a variant written inench which features in the Turkic inscriptions. Nevertheless, only if we accept a conjecture as certitude can we establish an “infallible” etym ology for the tribe nam e Jeno. The duality o f tarjan and tarkhan (Tarkany) corresponds to the duality o f Jeleh and Ullo. The difficulties regarding the etym ology o f the other tribe nam es are even greater, and they rely on even more conjectures. I do not wish to get involved in the details, since the linguistic origin o f the tribal nam es is an inconclusive issue with regard to the conquering M agyars. The m em bers o f the M agyar tribal confederation were, or had been for some time by then, native speakers o f Hungarian. To sum up the picture which the authentic sources reveal: the H etm agyar (literally, ‘Seven M agyars’) tribal confederation was headed by the M egyer tribe whose original Hungarian nam e was M ajer or the “Turkic-sounding” Mejer. This was A rpad’s tribe, the tribe that led the Conquest. The confedera tion was a m ultiply differentiated system o f tribes. The other tribes: Nyek, Kiirtgyarmat, Tarjan, Jeno, Кёг and Keszi m ade up the second circle. The merging o f the Kurt and the Gyarm at tribes m ust have been tem porary only, and it served m ilitary purposes. The third circle within this set-up was the Khavar tribe w hich, led by one chieftain, was divided into three subtribes. The
352
From the Urals to the Carpathian Basin
fourth circle consisted o f the peoples the M agyars found on arrival in the C arpathian Basin and who joined the newcom ers. O f the peoples the M agyars cam e to m eet in the Carpathian B asin the various Slavic groups probably played an im portant role. Initially the M agyars referred to them as tots. The origin o f this ethnic nam e is debated. Either it is a M agyar adaptation o f the ethnic nam e Tat, used to designate people o f the C aucasus, am ong them Indo-Europeans, or it derives from the Indo-European *teuth (present in the Germ an ethnic nam e deutsch), via the O ld H igh Germ an diot ‘people’. For a long tim e the nam e served to com m only designate all Slavic peoples, and only in the 19th century was its m eaning lim ited to denote the Slovaks. The other m ajor groups the M agyars encountered in their new hom eland w ere the Avars, at that time undergoing Slavification. Their name has, perhaps, been preserved in the toponym Varkony. These Avars were bilingual, as w ere the Khavars; the second language o f the form er being Slavic. The conquering M agyars and their allies w ere led by the prince, by the suprem e com m ander or gyula, and the karha, w ho kept control over the peoples who had joined the M agyars or had been subjugated by them. We have no direct evidence regarding the p rince’s m ilitary train (m em bers, nam es, ethnicity) in the 10th century. Turkic analogies reveal that in the K hazar Em pire and elsew here the tarkhan title was, at tim es, used to denote such a group. The w ord tarkhan has erroneously been suggested to have m eant ‘blacksm ith’ in Turkic. Later in the course o f history, the w ord had the m eaning o f ‘tax-exem pt’, a status w hich craftsm en enjoyed. We m ight venture to say, therefore, that the Tarjan tribe was possibly involved in the p rince’s escort; the leader o f the escort m ight have been the chieftain o f the Tarjan tribe. In turn, the tribe m ight easily have adopted as its nam e the title o f its leader, w hich w as a com m on nam e-giving procedure am ong the Turkic peoples. This does not m ean, however, that every m em ber o f the Tarjan tribe w as also adm itted to the escort. All we know about the distribution o f the conquering tribes in their new hom eland is that following the Conquest none o f them had their ow n w ell-de fined areas. The Trans-Tisza region seems to have been inhabited by the Arpad dynasty, but they evidently soon occupied the region betw een the D anube and the Tisza, too. The conquest o f Transdanubia, w hich ended around 900, must have involved m ore directly the joined, new peoples. The dw ellings o f the conquering M agyar clans have traditionally been established by a com bination o f data from toponym ic research and the chronicles. This book cannot devote room to discuss these methods. The place nam es in question cannot have developed earlier than the 1 1th century and this happens to be the tim e when they first cropped up in the sources.
The M agyars in the Carpathian Basin
353
The question rem ains o f w hat the political organisation o f the peoples in the Carpathian Basin looked like, and how m uch o f their set-up did the M agyars adopt? The sources, w hich are rather taciturn regarding the matter, inform us that the zhupans kept control over local groups. T he M agyars kept this title, and set the zhupans to their own service. These becam e the spans. The ispan, a later M agyar title, derived from span, only it had different functions. In the south the Bulghars had a leading zhupan to control the other zhupans. The nam e o f this title in Hungarian was nandor span w hich m eant ‘Bulghar zhupan’, later derivations o f w hich include the titles nandorispan and nadorispan. Taking over existing political/organisational term s from the old dw ellers o f the Carpathian Basin was a relatively com m on phenom enon. The Hungarian word megye ‘shire’, for instance, is o f Slavic origin. The royal title kiraly ‘king’ was also directly taken from Slavic. Finally, going back to the personal name Charles the Great (Carolus M agnus). Krai in the G reek deed o f gift to the Nuns o f Veszprem Valley (before 1002 in a copy o f 1109) is the first authentic m ention o f the word. The sam e word can be read on the Hungarian crown, too (see above, p. 277). Estimations regarding the num bers o f the conquerors usually take the Jayhani tradition as a starting point. Here we can read that the leader o f the Magyars, the gyula, rides out with 20,000 horsem en. Few have noticed the fact that the sam e source claim s that the K hazar’s suprem e com m ander rides out with 10,000 horsemen. It is highly unlikely that there w ere tw ice as many Magyars as Khazars. A Persian source, the history o f Husros I, talks o f 50,000 Khazars around 567, not counting the wom en, children and servants, or the members o f the leading clans am ounting to 3000. Such figures m erely serve to indicate proportions, and in any case, the sources w ere only ever concerned with the num bers o f active warriors. This m ust have m eant a total o f 300,000 with the Khazars. Given this ratio, the num ber at the tim e th с Jayhani tradition was written w ould have been 2 0 0 ,0 0 0 , give or take a little; plus the joined tribes and those peoples that populated the Carpathian Basin w hen the M a gyars arrived. By the m iddle o f the 10th century this region m ust have con tained around h a lf a m illion inhabitants. The individual tribes did not have clearly-m arked land frontiers, but the Magyars soon staked out the lands o f their tribal confederation as a whole. Archaeological evidence has revealed that they posted garrisons, w hich served as the first line o f defence, to the outer slopes o f the Carpathian M ountains. The southern, northern, northwestern and parts o f the w estern borders were guarded by m arches and m arch guards. This also attests to the fact that the Arpad dynasty was still in control o f the tribal confederation, although they did not play an im portant part in the incursions on the west.
354
From the Urals [a the Carpathian Basin
Certain clans within m any tribes would keep up their nom adic way o f life, and would com m ute betw een w inter and sum m er dwellings. It is almost certain that these interior m igrations triggered pow er-related conflicts. This situation had begun to deteriorate in the early 10th century. A fter 955, the lost battle o f Lech, the collapse o f the system o f the tribal confederation accelerated, the tribes disintegrated and settled apart. To avoid significant changes endangering stability, the prince had to bank on the clans. Relying partly on the p rin ce’s m ilitary escort and partly on the leaders o f the mighty clans, the ruling princes, Taksony and Geza (father o f Saint Stephen), set out to reorganise the M agyars o f the Carpathian Basin.
3. SOCIAL STRUCTURE Follow ing the year 628, the Bishop o f A lbania conferred with the Khazars several times. Describing one such occasion, an A rm enian source enum erates the m em bers o f the Khazar delegation. The list proceeds from the highest in rank to the lowest. Accordingly, the source m entions sovereigns, princes, noblem en, m ilitary com m anders and various clan patriarchs. It is very prob able that this stratification reflected contem porary Persian conditions rather than the actual K hazar system; still, it gives some idea o f the set-up o f the K hazars’ leading strata. The Turk texts differentiate betw een hereditary and bestow ed peerage. Hungarian archaeological finds o f the C onquest period reflect an econom ically highly stratified society. Econom ic status, however, did not in every case have a bearing on social standing. If we isolate a relatively hom ogeneous group o f titles— that o f the great Turk inscriptions o f Central A sia— and ignore titles w hich evidently apply to non-Turks (e.g. serve to denote Tibetan and Chinese persons), disregard one-off cases o f personal nam es deriving from peer names, and overlook functions such as ‘sp y ’, ‘envoy’ (also, analytic titles like ‘head o f the arm y’, etc.), we are left w ith some thirty titles altogether, give or take a few. To these we can add the com pound titles, that is, w hen a title com prises multiple elem ents. Thus we can establish that the num ber o f titles actually featuring in the sources on contem poraneous and sim ilar tribal organisations (the Second Turk Khaghanate and the Uighur Khaghanate) adds up to around forty. A closer look at the K hazar titles o f the time reveals that, although the phonetic forms m ay vary, practically every one o f them has an equivalent am ong the East Turks. Even the Danube Bulghars stuck w ith the sam e system, and had only one or two new titles com pared to those in the Turk inscriptions in Central Asia.
The M agyars in the Carpathian Basin
355
It would be expected, then, that the conquering M agyars’ organisation, too, featured a fair num ber o f hereditary and bestow ed peerages. However, the fact o f the m atter is that there are m ore Avar titles, recorded in the w estern sources, than Magyar. Avar exam ples include khaghan, khapkhan, tudun, yugurm sh, khatun, etc. The Latin sources tend to use the stereotypical rex, regulus, dux titles. The M agyar gyula, karha, span and ispan have been dealt with above. The title gyula has only survived in a personal name. M ention m ust be made of the title o f ban which entered the Hungarian language from the Turkic, via Balkan Slavic m ediation. The few still existing Arpadian nam es which pre served title nam es also deserve attention, for instance Jeleg and Ullo (elig ‘king’), Jeno (ineg ‘minister, counselor’), as well as Gyeucsa (pronounced dyeucha'Jevu < ja b g u ‘ru ler’ + -csa, a Hungarian suffix), and Bela (< bila < buyla < boyla ‘m in ister’). The identification o f Zalta, quoted by Porphyro genitus, with the Arab sultan title (via Turkic m ediation) sounds unconvinc ing, not to mention the place names with Solt which have been suggested to be also related in some way. We shall not discuss those place nam es which are positively know n to have derived from personal nam es, and w hich them selves have derived from Turkic titles (like, for instance Inancs). These names occurred relatively late, therefore their uncertain chronology m akes them unserviceable. For instance, they m ight have derived from eastern immigrants after the C onquest period. The nam e o f A rpad’s great-grandson, Termecsti, originates from the Turkic-Avar word terem ‘hall ’. The present-day title would be *termes ‘w arden’, but instead we only h a v e palotas ‘palace w arden’ and udvarnok ‘court w arden’ in Hungarian, both o f which are o f Slavic origin. Two other words o f the com m on H ungarian vocabulary belong to the category o f words denoting nobiliary titles. The Hungarian bo ‘plentiful’ undoubtedly derives from the Turkic beg ‘chieftain’, ‘b ey ’; the sem antic change it under went was sim ilar to that o f the title doge which ended up as the Hungarian diis ‘abundant’. The H ungarian ur ‘lord’ derives from the Turkic uri ‘youth’ whose meaning o f a ‘noble, young gentlem an’ was only taken overby the H ungarians at the time o f H ungarian-Turkic interaction. The Hungarian language also tagged a suffix onto this ur, o f Turkic origin, to create uruszag > orszag ‘country’, ‘realm ’ as in M agyarorszag ‘H ungary’. The m odel o f sem antic history is like w hat happened with the Hungarian word m alac ‘p iglet’. Only its m eaning ‘y oung’ exists in the Slavic languages, and only in Hungarian did it develop its sem antics from ‘young anim al’ to ‘young p ig ’. The title urum ‘heir to the throne’ (which strongly resem bles the Hungarian uram ‘m y lord’) in Cinnamus, a 12th-century Byzantine author, is probably a suffixed Turkic derivation o f the sam e Turkic word, and not a Hungarian variant with a Hungarian personal suffix. The origin o f the Hungarian word in ‘p o o r’ (which has only survived in inseg ‘penury’) is yet unknown. The suggested etym ology
356
From the Urals to the Carpathian Basin
for the word jobbagy ‘serf’ is not very satisfactory. Supposed to have origi nated from the word jo ‘good \jo b b ‘b etter’ sim ilarly to the G reek aristokratia or the Latin optimates, in the M iddle A ges the w ord actually used to denote the m em bers o f the leading social stratum and ‘villein’ only from the 16th century. The suggested addition o f the dim inutive suffix -gy, however, does not sound very convincing. The Hungarian word hadnagy (today ‘second lieutenant’) cropped up in the sources surprisingly late (1213), although it m ight reflect the A ncient Hungarian w ord denoting the ‘head o f the clan ’, due to the fact that the Hungarian had ‘arm y’ had an earlier m eaning o f ‘clan’, ‘extended fam ily’, and nagy ‘great’ also m eant ‘head o f ’. The leading stratum o f the conquering M agyars was backed by groups o f armed, free people. Their m ain task was participating in raids, but naturally they w ere also in charge o f guarding the anim als, and o f nom adic agriculture, too. In the 1Oth century, al-Istahri w rites o f the K hazars that their pagan groups would perm it trading in their own and fellow kinsm en’s children as slaves, w hile the Jewish, the M uslim and the C hristian groups forbade this custom. The Jayhani tradition claim s that the Etelkoz M agyars w ould sell their cap tured Slavic slaves to the Byzantines at Kerch. This custom m ust have lived on, m any decades after the Conquest. Actual work, especially agricultural, was alm ost certainly perform ed by servants. The status o f servants in this rather “sourceless” period is very difficult to establish, and in any event status was subject to the current circum stances. Ibn H aukal w rites that a K hazar m erchant adopted his slave w ho soon assum ed an im portant role in the business. W hen the m erchant died, the adopted son laid claim to the inheritance on the grounds o f being the m erchant’s son. However, the deceased’s real son quite unexpectedly appeared from lands afar, and stood up for his rights. The judgem ent ruled that the slave should pass to the m erchant’s real son. A lbeit anecdotal and sounding like a legal parable, the story typifies the instance o f adopted slaves integrating into the family, and the Hungarian w ords csalad ‘fam ily’ and cseled ‘servant’, both deriving from the same Slavic chelyad, provide a fine correspondence. The M agyar clan was a larger unit than an “extended” family. A ny clan w ould consist o f anything from 5 to 20 related extended fam ilies. The clan patriarchs played a significant role in the clan cult, in settling disputes and in the division o f land, booty and work. Clans were founded on the principle o f blood descent, but there was m uch cross-breading due to the adoption o f w om en o f other clans, or o f outsiders. A doption, however, w as usually limited to m en w ho were needed for labour in anim al-rearing. O utsiders m ight be the orphaned m em bers o f other clans, and also non-M agyars. A dm ittance into a clan had its own ritual, by w hich the adopted person swore to accept the clan cult.
The M agyars in the Carpathian Basin
357
The principle o f patrilineal descent operated in the clans. Turkic and M ongolian nom ads held that one’s bones were patrilineal and blood m atrilineal. A ccordingly, the words denoting ‘bone’ in the Turkic languages more often than not also m ean ‘clan’. Ibn Haukal reports on the above belief. An inheritance dispute was settled thus: to decide who the real son o f a deceased wealthy m erchant was, they exhum ed one o f his bones. The false and the real son were then required to drip blood on the bone. The boy whose blood trickled o ff was lying, while the one whose blood the bone absorbed was truthful. The sm aller unit constituting the clan, the “extended” family, was founded on the cohabitation o f three or four generations. The size o f this unit was incidental, but also largely depended on the num ber o f anim als they could rear together, and on the conditions o f the w inter dwellings. Such was the stability o f the clan system that continued to operate even after the tribal confederation had disintegrated. The system underwent changes, but nevertheless resem bled the original set-up. Close ethnographic analogies suggest that the clans formed smaller units w hich even nowadays are called had in som e Hungarian regions. The position o f craftsm en was special. M ost o f the ironm asters, im plem ent and w eapon m akers m ust have entered the C arpathian Basin together with the conquering M agyars. As regards designations, only the Hungarian nam es acs ‘carpenter’ and szucs ‘furrier’ are o f Turkic origin. The form er was a m aker o f sators ‘ten ts’ and karos ‘stakes’, while the latter used a gyuszii ‘thim ble’. These words are also Turkic loan words. The furrier was an expert in working with leather. The local Slavs gave blacksm iths (kovdcs), potters (gerencser), weavers (takacs) and turners (esztergalyos), and the word konyar, o f Slavic origin, suggests a horse-related occupation, while teszar worked with wood (see Acsteszer). The Hungarian language borrow ed these w ords from local Slavic craftsm en. This does not m ean that these crafts were unknow n to the conquering M agyars. A vasvero, literally, ‘iron b eater’, had a different tech nique from the blacksm ith; the csikos ‘horseherd’ and the ju h a sz ‘shepherd’ (all nam es o f Finno-Ugrian origin) worked in different anim al-rearing system s from the konyar and the pasztor (names o f Slavic origin). Thus, very often the old M agyar craft expressions w ere preserved, alongside a brand new craft lingo. The Hungarian otvos ‘sm ith’ was an expert in iron forging (otteveny), yet he probably encountered new techniques and tools. The archaeological excavations o f iron furnaces from around the Conquest period have revealed that the traditions o f iron forging in the Carpathian Basin were unbroken; m oreover it is likely that pre-C onquest ironsm ithing in Transdanubia continued throughout the 10th century, and also appeared in the Upper Tisza region as from the 11th century. Scholars hold that there was no equivalent in H ungary to the types o f furnaces the K hazars used, hence it is
358
From the Urals to the Carpathian Basin
likely that the conquering M agyars did not bring the technology and the ironm asters along w ith them. The w eaponry and accoutrem ent o f the conquering M agyars w ere m ade by craftsm en. These are readily reconstructible from objects unearthed from graves dating back to that period. The m ost im portant piece o f equipm ent was the 1.1 m etre-long, recurved, or reflex, com posite bow. The flexibility o f its w ooden staves was enhanced by tw isted bundles o f stag sinew, or plates o f horn. The parts exposed to the heaviest wear, that is the nocks, w here the bowstring was attached to the limbs, and the nocking point, w here the arrow was placed for shooting, were protected by plates o f bone. Some parts were even covered w ith leather. Technical reconstruction has dem onstrated that the
Figure 69 R econstructed Conquest-period m an’s costum e (after G yula Laszlo)
The M agyars in the Carpathian Basin
359
Figure 70 Reconstructed harness
M agyar’s bows could fire arrows fairly accurately to a distance o f approxi mately 6 0 -7 0 metres. They kept their bows in quivers. The type o f quiver depended on w hether the bow was relaxed or braced. Arrows w ere kept in arrow quivers. The tips o f arrows with w ooden shafts were m ade o f iron. The variously shaped and w eighted arrow heads w ere designed for different purposes. There were so-called “w histling” arrows (which had a sm all bore running through the head to produce a whistling sound when they w ere fired), and incendiary arrows, used in night-tim e com bat, or to set w ooden buildings on fire. The M agyars’ spears included close-com bat spears and signalling spears which served to direct m ilitary actions. W here visibility w ould permit, the latter were pressed into the ground and w orked like “sign boards” . The equipment also consisted o f a sabre which w as double-edged at the point with a fuller running down its entire length. Slightly arched, the blade w as approxi mately 0.7 -0 .9 m etres long. The hilt was m ade o f w ood and w as covered with leather, and the cross-guard would often have an ergonom ic design to allow for a better grip. W hen out o f use, the sabre was kept in a scabbard whose lockets were finely om ated. We have much know ledge (see p. 119) about the famous sword o f Kiev w hich attests to a M agyar influence (see Plate I). Other weapons included the com bat axe which was m ore often than not two-handed. No mace has yet been found.
360
From die Urals to die C arpadiian Basin
M ilitary clothing included leather (rather than iron) arm our under which the M agyar warriors wore linen or hem p-cloth underwear, a thin leather shirt and trousers which could be tucked into the boots. To cover all, they wore a gown which reached the knees. The Hungarian word barsony ‘v elv et’, of Turkic origin, might have referred to the m aterial o f this gown. On top o f this, the M agyars wore a buckled belt om ated w ith various metal studs and mounts. A w arrior would w ear his sabre and arrow -quiver on the left-hand side, and the bow -quiver and a chased or em bossed sabretache, in w hich he carried his flint and steel, on the right. The m any archaeological finds o f C onquest-period saddles enable us to reconstruct this equipm ent well (see Plate XVI), and we also have a highly detailed description o f a 10th-century Turkic saddle. We have been able to reconstruct the M agyars’ harness and their usually pear-shaped stirrup, too (see Figure 70). M erchants are known to have accom panied the M agyars to the Carpathian Basin. M ost o f the coins found in Conquest-period graves m ust have come from the booty obtained in raids, from w ar indem nities, taxation and ransoms. But we have also come across currencies attesting to long-distance trade, such as Arab dirhems. With objects it is very difficult to ascertain w hether they were brought along by the M agyars, were robbed in the course o f incursions, or w hether they entered the country via trade channels. W ith objects bearing the traits o f craftsm en o f the neighbouring areas (M oravian ironwork, for instance), it is nigh-im possible to establish w hether they w ere purchased or looted, or w hether the M agyars brought along with them the craftsm an who, having settled in Hungary, continued his craft in the m anner and style he had learnt back at home. For instance, the hilt o f the fam ous Vienna sabre had fragm ents o f whale-skin on it, w hich were shown to have originated from a species indigenous to the Indian Ocean. Also, the raw m aterial o f many m other-of-pearl ornam ents came from the A driatic Sea, and m any other objects, too, travelled to H ungary from afar.
4. ECONOMY The debate regarding the M agyars’ econom y flares up from tim e to time. Opinions tend to polarise according to w hether the M agyars are seen as having been either pastoral nom ads, or sem i-nom adic. I shall attem pt to draw the picture which the sources, as well as Turkic and M ongolian analogies, suggest. The M agyars’ anim al husbandry is well characterised by the ratios of livestock. In the M ongolian Republic, which has m any m ore geobotanical regions, the country-wide distribution o f livestock in 1967 was as follows:
The M agyars in the Carpathian Basin
Figure 71 Make-up of livestock in Mongolia
361
IS □ □
362
From the Urals to the Carpathian Basin
sheep 23% , goat 9%, cattle 24%, horse 30% , and cam el 14%. The differences o f these ratios will greatly vary, o f course: naturally, the ratio o f cam els is significantly sm aller in the colder northern areas. But, m ore im portantly, the distribution o f horses in the steppe areas was as high as 45% (for instance, in the area o f Ulanbator). The ratios were significantly different in the grovy steppe o f West M ongolia where the ratio o f sheep was about 30% , horse 25%, and cattle 25%. The M agyar pastoralists, too, counted their livestock by m eans o f the system o f standard anim als w here 1 standard anim al (b o d in M ongolian) = 1 cow or 1 horse or 7 sheep or 10 goats. A camel was counted as one and a h alf standard anim als. The distribution o f anim als, in term s o f standard anim als, was, in the very best grazing areas, over 12 per square kilometre. M ongolia’s high density areas surprisingly coincide with the centres o f the A ncient Turkic and M ongolian em pires, the valleys o f the Rivers O rkhon and Selenga. The density in the m ore sheltered, grovy steppe areas w as 8-12 standard anim als to a square kilometre, and 4 -8 in the open steppe regions. W hat this means, in effect, is that even under m odem M ongolian circum stances, an area o f ten square kilom etres in the m ost sheltered grazing places, was able to provide the m eans o f subsistence for 120 standard anim als. Based on the bone finds unearthed in m ediaeval Eastern European archae ological sites, the distribution o f livestock was as follows. I. foresty regions: cattle 31%, sheep and goat 20%, pig 39% (D niester and K am a region 36%) and horse 10%; II. grovy steppe region: cattle 39% , sheep and goat 24% , pig 26% , and horse only 11%; 111. open steppe region: cattle 23% , sheep and goat 65% (w ith sheep reaching 54% in places), pig 1-2% , and horse 11.5%. The ch ief difference betw een the Eastern European archaeological data and m od em M ongolian figures is in the ratio o f horses. Even in the northernm ost hilly regions o f M ongolia, the ratio ofhorses is never less than 20% . O bviously this is rooted in history; however, archaeologically speaking, horses are more problem atic than the other dom estic anim als, due to burial custom s and the system o f beliefs. Conquest-period M agyar sites show an interesting distribution. In some places the horse-stock peaks at 25-35% , in others at only 8-10% . The cattlestock is significant: 25-35% , with sheep surprisingly low, a m ere 16%. There are sites w here the ratio o f pigs reached 15-20% o f the total dom estic animals. These proportions are not necessarily a reflection o f the actual state at the time, on account o f the fact that burial custom s greatly influenced the variety o f bones that survived. Assessing the highly dispersed figures is a difficult task, yet there is a strong hint o f the presence o f two structures o f anim al husbandry, one o f w hich places the em phasis on breeding horses, w hile the other on rearing pigs, w ith a sm aller horse-stock. This w ould suggest that the M agyars’
The M agyars in the Carpathian Basin
363
econom y was indeed mixed, with a concurrence o f an arable and anim al-rear ing econom y and a nom adic system. However, in larger areas even the nom adic economy included a unique type o f agriculture alongside animal husbandry (described above, see pp. 142— 145). It cannot be m ere chance that so m any Hungarian pig-rearing term s are o f Turkic origin (see p. 110), and that the proportion o f pigs in Conquest-period archaeological sites is surprisingly high. The M agyars m ust have brought this branch o f anim al husbandry along with them even if they hardly drove any pigs across the passes o f the Carpathian M ountains. The M agyars did not take the pigs with them on their regular annual alternating pasturing m igrations; we can only guess, therefore, that they kept the herds near the w inter dwellings, drove them into oak forests to feed on acorns, and that substantial num bers o f the fam ily stayed behind to guard them and also to cultivate the land. Yet the M agyars m ust have included groups which were involved in rearing sheep to a greater extent than pigs. This m ixed anim al-rearing culture m ust have m eant a transition from nom adism to settling down. Conquest-period M agyar agriculture can be reconstructed with the help o f linguistic and archaeological sources. W ithout doubt, the M agyars had known and practised tillage farm ing before they entered the Carpathian Basin, and they are known to have yoked oxen to their ploughs. Their light, wooden ploughs w ere portable and could be loaded onto wagons; they m ight have been brought with the M agyars from the Etelkoz. The M agyars had com e to know Slav-type ploughs even before the Conquest, and it was these very same models that they found on arrival in the Carpathian Basin. For reaping wheat they used scythes (sarlo, o f Turkic origin), and ground (orol o f Turkic origin) the harvest w ith hand-mills. The fact that the Hungarian w ord arok ‘ditch’ features even in the earliest w ritten Hungarian text m akes it alm ost certain that the ancient M agyars had an irrigation system o f som e description. The spade was another im portant agricultural tool. We have already discussed the M agyars’ viniculture which they had brought along with them (see p. 141). O f course, only those groups o f M agyars who pursued an arable and anim al-rearing econom y produced wine also. On entering the Carpathian Basin, the M agyars found a significantly dif ferent Avar-Slav agriculture which they com pletely adopted in a 100-150 years. Careful consideration o f the excavation data leads us to believe that the M agyar econom y was m ultiply differentiated. Am ong the nom adic system o f alternating pastural and arable and anim al-rearing econom ies a w ide array o f transitional forms existed which further developed after the Conquest.
364
From [he Urals to the Carpathian Basin
This is how the Jayhani tradition writes o f the M ajgars: “The land o f the M ajgars abounds in w oods and waters, the soil is m oist, and it has plenty o f arable land.” This held true for both the Etelkoz and the C arpathian Basin.
5. RELIGION, LORE, CULTURE The M agyars brought to the C arpathian Basin their traditions w hich the K hazar Em pire and Turkic, Slavic and B yzantine neighbours had shaped. It w ould be convenient to infer from the etym ology o f the H ungarian words taltos ‘sham an’ and boszorkany ‘w itch’ that the first was a representative o f the Finno-U grian system o f beliefs, and the second o f the Turkic. This w ould be, m ethodologically speaking, improper, and this is ju st the kind o f approach that should be avoided. It w ould be like claim ing that the m eaning o f the texts printed on a placard and a poster respectively should be differentiated on the grounds that ‘placard’ derived from the Old French placquart and ‘p o ster’ from the Latin postis, w hich is evidently an obscure argum ent. Also, the function o f the above M agyar characters o f folk beliefs underw ent substantial change after the Conquest. The docum ents o f m ediaeval w itchcraft trials pro vide am ple evidence o f this. Yet the M agyars m ust have adopted the word boszorkany ‘w itch’ from Turkic for good reason. It denoted som ething that had not existed before; som ething whose Turkic m eaning m ust have been im portant. The word entered the com m on H ungarian w ord-stock, and as sum ed new m eanings. A good exam ple is the history o f the H ungarian word boles ‘w ise m an ’, o f Turkic origin. The root, biigii, m eans ‘know ledge’ and especially the ‘knowledge o f supernatural things’. H aving entered the M on golian language, the w ord took the m eaning o f ‘m ale sham an’. Used attribu tively, the w ord w ould m ean ‘ho ly ’, as in Turkic translations o f old Chinese texts w here w e can read about the Biigii Kungfuci ‘Holy C onfucius’. This root word features in m any personal nam es, too. The derivation, biigiichi, m eant ‘w izard ’, ‘m edicine-m an’ in m any Turkic languages, and in M ongolian ‘sha m an’. On entering the Hungarian language, the Turkic biigiichi becam e, by process o f system atic change, bucs > bocs > boles ‘w ise’. But the basic word biigii also entered the Hungarian, surviving only in the verbs biivol ‘to b e w itch’, elbiivol ‘to charm ’ and in the noun com position biibaj ‘charm ’, ‘attractiveness’. The boles ‘wise m an ’ was a person w ho knew and practised bit ‘m agic’. These words took different courses o f developm ent, but their etym ologies shed light on the fact that they are actually related. The person w ho originally had supernatural know ledge sim ply becam e a person who had know ledge and ‘w isdom ’.
The M agyars in the Carpathian Basin
365
The ja va s ‘w itch-doctor’ and the orvos ‘m edicine m an, today: physician, doctor’ w ere also m em bers o f the ancient M agyars’ system o f beliefs. Javas com es from the ancient Finno-Ugrian word-stock, and orvos is, perhaps, o f Iranian origin (as for instance the Sakha aruva, arva ‘m edicine’ < ‘p lan t’, Indo-European *ar- ‘to sow ’). The Turkic arva- m eant ‘to practise m agic, w izardry’, arvish ‘w izardry’ and arvishchi ‘w izard’, but the Finnish language has sim ilar words, like arpa ‘oracle stick’. The Hungarian baj ‘grace’, w hich originally m eant ‘m agical bond’, is o f Slavic origin (and not Turkic), but its m eaning in the Slav language changed under O ld Turkic influence. The word varazs ‘charm ’ is also o f Slavic origin. Javas and orvos, then, had co-existing functions, both w ere healing folk, the latter perhaps the m edicine m an know n from other cultures. There is no equivalent in present-day H ungarian to the ‘rainm aker’. It is likely, however, that along with the taltos, the boszorkany, the boles, the ja v a s and the orvos, the rainm aker also featured am ong the conquering M agyars. Chinese sources talk about the Turkic nom ads whose tactics included the production o f a storm or rain. Various other sources confirm this. For instance, a M uslim traveller, a certain Tainim ibn Bahr, wrote in 821 about the Uighurs: “Am ong the m iracles o f the land o f the Turks are pebbles which they possess, and with w hich they can create rain, snow and cold at will. The history o f these pebbles in their possession is well known and w idespread, and no Turk has any doubts about them. And these pebbles are prim arily in the possession o f the king o f the Uighurs, and their other rulers do not possess any.” Gardizi, too, speaks about the rainm aking stone, and the story he relates is about the Oghuz, the Kharlukhs and the Khazars fighting for the possession o f such a stone. Yet, another source claim s that the Oghuz and the Turks clashed for the same reason. The m agicians o f the “Caucasus Huns” were able to produce rain and storm s from the shelter o f their oak tree consecrated to the deities. The nam e o f the rainm aker in Turkic was yatchi or ja tch i. The root w ord, у at, according to Khashghari, m eant ‘magic m ade w ith stones to create rain and tem pest’. In all probability these rainm aking stones in the graves o f the Conquest-period M agyars will be identified shortly. These different forms o f beliefs were not singularly characteristic o f one people only, local versions occurred practically everywhere. Hence w e cannot venture further than establishing that the archetypical forms o f folk belief existed am ong the M agyars, too. These included the veneration o f holy shrines and trees. Such elem ents are to be found in the system o f beliefs o f the Volga Finno-Ugrian and Turkic peoples (for instance the C huvash K erem ets), as well as in the w ritten sources o f the Khazars. Holy objects also included protective am ulets, m any o f w hich bore inscrip tions. The Hungarian word betu ‘letter (sym bol)’ exists in C huvash in the form
366
From гhe Urals to the Carpathian Basin
o f petii, whose second m eaning is the equivalent o f ‘am ulet’ which used to m ean ‘protective w riting’. A pendant with a G reek-letter C hristian prayer inscribed on it came to light from the C onquest-period grave at Piliny. The M agyars’ graves contained a num ber o f other protective objects, too, such as a small saw, a set o f b ea r’s fang and animal bones. Sim ilar protective objects w ere found in excavations in the form er K hazar Empire. M ovses Dashuranci speaks o f the “Caucasus H uns” having as a holy tree the oak which they consecrated to the god o f the heavens. Therefore, when these Caucasus Huns were converted to Christianity, the old sym bol was also converted and a cross was carved out o f the holy oak tree. The tree consecrated to the god o f the heavens, Tengri, is, o f course identical w ith the tree o f worlds w hich leads to the shaman other world. K hashghari was a follower o f the M uslim faith, but nevertheless gives a detailed account o f the pagan Turkic world in the typical m anner o f his time. In his report o f 1074, he calls Allah by his Turkic nam e, Tengri (as opposed to a near contem poraneous source, the K hutadghu bilig). H aving established this, and cited some sayings and poem s related to Tengri, he continues thus: “the infidels— m ay God destroy them !— call the sky Tengri, also anything that is im posing in their eyes they call Tengri, such as a great m ountain or tree, and they bow down to such things. [...] We take refuge from error in G od” (translated by D ankoff and Kelly). This account reveals a syncretism o f sham anism and Tengrism. The world in the sham an’s cosm ic concept is tripartite; it rests on the threesom e o f the other world, this w orld and the nether world. The tree o f worlds connects these worlds. Depictions o f this tree o f worlds have come to light from Conquestperiod excavations, on for instance the w om an’s harness found in a w om an’s grave at Szakony. Other symbols o f the cosm ic w orld concept, such as the sun and the moon, can also be found on archaeological finds from the Conquest period. The m em ory o f holy trees and holy groves are preserved by m any place nam es. Anonym us held that the silva Igfon, that is, the Igyfon forest, was located betw een the M aros and the Szamos. The earliest H ungarian written laws m ention sacrifices held near springs, stones and fountains. The igy constituent in the nam e Igyfon is identical w ith the egy ‘h o ly ’ (a w ord o f Turkic origin) in the Hungarian word egyhaz ‘church, holy h ouse’, and the second constituent fo n derives from the old Hungarian word m eaning ‘densely w oven’ (i.e. Igyfon, a holy, dense forest). The word m eaning ‘h o ly ’ also appears in the Hungarian word iinnep ‘festival’, ‘holiday’ (< id nap ‘holy d ay’). The ethnographic sources, discussed in C hapter II. 6 , include Ibn F adlan’s description o f the gods o f the Bashkirs. They had the sun as one, but the god o f the heavens was even greater (see p. 148). This account o f Ibn Fadlan is
The M agyars in the Carpathian Basin
367
applied to the Khazars by the Persian Am in Ahm ed Razi, a 14th-century compiler. M ention has been made o f the adoration o f a sun-god (see pp. 151, 191). As yet, we have been unable to track down the passage o f the Hettite nam e o f the sun-god Ishtenu, from its Hatti origin, via the Caucasus M ountains and the Khazars, to the pre-Conquest-period M agyars (isten in m odern Hungarian stands for ‘god’). However, as we have seen (pp. 187-191), a num ber o f sim ilar itinerant cultural terms, such as the cultic objects o f balta ‘axe’ and szekerce ‘adze’, w andered to the M agyars from afar. It has been the subject o f m uch debate w hether the real or sym bolic w ound ing o f the cranium — so-called trepanation— actually served as a remedy, or whether it was related to beliefs. Am ong the data cited in support o f this is the Hungarian word agyafurt, now m eaning ‘cunning’, but derived from agy ‘brain’, ‘skull’ + f u r ‘to drill, to bore’. Only a couple o f personal nam es (such as Lehel and Lei; related to the words lehel ‘breathe’ and le le k ‘soul’) and later sources offer some insight into the Arpad-era M agyars’ concept o f the soul. Still, we have good reason to believe that their system o f beliefs included the “life-soul” which ceased to exist after death, as well as the soul w hich departed at death, and was reincarnated in the form o f a bird, for instance. Im personated nature, the veneration o f rem arkable natural phenom ena, healing persons who also practise hom eopathic m agic, their w hole system o f beliefs, and the tripartite division o f the world are some o f the elem ents o f the M agyars’ religion. The great w orld religions— prim arily Christianity— w ere built upon and com bined with these. As we have seen (p. 233), the K hazar Em pire, too, had Christian groups. We know o f K hazar princesses, m arried to Byzantine em perors and Caucasian rulers, who were christened; and also o f Byzantine and A rm enian princesses in the K hazar court. Christian com m unities existed in the Crim ea and the Caucasus, too. Gothic and Byzantine C hristianity flourished alongside the Khazars in the Crimea. It is very hard to believe that Cyril and M ethodius were the only C hristians to have m et the M agyars prior to the Conquest. Although not m entioned in the sources, the M agyars living in the Etelkoz m ust have sent several envoys to the Eiyzantine court. We do not know w hether any pre-Conquest princely M agyar youths o f the Etelkoz w ere educated in the Byzantine court— Khuvrat, the ruler o f the Bulghars in any case, is know n to have been. Nevertheless, based on authentic evidence, we can fairly safely establish that the pre-C onquest Arpadian M agyars had heard about C hristian ity even before they crossed the Carpathians. M oreover, it is possible that certain M agyar groups were actually converted. Going by present-day H un garian term s lik egyasz ‘bereavem ent’, erdem ‘m erit’, torveny ‘law ’, bocsanat ‘pardon’, bucsu ‘indulgence’, o r o k 1eternal’, bun ‘sin \ g y o n ‘confess’, igy/egy
368
From the llra ls to the Carpathian Basin
‘ho ly ’, w hich are o f Turkic origin, our suspicions are further confirm ed. It has been argued that these w ords w ere borrow ed after the C onquest from the local Bulghar-Turks who lived to the south o f the R iver M aros, due to the m ediation o f the Transylvanian gyula s daughter Sarolt (Saint Stephen’s m other) and her environm ent. This view is difficult to uphold. It is certain, however, that shortly after the Conquest the local lords o f the southern regions began a series o f visits to Byzantium ; Gyula and Bulcsu w ere baptised; B yzantium sent a m issionary bishop to Hungary, nam ely Hierotheus, w ho arrived am ong the M agyars around 952 as a m onk, and was shortly ordained B ishop o f Turkey [Hungary]. Although the Danube B ulghars had been baptised around 865, the language o f the liturgy was G reek and later Slavic. There is one Hungarian w ord w hich com es from this Bulghar-Turkic ecclesiastic term inology: kep ‘im age’. The original Turkic m eaning was ‘shape’, ‘form ’, ‘im age’, ‘likeness’. This existed am ong the D anube Bulghars w here it m eant ‘pagan id o l’, ‘pagan im age’. Tagged w ith a Slavic affix, its derivation kapiste, w as the nam e o f the pagan sacrificial shrine, which attests to the fact that the w ord m ust have had a pagan m eaning related to religious faith. The earliest occurrence o f the Hungarian derivation is in Christian texts (from ca. 1315: “szent oltarun kuner kepeben” [in form o f bread on the holy altar]). The other part o f the Christian vocabulary o f the M agyars was o f Slavic origin. Words including kereszt ‘cross’, karacsony ‘C hristm as’, szent ‘saint’, pap ‘p riest’, barat ‘m onk \ p o k o l ‘hell’, w ere borrow ed from the G reek-rite Slavs, as opposed to apaca ‘n u n ’, apat ‘abbot’ w hich entered the Hungarian language from the Latin-rite Slavs. To this w ord-stock was added later the standard Latin terminology, as well as a num ber o f Latin term s via Germ an and Italian m ediation. Traces o f Christianity are to be found in the archaeological finds o f the C onquest-period M agyars. The sabertache plate o f B ezded w hich features an intertw ining cross and a tree o f life is a “pet” reference o f scholars. From the 10 th century crosses m ade o f cast bronze or sheet m etal began to appear in the M agyars’ graves. We have m uch know ledge o f the burial custom s o f the M agyars. The deceased was laid on his or her back, with the legs facing east, according to the Turk tradition (see p. 138). The hands w ere rested on either side o f the body, or w ere placed on the pelvis (see Figures 24 and 25 on p. 137 and Plate II/3). A sim ple coffin is apparent (the Hungarian w ord koporso ‘coffin’, o f Turkic origin, had the m eaning ‘casebox’). The bottom o f it was lined with m oss or cloth. Covered with leather, they placed the bones o f the deceased’s horse at the foot o f the grave. O ccasionally they w ould stu ff the flayed horse-skin with hay or foliage.
The M agyars in the Carpathian Basin
369
Around 681, the Albanian bishop Israyel notes o f the so-called “Caucasus Huns” (actually the Khazars) that they sacrificed a horse to the god o f the heavens, Tengri; and that they burned the horse. It well m ight be that the rem ains o f the burned anim al and its skin were interred with the deceased to as a synechdochic sym bol (pars pro toto). In spring 900, the Bishop o f Salzburg, Theotmar, w rote a letter to Pope John IX, in reply to the accusations o f the M oravian clergy. The M oravian bishops w ere seeking to gain independence for their own M oravian bishopric, and to break o ff from Salzburg. Theotm ar quotes from the com plaining letter the M oravians had w ritten to the Pope: “As it happens, the said Slavs accuse us o f offending the Catholic faith w ith the M agyars, o f having pronounced oaths over dogs and w olves and other unholy pagan creatures, and o f m aking peace with them; also o f paying them m oney to go to Italy: if w e w ere to discuss our matters in your holy presence, then, before the alm ighty Lord who knows all things even before they happen, and before your apostolic holiness, their falseness w ould com e to light and our innocence gain proof. B ecause they [the M agyars] have always threatened Christians in lands afar, have devastated them, and yet we only gave them w orthless m oney and linen clothes to pacify their w ildness to keep aw ay their raids [...] M any tim es for m any years have they [the M oravians] com m itted the sin w hich they now falsely accuse us with. They took unto their people large num bers o f the M agyars, and shaved their own false C hristians’ heads in the fashion o f these M agyars, and sent them [the M agyars] against our Christian peoples.” Evidently, the text refers to the campaigns o f 881 and 894 when, in league with Svatopluk, the M agyars launched an attack against the Franks (see pp. 331-332), and the cam paign o f 892, fought in alliance with Am ulf. It m atters little w hether the M agyars swore by dogs or wolves, the description nevertheless, gives a good idea about the customs o f the A rpadian M agyars. The fact that the conquering M agyars were fam iliar with C hristianity does not m ean that a significant a num ber o f M agyar groups, or the leading strata actually practised it. There is not a shred o f support to confirm this; m oreover, there is one thing that contradicts the idea. Conversion to any o f the world religions always involves the appearance o f literacy; the Word m ust be com m itted to writing. W riting am ong the pre-C onquest M agyars could have been anything from H ebrew to Greek. We cannot dism iss the suggestion that there w ere people am ong the M agyars who knew one or m ore o f these scripts, and m any o f them m ight even have acquired some Latin during the course o f the raids on the west. Or, the M agyars m ight have had clerk prisoners or servants. But writing did not spread. A runiform script, however, w as probably fairly w idely used. As will be pointed out in C hapter XVI, the Szekely runiform script was w ithout doubt related to the other Eastern European
370
From гhe Urals со the Carpathian Basin
runiform scripts, even if we are not sure how. Also, it is clear that runiform letters o f Slavic origin entered the Szekely runiform script, w hich could hardly have happened before the Conquest. Thus, the M agyars entered the Carpathian Basin with a sim ple runiform writing w hich cannot, perhaps, have differed greatly from the Szarvas-N agyszentm iklos writing.
NOTES The institution o f the sacral and dual kingship, as that o f the Hungarians, has been much debated recently by Gyorffy, C zegledy and Krist6 (see G yorffy 1955, 1973, 1984, 1993a, 1993b; C zegledy 1966, 1974,1975; Krist6 1980, pp. 2 0 7 - 2 2 8 ,1993a, 1993b, 1996; as w ell as Bakay 1978; DUmmerth 1971,1977; Негёпу i 1982; Makk 1985; T 6 th 1988; and Uhrman 1 9 8 7 -1 9 8 8 ). I have also dealt with the problem in m y inaugural lecture as an ordinary m ember at the Hungarian Academ y o f Sciences, delivered on 12th February 1996. The text is now in print and w ill be published in 1999. The idea came from C zegledy w ho referring to a work o f Roheim written in 1917 elaborated a theory. Czeglddy projected the system o f the Khazars on the Magyar structure and under his authority the existence o f the sacral kingship am ong the Hungarians o f the Conquest gradually becam e generally accepted. However, GyOrffy considered Kursan to have been the sacral king, w hile according to Kristo, Kursan should be regarded as g yula, that is, the sacral king must have been A lm os and his son Arpad. The fact that there never existed a sacral dual kingship among the Magyars was recently also argued by K eszi (1995) with a German summary. For Ibn Rusta’s text, published by Zim onyi, see Krist6 (1995). I have quoted Gardizi from Martinez (1982). A comparative description o f the sacral kingship with untenable conclusions is Waida (1976). The H etm agyar ‘Seven M agyars’ issue has been discussed in Nem eth (1991, pp. 18 1 -2 1 3 ) which, by now, has a great many untenable arguments. For the group o f Turkic tribes with “numerical” names, see Czeglddy (1963). S. L. T 6 th (1995) is a very recent account o f the location o f the M agyars between 895 and 899, com plete with an overview o f past opinions. Porphyrogenitus’s sentence about the bilingualism o f the Khavars and the M agyars is analysed by M oravcsik in Jenkins (1962, p. 150; see also, M oravcsik 1984, p. 46, note 33). Suggestions for the numbers o f the conquering Magyars vary greatly. For more recent opinions see K ovacsics (1995b), G y 6 rffy (1995), and Kristo (1995a). N o truly reliable source exists in support o f this issue. The size o f the Khazar population is discussed in Ludwig (1982, pp. 2 1 1 -2 2 3 ). For the religion o f the Khazars and its sources consult Dunlop (1954) and Ludwig (1982). The distribution o f livestock in the individual M ongolian regions is excellently illustrated by a map in BadamZav (1966), an adaptation o f which I have included in this book (see Figure 71 on p. 361). For M ongolian animal rearing, see Rona-Tas (1961a), its Russian edition ap peared in 1964, there exist also a Polish and a Japanese translation o f the book. For the livestock figures o f South Russia and the Conquest period, see J in os M atolcsi’s archaeozoological chapter in Hajdii-Krist 6 -R ona-Tas (1976, vol. 1/1, pp. 164 -2 0 1 ). Krist6 , w ho claim s that the M agyars were pure nomads, does not discuss the make-up o f livestock o f the conquering M agyars (Kristo 1995b); he is mainly concerned with reconstructing the M agyars’ way o f life
The M agyars in the Carpathian Basin
371
via the written sources. Krist6 does not take into consideration the fact that the Arpadian Magyars brought several types of economic systems with them, and insists on the Magyars being nomadic pastoralists. It must be said, however, that as regards nomadism, his viewpoint has significantly changed in comparison to his earlier works. I have discussed the old Avar-Magyar title termes in connection with the name Termecsii in R6na-Tas (1995c and 1996g), theorigins ofthe Hungarian word ur ‘lord’ in R6na-Tas (1994a, 1995b). The comments of Kara (1995) on the etymology of the Hungarian ur attest to scanty knowledge regarding the natureofwordborrowing, andreflect thepractice of forming linguistic assumptions with only the help of dictionaries. The Hungarian word for ‘piglet’, malac, has the meaning ‘piglet, pigling’ in none of the Slavic languages, where it means ‘young’, and the semantic narrowing occurred in Hungarian. The same happened with the Turkic word uri. Benko (1998) questioned my suggested etymology of Termachu. He did not take into account that the second vowel, which is a closed e in Hungarian, was not a closed one in the original language. The present Hungarian structure of the word is a relatively late development. Thus the Greek transcription of the second vowel by alpha is possible. See more about old Magyar traditions of iron forging in GomOri (1994), with further bibliography. For the particulars of Magyar agriculture, see Balassa (1994). Dienes (1972a, 1972b) and Balint (1976) discuss the relevant archaeological finds, Fodor (1994, p. 54) the sword of Kiev. Molnar (1993, English version: 1994) writes about the rainmaker. The finds of Piliny and other Conquest-period objects are presented in Dienes (1972a, p. 48). Data about the religious world of the “Caucasus Huns” were collected by Ludwig (1982, pp. 305-332). The best Khashghari edition is by Dankoff-Kelly (1982-1985). For the ornaments on the woman’s harness of Szakony, see Dienes (1972a, p. 49). Read about symbolic or real trepanation in Ery (1994, p. 222), and about the Hungarian word agyafurt ‘cunning, literary: bored brain’ and its literature in Pais (1975, pp. 7-10). Consult Dienes (1978), Mesterhazy (1994) for an interpre tation of the archaeological matter of the Conquest period with special emphasis on the reconstruction of the system of beliefs. These works, however, cannot always avoid the pitfall ofover-interpreting the data. On theoeuvre of Dienes see the obituary written by Bona (1997a). An attempt at the reconstruction of the old Hungarian world of beliefs with help of written Latin sources is Szegfii (1996). Taking up theresearch of Melich, Ndmeth (1940a) holds theHungarian Christian word-stock (of Turkic origin) to be Danube Bulghar. The literature of the Hungarian word terem ‘womb (ofVirgin Mary)’, which has themeaning ‘tabemaculum’ in theCodex Cumanicus, is discussed in Ligeti (1986, pp. 275-276), and Rona-Tas (1995c, 1996g). For the oath over a dog see B&lint. (1971), Vajda (1979) and Sinor (1992). For Theotmar’s letter see the Hungarian translations in GySrffy (1975c, pp. 217-223), Kristo (1995d, pp. 185-187), the text in Gombos (1937, pp. 1469-1747), and LoSek (1997).
X. THE INTEGRATION OF THE MAGYARS WITHIN EUROPE
The chapter heading is not a m istake. We are looking at the integration o f the M agyars not into Europe, but within Europe. As we saw in the previous chapters, the M agyars have always lived in Europe, if we take as the border o f Europe the geographical border o f the continent today, that is, the Ural M ountains. We have to relate how the M agyars adapted to historical circum stances at least three times. Firstly when they transferred to a highly developed nom adic life style in such a way that preserved their ethnic identity; secondly when they preserved their tribal structure and further developed their ethnic identity on the K hazar-controlled steppe where the m ajority o f the population was Turkic; and thirdly when they arrived in the Carpathian Basin where once again they were able to adapt to circum stances and survive. Thus the M agyars have integrated within rather than into Europe. These very facts, however, do not give an answ er to the questions “how ?” and “w hy?”. The question is endlessly raised as to how and why the M agyars survived, when the Huns, the Avars, the Alani, the Cum ans or the Pechenegs all disappeared o ff the historical stage, or sunk in the sea o f peoples— espe cially w hen the M agyars’ history and language m arked them out sharply from anything that is usually called European. In this section o f my book I would like to point out certain facts relating to this.
1. THE INTEGRATION OF PEOPLES, THE TYPES OF ETHNIC CHANGE IN THE M ID D LE AGES The integration o f peoples with each other takes m any different forms, so that no com plete typology can be drawn up from any one point o f view. However, it is worth considering certain types that m ay be o f im portance to us, since in their sim ilarities and differences they both help us to understand M agyar history and show points o f contact in time and space.
374
From the Umls to the Carpathian Basin
It has long been noticed that there w ere considerable differences in the way nom adic tribes were integrated in China and around the M editerranean sea. The relations betw een the nom adic tribes and China varied greatly. If we start with the Turks, who play an im portant role in the history o f the M agyars, too, we discover that the integration o f som e o f the Turkic tribes into C hina began early on. On the gravestone o f the Khaghan Kol Tegin, who died in 732, we can read: [After the fam ous khaghans died] “then the younger brothers succeeded to the throne and the sons succeeded to the throne. But, apparently the younger brothers did not resem ble their elder brothers, and the sons did not resem ble their fathers. Consequently, unw ise khaghans succeeded to the throne, bad khaghans succeeded to the throne. The buyruks, too, w ere unw ise and bad. Since lords and peoples were not in accord, and the Chinese people w ere wily and deceitful, since they were tricky and created a rift betw een younger and elder brothers, and caused the lords and the people to slander one another, the Turkic people caused their state w hich they had established to go to ruin, and their khaghan whom they had crowned to collapse. Their sons w orthy o f be com ing lords becam e slaves, and their daughters w orthy o f becom ing ladies b ecam e servants to the C hinese people. The T urkic lords abandoned their Turkic titles. Those lords who were in China held the C hinese titles and obeyed the Chinese em peror and gave their services to him for fifty years. For the benefit o f the Chinese, they went on cam paigns up to the land o f the Bukli [Korean] khaghan in the east, where the sun rises, and as far as the Iron Gate in the west. For the benefit o f the Chinese em peror they conquered countries.” (Translated by Talat Tekin.) Indeed the Chinese divided the Turks, taking the m ajority into their service. The process o f Turk assim ilation thus began, their political autonom y was lost, and their ethnic autonom y was endangered. A group o f Turks, led by Kol Tegin and his minister, Tonyukhukh escaped from here and established the Second Turk Khaghanate. Others assim ilated into the C hinese population. It is inter esting that one o f the main arguments in favour o f independence w as that trade w as better carried out from a distance, w hereas being close to the Chinese softened them. The w arning was expressed in the follow ing term s: “If you stay in the land o f Ottiken and send caravans from there, you w ill have no trouble. If you stay at the Ottiken M ountains, your power, Turk people, will last for ever and you will live w ell!” The K hitais— a people speaking an old M ongolian language— offer another type o f integration. As w e have seen (pp. 253-255), they are m entioned as early as the 5th century, but reached the zenith o f their pow er in the 1Oth century. At the same time as the M agyars entered the C arpathian Basin, the K hitais occupied the north o f China. In 916, their ruler took the title o f em peror
The integration o f the M agyars w ithin Europe
375
and founded the Liao dynasty, which ruled until 1114. During this period, in southern C hina sm aller dynasties seized control one after the other, w ithout being able to m aintain their position. The rule o f the Khitai dynasty was brought to an end by another nom adic people, the Jurchens, relations o f the Manchus. T hey founded the Jin, or Golden dynasty. M any o f the Khitai had already becom e assim ilated Chinese under the Liao dynasty, but one group resisted this process and headed west, where they appeared as Black Khitais, or Kharakhitais. In 1125, their ruler took the title o f Gurkhan, and they ruled from the Caspian Lake to the Chinese border. The Chin dynasty was brought down by the descendants o f Chingis Khan, the M ongols. Unlike previous nom ads, who had seized only parts o f China, the M ongols brought the whole o f the country under their control, and established the Yuan dynasty. A t this point the assim ilation o f the M ongols began, although Chinese culture also developed continually under the mutual influence o f M ongolian and barbarian dynasties. W hen the M ing dynasty ended the rule o f the Yuan dynasty in 1368, the M ongols m ade four different choices. Some withdrew to the steppes with the last M ongol emperor. Some remained in the Chinese Empire in com m on areas in tight-knit groups, where they acted as a border defence against their form er M ongol brothers. A third group rem ained together, but having lost any connection with the other M ongols and m ixing with non-Chinese groups, they abandoned first their language and then their culture, but did not lose their identity. And finally, individual families m oved into Chinese-dom inated areas, and while they com pletely assim ilated, their m em bers preserved, up to the 20th century, the knowledge that their forefathers were Mongols. We can see that the Chinese attem pted to assim ilate foreigners in different ways. This was not as successful as it m ight seem from the outside. The pro cess took m any forms, was slow, and produced varying results. Nonetheless, because o f the undifferentiated nature o f Chinese society the integration o f foreigners was a one-way process, from which the only escape was to break away and go back to the steppes. The Rom an Empire developed differently. It, too, was faced with roving tribes. It beat back the early attacks o f the Germ anic tribes, and those o f the Huns, too. But the Germ ans kept appearing again and again. The change in R om an-G erm anic relations, which w ent back a long way, cam e in the 5 th century. It is worth exam ining these events in some detail, if we want to un derstand the question o f how and w hy the M agyars integrated within Europe for the third time. From the 5th until the 13th century a new w orld em erged which can roughly be divided up into three main periods. The first period, from the 5th to the 9th century, consisted o f the ‘civilised’ Rom ans pitted against the ‘barbarian’
376
From the Urals to the Carpathian Basin
Germ ans. The decisive factor was the fact o f belonging to the Rom an Empire. This brings to m ind Saint P au l’s proud declaration, civis romanus sum ‘I am a Rom an citizen’. B elonging to the Roman Em pire was m ore im portant than one’s real origins or social situation. However, Rom e was increasingly de pendent on the Germ ans for their m ilitary defence. First Odoaker, and then Theoderich took over real pow er in Rome. B ut Rom e preserved its autonomy, and Theoderich ruled formally on b ehalf o f the Rom an Emperor. The Rom ans and the Germ ans settled into a kind o f cohabitation. In the m eantim e, the contrast betw een civilised and barbarian w as trans formed. W ulfila christened some o f the Goths and spread C hristianity am ong the G erm ans in their own language. For a tim e it seem ed that the civilised-bar barian contrast w ould become transform ed into one betw een orthodox C hris tianity and the Arian doctrines held by the Germ anic tribes. A t the sam e time, this world came under new attack from the outside. From the north the Norm an-Vikings, in the C arpathian Basin the Avars, and in the Balkans the Turkic Danube B ulghars came into conflict with the R om an-G erm an world. From the south the A rabian great pow er em erged, bringing Persia under its pow er and m oving into the Iberian Peninsula. W hile it seem ed as though the centre o f gravity in the Rom an Em pire was m oving to Constantinople, the Germ anic tribes underw ent a total transform a tion betw een the 5th and 9th centuries. Some o f the tribes, like the Franks and the Saxons, survived, but now territorial ties and loyalty to the ruling houses played an increasing role in questions o f identity. The contrast between orthodoxy and A rianism did not prove to be a lasting one. The religious and non-religious organisations developed and operated in parallel. The m em bers o f the M erovingian dynasty played a m ajor role in this process, espe cially C lovis (483-511), who used the support o f the Catholic orthodoxy to strengthen his pow er over the Arians. The universal C hurch and the dif ferent peoples— gentes— created a com m unity system in which, alongside the sharing o f m orals, custom s and institutions (mores, leges, instituta), there was also a shared w ritten language, Latin. The introduction to the Lex salica in the m iddle o f the 8 th century m entions that the fact that God chose the Frank people, gens francorum, to defend Christianity, bestow s a legitim acy on that people. The b elief in the king and b elief in G od (fidelis regis, fidelis Dei) becam e one and the same thing. The word francus m oved tow ards m eaning ‘free’, referring to the free w arriors o f the em pire, i.e. it gained a sociological rather than a tribal or ethnic content. O ut o f the tribal nam e w as form ed a standard word in English: fra n k used to m ean ‘free’, and then ‘generous, liberal’, and today it m eans ‘sincere, candid’. A t the end o f the process a Romanised w orld existed next to a Germanised world. In Italy, France and the part o f the Iberian Peninsula not occupied by
The integration o f the M agyars w ith in Europe
377
the Arabs, neo-Rom an cultures developed, absorbing the different eastern and western G erm an groups as well. The local Germ an groups betw een the Rivers Rhine and Elbe rem ained in the majority, as can be seen in the balance o f power, too. The local population was hardly aware o f this, and the awareness o f the clerical literati can only be caught in slips o f the tongue or in m istakes over term inology or content. Tribal origins were barely a memory, and groups were differentiated according to their social position and their lords. The crowning o f Charles the G reat (Charlem agne) in Rom e at Christm as 800 m arked the end o f an ethnogenetic process: the end o f the integration o f the Germ an tribes, so that the firm pillars o f the new European order in the 9 th -11th centuries could take shape. It is not m ere chance that while the ethnic nam e Frank m oved from m eaning ‘bold’ to ‘free’, at the other end o f Europe we can observe exactly the opposite process. The nam e for the Slavic peoples, Sclaveni, Slavii, sim ilarly takes on a social dim ension, m eaning ‘slave’. In Latin we find sclavus, in Germ an Sklave, and in English slave. The Norm an-V iking traders played a role in this, but so did the Pechenegs and the M agyars. N aturally this process only lasted until the Slavic tribes formed units w ielding their own political power. The ethnic nam e Croatian first appears in 852, while the nam e Serb ap peared som ew hat earlier. It is worth paying attention to the appearance o f these tribal nam es, since the Slavs had been around for centuries by then. Their names always referred to the geographical surroundings: Polyane (the plane dwellers), D erevlyane (the woodlanders), Timochani (the people living by the River Tim ok), etc. This is actually how the prim ary R ussian chronicle ob served the events: “ Since they settled down by the R iver M orava, they called themselves M oravians” (sedosha na retse imenem M orava i prozvasha sya Morava). In C hapter 31 o f the work by Constantine Porphyrogenitus, he writes of the C roatians that they were descendants o f the pagan w hite C roatians who used to live in the area behind Turkia, i.e. Hungary. He w rites in a sim ilar fashion about the Serbs in the next chapter. He adds that in the tim e o f Heraclius some o f them were settled down in the Balkans. “This part behind Turkia” , w hich the Serbs called Boyki in their ow n language, covered the area o f present day Saxony, where Slavs still live, in Cottus and B autzen near Lausitz. This region was formerly Czech, and the people who now live here, the Sorbs, speak a w estern Slavic language. A nother interesting point is that the Em peror said o f the ethnic nam e ‘Serb’ that in the language o f the Rom ans this means ‘slaves’ (douloi). W hatever the origin and etym ology o f the word Serb is (som e claim it is Iranian in origin), here again we find the interpretation o f ‘Slav as slav e’, even if this is popular etymology, referring to the Latin word servus.
378
From the llra h to the Carpathian Basin
One group o f the Slavs was organised by the Frank trader Sam o in the first h alf o f the 7th century, in the region that later becam e M oravia. A nother Slavic group was organised by the Turkic Danube Bulghars. We have seen how varied were the origins represented by the Pannonian Slavs under Pribina and Kocel. The Kievan Rus Empire organised by N orm an-Viking arm ies cam e about in the sam e pattern. In some ways we could say the Slavic ethnic form ation was delayed by foreigners. The developm ent o f the Slavic peoples was com pleted by the 9th century. The M oravians and Czechs, organised on the Frankish m odel, the Serbs and the Croatians, organised by the ‘w hite’ C roatians and ‘w h ite’ Serbs, the Bulghars, organised by the Turks and increasingly assim ilated as Slavs, and the Kiev Slav region, organised by the Vikings and indeed populated by Slavs, were joined in the second h alf o f the 1Oth century by the Poland o f the Piasts. Out o f tribes organised on a territorial basis developed peoples with ties to the ruling dynasties o f each territory. The third type o f developm ent is represented by the Turkic peoples in Eastern Europe. The usual system am ong the Turks was a confederation of tribes, as we can see on the Inner Asian steppes am ongst the Turks, the Uighurs, the Khirghiz, the Kharlukhs and other Turkic tribes. This held true for Eastern Europe as well. The strength and historical advantage, and at one and the same time the disadvantage o f this tribal confederation, lay in its flexibility (see above, pp. 341-348). This structure allowed new tribes to join, enabled the gradual developm ent o f loyalties, and m ade it possible to change the structure quickly and easily. W hile new nam es constantly appeared, in m any cases the sum o f the tribes changed little. New tribes cam e to the fore, others were pushed to the back, some sections joined other confederations, but taken as a whole, the m ass rem ained unchanged. These confederations m anaged to consolidate their pow er in different ways. One was the construction o f a nom adic empire, o f which the K hazar Empire is an example. In this case, the tribal confederation found itself in circum stances which m ade the concentration o f pow er essential. The internal tribal structure was suitable for this from the beginning, since its needs could be fulfilled outwardly. M arauding cam paigns aim ed at trade were a common feature o f the Early Turkic period. The aim o f these was partly to force exchanges, and partly to seize treasures. The K hazars were able to conduct constant cam paigns to the north against the Slavs and to the south against the Caucasian peoples. At the same time, control over the trade routes also brought in wealth, giving the tribes an interest in trading with B yzantium and the Arabs. The tem porary decrease in pressure from the east contributed to the con solidation o f power, and in the course o f time the dual kingship settled into a
The integration o f the M agyars w ith in Europe
379
stable form and took on the proportions o f a cult. However, this underm ined the tribal confederation. The role o f the sacral king, the w ar ruler, the armed retinue, the hiring o f m ercenaries and the settling o f the people on the land all robbed the tribal confederation o f any role or weight. The representatives o f the confederation came into conflict with the centralised authorities, and once defeated, left the empire. The Avars and the Danube Bulghar Empire offer two exam ples o f how tribal confederations can develop in a different direction. In both cases a leading stratum emerged. Once again, m arauding trade cam paigns played a significant role, but here the direction was only the Byzantine and the R om an-G erm an world. However, there was little long-term future in this, both because the counter-attacks came from an enemy that enjoyed far greater reserves, and because Byzantium and the Frankish rulers could exercise a far w ider degree o f political diplomacy. A decisive feature o f the cam paigns was the fact that the armies were only in part Bulghar and Avar: a considerable proportion was Slav. This strengthened the position o f the Slavs within both empires. In the case o f the Bulghar Empire, the process resulted in the Slavification o f the Bulghars, although the fram ework o f the state was preserved with the help o f the conversion to Christianity. The basis was laid down probably by the laws of Krum (803-814), which did not survive them selves, but can be pieced together from later references, which suggest that this was the period when power was centralised and tribal power curtailed. The sam e Slavification took place am ong the Avars, too, but the conversion to Christianity took place too late, and the frames o f the empire could not be saved. We can distinguish three main types o f integration in m ediaeval Europe: the German, the Slavic and the Turkic types. The Germ an and the Slavic types were sim ilar in that they w ent beyond the tribal structure: the differentiated society was able to organise on the basis o f territorial com m unities and loyalties, and could develop a new system o f taxation. The difference between the two lay in the fact that w hile the Germ ans dism antled the tribal structure, the Slavs either did not develop it, or if they did, outside forces soon brought it to an end. In both societies clanhood played a m ajor role, as the clan repre sented a unit both in terms o f taxation and m ilitary service. The Turkic type was based on the tribal confederation. This had both advantages and disadvantages. Overall, it obstructed m ediaeval ethnic devel opment, since it underm ined loyalties. Let us disregard here other examples o f m ediaeval Turkic developm ent, such as those o f the Volga Bulghars, the Seljuks, the Cumans, and the Ottomans. Let us also disregard those types represented by the “sea nom ads”, the northern peoples such as the Vikings, who bore some sim ilarity to the Turks. In the three developm ent processes in 9 th -10th-century Eastern Europe that we have examined, that o f the Khazars,
380
From the Urals to the Carpathian Basin
the Danube Bulghars and the Avars, we could see that the processes o f the ending o f the tribal confederation and the developm ent o f territorial pow er had already begun. This, and only this, could m ake it possible to introduce a new system o f taxation, and in this way ensure that the system w ould not collapse due to any fiasco o f a m arauding campaign. A system w here the main source o f incom e was collected or forced out o f foreigners quickly reached the brink o f collapse. The break-up o f the tribal confederation o f the three Turkic peoples took place in different ways: the K hazars and the Bulghars m anaged to consolidate their empires “from above” , in both cases sanctifying this by converting, the Khazars to Judaism , and the B ulghars to Christianity. For the Avars this did not succeed, even though their leaders w ere christened before the Bulghars. N onetheless, both peoples were gradually assim ilated into the Slavic peoples. It was the Bulghars that delivered the fatal m ilitary blows to the foundations o f the Avar Empire.
