Phrasal Verbs Tesis

Phrasal Verbs

Topics in English Linguistics 78

Editors

Elizabeth Closs Traugott Bernd Kortmann

De Gruyter Mouton

Phrasal Verbs The English Verb-Particle Construction and its History

by

Stefan Thim

De Gruyter Mouton

Doctoral dissertation, Friedrich-Alexander-Universität Erlangen-Nürnberg (originally submitted under the title “Phrasal verbs in transition: a diachronic and comparative study of the English verb-particle construction”)

ISBN 978-3-11-025702-1 e-ISBN 978-3-11-025703-8 ISSN 1434-3452 Library of Congress Cataloging-in-Publication Data A CIP catalog record for this book has been applied for at the Library of Congress. Bibliographic information published by the Deutsche Nationalbibliothek The Deutsche Nationalbibliothek lists this publication in the Deutsche Nationalbibliografie; detailed bibliographic data are available in the Internet at http://dnb.dnb.de. ” 2012 Walter de Gruyter GmbH, Berlin/Boston Printing: Hubert & Co. GmbH & Co. KG, Göttingen 앝 Printed on acid-free paper 앪 Printed in Germany www.degruyter.com

Acknowledgments This book is the published version of my doctoral dissertation at the Friedrich-Alexander-Universität, Erlangen (2009). I am greatly indebted to Angelika Lutz, my Doktormutter, for her invaluable criticism, advice and encouragement throughout the writing of the dissertation. For various kinds of professional support I also wish to thank Minoji Akimoto, Renate Bauer, Hartmut Burmeister, Claudia Claridge, David Denison, Klaus Dietz, Philip Durkin, Marion Elenbaas, Teresa Fanego, Mechthild Habermann, Thomas Herbst, Andrew James Johnston, Dieter Kastovsky, Lucia Kornexl, Svenja Kranich, Bettelou Los, Robert Mailhammer, Ferdinand von Mengden, Donka Minkova, Horst Haider Munske, Eva-Maria Remberger, Hans Sauer, Herbert Schendl, Hildegard L.C. Tristram, Theo Vennemann, Ilse Wischer and Nuria Yáñez-Bouza. During the very early phase of the dissertation I stayed as a visiting scholar at the Research Unit for Variation and Change in English at Helsinki University, and I would like to thank Terttu Nevalainen and her colleagues, especially Arja Nurmi, Helena Raumolin-Brunberg and Matti Rissanen. I gratefully acknowledge funding for the stay from the German Academic Exchange Service (DAAD). Elizabeth Traugott has been a marvellous editor, who deserves the highest praise for the feedback she has provided. All remaining errors are of course mine. And last, but by no means least, I thank my family and friends for putting up with me for all the years, and for their love and support.

Berlin and Erlangen, August 2012

Stefan Thim

Contents Acknowledgments List of figures List of tables Abbreviations 1. 1.1. 1.2. 1.3. 1.4.

Introduction Aims The term phrasal verb Contrast, comparison, history Structure of this study

2. 2.1. 2.2.

Present-day English and other Germanic languages Phrasal verbs: some examples Semantic characteristics 2.2.1. Three semantic types 2.2.2. Compositional constructions 2.2.3. Aspectual constructions 2.2.4. Idiomatic constructions Syntactic characteristics 2.3.1. Transitivity and serialization 2.3.2. Particles and prepositions 2.3.3. Phrasal-prepositional verbs Further observations 2.4.1. Cranberry verbs 2.4.2. Nominalisations 2.4.3. Other word formations 2.4.4. Prefix verbs and phrasal verbs 2.4.5. ‘Group-verbs’ etc. 2.4.6. Replaceability by a simple verb 2.4.7. Variation and style Verb-particle constructions in other present-day Germanic languages 2.5.1. Basic word order 2.5.2. Particle position

2.3.

2.4.

2.5.

v x xi xii 1 1 2 3 8 10 10 11 13 14 16 19 20 21 26 28 30 30 30 34 34 36 40 42 45 47 48

viii

Contents

2.6.

Phrasal verbs as periphrastic word formations 2.6.1. Verbs plus particles? 2.6.2. Periphrastic word formation 2.6.3. Further pros and cons 2.6.4. Phrasal verbs as constructions Conclusion

2.7. 3. 3.1.

3.2.

3.3. 3.4. 4. 4.1.

4.2.

4.3.

4.4. 5. 5.1.

55 56 62 67 69 72

The development of postposed particles Preverbs 3.1.1. Preverbs in non-Indo-European languages 3.1.2. Preverbs in Indo-European The development of English word order 3.2.1. Word order in earlier Germanic 3.2.2. Word order in Old English 3.2.3. The rise of Modern English word order The position of the particle in medieval English Conclusion

74 75 78 81 89 89 93 100 103 115

Writing the history of the phrasal verb A classic study: Kennedy (1920) 4.1.1. The ‘rise’ of the phrasal verb 4.1.2. Colloquiality, informality, nativeness 4.1.3. The impact of Kennedy’s study Some textbooks and language histories 4.2.1. The pitfalls of history 4.2.2. Coverage in CHEL I and II Lexicographic coverage: a characteristic example 4.3.1. Bosworth–Toller and other older dictionaries 4.3.2. The Middle English Dictionary 4.3.3. The Oxford English Dictionary 4.3.4. The Dictionary of Old English 4.3.5. Concluding remarks on the historical dictionaries Conclusion

117 117 118 120 123 124 124 126 131 132 135 138 140 143 143

Word formation, sound change and semantics Changing prefix inventories in English 5.1.1. The Old English prefixes 5.1.2. The prefixes in Middle English and beyond

145 145 146 153

Contents

5.2.

ix

Preverbs and particles in medieval English 5.2.1. On sound change and word formation 5.2.2. Some comparative evidence 5.2.3. Prefix variation in Old English 5.2.4. Particle semantics in medieval English 5.2.5. Some conclusions An outlook to modern English 5.3.1. Etymology and integration 5.3.2. Phrasal verbs in 15th- and 16th-century English Conclusion

158 158 165 171 176 183 185 185 192 195 197 197

6.3. 6.4.

Frequency, style and attitudes Counting phrasal verbs 6.1.1. Quantitative long-term developments from Middle English to the 20th century 6.1.2. Early Modern English frequencies 6.1.3. Relative frequencies of particles 6.1.4. Quantitative long-term developments Style and attitudes 6.2.1. Text type and frequency 6.2.2. Pre-1800 evidence for colloquiality? 6.2.3. An example: Samuel Johnson 6.2.4. James Cook rewritten and John Dryden revised The colloquialization conspiracy: a first suggestion Conclusion

197 205 210 211 214 215 218 221 226 233 245

7. 7.1. 7.2.

Conclusion Summary Outlook

247 247 252

5.3.

5.4. 6. 6.1.

6.2.

References

255

Index

293

List of figures Figure 2-1 Figure 2-2 Figure 2-3 Figure 3-1 Figure 3-2 Figure 3-3 Figure 5-1 Figure 5-2 Figure 6-1 Figure 6-2 Figure 6-3

Semantic classification of phrasal verbs Subtypes of complex predicates The transitive verb-particle construction with its two allostructions Paths of adpositions and preverbs Positional changes and their causes Development of the order ‘verb-particle’ from early Old English to late Middle English

13 65 72 88 104 112

Semantic development of verb-particle constructions Etymologies of phrasal verbs (types) in English letters, 1450–1600

193

Development of phrasal verbs (Konishi 1958) Phrasal verbs in English plays (Spasov 1966) Development of phrasal verbs, 1640–1740

199 201 206

183

List of tables Table 2-1 Table 2-2

Phrasal verbs and prepositional verbs compared Group-verbs in present-day English

Table 3-1 Table 3-2

Word order in early West Saxon OV and VO orders in English

97 101

Table 5-1 Table 5-2 Table 5-3

Old English verb prefixes and their etymologies Old English verb prefixes: overview Modern English verb prefixes

151 152 154

Table 6-1 Table 6-2

Phrasal verbs in two corpora Phrasal verbs in the Corpus of Early English Correspondence Phrasal verbs per 1,000 words in three corpora Relative frequencies of particles in three corpora Frequencies of phrasal verbs in nine corpora

209

Table 6-3 Table 6-4 Table 6-5

27 37

210 210 211 213

Abbreviations Short titles of editions For full bibliographic information on editions used see the list of references at the end of the book. Where examples are retrieved from the Dictionary of Old English (DOE) or the Dictionary of Old English Corpus, the references follow the DOE conventions. ÆCHom I = Ælfric’s Catholic Homilies: The first series (Clemoes 1997) ÆCHom II = Ælfric’s Catholic Homilies: The second series (Godden 1979) ÆHom = Homilies of Ælfric: A supplementary collection (Pope 1967–68) ÆLS = Ælfric’s Lives of the Saints (Skeat 1881–1900 [1966]) Ancient Laws (Thorpe) (Thorpe 1840) Ancr (Nero) = The English Text of the Ancrene Riwle (Day 1952) ApT = The Old English Apollonius of Tyre (Goolden 1958) AV = The Authorised Version of the English Bible 1611 (Wright 1909) Bede = The Old English Version of Bede’s Ecclesiastical History of the English People (Miller 1890–1898) Beowulf (Dobbie 1953) BlHom = The Blickling Homilies (Morris 1874–1880 [1967]) Chaucer (Benson 1988) [GP = General Prologue; Mel = The Tale of Melibee; MT = The Miller’s Tale] ChronA = The Parker Chronicle (Plummer 1892–1899 [cf. Bately 1986]) ChronE = The Peterborough Chronicle (Plummer 1892–1899 [cf. Irvine 2004]) ChronF = The Domitian Bilingual (Dumville 1995) CP = King Alfred’s West Saxon Version of Gregory’s Pastoral Care (Sweet 1871) Ex = Exodus (Krapp 1931) GD = The Old English Translation of Gregory’s Dialogues (Hecht 1900– 1907) Gospels = The Old English Version of the Gospels (Liuzza 1994) Herbarium = The Old English Herbarium and Medicina de Quadrupedibus (de Vriend 1986) LawIne = Ine’s Laws (Liebermann 1903) LS 5 (InventCrossNap) = Invention of the Cross (Napier 1894) Mk = Mark (CCCC 140) in The Four Gospels in Anglo-Saxon, Northumbrian, and Old Mercian Versions (Skeat 1871–1887)

Abbreviations

xiii

MtGl (Ru) = Matthew (MS Rushworth) in The Holy Gospels in AngloSaxon, Northumbrian, and Old Mercian Versions (Skeat 1871–1887) Or = The Old English Orosius (Bately 1980) PsGlG (Rosier) = The Vitellius Psalter (Rosier 1962) St Marg = Seinte Marherete: Þe Meiden ant Martyr (Mack 1934) Tatian (Sievers 1892) T-Chron [B and F] = The Anglo-Saxon Chronicle (Thorpe 1861 [cf. Taylor 1983 and Dumville 1995]) WHom = The Homilies of Wulfstan (Bethurum 1957) Wulfila = Die Gotische Bibel (Streitberg 2000 [1919]) Grammatical abbreviations Grammatical category labels in interlinear morphological glosses are printed in small capitals (‘GEN’ for ‘genitive’, etc.). The interlinear glosses follow the Leipzig Glossing Rules and provide relevant structural information rather than complete morphological descriptions. * ? 1,2,3 abl acc adv art aux comp cop dat dem det fem gen imp inf int N neg

ungrammatical / unattested acceptability doubtful first, second, third person ablative accusative adverb article auxiliary complementizer copula dative demonstrative determiner feminine genitive imperative infinitive interrogative noun negation, negative

nom NP O part perf pl pres pret prt pst ptcp rel S sg subj trans V V-1 V-2 V-3 V-F VP

nominative noun phrase object partitive perfect plural present preterite particle past participle relativizer subject singular subjunctive transitive verb verb first verb second verb third verb final verb phrase

xiv

Abbreviations

Data from the British National Corpus (BNC) Data cited in this study have been extracted from the British National Corpus (version 3, 2007, ), distributed by Oxford University Computing Services on behalf of the BNC Consortium. All rights in the texts cited are reserved. Quotations taken from the BNC are followed by the text identifier and the sentence number (e.g. Eke out the little pleasures.A5X_39).

Chapter 1 Introduction 1.1.

Aims

Phrasal verbs have long been regarded as being among the most characteristic features of the English language. As early as 1712, Michael Mattaire in his English Grammar described the basic syntactic peculiarities of the English verb-particle construction. A few decades later Samuel Johnson and Robert Lowth – one the most influential lexicographer and the other the most influential grammarian of the 18th century – directed their attention to the phrasal verb. Johnson writes in the Preface to A Dictionary of the English Language (1755): There is another kind of composition more frequent in our language than perhaps in any other, from which arises to foreigners the greatest difficulty. We modify the signification of many verbs by a particle subjoined; as to come off, to escape by a fetch; to fall on, to attack; to fall off, to apostatize; to break off, to stop abruptly; to bear out, to justify; to fall in, to comply; to give over, to cease; to set off, to embellish; to set in, to begin a continual tenour; to set out, to begin a course or journey; to take off, to copy; with innumerable expressions of the same kind, of which some appear wildly irregular, being so far distant from the sense of the simple words, that no sagacity will be able to trace the steps by which they arrived at the present use. (Johnson 1755: n.pag.)

Thus, while the term phrasal verb appears to be a coinage of the first half of the 20th century, the construction itself has attracted linguistic attention for the last 300 years or so. In the 20th century phrasal verbs came to be one of the favourite topics not only of grammarians and lexicographers, but also of authors of popular style guides; only few properties of English are commonly seen as more typical of the language than the phrasal verb. The above quotation from the Preface to Johnson’s dictionary shows that this view goes back at least to the middle of the 18th century, when many of the modern notions concerning the distinctive properties of the English language were first expressed. So powerful have these notions been that one obvious question has in fact rarely been asked, namely: how ‘English’ are the phrasal verbs really? One aim of the present study is to explore this question. The other aim is to trace the evolution of the Modern English phrasal verb from its early history up to the present. Obviously, both aims are closely connected,

2

Introduction

since the peculiarities of the Modern English construction are bound to remain indistinct unless one is ready to adopt a historical, contrastive and cross-linguistic perspective. The dichotomy of synchrony vs. diachrony has led to deplorable limitations of linguistic interests and insights throughout much of the 20th century. Over the past years, the traditional dichotomy has been transcended by a large body of usage-based functionalist, variationist and typological research, whose stance has been summed up by Bybee (2010: 104) as follows: “Language change is not just a peripheral phenomenon that can be tacked on to a synchronic theory; synchrony and diachrony have to be viewed as an integrated whole”. Thus the wide scope of the present study does not exclude the language of today, which I regard as essentially situated in a historical variational space (see e.g. Oesterreicher 2001 and the references provided there). In the following chapters a description of the main characteristics of the verb-particle construction in present-day English will be taken as a starting point from whence the development of its structural properties will be sketched in the light of previous research and of traditional attitudes. I have come to believe that many of the problems of analysis which have beset the discussion of phrasal verbs for a long time are best solved within a constructional framework, not least if historical developments are to be accounted for as well; the present study is decidedly functionalist (cf. e.g. Dik 1997, Traugott 2003a, Croft 2006, Fischer 2007, Haspelmath 2008 and 2010, Bybee 2010 and Smirnova & Mortelmans 2010). But evidence from studies adhering to other linguistic persuasions will be taken into account wherever this is appropriate for the discussion at hand. Recent monographs containing discussions of various aspects of the phrasal verbs include Claridge (2000), Dehé (2002), Hampe (2002), Gries (2003), Cappelle (2005), Elenbaas (2007), Waibel (2007) and Matsumoto (2008). Bacchielli (1999) and Dehé (2003–) provide select bibliographies. 1.2.

The term phrasal verb

The term phrasal verb is rarely ever used except with respect to English, where it is sometimes applied not only to phrasal verbs as described by Samuel Johnson in the quotation at the beginning of this introduction, but also to other verbal constructs, most notably prepositional verbs (cf. the discussion below, Chapter 2.3.2). It seems that the term was originally coined with respect to English alone and it is probably first found in Smith (1925), where it is attributed to one of the editors of the OED, Henry Bradley (1845–1923):

The term phrasal verb

3

The term ‘phrasal verb’ was suggested to me by the late Dr. Bradley; not, as he wrote, that he was satisfied with it, or would not welcome any alternative that he could feel to be an improvement. But, as he said, one cannot write of these verbs without some workable description; and although the word ‘phrasal’ is perhaps objectionable in formation, it fills a want, and is sometimes indispensable. (Smith 1925: 172, fn.1)

Characteristically, the term is introduced in a chapter on ‘English idioms’, where Smith calls the phrasal verbs “one of the most striking idiosyncrasies of our language”, despite adding the observation that they “correspond to the compound verbs in synthetic languages” (Smith 1925: 172). But the type of construction discussed in the present study also goes by a rather large number of other names. This is ultimately due to the fact that it straddles the conventional boundary of morphology and syntax and that it has no place in Latin school grammar. Consequently, there is no traditional term for it. Together with the tendency in linguistics to coin new terminology this has led to a plethora of designations, e.g. verb-adverb combination, particle verb, verb-particle combination, verb-particle construction, discontinuous verb, merged verb, separable verb, two-word verb, separable compound, poly-word verb, etc. (cf. the long list provided by Carstensen 1964: 306–308). Of course, the choice of terminology often reflects theoretical preferences, or characteristic features of the authors’ linguistic classifications. For a long time the interest in the history of phrasal verbs was a rather marginal area, though, and most studies have tended to concentrate on a small number of topics. Thus, so far some points – like the relative position of verb and particle in Old English – have received a lot of attention, while others have received considerably less attention. The first major study of the Modern English phrasal verb and its history is a short monograph by Kennedy (1920) on the ‘verb-adverb combination’, although in retrospect Harrison’s (1892) study of ‘separable prefixes’ in Old English can perhaps be taken to mark the beginning of the diachronic research tradition in English historical linguistics. 1.3.

Contrast, comparison, history

Although from the very beginning the use of the term phrasal verb implies that the construction is distinctively English, there are comparable verbal constructions in other languages, as already noted by Smith (1925). The most obvious parallels can be found in other Germanic languages. Cf. e.g. present-day German aufgeben ‘give up’, which, like its English translation,

4

Introduction

consists of a particle (auf, cognate to up) and a verb (geben, cognate to give): (1)

Present-day German Alexander gab das Cellospielen auf Alexander gave the cello:playing

up

‘Alexander gave up playing the cello.’ But neither syntactically nor semantically are there always one-to-one correspondences. Cf. e.g. the German verb aufmachen ‘open’ (particle auf and verb machen, cognate to make) in the following example: (2)

Present-day German Wenzel sagt dass Eva

die Tür

aufmachen wird

Wenzel

the door

up:make:INF AUX:3SG

says

COMP

Eva

‘Wenzel says that Eva will open the door.’ In the modern Germanic languages the distribution of pre- and postposition of the particle is entirely rule-governed. In those Germanic languages where the particle may either follow the verb as in the first example or precede the verb as in the second example it has long been customary to call such particle verbs separable prefix verbs, a term which obviously cannot be applied to present-day English, where the particle is always separated and behind the verb. In studies with a comparative focus the more neutral term particle verb (or verb-particle construction) is now well-established. Consequently, in this study this term will be used from a comparative and contrastive point of view, while the term phrasal verb will be reserved for the Modern English verb-particle construction and its peculiarities, especially where the discussion is restricted to English. In Old English, the particle may occur before or after the verb, as in the present-day Continental West Germanic languages, where the alternation between pre- and postposed particles has been historically more stable than in English. Cf. the repeated use of Old English ut-gan ‘go out’ in the following example (note that eod- is the regular suppletive preterite of gan ‘go’ in Old English): (3)

Old English (ÆCHom II, 1 [012700 (10.256)f.]) Gað ut of ðam ofne and cumað go:IMP.PL out of

the

oven and

to me

come:IMP.PL to me


Hi

ðærrihte

ut

eodon

they

immediately

out

go;PRET:PL

5

‘Come out of the oven and come to me; they immediately came out.’ Thus the primary syntactic development in the English construction is one towards almost exceptionless postposition of the particles. And while many details of the development have not been fully explained, there is now general agreement that this development is connected to the changes in the basic word order in the history of English. But the normally post-verbal position of the particle in present-day English also has close correspondences in other present-day Germanic languages. Taken together, such parallels in a group of genetically related languages can be regarded as a clear indication of common historical origins. Although shared features may in principle also be the result of convergence and contact, in the present case common ancestry is the only viable explanation. Similarly, the semantic properties of the particle verbs in the various present-day Germanic languages are very much alike. However, in this respect in all the languages the modern stages differ significantly from their earliest stages, where typically compositional combinations of verbs of motion and spatial particles are found, e.g. Old English forþfēran (literally: ‘travel/move away or by’): (4)

Old English (LS 5 (InventCrossNap) [53]) & ferde forþ on his weig and went forth on

his way

‘and went forth on his way’ Although the non-spatial and non-compositional combinations are slower to emerge, they are also already found in Old English. Thus forþfēran is frequently found in the figurative sense ‘die’: (5)

Old English (LawIne [38]) & fere se ceorl forð and go:PRES.SUBJ.SG the man

forth

‘and if the man dies’ Cf. the semantically corresponding phrasal verb pass away in present-day English. This is of course an instance of the readily intelligible and crosslinguistically wide-spread metaphor ‘life is a journey’ (cf. e.g. Lakoff 1993 and the overview in Croft & Cruse 2004: Ch. 8, with further references), used euphemistically. In fact, all non-literal senses of the verb-particle

6

Introduction

constructions may be reasonably assumed to have developed out of such metaphorical (or metonymical) uses. The observation that forþgān with its equivalent literal meaning ‘go forth’ does not seem to be used in the sense ‘die’ suggests that this metaphorical use of forþfēran is lexicalized in Old English (note that forþgān in its turn is also used as a gloss of Latin procedere, progredi, praeterire, etc.); cf. DOE s.vv. forþ and forþ-gān. The particle in Old English may either precede or follow the verb, cf. example (3) above. In general, this is entirely independent of the meaning of the Old English verb-particle construction, as a comparison of the following two examples to the preceding examples (4) and (5) shows: (6)

Old English (ÆCHom I, 10 (G) [260.65]) þe big sume weige sæt þær se hælend REL

by

some way

sat

there the Saviour

forð ferde forth went

‘who was sitting by the way where the Saviour passed by’ (7)

Old English (WHom 20.2 [9]) feower geara fæce ær he

forð ferde

four

forth went

years’

time

before he

‘four years before he died’ On the whole, the semantic developments in the Germanic languages are rather similar and follow the same paths; consequently, the inventories of particles in the various Germanic languages show a considerable, nonaccidental etymological and semantic overlap. Moreover, there is a kind of competition between particle verbs and prefixed verbs (i.e. verbs with ‘inseparable prefixes’), which in all the Germanic languages represent an older type. At closer inspection, though, this older type turns out to be the likely result of even older verb-particle combinations with the particles in preverbal position, i.e. typical instances of Indo-European preverbs in a position predictable from the basic verborder in Germanic. Considering the high probability of OV as the basic word order in Proto-Germanic this is, again, not particularly surprising. In some of the Germanic languages there is a strong tendency for this older type to recede, most notably in the Scandinavian languages, while this tendency is considerably weaker in the Continental West Germanic languages. In English, a further complication lies in the abundant influx of borrowed verbs from French and Latin in the centuries following the Norman Conquest. Quite a few of these are partly synonymous with particle verbs, while the new verbal prefixes borrowed into the language via these loanwords have traditionally been taken to seal the demise of many of the older inherited prefixes. This has been a remarkably confused discussion for a


7

long time. But nevertheless it is worth being spelt out in some detail, since it is closely connected to the question of the lexical status of the English verb-particle construction from Old English to present-day English, in particular with regard to its stylistic classification. In this context etymological considerations come into play and one must seek an answer to the question why the majority of English verb-particle constructions contain monosyllabic verbs of Germanic descent, while there seem to be restrictions on the use of borrowed and/or polysyllabic verbs. At this point a brief note on the use of period labels in this study may be appropriate. The periodisation of languages into ‘old’, ‘middle’ and ‘new’ periods, or even more subtle distinctions, is a tricky issue (cf. Lass 2000). For English the older threefold division has now been largely replaced by a fourfold division into Old English, Middle English, Early Modern English, Late Modern English. Usage as to the extent of these periods may differ considerably. This is particularly true with respect to the distinction between Early Modern English and Late Modern English (the most recently introduced term), which has been drawn in the literature variously between the end of the 17th and the beginning of the 19th century. Likewise, there is no consensus when ‘present-day English’ begins and whether or not it is part of Late Modern English (cf. e.g. Görlach 1994: 8 and Beal 2004: xi– xiv). But the traditional distinction between Old and Middle English is also problematic, despite the seemingly straightforward extra-linguistic watershed marked by the Norman Conquest (cf. Lutz 2002a). In the present study the periodization is as follows: ‘Old English’ (all English texts before 1100), ‘Middle English’ (up to 1500, subdivided into early Middle English up to 1340 and late Middle English after 1340, cf. Horobin & Smith 2002), ‘Early Modern English’ (up to 1700), ‘Late Modern English’ (up to 1945, following Beal 2004, and thus excluding present-day English). Where finer distinctions are not necessary, ‘medieval English’ will be used as a cover term for Old and Middle English, while ‘Modern English’ is used as a broad cover term for post-medieval English (as opposed to either medieval English or to other Germanic languages). This distinction seems useful in particular with respect to the history of the phrasal verbs, where most research is concerned with either the medieval or the modern period, but rarely with both and where, moreover, the research issues in the former differ markedly from those in the latter.

8

Introduction

1.4.

Structure of this study

The study is organized as follows. Chapter 2 is devoted to a discussion of particle verbs in present-day English and in other Germanic languages and serves as a point of reference for the ensuing historical and cross-linguistic explorations. First the basic semantic and syntactic characteristics of phrasal verbs in present-day English are outlined and criteria for distinguishing phrasal verbs from other verbal structures are established. Also, a number of traditional characterizations of the phrasal verb will be called into question. In a contrastive survey of the other contemporary Germanic languages it is then investigated to what extent the English phrasal verb is structurally remarkable from a synchronic perspective, and what systematic parallels there are across the modern Germanic languages. In the remainder of the chapter arguments are put forward in favour of analysing phrasal verbs as periphrastic word formations in a constructional framework. Chapter 3 is concerned with the history of the Modern English postposed particles. A cross-linguistic and comparative discussion of preverbs reveals that the Old English prefix verbs typically represent a more advanced stage in the development of complex predicates, where formerly independent preverbal particles have fused with verbal stems, while the Old English ‘separable prefix verbs’ represent a younger and less advanced stage of what is essentially the same path of development. The remainder of the chapter contains to a discussion of changes in English word order and their connection to the decline of particles in preverbal position. Chapter 4 critically examines the ways in which the history of the phrasal verbs has been portrayed in the English research tradition of the 20th century, where many traditional beliefs about the construction have hardly ever been put to the test. I will also examine the connections of this to the treatment of particle verbs in the historical dictionaries, which can be shown to provide insufficient and often misleading coverage. In Chapter 5 the interplay of phonological, morphological and semantic reasons for the loss of native prefixes in medieval English is explored and once again comparative evidence from other Germanic languages is taken into consideration. For a satisfactory and coherent account of the history of the phrasal verbs, it will be necessary to analyse their development in relation to the fate of the native prefixes and in relation to the transfer and integration of borrowed lexical material in the centuries after the Norman Conquest. Building on the discussion of preverbation in Chapter 3, it is shown that prefix verbs and particle verbs tend to develop along functionally equivalent lines. In this context the vexed question is discussed to what extent the older prefixes were replaced by particles and borrowed pre-

Structure of this study

9

fixes respectively, and how the characteristic etymological and semantic properties of the Modern English phrasal verbs can be explained. In Chapter 6, the historical reasons for the common perception of phrasal verbs as particularly ‘English’, ‘colloquial’ and ‘informal’ are explored. The first part of the chapter offers a critical survey of quantitative approaches to the development of phrasal verbs in the Modern English period. In the remainder of the chapter it is demonstrated that up to the end of the 18th century attitudes towards the phrasal verb are neutral, and this observation is confirmed by a discussion of a number of 17th- and 18thcentury texts. Consequently, I will argue that the common perception of the construction type is rooted in the English normative tradition in a more complex way than has hitherto been assumed. Chapter 7 provides a summary of results and of open issues, and an outlook to future research.

Chapter 2 Present-day English and other Germanic languages The present chapter provides a descriptive reference point for the historical analysis of phrasal verbs. The focus will be on their central semantic, syntactic and stylistic properties in present-day English and on their overall role in the lexicon, and it will be argued that phrasal verbs are best treated as periphrastic word formations, particularly if cross-linguistic evidence is taken into account, as a brief contrastive examination of similar, and indeed cognate, constructions in other present-day Germanic languages will show. 2.1.

Phrasal verbs: some examples

Phrasal verbs are made up of two components: a verb and a particle which is typically homonymous with an adverb or a preposition (for significant modifications of this preliminary characterisation, see Section 2.6 below). In present-day English, phrasal verbs show a number of distinctive semantic, syntactic and prosodic characteristics. Some of these characteristics can be seen in the following examples (all from the BNC): (1)

He nearly gave up, not knowing what to do next.BM0_104

(2)

I gave up the job there and then and headed for Brazil.APC_2470

(3)

I can see why you gave the job up!HYU_264

(4)

I thought that I was being stupid, so I gave it up.B0U_368

(5)

Eventually he gave it up, stood up and put on his hat.B0U_368

(6)

Beneficial insects such as ladybirds, horseflies and lacewings are encouraged, not killed off, so that they eat up harmful aphids.BN4_1763

(7)

If we eat out my favourite meal is oysters and caviar followed by asparagus with melted butter.CEK_4661

(8)

This other bloke came in and joined ‘im.CKE_1916

(9)

The hours burned by on the green screen, and when they finally flopped in Lucy’s living room, sipping brandy, a key turned in the lock and in came the charmingly dishevelled one and only son.A0L_1418

Phrasal verbs: some examples

11

(10) If he’d been going to hang up on me, something made him change his mind and could be it was something I said.FAP_3057 Examples (1)–(4) contain the phrasal verb give up in different syntactic configurations. It may be used intransitively as in (1), or transitively as in (2)–(5). In the transitive constructions, the object may follow the particle as in (2) or precede it as in (3); pronominal objects, however, usually precede the particle as in (4) and (5). Semantically, give up is clearly idiomatic, since its meaning cannot be inferred from the individual lexical meanings of its components give and up. However, as the other examples show, the degree of idiomaticity may vary – compare give up in (5) with stand up and put on in the same sentence and with kill off and eat up in (6) and eat out in (7). In example (8) at last, come in is entirely compositional from the meaning of the simple verb come and the directional particle in. Directional particles like in in (8) can also occur in sentence-initial position, as the inversion construction in (9) shows. Finally, verb and particle may be followed by a preposition and form more complex constructions, the ‘phrasal-prepositional verbs’, as in (10). Phrasal word formation is a productive process in English. But while the number of verbal elements in the construction is unrestricted, the number of particles is rather small, and all of them are homonymous with prepositions or with spatial adverbs. A list for present-day English would at least include the following particles: aback, aboard, about, above, across, after, ahead, along, apart, around, ashore, aside, astray, asunder, away, back, behind, by, down, forth, forward(s), home, in, off, on, out, over, past, round, through, to, under, up (cf. e.g. the lists in Cowie & Mackin 1975, McArthur & Atkins 1992 or Quirk et al. 1985: § 16.2). But not every phrasal verb will show all of these typical characteristics exemplified so far, and not all of the characteristics are observable in the above examples, as a more systematic look at the English phrasal verb will show. 2.2.

Semantic characteristics

Verb-particle combinations may be highly polysemous; quite commonly, the meanings range on a cline from purely compositional to highly idiomatic: (11) We’d better take in the children’s toys. (12) They supplement their income by taking in students.

12

Present-day English and other Germanic languages

(13) I’ve taken in your trousers, because they were too loose. (14) Grammar takes in syntax and morphology but not phonology. (15) I thought we might take in a show after dinner. (16) I was too tired to take in what she was saying. (17) I’m not surprised he was taken in: he’s as gullible as a child. These examples from Huddleston & Pullum (2002: 284) are arranged according to their relative compositionality. Such clines are commonly the result of linguistic change, with older and more recent forms continuing to co-exist. A direct comparison of take in ‘carry inside’ in (11) and take in ‘deceive’ in (17) shows the well-known development from concrete to abstract meaning; cf. the further discussion of lexicalization processes below. Not all phrasal verbs show the full range of idiomaticity as take in in these examples – some will be purely compositional in all uses while others will appear as non-compositional combinations only. These can be assumed to have undergone a lexical development from compositional to non-compositional, with the earlier, compositional meanings lost, while the later non-compositional meanings fossilize (a more detailed treatment of this topic will be provided in the historical discussion in Chapters 3 and 5 below). In another widespread type of phrasal verbs, the particle may function as an aspectualizer, e.g.: (18) Newcomen got round this difficulty by fitting a leather skirt on top of the piston, this being kept supple by filling it up with water.EED_299 (BNC) (19) While Charles listened to all this good advice, he drank up his glass of champagne and felt a bit better. ACE_2531 (BNC) (20) Abraham talked on, not noticing her lack of attention.GW8_176 (BNC) (21) She chatted away, her hands illustrating her words.CDX_1644 (BNC) At first sight the different semantic types exemplified here may seem rather random and disordered, but they can be divided into three major semantic categories.


13

2.2.1. Three semantic types Although it is not always possible to draw clear-cut distinctions, it seems useful to distinguish between three, albeit somewhat idealised, semantic types of phrasal verbs, which have been characterized as ‘literal’, ‘aspectual’ and ‘non-compositional’ (or ‘idiomatic’); cf. the tripartite semantic division in König (1973: Ch. 9.4), who takes up comparable earlier classifications by Bolinger (1971), Fraser (1965 and 1966), Live (1965) and Makkai (1972) but adds the caveat that it is often difficult to distinguish between adverbial, aspectualizing and idiomatic uses (König 1973: 98). However, this threefold categorization is better subdivided as shown in Figure 2-1, since both the combinations with a directional and the combinations with an aspectualizing particle are semantically compositional and contrast with the non-compositional combinations whose meanings cannot be inferred from their parts. In keeping with much of the literature, the compositional type with directional particles will here be simply referred to as ‘compositional’, while the compositional type with aspectual particle will be referred to as ‘aspectual’. With the non-compositional combinations, it is by definition not possible to assign particular meanings to the particles. In the following sections, each of these types will be dealt with in some more detail and thereby (if only implicitly) treated as if they were separate categories. At closer inspection, though, it turns out that both the compositional vs. noncompositional and the directional vs. aspectual particles are more properly to be seen along clines reflecting both their synchronic meaning and their diachronic development (for a comprehensive discussion of particle semantics in present-day English, see Cappelle 2005: Ch. 8; cf. also Dirven 2001). verb-particle combination

compositional

directional particle

aspectual particle

Figure 2-1. Semantic classification of phrasal verbs

non-compositional

14


Another well-known semantic classification, proposed by Bolinger (1971), will not be used here (for a discussion of Bolinger’s model, see Hampe 2002). In his discussion of ‘stereotyping’, Bolinger makes a twofold distinction: (i) between ‘first-level metaphor’, where the literal (‘adverbial’) meaning of the particle is changed, and ‘second-level metaphor’, where the meaning of the whole phrasal verb is non-literal, and (ii) between ‘first-level stereotype’, where the meaning of the combination is additive, and ‘second-level stereotype’, where the meaning of the combination cannot be inferred from the meaning of its parts (Bolinger 1971: 113– 114). Bolinger’s model, which has not met with widespread acceptance, has a number of inherent problems. These include his choice of terms (e.g. it is usually metonymy rather than metaphor that plays a role in the semantic changes in question, cf. the discussion in Chapter 5 below) but also his choice of categories (which e.g. explicitly include figurative use, a process not specific to phrasal verbs, but not the aspectual combinations). 2.2.2. Compositional constructions In the compositional constructions, the verb combines with a directional particle and the whole construction is transparent from the meaning of its constituents, e.g.: (22) Well it reminds me when I was in a shop on the High Street for many years and a little boy and girl came in with a, with an Alsatian dog, a puppy.KM3_748 (BNC) (23) Fold forward and remove the four bolts which go into the floor and carry the seat out.AN2_1732 (BNC) The formation of such compositional constructions is a process whose productivity can be illustrated by the exchangeability of verbs and particles (cf. the similar examples discussed in Jackendoff 2002: 74): (24) George

tossed took put carried threw

the food

up. in. away. back. out.

In such syntagms, the paradigmatic insertion of any verb and any particle seems possible, as long as the combination of verb and particle allows an


15

interpretation of motion through space, with the particle expressing the direction and the verb expressing the kind of the verbal action. Characteristically, the directional particles in the compositional constructions can be replaced by directional prepositional phrases: (25) George carried the food into the house. The directional particles can be fronted, cf. the inverted order in (26) with the particle in sentence-initial position. Again, directional prepositional phrases can also occur in this position, cf. (27). (26) Then the door opened, and in came Felix, Sophie, and Agatha.H8G_451 (BNC) (27) Into the shop came a young and very hot couple, leaving their bicycles outside.H9Y_83 (BNC) Particles preceding the object can never be replaced in such a way, though. Thus George carried in the food is possible but *George carried into the house the food is impossible. It deserves to be pointed out that only the compositional combinations show the full range of syntactic properties typical of phrasal verbs, and that the restriction of the syntactic possibilities clearly goes along with different semantic properties. This is why an exclusion of the compositional combinations from the phrasal verbs seems problematic. Quirk et al. (1985: § 16.2 et passim), who would seem to advocate such an approach, do not provide any kind of explanation why the syntactic properties they list are also (and only) possible with other kinds of combinations of verb and particle. In fact, by excluding the compositional combinations from the phrasal verbs by definition (since phrasal verbs are ‘multi-word verbs’ and these in turn are defined as ‘idiomatic’), Quirk et al. (1985) somehow fail to provide an account of their syntactic properties at all, which they do not discuss elsewhere either. Cf. the rather self-contradictory discussion by one of the co-authors of the grammar: Multi-word verbs are combinations of verbs with other words that form an idiomatic unit, inasmuch as the meaning of the combination cannot be predicted from the meaning of the parts … In free combinations [treated as a sub-category of phrasal verbs, which in their turn are treated as a sub-category of multi-word verbs, ST], the verbs and the particles are both transparent in meaning. (Greenbaum 2000: § 11.18)

16


The problem here is connected to the dubious category ‘multi-word verb’ (cf. Section 2.4.5 below). 2.2.3. Aspectual constructions As is well known, the enormous literature on aspect and aktionsart is characterised by considerable terminological confusion (cf. e.g. Brinton 1988: Ch. 1 and Bußmann 2008: s.vv., and the references there; see also Comrie 1976, the contributions in Vetters & Vandeweghe 1991 and Binnick 2001). Kortmann (1991) suggests drawing the distinction along the following lines: ASPECT: grammatical category; non-deictic; concerned with situation-internal time; presentation of some situation as incomplete/in progress/existent (‘from within’) or complete (‘from without’) at a given point/period in time; AKTIONSART: lexical category; non-deictic; concerned with situation-internal time; temporal constitution inherent in the meaning of the verb (whether simplex, complex, or verbal syntagm) or predicate. (Kortmann 1991: 19)

Here these definitions are adopted, while the term aspectualizer is used as a cover term for both aspect and aktionsart marking, not least since the observation that the difference between aspect and aktionsart is one of grammatical vs. lexical coding along a (synchronic and diachronic) cline appears to be very much in favour of Sasse’s proposal to altogether abandon the term aktionsart. This would also seem to fit in with the constructional approach taken here (cf. Section 2.6.4 below): Aspectuality is always a matter of the correlation of lexical semantics and TAM [viz. tense–aspect–mood] categories, and can be ordered along a continuum from zero lexical and maximal grammatical distinctions to maximal lexical and zero grammatical distinctions. (Sasse 1991: 44)

Consequently the use of aktionsart should be seen as a mere shorthand for ‘lexical aspect’ (for a comprehensive discussion of the aspectual impact of the particles, see Cappelle 2005: Ch. 8). Aspectual constructions might be treated as a sub-group of the compositional constructions, since their meaning is usually fully transparent and readily understandable ad hoc formations are possible, e.g.: (28) And having another baby to use the clothes up seems a little extravagant.K4P_1490 (BNC)


17

But the particles in these constructions are not directional but aspectual and they typically mark telic aktionsart, as shown by Brinton (1985), cf. e.g. (29) and (30): (29) He used our supplies. (30) He used our supplies up. Clearly, the difference between (29) and (30) with the added particle up is that in the second sentence the verbal event of the first sentence is presented as directed towards a final stage that is not expressed by the simple verb (although completion may in principle be part of the meaning of a simple verb, e.g. He finished our supplies), i.e. the particle introduces “the concept of a goal or an endpoint to durative situations which otherwise have no necessary terminus” (Brinton 1985: 160). Thus completely in (31) only serves to intensify the particle in a largely synonymous sentence (cf. Peters 1993 on intensification): (31) He used our supplies completely up. An apparently redundant use of aspectualizing particles is possible, as in (32) or (33), where the aktionsart meaning of the particle is already present in the simple verb: (32) Chico finished up his drink.HTU_3724 (BNC) (33) … so they didn’t come, and I didn't go out and Gemma came up and I was sitting there talking and they had a drop, drop of wine and I had one with erm and Gemma finished her biscuits up …KC2_1222 (BNC) This is not normally encountered with the directional particles, i.e. pleonastic constructions (as in *He entered the room in) are less common, although not impossible: (34) The hugely distended stomach had to be decompressed before it could be returned back into the abdomen, and the defect in the left hemidiaphragm (6×5 cm) was repaired with 2-;0 silk sutures.FT2_1308 (BNC) Conversely, fronting of the particle is not possible with the aspectualizing particles (e.g. *Up he ate).

18


Quite clearly it is up that is the most central aspectualizer among the particles, both in terms of frequency and of meaning. Besides up, a rather restricted set of other particles can be used as aspectualizer, e.g. down as in the following example: (35) Spitting it on to the floor, he raised the half-empty bottle to his lips and drank down the fiery vodka in great gulps, as if to drown the useless curses which rose in his throat.CDA_602 (BNC) (36) She swallowed down a mouthful of wine.JY9_1827 (BNC) As these examples with down indicate, the distinction between compositional and aspectual may be somewhat blurry, since the particle in (35) and (36) is both directional and aspectual. In fact, this overlap provides the context for the development of aspectualizing particles out of spatial ones; cf. the discussion in Chapter 5 below. The particles out, over and through are also found as telic aspectualizers, e.g.: (37) With that beat I needed a really stomping guitar line to go with it so I worked it through in my head and then worked out the chords on the piano.C9L_2186 (BNC) (38) Bill Murray spent £50,000 on setting up his restaurant at Telegraph Hill, near Exeter, Devon, two years ago but said the business started to go downhill when he handed it over to a manager to run.A0C_167 (BNC) Not all aspectualizing particles are telic, though. The two particles on and along, for example, may function as continuative (i.e. atelic) aspectualizers, e.g.: (39) Abraham talked on, not noticing her lack of attention.GW8_176 (BNC) (40) In the end, Mungo reasoned that the old man had probably been driving along, had somehow caught a glimpse of him, and had taken a short cut from the road.ACV_152 (BNC) Away is near-synonymous to on, too, but in addition it tends to intensify the verbal event, e.g.:


19

(41) Breeze talked away for all she was worth as she cut bread-andbutter in the draughty old kitchen, but she knew that her sister wasn’t really listening.BMU_665 (BNC) To sum up, in the aspectual combinations the particles are used in a semantically transparent way, which may overlap considerably with the ‘literal’, i.e. directional, use of the particles. Another specific characteristic of the aspectual combinations is the occurrence of ‘pleonastic’ combinations, where the aspectual value of the particle is also part of the aktionsart of the verb alone. 2.2.4. Idiomatic constructions Brinton & Traugott (2005: 32) point out that the term lexicalization has been used variously in the literature. In its broadest sense, the term may refer to synchronic word formation processes, more narrowly to fusion with decreasing compositionality but also to processes of separation with increasing autonomy. Here the term will be used in a more narrow sense, but broadly enough to encompass institutionalized uses of phrasal verbs which are not idiomatic, and I will refer only to those lexicalized combinations as idiomatic whose meaning is non-transparent (on idiomatization and lexicalization cf. e.g. Brinton & Traugott 2005 or Bußmann 2008: s.vv. and the references there, and see the discussion of idiomaticity in Chapter 6 below). Thus the idiomatic constructions are different from the two preceding groups in that their meaning cannot be inferred from the meaning of their elements; they belong, quite unambiguously, to the lexicon, as a few examples suffice to show: (42) My husband actually said to me that giving up smoking was easy because he’s done it plenty of times.JJP_385 (BNC) (43) In the following extract we see that an equally offensive act is one in which a soft teacher tries to assert authority, but when challenged gives in.ECN_742 (BNC) (44) Farmers, sailors, and chemists get by perfectly well on the basis of everyday experience, without recourse to Aristotelian logic.ABM_469 (BNC) (45) He could not make it out, nor could he trust his own memory.BNF_1301 (BNC)

20


(46) After she hangs up on Mark, Martha takes a deep breath and dials a London number.HGU_3802 (BNC) For the idiomaticity of a construction there is some syntactic evidence. With the idiomatic constructions, the positional variants typically tend to be more restricted, cf. (45) above and the following example: (47) *… and out he made it. This last property is shared with the aspectual combinations, although the reason for this is more evident in the case of the present category: since no clearly identifiable meaning can be ascribed to the particle, it is unlikely to be focalised in an inversion. Similarly, coordination of either of the elements in idiomatic phrasal verbs usually results in zeugma, as in (48) and (49), while it is common with compositional combinations, as in (50): (48) ??He could not make it out or up. (49) ??He could not make and carry it out. (50) Oh well I think I’ll just play with the tab and make it pop in and out a few times.H61_413 (BNC) Idiomatic phrasal verbs are certainly the type that has attracted most attention, in particular in the more popular literature (e.g. in style guides, etc.) and in the teaching of English as a foreign language (cf. the discussion of this in Section 2.5 and in Chapter 6). But clear-cut boundaries between the three semantic classes are virtually impossible to draw, since very often literal, aspectual and idiomatic combinations are highly contiguous. Moreover, it is worth pointing out that even though non-compositionality may be the prototypical semantic characteristic of the phrasal verbs, it is only in the compositional (transitive) phrasal verbs where the full range of syntactic properties is displayed. 2.3.

Syntactic characteristics

The basic syntactic characteristics of the phrasal verbs are well-known and have been amply described in the literature. The syntactic description in this section draws on the criteria long established for the analysis of phrasal verbs in the classic modern studies on the subject, starting with van Dongen (1919) to Wood (1956), T.F. Mitchell (1958), Live (1965), Fraser (1965,


21

1966, 1970, 1974, 1976), Bolinger (1971), Sroka (1972) and Pelli (1976); for a critical overview of the pre-1970s classifications, see Lipka (1972: Ch. 1 and the relevant section in his bibliography) and Carstensen (1964), who provides a useful concise examination of the earlier literature and discusses the criteria for phrasal verbs established there. More recent approaches operating with small clauses and intransitive prepositions, which challenge the notion of phrasal verbs in English, will be discussed towards the end of the present section. 2.3.1. Transitivity and serialization Phrasal verbs may be intransitive, as in the following examples (cf. Jackendoff 2002: 69–73): (51) Your children will grow up. (52) The whole house blew up. (53) My mother freaked out. The intransitive phrasal verbs given as examples in these sentences clearly belong to the class by virtue of their idiomatic meaning. But intransitive constructions of the kind He went away will also be regarded as phrasal verbs in this study, although viewing these intransitive non-idiomatic cases in isolation one could argue in favour of an analysis as verbs with adverbial complements rather than phrasal verbs, where the particle away would be analysed analogous to to the National Gallery in He went to the National Gallery; cf. König’s (1973: § 9.1.2) examination of the fuzzy distinction between verb particle and directional adverb. Studies of the phrasal verb are often remarkably reluctant to discuss this issue, which is frequently decided on by fiat. But although an exclusion of this type can make sense from a syntactic point of view, a study with a focus on the phrasal verb (and in particular one concerned with the history of the construction) should, to my mind, include this type, also because it provides the diachronic input to the development of aspectual and idiomatic meanings. With transitive phrasal verbs, if the object is a full noun phrase the object may come either before the particle, as in examples (54)–(56), or after the particle,1 as in examples (57)–(59) (i.e. ‘joined’ vs. ‘split’ order in terms 1

Note that this and similar statements are used here only as convenient descriptions of the order of the elements, which are handier to use than, e.g.,

22


of Lohse, Hawkins & Wasow 2004; other current designations include ‘continuous’ vs. ‘discontinuous’ order, but also less readily intelligible distinctions like ‘construction0’ vs. ‘construction1’ used by Gries 2003). (54) I can put out the announcement.G0P_253 (BNC) (55) They never blew up the houses.ANU_761 (BNC) (56) He had left out Peter. (57) I can put the announcement out. (58) They never blew the houses up. (59) He had left Peter out. However, if the object is a pronoun, it will normally precede the particle: (60) I put it out. (61) They blew them up. (62) … he had left him out …H9D_3243 (BNC) (63) *I put out it. (64) *They blew up it. (65) *He had left out him. It seems that the position of the object is connected to a number of factors, including its weight. Thus Quirk et al. (1985: 1154) state that the “particle tends to precede the object if the object is long, or if the intention is that the object should receive end-focus”; cf. the following examples (based on Jackendoff 2002): (66) Lila looked up the answer to the question that was on everyone’s mind.

“the order may be V O prt or V prt O”. They are not meant, however, as a commitment to what constitutes the ‘normal’ or the ‘underlying’ order, let alone an implicit commitment as to whether it is the particle or the object that ‘moves’; i.e. in the present context the statement that the particle may precede the object and the statement that the object may follow the particle are interchangeable.


23

(67) *Lila looked the answer to the question that was on everyone’s mind up. (68) Lila looked it up. (69) *Lila looked up it.2 (70) He left out hím (not hér). The position of the objects in the example (5) above and in similar instances can thus be accounted for in principle; a more precise description of the factors governing the choice between the two positional variants is, however, considerably more intricate in detail and no analysis has so far gained general acceptance. Generative approaches in particular have been very much concerned with accounting for the serialization alternatives typical of the transitive construction and there the issue, being a major touchstone for theory-internal skirmishes, seems far from resolved (cf. e.g. Dehé 2002 or Farrell 2005). Among the detailed studies devoted to the subject of particle placement in present-day English, Dehé (2002) argues that the ‘neutral’ order is ‘verb-particle-object’ and that the choice of one order rather than the other depends on the “news value” (Dehé 2002: 77) of the object, while Gries (2003) describes particle placement as a constructional alternation which depends on a number of variables and which can be explained by a statistical multifactorial analysis of these variables. Gries in particular provides an exhaustive list of the criteria suggested so far in the literature; but cf. also Cappelle (2009), who argues in favour of a free variation in an ‘allostructional’ analysis (see Section 2.6 below). Moreover, prepositional complements may not intervene between the verb and the particle, thus: (71) Sim ran away to the city. (72) Please look out for Harry. (73) Jill grew up into a strong woman. (74) *Sim ran to the city away. (75) *Please look for Harry out. (76) *Jill grew into a strong woman up. 2

This is a possible construct, but not in the intended sense; cf. Section 2.3.2 below.

24


This remains true in nominalisations (cf. Section 2.4 below): phrasal verbs can be nominalised, but in nominalisations, a prepositional complement always follows the particle, thus: The rapid looking up of the information is important or The prompt sending out of the reports is commendable, but not *The rapid looking of the information up is important and *The prompt sending of the reports out is important; “although of-NP is the counterpart of the direct objects in such nominals, it behaves as a PP with respect to particle position, just the way it looks” (Jackendoff 2002: 72). An adverbial modifier like right or completely can precede the particle. However, this is only possible if the particle follows the object (see Jackendoff 2002: 70–73 for more examples and for a more detailed discussion), e.g.: (77) I’ll look the answer right up. (78) Please shut the gas completely off. (79) *I’ll look right up the answer. (80) *Please shut completely off the gas. Other syntactic characteristics are noticeably tied to the semantic makeup of the verb-particle combination and involve fronting of the particle and growing rigidity of word-order. Thus the particle may immediately precede the verb, as in (81); cf. also (9) above. (81) A tap is turned, a pressure of 120 pounds per square inch applied to the piston, and up goes the end of the coach as easily as though it were a doll’s house.B2S_161 (BNC) In such sentences with a fronted particle, the subject follows the verb if it is a full noun phrase. This is one of the noticeable exceptions from subjectverb order in declarative sentences in present-day English. If the subject is a pronoun, though, it will immediately precede the verb: (82)

(and …) up it goes.

But fronting of the particle is of course only possible in compositional combinations with a directional particle. Just as the placing of ‘heavy’ noun phrases is evocative of the Law of Growing Members, the occurrence of other positional variants could be characterized as predictable from Behaghel’s First Law, since in these cases growing semantic compositionality goes along with syntactic adjacency


25

(Behaghel 1909, 1930 and 1932: § 1426). Quite clearly in many cases the syntactic properties depend on the semantics of the phrasal verb. Cappelle (2009) argues that apparently the degree of lexicalization plays a role in the choice between joined and split order, with a tendency for ad hoc formations to appear in split order, while idiomatic combinations show a clear tendency to appear in joined order. But this tendency is mainly a correlation between idiomatic combinations and joined order, while there is no comparably strong correlation between compositional combinations and split order (cf. Lohse, Hawkins & Wasow 2004: 256). On the whole, the positional variability of verb and particle decreases with the degree of idiomatization, both absolutely, as in the case of particle fronting, and relatively, as with respect to the alternative between joined and split order. Some phrasal verbs only allow one sequence of elements, especially in those cases where the phrasal verb and the object form an idiomatic phrase. Thus in give up hope the particle normally precedes hope as in (83), while the acceptability of (84) is highly doubtful; conversely, in laugh one’s head off the particle will be placed in final position as in (85), while the order in (86) is again highly doubtful: (83) By the time I had almost given up hope, a telegram arrived.G3B_1882 (BNC) (84) ??By the time I had almost given hope up, a telegram arrived. (85) By this time Irene was emitting a steady gurgle of contentment, when she wasn’t laughing her head off.FYV_985 (BNC) (86) ??By this time Irene was emitting a steady gurgle of contentment, when she wasn’t laughing off her head. This is confirmed by a search of the BNC, where sequences as in (84) and (86) are not attested (cf. e.g. Quirk et al. 1985: § 16.4n. for more examples). Other idiomatic phrases contain passive phrasal verbs and occur in the passive only, e.g. fed up with something: (87) I’m convinced half of them don’t actually open, they’re just designed as offensive weapons for use by psychopaths fed up with not getting a seat on the tube.HWL_938 (BNC) There is nothing particularly remarkable about this: syntactic restrictions are a well-known property of idioms (cf. e.g. Burger 2007: 16–32), which can be observed here in idioms containing phrasal verbs.

26


2.3.2. Particles and prepositions A distinction between phrasal, prepositional and phrasal-prepositional verbs was first suggested by T.F. Mitchell (1958) and has since found its way into many reference accounts of English; for a fuller description see e.g. Quirk et al. (1985: Ch. 16), Biber et al. (1999: Ch. 5) and Huddleston & Pullum (2002: Ch. 4; but note that the classification there diverges considerably from other reference grammars). The discussion in Quirk et al. (1985: § 16) is somewhat diffuse and fails to provide a coherent syntactic analysis, but it is quite valuable as a repository of examples (cf. the critical evaluation by Mahler 2002, with further references, and the arguments put forward against prepositional verbs by Klotz 2000: 53–60). Here prepositional verbs will not be treated as a sub-category of phrasal verbs. This conflation is sometimes encountered in English linguistics and lexicography, where one can find a distinction between ‘phrasal verbs with prepositions’ (i.e. prepositional verbs) and ‘phrasal verbs with adverbs’ (i.e. phrasal verbs). Nevertheless a comparison of the phrasal verbs with the prepositional verbs serves to illustrate their specific syntactic properties. At first sight the underlined strings in the following sentences may seem to be quite similar: (88) The following day the then Indian Prime Minister, Rajiv Gandhi, called on the King and they exchanged views on bilateral matters.HKS_3008 (BNC) (89) The silence returned, and he switched on the light.CKB_368 (BNC) However, a closer look reveals that the two examples contain constructions which differ considerably. In (88), on is a preposition and may be analysed as part of a prepositional verb call on (but see the caveat below). In (89), on is a particle that belongs to a phrasal verb switch on. The differences are highlighted by a number of well-known contrasts exemplified in Table 2-1. The prepositional verb may be analysed either as She looked at the book or as She looked at the book – some of the permutations in the table above could be taken to support the first analysis, others to support the second. However, in either case (and in the first case it might be advisable to abandon the term ‘prepositional verb’ altogether), at would be analysed as a preposition. But the particle of the phrasal verb is clearly distinct from a preposition, despite similarities as in a. and b.; the positional characteristics illustrated in c. to f. have already been discussed above.


27

Table 2-1. Phrasal verbs and prepositional verbs compared a. b. c. d. e. f. g. h. i.

phrasal verb She looked up the number. (… the number) which she looked up *She looked carefully up the number. She looked the number up. She looked it up. *She looked up it. *(… the number) up which she looked *It was up the number that she looked. Which number did she look úp?

prepositional verb She looked at the book. (… the book) which she looked at She looked carefully at the book. *She looked the book at. *She looked it at. She looked at it. (… the book) at which she looked It was at the book that she looked. Which book did she lóok at?