2. THE THIRD INTEGRATION OF THE MAGYARS The M agyars fitted once into the nom adic world, once acclim atised to the K hazar Em pire, and once accom m odated them selves to the new circum stances created by the Conquest. As we have seen, the key to the different European integrations, and thus to the em ergence o f m ediaeval peoples in Europe, was how they achieved a European system organised on a territorial basis, socially stratified and based on feudal loyalties. Each o f these factors in different ways represented an end to the system based upon the tribe and the tribal confederation. This took place am ong the M agyars, too. M any o f the details o f the process m ay rem ain enlightened guesswork, but in the light o f the preceding analyses, we are justified in drawing the outlines o f the process. During the C onquest the M agyars appear as the Seven M agyar tribal con federation. We know that there were not only seven tribes, and that they were not only M agyars. However, before the C onquest the pow er relations within the confederation were decisively shaped first by its dependence upon the K hazar khaghanate, then its gaining o f independence, later by the joining o f the K havars, and finally by the House o f Alm os and Arpad. An im portant point in the developm ent o f the M agyar peoples was noted by Constantine Porphyrogenitus, w ho wrote: “These eight clans (okto geneai) o f the Turks do not obey their own particular princes, but have a joint agreem ent to fight together w ith earnestness and zeal upon the rivers, w her ever w ar breaks out. They have for their first c h ief (kephale prote) the prince who com es by succession o f A rpad’s family, and two others, the gylas [gyula]
The integration o f the M agyars w ith in Europe
381
and the karhas, who have the rank ofjudge: and each clan (genea) has a prince” (translated by Jenkins). Thus instead o f w riting that the tribal princes, who had no pow er in times o f peace, lead each tribe into battle, he says they did not obey their tribal princes, and in times o f war fought together, and inciden tally adds that every tribe had its arkhon. This would suggest that the tribal confederation no longer had a m ilitary role in the m iddle o f the 1Oth century. The very rapid break-up o f the tribes and the M agyar tribal confederation is indicated by the so-called “tribal place nam es” that are found in the earliest Hungarian docum ents, i.e. those place nam es that consist o f one o f the nam es o f the seven tribes, unadulterated by any affix (M egyer, Gyarm at, Nyek, Kiirt, etc.). These place nam es are in great num ber scattered over the Carpathian Basin. These point to the fact that in their surroundings, their tribal background was considered to be sufficient identification. The dispersing settlem ent could have happened after 1000 A I), i.e. on the orders o f Saint Stephen, but the facts suggest otherwise. The spreading out o f com m unities suggests that there was som ething to spread out for— that there was a m oving force behind this action: the will o f the central pow er to break up the tribal confederation, and the process began before 1000 A D . In the literature on the M agyars’ incursions on the w est I did not find any reference to the participation o f the Slavic populations o f the C arpathian Basin, as was carefully docum ented about the Danube Bulghars and the Avars. This m eans that they did not participate in m ilitary power, although it is true that there had not been m uch tim e for this. Thus Slavic leaders did not reach im portant m ilitary positions. This was a m ajor factor in the fact that the M agyars, unlike the Bulghars and the Avars, did not becom e assim ilated. As we have seen, a key question in the types o f ethnic developm ent w as the change in loyalties. Tribal affiliations were o f special im portance where, like on the steppes or am ong the Germ ans, the neighbours belonged to the same people and the enem y spoke the sam e language. The M agyars on the K hazar steppe and later in the Carpathian Basin, however, faced enemies w ho neither spoke Hungarian nor had M agyar traditions. The dissatisfied groups left the tribal confederation and w ent o ff w ith the Bulghars to the V olga-K am a region. As the tribal confederation as an organisation rapidly disappeared in the Carpathian Basin, the people once again emerged. The notorious raids o f the M agyars w ere hardly organised on a tribal basis. In the note by Ibn Hayyan about the 942 cam paign o f the M agyars in Spain is an interpolation about the seven leaders o f the M agyars, w hich goes back to focal sources from an earlier period which had probably drawn on the B yzantine sources (see pp. 73 and 277). The reasons w hy the conquering M agyar people reached the threshold o f creating a state and was quickly able to fall in with European developm ent are
382
From the Urals to the Carpathian Basin
m any and com plicated, and are interwoven w ith historical chance. But some o f the reasons can be picked out: the rapid destruction o f the tribal confedera tion, the exclusion o f the native population from initial m ilitary activity, and the fact that the vast m ajority o f the conquering peoples spoke Hungarian. In the M iddle Ages, the ability to speak m ore than one language was m ore natural than we think today. Not only did the Khavars learn the language o f the M agyars, but som e Slavic groups are also known to have joined the invading M agyars before the Conquest. M any o f the M agyars are know n to have spoken the language o f their neighbours, and o f som e o f the local peoples. However, the know ledge o f languages and linguistic identity are not the same thing. T hat the M agyar people and the Hungarian state developed in East-Central Europe in so short a time is undoubtedly partly explained by the fact that on the eve o f the C onquest there w as no unified and organised political form ation in the Carpathian Basin, and that its population at the tim e lay on the edge o f other political units, that o f the M oravians, the Franks, the Bulghars, the C roatians and the Serbs. A nother factor lay with the Slavified Turkic Avars, who hoped that the M agyars w ould help them revive the Avar Em pire. The Turkic chiefs o f the Bulghar Zhupas also played a role in the building o f the Hungarian state, hoping their new m asters w ould support them against the local Slavic population. These reasons m ay represent only a part o f the historical picture, but can help reveal the enigm a o f how the conquering M agyars w ere able to form their own state in less than one hundred years. The form ation o f the Hungarian state was not an isolated event in m ediaeval Europe. At alm ost the same time, new states em erged on the fringes o f the Caroiingian Europe such as N orw ay (986), Denm ark (960-976), Poland (after 960), the Czech Kingdom (921-935) and the K ievan Rus (from 983). The events o f the years 9 73-997 in Hungary, as w ell as 1001, the coronation o f the first Hungarian king, Saint Stephen, m arked not only the foundation o f a new state, but also led to the birth o f a new historical region. The form ation o f ethnic groups did not com e to an end, but proceeded under fundam entally different circum stances. Thus, a historical transition had ended, providing place for new forces and new formations.
The integration o f the M agyars w ith in Europe
383
NOTES
For the translations of Turk inscriptions see Tekin (1968). The best source on the Khitai is still Wittfogel-Feng Chia-Sheng (1949), even if one might sometimes disagree with their conclu sions and interpretations of the linguistic material. There is considerable literature on thedevelopment andthe ethnogenesis ofthe“barbarians”. Of the international literature, I refer here only to the works cited in the Notes to Chapter I. I gave a lecture on the types of ethnic development at the Hungarian Academy of Sciences’ Committee for Magyar Proto-History conference in 1995, thetext ofwhich appeared in Magyar Tudomany (Rona-Tas 1996c) and an extended version in the volume containing the papers of the conference (see R6na-Tas 1997c). The unsuffixed tribe names of the Hungarian tribal confederation as place names (as e.g. Megyer, Kir, Keszi, etc.) are dispersed in the Carpathian Basin in a great number. This fact has been recognised and widely discussed in the literature. There is agreement on their antiquity and on most questions of details. The only point yet discussed is whether the dissolution of the tribal system began only after 1000 or much earlier. This tangles the question of the eventual role of the tribes and the tribal system in forming the new Hungarian state. The “Turkic” tribal system began to dissolve before the Conquest. The process accelerated in the middle of the 10th century. The organisation of the new Hungarian state was based on the “Finno-Ugrian” clan system and this was one of the reasons why the Hungarians as Finno-Ugrians did not disappear. On the tribal names and their surfacing as place names see Homan (1938), and Kristo-Makk-Szegfii (1973).
XI. SUMMARY OVERVIEW
We have attem pted to place the history o f the conquering M agyars in a new framework. The m ain tools o f this endeavour have been the theoretical base, stricter methodology, broader source m aterials and a broader historical con text. We defined the concept o f people. A people is the group o f individuals who share a com m on sem iotic system or code, and whose m em bers conscientiously differentiate them selves from other peoples, and have a com m on self-desig nation, an ethnic name. The m ost im portant com ponent o f these shared systems is language. However, anyone can becom e a m em ber o f a linguistic community by changing his or her language. O ther factors, such as the knowledge o f shared lineage, shared territory, shared political organisation, shared religion and religious beliefs all contribute to the developm ent o f a people. I f we consider the M agyars as a whole, there is no reason to assum e a change o f language. However, num erous other groups did becom e a part o f the Hungarian people by changing their language. We set the limits o f ancient M agyar history by defining it as beginning in the 1st m illennium BC when the M agyars left the com m unity o f related peoples. Thus we can talk about independent M agyar history from the 5th or even the 8th century BC. This early period comes to a close when the Hungarian state is founded in 1001 AD. We have given a broad interpretation to the sources on ancient M agyar history, drawing not only on w ritten sources, but also on the findings o f linguistics, anthropology, archaeology and physical anthropology. With the written sources we have endeavoured to draw on the latest critical interpreta tions. In our use o f linguistic sources we have been especially critical, since, in this respect, Hungarian research is fifty years behind the times. We have indicated the m ethodological problem s and limits in em ploying the results o f archaeology, ethnography and physical anthropology, and have attem pted to meet the m ethodological dem ands o f using such varied sources. In our initial examination o f the Conquest and the historical events that preceded it, we excluded the M agyar chronicles as secondary and later sources, drawing upon
386
From the Urals to the Carpathian Basin
them only when it fitted in with the prim ary sources. In areas w here contem porary sources did not give enough direction, w e used contem porary analogies to form hypotheses. A t the sam e tim e we tried to keep to a m inim um the num ber o f assum ptions, preferring to state where the picture was not clear or where there was no definite answ er to questions— rather than constructing a shaky house founded on guesswork. We presented the history o f related peoples speaking Finno-U grian lan guages, and we have paid m ore attention than before to the history o f the neighbouring peoples, especially the Eastern European and the nomadic peoples o f the steppes. We have attem pted to present early M agyar history as a part o f the history o f Europe, or m ore particularly o f Eastern Europe and the steppe. M oving on to the history o f the conquering M agyar people, we first o f all exam ined the nam es o f the M agyars before and at the tim e o f the Conquest. We established that the nam es prior to the C onquest reflected the nam es given by the M agyars’ 9th-century neighbours, rather than the period betw een the 6 th and 9th-centuries, as had hitherto been believed. We ascertained that the Byzantines called the M agyars Turks, the Slavs called them Onughurs, the Volga Bulghars Bashkirs, and m ost probably the Khazars M ajgars. The origin and use o f the ethnic name Savarti A sfali are unclear. C ertainly by the 7th century, but probably m uch earlier, too, the M agyars called them selves M agyars or m ore precisely M agyers. This nam e originally was form ed out o f two tribal nam es, the M anys (later M agy) and the Er. Foreigners have called them M agyars since the Tatar invasion, probably influenced by the interpreters who cam e from the M agyars into the Volga region. The proto-M agyars cam e from the region betw een the Ural M ountains and the Kama. In the region o f the U grian Urheimat deforested land cultivation, prim itive anim al-keeping, and bronze crafts all played a m ajor role alongside hunting on horseback in the life o f this com m unity o f closely related peoples. The history o f the Ugrian community, which was shorter than we previously thought, and the ancient history o f the independent peoples that em erged from that com m unity all took place in the w ooded region o f the Urals and around the Volga. W hile the proto-M agyars had long and very close relations with the Iranians, the independent M agyars had in the first period only trade relations with them, while having a loose but far from negligible contact with the Perm ian peoples, ancestors o f the Zyryans and the Votyaks. It w as during this period that the M agyar people was form ed out o f the M anys and Er groups. The independent M agyars m oved southwards, once again developing close relations with the Iranians, who controlled the steppe. A t this point the M agyars began the transition to a nom adic way o f life, a process that we believe took m uch longer than was previously thought. For centuries, the
Sum m ary overview
387
M agyars lived in the southern region o f the Urals in such a w ay that some groups cultivated the land, reared anim als and hunted on horseback, while others participated m ore folly in the steppe life, joining the nom adic m ilitary campaigns o f the Iranians, and subsequently those o f the Turks. The Hunnish period, w hich fitted in betw een the Iranian and the Turkic periods in the 3rd^-th centuries, m ay have affected the M agyars to some extent, but a more essential change came in the 5th century, when the steppe was under the control o f the Turkic tribes, or m ore precisely, o f the different O ghur groups. Even then, the shift to a nom adic way o f life was gradual, rather than sudden. The decisive change came only at the end o f the 6th century when the M agyars m oved from the southern Urals to the K uban-M aeotis region, which had been vacated by the Onoghur Bulghar-Turkic groups. They rem ained there until the end o f the 7th century. We presented a new picture o f O noghur-B ulghar history. The centre o f K huvrat’s B ulgharia was situated around the River Dnieper, and m ore broadly the Rivers D niester and Don. In the 7th century, the M agyars took part in the Khazar defeat o f the Bulghar Empire. The Bulghar tribes w est o f the Dnieper moved to the Lower Danube, the C arpathian Basin, and the two shores o f the Adriatic Sea. Those who rem ained m oved northeast to the banks o f the Don. The M agyars occupied the region between the Dnieper and the Low er Danube (the Etelkoz). Contrary to earlier belief, the latest research leads us to think that Levedia was not an independent, form er Urheimat occupied by the M agyars; rather, the tribal territory o f Prince Levedi, within the borders o f the Etelkoz. The M agyars lived here, between the D nieper and the Low er Danube, from the end o f the 7th century until the Conquest. Their close but varying rela tions with the Khazars, the Bulghars, the Alani, the Byzantines and later the Pechenegs and the Slavs fundam entally shaped the conquering M agyars. In this period the M agyars becam e increasingly independent, and by the second half o f the century they had becom e an autonom ous and significant European power. We paid particularly close attention to the history o f the Pechenegs— how they positioned them selves at the end o f the 9th century— and the events leading up to and surrounding their attacks on the Magyars. Leading the changes that took place in the M agyar tribal confederation were the leaders o f the w estern h alf o f the region, the A lm os-A rpad clan which gradually took over power and assum ed the role o f supreme commander. Relying partly on their m ilitary entourage, and partly on the Khavars, who transferred allegiance from the Khazars, they dethroned the last leader o f the eastern lands close to the Khazars, Levedi, who had been com pletely stripped o f power. They succeeded in winning the subsequent recognition o f the
388
From the Urals to t!w Carpathian Basin
surrounding powers, including the Khazars. There is no indication in the sources that the special set-up o f the dual kingship, the sacral kingship, developed after the Conquest. M aking the m ost o f the p a x khazarica at the end o f the 8 th century, the M agyars established close relations w ith the Turkic and Slavic peoples in the Carpathian Basin and the Balkans, and w ith the Franks. We endeavoured to place the events o f the C onquest in the broader context o f the 9th-century history o f Eurasia from the eastern fringes o f the steppe to the Franks. In the background o f the C onquest the m ain pow er divide o f the period was betw een the F rank-B ulghar-P echeneg coalition and the em erging B yzantine-M oravian axis, although the reality was far m ore com plex than this. The conflicts between the Franks and M oravians and later the Avars, and betw een the Byzantines and the B ulghars were constantly changing. Also exercising strong political influence on events w ere the Papacy, the C onstan tinople patriarchy, the A rchbishop o f Salzburg, and the B ishop o f Passau, who used political tools in their rivalry to convert the peoples o f the region. We exam ined the peoples o f the Carpathian Basin, the short life-span o f Slavic rule in Pannonia and the Slavification o f the Avars, the latter a slow process that had not been com pleted by the time o f the Conquest. From the beginning o f the 9th century the M agyars played an increasingly im portant role in these events. We divided the C onquest itself into three periods follow ing the battles o f 894. The first is from 895 until 898, the second from 899 to 900, and the third lasting up to 902. The m ain body o f the M agyars cam e from the northeast, travelling along the R iver Tisza to attack the Bulghars with the aim o f occupying the strategi cally crucial salt m ines under B ulghar control in the M aros region. O ur sources do not allow us to say for certain w hether they spent the w inter o f 894-895 in the Upper Tisza region, w hether they w ithdrew to the Etelkoz and crossed the Verecke pass again in the spring, or again w hether some encam ped for the w inter around the Tisza and w ere joined by others in the spring. W hat w e can say is that in spring 895, the main body o f the M agyar arm y w as in the eastern h alf o f the C arpathian Basin. The Byzantines did not support the m ilitary cam paign conducted from here by attacking the B ulghars from the south. The Bulghars defeated those from the Etelkoz attacking the Low er Danube, while the Pechenegs attacked the M agyars in the Etelkoz. Those that rem ained fled to jo in the m ain M agyar arm y in the C arpathian Basin. Between 895 and 898, the w estern border o f the M agyar forces v/as the R iver Tisza, sending at m ost reconnaissance parties foraying into the region betw een the Tisza and the Danube. The international situation favoured the M agyars. The M oravians under M oym ir II, trusting in the support o f the
Sum m ary overview
389
M agyars, set up their own political and independent church structures. The Franks, however, in league with the Czechs— M oym ir’s brother Svatopluk II and the Frankophile ruler o f Transdanubia, B raslav— endeavoured to prevent them. M aking the m ost o f this fraternal strife A m ulf, the Frank ruler who had been crowned in 894, severely weakened the M oravian Empire. He then m ade a deal w ith the M agyars as part o f his plans to extend his pow er to Italy. Consolidating the position and structure that they had achieved so far, the M agyars sent troops to attack northern Italy in 899-900. Berengar, who held sway there, bought the withdraw al o f the M agyars by offering food and hostages. In D ecem ber 899 A m u lf died. The returning arm y from Italy and the m ain M agyar arm y crossing the Danube caught the M oravians in a pincer movement, who were busy devastating Transdanubia, and occupied that re gion. By the end o f 900 the Bavarians were building defensive lines on the River Enns. The M agyars, who were consolidating their control over the Carpathian Basin, sent peace delegations in order to secure the gains they had made. These were rejected by the court o f Louis the Infant, A m u lf’s successor, whereupon the M agyars occupied the K isalfold (N orthw est Plain) and pressed westwards along both banks o f the Danube. Their advance was halted by the arm ies o f the Bishop o f Passau. By 902, the M agyars finally occupied M oravia, which marked the end o f the m igration and accom panying m ilitary actions that Hungarian historians term the H onfoglalas, literally m eaning ‘hom eland occupation’, but translated as ‘C onquest’, for w ant o f a better word. The Conquest was led by the M egyer tribe, headed by the H ouse o f Alm os and Arpad. We have put forward the view that Kusal (Kursan, Kiisen) was a member o f this dynasty, and probably A rpad’s brother. The historical circum stances and the lack o f tim e m eant that the dual kingship could not com e into being again after the dethroning o f Levedi, and the sacral kingship even less so. The M agyar tribal confederation was a m ultiply differentiated system, in which the tribe o f the prince played a prom inent role, w ith the other tribes being subordinate to the leading tribe at various levels. The seven M agyar tribes were jo in ed by three K havar tribes that unified to form one tribe, and probably other groups also joined them. O f the tribe nam es, only the etym ol ogy o f Jeno and Tarjan is certain, but the linguistic origin o f tribe nam es tells us nothing about the ethnic and linguistic m ake-up o f the tribes. The M agyar tribal confederation undoubtedly spoke Hungarian, and the M agyarisation o f the Khavars occurred very quickly, while some groups o f the M agyars were for a while bilingual, speaking Turkic as well. After the Conquest, the M agyars were able to preserve their ethnic and linguistic identity, because the people in the Carpathian B asin w ere not unified; Slavic and Slavified Avar groups, Frank and B avarian settlers, and
390
From the Urals to the Carpathian Basin
perhaps other fragments o f people from earlier periods lived in a m ottled community. It was not ju st the lack o f ethnic and political unity am ong the Slavs that m eant they could not assim ilate the M agyars in the sam e way as they had the Bulghars and the Avars; they were also excluded from any m ilitary role in the M agyars’ cam paigns beyond the region. This shut them out from any positions o f power and hindered any organised defence o f the C arpathian B asin’s Slavs. The conquering M agyars were led by the prince, by the suprem e com mander, gyula, and the karha, who kept control over the peoples who had joined the M agyars or had been subjugated by them. Initially, the conquerors left intact the political and social structure o f the Carpathian Basin, the main features being a county system formed by small Slavic groups, with Avar and B ulghar zhupans at the head o f the counties. The head o f the B ulghar zhupans was the ‘N andur zhupan’ which m eant ‘Bulghar zhupan’. This title was taken by the M agyars, and becam e used as nadorispan. M agyar agriculture was characterised by the cohabitation o f a strong nom adic anim al-rearing group that em ployed ‘m igrating cultivation’ and an anim al-rearing group that also cultivated the land. This settled upon the local farm ing o f the Avars and Slavs. In term s o f culture, there was Sham anism as well as Tengrism and num erous elem ents o f the main world religions which the M agyars had gained acquain tance with in the Etelkoz, in the vicinity o f the Khazars and Byzantium . We have a good picture o f their m aterial culture and their customs. It is almost certain that they also had runiform writing. The main question concerning the Conquest is the unparalleled, uniquely efficient integration o f the M agyars, which took place very quickly— histori cally speaking. The M agyars adapted to historical circum stances three times in the 1st m il lennium. First o f all when they made the transition to nom adism , com prising the rearing o f large anim als, on the border o f the steppe and the forests; secondly when they integrated into the econom ic and political system o f the K hazar Em pire; and thirdly when they abandoned their tribal confederation in the Carpathian Basin and integrated into the European feudal system. We described three main European paths towards m ediaeval ethnic form a tion. A com m on feature o f the German, Slavic and Turkic types was that the tribe ceased to exist as a political and econom ic unit; empires or states were formed on a territorial basis, or the people as such ceased to exist. However, there were m ajor differences in the processes o f developing as a people. The Turkic path was characterised by the tribal confederation w hich among the K hazars developed into an empire, but was not able to end com pletely the tribal system. Am ong the Bulghars, in parallel to the process o f ethnic
Sum m ary overview
391
assim ilation, central pow er was able to end the tribal confederation and lay the foundations o f a state, and this was strengthened by conversion to C hris tianity. The Avars failed to end the tribal system and lay the foundations o f a state due to internal and external factors. Several factors facilitated or ensured the successful developm ent o f the M agyars in the Carpathian Basin. O f these, we should m ention the politi cal transform ation that took place over a long— longer than previously as sumed— and relatively quiet period in the Etelkoz, in which the dethroning o f the House o f Levedi was just one episode. This period m arked the beginning o f the end o f the tribal organisation, although at this time the process followed the Khazar m odel. This continued after the Conquest, but under different conditions. A dapting the Slavs’ territorial organisations while excluding them from m ilitary pow er proved to be an effective com bination. The divisions among the local Slavs, the particular role o f the Slavified Avars, the internal problem s o f the Danube Bulghars, the collapse o f M oravian rule, Frankish ambitions in Italy, the new policy o f the M acedonian dynasty in Byzantium , the internal problem s o f the Papacy, the appearance o f the Czechs and the forming o f Kievan Rus all represented the historical context. From the end o f the Conquest around 902 to the establishm ent o f the Hungarian state by Saint Stephen in 1001, there was a period o f about one hundred years. The position o f the M agyars at the time o f the C onquest m erely laid the basis for their effoits to make the m ost o f their opportunities in spite o f every internal and external difficulty and in spite o f their cam paigns and defeats. How they m anaged to do so, and in this way create a European state, may be the subject o f another book.
PART FOUR
RECENT RESEARCH AND STUDIES
XII. AN OVERVIEW OF THE STUDY OF ANCIENT HUNGARIAN HISTORY
The research of ancient Hungarian history goes back to the beginnings of Hungarian historiography itself. Due to the limitations of space, I cannot give a survey of research history here, but merely a summary of the research of the past few decades. I am fully aware of the presence of bias both in the selection of the titles and in my comments. I shall only give thought to those issues relevant to the subject of this book; in doing so, however, I cannot be ex haustive. This chapter is more or less identical with the preface of my book A korai magyarsag tortenete [The ancient history of the Magyars], Since I am writing here of the research work of others and myself, I am giving the bibliographical references in the main text. Ever since my college years I have been interested in the ancient history of the Magyars, and in the emergence of Hungarian culture. My profes sors—Gyula Ortutay, Istvan Talasi, Lajos Ligeti and Gyula Nemeth—con vinced me that unless I really became involved in the study of the broader context, I would never achieve even the slightest result. I was undaunted, however, by the fact that direct involvement in this field was impossible for two reasons. Especially in the 1950s and 1960s in Hungary, the research of ancient Hun garian history was a very haunted field indeed. Ideological expectations, political prohibitions and justified and unjustified fears put huge pressure on scholars, rendering their work well-nigh impossible, and hindering genuine progress. Packed with tragicomic scenes, the so-called Erik Molnar debate would make a separate chapter in scientific history. The Hungarian Society for Linguistics held a debate on Hungarian prehistory on 1st December 1953 in the auditorium of the Lorand Eotvos University which was attended by many. However, the most significant talks and remarks were not published until two years later (A m agyar... 1955). The other obstacle to research was the unhealthy atmosphere that prevailed in Hungarian scientific life; the scientific scene was controlled by a few monopolist factions.
396
Recent research and studies
It was no m ere chance, then, that until the latter h a lf o f the 1960s little happened in the field o f ancient Hungarian historic research. Questions and ideas were conceived in a constrained situation. Scholars from m any countries came up with the idea o f writing a new European m ediaeval history. Eastern European efforts were led by scholars o f the G erm an D em ocratic Republic who went beyond the standard procedure o f scientific co-operation and got the non-scientific authorities involved, too. The resulting w ork should have been the Enzyklopadie der Friihgeschichte der europdischen Volker which had originally sought to give the M arxist version o f the entire history of m ediaeval Europe, but latterly contented itself w ith a m ore m odest Eastern E uropean survey. But it very soon transpired that the Eastern European scholars, encouraged to collaborate by the central com m ittees o f their respec tive com m unist parties, were unable even to share out the headwords. So the editors m ade a strict and sim ple— yet com pletely unscientific— decision. Every country could write only those headw ords which related to m ediaeval events that occurred within the present-day boundaries o f that country. The story o f this stillborn scheme would certainly m erit a short study; still, some scholars were inspired by it. However, it transpired that despite all o f the com m only shared ideological constraints, the researchers o f the different eastern countries disagreed even on the m ost fundam ental issues. The use of a com m on term inology was sim ply out o f the question therefore; moreover, so was a m utual discussion o f the terminology. The Hungarian scholars seized the first opportunity to quit, and m ade an abortive attem pt to publish a Hungarian encyclopaedia. Saving what they could, they com piled and edited the ready articles, and finally in 1994 published the Lexicon o f Ancient Hungarian History (Kristo 1994c). Sim ultaneously attem pts were m ade to bring together the scholars o f two countries to discuss m utual problem s. Although such scientific m eetings were unsuited for the developm ent o f a com m on language, there were some rem arkable attem pts. For instance, betw een the 4th and 6 th o f M ay 1971 a C zech-S lovak-H ungarian m eeting was organised in Bratislava, with the aim o f establishing the conceptual-historical background o f the historical form a tion preceding the m odern nation (S pira-Szucs 1972). Jeno Sztics’s paper Л “N em zetiseg ” es “nem zeti ontu d a t” a kozepkorban. Szem pontok az egyseges fo g a lm i nyelv kialakitasahoz [“N ationhood” and “national identity” in the M iddle Ages. A spects for the developm ent o f a uniform term inology] (S piraSzucs 1972, pp. 9 -1 7 ) was published from the proceedings o f the meeting. A lthough this was not the first article Sziics had w ritten on the sam e issue (see Szucs 1966, pp. 245-269), we are aw are today that his w ork m eant a turning point in Hungarian research. Sziics closed his dissertation in 1970 and pub
An overview o f die study o f ancient Hungarian history
397
lished his w ork in 1971 (Szucs 1971). He was a follow er o f W enskus, and gave a new theoretical context to the related issues. Formerly, scholars had focused on the origins o f nation and people. Gyula N em eth’s work, A honfogla/o m agyarsag kialakulasa [The form ation o f the conquering M agyars], written in 1930, reflects a significant change o f attitude, on account o f its focusing on form ation, as opposed to origin. A m onolineal proto-history was superseded by a com plex ethnic history which gave consid eration to the fact that the history o f a language did not (necessarily) coincide with the history o f its speakers— not even given that the m ost im portant source o f the historical reconstruction process is language. “Only ‘in general term s’ does the research o f the origin o f a language shed light on the origins o f the people; due to the fact that the concept o f linguistic affiliation is not identical with ethnic affiliation,” N em eth wrote in 1930. Szucs w ent farther, and considered the study o f the history o f ethnic identity to be im portant. In the 1970s and 1980s he put much effort into his chosen subject, and his results appeared in a num ber o f publications. He considered ethnicity to be a subject o f ethno-sociology, which could not be understood w ithout the reconstruction o f the phenom ena and history o f ethnic identity. In Szucs’s view the Rom an Empire, the early “barbarians” , the Germ ans and the later steppe peoples— among them the conquering M agyars— represented three types o f ethnic self-reflection. His m any-sided research provides a social historical basis for the study o f these historical groups. Szucs’s dissertation appeared long after his tragic death in 1988. Together with an unfinished paper, it was published in 1992 bearing a more suggestive title than the original: A m agyar ostortenet vazlata [An outline o f Hungarian proto-history] (Szucs 1992). In 1970, I subm itted to the Hungarian Scientific Degree C om m ittee m y academic doctoral thesis entitled A z altaji nyelvrokonsdg vizsgdlatdnak alapja i (A nyelvrokonsdg elmelete i s a csuvas-m ongol nyelvviszony) [The foun dations o f the research o f Altaic linguistic relationships (A theory o f linguistic relationship and the C huvash-M ongolian linguistic relations)]. W hile Szucs had sought to clarify the term inology from an ethno-historical viewpoint, I aimed at doing the same in the field o f linguistic relationships. The chapters on linguistic relationship were published in 1978 (see Rona-Tas 1978a). W hile S zucs’s research did not cause much o f a stir (see my rem arks about it in Rona-Tas 1985c, pp. 133-136), another scientist’s new theory provoked a real storm. In 1969, Gyula Laszlo propounded his new theory at a talk he gave in Budapest. He published a sum m ary o f his ideas in two studies in 1972, and in a small brochure in 1973 (see Laszlo 1970a, 1970b, 1973). The theory o f a double conquest contends that when the M agyars entered the Carpathian Basin in 895, they cam e across significant M agyar groups who w ere supposed to have com e to their new hom eland in 670. The second m igratory wave o f
398
Recent research and studies
M agyars is supposed to have “avoided” the M agyars o f the first Conquest. L aszlo’s theory was debated at the 4th C ongress o f Finno-U grian Scholars in 1975 (published in Laszlo 1975), and in 1978 he published a new sum m ary (Laszlo 1978, 1982). The initial clam orous response to the theory calm ed down after a while. Several publications appeared in response to the new theory, but m ost o f these dism issed L aszlo’s views. Laszlo replied to many, for instance in the Szom bathely periodical E letunk [Our life]. Later a collec tion o f his w ritings from this paper, together with an interview, w ere published in several editions (Laszlo 1987). The last notew orthy contributions to this debate w ere w ritten by Im re Boba (1982-83), G yula K risto (1983a), Istvan Bona (1984a) and Csanad Balint (1989a). As a w hole the theory was not accepted. It m ust be said for Gyula Laszlo, however, that he drew the line betw een the early and late Avar cultures. Thus, albeit indirectly, he helped to m ake clear a num ber o f issues o f the M agyar Conquest. Essentially at the same tim e that Gyula Laszlo propounded his theory o f a double Conquest, G yula N em eth (Julius № m eth in his publications other than in Hungarian) cam e out with a revision o f the views o f G om bocz. In 1912, G om bocz (1912, pp. 201-208) established that M agyar-T urkic interaction took place in the V olga-K am a region, and contended that the Chuvash-type Turkic loan words in Hungarian m ust have entered the language at this local ity, betw een 600 and 800. Even in this w ork he did not deny that the Kama M agyars m ight have w andered north from their southern hom eland. Gom bocz later m odified this theory, and claim ed that M agyar-T urkic interaction took place farther south, in the Caucasus region, and a lot earlier. Published in 1917, his first study was followed by m any others (G om bocz 1917-1927, 1920, 1921). N 6 m eth (1966a, 1966b, 1966c), in two articles in Hungarian and one in G erm an, added five new B ashkir-M agyar tribal nam e parallels ^ у ё к , Gyula, Ker, Keszi, M agyar) to P auler’s formerly accepted list o f two (G yarm at, Jeno). Based on these parallels he revised G om bocz’s latter theory, and suggested that the M agyar Urheimat was actually in the K am a-B elaya region (where G om bocz had initially put it), and that the M agyars did not becom e politically involved w ith the K hazar Em pire until rather late, “towards 800” . N 6 meth (1972, pp. 70 -7 4 ) further elaborated his theory, and included it in his m ono graph about Gom bocz. N em eth died in 1976, w ith an unpublished m onograph in his m anuscript legacy. N em eth bequeathed m e this, as well as his other unpublished m anuscripts and notes. These are available today at the library o f the D epartm ent o f Altaistics o f the Attila Jozsef U niversity in Szeged, Hungary. W ritten about the form ation o f the M agyars, the m onograph is an alm ost com pletely revised version o f N em eth’s original 1930 work. It was finally published, with great care and scrupulous philological attention, by my
An overview o f [he study o f ancient Hungarian history
399
pupil, A rpad Berta, in 1991 (see N em eth 1991, on the B ashkir-H ungarian tribal nam es, pp. 182-215; on Nemeth, see Rona-Tas 1990c). The reactions to Gyula N 6 m eth’s views were mixed. A form er student o f his, Laszlo Rasonyi (1 9 6 4 ,1976b), set out enthusiastically to find M agyar toponym s in Bashkiria. Unfortunately his results cannot be accepted. № m e th ’s younger students clearly saw the im mense problem s related to parallels o f tribal and place nam es. Ligeti (1964b), for instance, had already expressed his doubts about the tribal nam es G yarm at and Jeno. V asary (1976, 1985) and M andoky-K ongur (1976, 1986, 1988) subjected the problem s to careful scru tiny. 1 personally discussed two different aspects o f the Bashkir question (Rona-Tas 1982c, 1987c). Berta (1989, 1990a, 1990b, 1991, 1992a, 1992b) elaborated a new system for the explanation o f M agyar tribal names. Istvan F o d o r’s w ork brought a breath o f fresh air to research. Until the end o f the Second World War, Hungarian scholars had been barred from entering archaeological sites in the Soviet Union (save one forgettable visit to the Ukraine), and had consequently based their research on the works o f Talgren (1914, 1919). However, following the Second World War, Soviet periodicals and various other publications becam e more w idely available, and as from the latter h alf o f the 1950s Hungarian archaeologists w ere allowed to travel to the Soviet Union to explore the latest finds. Istvan F odor’s activity was outstand ing in this respect. Setting out from Khazan in 1967 (alw ays collecting local publications) he focused on sites where Finno-U grian finds w ere thought to have been excavated. In the Soviet U nion archaeological w ork was associated with the construction o f the hydroelectric pow er plants on the R iver Volga. A fraction o f the im m ense construction budget (large sums o f money, even so) was allocated to archaeological excavations. W hat this m eant in term s o f the research o f ancient Hungarian history was that substantial am ounts (on an international level) w ent to the archaeological research o f peoples historically related to the M agyars. Khazan archaeologists greatly benefited from this opportunity and launched significant archaeological excavation schemes through out the vast region from the Urals to the right bank o f the Volga. It was partly these new findings that Istvan Fodor (1973) published in his survey entitled Vazlatok a fin n u g o r ostortenet regeszetebol [Sketches from the archaeology o f Finno-Ugrian prehistory]. Published in 1975, F odor’s m onograph (Fodor 1975; in German: 1982a; in English: 1982b) aim ed at giving an overall idea about the origins and m igrations o f the M agyars, and the Conquest. The chapters on the ancient periods in this book are especially enlightening. Two archaeologists, B ela Kurti and C sanad Balint, published an excellent survey o f the archaeological cultures thought to be related to the Finno-Ugrians, the M agyars and the other steppe peoples they m ight have m aintained connections with (H ajdu-K risto-R ona-T as 1976, vol. I, ch. 1).
400
Recent research and studies
Istvan Fodor propounded his historical concept in an academ ic degree thesis w hich he defended in 1982. Fodor claim ed that the M agyars w andered from their W est Siberian Urheimat to the K am a region in the 5th century. Leaving behind a small group, the greater part o f the M agyars m igrated from this region in the early 8 th century. Being one o f the official exam iners o f F o d o r’s thesis, I com m ented (Rona-Tas 1988e, pp. 153-165) on the excellent qualities o f his work, the fresh findings he had operated with, his new approach and high scientific standards; but took issue w ith F o d o r’s arguments. Fodor had constructed his paper partly on Istvan D ienes’s (1963) reasoning, w hich claim ed that parallels o f the funeral shrouds found in C onquest-period graves can only be found am ong the Ob-U grians. Fodor held that the use o f such shrouds gained ground in the K am a region in the 8 th century am id the Perm ian peoples (am ong others), w hich was w here the M agyars had brought them from to the C arpathian Basin. N ot counting the chronological problem s o f the issue, one cannot fail to acknow ledge the inter-ethnic character and the spread o f these shrouds. We do not know, however, w hich ethnic group or joined people used them in H ungary (see M. B enko 1992/93). The finds o f the graves excavated by E.A. H alikova at B olshie Tigani and outside Tankeyevka (see Halikova 1976, 1978; H alikova-K azakov 1977; H alikova-H alikov 1981) have a fair num ber o f H ungarian parallels, and Fodor (1977), too, holds that these locations w ere the burial sites o f the M agyar ethnic group. Yet by 1977, Fodor realised that— for chronological reasons— the cem etery cannot have been the last burial site o f the M agyars before they set o ff w estwards. Rather, it m ust have served those that stayed behind, on account o f the fact that the Tankeyevka cem etery was first used in the 9th century, and the sam e population continued to bury its dead there w ell into the 11th century. However, Fodor did not give consideration to the possibility o f M agyars (previously a satellite people o f the Volga Bulghars), who had at that tim e arrived from the south, being buried there. In m y com m ents I gave a detailed exposition, arguing that the num ism atic m aterial unearthed in these sites attests to the fact that the Volga B ulghars m ust have reached the Bolshie Tarhani area (which is where the earliest Bulghar-Turkic finds w ere exca vated) in the late 8 th century (see G ening-H alikov 1964), and only around 900 did they reach the K am a region (Sm irnov 1951; Fahrutdinov 1975; Kazakov -S taro stin -H alik o v 1987). I f M agyar-B ulghar-T urkic cohabitation took place in the V olga-K am a region, F odor’s chronology w ould be ju st too short— on account o f the fact that M agyar-T urkic interaction cannot have begun in the latter h a lf o f the 8 th century. The great influence on the M agyars’ language and culture w ould seem to contradict this, however. But assum ing that a group o f the M agyars, a satellite people o f the Volga B ulghars, m arched north, fits the picture.
An overview o f die study o f ancient H ungarian history
401
Fodor refers to an article of Karoly Czegledy (1976). Unfortunately, the study in question offers two arguments only: the above-mentioned Bashkir theory of Gyula №meth, and Fodor’s viewpoint. The year 1976 marked the beginning of the publication of a five-volume series of university textbooks (Hajdu-Kristo-Rona-Tas 1976, 1977, 1980, 1982), edited by Peter Hajdu, Gyula Kristo and myself. In these volumes we gave a broad meaning to the term “source material”, so as to include all knowledge and data which we felt had direct or indirect bearing, as a source, on the ancient history of the Magyars and their language. We also admitted anything indispensable for the evaluation of these sources. Half of the first volume is devoted to the archaeological, archaeozoological, anthropological and ethnographic sources, and half to the linguistic and written sources. The second volume contains illustrations of the archaeological sources, and the indices. The third volume discusses numismatics, and the different systems of writing and of chronology. The fourth volume is devoted to the material and methods of the natural sciences—hence chronology and age-dating, historical geography, climatology, phytogeography, zoogeography, soil sci ence, geophysics, the map sources, and a small lexicon of terminology. We supplied every volume with a bibliography to facilitate farther research. In retrospect, it must be said of this giant work, which extended to almost every branch of science and involved the work of some sixty Hungarian scholars and scientists, that its standards are uneven, and it cannot be denied that some parts of it were written in haste. Notwithstanding these shortcomings, the series was soon to become an invaluable study tool and research reference, and it was widely used in Hungary and abroad in the teaching of Hungarological studies. Concurrent with the production of this handbook, the University of Szeged offered students courses in ancient Hungarian history, and the Csoma Korosi Society organised lectures in proto-history. My lectures were collected in manuscript form by the Csoma Korosi Society (Rona-Tas 1979). In them, I give a detailed interpretation of the concept of “source material”, an account of research methods, of the history of the natural geography of Eurasia, of the designations of the Magyars, their formation and migrations. I also discuss my views regarding the southern origin of the Magyars found in Magna Hungaria, in the Volga-Kama region. In 1973, we organised a national conference on the issues of Hungarian proto-historical research. The volume entitled Magyar ostdrteneti tanulmanyok [Studies in Hungarian proto-history] (Bartha-Czegledy-Rona-Tas 1977) contains the talks given at this conference, and a handful of other studies. The book sought to give an overview of the state of research. In my article on the character of Magyar-Bulghar-Turkic interactions I examined the unique
402
Recent research and studies
phenom enon w hereby the greater part o f H ungarian words o f Turkic origin feature in the M ongolian language, too. I sought to give an explanation, and at the same tim e I outlined the highly debated issues o f the relationship am ong the Altaic languages (Rona-Tas 1977b). Referred to as the “silver book” , on account o f the colour o f its cover, this book reports on the 1973 sym posium (Ecsedy 1977) and on the work o f the Szeged Proto-H istory R esearch Group (see Rona-Tas 1977a). I shall highlight only one article here, about a field which has seen much progress recently. The origins o f the m elodies o f Hungarian folk m usic have long been the focus o f research. Bela Bartok and Zoltan Kodaly were particularly interested in these issues. Bartok came to know some Cherem is folk songs w hich Yrjo W ichm ann’s Hungarian wife had collected back in 1906. B artok im m ediately recognised the sim ilarity with the Hungarian pentatonic “fifth-shifted” m e lodic construction. K odaly becam e acquainted with a further 311 Cherem is folk songs, published by V. M. Vasilyev in 1919, which, together with the Chuvash folk songs collected by Lach and M aximov, he analysed in 1937. This vast stock o f folk songs revealed that one o f the oldest strata o f Hungarian folk m usic displayed profound sim ilarity with Turkic, and especially Chuvash folk music. However, Kodaly (1937) found the closest parallel with Cheremis m elodies on the basis o f the data available to him at the time. In 1947, Kodaly (1947) published a small booklet containing Chuvash folk songs, claiming that the only explanation for the rem arkable M agyar-C huvash sim ilarities was that the forebears o f the M agyars and the Chuvash, the B ulghar-Turkic peoples m ust have lived together prolongedly in the Caucasus region. Kodaly w as fully aware that his results were prelim inary, so in 1958 he sent a student o f his, Laszlo Vikar, to the Volga region to collect folk m usic am ong the Finno-U grian and Turkic population. V ikar’s w ork was m ost successful. To gether w ith a num ber o f articles, he published m onographs on Cherem is (V ikar-B ereczky 1971), Chuvash (V ikar-B ereczky 1979) and Votyak (V ikarBereczky 1989) folk music respectively. V ikar’s on-location research persuaded him that the correspondences be tween Hungarian and Cherem is folk m usic were secondary. The m usic o f the so-called “M ountain C herem is” groups stood closest to H ungarian, and the farther north or east he went, the lesser the sim ilarities were. The South Cherem is groups are under a strong Chuvash influence, and the C herem is folk m elodies w hich Vikar had known before his visit (southern ones) had acquired their traits calling Hungarian folk m usic to m ind as a result o f this Chuvash influence. Thus, the only argum ent supporting Hungarian folk m usic’s FinnoUgrian origin was rejected. At the sam e time, Vikar faced a unique, deceptive duality. How was it possible that while the Hungarian language was of Finno-U grian origin, Hungarian m usic displayed Turkic characteristics? In
An overview o f the study o f ancient Hungarian history
403
effect, K odaly him self had given the answ er in the epilogue o f his 1947 publication o f folk melodies. The Hungarian m elodic w orld was transform ed under Turkic influence, sim ilarly to m any areas o f the M agyars’ animal husbandry, agriculture, spiritual and social culture. I appraised V ikar’s book on C huvash folk songs with regard to proto-history in an article (Rona-Tas 1980c). I reckoned that it was rem arkable that, o f the C herem is m elodies, the South Cherem is ones were m ore closely related to Hungarian m usic, and that o f the C huvash m elodies those that were sung in the northern regions o f the Chuvash language area, that is, near the border between the tw o groups, also featured close similarities. We are witnessing, then, a m elodic world that em erged from a Finno-U grian-Turkic interaction by secondary processes. 1 called Gyula L aszlo’s attention to this, knowing that he was looking out for arguments in support o f his own theory o f num erous dualities in ancient M agyar culture. Alas, to no avail. Vikar continued his research and in 1988 he defended his doctoral thesis on the m usic o f the Finno-Ugrian and Turkic peoples o f the V olga-K am a region (see Vikar 1993). His m onograph on the Tatar folk m usic o f Khazan will appear in 1999. V ikar’s work is closely connected with Janos S ipos’s (1994, 1994-1995) recent collections o f folk music in Turkey and Khazakh-
stan. Pal Liptak wrote a survey o f physical anthropological research in the Szeged textbook and in the “silver book” . Liptak investigated the historic aspects o f the types which he had established by classifying bone dim ensions, skull shape and proportions. In his m ost recent study he sought to place these in a proto-historic context, and to establish connections betw een the bone m atter o f the Carpathian Basin and that o f Eastern Europe. He discussed the issue o f the bones o f the leading M agyar strata differing from laym en’s bones (for the principles and m ethodology, see Liptak 1971; see also Liptak 1977, 1983). The researches o f Liptak and his school are highly limited by the paucity o f C onquest-period skeletal finds. A mere 199 sets o f m ale bones, and 154 female ones have been identified at the grave sites. Recently K inga Ery (1983, 1994, 1995) gave an overview o f the state o f research. Nemeskeri and his colleagues had long included serology in their research in physical anthropology. In 1990, however, they took a com pletely new approach. In some o f his earlier articles, in M agyar Tudomany [Hungarian Science], as well as at the 1990 Finno-Ugrian Congress in Debrecen, Tamas Tauszik (1990a, 1990b) proposed a new theory. He based his research on the m ethodology developed by Hideo M atsum oto (1988; for further bibliography, see also Tauszik 1990a, 1990b). This m ethod is founded on the 28 m arkers identified in the gam m a-globulin com ponent o f blood. M atsum oto not only developed a technique for identifying these m arkers, but also sought to group
404
Recent research and studies
them. He found that the co-occurrence o f certain m arkers was characteristic o f the so-called “greater” races, that is, the M ongoloid, the Europoid and the Africanoid anthropological subspecies. Subjecting to close scrutiny the oc currence and geographical distribution o f m arker groups, M atsum oto was keen to draw inferences regarding the origins o f the Japanese. His laboratory, however, was international, and processed blood sam ples from Hungary, too. Tauszik’s new approach caused quite a stir. In 1 9 8 9 ,1 visited M atsum oto in Osaka, and thoroughly discussed the historical applicability o f his data and m ethods. Follow ing m y visit I w rote a sum m ary o f the debate about the m ethod, and expressed m y ow n view s on the issue (Rona-Tas 1990a, 1990b). In m y opinion, the exam ined synchronous data w ere unserviceable for the straightforw ard explanation o f historical processes. M atsum oto held that the M ongoloid race had its origins in the northern B aikal area, from w here it spread throughout Eurasia. He based this conclusion on the study o f the blood sam ples o f the B uryat people who live north o f Lake Baikal. However, these Buryats have only been living in this area since the 16th century. M atsum oto term ed one specific group ‘U ralian’ on account o f the fact that the blood sam ples had com e from a hospital in the Ural M ountains. H ungarian scholars next tried to draw inferences regarding the Uralic peoples. Tauszik observed that a 5% group o f the M agyars, w idely distributed throughout the country, featured the traits o f the M ongoloid subspecies. One h a lf o f this group shared the features o f the N orth M ongoloid group, the other o f the South M ongoloid. We do not know, however, how this 5% (or rather, 2 x 2.5% ) M ongoloid group m ingled w ith the M agyars. Did it happen at once or in several “instalm ents” ; did a higher ratio reduce gradually to 5%, or w as it a sm aller ratio that gradually increased to the current level? A lthough the m ethod has proved unsuccessful (as yet), scholars should continually update their know ledge o f the results o f the natural sciences. F or instance, great progress has been made in the tests determ ining the blood properties o f a deceased from fossil bones (see, for instance Lengyel 1975). O n 27th April 1983 Iren Juhasz, archaeologist o f the M useum o f Szarvas, brought to light a bone needle-case at an Avar cem etery near Szarvas. The case had strange runiform writing on it. Juhasz (1983, 1985) published her find in the 1983 volum e o f A cta Archaeologica. The find dates back to the Late Avar period, and Juhasz holds it to be a late-8th-century work. Two things were clear from the outset: the w riting was quasi-identical w ith the runiform inscription on the Nagyszentm iklos treasure; and there w as a possibility o f the find contributing to solving the riddle o f the origin o f the Szekely runiform script, hence answ ering a num ber o f questions related to ancient M agyar history. A rchaeology had long known runiform inscriptions dating from the Avar era (for an overview and a bibliography, see V asary 1972); this inscrip-
An overview o f the study o f ancient Hungarian history
405
tion, however, featured at least 59 individual symbols. The next task was to prepare a thorough palaeographic publication. C om m issioned by Juhasz, I subjected the find to a series o f m icroscopic tests, and w ith the help o f some good photographs I published the inscription (Rona-Tas 1985a, see Figure 22 on p. 130). Several attem pts were m ade to decipher the text (by Janos Harm atta, G abor V ekony and myself), but the truth is that none o f these attem pts w ere convincing enough. It seem s that deciphering this inscription will have to w ait for som e time. However, a whole series o f objects with short runiform inscriptions came to light, m ost o f which were published by Janos Harm atta (1983a-c, 1984a, 1984b, 1985a, 1985b, 1988, 1992c). G abor Уёкопу (1987a, 1987b) had a go at deciphering som e o f these inscriptions. In 1992 in Szeged, we organised a small conference on the runiform inscriptions o f the C arpathian Basin (see Sandor 1992b), at which two new inscriptions w ere presented. Iren Juhasz (in Sandor 1992b, pp. 15-19) found a new, shorter inscription at this sam e site o f Szarvas. Istvan Dienes established that the inscription on the flap o f a bone quiver unearthed at the excavation at H om okm egy-H alom (near Kalocsa, Hungary) was w ritten during the Conquest period (Dienes 1992,1994); Gabor Уёкопу (in Sandor 1992b, pp. 4 1-49) m ade an attem pt to read it in Turkic. Even if there is no consensus betw een scholars concerning the linguistic aspects o f the inscriptions, there is unanim ous agreem ent about the signifi cance o f the new finds w ith regard to the history o f Hungarian literacy. Scholars felt that the research o f the history and form ation o f Central European m ediaeval scripts had com e to a turning point. This called for the reconsid eration o f the origins o f Hungarian literacy (see Rona-Tas in Sandor 1992b, pp. 4 1 -49), and placing the Nagyszentm iklos and Szarvas w ritings in a broader, Eastern European historic context (Rona-Tas 1988b). The classification o f Eastern European runiform scripts, as well as the new finds, necessitated the analysis o f the system and origins o f the Turk runiform inscriptions o f Central Asia. Scholars were able to determ ine the phases o f the formation o f Turk runiform writing, and established the Sem itic origins o f its very first forms (Rona-Tas 1987b). The new system atisation o f C entral Asian Turk runiform inscriptions and the discovery o f new sam ples substantially changed the situation o f research. M ention m ust be m ade o f the works o f D .D . Vasilev (1983a, 1983b), I.L. K yzlasov (1990) and I.V. K orm usin (1975), as well as S.G. K lyashtom yj who discovered m any new Turkic inscriptions in M ongolia, the m ost im portant am ong them being those o f Bugut, Tez and Terh (see Rona-Tas 1996d). Given these circum stances, keen interest in the Szekely runiform script developed internationally. Istvan V asary (1974) gave an excellent survey o f the state o f the art which served as a very good starting point. The 8гёке1у
406
Recent research and studies
runiform writing reached its first peak in the 15th century at the court o f the Hungarian King M atthias I. N ew results have been published on the earliest occurrence o f this script, dating back to the M atthias era (Rona-Tas 1985, 1986). Klara Sandor gave a new and thorough philological analysis o f the longest Hungarian runiform inscription, now kept in Bologna, in a m onograph pub lished by the Szeged w orkshop (Sandor 1991). Geza Ferenczi (1992) gave an overview o f the research o f the old and latest discovered w ritings at a sym posium in Szeged in 1992, and the new trends o f research w ere sum m a rised in studies by Ferenc Kosa (1992) and Klara Sandor (1992a). Because first m ention o f the Szdkely runiform w riting occurs in the H ungarian K ezai’s C hronicle only, w e came to the conclusion that the tim e span separating the Szekely runiform script and earlier scripts had significantly dim inished, and we saw hope o f a turning point in the research o f the origins o f the Sz6kely w riting in the near future. Little headw ay has been made in the study o f the w ritten sources. It is a w ell-know n fact that even at the tim e it was published, in 1900, A m agyar honfoglalas kiitfoi [The sources on the M agyar Conquest] was not quite up to par, scientifically speaking. The situation is best w ith the B yzantine sources. M oravcsik’s fundam ental w ork o f 1943 saw tw o new editions in 1958 and 1983 (see M oravcsik 1983). The new edition o f Constantine Porphyrogenitu s’s work published in collaboration with Jenkins (see M oravcsik-Jenkins 1967), is a significant international scholarly work. M oravcsik’s other works, as w ell as his Introduction contain m any novelties (M oravcsik 1976). In 1984, his students published a critical edition o f B yzantine sources relevant to the ancient history o f the M agyars (M oravcsik 1984). Follow ing the death o f M oravcsik (in 1972), Samu Szadeczky-Kardoss, Ferenc M akk and Terez Olajos continued to explore the Byzantine sources. The investigation o f the Hungarian crown and its inscriptions presents a unique but im portant chapter in the research o f the Greek sources. M ention m ust be m ade here o f Gyula K risto’s (1980) m onograph in which, based on Constantine Porphyrogenitus, he gives a survey o f the ancient history o f the M agyars from the 830s until the foundation o f state in 1001. K risto set 830, the year in w hich the Fortress at Sharkel was constructed, as his starting point; this and his discussion o f the events leading up to that date rest on the research o f Fodor and Czegtedy. It has been suggested that the Fortress at Sharkel was built against the M agyars; however, there is little evidence to support this. C ertainly the w ell-inform ed Byzantine sources do not contain such an assertion. The claim is rooted in an erroneous interpretation o f one o f Ibn R u sta’s descriptions. I com m ented on the rectification o f this error in a
An overview o f the study o f ancient Hungarian history
407
talk, and my pupil, Istvan Zimonyi, discussed the matter in detail (see Kristo 1995d, p. 34, note 50). While the research of the Byzantine sources is a very active field indeed, the same cannot be said for the Latin sources from outside Hungary. Since Gombos (1937-1943) published his practical overview, source research has made almost no headway whatsoever. Peter Kiraly’s (1987) compilation is an exception. He collected a group of personal names like Ungarus, Hungaer, etc. which can be found in the 8th—9th centuries, in the pre-Conquest-period western monastery accounts and charters. Kiraly, probably with good reason, reckons that some of these may have a bearing on the homeland of the Pannonian Turkic Onoghurs. A very small team devotes most of its time to researching the Latin sources of Hungary. This does not mean, however, that no works were written about the Latin sources from outside Hungary that might modify current knowledge. In 1984, I gave a talk at the Rhine-West phalia German Academy in which I made an attempt to reconsider two sources (Rona-Tas 1988a, pp. 275-310). One of them was the letter written prior to 923 which derives the Hungarians’ designation from the Old High German Hungar which, similarly to English, means ‘hunger’. The other is a source dated 1071 which speaks of a gift presented by Anastasia (the widow of the Hungarian King Andrew I) to the Bavarian Prince Otto. The queen refers to this gift as ‘the sword of Attila’. Naturally, I was neither the first nor the last to consider the matter. We owe many excellent studies in the field of source processing and interpretation to Tamas Bogyai, Szabolcs Vajay, Imre Boba and, of course, Gyorgy Gyorffy. Their works, however, fall outside the chronological limits of this survey. Little progress has been made in the publication of the Oriental (Arabic, Persian, Armenian, Georgian, Chinese, Tibetan and other) sources. In 1969, a research group was set up at the Department of Arabic Studies of the Lorand Eotvos University, initially to focus on the Arabic and Syrian sources. The team was extended to include experts of other fields, and in 1977 Karoly Czegledy, the director of the group, decided that the publication of their results could go ahead (see Czegledy 1977). Initially planned to consist of six volumes, the series was extended with volumes containing the Byzantine, Latin, Slavic, Tibetan, Chinese, Turk and Iranian sources, due to be published starting 1978. We know today that nothing came of this ambitious scheme. However, many works have appeared on these sources and their various aspects. Odon Schutz wrote about the Armenian sources, and Margit Biro of many of the Georgian ones. The 7th-10th-century Tibetan sources are espe cially important regarding background information. G iza Uray’s works in this field are of outstanding significance (see Rona-Tas 1992-1993, for Uray’s bibliography, see Steinkellner 1991, pp. xv-xxxiv). The Chinese sources were
408
Recent research and studies
treated in m any books and studies by Ildiko Ecsedy. M any Slavist scholars discussed the problem s o f the Cyril and the M ethodius legends (see Kiraly 1974; H. Toth 1981). Even the results o f the above-m entioned pieces o f research cannot com pen sate for the series which was never published. But that is only one side o f the problem . The tense circum stances around the publication o f the series hin dered further research and the continuation o f the scheme. Consequently, the first volum e o f M ihaly K m osko’s collected w orks w as published only in 1997 (see K m osko 1997, Zim onyi 1991). These survived in m anuscript form and w ere for a long tim e locked away. A G erm an edition by Istvan Zim onyi and H ansgerd Gockenjan is going soon to press. N othing m uch happened in the field o f publishing critical editions o f the foreign sources until the publication o f a highly popular book (in two editions), designed to m eet the general public’s long-felt w ant (Gyorffy 1975c— first edition 1958). One o f the oddities o f this book was, however, that Gyula N em eth’s nam e appears on the list o f contributors on the book title-page, although he is not represented with his source translations either in the first, or in the second edition; moreover, he did not write a single line in the book. The Turk inscriptions o f Orkhon were translated by Zsuzsa Kakuk, who also features on the title-page. Only one o f the rich and as yet unexplored body o f Islam ic sources has reached Hungarian researchers, nam ely a recently published source about the M agyar raids in Spain, w ritten by an 1lth-century author around 1075 (C hal m eta 1976, 1979). U nfortunately C halm eta’s edition is not a critical one; it lacks, for instance, the analysis o f the source, and neither does it discuss the peculiarities, or attem pt to reconstruct the antecedents o f the unique, 13 th-century Rabat m anuscript. The book does not give thought to the place nam es other than those related to the M editerranean; the spelling o f the author and the copyist was not considered— to mention but a few problem s which m ake the edition im possible to appraise w ith regard to the M agyars. C onse quently, the research did not in the least bit benefit from the debate (see Czegledy 1979, 1981; Gyorffy 1993a) provoked in H ungary by the data contained in the book. A nother Arabic work— edited by Dubler (1953) and also published by T. Lewicki (1938,1956) and Hrbek (1955a, b) with comm ents on its Hungarian aspects— m et an even stranger fate. Abu Ham id al-G am ati (bom in Granada) arrived in Hungary from the land o f the Volga Bulghars and spent three years am ong the Hungarians between 1150 and 1153. The H ungarian translation of his w ritings was based on the 1971 Russian edition. The foreword, the introductory study, the historical com m ents and the notes w ere translated from the Russian, and the Arabic text was translated by Tamas Ivanyi. The transla
An overview o f the study o f ancient Hungarian history
409
tion was collated with the original by Karoly Czegledy who also supplem ented the notes. However, nothing was contributed to the discussion o f text phi lological problem s (see B olsakov-M ongajt 1985). The bibliography lacks H rbek’s (1955b) article which appeared in A cta Orientalia Hung., although in his notes covering four and a h a lf pages Czegledy keeps referring to H rbek’s works. Evidently this work does not put an end to philological problem s. I made an attem pt to solve one such problem in connection with A bu H am id al-G am ati’s work (see Rona-Tas 1992b, pp. 221-226; in English: 1994b). Correctly analysed, the text sheds light on the M agyars’ designation ‘B ashkir’, and on the nam es the Volga Bulghars called the Volga M agyars. This brings us to one o f the key issues o f ancient Hungarian history. One difficulty with the Oriental sources is identifying the persons, events and places which appear in the different sources. Karoly Czegledy has made significant progress with his research works in this field. His articles written in Hungarian w ere republished in 1985 in a collected edition (Czegledy 1985). His book about the sources relating the early m igrations o f the nomadic peoples (C zegl 6 dy 1969) was translated into English by R G olden (Czegledy 1983a). There are three ways to approach ancient Hungarian history: from the beginnings, the contem poraries, and the ensuing periods. Taking a closer look at the ensuing periods— that is, the Conquest and the Arpad era follow ing the foundation o f state— we find that both the narrative sources (prim arily the Hungarian chronicles) and the deeds can offer us inform ation regarding the preceding, w ell-nigh “sourceless” period. Edited by Gyorgy G yorffy (1963— 1998), the first four volum es o f a series o f seven contain a w ealth o f inform a tion on the geographical sources. In the context o f the Hungarian language, many geographical and personal nam es preserve the phenom ena o f past ages. 9 th -10th-century M agyar tribe, clan and personal nam es can only be recon structed from the Hungarian deed sources, increasing in num ber from the 11th century on (K risto-M akk-Szegffi 1973-1974). A t the same time these deeds supply a background to the Hungarian narrative chronicles which, in turn, cannot be interpreted w ithout knowledge o f the related deeds and charters. The source value o f the m ediaeval Hungarian chronicles is a hotly debated issue, especially regarding those chronicles that are related to the C onquest and the 9th and 10th centuries. Following Csaba C sapodi’s (1978) sum m ary the debate about Anonym us (Gyorffy 1988) flared up again. Gyorffy recon sidered m ost o f his earlier articles and, having supplem ented or added epi logues to them, and indicated the parts affected by his changes o f view, he published a collected edition o f them (Gyorffy 1990, 1993b). Unfortunately, the Turkological parts are, in m any respects, outdated. Gyorffy sum m arised
410
Recent research and studies
his view s on proto-history in his preface to the second edition o f Gyorffy (1975c). This survey w ould not be com plete w ithout m ention o f four w orks which consider the history o f the word-stock o f the Hungarian language. These are A magyar nyelv torteneti-etimologiai szotara [A historical-etym ological dic tionary o f the Hungarian language] (Benko 1967-1984), A Magyar szdkeszlet finnugor elemei [The Finno-Ugrian elem ents o f the Hungarian vocabulary] (L a k o -R id e i 1967-1978), the Uralisches etymologisches Worterbuch (Redei 1988-1994) and the Etymologisches Worterbuch des Ungarischen (Benko, L. 1992-1995). The Turkic sections o f the first w ere com piled by Zsuzsa Kakuk and revised by Lajos Ligeti; in the other three the Turkic, as w ell as the M ongolian and M anchu connections, were com piled by the editors and revised by myself. The editors often followed m y suggestions in unresolved questions; at other times, quite naturally, they w ent along with their own ideas. This publication w ork and the rapid progress o f Turkological research necessitate the com pilation o f a m onograph on the Turkic loan w ords in the Hungarian language. Lajos L igeti’s last w ork deserves a special place in the research o f ancient H ungarian history (Ligeti 1986). The book is difficult to read for scholars unfam iliar with Turkology or the history o f Central Asia. M eticulously elaborated chapters which sum m arise the work o f a lifetim e intertw ine with extem poraneous ideas which came to the author’s m ind when w riting the book— all o f this m akes it hard to understand for the uninitiated. Yet once anyone sets foot in this field, L igeti’s m onograph sim ply cannot be avoided. Since the publication o f the Hungarian original o f m y book in 1996 many works appeared on the origin and the early history o f the M agyars. M ost o f them will rem ain inaccessible to the non-H ungarian reader. I shall mention here therefore only three works. The dissertation published in Italian o f Di Cave (1995) gives a good overview on the Hungarian research. The m ono graph o f K risto (1996) is not a translation o f one o f his works but a new one. For those who w ould get an introduction into his views this w ork can be recom m ended. Less fortunate is the book o f Laszlo (1996), w hich was originally addressed to the Hungarian reader and the, otherw ise excellent, translation could not help to m ake fam iliar the reader w ith the context o f the problem s. It has a second part titled Reader on the Magyars o f the ConquestPeriod Hungary on pages 173-363 w hich contains English translation of sources, o f scholarly and less scholarly literature, and even o f poem s. The extracts are published w ithout any com m ent or context. For those who are interested in the rich w orld o f the recently deceased scholar, w ho m ade a great im pact on Hungarian scholarship and way o f thinking, m ay read this selection with interest.
An overview o f the study o f ancient Hungarian history
411
NOTES On the theory o f a double conquest, see my remarks in Rona-Tas (1980d). I would also like to call attention in this context to a paper which was prepared for the Seminar o f the Department o f Altaic Studies at the University o f Szeged (Madaras 1975). On the change o f G om bocz’s opinion, see my remarks in R 6 na-Tas (1 9 9 Id, pp. 199-205). In connection with the Hungarian H oly Crown, let me note that the US Government gave it back to Hungary in 1979. On the materials o f a conference on the Crown, which follow ed this event, see S tudien z u r M a c h tsy m b o lik ... (1983).
Xill. THE LEVEDI QUESTION AND THE EARLIEST HUNGARIAN CHRONICLE
One of the most widely discussed issues of the ancient history of the Magyars is the question of when the Magyars arrived in their last Urheimat before moving to the Carpathian Basin. As I have pointed out above (pp. 215-219, 325-326), this must have happened around 670 when the greater part of the Bulghars moved to the west and a smaller part evaded to the north after having been defeated by the Khazars and the Magyars. The Magyars occupied the region formerly ruled by the Bulghars between the Dnieper and the Lower Danube which was later called Etelkoz in Hungarian, that is Mesopotamia, the “Region between the Rivers”. The word Etil or Etel is of Turkic origin, it had the meaning of‘river’, and in many cases ‘the River’, that is, the greatest river in the region. It used to be the Turkic name for the Volga, but other great rivers were also called Etil (e.g. the Dnieper) by peoples speaking Turkic dialects (see p. 434). The word was borrowed by Ancient Hungarian, and remained in use until the Middle Hungarian linguistic period, when it disappeared. The second part of the word -kiizu (or -koz in modern Hungarian) has the meaning of‘intermediate space’, and is an old Hungarian word of Finno-Ugrian origin. When the Magyars arrived in the Etelkoz is heavily debated. Most scholars consider a relatively late arrival and suggest the second half or end of the 9th century. When a significant, politically organised group of people moves to a new homeland, there are three feasible outcomes. (1) People В pushes people A out of its land. In this case people A turns up in a new place, and the sources record its migration. (2) People A moves (for whatever reason) and evacuates the region which is then occupied by the newcomer, people B. In this case, too, the sources tell us something about the migration of people A. The causality, albeit different in the two cases, may appear more puzzling to the contemporary chroniclers than to the modern-day historian. (3) People В occupies the territory inhabited and ruled by people A. The latter does not move away, but remains under the suzerainty of people B. In such cases peo ple A disappears from the sources altogether. Of course, these examples are simplified, and there must have existed mixed cases when one part of peo ple A moved away and another part remained under the rule of the newcomer
414
Recent research and studies
people B. For instance, this happened with the Bulghars (see p. 220). It is, however, very im portant that betw een 670 and the M agyar C onquest the otherw ise well-inform ed Byzantine sources do not m ention w estw ard move o f any people, and neither do the sim ilarly w ell-inform ed Latin chronicles. If the M agyars had arrived in the Etelkoz in the second h alf o f the 9th century, they w ould have to have driven out the people living there, or filled up the locality o f the people who left, or conquered people w ho rem ained here. None o f these three alternatives is reported in any o f the sources. The ancestors of the later Volga Bulghars m oved further northw ard after 737 (see pp. 220-227). Therefore, we have only two options left. (I ) The M agyars occupied the land o f the B ulghars between 670 and 680. One part o f the B ulghars moved westward, and was to become the Danube Bulghar group, or those Bulghars who joined the Avars, or the group that m oved to Dalm atia and even Italy; the other part o f the Bulghars fled to the northeast, and later m igrated further to the V olga-K am a region. (2) The future Volga Bulghars rem ained in their old hom eland, came under K hazar authority, and the M agyars occupied their place only after the Bulghars had m oved to the northeast. This w ould m ean that the M agyars occupied the Etelkoz in the 740s. The first scenario seems to be more plausible, but the second cannot be ruled out either. In any case, the arrival of the M agyars in the Etelkoz m ust be dated m uch earlier than was previously suggested. The S lavs’ designation o f the M agyars prior to the C onquest was Ungri (see pp. 282-287) which is a derivation o f the nam e Onoghur-{Bul ghar). This could only have been possible in the Etelkoz. Two hundred years elapsing betw een the O noghurs’ m igrating w est in 670 and the arrival o f the M agyars would m ake this designation inconceivable. The linguistic analysis o f early Turkic loan words in the Hungarian language adds to the arguments supporting the early arrival o f the M agyars in the Etelkoz. The only source hitherto cited in favour o f the late occupation o f the Etelkoz is the work o f Porphyrogenitus, which again brings us to this frequently dis cussed report and its sources. The question arises: w hat w ere the sources of the report about the M agyars? Different ones, evidently. Let us take a look at one o f them. The Hungarian prim ary chronicle or gesta was w ritten some 270 years after the Conquest. It is certain, however, that at the court o f the princes and later o f the kings, the traditions were known even before the first w ritten version appeared. At the royal courts chroniclers w ould perform the history o f the royal clan. The perform er usually relied on his m em ory when giving his rendering o f the historical traditions, but it is equally possible that he had recourse to some written source, such as a drawing, a rune, or a colourful, canonical version o f the history. We know o f m any such stories from the Eurasian peoples. The epic traditions o f the Turks and M ongols can offer some
The Levedi question and the earliest Hungarian chronicle
415
inform ative parallels regarding the passing down o f such historical traditions. Many sim ilar traditional texts have been recorded and analysed, from the famous 13th-century Secret History o f the M ongols to 20th-century Bashkirian Shejeres or clan legends. The 1 lth-century author(s) o f the first H istory o f the H ungarians (Gesta Hungarorum), w ritten under Bela I (1060-1063) or Solom on (1063-1074), must have drawn on several sources. Such sources included the western traditions o f chronicle-writing. These m ight have served as exam ples, since they too related stories about kings, and served to historically legitim ise the power o f royal clans or archbishoprics. The European chronicles contem po rary to the Hungarian ones were packed with samples o f reasoning which the Hungarian authors m ight easily have used to justify the pow er o f their own rulers. Thus the western sources gave the Hungarians some idea o f how a chronicle should be written and what elem ents it should or should not include. The raw m aterial, which next had to be slotted into the “ready-m ade” formula, was taken from the living, old Hungarian nobiliary and royal traditions. The chroniclers sought, though not always successfully, to weed out all elements concerning the M agyars’ pagan past, as well as anything else that did not fit the new European model. We know very little o f the historical traditions relating the times before and after the Conquest. A num ber o f parallels, however, give us some idea o f what they m ust have been like. Thus, for instance, The Secret History o fth e M ongols gives a description o f how Chingis khan defeated the Tangut people. It then continues: “After he had plundered the Tangut people and, m aking Iluku Burkan change his nam e to Shidurku (the Subdued), had done away with him; and after he had exterm inated the Tangut people’s m others and fathers down to the offspring o f their offspring, maiming and tam ing Chingis khaghan gave the follow ing order: »W hile I take the meals, you m ust talk about the killing and destruction o f the Tangut, and say ‘M aim ed and tamed, they are no m ore’«” (from C hapter 268, translated by Igor de Rachewiltz). It is historical fact that the Tanguts, “whose every member had been killed”, shortly afterwards rebelled against the khan, and it was in the second w ar against the Tanguts that Chingis eventually died. H ow ever, the bard or chronicler had to perform the deeds o f Chingis khan during a sumptuous repast. The H ungarian Anonym us, too, talks o f the glorious deeds being recited during meals (Chapter 22). The H istory o f the Hungarians has not come down to us, but we have a number o f clues that tell much about its character. The tradition which Termecsu and Bulcsu recounted to the Byzantine court m ust undoubtedly have originated from a sim ilar royal court history. Such as, for instance, the Levedi story. Here is the story in full:
416
Recent research and studies
From D e adm inistrando imperio, Chapter 38: O f the genealogy o f the nation o f the Turks, an d w hence they are descended. “The nation o f the T u rb had their old dw elling next to K hazaria, in the place called Levedia after the nam e o f their first voivode, w hich voivode was called by the personal name o f Levedi, but in virtue o f his rank was entitled voivode, as have been the rest after him. N ow in this place, the aforesaid Levedia, runs the R iver Chidm as [Hidmas], also called Chingilous [Hingilus]. They w ere not called Turks at that time, but had the nam e ‘Savarti A sfali’, for som e reason or other. The Turks were seven clans, and they had never had over them a prince, either native or foreign, but there w ere am ong them ‘voivodes’, o f whom the first voivode was the above-m entioned Levedi. They lived together with the Khazars for three years, and fought in alliance w ith the K hazars in all their wars. Because o f their courage and their alliance, the khaghan-prince o f Khazaria gave in m arriage to the first voivode o f the Turks, called Levedi, a noble K hazar lady, because o f the fame of^iis valour and the illustriousness o f his race, so that she m ight have children by him ; but, as it fell out, this Levedi had no children by this sam e K hazar lady. Now, the Pechenegs who w ere previously called ‘K angar’ (for this ‘K angar’ was a name signifying nobility and valour am ong them), these, then, stirred up w ar against the Khazars and, being defeated, were forced to quit their ow n land and to settle in that o f the Turks. A nd when battle was joined betw een the Turks and the Pechenegs who w ere at that tim e called ‘K angar’, the arm y o f the Turks was defeated and split into two parts. One part w ent eastw ards and settled in the region o f Persia, and they are to this day called by the ancient denom ination o f the Turks ‘Savarti A sfali’; but the other part, together with their voivode and ch ief Levedi, settled in the w estern region, in places called Etelkoz, in w hich places the nation o f the Pechenegs now lives. A short w hile afterwards that khaghan-prince o f K hazaria sent a m essage to the Turks, requiring that Levedi, their first voivode, should be sent to him. Levedi, therefore, cam e to the khaghan o f Khazaria and asked the reason w hy he had sent for him to come to him. The khaghan said to him: »We have invited you upon this account, in order that, since you are noble and wise and valorous and first am ong the Turks, we m ay appoint you prince o f your nation, and you m ay be obedient to our word and our com m and.« But he, in reply, m ade answ er to the khaghan: »Your regard and purpose for m e I highly esteem and express to you suitable thanks, but since I am not strong enough for this rule, I cannot obey you; on the other hand, however, there is a voivode other than me, called A lm outzis [Alm os], and he has a son called Arpad: let one o f these, rather, either that A lm outzis, or his son Arpad, be m ade prince, and be obedient to your word.« T hat khaghan w as pleased at this saying, and gave some o f his m en to go with him, and sent them to the Turks, and after they had talked the m atter over
The Levedi question and die earliesr. Hungarian chronicle
417
w ith the Turks, the Turks preferred that Arpad should be prince rather than Alm outzis, his father, for he was o f superior parts and greatly adm ired for wisdom and counsel and valour, and capable o f this rule; and so they m ade him prince according to the custom, or ‘zakanon’, o f the Khazars, by lifting him upon a shield. Before this Arpad the Turks had never at any tim e had any other prince, and so even to this day the prince o f Turkey [i.e. Hungary] is from his family. Some years later, the Pechenegs fell upon the Turks and drove them out with their prince A rpad.” (Translated by Jenkins.) This story has a strong elem ent o f truth to it. Shortly before the Conquest a change o f dynasty occurred am ong the M agyars. The original leading clan, Levedi’s, was superseded by the new prince A lm os’s son Arpad. This histori cal fact requires some explanation, however. The appearance o f a new dynasty always brought about a crisis o f legitimacy. The new ruler— who had come to pow er by force or possibly with the help o f allies— needed to explain what happened to the previous clan, why its rule had com e to an end, and on what grounds had he put forward his claim to the throne. At that time, the legitim acy o f pow er in the steppe m eant being recognised by the K hazars. W hat the story o f Levedi tells us is that Levedi relinquished his claim to the throne voluntarily, and that the Khazars followed his recom m endation by w arranting Alm os or his son, Arpad, the right to power. The Secret H istory o f the M ongols relates a sim ilar story. A fter the death o f Chingis khan, Ogodei took over power. (Ogodei is know n as O khtay khan from the history o f the Tatar devastation o f Hungary in the 13th century.) He was not the eldest son, and he ascended to pow er after some struggle. This is the reason why a supplem entary chapter was inserted into The Secret H istory o f the M ongols w hich tells us that Ogodei had actually been appointed ruler in the lifetim e o f Chingis khan, and that O godei’s two older brothers Jochi and Chagatai had voluntarily relinquished, for the benefit o f Ogodei, their claim to the throne o f the great khan. The same chapter contains a peculiar section, where O godei speaks o f his fear o f begetting an inapt son: “ ‘Later, if per chance some am ong m y descendants will be born so w orthless that / Even if one w rapped them in fresh grass / They would not be eaten by an ox; / Even if one wrapped them in fat, / They would not be eaten by a dog, will they not— [like the unskilled hunter]— miss the elk breadthwise just as the rat lengthwise.’ So he spoke and at his words Chingis khaghan declared as follows: ‘If Ogodei speaks such words, that will do’” (Chapter 255, translated by Igor de Rachewiltz). So he then turns to Tolui, the youngest prince, who assures his brother o f his loyalty. Eventually, however, after O godei died and the M ongols devastating Hungary hurried hom e to sort out m atters o f power, the throne was not passed down along O godei’s branch, as w ould have been expected, but rather, Tolui’s son was elected great khan. The entire Ogodei
418
Recent research and studies
story is a later interpolation which served purely the legitim acy o f the Tolui branch. Porphyrogenitus’s text itself contains some giveaw ays suggesting that we are looking at a legitim isation story originating from the chronicles o f the House o f Arpad. A t the very outset the Em peror establishes that the Turks [M agyars] “had never had over them a prince, either native or foreign, but there w ere am ong them ‘voivodes’”, and that Levedi was the first am ong the voivodes. The ‘first’ (protos) in this instance could have the m eaning ‘first in tim e’. But why was he given a K hazar wife? In this respect, the steppe custom s did not significantly differ from the contem porary European ones. Chinese sources claim that the Turks rebelled against their lords, the R uanruan, because their ruler had refused to m arry his daughter to the leader o f the Turks. Then, when A rpad is lifted upon a shield, the text reads: “Before this Arpad the Turks had never at any tim e had any other prince, and so even to this day the prince o f Turkey [i.e. Hungary] is from his family.” This undoubtedly sounds like an interpolation, and it is very likely that it was taken from the oral traditions o f the Gesta Hungarorum passed down at the court o f Arpad and recited during feasts. The part w hich relates Levedi facing up to his incom petence, and recom m ending Alm os or A rpad instead o f him self, lacks even the sm allest fragm ent o f credibility. L evedi’s w ife m ight have been infertile, but at that time polygam y was com m on practice am ong the leading strata. Ibn Fadlan claims that the K hazar khaghan had tw enty-five wives, as well as sixty concubines and slaves. And even if Levedi truly had no son, his brother or his b rother’s son w ould have succeeded him as prince. In the light o f this, the least we can say is that the Levedi story was mere A rpadian convention pertaining to an earlier event, and that all the confusion and m isinterpretations o f text and context are caused by this interpolation. T his sam e tradition recounts the Savarti Asfali story. This is w hat the text claims: (1) The M agyars used to have a home. (2) This place was called Levedia. (3) It was nam ed after their first voivode, Levedi. (4) In this land runs the R iver Hidm as, also called Hingilous. (5) The M agyars w ere not called Turks at that time, but, for some reason or other, had the nam e ‘Savarti A sfali’. ( 6 ) The Turks consisted o f seven clans, and they had never had over them a prince, either native or foreign. This story features elem ents w hich cannot have originated from the M a gyars. ‘L evedia’ could hardly have been a country or place nam e. It is unlikely that a place is nam ed after a person who, having been defeated, left it. In those days, it was com m on practice am ong the M agyars to nam e sm all localities after persons (we know plenty o f these); however, the -a or -ia suffix is not a Hungarian one, and significantly large areas were never nam ed after persons
The Levedi question and the earliest H ungarian chronicle
419
by the Magyars. Consequently, the designation ‘Levedia’ cannot be Hungar ian. But neither could the claim that the Magyars were, for some reason or another, called Savarti Asfali have been based on a Hungarian account. Firstly, because if it had been an established fact and the Magyars had previously called themselves by this name, they would surely have given some explana tion for it. At least the name kangar is fully explained in the very same chapter: the Pechenegs were previously called ‘Kangar’ “for this ‘Kangar’ was a name signifying nobility and valour among them”. Secondly, on account of the Greek word A sfali (asphales ‘strong’, ‘reliable’). And thirdly, it is utterly impossible that the Magyars should have claimed that they used to be known by a different name. The identicalness of a people’s self-designation is vital to ethnic identity. Savarti A sfali and the like are foreign-type names which the peoples they designated never actually used themselves. Consequently, we can safely assert that the Savarti Asfali story is of Byzantine origin. The story falls into the category where a newly emerged people is connected with an existing, known one (which so often happened with the Magyars in the western chronicles), on the grounds that the report recounts about an eastern group of Magyars. The correct translation of this part of the text would be this: “One part was settled in the east [i.e. lives there, and not went eastwards a n d settled] The Greek original does not have the word ‘went’ in it] in the region of Persia, and they are to this day called by the ancient denomination of the Turks ‘Savarti Asfali’; but the other part, together with their voivode and chief Levedi, was settled in the western region, in places [the Greek text uses the plural] called Etelkoz, in which places the nation of the Pechenegs now lives.” So the Magyar envoys informed the Emperor that one part of the Magyars lived in the east. After the Conquest is described, the text continues thus: “To the aforesaid nation of the Turks that settled in the east, in the regions of Persia [literally, ‘towards the borders of Persia’], these Turks aforesaid who live toward the western region still send merchants [pragmateutas] who look them up, and often bring back official messages from them.” Again, this comes from the Magyar report of Termecsii and Bulcsii. We know that even in the 12th century the Magyars travelled to and from their eastern relatives, and that they maintained trade links, too. The question arises, then: what is this Greek report doing in the story, and why were the East Magyars called Savarti Asfali? The key to the mystery is a work attributed to Porphyrogenitus, but actually finished after his death, entitled D e ceremoniis aulae Byzantinae [Of Byzantine court ceremonies]. This text refers to the Sevortii (Sevortioi) who are said to consist of three tribes. This Armenian-named tribe (or whose name is interpreted to be Armenian) is
420
Recent research and studies
known to m any Arabic authors (M asudi, Dim aski), as w ell as the A rm enian sources. These write that the Sevortii lived to the north o f the Caucasus, in the region o f the R iver Kur. This people’s nam e in the A rm enian sources is Sevordik, literally ‘O f Black D escent’. The B yzantine source translates the nam e Sevortii, and explains that these were the ‘Black B oys’ (m aura paidia)— w hich goes to show that the Byzantines knew this A rm enian popular ethnic name etymology. The sources m ention them in the 8 th century, but it is possible that the tribe survived even into the 10th. There is no guarantee, naturally, that this designation is indeed Arm enian; we only know for certain that it was read as Arm enian, or according to a popular ethnic nam e etymology. ‘B lack’ and ‘W hite’ frequently feature in Turkic (and early M agyar) namegiving. In this iconography white stood for ‘new ’, ‘hig h ’, ‘no b le’ w hile black usually m eant ‘old ’, ‘retarded’ or ‘sim ple’. In Turkic the black people (kara boduri) were the commoners. The Slavic prim ary chronicle m entions Black M agyars (Ugri) and W hite M agyars alike. In the chapter on the Pechenegs the Em peror writes that ‘K angar’ was not a com m on designation for all o f the Pechenegs, but only for the people o f three provinces. These were m ore valorous and noble than the rest. In the chapter on the M agyars, Porphyrogenitus claims that the Pechenegs w ere “previously” called ‘K angar’. W hat exactly he m eant by “previously” is a good question. Researchers m ust have been puzzled by the peculiar course ethnic nam es took in those days. The ethnic nam e ‘K angar’ has been shown to have existed in the Caucasus region as early as in the 6 th century, before the em ergence o f the Turkic peoples. It cannot be inferred from the Syrian sources w hich report on the Kangars exactly w hat their ethnicity and language was. Possibly they encountered the Pechenegs later, but there is no evidence to support this. If, however, such a tribal nam e cropped up am ong the Pechenegs, the Pechenegs would certainly have associated it with their word kongor ‘brow n (colour of a horse)’. In w hich case we are again looking at the case o f an ethnic name established by m eans o f a popular etymology, as in the case o f the ethnic name Turk (see also pp. 279-281). Based on the com parison o f the above analysis o f the Pecheneg story with the sources, I feel that the reconstructed events are a lot sim pler than was previously supposed. The fact that the “Pechenegs who were previously called ‘K angar’” attacked the M agyars m eans that those three Pecheneg tribes who attacked first were, so the other chapter tells us, called ‘K angars’. This attack m ust have shaken L evedi’s power. Alm os and A rpad’s clan m ust have then seized the opportunity and taken over power. The sources are silent on when this happened exactly— unless we take for granted R egino’s inform ation of 889, according to which the M agyars were driven from their old place by the Pechenegs (Pecinaci). This inform ation has little chronological value, owing
The Levedi question and the earliest Hungarian chronicle
421
to the fact that it was custom ary in those days to include the inform ation about the new ly em erged peoples— somewhere in the annals. The fact that the year 889 happens to be ‘som ew here’ is o f no real consequence regarding the ac tual chronology o f events. It was probably when Regino first heard o f the M agyars— perhaps apropos o f the change o f dynasty. It does not mean, therefore, that every event under the heading 889 actually happened in that year. In any case, it is rem arkable that the w ell-inform ed Regino (who natural ly cites every thinkable m ediaeval rhetoric figure about the M agyars) only knows o f a single Pecheneg attack. The fact that the Pechenegs set out in 894 to launch their decisive attack on the M agyars and the attack happened in 895, does not m ean that they did not m ake raids on the Etelkoz M agyars earlier for booty or reconnaissance. The Em peror inserted the account o f the Pecheneg-K angar attack in b e tween the part about the East M agyars and the part about L evedi’s “renuncia tion”. To remain on the more or less definite side o f our sources, all that can be said o f this is that the first P echeneg-K angar attack m ust have occurred under the rule o f Levedi, and the second after the A rpadian dynasty had taken over rule. However, there was no mention o f this in the Gesta Hungarorum.
NOTES The part o f the text referring to Levedia and Etelkoz, and to Levedi, has been analysed recently by a number o f scholars. See the bibliography about the Etelkoz (p. 324). I insert here one o f the key parts o f the Greek text and the translation from M oravcsik-Jenkins (1967, p. 170, line 2 6 - p . 172, line 31): K a t t o (lev e v (le p o q 7ipoq a v a i o X n v ei<; t o -rife П е р т 5 о < ; (le p o q катф кт|ст£У , o i кой (lEXpi то и v u v к а т а r ^ v tcov T ou p K to v r ip '/c d a v e T to v u iiia v kocXouvtcxi £a|3cxpTOi а о ф а Х о ;, то 8£ sT e p o v ( is p o q e iq t o 6 \) t\k o v к а т ф к г |а е jjip o c , a | i a m i т ф р оероб со a i n w v к а л а р х л у ф AePe5i
“One part went eastwards and settled in the region o f Persia, and they to this day are called by the ancient denomination o f the Turks ‘Sabartoi asphaloi’; but the other part, together with their voivode and ch ief Lebedias, settled in the western region, in places called Atelkouzou, in which places the nation o f the Pechenegs now lives.” (M oravcsik-Jenkins 1967, pp. 1 71-173.) As can be seen from the Greek text, there is no word which would mean “went”. The text says that one part o f them was settled [lived] (katokesen) eastwards in the region o f Persia, and they are called by the ancient denomination o f the Turks Savarti Asphali, but the other part, together with their voivode and chief Levedias was settled [lived] (katokese) in the western
422
Recent research and studies
region, in places called Atelkuzu, in which places the nation o f the Pcchenegs now lives (katoikei). A s was pointed out by Harmatta (1985c, pp. 4 6 -4 7 ) the verb katoikeo is used in the DAI (according to him always, according to me in most cases) in the sense ‘to live somewhere, to be settled som ewhere’ and the meaning ‘to go som ewhere, to migrate som ew here’ is expressed, among other verbs, by kataskeno. There is one exception to this in the follow ing text (M oravcsik-Jenkins 1967, p. 172, line 61 - p . 174, line 65): Ei<; 8s
t o KaTaaicevwaav t o rcpoppriOev ei3vo<; t c o v ToupKtov ярое; a
v a T o t a i v ei<; та
Tfiq nepoi8oq дёрг| дехР1 той vuv лрсгуцатеитск; anooTeX/.ouoiv outoi
oi
npoq t o
Sutikov liepoqoiKouvTeqTtpoppri^evTeqToupKoi, m i P^enoumv сш тойс;, m l anoK'pioaq лара auTdjv rcpoi; айтой^ лоХАакц «7tOKop.i£ownv.
“To the aforesaid nation o f the Turks that settled in the east, in the region o f Persia, these Turks aforesaid, w ho live toward the western region still send merchants w ho look them up, and often bring them back official m essages from them.” (M oravcsik-Jenkins 1967, pp. 173-175.) But in this text the verb kataskeno is equivalent to the verb oikeo in the next sentence. And in fact also here, the author says that to the Turks (Hungarians) w ho live in the east those Hungarians who live in the w est send merchants with official assignment (pragm ateutds), etc. The preposition eis ‘into’ also has the meaning o f en ‘in’ in Late Greek. The results o f Harmatta were questioned by Vekony (1986, pp. 4 3 -4 5 ). He cited a handful o f places where the verb katokeo means ‘to go som ew here’. Olajos (1995, pp. 4 4 -4 6 ) discussed the issue recently. She has shown that many o f the places quoted by Vekony do not pertain to our question, and only three cases have, according to her, the meaning ‘to go over, to settle som ewhere in ’. Olajos argues that the aoristus indicativi (katoksen ) reflects an aspect which is “incohative, ingressive, metaptotic” (Olajos 1995, p. 45) and thus it cannot mean that the Hungarians lived there, but only that they moved there. Her view s were unconvincing to Harmatta (1 996a, 1998). Harmatta stressed that the Greek aoristus is neutral to the aspect, and thus its use cannot be an argument in favour o f the claim o f Olajos. I think that the mislead ing point was the translation o f M oravcsik-Jenkins, who included the word “went” in their translation without giving any reason or calling for attention. M ost o f the more recent editions o f the DAI follow here the authority o f M oravcsik-Jenkins. For information on the Savarti issue, see Chapter VI.4 o f this book (pp. 2 8 8 -2 8 9 ), and the notes on p. 313.
XIV. HISTORICAL TRADITIONS, ATTILA AND THE HUNNISH-MAGYAR KINSHIP
The notion o f H unnish-M agyar kinship appeared gradually in the Hungarian chronicles, and it was soon to become part o f the M agyars’ historical con sciousness. There is no m ention o f the H unnish-M agyar kinship in Anonymus, and the chronicler never even m entions the nam e ‘H un’. Instead, he claims that the M agyars were descended from the Scythians. M agog was the son o f the Old Testam ent Japhet, and M agog’s progeny begot the “valorous and im m ensely powerful King A ttila” who m arched to the land o f Pannonia heading the people that came out o f the land o f the Scythians. Later, the Magyars “heard from trickling tidings” that this land (Pannonia) was the land o f King Attila, from whose line Almos, A rpad’s father, was descended. Anonymus holds M agog to be the M agyars’ nam esake and ancestor, and Attila appears as the historical personality legitim ising the Conquest. Deriving the M agyars’ designation from M agog was not A nonym us’s idea. As we have seen above (p. 282), in his letter written before 923 to Dado, the Bishop o f Verdun, a Germ an m onk m akes mention o f the rum ours regarding the M a gyars’ descent from the progeny o f Gog and Magog. In his work entitled Pantheon (ca. 1189), Gottfried o f Viterbo claims— drawing on Isidorus o f Seville (565-636)— that the Goths had acquired their name from their ancestor Gog. It has been often discussed that Gottfried o f Viterbo holds the people o f Scythia to be nam ed after M agog. The view that G ottfried o f Viterbo, chaplain and scribe to Frederick Barbarossa, m aintained contacts with Anonym us, has recently gained currency. In any case, we know that Frederick Barbarossa arrived in Hungary in late M ay 1189, and was received cerem oniously by the Hungarian King Bela III who probably also accom panied him to Buda. In the same year Frederick B arbarossa’s envoys went to the court o f Вё1а III in Obuda. Anonym us rem arks that Attila set up his seat in Buda, renovated the buildings he had found there and erected a protective enclosure wall. The Magyars called this Budavar ‘Buda C astle’ (Buduuar) in their ow n tongue, while the Germ ans referred to it as Ecilburg (Ecilburgu). This G erm an name is actually none other than A ttila’s castle in the Nibelungen (Etzelen biirge). All o f the contem porary sources relating Frederick B arbarossa’s m arching
424
Recent research and studies
across H ungary and his m eeting with the Hungarian king agree that the royal appointm ent was held in “A ttila’s castle” . This attests to the fact that by the end o f the 12th century the descent from Attila was know n to the royal courts o f Hungary. To this tradition w as added the H unnish-M agyar kinship “theory” o f the Hungarian chroniclers, who had drawn on elem ents o f the w estern sources. Com parative analysis o f the Hungarian chronicles and the contem poraneous European sources has re vealed that not only the idea o f a H unnish-M agyar kinship, but also the arguments supporting it and the actual phrasing originate from the western sources. This is one detail which Hungarian historians tend to agree on. However, the descent from Attila and the notion o f a H unnish-M agyar kinship m ust be treated as separate issues. This descent theory has a great many aspects. The A rpadian rulers ’ claim to be descended from A ttila is not identical with the notion o f the M agyars’ descent from the Huns. In the M iddle Ages the rulers, the nobility, the feudal lords and the people w ere not necessarily o f the sam e stock. M oreover, w e m ust separate the actual facts o f lineage from beliefs, and the actual facts from the facts that the m ediaeval H ungarians knew. In our view the actual facts, the knowledge o f and claim s about these facts are three different things. The contem porary sources, o f course, never m ade such a distinction. The descent from Attila served to justify the A rpadian dynasty’s claim to the throne. The question arises o f w hy the descent from A ttila furnished the A rpadian dynasty’s legitimacy, given A ttila’s historical reputation o f being the “sword o f G od”, G od’s punishm ent or, at any rate, the “scourge” o f the Christian world (flagellum D e ift The question has been w idely dealt with. To sum up a w ealth o f research, it can be said that at the tim e the M agyars’ historical traditions emerged, the European tradition’s notion o f “enem y o f the faith” was replaced by the less severe “ instrum ent o f G od”. This m ade this legitim isation acceptable. However, it still does not follow w hy they referred to Attila. For that m atter they m ight have traced back the A rpad dynasty’s lineage to the Romans, the Troyans (A nonym us actually m entions them), Charles the G reat or anyone else. There are tw o reasons for the M agyars’ linking them selves with Attila. Firstly, the fact that Attila had his seat in Pannonia was in itself a good enough reason. Also, it follow ed from the order o f the steppe w here A ttila was the standard source o f legitim acy until Chingis khan. Thus the w estern sources provided the H ungarians w ith a literary tradition, and accepted their reasoning, but did not claim (lacking the grounds) that the H ungarian ruling dynasty had actually descended from Attila. This follow ed later, after the em ergence o f the Hun story, in Thom as o f Spoleto (died ca. 1268), and in the Hungarian chronicles, forem ost in de K eza’s Chronicle (1282-1285).
Historical traditions, Attila and the H unn ish-M agyar kinship
425
The descent from Attila was an idea rooted in the Arpadian traditions even before Anonymus. The Hungarian King Andrew I died in 1060. His widow, Anastasia, of Kievan origin, and their son Solomon, were forced into exile. They fled to the Frankish Emperor Henry IV. The Frankish armies interfered in this power struggle, and supported Solomon and his allies against the pretender Вё1а, who could not cope with his royal duties, and died when marching against the Frankish forces. Supported by the Franks, Solomon was crowned king at Szёkesfehёrvar in 1063. On this occasion the queen mother, Anastasia, made a gift of a sword to the Bavarian Prince Otto II in gratitude for having helped Solomon to the throne. A contemporary chronicler, Lampert of Hersfeld, gives a detailed account of the sword in his annals (1077-1079). The sword came to view when the Emperor Henry IV stopped over in Hersfeld in 1071, on his way to Mainz. In Hersfeld, one of the Emperor’s knights fell from his horse and suffered a lethal wound from his own sword. The Emperor then bestowed to the monastery of Hersfeld 30 peasant-estates in return for their prayers for the memory of the unfortunate knight. Lampert mentions that the sword that wounded the knight was identical with the one Anastasia had presented to Otto, which was none other than Attila’s sword. This comment, too, has been much discussed. After the mention of Attila’s sword, by way of an explanation, Lampert inserted Jordanes’s story of the sword. This story relates how a peasant boy accidentally came across it. There is no doubt about the fact that Anastasia presented her benefactor with the sword, referring to it as being Attila’s. Naturally, we do not know what Anastasia actually thought about the sword; whether it was really Attila’s or whether it was just pure imagination. It is certain, though, that Lampert of Hersfeld did not invent the story himself. The fact is that the story relates the account of an eyewitness of the Emperor’s offering a gift to the monastery. Lampert of Hersfeld would never have risked discrediting the Emperor’s deed by including in his account a story which even in those days could be proved to be false. Yet we know that in 1058 Lampert of Hersfeld visited Hungary, and in the course of this visit he might have acquired knowledge of the living royal Hungarian tradi tions. The source of his annals (written down 14 years later) was first-hand; but had his account been fabricated, the Frankish lords might have verified its falsity at any time—a risk not to be taken. The gift of the sword as being Attila’s cannot have been derived from either the western, or the eastern (for instance, Kievan) sources; there is certainly no trace of this, and neither is it very probable. All we can say is that the Attila tradition enjoyed a vogue in the Hungarian royal courts at least 150 years before Anonymus, and over 150 years after the Conquest. The chronicles relate the Magyar royal clan’s totemistic legend of origin. Mediaeval people felt no contradiction between this story, which claims that
426
Recent research and studies
A lm os’s clan was descended from a m agical haw k (turul), and the M agyars’ descent from Attila or Japhet’s son Magog. The different legitim ating ele m ents finely com plem ented each other. The Bible is full o f sim ilar contradic tions, but we might also refer to the N orm an chronicler W illiam o f Jumiege who arrived in England with W illiam the Conqueror. Very quickly he had to assign the new royal dynasty a place in the European royal lineage system. He needed, on the one hand, to establish connections w ith the glorious Roman em perors— who, so Virgil claimed, originated from the Troyans— and, on the other hand, he needed the C hurch’s blessing too; so he boldly claim ed that W illiam the C onqueror him self had descended from the Troyans, w hile at the sam e time, he traced back the new k in g ’s lineage to Japhet’s son, M agog. The defeated Anglo-Saxons had to keep up with this extraordinary genealogy. An early Irish-E nglish chronicle (which had m any versions in circulation by 1168) tells us about the conquest o f Ireland. The author(s) sim ply inserted into M agog’s family tree a Scythian person called Partholon who “cam e unto the isle o f the Irish 1300 years after the deluge” . The m essage was that the Irish had a longer line o f descent— hence older rights— than the N orm ans. In his chronicle started betw een 1136 and 1138, G alfred o f M onm outh writes that Brutus, the son o f the Troyan A eneas’s son Ascanius, founded Britain (the etym ology o f the name differs little from the H ungarian etym ologies A nony mus com es up with), but on arrival on the island he m et a giant named G oem agog whom he encountered in a duel. A nother author holds this Goem agog to be the forebear o f the Danes and the Norm ans. We refer to these genealogies to dem onstrate that such descent traditions were intrinsically part o f m ediaeval European thought. The next question is where the M agyars had recourse to the idea of originating their people from Attila. We know that the Danube B ulghars’ list o f rulers was headed by an Attila (Avitohol and his successor Im ik, A ttila’s son). As has been pointed out, the M agyar Conquest was partly triggered by a jo in t P echeneg-D anube Bulghar attack; it w ould seem logical, therefore, that the real “argum ent” was about whose royal dynasty was “really” descend ed from Attila, or whose lineage was m ore distinguished on the descent chart o f A ttila’s progeny. Thus the Attila tradition did not m ean that the Hungarian royal dynasties were actually descended from Attila, but rather, that they considered them selves to be his descendants; or m ore precisely, their self-proclaim ed lineage was a chapter o f 1 0 th -12th-century royal court traditions. This form ed part of the m ultiply transform ed “rom ance” w hich served as a source for the Hungar ian prim ary chronicle. We can assume, although there is no evidence to support it, that the Attila tradition had existed am ong the M agyars o f the Conquest.
Historical traditions, Attila and the H unn ish-M agyar kin sh ip
427
At the end o f the 12th century, Anonym us only m entions Attila, the lord o f the people that “came out from the land o f Scythia” . However, a H unnish chronicle— a story o f the H unnish descent o f the M agyars— gradually emerged in the 13th century. A certain Hunor cropped up alongside Magor. But that is another story which is not related to the ancient history o f the M agyars.
NOTES The issues o f the Hunnish-M agyar kinship were dealt with by many authors. From the more recent literature G yorffy (1948, 1993b), Szucs (1973, 1985), Krist 6 (1983b, 1996) are impor tant; see also Korde, in: Krist6 ( 1994c, pp. 2 7 4 -2 7 5 ). For the historiographical and ideological aspects o f the H un-M agyar question, see Rona-Tas (1991 d). For the Hungarian story o f Attila’s sword, see Rona-Tas (1988a, pp. 112-114). For genealogical theories in m ediaeval Europe, see R6 na-Tas (1978a, pp. 4 5 -4 6 ).
XV. THE EAST MAGYARS, THE BASHK1 RIAN TRIBAL NAMES AND YUGR1A
The question o f the so-called “Bashkirian Urheimat'’ in the ancient history o f the conquering M agyars deserves special attention. Due to the lim itations o f space, we cannot give a survey o f its literature, which would undoubtedly fill libraries. We can m erely attem pt to sum up the facts and dispel som e m iscon ceptions. The sources unanim ously confirm that on his first journey to the east in winter 1235-1236, the Hungarian friar Julianus m et a M agyar w om an in the capital o f the Volga Bulghars. She explained to him that the M agyars’ Julianus had set out to find lived two days’ w alk from there. “And he findeth these by the great R iver Etil.” The doubts regarding the authenticity o f the story recorded by Friar Riccardus have been proved to be unfounded. E til is the nam e o f the River Volga; however, in those days they held the Kama to be the upper course o f the Volga, and thought that the river today called Volga flowed into it. The Bashkirian clan legends share these views, and even in the early 20th century A shm arin wrote that the Chuvash o f the Chistopol region called both the Volga and the Kam a A dil (< Etil). In 1237, when the Tatars devastated Volga Bulgharia, the M agyars o f the Kama scattered, crossed over to the w est bank o f the Volga, and m any o f them m oved even farther. As we have seen (see p. 298), m em ories o f this scattering have been preserved in a num ber o f place names. An epitaph unearthed in Chistopol (near the Kama) distinguishes betw een father and son by tagging onto the father’s name the ethnic nam e M agyar (see pp. 301-302). The cemetery excavated outside Bolshie T'igani is also important, because it contains grave goods, and features burial custom s sim ilar to the Arpadian M agyars’ (see pp. 120-122). Also, we have seen exam ples o f the M agyars being called Bashkirs by the Volga Bulghars (see pp. 289-294). These facts and finds were a m agnet for researchers, and quite under standably the view gained wide currency that the prolonged presence o f Magyars in the V olga-K am a-U ral region between the 10th and 13th centuries meant that this was in fact the M agyar Urheimat from where they set out to conquer the Carpathian Basin.
430
Recent research and studies
Am ong m any arguments, it was hinted at the end o f the last century that M agyar tribal nam es could be found am ong the Bashkirs. For a long tim e only the correspondences o f two M agyar tribal nam es, the G yarm at and the Jeno, were discussed, but later attempts sought to connect other M agyar tribal nam es with B ashkir ones. The truth is that even if these correspondences o f tribal nam es w ere correct, they w ould not tell us m uch about the form er hom eland o f the M agyars. The w anderings o f nom adic tribes are very rarely traceable w ith the help o f the w ritten sources; moreover, there is no way o f know ing w hen and how sim ilar tribal nam es entered today’s Bashkiria. N evertheless, it is w orth investigating w hether or not these nam es indeed refer to identical tribes; also, it is worth finding out m ore about the history o f the nam es, and o f the tribes they designate. These issues are worth looking into regardless o f w hether the etym ology o f these tribal names is correct or not. The oldest B ashkirian-M agyar tribal nam e correspondence was established betw een the Bashkirian Yurmati and the M agyar Gyarmat; and the Bashkirian Yeney and the M agyar Jeno. The first raises severe phonetic problem s. The w ord-final in the H ungar ian m ight have disappeared, its absence w ould not m ean any problem ; it is striking, though, that Porphyrogenitus does not add it on ( kurtugermatos ), yet K urt does take on a word-final -u. The absence o f a w ord-final -i can be explained by the fact that the w ord is disyllabic, or that it w as thus “processed” by the G reek language. The difference o f the vowels o f the first syllable, however, cannot be accounted for. The G reek form reflects the form Yermat w hich could be the transliterated form o f the M agyar Jermat or even Gyermat. The /е/ occurring in the first syllable o f the oldest Hungarian form m ay have two explanations. Either it featured in the original tribal nam e, in w hich case Jarmat > Gyarmat was brought about by processes o f late assim ilation; or Jarmat (with a) was the original form, and the e was secondary. Again, this could be due to assim ilation, or m erely a transcription problem . The applica tion o f some orthographic procedure was necessary to transcribe the wordinitial j-, in order for the Greek g- to be read as_y- or j-. The first syllable o f the M agyar nam e cannot have had an о or an u. The nam es G orm ot and Gurm ot, w hich feature in the early H ungarian sources, are Latin transcriptions; the reason w hy they nevertheless contain these vowels is due to the inability o f the Latin orthography to reproduce the H ungarian labial /а/. For w ant o f a better alternative, they transcribed this sound w ith the letter o, as in A nonym us’s rendering o f ‘M agyar’ ( Moger ). Such early Hungarian data do not point to the existence o f a form *Gyormot; the G reek data, too, contradict this. However, the и in the present-day B ashkirian language goes back to an early o. The tribal nam e Yurmati was * Yormati, then. The word-
The East Magyars, the Bashkirian tiib a l names and Yugria
43 1
initial consonant is also a problem. If we are indeed looking at the Hungarian tribal name Gyarmat, then the Bashkirian language—which lacks the wordinitial /j/ sound—would surely have replaced it by some other consonant, most probably by ch. This /ch/ shifted to /s/ in vernacular Bashkirian words and in the early loan words, after the old /s/ had changed to /h/: Old Bashkirian */ch/ > Ы Old Bashkirian */s/ > /h / Consequently, if the Hungarian Jarmat was borrowed early on, it would now sound *Sarmat in modem Bashkirian. Assuming that the old Bashkirs were less aware of the mysteries of phonetic history than today’s scholars, the replacement of the Hungarian j - with у - is improbable. It is unlikely that they knew that the Hungarianj - went back to the samej - that occurred in the Turkic languages (where, even earlier, it used to be y-) and so, regardless of the fact that they heard the sound /j/, they replaced it with y-. If, however, y - was the original word-initial sound, the Hungarian language would have preserved it on account of the fact that it accepted that consonant in a word-initial position. Thus the adopted tribal name would have been *Yarmat. The only possible assumption would be that the ancestors of the Yurmati maintained close ties with a y'-Turkic group which had the form Jarm ati, which then became Jorm ati in that language. Then they adopted the tribal name, but “processed” the phonetic difference by systematic analogy (reasoning that “when they utterjo k , we say у ok, so if they say Jormati, we must pronounce it Yormati”). Such a neighbouring people could only have been a Turkic-speaking one. Under the given circumstances, the Volga Bulghars are the only option. It is remarkable that the Bashkirs have clan names like Yurma, Yurmi and Yumran (and the latter also has a variant Yumuran). Yurmat is the left tributary of the Bashkirian River Ik; Yurmit the left tributary of the River Metev; Yurmi the right tributary of the River Cheremshan and also of the Usen which flows into the Ik; also, the Bashkirs have a brook called Yurmi-az in the Biri district, and a rivulet named Yurmash which is tributary to the River Ufa. It is hardly conceivable that these river names originated from tribal names. It would be possible vice versa, but would have to be proved. Recently, Bashkirian scholars published the Yurmati tribe’s legend of origin. There are two interesting points about it. One is that this group traces its lineage back to the Golden Horde, and accordingly, the story begins with Chingis khan. The other is that in the 15th century the Yurmati group moved south from the Shishme and the Zey regions to the area of today’s Sterlitamak.
432
Recent research and studies
N ow here does the text allude to the tribe having connections with, for instance, the Volga M agyars. The Hungarian Jeno goes back to the Old Hungarian form Yeneh (Genah, in G reek transliteration). In this latter the w ord-initial y- (the letter j is pronounced [y] in Hungarian) is in fact a protheticy- which derives from the old form *ineg (see p. 351). In the Bashkirian language the word-final -eg of original words and old borrowings shifted to -ey. So, if either the Hungarian or any Turkic language had the form *Yeneg, it would have been transformed to Yeney in the Bashkirian by m eans o f system atic change. It is striking that w hile Yurmati has a widespread fam ily o f nam es (Yurma, Yurmi, Yumran, etc.), Yeney does not; not even am ong the w ater names. B y analogy with the Turkic ilteber title featuring as yiltever (with a prothetic y-) in the Volga B ulghar language around the 10th century (see p. 226), it m ight be assumed that the Turkic *ineg existed in the form *yeneg am ong the Volga Bulghars. The title nam e, perhaps, becam e a tribal name, then. One o f the subtribes o f the Tabin tribe is called K ese Tabin in Bashkirian. K ese in Bashkirian has the m eaning o f ‘sm all’ (etym ologically the Hungarian kicsi ‘sm all’ and the Bashkirian word have the same origin); the Turkic kichi becam e kese in the Bashkirian language by process o f system atic change. Thus, K ese Tabin m eans ‘Small T abin’. The adjective sm all occurs among other Turkic peoples, too. Com parison with the M agyar tribal nam e Keszi is im possible both phonetically and morphologically. The nam e Nagm an, one branch o f the Bashkirian Uhergen tribe, derives from a personal nam e o f Arabic origin. Com parison w ith the H ungarian tribe nam e N yek (via the putative form s *Nekimen > *Negmen > *Nagman > *Nagman) poses insuperable phonetic difficulties. The ethnic nam e M isher is also supposed to be connected to the Magyars. M isher is the name o f a group o f Tatars in the Volga region. A Hungarian M ejer > Bashkirian M isher shift has been suggested, but the H ungarian Mejer could only have becom e *Miser in the Bashkirian language (via the form *Mecher); however, no such form exists. The *Mecher form m ight have transform ed to M isher in certain Tatar dialects— but only very late and strictly in a few dialects only, and not in Bashkirian. As has been pointed out above (see p. 306), the ethnic name M isher derived from the ethnic nam e Meshcher. In spite o f the etym ology o f the M agyar tribal nam es, o f w hether they derive from com m on words, and o f the original language o f these com m on words, only the parallel o f the Hungarian Jeno tribe is flawless. The B ashkirian tribal nam e Yeney can be o f M agyar or Volga Bulghar origin. Only provided the Bashkirian Yurmati derived via Volga Bulgharian can it be linked with the Hungarian tribal name Gyarmat. The correspondences o f all other tribal names pose insurm ountable difficulties.
The East Magyars, the Bashkirian triba l names and Yngria
433
Some geographical names are usually referred to in support o f Bashkirian-M agyar correspondences and the Bashkirian Urheimat theory. These can be grouped threefold. The first group contains nam es like Magash, w hich has been identified w ith the Hungarian word magas m eaning ‘hig h ’, ‘tali’; how ever, the word is o f Iranian origin, and means ‘m osquito’. Since M agash also occurs as the nam e o f a stream — where the m eaning ‘h ig h ’ or ‘ta ll’ can be excluded— it seems as though we are looking at a designation like Kom arom (a town in northw est Hungary) which derives from the Slavic kom ar ‘m os quito’. In the 1Oth century, M asudi wrote o f the A lanians’ capital being called Magash. The 10th-century author o f H udud al-alam discusses at length the mosquitoes in the country o f the Alanians. The B ashkirian M agash belongs to that group o f geographical nam es which originate from the Iranian tribes, and which resem ble a Hungarian word by accident only. The second group o f Bashkirian geographical nam es includes words like Kondoros. K undur in the river nam e Kundurcha m ay etym ologically be related to the word kondor in the Hungarian river nam e Kondoros. The root word o f this river nam e does not exist as an independent word in the H un garian, but is a Chuvash-type equivalent o f the Turkic kunduz and the Bashkirian kondoz ‘beaver’, and postulates the form *kundur. The word-final -sh suffix in Kondoros is purely Hungarian; the root o f the B ashkirian river name Kondorosh, however, is Volga Bulghar, not Old Bashkirian. M odern Bashkirian has two plant names: kondorak ‘Agrostis (a w ater plant)’ (=Tatar kondirak); and kondorsok which goes back to the form kondorchuk, and is likewise the nam e o f a water plant. These plant nam es would translate as ‘beaver-grass’. The -sh suffix is very frequent in the Bashkirian language. Thus the root o f Kondorosh is a geographical name originating from the Volga Bulghar period, and only indirectly has a bearing on sim ilar sounding H un garian place nam es— only in that the roots in both languages are loan words. The third group o f geographical names includes words like Bikesh which, in Russian transcription, also has the form Bekash. It is a village nam e which has the above-m entioned Bashkirian -sh suffix tagged on and w hich derives from the Bashkirian bik ‘lord’ (Old Turkic bek < berk, and not beg). The place name bears a rem inder o f the one-tim e ow ner o f the village, and has nothing to do whatsoever with the Hungarian word bekas ‘o f frogs’. B ikkol is a frequent Bashkirian place name. The second constituent, kol, m eans ‘servant’, ‘villein’, and the name thus m eans ‘B ik ’s (lord’s) servant’. This very strongly reminds us o f a highly widespread Turkic-type personal nam e pattern (for instance M ahm udkuli ‘M uham m ad’s servant’, etc.), m any o f which becam e geographical nam es. The seeming sim ilarity o f the Bashkirian Bekash and the Hungarian Bekas is very m isleading; and the geographical nam e belongs to a recent group o f geographical nam es o f Bashkirian origin.
434
Recent research and studies
Weighing up the evidence, it can be said that no Hungarian toponym s have been located in the V olga-K am a region. Finally, it m ust be m entioned that by now we have a very thorough know l edge o f Bashkirian language. M any o f us have investigated the possibility o f a group o f Hungarian loan words in Bashkirian, or a H ungarian influence in the language. Research results have been negative so far. The B ashkirian language belongs to the K ipchak language group w hose speakers m ust have arrived to the Belaya region in the 13th century. The B ashkirian nam e o f the Belaya is Agizel, that is, ‘W hite R iver’; whose izel constituent, m eaning ‘great riv er’, is etym ologically identical with the word E til that features in the Hungarian sources, too. The Bashkirian sound signified by the letter z is not unlike the English voiced interdental fricative / 6 /] (as in the definite article the). This sound has two origins. In Old Bashkirian words it goes back to the oldz; in foreign words, as well as in Bashkirian words in intervocalic position, it occasionally corresponds to an earlier d (Turkic adash ‘nam esake’ > B ash kirian azash, Russian beda ‘poverty’ > Bashkirian biza). A t the sam e time, the regular correspondence o f the old Turkic -d- is -y- in the B ashkirian (Turkic adak ‘leg ’ > Bashkirian ayak). However, in cases w hen in m odern Bashkirian the old Turkic -d- corresponds to a z, it can always be shown that the d is a secondary derivation, not original (adash ‘nam esake’ < atdash). This is also the case w ith A k Izel > A gizel form o f the river nam e Belaya. The Bashkirian Izel goes back to the form Edil, and not Etil. The river nam e E til and the com m on word etil (cf. the Tatar Idel or the m odem C huvash A dil) shifted to E dil in the Volga Turkic languages, prim arily in Volga Bulgharian. This means that the forebears o f the speakers o f the Bashkirian language cannot have arrived to the region where the rivers were no longer called etil, but edil earlier than the 13th—14th centuries. The old -t- was preserved in the B ashkirian (batir ‘h ero ’ > Bashkirian batir, otuz ‘thirty’ > utiz). The B ashkirian historical tradition, too, suggests that they m oved to their present-day areas after the Tatar invasion. N aturally this applies only to the B ashkirian-speaking ethnic ity, and not to the ethnic name. The Bashkirs m ust have adopted their original ethnic nam e B achgird som ewhere in the southern Urals. The so-called Yugria question is frequently m entioned in connection with the history o f the East M agyars. The region between the K am a and the Urals is referred to as Yugria by the early Russian sources, as o f the 11th century. The Arab sources (for instance, M arvazi) use the form Yura w hich goes back to the form Yugra. The sources claim that this people (or population o f a country) lived “behind” the Bulghars. Som e scholars hold the w ord ending to be a Slavic suffix, and figured that Yugri was a Slavic plural form which originated from the earlier form o f Ungri. The word-initial y - is supposed to have developed in the R ussian or the Zyryan language. However, this Ungri
The East Magyars, the Bashkirian tribal names and Yugria
435
rings a bell: it is supposed to be identical with the M agyar Ongri, and its Slavic derivation (Ongre > Ugri] see p. 284). Thus the designation Yugria is supposed to preserve m em ories o f a M agyar presence in the Kama region. Provided the nam e Yugria is indeed a derivation o f the O noghur ethnic name (or its variant applied to the M agyars), there is a problem , because we established above (see pp. 282-287) that *Ungri was the S lavs’ designation for the M agyars. The Slavs o f the 11th century could only have called the M agyars o f the K am a region Ugri if they had known about the K am a M agyars ’ kinship with the M agyars o f the Carpathian Basin. Consequently, it happened the other way round com pared to the Bashkirian designation by which the Volga Bulghars called the East and West M agyars. We cannot positively rule out this possibility; however, nowhere does the nam e Yugria have variants such as *Yonugria, *Yongria or *Yungria. It would be difficult to explain why the Arabs referred to the area in question Yura (< Yugra), given that they usually called the M agyars Turk, Bashkir, M ajgar or Majar. However, an ethnic nam e Ugur would have become Ugri in Russian, and a secondary -y might have been tagged on later. Consequently, we can establish that an old Slavic form *ongre did become Ugri later; but not all Ugri forms go back to the form *ongre. O riginating from the form Ugur would also be acceptable. After all, if there existed an ethnic nam e *Yugur, it, too, would have become Yugri (in the plural) in Old Russian. W ho were these people? We are in the dark regarding this matter. It is remarkable, however, that while the nam es o f the Bulghars o f the steppe, the Carpathian Basin and the Volga Bulghars are always co-m entioned with the ethnic nam e O noghundur or Onoghur, we have no such inform ation about the Volga Bulghars. It is possible that an U ghur tribe m oved north together with the Volga Bulghars and the M agyars. It is not very likely that the name Yugria conceals a m etathetic vari ant o f the Uighur ethnic nam e (but: see the nam e o f the Yellow Uighur, the Ydghur). We have Chinese sources dating from the 10th century w hich tell us that the Ughur group— formerly m entioned in connection with the Ruanruans (that is, the Asian Avars)— whose splinter group joined the M ongolian-language Khitai, changed its nam e from Ughur to Yughur, evidently by process o f inside developm ent (see pp. 210,255). The annals o f the Khitai Liao dynasty claim ed that they lived afar, and the Khitai sent people in exile am ong them. They lived in very cold regions, were fishermen, and sent furs to the court o f the Liao emperor. M asudi (who died ca. 956) nam ed a group called Y ighur am ong the Kimeks living in the Irtysh region. The shah o f Khwarezm reported in 1182 that the Yughur-zadaghan o f the surrendering K ipchak chieftains visited his court.
436
Recent research and studies
The nam e o f Yugria, then, conceals a 1O th-11th century form o f the ethnic nam e Ugur. Traces o f this in West Siberia have been found in other sources. Thus Yugria w as only connected to the M agyars in so far as their nam e U ngar conceals the form o f On Ugur ‘Ten O ghurs’.
NOTES For tw o Julianus-related texts, see D6rrie (1956), and GOckenjan-Sweeney (1985). On the most recent Hungarian translations, see GyOrffy (1965, pp. 3 9 -5 1 ). For the doubts raised about Julianus’s reports, see Sinor (1952). Ashm arin’s data are included in his dictionary (ASmarin II, 1929, p. 143). For the latest literature on the Bashkir question and the Bashkir tribal nam es, see p. 313 o f this book. V&s&ry (1985) discusses the linguistic aspects o f Bashkirian-M agyar connections, together with further bibliography. A summary o f the Yugria issue is given in Zsirai (1930) and VAs4ry (1982). In an as yet unpublished study I discuss the issues o f the U ghur or Yughur tribe which lived am ong the Avars and the Khitai (see som e o f these results in R6na-Tas 1996a).
XVI. THE SZEKELY RUNIFORM SCRIPT
Written around 1282-1285, Simon de K eza’s Latin chronicle is the first to mention the Szekely runiform writing. De K eza claim ed that the Szdkely were the offspring o f the Huns and that after the Conquest they joined the M agyars. However, after the conquest o f Pannonia they did not follow them to the Pannonian plain, but were granted estates in the m ountainous borderlands, together with the Vlachs, where m ingling with the V lachs, “it is said they used their letters” . The next m ention o f the w riting can be found in the late-15thcentury author T huroczy’s chronicle. Thuroczy w rites this about the 8гёке1у: “they have not yet forgotten the Scythian letters: they do not com m it these to paper in ink, but deftly incise them on sticks, in the m anner o f runes.” He remarks in the preface to his chronicle: “For even in this age o f ours, part o f the very sam e nation, which lives in the region o f Transylvania, knows how to incise letters into wood, and w ith these runes, to w rite.” The oldest know n Szekely runiform inscriptions date back to the latter h alf o f the 15th century. The Hom orodkaracsonfalva inscription is pre-1495, while the Szekelyderzs inscription, carved in a brick, is roughly the sam e age (it has been dated to 1431; however, this cannot be proved). The earliest runiform alphabet appears on the cover o f a 1483 incunabulum o f the palace library at Nikolsburg (M ikulovo, Bohemia). The cover is a later addition to the book which was acquired by Filipec (form erly the Bishop ofV arad) after the chapter o f Olomouc ordained him bishop in 1484 at the instigation o f the Hungarian King M atthias I. Scratched on a prison wall by a Szekely prisoner, the Istanbul inscription bears the date 1515. The Felsoszem ered (Slovakia) inscription dates back to 1482, but its runiform character is uncertain. M arsigli, the Italian military engineer who worked in Transylvania, copied a carved calendar o ff a cane in 1690 and took it hom e with him to Bologna. The original seem s to have been created in 1450. It is known today as the Bologna runiform in scription. In 1598, Janos Telegdi wrote a short m onograph about the Szekely runiform writing, o f w hich only copies survived. The book containing also an alphabet becam e highly fashionable in later ages, and it significantly contrib uted to a new renaissance o f 8гёке1у runiform writing. Presently, we know o f
438
Recent research and studies
N ik o lsb u rg
1.
2.
3. 4. 5.
1 X T Ф f
a
[a]
b
b
[b]
ch
с
[c]
encz
enc
[ents]
eczk
eck
[etsk]
cs
[t§]
4 X
eb ecz
ecz, ech (ecs)
7.
Ю
encz, ench (encs)
r X 1
ed
11 .
2
e
12.
0
f
13.
A и /
egh (eg)
10.
14. 15.
ey (ej)
Pt
czy
[ents]
7 © A
eg (egy)
18.
X X
eh
19.
г
i
20.
г
ac (ak)
22.
ets
ed
ei e
17.
X A
eb
4 X
f
d
and
eng (engy)
21.
И
t
© Ж X
16.
a
ecz
M
9.
T ra n scrip tio n
a
а
6.
8.
M o d ern H u n g arian tra n slite ratio n
Telegdi
M isk o lczi
athy (aty)
f
eg gy
7 j
[d]
and
[and]
j
Ш
3C e e ® f f A gh g gy gy f
[e]
[f] [g]
[dj] [endj]
X
ty ty
ety
ech ech
X 1 Л/
ah i ek
A
el
HI
hh
[hh]
X
h
h
[h]
f A-
i
i
[i]
к
к0 >ack)
к (back)
unk
[unk]
1
[1]
v n c(u n k ) l
d
A
1
Figure 72 The earliest alphabets of the Sz6kely runiform script I
439
The Szekely run ifo rm script
N ikolsburg
M iskolczi
M o d em H u n g arian tra n slite ratio n
Telegdi
T ra n scrip tio n
23.
0
ely (ely)
0
ely
0
ly
ly
[j]
24.
*
m
rn
a
m
m
[m]
25.
•)
n
n
0
n
n
[n]
26.
D
eny (eny)
£ 3 D
eny
CD
nny ny
[nj]
27.
V
nye (nj)
nj
[nj]
28.
0
0
0
о
0
0
[0]
ep
Э
ер
P
P
[p]
mp
[mp]
29. 30.
t) 3
emp
*
31.
/\ \/
ek
32.
w
r
33.
Y
34.
A
35.
О
ак
О
к
к (front)
к (front)
и /
ег
H
r
r
[r]
|s, e(~ch (ezs)
у
ezf
r
ss
zs
m
Too
л
ef
Л
s
s
Й
1
eth (et)
ч
et
T
t
t
[t]
36.
лр
enth (ent)
ent
[ent]
37.
W
v(u )
38.
n
ew (ev)
39.
* П I
её (eo, 6)
40. 41. 42.
\г
43.
м м О: ч I й
u
Ю
u
u
M
eu
/Л
V
V
[v]
6
X
0 ii
0 ii
V fz
sz
a w
[o ]
z
[ii] M И
ezt
zt
[zt]
w (Q) s (sz) ez
e|z ez
h И
z
44.
y\
eft
St
[st]
45.
ф
us
us
[us]
Figure 73 The earliest alphabets of the Szekely runiform script II
440
Recent research and studies
some 14-15 runiform inscriptions w hich date back to the tim es before Telegdi, or are independent o f it. I cannot accept the view that the whole script was in fact invented at the court o f the Hungarian King M atthias I. However, we have am ple evidence that it w as know n in his court and that it was very popular. It m ight have been brought into fashion there, and new characters m ight have been added to it then. The origin o f this script is still vague. O nly a few features have been identified about the writing itself. The letter forms w ere greatly influenced by the fact that they were incised. The writing was read left to right and the vowels were usually only added in if they signified long ones. These latter traits m ight suggest a connection with the Sem itic scripts. Yet the Sz 6 kely script has a num ber o f ligatures and contractions which is characteristic o f m ediaeval Latin writing. This m akes it certain that the Szekely runiform script developed (further) under the influence o f Latin writing. We know the script already existed in the 13th century and we have know ledge about its letters, styles and orthography dating from the 15th century. Therefore, we should attem pt to reconstruct its m ajor phases o f development. Only two very sim ple Szekely letters (s and n) m ake a flaw less m atch with the East Turk runiform writing. Four letters (a, e, o ,f) are positively rooted in G reek writing, via Slavic m ediation; sim ilarly h and 1, to a lesser certainty. It cannot be established w hich the m ediating Slavic alphabet was, because the sym bols o f a an d / w ould suggest the Cyrillic, while e and о the Glagolithic script. O w ing to the fact that G lagolithic w riting (after 870) cam e earlier than Cyrillic (betw een 893 and 927 at the court o f the B ulghar ruler Sim eon), it is possible that the letters o f the Szekely alphabet in question go back to some m utual period. In any case, the early historical period o f Szekely w riting, when Greek letters entered the alphabet, cannot be earlier than the late 9th century. The fact that both the sym bols d e n o tin g /a n d h were borrow ed from another alphabet suggests a language which lacked these consonants altogether. Thus it w as not before the language needed these letters that it borrow ed them from another language. Both the Slavic and the M agyar languages w ere like that. The 10th-century H rabar writes that the Slavs originally used a runiform script, follow ing which they shifted to the m ixed usage o f G reek and Latin letters. Then came Cyril who invented “Slavic” writing, that is, Glagolithic. Thus, it is probable that the addition o f h and / served Slavic needs. The consonants h and / are secondary in Slavic, and are not part o f the initial phonetic range. They can be “age- dated” by observing their relative position in the Glagolithic alphabet: they stand near the very end, side by side. The sym bols denoting a, e and о were probably borrow ed because the initial script
The Szekely ru n ifo rm script
441
had a different system for signifying vowels, or its consonants did not resem ble the borrow ed ones. It is rem arkable that in the Szekely script the symbols o f the consonants sh and 1, r and z are related. It appears that the symbol o f sh is a derivation o f the Greek lam bda, and the sym bol o f / is formed by the addition o f two diacritical marks to sh\ while the symbol o f r is simpler, and z is derived from r plus two diacritical marks. The sym bol o f the Hungarian letter g y (as in the B ritish English ‘duke’) is the letter d plus a diacritical mark. The sym bol o f the Hungarian letter ny (as in the B ritish English ‘new ’) is clearly a com bination o f the sym bols o f n and y. These points refer to the internal developm ent o f the writing. R econstructing the internal developm ent o f an alphabet enables us to restore the older, sim pler alphabet com prising few er sym bols. In any case, the reconstruction reveals that the Szekely runiform writing underw ent significant changes betw een the 10th and the 15th centuries. It first encoun tered Slavic writing, and then Latin literacy. It is a w ell-know n fact that a runiform script was em ployed in the Carpathian Basin in the Late Avar period. Several dozens o f samples o f it have come down to us, and it has been term ed N agyszentm iklos-Szarvas runiform script after the two locations w here the m ost significant samples were brought to light. We also know that this writing belongs to the family o f Eastern European scripts (see pp. 404-406) o f which m ore and m ore specim ens are being found in the region o f the form er K hazar Empire. It m ay be due to chance that only one C onquest-period object has been unearthed with a short inscription written in this Eastern European writing (from a site at Hom okm egy-H alom , Kalocsa, Hungary). Nevertheless, we are still lacking two or three links connecting the Szekely runiform script with Late Avar period and other Eastern European scripts. Hopefully, these m issing links will be found shortly. The Hungarian words ir 4o w rite’ and betu ‘letter (character)’ are o f Turkic origin, therefore it would be expected that the M agyars adopted the Turk variant o f the Eastern European runiform script. This is supposed to have been supplem ented w ith the letters via Slavic m ediation. It is possible that those missing links should be looked for on the notches on bricks o f llth -1 3 th century Hungarian churches. W hether the A rpadian conquerors brought the writing with them, or adopted the script in the C arpathian Basin we do not know; however, it is certain that before the foundation o f state and the adoption o f Latin literacy, the M agyars had their own writing. The Latin alphabet superseded it, but fond m em ories o f it lived on into the Renaissance. The old Hungarian writing was revived, as were the Germ anic runes in the Scandina vian courts. It could be possible that in the 15th century the Szekely alphabet was artificially transform ed to som e extent, which hinders today the recon-
442
Recent research a m i studies
Figure 74 The runiform inscription o f H om okm egy-H alom (near Kalocsa, South Hungary) from the tim e o f the C onquest
Figure 75 The four sides o f the Szarvas needle-case
The Szekely ru n ifo rm script
443
struction o f the oldest letter forms and orthography. It is certain, however, that the 8гёке1у runiform script was directly related to the M agyars o f the C on quest period. This brings us to the problem o f the Sz 6kelys. First m ention o f this group dates back to the 12th century. The descriptions o f the battles on the Rivers Olsava (1116) and Lajta (1146) respectively speak o f the Б г ё к е ^ , togeth er with the Pechenegs, as forming the advance guard o f the M agyar forces. Their first m ention in a docum ent is an entry in the 1217 register o f Varad which talks o f a szekelyszaz ‘Hundred 8гёке1у’ (centuriatus Sceculzaz) unit. A charter o f 1250 tells us that they had participated in a m ilitary cam paign o f the count o f Szeben in 1210. Toponymic data point to the fact that the Szekelys must have served as border guards in the 11 t h - 12th centuries. The Bihar county Б г ё к е ^ m igrated to their later settlem ent areas as o f the 12th century. They reached the Carpathians by the 13th century. Follow ing m any internal m igra tions and wars, at the end o f the 13th century the areas collectively term ed Szёkelyland began to emerge. Their uniquely autonom ous system o f szeks ‘settlem ents’ served to m aintain their independence, while the other 8 гёк е 1у posts in West Transdanubia, the O ^ g , in M oson, near B ratislava and in Baranya county soon assim ilated into the M agyar population. The 8гёке1у’з lan guage shows no trace whatsoever o f Turkic origins, as has been suggested. Their language preserves old Hungarian characteristics. The only argum ent in support o f their foreign origin is that according to the old nom adic system it was the m ost recently joined peoples that served as border guards and were sent to the front lines o f battle. This, in itself, is sim ply not convincing enough. It was indeed nom adic practice to get the m ost recently joined peoples to perform the m ost dangerous tasks, but one cannot say that those groups who were sent to do risky jobs were always newcom ers. The fact that this runiform writing only survived am ong the 8 гёке 1уз does not shed m uch light on their history. The writing known today is alm ost certainly a M atthias-period trans formed version, and the unique historical circum stances m ust have contrib uted to the survival o f the script after the 15th century and later. Yet the fact that de К ёга m entions the special writing o f the Б г ё к е ^ as early as around 1285, suggests that the Szekelys or Szёkelyland played an im portant part in preserving the script.