The differences in g. and h., involving pied-piping and clefting, are noteworthy since they provide further evidence that the particle and the object do not form a prepositional phrase. Prosodic evidence for this is added in i.: prepositions are normally unstressed, except for contrast. On the basis of these differences, tests for the distinction between phrasal and prepositional verbs can be established (see, e.g., Quirk et al. 1985: § 16.16). Thus He turned on his supporters can be read as either containing a phrasal verb ‘excite’ or a prepositional verb ‘attack’ (Quirk et al. 1985: § 16.6 note a.): (90) He túrned ón his suppórters. ‘He excited his supporters.’ (91) He túrned on his suppórters. ‘He attacked his supporters.’ The prosodic differences indicated by the acute accents show that turn on ‘excite’ in example (90) is a phrasal verb, while turn on ‘attack’ in (91) is a prepositional verb; the syntactic properties of each are according to the respective options which are set out in Table 2-1. At this point a brief note on terminology seems appropriate. It has been variously suggested analysing the particles of phrasal verbs as ‘intransitive prepositions’ (cf. e.g. Aarts 1997: Ch. 9), and this notion has found its way into Huddleston & Pullum’s grammar (2002), where one particularly unfortunate choice of both terminology and analysis can be found. The authors first abolish the term ‘phrasal verb’ (as applied to both phrasal and prepositional verbs), pointing out that “they do not form syntactic constituents belonging to the category verb” (Huddleston & Pullum 2002: 274), only to introduce a new category ‘verb-particle-object construction’, viz. transitive compositional phrasal verbs. Intransitive compositional phrasal

28


verbs are categorized by them as ‘verbs with intransitive prepositions’. Both categories may then be lexicalised as ‘verbal idioms containing intransitive prepositions’ (Huddleston & Pullum 2002: Ch. 4, § 6.3), a term which is also meant to include combinations with aspectualizing particles. But these can hardly said to be always lexicalized and the analysis of their particles as prepositions would be in need of further explanation. The whole analysis follows from the authors’ determination to apply the concept of ‘intransitive preposition’ to the analysis of English (Ch. 7, § 2.4), which in this case at least results in an extremely unwieldy classification. But such analyses entirely fail to account for the syntactic differences exemplified in the present section; cf. e.g. Standop’s (1999: 253) critique of Aarts’ proposals (1997: Ch. 9). The most detailed discussion of the issue is by Cappelle (2005; see also 2004), who reaches the following conclusion: “(i) particles have different distributional properties from PPs, (ii) they cannot be simply analysed as reduced PPs, and (iii) they do not always have the same meaning as formally related directional PPs. Calling them ‘prepositions’ blinds us to these facts” (Cappelle 2005: 101). All in all there is little benefit to be derived from replacing the term particle by intransitive preposition, especially from a historical point of view (cf. the discussion in Chapter 3 below) and consequently the term will not be used in the following. Similarly, approaches will be discounted here which further do away with phrasal verbs by postulating the existence of small clauses (as in den Dikken 1995 and Svenonius 1994, besides Aarts 1989 and 1992); cf. the arguments against a small clause analysis in Booij (1990). 2.3.3. Phrasal-prepositional verbs Phrasal-prepositional verbs are phrasal verbs with prepositional complements. Normally the term is reserved for constructions showing a certain degree of idiomaticity, e.g.: (92) English winemakers are always having to put up with this kind of jibe, despite possessing a viticultural heritage which dates back to the Romans.ECT_3201 (BNC) (93) Ruth had called to see Mrs Johnson, looked in on one or two of the other neighbours, and was halfway up the hill to the churchyard before Maria caught up with her.CB5_1015 (BNC)


29

Like the simple prepositional verbs, the phrasal-prepositional verbs can form a passive (although the acceptability of sentences containing such passives may vary): (94) On the other hand, public relations is often looked down on by the media as messengers with stories that hold no interest and which will be presented in the wrong way at the wrong time and in the wrong situation.EVF_1281 (BNC) (95) There is, however, a widespread opinion that this is not the case, and that illness in old age is ‘just one of those things to be put up with’.B01_688 (BNC) In prepositional passives, the subject corresponds to the ‘prepositional object’ in the active sentence (cf. Quirk et al. 1985: § 16.9 and Huddleston & Pullum 2002: § 6.3.2) and phrasal-prepositional verbs will normally show preposition stranding as in (94) and (95). This is also true of relative constructions: (96) He said it as if books were some terrible bore which everybody had to put up with but which weren’t of the slightest use.B0U_1411 (BNC) Like the simple phrasal verbs, some phrasal-prepositional verbs take a direct object, e.g.: (97) Could the real purpose of Sandra’s visit to the doctor’s be to fix her up with contraception?ANY_196 (BNC) (98) So let me in on the secret.G1S_346 (BNC) These transitive phrasal-prepositional verbs cannot occur in the prepositional passive (cf. Huddleston & Pullum 2002: § 6.3.2), but passives where the direct object in the active sentence corresponds to the subject in the passive sentence are possible, as in (99); in (100) the corresponding active construction is given, in (101) an example attested in the BNC: (99) John was astounded to learn Carol had entered him in the competition only mum Linda was let in on the secret.K4C_724 (BNC) (100) (… she) had let mum Linda in on the secret. (101) There was no way I was going to let my children in on this one.CB8_1127 (BNC)

30


2.4.

Further observations

This section concludes the basic descriptive overview of the present-day English phrasal verb and points to a number of additional peculiarities which will be relevant in the further discussion. 2.4.1. Cranberry verbs Sometimes the verb element of a phrasal verb occurs only as part of that particular phrasal verb, i.e. there is no corresponding simple verb, e.g.: (102) To finish the planting scheme, there are tiny species which can eke out an existence in a teaspoonful of soil.ACY_487 (BNC) For these (cf. also e.g. mete out or fob off) the term cranberry verbs, in allusion to the morphological term cranberry morphs (cf. e.g. Fleischer 2000: 893), might be appropriate. Such lexicalised cranberry verbs are similar to a different type of phrasal verbs, which, however, are typically fully transparent, e.g.: (103) At the moment, there is no suggestion, no-one can provide any evidence that Andrew Gilligan has sexed up the story. () Cf. Section 2.6.1 below for a more detailed discussion of this particular type of phrasal verb. But although eke out and sex up are similar to each other with respect to the fact that there are no corresponding simple verbs *to eke or *to sex (although there is a simple verb sex, this is unlikely to be the basis of sex up; cf. OED s.v.), they are otherwise different. This difference can be seen in the semantic make-up of such verbs: while the first type is always lexicalized and semantically opaque, the second type is just the opposite, since it is typically a transparent derivation. 2.4.2. Nominalisations Phrasal verbs can be nominalised, but some of these nominalisations seem quite irregular compared to other word formations in English, and some are of limited productivity. These are the options (for a phrasal verb V prt with V ‘verb’ and prt ‘particle’):


V prt V prter Ver prt Ver prter Ver prterer prtV prtVer Ving prt

31

e.g. drop-in e.g. drop-inner e.g. dropper-in e.g. dropper-inner e.g. dropper-innerer e.g. outbreak e.g. bystander e.g. dropping in

The formations with -er are, like other deverbal -er suffixations, typically (but not necessarily) agent nouns, while the suffixless formations are typically (but again not necessarily) action nouns; that is to say, the semantics of these formations are entirely in line with the semantics of other (nonphrasal) deverbal nominalisations in English. The morphological analysis of these patterns is open to dispute, and some or even all of them would not be regarded as nominalisations of phrasal verbs by some authors; cf. the discussion in Schmid (2005: 128–130) and see Marchand (1969) for a discussion of the various types (which however is scattered throughout the handbook: §§ 2.35–2.48 ‘combinations with locative particles as first elements’, § 2.49.6 ‘type looker-on’, §4.30 ‘-er suffixation’, § 5.9.1 ‘type showoff’, § 5.9.2 ‘type blackout’). The various positions of the suffix -er in the formation of agent nouns is particularly remarkable; but it is curious that despite the proliferative literature on the phrasal verbs this has rarely been the subject of specialized studies. Indeed, so far no one has investigated in detail the factors influencing the choice of one of these constructions over the others, nor are the reasons for the existence of these constructional alternatives entirely clear (but cf. Chapman 2008 and Denison 2008). Bauer (1983: 288–289) summarizes the issue as follows: There is great difficulty in forming -er subject nominalizations from phrasal verbs, and there are five competing patterns. The first is to add -er to the verbal base before the particle, as in dropper-in, finder-out, seer-off. The second, a very rare one, is to add -er to the particle, as in come-outer. The third is to prepose the particle and add the -er to the verb, as in onlooker … The fourth is to add the -er to both the verb and the particle, as in breaker-inner, cleaner-upper. The fifth is to omit the particle completely, as in waiter (from to wait on somebody) … Of these, the first and fourth tend to feel very clumsy, and as a result tend to be used mainly in colloquial speech; the second, as I said, is very rare, because the suffix seems to be added in the wrong place; the third and the fifth both lose the coherence of the verb + particle unit: in particular the fifth would probably not be connected with the phrasal verb [I have therefore not included

32

Present-day English and other Germanic languages it in the list above; cf. the discussion in Section 2.3.2, ST]. Both the awkwardness of the patterns and the fact that there is such a wide choice of patterns provide excellent reasons for avoiding the problem if possible, and the conversion option allows the whole problem to be avoided. One result is that the converted form drop-in sounds less colloquial than the -er form dropper-in.

As Bauer’s discussion shows, this topic deserves closer examination. Moreover, in very many cases speakers do not seem to feel comfortable with any of the alternatives involving -er, as the following example from an internet chat on the subject of dropping in on people shows: (104) I think cel [sic] phones have made dropping in obsolete. I always call first as do all my friends. Even the one friend I’ve had for ages who has always been a dropper-in (drop-inner? dropper-inner?) now calls first. () Perhaps this kind of insecurity reflects the fact that phrasal verbs are at an interface3 of syntax and morphology, and that they seem to constitute constructional units: with other verbs (simple and compound), suffixes of deverbal nominalisation are attached4 to the verbal stem, preceding the inflectional ending, and the normal order is ‘lexical stem – derivational suffix – inflectional suffix’. Thus each such -er-suffixation creates an irregularity: ‘v-er prt’ places the derivational suffix in an infixal positional highly uncommon in English (in a purely syntactic analysis, this is, of course, not an irregularity; but then the alternative positions of the suffix would remain unaccountable); ‘v prt-er’ does not attach the deverbal suffix to the verbal stem; in ‘v-er prt-er’, the suffix is both behind the verbal stem and at the end of the whole verb (i.e. regular), but it also combines both irregularities,

3 4

I take the notion of interface to be merely a convenient metaphor; cf. the discussion below. Here and elsewhere this term is meant to refer to linguistic structure and not to processes, in line with Haspelmath’s (2002: 21) caveat: “it is often convenient to describe morphological patterns as if they were the results of events … Linguists use such process terms very often. They talk about elements ‘being affixed’ to bases, about a complex word ‘being derived’ from (i.e. built on the basis of) a simpler one, or about one affix ‘replacing’ another one. It is important to keep in mind that these process terms are purely metaphorical, and that they do not refer to any actual events or processes”.


33

plus an irregular double suffixation. This is even more so the case with triple -er suffixation, e.g.: (105) I give up on so many things when they’re not going my way. I am a giver upperer in the worst way. (http://birdynumnum.livejournal.com, entry for 2008-04-21) The other types of nominalisations are likewise subject to restrictions. The pattern ‘v prt’ is now restricted to action nouns, where it is quite productive. Otherwise the use of particles in pre-position occurs mainly in lexicalised formations and in ad hoc formations on the basis of compositional (directional) phrasal verbs, while other present-day English ad hoc formations will typically show a particle in post-position, as in example (105); cf. also Section 2.4.4 below. In the highly regular derivations in ‘-ing’ and ‘-ed’, the suffix is always attached to the verbal element, e.g.: (106) This engagement, in the commencement of his second year at the bar, and the dropping in of occasional fees, must have raised his hopes; and he now abandoned the scheme of becoming a provincial barrister. () (107) The dropped-in classes then provide any required extra capabilities and help fulfill at least the first two requirements listed earlier. () (108) Believe it or not this is the broken up lathe bed in this wheelbarrel. () But as example (108) shows it is clearly the past participle form of the phrasal verb that provides the basis for such word formations, and not a distinct and fully lexical suffix -ed (in that case we would expect *breaked up). If the -ing-formations are taken to be similarly based on the -ing form, then the position of the suffix with -ing- and -ed- formations is not particularly noteworthy and forms like *drop-inning are not to be expected.

34


2.4.3. Other word formations Phrasal verbs can also be involved in word formations with -able (and -ability), but again it seems that speakers feel slightly uncomfortable with such word formations, e.g.: (109) Going to miss him so much, but we’re pretty much unbreakupable (haha) by this point so we’re sweet. () (110) What you got with the uber was a sagging, hard to solo on, hard to cut in a mix (because it has no high mids which is what marshall is all about), and unbreakupable (new word) bland clean channel. () But unlike noun formations, such adjective formations never seem to show final position of the particle, i.e. ??breakableup is quite unlikely. In fact, not even the depths of the world wide web yielded a single instance of this and similar constructions. 2.4.4. Prefix verbs and phrasal verbs The relationship of prefix verbs and phrasal verbs plays a major role in the discussion of the history of the phrasal verbs (cf. the following chapters). In present-day English verbal prefixation and the formation of phrasal verbs are in general separate processes. However, it is worth pointing out that a number of verbal prefixes are homonymous with particles of phrasal verbs, and in a few instances the prefix verb and the phrasal verb are synonymous, e.g.: (111) His remarks, downplayed by the UK Foreign and Commonwealth Office, provoked an uproar among UK right-wing Conservative politicians.HLG_1148 (BNC) (112) In July 1985 two serious disturbances occurred in Handsworth, but both were played down and went unreported in the media.AS6_303 (BNC) Very often, however, the prefix verbs will not be replaceable by corresponding phrasal verbs and vice versa, e.g.:


35

(113) We’ve just been overcharged for this sandwich.K26_544 (BNC) (*We’ve just been charged over for this sandwich.) (114) The words came over clearly but cautiously.BNN_60 (BNC) (*The words overcame clearly but cautiously.) Note that (111) and (113) are structurally completely equivalent, also with regard to stress (cf. LPD s.vv. downplay and overcharge, where both are shown to carry primary stress on the second and secondary stress on the first element). It has been argued that verbs labelled ‘prefix verbs’ in the present section are instances of verbal compounding (cf. e.g. Marchand 1969: 108– 121). Adams (1973: 113–127) treats such formations in a chapter on ‘compounds containing particles’, but then explicitly labels out- a prefix, while the status of some elements discussed in that context remains somewhat vague (see also the more detailed discussion in Schmid 2005: 128–130). Since the first elements in these formations are often semantically clearly and systematically distinct from the homonymous adverbs and since they do not normally carry primary stress, they will be regarded here as prefixes – although a clear distinction between compounding and affixation is notoriously difficult to draw (cf. e.g. ten Hacken 2000). Marchand (1969: 108) claims that on the whole these prefix verbs are not a productive class in present-day English any longer: “With verbs, combinations are no longer freely possible”. But Marchand’s discussion is not free from inconsistencies. It is hard to see why he treats out-, over- and under- separately as elements of compound verbs (1969: 96–100) while these elements are then again treated besides down-, in- and up- as instances of what he classifies as combinations with locative particles as first elements (the other particle elements after-, back-, by-, forth-, off-, on- and through- do not combine productively with verbs in present-day English, cf. Marchand 1969: 108– 121). Formations with down-, out-, over- and under- are clearly productive (e.g. downplay, outclass, overcharge, underestimate) and such prefixes are semantically distinct from the homonymous adverbs, but not necessarily from the particles of phrasal verbs, as examples (111) and (112) show (cf. the more detailed discussion of their productivity by Scheible 2005 and Schröder 2008). Somewhat confusingly, Marchand characterizes the ‘preparticles’ in his ‘compound verbs’ as follows: Derivatively [!] productive verbal particles, however, are not used with any of the meanings the particles have as independent adverbs … This attitude of the language towards full words as preverbs is probably one of the reasons why the types underline and overshadow, where the particle has a locative meaning

36

Present-day English and other Germanic languages (weakened as it is) have long ceased to be productive. Another reason is that locative adverbs have come to be placed after the verb. Even the weakened locative meaning of the preceding types seems to have made the particle undesirable for verbal composition. Such a reduction of the full word value brings the locative particles nearer to prefixes with which they also have the stress pattern in common. (Marchand 1969: 100)

Bauer (1983) cites outachieve, overachieve, overbook, overeducate and overmark as “[r]ecent examples” (Bauer 1983: 208–209). Adams (2001: 100–109), however, claims that although the group of verbs like, e.g., download, upgrade etc. looks deceptively similar, these word formations are probably best regarded as denominal conversions rather than prefix or compound formations. Verbal prefixation with up- is decidedly archaic and does not really seem to exist outside lexicalized formations (e.g. uphold, uproot, upset), although many formations of this type seem to date from as late as the 19th century (e.g. upbreak, uplift, uptake; cf. Marchand 1969: 121). On the whole, however, the formation of prefix verbs with prefixes homonymous to the particles of the phrasal verbs is a relatively marginal process in present-day English, especially when compared to the formation of phrasal verbs, although this finding is easily obscured by the considerable number of lexicalized instances of such prefix verbs. 2.4.5. ‘Group-verbs’ etc. It has been variously suggested that the phrasal verbs should be analysed as a subset of a larger class labelled ‘group verbs’ or ‘multi-word verbs’ (classic proposals include Cowie & Mackin 1975 and Vestergaard 1977). Although it is certainly sensible to study the development of the phrasal verbs in regard to the development of various kinds of composite predicates in English,5 it is worth pointing out that the complex verbal constructions

5

Brinton & Akimoto (1999: 2) distinguish between two main types of composite predicates, namely phrasal verbs and complex verbs, which are characterized by “a tripartite structure consisting of a verb of general actional meaning …, the indefinite article and a deverbative noun”. It is often observed (following Bolinger 1971: 45) that what these constructions tend to have in common is their analyticity and semantic ‘spreading’, i.e. both syntax and semantics are unpacked from one into two or more words (cf. e.g. Brinton 1996, where this is identified as ‘increasing segmentalization’). For a


37

Table 2-2. Group-verbs in present-day English (according to Denison 1998: 222) [Od ‘direct object’, Oprep ‘prepositional object’, 2nd prt ‘second particle’] Od –

Oprep –

2

Label intransitive phrasal verb transitive phrasal verb

+

–

–

3

prepositional verb

–

+

–

4

phrasal-prepositional verb

–

+

+

5

+

+

+

6

+

+

–

7

–

–

+

8

+

–

+

1

2nd prt Examples – EAT out, WISE up CLEAN sth. out, MESS sth. up INSIST on sth., DEAL with sth. HANG up on sb., GET away with sth. TAKE sth. out on sb., PUT sth. over on sb. TAKE sb. for sth., SUSPECT sb. of sth. (COME on over, GET back in) GET sth. over with, (READ sth. back out)

variously referred to as ‘group-verbs’, ‘complex verbs’, ‘multi-word-verbs’ etc. constitute a syntactically, semantically and stylistically heterogeneous group whose members do not have more in common than the fact that they may be seen as verbal constructs consisting of more than one word (in fact, for the complex predicates it is indeed not even necessary for any of the elements to be a verb). This is reflected in the fact that there is no consensus as to what ought to be included among such a group of verbal constructions. Let us therefore have a quick look at some classifications of group-verbs and then turn to the reasons why phrasal verbs will not be treated as a subset of them in the present study. E.g., Denison (1998: 221) defines group-verbs as “multi-word lexical item[s] with verbal function” and suggests a classification of group-verbs in English as shown in Table 2-2 (cf. also Denison 1981 and 1984). Note that this classification rests somewhat arbitrarily on syntactic criteria alone (kind of object and number of particles) and that it excludes some types of constructions others would discuss under this heading, notably constructions containing light verbs of the kind have a drink. The numerous problems connected to this highly critical discussion of the research tradition on ‘group verbs’ and related topics, see van Pottelberge (2001).

38


classification will not be discussed here in detail; suffice it to point out that it is quite dubious whether all these constructions should be identified as lexical items: e.g., one could argue that suspect someone of something is not one lexical item, but rather an instance of the lexical item suspect with the two complements someone and of something, since there are hardly two different verbs in He suspected murder and in He suspected the gardener of the murder. On the whole, this chart does little more than to show that English verbs may be followed by particles, objects and prepositions, and it confuses the periphrastic word formation observable in the phrasal verbs (cf. the discussion in Section 2.6 below) with the valency properties of the prepositional and other verbs listed in the table. In Claridge (2000: Ch. 3), a sixfold division among multi-word verbs is drawn: phrasal verbs, prepositional verbs, phrasal-prepositional verbs, verb-adjective combinations (e.g. break open), verbo-nominal combinations (e.g. take a walk), and verb-verb combinations (e.g. let go). But the reason for treating them together is based on the mere intuition that these sub-classes exist and that they must somehow be connected: I would like to draw attention to a general problem concerning all the categories above … It is actually a (vicious?) circle we are moving in: we all know in some way what, e.g., a phrasal verb is, but a full and theoretically adequate proof of this intuitive knowledge seems impossible. If in doubt, I will therefore trust my intuitions more than I will trust any kind of test. (Claridge 2000: 40– 41)

The term ‘multi-word verb’ itself is taken from Quirk et al. (1985: § 16) and it really seems that the ultimate reason for discussing the constructs in question together is provided by the classification in Quirk et al. (1985), to which Claridge’s classification corresponds (cf. also the critique in Section 2.2.2 above). Claridge (2000: 42–44) mentions four arguments in favour of a distinct class of multi-word verbs: i.

the multi-word verbs are all representative of the analytic trend in English; ii. they have often been assigned to the ‘idiom section’ while they should really be treated as a separate class; iii. some of them share other features like lexicalization or the development of aspectual meanings; iv. there are synonymous simple verbs for most multi-word verbs. But these arguments do not provide a sufficient basis for regarding labels such as ‘group verbs’ etc. as more than a convenient shorthand for diverse constructions which may but need not share a set of common properties.


39

E.g., it is questionable whether come in (e.g. She came in) or let go (e.g. She let go) are indicative of ‘analytic drift’: what would be the more synthetic alternative in Old English? Moreover, it will be shown in Chapters 3 and 5 below that verb-particle combinations could even be regarded as a first step towards greater syntheticity rather than analyticity. But even if ‘group verbs’ were all to be seen as instances of analytic drift, these constructions develop from widely divergent sources, and historical studies in particular would seem to be well advised to focus on these individually different developments rather than lumping them together on the basis of a label whose explanatory value is quite dubious (unless one would be prepared to regard drift as an independent force causing linguistic change – for which there is not the slightest evidence). Also, the observation that constructions are not really idioms cannot possibly be a criterion for establishing a separate class and the logic of this criterion is in fact circular (‘multi-word verbs’ which have been assigned to the idiom section and do not belong there are, hence, ‘multi-word verbs’). Lexicalization is obviously not a property of the compositional phrasal verbs, and neither is aspectualization a property of the prepositional verbs. (For a discussion of the unsuitability of the existence of synonymous simple verbs as a criterion, see the next section.) The only reason to justify discussing some of these different verbal constructions together would seem to be a focus on specific diachronic processes, like e.g. lexicalization or aspectualization, which characterize some members of each of these construction types (see the overview provided by Traugott 1999 and the discussion in Chapter 5 below). Moreover, the major common property of the ‘multi-word verbs’ is idiomaticity, cf. e.g. the classification in Quirk et al. (1985: § 16.2) or also in Biber et al. (1999: § 5.3); this seems to have been first suggested in Kruisinga (1925) and Poutsma (1926). But proposals to define phrasal verbs as a subset of multi-word verbs are seriously hampered by the fact that idiomatic phrasal verbs are just one of the three major semantic categories of phrasal verbs. Synchronically, these share specific syntactic characteristics (cf. the outline above); historically, aspectual and idiomatic phrasal verbs develop on the basis of the compositional ones (cf. Chapter 5 below). Here phrasal verbs will therefore be treated independently; the underlying historical processes may be decidedly more diverse than the wholesale reference to ‘analytic drift’ and similar notions may imply. Moreover, except for the ‘prepositional verbs’ such ‘multi-word-verbs’ are very common in quite a few synthetic languages, and they exist for example in Greek and Turkish (cf. also the discussion of preverbs in the agglutinative Kartvelian and Finno-Ugric languages in Chapter 3 below). One might perhaps best view these constructions as being connected by resemblances

40


which may but need not apply to all of them and which predictably result in rather divergent classifications. 2.4.6. Replaceability by a simple verb One criterion that is frequently encountered in characterizations of the phrasal verbs is their replaceability by a simple synonymous verb. It was already shown by Bolinger (1971) that this belongs with a number of criteria which are of little or no value for a definition of the phrasal verb. For a summary of the communis opinio on the etymology and semantics of the phrasal verbs, cf. e.g. Dixon (1991: 275): The vast majority of phrasal verbs are based on monosyllabic roots of Germanic origin, almost all belonging to the types MOTION (e.g. bring, carry), REST (sit, stand), AFFECT (cut, kick, scrape), GIVE (give, get, have), MAKING (make, let), or the grammatical verbs be and do. The resulting phrasal verbs are distributed over a wider range of types; some of them have quite abstract and specialised meanings.

McArthur (1992: s.v. phrasal verb) gives examples such as (115) and (116): (115) They used up/consumed all the fuel. (116) The soldiers moved forward/advanced. In such pairs, the phrasal verbs (here: use up and move forward) are said to contrast with Latinate simple verbs (here: consume and advance), which are commonly characterized as more learned and stylistically elevated than their phrasal counterparts of Germanic origin; cf. Chapter 5 below for a more detailed discussion of the impact of the etymological provenance of a simple verb on its selection as part of a phrasal verb. For the time being it may suffice to point out that the two arbitrarily picked examples from McArthur (1992) indicate that the story must be more complex than that, since the four simple verbs use, move, consume and advance are all, in fact, from French. It is by no means possible to find synonyms (or near-synonyms) for all phrasal verbs; as a criterion, this “includes both too little and too much”, as Bolinger (1971: 6) observes, who adduces the following counter-examples: (117) The plane took off.


41

(118) He broke out with a rash. (119) He hauled off and hit me. In (117), departed would be too unspecific; in (118), erupted would be ludicrous; for (119), Bolinger (1971: 6) concludes, there is “no synonym that I am aware of, unless we admit ‘He upped and hit me’”. Moreover, due to the replaceability of to eat dinner by to dine, of to make a mistake by to err etc. this criterion would yield rather undesirable results. Even worse, and unnoticed by Bolinger, the staple examples for synonymous pairs of simple and phrasal verbs obscure the fact that simple synonyms are really the exception rather than the rule for both the compositional and the aspectual combinations, cf. e.g. an example like (116) above and the following phrasal verbs: (120) The soldiers rode forward. (121) The soldiers pushed forward. (122) The paper floated down. (123) We scrambled up. Perhaps one could come up with synonymous Latinate simple verbs, but these would be unlikely to be an established part of the English lexicon. Considering the replaceability of some phrasal verbs by simple Latinate verbs as a property of the phrasal verbs means putting the cart before the donkey. Rather, the wealth of synonyms is a characteristic property of the English lexicon which is ultimately due to the extent of borrowing from French, Latin and other languages in Middle and Early Modern English and not a defining property of the phrasal verbs alone (cf. the discussion of this aspect from a historical point of view in Chapters 5 and 6 below). Bolinger (1971: 17) also discusses a rather ingenious way of defining phrasal verbs: “Phrasal verbs can be defined by simply listing them”, but he adds immediately: “it has never been suggested that such a list might be exhaustive. For one thing, it is continually being added to”. Of course this observation is simply a consequence of the fact that phrasal verb formation is a productive process in English (cf. Section 2.6 below), apart from the obvious difficulty that such a list would also have to be established on the basis of criteria of some kind (see e.g. the discussion of productivity in Bauer 2003a: Ch. 5 and the overview in Koefoed & van Marle 2000 and the references there). I will likewise entirely exclude a discussion of the passivization test, which can be useful in dealing with prepositional verbs, in

42


order to distinguish them from less lexicalized combinations of verb and prepositional phrase, but with the phrasal verbs, “the test hardly does more to than confirm transitivity, which is not likely to be in doubt” (Bolinger 1971: 7). 2.4.7. Variation and style Variation and style of the phrasal verbs in present-day English have only rarely been subject to in-depth studies, and it is remarkable that up until very recently linguistic research has by and large contented itself with reproducing the often impressionistic assessments popularized in style guides since the beginnings of the 20th century. Bolinger (1971: 17) finds these British English examples (taken from Hill 1968) not acceptable in his own American English: (124) We have a comedian here who takes the Prime Minister off wonderfully. (125) They became very poor, so they sold out their big house and went to live in a cheap hotel. Bolinger’s observation points to two related aspects: First, phrasal verbs may have variety-specific properties in different varieties of English (cf. also Schneider 2004 on the structural nativization of phrasal verbs in world Englishes). Second, there are phrasal verbs in British English which do not occur in American English. But it has been suggested repeatedly that the use of phrasal verbs in British English is somehow due to influence from American English, cf. e.g. the remarks s.v. phrasal verb by Gower (Fowler 1965) and by Burchfield (Fowler 1996). Such a suggestion may be quite dubious. Indeed, such an assumed American influence is extremely unlikely before the 20th century, whereas phrasal verbs had been used widely in British English before that time, and even before the very existence of the American continent was known to the English (cf. Chapters 3 to 6 below). Thus that presumptive American influence is perhaps better interpreted as concomitant with the notion that phrasal verbs are colloquial, rather than as a factor in the development of phrasal verbs in British English (cf. the discussion in Chapter 6 below). In their discussion of phrasal verbs, Quirk et al. (1985) repeatedly point out their informality, e.g.: “phrasal verbs are usually informal” (§ 16.3). But among the examples they provide (§§ 16.3–4) are go astray, reel back, touch down (of a plane), or turn on (the light), etc. None of the examples


43

they give seems to be particularly colloquial, nor would such an impression be supported by a search of the BNC; cf. the following examples: (126) Blasts from weapons went astray or were turned back to their sources so that the warlock seemed to be using his/her assailants as puppets to fight themselves.CM4_2611 (BNC) (127) The years reeled back, her old saggy breasts seemed to respond and she crossed the room without realising it, wrapping both her children in the Chinese spread and taking them into her arms like a mediaeval St Anne with Virgin and Child.A6J_1431 (BNC) (128) The plane touched down at Hangchow, where we were exhorted to eat a rather mediocre dinner at the airport before continuing our flight.KAL_65 (BNC) (129) Wandering out again like an unhappy ghost, she went into her workroom, turned on the light and stared at the bench.H9V_2134 (BNC) In these instances one is reminded of Bolinger’s discussion of the replaceability of phrasal verbs by synonymous simple verbs (cf. Section 2.4.6 above), which here again would produce semantically or stylistically unsatisfactory results. This is not to say that the claim that many phrasal verbs in present-day English tend to be colloquial would be altogether wrong. Quite clearly, very many of them do belong to the more informal or ‘colloquial’ registers of the language, as in the following two examples (see Chapter 6 below for a critical discussion of ‘colloquiality’ as a linguistic concept): (130) I said, “Oh for Chris’sakes, shut up I know you’re trying to be nice to me and help me, but please don’t because you’re just giving me a headache”.FR5_2416 (BNC) (131) But once they were engaged, there was nothing she could do to turn him on.K54_2040 (BNC) The view that many phrasal verbs in present-day English tend to be colloquial also seems to be supported by the figures for their occurrences in the four registers ‘conversation’, ‘fiction’, ‘news’, and ‘academic prose’ given in Biber et al. (1999): The distribution patterns of phrasal verbs closely matches that for lexical verbs generally … except that academic prose has fewer than would be expected. Thus, rather than being a marked feature of conversation, phrasal verbs are

44

Present-day English and other Germanic languages notably rare in academic prose. In their place, academic prose shows a much greater reliance on derived verbs and more specialized verbs generally. (Biber et al. 1999: 409)

And similarly for the most common phrasal verbs: Overall, conversation and fiction show much greater use of the most common phrasal verbs than news and academic prose. The difference is especially noteworthy for intransitive phrasal verbs, which are extremely common in conversation and fiction, but extremely rare in news and academic prose. One reason for this is that most phrasal verbs are colloquial in tone. (Biber et al. 1999: 409)

But a closer look at Table 5.14 in Biber et al. (1999), which provides an overview of the most common phrasal verbs in the four registers, raises the question whether it is always colloquiality that accounts for the relative lack of phrasal verbs in academic prose and of the most common phrasal verbs in both news and academic prose (since Biber et al. 1999 visualize their findings with dots and beams only, it is not possible to reproduce their exact figures). One reason why, for example, get up or sit down are much more frequent in conversation and fiction than in news and academic prose is clearly to do with their meaning and not with their colloquiality, since academic texts referring to people getting up or sitting down will predictably be rather rare. This seems also corroborated by the inverse observation that make up and carry out are much more frequent in news and academic prose than in conversation and fiction: the occurrence of these phrasal verbs is clearly determined by subject matter (“carry out an experiment” etc.), not by style. Among the frequent phrasal verbs, take up, take on, look up, set up, take off, take over and turn out are also relatively rare in conversation. If these were ‘colloquial’ phrasal verbs, one could expect them to be frequent rather than infrequent in conversation. With others, e.g. shut up or come on, the connection to colloquiality is more plausible. But with very many of the remaining ones, the figures are quite inconclusive. Thus find out is just as frequent as give up in the ‘news’ register, but much more common in the ‘conversation’ register; this may but need not point to a stylistic difference. More importantly, the overall infrequency especially of the intransitive phrasal verbs in academic prose need not be connected to their avoidance as being colloquial, but rather with the preference of formal Latinate verbs, and this would not necessarily entail the conclusion that the phrasal verbs are informal. The question whether these frequencies in present-day English have anything at all to do with colloquiality will be addressed in the historical discussion in Chapter 6 below.


45

In many instances phrasal verbs are stylistically neutral, as is the case in very many of the examples given so far: turn on the light or look up a word in the dictionary can by no stretch of the imagination be labelled ‘informal’. In some cases they may even be distinctly formal, like call forth in the following example: (132) Each of these is lexically distinct, in that it has, for instance, different typical contrasts (e.g. long-sleeved for (a) and nude for (b)), and the two readings are called forth by different types of context.FAC_876 (BNC) Marks (2005c) concludes rather sensibly from a lexicographer’s perspective: “Some phrasal verbs are informal, and some are formal, but most are neutral; in this respect they are not different from other categories of vocabulary”. Both the characterisation of phrasal verbs as colloquial and the claim that they can be substituted by (typically more formal) simple verbs at best only describe tendencies, which are far from being unequivocal defining properties. 2.5.

Particle verbs in other present-day Germanic languages

In all present-day Germanic languages there are verb-particle constructions which are very similar to the English phrasal verbs, cf. the following examples (objects are underlined): (133) Swedish (Braunmüller 1999: 65) Vi målade över tapeten med grön färg. ‘We painted over the wallpaper with green paint.’ (134) Danish (Braunmüller 1999: 65) Vi malede tapetet over med grøn farve. ‘We painted the wallpaper over with green paint.’ (135) Norwegian (Askedal 1994: 262) Boka kjem ut i neste veke. ‘The book comes out next week.’ (136) Icelandic (Thráinsson 1994: 175) Fjöldi manns tók bækurnar fram. ‘Many people took the books out.’

46


(137) Faroese (Barnes & Weyhe 1994: 211) Hann las brævid upp. ‘He read the letter out.’ (138) Dutch (Booij 2002b: 21) Hans belde zijn moeder op. ‘Hans rang his mother up.’ (139) West Frisian (Hoekstra 2001: 93) De plysie siket it hûs troch. ‘The police search the house through.’ (140) German Iss die Qualle auf. ‘Eat the jellyfish up.’ (141) Yiddish (Jacobs, Prince & van der Auwera 1994: 407) Ikh heyb on. ‘I heave on (i.e. start).’ The similarities are quite striking and point to their shared historical origins which will be discussed in detail in the historical account in Chapter 3. From a contrastive point of view, there are such remarkable parallels in principle on the one side and variation in detail on the other. Let us therefore have a brief exemplary look at the major parallels and discrepancies in order to place the English construction within its wider Germanic context. The discussion in this section is restricted to the standard varieties of the languages in question; Luxembourgeois and Pennsylvania German will not be discussed here, but the general observations in this section also apply to these two languages. The three main semantic types of phrasal verbs discussed above for English (i.e. compositional, aspectual and idiomatic combinations) may indeed be found in each of the present-day Germanic languages, as examples (133)–(141) indicate. Although the examples are restricted to one per language, every present-day Germanic language has all three semantic types (cf. e.g. the contributions in König & van der Auwera 1994). Thus example (136) from Icelandic is compositional, with the corresponding English gloss ‘take out’, while example (140) from German is aspectual, again with the corresponding English gloss ‘eat up’. Example (141) from Yiddish is idiomatic and would yield a literal translation *‘heave on’. For a semantic account of the particles in one Germanic language, see e.g. the description of German ‘separable prefixes’ in Fleischer & Barz (1995: § 5.3.3.3).


47

2.5.1. Basic word order As regards element order, there is significant variation among the Germanic languages, both with respect to the relative order of verb and particle and with respect to the relative order of the particle and an object. Quite clearly, though, in all the Germanic languages the relative order of verb and particle is connected to the basic word order (cf. Holmberg & Rijkhoff 1998). Discussing basic word order is beset with pitfalls, since any statement depends on what elements are seen as constitutive of word order, in what kind of clauses, and on the basis of what kind of theoretical framework, so that a simple collection of descriptive data for different languages can turn into a lengthy journey. To take one example, reference grammars of Modern German (e.g. Eisenberg 2006, Engel 1988, Helbig & Buscha 2001, Sitta 1998) describe the position of elements in the clause primarily with respect to the sentence brace, which in declarative main clauses with a finite main verb and a ‘separable prefix’ (i.e. a verb particle) is established by the verb and the particle; i.e., these grammars do not explicitly describe the position of the particle in such clauses, but rather the position of all other elements relative to verb and particle (for contrastive and typological discussions of the sentence brace, word order and particle verbs in present-day German, see the contributions in Lang & Zifonun 1996, with further references). The brief discussion here focuses on the position of the verb in declarative main and subordinate clauses, especially with regard to the V-2 phenomenon; cf. e.g. Harbert (2007: 398): “the V-2 phenomenon is the requirement, apparently holding under at least some circumstances in all of the GMC [viz. Germanic] languages, that the final verb of the clause be no further from the beginning of the clause than second position (not counting conjunctions)”. Holmberg & Rijkhoff (1998: 79) characterize V-2 as “perhaps the most salient ‘special’ feature of the Germanic languages, distinguishing Germanic from all the other modern European languages”. Among the present-day Germanic languages, Afrikaans, Dutch, Frisian and German may be characterized as V-2/V-F, i.e. (leaving aside the details of the sentence brace and various exceptions, which, however, do not affect the validity of the general distinction) in declarative main clauses the finite verb is in second position (V-2), but in subordinate clauses the verb is in final position (V-F). The other Germanic languages may be characterized as V-2 or V-3, i.e. the finite verb either must follow the first element immediately (V-2), or there may be another intervening element (V-3). The Continental West Germanic languages (i.e. all except for English) are often labelled ‘OV’ languages (since they have the order ‘object-verb’ in all clauses not affected by the V-2 rule), while English and the Scandinavian

48


languages may be labelled ‘VO’ languages (since practically all main and subordinate clauses only have the order ‘verb-object’; for Yiddish, cf. fn. 6 below). As is well-known, English is rather strictly V-3 and ‘SVO’ (i.e. the verb may normally be preceded only by the subject and by an additional adverbial: the verb may be the second or the third element, but it must follow the subject, and consequently English is the only Germanic language without pervasive V-2); the Scandinavian languages, though, are normally V-2 (i.e. the subject need not precede the verb, but the verb must not be preceded by more than one element: the verb must be the second element; but subordinate clauses tend to be V-3; cf. Askedal 2005). But the topic is rather intricate in detail and, as Harbert (2007: 398) points out, “has been treated as the central problem of GMC syntax in recent years, and its analysis has been the subject of a vast literature”; cf. e.g. Askedal (1995), Braunmüller (1982 and 1999), König & van der Auwera (1994) and Harbert (2007: 398–415). 2.5.2. Particle position In those languages where the word order is strictly VO (V-2 or V-3), the particle will normally follow the verb, while in the languages with a mixed VO/OV (V-2/V-F) order, the particle will precede its verb in final position but follow its verb in non-final position. By inference, one way of summarizing the positional properties of the particles in all Germanic languages could be to observe that the particle will normally precede its verb when it is final, but follow the verb in all other cases.6 Thus Afrikaans, Dutch, Frisian and German show preverbal position of the particle in subordinate clauses and, more generally, wherever the verb is clause-final; cf. the following examples from Afrikaans (i) and their German equivalents (ii): 6

One exception to this appealingly simple rule (which, it should be stressed, is an observation and not an explanation), can be found in Yiddish, where we find sentences such as Er vet avek-šikn dem briv ‘He will off-send the letter’ (but Er šikt avek dem briv ‘He sends off the letter’; examples from Jacobs 2005: 239, following den Besten & Moed-van Walraven 1986), where the exceptional order may be interpreted as one of several remnants of earlier OV. But be that as it may, Yiddish with its rather special history among the Germanic languages, especially with respect to its contact languages, may be the exception that proves the rule; cf. e.g. Jacobs, Prince & van der Auwera (1994) and in particular Jacobs (2005).


(142) i. Ek skakel ii. Ich schalte I

turn

49

die lig nou af. das Licht jetzt aus. the light

now

off

‘I’m turning the light off now.’ 7 (143) i. Ek het vergeet om ii. Ich habe vergessen, I

have forgotten

die lig af te skakel. das Licht auszuschalten.

(for)

the light

off-to-turn

‘I forgot to turn off the light.’ (144) i. Ek skal die lig afskakel. ii. Ich werde das Licht ausschalten. I

will

the light

off-turn

‘I’ll turn off the light.’ The traditional term separable prefix verb for the particle verbs in these languages is rather unfortunate for a number of reasons. These reasons have been listed succinctly by Booij (2002a and b) for Dutch, but they apply equally to Afrikaans, Frisian and German: “The basic reason why SCVs [viz. ‘separable complex verbs’, which include particle verbs] have to be considered as word combinations, and not as prefixed words, is that they are separable” (Booij 2002a: 206). By the same token, Mahler’s (2002) proposal to analyse the particles as affixes (prefixes in Old English and suffixes in Modern English) should be treated with a good degree of caution, especially since his evocation of functionalist principles is, to my mind, not a satisfactory argument in favour of adding to the great terminological confusion by labelling an element which may stand separated before or after the verb as ‘prefix’, a term commonly (and sensibly) reserved to bound morphemes in pre-position (cf. e.g. the definition by Marchand 1969: § 3.1.2.1). Mahler’s (2002: 527) claim that particle verbs consist of a free lexical morpheme and a bound derivational morpheme is not only empirically dubious (e.g. eke out, come in: eke is not free, and in is not bound) but also methodologically: For one thing, it is not sufficiently justified by the observation that ‘prefix’ is a functional and ‘particle’ a structural term so that the derivational function of affixes can be structurally realized by particles since it does not follow from this that the function of a particle needs to be ‘prefix’. It also introduces implicitly the assumption that tmesis is a 7

The Afrikaans examples are from Donaldson (1993: 374), their German translations are mine; note the partly different orthographic conventions in the two languages when the particle precedes the verb. For parallel examples from Frisian, see e.g. Tiersma (1985: 109).

50


historical process operating on verbs and their prefixes rejected in IndoEuropean linguistics for the same reason: prefixes develop out of particles, not vice versa (cf. e.g. Baldi 1979 and the relevant studies discussed in Chapter 3 below, especially Kuryłowicz 1964: 171 and the overview in Fritz 2005: 22–35). In the Germanic languages, the particle in preverbal position differs from an inseparable prefix in that it is stressed; also, the infinitival particle TO and the past participle prefix GE-8 stand between the particle and the verb. This order is a further strong argument against analysing the particles of the ‘separable prefix verbs’ as prefixes since the normal order of affixes in any kind of derivation is inflectional prefixes – derivational prefixes – root – derivational suffixes – inflectional suffixes; “derivation creates lexemes, inflection creates forms of lexemes”, as Booij (2007: 71) reminds us. Bybee (1985) provides ample cross-linguistic evidence for this observation, which confirms Greenberg’s relevant universalist claim (1966); for a cross-linguistic discussion see also Hall (2000). Thus we have, for example, two prosodically distinct German verbs ´überfahren ‘pass over, traverse’ as in (145) with a ‘separable’ prefix, i.e. a verb-particle construction, and über´fahren ‘to run someone down (with a car, bike etc.)’ as in (146) with an inseparable prefix, with different meaning and syntax: (145) Er hat versucht, früh überzufahren. he

has tried

early over-to-drive

‘He tried to pass over early.’ (146) Er hat versucht, einen Hund zu überfahren. he

has tried

a

dog

to over-drive

‘He tried to run down a dog.’ Likewise, similarly parallel prefix and particle verbs are always possible in Swedish, where the prefix verbs tend to be figurative, e.g. Skogsarbetarna bröt av alla grenar och kvistar ‘break off (branches and twigs from a tree)’ vs. Polisen avbröt diskussionen ‘break off (a discussion)’ (examples from Braunmüller 1999: 77; cf. also Svenonius 1996: 19 for examples from Danish and Norwegian). But, again, this corresponds to English and the other West Germanic languages, where prefix verbs of the type outrun also exist. Moreover, this is entirely in line with cross-linguistic evidence from 8

The small capital letters here indicate reference to the infinitival particles cognate to English to and the past participle prefixes cognate to German gein all Germanic languages.


51

non-Indo-European languages, where similar pairings are well-attested (cf. e.g. the discussion of Georgian preverbs by Harris & Campbell 1995: 95; for a description of the usage in Swedish and for a comparison between English and Swedish, see also Lindelöf 1935: 258–261). Frisian is reported to prefer verb-particle constructions to inseparable prefix verbs also in cases where German and Dutch have the latter, e.g. Frisian ´oerride as in in hûn oer te riden ‘to run down a dog’ vs. German über´fahren as in (146) above and Dutch over´rijden as in een hond to overrijden; cf. Hoekstra (2001: 93). Theoretical implications of the relation between meaning and separability are also discussed by van der Auwera’s (1999), with respect to verbal prefixation in Dutch. In English and the North Germanic languages, on the other hand, the particle always follows the verb, with some minor exceptions like the fronting of particles discussed above, cf. examples (9), (81) and (82) (for the order of verb and particle in the Scandinavian languages, see examples (133)–(137) above and (148)–(153) below). In this respect, word order patterns in these languages are identical. As in all the Germanic languages, though, (inseparable) prefix verbs may occur, and especially in the Scandinavian group there may be considerable overlap between these and the particle verbs. Cf. e.g. the Norwegian example (135) from Nynorsk with a postposed particle and the following synonymous example from Bokmål (Askedal 1994: 262): (147) Boken book

utkommer i neste uke. out-comes

in next week

‘The book is going to come out next week.’ Moreover, there is some remarkable cross-Scandinavian variation with respect to the order of particle and object. In Norwegian and in Icelandic the rules are just like the rules in English. Thus we find the familiar English pattern also in sentences such as (148)–(151) from Norwegian (i) and from Icelandic (ii) (examples (148)–(153) are from Svenonius 1996: 11–18): (148) i. Vi kastet ut ii. Við hentum út we threw

out

hunden. hundinum. dog

i. & ii. ‘We threw out the dog.’

52


(149) i. Vi kastet hunden ut. ii. Við hentum hundinum út. we threw

dog

out

i. & ii. ‘We threw the dog out.’ (150) i. Vi kastet den ii. Við hentum honum we threw

it

ut. út. out

i. & ii. ‘We threw it out.’ (151) i. *Vi kastet ut ii. *Við hentum út we

threw

den. honum.

out it

(Cf. English *We threw out it.) But in Danish (iii) and Swedish (iv), the positional options are more restricted; cf. (152) and (153), and also examples (133) and (134) above: (152) iii. Boris

flyttet

møblene

Boris

moved

furniture

around

numret

upp.

number

up

iv. *Johan skrev Johan

(153) iii. *Boris iv. Johan iii. iv.

wrote

rundt.

flyttet rundt møblene. skrev upp numret.

‘Boris moved the furniture around.’ ‘Johan wrote down the number.’

According to Svenonius (1996: § 4.3), the rules in Faroese seem to be similar to those in Danish (but Holmberg & Rijkhoff 1998: 86 state they are similar to those of Icelandic; considering the genealogical proximity to Icelandic and the heavy influence from Danish none of these statements seems implausible, but Faroese positional syntax is not a particularly wellstudied topic, cf. Braunmüller 1999: § 5.5.1 and in particular Thráinsson et al. 2004). In Danish only the order object–particle is possible, but in Swedish only the order particle–object is possible, even in cases where it is not normally possible in English, cf. the following Swedish examples (from Holmes & Hinchliffe 1994: § 656): (154) Skriv write

upp

det!

up

it

‘Write it down!’


(155) Jag I

ringer upp

dem.

ring

them

up

53

‘I will ring them up.’ (156) Låt let

mig

stänga

av

me

close

off it

den.

‘Let me switch it off.’ That is to say, the English rule that requires pronoun objects to stand between the verb and the particle does not exist in Swedish, where the particle precedes all kinds of objects. Similarly, the serialization rules in the Continental West Germanic languages (except for Yiddish, where the straightforward identification as a West Germanic language is problematic) are not completely alike, despite the correspondences in principle. For example, a number of serializations in Dutch (i) are not possible in German (ii) (examples from van der Auwera 1995: 88; cf. there for examples of some further differences): (157) i. … om het boek … for

the book

terug te sturen back to send

ii. … um das Buch zurückzuschicken … for

the book

back-to-send

i. & ii. ‘… in order to send back the book’ (158) i. … om het boek … for

the book

terug te kunnen sturen back to can

send

… om het boek

te

kunnen

terug sturen

… for

to

can

back send

the book

ii. … um das Buch zurückschicken zu … for

the book

back-to-send

to

können can

i. & ii. ‘… in order to be able to send back the book’ (159) i. … dat ik … that I

hem op

wilde

bellen

him

wanted

ring

up

… dat ik

hem wilde

… that I

him

ii. … dass ich ihn … that I

him

opbellen

wanted

up-ring

anrufen

wollte

on-call

wanted

i. & ii. ‘… that I wanted to ring him up’ But although these examples may be taken as evidence that the separable preverbs are less free in German than in Dutch, as van der Auwera (1995: 87) observes, it is worth pointing out that the relative order of verb, particle

54


and object remains predictable from their general positional properties in Germanic. To conclude, the verb-particle construction is a common property of the Germanic languages, with considerable semantic and syntactic correspondences between the different members of the Germanic family: the semantic and syntactic properties of the English phrasal verbs are in essential aspects identical to those of the particle verbs in the other Germanic languages. Moreover, as will be argued in Chapter 6 below, their stylistic properties are likewise very similar. Studies of the English verb-particle construction would be well advised to bear these observations in mind, and to distinguish between general Germanic and specific English characteristics of the phrasal verbs. Although this may sound like a complete truism, very many accounts of the English phrasal verbs are seriously hampered by the misconception that the phrasal verbs are particularly ‘English’. Similarly, Olsen (1997) states: Alle germanischen Sprachen haben Partikelverben ausgebildet. Insofern stellen Partikelverben ein gesamtgermanisches Phänomen dar, einzelsprachliche Ausprägungen der Partikelverben dürfen daher nicht als eine isolierte Spracherscheinung behandelt werden … ein zu eng gefaßter Blick [kann] gerade in bezug auf das Englische zu verzerrten Ergebnissen über das Wesen der Partikelverben in dieser Sprache verleiten. (Olsen 1997: 45) [‘Particle verbs have developed in all Germanic languages. Thus particle verbs are a common Germanic phenomenon, and the particle verbs in any of these languages should not be treated as an isolated linguistic characteristic … in particular with respect to English, a focus which is too narrow may lead to distorted results regarding the nature of the particle verbs in the language.’]

One striking aspect in this connection is to do with English lexicography. There is a considerable number of specialized dictionaries of ‘phrasal verbs’ (cf. Herbst & Klotz 2009; the inverted commas are meant to indicate that often in these dictionaries the term also covers prepositional verbs and even other kinds of verbal constructions), usually aimed at learners of English as a foreign language, most notably vol.1 of Cowie & Mackin’s Oxford Dictionary of Current Idiomatic English (1975, new ed. 2006 as Oxford Phrasal Verbs Dictionary for Learners of English), but also e.g. Meyer (1975), McArthur & Atkins’ Dictionary of English Phrasal Verbs and their Idioms (1992), Longman Phrasal Verbs Dictionary (2000), Macmillan Phrasal Verbs Plus (2005) Cambridge Phrasal Verbs Dictionary (2006), etc. These dictionaries very much reflect (and promote) the view that the phrasal verb is a separate and particularly difficult area of the English lexicon, which is quite characteristic of the language. As will be argued


55

below, it is only this kind of evaluation of the construction that is typical of English, rather than the construction itself (cf. the discussion in Chapters 5 and 6 below). Promoting such a view has, of course, also a commercial aspect, considering the enormous market for reference tools for English as a foreign language. But at closer inspection such dictionaries may turn out to be quite superfluous, since lexicalized phrasal verbs have their place in the general dictionaries of the language (including learners’ dictionaries) – there are no comparable dictionaries of particle verbs in the other Germanic languages, and in none of them this seems to constitute a lexicographic gap. 2.6.

Phrasal verbs as periphrastic word formations

How, then, should the English phrasal verbs best be analysed? In the remaining part of this chapter it will be argued that an effective approach for a study like the present one is to treat verb-particle constructions in English, as in the other Germanic languages, in terms of word formation, namely as periphrastic word formations, which in a typological and historical perspective form a subset of the widely attested class of complex predicates. This approach has a threefold advantage. The invocation of periphrasis takes into account the fact that phrasal verbs are not ‘just’ ‘words’ (cf. the discussion of typical and atypical words in Schmid 2005: 25–28). This is obvious in English, where it is a minimum requirement for a phrasal verb that it consists of a ‘verb’ and a ‘particle’, which are always separated, but it also applies to the other Germanic languages. The argument is slightly more complex for the present-day Continental West Germanic languages, where particle verbs may sometimes look like ‘one word’ (cf. the examples in the preceding section); but there, too, a good case can be made against the classification of particle verbs as words (more precisely: as ordinary word formations), cf. the discussion below. In the English tradition the syntactic status of the phrasal verbs as ‘colligations’ has traditionally been stressed (for the use of the Firthian term with respect to phrasal verbs, see T.F. Mitchell 1958). The semantics of the phrasal verbs has often been whole-heartedly assigned to the field of idiomaticity (or of phraseology); cf. Sinclair’s well-known criticism that “the decoupling of lexis and syntax leads to the creation of a rubbish dump that is called ‘idiom’, ‘phraseology’, ‘collocation’ and the like” (Sinclair 1991: 104). The image of the ‘rubbish dump’ is regularly invoked in corpus- (that is, electronic corpus-) oriented studies of the phrasal verb, and it usually leads to some very general kind of statement that phraseology, and hence the phrasal verbs, are worthy of our linguistic attention, cf. e.g. the

56


introductions (et passim) in Claridge (2000) or Waibel (2007). But although it is certainly true that phraseology is worthy of linguistic attention and that discarding the phrasal verbs to the idioms section seems to have had a rather detrimental effect on their study for long, it might be more important to point out that the very treatment of phrasal verbs in the English tradition is based on a complete linguistic and cultural misperception which stresses the idiosyncrasy and the particular Englishness of the construction type (see Chapter 6 below), while actually little of this can be convincingly argued to be the case (cf. the discussion of the Germanic parallels in Section 2.5 above and in Chapter 3 below; for a general discussion of the similarities and differences between word formation and phraseology, see Barz 2007). Consequently the decision to treat phrasal verbs as periphrastic word formations also underlines the observation that they are not just ‘idioms’. Rather, both compositional subtypes can be described as absolutely regular and largely predictable, and they are structurally and semantically quite comparable to other, indisputable types of word formation. The varying degrees of schematicity of such morphosyntactically complex structures indicate different stages in their constructionalization (on phrasal verbs as constructions, see Section 2.6.4 below). The idiomatic phrasal verbs can then be regarded as equivalent to other idiomatic instances of word formation, where non-compositionality is widely encountered in lexicalized formations. And last, discussing the phrasal verbs as periphrastic word formations on a par with the particle verbs in the other Germanic languages opens up comparative and diachronic perspectives which remain closed to studies stressing their lexical and syntactic idiosyncrasies, in particular with respect to the long-term development in English, where the conditions in Old English are quite comparable to those in the present-day West Germanic (OV-) languages; cf. the discussion in the following chapters. In fact, many issues surrounding the development and the status of the phrasal verbs in English will turn out to be wild goose chases once the phrasal verbs are described as more regular and less exotic. 2.6.1. Verbs plus particles? First let us briefly illustrate why phrasal verbs cannot simply be regarded as syntactic combinations of ‘verbs’ and ‘adverbial particles’, even if one confines the analysis to present-day English. Defining the elements of the phrasal verb by referring to their function within the construction is indispensable if one wishes to account satisfacto-


57

rily for the productivity of the construction. For example, in the following short text an incident which led to the ad hoc formation of a phrasal verb terrace out is reported: (160) On a teacher-training course held by a lakeside a few years ago, the tutors gathered every evening for beer on a terrace overlooking the lake. After a few days of this routine, I asked one of my colleagues if he was coming down to the terrace and he replied ‘Oh, I don’t know, I think I’m just about terraced out’. (quoted from Marks 2005a) Claiming that this is a formation of the kind “[terrace (lexical verb)] + [out (particle)]” is rather unsatisfactory. It presupposes a lexical verb terrace as one part of the construction. But such a verb does not exist, nor is it clear what should be its meaning (there is, in fact, a verb terrace, which, however, cannot provide the basis for this phrasal verb, cf. OED s.v. terrace v.). In the context of (160) that verb could only be explained as a conversion of the noun terrace. Thus, in that case one would have to posit a conversion of the noun terrace into a verb terrace with no clearly identifiable sense before the formation of the phrasal verb, where the whole ad hoc formation would adopt a readily intelligible compositional meaning virtually out of nowhere. It seems considerably more plausible (and, what might be crucial, more economical) to abandon the idea of a lexical verb terrace as part of the above construction altogether, and to state that terrace here simply functions as the verbal element in a verb-particle construction. In other words, it is a ‘verb’ only by dint of being part of the phrasal verb. And terrace out is not an isolated instance, as the following examples show; here the first two are well established in the language, while the third is more likely to be a nonce formation: (161) He completely sexed up his appearance and attitude. (http://iamboigenius.com/trey-songz-passion-pain-pleasure-albumreview/) (162) A lot of heavy manual labour and chiropody also happens after meals, but no one bigs up those connections. (The Observer, 18 September 2011) (163) Furnitured out …: I’m absolutely exhausted just now because we’ve spent seven hours looking at furniture and household items to order for the flat. (http://newtonstheories.blogspot.com/2006/08/furnitured-out.html)

58


Similarly, it doesn’t take long to think up contexts for the following, at first sight slightly absurd, ad hoc formations: I have homered myself up. We garden-gnomed away. She has happied me in. It comes as no surprise that it is impossible to give an exhaustive list of the verb elements found in the phrasal verb: the formation of phrasal verbs is a productive process in present-day English whose application is not restricted to the word class ‘verb’. Cappelle (2007) has therefore argued that phrasal verbs should rather not be regarded as simple combinations of ‘verbs’ and ‘particles’ and sums up the issue as follows: While it is true that phrasal verbs always contain an element capable of carrying verbal inflection, this element need not actually be a verb outside the combination. Indeed, many phrasal verbs are based on an element from another word class – a noun … or an adjective … By allowing non-verbs to fill the slot, the English language sanctions many new creations, some of which are here to stay … while others have a more transient existence. (Cappelle 2007: 43)

Cappelle argues in favour of analysing such phrasal verbs as constructional idioms (cf. below), and clearly a pattern like [to be _ -ed out] ‘to have had enough of _’ is productive in present-day English, cf. e.g. a sentence like I’m coffeed out ‘I have had enough coffee’. It follows from these observations that the elements of the construction are to be defined functionally, i.e. the ‘verb’ element of a phrasal verb is whatever may be morphologically marked as finite in the construction, while the verbal status is conferred through combination with the particle. Thus phrasal verb formation does not involve the prior conversion of simple words belonging to other word classes into verbs: functionally speaking, the process in question here is derivation. Moreover, there are other aspects which are inconsistent with a syntactic approach to the phrasal verbs. For example, such an approach leaves completely unspecified what the particle actually is (except for being indeclinable, which is the defining property of all particles). As an adverbial complement, its positional properties would be different from all other such elements in English, so that one would have to posit an adverbial sub-class ‘particle complement’ (or the like), which would take us back to the start. The semantic properties of the particles would be quite different from other English adverbs. Even in the compositional combinations the particles are semantically rather abstract and connected to the internal spatial or temporal constituency of the verbal event (i.e. they indicate direction or aspect). Nevertheless it seems clear that typically the compositional combinations of directional particles and verbs are more akin to compounds than to affixations and that we can observe the well known


59

cline from compounding to derivation (as in the clearly derivational function of the particles in certain constructs discussed above) in phrasal verb formation as well (see Section 2.6.2 below). The common observation that compositional phrasal verbs can often be replaced by simple verbs is also of interest here: the meaning encoded by the particles clearly tends to be so general that it is found incorporated in other, common, simple verbs. And although non-compositional constructions could be assigned to phraseology primarily because of their intransparent semantics (which however is also a common feature of lexicalized word formations), their internal constituency, which is rule-governed and predictable, would render their classification as belonging to phraseology rather uncommon. Cf. the discussion in Barz (2007), or the array of topics covered in Burger’s (2007) introduction to phraseology, where it would be quite difficult to tell where the phrasal verbs (or, since Burger concentrates on German, the German particle verbs) could be fitted in. The claim that particles of phrasal verbs are functionally equivalent to affixes receives further support from another observation, which has already been mentioned above (cf. Section 2.4.4): there is a considerable overlap between the postposed particles in phrasal verbs and a number of homonymous prefixes in prefix verbs: down, out, over, under and to some extent also in and up may be synonymous as prefixes and particles (cf. the examples in 2.4.4). That is to say, seen from a functional perspective, it may be quite irrelevant whether the particle is preposed (as an affix) or postposed (in a phrasal verb), and thus it may well be argued that this remains true in those cases where prefixes identical to the particles do not exist any longer. The basic processes in question are compounding and derivation, while the obligatory postposition of separated particles in the phrasal verbs is the result of long-term changes in English serialization (cf. the discussion in Chapters 3 and 5 below). In addition to the derivational function discussed so far, the particles of phrasal verbs have another property which is commonly associated with derivational affixes, namely their capacity to effect changes in the transitivity (or rather: the argument structure) of a verb (cf. e.g. Cappelle 2007 and the references given there). This may be seen as a strong argument in favour of regarding the particles as functionally equivalent to derivational affixes (cf. Booij 2002a–b). If the phrasal verbs are regarded as a kind of word formation, this is only to be expected, since it is a well known fact that the valencies of derivations are not necessarily identical to those of their bases. Cf. e.g. the overview in Haspelmath & Müller-Bardey (2004); in the present context, their remarks on the general features of valencychanging morphology are of particular interest:

60

Present-day English and other Germanic languages Valency-changing categories generally have many of the properties that are considered as characteristic of derivation … they often exhibit formal and semantic idiosyncrasies and arbitrary restrictions on productivity … combinations of the verbal root and the valency-changing morpheme show a tendency toward lexicalization … Finally valency-changing can sometimes be combined with other word-classes, especially adjectives and nouns. (Haspelmath & Müller-Bardey 2004: 1139)9

E.g., the verb bark does not select animate objects (i.e. *bark someone is impossible), while the phrasal verb bark off does, cf. the following example:10 (164) Could coyotes have come down from the hills and attacked him? Possibly. But in the past, he and his canine buddy always barked them off. () The derivational force of the particle in bark off is underlined by a comparison with German, where the phrasal verbs would be translated by an (inseparable) prefix verb: bellen ‘bark’ vs. (jemanden) verbellen ‘bark (someone) off’. Similarly, dream (one’s life, the day, etc.) away also corresponds to a prefix verb with ver-, while other phrasal verbs correspond to (separable) prefix verbs in German, e.g.: (165) He is just slowly dreaming his life away. () 9

10

Their explanation for this runs as follows: “That valency changing categories are located more toward the derivational end on the derivational-inflectional continuum is closely related to their function … In Bybee’s (1985: 13) terms, valency-changing categories are highly relevant to the verb’s meaning, i.e. their semantic content directly affects the semantic content of the verb stem … Highly relevant meanings like those that change valency are most likely to be derivational” (Haspelmath & Müller-Bardey 2004: 1139–1140; their emphasis). Typical instances of the type of phrasal verbs in question have received a lot of attention in Construction Grammar (cf. Section 2.6.4 below), while some generativists would regard [someone off] as a small clause; but cf. e.g. We danced the night away, where a small clause analysis positing [the night away] would be decidedly weird, in particular in view of the possibility of We danced away (all night).