444
Recent research and studies
NOTES For the latest research results of the Turkic and Hungarian runiform script, see pp. 404-406 of this book, and Sandor (1991, 1992a, 1992b); for the background see Rona-Tas (1987b, 1988b, 1992a, 1992c). V asary (1974) and Gobl-Rona-Tas (1995) give a good overview of the research history. More about the Hungarian words ir ‘to write’ and betu ‘letter (character)’ can be found in Rona-Tas (1992c). I am not going into the discussion of the problem of the origin of the Szekely. There are no conclusive arguments in favour that the Szekely group would be of non-Hungarian origin.
APPENDIC ES
BIBLIOGRAPHY Theforeign words and names in the titles o f the bibliography appear below in scientific transliteration
A magyar ostortenet kerdesei [The questions o f Hungarian proto-history], A M agyar N yelvtudom anyi Tarsasag vitaiitese, 1953. decem ber 1. B uda pest 1955. ABAEV, V.I. ( 1 9 5 8 -1 9 8 9 ) , Istoriko-etim ologiceskij s lo v a r ’ osetinskogo ja zyka I - I V [Historic-etym ological dictionary o f the Ossetic language I-IV ], M oskva - Leningrad. ADAMS, D.Q. (1988), Tocharian historical phonology a n d morphology. New Haven [cf. comments o f RlNGE, D.A. (1990), Language 66, 400-408]. AFANAS’EV, G .E . (1993), D onskie alany [Alan people living on the Don], Moskva. ALFOLDI, A . (1932), A tarchan mdltosagnev eredet 6 rol [On the origin o f the title tarkhan]. M agyar N yelv 28, 205—220. ALTHEIM, F. - STIEHL, R. (1967), Die A raber in der Alten Welt, IV. Neue Funde. Berlin. ANDERSON, B. (1 9 9 0 ), Im agined communities. Reflections on the origin and spread o f nationalism. L ondon-N ew York. ANTTILA, R . (1 9 8 9 ), H istorical and comparative linguistics. A m s te r d a m P h ila d e lp h ia . ARTAMONOV, M .I. (1 9 6 2 ),Istorijahazar [History o f the Khazars]. Leningrad. ASMARIN, N.I. (1929), Slovar ’ cuvasskogo ja zyka II [Dictionary o f the C hu
vash language II]. K azan’. D. (1966), Btigd Najram dah M ongol A rd Uls gazarziijn atlas [The atlas o f the M ongolian P eople’s Republic], Ulaanbaatar. BAILEY, H.W. (1979), D ictionary o f Khotan Saka. Cambridge. BAKAY, K. (1967), Archaologische Studien zurF rage der ungarischen Staatsgrundung. Budapest. BAKAY, K. (1971), Scythian rattles in the Carpathian Basin a n d their eastern connections. Budapest. BAKAY, K. (1978), A m agyar allamalapitas [Foundation o f the Hungarian state], Budapest. B A D A M ZA V ,
448
Appendices
BALASSA, 1.(1994), A m agyar foldm uvetes em lekei a 9-10. szazadban [Relics o f H ungarian agriculture o f the 9th—10th centuries], in: K OvA cs (1994), 235-246. BALINT, Cs. (1971), A kutya a X -X II. szazadi m agyar hitvilagban [The dog in 1 0 th -12th-century Hungarian religious beliefs]. A M ora Ferenc Mitzeum Evkdnyve, 295-315. BALINT, Cs. (1976), A m agyar ostortenet regeszet 6 nek keleti kapcsolatai es elo zn ^ n y ei, a Karpat-m edence V I-IX . szazadi regeszeti-neptorteneti kerdesei a honfoglalaskor [The eastern connections and prelim inaries o f the archaeology o f Hungarian proto-history. Issues o f the archaeology and genealogy o f the C arpathian B asin in the 6 th—9th centuries], in: H ajd u KRISTO - R o n a -T as (1976-1982) 1/1, 24-163 and II (Tables). BALINT, CS. (1984), review on WERNER (1984). A cta A rchaeologica Hung. 36, 263-269. BALINT, Cs. (1988), N ochm als tiber die Identifizierung des Grabes von Kuvrat. A cta O rientalia Hung. 42, 377-389. BALINT, CS. (1989a), Exkurs: die “doppelte ungarische Landnahm e” : D ieA rchaologie der Steppe: Steppenvdlker zw ischen Volga undD onau vom 6. bis zum 10. Jh. W ien-K oln, 233-235. BALINT, CS. (1989b), D ie Archaologie der Steppe. W ien-K oln. BA r CZI, G. (1951), A Tihanyi A patsag alapltolevele m int nyelvi em lek [The deed o f foundation o f the Tihany A bbey as a linguistic m onum ent], B uda pest. BARCZI, G. (1952), Torokjovevenyszavaink legregibb retegenek kerdeseihez [On the questions o f the ancientm ost layers o f Turkic loan w ords in the H ungarian language], A z MTA I. Nyelv- es Irodalom tudom anyi Osztalyanak K o zlem in yei 2, 347-359. BARCZI, G. (1958), A m agyar szdkincs eredete [The origin o f the Hungarian word-stock]. Second, enlarged edition, Budapest. B A rc zi, G. (1965), A propos des vieux mots d ’em prunt turcs du proto-hongrois. A cta O rientalia Hung. 18,47-54. BARCZI, G. (1971), Le traitement s et de z turcs dans les mots plus anciens mots d’emprunt turcs du hongrois, in: LIGETI (1971), 383-390. B A rc zi, G. (1972), Quelques conclusions tirees de l ’etude des plus anciens m ots d ’em prunt turcs du hongrois. A cta Orientalia Hung. 25, 383-390. BARTHA, A. (1987), A m agyar honalapitas [The foundation o f the M agyar hom eland]. Budapest. BARTHA, A.-CZEGLEDY, K . - R o n a -T a s , A. (Eds) (1977), M agyar ostdrteneti tanulm anyok [Studies in Hungarian proto-history]. Budapest. BASKI, I. (1986), A prelim inary index to R asonyi s O nom asticon Turcicum. Budapest.
Bibliography
449
BAZIN, L. (1981-1982), Pour une nouvelle hypothese sur l ’origine des kha zars. M aterialia Turcica 7 -8 , 51-71. BELYH, S. (1996), Esco raz ob ё й ю ш т е ar. Finno-ugrovedenie 3, 86-93. BENKO, L. (Ed.) (1967-1984), A m agyar nyelv torteneti-etim ologiai szotara I - I V [The historical-etym ological dictionary o f the Hungarian language I-IV ], Budapest. BENKO, L. (1984), Z ur Geschichte des Ungam tum s vor der Landnahm e im Zusam m enhang m it Leved und Etelkoz. A cta Linguistica Hung. 34, 158— 173. BENKO, L. (1985), A m agyarsag honfoglalas elotti toilenetehez. Leved ёв Etelkoz kapcsan [On the pre-Conquest history o f the M agyars, apropos o f Leved and Etelkoz]. A M agyar N yelvtudom anyi TarsasagK iadvanyai, 172, Budapest, 7-37. BENKO, L. (Ed.) (1992-1995), Etym ologisches Worterbuch des Ungarischen / - / / / . Budapest. BENKO, L. (1995), A helynevek szerepe az Arpad-kori n 6pessegtort 6 neti kutatasokban [The role o f toponym s in the genealogical research o f the Arpad period], in: KOVACSICS (1995b), 96-106. BENKO, L. (1998), Hogyan is hivtak Arpad dёdunokajat? [W hat was A rpad’s great-grandson called?]. M agyar N yelv 94, 2, 157-165. BENKO, M. (1992/93), Burial masks o f Eurasian mounted nom ad peoples in the m igration period (1st m illennium AD). Acta Orientalia Hung. 46, 113-131. BERECZKY, G. (1992), Grundziige der tscheremissischen Sprachgeschichte II, Szeged. BERECZKY, G. (1994), Grundziige der tscheremissischen Sprachgeschichte I, Szeged. BERECZKY, G. - DOMOKOS, P. (Eds) (1983), Uralisztikai tanulmanyok. H ajdu P eter 60. sziiletesnapja tiszteletere [Studies in Uralistics. In honour o f the 60th birthday o f Р ё 1ег Hajdu]. Budapest. BERTA, A. (1989), Uj Уё1етёпу torok eredetti torzsneveinkrol [New views on M agyar tribal nam es o f Turkic origin], Keletkutatas, Spring, 3-17, 121- 122 . BERTA, a . (1990a), Historische Zeugnisse der ungarischen Stam m esnam en,
in: BRENDEMOEN, B. (Ed.), Altaica Osloensia. Proceedings o f the 32nd m eeting o f the Perm anent International Altaistic Conference. Oslo, June 12-16, 1989. Oslo, 2-38. BERTA, A. (1990b), Ungarische Stam m esnam en ttirkischen Ursprungs. UralAltaische Jahrbucher, Neue Folge 9, 31-37. BERTA, A. (1991), Torok eredetu torzsneveink [M agyar tribal nam es o f Turkic origin], N yelvtudom anyi K ozlem enyek 92, 3-40.
450
Appendices
BERTA, a . (1992a), A tiirk hadi m uszavak egyik jelentestani csoportja [One sem antic group o f Turkic m ilitary term s], K eletkutatas, Fall, 20-26, BERTA, A. (1992b), Die chasarische B enennung der U ngam , in: FRAGNER, C h . - SCHWARZ, K. (Eds), F estgabe an J o s e f Matuz. O sm anistik-Turkologie-D iplom atik. Berlin 1992, 7-11. BERTA, A. (1994), Zum Wandel des auslautenden -G im Kiptschakischen. Jour nal o f Turkology 2, 163-195. BE$EVLIEV, V. (1963), D ie Protobulgarischen Inschriften. Berlin. BE§EVLIEV, V. (1979), Parvo-balgarski nadpisi [Proto-Bulghar inscriptions].
Sofija. BEUMANN, H. - SCHRODER, W. (1985), Frtihm ittelalterliche E thnogenese im Alpenraum . Sigmaringen. BIRO, м . (1976), Gniz forrasok [The A rm enian sources], in: HAJDU - KRISTO - R o n a -T as (1976-1982), 1/2,274-278. ВОВА, I. (1967), Nomads, N orthm en and Slavs. Eastern Europe in the Ninth Century. The Hague. ВОВА, I. (1982-1983), A tw ofold conquest o f H ungary or “ Secundus ingressus” . Ungarn-Jahrbuch 12. Zeitschrift fur die K unde U ngam s und verwandte Gebiete, 23-41. BOBA, I. (1984), “Abodriti qui vulgo praedecenti vocatur” pro “m arvani praedecenti”? Paleobulgarica 8 , 29-37. BOBA, I. (1991), The federal structure o f the earliest H ungarian polity: The role o f the “C ovenant o f Blood”. U ral-Altaische Jahrbiicher [USA], 63, 99-122. BOBA, I. (1993), Awaren in M itteleuropa. Ural-Altaische Jahrbiicher [USA], 65, 4 6 -72. BOGYAI, T. VON (1960). Die Kirchenorte der Conversio B agoariorum et Carantanorum . M ethoden und M oglichkeiten ihrer Lokalisierung. Siidostfo rschungen 19, 52-70. BOJKO, K.A. (1977), Arabskaja istoriceskaja literatura v Ispanii (V H I-pervaja tre t’ XI v.) [Arab historical literature in Spain (from the 8 th to the first third o f the 11th centuries)]. M oskva. BOL$AKOV, O.G. - MONGAJT, A.L. (1971), Putesestvie A bu H am ida Al-G arnati v Vostocnuju i C entral’nuju Evropu (1131-1153) [Abu H am id al-Garn ati’s journey in East and Central Europe (1131-1153)], M oskva. BOL§AKOV, O.G. - MONGAJT, A.L. (1985), A bu-H am id a l-G am ati utazasa K elet es K ozep-Europaban 1131-1153 [The journey o f A bu-H am id alG am ati in East and Central Europe, 1131-1153]. Budapest. BONA, I. (1968), A briss der Siedlungsgeschichte U ngam s im 5 -7 . Jahrhundert und die Awarensiedlung von Dunaiijvaros. Archeologicke rozhledy 20, 605-618.
Bibliography
451
BONA, I. (1971), Ein Vierteljahrhundert V olkerw anderungszeitforschung in Ungarn (1945-1969). A cta Archaeologica Acad. Sc. Hung. 23. BONA I. (1984a), А кёвб avar korszak [The late Avar period], in: SZEKELY (1984), 327-329. B o n a , I. (1984b), A z avarok. A Karoling Pannonia [The Avars. The Carolingian Pannonia], in: SZEKELY (1984), 310-369. BONA, I. (1985), Die Verwaltung und die Bevolkerung des karolingischen Pannoniens im Spiegel der zeitgenossischen Quellen. M itteilungen des Archaologischen Instituts der Ungarischen Akadem ie der Wissenschaften, 14,149-160. BONA, I. (1988), Die Geschichte der Awaren im Lichte der archaologischen Quellen. Settim ane di studio del Centro italiano di studi sull alto medioevo, 35 ,4 3 7 -4 6 3 . BONA, I. (1990a), Beitrage zum asiatischen Ursprung der aw arenzeitlichen partiellen Pferdebestattungen. Wosinsky M or M uzeum Evkdnyve. 15, 113-124. BONA, I. (1990b), Hunnen, in: FLECKENSTEIN, J. (Ed.), Lexikon des M ittelalters. Bd. 5. M unchen, 222-224. BONA, I. (1990c), I Longobardi in Pannonia, in: MENIS, G.C. (Ed.), Longo bardi. M ilano, 14-73. BONA, I. (1990d), Byzantium and the Avars: the archeology o f the first 70 years o f the Avar era, in: A USTIN, D. - ALCOCK, L. (Eds), From the Baltic to the B lack Sea. Studies in M edieval archeology. London, Boston etc., 113-117. BONA, I. (1990e), Z eit des ungarisch-slaw ischen Zusam m enlebens, in: KOPECZI, B. (Ed.), K urze Geschichte Siebenbiirgens. Budapest, 895-1172. BONA, I. (1990f), V o lk e r w a n d e r u n g u n d F r iih m itte la lte r s (271-895), in: KOPECZI, B . (Ed.), K urze Geschichte Siebenbiirgens, B u d a p e st, 62-106. BONA, 1. (1991), Das Hunnenreich. B udapest - Stuttgart. BONA, I. (1993), “Barbarische” Nachahm ungen von byzantinischen Goldmiinzen im Awarenreich. Rivista Italiana di N um ism atica 95, 529-538. BONA, 1.(1994a), Az Avar Birodalom vegnapjai. Vitak 6 s uj eredm enyek [The wane o f the Avar Empire. Argum ents and new results], in: KOVACS (1994), 67-75. BONA, I. (1994b), From Dacia to Transylvania. The period o f the great m i grations (271-895), in: B O N A , I. - KOPECZI, B. et al. (Eds), H istory o f Transylvania. Budapest, 62-106. BONA, I. (1994c), The H ungarian-Slav period (895-1172). in: B O N A , I. KOPECZI, B. et al. (Eds), H istory o f Transylvania. Budapest, 109-177. BONA I. (1997a), Dienes Istvan (1929-1995), in: A N yiregyhazi Josa Andras Muzeum Evkonyve 37-38. Nyiregyhaza, 11-33.
452
Appendices
BONA, I. (1997b), A honfoglalo m agyarsag regeszeti hagyatekanak tarsadalom torteneti tanulsagai [The socio-historical lessons o f the archaeological m aterial o f the conquering M agyars], M agyar Tudomany 12, 1451-1461. BOOR, C, DE (Ed.) (1903), Excerpta de legationibus. Berolini. BOOR, C. DE - WlRTI-I, P. (Eds) (1972), Theophylacti Sim ocattae Historiae. Stutgardiae. BROMLEJ, J.V. (1973), Etnos i etnografija [Ethnos and ethnography], Moskva. BROMLEJ, J. V. (1976), Etnosz i s neprajz [Ethnos and ethnography], Budapest. BUDENZ, J. (1873), Jelentes V am bery A. m agyar-torok szoegyeztet 6 seirol [Report on Arm in V am bery’s word correlations], N yelvtudom anyi Kozleт ёпуек 10, 67-135. BURY, J.В. (1912), ^ history o f the E ast Rom an Em pire fro m th e fa ll o f Irene to the accession o f Basil I. (802-867 AD). London. BYNON, T h. (1977), H istorical linguistics. C am bridge-L ondon-N ew YorkM elboum e. CANETTI, E. (1977), D ie gerettete Zunge. Geschichte einer Ju g en d (3. Aufl.). M tinchen - Wien. CANETTI, E. (1982), A m egorzott nyelv. E gy ifjiikor tortenete [The preserved language. History o f a period o f youth]. Budapest. CERNYH, E.N. - KUZ’MINYH, S.V. (1994), Die alteste M etallurgie Nordeurasiens, Problem e der W echselbeziehungen zw ischen den erzeugenden Zentren. Journal de la Societe Finno-O ugrienne 85, 21-39. C h a l m e t a , P. (1976), La m e d ite rra ^ e occidentale et al-A ndalus de 934 a 941: les donnee d ’lbn Hayyan. Rivista degli Studi Orientali 50, 337-351. CHALMETA, P. (Ed.) (1979), Al-M uqtabas (V) de Ibn Hayyan. M adrid. CHAVANNES, E. (1897a), Voyageurs chinois chez les khitan et les joutchen I. Journal Asiatique 1, 377—442. CHAVANNES, E. (1897b), Le nestorianism e et l’inscription de K ara Balgassoun. Journal A siatique 1, 43-85. CHAVANNES, E. (1900), D ocum ents sur les tou-kiue (turcs) occidentaux. Paris. CLAUSON, G. (1956), A note on Qapqan. Journal o f the R oyal A siatic Society 124-138. CLAUSON, G. (1962), Turkish and M ongolian Studies. London. CLAUSON, G. (1972), An etym ological dictionary o f pre-thirteenth-century Turkish. Oxford. CROSS, S.H. - SHERBOWITZ-WETZAR, O.P. (1953), The Russian primary chronicle. Laurentian Text. Cambridge, Mass. CS. SOS, A. (1973), D ie slawische Bevdlkerung Westungarns im 9. Jahrhundert. Miinchen. CSAPODI, Cs. (1978), Az Anonym us-kerdes tortenete [The history o f the Ano nym us question]. Budapest.
Bibliography
453
CSONGOR, В. (1993), A kinai fo rra so k az azsiai avarokrol [Chinese sources
on the A sian Avars]. Budapest. CSONGOR, B. - FERENCZY, M. (1993), A kinai nevek es szavakm agyar atirasa
[The transcription o f Chinese names and words in Hungarian], Budapest. CZEGLEDY, K. (1950-1951), Zur M esheder Handschrift von Ibn Fadlans Reisebericht. A cta Orientalia Hung. 1, 217-260. CZEGLEDY, K. (1956), A m agyarnepnev le g r e g ib b e lo fo r d u la s a i a fo r r a so k ban [T h e e a r lie st o c c u r r e n c e s o f th e e th n o n y m M agyar in th e s o u r c e s ], in: Pais Dezso Emlekkonyv. B u d a p e st, 373-385. CZEGLEDY, K. (1959), A szavard к ё ^ ё в Thury Jozsef elott es utan [The Savard question before and after Jozsef Thury]. M agyar N yelv 60,373-385. CZEGLEDY, K. (1963), A regi torok torzsek num erikus felosztasanak kerdёsёhez [On the question o f the num erical division o f the ancient Turkic tribes], M agyar N yelv 59, 456—461. CZEGLEDY, K. (1966), Das sakrale Konigtum bei den Steppenvolkem . Numen. International Review fo r the H istory o f Religions 13, 14-26. CZEGLEDY, K. (1969), N om ad пёрек vandorlasa N apkelettdl N apnyugatig [The m igration o f nom adic peoples from Orient to Occident], B u d a p e st. CZEGLEDY, K. (1971), Pseudo-Zacharias Rhetor on the N om ads, in: LIGETI, L. (Ed.), Studia Turcica. Budapest, 133-148. CZEGLEDY, K. (1972), On the num erical com position o f the ancient Turkish tribal confederations. A cta Orientalia Hung. 25, 275-281. CZEGLEDY, K. (1973), Gardizi on the history o f Central Asia (145-780 AD). Acta Orientalia Hung. 27, 257-267. CZEGLEDY, K. (1974), A szakralis kiralysag a steppei nepeknel (a kazaroknal es a m agyaroknal) [The sacral kingship am ong the nom adic peoples (the Khazars and the M agyars)]. M agyar N yelv 70, 11-17. CZEGLEDY, K. (1975), Arpad es Kurszan (Az Arpad-haz m egalapitasahoz) [Arpad and Cursan (On the founding o f the Arpad dynasty)]. Zalai Tiikdr 2 ,4 3 -5 8 , in: CZEGLEDY (1985), 113-128. CZEGLEDY, K. (1 9 7 6 ), Etim ologia ёв filologia (B olgar-torok jбvevёnyszavaink atvetetenek torteneti hattererol) [Etym ology and philology (on the historical background o f the adoption o f B ulghar-Turkic loan words in the Hungarian language)]. N yelvtudom anyi E rtekezesek 89 , 8 2 - 8 9 . CZEGLEDY, K. (1977), A m agyar ostortenelem irott forrasainak kesziilo uj kiadasarol [On the new edition o f the written sources o f Hungarian prehis tory], in: BARTHA - CZEGLEDY - R O N A -T A S (1977), 323-325. CZEGLEDY, K. (1979), Uj arab forras a m agyarok 942. evi spanyolorszagi kalandozasarol [New Arabian source on the M agyar raids on Spain in 942]. M agyar N yelv 75, 273-282.
454
Appendices
CZEGLEDY, К. (1981), M 6 g egyszer a m agyarok 942. evi spanyolorszagi kalandozasarol [Again on the M agyar raids on Spain in 942]. M agyar N yelv 7 7 ,4 1 9 —423. CZEGLEDY, K. (1983 a), From East to West: The age o f nom adic m igrations in Eurasia. Translated by P. Golden. Archivum Eurasiae M edii A evi 3 ,2 5 -1 2 5 . CZEGLEDY, K. (1983b), Zur Stam m esorganisation der tiirkischen Volker. Acta Orientalia Hung. 36, 89-93. CZEGLEDY, K. (1983c), Zur Geschichte der Hephtaliten. A cta A ntiqua Hung. 2 8 ,2 1 4 -2 1 7 . CZEGLEDY, K. (1985), M agyar ostdrteneti tanulm anyok [Studies in H ungar ian proto-history], Budapest. CZEIZEL, E. (1990), A m agyarsag genetikaja [The genetics o f the M agyars], Debrecen. DANKOFF, R. (1972), K asgari on the tribal and kinship organization o f the Turks. Archivum Ottomanicum 4, 23-43. DANKOFF, R. - KELLY, J. (1982-1985), M ahm ud al-K asgarl Compendium o f the Turkic dialects. Harvard University Printing Office. DARKO, E. (1924), Z ur Frage der urm agyarischen und urbulgarischen Beziehungen. K orosi Csoma-Archivum 1, 292-301. DEER, J. (1945-1946), A IX. szazadi m agyar tortenet idorendjehez [On the chronology o f 9th-century Hungarian history], Szazadok 7 9-80, 3-20. Di CAVE, C. (1995), L'arrivo degli ungheresi in Europa e la conquista della patria. Fonti e letteratura critica. Spoleto. DIENES, I. (1963), Honfoglaloink halottas szokasainak egyik ugor kori eleтёгб1 [On one Ugrian elem ent o f the burial custom s o f the conquering M agyars]. Archaeologiai Erteslto 90, 108-113. DIENES, I. (1972a), A honfoglald magyarok [The conquering M agyars]. Bu dapest. DIENES, I. (1972b), The H ungarians cross the Carpathians. Budapest. DIENES, I. (1978), A honfoglalo m agyarok telekhiedelm einek regeszeti bizonysagai [The archaeological evidence o f the spiritual beliefs o f the con quering M agyars], in: HOPPAL-ISTVANOVICS (1978), 28-39. DIENES, I. (1992), A Kalocsa kom yeki Rovasem lekrol [On the K alocsa runi form inscription], in: SANDOR (1992b), 31-40. DIENES, I. (1994), Landnahm ezeitliche K erbinschrift aus dem G raberfeld von H om okm egy-H alom in der Um gebung von Kalocsa. Folia Archaeologica 4 3 ,1 6 7 -1 7 9 . DOBLHOFER, E. (1955), Byzantinische D iplom aten und ostliche Barbaren... ausgewahlte Abschnitte des Priskos und M enander Protektor... Graz. DOERFER, G. (1963—1975), Turkische undm ongolische Elem ente im Neupersischen I-IV . W iesbaden.
Bibliography
455
DONNER, К. (1924), Zu den altesten Bertihrungen zw ischen Sam ojeden und Ttirken. Journal de la Societe Finno-O ugrienne 40, 3-42. DORRIE, H. (1956), Drei Texte zur Geschichte der Ungarn und M ongolen. Gottingen. DOWSETT, C.F.J. (1962), H istory o f the Caucasian Albanians. London. DUBLER, C.E. (1953), A bu H am id el Granadino у su relacion de viaje p o r tierras Eurasiaticos. Madrid. DUMMERTH, D. (1971), Almos fejedelem mi'tosza 6 s valosaga [Prince Almos: myth and reality], Filoldgiai K ozlony 17, 404-430. DUMMERTH, D. (1977), A z Arpadok nyomaban [On the trail o f the Arpads], Budapest. DUMMERTH, D. (1986), Almos az aldozat [Almos, the sacrifice], Budapest. DUNLOP, D.M. (1954/1967), The history o f the Jew ish K hazars. Princeton. ECSEDY, I. (1976), K in a i fo r r a so k [T h e C h in e s e s o u r c e s ], in: HAJDU - KRISTO - R O N A-TA S (1976-1982), 1/2, 283-305. ECSEDY, I. (1977), A Korosi Csoma Tarsasag ostorteneti konferenciaja (1973. novem ber 21 -2 3 .) [The prehistory conference o f the Csoma Korosi Society, in: B a r t h a - C z e g l e d y - R o n a -T a s (1977), 327-332. ECSEDY, I. (1979), N om adok es kereskedokK ina hatarain [Nomads and m er chants on the borders o f China], Budapest. ECSEDY, I. (1983a), A nom ad gazdasag es tarsadalom probl 6 mai a tortdnelemben es a kutatasban [The problem s o f nom adic econom y and society in history and in research], in: TOKEI (1983), 15-40. ECSEDY, I. (1983b), A z e u r a z sia i lo v a s n o m a d o k v ila g a m in t a fo ld m u v e s k o z o s s 6 g e k ‘Ч б й ё п е 1 т 1 k o m y e z e t e ” [T h e w o r ld o f th e E u r a sia n e q u estr ia n n o m a d s as th e “h is to r ic a l e n v ir o n m e n t” o f a g ricu ltu ra l c o m m u n itie s ], in: TOKEI (1983), 41-50. ECSEDY, I. (1987), A kinai allam kezdetei [The beginnings o f Chinese state
hood], Budapest. EGGERS, M. (1995), ‘Das Grossmahrische Reich ’. Realitat oder Fiktion. Eine
N euinterpretation der Quellen zu r Geschichte des Mittleren D onauraum es im 9. Jahrhundert. Stuttgart. EGGERS, M. (1996), D as Erzbistum des Method. Lage, Wirkung undN achleben der kyryllomethodianischen Mission. M unchen. ELTER, I. (1981), ^ h a n y m egjegyz 6 s Ibn Hayannak a m agyarok 942. evi spanyolorszagi kalandozasarol szolo tudositasahoz [R em arks on Ibn H ayyan’s report on the M agyar raids on Spain in 942]. M agyar N yelv 77, 413-419. EMMERICK, R.E. (1979), A guide to the literature o f Khotan. The Reyukai Library, Tokyo.
456
Appendices
EMMERICK, R.E. (1984), R esearch on Khotanese: A survey (1979-1982), in: SKAMKOWSKI, W. - TANGERLOO, A. VAN (Eds), M iddle Iranian Studies. Leuven, 127-145.
EMMERICK, R.E. - SKJAERVO, P.O. (1982, 1985), Studies in the Vocabulary o f K hotanese / - / / . Wien.
ENGEL, P. (1990), Beilleszkedes Europaba, a kezdetektol 1440-ig. M agyarok Eurdpaban I [Integration into Europe, from the beginnings to 1440. The M agyars in Europe I]. Budapest. ERDAL, M. (1993a), Around the Turkic “A pple”. The Journal o f Indo-Euro p ea n Studies 21, 27-36. ERDAL, M. (1993b), Die Sprache der w olgabulgarischen Inschriften. W ies baden. ERDELYI, I. (1982), A z avarsag es K elet a regeszeti fo rra so k tiikreben [The Avars and the O rient as reflected by the archaeological sources], Budapest. ERDELYI, I. (1986), A m agyar honfoglalas es elozm enyei [The M agyar C on quest and its prelim inaries]. Budapest. ERY, K. (1983), Com parative statistical studies on the physical anthropology o f the C arpathian Basin population between the 6 th—12th centuries AD. A lba R egia 20. Szekesfehervar, 89-141. ERY, К . (1994), A Karpat-m edence em bertani кёре a honfoglalas koraban [An anthropological view o f the Carpathian Basin in the C onquest period], in: KOVACS (1994), 217-224. ERY, K. (1995), A honfoglalas es az A rpad-kor nepessegenek em bertani vazlata [An anthropological survey o f the population o f the C onquest and the A rpad period], in: KOVACSICS (1995b), 127-130. FAHRUTDINOV, R.G. (1975), A rheologiceskiepam jatniki volzsko-kam skij Bulgarii i ее territorija [Archaeological findings in V olga-K am a Bulgharia and her territory]. K azan’. FASMER [Vasmer], R. (1964-1973), E tim o lo g ic eskijslo va r’ russkogojazyka I-IV, p ere vo d s nemeckogo i dopolnenija O.N. Trubaceva [An etym ological dictionary o f the R ussian language. Translation from G erm an and supple ments by O.N. Trubacev I-IV ]. M oskva. FEJGINA, S.A. (1972), Istoriografija jevrejsko-hazarskoj perepiski X.v., in: PASUTO, V.T. (Ed.), F eodal’naja Rossija vo vsem im o-istoriceskom processe. F eudal Russia in the process o f w orld history. The historiography o f H ebrew -K hazarian correspondence in the 10th century. M oskva. FERENCZI, G. (1992), A szekely rovasiras Erd61yben m a letezo em lekei. [Ex isting examples o f Transylvanian Szekely runiform script], in: SANDOR (1992b), 51-67. FERJNCZ, I. (1976), Szlav forrasok [The Slavonic sources], in: HAJDU K risto - R ona -T as (1976-1982), 1/2,200-210.
Bibliography
457
FINE, J.V.A. (1983), The early m edieval Balkans. A critical survey fro m the sixth to the late twelfth century. Arm Arbor. FLUSSER, D. (1980, 1981), The Josippon (Josephus Gorionides), vol. I. 2nd edition 1981, vol. II. 1980. Jerusalem. FODOR, I. [lin g u is t] (1961), T h e v a lid ity o f g lo t to c h r o n o lo g y o n th e b a s is o f th e S la v o n ic la n g u a g e s . Studia Slavica Acad. Sc. Hung. 7, 295-346. FODOR, I. (1963), A matem atikai m odszer nyelvtudom anyi alkalm azasanak hatarai: a nyelvi valtozasok lefolyasanak tem poja [The limits o f apply ing the m athem atical m ethod in linguistic studies: the speed o f language change], N yelvtudom anyi K ozlem enyek 65, 297-339. FODOR, I. [archaeologist] (1973), Vazlatok a fin n u g o r ostortenet regeszetebol [Sketches from the archaeology o f Finno-Ugrian prehistory], Budapest. FODOR, I. (1975), Verecke hires utjan... [“On Verecke’s im mortal pass”]. Budapest. FODOR, I. (1977), B o lg ar-to ro k jovevenyszavaink es a regeszet [Bulghar-Turkic loan words in the Hungarian language and archaeology], in: B a r t h a - C z e g l e d y - R o n a -T a s (Eds) (1977), 79-114. FODOR, I. (1982a), D ie grosse Wanderung der Ungarn vom Ural nach Pannonien. Budapest. FODOR, I. (1982b), In search o f a new homeland, the prehistory o f the H un garian people an d the Conquest. Budapest. FODOR, I. (1994), Leletek M agna H ungariatol az Etelkdzig [Finds from M agna H ungaria to the Etelkoz], in: KOVACS (1994), 47-65. F o d o r , I. (1995), Az osm agyarsag etnikai tudata [The ethnic identity o f the proto-M agyars]. N eprajzi Ertesito 77, 23-34. F o d o r , I. (Ed.) (1996a), “Oseinket felhozad...”. A honfoglalo m agyarsag. Kiallitasi katalogus [“( Peaks o f high C arpathian hills) / Once you gave our elders” . The Hungarians o f the Conquest. C atalogue o f an exhibition], Budapest. FODOR, I. (Ed.) (1996b), Kam a-videki rokonaink ostortenete. K iallitasi kata logus [The proto-history o f the relatives o f the M agyars in the K am a region. C atalogue o f an exhibition]. Tatabanya. FONT, M. (1994), Nyesztor-kronika [The N estor chronicle], in: KRISTO (1994c), 498. FRANKE, H. (1969). Bem erkungen zu den sprachlichen Verhaltnissen im Liao-Reich. Zentralasiatische Studien 3, 7—43. FRANKE, O. (1937), Geschichte des Chinesischen Reiches III. B erlin-Lepzig. FRIESINGER, H. - DAIM , F. (Eds) (1985), D ie Bayern und ihre Nachbarn. Berichte des Symposions der K om m ission fur Fruhm ittelalterforschung, 25. bis 28. Oktober, 1982, Stift Zwettl, Niederosterreich, Teil 2, Wien.
458
Appendices
H. - DAIM, F. (Eds) (1990), Typen der Ethnogenese unter besonderer В eriicksichtigung der Bayern. Berichte des Sym posions der {Com m ission fur Frahm ittelalterforschung, 27. bis 30. Oktober, Stift Zwettl, Niederosterreich, Teil 2. Wien. GABAIN, A. VON (1971), Friihe Zeugen der Scherengitter-Jurte, in: LIGETI (1971), 169-173. GAMKRELIDZE,T.V.-IVANOV, VYAC. VS. (1984),Indoevropejskijjazykiindoevropejcy / - / / [Indo-European language and the Indo-Europeans 1—11]. Tbilisi. GAMKRELIDZE, TH.V. - IVANOV, V.V. (1995), Indo-European a n d the IndoEuropeans. A reconstruction and historical analysis o f a proto-language a n d a proto-culture. Part I. Berlin. G a r ip o v , T.M. - KUZEEV, P.R. (1971), Baskirskie rodo-plem ennye nazvanija i ih altajskie paralleli [Bashkir tribal-clan nam es and their A ltaic parallels], in: SUNIK (1971). GELLNER, E. (1987), Nations and nationalism. Ithaca-N ew York. GENING, V.F. - H a LIKOV, A.H. (1964), Rannye bolgary na Volge [Early Bul ghars in the region o f the Volga], Moskva. GEREVICH, L. (Ed.) (1973), Budapest tortenete az oskortdl az Arpad-kor vegeig I [The history o f B udapest from prehistory to the end o f the Arpad era], Budapest. G ib b , H.A.R. - K r a m e r s , J.H. - L e v i -Pr o v e n q a l , E. - Sc h a c h t , j. (Eds) (1960-1993), The Encyclopaedia o f Islam. N ew edition, vols I—VII. L eiden-London. GlRTLER, R. (1982), “Ethnos”, “Volk” und soziale Gruppe. M itteilungen der Anthropologischen Gesellschaft, 112,42-57. GOBL, R. - R o n a -T a s , A. (1995), D ie Inschriften des Schatzes von N agySzentm iklos. Wien. GOCKENJAN, H.G. (1972), H ilfsvolker undG renzw achter im mittelalterlichen Ungarn. W iesbaden. GOCKENJAN, H.G. - SWEENEY, J.R. (1985), D er M ongolensturm. Berichte von Augenzeugen undZ eitgenossen (1235-1250). G raz-W ien-K oln. GOLB, N. - PRITSAK, O. (1982), Khazarian Hebrew docum ents o f the tenth century. Ithaca-London. GOLDEN, P.B. (1975), The people nwkrda. Archivum Eurasiae M edii A evi I, 21-35. GOLDEN, P.B. (1980), K hazar studies. A n historico-philological inquiry into the origins o f the K hazars 1-11. Budapest. GOLDEN, P.B. (1983), Khazaria and Judaism. Archivum Eurasiae M edii Aevi 3, 127-156.
FRIESING ER ,
Bibliography
459
G O LDEN , P.B. (1992), An introduction to the history o f the Turkic peoples.
Ethnogenesis an d state-form ation in medieval and early m odern Eurasia and the M iddle East. Wiesbaden. G O LDEN , P.B. (1993a), Rus, in: GlBB et al. (1960-1993), 618-629. G O LDEN , P.B. (1993b), Sakaliba, in: G lB B et al. (1960-1993), 871-878. GOMBOCZ, Z. (1912), D ie bulgarisch-tiirkischen Lehnw orter in der ungarischen Sprache. Helsinki. GOMBOCZ, Z. (1915), Arpadkori torok szem elyneveink [Hungarian personal names o f Turkic origin in the A rpad period], Budapest. GOMBOCZ, Z. (1917-1927), A m agyar oshaza 6 s a nem zeti hagyom any [The Hungarian Urheimat and the national traditions], N yelvtudom anyi Kozlemenyek 45 (1917-1920), 129-194, 46 (1923-1927), 1-33, 168-193 [Lec ture read on 8 th April 1918]. GOM BOCZ, Z. (1920), Bolgarok 6 s m agyarok [Bulghars and M agyars], Uj M agyar Szemle 2, 176-183. GOMBOCZ, Z. (1921), A bolgar kerdes es a m agyar hunm onda [The Bulghar question and the Hunnish legend o f the M agyars]. M agyar N yelv 17, 15-21. GOMBOCZ, Z. (1960), Honfoglalas elotti bolgar-torokjovevenyszavaink [Pre conquest Bolghar-Turkic loan words in the Hungarian language]. Published by Lajos Ligeti, Budapest. GOMBOS, A.F. (1937-1943), Catalogus fontium historiae H ungaricae I-IV .
Budapest. GOMORI, J. (1994), A 9-10. szazadi vaskohaszat [Iron forgery in the 9th—10th centuries], in: K O V A C S (1994), 259-269.
GREENFIELD, L. (1992), Nationalism. Five roads to modernity. Cambridge Ma - London. H. (1937), Le nom et l ’origine des hongroises. Zeitschrift der D eutschen M orgenlandischen G esellschaft 91, 630-642. GREGOIRE, H. (1938), L’habitat “p rim itif’ des M agyars et le Za(3apxoiaCTфоЛса. Byzantion 13, 267-278. GULYA, J. (Ed.) (1975), A vizim adarak nepe [The people of the water fowls],
GREGOIRE,
Budapest. GYO RFFY, G y . (1948), K ronikaink is a m agyar ostortenet [Hungarian chron
icles and Hungarian proto-history], Budapest. G y o r f f y , G y . (1955), Kurszan 6 s Kurszan vara, A m agyar fejedelem s 6 g kialakulasa 6 s Obuda honfoglalaskori szerepe [Kursan and K ursan’s castle. The emergence o f the Hungarian principality and the role o f Obuda in the Conquest period], Budapest Regisegei 15, 9-34. GYO RFFY, G y . (1959), Tanulmanyok a m agyar allam eredeterol [Studies on the origins o f Hungarian statehood], Budapest.
460
Appendices
GYORFFY, G y. (1963-1998), A z Arpad-kori M agyarorszag torteneti fo ld ra jza I - I V [The historical geography o f A rpadian H ungary I-IV ]. Budapest. GYORFFY, G y . (Ed.) (1965), N apkelet felfedezese. Julianus, Plano Carpini es R ubruk utijelentesei [The discovery o f the Orient. The travel reports o f Julianus, Plano Carpini and Rubruk], Budapest. GYORFFY, G y . (1971), A besenyok europai honfoglalasanak k 6 rdesehez [On the questions o f the European conquest o f the Pechenegs], Tortenelmi Szem le 1 4,281-187. GYORFFY, G y . (1973), Budapest tortenete az Arpad-korban [The history o f Budapest in the A rpad period], in: GEREVICH (1973), 217-349. GYORFFY, G y . (1975a), A uteur de l ’Etat des sem i-nom ades: le cas la Hongrie. Etudes historiques hongroises, I, 221-238. G y O rffy , G y. (1975b), The original landtaking o f the Hungarians. Budapest. G y Or f f y , G y . (Ed.) (1975c), A magyarok elddeirol es a honfoglalasrol [On the ancestors o f the M agyars and the Conquest], Second, enlarged edition. Budapest. GYORFFY, G y. (1977), Legenda ёБ valosag A rpad szen ^ly e koriil [Arpad: legend and reality]. Kortars 1, 103-113. G y O rffy , G y. (1980), Zur Frage der dem ographischen W ertung der papstlichen Zehntlisten. Studia H istorica A ca d Sc. Hung. 135. GYORFFY, G y . (1983), Wirtschaft und Gesellschaft der Ungarn um d ieJa h rtausendwende. M it einem Anhang: G esetze und Synodalbeschliisse Un gam s aus dem 11. Jahrhundert nach der Textausgabe von Levente Zavodszky. Budapest. GYORFFY, G y . (1984), Honfoglalas es megtelepedes [Conquest and settle ment], in: SZEKELY (1984), 577-650. GYORFFY, G y . (1985), Levedia es Etelkoz kerd 6 sehez [On the question o f Levedia and Etelkoz]. A M agyar N yelvtudom anyi Tarsasag Kiadvanyai, 172, Budapest, 3 -7 . GYORFFY, g y . (1988), Anonym us. Rejtely avagy tortenetiforras [M ystery or historical source?]. Budapest. GYORFFY, G y . (1990), A m agyarsag keleti elem ei [The Oriental elem ents o f the M agyars], Budapest. GYORFFY, G y . (1992), D iplom ata H ungarica antiquissim a 1000-1196. Bu dapest. GYORFFY, G y . (1993a), A 942. evi het vezer Ibn Hayyan feljegyzeseben [The seven chieftains o f the year 942 in Ibn H ayyan’s records], Budapest, 225-230. GYORFFY, G y. (1993b), K ronikaink es a m agyar ostdrtenet. R egi kerdesek— uj valaszok [Hungarian chronicles and Hungarian proto-history. N ew answers to old questions]. Budapest.
Bibliography
G Y O R FFY , G y .
461
( 1 9 9 3 c ) , Honfoglalo fejedelm eink es a vezerek. M itoszok, legalizalo tortenetek es egykoni feljegyzёsek [The C onquering M agyar princes and chieftains. M yths, legalising stories and contem porary records]. M agyar Tudomany 2 , 1 3 6 - 1 4 9 . G YO R FFY, G y . (1993d), V alasz [Kristo Gyulanak] [Reply (to Gyula Kristo)]. M agyar Tudomany 6 , 726-731. G YO R FFY, G y . (1995), A honfoglalok szama ёэ az A rpad-kor nёpessёge [The num bers o f the conquerors and the population o f the Arpad period], in: KO V A CSICS (1995b), 37—41. GYORFFY, GY. (1997), Arpad-kori oklevelek [Charters o f the Arpad era], B uda pest. GYORFFY, GY. -ZOLYOMI, B. (1994), A K arpat-m edence es Etelkoz kepe egy ё у е г ^ elott [The landscape o f the Carpathian Basin and the Etelkoz a m illennium ago], in: K o v A cs (1994), 13-37. H. TO TH , I. (1981), K onstantin-C irill is M etod elete i s m ukodese [The life and w ork o f C onstantine-Cyril and M ethodius]. Budapest. HAGN, TH. (1852), Urkundenbuch fu r die Geschichte des Benediktiner Stifles Kremsmiinster, seiner Pfarrereien undBesitzungen vom Jahre 777 bis 1400. Wien. HAJDU, P. (1962), F innugor nepek is nyelvek [Finno-Ugrian peoples and languages], Budapest. HAJDU, P. (1966), Bevezetes az urali nyelvtudom anyba (A m agyar nyelv fin n u g o r alapjai) [An introduction to Uralian linguistic studies (The FinnoUgrian bases o f the Hungarian language)]. Budapest. HAJDU, P. (1971), The Finno-U grian languages and peoples. London. HAJDU, P. (Ed.) (1976), Ancient cultures o f the Uralian peoples. Budapest. HAJDU, P. (1981), urali nyelviszet alapkerdesei [The rudim ents o f Uralian linguistic studies]. Budapest. HAJDU, P. (1994), Nicht-uralisches in den uralischen Sprachen. lncontri Linguistici. U niversita di Trieste, U niversita di Udine, 17, 59-79. HAJDU, P. -DOMOKOS P. (1978), Urali nyelvrokonaink [The U ralian linguistic relatives o f the M agyars]. Budapest. HAJDU, P. - DOMOKOS P. (1987), Die uralischen Sprachen u n d Literaturen. Budapest. HAJDU, P. - KRISTO, GY. - RONA-TAS, A. (Eds) (1976, 1977, 1980, 1982), Bevezetes a m agyar ostortenet kutatasanakforrasaiba 1/1, 1/2, II, III, IV [Introduction to the sources o f Hungarian proto-historical research], B uda pest. HAKIM ZJA N O V, F.S. (1978), Jazyk epitafij volzskih bulgar [The language o f the Volga Bulgharian inscriptions]. Moskva.
462
Appendices
H a k im z ja n o v , F.S. (1986a), Egy ujonnan felfedezett volgai bolgar sirfelirat [A recently discovered Volga Bulghar epitaph]. M agyar N yelv 82, 82. HAKIMZJANOV, F.S. (1986b), N ew Volga B ulgarian Inscriptions. A cta O rien talia Hung. 40, 173-177.
HALIKOVA, E.A. (1976), B ol’se-tiganskij m ogil’nik [The burial ground in Bolsie Tigani], Sovetskaja arheologija 2, 158-178.
HALIKOVA, E.A. (1978), Esce raz о probleme proishozdenija vengrov [Once more on the problem of the origins of the Hungarians], Sovetskaja arheolo gija 4 ,2 9 4 -2 9 8 .
HALIKOVA, E.A. - H a l ik OV, A.H. (1981), Altungarn an der K am a und im Ural. Budapest. HALIKOVA, E.A. - KAZAKOV, E.P. (1977), Le cimetiere de Tankeevka, in: ERDELYI, I. (Ed.), Les anciens hongrois et les ethnies voisines a Vest. Budapest, 21-223. HALIKOVA, J.A. (1973), Volgai Bulgaria es a X. szazadi M agyarorszag пёревsёge etnikai rokonsaganak кё^ёв ёЬ ег [On the ethnic kinship o f the popu lations o f Volga B ulgharia and H ungary in the 10th century], H ajdusagi M uzeum Evkdnyve. H ajduboszorn^ny. HAMILTON, J.R. (1955), Les Ouigurs a I ’epoque des Cinq Dynasties, d ’apres les docum ents chinois. Paris. HAMILTON, J.R. (1962), Toquz-Oguz et On Uygur. Journal Asiatique, 23-63. HARMATTA, J. (1962), Byzantinoturcica. A cta Antiqua Acad. Sc. Hung. 10, 131-150. HARMATTA, j. (1971), A z indoeuropai пёрек regi teleptitesteriiletei es vandorlasai [The ancient settlem ent areas and m igrations o f the Indo-European peoples], A M agyar Tudomanyos A kadem ia I. O sztalyanak K ozlem enyei 2 7 ,3 0 9 -3 2 4 . HARMATTA, J. (1977), Iraniak ёв fm nugorok, iraniak es m agyarok [Iranians and Finno-Ugrians, Iranians and M agyars], in: BARTHA - CZEGLEDY -
RONA-TAS (1977), 167-182. HARMATTA, J. (1979), Sogdian sources for the history o f pre-Islam ic Central Asia, in: Prolegom ena to the sources on the history o f pre-Islam ic Central Asia. Budapest, 153-165. HARMATTA, J. (1983a), Az avarok пуеЬ/ёпек к ё^ёв ёЬ ег [On the question o f the language o f the Avars]. A ntik Tanulmanyok 30, 71-84. HARMATTA, J. (1983b), Tiirk rovasirasos feliratok K elet-Europaban [Turk runiform inscriptions in Eastern Europe]. A ntik Tanulmanyok 30, 85-99. HARMATTA, J. (1983c), Egy avar rneltosagviselo pecsёtgyurйje [The signet ring o f an Avar peer], A ntik Tanulmanyok 30, 249-256. H a r m a tta , J. (1984a), A kom yei avar kori rovasirasos felirat [The Avar runiform inscription o f Kom ye], A ntik Tanulmanyok 31, 100-113.
Bibliography
463
HARMATTA, J. (1984b), Avar rovasirasos edenyfeliratok [Avar runiform in
scriptions on pots at Hatihabu]. Antik Tanulmanyok 3 \, 272-284. HARMATTA, J. (Ed.) (1984c), From Hecataeus to al-Hwarizmi. Bactrian,
Pahlavi, Sogdian, Sanskrit, Syriac, Arabic, Chinese, Greek a n d Latin sources f o r the history ofpre-Islam ic Central-Asia. Budapest. HARMATTA, J. (1985a), A Budapesti Torteneti M uzeum avar rovasirasos feliratai [The Avar runiform inscriptions o f the B udapest History M useum ], Antik Tanulmanyok 32, 46-56. HARMATTA, J. (1985b), Avar rovasirasos targyfeliratok [Avar runiform in scriptions on objects], A ntik Tanulmanyok 32, 57-72. HARMATTA, J. (1985c), Lebedia es Atelkuzu [Lebedia and Atelkuzu], A M a gyar Nyelvtudom anyi Tarsasag Kiadvanyai, 172, Budapest, 38-49. HARMATTA, J. (19 88), D e la question concernant la langue des avars. Inscrip tions runiques turques en Europe orientale. Ankara. HARMATTA, J. (1992a), Az onogur vandorlas [The O noghur migration], M agyar N yelv 8 8 , 257-272. HARMATTA, J. (1992b), Die Schrift bei den antiken Steppenvolkem . Vorlaufige M itteilung. A cta Classica Univ. Sc. D ebrecensis 28, 7-16. HARMATTA, J. (1992c), Rovasirasos feliratok avar szijvegeken. R ovasiras a Karpat-m edenceben [Runiform inscriptions on Avar belt-ends. Runiform inscriptions in the C arpathian Basin], In: SANDOR (1992b), 21-30. HARMATTA, J. (1993), A m agyar ostortenet kozep-irani forrasai [The M iddle Iranian sources o f Hungarian proto-history], M agyar N yelv 89, 336-337. HARMATTA, J. (1994), Nyugati ttirk uralom Kelet-Iranban [West Turk rule in East Iran], A ntik Tanulmanyok 36, 149-164. HARMATTA, J. (1995) [1996]. Sogdian inscriptions on Avar objects. A cta Orientalia Hung. 48, 61-65. HARMATTA, J. (1996a), K onstantinos Porphyrogennetos m agyar vonatkozasu muvei [The M agyar-related works o f Constantine Porphyrogenitus], in: K O V A C S - V ESZPREM Y (1996), 105-111. HARMATTA, J. (1996b) [1997], Turk and Avar runic inscriptions on metal belt-plates. A cta A ntiqua Acad. Sc. Hung. 31, 321-330. HARMATTA, J. (1998), A honfoglalas mai szemm el [A m odem outlook on the Conquest], M agyar N yelv 94, 129-151. HARMATTA, J. - M AROTH, M . (1985-1988), Z ur Geschichte der A rabisch Tiirkischen Beziehungen am Anfang des VIII. Jahrhunderts. A cta Antiqua Acad. Sc. Hung. 31, 441-446. HAUSSIG, W .H . (1953), Theophilakts Exkurs ttber die skythischen Volker. Byzantion 23, 275-462.
464
Appendices
HAZAI, G y . (Ed.) (1960), Sovietico-Turcica. Beitrage zur B ibliographie der tiirkischen Sprachw issenschaft in R ussischer Sprache in der Sowjetunion 1917-1957. Budapest. HEILIG, K.J. (1933), D er B rief des Rem igius von Auxerre um 900 liber Un gam , in: A g r o f K lebelsberg K uno Bee si M agyar Tortenetkutato Intezet Evkonyve 3, Budapest. HEISSIG, W. - KLIMKEIT, H.J. (1987), Synkretism us in den Religionen Zentralasiens. St. Augustin. HERENYI, I. (1982), A m agyar tбrzsszбvetsёg es torzsfoi [The M agyar tribal confederation and its chieftains]. Szazadok 116, 62-92. HODINKA, A. (1916), A z orosz evkonyvek m agyar vonatkozasai [Magyar-re
lated issues of the Russian annals], Budapest. HOLDER-EGGER, 0. (Ed.) (1894), Lam perti m onachi H ersenfeldensis opera, in: Scriptores Rerum Germ anicarum. H annover-Leipzig.
HOMAN, B. (1910), Zeitalter der orientalischen Quellen zur Urgeschichte der Ungarn. La revue Orientale, 28-40. H o m a n , B. (1917/1938), A m agyar пёр neve ёв a m agyar kiraly ci'me a
кбгёркоп latinsagban [The nam e o f the M agyar people and the title of the Hungarian king in the M iddle Ages], Torteneti Szemle, 6 , 129-158, 240-258, Reprint in: Tortenetiras es forraskritika. Budapest, 191-250. HOMAN, B. (1938), Szent Istvan [Saint Stephen], Budapest. HOPPAL, M. - ISTVANOVICS, M. (Eds) (1978), M itosz es tortenelem [Myth and history], Budapest. HRBEK, I. (1954), Novy arabsky pram en [New A rabic sources]. Ceskoslovenska Ethnographie 2:2, 157-175. HRBEK, I. (1955a), Arabico - Slavica. I. Abu Hamid al-A ndalusi und sein Werk Murib. A rchiv Orientalni 23, 109-135. HRBEK, I. (1955b), Ein arabischer B ericht uber U ngam (Abu H am id al-Andalusi al-G am ati 1080-1170). A cta Orientalia Hung. 5, 205-230.
HULTKRANTZ, A. (1993), Introductory remarks on the study of shamanism, Sham an 1,3-14. HYMES, D.H. (1960), Lexicostatistics so far. Current A nthropology 1, 3-44.
IVANYI, T. (1976a), M oham edan forrasok [The M oham m edan sources], in: HAJDU - KRISTO - RONA-TAS (1976-1982), 1/2, 211-240. IVANYI, T. (1976b), Szir forrasok [The Syrian sources], in: HAJDU - KRISTO -RONA-TAS (1976-1982), 1/2, 241-259. JANHUNEN, J. (1977), Sam oyed-Altaic contacts. Present state o f research.
Helsinki, 123-129. JANHUNEN, J. (1991), Ethnic death and survival in the Soviet North. Journal
de la Societe Finno-O ugrienne 83, 111-121.
Bibliography
465
JEN KIN S, R .J.H . (Ed.) (1962), Constantine Porphyrogenitus. De Adm inis-
trando Im perio vol. II. Commentary. London. JEREMIAS, Ё. - M A R O T H , М. (1977), Drei m itteliranische W orter fiir “Sonne” . A cta Antiqua 25, 221-226. JO HANSEN, U. (1987), Zur Geschichte des Schamanism us, in: H E ISSIG , W. KLIM KEIT, H-J. (Eds), Synkretism us in den Religionen Zentralasiens. W iesbaden, 8-22. JOKI, A.J. (1963), D er w andem de Apfel. Studia Orientalia 28, 1-17 (in Hun garian, in: GULYA (1975), 217-235. JOKI, A.J. (1973), Uralier und Indogermanen. D ie alteren Beriihrungen zwischen den uralischen und indogermanischen Sprachen. Helsinki. JUDAHIN, K.K. (1965), K irgizsko-russkij slovar’. [K irghiz-R ussian dictio nary], Moskva. JUHASZ, I. (1983), Ein aw arenzeitlicher N adelbehalter m it K erbschrift aus Szarvas. A cta Archaeologica Hung. 35, 373-377. JuhAsz, I. (1985), A szarvasi avar kori rovasirasos tutarto [The Avar-period needle case o f Szarvas with runiform writing]. M agyar Tudomany 30, 92-95.
Juh Asz , I. (1992), Ujabb rovasirasos emlek Szarvasrol [A new runiform inscription from Szarvas], in: SANDOR (1992b), 15-21. G.V. (1960), Vvedenie v bulgaro-tatarskuju epigrafiku [Introduc tion to the B ulgharo-Tatar epigraphy], M oskva - Leningrad. K A L D Y -N A G Y , G Y . (Ed.) (1976), Hungaro-Turcica. Studies in honour o f Julius Nemeth. Budapest. Ka r a , G y . (1995), Torok uri, magyar ur [Turkic uri', Magyar ur], M agyar N yelv 91, 65-68. K A R A CSO N YI, B. - SZEG FU , L . (1976), N yugati latin forrasok [The Western La tin sources], in: H A JD U - K R IS T O - R O N A -T A S (1976-1982), 1/2, 169— 186. KARLGREN, B. (1957), Grammata Serica recensa. Stockholm. K A ZA K O V , E.R - STARO STIN, P.N. - H ALIK O V, A.H. (1987), Arheologiceskie pam jatniki Tatarskoj ASSR [Archaeological relics in the Tatar A utonom ous Soviet Socialist Republic], K azan’. K A Z H D A N , A.P. - TA L B O T , А.М. (Eds) (1991). The O xford D ictionary o f Byzantium I—III. New York-Oxford. KELLY, J.N.D. (1988a), Reclams Lexicon der Papste. Stuttgart. KELLY, J.N.D. (1988b), The Oxford D ictionary o f Popes. Oxford. KESZI, T. (1995), Az ugynevezett m agyar szakralis kettos kiralysagrol [On the so-called M agyar sacral dual kingship]. Som ogyi M uzeum ok K ozlem enyei 11, 189-193.
JUSUPOV,
466
Appendices
KlRALY, P. (1974), A magyarok em litese a K onstantin- es a M etod-legendaban [M ention o f the M agyars in the Constantine and M ethodius legend], Budapest. KIRALY, P. (1985), Levedia-Etelkoz a szlavisztikai irodalom tiikreben [Levedia-Etelkoz in the literature o f Slavistic studies]. A M agyar N yelvtudom anyi Tarsasag Kiadvanyai, B udapest 172,49-87. KlRALY, P. (1987), A V III-IX . szazadiU ngarus, Hungaer, Hunger, Hungarius, Onger, W anger szem elynevek [The 8th-9th-century personal nam es Ungarus, Hungaer, Hunger, Hungarius, Onger, and Wanger], M agyar N yelv 83, 1 6 2 -1 8 0 ,3 14-331. K l j a S t o r n y j , S.G. - L lV § IC , V.A. (1972), The Sogdian inscription o f Bugut revised. A cta Orientalia Hung. 26, 69-102. KM IETO W ICZ, A. - KM IETO W ICZ, V.A. - L E W IC Z K Y , Т. (1985), Z rodla arabskie do dziejow slowianszczyzny III [Arab sources to the history o f the Slavs III]. Wroclaw. KMOSKO, M. (1997), M ohamedan irok a steppe nepeirdl. Foldrajzi irodalom 1/1 [M ohamm edan writers on the peoples o f the steppe. G eographical literature 1/1]. Budapest. KODALY, Z. (1937), A m agyar nepzene [Hungarian folk m usic], Budapest. KODALY, Z. (1947), Otfoku zene IV. 140 csuvas dallam [Pentatonic m usic IV. 140 C huvash m elodies]. Budapest. KOHN, S. (1881), H eber kutforrasok es adatok M agyarorszag tortenetehez [Prim aiy H ebrew sources and other data on the history o f Hungary], Buda pest. K O K O V C O V , P.K. (1932), Jevrejsko-hazarskaja perepiska v X v e k e [The Khazarian-H ebrew correspondence in the 10th century], Leningrad. K O M O R 6 C Z Y , G . (1992), Bezarkozas a nem zeti hagyom anyba [C o n fin ed by the national traditions], Budapest,
KORDE, Z. (1994), Anonym us, in: KRISTO (1994c), 50-51. K O R E N C H Y , E. (1972), Iranische Lehnw drter in den obugrischen Sprachen. Budapest. K O R E N C H Y , Ё. (1976), Irani nyelvtortenet [Iranian language history], in: H a j d u - K r i s t o - R o n a - T a s (1976-1982), 1/2, 85-103. KORMUSlN, I.V. (1975), К osnovnym ponjatjam tjurkskoj runiceskoj paleografii [On the basic idea o f Turkic runiform script]. Sovetskaja Tjurkologija 2, 25-47. KOSA, F. (1992), G ondolatok а згёке1у rovasi’ras kutatasanak lehetбsёgeirбl es m odjairol [Some thoughts on the possibilities and m ethods o f researching Szekely runiform script], in: SAn d o r (1992b), 69-77. K o s a , L. (1976) Gazdalkodas [Economy], in: H a j d u - K r i s t o - R o n a - T a s (1976-1982), 1/1,248-253.
Bibliography
467
K o sA ry , D. (1951), Bevezetes a m agyar tortinelem forrasaiba es irodalmaba I [Introduction to the sources and literature o f Hungarian history], Budapest. KOSZTA, L. (Ed.) (1995), Kelet i s N yugat kozott. Tortinelm i tanulm anyok Kristo Gyula tiszteletere [Between East and West. Historical studies in honour o f Gyula Kristo]. Szeged. KovAcs, L. (Ed.) (1994), H onfoglalas i s rig eszet [The Conquest and archae ology], Budapest.. KOVACS, L. - VESZPREMY, L. (Ed.) (1996), A honfoglalaskor irott fo rra sa i [The written sources o f the Conquest period]. Budapest. KOVACSICS, J. (1995a), A torteneti dem ografia valaszai es nyitott kerdesei az Arpad-kori nepesseg szam ara vonatkozoan [Answers and open questions o f historical dem ography regarding the population o f the A rpad period], in: KOVACSICS (1995b), 8-36. KOVACSICS, J. (Ed.) (1995b), M agyarorszag tortineti dem ografiaja I. A honfo glalas es az Arpad-kor n ip e ssig e [The historical dem ography o f H un gary I. The population o f the Conquest and Arpad periods]. Talks at the Conference o f Demographic History in B udapest on 6 February 1995, Budapest. KOVALEVSKIJ, A.P. (1956), Kniga Ahm eda ibn Fadlana о ego putesestvii na Volgu v 921-922 gg. [Ahm ed ibn F adlan’s book on his journey on the river Volga in the years 921-922]. H a r’kov. KRADER, L. (1963), Social organization o f the M ongol-Turkic p astoral no mads. The Hague. KRISTO, GY. (1976), M agyarorszagi latin forrasok [Latin sources in Hungary], in: H ajd u - K risto - R o n a -T as (1976-1982), 1/2, 186-194. KRISTO, G y. (1980), Levedi torzsszovetsegetol Szent Istvan allam aig [From the tribal confederation o f Levedi to the state o f Saint Stephen], Budapest. KRISTO, Gy. (1983a), N yelv 6 s etnikum. A “kettos honfoglalas”elm eleti alapjaihoz [Language and ethnicity. On the theoretical rudim ents o f the “double Conquest”]. Szegedi B olcsiszm iihely ’82. Szeged, 177-190. KRISTO, G y. (1983b), Tanulmanyok az Arpad-korrol [Studies on the Arpad period], Budapest. KRISTO, GY. ( 1993a), H onfoglald fejed elm ek; A rp a d es Kurszan [Conquering princes: Arpad and Kursan], Szeged. KRISTO, GY. (1993b), A honfoglalas evfordulojara kesziilve [Approaching the anniversary o f the Conquest], M agyar Tudomany 6 , 723-726. KRISTO, G y. (1994a), A K arpat-m edence i s a m agyarsag rigm ultja (1301 -ig) [The Carpathian Basin and the ancient past o f the M agyars (until 1301)]. Second edition, Szeged.
468
Appendices
KRISTO, G y. (1994b), K ronikakom pozicio [C hronicle com position], in: KRISTO (1994c), 381-382. KRISTO, G y . (Ed.) (1994c),
Korai magyar torteneti lexikon (9-14. szazad)
[Lexicon o f ancient Hungarian history (9th-14th centuries)]. Budapest. KRISTO, G y . ( 1995a), M agyarorszag lelekszam a az A rpad-korban [H ungary’s population figures in the A rpad period], in: KOVACSICS (1995b), 42-95. KRISTO, g y . (1995b), A honfoglalo m agyarsag eletm odjarol (Irott forrasok alapjan) [On the way o f life o f the conquering M agyars (B ased on the w ritten sources)]. Szazadok 129, 3-54. KRISTO, G y. (1995c), A magyar allam megsziiletese [The birth o f the H un garian state]. Szeged. KRISTO, G y . (Ed.) (1995d), A honfoglalas koranak irott forrasai [The written sources o f the Conquest period], Szeged. KRISTO, g y . (1996), Hungarian history in the 9th century. Szeged. KRISTO, G y . - M AKK, F. (1995), Az A rpad-haz uralkodoi [Rulers o f the A rpad dynasty], n.p. KRISTO, G y. - M AKK, F. - SZEGFU, L. (1973), A datok “korai” helyneveink ism eret 6 hez I. [Data on the “early” Hungarian place nam es I]. A cta Universitatis Szegediensis de A ttila Jozsef nom inatae, Acta Historica, 44. Szeged. KUZEEV, R.G. (Ed.) (1960), Baskirskie sejere [The genealogical table o f the Bashkirs], Ufa. KUZEEV, R.G. (1974), Proishozdenie baskirskogo naroda. Etniceskij sostav. Istorija rasselenija [The origin o f the Bashkir people. The ethnical com po nents. The history o f the settlement], Ufa. KYZLASOV, I.L . (1990), Drevnetjurkskaja runiceskajapis'm ennost’ Evrazii (Opyt paleograficeskogo analiza ) [Old Turkic runiform script o f Eurasia (The results o f the palaeographic analysis)]. M oskva. KYZLASOV, I.L. (1994a), Runiceskie p i s ’mennosti evrazijskih stepej [Runic records o f the Eurasian steppes], Moskva. KYZLASOV, I.L. (1994b), Drevnjaja p i s ’m ennost’ sajano-altajskih tjurkov. Rasskazy arheologa [The old w ritten records o f the Sayan-Altai Turks. The stories o f the archaeologist]. M oskva. KYZLASOV, L.R. (1965), О datirovke pam jatnikov jenisejskoj p is’mennosti [About the dating o f the w ritten Yenisei records], Sovetskaja Arheologija 3, 38-49. LABOV, W. (1994), Principles o f linguistic change. Internal factors. Oxford-C am bridge. LAKO, G y. - REDEI, K. (Eds) (1967-1978), A magyar szokeszlet finnugor elemei / - / / / [The Finno-Ugrian elem ents o f the Hungarian w ord stock I-III], Budapest.