(166) (Wer nach ewiger Jugend strebt,) verträumt (who seeks eternal youth)

away-dreams

61

sein Leben. his life

‘Whoever seeks eternal youth dreams away his life.’ () (167) Schmutzige Wäsche wirft dirty

clothes

er

weg.

throws he

away

‘He throws away dirty clothes.’ () Characteristically there is no one-to-one correspondence between phrasal verbs and German (inseparable) prefix verbs. More often than not, phrasal verbs correspond to ‘separable prefix verbs’ (i.e. German particle verbs), while the categorial overlap between the types is illustrated by instances such as throw away – German wegwerfen (particle verb), but also vergeuden (prefix verb) as in Er vergeudet seine Zeit (‘He’s throwing away his time’, where a translation with wegwerfen, although conceivable, would be unidiomatic). Likewise, compare run someone over – German (jemanden) überfahren (prefix verb, inseparable) and take/drive someone over – German (jemanden) überfahren (particle verb, separable). Note that *run someone is impossible, i.e. again the argument structure of the simple verb is different from the argument structure of the phrasal verb (and cf. the parallel behaviour of the prefix verb outrun [someone]). This is not to claim, however, that particle verbs (and in particular the English phrasal verbs) are simply a sub-category of ordinary word formation; quite clearly, phrasal verbs stand at an interface of different linguistic levels and this is why they could be treated from different angles in the literature – syntactic, morphological, lexical, phraseological (cf. Olsen 1997: 47, who points out that since they are ‘accessible’ to syntactic and hence non-morphological rules, particle verbs are exceptional words from a traditional point of view). But there are strong parallels to word formation processes, and from a historical, from a comparative Germanic and from a wider cross-linguistic point of view the formation of particle verbs is inextricably linked to unambiguous types of verbal word formation. Moreover, as a consequence of the functional parallels to word formation the relationship between verb and particle can be expressed conveniently in established morphological terms and also without terminological leaps resulting in diachronic and cross-linguistic pseudo-problems. What is perhaps the most persistent of such problems will be discussed in Chapters 4

62


to 6 below, namely the idea that there is something like a ‘rise’ of the phrasal verbs in the Middle English period. 2.6.2. Periphrastic word formation I will basically follow Booij’s suggestion, originally made for Dutch (1990, 2002a–b), to regard particle verbs as periphrastic word formations. That the particles are “more or less affixal in nature” had already been observed by Bolinger (1971: 112). But although such an approach has been suggested repeatedly – if at times only implicitly – it has never gained wider popularity in English linguistics (but cf. also Kruisinga 1932 and Mahler 2002). The standard accounts of English word formation usually have very little to say about the formation of phrasal verbs. Typically, the phrasal verbs are either tacitly excluded, or their exclusion is explained with reference to the separation of verb and particle. Thus Marchand (1969: § 3.1.4), in his classic study on word formation in English, comments briefly on what he describes as “the tendency to form postparticle verbs” from the 14th century onwards in his account of verbal prefixation, but deals only with deverbal nominalisations of the type a break-up etc. Otherwise he leaves out the formation of phrasal verbs altogether, just like e.g. Koziol (1972), Kastovsky (1982), Bauer (1983), and more recently Plag (2003) and Schmid (2005). Kastovsky (1982: 70) excludes the phrasal verbs from his discussion but uses the idiomatic phrasal verbs as an example in his discussion of ‘formatives’, i.e. ‘minimal formal units which can be isolated on the basis of their syntactic and/or phonological characteristics but whose meaning cannot be established’. Adams (1973) argues explicitly against including the phrasal verbs into her study of English word formation: Phrasal verbs, combinations of verbs and adverbs like to build up, to take over, and prepositional verbs, combinations of verb and preposition, like to laugh at are both word-like and phrase-like, and are in fact sometimes called ‘semicompounds’, for instance by Kruisinga (1932). Though they often form a semantic unit, and may be equivalent to a single-verb item … their constituents are interruptible by the inflections of the verb, and in the case of transitive phrasal verbs, by an object as well. But unlike some other ‘semi-compounds’ … these verbs are uniform in their syntactic behaviour, and are not subject to closer unification of their elements with the passage of time. (Adams 1973: 9)

The arguments against the treatment of phrasal words from a word formation perspective rest ultimately on the fact that in English the particles


63

are not adjacent to the verb. E.g. Allerton (2004: 97) finds “if they [viz. the phrasal verbs] did not so clearly consist of two words, they could even be accommodated quite nicely with compound lexemes”. However, this observation can be easily accommodated if the phrasal verbs are identified as periphrastic word formations. Periphrasis is a well-known morphological phenomenon, and once the objection to lexical periphrasis is abandoned, there is little reason to continue treating phrasal verb formation as outside the scope of word formation. As Haspelmath (2000: 661) points out, “insofar as derivation may also be regular, periphrasis within derivation is certainly not unimaginable, even though it is not normally called periphrasis” (see Haspelmath 2000 on periphrasis in inflection, derivation and syntax). Similarly, Booij (2010: 21) notes that “phrasal constructs may express morphological properties” and presents periphrasis as a major argument in favour of constructional approaches to morphology (see also Section 2.6.4 below, but cf. also Müller 2002, 2008 for a critical discussion of some of the relevant problems with respect to particle verbs in German and other languages). Strang (1962: 157) observes (in the context of a discussion of the English phrasal and prepositional verbs) that “just as we treat of complex forms in conjugation, we are compelled by lexical and other evidence to recognise the existence of units, functioning and conjugated as verbs, consisting of two or three words – which may not even always follow one another in unbroken sequence”, but does not follow up on this observation. Booij & van Kemenade (2003) express comparable views in their discussion of preverbs (cf. also Chapter 3 below), where they observe: Since PV-V [viz. preverb-verb] combinations express various aspectual notions, and have gained in frequency at the expense of the older bound aspectual prefixes, we might think of them in terms of a derivational type of periphrasis. Are we justified in extending the notion of periphrasis to word formation? Let us point out that, at a more general level, there are good arguments for locating certain syntactic patterns in the lexicon, although they are productive … Periphrasis may then be seen as a specific subcategory of construction, since the periphrastic forms compete directly with synthetic morphological forms: they must be used instead of a synthetic form for the expression of certain kinds of information … This is clearly the case for inflectional periphrasis. Is it also the case for PV-Vs: do they compete with derivational morphology? … it appears that preverbs have taken over the function of verbalizing prefixes … the development seems to warrant quite clearly the notion of derivational periphrasis. (Booij & van Kemenade 2003: 7–8)

Implicitly this has also often been acknowledged in the description of those Germanic languages where, unlike present-day English, preverbal particles

64


may immediately precede the verb (cf. e.g. Ronneberger-Sibold’s 2004 classification of the German particle verbs as word formations). More explicitly, Fleischer (2000: 892) concludes that the German particle verbs should be classified on semantic, functional and distributional grounds as derivations, especially when compared to the prototypical (bound) verbal prefixes of German. Considering the discussion in the preceding section, where the structural and semantic parallels of the construction in the different present-day Germanic languages were pointed out, it seems clear that Fleischer’s arguments for German are likewise valid for English, especially his observation that a derivational function is not necessary for all such particles to render their analysis as word formations plausible. From this point of view, the particles are strongly reminiscent of affixoids, intermediate between compounding and affixation, long discussed in German linguistics. In his critical assessment of the term ten Hacken (2000: 355) lists the following criteria put forward for the identification of an element as an affixoid: (i) increased productivity, (ii) decreased semantic specificity, (iii) etymological and formal link to an existing free stem. Interestingly, the term has now come to be rather avoided (most notably in Fleischer & Barz 1995), mainly because it establishes a static synchronic (pseudo-)category for a diachronic process (i.e. the cline between compounding and affixation). But see e.g. Malkiel (1978) on the “overlap between prefixation and composition” from a typological point of view and cf. the critical assessment in Schmidt (1987) and the discussion by Donalies (1999): this ties in nicely with the cross-linguistic evidence. From a typological perspective, phrasal verbs are ultimately a subset of the widely, though not universally, attested class of complex predicates, which have been discussed both in Indo-European studies and with respect to other, non-Indo-European languages. Ackerman & Webelhuth (1998) provide a cross-linguistic categorization of complex predicates as shown in Figure 2-2. They summarize their observations as follows: several languages possess predicates consisting of a verbal stem and some element which precedes it. The preceding element can be either bound like a standard prefix (or the first member of a compound) or is separable from the verbal stem under some syntactic conditions. Despite formal differences concerning separability of the pieces of the compositions, many of the same lexical semantic, argument valence, case government, and grammatical effects are evident irrespective of the prefixal (compound) or preverbal status of the element accompanying the verb. That is, in an intuitive sense we are confronted by the same phenomenon independent of whether we encounter a synthetic or an analytic expression type. (Ackerman & Webelhuth 1998: 13)


65

(complex) predicate α properties

inseparable-particle-predicate β properties

separable-particle-predicate γ properties

Figure 2-2. Subtypes of complex predicates (according to Ackerman & Webelhuth 1998: 6)

For a general discussion of complex predicates from a cross-linguistic perspective, cf. Ackerman & Webelhuth (1998), while Harris & Campbell (1995: 92–96) discuss the typical paths of development of preverbs (the cross-linguistic and comparative evidence will be discussed in more detail in Chapter 3 below). More specifically with respect to English, Olsen (1997) has pointed out that there are two kinds of complex verbs in English – ‘prefix verbs’ (with ‘real’ bound prefixes) and ‘particle verbs’. The bound verbal prefixes in present-day English (a-, be-, co-, contra-, counter-, de-, dis-, en-, fore-, inter-, mal-, mis-, out-, over-, pre-, re-, sub-, trans-, un-, under-, cf. e.g. Quirk et al. 1985: App. I) are treated exhaustively in the relevant handbooks (cf. the references above and cf. also the historical discussion in Chapters 4 and 5 below). In terms of word formation, phrasal verbs can then be analysed as compounds (compositional phrasal verbs: combination of a verb and a spatial particle), but also as derivations in those cases where the meaning of the particle is different from the meaning of the free adverb (in particular in aspectual combinations) or has a clearly derivational function, as in the examples (160)–(163) above. In the non-compositional combinations, the two elements are formatives of a semantically non-transparent, lexicalised complex construction. Sometimes linguistic wheels are reinvented. Carstensen (1964) provides a reference to a somewhat obscure publication by Shluktenko (1955), which, originally written in Russian in 1954 and published in a German translation in the GDR by the Society for German-Soviet Friendship in 1955 has predictably been rather neglected in studies of the English phrasal verb. Shluktenko’s study, it seems to me, could have easily become one of the classics in the field, had it been published in a more accessible place. Indeed, some of the issues raised in the article show that Russian scholarship in the 1950s was quite advanced in the discussion of phrasal verbs

66


(Shluktenko inter alia refutes the notions of ‘intransitive prepositions’ and of ‘prepositional verbs’ in English!), and therefore it is worth quoting Carstensen’s summary and his evaluation of Shluktenko’s position at length: Shluktenko … meint, daß es sich bei dem von ihm untersuchten Typ to stand up nicht um Verb+Adverb handele, sondern um eine Zwischenbildung zwischen Wort und Wortgruppe, für die O.S. Achmanowa die Termini Wortäquivalent und analytisches Wort vorgeschlagen habe. Shluktenko sieht in der Partikel überhaupt kein unabhängiges Wort, und er hat überzeugend dargestellt, daß diese zweite Komponente des Wortverbandes die Funktion eines wortbildenden Morphems erfüllt und dessen abstrakten Charakter besitzt, obwohl ihre Position nicht immer fest ist. Er stellt überdies eine große Ähnlichkeit zwischen diesen postpositiven Elementen und Präfixen fest, die für ihn wichtiger ist als die Unterschiede (die entgegengesetzte Position im Wort und die mögliche Trennung der Postposition vom Verb). Er nennt unsere Partikel daher ein postpositives Präfix, ‘…Teil eines analytischen Wortes, ein Morphem, das abstrahierte, funktionelle Bedeutung besitzt und das ein Hauptmittel zur interverbalen Wortbildung darstellt’ … Shluktenkos Darstellungen überzeugen sehr, obwohl es eigenartig anmutet, ein dem Verbum nachgestelltes Morphem als Präfix zu bezeichnen. Entscheidend ist die Überlegung, daß es sich um einen festen Bestandteil des Verbums handelt. Die Frage nach der Wortart ist auch damit irrelevant geworden. (Carstensen 1964: 327; cf. Shluktenko 1955, in particular 232–235) [‘According to Shluktenko the type to stand up he has studied is not a verb plus an adverb, but rather an intermediate formation between a word and a word group, for which O.S. Achmanowa suggested the terms word equivalent and analytic word. Shluktenko does not regard the particle as an independent word at all and he has shown convincingly that this second part of the word combination functions as a derivational morpheme with its abstract character, although it is not always positionally fixed. He also observes a great similarity between these postpositional elements and prefixes, which he regards as more important than their differences (different positions in the word and the separability of the postposition from the verb). He therefore calls our particle a postpositional prefix, ‘… a part of an analytic word, a morpheme with abstract, functional meaning and a major means of interverbal word formation’ … Shluktenko’s account is very convincing, although it appears strange that a morpheme following a verb should be called a prefix. What is decisive is that it is considered as an integral part of the verb. This as well renders irrelevant the question of word class.’]

For arguments in support of Carstensen’s scepticism towards identifying the particles as prefixes, cf. the terminological discussion above. Also, to give credit where it is due, T.F. Mitchell (1958: 103) in his influential article in which he establishes the distinction between ‘phrasal’ and ‘prepo-


67

sitional’ verbs points out: “It is the word-class approach that explains the tendency, for example, to regard the particle component of the English phrasal verb as either preposition or adverb rather than as of one grammatical piece with the verbal component”; but Mitchell goes on to analyse the construction as a ‘colligation’ on a par with the prepositional verb rather than, à la Shluktenko, as a periphrastic word formation. The present study, it should be obvious, would reject Mitchell’s use of ‘colligation’ (in its Firthian definition as “[m]utual accompaniment of grammatical categories”, cf. Mitchell 1958: 103 fn. 3) as a cover term for both ‘phrasal’ and ‘prepositional’ verbs as too diffuse and regard the former in terms of word formation and the latter in terms of valency. 2.6.3. Further pros and cons Some arguments against the analysis of phrasal verbs as periphrastic word formations have been already discussed. It seems to me that the functional parallels to other verbal word formations and the reference to the possibility of morphological periphrasis are strong arguments in favour of the view here suggested. However, a number of other arguments against considering phrasal verbs as periphrastic word formations could be put forward, and in the present section these will be discussed briefly. I will first summarize each argument and then provide a commentary. The first two arguments are based on observations on de-phrasal-verbal word formations (cf. Section 2.4 above): i.

The input to word formation is typically provided by words, not by phrases. But it may be provided by phrases. Thus de-phrasal-verbal word formation is not an argument in favour of the status of phrasal verbs as words (cf. Booij 2002b: 27). It follows that in the present context this kind of observation is not relevant. ii. Since internal affixation is normally impossible in English, the occurrence of -er suffixes attached to the verb element and followed by plural endings may be seen as an indication of the syntactic status of phrasal verbs, e.g. *hands-bag vs. runners-up (cf. Elenbaas 2007: 17). But this is only true of nominalisations in -er and -ing (cf. Jespersen 1936: §§ 2.51–2.52), where the plural -s always follows the derivational suffix (i.e. *runner-ups), irrespective of its position (cf. runuppers, not *runs-upper) while the affixless nominalisations are entirely regular (i.e. drop-ins, never *drops-in); that is to say, the grammatical suffix -s immediately follows the last derivational suffix

68


added to the base, as usual in English – the comparison to formations like handbag (but *handsbag) is irrelevant. Two other arguments to be discussed here are to do with stress and with the position of morphological heads: iii. In a typical English compound, the first element is stressed. But in phrasal verbs, the particle carries stress, and thus the phrasal verbs cannot be word formations (cf. Emonds 1993). – This is a schematic application of Chomsky & Halle’s (1968) compound stress rule, which is quite problematic in itself (cf. Olsen 2000: 899–900); moreover this argument tacitly implies that the stress rules of nominal compounds apply for verbal compounds. But since there are practically no unambiguous verbal compounds in English which could be used for a comparison (cf. Marchand 1969: 100), this argument is untenable. Schmid (2005: 137), in a brief overview of verbal compounding, mentions ‘particle compounds’ as the only instances: overlook or underline. Both verbs carry their primary stress on the second element. See also Plag (2003: 154–155), who argues that most formations which look like verbal compounds are, in fact, back-formations or conversions, except for copulative verb-verb compounds (e.g. stir-fry), where stress is often but by no means always on the first element. iv. As compounds, phrasal verbs would violate the Righthand Head Rule.11 Therefore, they cannot be word formations (cf. Elenbaas 2007: 17); a stronger version of this claim might be: phrasal verbs do not have a head at all, and thus they cannot be word formations (cf. Plag 2003: 136). – Again, this is a problematic conclusion on the basis of dubious premises. The Righthand Head Rule is not a universal linguistic law, but a simple observation and in many languages it does not apply, or at least not generally since even in the languages where righthand heads are the rule, there are well-attested exceptions; cf. e.g. Aronoff (2000: 200–201), Bauer (2003a: 177–182) and Booij (2007: 53–55 and 75– 78). In the context of the present study, Booij’s observations are particularly noteworthy:

11

The Righthand Head Rule dates back in substance at least to Jespersen (1905), while the term was coined by Williams (1981); see Aronoff (2000), who also provides a brief discussion of the concept ‘head’ (and the related concept ‘feature percolation’) in generative morphology. For a highly critical evaluation of the use of heads in syntactic theory, see Matthews (2007).


69

The generalization expressed by the RHR [viz. Righthand Head Rule] has a historical explanation. In languages with right-headed phrases, the univerbation of such phrases has lead to the emergence of right-headed compounds. Subsequently, some right constituents of compounds have developed into suffixes … Thus, such suffixes could start functioning as morphological heads. On the other hand, prefixes often derive from words functioning as left, non-head constituents of compounds or from preverbal adverbs. … Due to their origin from words in a (morphological or syntactic) non-head position, such prefixes do not function as heads either. (Booij 2007: 54–55)

Ultimately, all such arguments are connected to the basic fact that particle verbs are ‘phrasal’, and quite clearly a purely morphological treatment of these constructions would be inadequate (put differently, phrasal verbs violate ‘lexical integrity’, i.e. the generative notion that syntactic rules cannot operate on word formations; cf. e.g. Aronoff 2000: 199–200). 2.6.4. Phrasal verbs as constructions The functional arguments in favour of periphrastic word formation need to be complemented by an account for their other construction-specific properties, since the word-formational properties of the phrasal verbs do not account satisfactorily for their characteristic syntactic features. Regarding the phrasal verbs as (quasi-)word-formational constructions with specific syntactic features dispenses with the fundamentally fruitless issue whether they should be assigned to morphology or to syntax; this issue may be taken to be essentially a reflection of the traditional division of linguistic subdisciplines, with respective terminologies which are designed for the discussion of the ‘core’ phenomena of each such subdiscipline but which fail in the description of intermediate phenomena. The traditional notion of ‘construction’ has come to play a major role in a number of approaches grouped together rather loosely under the heading of Construction Grammar. This is not the place for a comprehensive introduction to Construction Grammar, let alone a discussion of the differences between various constructionist approaches. For a good overview and for further literature, see the contributions in Fischer & Stefanowitsch (2007), Stefanowitsch & Fischer (2008) and Lasch & Ziem (2011). For diachronic Construction Grammar, see Bergs & Diewald (2008) and Trousdale & Gisborne (2008). Booij (2010) is concerned with Construction Morphology and includes a discussion of ‘separable complex verbs’ in Dutch. The common denominator of the different approaches to Construction Grammar

70


is the basic tenet that linguistic signs, i.e. conventional pairings of form and meaning, are the central elements of linguistic structure; such pairings are called constructions. Thus the use of the term concurs with pre-generative discussions of linguistic structure, most notably perhaps with Paul’s views (see Paul 1995 [1920]: 109 et passim; cf. the discussion in Feilke 2007: 65– 66). The question whether a construction belongs with syntax, morphology, or the lexicon is of secondary importance only, since constructions are seen along a gradient from the lexical to the syntactic. That is to say, there is no strict categorial division between lexicon and syntax, but rather one of degree, which is reflected semantically in a continuum from ‘rich’ and ‘specific’ (lexical) to ‘abstract’ and ‘general’ (syntactic) meaning (cf. Bybee 1985).12 The implications of such a view of language for the description of phrasal verbs become obvious in the definition of constructions given by Goldberg (1995; see also Goldberg 2006): Phrasal patterns are considered constructions if something about their form or meaning is not strictly predictable from the properties of their component parts or from other constructions. That is, a construction is posited in the grammar if it can be shown that its meaning and/or its form is not compositionally derived from other constructions in the language. In addition, expanding the pretheoretical notion of construction somewhat, morphemes are clear instances of constructions in that they are pairings of meaning and form that are not predictable from anything else. (Goldberg 1995: 4)

More explicitly Bybee (2010: 77) singles out phrasal verbs as an example of expressions which “follow general grammatical patterns but have lexicalized conventionalized combinations”. Despite theoretical divergences, current Construction Grammar approaches can be characterized as mono-stratal, non-derivational and non-modular. The insistence on nonderivationality has led Gries (2003) in a study of particle placement to suggest the classification of joined and split order as two completely different constructions (viz. [V – prt – NP] and [V – NP – prt]) rather than one category ‘verb-particle construction’ (cf. Gries 2003: 140). From the per12

Construction Grammar thus ties in with other current linguistic trends which concentrate on ‘intermediate’ phenomena and processes, in particular the interest in grammaticalization and lexicalization (cf. e.g. the overview in Brinton & Traugott 2005, which, despite its title, is an introduction to both processes) and the strongly increased interest in phraseology (cf. Burger et al. 2007). For a discussion of Construction Grammar vis-à-vis phraseology, see Feilke (2007), and of Construction Grammar vis-à-vis grammaticalization, see Diewald (2007).


71

spective of Gries’ study, which focuses on the factors affecting the choice of one order over the other, this seems to make sense. However, I do not see how Gries accounts for the fact that the two orders are linked by affecting the same lexical units, and it does not seem satisfactory to claim that the sentences John picked up the book and John picked the book up do not contain the same verb-particle construction; Gries (2003: 141) says so explicitly. But he restricts the use of this term to particle placement (while otherwise using the terms ‘phrasal verb’ and ‘verb phrase’, without ever clearly defining them). In this case one can certainly agree that the two sentences above contain two different ‘verb-particle constructions’, but who (including generative linguists) would ever have claimed otherwise? Moreover, Gries somewhat downplays the fact that his multifactorial analysis cannot account for at least about one fifth (cf. Gries 2003: 123) of the orders in his small sample, i.e. there remains a sizeable portion of instances where the two orders are in free variation. Cappelle (2006: 18) has pointed out that the two orders may be seen as “variant structural realizations of a construction which is partially underspecified” and suggests analysing the positional alternatives as what he calls ‘allostructions’, obviously in analogy to allophones, -morphs, etc. Gries’s approach is, in effect, not too far away from Cappelle’s, since Gries in his introduction points out that the alternation is an instance of ‘allosentences’, i.e. “semantically equivalent but formally and pragmatically divergent sentence pairs” (Gries 2003: 1, with reference to Daneš 1964 [1966] and Lambrecht 1994). Cappelle (2006) provides a critical evaluation of both the generative view (viz. that one order is derived from the other, cf. e.g. Dehé 2002) and the extreme constructionist position (viz. that the two orders are unrelated constructions, cf. Gries 2003) and argues as follows: generative approaches suffer from a rather pointless search for a basic, underlying pattern from which the other pattern is then purported to be derived. Although the existence of an unmarked pattern and a more marked variant should certainly not a priori be excluded, there do not seem to be any convincing arguments for considering the continuous order as basic. Even if there were, there still would not be any cogent (theory-external) grounds for positing a transformational link between this ordering and the discontinuous ordering … Extreme constructionalism tries to do away with alternations by studying the constructions linked by any alternation in their own right … Just because the two orderings are not linked by a truly Chomskyan transformation, it does not necessarily mean that language users are not aware of their relatedness. (Cappelle 2006: 11–12)

72

Present-day English and other Germanic languages [VP, trans V {prt} NPdirect O {prt}]

[VP, trans V prt NPdirect O]

[VP, trans V NPdirect O prt]

Figure 2-3. The transitive verb-particle construction with its two allostructions (according to Cappelle 2006)

Cappelle points out that it seems odd to assign idiomatic expressions like throw the baby out with the bath water and throw out the baby with the bath water to two different, and unrelated, constructions (among the other examples he mentions make {up} one’s mind {up}, pull {up} one’s socks, put {out} feelers {out} or turn {back} the clock {back}): “the fact that even such an otherwise fixed idiom has two formal manifestations, or ‘allostructions’ …, adds weight to the assertion that acknowledging a link between, in this case, the V – Prt – NP ordering and the V – NP – Prt ordering seems the right thing to do” (Cappelle 2005: 462). But this is not the reintroduction of transformational rules by the backdoor. None of the two allostructions is regarded as basic; rather, the choice of one variant over the other may (but need not, as the unexplained instances in Gries’ corpus show) be conditioned by other, mainly discourse-functional factors. Cappelle represents the positional variants of the transitive phrasal verb as in Figure 2-3. Booij (1990: 61) thus characterises particle verbs as “phrases that are created in the lexicon” and, following Jackendoff (see e.g. Jackendoff 1997) regards them as ‘constructional idioms’, a term I will not adopt here because of its connection to Jackendoff’s particular views and the implied model of language. In line with Cappelle’s suggestion, I will therefore regard the English phrasal verb as one construction characterized by the properties outlined at the beginning of the present chapter, and the relative order of particle and direct object as allostructional. 2.7.

Conclusion

In this chapter it has been attempted to characterize the phrasal verb in present-day English with respect to its syntax, semantics and style, both from a language-internal perspective and from a contrastive perspective in the context of other contemporary Germanic languages. The presentation has been necessarily eclectic, in part because of the nature of the present study, which does not primarily aim at solving syn-

Conclusion

73

chronic problems in the analysis of present-day English, but also because of the vast literature on the topic, which can be covered only selectively even by studies focusing on the present-day language alone. However, I have tried to provide a sufficiently detailed description and also to emphasize a number of issues which are sometimes neglected in discussions of the phrasal verbs in present-day English and/or which to me seem to be misrepresented in the literature. In the last section I have put forward the view that verb-particle constructions can be characterised as periphrastic word formations on a cline from compounding to derivation and that their specific positional properties can be explained as allostructional variation. So far the discussion has been strictly synchronic, although the crosslinguistic comparison mainly with the other present-day Germanic languages has, implicitly, already opened up a diachronic perspective, which will be pursued in more detail in the following chapters.