Bibliography
469
L A szlo , G y. (1970a), A “kettos honfoglalasrol” [On the “double C onquest”]. Archaeologiai Ertesito 97, 161-190. LASZLO, G y. (1970b), K erdesek es feltevesek a m agyar honfoglalasrol [Q ues tions and claim s about the M agyar Conquest], Valosag 13, 1, 48-61. LASZLO, G y. (1971), Ostdrtenetunk legkorabbi szakaszai. A fm nugor ostor tenet regeszeti emlekei a Szovjetfoldon [The earliest phases o f our prehis tory. A rchaeological finds from Finno-Ugrian prehistory in the Soviet Union], Second edition. Budapest. LASZLO, G y. (1973), A honfoglalokrol [On the Conquerors], Budapest. LASZLO, G y. (1975), Die ungarische Landnahm e und ihre Vorereignisse (Uber die Forschungen zur Landnahm e der Ungam ), in: ORTUTAY, Gy. (Ed.), Congressus Quartus Internationalis Fenno-Ugristarum. Budapest, 195- 208. LASZLO, G y. (1 978),/4 "kettos honfoglalas” [The ‘double C onquest’]. B uda pest. LASZL6 , G y. (1982), A kettos honfoglalasrol [On the double Conquest]. His toric! 1 ,3 -4 . LASZLO, G y. (1987), Ostdrtenetunk [Our prehistory]. Third edition. Budapest. LASZLO, G y. (1990), Oseinkrol [On our ancestors], Budapest. LASZLO, Gy . (1995), A kettos honfoglalasrol [On the double C onquest], in: KOVACSICS (1995b), 122-126. LASZLO, G y. (1996), The Magyars. Their life and civilisation. Budapest. LASZLO, G y. - RACZ, I. (1 9 8 3 a ), A nagyszentmiklosi kincs [The N agyszent miklos treasure]. Third edition, Budapest. LASZLO, G y. - RACZ, I. (1983b), Der Goldschatz von Nagyszentmiklos. B u dapest. LASZLO, Gy . - RAcz, I. (1984), The Treasure o f Nagyszentmiklos. Budapest. LAUTERBACH, H. (1967), Untersuchungen zur Vorgeschichte der Protobulgaren, nach einem B ericht bei Theophanes, in: ALTHEIM - STIEL (1967), IV, 539-619. LENGYEL, I. (1975), Palaeoserology. Blood typing with the fluorescent anti body method. Budapest. L e w i c k i , T. (1938), W ^gry i m uzulm anie w^gerscy w swiete relacji podroznika arabskiego z XII w. Abu H am id al-A ndalusi al-G am atiego [Hungary and the Hungarian M uslim s in the light o f the records by the 12th-century Arab traveller Abu Hamid al-Andalusi al-G am ati], Rocznik Orientalistyczny 13, 106-122. LEWICZKI, T. (1949), Swiat slowianski w oczach pisarzy arabskich [The world o f the Slavs as seen by Arab authors]. Slavia Antiqua 2, 321-388. LEW ICK I, T. (1956), Zrodla arabskie do dziejow slowianszczyny [Arab sources to the history o f the Slavs]. W roclaw - Krakow.
470
Appendices
LEWICZKI, T. (1961), Znajom osc krajow i ludow Europy u pisarzy arabskich
IX i X w [European countries and people known by Arab authors in the 9th and 10th centuries]. Slavia A ntiqua 8 , 61-124.
LEWICZKI, T. (1962), Un temoignage arabe inconnu sur les slaves— de l ’an 720. Folia Orientalia 4, 319—331 . LEWICZKI, T. (1978), Les noms des hongroises et de la Hongrie chez les medievaux geographes arabes etpersans. Folia Orientalia 19, 35-55. LIGETI, L. (1935), Kina. M ult es je le n [China: past and present], Budapest. LIGETI, L. (1962), A m ongolok titkos tortenete [The secret history o f the M on gols], Budapest. LIGETI, L. (1964a), A m agyar nep m ongol kori nevei [Designations o f the M agyar people in the M ongolian period], M agyar N yelv 60, 385-404. LIGETI, L. (1964b), Gyarm at es Jeno [Gyarm at and Jeno], Nyelvtudom anyi E rtekezesek 40, 230-239. LIGETI, L. (Ed.) (1971), Studia Turcica. Budapest. LIGETI, L. (1976), A m agyar nyelv torok kapcsolatai es ami kortilottuk van [The Turkic connections o f the Hungarian language and related issues], M agyar N yelv 12, 11-27, 129-136. LlGETI, L. (1978), R 6 gi torok eredetu neveink I [Old Hungarian nam es of Turkic origin I]. M agyar N yelv 74, 257-274. LIGETI, L. (1979), Regi torok eredetu neveink II—IV [Old Hungarian names o f Turkic origin II—IV]. M agyar N yelv 75, 26-42, 129-141, 259-273. LIGETI, L. (1985), Levedia es Etelkoz [Levedia and Etelkoz], A M agyar N yelvtudom anyi Tarsasag Kiadvanyai, 172. Budapest, 57-76. LIGETI, L. (1986), A m agyar nyelv torok kapcsolatai a honfoglalas elott es az Arpdd-korban [Turkic connections o f the Hungarian language prior to the Conquest and in the Arpad era], Budapest.
L ih a Ce v , D.S. (Ed.) (1950), Povest’ vremennyh let I [Nestor chronicle]. Moskva-Leningrad. LIPTAK, P. (1971), Em bertan es em berszarmazastan [Anthropology and hu
m an genealogy]. Second edition, Budapest. LIPTAK, P. (1976), Antropologiai forrasok [The anthropological sources], in: H a j d u - K r i s t o - R o n a -T a s (1976-1982), I / l , 203-220 LIPTAK, P. (1977), A m agyar ostdrtenet kerdesei az antropologiai kutcitasok
alapjan [T h e q u e s tio n s o f H u n g a ria n p r e h isto r y in a n th r o p o lo g ic a l re se a r c h ], in: B a r t h a - C z e g l e d y - R o n a -T a s (1977), 231-242. LIPTAK, P. (1983), Avars and ancient Hungarians. Budapest. LIU MAU-TSAI (1958), Die chinesischen N achrichten zu r Geschichte der Ost-Turken (T'u-kue) / - / / . W iesbaden. LOSEK, F. (1997), Die Conversio Bagoariorum et C arantanorum und der B r ie f des Erzbischofs Theotmar von Salzburg. Hannover.
Bibliogi'aphy
471
LOTZ, J. (1969), A m agyar szo etim ologiajanak kerdesehez [On the etym o
logical issues o f the word M agyar]. N yelvtudom anyi K ozlem enyek 65, 385-389. LUDWIG, D. (1982), Osteuropaische Geschichte. Struktur und G esellschaft des Chazaren-Reiches im Licht der schriftlichen Quellen. Munster. LUDWIG, D. (1983), Chazaren, in: Lexikon des M ittelalters, Bd. 2. M iinchenZiirich, 1783-1788. LUKINICH, I. (1905), M enander Protektor tdrteneti m uvenekfennm aradt toredekei [Surviving fragm ents o f the historical work o f M enander Protector], Brasso. LYTKIN, V.I. - GULJAEV, JE.S. (1970), K ratkij etim ologiceskij s lo v a r ’ kom i ja zyka [Short etym ological dictionary o f Zyryan language], Moskva. MACARTNEY, C.A. (1930/1964), The M agyars in the N inth Century. C am bridge. MACKERRAS, C. (1972), The Uighur Empire according to the T ’ang Dynastic histories. A study in Sino-Uighur relations 7 4 4 -8 4 0 ,2 n d edition. Canberra. MADARAS, L. (1975), A kettos honfoglalas elm eletenek nehany telepiilestorteneti problem aja [Some settlem ent historical problem s o f the double Conquest theory], A cta Juvenum 8 . Sectio Historica, Szeged, 32-54. MAENCHEN-HELFEN, O.J. (1978), D ie Welt der Hunnen, D eutschsprachige A usgabe besorgt von R. Gobi. W ien-K oln-G raz. MAGOMEDOV, M.G. (1983), Obrazovanie Hazarskogo kaganata [The form a tion o f the K hazar Khaghanate], M oskva. MAKK, F. (1985), Kiknek az elen allott Levente? (Egy kosztantinoszi hely ertelmezesehez) [W hom did Levente lead? (On the interpretation o f a Constantine passage)]. Acta Universitatis Szegediensis de Attila Jozsef nominatae. A cta H istorica 82, 3-9. MAKK, F. (1995), Kiilfoldi forrasok es a m agyar tortenelem (X -X II. szazad) [Foreign sources and Hungarian history (10th—12th centuries)]. Acta U ni versitatis Szegediensis de Attila Jozsef nom inatae. A cta H istorica 102, 25-41. MAKK, F. (1997), Egy ostorteneti kezikonyv m argojara [Notes on a handbook o f prehistory], Aetas, 2 -3 , 161-188. MAKK, F. (1998), Gondolatok a m egjegyzesekhez. [Some thoughts on the remarks. R ejoinder to Rona-Tas 1998e], Aetas, 2 -3 , 227-237. M a l ja v k in , A.G. (1983), Ujgurskie gosudarstva v IX-XII vv [The Uighur states in the 9th—12th centuries], Novosibirsk. MALLORY, J.P. (1989), In search o f the Indo-Europeans. Language, A rcheol ogy and Myth. London. MANDOKY-KONGUR, I. (1976), M agyar eredetu torzsek a baskiroknal [Tribes o f M agyar origin am ong the Bashkirs], Tiszataj 30, 10, 4 1 ^ 4 .
472 M ANDOKY-KONGUR, I. (1986), Jeno es Yanay [Jeno and Yanay].
Appendices
Keletkutatas
S p rin g ,70-74. M ANDOKY-KONGUR, I. (1988), The question o f identifying the Hungarian
tribal nam e Jeno w ith the Bashkir ethnonym Yanay. Acta Orientalia Hung. 4 2 ,4 3 ^ 8 . MAROTI, E. - LAKATOS, P. (1976), Okori latin forrasok [A ncient Latin sources], in: HAJDU - KRIST6 - R o n a -T a s (1976-1982), 1/2, 160-169. MARQUART, J. (1903), Osteuropaische und ostasiatische Streifziige. Leipzig. MARTINEZ, A.P. (1982), G ardlzl’s two chapters on the Turks. Archivum Eu rasiae Medii Aevii 2, 109-217. MATSUMOTO, H. (1988), Characteristics o f M ongoloid and neighbouring populations based on the genetic m arkers o f hum an im m unoglobulins. Human Genetics 8 0 ,2 0 7 -2 1 8 . MELICH, J. (1925-1929), A honfoglalaskori Magyarorszag [H ungary in the C onquest period]. Budapest.
K.H. ( 1951), The Oriental elements in the vocabulary o f the oldest Russian epos, the Igor Tale "Slovo о Polku Igoreve ”. Supplem ent to Word.
M ENGES,
New York. MENGES, K.H. (1968), Tungusen undLjao. Wiesbaden. M ENGES, K.H. (1979), Vostocnye elementy v ‘Slove о polku Igoreve’ [Eastern elem ents in the Igor Tale]. Leningrad. MESTERHAZY, K. (1994), Tobbgyokeru osi vallasunk em lekei [Relics o f the heterogeneous ancient religion o f the M agyars], in: KovAcs (1994), 195-205. MESZAROS, E. (Ed.) (1996), Unnepi konyv Mikola Tibor tiszteletere [Festive book in honour o f Tibor M ikola], Szeged. MINORSKY, V. (1937), Hudud al-alam, the regions o f the world. London. MIQUEL, A. (1973), La geographie humaine du monde musulman ju s q u ’au
milieu du l i e siecle. Vol. 1: Geographie et geographie humaine dans la literature arabe des origines a 1050. Paris-L a-H aye. M OCSY, A. (1984), A rom ai kor [The Rom an age], in: SZEKELY (1984),
195-264.
Idojaras-varazslas Belso-Azsiaban [W eather m agic in Inner-Asia]. Budapest. M OLNAR, A. (1994), Weather-magic in Inner-Asia. Bloom ington. MOOR, G y. (1953), A m agyar nepnev eredetenek kerdesehez [On the question o f the origins o f the ethnonym Magyar]. Nyelvtudomanyi Kozlemenyelc 54, 75-95. ' MORAVCSIK, GY. (1926-1932/1967), B yzantinische H um anisten iiber den Volksnam en tiirk. Korosi Csoma Archivum 2, 3 8 1 -3 8 4 . R eprint, in: MORAVCSIK (1967), 383-385. M OLNAR, A. (1993),
Bibliography
473
MORAVCSIK, G y. (1 929-1930/1967), Die archaisierenden N am en der Ungarn in Byzanz. Byzantinische Zeitschrift 30, 247-253. Reprint, in: MORAVCSIK (1967), 320-325.
MORAVCSIK, GY. (1930a), Zur Geschichte der Onoguren. Ungarische Jahrbticher 10, 53-90. MORAVCSIK, GY. (1930b), Az onogurok tortenetehez [On the history o f the Onoghurs], Budapest. MORAVCSIK, G y. (1931), Arpad 894. evi vez 6 rtarsanak neve [The nam e o f A rpad’s co-chieftain in the year 894]. M agyar N yelv 27, 84—89. MORAVCSIK, GY. (1958/1983), Byzantinoturcica / - / / , zw eite neubearbeitete Ausgabe, Berlin 1958, N eudruck 1983. MORAVCSIK, GY. (1967), Studia Byzantina. Budapest. MORAVCSIK, GY. (1976), Einfuhrung in die Byzantologie. Budapest. MORAVCSIK, G y. (1983), Byzantinoturcica 1-11, 3. unveranderte Auflage. Berlin. MORAVCSIK, GY. (1984), A z Arpad-kori m agyar tortenet bizanci fo rrd sa i [The Byzantine sources related to Arpadian Hungarian history], Budapest. MORAVCSIK, G Y .- JENKINS, R.J.H. (1967), Constantine Porphyrogenitus. De Adm inistrando Imperio. Greek text edited by Gy. M oravcsik, English translation by R.J.H. Jenkins. New, revised edition. Dum barton Oaks. MORIYASU, T. (1980), La nouvelle interpretation des m ots Hor et Ho-yo-hor dans le m anuscrit Pelliot tibetain 1283. A cta O rientaliaH ung. 34,1 7 1 -1 8 4 . MUNKACSI, B. (1901), Arja es kaukazusi elem ek a fm n -m a g ya r nyelvekben [Aryan and Caucasian elements in the Finno-Hungarian languages], B uda pest. N EM ETH , GY. (1927), A tiirk nspnev [The ethnonym Turk], M agyar N yelv 23, 271-274. NEMETH, GY. (1930), A honfoglald m agyar sag kialakulasa [The form ation o f the C onquering M agyars], Budapest. NEMETH, G y. (1934), A tбrбksёg oskora [Prehistory o f the Turkic peoples], in: B erzeviczy Emlekkonyv, Budapest, 158-174. NfiMETH, [GY.] J. (1938), Das Volk m it den scheckigen Pferden. K orosi Csoma Archivum I. Erg. Bd. 345-352. NEMETH, GY. (1940a), A m agyar keresztenyseg kezdete [The beginnings o f Christianity in Hungary], Budapesti Szem le 256, 14-30. NEMETH, [Gy.] J. (1940b), La question de l ’origine des Sicules. Archivum Europae Centro-Orientalis VI, 208-241. N EM ETH , [Gy.] J. (1941), Die Volksnamen quman und qun. K orosi Csoma Archivum III, 95-109. NEMETH, [G y.] J. (1942-1947), Problem e der tiirkischen Urzeit. Analecta Orientalia M em oriae A lexander Csoma de Koros Dicata, 57-102.
474 NEMETH, [G y.] J. (1949),
Appendices
Le systeme des noms des peuples turcs. Actes du
XXIe Congres International des Orientalistes. Paris, 174. (1959), Eine Worterliste der Jassen, der ungarischen Alanen. Berlin. NEMETH, [G y.] J. (1964), La provenance du nom Bulgar (Pohodzenija nazvi «Bolgari»). Onomastica No. 28, 12 p. NEMETH, [Gy.] J. (1965), Kereit, Kerey, Giray. Ural-Altaische Jahrbiicher 3 6 ,3 6 0 -3 6 5 . NEMETH, G y. (1966a), M agyar torzsnevek a baski'roknal [M agyar ethnonym s am ong the Bashkirs]. Nyelvtudomanyi Kozlemenyek 6 8 , 35-50. NEMETH, G Y . (1966b), A baskir fo ld i m agyar oshazarol [On the M agyar Urheimat in Bashkiria], Elet es Tudomany 31, 596-599. NEMETH, [G y.] J. (1966 c), Ungarische Stam m esnam en bei den Baschkiren. Acta Linguistica 16, l - 2 1. NEMETH, [G y.] j. (1969), Der Volksname K arluk und seine sem antische Gruppe. Acta Linguistica 19, 13-18. NEMETH, [G y.] J. (1971a), M agyar und Miser. Acta Orientalia Hung. 25, 293-299. NEMETH, [G y.] J. (1971b), N o m s e th n iq u e s tu rcs d ’o r ig in e to te m is tiq u e , in: LIGETI (1971), 349-359. NEMETH, G y. (1972), Gombocz Zoltan [Zoltan Gom bocz], Budapest. NEMETH, [Gy.] J. (1975), Tiirkische und ungarische Ethnonym e. Ural-Altai sche Jahrbiicher 47, 154—160. NEMETH, [G y.] J. (1982), The m eaning o f the ethnonym Bulgar, in: RONATAS, A. (Ed.), Studies in Chuvash Etymology I, 7-13 . NEMETH, G y. (1990), Tordkok es magyarok / - / / [Turks and M agyars 1-11]. Budapest. NEMETH, GY. (1991), A honfoglalo magyarsag kialakulasa [The form ation o f the Conquering M agyars], Second, enlarged and abridged edition. B u dapest. OLAJOS, T. (1969), A datek a /Н/ung/a/ri/i/ nepnev es a kesoi avar etnikum tortenetehez [Addenda to the ethnonym /Н/ung/a/ri/i/ and the history o f the late Avar ethnicity]. Antik Tanulmanyok 16, 87-89. OLAJOS, T. (1994), Source byzantine inobservee concem ant la protohistoire du peuple hongrois, in: SCHOORS, A. et VAN D EU N , P. (Eds), Philohistor Misc. in honorem Caroli Laga septuagenarii. Leuven, 441. OLAJOS, T. (1995), N ehany 6 szrevetel a “De adm inistrando im perio” m agyar vonatkozasu reszleteihez [Some rem arks on the M agyar-related details o f “De adm inistrando im perio”]. Magyar Nyelv 91, 44-52. NEMETH, [Gy.] J.
Bibliography
475
OLAJOS, T. (1996), A 9. szazadi avar tortenelem re vonatkozo gorog forrasok [Greek sources on the history o f the Avars in the 9th century], in: KOVACS - VESZPREMY (1996), 91-10.3. OSTROGORSKY, G. (1993), History o f the Byzantine state. Oxford. P ais, D. (1953), Nyek torzsneviink es ami koriilotte lehetett [Our tribe name Nyek and related issues]. Magyar Nyelv 49, 278-293. PAIS, D. (1975a), A magyar osvallas emlekeibol [On the relics o f ancient M a gyar religion], Budapest. PAIS, D. (Ed.) (1975b), Anonymus Gesta Hungarorum. Bela kiraly jegyzoje a magyarok cselekedeteirol [A nonym us’ G esta Hungarorum . King B ela’s scribe on the deeds o f the M agyars]. Budapest. PALOCZY HORVATH, A. (1989), Besenyok, kunok, jaszok [Pechenegs, Cumans, Yas], Budapest. PAULER, Gy. - S zilA g y i, S. (1900) A magyar honfoglalas kutfoi [The w ritten sources o f the M agyar Conquest], B udapest (Reprint 1995). PAULY, A.F. - WISSOWA, G. (1894-), Realencyclopadie der classischen Altertumswissenschaft. Neue Bearbeitung von G. W issowa, I-X X IV , Supple ment, I-. Stuttgart. PELLIOT, P. (1915), L’origine de T ’ou-kiue, nom chinois des Turcs. T'oung Pao, 687-689. PlGULEVSKAJA, N.V. (Ed.) (1957), Teofilakt Simokatta, Istorija [Theophylaktos Sim okattes, History], M oskva. PINKS, E. (1986), Die Uiguren von Kan-chou in der friihen Sung-Zeit (9 6 0 1028). W iesbaden. PLETNEVA, S.A. (1976), Hazary [The Khazars], Moskva. PLETNJOWA, S.A. (1978), Die Chasaren. Mitelalterliches Reich an Don und Volga. Leipzig. POHL, W. (1988), Die Awaren. Ein Steppenvolk in Mitteleuropa 567-822 n. Chr. Mtinchen. POPPE, N. (1954), Ein altes K ulturwort in den altaischen Sprachen. Studia Orientalia 19, 23-25. PRITSAK, O. (1953-1954), Die Karachaniden. Der Islam 31, 16-68. PRITSAK, O. (1975), The Pecenegs. Archivum Eurasiae Medii Aevii 1, 211— 235. PRITSAK, O. (1977), Die bulgarische Fiirstenliste und die Sprache der Protobulgaren. W iesbaden. PRITSAK, O. (1978), The K hazar K ingdom ’s conversion to Judaism . Harvard Ukrainian Studies 2, 280-281. PULLEYBLANK, E.G. (1991), Lexicon o f reconstructedpronunciation in Early Middle Chinese, Late Middle Chinese and Mandarin. Vancouver.
Appendices
RACHEWILTZ, I. DE (1984, 1985), The secret history o f the M ongols. Papers on F ar Eastern H istory 30, 31, 81-160, 61-93.
RACHEWILTZ, I. DE (1994) The secret history o f the M ongols: Some funda m ental problems. Bulletin, The IA M S N ew s Inform ation on Mongolian Studies 1993, 2-1994, 1. Ulaanbaatar, 3-10. RAMSTEDT, G.J. (1947), The relation o f the Altaic languages to other language groups. Journal de la Societe Finno-O ugrienne 53, 15-26. RAMSTEDT, G.J. (1957), Einfuhrung in die altaische Sprachw issenschaft I. Lautlehre. Helsinki. Ra SONYI, L. (1961), Les noms de nombre dans I 'anthroponymic turque. Buda pest. RASONYI, L. (1964), Ba§kurt ve M acar yurtlarm daki ortak cografi adlar tizerine [On the com m on geographical nam es in B ashkiria and Hungary]. X. Tiirk D ili K urultayinda Okunan Bilim sel Bildiriler. Ankara, 105-112. R a SONYI, L. (1976a), The psychology and categories o f nam e giving among the Turkish peoples, in: KALDY-NAGY (1976), 207-224. Ra SONYI, L. (1976b), Azonos foldrajzi nevek a baskir es a m agyar foldon (Com m on geographical nam es in Bashkiria and Hungary). M agyar Nyelv 7 2 ,4 8 -5 3 . REDEI, K. (1964), Vannak-e az elom agyar-perm i erintkezesnek nyelvi nyomai? [Do pre-M agyar-Perm ian connections have linguistic traces?]. Nyelvtudomanyi Kozlemenyek 6 6 , 253-261. R e d e i , K. (1969), Gibt es sprachliche Spuren der vorungarisch-perm ischen Beziehungen? Acta Linguistica Hung. 19, 321-334. REDEI, K. (1979), Szofejtesek. 184. m agyar mese “M archen, Fabel” [Ety m ologies. No. 184. Hungarian mese “M archen, Fabel”]. N yelvtudom anyi K ozlem enyek 81, 353-354. REDEI, K. (1986), Zu den indogermanisch-uralischen Sprachkontakten. Wien. REDEI, K. (1988), Die altesten Indogerm anischen Lehnw orter der uralischen Sprachen, in: SlNOR (1988), 638-664. REDEI, K. (Ed.) (1988-1994), Uralisches etym ologisches W orterbuch /- //. Budapest. REINHART, J. (1990), Die tocharischen Entlehnungen im Altaischen und die Chronologie der tocharischen Lautgesetze, in: Tocharische Akten derF achtagung der Indogerm anischen Gesellschaft, Berlin, Septem ber 1990, 7 3 92. RONA-TAS, A. (1959), Som e data on the agriculture o f the M ongols, in: Opuscula Ethnologica M emoriae Ludivici Biro Sacra. Budapest, 443-472. RONA-TAS, A. (1961a), N om adok nyomaban. Etnografus szem m el M ongoliaban [On the tracks o f the nomads. An ethnographer in M ongolia], Bu dapest.
Bibliography
477
R O N A-TA S, A. (1961b), Notes on the Kazak Yurt o f West M ongolia. A cta O ri
entalia Hung. 1 2,79-102. R O N A-TA S, A. (1963), Felt-m aking in Mongolia. A cta O rientalia Hung. 16,
199-215. RONA-TAS, A. (1964), Po sledam kocevnikov. M ongolija glazam i etnografa. P erevod s vengerskogo [On the track o f nomads. M ongolia in the eyes o f the ethnographer. Translation from Hungarian], M oskva. R O N A-TA S, A. (1970), Az altaji nyelvrokonsag vizsgalatanak alapjai. (A nyelvrokonsag elmelete es a csuvas-m ongol nyelvviszony) [Rudim ents o f the study o f Altaic linguistic relationship. (A theory o f linguistic relation ship and Chuvash-M ongolian linguistic relations)]. D Sc dissertation, B u dapest. (M anuscript.) R o n a -T a s , A. (1971), A csuvas szovegi gutturalis zarhangok tortenetehez [On the history o f word-final guttural plosives in the C huvash language], N yelvtudom anyi Kozletnenyek 73, 198-207. RONA-TAS, a . (1972), The birth o f m odern anthropology in M ongolia. The M ongolia Society Bulletin 11:1,5-12. RONA-TAS, A. (1973), Did the proto-A ltaic people know the stirrup? Studia M ongolica 13, 169-171. RONA-TAS, A. (1974a), Tocharische Elem ente in den altaischen Sprachen? Sprache, Geschichte und K ultur der altaischen Volker. Protokollband der XII. Tagung der Perm anent International A ltaistic C onference 1969 in Berlin. Berlin, 499-504. RONA-TAS, A. (1974b), Obscee nasledie ili zaim stvovanija? (K problem e rodstva altajskih jazykov) [Com m on inheritance or borrow ing? (To the problem o f the relationship o f the Altaic languages)]. Voprosy Jazykoznanija 2, 31-45. RONA-TAS, A . (1977a), Beszamolo a Szegedi Ostorteneti M unkakozosseg m unkajarol [Report on the work o f the Szeged Proto-history Research Group], in: BARTHA - CZEGLEDY - R O N A-TA S (1977), 333-335. R ONA-TAS, A. (1977b), A m agyar-bolgar-torok erintkezes jelleg e [The n a ture o f M agyar-B ulghar-T urkic interaction], in: BARTHA - CZEGLEDY ~ R O N A-TA S (1977), 267-276. R ONA-TAS, A. (1978a), A nyelvrokonsdg. K alandozasok a torteneti nyelvtudomanyban [Linguistic relationship. Adventures in historical linguistics]. Budapest. R O NA-TAS, A. (1978b), Bevezetes a csuvas nyelv ismeretebe [Introduction to Chuvash language studies], Budapest. RONA-TAS, a . (1979), A m agyarsag ostdrtenete [The proto-history o f the M agyars]. Korosi Csoma Tarsasag (M anuscript). Budapest.
478
Appendices
RONA-TAS, A. (1980a), The place o f agriculture in the ethnography o f the M ongols. Etnografija Polska 2 4 ,4 5 -5 1 . RONA-TAS, A. (1980b), On the earliest Sam oyed-Turkish contacts, in: Congressus Quintus Internationalis Fenno-Ugristarum III. Turku, 3 7 7 - 385. RONA-TAS, A. (1980c), Review o f VIKAR-BERECZKY (1979). Nyelvtudoma nyi Kozlemenyek 80, 421-425. RONA-TAS, A. (1980d), Rem arks on Laszlo (1975), in: Congressus Quartus Internationalis Fenno-Ugristarum II, Budapest, 90-91. RONA-TAS, A. (1981), The character o f the H ungarian-B ulgaro-Turkic rela tions, in: KALDY-NAGY, G y . (Ed.), Studia Turkologica 5, 119-128. RONA-TAS, A. (1982a) A kazar nepnevrol [On the ethnonym Khazar], Nyelv tudomanyi Kozlemenyek 84, 349-380. RONA-TAS, A. (1982b), On the history o f the Turkic and Finno-U gric affri cates. Acta Orientalia Hung. 3 6 ,4 2 9 -4 4 7 . RONA-TAS, A. (1982c), Ostortenetunk 6s a baskir kerdes. A Jurm atiak tor tenete [Our proto-history and the Bashkir question. The history o f the Yurmati]. Tiszataj 8, 52-61. RONA-TAS, a . (1982d), The periodization and sources o f the C huvash lin guistic history, in: R o n a -T a s (1982e). RONA-TAS, A. (Ed.) (1982e), Chuvash Studies. B udapest-W iesbaden. RONA-TAS, A. (1983a), Ujabb adatok a kazar nepnev tortenet6hez [New data on the history o f the ethnonym Khazar]. Nyelvtudomanyi Kozlemenyek 85, 126-133. RONA-TAS, A. (1983b), “Rokonsagszerii osi kapcsolat” . N em eth G yula nezetei az urali es torok nyelvek viszonyarol [“A kindred-like ancient contact”. G yula N em eth’s views on the relationship o f Uralian and Turkic languages], in: BERECZKY - DOMOKOS (1983), 331-346. RONA-TAS, A. (1983c), De hypothesi Uralo-Altaica. Memoires de la Societe Finno-Ougrienne 185, 235-251. R o n a -T a s , A. (1984), A keleti m agyarok egy kairoi geniza-lev61ben [The East M agyars in a geniza-letter from Cairo], in: SCHEIBER, S. (Ed.), Evkdnyv 1983/84. Budapest, 293-296. R o n a -T a s , A. (1985a), A szarvasi tutarto felirata [The inscription o f the Szarvas needle case], Nyelvtudomanyi Kozlemenyek 87, 225-248. RONA-TAS, A. (1985b), Wiener Vorlesungen zur Sprach- und Kulturgeschichte Tibets. Wien. RONA-TAS, A. (1985c), Az eredetkerdes es a tarsadalm i tudat [The question o f origin and the social consciousness], Valosag 1985:8, 103-106 = RONATAS (1995a), 133-136.
479
Bibliography
RONA-TAS, a . (1985-1986), A m agyar rovasiras es a M atyas-kori hum anizmus [M agyar runiform script and hum anism in the era o f King M atthias]. N eprajz i s Nyelvtudomcmy 29-30, 173-179. R O N A-TA S, A. (1986), A m agyar nepnev egy 1311-es volgai bolgar sirfeliraton [The M agyar ethnonym on a Volga Bulghar epitaph o f 1311]. M agyar N yelv 82/1, 78-81.
A. (1987a), M a te r ia lie n -KLIM KEIT(1987), 33-45.
R 6N A -T A S ,
zu r a lten R e lig io n d er T iirk en , in:
HEISSIG
RONA-TAS, A. (1987b), On the developm ent and origin o f the East Turkic “runic” script. A cta Orientalia Hung. 41, 7-14. R O N A-TA S, A . (1987c), Istak, in: URAKSIN, Z .G . (E d .), Baskirskaja itnonimija [Bashkir ethnic names], U fa , 49-58. R O NA-TAS, A. (1988a), Ethnogenese und Staatsgriindung. Die tiirkische K om ponente in der Ethnogenese des Ungartum s, in: Studien zur E thno genese vol. 2. Opladen, 107-142. RO N A-TA S, A . (1988b), Problem s o f the East European scripts with special regard to the newly found inscription o f Szarvas, in: Popoli delle Steppe: Unni, Avari, Ungari, Spoleto 2 3-29 aprile 1987. Spoleto, 483-511. RONA-TAS, A. (1988c), Altaic and Indo-European. M arginal rem arks on the book o f Gam krelidze and Ivanov. A cta Orientalia Hung. 42, 391-404. R O N A-TA S, A. (1988d), T u rk ic in flu e n c e o n th e U r a lic la n g u a g e s , in: SlNOR (1988), 742-780. RONA-TAS, A. (1988e), Opponensi velem eny Fodor Istvan: A honfoglalo m agyarsag keleti gyokerei rim u kandidatusi ertekezeserol [O pponent’s view on Istvan F o d o r’s academ ic thesis “The Oriental roots o f the conquer ing M agyars”], in: A Nyiregyhazi Josa A ndrds M uzeum Evkonyve 21-23 (1978-1980). N yiregyhaza, 162-128. [Appeared in 1988.] R o n a -T a s , A. (1989), Nep es nemzet. Vitak fogalm akrol Europaban. Szucs Jeno ет1ёкёпек [People and state. European debates on the concepts. In m em ory o f Jeno Szucs], Valosag 6 , 1-13. RO NA-TAS, A . (1990a), A historical evaluation o f the recent genetic research. Congressus Septimus In te r n a tio n a l Fenno-U grist arum 1. Sectiones plenares et Symposia. Debrecen, 96-100. RO NA-TAS, A. (1990b), M e g j e g y z e s e k a le g u ja b b g e n e tik a i v iz s g a la to k tort6n eti fe lh a sz n a lh a to sa g a r o l [R e m a r k s o n th e h is to r ic a l u s a b ility o f th e la te st g e n e tic r e se a r c h ]. M agyar Tudomany 90, 918-924. RO NA-TAS, A. (1990c), N im eth Gyula [Julius № m eth], Budapest. RONA-TAS, A. (1991a), An introduction to Turkology. Szeged. RONA-TAS, a . (1991b), Bartha Antal: A m agyar nep ostortenete [Antal Bartha: The proto-history o f the M agyar people]. M agyar N yelv 87, 224-230.
480 R O N A -T A S, A. (1991c), R e v ie w on POHL (1988).
Appendices
Acta Orientalia Hung. 45,
165-166. R O N A -T A S, A. (1991d), N yelvtortenet es ostortenet, in: K lNC SES N a g y , Ё.
(Ed.), Ostortenet es nemzettudat 1919-1931 [Language history and proto history]. Szeged 65-71 = R o n a -T a s (1995a), 199-205. R o n a -T AS, A. (1992a), A m agyar l'rasbeliseg torok eredetehez [On the Turkic origins o f Hungarian literacy], in: S A n d o r (1992b), 9-14. R O N A -T A S, A. (1992b), Csuvas ‘nagy, oreg’. Em lekezesul a turkologia egyik nagy 6 regjere, Halasi Kun Tiborra [Chuvash ‘grand’, ‘eld er’. In m em ory o f one o f the great elders o f Turkology, Tibor Halasi Kun]. Keletkutatas Autum n 121-126. R o n a -T a s , A. (1992c), A m agyar l'rasbeliseg tortenetdhez {ir es betu szavaink etim ologiajahoz) [The history o f Hungarian literacy (on the etym ology o f the Hungarian w ords ir ‘w rite’ and betu ‘letter’, ‘character’], in: SANDOR (1992b), 9-14. R o n a -T a s , A. (1992-1993), In m em oriam Geza Uray (1921-1991). Acta Orientalia Hung. 93-96. R O N A -T A S, A. (1993), A honfoglalas kori magyarsag. A kadem iai szekfoglalo 1991. junius 10. [The M agyars o f the C onquest period. A cadem ic inaugural lecture, 10 June 1991]. Budapest. R o n a -T a s , A. (1 9 9 4 a ) Тбгбк eredetu-e ur szavunk? (A szegedi turkologiai m uhelybol) [Is the Hungarian word ur ‘lord’, ‘noble’ o f Turkic origin? (From the Szeged Turkology W orkshop)]. Magyar Nyelv 90, 138-146. RONA-TAS, A. (1994b), An old Turkic nam e o f Kiev. Acta Orientalia Hung. 47, 175-180. R O N A -T A S, A. (1995a), A magyarsag korai tortenete (Tanulm anyok) [The early history o f the M agyars (Studies)]. Szeged. RONA-TAS, A. (1995b), Ujabb adatok иг szavunk eredetdhez [R ecent data on the origin o f the Hungarian word iir ‘lo rd ’, ‘no b le’]. Magyar Nyelv 91, 63-65. R O N A -T A S, A. (1995c), H ogyan hivtak A rpad dedunokajat? [W hat was Arp ad ’s great-grandson called?], in: KOSZTA (1995), 417—430. R O N A -T A S, A. (1995d), Hungarian baj ‘beauty, charm ’. Acta Orientalia Hung. 4 8 ,4 3 7 -4 4 1 . R O N A -T A S, A. (1996a), Ugor, ogur or ugur ? R em arks on the nam e “FinnoU grian”, in: MESZAROS (1996), 265-269. RONA-TAS, A. (1996b, second edition 1997), A honfoglalo magyar пёр. Bevezetes a korai magyar tortenelem ismeretebe [The M agyars o f the C on quest. An introduction into the ancient history o f the M agyars], Budapest. R O N A -T A S, A. (1996c), A h o n fo g la la s e lo z m e n y e i [T h e p r e lim in a r ie s o f the C o n q u e st] Magyar Tudomany 8 , 819-926.
Bibliography
481
R 6 n a -T a s , A. (1996d), A m agyar o s t c ^ n e t hatterenek keleti torok (ttirk)
forrasai [The East Old Turkic (Turk) sources o f the background o f the H ungarian early history], in: KOVACS - VESZPREMY (1996), 11-22. RONA-TAS, A. (1996e), A m agyar ostortenet hatterenek tibeti forrasai [The Tibetan sources on the background o f the Hungarian early history], in: KOVACS - VESZPREMY (1996), 29-34. R o n a -T a s , A. (1996f), The m igration and landtaking o f the M agyars. H un garian Q uarterly 37 Winter, 37-41. R o n a -T a s , A. (1996g), An “Avar” word: terem, in: BERTA, A. - B r e n d e MOEN, B . - SCHONIG, C. (Eds),Sym bolae Turcologicae. Uppsala, 181-188. R o n a - T a s , A. (1997a), Nom ad sator, arok es кегйёв [Nom ad yurt, trench and enclosure], in: KovAcs, L. - PALADI-KOVACS, A. (Eds), H onfoglalas i s neprajz. B u d ap est, 173-183. R ONA-TAS, A. (1997b), The m igration o f the Hungarians and their settlem ent in the Carpathian Basin, in: Berta, A. [Ed.], H istorical a n d linguistic inter action between Inner-Asia a n d Europe, Proceedings o f the 39 th Perm anent International A ltaistic Conference (PIAC), Szeged, Hungary, 16-21 June 1996. Szeged, 243-253. R 6 n a - T a s , A. (1997c), А пёррё valas a legiijabb kutatasok fёnyёben [The formation o f the ethnos in the light o f the latest researches], in: KOvAcs, L. - PalAdi-KOvAcs, A. (Eds), Honfoglalas es neprajz. Budapest, 23-32. R 6 n a -T a s , A. (1997d), Torok nyelvi hatasok az osm agyar nyelvre [Turkic in fluence on the ancient Hungarian language], in: KOVACS, L. - VESZPREMY, L. (Eds), H onfoglalas i s nyelveszet. Budapest, 49-60. R 6N A -T A S , A. (1998a), Hungarian as a source o f the history o f the Turkic languages, in: DEMIR, N. - TAU BE, E. (Eds), Turkologie heute - Tradition undPerspektive. M aterialien der dritten Deutschen Turkologen-Konferenz, Leipzig, 4 -7 . O ktober 1994, W iesbaden, 261-265. RONA-TAS, A. (1998b), The reconstruction o f Proto-Turkic and the genetic question, in: JOHANSON, L. (Ed.), The Turkic Languages. London, 67-80. RONA-TAS, A. (1998c), Turkic writing system s, in: Johanson, L. (Ed.), The Turkic Languages. London, 126-137. RONA-TAS, A. (1998d), Western Old Turkic, in: JOHANSON, L. (Ed.), The Mainz M eeting. Proceedings o f the Seventh International C onference on Turkish Linguistics, August 3 -6 ,1 9 9 4 . W iesbaden, 619-626. R 6N A -T A S , A. (1998e), Folytassuk a vitat. M egjegyzesek M akk Ferenc konyvbiralatahoz [Let us continue the discussion. Rem arks on the book review o f Ferenc M akk (1997)]. A etas 2 -3 ,2 1 6 -2 2 6 . R 6N A -T A S , A. - FODOR, S. (1973), Epigraphica Bulgarica. A volgai bolgartorok feliratok [The Volga-Bulgharian inscriptions], Studia Uralo-A ltaica 1. Szeged, 190 p.
482
Appendices
ROUX, J.-R (1956), Tangri. Essai sur le viel-dieu de peuple altai'ques. Revue de I ’H istoire des Religions 149, 150, 49-82, 197-250, 27-54, 173-212. ROUX, J.-R (1958), Notes additionelles a Tangri. R evue de I'H istoire des R eligions 154, 48-50. ROUX, J.-R (1962), La religion des turcs de l ’Orkhon des V IIе et V IIIе siecles. R evue de I ’H istoire des Religions 161, 1-24, 199-231. SA L A M O N , A . - ERDELYI, I. (1971), D as volkerwanderungszeitliche Graberf e l d von Kornye. Szeged. S A N D O R , K. (1991), A Bolognai R o va sem lik [The B ologna runiform inscrip tions]. Szeged. SANDOR, к . (1992a), А згёке1у rovasiras atv 6 tel 6 nek lehets 6 ges utjairol es m odjairol [The possible channels and procedures o f adopting the Szekely runiform script], in: SANDOR (1992b), 79-91. SANDOR, K. (Ed.) (1992b), Rovasiras a K arpat-m edenciben [Runiform script in the Carpathian Basin], Szeged. SCHEIBER, A. (1983), Jewish inscription in H ungary fro m the 3rd century to 1686. Budapest. SCHREINER, R (1985), Theophylaktos Simokates. Geschichte. Stuttgart. SCHUTZ, O . (1976), O r m e n y fo rra so k [T h e A r m e n ia n s o u r c e s ], in: H A JD U KRISTO - R O N A -TA S (1976-1982), 1/2, 260-273. SCHUTZ, O. (1995), H ajk i s B el harca. A r ig i orm eny irodalom kincsestara [The feud o f H ajk and Вё1. A treasure trove o f old A rm enian literature]. Budapest. Se n g a , T. (1990), The Toquz Oghuz problem and the origin o f the Khazars. Journal o f Asian H istory 24, 57-69. S E N G A , T. (1993), A besenyok a 8 . szazadban [Pechenegs in the 8 th century]. Szazadok 127, 503-516. SEZ G IN , H. (1967), Geschichte des arabischen Schrifttum s, vol. I. Leiden. S lN E H U U , M. (1980), Orhon-Selegijn runi bicgijn sine dursgal [New findings with runiform script from the region o f the R ivers O rkhon and Selenga]. Ulaanbaatar. SlNOR, D. (1952), Un voyageur du treizem e siecle: le D om inicain Julien de Hongrie. Bulletin o f the School o f O riental and African Studies 14, 589—
602. SlNOR, D. (1963), Introduction a I'etude de I ’Eurasie Centrale. Wiesbaden. SlNOR, D. (Ed.) (1988), The Uralic languages. Description, history and fo r eign influences. Handbuch der Orientalistik. Achte A bteilung, Handbook o f Uralic Studies. Leiden.
SlNOR, D. (1992), Taking an oath over a dog cut in two, in: BETHLENFALVY, G. - BIRTALAN, A. - SARKOZI, A. - VlNKOVICS, J. (Eds), A ltaic Religious
Bibliography
483
Beliefs an d Practices. Proceedings o f the 33rd M eeting o f the PIAC, B uda pest, 301-307. SlNOR, D. (1995), Western inform ation on the K itans and some related ques tions. Journal o f the Am erican O riental Society 115, 262-269. SlNOR, D. (1997), The Uighur Em pire o f M ongolia, in: SlNOR, D. (Ed.), Studies in M edieval Inner Asia. A ldershot-B rookfield, vol. IV, 1-29. SlPOS, J. (1994), Torok es m agyar siratok [Turkic and M agyar mourners]. Keletkutatas, Fall, 46-58. SlPOS, J. (1994-1995), Torok nepzene / - / / [Turkish folk m usic 1-11]. B uda pest. SKOLD, H. (1925), D ie ossetischen Lehnw orter im Ungarischen. Lund. SM IRNOV, A.P. (1951), Volzskie B olgary [Volga Bulghars]. Moskva. SPIRA, G y. - SZUCS, J. (Eds) (1972), N em zetiseg a feudalizm us koraban [Nationhood in the feudal era], Budapest. STEARNS, M. (1978), Crimean Gothic. A nalysis and etym ology o f the Corpus. Saratoga. STEINKELLNER, E. (Ed.) (1991), Tibetan history and language. Studies dedi cated to Geza Uray on his seventieth birthday. Wien. Studien zur Ethnogenese 1-11. Abhandlungen der Rheinisch-W estfalischen Akademie der W issenschaften. Opladen 1985, 1988. Studien zur M achtsymbolik des mittelalterlichen Ungarn. Insigniae regni Hungariae I. B udapest 1983. SUNIK, O.P. (Ed.) (1971), Problem y obscnosti altajskih ja zyko v [The problem o f langue com m unity in the group o f Altaic languages], Leningrad. SWADESH, M. (1950), Salish internal relationship. International Journal o f American Linguistics 16, 157—167. SZADECZKY-KARDOSS, S. (1970), Onogouroi, in: Paulys Realenzyclopadie der klassischen Altertumswissenschaft. Supplem entum band 12, cols, 9 0 2 906. Stuttgart. SZADECZKY-KARDOSS, S. (1972), Ein Versuch zu r Sam m lung undchronologischen Anordnung der griechischen Quellen der Awarengeschichte nebst einer Ausw ahl von andersprachigen Quellen. Szeged. SZADECZKY-KARDOSS, S. (1976), Gorog es bizanci forrasok [The G reek and Byzantine sources], in: HAJDU - KRISTO - R O N A-TA S (1976-1982), 1/2, 119-159. SZADECZKY-KARDOSS, S. (1992), A z avar tortenelem fo rra sa i I [The sources of Avar history I]. Szeged 1992. SzAle, L. (Ed.) (1996), Honfoglalas [The Conquest], 1100 teny 6 s adat [ 1100 facts and data, a supplem ent to M agyar Hirlap]. Budapest. SZEGFU, L. (1994), Latin nyelvu forrasok [The Latin sources], in: KRISTO (1994c), 397-399.
484
Appendices
SZEGFU, L. (1996), Osi szellemi droksegunk I. Gondolatok az osi magyar hitvilagrol [Our ancient heritage o f beliefs. Thoughts on the ancient Hun garian w orld o f beliefs], Szeged.
SZEKELY, GY. (Ed.) (1984), Magyarorszag tortenete. Elozmenyek es magyar tortenet 1242-ig [The history o f Hungary. Prelim inaries and Hungarian history to 1242], I/1-I/2. Budapest.
SZEKELY, G y. (1995), A honfoglalas kori m aradvany nepek a Karpat-m edenceben (roman es nem et elm eletek) [Conquest-period rem nant peoples in the Carpathian Basin (Rom anian and Germ an theories)], in: KOVACSICS (1995b), 106-121. SZENTPETERY, E. (Ed.) (1937-1938), Scriptores rerum Hungaricarum tem pore ducum regumque stirpis Arpadianae gestarum / - / / . Budapest. SZOKE, B.M . (1994), A Karpat-m edence a 9. szazadban [The Carpathian Basin in the 9th century], in: KOVACS (1994), 77-84. SZORENYI, L. (1993), Hunok es jezsuitak, fejezetek a magyarorszagi latin hosepika tortenetebol [Huns and Jesuits, chapters from the history o f the Latin heroic epics in Hungary], Budapest. SZUCS, J. (1966), A m agyar szellem tort 6 net nem zet-koncepciojanak tipologiajahoz [On the typology o f the nation-concept o f the H ungarian history o f ideas], Tortenelmi Szemle 9, 245-269. SZUCS, J. (1971), “Gentilizm us” . A barbar etnikai tudat kerdese: tezisek es vita [“G entilism ”. On the question o f barbarian ethnic identity: Theories and debate]. Tortenelmi Szemle 14, 188-211. SZUCS, J. (1973), Tarsadalom elnrelet, politikai teoria es tortenetszem lelet Kezai Sim on Gesta H ungarorum aban [Social theory, political theory and historical view point in de K eza’s “Gesta H ungarorum ”]. Parts 1-2. Szaza dok 107, 569-643, 823-878. SZUCS, J. (1981), Nation und Geschichte. Budapest, 378. SZUCS, J. (1983), The three historical regions o f Europe: An outline. Acta Historica Hung. 29, 131-184. SZUCS, J. (1985), Torteneti eredetkёrdёsek es nem zeti tudat [Questions o f historical origin and national identity]. Valosag 28/3, 31-49. SZUCS, J. (1986), Sur le concept de nation: reflexions sur le theorie politique m edievale. Actes de la Recherche en Sciences Sociales 64, 51-62. SZUCS, J. (1990), Die drei historischen Regionen Europas. Frankfurt. SZUCS, J. (1992), A magyar nemzeti tudat kialakulasa. K et tanulmany a kerdes elotortenetebol [The developm ent o f the H ungarian national iden tity. Two studies on the prelim inaries], Szeged. TALGREN, А.М . (1914), A kelet-oroszorszagi bronzkorszakrol [On the Bronze Age o f East Russia]. Archeologiai Ertesito 34, 78-83.
Bibliography
485
А.М . (1919), L'epoque dite d ’A nanino dans la Russie Orientale. Helsinki. TA U SZIK , T. (1990a), G enet ikai vizsgalatok es a m agyar sa g tdrtenelm e [Genetic research and the history o f the M agyars], M agyar Tudomany, 8, 904-918. TA U SZIK , T. (1990b), Human genetics and the Hungarian history. In: Congressus Septim us Internationally Fenno-Ugristarum, vol. I.A. Debrecen, 2 0 3 232. TEG Y EY , I. - V E K E R D Y , J. (1965), Bevezetes az indoeuropai nyelvtudomanyba [Introduction to Indo-European linguistic studies], Budapest. TEKIN, T. (1968), A gram m ar o f Orkhon Turkic. Bloomington. TALG REN ,
TEKIN, T. (1987), Tuna B ulgarian ve dilleri [Danube B ulghars and their lan guages], Ankara. (1988), Volga bulgar kitabeleri ve Volga B ulgarcast [Volga Bulghar inscriptions and the Volga Bulghar language]. Ankara. THALLOCZY, L. (1896), A dalek az o-hit tortenetehez M agyarorszagon [Addendum to the history o f orthodox religion in Hungary], Szazadok 30, 199-206. THOMAS, F.W. (1951). Tibetan literary texts a n d docum ents concerning Chinese Turkestan. Part II. Documents. London. THOMAS, W. (1985), D ie Erforschung des Tocharischen (1960-1984). Stutt gart. THOROCZKAY, G. (1994-1995), Az Anonym us-kerdes attekintese [A survey o fth e A nonym us question]. F o n s (\) 199 4 /2 ,9 3 -1 4 9 ,(2 ) 1 9 9 5 /2 ,117-173. TOGAN, Z.V. (1939), Ibn Fadlan's Reisebericht. Leipzig. TOKEI, F. (Ed.) (1983), N om ad tarsadalm ak es allam alakulatok [Nom adic societies and forms o f statehood]. Budapest. TOMKA, P. (1994), 9. szazadi nepesseg a Kisalfoldon [Ninth-century popula tion in the K isalfold region], in: KOVACS (1994), 99-107. TOTH, S.L. (1988), A z etelkozi m agyar-besenyo haboru [The M agyar-P echeneg w ar in the Etelkoz], Szazadok 122, 541-576. TOTH, S.L. (1993), A 9. szazadi m agyar torzsszovetseg szervezete es ktilkapcsolatai (838-896) [The organisation and foreign relations o f the 9thcentury M agyar tribal confederation (838-896]. Ph.D. thesis, Szeged. Т бтн, S.L. (1995), A honfoglalas idopontja [The date o f the Conquest], Acta Universitatis Szegediensis de Attila Jozsef nominatae. A cta H istorica 102, 3-10. Т б тн S.L. (1996), A konstantinosi “Turkia” ertelm ezesehez [To the under standing o f “Turkia” o f Constantine]. M agyar Nyelv, 92, 54-63.
TEK IN, T.
486 TOYNBEE, A. (1973),
Appendices
Constantine Porphyrogenitus and his world.
L on
d o n - N e w Y o r k -T o r o n to . UHRM AN, I. (1987-1988), A z osm agyar kettos fejedelem sdg kerdesehez [On
the question o f the ancient M agyar dual kingship]. Tortenelmi Szemle, 206-226. URELAND, STURE, P. (Ed.) (1985), Entstehung von Sprachen und Volkern, Giotto- und ethnogenetische Aspekte europaischer Sprachen. Tubingen.
VAJAY, Sz . DE (1968), Der Eintritt des ungarischen Stammebundes in die europaische Geschichte (862-933). M ainz-M unchen. VAJDA, L. (1979), Ruchlose und heidnische Dinge, in: Explanationes FennoUgricae in Honorem Hans Fromm. M iinchen, 373-403. VAMBERY, a . (1860), Tarikhi Engurusz, azaz M agyarorszag tortenete czim u torok kezirat ism ertetese [A presentation o f Tarikhi Engurus, that is, the History o f Hungary, a Turkish m anuscript], Magyar Akademiai Ertesito, A Nyelv- es Szeptudomanyi Osztaly Kozlonye 1, 360-362. V a m b e r y , A. (1870), M agyar es torok-tatar szoegyezesek [M agyar and T urkic-T atar w ord correspondences]. Nyelvtudomanyi Kozlemenyek 8 , 109-189. VAMBERY, a . (1877), A torok-tatar nyelvek etimologiai szdtara [Etym ologi cal dictionary o f T urkic-Tatar languages]. Budapest. VAMBERY, H. (1878), Etym ologisches W orterbuch der turko-tatarischen Sprachen. Leipzig. VAMBERY, A. (1882), A magyarok eredete [The origin o f the M agyars], First and second editions. B udapest (G erm an edition: Leipzig). VAMBERY, A. (1914), A magyarsag bolcsojenel, A magyar-torok rokonsag kezdete es fejlodese [At the cradle o f the M agyars. B eginnings and devel opm ent o f M agyar-T urkic relationships], Budapest. VAN DEN BERGHE, P. (1987), The ethnic phenomenon. New York-W estport-L ondon. V A R A D Y , L. (1989), Revision des U ngam -Im age von K onstantinos Porphyrogennetos. Byzantinische Zeitschrift 92, 22-58. V A sA ry, I. (1972), Runiform signs on objects o f the Avar period (6th—8th centuries AD). Acta Orientalia Hung. 2, 335-347. V A sA ry, I. (1974), A m agyar rovasiras. A kutatas tortenete es m ai helyzete [The M agyar runiform script. The history and present-day state o f re search], Keletkutatas, 159-171. V A sA ry, I. (1976), The Hungarians or M olars and the M escers/M izers o f the M iddle Volga Region. Archivum Eurasiae Medii Aevii I, 237-275. V A sA ry, I. (1977), A volga-videki m agyar toredekek a m ongol kor utan [Rem nant M agyars in the Volga region after the M ongol era], in: BARTHA - C z e g l e d y - R o n a -T as (1977), 283-290.
Bibliography
487
VASARY, I. (1982), The “Yugria” problem , in: R6 n a -T as (1982e), 247-257. VASARY, I. (1985), The linguistic aspect o f the “B ashkiro-H ungarian com plex” . Archivum Eurasiae Medii Aevii V, 205-232. VASARY, I. (1 9 93a),/l regi Belso-Azsia tortenete [The history o f ancient C en tral Asia]. Szeged. VASARY, I. (1993b), Nogai, in: GlBB et al. (1960-1993), VII, 85-86. VASILEV, A.A. (1984), History o f the Byzantine Empire / - / / . W isconsin. VASILEV, D.D. (1983a), Korpus tjurkskih runiceskih pam jatnikov bassejna Jeniseja [The findings with Turkic runiform script that came to light in the basin o f the R iver Yenisei]. M oskva. VASILEV, D.D. (1983b), Graficeskij fo n d pamjatnikov tjurkskoj runiceskoj pis'mennosti aziatskogo areala [Graphic elem ents o f the findings with Turkic runiform script in the Asian area]. Moskva. VASILEV, D.D. (1994), Versuch zur Losung der K erbinschrift aus der Umgebung von K alocsa im Spiegel der eurasiatischen Parallelen. Folia Archae o l o g ic a l, 181-191. VEKONY, G . (1986), Levedia m eg Atel 6s Kuzu [Levedia and A tel and K uzu]. Magyar Nyelv 82, 41-53. V EK O N Y , G. (1987a), Spatvolkerwanderungszeitliche K erbinschriften im Karpatenbecken. Acta Archaeologica 39, 211-256. VEKON Y , G. (1987b), Keso-nepvandorlaskori rovasfeliratok a Karpat-medenceben [Runiform inscriptions in the C arpathian Basin o f the late m igra tory period]. Szombathely. VEKONY, G. (1992), Varazsszoveg a halom i honfoglalas kori tem etobol [Magic words from the C onquest-period grave o f Halom ], in: SA N D O R (1992b), 41-49. VERNADSKY, G. (1939), Lebedia. Studies on the M agyar background o f Kievan Russia. Byzantion 14, 179-203. VIKAR, L. (1969), La transposition de quinte des tschouvaches. Bulletin de Г Institute de Musique 13, Sofia. VIKAR, L. (1976), The m usic o f the Finno-Ugrian people, in: HAJDU, R (Ed.) (1976), 319-336. VikAr, L. (1993), A volga-kamai finnugorok es torokok dallamai [The melodies o f the Finno-Ugrians o f the V olga-K am a region and the Turks]. Budapest.
VIKAR, L. - B ER ECZKY, G. (1971), Cheremis folksongs. Budapest. V ik Ar , L. - BERECZKY, G. (1979), Chuvash folksongs. Budapest. VIKAR, L. - BERECZKY, G. (1989), Votyakfolksongs. Budapest. WAIDA, M. (1976), Notes on sacred kingship in Central Asia. Numen. Inter national Review fo r the History o f Religions 23, 179-190.
488
Appendices
(1961), Stam m esbildungund Verfassung. D as Werden derfriihm ittelalterlichen Gentes. K oln-G raz (Zweite A uflage 1977). WERNER, J. (1984), D er G rabfund von M alaja Perescepina und Kuvrat, Kagan der Bulgaren. M unchen. W H IT B Y , M. (1988), The Em peror M aurice and his Historian: Theophilact Sim ocatta on Persian and Balkan warfare. Oxford. W H IT B Y , M. and М. (1986), The history o f Theophilact Simocatta, an English translation with introduction and notes. Oxford. W lN D E K E N S , A.J. VAN (1976), Le tokharien confronte avec les autres langues indo-europeennes I, L aphonetique et le vocabulaire. Louvain. W ITTFO G EL, K.A. - FENG C H IA -SH E N G (1949), H istory o f Chinese society. Liao (907-1125). Philadelphia. W O L FR A M , H. (1979a), Conversio bagoariorum et carantanorum. D as Weissbuch der Salzburger Kirche iiber die erfolgreiche M ission in Karantanien u n d Pannonien. W ien-K oln-G raz. W O L FR A M , H. (1979b), Geschichte der Goten, von den Anfangen bis zur Mitte des sechsten Jahrhunderts. E n tw u rf einer historischen Ethnographie. M unchen (Zweite Auflage 1980). WOLFRAM, H. (1996), Landnahm e, Stam m esbildung und Verfassung. Uberlegungen zu “Vortrage und Forschungen 41,1 und 2” . D eutsches A rchivfur E rforschung des M ittelalters, 161-169. W O L FR A M , H .-D a im , F. (Eds) (1980), D ie Volker an der mittleren und unteren D onau im 5. und 6. Jahrhundert. Berichte des Sym posions der Kommission f u r Fruhm ittelalterforschung, 24-27. O ktober 1978, Stift Zwettl, Niederosterreich, Wien. W O LFR A M , H. —PO H L, W. (Eds) (1990), Typen der E thnogenese unter besonderer Beriicksichtigung der Bayern. Berichte des Sym posions der Kommission f u r Fruhm ittelalterforschung, 27. bis 30. Oktober, 1986. Stift Zwettl, Niederosterreich, Teil 1. Wien. W O L FR A M , H. - SCHW ARCZ, A. (Eds) (1985), D ie Bayern und ihre Nachbarn. Berichte des Sym posions der Kom m ission f u r Fruhm ittelalterforschung, 25. bis 28. Oktober, 1982, Stift Zwettl, N iederosterreich, Teil 1, Wien. ZIEGLER, K. - SO NTH EIM ER, W. (Eds) (1979), D er K leine P auly I-V, Lexikon d erA n tike in f u n f Banden. M unchen. Z lM O N Y I, I. (1990), The origins o f the Volga Bulghars. Szeged. Z lM O N Y l, I. (1991), Km osko M ihaly ёв a m agyar ostortenet [M ihaly Kmosko and Hungarian prehistory], in: K lN C S E S N A G Y , Ё . (Ed.), O stortenet es nem zettudat 1919-1931 [Prehistory and national identity 1919-1931]. Szeged, 89-93. W ENSK U S, R .
Bibliography
489
ZlMONYl, I. (1994), Mohamedan forrdsok [The M oham m edan sources], in: KRIST6 (1994c), 461-462. ZSIRA1, M. (1930). Jugria [Yugria], Budapest. ZSOLDOS, A. (Ed.) (1995), A m agyar honfoglalas a kiilfoldi tort6neti'rasban [The M agyar C onquest in foreign historiography]. M agyar Tudomany 40, 1401-1511.