Chapter 3 The development of postposed particles What has traditionally been referred to as the ‘rise’ of the phrasal verbs is, syntactically, the loss of positional variability of the particle in relation to the verb. In Modern English the particle always follows the verb, with some well-described exceptions. But in Old English and to some extent still in Middle English, the particle may both precede the verb, as in example (1), and follow the verb, as in example (2): (1)

… swa swa se blinde clypode, þe big sume weige sæt þær se hælend forðferde ‘… as the blind man cried, who was sitting by the way as the Saviour passed by’ (ÆCHom I, 10 (G) [260.65])

(2)

… & ferde forþ on his weig ‘… and went forth on his way’ (LS 5 (InventCrossNap) [53])

Just as in Modern Dutch or German (cf. Chapter 2 above), this is clearly connected to the factors clause type and finiteness, although compared to the Modern West Germanic V-2/V-F languages particle placement in Old English appears to be more flexible and less rigidly governed by other syntactic properties of the clause. This is also the case in the older stages of the other Germanic languages (cf. the discussion below). Consequently the situation in Old English appears to be comparatively messy. In fact, this is a well-known characteristic feature of Old English syntax and in particular word order, which has been pointed out often: although it is not too difficult to observe general tendencies or ‘rules’, it is then just as easy to find a bewildering array of counterexamples and apparent exceptions. This, of course, is not to say that Old English word order is ‘free’. Such a claim would be entirely unfounded in view of the regularity of certain orders and of the apparent impossibility of others (cf. the discussion in Section 3.2.2 below). Nevertheless the Old English particle verbs behave in many ways essentially identical to the particle verbs in other Germanic V-2/V-F languages. In the following the discussion from Chapter 2 will be continued, where contrastive evidence from other present-day Germanic languages

Preverbs

75

was examined and the regularity of the situation in present-day English was shown. This chapter, then, will be devoted to a description of the positional properties of verbs and particles in Old and Middle English and of the syntactic changes involved in the loss of preverbal position of the particles. I will argue that there are good arguments in favour of postulating a distinct class of Old English particle verbs which are not just free combinations of verb and adverb or preposition, and which are diachronically related to (inseparable) prefix verbs. This approach seems even more plausible if diachronic and cross-linguistic observations are taken into account, which we will therefore turn to first. 3.1.

Preverbs

The Old English particles analysed by some authors as separable prefixes are also used in contexts where it may seem more appropriate to analyse them as prepositions or postpositions, or as adverbs. The foundation for the discussion was laid in a seminal article by Mitchell (1978), the title of which deserves being quoted here: “Prepositions, adverbs, prepositional adverbs, postpositions, separable prefixes, or inseparable prefixes, in Old English?”. The article demonstrates the enormous difficulties in arriving at a clear-cut classification of the elements in question (cf. the extensive bibliography in Mitchell 1978 and the discussion in Mitchell 1985: §§1060– 1080). Although Mitchell prefers the use of ‘pre-/postposition’ and ‘adverb’ (and of ‘prepositional adverb’ when referring to a particle’s potential of being used both as an adposition and an adverb), he concludes: For we have reached the boundaries where the kingdoms of the preposition, the adverb, the separable prefix, and the inseparable prefix, meet and melt into one another. We had better be micle mearcstapan here, not insisting on any one of the four terms but recognizing that, as Campbell puts it (OEG [viz. Campbell 1959], § 78), ‘we have in fact a system of separable prefixes like those of Dutch and German’ – in embryo, at any rate. (Mitchell 1978: § 50)

From a diachronic and comparative perspective, Mitchell’s observation, “the question whether or not Old English has ‘separable prefixes’ is largely terminological” (1978: § 21), is of little help, though. It fails to take into account the cross-linguistically well-attested category of preverbs, which would help to explain the apparent categorial indeterminacy of at least some Old English ‘particles’, ‘prefixes’, ‘adverbs’, ‘prepositions’ and ‘postpositions’. Despite its enormous erudition, the discussion serves to

76

The development of postposed particles

obscure the issue in this respect almost as much as it elucidates it, not least since Mitchell also discusses the position of prepositions and adverbs which quite clearly are not ‘separable prefixes’ in relation to their adjacent noun phrases rather than to clause type. The terminological confusion discussed by Mitchell is to some extent the result of category confusion, as Denison (1981: 49–50) points out, who suggests using ‘prefix’, ‘adverb’, and ‘preposition’ as mutually exclusive occurrences of ‘particles’ (which he prefers as the cover term for all three to Mitchell’s ‘prepositional adverb’). But Denison’s discussion sidesteps some of the relevant issues since it appears to be based on the assumption that if there are good reasons against the classification of elements as separable prefixes, there can be only adverbs and prepositions left. This approach implies essentially a very traditional parts-of-speech analysis, which is bound to yield inadequate classifications for most languages (cf. Haspelmath 2010). With regard to Old English, it is unsatisfactory to classify elements as either ‘adverbs’ or ‘prepositions’ on the basis of their occurrence in different clause types. Preverbal position typically occurs in subordinate clauses and postverbal position typically in main clauses, but otherwise there is no discernible difference between such particles. Denison (1981: 57–60) therefore concedes that he cannot offer a satisfactory analysis of what he must call ‘doublets’ in his framework, viz. cases like wiðfon – fon wið etc., where he must either posit synonymous doublets, or regard compounds and collocations as forming a cline, or abandon his terminology. Rejecting the third alternative, he states that the assumption of separable prefixes would not help to remedy the dilemma and suggests tentatively the adoption of the second alternative for a synchronic analysis and the first for a diachronic one, which “may even represent the intuitions of Anglo-Saxon speakers” (Denison 1981: 60). In sum, although Denison’s arguments against the classification of the particles as prefixes can be accepted as valid, his further conclusions need not be accepted, even if that leaves the particles in question without a more specific traditional label. Fischer et al. (2000: 182), in their discussion of verb-particles in Old and Middle English, conclude: “As it is not our intention to try to resolve this terminological puzzle, it seems best to use ‘particle’ as the most neutral term”. But the matter is far from being purely terminological and from a lexical point of view it matters considerably whether or not there are ‘separable prefix verbs’ like those of Dutch or German in Old English (cf. the discussion of the coverage in the historical dictionaries in Chapter 4 below). In comparative linguistics the particles have long been characterized as preverbs, a term which is particularly useful because it serves to avoid the rather fruitless terminological discussion of word-class status and under-

Preverbs

77

lines the diachronic and cross-linguistic peculiarities of verb particles and their development. Let us therefore now turn briefly to a more general discussion which takes into account the evidence for preverbs in other languages. Booij & van Kemenade (2003) give the following general characterisation of preverbs: The notion ‘preverb’ is a traditional descriptive notion in Indo-European linguistics. It refers to morphemes that appear in front of a verb, and which form a close semantic unit with that verb. In many cases, the morpheme that functions as a preverb can also function without a preverbal context, often as an adverb or an adposition. Most linguists use the notion ‘preverb’ as a cover term for preverbal words and preverbal prefixes. The preverb may be separated from the verb whilst retaining its close cohesion with the verb, which is called ‘tmesis’. It may also develop into a bound morpheme, that is, a prefix inseparable from the verb, with concomitant reduction of phonological form in some cases. If the preverb has become a real prefix, we may use the more specific notion of ‘complex verb’, whereas we take the notion ‘complex predicate’ to refer generally to multi-morphemic expressions with verbal valency … different types of complex predicates represent various kinds of mismatches in the syntactic and morphological coding of complex events and verbal valency, and thereby challenge our view of the architecture of the grammar, and the relation between syntax, morphology, and the lexicon. (Booij & van Kemenade 2003: 1)

The occurrence of preverbs is by no means restricted to the Indo-European languages (on preverbs in Indo-European, see Hackstein 1997 and 2003 and the references provided there). Indeed, preverbs may be found in very many typologically quite distinct and genetically unrelated languages. Moreover, despite their traditional name (which, as we shall see below, is connected to the serialization rules of Proto-Indo-European and its oldest attested languages), preverbs may also occur in postverbal position in some languages, as parts of ‘separable complex verbs’. The discussion by Booij & van Kemenade (2003), alongside the articles collected in the same volume, provides further literature and a useful overview of preverbation. The term ‘separable complex verb’ suggested by Booij & van Kemenade will not be used here, though, because as will become clear below it implies putting the cart before the horse with respect to the development of such constructions. Incidentally, note that I am not claiming that preverbs ‘exist’ as a category, which would be pointless since linguistic categories do not really exist anyway – see Haspelmath (2007). But the assumption of that category may provide fruitful cross-linguistic insights on the development in English, despite Haspelmath’s scepticism on the issue of comparing structures. Below, I illustrate preverbs in non-Indo-European languages and then turn to their development in Indo-European.

78


3.1.1. Preverbs in non-Indo-European languages A good example of the typical processes at work in the development of verb-particle constructions is the development of verb-particle constructions in the Kartvelian languages (cf. Harris & Campbell 1995: 94–96). Kartvelian is unlikely to be genetically related to Indo-European (or to any other language family), and for its earlier stages there is a considerable body of historical evidence from Georgian (see Harris 1991; the diachronic syntax of Kartvelian is discussed by Harris 1985). In the Kartvelian language Svan there are eight different preverbs (cf. Deeters 1930: 15–24; for a brief characterisation of the two languages to be discussed here, Georgian and Svan, see Campbell 1991). Four of those are (inseparable) prefixes, while the other four are adverbial particles which may occur before or after the verb: sga ‘in’, ži ‘up, on’, ču ‘down, under’, and ka ‘out, from’. In preverbal position, these particles may either precede the verb immediately, or be separated from it by a number of unstressed elements, typically other particles; the latter case is an instance of tmesis.1 Thus we find both the position in (3) and the one in (4) (the examples are from Deeters 1930: 17 and 15; see also Harris & Campbell 1995: 94). (3) (4)

ačad

sga

s/he.go

in

sgāčad

[< sga ačad]

in=s/he.go

The Svan prefixes, on the other hand, always occur as affixes, immediately adjacent to the verb. In this position they have been subject to morphological fusion and to substantial phonological changes. That the particles have also undergone partial fusion in preverbal position testifies to the regularity of the morphonological development (cf. Harris 2003: 64 and the discussion below). Accordingly, where both types of preverbs precede a simple 1

The term ‘tmesis’ is used here only as a shorthand for preverbal positions of preverbs which do not precede the verb immediately, and it is not used in the traditional sense, where it implies that the elements between the preverbs and the verbs interrupt an originally unbroken sequence; cf. e.g. Bußmann (2008: s.v.) and Fortson (2004: 140) for the history of the term, which is originally based on the notion that the occurrence of tmesis in archaic Homeric Greek is somehow deviant from the situation in later classical Attic. Cf. also the discussion of tmesis and Wackernagel’s Law in Latin in Section 3.1.2 and in Gothic in Chapter 5 below.

Preverbs

79

verb, the sequence is always particle–prefix–verb, as is to be expected. The overall development in Svan is summarized by Harris & Campbell (1995) as follows: One set of adverbs in Svan … were reanalyzed as optionally separable elements of the verb, contributing to the meaning of the verb. Frequently, but not always, they occurred immediately before the verb or separated from the verb by another element … Phonological change made these elements proclitic to the verb when immediately pre-verbal … Earlier the same changes had taken another set … along the same path from adverb to verbal proclitic. (Harris & Campbell 1995: 95)

The higher age of the prefixes is also witnessed by the fact that they have undergone semantic changes which have resulted in a partial loss of spatial meaning (cf. Deeters 1930: 19–24). The development of such semantic changes can be observed in genetically related Georgian. Different to Svan, in Old Georgian the particles occur in preverbal position only; their status as particles (rather than prefixes) in Old Georgian is, however, clear from their other syntactic characteristics (cf. Harris & Campbell 1995: 95). The three main semantic types as discussed in Chapter 2 above for the present-day English verbparticle construction can be found and the overall development goes from preverbal literal (i.e. spatial) particles in Old Georgian to prefixes in Modern Georgian. Contrastive observations with respect to the parallel developments in Kartvelian and in Indo-European are discussed in Schmidt (1969), where evidence for the adverbial etymologies of the Kartvelian preverbs is also provided (cf. also Cherchi 1994 and 1997 and the discussion in Harris 2003: 63–66). In particular with respect to the development of verb-particle constructions in English, it is remarkable that some of these prefixes may be either phonologically reduced or semantically changed, so that the conclusion that morphological and semantic changes may occur independent of each other could be justified (cf. the discussion of Old and Modern Georgian forms and meanings in Harris & Campbell 1995: 96). This observation is particularly noteworthy in view of widespread claims to the contrary in the grammaticalization literature (cf. e.g. Bybeee & Dahl 1989: 65), although of course the general tendency for a co-occurrence of morphological and semantic reduction is hardly to be doubted (cf. the discussion of grammaticalization and preverbation by van der Auwera 1999 and the general discussions in Bybee 1985 or Hopper & Traugott 2003). There are very similar typological parallels in other language families which are genetically unrelated to both Kartvelian and Indo-European. For example, van den Berg (2003) discusses preverbs in Dargi, an East

80


Caucasian language. Kolehmainen (2006), in the context of a contrastive study of particle verbs and prefix verbs in Finnish and German, deals with preverbs in Finno-Ugric languages. The following examples show the use of particle verbs in Estonian (Ackerman 2003: 14): 2 (5)

Kütt

laskis

hunter

shoot:PRET.3SG rabbit:PART

jänest.

‘The hunter shot a rabbit.’ (6)

Kütt

on

jänese

hunter

COP.3SG

rabbit:GEN down

maha

lasknud. shoot:PERF

‘The hunter has shot (down) the rabbit.’ (7)

Kütt

laskis

hunter

shoot:PRET.3SG rabbit:GEN down

jänese

maha.

‘The hunter shot (down) the rabbit.’ Similar structures can be found in Finnish and Hungarian, and although their recognition in the literature seems to depend to a considerable extent on the linguistic traditions of describing the individual languages, their structural and semantic properties again match those of other languages with verb particles (cf. the discussion in Kolehmainen 2006: 214–220 and the references given there). Depending on the respective traditions, the identification of preverbs in Finno-Ugric is in part a terminological issue, along the lines familiar from the discussion of Old English and other Germanic languages (cf. also Kiefer & Honti 2003). While the particle verbs in Hungarian are traditionally referred to as ‘prefix verbs’ and in Estonian as ‘compound verbs’, they have been subject to only very little attention in Finnish, where such constructions appear to have been rather neglected and treated as a phraseological phenomenon, if at all. But the striking parallels across the various Finno-Ugric languages as discussed by Kolehmainen (2006) make it quite likely that despite these terminological differences the particle verb is a common inherited property of the language family, comparable to the situation in Germanic. However, in the past for each of these languages language contact was claimed to have played a role in the development (for Finnish: with Swedish and German, for Estonian and for Hungarian: with German). But such proposals are highly speculative; e.g. in Hasselblatt (1990) the presumptive calques announced in the title of the 2

Adapted from Lavotha (1960: 104); I have slightly altered the glosses according to Lavotha (1973: 107). For further literature on Estonian, see Kolehmainen (2006: 217–219).

Preverbs

81

study turn out to be structural and semantic parallels in German and Estonian, which are assumed but by no means shown to be borrowings in Estonian on the basis of the observation of German influence elsewhere in the language. Most inseparable prefix verbs in the Finno-Ugric languages indeed seem to be Germanic calques, where the prefixal elements are unproductive. Wälchli (2001) discusses preverbs in Latvian and Livonian as part of a Baltic areal continuum and Kiefer & Honti (2003) take preverbation to be a Slavic-Uralic-Georgian areal phenomenon. In the light of the comparative evidence presented by Kolehmainen, it seems quite unlikely, though, that the particle verbs in these languages are the result of language contact in historical time. For a general discussion of the methodological issues involved in the assumption of language contact as an explanation of linguistic change, see Lass (1997: Ch. 4) and also the discussion in Chapter 5 below. The reasons why in some languages there are verb-particle constructions while they are absent in others are not particularly well understood – it seems that this issue has not been dealt with in detail at all in the linguistic literature so far. But on the basis of the evidence put forward so far, it is clear that the emergence of such constructions is a cross-linguistically widely attested instance of convergent evolution (cf. Lass 1997: 119) which is observable in several unrelated languages. The Australian Aboriginal language Warlpiri provides an extreme case of preverbation (cf. Hale, Laughren & Simpson 1995). Warlpiri has only about 120 simple verbs and makes extensive use of preverbation to form new verbs. Semantically, these verbs again range from fully compositional to fully idiomatic. Syntactically, the preverb either immediately precedes the simple verb or it is separated from it by an unstressed auxiliary (i.e., again, tmesis), but since word order in Warlpiri is quite free, the preverb may also follow the verb, especially if the preverb is highly productive and if the construction is noncompositional (see also Nash 1982 and 1986 and Simpson 1992). The parallels to the Germanic languages are quite striking, even more so since Warlpiri is typologically very different and since in this case language contact, sprachbund and similar explanations are completely out of the question. 3.1.2. Preverbs in Indo-European In the Indo-European languages, we find ample evidence for the emergence of prefix verbs and verb-particle constructions out of adverb-verb syntagmata. For Indo-European, verb-particle combinations have been discussed

82


by Kuryłowicz (1964: Ch. 7), Watkins (1963 and 1964) and Pinault (1995) (cf. also the general bibliographical discussion of Indo-European syntax by Fritz in Meier-Brügger 2003: IV.A and Fritz 2005 on Homeric Greek, including a brief overview of older Indo-European research on preverbs). In the present-day Romance languages such constructions do not seem to be widespread (cf. e.g. Harris & Vincent 1988), but Masini (2006) and Iacobini & Masini (2006) describe verb-particle constructions in Italian (in particular in the Northern Italian dialects). Dufresne, Dupuis & Tremblay (2003) discuss particle verbs in Old French, and Vincent (1999) explores the development from Indo-European via Latin to the earlier stages of the Romance languages. The contributions in Rousseau (1995) provide an overview of preverbs in the languages of Europe; see also Booij & van Kemenade’s (2003) general introduction to the topic from a cross-linguistic perspective. C. Lehmann (1995) summarizes the typical process as follows: Indo-European languages testify abundantly to the following alternative: an adverb which mediates between a verb and an NP may find either its relationship to the NP or the one to the verb tightened. In the former case, it becomes an adposition, in the latter, a preverb. It is true that there are IndoEuropean languages such as German, where the same elements may function now as adverb, now as preposition, now as preverb; and we cannot exclude the possibility that such a situation obtained in Proto-Indo-European as well. Nevertheless, the evidence of the earliest Indo-European languages, Hittite and Vedic, suggests that there was a class of elements whose primary function was that of an adverb, and, as it appears, in a diachronic sense, could function also as either adpositions or preverbs. (C. Lehmann 1995: 98)

Haase (2001: 737–739), in his discussion of the typical development of case endings out of such nouns, assumes the following path of grammaticalization: ‘relational noun > localizer > adposition > local case (> ø)’; cf. also the discussions by Bybee (1985) and by C. Lehmann (1983 and 1995) and the overview in Hopper & Traugott (2003). In Hittite, such adverbs are apparently still “completely ambivalent” (C. Lehmann 1995: 98; see also the discussion in Starke 1977: 127–131). Etymologically the Hittite preverbs are nouns, and so are at least some of the preverbs in many languages; cf. also Wilhelm’s (2002) discussion of the emergence of preverbs between the oldest and younger attestations of Hittite and the overview and further literature given in Hoffner & Melchert (2008: 294– 301). Even if the arguments against the assumption of three distinct categories adverb, preverb and postposition are valid (which e.g. Hoffner & Melchert 2008 doubt, though), it is remarkable that the serialization rules for such particles and verbs appear to be largely identical – including

Preverbs

83

tmesis with intervening negators, pronouns or locatival phrases – to those in languages with unambiguous preverbs (cf. Hoffner & Melchert 2008: 296). For a brief overview of Hittite and its position among the IndoEuropean languages, see Fortson (2004: Ch. 9); parallel developments in Tocharian and other Indo-European languages are discussed in Hackstein (1997 and 2003). Vedic represents a more advanced stage of the development than Hittite. In Vedic there is already a strong tendency for the particle to attach to and in certain contexts even to fuse with the verb (C. Lehmann 1995; on the relation of particles and adverbs in Indo-European and the problems of reconstruction connected to this somewhat neglected topic, see Dunkel 1992, on the issue of distinguishing between preverbs and prepositions in Indo-European, see Kuryłowicz 1964: 171–178, and for lists of Proto-IndoEuropean adverbs and prepositions and their cognate forms in various languages, see e.g. Beekes 1995: 218–223). In Latin such preverbs have either become prefixes, via cliticization to and subsequent fusion with the verb, or prepositions, e.g.: (8)

Caesar

milites

castris

educit.

Caesar[NOM.SG]

soldier:ACC.PL

camp:ABL.PL

out-lead:3SG.PRES

‘Caesar leads the soldiers out of the camp.’ (9)

Caesar

milites

ex

castris

ducit.

Caesar[NOM.SG]

soldier:ACC.PL

out

camp:ABL.PL

lead:3SG.PRES

‘Caesar leads the soldiers out of the camp.’ Cf. the discussion by C. Lehmann (1995: 99); in earlier records of Latin, though, there is evidence for the stage preceding the fusion of preverb and verb, as the following remark by the grammarian Festus on the language of early Latin prayers testifies: (10) Sub vos placo in precibus fere cum digitur significat id quod supplico … ob vos sacro in quibusdam precationibus est pro vos obsecro ut sub vos placo pro supplico. [‘Sub vos placo, which is said usually in prayers, means the same as supplico … ob vos sacro in certain prayers compares to vos obsecro as sub vos placo to supplico.’] This statement was already discussed by Delbrück (1893: 665); cf. also Booij & van Kemenade (2003) and Vincent’s (1999) account of the later development of preverbs and prepositions (with a focus on the latter) out of

84


adverbial particles in the Romance languages and its theoretical implications. Watkins (1964: 1039) points out that “[b]y the time of our earliest extended records, the univerbation of P [preverb] and V [verb] is an accomplished fact … But clear traces remain of an earlier stage preceding the univerbation; that commonly described as ‘tmesis’”, and he explains the position of the cliticized intervening elements as due to Wackernagel’s Law. The development of prefixes out of originally independent words depends crucially on word order, though; cf. the examples (3) and (4) above from Svan, where this is nicely shown: quite naturally, immediate adjacency is a necessary condition for morphological fusion. This observation is already clearly expressed in the older literature on Indo-European morphosyntax, cf. e.g. the classic discussion of preverbs and prepositions in Indo-European by Delbrück (1893), who stated: Aus meiner Darstellung dürfte sich ergeben haben, dass es die ältere Aufgabe der Präpositionen war, die Handlung des Verbes nach Massgabe des ihnen innewohnenden Sinnes näher zu bestimmen. Trat nun zu dem so bestimmten Verbum ein geeigneter Kasus, so konnte sich zwischen ihm und der Präposition ein näheres Verhältnis entwickeln … Da die traditionelle Wortstellung im Indogermanischen die gewesen sein wird, dass die Präposition vor dem Verbum stand, vor ihr aber der Kasus (da ja das Verbum gewohnheitsmässig am Satzschluss stand), so ergiebt sich als natürliche Stellung der Präposition die Stellung hinter dem Kasus, den sie bestimmt … Die Präp. war in der Ursprache im Hauptsatze jedenfalls nicht mit der Verbalform, zu der sie innerlich gehört, verschmolzen … Dieser Zustand ist im Altindischen geblieben, in anderen Einzelsprachen aber hat sich allmählich eine Annäherung der Präp. und der Verbalform vollzogen, so dass im nachhomerischen Griechisch, im Lateinischen, Germanischen, Baltisch-Slavischen die Verbundenheit der regelmässige Zustand ist. (Delbrück 1893: 664–665) [‘It should have become clear from my discussion that it was the older task of the prepositions to modify the action of the verb according to their inherent meanings. When a verb which was modified in such a way was then accompanied by a suitable case, a closer relationship could develop between it and the preposition … Since it will have been the traditional word order in IndoEuropean that the preposition preceded the verb and the case preceded the preposition (since the verb would usually be at the end of the sentence), the natural position of the preposition would be after the case it governs … At any rate in main clauses of the proto-language the preposition had not merged with the verb to which it inherently belongs … This state has remained in Old Indic, but in other languages the preposition and the verb by and by have drawn closer to each other, so that they are regularly united in post-Homeric Greek, in Latin, in Germanic, and in Balto-Slavic.’]

Preverbs

85

Despite his slightly confusing choice of terminology (see also W.P. Lehmann 1993: 227–228), Delbrück’s discussion leaves little doubt as to the basic processes deducible from the older Indo-European sources (cf. also Hirt 1928: § 23). In his discussion of ‘univerbation’ (i.e. the development from preverb to prefix) and word order change in Vedic and in Insular Celtic, Hock (1991: 338) notes that ‘univerbation’ is “the morphological counterpart of Behaghel’s Law, by which elements forming a single lexical unit become a single word”. Hackstein (1997) has shown that Delbrück’s assumptions are supported by more recent evidence from Tocharian, which like Hittite was unknown to Delbrück. Although in the present context only the evolution of preverbs is of immediate interest, Delbrück’s diachronic comparative observations go a long way towards explaining the messy situation in Old English which seems to have puzzled Mitchell (1978) so much; the Old English data necessarily yield an unwieldy synchronic analysis with a fair degree of layering, which proves to be a snapshot of a fairly orderly and widely-attested diachronic process (cf. Hopper 1991 and e.g. the overviews in Hopper & Traugott 2003: Ch. 5 and in Brinton & Traugott 2005: 143 on layering). Characteristically, Mitchell’s paper, despite providing more than ample references to the older Anglicist literature on his topic (cf. Mitchell 1978: fn. 2), does not contain a single reference to the comparative tradition, let alone more theoretically-inclined studies. Delbrück’s traditional comparative account serves to explain the emergence of postpositions and of prefixes out of frequent syntactic strings ‘oblique noun phrase – adverbial particle – final verb phrase’, where for the emergence of postpositions the reanalysis of the noun-particle sequence as a prepositional phrase is relevant, and for the emergence of prefixes the morphological fusion of the preverb and the verb. The two processes can be schematically represented as follows: (11) N.OBLIQUE particle V > [N.OBLIQUE adposition] V > N.OBLIQUE [preverb V]

> N.OBLIQUE [prefix-V]

Of course in principle the verb may co-occur with either the oblique noun phrase or the adverbial particle alone (but cf. Behaghel’s claim, contra Delbrück, that the particles originally belong with the noun rather than the verb, 1924: § 518). While Delbrück remarks on the positional requirements for the process and in particular on its connection to the usual order of elements in Indo-European (cf. the quote above), the wider implications of these observations for the developments of preverbs in the Germanic languages seem to have been made explicit only after the relative ordering of elements had become a major topic of typological research. Following

86


Greenberg’s pioneering study (1966), in particular W.P. Lehmann (1973) and Vennemann (1974a) suggested the correlation of word order parameters with ideal language types (cf. the overview and further literature in Askedal 2001 and see Siewirska 1993 for a cross-linguistic discussion of factors relevant to the establishment of word order; for a highly critical discussion of the generalizations proposed by Lehmann and Vennemann, see Comrie 1981: Ch. 4). There can be little doubt that the basic word-order in Proto-IndoEuropean was OV (for reasons for the reconstruction of Proto-IndoEuropean as an OV language and a detailed discussion of the connected methodological issues, see W.P. Lehmann 1993: 187–233). This is clearly reflected in some of the more archaic daughter languages, in particular in their oldest attestations (despite the arguments put forward by Friedrich 1975). Consequently, from a typological point of view, one would expect the use of postpositions rather than prepositions, and indeed the most archaic attestations of Indo-European languages provide evidence for the early emergence of postpositions, either quite regularly (as in Hittite), or in archaic traces (as e.g. in Greek and Latin). See Gamkrelidze & Ivanov (1995: Ch. 6) and cf. e.g. the overview in Fortson (2004: Ch. 7 [for the emergence of adpositions] and Ch. 8 [for word order]) and the references given there. On the interplay of prosody, syntax and semantics in the development of Indo-European prepositions, see Hackstein (2011). Note that the development of preverbs and the underlying diachronic assumptions made in the present section fully meet Clackson’s (2007: 159) prerequisites for syntactic reconstruction discussed in his critical assessment of the limits of syntactic reconstruction. Evidence of archaic OV is, of course, also visible in the Germanic languages, whose history is characterized by a long-term development towards VO structures (for discussions of the basic word order of Proto-Germanic, see e.g. W.P. Lehmann 1972, Hopper 1975, Braunmüller 1982, Gerritsen 1984, Lass 1994, Faarlund 2001 and Harbert 2007, among others). In the present context it needs to be remembered that this development implies a major change in the conditions which are relevant for the development of preverbs. This observation supplies the essential explanation for the developments in English, in particular in comparison to the other Germanic languages. Wherever OV order disappears, the adverbial particles will tend to follow rather than to precede the verb. Synchronically, this is reflected by the serialization rules in the present-day Germanic languages, as the discussion in Chapter 2 above has shown. Diachronically, there is cumulative and convergent evidence for archaic OV structures (cf. the discussion in Hock 1991: 609–621) with intervening adverbial particles, which continue

Preverbs

87

to intervene after the change to VO; so far I have not come across a fully satisfactory discussion (let alone explanation) of this observation. For a general overview of the issues involved in the discussion of such implicational typological tendencies and their relationship to the explanation of syntactic change, see Dryer (1995), while Harris & Campbell (1995) provide a rather critical discussion of typological approaches towards the explanation of word order change (1995: Ch. 8), but their discussion does not invalidate the observations of the present section. Thus both the development of preverbs into prefixes and the later development into postverbal particles in verb-particle constructions appear to be connected to more general developments in word order. The typical paths of development of preverbs and of adpositions discussed so far can, then, be summarized as in Figure 3-1 below, in which the dotted arrows indicate developments connected to word order change; elements characteristic of different stages may co-exist due to layering. This path of development clearly shows typical characteristics of grammaticalization processes (cf. e.g. Bybee, Perkins & Pagliuca 1994, C. Lehmann 1995 and Hopper & Traugott 2003). Like the development of adpositions, the emergence of preverbs goes along with decategorialization and loss of syntactic freedom of the original adverbial particles, whose semantics becomes more abstract and relational, although the original meaning tends to persist for a long time, at least in traces. Conceivably a long-term result of this, the concomitant prosodic weakening may lead to phonetic attrition and even complete loss (cf. also the discussion in Chapter 5 below). On the reconstruction of (Indo-European) syntax and the relationship of adpositions and preverbs, see also Hackstein (1997); cf. also the overview and discussion of a number of different models accounting for the paths of development of preverbs in Kolehmainen (2006: 152–162). Figure 3-1 is not exhaustive with respect to the further development of postpositions, which may also become inflections later on (cf. e.g. the discussion of this grammaticalization path by C. Lehmann 1985, where the interplay of synchronic variation and diachronic change is also focused on). For the later development of English prepositions, see Lundskær-Nielsen (1993) and Iglesias-Rábade (2011) and the references provided there. In the context of the present chapter, Lehmann’s (1985: 312–314) argument that the loss of one grammaticalized element cannot be taken to explain the grammaticalization of another, functionally equivalent, element seems particularly noteworthy, which, mutatis mutandis, also holds in the lexicon for the loss of prefixes and the development of verb-particle constructions in the case at hand.

88

The development of postposed particles [O] – adverbial particle – [V]

adposition

preverb

postposition

prefix

preposition

ø

postverbal particle

OV

VO

Figure 3-1. Paths of adpositions and preverbs

The native prefixes of English, as those of other languages, reflect an earlier OV stage, while the postverbal particles are characteristic of the more recent VO stage, which was reached in English by the late Middle English period. This observation ties in nicely with the well-known traditional claim that the order of elements in word formations reflects the word order in clauses; typically, the former can be observed to be more conservative than the latter and has indeed often been used for its reconstruction. But, as Harris & Campbell (1995: 200–210) show, this hypothesis has never been subject to closer empirical scrutiny in a range of languages with a long history of transmission, and it turns out on closer inspection to be beset with methodological pitfalls: There is reason to believe that in at least some instances a compound pattern may represent a borrowed pattern, a word order more than one stage back, or a sequence created for the purpose of distinguishing compounds from their phrasal equivalents. On the basis of these considerations, we conclude that compounds do not necessarily reflect the phrasal word order of the immediately preceding period. Compounds are therefore not a reliable guide to reconstruction. (Harris & Campbell 1995: 210)

The diachronic development of preverbs in English can be seen in the wider context of a cross-linguistically well-attested path of development (cf. the more detailed discussion in Chapter 5 below), and the continued existence of ‘separable prefixes’ with their mixed pre- and postverbal positions in the present-day Continental West Germanic languages turns

Preverbs

89

out as the expected situation in mixed OV/VO languages. For a discussion of preverbs in Proto-Germanic and in the earliest attestations of Germanic, see in particular Hopper (1975: 40–43). The development in German is discussed by Harnisch (1982) and Hinderling (1982), the development in North Germanic is discussed by Faarlund (1995). More generally, the topic is also discussed in Erben (2006: 136–139), cf. there for further literature. For a critical assessment of older approaches to the syntactic description of Old and Middle English preverbs (in particular with regard to word-order), see Denison (1981: 111–132); de la Cruz (1977) provides a valuable overview of the Indo-European parallels of the development in Germanic and specifically in English. In the light of the evidence discussed so far, it does come as quite a surprise that even a study as historically aware and cross-linguistically well-informed as Harbert (2007) presents the particle verbs as a particular Germanic phenomenon, and among the particle verbs in the Germanic languages the phrasal verbs as a particularly English one. This is no doubt the case because Harbert relies on the often unfounded and ill-conceived information from studies concentrating on English alone (cf. Harbert 2007: 36–40 and 366–367). But of course the identification of cross-linguistic patterns of change cannot act as a satisfactory substitute for a more detailed description of the development in English, and this is what we will turn to next.

3.2.

The development of English word order

The changes in the position of the particle, then, ought to be seen in connection to other positional changes in the clause, which will be sketched in the following as far as they are relevant to the development of the English verb-particle construction. 3.2.1. Word order in earlier Germanic Despite significant disagreement in detail, there is a broad consensus in the literature that the word order in the earliest attestations of Germanic is a mixed OV/VO type. This observation applies not only to the order of finite verbs and objects, but also to that of other kinds of head-dependent relations (such as the relative order of nouns and adpositions touched upon in the preceding section). As in the preceding chapter, I am not going to engage here in the discussion of ‘basic word order’ in the generative sense, where one order is assumed to be derived from another in the synchronic

90


analysis; consequently it should be clear that terms like ‘V-2 shift’ will be understood here as historical processes rather than as synchronic operations on another, underlying order. For an overview of the relevant issues in generative grammar (as far as the history of English is concerned) and for further literature, see Denison (1993), Fischer et al. (2000) and Roberts (2007). For arguments against explaining all results of historical change by way of generative synchronic rules or constraints and in favour of considering typological parallels and historical and functional explanations, see e.g. Fischer (2008). The most common positions of the finite verb in Old English are V-2 and V-F, depending to a significant extent – although by no means exclusively – on clause type, cf. the discussion of V-2/V-F orders in the presentday Germanic languages in Chapter 2 and the general references given there. The explanation of the change from the order of clause elements found in Old English to the rather strict SVO order in present-day English may well be the most hotly contested area of English historical syntax, and the mere description of the situation in Old English provides formidable problems. The discussion in the present section will therefore be restricted to a brief sketch of some main lines of development and to a review of the literature, as far as this appears relevant to the history of the phrasal verbs. Consequently the general changes in word order will be treated here as background information (as far as this is here possible; Denison’s remark, “[i]t is indeed a difficult subject, both from the point of view of the data and of assessing all the different and often incompatible studies which have been published” [1987: 139], still holds true). Neither their full description nor their explanation is the object of the present study; for an overview of the discussion and for further literature, see Traugott (1992), Fischer (1992), Denison (1993), Fischer et al. (2000) and Fischer & van der Wurff (2006). The issue whether the establishment of VO order is connected to the history of V-2 will be left aside here. A number of fairly influential (mainly generative) studies of the development of English word order, however, concentrate on the loss of OV and what they regard as the rise and fall of V-2 as essentially separate phenomena (cf. e.g. the division of the topic in Fischer et al. 2000 and the discussion of competing accounts in Denison 1993). This appears to be justified by the occurrence of VO orders which are not strictly V-2 and vice versa, but in my view it neglects the overwhelming frequency of orders which are both VO and V-2 (as long as they are not V-F) and the chronological parallelism of both developments. The arguments put forward in Fischer et al. (2000: 83) are, to my mind, not entirely persuasive, since they take the simultaneous synchronic occurrence of V-2 and what may be regarded as OV in the ‘mixed type’ present-day


91

Germanic languages as evidence against the historical connection between V-2 and VO. It may be more appropriate to view both the establishment of V-2 and the loss of OV as subject to the complex interplay of discourse factors such as topicalization, verb-fronting and extraposition of heavy objects, and to look into the subsequent grammaticalization of frequent orders as central to the development. For a general account of Old English word order along these lines, cf. Traugott (1992: 273–281) and e.g. the more specific proposals in Strang (1970) and Stockwell (1977), or the more comprehensive studies by Fourquet (1938), Reszkiewicz (1966) and Kohonen (1978); see also Hinterhölzl & Petrova (2009). Note also that all attempts to describe (let alone explain) Old English word order on the assumption that there is a ‘base order’ OV and an additional V-2 constraint fail to account in a satisfactory way for very many details, cf. the comments by Denison (1987: 153). But since this hardly makes a difference to the discussion of the verb-particle construction, the issue will for now not be pursued any further. Between the earliest attestations of (Northwest) Germanic and Old English, there are some significant changes in the relative frequencies of the word order patterns. In the Northwest Germanic runic corpus (3rd to 7th centuries AD), there are both V-F and VO orders, as in the following examples (12)–(14) from the early 5th century (the present syntactic discussion of the earliest evidence is based on Lass 1994: 217–224; for the sake of convenience I have followed the regularized presentation of these examples in Antonsen’s study of the runic inscriptions 1975: 24–25, for a closer transliteration, see Antonsen’s discussion of the corpus of inscriptions 1975: 29–89). (12) [me]z me:DAT

Woduride

staina

W:DAT

stone:ACC three daughters:NOM made:PL

þrijoz dohtriz

dalidun

‘For me, Woduridaz, three daughters made the stone.’ (13) ek I

Hlewagastiz Holtijaz

horna

tawido

H:NOM

horn:ACC

made:SG

‘I, Hlewagastiz Holtijaz, made the horn.’ (14) ek I

Hagustaldaz hlaiwido

mago

H:NOM

son:ACC my:ACC

buried:SG

minino

‘I, Hagustaldaz, buried my son.’ On the basis of the runic evidence it can thus be concluded that VO is already in existence in the earliest Germanic texts, as shown by example (14). Due to the peculiarities of the runic evidence, the identification of the

92


position of the verb with regard to subject and object is possible in 34 instances only (Antonsen 1975: 24–25). The material in the corpus is entirely prepositional but otherwise largely postmodifying; cf. Antonsen (2002: Ch. 13), where the usefulness of the runic evidence for the study of early Northwest (rather than just ‘North’) Germanic is evaluated. V-F as in examples (12) and (13) appears to be the clearly dominant order: V-F order occurs in about 70 per cent and VO is found in about 20 per cent of all clauses, while the remaining ten per cent are VSO (excluding verb-initial imperatives); there is no evidence for subordinate clauses in the corpus. But two out of three possible instances of VSO are without a subject, as in example (15) – which somehow begs the question as to their interpretation. (15) tawo

laþodu

prepare:1SG invitation:ACC

‘I prepare an invitation.’ On the whole, these observations tie in nicely with OV as the reconstructed order for Proto-Germanic (and also for Indo-European), and with the long term-developments in the histories of the individual Germanic daughter languages. The positional changes can be observed to be fully on their way already in the somewhat later North Germanic runic inscriptions, where V-2 has become the frequent type (cf. Antonsen 1975: 25). But it would be misleading to claim that the messiness of the situation in Old English contrasts with clear-cut word order rules in the other early Germanic languages. Rather, in all of them the regularities of V-2 and (where retained) V-F increase in the course of their histories, which lends some support to the notion that here in general more rigid orders represent more advanced stages in the grammaticalization of pragmatically regulated discourse orders; cf. the discussion below. The loss of OV occurs in the various Germanic languages to different degrees. While some, like Dutch and German, have residual OV orders in subordinate clauses, others, like Yiddish and Icelandic, are V-2 throughout (cf. the discussion of ‘asymmetric’ and ‘symmetric’ V-2 languages in Harbert 2007: 400–404). For a brief overview of the history of analysing V-2 orders in Old English and other Germanic languages (which goes back to the 19th century), see e.g. Stockwell (1990: 92, fn. 6).


93

3.2.2. Word order in Old English Despite the plethora of studies dealing with the development of Old English word order, only little attention has been given to the word order of the earliest English texts (which is to some extent surely to do with the fact that many of the earliest texts do not really lend themselves easily to the study of word order). In the earliest Old English texts there appears to be still a considerable amount of V-F clauses (both main and subordinate), but there is also a general increase in VO structures and also in V-2, as in the other Germanic languages. And as in the other Germanic languages, that increase is more marked in main clauses, while V-F tends to remain in subordinate clauses. In the earliest Old English texts, Lass (1994: 221) observes a “syntax not far removed from NWGmc [viz. Northwest Germanic], though with a somewhat different distribution of certain order types”, and towards the early 9th century he observes a more distinct “drift away from dominant OV order” (Lass 1994: 224); cf. e.g. the discussions of the syntax of ‘Cædmon’s Hymn’ from the early 8th century by Lass (1994) and by Nielsen (1998), where the language is shown to be very much of a transitional type from OV to VO. The following example from Alfred’s preface to the Cura Pastoralis shows some of the typical word orders in early West Saxon (late 9thcentury): 3 Ða gemunde ic hu sio æ wæs ærest on Ebriscgeðiode funden & eft, ða hie Creacas geliornodon, ða wendon hie hie on heora agen geðiode ealle, & eac ealle oðre bec. & eft Lædenware swæ same, siððan hie hie geliornodon, hie hie wendon ealla ðurh wise wealhstodas on hiora agen geðiode. & eac ealla oðra Cristna ðioda sumne dæl hiora on hiora agen geðiode wendon. [‘I also remembered then that the law was first composed in Hebrew and that later, when the Greeks learnt it, they translated it completely into their own language, and also all other books. And afterwards the Romans, too, when they had learnt them, had them all turned into their own language through wise translators. And likewise all other Christian peoples translated some part of them into their own languages.’]

3

Quoted after the normalized version in Mitchell & Robinson (2001), except for the Tironian sign (‘&’ in the example), which I have used to replace the editorial ond (since it is possible that this sign is not always used as an abbreviation for a conjunction at all, but rather simply marks the beginning of a new sentence in the manuscript; cf. Mitchell 1985: § 1724).

94


This passage is quite instructive since it contains three times the phrase (hie) wendon on hiora agen geðiode ‘(they) translated [a text] into their own language’, but each time with different positions of verb and object marked by underlines in the following examples (in (16) the underlined pronoun hie is acc.sg.fem., anaphoric to æ, in (17) it is acc.pl., anaphoric to bec): (16) ða then

wendon hie

hie

on heora agen geðiode ealle

turned

it

on

they

their

own

language all

‘Then they translated it completely into their own language.’ (17) hie

hie

wendon ealla ðurh

wise wealhstodas

they

them

turned

wise

on

hiora agen geðiode

on

their

all

through

translators

own language

‘They had them all turned into their own language through wise translators.’ (18) &

eac ealla oðra Cristna ðioda sumne

and also all

dæl

other Christian peoples some:ACC part[ACC]

hiora

on hiora

agen geðiode

wendon

DEM.GEN.PL

on

own

turned

their

language

‘And likewise all other Christian peoples translated some part of them into their own languages.’ Bean (1983), in her study of the development of Old English word order, takes such examples as evidence that the picture of the overall development from OV to VO is considerably blurred by stylistic factors, and she takes the three different orderings to be a “means to avoid monotony” (Bean 1983: 123). But, as Denison observes in his discussion of Bean’s material, the orderings are “far from random” (Denison 1987: 141; for a fuller appraisal of Bean’s study and its methodological and factual shortcomings see Denison 1987 and 1993: 47–48). Rather, they represent what may be expected as being entirely normal in Old English – and not for stylistic, but for syntactic reasons. Thus (18) is V-F, which at that time is the normal order in subordinate clauses, but which is also quite common in clauses introduced by and, ac or ne, as in the example (see Mitchell 1964 for a full discussion of the special status of such clauses with regard to word order in Old English and for a thorough criticism of earlier studies which fail to take this into account). This is an instance where the word order in Old English deviates noticeably from the word order in the present-day Germanic V-2/V-F languages (cf. the more extensive overview in Mitchell 1985: §§


95

1719–1732 et passim and Denison 1987: 144–146 for the possible implications of this peculiarity). Conversely, a typical V-2 context is provided by (16), where the verb follows the sentence-initial adverb þa (for a discussion of word order, clause type and the role of sentence-initial elements, see Andrew 1940; for a discussion of þa with respect to clausal syntax, see in particular Blockley 2001 and van Kemenade & Los 2006 and the references given there). In principle V-2 is independent of the length of the sentenceinitial element, e.g. (cf. Traugott 1992: 274): (19) On þæs on

caseres

DEM.GEN.SG

dagum

þe

wæs gehaten

emperor:GEN.SG day:DAT.PL REL was

called

Licinius | wearð astyred mycel ehtnys

ofer þa Cristenan

Licinius

over

was

stirred

great

persecution

the Christians

‘In the days of the emperor who was called Licinius | [there] was stirred a great persecution over the Christians.’ (ÆLS, Forty Soldiers [000300 (4)]) More typically, though, there is a very strong tendency for ‘light’ (short, given) elements to come early and for ‘heavy’ (long, new) elements to come late (cf. Reszkiewicz 1966 and Kohonen 1978). This view seems to be accepted in principle by anyone working on the topic. But although this is in line with Behaghel’s Laws (cf. Behaghel 1909 and 1924) and although this appears to be intuitively plausible, the exact nature of ‘lightness’ vs. ‘heaviness’ has to date rarely been pinned down (cf. the discussion in Denison 1993: 39–41 and Pintzuk & Taylor 2006). With regard to clauseinitial positions, this is most noticeable with pronouns, which will precede the verb even in cases where a corresponding full noun phrase would be found after the verb. Thus (18) provides another typical order, in which a light pronominal object precedes the verb in what is arguably still a marginal instance of a V-2 clause. Full nominal objects, on the other hand, are normally postverbal in main clauses, while in subordinate (and coordinate) clauses they may both precede and follow the verb. This had already been commented on by Mitchell (1964: 119); for a full analysis of pronouns as ‘syntactic clitics’, see van Kemenade (1987), but cf. also Stockwell (1990). But there are quite a few cases which resist neat interpretation along the line sketched here, e.g.: (20) Ðillice word Maria heold such

words Mary

held

aræfniende on hyre heortan. pondering

on her

heart

‘Mary kept and pondered such words in her heart.’ (ÆHom I, 2 [009300 (197.214)])

96


It thus seems clear that the word order in Old English is neither entirely free nor as restricted as e.g. the word order in present-day Dutch and German, despite the general similarities. Maybe neither V-2 (or VO) nor V-F are rules of Old English grammar – to date there is no account of Old English syntax which would succeed in accommodating these tendencies with the countless possible (and indeed attested) exceptions; a look into any of the desperate attempts (especially those in which V-2 and OV are taken to be ‘real’) to formulate satisfactory rules is instructive. It is indeed much more likely that Old English word order had better be explained as the interplay of a variety of factors and that consequently a multi-factorial analysis of the attested corpus would yield better results than the search for tidy little trees. In principle this suggestion was made already by Denison (1987: 155), but so far little has been done in this direction; for a discussion of the role of pragmatic discourse factors in the development of English word order and for more recent literature, see Seoane (2006). In a short representative sample of early West Saxon of the late 9th century (Denison 1987), the most common word orders in main and subordinate clauses respectively were as shown in Table 3-1.4 In this sample, V-F is highly typical of subordinate clauses and V-2 is highly typical of main clauses, but V-2 also occurs in subordinate clauses where the first element is not the object (note how the figures would be distorted if main clauses introduced by and, ac or ne were not counted separately). V-F in main clauses only occurs if they are introduced by a conjunction. Moreover, main clauses without an overt subject always have the verb in first position in the sample (as e.g. Foron þa up be Temese [ChronA 894] ‘[They] travelled then up along the Thames’), while in the other clauses with the verb in initial position it is immediately followed by the subject (as e.g. Wæs Hæsten þa þær cumen mid his herge [ChronA 894] ‘Then Hæsten had come there with his army’). On the basis of such evidence, it is justified to conclude that in main clauses the verb rather consistently precedes the object (consonant with both V-2 and VO orders) as the unmarked order, just as V-F is the unmarked order for subordinate clauses.

4

Denison’s sample consists of the entries for 892–898 in the Anglo-Saxon Chronicle, MS A (Bately 1986). The table disregards a number of rarer orders in the sample; for a more detailed description of the sample and the principles on which the analysis was carried out, see Denison 1987: 142–144, fnn. 6–7); ‘X’ stands for any element except subject and verb, ‘X≠O’ further excludes objects.


97

Table 3-1. Word order in early West Saxon (according to Denison 1987: 143) main clauses

and/ac/ne …

subordinate clauses

VSX

7

2

–

X≠OVS

47

3

4

SVfiniteXVnon-finite

1

5

8

SVX

13

11

17

SXV

–

10

53

SXVX

–

–

11

(others)

5

9

3

Further evidence for this claim is provided by potentially ambiguous sentences like & þa burgware hie gefliemdon [Chron A 894] ‘and the citizens put them to flight’, where subject and object are morphologically indistinct and the order SOV (rather than OSV) is used, despite the otherwise observable tendency for pronouns to be placed early in the clause (cf. Denison 1987: 147). As a result of the emergence of V-2 in earlier Germanic, Old English shares with the other West Germanic daughter languages the clausal brace, whereby finite and non-finite verbs become separated, e.g.: (21) Forðon we sceolan mid therefore we must:PL

with

ealle mod & all

mægene to Gode

mind and power

to God

gecyrran turn:INF

‘Therefore we must turn to God with all our mind and power.’ (BlHom 97. 26) There is a decided lack of studies of the clausal brace in Old English; Mitchell (1985) does not have a lot to say about it (“Intervention by other elements produces patterns which are certainly not acceptable in MnE [Modern English] but which certainly do not fall with OE examples of S …V, which requires S (…) Vv”; Mitchell § 3910), and the few pertinent titles tend to have been written by scholars with a background in Germanic philology. Implicitly, the topic is of course dealt with in many discussions of V-2 vs. OV orders. Yet a description on the basis of the clausal brace might actually be superior with respect to descriptive adequacy (cf. Haspelmath 2010). The clausal brace can be seen as the result of the finite

98


verb no longer being in final position, while the non-finite verb remains in its old position at the end of the clause. And although it seems plausible to assume that it is a result of the V-2 development, it is not a necessary one, since the brace is hardly ever found in the North Germanic languages, which are all V-2; it is tempting to speculate that there is a connection between this observation and the loss of OV in these languages (cf. Vennemann 1974b and 1984, Stockwell 1977 and the overview in Lass 1994). Moreover, in Old English the clausal brace is not obligatory, e.g.: (22) We willað secgan eow sum bigspell we want:PL tell:INF you some example

‘We want to tell you an example/a parable.’ (ÆCHom I 14.1212.6) Fischer et al. (2000: 142) observe that the position of object pronouns after the non-finite verb, as in this example, appears to be only possible from late Old English onwards. It has been suggested that exbraciation is the relevant factor in Old English (cf. Stockwell 1977), and indeed this would fit in very well with the nearly complete loss of the clausal brace after the Old English period and also with the long-term tendency for late position of objects, which is already observable in Old English. Conversely, the first position in V-2 clauses is most frequently occupied by þa or by the subject. Cf. Stockwell (1977 and 1984) for a plausible scenario for the factors relevant to the emergence and later development of the Old English order; note also Stockwell’s interesting hypothesis that the apparent operation of V-2 in early Old English may be actually quite deceptive and may be the result of a thematic fronting of verbs. This would provide a good explanation for the many apparent ‘exceptions’ to V-2 and also tie in with Vennemann’s (1984) observation that V-2 and the sentence brace become fully grammaticalized only later (and never quite in English); but cf. also Stockwell & Minkova (1991). Van Kemenade (2003) argues that the position of short sentence-initial adverbs like þa and of subjects is completely independent of the position of the finite verb and suggests that the exclusive occurrence of negative-initial V-F clauses in the older poetic records provides evidence for a more archaic (i.e. less advanced V-2) word order pattern in these texts. Here again the finite verb undergoes positional change, while the position of other elements remains fairly stable and regulated by weight (cf. the positional differences between pronominal and full subjects with respect to the sentence-initial adverbs discussed in van Kemenade 2003: 363–366, and cf. Lenker 2010 for a detailed diachronic discussion of the use of þa and other adverbial connectors).


99

It is, of course, easy to see how this kind of distribution may lead eventually to the SVO order of present-day English, where both the postverbal position of the object and the preverbal position of the subject are fully grammaticalized, but the exact details of the emergence of SVO must still be characterized as rather obscure. It is possible that the fixing of word order is connected to the loss of inflectional morphology; however, this loss cannot account for the emergence of SVO rather than V-2 (or any other order, for that purpose), as the comparison with other present-day Germanic languages with a comparably poor inflectional morphology shows. Similarly, the loss of inflectional morphology may well be a necessary condition for the development towards fixed word order, but it is by no means a sufficient one, as the comparison with Icelandic shows; cf. e.g. the discussion of the development of English direct vs. indirect object marking and the loss of post-head genitives by Allen (2006), who also considers comparative evidence from Faroese. The development of the V-2/V-F distinction has direct consequences for the development of particle placement, as Vennemann (1984) has observed: the clausal brace owes its existence to the coexistence of verb-second and verblate syntax. Since this coexistence continues in German, enhanced by the strengthening of verb-late syntax to verb-last syntax in coordinate clauses, the clausal brace has always thrived and even gained length in German. Contrariwise, since verb-late syntax was given up in Scandinavian and, for a intermediate period, in English in favor of generalized verb-second syntax, the clausal brace lost its support, Behaghel’s First Law made its power felt, and the brace has been on the decline for centuries, with exbraciation, i.e. rightward transposition from the brace, as the mechanism of brace reduction. In English, where the assimilation of subordinate clauses to main clauses was executed with the greatest thoroughness, the reduction of the clausal brace has been carried out to the greatest extreme of all the Germanic languages, with only traces surviving in pronominal object embracing, unmarked indirect object placement, and simple adverb embraciation. (Vennemann 1984: 634)

Vennemann’s observation that the clausal brace becomes more rigid in the course of time (although with noticeable diastratic variation) receives further support from Schmidt’s (1998) and Roelcke’s (1998) observations on the development of the clausal brace in German and its varieties (cf. Schrodt 2004: §§ S 183–284 on the clausal brace in Old High German). The connection between verb placement and particle position is, of course, what might have been expected on the basis of the striking correspondences between these factors found in the present-day Germanic languages. Here it becomes clear, though, that this connection is ultimately not fortuitous but

100


epiphenomenal to historical changes in the positions of verbs and object (for an account of the reverse change in Chinese from SVO to SOV, which was concomitant with the emergence of postpositions, verbal suffixes and agglutinative tendencies, see Li & Thompson 1974). In the light of such observations, it becomes very questionable indeed whether there is such a thing as the ‘rise’ of postverbal particles, as some discussions of the topic suggest, and this question will be dealt with shortly. 3.2.3. The rise of Modern English word order But let us first return briefly to the more general long term developments in English word order. Table 3-2 below shows an early and rough attempt at describing the overall tendencies in the development of English word order (Fries 1940; percentages in the second column sic). Fries does not elaborate on his database (neither on its size nor on the texts included) but since his observations are in principle confirmed by later research on the changes in word order it seems admissible to reprint his figures here as an illustration of the long-term development. More reliable figures are difficult to procure, for the reasons identified by Fischer (1992: 371–372): “It will not be possible to provide exact percentages and tables since the amount of research devoted to Middle English word order is still scanty and, more problematically, shows great methodological variation”. Thus, since reliable in-depth studies of word order in Middle English are surprisingly scarce, much of the discussion of the developments presented here will remain somewhat vague, not least also because of the considerable complexity of the Middle English dialect situation. The generative discussions of some of the relevant changes, by Kroch & Taylor (2000), Trips (2002), Roberts (2007) or S. Fischer (2010) all rely on a rather small set of data available through electronic corpora. Fries’ table shows very clearly, though, that by the end of the Middle English period OV order is almost completely lost. But, as we have seen in the preceding discussion of Old English word order, it is necessary to distinguish between different clause types, since the loss of OV is already well under way in Old English main clauses, where it is only left as a (quantitatively) marginal word order, which may be restricted to its use as a focusing device. In fact, already in early Middle English the use of OV may be considerably more restricted than Fries’ figures indicate. For the first half of the 12th century, Mitchell (1964) in his study of word order in the Peterborough Chronicle finds that in subclauses 28 per cent of the word orders would be impossible in present-day English in the First


101

Table 3-2. OV and VO orders in English (according to Fries 1940: 201) c.1000

c.1200

c.1300

c.1400

c.1500

OV

52.5%

52.7%

>40%

14.3%

1.87%

VO

47.5%

46.3%

<60%

85.7%

98.13%

Continuation (1122–1131), and only 12 per cent in the Second Continuation (1132–1154), in stark contrast to a mere 41 per cent of ostensibly ‘modern’ word orders in late Old English. Similar results have been reported by Palmatier (1969) for the Ormulum (late 12th century), while MacLeish (1969) finds that in late Middle English the brace construction is extremely rare in main clauses while it does occur in subordinate clauses, especially with pronominal objects. But Palmatier and MacLeish do not draw finer distinctions between various types of clauses, despite Mitchell’s (1964) demonstration that these distinctions are of essential importance in medieval English. With all these studies it is rather difficult to use them for more than the general impressions. Criticizing Clark’s claims as to the modernity of the syntax of the Second Continuation of the Peterborough Chronicle (Clark 1958: lxvi), Mitchell is rather keen on stressing the continuity in word order between Old English (cf. also the remarks in Mitchell 1985: §§ 3948–3951) and the language of the Peterborough Chronicle and leaves it largely undiscussed whether his ‘old’ orders in the Peterborough Chronicle are really old or just orders which appear to be old but really follow different and more modern ordering principles (cf. the discussion in Fischer 1992: 372–373 and also the criticism in Denison 1993: 34–35). Likewise, Stockwell & Minkova (1991: 368) draw attention to the inherent contradiction between Mitchell’s claim that the syntax is archaic and his observation that the incidence of ‘old’ orders in subordinate clauses has decreased significantly. Thus while V-F is already marginal in Old English main clauses, it becomes rapidly marginal in Middle English in all kinds of clauses. Already Reszkiewicz (1962) in his study of word order in the late Middle English Book of Margery Kempe had observed that SOV occurs only once in a subordinate clause with a pronominal object, while SvO(V) and OvS(V) (i.e. verb second with clausal brace) are still frequent with all types of objects (Reszkiewicz 1962: 39). On the whole, it seems that the deviations from SVO (V-2, SOV, OSV, etc.) increasingly come to be used as stylistically marked and archaic orders; cf. the overview in Fischer (1992: 372–375). Fischer summarizes the Middle English development as follows:

102


These facts reveal a clear tendency towards SVO in both main and subclauses. The vestiges of SOV order that remain give us some idea about the underlying causes of the development. The far greater frequency of preverbal pronominal objects was already apparent in Old English and has been connected with ordering principles related to weight (cliticization) and theme-rheme structure, which are themselves presumably interrelated … Thus, NPs that are light (pronominals) and represent given information (pronominals are usually anaphoric) tend to occur early in the phrase, whereas heavy NPs and NPs containing new information tend to occur late. The question is why this multivariable but pragmatic Old English system … was slowly being replaced by an almost invariable one in the later periods. (Fischer 1992: 373–374)

It is remarkable, though, that V-2 continues for a considerably longer time than all the other variants deriving from Old English. But OV orders are still current at the end of the 14th century, e.g.: (23) … the hooly blisful martir for to seeke, that hem hath holpen whan that they were seeke. (Chaucer, GP, 18) ‘… to seek the holy blessed martyr, who (has) helped them when they were ill.’ (24) A clerk hadde litherly biset his whyle, but if he koude a carpenter bigyle (Chaucer, MT, 3300) ‘A scholar had wasted his time, unless he could deceive a carpenter.’ (25) … by whiche thynges I may my persone and myn hous so kepen and deffenden that myne enemys shul been in drede myn hous for to approche. (Chaucer, Mel 1334) ‘… thus I can protect and defend myself and my house in such a manner that my enemies will be afraid of approaching my house.’ As these examples indicate, where OV is still found it tends to occur with object pronouns and in subordinate clauses, as in (23), and there are also subordinate clauses with full objects with a clausal brace, as in (24). As (25) shows, the clausal brace may also occur in main clauses. The use of OV is not restricted to poetic language, as these examples from Chaucer may misleadingly indicate (cf. MacLeish 1969 for more examples of the full range of positional variants in late Middle English). However, OV becomes rare in English early in the Early Modern English period, except for poetic registers (V-2 still exists in a number of present-day English constructions whose historical continuity to Old English is clearly visible, while the emergence of dummy subjects coincides with the loss of V-2, cf. e.g. Fischer 1992: 376). For the persistence of V-2 patterns in Early


103

Modern English, see e.g. Jacobsson (1951) or Bækken (1998), for a discussion of present-day English V-2 constructions which are not the direct historical reflexes of earlier V-2, see Stockwell (1984); cf. also Nevalainen (1997) on Early Modern English developments. Trips (2002) provides an interesting in-depth study of the development from OV to VO in the early Middle English dialects, where she is rather quick to attribute the changes to Scandinavian influence. Her argument rests largely on anachronistic parallels in the later Scandinavian languages, including Old Norse, and the historical input available from Old English is consistently downplayed; this criticism also goes for other syntactic studies positing language contact with Old Norse as a relevant factor, which tend to present the language contact scenario in such a way that is fits their contact agenda, e.g. Kroch & Taylor (2000) or Kroch, Taylor & Ringe (2000). On the whole, old V-2 disappears last in clauses introduced by short adverbial elements, where it was most regular in Old English (cf. the typical Old English orders in main clauses introduced by þa, þære, ne, etc.), which lends some support to the notion that SVO is the result of the generalization of frequent patterns. 3.3.