CHART OF RULERS
Byzantine emperors Based on Fine (1983)
Constantine I Constantius Julian Jovian Valens Theodosius I Arcadius Theodosius II Marcian Leo I Leo II Zeno Basiliscus Zeno (again) Anastasius I Justin I Justinian I Justin II Tiberius I Constantine Maurice Phocas Heraclius Constantine III and Heraclonas Heraclonas Constans II Constantine IV Justinian II Leontius Tiberius II Justinian II (again) Philippicus Anastasius II Theodosius III Leo III Constantine V Leo IV Constantine VI
3 2 4 -3 3 7 337-361 3 6 1 -3 6 3 3 6 3 -3 6 4 3 6 4 -3 7 8 3 7 9 -3 9 5 395^108 408—450 4 5 0—4 5 7 4 5 7 -4 7 4 474 4 7 4 -4 7 5 4 7 5 -4 7 6 476-491 4 9 1 -5 1 8 5 1 8 -5 2 7 5 2 7 -5 6 5 5 6 5 -5 7 8 5 7 8 -5 8 2 5 8 2 -6 0 2 6 0 2 -6 1 0 610-641 641 641 6 4 1 -6 6 8 6 6 8 -6 8 5 6 8 5 -6 9 5 6 9 5 -6 9 8 6 9 8 -7 0 5 705-711 711-713 7 1 3 -7 1 5 7 1 5 -7 1 7 717-741 7 4 1 -7 7 5 7 7 5 -7 8 0 7 8 0 -7 9 7
492
Appendices Irene Nicephorus I Staurakios M ichael I Rangabe Leo V M ichael II Theophilus M ichael III Basil I Leo VI, the Wise Alexander Constantine VII, Porphyrogenitus Romanus I Lacapenus Romanus II Nicephorus II Phocas John I Tzim iskes Basil II
7 9 7 -8 0 2 802-811 811 811 -8 1 3 8 1 3 -8 2 0 8 2 0 -8 2 9 8 2 9 -8 4 2 8 4 2 -8 6 7 8 6 7 -8 8 6 8 8 6 -9 1 2 9 1 2 -9 1 3 9 1 3 -9 5 9 9 2 0 -9 4 4 9 5 9 -9 6 3 9 6 3 -9 6 9 9 6 9 -9 7 6 9 7 6 -1 0 2 5
Danube Bulgar rulers Based on Fine (1983)
Isperikh (Asparukh) Tervel (unknown) Sevar (Sever)
6 7 0 s-7 0 1 7 0 1 -7 1 8 7 1 8 -7 2 5 7 2 5 -7 3 9
Rulers during “Time o f Troubles" Kormisos Vinekh Telec Sabin (? Sebin) Umar Toktu Pagan Telerig Kardam Krum Omurtag Malamir Persian, Persiam (=Malamir?) Boris I Vladimir Simeon Peter Boris II
7 3 9 -7 5 6 756-ca.761 ca.761-ca.764 c a .7 6 4 -7 6 7 767 7 6 7 -ca .7 6 9 ca.770 ca .7 7 0 -7 7 7 ca.777-ca. 803 8 0 3 -8 1 4 814 -8 3 1 8 3 1 -8 3 6 8 3 6 -8 5 2 8 5 2 -8 8 9 8 8 9 -8 9 3 8 9 3 -9 2 7 927 -9 7 1 9 67 -9 7 1
493
Chart o fn ile rs
Samanid rulers 8 1 9 -8 6 4 8 6 4 -8 9 2 8 9 2 -9 0 7
Ahmad I Nasr I, ibn Ahmad Ism a’il (E sm a’il) I, ibn Ahmad
Frankish rulers. The Carolingians Pepin III the Short Charles the Great (king) (em peror) Louis the Pious Lothar I (M iddle F rank) Charles II the Bald ( West F rank) Louis o f Germany (E ast F rank ) Carloman Louis III Charles III the Fat Arnulf o f Carinthia (em peror) Louis the Infant
7 5 1 -7 6 8 7 6 8 -8 1 4 8 0 0 -8 1 4 8 1 4 -8 4 3 8 4 3 -8 5 5 8 4 3 -8 7 7 8 4 3 -8 7 6 died 880 died 882 8 8 1 -8 8 7 8 8 7 -8 9 9 896 900-911
M uslim rulers, the caliphs Muhammad Hegira A bu Bakr ‘ Umar I ‘ Utman (Othman) ‘All
ca. 5 7 0 -6 3 2 622 6 3 2 -6 3 4 6 3 4 -6 4 4 6 4 4 -6 5 6 656-661
llmayj/ads M u'aw iya Yazide 1 Marwan I 'Abd al-Malik Walid I Sulayman ‘ Umar II Marwan II
6 6 1 -6 8 0 6 8 0 -6 8 3 6 8 4 -6 8 5 6 8 5 -7 0 5 7 0 5 -7 1 5 7 1 5 -7 1 7 7 1 7 -7 2 0 7 4 4 -7 5 0
Appendices
494
Abbasids al-'Abbas al-Saffah aI-Man$Cir al-Mahdl M usa al-Hada HarQn al-RaSid al-Amin al-M a'm un al-M u‘ ta?im al-Watiq al-Mutavvakkil al-Muntasir al-Mustain al-Mutazz al-Muhtadi al-M u‘ tamid al-M u”tadid al-Muqtafi al-Muqtadir al-RadT
7 5 0 -7 5 4 7 5 4 -7 7 5 775 -7 8 5 7 8 5 -7 8 6 7 8 6 -8 0 9 809 -8 1 3 813 -8 3 3 8 3 3 -8 4 2 8 4 2 -8 4 7 847-861 8 6 1 -8 6 2 8 6 2 -8 5 5 8 5 5 -8 6 9 869 8 7 0 -8 9 2 8 9 2 -9 0 2 9 0 2 -9 0 8 9 0 8 -9 3 2 9 3 4 -9 4 0
Moravian riders M oymir I Rastislav Svatopluk 1
8 3 0 -8 4 6 8 4 6 -8 7 0 8 7 0 -8 9 4
M oymir II т. Svatopluk II
8 9 4 -9 0 2
CHRONOLOGICAL INDEX (Dates occurring in the text)
Before Christ to 20 000 20 000 -5 0 0 0 to 10 000
Palaeolithic A ge M esolithic A ge last great glacial period (Wurm)
10 000 -7 0 0 0 7000-5000 7th—6 th millennia 6 th—5th millennia 6 th—5th millennia 5000-500 circa 5000 5th—4th millennia 5th—4th millennia 4th—3rd millennia circa 4000 circa 4000 4000-800 3rd millennium circa 3000 circa 2000 circa 2000 circa 2000 2nd millennium 2 nd—1st millennia
Lower H olocene 1 17 Lower H olocene 2 17 N eolithic A ge in the river-valley cultures 22 Indo-European proto-language 2 2 ,1 8 8 taiga spreads in the Ural region 34 M iddle H olocene 17 Neolithic A ge in Northern Eurasia 20 Uralic linguistic com m unity 22 Indo-European proto-language dissolves 188 Uralian peoples splinter apart 318 Uralian U rheim at breaks up 318 F in n o -U g ria n sp e el o fffro m the Uralian com m unity 96 Ancient Iranian period 96 Iranian and Finno-Ugrian interaction 96 imported bronze enters the steppe 20 Ugrians peel o ff from the Finno-Ugrians 9 6 ,3 1 8 earliest written Indo-European sources 187 Hettite Empire emerges 187 core o f ancient China evolves 85 dominance o f an arable-animal rearing econom y among the Finno-Ugrian people 318 locally produced bronze in the Carpathian Basin 20 Hammurabi 23 Hettite texts 187 Shang-Y in dynasty in China 85 Indo-European community arrives in the Indian subcontinent 195 beginnings o f the history o f the Armenian people 76 early Turkic-Tocharian interaction 194 the steppe zone withdraws towards the north 320 first traces o f the Kimmers 195 Zhou dynast)'in China 85 iron enters the steppe 20 Scythians emerge 1 9 5 ,3 0 8 beginning o f independent M agyar history 22, 97, 385
1900-1700 18th century circa 1640 circa 1600 1500 1st millennium 1st millennium 1st millennium 1000-800 11 th—3rd centuries 8th century 8th—7th centuries 8th—5th centuries
20 20 16
496 8 th - 2 nd centuries 835 6 th century mid- 6 th century circa 1st millennium 1st millennium circa 520 507 507 507 504 504 centuries вс 500 425 4th-2nd centuries 330 328 318 3rd-2nd centuries
3rd century 3rd century 3rd century вс 7th century a d 247 2 0 9 -1 7 4 206 2 nd century from 2 nd century circa 2 nd century 175 1 45-186 1st century 1st century 92 circa 90 64 B C -I9A D 36 63 BC-14 AD
Appendices 97, 9 9 ,3 1 8 Old Iranian period 195 Assyrian sources mention the M ede people 20 iron spreads in the steppe 195 Achaemenid dynasty founds the Persian Empire 196 Khwarezm and Sogdiana are populated by Iranian people 320 the steppe zone shifts north 195 Darius the Great makes Persia a world empire 23 Cleisthenes o f A thens’ constitution 23 K ong Fuzi introduces his reforms in China 23 A eschylus’ dramas 23 birth o f Athenian democracy 23 Roman Republic is founded 103 Turkic languages split 17 Upper H olocene 45 death o f Herodotus 175 Kelteminar culture in the Ural 195 Alexander the Great destroys the Persian Empire 195 Alexander the Great marries the Sogdian Princess Roxana first mention o f the Xiongnu in the Chinese sources 203 Scythians are superseded by the Sarmatians in the steppe 1 9 6 ,3 0 8 ,3 1 0 20 iron spreads to Transbaikal areas o f Siberia 253 first mention o f the Xianbei in Chinese sources 6 3 ,9 7 Middle Iranian period 196 Eastern Persia is occupied by the Parthians 204 Maodun, the first significant Xiongnu ruler 85 reign o f the Han dynasty begins in China 78 Arsaces I, founder o f the Parthian Empire M iddle Iranian loan words enter the Hungarian language 97, 101, 320 192 one group o f the Yuezhi m oves west 192 Yuezhi are defeated by the Xiongnu 89 Sima Qian, Chinese historian 196 Parthians rule southeast o f the Caspian Lake 196 M edes rule southwest o f the Caspian Lake history o f the early Han dynasty is completed 89 Sima-qian writes Shiqi, the first Chinese historical work 89 200 Strabo 204 Chinese conquer the western Xiongnu Empire 47 Roman Empire under Augustus
After Christ from 9 25-2 21 227 1st century circa 91
West Han dynasty East Han dynasty Sasanian dynasty took up power Yazigs live in the Carpathian Basin groups o f the Xiongnu migrate west
85 85 197 310 204
497
Chronological index 1st century 1st century
first centuries 1st century 98 2nd-3rd centuries 107-114 circa 150 151-178 3rd century 3rd century 3rd century 3rd century 3rd century second hatf o f 3rd century 3rd—4th centuries 217-277 227 227 227-241 circa 230 to 264 289 end 3rd century 4th century 4th century 4th century circa 300 301 309-379 348 around 350 mid-4th century circa 350 circa 350 circa 350 circa 350 circa 360 after 370 375 383 385 386 395 early 5th century early 5th century
Chinese counter-attack against the Xiongnu Tocharian people settle in the region o f Turfan Finnish tribes settle in their present-day regions Yazigs lived in the Carpathian Basin Tacitus writes G erm ania Sogdian texts describe the history o f the steppe maps o f Marinus first groups o f Goths appear north to the Black Sea P to lem y's m ap o f the world groups o f Jews wander throughout the Roman Empire the first Georgian state the metal stirrup appears in eastern Asia Goths conflict with the Roman Empire first mention o f the ‘Khaghan’ title in Chinese sources
204 194 183 310 184 6 3 ,6 4 45 203 45 90 79 103,251 203 213
advance o f the Goths brings Scythian rule to an end 308 Hunnish period in the steppe 320, 387 M ani’s religion, Manichaeism gains state support 197 Sasanian dynasty took up power 197 Sasanian rule in Persia begins 63 Ardashir I, ruler o f Persia 63 the G oths’ settlement in the Lower Dnieper region is completed 203 China is divided into three parts 85 H istory o f the Three K ingdom s is completed 89 end o f Goth incursions on the Balkan 203 the ethnic name Hun is used 309 Huns set o ff towards Europe 320 nomadic peoples appear in the Volga and Ural region 320 history o f the late Han dynasty is written 89 Christianity becom es a state religion in Armenia 76 Shapur II, ruler o f Persia 197 Visigoths (West Goths) are baptised and settled in the Roman Empire 203 King Ermanrich ruled the Goths 181 Huns extend their power to the Southern Ural region 325 new influx o f peoples into the former Sogdiana 197 Huns set o ff towards Europe on the Khazakh steppe 210 ,3 2 1 Turkic peoples appear on the H uns’ trail 211 ,3 2 1 the Var and Chyon tribes attack Persia 197 Ostrogoths (East Goths) establish an empire between the Rivers Dniester and Don 203 Ostrogoth Empire is overthrown by the H unnish-Alanian alliance 203 Huns cross the River Volga 206 death o f Wulfila 203 Ruanruan appear in today’s M ongolia 213 the Tuobas found the Wei dynasty 253 Huns reach the north Caucasus 206 beginnings o f M a g y a r-O g h u r connections 321 Ruanruan Empire 213
498 5th century 5th century from 5th century from 5th century circa 400 4 0 0 -6 0 0 between 4 10 and 420 434 437 440-461 441 445 circa 4 4 8 -4 4 9 circa 450 4 5 0 -4 5 7 451, spring 452 453 mid-5th century mid-5th century second half o f 5th century after 460 circa 463 476 483-511 488 from 6 th century circa 6 th century 6 th century first half o f 6 th century circa 506 527 5 2 7 -5 6 5 530 540 5 4 1 -7 0 0 546 5 th—7 th centuries 6 th century 6 th century 6 th century 6 th century 6 th century m id- 6 th century m id- 6 th century-630
Appendices turn in Roman-Germ anic relations Khitais are mentioned by the sources ethnic name ‘Onoghur’ crops up Roman-Germanic cohabitation centre o f the Hunnish Empire in the region between the Rivers Danube and Tisza and in Pannonia initial phase o f the M iddle Chinese linguistic period birth o f the rethor Priscus Bleda and Attila com e to power Huns defeat the Germanic Burgundians Pope Leo I Attila challenges Byzantium Attila Great King o f the Huns rethor Priscus at the court o f Attila Avar Khaghan defeats the Ughurs Marcian, Roman emperor Attila and allies attack Visigoths (West Goths) Attila turns against Italy Attila’s death, disintegration o f the Hunnish Empire Hephthalite or Ephthalite dynasty Oghur Turks set o ff w est from Khazakh steppe role o f On-Oghur Bulghars increases in Eastern Europe Chuvash-type language present w est o f the Urals Oghur/Ughur groups extend their power to the south and west o f the Ural Mountains end o f the Western Roman Empire C lovis o f the M erovingian dynasty chronicle o f the Song dynasty is completed Syriac chronicles stirrup spreads in Eastern Europe Kushnarenkovo culture Sabirs in Eastern Europe triumph o f Sabir hegem ony in Volga region ‘Slav’ name appears in Procopius Emperor Justinian annals o f the Southern Chi are completed description o f the world by Cosm as Indicopleustes the peak o f Tuyiihum state Tiele or Ughur people revolt against the Ruanruan Otilken, the holy mount o f the Turks Muayeris (M ouageris) evidence o f the ethnic name ‘Kangar’ in the Caucasus first mention o f the Alani in the sources changes in archaeological finds o f the V olga-K am a region ruling dynasty in Yemen converts to Jewish faith Jordanes com pletes his G etica Slavs first appear in the Balkans
375 253, 374 285 3 7 5 ,3 7 6 206, 207 88
51 207 207 207 207 207 51 211
207 207 207 208 197 3 2 1 ,4 1 4 112 212
209, 321 46 376 89 75 110 121
2 1 2 ,3 2 7 212
240 48,211 89 51 253 213 10
297 420 200
215 231 55 286
Chronological index circa 552 before 552 553 553 553-572 circa 554 555 555 circa 555 after 556 557 558 552-558 558-559, winter circa 560 561 562 562 565 565-636 566 567 567 568 568, Easter 570s 570s 572-580 574 578 circa 580 580 between 566 and 581 582 582 582 circa 583 circa 583 587 until 591 6th century end 6 th century end 6 th century
499 fall o f the Ruanruan or Asian Avar Empire; beginnings o f the First Turk Khaghanate; Khaghan Bumin 80,213, 2 7 8 ,3 1 6 200 Procopius mentions the Alani 214 death o f Khaghan Bumin Procopius com pletes his H istory 59 214 Mugan Khaghan 89 annals o f the Touba-Wei dynasty are completed steppe history o f Pseudo Zachariah 75, 213, 229 2 6 1 ,3 1 6 Avars approach the Caucasus Avars overthrow the Sabirs 213 East Turks conquer the Turkic peoples o f the Eastern European steppe 80 214 Istemi and Husrau Anoshirvan join forces 112 Sabirs are defeated by the Avars 5 1 ,2 8 2 Agathias com pletes his work Byzantine Emperor Justinian turns to the Utrighur 11 261 Khaghan Bayan, Avar ruler 213 Frankish rulers fight the European Avars Avars reach the Lower Danube region 2 6 2 ,3 1 6 Avar armies fight: at the River Elba with Sigibert I 261 262 Justin 11 refuses to com ply with the Avars’ requests 423 Isidorus o f Seville 261 Avars defeat the Frankish armies together with their Avar allies, the Longobardi overpower 123, 262 the Gepidae history o f Husros I estimates Khazar population to be 353 fifty thousand led by the Sogdian Maniah, a Turk envoy travels to Byzantium 214 ,2 8 1 Longobardi cede the entire Carpathian Basin to the Avars 123, 262 112 Sabirs disappear from view in the sources ceaseless Avar attacks on Byzantium during its military 262 campaign against Persia 80, 279 inscription o f Bughut Avars overpower the Byzantine Emperor Tiberius 262 death o f world chronicle-writer Johannes Malalas 51 death o f Cassiodorus 51 Sui dynasty unites China 85 5 1 ,2 8 2 Turks are mentioned in Theophanes Bizantius 240, 262 Avars occupy Sirmium Menander Protector com pletes his work 51,211 282 Agathius dies death o f the Avar Khaghan Bayan 262 peoples fleeing from the western Turks seek Avar refuge 262 298 Stronghold o f Matzaron is mentioned Byzantium is under growing Avar and Slav pressure 262 mention o f the Kapaghan title o f the Tuyiihuns 253 217 Bulghars migrate from the Kuban region the M agyars m ove fro m the south Urals to the D o n -K u b a n region 325
500 end 6 th century end 6 th century early 7th century early 7th century early 7th century early 7th century early 7th century 7th century 7th century 7th century 7th—10th centuries 7th~8th centuries 600 6 0 0 -8 0 0 early 600s 602 6 10-641 until 620 620s 620s 6 2 0 s-6 3 0 s 622 626 6 2 6 -6 2 8 627 627 after 628 circa 629 circa 630 630 631 632 635 635 636 636 6 3 6 -6 5 6 637 638 640 640s 641 642 after 642 644—646 mid-7th century
Appendices 240 major Slavonic settlements in the northeast Balkans 322 changes due to the weakening o f the western Turks 276 M aurice’s compendium on strategy 2 1 5 ,3 2 6 Bulghars proclaim independence in the Dnieper region 82, 280 rise o f the Old Tibetan Empire 241 influx o f Slavs into the Adriatic region 242 growing significance o f Slavs in Transylvania 231 Tengrism is the Khazars’ religion 241 Slavs’ significance increases alongside the Avars’ 286 Slavs call the Magyars White Ugrians Old Tibetan sources 82, 280, 407 321 independent Oghur groups still dw ell in South Siberia 240 Byzantium re-conquers Sirmium Turkic loan w ords enter H ungarian 398 Oghur groups regain independence 215 52 death o f the Emperor Maurice Heraclius, emperor o f Byzantium 52, 230, 240, 286 Avar and Slavonic military campaigns against 262 Byzantium Frankish merchant Samo organizes Slavs 2 4 1 ,3 7 8 241 independence o f the dukes o f Serbia and Croatia 242 Slavonic groups in Dalmatia, Bohem ia and Slovakia 64 Muhammad flees from M ecca to Medina: the Hegira 2 4 1 ,2 6 2 Persian-A var-Slavonic alliance against Byzantium 282 The Georgian Chronicle calls the Khazars ‘West Turks’ Heraclius defeats the Persians with Khazar aid at Tiflis 5 2 ,2 3 0 ,2 6 2 ,3 2 5 Muhammad entrenches his strongholds to fight o ff the hostile Arabs 296 354 talks o f the Caucasus Albanian Bishop with the Khazars 241 Slav-A var war 80, 230 Western Turk Empire collapses 262 Byzantine-Frankish diplomatic connections 262 one group o f Bulghars seeks refuge with the Bavarians 64 death o f Muhammad 64 Arab Caliph Omar occupies Damascus Khuvrat rebels against the Avar ruler 215, 219, 322, 326 64 Omar occupies Jerusalem and Ctesiphon annals o f the Liang, Chen and the Northern Chi dynasties, and o f the Zhou dynasty are completed 89 annals o f the Sui dynasty on the forebears o f the Turk peopl e 281 Arabs overthrow the Sasanian dynasty 63 Theophilactus Simocattes com pletes his World H istory 211 Arabs penetrate into Armenia 76, 78 Khuvrat and his allies defeat the Avars 123 death o f the Emperor Heraclius 219 64 Arabs deal the Sasanian Empire a decisive blow Onoghurs in the Carpathian Basin 284 H istory o f the Yin dynasty is completed 89 nascent independent Tibetan literacy 83
501
Chronological index 650 after 650 mid-7th century second half o f 7th century 650-670 651 652-653 656 659 661 circa 670 circa 670 670s circa 670 678 679 after 679 circa 680 680-740 681 682 687 695 698 early 8 th century end 7th century-895 around end o f 7th century early 8 th century early 8 th century 8th century 8th century 8th—9th centuries 8th—9th centuries 8th—9th centuries 8th—9th centuries circa 700 circa 700 circa 701 705-711 711 711
first year in the Old Tibetan annals death o f Khuvrat Khazar rule extends to the Crimean Peninsula
84 219 230
runiform script diffuses in the steppe 80 Khazars obliterate the Bulghar Empire 230, 326, 387 Western Turk tribes are mentioned in the annals o f the Tang dynasty 225 Khazar ruler takes up the Khaghan title 230, 282 Ali fights Muhammad’s widow 64 a historical summary o f the Northern Chinese dynasties is compiled 89 M uaviya founds the Umayyad dynasty 64 K hazars defeat the O noghundur-B ulghars w ith M agyar aid, a n d establish an em pire 123, 219, 230, 262, 322 Onoghur-Bulghars arrive to the Lower Danube 286 Bulghar Turks settle with the Slavs in the Balkans; other groups m ove in with the Avars 220, 242 M agyars take up the p la ce o f the B ulghars betw een the R ivers D nieper a n d D anube 326, 413, 414 history o f M ichael o f Syria 74 Avar envoys journey to Constantinople 262 the A rm enian G eography mentions the Magyars as Turks 282 Elteber Alp marries the daughter o f the Khazar king 226 Second Turk Khaghanate 80 Albanian Bishop Israyel on the horse sacrifice rite o f the Huns (Khazars) o f the North Caucasus 2 3 1 ,3 6 9 Kuver and his groups depart from the Avar Empire 263 Armenian chronicle 78 Byzantine Emperor Justinian II is exiled to Kherson 230 66 Arabs occupy Carthage Byzantine colonies still exist in Crimean Peninsula 327 M agyar p eo p le live in the E telkoz 387 Turkic and Slav bilinguality is widespread in the Carpathians 228 Old Turkic inscriptions 80, 82, 278 self-designation o f the M agyars is ‘M a g y a r ' or ‘M a g y e r ' 307 late Lomovato culture 122 Arabian sources mention the Alani 200 Arabs crush the Visigoths (West Goths) 203 Slavonic name ‘Ungri’ is adopted by the Slavs’ 2 8 1 ,2 8 6 neighbours Turkic phonetic change in szolo type words 36 k- > h- consonant change in the Khazar language 39, 226 Geographus Ravennas com pletes his C osm ographia 56 124 Avar Empire regains strength death o f Asparukh 227 Justinian reclaims his throne with Bulghar aid 230 230 Philippicus Bardanes drives Justinian away Arabs reach the Indus River valley 66
502 7 1 7 -7 1 8 7 2 0 -7 2 5 720 or 732 723 724 725 726 7 2 6 -8 4 3 730 circa 730 7 3 1 -7 3 6 T il 732 732 7 3 3 -7 4 6 735 7 3 5 -7 5 6
1Ъ1
737 737 739 739 740 circa 740 750 750 after 750 circa 750 circa 750 m id- 8 th century m id- 8 th century second half o f 8 th century mid-8th - 10 th centuries 8 th—9th centuries 751 751 751 7 5 3 -7 5 4 755 7 5 9 -7 6 0 760 760
Appendices Arabs besiege Constantinople Turk inscription o f Koli chor Turk inscription o f Ong Arabs occupy Balanjar Turk inscription at Ihe Ashete Arab campaign against the Khazars led by Jarrah Turk inscription o f the minister Tonyukhukh dom estic struggles in Byzantium Khazars defeat the Arabs under Ardabil Bulan, the first converted Jewish Khazar ruler ‘Ungarus’ name crops up Emperor Leo I ll’s son marries the daughter o f the Khazar Khaghan Charles Martel halts the Arabs at Poitiers funeral and inscription o f the Turk Khaghan 8 1 ,8 3 , Kol Tegin Byzantine bishopric list Khaghan B ilg e’s inscription Taihir inscription Marvan leads campaign against the Khazars; the Khazar ruler is forced to convert to Islam; a large section o f Alani escape to the north 201, 202, 220, 223, M agyars a n d B ulghars m ove north a long the R iver Volga 306, 322, Arabs defeat the Tiirgesh ruler o f the Turkic Gandhara dynasty relinquishes the
227 81 81 220, 224 81 201 81 258 230 232 285 230 6 6 , 230
148,374 52 81 81
229, 230 3 2 7 ,4 1 4 230
228 throne to his son, Kesar 227, 242 first Bulghar dynasty dies out 46 Byzantium halts the advance o f the Arabs 80, 252 Uighurs overthrow the Second Turk Khaghanate 66 , 327 Arab Abbaside dynasty com es to rule 81 Turk inscription o f Tez II 229 Uighurs finally take over power from the Turk 81 Uighur runiform inscriptions 221 future Volga Bulghars reach the southern bend o f the Volga 229 an Uighur Khasar tribe is mentioned by the Chinese sources 376 Lex Salica is compiled 127, bone needle-case o f Szarvas Saltovo-M ayak culture eastward advance o f the Slavs Byzantium surrenders Ravenna Pepin III the Short is crowned king o f all Franks in the battle on the River Talas the Arabs and Tibetans 6 6 , 82, defeat the Chinese Turk inscription at Terh Uighurs defeat the Kharlukh at the River Selenga M oyin chor’s Turk inscription Khazar tarkhan R as’ military campaign against the Alani Arab-Khazar alliance
264 139 320 46 56 230 81 234 81 201 230
Chronological index 761 762 763 763 764 until 764 768-814 768 775 -7 8 0 111 111 7 86-809 787 787-789 788 790 790 791 795 796 7 97-809 798 after 798 end 8th century end 8th century end 8th century end 8th century end 8th century 9th century 9th century 9th century 9th century 9th century 9th century 9th century 9th century 9th century early 9th century early 9th century early 9th century early 9th century early 9th century early 9th century 800 Christmas circa 800
503 ‘Hungaer(us)’, ‘Hunger(us)’, and ‘H ounger(us)’ family names 285 Uighur Khaghan converts to Manichaean faith 82, 252 Tibetan forces occupy the capital o f China 82 Turk inscription o f Sevrei Sumun 81 Tibetan Zhol inscription 84 Arab-Khazar fights 231 rule o f Charles the Great 56 Avar envoys visit Charles the Great 263 Leo IV (the “Khazar”) o f Byzantium 230 founding charier o f the monastery o f Krems, with the ‘zhupan’ title (jo p a n ) 58 Telerig seeks refuge in Byzantium, where he is baptised 227 Harun al-Rashid Arab Caliph 6 6 , 232 Crimean Goths under Khazar rule 203 coin at the Bolshie Tarhani cemetery 1 21 ,4 0 0 Franks cross the River Enns 263 Avar-Frankish negotiations on border issues 263 death o f Levond 78 Charles the Great leads the Franks into battle against the Avars 1 2 6 ,2 6 3 ,2 8 4 Crimean Goths under Byzantine rule 203 Avar tudun is baptised at the Frankish court 2 6 3 ,2 8 5 ethnic name ‘Hungarius’ crops up 285 the last Khazar attack against Transcaucasia 231 Khazars successfully conclude the battles with the Arabs 328 Georgian work, the M artyrdom o f A b o is written 79 Slavonic centre in Pannonia 126 M agyars fo rm a lly still belong under K hazar suzerainty 328 follow ing internal fights, the Khazar nobiliary stratum converts to Jewish faith 328 last use o f the N agyszentm iklos Treasure neighbours o f the Slavs call the Magyars Ungri 286 PaxK hazarica 2 0 3 ,3 2 8 .3 8 8 Rus influence in the Central Volga region 182 Theophanes 5 2 ,2 8 2 end o f the emergence o f the Slavonic peoples 378 Nicephorus 52, 220 Karyala is mentioned in the sources 185 ethnic names ‘Onger’ and ‘U ngerus’ crop up 285 the d ow nfall o f the H ouse o f Levedi 329 Slav efforts to “ reoccupy” Peloponnisos and Hellas 257 Oghuz live in the Lake Aral region 234 234 Kimeks live by the River Irtish Kharlukhs at the Rivers Talas and Chou 234 strengthening o f the Serbs 242 use o f the ‘Turk’ name becom es restricted in the Byzantine sources to the Magyars 282 Charles the Great is crowned emperor 5 6 ,3 7 7 Khazar ruler Obadiya bolsters Judaism 232
504 circa 800 until 802 8 0 3 -8 1 4 804 805 810 or 821 811 811 812 812 813 813 -8 3 3 814 814-831 818 circa 820 820 -8 2 3 821 8 2 1 -8 2 2 822 822 826 827 829 829 circa 830 8 3 0 s-8 4 0 s 8 3 0 -8 4 6 832 832 833 838 circa 839 8 3 9 -8 5 0 840 circa 840 8 4 0 -8 5 0 840 after 840 842 843 846 8 4 6 -8 7 0 8 4 6 -8 4 7 8 4 6 -9 0 2
Appendices Khavar people flee from the Khazars to the Etelkoz 322, 328 Avars hold onto their positions as far as the Vienna forest 263 reign o f Krum in Bulgharia 379 Avar Khapkhan Theodorus visits the Frankish court 263 Avar Khaghan Abraham is baptised in the River Fischa 263 Uighur inscription at Kharabalghasun I 81 Avars in the Frankish court 263 Byzantium occupies Pliska 227 trade restrictions to Bulghar merchants by Leo the Wise 258 ethnic name ‘Wanger’ crops up 285 Theophanes’ great C hronographia is completed 55, 282 Caliph al-M a’mun, son o f Harun al-Rashid 224 death o f Krum, Bulgharian ruler 227 Omurtagh takes a firm grip on central power 227, 242 death o f the chronicler Theophanes 52 264 major changes in the population o f Transdanubia Thomas-rebellion 258 Tamin ibn Bahr writes about the Uighurs’ magical rain-making stones 365 Tibetan-Chinese peace treaty 84 Slavonic envoy from Moravian territory visits the Frankish court 243 Avar Khaghan and his train at the Frankfurt assem bly 263 Crete occupied by Arabs 258 Bulghars storm the Timochani in Sirmium, and eastern Slavonia 242 death o f the patriarch Nicephorus 52 eastern Frankish borders are dispersed 263 Omurtag extends his power to the River Maros 126, 242 sources mention Rus Khaghanate 246 Moymir, the first Moravian ruler 243 Uighur tribal confederation breaks up 252 Malamir concludes a peace with the Franks 243 Ratbod introduces Pribina to Louis o f Germany 243 Khazar fortress at Sharkel is erected with Byzantine help 324, 328, 406 M agyars at the L ow er D anube 329 Bulghar military campaign against the Serbian chieftain Wlastimir 242 Uighur inscription o f Suji 81 Franks accord Pribina an estate by the River Zala 133, 243 castle com plex at Mosaburg (Zalavar) 133 Khirghiz defeat the Uighurs and bring an end to Kharlukh rule 8 0 ,2 5 2 Transdanubia under Slavonic rule 264 anti-Buddhist revolt in the Old Tibetan Empire 83 internal fights end in Byzantium 46, 258 Pribina is accorded new estates in Burgenland 243 Rastislav, ruler o f Moravia 243 first work o f Ibn Hordadzbeh 68 , 296 Tankeyevka cemetery 121, 222, 400
Chronological index 847 -8 5 8 and 8 6 7 -8 7 7 around 847 mid-9th century mid-9th century mid-9th century after mid-9th century
505
Ignatius is the Patriarch o f Constantinople 259 al-Khwarezmi died 68 Rus campaign to the Caspian Lake region 231 Avars’ political establishment collapses 264 Slav settlers in the region o f the Rivers Upper Tisza and Szam os 266 Slavs are positively known to have called the Magyars ‘Ungri’ 286 after mid-9th century important changes in the history o f the Khazars 257 850 church is consecrated at Zalavar in honour o f the Virgin Mary 243 between 850 and 920 Volga Bulghars reach the region o f B olshie Tigani 121 852 ethnic name ‘Croat’ crops up 377 852-889 Boris, Bolghar ruler 6 0 ,2 2 8 854 M uslim Khazars and Alani settle in Islam territories 201 858-867 Pope N icholas I 244 859-873 Aldwin, Bishop o f Salzburg 56 860 M ethodius may perform Slavic liturgy 260 860s Transdanubia is called Sclavinia in the sources 264 circa 860 Khazars’ conversion to Judaism ends 233 circa 860 Rastislav gains independence from the Franks 243 860, 8th May Louis o f Germany’s deed o f gift mentions the Vangars 58, 285 861 Constantine (Cyril) and Methodius proselytise among the Khazars; they might have met the Magyars in the Crimea 233, 2 8 6 ,2 9 6 , 329 862 Rastislav sends to Constantinople for m issionaries 243 862 in alliance with Rastislav, the Magyars encounter the Franks 329 862 Carloman rebels against his father, Louis o f Germany 331 863 Synod o f Lateran 259 863 Constantine and M ethodius at the seat o f the Duke o f Moravia 6 0 ,2 4 4 864 Druthmar o f Aquitania’s work written in Westphalia 232 864 Louis o f Germany defeats Rastislav under Ddveny (D evin) 244 865-868 Danube Bulghars are baptised 3 2 9 ,3 6 8 867, 13th N ovem ber Pope N icholas I died 259 867 deposition o f Patriarch Photius 259 867 Byzantium finally breaks away from the Roman Church (1054: official rupture) 46 867 Constantine and Methodius travel to Rome 245 867-872 Pope Adrian II 245 868 Byzantine-Bolghar peace treaty 244 869 Fourth Council o f Constantinople 259 869 Bolgharia joins Byzantine Christianity 2 9 6 ,3 2 9 869 Constantinus (Cyril) dies in Rom e 60, 2 4 5 ,2 8 6 circa 870 parts in al-Jayhani’s report on t h e ‘Majghars’ 6 9 ,2 9 5 ,3 2 8 870 Arabs occupy Malta 258 870 Rastislav is blinded by the Franks 245 870 Conversio Bagoariorum et Carantanorum is completed 56, 116, 264 circa 870 Methodius writes the Constantine Legend in Rom e 60 after 870 G lagolithic writing em erges 440
506 872 872 872 8 7 2 -8 8 2 8 7 3 -9 0 7 874 875 876 877 circa 879 8 8 0 -8 8 2 880 880 881 88 1 -8 8 2 after 882 882 885 885 885 8 8 5 -8 8 6 885 -1 0 4 5 end 885 886
8 8 6 -9 1 2 887 888
8 8 8 -9 2 4 889 891 891 891 892 892 892 8 9 2 -8 9 8 8 9 2 -8 9 9 early 893 893 893 8 9 3 (8 9 4 ) 8 9 3 -9 2 7 8 9 3 -9 2 7 894
Appendices 258 Paulikians’ m ovement is put down 260 Hadrian dies 72 al-Jakubi’s World H istory completed 245, 260 Pope John VIII 56 Theotmar, Archbishop o f Salzburg 245 Moravians conclude a peace with the Franks at Forchheim 260 Charles the Bald gets the crown 246 death o f Louis o f Germany 260 Charles the Bald dies 247 death o f Rurik 247 O leg seizes power over the Rus 246 M ethodius is cleared o f charges in Rome 246 death o f Carloman together w ith the K havars, the M agyars m eddle in M ora via n -F ra n kish fig h ts under Vienna 56, 329, 331, 369 286 Methodius might have met the Magyars at the Danube 286 the neighbours o f the Slavs call the Magyars Ungri 331 Byzantine m ission conferred with the Magyars Pope Marinus I dies 260 death o f Methodius 60, 249, 286 Svatopluk concludes a peace with Arnulf, heir to the throne 246 Ibn Hordadzbeh com pletes his abridged work 68 , 296 reign o f the Armenian Bagratid dynasty 76 M ethodius’ disciples are driven out o f Moravia 61 Leo the Wise deposed Patriarch Photius for Stephen 260 reign o f Leo VI the Wise 5 3 ,2 5 8 246 Charles III the Fat is forced to give up his throne death o f Charles the Fat 260 Berengar becom es successor o f Charles the Great in Northern Italy 2 6 1 ,3 3 6 R egino’s information on the Magyars being expelled 5 7 ,3 3 0 ,4 2 0 ,4 2 1 260 Pope Formosus is elected G uido’s coronation as Roman Emperor 260 solar eclipse is observed in Byzantium 331 death o f the chronicler al-Baladzuri 72 242 Bulghars control the salt mines along the River Maros A rnulf attacks Svatopluk with Magyar aid 2 4 6 ,2 6 1 ,3 3 1,369 Lambert o f Spoleto rules in Northern Italy 260, 336 Regino, Abbot o f the monastery o f Priim 57 Samanid ruler Esm a’ il ibn Ahmad com es to power 256 ,3 3 0 Slavonic Old Bolgharian becom es official language in Bolgharia 60 P echenegs are defea ted by the Oghuz, a n d set out west, fr o m the Ural passage 257, 324, 331 death o f Stephen, the Patriarch o f Constantinople 331 Simeon, Danube Bulghar ruler 60, 228, 258 beginnings o f Cyrillic writing 440 N icetas Scleros m eets A rp a d a n d K ursan 344
507
Chronological index 894 894 894 894 894 894-896 895-902 895 895 895, spring 895, autumn 895-898 896 897 898, 15th October 898 898-899 898-900 899 899 899 899-900 899, December 899-900 last decade o f 9th century end 9th century end 9th century beginning o f 10 th century early 10 th century early 10 th century early 10 th century early 10 th century 10th century 10th century 10th century 10th century 10th century 10th century 10th century 10th century 10th century
successful Mag)>ar attack against the Danube Bulghars 332, 337, 388 Pechenegs cross the River Volga 343, 331, 421 Svatopluk sends envoys to the M agyars, a n d attacks the Franks w ith M agyar aid 332, 369 death o f Svatopluk 246 Arnulf is crowned king 2 6 1 ,3 3 6 ,3 8 9 Bulghar-Byzantine war 5 3 ,6 1 the H ungarian C onquest xviii. 3 3 2 -3 3 8 Pechenegs attack the M agyars 239, 331, 387, 421 Magyars feature as Avars in the Fulda Annals 335 m ain M agyar arm ies in the eastern side o f the C arpathian 3 3 2 ,3 3 4 Basin, fro m where they atta cked the B ulghars M agyars enter the C arpathian B asin fr o m the Verecke p a ss 397 fir s t p h a se o f the M agyar C onquest; w estern fro n tier: the R iver Tisza 324, 334, 336, 388 Arnulf is crowned emperor by the Pope 246, 261, 336 al-Yakubi dies 68 death o f Lambert o f Spoleto 2 6 1 ,3 3 6 A r n u lf concludes a p a c t w ith the M agyars 389 bishops and archbishops are ordained on Moravian soil 336 Pope John IX 261 Philotheus refers to the Magyars as Turk 53 M agyars se n d m ajor fo rc e s to Italy 337 Arnulf, too, enters Moravian conflict 337 M agyars together with A r n u lf defeat B erengar 2 6 1 ,3 8 9 337, 389 death o f Arnulf sec o n d p hase o f the M agyar C onquest: the conquering o fT ransdanubia 334, 345, 388 peak o f Khazar Empire Slavs’ neighbours, too, refer to the Magyars as 'U ngri’ Greek letters enter the Szdkely runiform script
257 286 440
253, 374 Khitai dynasty com es to the zenith o f its power Bulghar Turks in the Volga-K am a-Cherem shan region 68 к > h, d > z phonetic change in the Volga Bulghar language 226 final -g disappears in back vocalic words from the Volga Bulghar language 226 Volga Bulghars speak a Chuvash-type language 226 migrating east from the Iberian Peninsula, the Jewish groups abandon their Hebrew tongue to Yiddish 7 ,9 0 Kievan Letter 289 prothetic y- appears in Volga Bulghar 432 the ethnic name *B achgird 290 Hrabar claim s the Slavs originally used runiform script 440 Slavisation o f the Meryas begins 183 al-Istahri 7 1 ,3 5 6 al-Dinavari 72 Magyars are referred to as Huns 309
508 10 th century 10 th century 10 th century 10 th century 10 th century 10 th century
from first half o f 10 th century 10th—13th centuries 10th-15 th centuries circa 900 circa 900 900s 900 9 0 0 ,29th June 900, spring end 900 9 0 0 -9 0 2 900-911 902 902 904 904 9 0 5 -9 5 9 906 until 907 907 9 0 8 -9 3 2 circa 911 912-961 9 1 3 -9 4 5 circa 913 9 1 3 -9 5 9 915 916-1114 9 2 0 -9 4 4 921 921, 2 nd April 9 2 1 -9 3 5 9 2 1 -9 2 2 922 922, 12 th May before 923 923
Appendices hey-day o f Armenian historiography Masudi maintains that the capital o f the Alani was called Magash Permian-Bulghar-Turkic interaction Jewish community in the Carpathian Basin craft o f iron smelting was uninterrupted in the Carpathian Basin ethnic name o f the Ughur group changes to Yughur in the Chinese sources ‘Ungri’ variant o f the ethnic name in Byzantine sources
Magyars in the Volga-Kama-Ural region
76 433 180 287 357 435, 436 283 15, 429 441 246
transformation o f the Szdkely runiform script Slavs reach the line o f the N izhniy N ovgorod-R yazan Volga Bulghar ruler in alliance with Khwarezm 222 Volga Bulghars reach the River Kama 68 , 121 ,4 0 0 letter o f Theotmar, Bishop o f Salzburg, to Pope John IX 56, 369 337 Magyars attempt to penetrate into Venice
returning from Italy, the Magyar army conquers Transdanubia Bavarians erect a fortress against the Magyars on the River Enns end o f the third period o f the Magyar Conquest coins o f Louis the Infant Arethas refers to the Magyars as ‘Turkoi’
3 3 7 ,3 5 2
3 3 7 ,3 8 9 3 3 4 ,3 8 8 , 391 31 53, 276 Magyars conquer Moravia 3 3 8 ,3 8 9 Leo VI the Wise writes his Taktika 53, 275 Kursan is entrapped by the Bavarians at the River Fischa 345, 347 Constantine VII Porphyrogenitus 53 Regino: Synodalibus causis 11 reign o f the Tang dynasty 8 5 ,8 7 Rurik attacks Byzantium 247 Caliph al-Muktadir 69 death o f Ibn Hordadzbeh 6 8 , 296 Abdul-Rahman III, ruler o f Cordova 91 reign o f Igor 62, 247 Jayhani sets out to write his work 7 1 ,2 9 6 reign o f Constantine VII, Porphyrogenitus 46, 53 R egino’s death 57 Khitai Liao dynasty 210. 2 5 3 ,3 7 5 Romanus 1 Lacapenus converts the Jews by force 232 forays led by Tarhos and Bogat 346 Arabic envoys set o ff to the Volga Bulghars 69 formation o f Czech Kingdom 382 Ibn Fadlan’s travels 121,220, 2 2 3 ,2 3 5 ,2 8 9 one group o f Pechenegs still lives in its old territories 235, 331 Ibn Fadlan at the court o f the ruler o f the Volga Bulghars 224, 291, 294 letter written to Dado, Bishop o f Verdun 58, 282, 407, 423 death o f the chronicler al-Tabari 72
509
Chronological index 924 924 924-925 926 circa 927 circa 930 932 932 932 934 937 circa 940 941 942 943 943 943-973 944 944 945 945 circa 945 circa 9 4 5 -9 5 0 947 949 mid-10th century second half o f 10th century second half o f 10th century around 950 circa 950 circa 950 around 952 circa 952 circa 955 955 955-960 circa 956 957 960 960-976 after 960 961-972 around 960 before 962 969
N aum ’s record o f his life forays led by Salard N icholas M isticus refers to the Magyars as ‘West Turks’ Ekkehard’s note on an individual named Ungar Byzantine court sends Gabriel to the Magyars report o f Ibn Rusta on the ‘Majghars’ Masudi claims the Alani were Christians until 932 Alani expel their bishops the ‘al-wanda’.riyyat and the Burghars lead a joint military campaign against Byzantium death o f al-Balhi al-Jayhani is succeeded by his son as minister Sepher Yosippon is completed in Italy death o f al-Jayhani’s son Magyar forays into Spain report o f Abu D u laf’s diplomatic trip Jews flee to Khazaria Ibn Haukal journeys the entire Islamic world death o f Saint Vazul exile o f Romanus Lacapenus earliest annals o f the Tang dynasty are completed death o f Igor Pechenegs live in and to the east o f the Etelkoz Porphyrogenitus completes his work Magyar raids in Italy led by Taksony Liudprand visits Constantine in Byzantium as an envoy population o f the Carpathian Basin is around half a million the ethnic name Ungroi appears in Greek sources
61 346 282 285 346 295 202 202 291 7 0 ,2 8 9 69, 296 91, 286 6 9 ,7 0 73,381 70 243 71 53 53 89 247 2 3 7 ,2 3 8 54,201 346 57 353 287
crucifix appears in Magyar graves 368 Bulcsii holds the title gyu la 347 Porphyrogenitus mentions the Lenzen tribe 242 Termecsii and Bulcsii visit the Byzantine court 116, 345 final version o f O n P eoples and On P rovinces 49 m issionary bishops are sent to H ungary fro m B yzantium 368 Taksony heads the leading tribe 346 death o f Bulcsii, Csaba, Lei and Sur in the battle on Lech Plain 346, 354 letter o f Hasday ibn Shaprut calls the M agyars‘Hungrin’ 287 death o f Masudi 73,201,290,435 Igor’s widow converts to Christianity 247 Song dynasty reunites China 8 5 ,8 7 formation o f Denmark 382 formation o f Poland 382 Kiev ruler Svyatoslav defeats the Khazars 247 letter o f the Khazar Khaghan Joseph Cyrillic Danube Bulghar inscription
253 228
510 972 972 973 976 9 8 2 -9 8 3 983 986 988 988 992 circa 1000 1001 11 th century 11 th century 1 1th century 11 th century
Appendices death o f Liudprand, Bishop o f Cremona Svyatoslav is murdered by the Pechenegs on Byzantine request al-Biruni is born in Khwarezm al-Jayhani’s grandson becom es minister H udtid al-alam is com piled 71, 74, formation o f Kievan Rus formation o f Norway Vladimir, ruler o f Kiev, is baptised Ibn al-Nadim reports on the library o f al-Mamun Leon D iaconus refers to the Magyars as Huns and Scythians death o f Mukaddasi the coronation o f Saint Stephen 5, 382, chronicle written for the kings o f the House o f Arpad Old Russian primary chronicle is com piled 62, al-Bakri
57 247 73, 202 69, 296 202, 295 382 382 62, 247 224 53 71 385, 406 58 2 8 6 ,4 3 4 71
Ibn H ayyan’s work 340 Slavisation o f the M eshchers and Muromas begins 183 Magyar sources containing personal names 59 11 th century Russian chronicles refer to the Kam a-Ural region as Yugria 435 1 1th century 11 th century iron smelting appears in the Upper Tisza region 357 before 1002 Deed o f gift to the nuns o f Veszpr 6m Valley 353 11 th—12 th centuries further traces o f Eastern European runiform script 441 appear on bricks o f churches 11 th—12 th centuries Szekelys guard the Magyar borderland 443 Zyryan-Russian and Volga Bulghar connections 11th—13th centuries 180 1025-1081 Lampert o f Hersfeld 57 beginnings o f the Russian annals 62 1037-1039 74, 295 1050-1052 Gardizi’s M ost E xcellent R eports Deed o f Foundation o f Tihany Abbey 108, 277 1055 1058 Lampert o f Hersfeld in Hungary 57, 425 1060 N ew Tang annals are com pleted 89 1060 Andrew I dies 425 1060-1063 Bela I 415 1063 the Seljuk Toghril Khan died 256 1063 Otto, Duke o f Bavaria, helps Salamon ascend to the throne 57 1063-1074 reign o f Salamon 4 1 5 ,4 2 5 between 1067 and 1075 Greek part o f the Hungarian royal crown is made 277 1069 Turkic Kutadghu B ilig ‘W isdom o f Royal G lory’ 135,256 1070s Magyar primary chronicle is committed to writing 338 Frank King Henry IV visits Hersfeld monastery 57, 425 1071 1071 story o f the giving away o f Attila’s sword is mentioned 407 circa 1074 Khashghari’s dictionary 1 3 5 ,2 2 5 ,3 6 6 1074-1077 King Geza I 277 circa 1075 Ibn H ayyan’s summary o f the sources concerning the Magyar raids on Spain 277, 408 1076 death o f Ibn Hayyan 73 1077-1079 annals o f Lampert o f Hersfeld 57, 425 1 1th century
511
Chronological index 1080 1080 1086 1094 1095-1116 12th century 12 th century 12th century 12 th century 12 th century 12 th century 12 th century 12 th century 12th century 12th century 12th century 1112
1116 circa 1120 1122
1125 1127 ИЗО 1130-1140 circa 1136-1138 1146 1150 1150-1153 1153 1155 1162 circa 1165 1168 1170 circa 1170 1182 circa 1189 1189, end May 1196,23rd April 1199 end 12 th century 13th century 13th century 13th century 13th century 13th century
new Armenian kingdom is established in Cilicia Abul-Hamid al-Gamati is bom in Granada personal name ‘M oghurdi’ preserved in a charter death o f al-Bakri reign o f King Coloman Yugrians live on the western side o f the Urals Byzantine source translates asszony as ‘princess’ Judah ben Barzillai quotes from the letter o f Joseph Volga Bulghars refer to the Magyars as Bashkirs ethnic name ‘Magyar’ is mentioned in Slavonic sources o f Greek origin Ahmad Tusi, Najib Hamadani manuscript o f the M ethodius legend first heyday o f the Georgian kingdom Cinnamus recorded the word urum , meaning ‘heir to the throne’ variant o f the ethnic name ‘M agyer’ is still in use first mention o f the Szekelys Kharakhitai bring Khirghiz hegem ony to an end in the steppe description o f the battle on the River Olshava M arvazi’s work al-Gamati arrives in Baghdad Kharakhitai ruler becom es gurkhan Prince A lm os died Hebrew epitaph in Austria personal name ‘M oghurdi’ occurs in the copy o f a charter o f 1086 chronicle o f Galfred o f Monmouth description o f the battle on the River Lajta al-Gamati leaves the Volga Bulghar capital Abul Hamid al-Gamati visits Hungary al-Gamati leaves his son behind in Hungary al-Gamati writes his travel book in Baghdad al-Gamati writes his second book death o f Idrisi Conquest o f Ireland, the early Irish-English chronicle death o f al-Gamati in Damascus Benjamin o f Tudela’s travel book Shah o f Khwarezm reports on the Yugur-Zadagan P antheon by Gottfried o f Viterbo Frederick 1 Barbarossa arrives in Hungary death o f Bdla 111 Michael o f Syria com pletes his chronicle chronicle o f Anonym us Votyaks under the influence o f the Khazan Tatars first Finnish linguistic records first Estonian sporadic linguistic records M ongolian loan word borrowing in the Volga region Bashkirs arrive in the Belaya region
76 291 298 7 1 ,2 9 5 3 3 8 ,3 3 9 179 328 91 293 298 70 6 1 ,2 8 6 79 355 303 443 80 443 71 291 375 226 287 298 426 443 291 72, 408 292 292 292 72 426 72, 292 202
435 423 423 59 75, 230 5 9 ,4 1 4 ,4 2 7 180 183 184 181 434
512 13th 13 th 13th 13th
Appendices century century century century
1210
1213 1217 1221-1222
1225 1229 1235-1236, winter 1235-1236 1237 1239 circa 1240 1241-42 1245 1246 1 247-1252 1247 second half o f 13th century after 13th century 1259-1326 1260 1268 after 1280 1281-1360 1282-1285 1282-1285 1238 1286 14th century 14th century 1311 1311 1311, 9th December circa 1315 1318 1321 1331 1333 1340s
beginnings o f Cherem is-Chuvash interaction Estonia is occupied by German and Danish armies beginnings o f the M ongolian Empire manuscript copy o f Ibn H ayyan’s work in Rabat Szekelys participate in the military campaigns o f the count o f Szeben Hungarian word hadnagy ‘army com mander’ crops up in the sources the ‘register’ o f Varad mentions the Szdkely-Saxon alliance campaigns o f the M ongolian commander Siibotei first mention o f the place name M ogorsciget death o f Yakut, the geographical writer 72, Friar Julianus meets the eastern Magyars 1 2 0 ,3 0 2 , M ongolian raids on the Volga Magyars in the Volga region 180, M ongols destroy the Volga Bulgharia M ongols destroy the capital o f the Alani first version o f the Secret H istory o f the M ongols is completed 290, 300, Tatar (M ongol) invasion o f Hungary 90, second settling o f the Cumans into Hungary G uyiik’s letter to Pope Innocent IV Plano Carpini’s work Plano Carpini delivers a letter from Giiyiik written in Persian
181 184 272 408
Magyar personal names among the slaves sold in the Crimea Yugrians migrated east Osman, founder o f the new empire al-Juzjani com pletes his historical work, while Juwayni is still working on his own 74, Thomas o f Spoleto died M ongolian Yuan dynasty rules China Volga Bulghar epitaphs 113, K ezai’s chronicle 58, first mention o f the Szekely runiform script in de Keza 440. death o f Juwayni death o f Bar Hebraeus the centre o f the Yugrians is in the Ob region Bishop Stephen o f Perm com m issions an alphabet Rashid ud-Din mentioned Majar, Bashgird and Kelar in his work Volga Bulghar inscription with the ethnic name ‘M agyar’ 120,301, death o f Majar Redzhep’s son Ismail earliest occurrence o f the Magyar word k ip ‘picture’ death o f the Persian chronicler Rashid ud-din Abul Fida reports on the city o f Mazhar death o f Abul Fida first occurrence o f the place name Tokaj ethnic name Ar mentioned on a Volga Bulghar inscription
302 179 272
443 356 443 300 298 290 429 298 429 202
415 417 202
300 293 300
300 424 254 226 437 443 74 75 179 180 300 302 301 74 301 72 141 305
513
Chronological index 1347 1358 1368 1379 15th century 15th century 15th century 15th century from 15th century 1431 circa 1450 1469-1489 1473 1482 1483 1483 1484 1488, 29th July before 1495 end 15th century 16th century 16th century 16th century 1515 1551 1560-1562 1593 1598 17th century 17th century 1611 1632 1634 1644-1912 1690 1692 18th century 18th centurv 1722 1736 1786, 2nd February 1791 1799
Danes sell o ff their Estonian territories to the Germans the H ungarian Illum inated C hronicle rule o f the Yuan dynasty is ended by the M ing dynasty in China Russian chronicles mention the land o f the Ar the advance o f the Ottoman Turks copies o f the manuscript o f the Constantine legend artificial transformation o f the Germanic runiform script Bashkirian Yurmati tribe m oves from the Shish and Zey region to the Sterlitamak area ethnic name ‘Magyar’ crops up in the Turkic and Greek sources Sz^kelyderzs inscription Bologna inscription mention o f the city and the princes o f the Ar people B uda Chronicle Felsoszemered inscription earliest Szekely alphabet written on the cover o f the Nikolsburg incunabulum first mention o f the Mocharin people Philipec is elected bishop Ivan I ll’s letter to the Hungarian King Matthias Homorodkaracsonfalva inscription Thur6 c zy ’s chronicle mentions the Szdkely runiform script intensive Russian influence in the middle Volga region the language o f the Ostrogoths dies out first continuous Finnish and Estonian texts Istanbul runiform inscription Khazan is occupied by the Russians Ogier Ghislain de Busbecq’s Crimean Goth glossary Amin Razi’s work Janos T elegdi’s monograph on the Szekely runiform script Estonians under Swedish authority dw ellings o f the Voguls are found as far as the River Kama first edition o f De adm inistrando im perio founding o f Tartu university Tallin printing press reign o f the Manchu Qing dynasty in China Marsigli copies a calendar carved into a wooden stick Witsen claims the Voguls lived around Perm in Bashkiria Eastern Cheremis splinter o ff Teryuhans are Russianised Peter the Great journeys the Volga region a manuscript claims that the Voguls live by the River Chusovaya Jones promulgates his Indo-European linguistic affinity theory French national assem bly on the nation discovery o f the Nagyszentm iklos Treasure
184 58 85, 375 305 301 60, 286 441 431 301 437 437 305 58 437 437 298 437 179 437 437, 440 182 203 183, 184 437 182 203 70 437 184 305 49, 54 184 184 85 437 305 181 183 301 305 171 14 131
INDEX OF PROPER NAMES
The index contains the short titles and the authors o f the sources discussed in the book, as well as notable historical individuals, geographical and political names. We have italicised abbreviated titles. With most Arabian authors we have included the spelling given by the Encyclopaedia o f Islam, in parentheses. Chinese ethnic and tribal names are given in pinyin transliteration. Ethnic and tribal names are given in the Index of Words, Ethnic and Tribal Names. Aachen
Adil see Etil, river
263
Abbasid Caliphate Abbasid dynasty Abdallah
Adrian II, Pope
73, 327
Adriatic Sea
66,9 1
Aeneas
224
Abdul Rahman III, ruler o f Cordova Abraham, Avar Khaghan
91
263
Abu Bakr (Abu Bakr Muhammad ibn Muzaffar ibn M uhtag)
2 2 0 ,2 4 1 ,3 2 6 ,3 6 0 ,3 8 7
426
A eschylus
23
Aethicus (Ethicus) Ister Aetius
Abu D u laf (M is' ar ibn Muhalhil al-Hazragl 70
Abu Hamid al-Garnati (Abu Hamid al(jarnatl; M uham m ad ibn ‘A b d al-
56
207
Afghanistan
64
al-Yanbu‘I Abu Dulaf)
434
64, 70, 146, 212, 330
Agapius (Mahbub ibn Kustantin) Agathias Agathon A gizel
52 434
Rahman ibn Sulayman al-M azini al-
Ahmad Tusi (Ahmad TQsT)
Qaysl)
Ak Izel
72, 165, 290, 291, 292, 293,
Abul Fida (Ism a‘ il ibn 'A ll ibn Mahmud
Alania
200
2 0 1 ,2 3 7
ibn 'AyyOb, al-Malik al-Mu’ayyad 'Imad
Alanian kingdom
al-DIn, 'A b u ’l-Fida)
Alatir, river
72,3 0 1
Abul Gazi (A b u ’l-Ghazi Bahadur Han; 'A rab M uham m ad Khan) (Bahadur Khan)
Aldebro
2 7 5 ,2 9 0
Acsteszer Adiabene
1 9 5 ,1 9 7
2 0 1 ,2 0 2
182
Albania, Caucasian Albula, river
Achaemenid dynasty
Aldw in
70
434
Alani Empire
3 0 5 ,3 1 3 , 408, 409
73
51
79, 354
283
154 5 6 ,2 4 4 6 , 75
357
Alexander I the Great
231
A lexa n d er the G reat Syrian L egends
75
516
Appendices
Alexandria
298, 303, 340, 341, 344, 348, 366, 409,
45, 64, 291
A lfbld (Great Hungarian Plain)
141, 155,
158, 264 A li
415, 423, 424, 425, 426, 427, 430 A nsila
208
Antapodosis
6 4 ,6 6
57
Almatai
217
Aquileia 207, 242
A lm ich
225
Ar (Ar), land o f
Alm il
225
Ar Hoja
A lm is
225
Arabian Empire
A lm ish ibn Shelkey
Almos
224, 225
305 6 7 ,2 4 6
Arabian Peninsula
59, 225, 330, 344, 345, 370, 380,
387,
305
389, 416, 417, 418, 420, 423, 426
Ardabil
64
66 , 197, 202, 234, 256
Aral, Lake
230
Almos-Arp&d clan
Ardashir 1, Persian ruler
Almos dynasty
Arethas 53, 276
380
226
Arm enia
Alm uch
226
Armenian Empire
Alm ush
226
Armenian Kingdom
A lm outs
Alp Elteber
226
A lpine Slavs
242 19, 20, 8 1 ,2 1 4 , 281
A m erigo Vespucci
273
Arpdd era
5 9 ,1 3 4 ,1 5 9 ,1 6 4
Arp4d, House o f
78, 215 5 7 ,4 0 7 ,4 2 5
Arsaces 1 Arsk
46 71, 291 5 7 ,2 0 8 ,4 0 7 ,
304
Ascanius Ashina
425
51
A sia 344
Annales gemmeticenses
2 8 0 ,2 8 1 1 9 ,1 4 3 ,1 6 6 ,2 0 6 ,2 5 1
Askold
344
Annales iuvavenses maximi Annales sangallenses maiores
344
Annals o f Salzburg s e e Annales iuva venses maximi Anonym us, Hungarian chronicler
23, 36
247
Aspandiat 56, 347
305
426
A sia Minor 285
xvi, 58, 329, 339
78
Aru, country o f
Andrew I, king o f Hungary
Annales heremi
55, 5 9 ,2 2 8 ,2 5 5 ,3 4 5 ,3 5 2 ,
353, 424
216
Annales alamcmnici
420, 423
Arp&d dynasty
Anastasia, Hungarian Queen
Anekdota
59, 115, 116, 133, 226, 276, 277,
4 1 7 ,4 1 8 ,
19
Ananias o f Shirak
Andalusia
Arpiid
355, 367, 370, 380, 3 87, 389, 409, 416,
197, 214, 295
Amur-Usuri relict
A natolia
76
330, 335, 339, 344, 345, 346, 347, 351,
Am in (Ahm et) Razi 70, 367
Anakopia
77
389
3 4 ,2 7 2
Amu Darya
76, 78, 201
A rnulf 246, 260, 261, 331, 336, 337, 369,
Altai M ountains America
63
Asparukh
151 62, 215, 2 1 6 ,2 1 9 , 227
AsszonyfOlde
328
Asszonyndpe
328
Atelkouzou (see also EtelkOz) xv, xvi,
422
5 6 ,5 8 ,5 9 ,1 1 5 ,1 6 4 ,2 2 6 ,2 7 5 ,2 9 3 ,2 9 4 ,
Athens
6
238, 288,
517
Index o f proper names al-Athir
282
Atil, river Attila
al-Balhi (A bu Zayd Ahmad ibn Sahl al-
237, 287, 295
Balhl)
9, 10, 51, 55, 57, 59, 61, 62, 164,
203, 207, 209, 309, 407, 4 2 3 ,4 2 4 ,4 2 5 , 426, 427
7 0 ,7 1 ,2 8 9 ,2 9 0 ,2 9 3
Balkan M ountains Balkans
45, 46, 52, 123, 203, 240, 241,
257,
262, 286, 322, 331, 376, 377, 388
Aufi (M ubammad-i ‘Awfi)
71
Balkhash, lake
Augustus, Roman emperor
47
Baltic Sea
Austria
2 4 3 ,2 6 2 , 265, 337
Avar Empire
Bandur
56, 115, 1 2 3 ,1 2 4 ,1 2 6 ,1 3 3 ,
158, 215, 220
204
207
49
Baranya county
116 ,4 4 3
211, 220, 241, 262, 263, 284, 309, 329,
Barbarossa, Frederick, emperor
380, 382
Barents Sea
Avar era
159
Aventinus Avesta
BarHebraeus (Gregorios Barhebraeus, Abu’l-
3 4 4 ,3 4 6
Farag Yaybya ibn Hartin)
303
Avitohol
Bari
6 1 ,4 2 6
Azerbaijan
Barsilia 209, 216,
328
B&cs-Kiskun county, Hungary
Bactros, river
154
212
66 , 67, 6 8 , 69, 70, 71, 73, 224,
291, 292 Bahman Yast Baikal, lake Bakath
Basil, Saint
7 6 ,7 9
377
Bavaria
5 7 ,2 4 6 ,2 6 2
Bayan, Khaghan
262
433
Bekash
63
53
Bautzen
Вёкйв
Bagratid dynasty
181, 293, 3 0 6 ,3 1 3 , 399, 430
Basil 1, Emperor o f Byzantium 258,259,260
Bactria 1 9 6 ,2 0 0 ,2 1 2
Baghdad
235
Bashkiria 23
141
237
Bashjird
Babylonian Empire
69
200
Bart6 k, B61a Baruh
171
75
259
Bars al-Saqlabi
201
Azov, Sea o f (see a lso M aeotis)
Babel
423
179
433
Вёкёв county, Hungary
8 1 ,4 0 4
Bekker
2 1 2 ,2 4 8
127, 154
49
Bdla I, king o f Hungary
415
al-Bakri (Abu 'U bayd al-Bakri, ‘Abd A l
B£la III, king o f Hungary
xv, 59, 423
lah ibn 'A bd al-‘ A ziz ibn Muljammad
B61a IV, king o f Hungary
202, 328
b .A yyu b )
Belar
Baksh
7 1 ,7 3 ,1 6 5 ,2 7 7 ,2 9 5
Belaya, river
293
al-Baladhuri (Ahmad ibn Yabya ibn 6 abir ibn D aw ud al-Bala
220, 224, 230, 282
Balasaghun
256
Balaton, lake Bald Hill
293
141
1 1 6 ,1 3 3 ,2 7 7
72, 282
Belgrade Belorussia
1 8 2 ,3 9 8 ,4 3 4
1 1 5 ,3 4 6 246
Benedict IV, Pope 261 Benedictus Polonus
290
Benjamin o f Tudela
202
Benvenuto
123
Appendices
518 Berengar I, king o f Italy
137, 261, 336,
Berengar II, king o f Italy Bersilia
57, 137, 336
Beshbalik
Boris, Bolghar ruler
Borsod-Abauj-Zemplen county, Hungary 137, 154
433
Bilge Khaghan
Bosphorus
148
B ilge K haghan inscription Billa
Boyki
81
al-Biruni (A b u ’l Raybamni Mubammed ibn Abmad al- Blrunl)
Black Adil
73, 202
Brutos, river 237 426
Bucharest-Tei
211
118
B uda Chronicle
2 1 6 ,2 1 7
Black Khitai see Kharakhitai
Buda
Black Sea
Budapest
20, 36, 45, 46, 56, 111, 203,
220, 237, 278, 295, 296, 320 Blatnica
3 9 6 ,4 4 3
Britain 426
Brutus
134
Black Bulgharia
3 3 6 ,3 3 7 ,3 8 9
Bratislava
115,293
Biyelo-Brdo
237
377
Braslav
293
Bilyar
126
58
423 1 1 9 ,1 3 4 ,1 5 4
Budavar
423
Buddha
147
Bleda
207
Bug, river
B ocsa
219
Buga
Bogat
2 1 6 ,2 1 7 ,2 3 7
346
B ughut inscription
201
B ogou see Bug, river
Biigii Kungfuci
Bogii Khagan
Bukhara
Bohem ia Bokli
252
437
374
Bologna
437
64, 80, 279, 405
364
66 , 69, 70, 295
Bulan
232
Bular
115,293
Bulcsu
337, 406, 437
B ologna R uniform Inscription
54, 276, 345, 346, 347, 368, 415,
419
B olshie Tarhani 121, 122, 222, 248, 400
Bulghar-Byzantine War
B olshie Tigani
Bulghar Empire, Danube
120, 121, 122, 123, 167,
53 62, 112, 113,
200, 242, 379
400, 429 Bonifatius, Pope 261 B ook o f G o ld F ields a n d P recious Stone M ines
73
60, 41, 228, 244,
2 4 5 ,2 5 9
221
Bihar county, Hungary 443 Bikesh
63
B o o k o f W arning a n d C orrection
76, 79, 203, 426
Bidenga
B ook o f R ulers
74
252 368
B ible,
67, 6 8 , 296
B ook o f the E m bellishm ent o f the R eports
2 2 4 ,2 2 9
Bezded
B o o k o f R outes a n d K in g d o m s (K itab alm asaliq val-m am aliq)
3 3 7 ,3 8 9
72
112, 219, 220,
230, 326, 387 Bulghar Empire, Volga
73
B o o k o f R oger
Bulghar Empire, Great
B ulghar regal list
113
224, 226
519
Index o f proper names Bulghar, town Bulgharia 258,
388,
115, 293
60, 227, 228, 237, 245, 246, 259, 2 6 0 ,3 2 2 ,3 3 1 ,3 8 7
Bulgharios
389, 390, 391, 397, 400, 403, 405,
4 1 3 ,4 2 9 , 435, 441 Carpathian Mountains
19, 105, 118, 319,
3 3 5 ,3 5 3 ,3 6 3 ,3 6 7
2 0 0 ,2 2 9
Bulyar
115
Carpathians see Carpathian Mountains
Bumin
214
Carthage
B undahishn Buqaogul
63
66
Caspian Lake 2 8 ,1 1 9 , 195,196, 197,200, 2 0 1 ,2 0 2 , 2 3 1 ,2 5 7 , 2 8 2 ,3 7 5
149
Burgar see Bulghar
Cassiodorus
Burgenland
Cathay
243
Busbecq, Ogier Ghislain de Butaul zoapan
203
131
Buy la inscription
Caucasian Albania
79, 354
Caucasian Mountains
127
2 0 1 ,3 6 7 ,4 4 3
20, 36, 45, 6 6 , 75, 78, 79, 80,
Caucasus
Buyla zoapan 131
117, 119, 132, 151, 1 7 1 ,2 0 0 , 2 0 1 ,2 0 2 , 46, 48, 6 8 , 80, 200,
Byzantine Empire
206, 209, 212, 213, 216, 217, 220, 224, 229, 230, 261, 288, 316, 321, 322, 349,
2 2 7 ,3 1 0 ,3 2 5 ,3 2 7 Byzantium
5 1 ,5 5
253
11, 46, 52, 57, 64, 66 , 68 , 74,
91, 206, 207, 209, 212, 214, 215, 227, 228, 230, 231, 232, 235, 240, 241, 244, 246, 247, 258, 259, 260, 262, 263, 277,
367,
398, 420 xvii, 20, 63, 6 8 , 84, 86 , 87,
Central Asia
90, 204, 224, 228, 230, 236, 252, 272, 2 8 2 ,3 5 4 ,
4 0 5 ,4 1 0
282, 291, 296, 327, 328, 331, 332, 346,
Central Empire (see also China)
3 6 8 ,3 7 8 ,3 7 9 ,3 9 0 , 391
Central Eurasia
Central Ural region
20
Central Volga region Caesar (Julius Caesar) Cairo
228
7 3 ,9 1 ,2 8 9
Campus Mauriacus Carinthia
90 207
1 2 6 ,2 6 4
246, 2 6 0 ,3 3 1 ,3 4 4
Carolingian Empire Carpathian Basin
56
4, 1 3 ,2 0 ,2 9 ,4 1 ,4 5 ,5 7 ,
19
216
417
Charles I the Great 263,
56, 57, 124, 243, 251,
273, 284, 285, 336, 353, 377, 424
Charles II the Bald
Carloman (Karlmann), Frankish ruler 243, 244,
Ch'adar Bulghar Chagatai
Caliphate o f Cordova
85
1 7 ,1 8
Charles III the Fat
260 246, 260, 286
Charles Martel, M erovingian ruler
66,
230 Chen dynasty
89
105, 114, 115, 117, 118, 119, 120, 123,
Cherdaklin
125, 132, 134, 138, 141, 155, 156, 159,
Cherdyn district
167, 220, 242, 249, 253, 254, 256, 261,
Cheremshan, river
262, 263, 265, 266, 270, 284, 287, 293,
Chichek
305 179 122, 181, 222, 431
230
300, 301, 302, 310, 322, 324, 326, 327,
Chi dynasty
329, 331, 332, 336, 337, 338, 339, 348,
China
89
8 ,1 0 ,2 0 , 29, 82, 84, 85, 86 , 87, 8 8 ,
349, 352, 353, 354, 357, 360, 363, 364,
89, 90, 203, 204, 205, 214, 252, 253,
370, 373, 374, 376, 381, 382, 383, 387,
2 5 5 ,3 7 4 , 375
Appendices
520 Chinghis Khan Chistopol
9, 1 0 ,2 0 8 ,2 2 9 , 253, 272
Chronicle o f the Popes Chronicon pictum Chronographia Chu, river Chusdal
Chussal
Crete
52
258
3 0 2 ,3 2 7 , 329, 367 Crimean Peninsula 233, Csaba
344
46, 203, 206, 230,
286, 295, 327 346
Csongrad county, Hungary
344
Cilicia, province o f
Ctesiphon
76
Curzan
355
C leisthenes o f Athens
Cyril
23
C lovis, M erovingian ruler
1 3 7 ,2 6 6
64
344 60, 164, 233, 244, 245, 249, 258,
260, 2 8 6 ,2 9 6 , 313, 329, 3 6 7 ,4 0 8 ,4 4 0
376
Cyril, Legend o f 60, 286
371
C olom an 1 Beauclerc, king o f Hungary 338, 339 Columbus
57
Crimea 2 0 1 ,2 0 3 ,2 3 1 ,2 3 2 ,2 3 3 ,2 4 8 ,2 9 6 ,
344
Codex Cumanicus
51
56
377
Cremona
179 ,3 0 5
Cinnamus
Cosmographia Cottus
53
xv
344
Chussala Chussol
344
204
Chusovaya
90
Cosm as Indicopleustes
3 0 1 ,3 0 2 ,3 1 2 ,3 1 3 ,4 2 9
Chronicle o f George the Monk
Cordova Caliphate
Czech Kingdom
382
Czech Republic
242
275
Confucius see K ong Fuzi Constantine (Cyril) see Cyril
Dacia
Constantine Legend see Cyril Legend
Dado, Bishop o f Verdun
Constantine VII Porphyrogenitus
46, 51,
158
53, 54, 55, 57, 212, 216, 230, 276, 288,
Dalmatia
345
Dam ascus
Constantinople
53, 61, 116, 227, 230,
1 2 3 ,2 4 2 ,2 4 4 ,4 1 4 64, 66 , 72
Danapris see Dnieper
243, 244, 245, 258, 259, 260, 262, 263,
Danastris see Dniester
286,
Danube
329, 376, 388
58, 164, 282,
3 1 3 ,4 2 3
46, 134, 154, 200, 207, 2 1 5 ,2 1 6 ,
Constantinus (Cyril) see Cyril
227, 229, 238, 240, 242, 262, 285, 286,
Continuation o f the Chronicle by George
287, 295, 296, 326, 328, 331, 332, 335,
the Monk see Georgius monachus con tinuatus
3 3 7 ,3 3 8 , 346, 352, 3 8 7 ,3 8 8 , 3 8 9 ,4 1 3 Danube Bulghar Empire see Bulghar Em
Conversio bagoariorum et carantanorum,
pire, Danube
(Conversion o f the Bavarians and Car-
Danube-Tisza, land between
antans)
Dar al-Babunaj
56, 1 16,264, 266
Conversion o f the Bavarians and Caran-
D ar-iA lan
tans, see Conversio Bagoariorum et Ca
Darial, Pass
rantanorum
Darius
Cordova
7 3 ,9 0
Dashuran
232
2 0 1 ,2 0 2 2 0 1 ,2 0 2
195 79
12 6 ,1 5 4
521
Index o f proper names
Dashuranci see M ovses Dashuranci
Dunaszerdahely
De adm inistrando im perio (G overning o f
Dunhuang
em pire)
45, 46, 49, 54, 164, 237, 239,
154
8 1 ,8 2 ,8 3
Dvina, river
1 7 8 ,1 7 9
276, 3 4 0 ,3 4 5 ,4 1 6 De cerim oniis aulae B yzantinae
54, 419
De origine actibusque G etarum
55
D eed o f F oundation o f Tihany A b b ey
Early Han Dynasty 108,
D eed o f g ift to the N uns o f Veszprem Val ley
Demeter, Saint
Dentumoger
D 6veny
East Crimea
132 327 xvii, 6 8 , 72, 78, 110, 112,
113, 120, 1 3 9 ,2 1 5 , 2 5 1 ,2 5 2 ,3 2 7 ,3 7 8 ,
241
379,
382
Derbend
103, 192
East Europe
353
Denmark
East A sia
East Bulgharia
277, 278
386, 403
East European steppe
341
East Han dynasty
2 0 1 ,2 8 2
East Hungary
244
Diakonus, Leon
89
346
East Turkestan
53
158, 206, 208
85
81, 196
Dietmar see Theotmar
E aster Chronicle
52
Dimaski (Sam s a!-DIn Abu 'A bd Allah
Eastern Anatolia
187
Muhammad ibn Abl Talib al-Dima5qi)
Ecilburg
420
Edayab
231
Edessa
7 4 ,7 5
al-Dinavari (A bu Hanlfa Ahmad ibn Daw ud al-Dmavvarl) Dir
Edil
72
Eger
247
Dnieper (Danapris), river
201, 203, 216,
423
434 154
Egypt
2 9 ,1 9 5 ,3 0 2
217, 219, 220, 237, 238, 295, 322, 326,
Ekkehard
327,
Elba, river
328, 3 8 7 ,4 1 3
Dniester (Danastris), river 237, Dobruja
203, 216, 2 17,
Elista
285 2 4 2 ,2 6 1 ,3 7 7
132
322, 362, 387
England
132
Enns, river
Don, river
114, 132, 155, 196, 203, 207,
209, 216, 220, 231, 238, 295, 327, 328, 387 Donets, river
D on-Volga elbow
Drava, river
202 221
263
Ermanrich, king o f the Goths
181, 203,
Esmail ibn Ahmad, Samanid ruler
31,
330 Esperuh see Asparukh
243, 264
Estonian Republic
243
184,185
Etel kai Kouzou see Etelkoz
Druthmar o f Aquitain D uch’i Bulghar
232
Etel see also Etil Etelkoz
216
Dula, Alanian ruler Dulo dynasty
263, 337, 338, 389
Erik, prince o f Friaul 206
202, 322, 328
D onets-D on region
Drava region
91
227, 328
62, 227
xxi, 40, 117, 123, 156, 217, 219,
238, 239, 242, 257, 286, 288, 322, 324, 325, 326, 327, 329, 330, 331, 334, 338,
522
Appendices
Etelkoz (continued) 339, 356, 363, 364,
Gabriel, cleric
346
387, 388, 390, 391, 413, 414, 419, 421,
Galanta
422
Galfred o f M onmouth
Etelkiizii see Etelkoz
154
Gandhara
426
228 8 , 80, 252
Ethicus se e Aethicus
Gan-su, province
Ethiopia
Gardizi (A bu Sa'Td 'A bd al-H ayy ibn al-
231
Etil, river 2 1 1 ,2 2 0 ,2 3 5 ,3 2 8 ,4 1 3 ,4 2 9 ,4 3 4
Dahhaq ibn M ahmud Gardizi)
Eurasia
150, 165
xv, 16, 17, 22, 82, 146, 153, 227,
2 5 1 ,3 8 8 , 4 0 1 ,4 0 4 Eurasian steppe Europe
ai-Gamati see Abu Hamid al-Garnati
147
Gaznavid
1 9 ,2 3 ,3 4 ,4 0 ,5 8 ,6 4 ,9 0 ,9 1 ,1 0 3 ,
Genah
7 3 ,7 4
3 5 0 ,3 5 1 ,4 3 2
160, 203, 205, 206, 2 0 7 ,2 0 8 , 210, 213,
Geobitsha
214, 230, 251, 285, 320, 373, 375, 377,
Geographus Ravennas
56
378, 380, 382, 386, 396, 427
Geography, A rm en ia n
78, 216
E xcerptum de karentanis E xordia S cythica
56
56
E xtract (Muqtabas) E zekiel, B ook o f
73
282
G eorgia
306
79
G eorgian C hronicle
79, 282
Georgius M onachus
54
G eorgius m on a ch u s contin u a tu s Georgius Pisides
Palicsi
G erm ania
344
Farkaszug
Pejtir county, Hungary Fergana
81
Filipec
437
Finland
437
183 ,1 8 5
263, 345 259, 260, 261
9 0 ,9 1 ,2 3 1 ,3 7 6 57, 124, 243, 259, 260,
228 118,119
6 6 ,2 3 0
G oem agog
426
75, 272, 2 8 2 ,2 8 3 ,4 2 3
Golden (Jin) dynasty
F u ld a A nnals
Gormot
375
431 214
430
Gottfried o f Viterbo
423
G overning o f em pire see D e adm inistrando
57 200
Gibraltar
215, 277
273, 354
Golden Mountain
261
From Kesar
Fulin
Gdza I, king o f Hungary
Golden Horde
2 6 1 ,3 3 2
Fulda
135
G etica see D e origine
G og
263
Frankish Empire
Frumu$ica
229
G iza , prince
245
Formosus, Pope
Friaui
G eser epic
gyars
Fischa river
Frankfurt
229
Gesztered
183
Forchheim
Ge-sar
7 ,2 3 1 ,2 7 4
G esta U ngarorum see H isto ry o f the M a
Finland, G u lf o f
France
154
55, 334
52
184
Germany
137
Felsoszem ered
71, 74,
5 7 ,3 3 5
im perio G ozw in
245
523
index o f proper navies
Granada
2 9 1 ,4 0 8
Hellas
257
Great Hungary see Magna Hungaria
Henry IV, German king
Great Moravia
Hephthalite Empire
249
Great Wall o f China Greece
85
Heraclius, Byzantine emperor
257
230,
Greek-Roman Empire Groze$ti
46
118
Giirgench
260
Het<5ny
293
Gurkhuman
293
Gyarmat
3 4 0 ,3 5 0 ,3 5 1 ,3 7 0
H eves county, Hungary
3 5 0 ,3 5 1 ,3 8 1 ,3 9 8 ,3 9 9 ,4 3 0 ,
277
154
217
Hidmas, river
4 3 1 ,4 3 2 Gyeicha
293
Hibil, river 300
G yeovicha
4 5 ,5 5 ,5 6 ,1 9 6 ,2 1 1 5 7 ,4 2 5
H£tmagyar
430
GUyiik khan
14
Hersfeld
234
Gur Kermen
Gurmot
Herder
4 1 6 ,4 1 8
Hierotheus
368
Himalayas
82
Hingilus, river 416, 418
277
Hippie Mountains
216
Gyermat
429
GyeUcha
277
Historia Ottonis 57 History o f Albania 79
GyeQcsa
2 7 3 ,3 0 6 ,3 5 5
History o f Armenia
78
History o f Husros I
353
Gyevicha
Gyula
277
51
History o f Procopius
277, 398
History o f Saints and Martyrs History o f the Magyars Hadrian II, Pope Hagar
245, 259, 260
23
Han dynasty
85, 89
Hangai Mountains
421
Hoirosphactes, Leon
39, 114
Homer
Hara siv
39
1lom okm egy-H alom
Harun al-Rashid, caliph Harun ibn Yahya
Hasday ibn Shaprut
al-Hazari
66 , 232
68
9 1 ,2 0 2 , 2 8 7 ,3 1 3
187 150
6 9 ,7 0
Hazari see al-Hazari Hazarig
229
Hebraeus see Bar Hebraeus
Hrabar
81
1 8 9 ,3 0 9
Homor 6 dkardcsonfalva Hovratu
72
334
Hoitu Tamir inscription
10
Hara siu
Hazarhakhan
76
5 8 ,5 9 ,4 1 5 ,4 1 8 ,
History o f the Muslim Conquests History o f the World 2 10
287
Hammurapi
Hatti Empire
52, 219,
240, 262, 286, 377
Herodotus
Guido III, Pope
57, 425
199
133, 405, 441, 442 437
217 see also Khuvrat 440
Hua countiy
248
Huangdi, emperor o f China Huang-he, river
Hudud al-alam
23
85 71, 74, 164, 202, 215,
225, 234, 235, 238, 248, 249, 257, 295, 433
524
Appendices
H ugo o f Provence Hun
IbnR usta(A bu 'A ll Ahmad ibn 'U m aribn
137
6 4 ,2 0 5 ,2 1 1 ,2 4 8
Hungaer Hungar
225, 232, 233, 295, 343, 370, 406
287
Hungary
68 , 71, 72, 74, 165, 200, 202,
Rusta)
285
Idel
7, 15, 34, 54, 55, 58, 72, 90, 91,
105, 111, 119, 120, 122, 127, 133, 137,
434
al-Idrisi (Abu ‘ Abd Allah Muhammad alIdrisI)
72, 165
140, 141, 161, 162, 1 6 3 ,2 0 2 ,2 3 4 , 273,
Igfon, forest o f
287,
Ignatius, patriarch
259, 260
347, 350, 355, 360, 368, 377, 382, 405,
Igor, ruler o f Kiev
6 2 ,2 4 7
407, 4 0 8 ,4 1 7 ,4 1 8 , 424, 425, 441
Ihe A shete inscription
Hunila
292, 298, 300, 302, 308, 345, 346,
208 51, 205, 207, 208
Ik, river
2 7 2 ,3 2 8
H vaday nam ag Hyonite tribe
214
63
Iluku-Burkan Illyria
195
355
India
282
Indian Ocean Iberia
340
360 66
Indus valley
Iberian Peninsula
7, 71, 73, 90, 230, 376
Ibn Fadlan (Aljmad ibn Fadlan ibn al-‘ A b bas ibn RaSid ibn Hammad)
39, 68 ,
Ingul, river
118 ,2 1 7
Ingulets, river Inner A sia
217
20, 278, 280
69, 70, 106, 113, 121, 147, 148, 150,
Innocent IV, Pope
165, 220, 223, 224, 225, 226, 232, 233,
Insar, river
234, 235, 248, 289, 291, 294, 313, 366,
Iran
418
Ireland
Ibn al-Fakih (Abu Bakr Abmad ibn M u
426
Irene, empress o f Byzantium Irgiz, river
madani)
Irnik
Ibn Hauqal (A b u ’ l-Q asim ibn 'A ll alNasibl, ibn Havvqal)
7 0 ,7 1 ,2 8 9 ,3 5 6 ,
357
Hussayn ibn Hayyan)
73, 165, 277,
3 4 0 ,3 8 1
235
Iron Gate
374
Irtysh region
435 234
Isaac, Avar ruler Ishtemi
263
2 1 4 ,2 1 5
Isidorus o f Sevilla
Ibn Hordadzbeh (Abu ’ 1-Qasim‘ Ubayd A l
xvi
Ismail ibn Ahmad see Esmail ibn Ahmad
lah ibn ‘Abd Allah ibn Hurradadbih)
Ismail, son o f Majar Rejep
6 8 ,7 1 ,7 4 , 2 9 6 ,3 0 7
Israel, Albanian bishop
Ibn al-Nadim (Ibn al-Nadim A b u’ I Farag
301
231, 369
al-Istahri (Abu Ishaq Ibrahim ibn Muham
MQbammed ibn A bl Ya‘ qub Ishaq al-
mad al-Istahri al-Farisi)
Warraq al-Bagdadi)
3 0 4 ,3 5 5
224
230
6 2 ,4 2 6
Irtysh, river
Ibn Hayyan (Abu Hayyan ibn H alaf ibn
300
182
330
hammad ibn Ishaq ibn al-Faqlh al-Ha6 8 ,2 3 2
xv, 58
415
283
Inancs
210
82
5
Illum inated Chronicle
Husrau Anoshirvan, Persian emperor
Hwkurithou
81
Ihe H oshutil inscription
Hunnish Empire Hunor
366
71, 148, 289,
525
Index o f proper names Istanbul Istria
437
Julianus, friar
283
Justine II, emperor o f Byzantium
Istros, river
346
Justinian I, emperor o f Byzantium
Istvan I, Saint, king o f Hungary see Stephen, Saint Italy
346,
3 6 9 ,3 7 6 ,3 9 1 ,4 1 4
Ivan III, czar
Jutocsa
Juwayni (Cjuwaym, ‘ A lu ’ al-Din ‘ Ata-
al-Guzganl)
4 2 3 ,4 2 6
Jarmat
431
Jarmati
154
Kalankatuyk
Japhet
Kali
Kalocsa
68
(jarmi)
133, 405, 441, 442
Kama, river
201
317, 362, 386, 298, 400, 429, 434, 435
Muljammad ibn Aljmad)
'A b d A llah 69, 70, 71,
Kharakhum Karashar
72, 74, 151, 165, 295, 296, 307, 327,
Kara su
328,
Kara sub
329, 330, 343, 344, 353, 356, 364 66
Jebel al-Tarik
64
Jiakann Mountains
39
Karos
3 1 ,1 3 5
Kartli
79
K artlis chovreba
216
K enezlo
Johannes Malalas John VIII, Pope John IX, Pope
273
Ker
51
Kerch
2 4 5 ,2 6 0
Kerj
5 6 ,2 6 1 ,3 6 9
5 1 ,5 5 ,1 8 1 ,1 8 2 , 229, 425
Joseph, Khazar Khaghan
80
Judah ben Barzillai
91, 202, 215,
3 5 0 ,3 5 1 ,3 8 3 ,3 9 8 2 9 5 ,3 2 7 ,3 2 9 ,3 5 6 295
K ese Tabin K eszi
432
3 5 0 ,3 5 1 ,3 8 3 ,3 9 8 ,4 3 2 178
deK eza, Simon 91
175
135, 137
Ket, river
233, 2 8 7 ,3 1 3
79
Keltaminar culture
8 2 ,2 5 4 ,3 7 5
417
Jordanes
192 39
Karlmann see Carloman
430
Jin dynasty
235
Karl (Charles, the Great)
295
Jerusalem
28, 94, 111, 120, 121, 122,
139, 179, 1 8 0 ,2 1 1 ,2 2 2 , 289, 3 0 1 ,3 0 5 ,
Jayhani (al-Gayhanl, A b u
Jayhun
79
347
Kalman see Coloman I Beauclerc
431
al-Jarmi (M uslim ibn A bi M u slim al-
Juanser
74
180
224
Jochi
290, 300
'Utm an ibn Sirag al-DIn Muhammad
Kal
Jeno
230
345
al-Juzjani (M inhag al-Dm ‘ Abu ‘ Umar
180
Izhevsk
Jarrah
11,48,
51,211
M alik ibn Muhammad)
179
434
Izh, river
Jafar
262
Justinian II, emperor o f Byzantium
20, 57, 90, 91, 123, 203, 207, 220,
259 260, 261, 259, 283, 286, 336, 337,
Izel
120, 182, 293, 302, 429
xv, 58, 437, 443
de K e za ’s Chronicle
xv, 406, 424
526
Appendices
Kinel, river
66
Khabul
Kharabalghasun inscription
64, 81, 149,
2 2 1 ,2 9 1
Kisalfdld
1 5 4 ,2 6 6 ,3 3 8 ,3 8 9
Kitab al-masaliq val-mamaliq see Book of
168 Kharakhanid Empire Kharakhitai
Routes and Kingdoms
256
2 5 6 ,3 7 5 201
Khasahia
Khasar 228, 252, 272 Khashghar
256
K hashghari (MabmOd al-Ka5gari
106,
Kitab al-tanbih
291
Klyazm a, river
183
K nossos
23
Koblenz
57
Koli chor inscription
135, 225, 2 3 5 ,2 4 9 , 256, 290, 365, 366,
KOI Tegin
371
Kol Tegin inscription
Khass Hajib
256
K olo
Khazakhstan
2 0 ,8 1 ,4 0 3
Komarom
433
Kondoros
433
Khazakh steppe Khazan
210
1 8 0 ,2 0 8 ,3 0 4 ,4 0 3
Khazar Empire
39, 69, 70, 9 1 ,1 1 2 , 166,
Kondorosh
433
K ong Fuzi (Confucius) K opaszhegy
141
296, 322, 327, 328, 331, 3 4 7 ,3 4 9 ,3 5 2 ,
Kophis, river
216
3 6 4 ,3 6 6 , 3 6 7 ,3 7 8 ,3 8 0 , 390, 398, 441
Koran se e Q ur'an
Khazar Sea see Black Sea
Korax
216
Khazaria
2 3 2 ,2 3 3 ,2 3 7 ,2 3 8 ,4 1 6
Korea
204
Khazarig
201
KOrfls
2 6 6 ,3 4 6
Kherson
201, 203, 230, 23 7, 295, 327
Koszorudiilo
Khitai
8 1 ,2 5 3
248, 253, 254, 255, 256, 375
Khitai dynasty
210
Kotragos Kotsel
137
1 6 7 ,2 4 4 ,2 4 5
Krak’s learn
216
252
Krems
58, 164
Khotan
6 4 ,1 9 6 ,2 7 9
Krilos
118
366
Khuvrat 6 2 ,1 1 2 ,1 2 3 ,2 1 5 ,2 1 6 ,2 1 7 ,2 2 6 ,
Krum, Bulghar ruler Kuar
Kuban region
387
Kuban, river 68 , 69, 70, 73, 196, 215, 234,
257,
301, 435
al-Khwarezmi (Mubammad ibn M usa al68
Huwarizml) K iev
6 2 ,9 1 ,1 1 9 ,1 5 2 ,2 3 1 ,2 4 2 ,2 4 6 ,2 4 7 ,
286, 2 8 9 ,2 9 1 ,2 9 3 ,3 1 3 , 359 Kievan Rus
378, 382
K iev Principality Kilyna, river
221
249
227, 242, 258, 263
151
227, 230, 248, 262, 285, 322, 326, 348,
Khwarezm
23
220
K hocho
Khutadghu bilig
81, 148, 166, 297
214
203, 220, 2 2 2 ,2 3 1 , 233, 2 5 7 ,2 8 2 ,2 8 9 ,
Khirghizia
81
83
326, 387 111, 132, 215, 216, 217,
2 1 9 ,3 2 2 Kubu (Koubou), river Kucha
192
Kudama
6 8 ,7 1
Kudimkar
179
Kuma, river
301
Kunb&bony
219
Kundajik
328
297
Kundurcha, river
2 9 1 ,4 3 3
527
Index o f p rop e r names Kuphis, river
2 1 6 ,2 1 7
Kup’ iB ulghar Kur, river
420
Kurayish
64
Kurt
tium
185
335, 344, 345, 346, 347, 370, 389
Lerida
277
Levedi
238, 2 8 8 ,3 2 5 , 326, 3 3 0 ,3 4 9 ,3 8 7 ,
389,
415, 416, 417, 419, 420, 421
Levedi, house o f
3 5 0 ,3 5 1 ,3 8 1 ,4 3 0
Kiirtgyarmat
46, 53, 164, 258, 260, 275, 276,
309, 331
Kurland Peninsula Kursan
L eo VI the W ise, em peror o f B yzan
216
Levedia
305, 350, 351
123, 324, 325, 326, 338, 387,
Kusal
344, 345, 389
Kusan
344, 345
Levond
Kiisel
345
Lex salica
Ktisen
345, 389
Liang dynasty
Kushnarenkovo Kussal
4 1 6 ,4 1 9 ,
121, 122
344
Kutadgu bilig Kuver
376
Liao dynasty
85, 87, 253, 254, 255, 375
262
Lithuania
9 1 ,2 3 2
237
Liudprand
Kuybishev
121
Liuntika
Ladoga, lake Lajta
185
Late Han dynasty Latium
272
Latvia
185
122
57
330 137
Louis II, ruler o f Italy
259
Louis III, the Infant, East Frankish ruler 3 1 ,2 4 6 , 3 3 7 ,3 8 9
89
58, 164, 2 4 3 ,2 4 4 ,2 4 6 ,
260, 285, 3 3 1 ,3 4 4 Lower Danube
2 6 2 ,3 1 6 ,3 2 9 ,3 3 1 ,3 3 4 ,
3 3 5 ,3 4 4 ,3 4 5 ,3 8 7 , 3 8 8 ,4 1 3
377
Lower Oka
74
Legend o f M ethodius
183
Lower Pannonia
57, 346, 354
Legend o f C onstantine
161
Lomovato
Louis o f Germany
Lechfeld, battle o f
Lehel
2 6 1 ,3 3 6
259
Lebanon
123
Lothar II, 57, 164, 259, 425
Lambert, North Italian ruler
Lausitz
Lombardy
Lotharingia
443
Lateran
57 330, 335, 345
Liupold o f Merseburg
413
Lambert o f Hersfeld
89 254
Kuvu, river
Kiizii see Etelkoz
421
78
Liao Empire
Linz
135
2 4 1 ,2 6 3
3 2 9 ,3 3 0 ,3 9 1
Loyang
60
Luitpold
61
56
85 338
367 3 4 6 ,3 6 7
Lena, river
M acedonia
81
Leo I, Pope
Madara
207
Leo III, emperor o f Byzantium
230
Leo IV “the Khazar”, emperor o f Byzan tium
230
M aeotis
258, 309
331 209, 216, 282, 283, 291, 328,
387 Magari
298
528
Magash 202,433 Magery 298 Magna Bulgaria see Bulghar Empire, Great 180 Magna Hungaria 52, 180, 293,401 Magog 78, 272, 282,283, 423, 426 Magor 272,328 Magyarorszag 355 Mahbub ibn Kustantin see Agapius Mahmudkuli 433 Mainz 425 Majar 301,306 Majar Rejep 301 Majarkay 301 Majartai 301 Makarach 297 Makdisi see al-Muqaddasi Mala Pereshchepino 217 Malamir 227,228,243 Malta 258 al-Mamun(al-Ma ’тип ibnHarun al-RaSkl), caliph 224 Mani 63 Maniakh 214 Maodun, Xiongnu ruler 204 Maqdisi see Muqaddasi Mar Jakob ben R. Hannukah 289 Marcianus, Roman emperor 207 Margaret, Saint 328 Marinus I, Pope 260 Marinus of Tyros 45 Maros, river 126,242,266,346,366,368, 388 Marseille 90 Marsigli 437 Martianus II, Pope 260 Martyrdom o f Abo 79 Marvan 201,202,220,229 Marvazi (Saraf al-Zaman Tahir Marwazi) 71,165,434 Mary, Virgin 264, 328
Appendices
Mashad 70, 165 Master P. 59 al-Masudi (al-Mas' udi, Abu *1-Hasan' AIT ibn al-Husayn) 73,106,149,168,201, 202, 215, 232, 233, 290, 291, 420, 433, 435 Matthias, king of Hungaiy 179,406,437, 440 Mattsee 58,285 Matzaron 298 Maurice, emperor of Byzantium 52, 53, 200,276 Mayak 132 Mayna, river 122,222 Mecca 64,292 Mede Empire 199 Medina 64,296 Mediterranean Sea 408 Megyer 255,306,351,381,383,389 Melas, river 211 Menander Protector 11, 40, 51, 52,211, 213,214, 282 Merovingian dynasty 56,376 Mesopotamia 20,22,195 Metev, river 431 Methodius 60, 61, 164, 233, 244, 245, 246, 249, 258, 260, 286, 296, 313, 408 Methodius, Legend o f 61, 286 Meursius 49,54 Mezen 179 Mezoseg 124 Michael I, emperor of Byzantium 227 Michael III, emperor of Byzantium 258 Michael of Syria 75, 200, 201, 229,325 MikulCice 126 Mikulovo 437 Mindszent 137 Ming dynasty 85, 375 Miracles o f Saint Demeter see Saint Deme ter's Miracles Miracles o f Saint George
53
Index o f proper names Mitrovica
529 Muchar
245
Mochar
2 9 8 ,3 0 6
Modena
49
Muhammad
M ogorsciget Mogurdi
1 0 ,6 4 , 72, 256, 296
lah Mubammad ibn Ahmad ibn Abi
298
Moksha, river
Bakr al-Muqaddasi) al-Muqtabas (M uqtabis)
182
71 7 3 ,1 6 5
al-M uqtadir (al-M uqtadir b i-'lla h ), c a
119
M oldovan-Podolan loess ridge Mon Kermen (see also K iev) Mongolia
214
al-Muqaddasi (Sam s al-Din A bu 'A bd A l
298
Mohamed see Muhammad
Moldavia
2 9 8 ,3 0 6
Mugan khaghan
liph
158
142, 143, 210, 213, 253, 361,
362 Mongolian Empire
221 Muruj adz-dzahab M ycenae
7, 10, 272, 302
Mongolian Republic
69
Murom castle (M uromskoe gorodishche)
293
291
189
81, 162, 360
Monomakh see Vladimir Monomakh Morava, river Moravia
N&gman
377
6 0 ,6 1 , 244, 249, 3 3 1 ,3 3 2 , 336,
132, 147, 167, 264, 266, 370, 404, 405,
338, 378 Moravian Empire Mordia
441
246, 332, 336, 389
Nagytarcsa
279
Morocco
432
Nagyszentm ikl 6 s 1 2 6 ,1 2 7 ,1 2 8 ,1 2 9 ,1 3 1 ,
146
154
Najib Hamadani
70
Mosaburg see Zalavar
Ndndorfejdrvir
Moses o f Horene see M ovses Horenaci
Narim, river
Moses, five books o f 232
Nasr ibn Ahmad, Samanid ruler
Moson
443
Naum
Mosul
231
Neanderthal
Mouageris
Movses Dashuranci
79, 151, 231, 366
Movses Horenaci (M oses o f Horene)
78,
Moyin chor inscription
Nehavend
64
Nekim en
432
81
Moymir II, ruler o f Moravia 337
Nestor, monk
79
Moymir I, ruler o f Moravia
Mozhar
16
432
Nestor Chronicle
215 Movses Kalankatvaci
62
62
Nibelungenlied see Song o f the Nibelungs 243, 244 246, 336,
Nicephorus
5 2 ,2 1 6 ,2 1 7 ,2 2 0
Nicephorus I, emperor o f Byzantium Nicephorus Phocas
2 9 8 ,3 0 6
N icetas Scleros
298
N icholas I, Pope
Mozharka Mozharovo
298
Mroveli, Leonti Mu’awiya
64
Muayeris
297
Nikolsburg
259
331, 344 244, 259
N icholas M isticus 79
6 9 ,3 3 0
61
N egm en
297
284
178
5 3 ,2 8 2
437
Nikon, igumen o f K iev N ikopsis, river
201
62
227
530
Appendices
N isibis Nitra
O ng inscription
74
Ong khan
1 5 5 ,2 4 3 ,2 4 4 ,2 4 6 ,2 6 6
Nizhni N ovgorod N oah
246
81
272
Onggiit
272
Oniega, river
9
North Afghanistan North Africa
North Greece North Italy
Onoghur Bulghar Empire
74
146
North Caucasus
232
56
drhalom
168
82, 297, 362
Oroszpatak
208
Orseg
81
Northeast Hungary
168
Northern Chi dynasty
89
Northern Mountain Range (Hungary) Northern Zhou dynasty
154
443
Osman
272
O ssetia
201
Ostmark
57
Oswald, bishop o f Pannonia
8 9 ,1 6 8
Otto I, Bavarian prince
382 180 ,2 4 7
Otto I, Saxon ruler
Noyon-ula
208
Ottoman
1 6 6 .3 5 1 ,3 8 1 ,4 3 2
Nyi'regyhaza
Otiiken
10,3 7 4
Oxford
91
10, 91
197
1 7 ,1 7 8 ,1 7 9 ,3 1 7 ,3 2 0
Obadiya, Khazar ruler Obuda
57
Ottoman Turkish Empire
273
244
57, 407, 425
16 2 ,2 7 2
O xos, river Ob, river
82, 408
278
Novgorod
N yek
123
O rkhon Turk inscriptions
23, 259, 262, 336, 337
North Tocharistan
Norway
Onogoria
Orkhon river
23
North M ongolia
185
232
423
Odoaker, Germanic chieftain
376
56, 116, 124, 126, 166, 203,
207,
227, 243, 244, 245, 249, 264, 283,
285, 309, 332, 344, 388, 424, 437
O fo see Ufa OgOdei, M ongolian ruler Oka, river
Pannonia
Pannonian Downland
417
1 8 2 ,1 8 3 ,2 4 6
P annonian L egends
Okhtay khan see Ogodei
P antheon
423
O ld Testam ent
Papagaia
201
282, 423
O ld Tibetan A n n a ls
84
O ld Tibetan C hronicle Old Tibetan Empire O leg, ruler o f Kiev Olhontor Bulkar O lom ouc
Paris
Partholon
247
Passau
2 1 5 ,2 1 6
56, 244. 245, 264, 338, 388, 389
Paul, Saint
Omurtagh, Bulghar ruler
376
Pecheneg country
443
Omar I, Arabian caliph
199
426
Patsinakia see Pecheneg country
437
Olshava, river
49
Parthian Empire
84 82, 83
155 60
64
Pechora, river
227, 228, 242
O ne T housand a n d One N ights
67
Peking (Beijing) Peloponnisos
201, 234, 237
79
257
88
531
Index o f proper names Penjikent
Pribina
252
Peoples, O n
1 3 3 ,1 6 7 ,2 4 3 ,2 4 4 ,2 6 4
Priskos the Rhetor
49
Pepin the Short, Frankish ruler
56
229,
Perm
180 ,3 0 5
Probota
Persia
52, 64, 74, 80, 195, 196, 197, 202,
Procopius
118 5 1 ,2 4 0
206, 210, 214, 230, 231, 232, 238, 262,
Provence
2 8 8 ,3 2 5 , 3 7 6 ,4 1 9 , 4 2 1 ,4 2 2
Provinces, On
Persian (Presiam), Danube Bolghar ruler
Prut
227 Persian Empire
Petronas
237 118,132 74
Pseudo M ovses Horenaci Pseudo Zakarias
Pugur
78, 225
75
45, 67, 68 , 78, 215, 16
Ptolemy
185 215
Phanagoria
229
232
Philippicus Barnades Philotheus
230 Qing dynasty
53
Photius, patriarch Phrom Ge-sar Piasts
301
231
Petrozavodsk Peuke limne
49
5 7 ,3 3 0
Pseudo Ibn Muqqafa
154
Peter I the Great, Russian czar
260
Przemysl
1 9 5 ,1 9 6 ,1 9 7
Pest county, Hungary
258, 259, 260
228
85
al-Quds (Jerusalem) Q u’ran
71
301
378
Piliny Pisa
Priim
5 1 ,7 5 , 207, 208, 209,
248
2 1 6 ,3 6 6 56
Piwki
Raba, river
215
Plano Carpini Pliska
Rabat 290, 293, 300
6 6 ,2 3 0
Poland
Pontus steppe
272
90, 92
122, 135
Ras Tarkhan
9 1 ,2 3 2 ,3 7 8 ,3 8 2
Polo, Marco
Radanite house o f trade Rakamaz
227
Poitiers
264
7 3 ,4 0 8
200, 201
Rashid ud-DIn
290, 300
Rastislav, Moravian ruler
285
243, 244, 245,
246, 329, 331
Porphyrogenitus (see also Constantine VII Porphyrogenitus)
54, 201, 202, 215,
216, 217, 235, 237, 238, 242, 288, 289,
Ratbod, Frankish marquis Ravenna
243
4 6 ,5 6 ,2 2 0 ,2 6 1 ,3 2 6
Regensburg
58, 244, 245
296, 298, 305, 306, 326, 330, 334, 335,
Regino
340, 346, 348, 350, 370, 377, 380, 406,
R egions o f the W orld see H u d u d al-alam
4 1 4 ,4 1 8 ,4 1 9 ,
R elatio de legatione C onstantinopolitana
Porto
420, 430
259, 260
P o ve st' vrem ennyh let Pozsony
56
Presiam see Persian Preslav
61,3 3 1
1 1 ,5 7 ,2 8 3 ,3 3 0 ,3 3 2 ,4 2 0 ,4 2 1
57 62, 165
R epaym ent see A ntapodosis R eport on the route o f Salam the In ter p re te r
68
R esponses to the B urghar kin g
224
532
Appendices
Rhine, river
7 ,9 0 ,1 1 9 ,2 0 7 ,3 7 7
Riccardus, frater
Rogerius II, king o f Sicily Roman Empire
72
1 3 ,4 7 ,9 0 ,2 0 3 ,2 0 6 ,2 0 7 ,
23
60, 228, 244, 245, 246, 258, 259, 2 6 1 ,2 7 2 ,2 9 1 ,3 3 6 ,3 7 6 , 377
Rostov
183 183
14
Roxana
2 4 1 ,3 7 8
Sarmatia
262
216
368
Sasanian dynasty
63
Sasanian Persia see Persia
Satumia
Ruanruan Empire
211
Sava, river
69
332
Rum Sea see Black Sea
Saxonia
Rurik, ruler o f K iev
Sayan Mountains
Rus Khaghanate
247
246
377
Rus see Kiev
Schechter Text
Russia
Sclavinia
Russian Primary Chronicle Ryazan
62
246
177, 178, 318
11 9
Scandinavia
117, 1 2 2 ,1 8 2 , 237, 246, 285
256
283
Sausan al-Rasi
290
46, 5 2 ,6 4 ,1 9 7 ,
206 Satuk Bughra Khan
195
Rubruk
152
232
Sasanian Persian Empire
Rostov-Suzdal Principality Rousseau
Samkars
Sarolt
5 3 ,5 4 ,2 3 2 ,2 6 1
260,
71
Sankt Florian
Romanus Lacapenus, emperor o f Byzan
Rom e
2 2 1 ,2 3 1 ,2 9 1
Sambat (K iev)
Samo
375, 376, 397 Roman Republic
tium
Samara
Samarkand
429
91
264
Scythia
4 2 3 ,4 2 7
Sebeos
78
Secret History o f the Mongols 290, Sabatai
287
Sahname Saint Basil
293, 3 0 0 ,4 1 5 ,4 1 7
Selenga, river
63 53
Saint Demeter
52
Saint Gallen Abbey Saint Petersburg
52, 241
285
196
Sepher Yosippon
91, 286, 287, 313
237
Seretos, river Sevastopol
185
237 203
Sevrei sumun inscription
Saksin
2 9 1 ,2 9 2
Shang dynasty
Salaho
243
Sh apu rll
Salamon, king o f Hungary Salard
5 7 ,4 1 5 ,4 2 5
346 1 3 9 ,1 4 0 ,2 0 2 , 328
56, 242, 244, 245, 263, 264,
347, 369, 388 Samandar
81
8 5 ,8 6
197
Sharkel, Fortress at
114, 201, 216, 231,
296, 324, 328, 406
Saltovo-M ayak Salzburg
234, 362
Seleucid dynasty Seret, river
Saint Demeter's Miracles
51, 273,
Shishm e, river
2 2 0 ,2 3 0
Samanid dynasty
Shine usu inscription Shiqi 89
69, 330
Siberia 319
81
2 2 2 ,4 3 1
19, 20, 121, 132, 212, 288, 317,
533
Index o f proper names Sicily
72
Strabon
Sigibert I, Frankish ruler Siktivkar
179
Silk Route
Silvester, abbot
Subbotici
62
Stibotei
Silzibulos, Turk ruler
214
53, 61, 258, 331,
335, 440
Simson Yehudah Singul, river
Sur
216
81
71
346
Sura, river Suzdal
240, 242, 245, 262, 266, 346
Sisola, river
118
8 5 ,2 8 1
Suj. inscription Sukrullah
289
124
Sirmium
300
Sui dynasty
Simeon, Bulghar ruler
Si 6 , river
122
118,119
Sudova Vishnya
89
334,
52
Strazsadomb
2 9 ,1 9 2 ,2 1 4
Sima Qian
2 0 0 ,2 0 2 ,2 1 1 ,2 1 6
Strategicon
261
182
183
Svatopluk 1, ruler o f M oravia
Slavonia
242
Slovakia
1 5 9 ,2 4 2 ,2 7 8 ,4 3 7
Sogdiana
196, 200, 212
Svatopolk, ruler o f Kiev
Sok, river
291
Sviyaga, river
Svatopluk II, ruler o f Moravia
Solitudo avarorum Solt
264
Sw eden
South A sia
Syria
South Siberia
2 5 3 ,3 2 0
South Transylvania
Spain
gary Szakony 89
3 6 6 ,3 7 1
Szarvas
7 0 ,1 1 7 ,1 8 4
266, 366
127, 130, 131, 132, 167, 264,
370, 404, 405, 4 4 1 ,4 4 2
7, 6 6 , 73, 91, 92, 165, 203, 277,
Szeben
443
Szekelyderzs
2 8 7 ,3 4 0 ,3 8 1 ,4 0 8 Spoleto
154
Szam os, river
89
Southern Chou dynasty Soviet Union
1 9 7 ,2 1 2
6 4 ,2 1 7 ,2 5 8
S z a b o lc s-S z a tm a r-B e r eg cou n ty, H un
335
Southern Chi dynasty
119
Syr Darya
207, 423
274, 370
231, 247
Synodalibus causis 11
162
South Russia
62
221
Svyatoslav, ruler o f K iev
85, 87, 88 , 8 9 ,2 5 4
Song o f the N ibelungs
246, 331,
3 3 6 ,3 8 9
355
Song dynasty
244, 245,
246, 336, 369
179
Szekelvland
1 2 3 ,2 5 9 ,2 6 0 ,2 6 1
Stephen o f Constantinople
Szekesfehcrvar
260
Stephen Saint I, King o f Hungary
437 443
4, 5,
425
Szemered see Upper Szemered
1 3 4 ,2 1 5 , 273, 277, 339, 368, 381, 382, 391 Stephen V (VI), Pope
260
Stephen VI (VII), Pope Stephen, Bishop o f Perm Stephen, Patriarch Sterlitamak
al-Tabari (Abu G a'far Muhammad ibn Ga-
261
331
1 2 1 ,1 2 2 ,4 3 1
180
rir at-Tabarl) Tacitus
184
Tactics
53
Taihir inscription
72, 282
81
534
Appendices
Taimyr Peninsula Taksony
178
Theodora
3 4 6 ,3 5 4
Theodorus II, Pope
6 6 , 82, 166, 204, 230, 234,
Talas, river
2 3 0 ,2 5 8
330
261
Theodorus, Avar ruler Theodorus Sincellus
Talas Valley
Talmud
166
263 52
Theophanes Bizantius
51, 52, 55, 215,
2 1 6 ,2 1 7 ,2 3 2 , 282
232
Tamatarhan
201
Theophilactus Sim ocattes
Tamin ibn Bahr
Tang Annals
365
89
Theotmar (Dietmar), archbishop o f Salz
85, 86 , 88 , 89, 146, 225
Tang dynasty
burg
56, 369
Tangri han see Tengri Khan
Thessaloniki
Tangshu
Thewrewk
89
Tankeyevka
Taraz (Talas) river
52, 244, 258 278
Thom as o f Spalato
1 2 1 ,2 2 2 ,4 0 0
Taraz, town
330
Thrace
2 1 5 ,2 5 8
Three kingdoms
330
424
89
Tarentum
259
Thurk
Tarhanka
222
Thuroczy's chronicle
437
Tianshan Mountains
204
Tarhos
346 66
Tarik Hill Tarim Basin Tarj6 n
Tarkacsu
193
2 6 1 ,2 8 3
Tiberius, East Roman emperor Tibet
Tiflis
Tatar Kalmayur
211
Tim, river
183
178
Tim isis see Temes
178
Tegin al-Turki
69
Timok, river
Telegdi, Janos
437, 440
Tisza, river
Telerig, Bulghar ruler Temes, river Tem eskoz
227
242
Teremhegy
256
Togorma Tokaj
1 5 1 ,3 6 6 116
Terh inscription
242, 264, 377 126, 159, 2 0 7 ,2 6 2 , 332, 336,
352, 388 Toghril
346
Tengri Khan
82
2 0 1 ,2 3 0
Til, river
305
Tatar R epublic
262
8 3 ,8 4 ,2 5 4
Tibetan plateau
345 351
Taz, river
278
Tiber, river
2 5 4 ,3 5 1 ,3 5 2 ,3 8 9
Tarkdny
51, 211, 212,
248
286
141
Tolui, M ongolian prince 8 1 ,2 2 8 ,4 0 5
Tonyukhukh
417
374
Termachu see Termecsii
Tonyukhukh minister's inscription
Termecsii
Toraye, gate
xxi, 54, 116, 167, 276, 345,
TorrMs see Termechii
3 7 1 ,4 1 5 Tervel, Danube Buighar ruler Teutgaer
227
285
Tez II inscription
200
Totila
208
Toulouse 81, 228, 405
Theoderich, Germanic chieftain
207
Toxus (Taksony) 376
346
Traianus, Roman emperor
345
81
535
Index o f proper names
Transbaikal area Transcaucasia
Uighur Empire
20 231
Transdanubia
126, 141, 154, 159, 244,
Transylvania 335,
1 9 ,9 1 ,2 1 7 ,2 1 8 ,2 3 9 ,2 4 9
Ukruh, river
346, 352 3 3 ,1 3 3 ,2 2 5 , 2 4 2 ,2 6 5 ,2 6 6 ,
336, 337, 344, 352, 357, 389, 443
Ulanbator Olio
Troyes
362
305
Umayyad dynasty
207
Trullos, river Tsimlyansk
201
3 5 1 ,3 5 5
Ulyanov
Trier 57
Ungar
237
287
Unguriya
Tuoba Wei dynasty
Upper Szemered
85, 89, 253
2 1 3 ,2 1 4 ,2 1 5 ,2 7 9 ,2 8 0 ,2 8 1
Turk Khaghanate, First
64, 80, 81, 214,
17, 22, 34, 45, 94. 96,
121, 178, 179, 208, 212, 234, 238, 239, 270, 289, 304, 317, 318, 319, 320, 322, 324, 325, 373, 386, 387, 399, 404, 434
255 Turk Khaghanate, Second
80, 81, 252,
2 8 1 ,2 8 2 ,3 5 4 ,3 7 4 al-Turki see Tegin al-Turki 8 , 64, 80, 82, 83, 279
Turkestan
Turkestan Basin Turkey
82
76, 187, 195, 302, 346, 368, 403,
Ural, river
2 0 ,4 5 ,2 3 5 ,2 9 1 ,3 2 0 ,3 2 5
Urals see Ural Mountains Urartu
76
Urgench
234
U sen, river Uzbek
431
272
Uzbek Khan
418 Turkia
2 3 7 ,2 7 6 ,3 4 5 ,3 7 7
Turkish Empire Turkmen
Vag, river
Turku, lake
277
Turuhan, river 278 346
Tver
185
Twrek
237
301 330
Tutis
U zia
272
272
Turkmenia
178
Var
155
2 1 0 ,2 1 1 ,2 1 3
Varad Varkony
4 3 7 ,4 4 3 2 1 3 ,3 5 2
Vasyugan, river Vatican
278
12 1 ,2 8 9
Ufa, river Uhergen
431 432
178
4 9 ,5 5
Venice
337
Verdun
5 8 ,1 6 4 ,2 8 2 ,4 2 3
Verecke, pass Ufa
266, 332, 334, 335,
Ural Mountains
278
Turk Empire
Turuk
437
3 5 7 ,3 8 8
8 0 ,8 2 ,1 9 2 ,2 1 2
Turk
292
Upper Tisza region
278
Turfan
64, 2 2 7 ,2 8 7 , 327
436
Ungriya
231
Tunhuang see Dunhuang
Turcu
64, 81, 82, 252, 253,
354 Ukraine
246, 264 Trans-Tisza
82, 252, 349
Uighur Khaghanate
332, 388
Veroce, county Veszprdm Valley Vidubec
62
134 353
536 Vienna
Appendices
Western Yin dynasty
5 6 ,3 2 9 ,3 3 1 ,3 4 8 ,3 6 0
Vienna Basin
263
Westphalia
Vienna Forest
263
White Lake
Vladikavkaz
W iching
201
Vladimir Monomakh, ruler o f Kiev Vladimir, ruler o f Kiev Vlastimir
62
62, 247
242
Vokil dynasty
Volga Bulghar Empire
232 185
260
William o f Jumiege
246
William the Conqueror
xvii, 426
Wisdom o f R oyal G lory
256
Witsen
62, 227
89
305
W orld C hronicle o f R egino
293
Volga B u lg h a r in sc rip tio n s
106, 167,
57
W orld C hronicle o f G eorgios M onachos 54
2 4 8 ,3 0 1 ,3 0 5 Volga Bulghar sites Volga Bulgharia
W orld H istory C olum n o f M ich a e l o f Syria
1 2 1 ,1 2 2
75
429
Volga, region
173, 180, 182, 212, 402
W orld H istory o f al- Yakubi
Volga, river
28, 90, 94, 111, 115, 117,
Worms
2 0 7 ,2 6 0
120, 122, 132, 181, 182, 183, 206, 209,
Wugur
229
214, 215, 216, 220, 221, 222, 231, 235,
Wuifila
238,
Wulka
291, 298, 301, 302, 304, 306, 319,
72
2 0 3 ,2 0 8 ,3 7 6 243
322, 324, 327, 331, 386, 399, 429, 432 V olga-D on, region
212
Volga-Kama, region
15, 208, 215, 226,
Xiongnu Empire
204
3 8 1 ,3 9 8 , 4 0 1 ,4 0 3 ,4 1 4 , 434 Volga-K am a-Cherem shan, rivers Volga-K am a-Ura) region Vungar
429
285
Vungrorum
68
Yaghlakhar
272
al-Yakubi (al-Y a'qubl, Ahm ad ibn Abi 285
Vyatka, river
Y a'qubi)
180
Vychegda, river
6 8 ,7 2
Yakut (Sihab al-Din A bu 'A bd Allah al-
179
Hamawi Yaqut al-R um l)
70, 71, 72,
148, 165, 290 Yarmati Wand.r
291
431
Yas province (Jaszsag)
Wei dynasty see Tuoba Wei dynasty
Yavdierdim
Wei, river
85
Yaxartes, river
West A sia
2 1 4 ,2 7 2
Yayik see Yeyik, river
West Han dynasty West Hungary West M ongolia
85
362
West Roman Empire West Siberia
Yeleh
237
179
200
3 4 5 ,3 5 1
Yellow River (H uang-he)
325
West Xiongnu Empire
Yazva valley Yazygia
46
3 1 9 ,4 3 6
West Turk Empire
197, 212
Yazikapan, province
243, 346
200
237
Yeloguy, river 204
Yemen
231
178
23, 85
537
Index o f proper names
Yeneh
3 5 1 ,4 3 2
Yurmati
Yeney
4 3 0 ,4 3 2
Yurmi, river
Yenisei, river Yermat
17, 19, 81, 166, 178, 179
4 3 0 ,4 3 1 ,4 3 2 432
Yurmi-az, brook
431
Zakarias, bishop
75
430
Yeyik, river Yighur
235
435
Yiltever, Bulghar king Yin dynasty
85, 89 85, 88 , 254, 375
Yuan dynasty Yuanshi
224, 226
300
Yughur
Zala, river
Yughur-zadaghan
Zalavar (Mosaburg)
Zalta 435
11
3 4 5 ,3 5 5
Zemarkhos Zey, river
214 431
434
Zhiguli, hills
Yugri
434
Zhizhi, Xiongnu ruler
221
Yugria
179, 434, 435, 436
Zhol inscription
Yujiulu
210
Zhou dynasty
Yura
434
Yurmash
237
Zikhia
Zvinitsa
84 8 5 ,1 4 8
201
Zventibald 431
126, 133, 167, 243,
264
Yugra
Yula province
213, 229
243
245,
4 3 5 ,4 3 6
Yugoslavia
Zakarias the Rhetor
244 228
204
INDEX OE WORDS, ETHNIC AND TRIBAL NAMES
This index contains the Hungarian and the foreign words, the designations o f languages, peoples, and tribes, as well as related proper nam es as they are dealt with in the book. Chinese ethnic and tribal nam es are given in pinyin transcription. The m eanings given are only for short reference. A bar A bdel
Alani
213
Abhazians
3 2 0 ,3 2 2 ,3 2 7 ,3 2 8 ,3 7 3 ,3 8 7
200
* a b lu /a p lu ' apple’
acs ‘carpenter’
107 ,3 5 7
agar ‘Hungarian’
Altaic
287
agyafurt ‘ciinning’
American
367 ,3 7 1
*ancha
303
Akkadian
Ancient Greek
2 2 9 ,2 7 3
Ancient Iranian
88
Alan-i-kaz(ar)
Ancient Permian
234
9 6 ,1 1 0 9 7 ,9 8 ,9 9 , 100, 104,
108, 1 7 4 ,3 5 0 ,3 5 6 ,4 1 3
111
111
alacs ‘spotted (o f an animal)’
275
Ancient Hungarian
189, 190
a k o l‘sheep pen’ al ‘false’
192
303
Ancient Chinese
2 2 9 ,2 7 3 ,3 0 9
189
134
*amlu ‘apple’
96, 98, 303
234
Akatsir
226
9 4 ,9 5 ,1 0 1 ,1 0 2 ,1 0 3 ,4 0 2
aluka ‘axe’ or ‘hatchet’
287
agyek ‘groin’
36, 110, 111, 192
alom ‘dream’
229
agyar ‘tusk’
274
alm a ‘apple’
107
107
2 7 4 ,2 7 5
Allem ande
agach eri
A katir
Aleman
434
184
agachchi ‘carpenter’
A jlad
2 3 3 ,3 6 9 ales ‘carpenter’
434
adash ‘namesake’
Agarenus
1 0 4 ,1 5 2 ,3 2 8 ,3 3 8 ,1 4 3
Albanian see also Causasian Albanian
110
a d a k ' leg ’
Alanian
192
abrak ‘fodder’
Aestii
196, 197, 200, 201, 202, 203, 206,
209, 214, 222, 229, 234, 248, 287, 291,
213
110
Ancient Turkic
98, 99 1 4 8 ,3 0 7
angga ‘a kind o f wild duck’
94
540
Appendices
200
A nglo-Saxon
426
as ‘ermine’
anjer ‘tusk’
99
asa g a l
225
dsegel
225
99
A sfali
2 8 8 ,4 1 9 ,4 2 1
1 9 6 ,1 9 7 ,2 0 0
Asian
1 6 2 ,2 0 0 ,2 0 2 ,2 8 0
Ant
240
a n zh a r ‘tusk’ Aors
apaca ‘nun’
368
Asian Avar
2 1 4 ,4 3 5
a pat ‘abbot’
368
Asian Hun
85, 204
A p fe l
192
A sii
apple
192
Assyrian
Aquitani
11
3 0 1 ,3 0 4 ,3 0 5
Ar
3 0 1 ,3 0 4 ,3 0 5
328,
338 Asut
300
a s z o k ' gantry’
96
d r ‘value’
195
asszony ‘lady’, ‘married w om an’
Ar
d r ‘tool’
200
110, 111
aszta g ‘stack (o f c o m )’
96
111
4 5 ,4 6 ,6 3 ,6 4 ,6 6 ,6 7 ,6 8 ,7 3 ,8 2 ,8 4 ,
Austrasii
11
146, 182, 201, 203, 212, 215, 220, 224,
Austrian
287
225, 227, 230, 232, 2 4 0 ,2 5 7 ,2 5 8 ,2 5 9 ,
Avar 13, 14, 30, 51, 52, 56, 75, 93, 112,
Arab
260,
263, 291, 294, 295, 296, 307, 322,
114, 116, 123, 124, 126, 131, 132, 133, 167, 199, 209, 210, 211, 212, 213, 214,
327, 328, 3 5 5 ,3 6 0 , 3 7 7 ,3 7 8 ,4 3 4 ,4 3 5 Arabian
3 0 1 ,3 7 6
216, 219, 235, 240, 241, 242, 251, 253,
Arabic
71, 74, 75, 106, И З, 173, 224,
256, 261, 262, 263, 264, 265, 266, 275,
225, 239, 256, 272, 290, 294, 2 9 7 ,2 9 8 ,
285, 305, 308, 309, 322, 326, 327, 335,
3 0 1 ,3 0 7 ,
352,
Arameic
4 0 7 .4 2 0 ,4 3 2
3 8 8 ,3 8 9 , 390, 391, 404, 414, 436, 441
6 3 ,7 9
ara n y ‘go ld ’
355, 373, 376, 379, 380, 381, 382,
97
a ya k ‘foot’
ara t ‘to harvest’
110
Archaic Chinese
88
arhon ‘ruler, prince’
Az 335, 346, 347, 348,
434
200
azash
434
Azha
253
381 a rk ‘chattels’
280 110
arkany ‘pole lasso’
b.j.n.y
291
46, 7 5 ,7 6 , 7 8 ,7 9 ,8 0 ,1 6 5 ,2 1 5 ,
b.j.nSk
291
216, 225, 282, 288, 303, 367, 407, 419,
b.jg.rd
291
420
bab ‘bean’
Armenian
arok ‘ditch’ a rp a ‘barley’
1 0 7 ,1 1 0 ,3 6 3 110,321
arpa ‘oracle stick’
365
drtany ‘(castrated) hog’ aruva ‘m edicine’ a r v a ‘m edicine’ As
2 0 0 ,3 0 5
365 365
110
111
B achgird
3 0 7 ,4 3 4
Bactrian
6 3 ,1 0 4 ,1 9 5 ,2 0 0
bagatur
178, 228
Bagirsik
213
Baioarii
11
baj ‘grace’ Bajghard
365 201
541
Index o f words, ethnic and tribal names Bajgir
308
Bajgird
307
Bajigit Bajna
8 4 ,2 5 2
Bastard
290
batir ‘hero’
2 9 0 ,3 0 0 201
Bajnak
Basm il
434
bator ‘brave’
201
Bavarian
Balkan Bulghar
Balkan Slav
262, 264, 266, 337, 345,
389
215
Balkan Onoghur
111,178
t o y ‘rich’
285
Bayan
239
115
2 2 0 ,2 6 1
b a lq a 1axe’ 189
b e-cha-nag ‘Pecheneg’
b a lla ' a x e ’
1 8 9 ,3 6 7
*becsu ‘cradle’
1 8 3 ,1 8 4 ,2 4 0
beda ‘poverty’
434
beg ‘chieftain’
233, 355, 374, 433
Baltic
Baltic Finnic
183,185
Baltic-Slav
bekas ‘o f frogs’
239
Balto-Finnic М л ‘title’
barack ‘peach/apricot’
111
111
barat ‘m onk’
110
barom ‘cattle’ 211
Barsilk
b erk ‘hard’ Bersil
225
bersil
224
Bersul
225
barsony ‘velvet’
116 239 110 , 116
111
h e r ‘paym ent’ 111
barazda ‘furrow’
belek ‘stam p’
belyeg ‘brandmark’
368
360
Bersiile
280, 433
224 224
Bascard
290
Besermyan
Bascart
290
betu ‘letter (character)’
Baschart
290
bash ‘head’ Bashghirt Bashghird Bashjirt
290
М и ‘bull’
110,321
birka ‘sheep’ 109, 111
290
bird ‘ju d ge’ Biscart
bivaly ‘bison’
290
biza ‘poverty’
Bashkhort
274
Black Khitai
121, 147, 149, 215, 274, 275,
Blandr
112
290
Bashkatur
Bashkir
111
434 375
234
289, 290, 291, 293, 294, 297, 300, 306,
blandur
307,
bo ‘wealthy, noble’
310, 312, 313, 366, 386, 429, 430,
4 3 1 ,4 3 5 , 436 Baskar Baskatur
112
290
290
290 290
111 , 133, 168,
3 6 5 ,4 4 1 ,4 4 4
bilincs ‘handcuffs’
Bashart Bashkar
298
290, 292, 293, 300
2 8 9 ,2 9 0
110, 321
116
Belorussian
2 2 4 ,2 2 6
ЬагапуЧ ат Ъ ’
Barsel
b eleg ‘stam p’
115,355
Baranjar
433
beklyo ‘fetter’ ‘hobble’
183
235
107
234 112, 355
bocs ‘sage (w ise m an)’ bocsanat ‘pardon’
367
bocsat ‘to pardon’
112
1 0 7 ,3 6 4
542
Appendices
*boesit
Bulgharian
106
b odun ka va rd i ‘brought the peop le to gether’
348
387
boji ‘lent’
112
bun ‘sin ’
Bolar
*burch
300
boles ‘w ise’
boleny ‘bison’ Bolghar 238,
buyruk
buza ‘wheat’
3 6 ,1 1 0 ,1 1 1 ,3 2 1
bors ‘pepper’
247, 248, 257, 258, 259, 262, 275, 276, 277, 278, 281, 282, 283, 287, 293, 295,
307
301, 306, 308, 309, 327, 329, 334, 344, 112
borlon ‘prison’
11, 46, 51, 53, 62, 64. 75, 112,
214, 217, 219, 224, 227, 230, 240, 244, 99
110
borso ‘pea’
1 0 6 ,1 1 0 ,3 2 1
114, 164, 182, 200, 2 0 1 ,2 0 3 ,2 1 1 ,2 1 3 ,
111
borong ‘be sorrowful/cloud over’
374
Byzantine
106, 110, 321
borona ‘harrow’
364
1 2 7 ,1 3 1 ,3 5 5
244, 245, 246
borji'i ‘ca lf’
boszorkany ‘w itch’
boyla
buyla
115
bor ‘w in e ’
boyarin
1 4 9 ,1 6 2 ,4 0 4
biivdl ‘to charm’
351
xviii, 13, 14, 60, 61, 113, 115,
bolyarin
112 ,3 6 7
2 1 3 ,2 9 1
Buryat
106 ,1 0 7
112, 114, 1 8 0 ,2 4 2 ,3 6 8 ,
307
Burghar
107. 112, 364, 365
bolcso ‘cradle’
93, 259, 346
Bulghar Turkic
345, 347, 355, 356, 364, 367, 379, 381, 1 1 2 .3 6 4 ,3 6 5
386, 387, 388, 406, 407, 415, 419, 420
115 2 2 8 ,3 5 5
bii ‘m agic’
Caesar
364
bfibaj ‘charm’
112, 364
biles ‘w ise m an’
364
bucsu ‘farewell’
106, 107, 112, 367
biigii ‘knowledge o f supernatural things’ 364
11 5 ,2 2 8
Carantan
11,2 6 4
Caroiingian
12 6 ,3 8 2
Caucasian Albanian Caucasian Hun eekla ‘beetroot’
biiguchi ‘wizard’ Biikli
374
Bular
115
107. 364
7
Celts
123
e h elya d ‘servant’
Bulghar Esegel
295
Bulghar se e also Danube Bulghar, Volga Bulghar
Celtic
xviii, 52, 68 , 69, 112, 115,
Cheremis 303, Chermis
123, 152, 1 8 2 ,2 0 1 ,2 1 3 ,2 1 5 ,2 1 6 ,2 1 7 ,
Chermish
220, 221, 222, 224, 226, 227, 228, 230,
Chinese
235, 238, 242, 258, 262, 263, 266, 276,
231
2 3 1, 365, 366, 369 111
305, 356
142, 180, 181, 182, 271, 298, 304, 305, 402, 403 181 181 xxi,
8 , 23, 33, 80, 82, 83, 84,
85, 86 , 87, 89, 103, 143, 146, 147, 148,
284, 285, 287, 290, 291, 295, 301, 305,
154, 165, 171, 172, 173, 192, 194,200,
308, 309, 322, 325, 326, 327, 330, 331,
203, 204, 2 1 1 ,2 1 3 , 225, 229, 230, 248,
332, 334, 335, 337, 338, 344, 345, 348,
252, 253, 255, 256, 275, 279, 281, 300,
353, 367, 378, 379, 380, 381, 382, 387,
302, 341, 354, 364, 365, 374, 375, 407,
388, 390, 400, 402, 413, 414, 435, 440
418, 435
543
Index o f words, ethnic and a ib a l names choban
Danube Bulghar
115
Chuvash
102. 103, 104, 105, 106, 107,
108, 109, 113, 114, 115, 116, 142, 152, 167, 180, 181, 182, 189, 194, 2 0 0 ,2 0 9 ,
126,
219, 228, 247, 248, 257, 258, 266,
296, 329, 332, 354, 368, 371, 376, 378, 3 8 0 ,3 8 1 ,3 9 1 ,4 1 4 , 426
211, 224, 225, 226, 228, 229, 239, 247,
Danube Magyar
261,
dara ‘groats’
281, 284, 293, 298, 301, 304, 306,
365, 307, 402, 403, 429, 433, 434 Cimmerians cinq ‘fiv e’
195
Deutsch
cinque ‘fiv e’
111
Digor
347
Crimean Croat
93, 116, 239, 241, 377, 378, 382
csaldd ‘fam ily’ Csango
3 0 5 ,3 5 6
csaszdr ‘emperor’
115,273
112
cseled ‘servant’ csep ‘thresher’
305, 356
1 0 0 ,1 1 1 ,3 2 1
diot ‘people’
352
Dirmir
213
disznd ‘pig’
107 ,1 1 0
cseresznye ‘cherry’ csiger ‘a bad w ine’ csikos ‘horseherd’
csodor ‘stallion’
355
donga ‘stave’
D ru-gu
280
diis ‘abundant’ 111
Dutch
110 , 111
dux
357
355
dvoryanin
115
110
csosz ‘field-guard’ cstitdrtok ‘Thursday’
111 112
152
East Frankish
East Iranian
152,281
172
East Magyar
Cuman
7, 105, 202, 274, 293, 302, 308,
East Slav
3 0 9 ,3 1 0 ,3 7 3 ,3 7 9 115,273 1 0 9 ,2 3 9 ,3 3 6 ,3 7 7 ,3 7 8 ,3 9 6
285
East Goths see Ostrogoths
cuckoo
Czech
355
9 8 ,3 0 9
1 11
csoroszlya ‘coulter’
czar
93
94 2 2 9 ,280
D ur-gyis
110
м е р и ‘thresher’
csorda ‘cattle’
111
did ‘walnut’
Dravidian
111
421
246
East Turkic
277, 2 8 2 ,3 2 5 ,3 4 1 , 3 4 7 ,3 5 4 ,
440 Eftaliyt
213
eg er ‘m ouse’ Dacian dada Danes Danish
310 173 426 184
304
104
doge ‘title’
33
csata ‘battle’
237 2 8 5 ,3 5 2
dinnye ‘m elon’
52
338
Croatian
377
deutscher M ensch
93
cirok ‘broomcorn’ cow ari
293
110
Derevlyane Dervlenin
93
61, 62, 107, 113, 115,
egy ‘one’
35
366
egyhaz ‘church’ ehsine ‘lady’
112, 366
328
ek- ‘to throw /sow ’
107
544
Appendices
eke ‘plough’
1 0 7 ,1 0 9 , 110
Finno-Ugrian
22, 34, 35, 37, 38, 94, 96,
elb uvol ‘to charm’
364
97, 100, 113, 122, 139, 166, 167, 173,
e lig ‘ruler’, ‘king’
355
174, 178, 181, 182, 183, 184, 189, 194,
elteber ‘title’
226, 347
246, 303, 304, 305, 306, 317, 318, 319,
em ‘w om an’
304
321, 324, 357, 364, 365, 383, 386, 399,
em b er ‘man’ ‘person’ Enets
274,
33, 37, 38, 63, 172, 192, 194,
fiv e
f o ‘head’
Englishman Ephthalite
3 0 4 ,3 0 5 ,3 1 1 ,3 1 9 ,3 8 6
93 100
304
Frank
56, 58, 243, 244, 245, 259, 262,
346,
376, 377, 382, 388, 389
French
e rk ‘strength’
е г о ‘strength’
112
f i i n f ‘fiv e ’
182,183
2 2 4 ,2 2 5 ,2 2 6 ,2 9 5
Estonian
93
281
Erzya Mordvin
112
eskii ‘oath’
7, 14, 37, 63, 281, 280, 283, 284,
304, 350
280
erkolcs ‘m orals’
100
f o g ‘tooth’
263, 266, 271, 304, 329, 331, 332, 338,
102, 304
erdem ‘merit’ . 112, 304, 367
Esegel
173
f o g - ‘to catch’
304 7 8 ,2 1 3 ,2 1 4 ,3 0 9
er ‘m an’
*erj
402, 4 0 3 ,4 1 3 fu l ‘b oy’
283, 303, 304, 340, 376, 434
English
Er
304
178
5 5 ,1 8 4 , 185
esztergalyos ‘turner’ European Avar eziist ‘silver’
357
2 1 3 ,2 1 4 98
g a cser ‘drake’
111
g a ra t ‘hopper’
111
Gazar
232
g a zd a g ‘rich’
98
g a zd ig ‘rich’
98
g e n j ‘treasure’
Falben ‘the pale’
274
100
f e j ' head’
96, 110
38
fe lk e zu ‘one-arm ed’
38
f e r j ‘husband’ *fi
34 304
304
Finnic Finnish
Gepidae
1 3 ,1 2 3 ,2 6 2 ,3 1 0 357
German xviii, 1 3 ,1 4 ,3 3 ,9 3 ,9 8 ,1 1 0 ,1 1 6 ,
285, 287, 303, 352, 368, 375, 376, 377, 379,
390, 423
Germanic
183, J84, 203, 206, 207, 246,
247, 264, 271, 285, 320, 375, 376, 441 Germans
fln k e ‘fiv e ’ Finn
46, 79, 94, 165, 200, 287, 407
163, 164, 184, 194, 273, 274, 283, 284,
9 7 ,1 0 0
fe lk a ru ‘one-armed’
fe n y o ‘pine, fir’
98
Georgian
gerencser ‘potters’
fe jn i ‘to milk (a co w )’ f e k ‘break’
111
g a b o n a ‘grains’
93
183 37 5 5 ,1 6 3 ,1 7 4 ,1 8 3 ,1 8 4
Germat Geta
7, 244, 248, 275 306
310
g h u rk h a n ‘title’ Ghuzz
375
2 3 5 ,2 5 6 ,2 8 7
545
Index o f words, ethnic and triba l names g orenv 'polecat’ Gothic
hadnagy ‘second lieutenant’
321
38, 55, 203, 207, 208, 248, 320
Goths
55, 206, 208, 308, 376 xvi, 6 , 11, 12, 46, 50, 52, 54, 60,
hacking ‘birch’
106
haghan ‘Khaghan’
39
h agy ‘to leave’
183
67, 79, 9 3 ,9 9 , 113, 131, 133, 152, 164,
hagym a ‘onion’
38
Greek
167, 180, 184, 195, 200, 204, 208, 216,
h a j ‘hair’
38
228, 244, 258, 276, 277, 278, 283, 287,
haj ‘fat’
184
288, 289, 297, 298, 304, 306, 308, 312,
hajdina ‘buckwheat’
347,
hajo ‘boat’
349, 356, 369, 406, 419, 4 2 1 ,4 3 0 ,
hakhan
440, 441 Greeks giila
6 ,6 1 ,7 6 ,2 6 0
112 112
gyam'i ‘suspicion’ 110
g y a p jii' w o o l’ Gyarmat
306, 312, 430
gyekeny ‘bulrush’
367
321
gyem ant ‘diam ond’
gyertya ‘candle’
gyla
302
110
gyeplo ‘reins’
106
‘axe’
Hanti
346 380
ha ra n g ‘b ell’
9 9 ,1 0 6
gyom or ‘stom ach’ gyon 'to con fess’
98
harom tiz ‘three tens’ hars ‘linden’ h at ‘six ’
190
Hattie
191 3 9 ,9 7 ,3 1 9
112,367
343, 347, 352, 353, 355, 367,
289,
3 9 ,1 0 6 ,2 3 4 226
7, 91, 190, 191, 202, 286, 287, 369
Hephthalite
63, 197
herjo ‘sparrow-hawk’ het ‘seven ’
390
195
96
ha zin g ‘birch’ Hebrew
xvi, 107
98
345
367
hazi ‘cadi’ 302
110
gyom ‘w eed’
106
h a v to m o b il' car’
gyogyul ‘to heal’
gyula ‘title’
189
179
h az ‘house’
gylas ‘title’
3 8 ,1 8 4
h a tty u " sw an’
347
368,
haluka
Hatti
273
gyila ‘title’
184
ha ll ‘to hear’
harm inc ‘thirty’
148
gyasz ‘bereavement’
*gyeu
3 8 ,1 8 4
hanbalik ‘capital’
gyart ‘m anu factu re’, ‘produ ce’, ‘fabri cate’
3 9 ,1 0 6 ,1 4 9 ,1 5 0 ,2 3 3
hal ‘to d ie’
gyalaz ‘to slander’
111
39, 97, 106, 108
hal ‘fish’
347
3 6 ,1 0 6
9 6 ,1 9 0 ,1 9 1
gyidol ‘to hate’
112
hetfo ‘M onday’
gyum dlcs ‘fruit’
39, 107, 110, 225
hetm agyar ‘Seven M agyars’
gyuru ‘ring’
1 0 6 ,1 0 8
gyuszii ‘thim ble’
357
356
hetti ‘seven ’
152 273, 370
340
hetiim ajer ‘Seven M agyars’
340
h etum ajeri ‘Seven M agyars’
340
hetum oger ‘Seven M agyars’
340
hab ‘foam ’
38
him ‘m ale’
h a d ‘arm y'
3 8 ,9 9 ,1 8 3 ,3 5 6 ,3 5 7
Hittite
3 8 ,2 7 1
36, 187, 188, 191, 192
546
Appendices
hitvany ‘contemptible’ ‘w orthless’ 39, 106
424, 425, 426, 427, 429, 430, 431, 432,
ho ‘sn ow ’
4 3 3 ,4 3 4 , 437, 441, 4 4 3 ,4 4 4
38
h o d ‘beaver’
h ungarius
3 9 ,9 7 ,3 1 9
H oi-yin lrgen ‘people o f the forest’
273
195
H ungri
hollo ‘raven’
38
hungrin
h om ok ‘sand’
39, 97, 106, 108, 319
hunni
hounger(us)
2 8 3 ,2 8 5
283 287 211
Hunnish
285
2 0 7 , 2 0 8 ,2 1 1 , 3 2 0 ,3 8 7 , 4 2 3 ,
424, 427
328
hua (M iddle Chinese: *hwar)
282, 313
hunger(us)
* h o k u z 'a x '
hsin ‘lady’
285
hungarus
hodu see h a d
248
huran ‘birch’
226
hiiker ‘o x ’
102
h u r o k ‘n oose’
1 0 6 ,1 0 8
hiikiz ‘o x ’
195
hiisz ‘tw enty’
38
h u k iir 'o x '
194
Hukurito Hun
195
13, 51, 53, 55, 75, 7 8 ,1 0 3 , 123, 203,
/ш г ‘to pull’
38
Hvar
151 ,2 4 8
H yon
197, 199
204, 205, 206, 207, 208, 210, 211, 264, 272. 275, 285, 305, 308, 309, 320, 321, 325,
373, 3 7 5 ,4 2 3 ,4 2 4 , 427, 437
hungaer(us) hungar Hungari
2 8 3 ,2 8 5
2 8 3 ,2 8 5 298
Hungarians
xiii, xvii, xviii, xx, 4, 11, 12,
ichirgu boyla ‘internal b o y la ’ id nap ‘holy day’ iga ‘yo k e ’
228
366
111
igyfegy ‘holy’
366, 367
ilig ‘ruler’, ‘king’
27, 33, 36, 37, 38, 39, 57, 7 2 ,9 2 , 93, 95,
ilteber ‘title’
9 7 ,9 8 ,9 9 ,1 0 0 . 101, 102, 1 0 4 ,1 0 6 ,1 0 7 ,
im a d ‘to adore’
111, 112, 113, 115, 116, 117, 119, 120,
Imek
351
432 112
234
133, 139, 141, 142, 152, 153, 154, 159,
Imi
161, 165, 171, 172, 173, 174, 176, 178,
Imniscaris
234 181
180, 181, 183, 184, 187, 190, 191, 192,
in ‘sin ew ’
95, 355
194, 195, 199, 207, 208, 215, 217, 225,
inag ‘friend’, ‘com panion’, ‘m ate’
228, 242, 255, 256, 259, 261, 272, 273,
inanch
275, 276, 277, 278, 280, 281, 282, 283,
Indian
284, 285, 287, 283, 294, 296, 298, 302,
Indie
303, 304, 305, 306, 307, 308, 309, 311,
Indo-European
313, 319, 320, 321, 324, 325, 328, 332,
351
351 6 7 ,8 3 ,9 7 ,1 9 5 ,2 9 7 195 22, 33, 35, 36, 76, 9 3 ,9 4 ,
95, 96, 111, 171, 174, 187, 188, 189,
334, 338, 339, 340, 344, 347, 348, 349,
190, 191, 192, 194, 195, 247, 271, 303,
350, 351, 353, 354, 355, 356, 357, 360,
3 1 8 ,3 5 2 , 365
363, 364, 365, 366, 367, 370, 381, 382,
Indo-Iranian
383, 385, 389, 391, 395, 396, 397, 400,
ineg
401, 402, 403, 404, 405, 406, 408, 409,
inench
410, 413, 414, 415, 418, 419, 422, 423,
ing ‘shirt’
9 6 ,1 8 9 ,1 9 5 ,3 0 3
3 5 1 ,3 5 5 ,4 3 2 351 108
Index o f wards, ethnic and tribal names
Inkeri
185
Innuits
je b u ‘title’
317
Iranian
355
97, 112, 133, 167, 441, 444
82, 96, 97, 98, 99, 101, 111, 113,
Jeno
280,
Jermat
407, 433
Irish-English
103
Jewish
426
111
430 2 1 5 ,2 7 7 ,3 5 5
7 2 ,7 5 ,9 0 ,9 1 ,1 5 0 ,2 0 2 ,2 3 2 ,2 3 3 ,
2 8 7 ,2 8 9 , 348, 349, 356
irk ‘divination rune’
280
jila ‘title’
110
iro ‘buttermilk’
225
3 1 2 ,3 8 9 ,4 3 0 ,4 3 2
je v u "title’
irenge 'stirrup’
107
je m is h ‘fruit’
je r k e ‘ew e’
307, 308, 318, 324, 365, 386, 387,
107, 225
je m ilc h ‘fruit’
194, 195, 1 9 6 ,1 9 7 , 208, 228, 239, 248,
1 5 0 ,3 2 9 ,3 4 3 ,3 4 7
j o ‘good’
104
100, 356
jo b b a g y ‘serf’
ishad ‘title’
150
jo le ‘title’
ishtanu ‘name o f g o d ’ Ishtek
108
je m ic h ‘fruit’
163, 166, 173, 178, 180, 187, 189, 192,
Iron
2 1 5 ,2 3 0
"jelek ‘cap’
insdg ‘poverty’ ir ‘to write’
547
1 9 1 ,2 4 7
jo p a n ‘title’
179
356
343 58
ju h a s z ‘shepherd’
isk o la ' sch ool’
9 3 ,1 1 5
j u la ‘title’
ispan ‘steward ( o f an estate)’
93, 115,
Jurchen
357
1 5 0 ,3 4 3 ,3 4 7
85, 254, 256, 375
3 5 3 ,3 5 5 ispinot ‘spinach’ Isten ‘G od’
93
191
Istvan ‘Stephen’ Italian
93
93, 115
1 4 ,3 4 ,5 3 ,9 3 ,2 8 7 ,3 6 8
izenge ' stirrup’ Izhor
111
kacsa ‘duck’
istrang ‘traces (o f a harness)’
103
k a d i' cod i’
106
kading ‘birch’
kadota ‘a part o f the plough’ k a d z i' cod i’
185
1 0 6 ,2 2 6
kadzing ‘birch’ kaghan-beh
106
150 111
kakas ‘cockerel’ ja b g u ‘title’
2 3 0 ,3 5 5
Jacobite
7 4 ,7 5
Japanese
101
kakat-szeg ‘an ail in the plough’ fa j/a ‘fish’
kalafa ‘to fo llo w ’
64
kalangya ‘haystack’
Jarmat
kalasz ‘ear (o f w heat)’
Jasz
104
Kalmiik
jo t- se e y a t-
Kamas
ja tc h i ‘rainmaker’ ja u sh ig ir ‘title’
233
ja va s ‘witch-doctor’ Jazygian
104
jeb g u ‘title’
365
230
111 111
149 178
kam on ‘m illet beer’ kcmca ‘mare’
365
111
184
ja ra t- see yarat4 3 0 ,4 3 1
184
3 9 ,1 0 6 ,2 3 4
208
111
Kangar
63, 234, 238, 288, 419, 420, 421
Kangli
300
kantar ‘bridle’
3 8 ,1 0 1 ,1 1 0
548
Appendices
kanyaro ‘m easles’
38
kerodzik ‘to ruminate’
kapiste ‘pagan sacrificial shrine’ кари ‘gate’
xvi, 38
Karagas
232
karam ‘(cattle)pen’
3 8 ,1 1 0
khabaroi
2 02
230,
94 38, 39, 321
111
katoikeo ‘to settle’
kavradi ‘brought together’
348 348
420
Kharakhanid
80, 266, 375
K hasar
300
1 1 5 ,2 1 3 ,2 2 8 ,2 2 9 ,2 7 3
khatun ‘title’ 329, 343, 347
kender ‘hem p’
1 0 0 ,1 1 0 ,3 2 1
кепуёг ‘bread’
346, 347, 352, 355, 390
84, 234, 252, 255, 256, 330,
3 6 5 ,3 7 8
152
110
kengyel ‘stirrup’
2 7 4 ,3 0 6 2 5 6 ,2 6 6
kharha ‘title’
110
228, 253, 263, 355 1 0 2 ,2 7 4
Kharakhalpakh
Kharlukh
k ed d ‘Tuesday’
253
khara bodun ‘black people, com mon peo
Kharakhitai
111
162
kecske ‘goat’
355
56, ! 35, 238, 252, 257, 296, 322,
327, 328, 329, 330, 331, 347, 348, 349, 3 5 0 ,3 5 1 ,3 5 2
368
Khazakh
110
272
Khavar
khavari
98
kephale prote ‘first head’
kerem et
348
226
kazal ‘haystack’
228
khara ‘black’
p le’
kavir- ‘to collect, assem ble’
Kerel
422
228
kaying ‘birch’
274 317
khapkhan ‘title’
422
kava- ‘to sw ell (o f a w ound)’
Kereit
Khamchadal
khapagan ‘title’
k a /o s k e n o ‘to m o v e so m ew h ere’
kavhan ‘title’
39, 150, 213, 214, 225,
233, 247, 263, 355, 374
K han ‘title’
202
kepe ‘stook’
347, 380, 382,
3 4 7 ,3 5 0
khalpakh ‘hat’
kasza ‘scythe’
k ep ’im age’
38
khaghan ‘title’
357
karvaly ‘sparrow-hawk’
Kelar
152
387 380
kende ‘title’
1 7 9 ,3 1 7
kez ‘hand’
184
karkundaj ‘title’
K echua
107
300
Khabar (see also Khavar)
karhas ‘title’
Kasog
432
K e tte d
183
Kartvelian
kese ‘sm all’
K et
178
kar о ‘stake’
228, 229, 273
Kesimir
368
K araim ‘the literate’
Karelian
kesar ‘title’
k esik ‘to be late’
karacsony ‘Christmas’
Karatai
368
38
kar ‘arm’
110, 111
kert ‘garden’
111
kapal ‘to h oe’
110
347 1 4 4 ,1 9 9 ,3 0 6 ,3 2 1
Khazan Tatar 346, 380
Khazar
281
39, 52, 53, 63, 6 6 , 6 8 , 69, 75, 78,
7 9 ,9 1 ,1 0 6 ,1 1 2 ,1 1 4 ,1 1 5 ,1 2 3 ,1 3 2 ,1 4 8 ,
229, 3 0 0
150, 151, 152, 165, 182, 2 0 1 ,2 0 2 ,2 1 5 ,
365
216, 219, 220, 221, 223, 224, 226, 228,
kereszt ‘cross’
368
229, 230, 231, 232, 233, 234, 235, 237,
549
Index o f words, ethnic and triba l names
246, 247, 249, 252, 257, 272, 273, 276,
kolovur ‘title’
282, 287, 288, 291, 295, 296, 297, 307,
kolyok ‘kid’
113,228 1 0 6 ,1 0 8 110
308, 310, 322, 324, 325, 326, 327, 328,
kolyu ‘pounder’
329, 330, 347, 348, 349, 350, 353, 354,
kom ar ‘m osquito’
356, 357, 365, 367, 369, 373, 378, 379,
Komi
380, 386, 387, 388, 390, 413, 416, 417,
kom lo ‘hop (the plant)’
418
k om ondor ‘(one type of) sheepdog’
Khazarous
298
Khirghiz Khitai
179,271 3 8 ,1 0 8 ,1 1 0
85, 87, 253, 254, 255, 256, 266,
kondor
64, 279, 280
433
Kondoros
433
1 0 4 ,2 7 9 ,2 8 0 ,2 8 1
Kondorosh
Khurturghur
213
kondorsok ‘a water plant’
168,
149,
2 3 3 ,3 4 3
433
kondoz ‘beaver’
209
konyar ‘horseherd’
Khutur Oghur
209
koporso ‘coffin ’
Khwarezmian
104, 196, 197, 222, 23 I
кд рй ‘churn’
kicsi 'sm all’ kifchak
10 7 ,4 3 2
98
korte 'pear’
2 3 4 ,3 3 1 ,4 3 5
koris ‘ash’
112
9 7 ,1 0 0
179, 180, 215, 232, 234, 235,
239,
272, 289. 293, 301, 302, 434, 435
kirdly ‘king’
1 10, 111
36
273, 277, 353
K rai ‘king’
273, 277, 293, 353
kulavuz ‘title’
113
Kun ‘Cuman’
274
107
ktinde ‘title’
Knbyna (?)
287
ktindti hakhan ‘title’
178
*kundur
k o k tu rk ‘Blue Turks’ kokeny ‘slo e ’
280
110
357
kudzat ‘plural o f k a d z i“cadi” ’
kis ‘sm all’
Koibal
38, 107, 110
kovacs ‘blacksm iths’
kincs ‘treasure’ Kipchak
111
kos ‘ram (the animal)’
308
A w ‘pain’
38
fo r r ‘w o lf’
107
Kimmer
110
korpa ‘bran’
kilenc ‘nine’
357
112,368
10 1 ,3 7 4
кого ‘w eed’
234
*kilchen Kimek
Korean
300
420
111
konkoly ‘corn cock le’
Khuturghur (Khutur Oghur)
Kibchaut
433
433
kongor ‘brown (colour o f horse)’
209
433
433
K hotanSaka
khut ‘power got from the heaven’
433
kondorak ‘Agrostis (a water plant)’ kondorchuk
3 7 4 ,3 7 5 ,3 8 3 .4 3 5
Khutur
110
kondirak ‘Agrostis (a water plant)’
80, 84, 168, 214, 2 5 2 ,2 5 3 , 378
Khotanese
433
343, 347 233
433
K undurcha
433
kunduz ‘beaver’
433
kolcson ‘loan’
107,112
kunta
koidok ‘navel’
109
kuolla
koldul ‘to beg’
112
t e r ‘belt, rank’
kolovros ‘title’
113
kurdash ‘belt-mate’
183 184 135 135
106
550
Appendices
Kurtugermatos Kutrigur kuu
Machagar
430
Machar
11,2 4 8
Machgar
184
kuula
M agash
184
kuvasz ‘(one type of) sheepdog’ kuvra- ‘to assem ble’ kuvrat- ‘to collect’ K valiz
297 300
110
348
M ager
298
M agere
6 1 ,2 9 8
Magyar
348
297 433
xiii, xv, xvii, xviii, 4, 5, 10, 11,
1 3 ,1 4 ,1 5 ,1 6 ,1 7 ,2 2 ,2 3 ,2 8 ,2 9 ,3 6 ,3 7 ,
213
3 8 ,4 3 ,4 5 ,5 3 ,5 4 ,5 5 , 56, 57, 58, 5 9 ,6 1 , landorispdn ‘palatine’ lang ‘flam e’
115
99
langos ‘a large, savoury fried doughnut’
120, 122, 132, 133, 134, 135, 136, 138, 139, 142, 143, 147, 151, 152, 153, 155, 159, 160, 162, 163, 166, 167, 183, 187,
99 Lanikaz
189, 192, 195, 199, 208, 212, 216, 220,
234
222, 225, 229, 233, 234, 235, 237, 238,
111
lapat ‘spade’
239, 242, 243, 246, 249, 251, 252, 254,
Lapp
183
Latin
xxi, 32, 33, 52, 54, 55, 57, 58, 61.
64, 93, 115, 164, 173, 183, 190, 194, 202, 229, 244, 245, 272, 283, 286, 287, 298, 300, 303, 308, 309, 340, 355, 356, 364,
368, 369, 376, 377, 407, 414, 430,
437, 4 4 0 ,4 4 1 Latvian
lelek ‘sou l’ Lengyel
368, 369, 370, 373, 374, 375, 377, 380, 381, 382, 385, 386, 387, 388, 389, 390,
184
391, 397, 398, 399, 400, 401, 402, 403,
Luwian
406, 408, 409, 410, 413, 414, 415, 417,
97
Longobardi
304, 305, 306, 307, 308, 309, 310, 311,
359, 360, 362, 363, 364, 365, 366, 367,
185
16 ‘horse’
294, 296, 297, 298, 300, 301, 302, 303,
351, 352, 353, 354, 355, 356, 357, 358,
237
Lithuanian
294, 296, 297, 298, 289, 290, 291, 293,
340, 341, 344, 346, 347, 348, 349, 350,
242
Lenzenin
285, 286, 287, 288, 289, 290, 291, 293,
332, 333, 334, 335, 336, 337, 338, 339,
367
242
Lenzen
273, 275, 276, 277, 278, 282, 283, 284,
324, 325, 326, 327, 328. 329, 330, 331,
367
111
len ‘flax’
255, 256, 257, 258, 264, 266, 271, 272,
312, 313, 317, 319, 320, 321, 322, 323,
184
lehel ‘breathe’
Livian
62, 6 8 , 73, 74, 92, 100, 117, 118, 119,
1 3 ,1 2 3 ,2 6 2 ,3 3 6
187, 188
418,
419, 420, 421, 423, 424, 425, 426,
427, 429, 430, 432, 435, 4 4 1 ,4 4 3 M agyer
3 0 3 ,3 0 7 ,3 8 6
m aharaja ‘a title’ m aa ‘land’ M aamees Macedonian
184 184
m aja
391
m acha, ‘tale, fable’
297
m airya ‘young man, young person’ 100
Maja Er 304
Majaer
307 307, 310
303
551
Index o f words, ethnic and triba l names Majagar Majar
m ees ‘person’
294, 297 2 9 8 ,3 0 0 , 301, 3 0 2 ,3 0 5 ,3 0 6 , 310,
313
M egeres
184
306
m egye ‘shire’
353
m ajaran
2 9 8 ,3 0 0
M egyer
Majarar
300
m eh ‘bee’
Majarat
300
Mejer
Majarin
298, 300
M ensch ‘man’
Majer
1 4 5 ,1 4 6 ,2 9 4 ,2 9 5 ,2 9 6 ,2 9 7 ,3 0 7 ,
3 1 0 ,3 1 3 ,3 4 3 ,3 4 7 ,3 6 4 ,3 8 6 , 435 Majgariya
Meri
36, 96
3 0 6 ,3 0 7 ,3 5 0 ,3 5 1 ,4 3 2
Merens
3 0 5 ,3 0 6 ,3 0 7 ,3 5 0 ,3 5 1
Majgar
305, 307, 350
183
m ert ‘mortal’, ‘person’
313
Merya 1 1 1 ,3 5 5 ,3 7 1
m ese ‘tale’
m am a ‘m other’
173
M eshcher
Managir
293
Mesher
307
Manch
1 0 0 ,3 0 4 ,3 0 7
M iddle Iranian
304
303 303
Misher
2 8 1 ,3 0 1
Mochar
3 4 8 ,3 8 6
M oger
1 8 1 ,2 7 1 ,3 0 3
m arta ‘mortal’, ‘person’
303
303
298
3 0 3 ,4 3 0
Moghul
272
M ogyer
294
Mojgar
294, 313
m aryah ‘mortal’, ‘person’
303
Moksha Mordvin
m a u ra p a id ia ‘black boys’
420
m olnar ‘m iller’
298
Meadow Cheremis
Mede
111
2 9 8 ,3 1 3
Mocharin
303
M ariy ‘person, husband, Cherem is’
Mecher
79
3 0 6 ,3 1 3 ,4 3 2
m iskarol ‘to geld’
319
m ard ‘person’
M iddle Turkic Mingrelian-Laz
319
Manycha-Er
63, 104, 165, 216, 228,
2 3 0 ,2 8 8
1 7 8 ,3 0 3 ,3 0 4 ,3 0 5 ,3 1 1 ,3 1 3
Mazarous
100, 413
97, 100,101, 104, 320
M iddle Persian
m anush ‘man’
Mari
1 11
303
m anu ‘m an’
Manys
181
M iddle Hungarian
m ansh kum ‘Vogul person’
Manycha
88
M iddle Chuvash
189
Manshi
46
M iddle Chinese
95, 96, 101
m angalica (a type o f pig) M ann
96
M iddle Greek
9 5 ,3 1 8
Manchu-Tunguzian
195
306
m ez ‘honey’
8 5 ,3 1 8 ,3 7 5 ,4 1 0
Manchu-Tunguz
Mandean
306
M eyeris
m ancha ‘tale, fable’ Manchu
304 1 8 3 ,3 0 6 ,3 1 3 ,4 3 2
M esopotamian
303
mon ‘great’ 181
3 0 6 ,4 3 2
M ongol
182 111
293
xvii, 75, 79, 85, 89, 90, 145 147,
154, 180, 181, 189, 202, 209, 229, 272,
195, 196
m edos ‘honey beer’
303
181, 183
m a la c ' pigling’
m an ‘great’
303
181
279, 298, 300, 301, 302, 308, 310, 312, 208
3 2 1 ,3 7 5 ,4 1 4
552
Appendices
M ongolian
95, 101, 102, 104, 115, 135,
Norman
246, 327, 426
138, 145, 146, 149, 150, 253, 254, 272,
Norman Viking
299,
no ‘w om an’
173
« « ‘w om an’
173
300, 320, 321, 360, 362, 364, 374,
375, 3 9 7 ,4 1 0 ,4 3 5 Morava
Nwkarda
377
Moravian
61, 109, 133, 243, 244, 245,
266, 329, 331, 332, 334, 336, 337, 360, 3 6 9 ,3 7 7 , 3 7 8 ,3 8 2 , 388, 389, 391
291
nyak ‘neck’
97
nyar ‘sum m er’ nyars ‘spit’
97
345
Mordens
181
nyereg ‘saddle’
Mordvin
1 8 1 ,1 8 2 ,1 8 3 ,2 9 8 ,3 2 2
nyolc ‘eight’
morior ‘to die’ mort ‘m an’
303 303
Kiosh ‘an exogam ic group’
303
Ob-Ugrian
38, 97, 142, 166, 174, 176,
178
ocsti ‘tailings’
178
110
Mozhar
2 9 8 ,3 0 3 ,3 1 3
odo ‘field ’
Mujgar
313
odo marij 305
*murj ‘pepper’ Muroma
9 7 ,3 1 9
98
303
mortos ‘mortal’, ‘person’ Motor
376, 377, 378
307
183
180
Oghor
211
Oghur
102, 103, 1 0 4 ,2 0 9 ,2 1 1 ,2 1 2 ,2 1 4 ,
murt ‘man’
2 7 1 ,3 0 3
215,
muzh ‘man’
303
2 6 1 ,2 8 4 , 3 2 1 ,3 2 5 , 387
229, 247, 248, 252, 253, 255, 256,
Oghur Turkic Oghuz
nador span ‘palatine’
115, 353
nadorispan ‘palatine’
115, 353, 390
Nagman
name
432
95
Nandor
2 8 4 ,3 9 0
Nandor
2 1 5 ,2 8 4 115, 353
34, 36
nemet ‘German’
273, 274
nemoy ‘dumb’ N enets
Nganasan
1 1 0 ,1 9 4 ,3 2 1
o k so ‘o x ’
1 9 4 ,2 4 7
okiir ‘o x ’
101, 102, 194, 247, 284
61 ‘sty’
110
Old Bolghar
246,
273, 274
Old French
N oghay
2 7 2 ,3 0 6 ,3 1 2
Nokarda
201
108 7, 364
Old German
178
Old Greek 80
60, 61, 107, 113,228,
284
Old Chuvash
95
Nine Oghuz (Tokhuz Oghuz)
121, 4 3 1 ,4 3 3 ,4 3 4
60
Old Church Slavic
273, 274
178
nev ‘nam e’
284
okor ‘o x ’
Old Bashkirian
2 1 5 ,3 2 9
narancs ‘orange’ nemets ‘German’
2 6 6 ,2 8 4 ,2 8 7 , 3 3 1 ,3 6 5
ok ‘arrow, tribe’
okiiz ‘o x ’ 101, 102, 194, 247, 284
nandorispan ‘palatine’ Nandur ‘Bulghar’
209
102, 209, 234, 235, 239, 256, 257,
38, 283 303
Old High German Old Hungarian 340, 432
283, 352 100, 108, 167, 279, 328,
553
Index o f words, ethnic and triba l names Old M ongolian Old Indie
254
Ossetian
303
200
O ssetic
9 8 ,1 0 4 ,1 8 9 ,3 2 4 ,3 2 8
Old Iranian
9 7 ,9 9 ,3 1 8
ostor ‘w hip’
Old Persian
189,303
Ostrogoths (East Goths)
Old Prussian
184
Old Russian
Ostyak
93
Old Slavonic
osztoke ‘prod’
189
Old Tibetan
84
Old Turkic
Ottoman
100, 101, 104, 106, 114, 135,
Olhontor Bulkar
3 0 1 ,3 0 8 ,3 0 9
Ottoman Turkish
2 0 9 ,2 7 3 ,3 4 1
dtvds ‘sm ith’ Oyrot
303
1 0 5 ,2 7 8
209, 248, 434
Otuz Tatar 110
357
1 0 5 ,3 0 1 ,3 0 2 ,3 7 9
otuz ‘thirty’
215
olio ‘kid (a young goat)’ onchara ‘tusk’
111
Ottoman Turk
189, 226, 2 8 0 ,2 8 1 ,2 8 4
Onger
3 8 ,9 9 ,1 7 9 ,3 0 3
otteveny ‘smelt m etal’
239
Old Syrian
203, 206, 231,
233
2 7 4 ,2 8 4 , 435
Old Slavic
96
357
8
285
Onggroi
283
ongon ‘(protective) amulet’
149
Pahlavi
6 3 ,2 1 6
Ongre
2 8 4 ,2 8 6 ,4 3 5
Paion
Ongri
435
pajta ‘stable’
Ongur
2 0 0 ,2 8 4
Onoghundur
309
Palaic
1 1 5 , 2 1 5 ,2 2 0 ,2 2 7 ,2 3 7 ,
284, 285, 435
palatinu s‘palatine’ Palmyrian
Onoghundur Bulghar
241, 242, 262, 286
111
187, 188 116
63
palotas ‘pal ace warden ’
On Oghur
2 0 9 ,2 1 1 ,3 4 1 ,4 3 6
p a lta ' axe’
Onoghur
51, 112, 123, 124, 139, 199,
p a l u k a ‘ax.e'
355
189 189
209, 211, 233, 248, 255, 273, 282, 284,
pancha ‘fiv e ’
285, 286, 308, 309, 313, 387, 407, 435
Pannonian Avar
219
Pannonian Slav
1 6 7 ,3 7 8
Onoghur-Bulghar
1 3 2 ,1 3 9 ,2 8 4 ,4 1 4
Onoghur-Khazar
133
pannonii
On Okh Bodun ‘Ten Arrow P eople’ Onughur
3 1 3 ,3 8 6
onychara ‘tusk’ orok ‘eternal’
98, 303 367
orol ‘to grind’ Orosut
papa ‘daddy’
173
pa raku ' axe’
189
224 189
parlag ‘fallow land’ p a r th a 'a x e ' 355
orvos ‘medicine man’ ‘doctor’ 272
368
p a r a th u ‘a xe'
110
orszag ‘country’, ‘realm’ os ‘ancestor’
11
p a p ‘priest’
p a r-sil
1 1 0 ,3 2 1 ,3 6 3
300
orsd ‘reel’
Osman
284
93
39
Parthian 112, 365
1 0 4 ,1 9 5 ,1 9 6 ,2 3 1 ,3 0 9
Pascatir
290
Pascatu
290
Pascatur
111
189
290
554
Appendices
p a szto r ‘shepherd’, ‘herdsman’
p a t ‘fiv e’
111, 357
222
p a szid y ‘bean’ 93 111
p a ta ‘h o o f’
Pecheneg
p o k o l ‘hell’
368
*pokU r' o x ’
194
*poktis ‘o x ’
194
Polish
111
p a tk o ‘horseshoe’
1 4 ,9 3 ,2 3 9 ,2 8 4
3 1 ,8 4 , 1 0 5 ,2 0 1 ,2 3 4 ,2 3 5 ,2 3 7 ,
Polovets
274
238, 239, 249, 257, 276, 287, 288, 291,
Polyane
2 4 6 ,3 7 7
293, 295, 308, 310 ,324, 325, 326, 327,
Proto-Turkic
108
330,
Proto-Uralic
94, 175
331, 332, 334, 335, 337, 346, 347,
373, 377, 387, 388, 416, 417, 419, 420,
pseudo-Avars
4 2 1 ,4 2 2 , 426, 443
p u n k d sd ‘Pentecost’
Pecinaci
420
p u rta " axe’
p ecu s, p e co ris ‘cattle’ *peku *peleku
189
p e n k i ‘fiv e’
93
p en kv e ‘fiv e ’
quinque ‘fiv e ’
93
p en ta -ko sta ‘fiftieth (day after Easter)’ 93
r-Turkic
p e n te k ‘Friday’
R aral
p e n ti ‘fiv e ’
9 3 ,1 5 2
93
102, 103, 104, 114, 209, 212 300
R em niscans
p e n tik ‘fifth (day)’
93
181 111
repa ‘beet’
298
111
retek ‘radish’
9 8 ,9 9 , 101, 139, 180, 1 8 2,319,
386,
400
Roman
7, 11, 103, 123, 124, 141, 200,
231, 304, 334, 376, 377, 379, 424
1 6 6 ,3 1 8 ,3 1 9 ,3 2 4
Romanian
1 7 3 ,3 1 0
Roxolani
196
34, 52, 63, 6 6 , 67, 71, 73, 131,
rozs ‘rye’
111
183, 195, 196, 2 0 1 ,2 1 2 ,2 1 3 ,2 1 4 ,2 2 5 ,
Ruanruan
Permyak Persian
179
234,
240, 257, 262, 275, 290, 294, 295,
296, 297, 300, 302, 353, 354, 367 p e t ‘fiv e ’
93
p e ta k ‘penny’
p e tti ‘am ulet’ p ila k k a ‘ax e ’
93
304
2 1 7 ,2 4 9 ,2 5 7 ,2 7 1
Russian 93
253, 2 5 5 ,3 7 5 , 435
33, 60, 93, 98, 165, 171, 177,
179, 180, 181, 182, 184, 185, 192, 195,
366
231, 239, 253, 271, 274, 278, 280, 286,
189
290, 291, 293, 298, 305, 306, 312, 350,
189
*pinkve ‘fiv e ’
248,
8 0 ,2 1 0 ,2 1 1 , 2 1 2 ,2 1 3 , 2 1 4 ,
R um eri ‘Greek person’ Rus
p e te k ‘five (-penny p iece)’
p ilk a ‘axe’
93
93
p e n ta ‘fiv e’
Permic
93
189
p e lk a ' axe’
Permian
115 93
p y en c h ‘fiv e ’
189
93
189
p y a ty ‘fiv e ’
p e le ku s ‘axe’
Peon
211
p u ya n ‘rich’
194
194
111
p o h a n k a ‘buckwheat’
93
377 russkiy m uzh
304
555
Index o f words, ethnic and tribal names Sabir
112, 200, 209, 211, 212, 213, 224,
2 4 8 ,3 0 9 ,3 2 1 ,3 2 5 ,4 3 3 ,4 3 4 Sabirian
212
Selkup
1 7 8 ,1 7 9 ,2 7 1
Semitic
22, 33, 63, 67, 74, 79, 190
Semitic-Hamitic
sa g ra k ‘goblet’
226
sepro ‘dregs’
sahrah ‘goblet'
226
Serb(ian)
Saka
64, 195, 196, 197, 248, 279, 280,
281,
sereda ‘W ednesday’ sereg 'army’
Sakaliba
2 3 5 ,2 3 9
serke ‘nit’
Saklab(i)
2 3 5 ,2 3 9
Sambat Sam oyed
177,271
Sapiri
104, 174, 177, 178, 318
246
106
Sara Oghur
sh a d ‘title’
233
shanyu ‘title’
sarkany ‘dragon’ sarlo ‘sick le’
419
sham lu ‘apple’
209 27
Sharaghur
209
110,363
Shara Oghur
209
shpan ‘title’
115
213
shupan ‘title’
300 357
skala ‘scale’
Sauromata
196
Sklaven
289
Slav
Savarti A sfali
212, 238, 288, 386, 418,
419, 421
Saxones
11 178
377
scola ‘sch ool’
9 3 ,1 1 5
2 7 6 ,3 0 8 , 3 0 9 ,3 1 0 , 4 2 3 ,4 3 7
s e g o r ‘ axe’ Seljuk
258, 262, 264, 273, 276, 277, 284, 285, 286, 287, 295, 310, 320, 327, 329, 332,
190 190
256, 266, 308, 309, 379
363, 365, 367, 369, 377, 378, 379,
3 8 6 ,3 8 7 ,3 9 0 ,3 9 1 ,4 3 5 ,4 4 0 slave
239
52, 53, 195, 196, 200, 204, 275,
securis ‘axe’
13, 1 4 ,5 2 ,5 6 ,6 0 ,6 9 , 116, 133, 134,
152, 163, 167, 182, 208, 228, 237, 239.
357,
sclavus ‘slave’
Scyth
212
274, 309, 376
Sayan Samoyed sclaveni
93
240
240, 241, 242, 243, 244, 245, 246, 257,
Savir see also Sabir Saxon
115
112
sir ‘tom b’
sator 'tent’ Savarti
192 204
1 9 6 ,3 0 8 ,3 0 9 ,3 1 0
Sarurghur
420
4 1 9 ,4 2 0
Sevortioi
209
Sarmata
Severyan
Sevortii
212
Saraghur
380
289
Sevordik ‘o f black descent’
83
sar ‘m ud’
Sasut
Severti
228
Sanskrit
110
Seven Magyar
Sam oyedic sam psi
300
serte ‘bristle’
152
152
112
106
Serkesut
274
111
60, 93, 239, 242, 377, 382
320
Saksa
94
Slavic
377 93, 110, 111, 113, 114, 115, 116,
123, 126, 132, 133, 152, 167, 184, 190, 255, 257, 262, 264, 265, 284, 286, 287, 352, 353, 355, 356, 357, 364, 365, 367, 377, 378, 379, 380, 382, 388, 389, 390, 407, 420, 433, 434, 435, 440, 441 slavii
377
556
Appendices
Slavonic
7 ,1 4 ,6 0 ,6 1 ,1 8 2 ,1 8 3 ,1 9 0 ,2 0 0 ,
Syarmas
228, 239, 240, 243, 245, 264, 271,
Syepir
208,
93, 239, 273, 280, 352, 396
Slovakian Slovene
280 2 4 6 ,2 8 6
Slovenian
212
Syrian
4 6 ,6 3 ,7 4 ,7 5 ,1 6 5 ,2 1 3 ,2 2 9 ,2 3 1 ,
407,
63, 64, 80, 82, 104, 195, 196,
197, 1 9 8 ,2 1 2 ,2 1 4 ,2 7 9 110 , 111
so m ‘dogberry’
so n b o ta ‘Saturday’ Sorb
225
Syipir
420 111
sza lm a ‘hay’
239
Sogdian
212
Syim esy
273, 276, 305 Slovak
181
152
2 3 9 ,3 7 7
sza rv ‘horn’
Southern Slavonic
108 1 9 0 ,2 3 9 ,2 4 0 ,2 6 4
96
sz a z ‘hundred’
96
szecska ‘ch aff’
111
Szdkely
so reg ‘a type o f fish ’
33, 1 15,225, 3 6 9 ,3 7 0 ,4 0 5 ,4 0 6 ,
437, 440, 4 4 1 ,4 4 3 , 444 szek er ‘w agon’
97
Sovar
2 1 2 ,2 2 4
s z e k e r c e 'hatchet, a x e ’
Soyot
178
sz e l ‘w ind’
106
szen a ‘hay’
111
sp a n ‘title’
9 3 ,1 1 5 , 353, 355
Spaniards Spanish
szen t ‘saint’, ‘sacred’
7
sparga ‘asparagus’ sp en d t ‘spinach’ Srem niscans sta rt ‘start’
110
93 208
93
siichii ‘honey beer’ sh u ku rru ‘lance’
183
226 226
110
sziics ‘furrier’
190
Tabin
432
Taigi
178
takacs ‘w eavers’ ta l ‘dish’
Suvaz
224
taltos ‘shaman’ 108
79
Sw edish
110 , 111
szu szek ‘wheat container’
2 1 2 ,2 2 4 ,2 2 6
Svan
109
sz u r ‘to filter (w in e)’
Suvar
su v e g ‘cap’
357
108
111
su ta ‘d oe’
152,191
sziinyog ‘m osquito’
su c h ig ‘honey beer’
355
3 6 ,3 9 , 107, 110, 111
szo m b a t ‘Saturday’
w heat’
stu d io ‘studio’
Sultan
97
szo lo ‘grape’, ‘vin e’
szo r ‘to throw’, ‘to separate tailings from
stra va ‘funeral feast’
Suomi
191
152
308
szd ‘w ord’
93
siillo ‘zander’
93, 367
szerda ‘ Wednesday ’ Szittya
93
93
Stephanus S tra n g
93
181
S te g r e if ‘stirrup’
1 9 0 ,3 6 7
szep tem b er ‘ Septem ber’
3 7 ,7 3 ,2 3 1
111
sza m o ro d n i ‘a kind o f vin e’
tanacs ‘counsel’ Tangut
1 8 3 ,1 8 4
357
97 364, 365 112
415
tanu ‘w itness’
112
111
557
Index o f words, ethnic and triba l names tarack ‘stolon’, ‘couch grass’
2 2 8 ,2 5 6 ,3 5 1 ,3 5 2 110
tarld ‘fallow land’ tarlo ‘stubble-field’ *tarmacs
Tocharian
96, 104, 189, 192, 193, 194,
195,
110
tarka ‘piebald’ larkhan ‘title’
Tat
111 , 280
111
tarho ‘curds’
247
tokhur ‘nine’
248
tokhuz ‘nine’
209, 248
Tokhuz Oghuz
321
80, 209, 273, 284, 340
tolm acs ‘interpreter’
116
tom eny ‘abundant’
352
Tatar 7 ,1 6 2 ,1 8 0 ,1 8 1 ,1 8 3 ,2 0 9 ,2 1 5 ,2 3 4 ,
tor ‘dagger’
357
432, 433, 434
torek ‘chaff’
278
Tavgi Samoyed
/e y ‘m ilk’
Tork
178
9 6 ,1 1 0
287
torm a ‘horseradish’
9 6 ,1 1 0
tor ok
112
te le k ' plot’
torontal ‘a predator bird’ torveny ‘law ’
tengri ‘god, sk y’
366
Tot
teprv ‘only, ju st’
273
troszk
term ecs
1 1 2 ,1 1 6 ,1 6 7 ,3 5 5 ,3 7 1
116
Termecsu term es
116,355
3 5 5 ,3 7 1
Teryuhan
183
teszar ‘carpenter’ Teuth
357
352 285
Teutonic
184 ,2 8 5
teve ‘cam el’
280
trpov
274
truk
279
triik
279
truka
279
truku
280
triiku
2 7 9 ,2 8 0 ,2 8 1 279 281
tudun
256, 263, 355
tujue (M iddle Chinese: thiirkvat) tiik o r ' mirror’
341 Thirty Tatar (O tuz Tatar)
341
Three K harlukh (U ch K harlukh) Three K hurikhan (O ch K urikhan) Tibetan
340 340
82, 83, 84, 228, 229, 252, 253,
2 8 0 ,3 1 2 ,3 5 4 ,
Timochani
Tulmach
287
Turgesh
8 4 ,2 8 0
Turgish
280
Turk
279
107
xvi, 10, 51, 52, 53, 56, 69, 70, 73,
407
80, 83, 8 4 ,1 1 0 , 111, 112, 116, 148, 178,
106, 110, 321
211, 212, 214, 215, 216, 225, 229, 237,
179, 182, 189, 194, 195, 200, 204, 209,
213
tilo ‘hemp-cutter’
242, 264, 377
tino ‘young o x ’ Tirpdk
309, 4 2 4 ,4 2 6
trpak
tturaka
110,321
Thirty O ghur ({Jtur Oghur, Utrughur)
Tiele
280
ttruka
Teutisch
321
112, 367
352
Troyan
xvi, 109
terem ‘hall’
110
278
Ten O ghur see O n O ghur
te r d 'k n e e '
1 1 2 ,2 8 7 112
112
tor ‘feast’
273, 301, 304, 306, 386, 403, 417, 429,
tehen ‘c o w ’
110
toklyo ‘year-old lamb’
2 7 3 ,2 8 0
110
238, 254, 256, 274, 275, 276, 277, 278, 279, 280, 281, 282, 287, 288, 289, 295, 304, 309, 310, 312, 316, 322, 326, 334,
558
Appendices
TuTk(continued) 3 3 5 ,3 4 6 ,3 4 8 ,3 4 9 , 354,
Uch Khurikhan
365, 367, 374, 378, 379, 380, 383, 386,
Udmurt
387,
« d « /‘to rest’
405, 407, 414, 416, 417, 418, 419,
4 2 1 ,4 2 2 ,4 4 1 TOrk Turkic
112
udvarnok ‘court warden’
2 7 8 ,2 7 9 ,2 8 0 ,2 8 1
Turki
340
1 8 0 ,2 7 1 ,3 0 3 ,3 0 4 ,3 0 5
53
Uggroi
xvi, xvii, xviii, 7, 8 , 10, 11, 27, 36,
38, 39, 5 2 ,6 3 , 71, 72, 74, 80, 81, 82, 83, 84, 94, 95, 96, 97, 101, 102, 103, 104,
355
111
« g a r ‘fallow land’ 283
Ughor
4 5 ,2 0 9
Ughur
209, 211, 212, 213, 214, 230, 261,
435, 436
105, 107, 109, 110. I l l , 112, 113, 115,
Ugr
131, 132, 133, 135, 138, 148, 149, 163,
Ugri
166, 167, 173, 179, 192, 194, 197, 199,
Ugrian
287 6 1 ,2 8 4 ,2 8 6 ,4 3 5 22, 96, 97, 98, 99, 110, 122, 139,
200, 201, 204, 208, 209, 210, 212, 215,
142, 176, 179, 190, 191, 286, 287, 305,
225, 226, 228, 229, 232, 239, 241, 247,
3 0 6 ,3 1 8 ,3 1 9 ,3 8 6 ,4 0 2
248,252, 253, 254, 256, 258, 264, 274, 276, 277, 278, 279, 280, 281, 282, 284, 285, 289, 290, 291, 294, 297, 301, 304,
8 , 63, 80, 82, 84, 209, 229, 234,
Uighur
235, 252, 254, 255, 266, 284, 342, 349, 3 6 5 ,4 3 5
306, 307, 3 1 2 ,3 1 9 , 320, 318, 321, 322,
iiker ‘o x ’
102
325, 327, 340, 341, 348, 349, 350, 351,
ukos ‘o x ’
194
Ukrainian
9 8 ,2 3 9
352, 355, 357, 360, 363, 364, 365, 368, 373, 374, 378, 379, 380, 383, 387, 388, 389,
390, 398, 400, 402, 403, 407, 410,
414, 420, 4 3 1 ,4 3 2 ,4 3 3 ,4 3 4 , 441, 444 Turkish
105, 141, 162, 274, 278, 279,
2 8 1 ,2 9 0 ,3 1 2 turkkut
279
Тигкб
278
Turkoi
5 3 ,2 7 6 ,2 8 1
Turkoman TurkU tiirkii tiirkiit
2 7 5 ,3 0 1
Tiirok Tiiriik tu ru l
2 7 8 ,2 7 9 ,2 8 0 ,2 8 1 279 110
2 7 8 ,2 7 9 2 7 8 ,2 7 9
iy«&‘hen’
U nc
272
Ungar
282, 283, 284, 285
Ungari
300
Ungarus
2 8 3 ,2 8 5
Ungerus
285
Ungor
309
Ungri
53, 58, 212, 255, 283, 284, 286,
Ungroi
434, 435
287
tirrnep ‘festival’
1 1 2 ,3 6 6
Unnoghundur-Bulghar
iino ‘c o w ’, ‘heifer’ ur ‘lord’ Uralic
2 5 6 ,4 2 6
tiizok ‘bustard’
237
3 1 0 ,4 1 4 ,
278
turd ‘cottage ch eese’
Ultin
110, 321
1 1 2 ,3 5 5 ,3 7 1
2 2 ,3 4 ,3 5 ,9 3 ,9 4 ,9 5 ,9 6 ,1 5 3 ,1 6 6 ,
1 7 3 ,1 7 4 , 176, 1 7 7 ,1 8 5 ,3 1 7 ,3 1 8 ,3 2 4 ,
321
404
110
uri
371
iirom ‘w orm w ood’ uborka ‘cucumber’ Uch Kharlukh
215
111
273, 340
321
tirii ‘sheep’, ‘wether ’, urum ‘title’
355
110, 3 21
559
Index o f words, ethnic and trib a l names
uruszag ‘country’, ‘realm’ Utrighur Utur
2 0 9 ,2 4 8
Uturghur Utur Oghur
301,
Volga Turkic
110
Votic
Vongur
iivecs ‘ew e ’
2 3 5 ,2 3 7 ,2 5 6 ,2 5 7 ,2 8 4 1 9 5 ,2 7 2
uzengu 'stirrup’ Uzian
201
Votyak
9 8 ,9 9 , 1 8 0 ,2 7 1 , 305, 386
Vungar
285
103
Vunundur
215
1 0 5 ,2 0 1 ,3 0 9
Vununtur
215
vajer ‘tusk’ Vanandar
110
328
1 9 7 ,1 9 9 ,2 1 1 ,2 1 3
Varangian
247
varazs ‘m agic’ Varkhon
365
2 1 0 ,2 1 3 ,3 0 5
vas ‘iron’
95 152
vasvero ‘»iron beater«’, ‘blacksmith’ 357 194
2 7 1 ,3 7 8 ,3 7 9
Visigoths
111
203
239, 242, 264, 322, 341, 347
West Turk
5 3 ,2 1 5 ,2 3 0 ,2 6 2 ,2 8 2
West Turkic
150, 1 6 6 ,2 1 5 , 277, 325
W hite Hun
378
197
White Serbs
378 286
255 215
Xianbei
2 1 3 ,2 5 3
Xiongnu
6 3 ,8 5 ,1 9 2 ,2 0 3 ,2 0 4 ,2 0 5 ,2 0 8 ,
2 1 3 ,3 2 0
437 284
Vlendur Bulkar Vodian
285
West Slav
Wulundur
villa ‘pitch-fork’
Vlendur
Wanger
185
Viking
Vlach
296
Wuguli
284
Vepsian
38
Wanandar
White Ugrian
194
*venezis
wafr ‘m an’
W hite Croatian
vasdrnap ‘Sunday’ veneris
246
98
valyti ‘trough’
Vengri
434
185
Vyatich
Var
167, 293, 298, 299, 300,
302, 306, 308, 311, 409, 432
209, 341
285
Uzbek
349, 379, 386, 4 0 0 ,4 0 8 ,4 0 9 ,4 1 4 , 429,
Volga Magyar
252
uuangar
291, 293, 294, 301, 310, 324, 326, 327,
4 3 1 ,4 3 2 ,4 3 3 ,4 3 4 ,4 3 5
2 0 9 ,3 4 1
Uturkar
Uz
355
1 1 ,2 4 8 ,2 5 2
7 8 ,2 1 5
vodzer ‘tusk’
Yabgu
255
Yabghu
185 98
2 2 9 ,2 7 3 ,2 7 7
yabloko ‘apple’
Vogul
3 8 ,1 7 8 ,1 7 9 ,3 0 4 ,3 0 5
yaghlakar
Vogur
201
Yakut
Volga Bulghar
1 5 ,6 3 ,7 0 , 113, 114, 120,
192
272
281
yarat- ‘to manufacture’
1 2 1 ,1 6 7 , 180, 181, 212, 220, 221, 222,
Yaolien
223, 224, 225, 226, 227, 248, 257, 287,
yas
255
200,202
148
560
Appendices
y a t- ‘magic to create rain’ Yazig
8 , 435
Yellow Uighur Yelyu
255 1 7 9 ,3 1 7
Yenisei Samoyed Yiddish Yighur
178
7 ,9 0
435
Yujielyu (M iddle Chinese hiukkulie)
yukagir Yula
3 0 7 ,3 1 7
150 178
Yurak Samoyed 233
y illa w a r ‘title’ y ilte v e r ‘title’
178
4 5 ,5 5
Yurmati 226
4 3 0 ,4 3 1 ,4 3 2
y u s ‘ermine’
200
224, 226, 347, 349, 432
234 435
z-Turkic
1 0 2 ,1 0 3 ,2 5 6 111
Yuezhi
192
za b ‘oat’
Yughur
435
zh u p a ‘territorial unit’
Yugri Yugria yugrush
255
255
Yiirka
y ilig ‘title’
Yoghur
255
Yugur
Yurak
435
347
Yimek
Yuguli
Yujueli
Yenisei Ostyak
y ila
365
2 0 2 ,3 0 9 ,3 1 0
179
zh u p a n ‘title’
1 7 9 ,4 3 5 ,4 3 6 256
y u g u ru sh ‘title’
Zikh Zyryan
263, 434
382
1 1 5,353, 390
2 0 0 ,2 0 1 ,2 8 7 98, 99, 166, 179, 180, 271, 303,
386, 434
LIST OF MAPS, FIGURES AND PLATES
Maps and figures 1. The southern frontier o f glaciation in Eurasia - p. 16 2. The geobotanical zones o f Central Eurasia - p. 18 3. Climatic, geohistorical, archaeological and language historical periods - p. 21 4. Biogeographical features o f the Uralic U rheim at - p. 35 5. The Roman Empire under Augustus (63 BC - 14
ad
) -
p. 47
6 . The Byzantine Empire under Justinian (5 2 7 -5 6 5 ) - p. 48
7. Transliteration and transcription o f Greek characters and clusters - p. 50 8 . The spread o f Islam until 750 - p. 65
9. Islamic dynasties in the 9th—10th centuries - p. 66 10. The Armenian Empire - p. 77 11. The Old Tibetan Empire and its contacts - p. 83 12. Central A sia and China under the Tang dynasty (6 1 8 -9 0 7 ) - p. 86 13. Central A sia and China under the Khitai Liao (9 0 7 -1 1 1 9 ) and the Chinese Song dynasties (9 6 0 -1 2 7 8 ) - p. 87 14. Wanderings o f the Jews expelled from Spain - p. 92 15. Relative locations o f the Uralic and the neighbouring peoples - p. 95 16. Conquest-period sites beyond the Carpathians - p. 118 17. Roman Pannonia - p. 124 18. Settled areas o f the Carpathian Basin in the 5th—6 th centuries - p. 125 19. The Buyla inscription on the Nagyszentm iklos plate - p. 127 20. Inscriptions o f the N agyszentm iklos treasure I (1 -6 ) - p. 128 21. Inscriptions o f the N agyszentm iklos treasure II (7 -1 4 ) and the tent (15) - p. 129 22. Inscriptions o f the four sides o f the Szarvas needle-case and their reconstruction - p. 130 23. Hair pin, Przemysl - p. 132 24. Drawing o f a Conquest-period grave. M indszent (Csongrad county, Hungary)— Koszorudiilo, Grave 2 - p. 137 25. Drawing o f a Conquest-period grave. K enezlo (Borsod-Abauj-Zemplcn eounty, Hungary)— Farkaszug, Cemetery I, Grave 11 - p. 137 26. The Saltovo-M ayak culture, the neighbouring archaeological cultures and the possible location o f major towns - p. 140 27. M ongolian tent roof-hoop p. 144
562
Appendices
28. Wooden framework o f a Khazakh tent - p. 144 29. Wooden framework o f a M ongolian tent - p. 145 30. Depiction o f a tent on a Chinese Tang dynasty lacquered bowl - p. 146 31. Physical anthropological groups o f the conquering Magyars - p. 155 32. Vegetation o f the Carpathian Basin and the EtelkOz (reconstructed by B. Z61yomi) - pp. 156-157 33. The skull measurements o f the two groups o f the conquering M agyars - p. 160 34. The Ural and the Kelteminar cultures in the 4th-2nd m illennia ВС - p. 175 35. The Ural U rheim at and the distribution o f Uralic languages today - p. 176 36. Typological classification o f the Uralic languages, after Peter Hajdii - p. 177 37. The Indo-European peoples’ Urheimat. V iew s since 1960 - p. 188 38. Ancient Indo-European languages, Hittite, Luwian and Palaic, am ong Anatolian and Middle-Eastern non-Indo-European languages - p. 188 39. The evolution o f Indo-European languages, after Gamkrelidze and Ivanov - p. 189 40. Isoglosses o f the Indo-European languages, after Anttila - p. 190 41. The main Indo-European languages and their early locations - p. 191 42. The hypothesised emergence o f the Tocharians, the Andronovo and the A fanasyevo cultures - p. 193 43. The Tocharians and their neighbours in the Tarim Basin - p. 193 44. The early distribution o f the Iranian peoples - p. 196 45. Developm ent o f the Persian Empire ( 6 th—5th centuries
вс)
- p. 197
46. Alexander the Great’s empire (3 3 6 -3 2 3 BC) and the Sogdians - p. 198 47. The Parthian and the M ede Empire (3 r d -lst centuries
вс)
- p. 199
48. China and the Xiongnu - p. 205 49. The split o f the Roman Empire in the 4th century, the Hun attack, and the Goths - p. 206 50. The Huns in Europe - p. 207 51. Early migrations o f the Turkic peoples - p. 210 52. Khuvrat’s grave - p. 217 53. Royal grave sites in the Ukraine steppe zone (7th century) - p. 218 54. Decipherment o f the inscription on Khuvrat’s signet ring - p. 218 55. Khuvrat’s Bulghar Empire (7th century) - p. 219 56. Bulghar migrations at the end o f the 7th century - p. 221 57. Bulghars, Magyars and Alani in the Khazar Empire (7th—9th centuries). The northerly migration o f the Volga Bulghars - p. 222 58. Ibn Fadlan’s journey - p. 223 59. Central A sia in the 9th—10th centuries - p. 236 60. The distribution o f the Slavs between 450 and 550 - p. 240 61. The distribution o f the Slavs between 600 and 700 - p. 241 62. P eoples in the Carpathian Basin in the 9th century. Avar and Avar-influenced Slavic archaeological sites in Transdanubia and Austria - p. 265
563
lis t o f maps, fig u re s and plates 63. Abu Hamid al-Gam ati’sjou m ey - p. 292
64. The distribution o f the splintered Volga Magyars after the M ongolian invasion - p. 299 65. The M agyars’ self-designations and external designations prior to the Conquest - p. 3 11 6 6 . Migrations o f the ancient Magyars - p. 323
67. The Magyar Conquest - p. 333 6 8 . Structure o f the N ine Oghuz tribal confederation -
p. 342
69. Reconstructed Conquest-period man’s costum e (after Gyula LSszIo) - p. 358 70. Reconstructed harness - p. 359 71. Make-up o f livestock in M ongolia - p. 361 72. The earliest alphabets o f the Sz^kely runiform
scriptI- p.438
73. The earliest alphabets o f the Szekely runiform script II - p. 439 74. The runiform inscription o f H om okm ^gy-H alom (near Kalocsa, South Hungary) from the time o f the Conquest - p. 442 75. The four sides o f the Szarvas needle-case - p. 442
Plates I. Hilt o f the sword o f Kiev II. 1. Golden eye-cerement, R akam az 2. Eye- and mouth-cerement, H ajdubdszorm eny-V id
3. Conquest-period grave, K unadacs III. Dated coins o f Louis the Infant and the Samanid ruler Esmail ibn Ahmad (9 0 4 -9 0 5 ) IV. Press-forged belt bosses, mounts and other finds from the early Avar period V. Belt mounts with griffin and sarmentose motifs, and other finds from the late Avar period VI. Conquest-period belt bosses, R etkozberencs-P aradornb VII. Conquest-period belt bosses, ring, and bone quiver-mount, R etkd zb eren cs-P a ra d o m b VIII.
1. Conquest-period belt bosses, belt boucle, sabretache plate, and trepan, Tiszaeszlar-B ashalom 2. Belt-end, Tiszaeszlar-B ashalom
3. Belt-end, N agyosz 4. Belt-end, Tuzser IX. Gilded silver mounts o f an “ever-ready” quiver, Karos X. Sabretache plates 1. E perjeske 2. Szolyva XI. Sabretache plates 1. A lsoverecke 2. B ezded
564 XII.
Appendices
M ounts from a sabretache, Ujfeherto-Micskepuszta
XIII. Sabretache plate, Karos XIV. Sabretache plate, Galgdc XV. Sabretache plate, Szolnok-Strazsahalom XV I. 1. Saddle, Izsak-Balazspuszta 2.
Saddle decoration, Soltszentimre
SOURCES OF MAPS, FIGURES AND PLATES
M aps M ost o f the historical maps published in this book were originally drawn for my university lectures. Their present form was completed by Andrea Fodor (DIM AP). For historical maps not listed below I made good use o f several historical atlases among which I would especially mention the D T V -A ilas zu r W eltgeschichte I—II, Deutscher Taschenbuch Verlag, 1982. The sources o f the follow ing maps are: 16. Fodor (1994) 17. M 6 csy (1984) 18. B 6 na (1984) 26. B ilin t (1989b) 31. Ery (1994) 33. Z61yomi, in: Gy6rffy-Z61yorm (1994) 37., 38., 41., 42., 43., 44. M allory (1989) 52., 53. Werner (1984) 58. Zim onyi (1990) 59. Kristo (1994c) 62. BolSakov-M ongajt (1985) 63 V£s4ry (1976) 66. P o h l(1 9 8 8 ) 71. Badam2av (1966)
Figures M ost o f the figures were sketched by m yself. I would like to mention the sources o f the follow ing figures: 19., 20.,21., 22 Gob!, in: G 6 b l-R 6 na-Tas (1995) 23. Fodor (1996a) 24. Csall&ny, in: K o v ic s (1994) 25. Josa, in: K o v ic s (1994) 27., 28., 29. R 6 na-Tas (1961a) 30. Gabain (1971) 32. Made by A k os R 6 na-Tas 36. Hajdii, in: H ajdu-D om okos (1978)
566
Appendices
39., 40. Mallory (1989) 69. Laszlo, in: K ovacs (1994) 70. K ovacs, in: Kov&cs (1994) 74. Juhasz, in: Sandor (1992b) 75. Varadi, in: G obl-Rona-Tas (1995)
Colour plates The colour plates are photos made by Jozsef Hapak. I ow e sincere thanks for the courtesy o f the follow ing museums for their kind permission to publish the photos o f the objects in their custody: Andras Jozsa Museum, Nyiregyhaza, Hungary: II. 1., VII, VIII. 1 2 . , 4., X. 1, XII Museum o f Hajdiiboszormeny, Hungary: II.2 Otto Herman Museum, M iskolc, Hungary: III., IX., XIII Hungarian National M useum, Budapest, Hungary: II.3, VI. 1., 2., 3.,VIII.3., X .2., XI. l.,2., XIV., XV., XVI Jozsef Katona M useum, Kecskemet, Hungary: X V I.l Plates 1.1,2. are from K ovacs (1994) Plates IV., V. are from H ajdu-K risto-Rona-Tas II (1997)
Hilt o f the sword o f Kiev
Plate II
1. Golden eye-cerement, R akam az
3. Conquest-period grave, Kunadacs
2. Eye- and mouth-cerement, H ajdiiboszorm eny-V id
Dated coins o f Louit
the Infant and the Samanid ruler Esmail ibn Ahmad (904-905)
Plate IV
Press-forged bell bosses, mounts and other finds from ihe early A var period
P la te V
Belt mounts with griffin and sarmentose motil's, and other finds from the late Avar period
Plate V I
P la te V II
Conquest-period belt bosses, ring, and bone quiver-mount,
R it k d z b e r e n c s - P a m d o m h
1. Conquest-period belt bosses, belt boucle, sabretache plate, and trepan, Tiszaeszlar-Bashalom
2. Belt-end, Tiszaeszlar-Bashalom
3. Belt-end, Nagyosz
4. Belt-end, Tuzser
Gilded silver mounts of an “ever-ready” quiver, Karos
Plate X
2.
S z o ly v a
Sabretache plates
1.Alsdverecke
2. Bezded
Mounts from a sabretache, Ujfehdrto-M icskepuszta
Plate X I I I
Plate XIV
Sabretache plate, Galgoc
Plate X V
Sabretache plate, Szobiok—Strdzsahalom
Plate XVI
2. Saddle decoration,
S o lt s z e n t im r e