The position of the particle in medieval English

The development of the position of the particle in verb-particle constructions can be seen as an immediate consequence of the development of the more general patterns of word order described so far (even if this claim had better be phrased somewhat more carefully, cf. the discussion of some of the unresolved issues below). In general the changing position of the particle is clearly connected to the loss of OV and the related changes described in the preceding section. Consequently, what appears to be a positional change of the particle (from ‘pre-’ to ‘postposition’, etc.), on closer inspection turns out to be a set of positional changes of the elements of the verb phrase and of the postpositional trends observable in objects. Approaches like Von Schon’s (1977), where the particle rather than the other elements of the clause is assumed to shift (Von Schon 1977: 222 assumes a ‘First Particle Shift’ to postverbal position after finite verbs in the 7th century and a ‘Second Particle Shift’ after non-finite verbs “some centuries later”) have been roundly criticized for their linguistic, methodological and philological shortcomings (see e.g. Denison 1981: 113–114 and 118–132) and will not be discussed here further.

104


O prt V v

>

v O prt V

(>

prt V v O)

>

v V O prt

OV

>

v V prt O v prt V O

VO V-2 extraposition v…V > vV

Figure 3-2. Positional changes and their causes

Based on the observation that the historical development of order within the clause is mainly to do with the changes in the position of the finite verb and the subsequent emergence of the clausal brace, which in its turn becomes dissolved by the exbraciation of brace-internal elements, the overall development of the orders of verbs and particles in relation to objects can be summarized as in Figure 3-2 (with the general positional changes below and their consequences for particle position above the dotted line; v ‘finite verb’, V ‘non-finite verb’). Although somewhat idealized in its neatness, this general path of development ties in completely with the cross-linguistic observations discussed above, the comparative reconstruction of changes in early Germanic word order, the historical evidence from other Germanic languages and the long-term changes observable in English (the regularities shown in the figure also apply when there is no object or no non-finite verb). Moreover, it is characterized by considerable explanatory elegance, since it derives the basic positional properties of the verbparticle construction from other, well-attested developments in the overall organization of English clauses, while it also accounts for the absence of certain patterns, like e.g. prt-O-V (cf. Hiltunen 1983a: 117); i.e., this is an explanation, since it identifies the relevant causal factors in the change – the fact that the reasons in the development of these factors in their turn are open to dispute is irrelevant in the present context. Although most treatments of the Old English particle verbs devote some space to listing the differences between particle verbs and inseparable prefix verbs (cf. e.g. the relevant sections in Denison 1981, Hiltunen 1983a, Mitchell 1985 or Fischer et al. 2000), all such criteria can be reduced to the simple observation that prefixes are unstressed and fused to the verbal stem, while particles are not (see also the discussion of this in Chapter 5 below). Problems of analysis are mainly to be encountered in cases of synchronic layering, where a preverb may be both a free particle and an


105

inseparable prefix (cf. Section 3.1 above). It seems that such criteria are usually of more relevance to the identification of prefixes than of particles (cf. e.g. Mitchell 1985: §§ 1072–1073), and the parallels to the situation in the present-day Continental West Germanic languages are obvious. However, the following differences between particle verbs and prefix verbs regarding word order are nevertheless worthy of being pointed out again, especially since they may serve to illuminate some of the characteristics of the Old English particle verbs which are not predictable in detail by a comparison with the present-day continental West Germanic languages: First, the negative particle ne and the infinitival particle to always immediately precede the finite verb; thus with preverbal particles the order is prt-ne/to-V (cf. examples (29) and (32) below), while the corresponding order with prefixes is ne/to-prefix-verb. Moreover, other elements as well may intervene between the preverbal particle and the verb, most notably modals, stranded prepositions and pronouns (cf. examples (30) and (31) below), while once again the prefixes are by definition inseparable; by contrast, cf. van der Auwera’s (1995) comparative discussion of preverbs in present-day Dutch and German. Indeed, the positional properties of a vast majority of verb-particle constructions found throughout the history of English confirm the validity of the diachronic path sketched above. Hiltunen (1983a) has studied the largest corpus of Old English texts with respect to particle verbs so far. His corpus contains twelve Old English and fifteen early Middle English texts with the aim of including “the chief variants of prose style between c. 800 and 1250” (Hiltunen (1983a: 31). With few exceptions, most later studies have tended to rely on the material available in Hiltunen’s study, which, although there is no reason to doubt the empirical soundness of his description, is in some respects somewhat unwieldy, especially since Hiltunen basically accepts Mitchell’s procedure (cf. e.g. Mitchell 1964, and still 1985) of distinguishing between different positional variants with reference to four frames of the type ‘S.V.’, ‘S…V’, ‘S…V…’ and ‘V.S.’. This, as we have seen above, does not quite capture the relevant factors for the description of the position of clause elements, although it can often be transposed into more appropriate terms. But many of his statistics are, due to this classificatory weakness, not as illuminating as one could wish, especially since they do not distinguish between finite and non-finite verbs. The following examples are taken from Hiltunen (1983a); the glosses are mine. In Old English the oldest order object–particle–verb occurs typically in subordinate clauses, e.g.:

106


(26) He … bebead he

implored

him, on Godes naman, þæt

he

him

he

on

God’s

ðone

cwelmbæran hlaf

DET.ACC.SG

deadly:ACC

name

aweg

loaf[ACC] away

that

bære carry:PRET.SUBJ.SG

‘He implored him, in the name of God, to carry away the deadly loaf.’ (ÆCHom ii 162.23) In subordinate V-F clauses, this pattern is extremely regular and there is hardly a single solid exception. Hiltunen (1983a: 116) finds only four instances with a postposed particle in his sizeable corpus. But of these one is not V-F (his example 15), one involves the adverb togædere (his example 17), which had perhaps better be not classified as a part of a verbparticle construction, and in the remaining two the particle is followed by a prepositional phrase, which might be a special case anyway (cf. the discussion in Fischer et al.: Ch. 6). As is to be expected considering the diachronic factors responsible for the ordering, the rule also holds for intransitive verbs, as in (27), and for those instances where the sequence prt–V is interrupted by an intervening element, as in (28)–(31): (27) for ðon

þa feawan

because of the few

hæfdon

þe

þær

REL

there out [away-]flown:PST.PTCP

ut

oþflugon

eft

had:PRET.PL afterwards

‘because of the few who had flown out from there afterwards’ (Or 96.3) (28) Uton we forþon

geþencean hwylc handlean

let.us we therefore remember

we him

what recompense we him

forþ to berenne habban forth to carry:INF have:PRES.PL

‘Let us therefore remember what recompense we have to offer him.’ (BlHom 91.13) (29) forðæm hio because

DEM

nanne swetne

wæsðm forð ne

no:ACC sweet:ACC fruit

bringð

forth NEG brings

‘because it does not bring forth sweet fruit’ (CP 341.23) (30) ær

he

before he

ut

wolde

faran to gefeohte

out wanted

go:INF to fight:DAT

‘before he wanted to go out to the fight’ (Or 232.4)


(31) &

þa ut

he

and then out he

gan

107

wolde

go:INF wanted

‘and then he wanted to go out’ (Bede 396.29) Since V-2 and exbraciation first and foremost apply in main clauses, the sequences v-prt-O and v-O-prt are typically found there, e.g.: (32) þa

ahof

Paulus up his heafod

there raised

Paulus up

his head

‘Paulus raised his head.’ (BlHom 187.35) (33) He nam up Sca Kyneburh & S. Kynesuið þe he

took

up

lægen in Castra

St. Cyneburh and St. Cyneswith REL lay:PL in Castor

‘He exhumed St. C. and St. C., who were buried at Castor.’ (ChronE 117.33 [963]) But non-finite verbs at the end of the clausal brace continue to have preverbal particles, e.g.: (34) þa then

nolde

he

adun

asceotan

NEG-wanted

he

down

shoot:INF

‘He did not want to fall down.’ (ÆCHom i, 170.25) (35) Wæs he was

he

ærest

up

ahefen

of þæm huse

first

up

taken

of the

house

‘He was first taken up from the house.’ (Bede 288.13) In V-F main clauses with the verb in final position, the particle will likewise be typically found in preverbal position, e.g.: (36) he he

mid mildheortnesse hine up ahof with compassion

him

up

took

‘He raised him with compassion.’ (ApT 18.9) As is to be expected from the general tendencies of word order change discussed above, this is fundamentally a relic order (also with regard to the verb), which is recessive already in Old English. Conversely, the spread of

108


VO orders in subordinate clauses goes along with the expected re-ordering of verb and particle, e.g.:5 (37) þæt he wearp þæt sweord onweg þæt he on handa hæfde that he threw

that sword

away

that he on hand

had

‘that he threw away the sword which he held in his hand’ (Bede 38.20) (38) for ðan þe se stream berð because

aweg Placidum

the stream carries away Placidus

‘because the stream carries Placidus away’ (ÆCHom ii, 160.5) (39) forðam þe him because

him

burston ut burst

butu his eagan

out both

his eyes

‘because both his eyes fell out’ (ÆLS i 422.120) And in clauses introduced by and, ac or ne, the order of verb and particle is quite regular according to the position of the verb. Thus we find a preverbal particle in (40), where the verb follows the object, but a postverbal particle in (41), where the verb precedes the object: (40) &

micele here huþe mid

and much

army booty with

him

aweg

them away

læddon led:PL

‘and took much war-booty away with them’ (ChronE 111.13 [943]) (41) &

efsones he

and soon after he

let him

ut

þurhc wærse red

let

out

through worse

him

advice

‘and soon after, on worse advice, he let him out’ (ChronE 276.12 [1140])

5

This touches upon a number of highly contested issues in historical syntax, which must be left undiscussed here. Especially in those approaches where the particle is believed to mark the base position of the verb evidence from subordinate clauses like the one given in these examples has led to some discussion over the question whether this is due to extraposition of objects (cf. van Kemenade 1987) or to V-2 (cf. Pintzuk 1991 and the brief overview in Fischer et al. 2000: 194–197); as the phrasing of my description shows, I tend to follow Pintzuk’s thorough analysis on this point.


109

There is, however, one relatively rare order which cannot be accounted for easily by the general processes outlined above. In a few V-2 clauses with a finite particle verb the particle occurs preverbally (e.g. example (42) from Hiltunen 1983a: 118, example (43) from Fischer et al. 2000: 187, and example (44) from Koopman 1985: 95). Elenbaas in her discussion of verbparticle constructions in the York-Toronto-Helsinki Parsed Corpus of Old English Prose (which contains 1.5 million words) only finds 24 clear examples of this order (cf. Elenbaas 2007: 191), while Hiltunen’s classification renders it impossible to draw any solid conclusions on the basis of his figures. (42) þæt heo onweg adyde þa gemynd þara that DEM away

did

DET

treowleasra

memory DET.GEN.PL faithless:GEN.PL

cyninga king:GEN.PL

‘that it removed the memory of the faithless kings’ (Bede 154.10) (43) Stephanus upastah þurh Stephanus up-rose

his blod gewuldorgebeagod

through his blood glory-crowned

‘Stephanus ascended, crowned with glory through his blood.’ (ÆCHom I, 3.56.31) (44) and aweg gelædde micelne dæl þæs and away led:SG

folces

to his rice

big:ACC part DET.GEN.SG people’s to his realm

‘and led away a great part of the people to his realm’ (ÆCHom II, 18.21) So far no entirely satisfactory explanation has been offered for the preverbal position of particles in V-2 clauses, and the discussions in the literature tend to be rather vague and adhoc-ish. The first principled account was attempted by Koopman (1985), who assumed that the verb and the particle are generated under a single node in the order prt–V and are moved together. However, this obviously creates problems with the analysis of the other orderings (cf. the critique of Koopman’s suggestion in Denison 1991: 36–38 and in Elenbaas 2007: 151–152). Another analysis, mentioned but not adopted by Fischer et al. (2000: 187) would consist in analysing such instances as (inseparable) prefixes. This however, is clearly unjustified in view of the separability of these combinations of verbs and particles in all other cases where they appear, while Elenbaas’ (2007: 191) proposed scenario of “the option of particle incorporation, by which the particle undergoes head-movement and attached [sic] to the left of the

110


(lexically decomposed) verb” merely rephrases the problem in theoretically more consistent terms. Her earlier suggestion that prt-V-O orders reflects ‘basic word order’ with an extraposed object (Elenbaas 2007: 166–167) appears to be historically more appealing. It can be pointed out, though, that these clauses very typically contain rather long elements after the verb, and for want of a better solution they had perhaps be seen best as V-F clauses with a postposed heavy object. This suggestion in fact ties in with the typical exbraciation of objects, cf. the discussion above and the description of weight ordering as a fundamental property of Old English syntax in Reszkiewicz (1966) and the quantitative analysis in Pintzuk & Taylor (2006). It seems that there is a tendency for Old English scribes to write particles and verbs as one word if the particle immediately precedes the verb, cf. e.g. example (43). This point, which has been noted for long but never been pursued in detail, is clearly in need of further investigation (cf. Gneuss 1973: 18). The issue is usually obscured by the modern editions. But why should the spelling of Anglo-Saxon manuscripts be relevant for a modern linguistic analysis? Although this kind of information is typically normalized in one way or the other by modern editors and subsequently in the electronic corpora as well (including the DOE Corpus, cf. the discussion in Chapter 4 below), here is a case where modern linguists have the opportunity to catch a glimpse of native speaker intuitions of a dead language. The evidence may turn out to be inconclusive on closer inspection, but simply ignoring it would be careless (note that otherwise the Anglo-Saxon scribes’ decisions on what constitutes a ‘word’ boundary are in general tacitly accepted unless there is good reason for not doing so). In this context it may also be relevant that Anglo-Saxon scribes tended to use spacing in Old English glosses to indicate the morphological structure of glossed Latin words, as Kornexl (2003) points out. Cf. also the discussions of the position of the role of textual evidence as the very basis of language history by Fischer (2004) and Lass (2004), and Fleischer’s (2009) exemplary analysis of palaeographical clues to prosody in Old High German manuscripts, which also includes a discussion of word separation. Minkova (2008: 26) takes an example like (43) as evidence that “up could be a prefix, or it could be a particle, and the word-order tests for OE allow both interpretations”. Indeed, based on this sentence alone this observations would be fully justified (in fact, an analysis of up as a prefix would presumably be more appropriate). But considering that there is elsewhere in the Old English corpus very little unambiguous evidence for prefix status of up, but a lot of evidence in favour of its particle status, Minkova’s claim does not appear to be fully justified (although the development of up into a


111

prefix is possible in principle, as its further history shows, cf. the discussion in Chapter 4 below). The long-term loss of preverbal particles, then, can be seen in Figure 3-3 below, which summarizes the positional developments observed by Hiltunen (1983a) in his study of verb-particle constructions in Old and Middle English. Although Hiltunen does not distinguish between finite and nonfinite verbs in the different clause types, the overall tendency is very clear from the percentages shown in the figure. The category ‘main clauses’ (‘principal clauses’ in Hiltunen’s terminology) also comprises coordinated clauses; consequently, the figures for the main clauses proper are really higher (cf. the discussion of word order in different clause types above and the figures for the individual clause types in Hiltunen 1983a: 106–110). Of course, these figures reflect the loss of V-F orders and the establishment of SVO by the end of the Middle English period. The findings of Elenbaas’ (2007) detailed quantitative study of the development of the particle position in the Middle English period based on the Penn-Helsinki Parsed Corpus of Middle English confirm the general trends noted here. Between the periods ‘M1’ (1150–1250) and ‘M3’ (1350– 1420) the percentage of postverbal particles in main clauses goes up from 93.3 to 98.4 per cent, in coordinate main clauses from 96.7 to 100 per cent and in subordinate clauses from 87.3 to 99.6 per cent (cf. Table 5 in Elenbaas 2007: 239, where the figures, however, are in part slightly at odds with those given in her subsequent discussion) – i.e., already at the beginning of the Middle English period postverbal position of the particles is clearly dominant in all clause types and it becomes the norm by the end of the period. In fact, on closer inspection it turns out that Elenbaas’ corpus hardly yields a single unambiguous example of a preverbal particle: there are practically no instances of preposed particles which are not immediately adjacent to the verb (there appear to be altogether only two tokens in her 1.2-million word corpus; each of these two occurrences is likely to be due to stylistic factors, cf. the examples given in Elenbaas 2007: 243 and 261). The majority of the remaining instances, where the particle is immediately adjacent to the verb, are either of doubtful relevance (e.g. close translations from Latin) or may in fact be analysed as prefix verbs. At any rate it seems beyond doubt that after the early Middle English ‘M1’ period there is hardly any solid evidence for preverbal particles in the corpus. Taking into account the corroborative evidence from earlier studies discussed above, there is therefore good reason for the assumption that by late Middle English preverbal particles have become marginal in English and that their occurrence is conditioned exclusively by stylistic rather than syntactic factors.

112

The development of postposed particles 100

92

80

87

58

60

–––– main clauses ........ subordinate clauses

44 40 33

20 13

0 eOE

lOE

eME

Figure 3-3. Development of the order ‘verb-particle’ from early Old English to late Middle English (according to Hiltunen 1983a: 111)

Already in the Second Continuation of the Peterborough Chronicle Denison (1981: 136) finds that “[i]n seventeen cases out of nineteen, the particle follows the verb” and that “[i]n both of the cases where it precedes the verb, the verb is finite and clause-final and the clause is a coordinate or subordinate one”. Contrary to Old English usage, the particle normally follows the verb in this 12th-century text, e.g.:6 (45) þat he that he

neure mare sculde cumen ut never

more should come

out

‘that he would never come out’ (ChronE 1140.45) (46) he suor …

þat he

he swore … that he

ealle his castles sculde iiuen up all

his castles should give

up

‘He swore … that he would give up all his castles.’ (ChronE 1140.40) The last example provides evidence for the sequence O-v-V-prt and is extremely difficult to classify; cf. Denison’s observation that this is about the only justification for Von Schon’s assumption of ‘particle shifts’ (1981: 137). This kind of evidence provides a formidable obstacle to neat correla6

Examples quoted after Denison (1981: 136–137), who following Mitchell (1964) analyses (45) as an instance of ‘S…V’, although this analysis somehow begs the question, since the clause could be just as well seen as an instance of SVO.


113

tions between general word order rules and the position of the particle, and in the literature there is a marked reluctance to discuss this observation in detail. One possible explanation may be that the position between the pronominal subject and the verb is a focus position in subordinate clauses; cf. the quantificational constituents neure mare in (45) and ealle his castles in (46), where this would be conceivable. (And cf. the corresponding reverse order in Old English, as e.g. in Heo of genimð þone scruf & þone teter [Herbarium 46.6.957, quoted after Elenbaas 2007: 166; lit. ‘it off-takes the scabies and the eczema’], where a long object is extraposed while the particle remains in preverbal position.) The implication of this proposal, though, would be that these clauses are quite modern to start with (with the order v-V-prt) and then turn into a pseudo-archaic type of S…V through topicalization; obviously, this suggestion is very much on the speculative side. In the more recent early Middle English Ormulum (late 12th century), Denison finds 512 verb-particle constructions altogether, of which only 28 (i.e. six per cent) have preposed particles; these appear almost without exception with non-finite verbs in subordinate clauses, i.e. where they would be also expected in Old English. But even more than in the Second Continuation of the Peterborough Chronicle, in the Ormulum this order is clearly exceptional, and it is possibly triggered by metrical factors (cf. Denison 1981: 180). Also, the Ormulum is from a different dialect area than the Peterborough Chronicle and it may very well be that this difference is due to diatopic rather than diachronic factors. And 300 years or so later, in a selection from the Paston Letters, Denison (1981: 139) finds no evidence at all for preposed particles except in a handful of formulaic participle constructions (e.g. a-bovyn wreten, etc.). In sum, all the orders typical of present-day English essentially already exist in Old English. This applies also to the marked orders with topicalized particle, which, however, are attested only rarely in Old English, e.g.: (47) Ut eode se sædere his sæd out went the sower

his seed

to sawenne to sow

‘Out went the sower, to sow his seed.’ (Mk (C) 4.3) (48) Up aræmde Abraham þa up

rose

Abraham

then

‘Up rose Abraham!’ (Ex 411)

114


Fischer & van der Wurff (2006: 191) claim that in Old English particles in initial position occur only in poetry, cf. example (48). And indeed an example like (47) (from Hiltunen 1983a: 121) cannot be taken as unambiguous counterevidence to that claim, since ut eode here translates Latin exivit (cf. Roberts 1936: 476). Whether this is sufficient ground for dismissing the example as irrelevant, as Roberts does, is hard to judge, though (for further examples, see Denison 1981: 140–143, who discusses some examples from Old English poetry and several examples from early Middle English onwards, noting that as in Modern English the particles are uniformly used literally, usually in combination with verbs of motion). With regard to the position of postverbal particles and objects, there is no evidence at all in Old and Middle English for pronominal objects which follow a postposed particle. This order may have been possible under the very restricted conditions where it is also possible in present-day English. However, considering the wealth of evidence available for Old English, it seems safe to conclude that otherwise this order was impossible in Old English. The lack of evidence for this ordering can be taken as further confirmation for the hypothesis discussed above that extraposition of heavy objects was an essential factor in the dissolution of the clausal brace observable in the history of English. In other words, the orders in the Middle English examples (49) and (50) are among the last relics of the clausal brace in English, while (51) is in line with the otherwise almost complete loss of the brace in English (cf. Section 3.2.2 above). (49) &

heo

and they

holden

hire up

held

her

up

‘and they held her up’ (Ancr. (Nero) 62.34) (50) &

hef

hire honden up

and raised her hands

up

‘and raised her hands’ (St Marg. (1) 22.9) (51) heo hef

up hire hond

she raised up

her hand

‘she raised her hand’ (St Marg. (1) 19.22) What makes this suggestion plausible is the comparison to German with its strengthening of the brace, where exbraciation as in example (51) only appears to be possible with very heavy objects. That is to say, with regard


115

to the position of particles, English has been extremely conservative throughout its history, while the order of other elements has been subject to considerable (and sometimes not fully explained) changes. But if we look at any Old English sentence with an S-V-O order, or any other order which would also be possible in present-day English, the position of the particle will be highly likely to be identical as well, cf. e.g. examples (33), (38) and (41) above, etc. The reasons for the apparent lack of complete synchronicity of the disappearance of preverbal particles and the establishment of SVO remain to be explained, though. It is probably fair to state that most syntactic studies of verb-particle combinations in Old and Middle English are primarily concerned with the relative positions of verb and particle as a means of establishing the ‘underlying’ word order; i.e., what is at stake there is not the history of the phrasal verbs but the history of English word order. But although many of these studies provide valuable insights into the syntactic aspects of the development, there has been a considerable degree of concentration in the literature on a few topics while others have been largely ignored. E.g., there are hardly any in-depth studies at all of the development of the positional properties of adverbials, despite Hopper’s (1975: 73–74) claim that in early Germanic the positional behaviour of directional complements is identical to that of preverbs. In other words, although the main changes in English word order are intrinsically connected to changes in the order of verbs and objects, it is possible – and indeed quite likely – that there are a number of changes affecting other clause elements. But as yet there is little research on such issues, and without it a satisfactory explanation of the lack of synchronicity may very well be beyond our reach. 3.4.

Conclusion

In this chapter it has been argued that throughout the history of English the position of the particle has remained essentially fixed, and that what appears as positional shifts of the particles is really epiphenomenal to positional changes of other elements of the clause, most notably the finite verb. From this perspective the positional characteristics of English particles are entirely in line with those of other Germanic languages, and the comparative observations made in the preceding chapter fully tie in with the evidence from earlier Germanic and from Old English discussed in the present one. There is very little reason to assume that the remaining problems of the history of particle placement, which are mainly to be found in late Old English and early Middle English, are an exception to this

116


observation, and these problems had best be regarded as problems connected to verb and object position in an intermediate stage from a language characterized by mixed V-2/V-F to a strict SVO language. Moreover, the development to be observed in the history of English ties in completely with the cross-linguistically well-attested development of preverbs and can therefore be characterized as highly regular in almost every respect, both from a comparative and a typological point of view.

Chapter 4 Writing the history of the phrasal verb The notion that the phrasal verbs are at the same time both typically Germanic and uniquely English is deeply rooted in the linguistic literature, where it has been traditionally taken for granted since the beginnings of the linguistic interest in the construction. The present chapter thus begins with a brief discussion of the first systematic synchronic and diachronic study of the phrasal verb, which has exerted a continual, if at times only implicit, influence ever since, as the subsequent discussion of the treatment of phrasal verbs in some widely-used handbooks and in the historical dictionaries will show. 4.1.

A classic study: Kennedy (1920)

More than by any other publication, the study of phrasal verbs has been shaped by Kennedy’s monograph on The Modern English verb-adverb combination (1920), which is the first in-depth account of the topic and which despite its title it is also strongly concerned with history.1 Strang (1970: 275) calls Kennedy’s study “the standard work on the subject”; cf. e.g. the more recent similar characterization by Fischer (1992: 398), among many others. As regards its influence, the only serious contender would probably be Bolinger’s (1971) seminal study of phrasal verbs in presentday English, which provides a considerably more sophisticated linguistic analysis than Kennedy’s study (which, however, is among the only 16 titles cited by Bolinger). But Kennedy sets the agenda for all subsequent research and he is the first to discuss virtually all the questions addressed in later studies up until today: e.g., the development of the English pattern from early Germanic onwards and the connection of the process to the long-term 1

Kennedy’s short bibliography very much reflects the little interest the topic had excited in the earlier literature. He refers to Koch (1891), Harrison (1892), Eitrem (1903), Webster (1909), Rickard, Shockley & Pratt (1909– 1910), Ellinger (1910), Kirkpatrick (1912), Curme (1914) and the NED (viz. OED1, incomplete in 1920); he does not mention the accounts by 18th century grammarians (cf. the discussion in Chapter 6 below). On the older grammatical traditions in German and Swedish, see Tiisala (2008) and McLelland (2008), who, focusing on various kinds of adverbs, also discuss verb particles and provide further literature.

118

Writing the history of the phrasal verb

typological shift from syntheticity to analyticity in English, the position of the particle, characteristic features of the verb elements like their etymologies and their length, and the functions of the particles. He also discusses the relation of phrasal verbs to synonymous hard words and, in a more speculative vein, their overall status in the lexicon. Not surprisingly, many points of Kennedy’s study are now superseded. But both some of his basic assumptions and some of his findings have rarely been questioned in later research. To my mind it should be viewed rather critically that there is an undercurrent of tacit assumptions about phrasal verbs in most subsequent 20th-century research. The present section will be deliberately selective, and the focus in the following will be on those aspects which have become the received wisdom in many studies of the phrasal verbs, although at closer scrutiny they turn out to be rather dubious. In particular, Kennedy discusses the development of the English verb-particle construction on the basis of the scanty earlier literature and of his own collection of data. But his study is almost completely ahistorical in some essential respects. He takes as his point of departure observations on contemporary English, which in itself is of course an entirely acceptable procedure. But these observations are then simply mapped onto the past and the question whether there were differences in the course of history is hardly ever raised. This applies to a few comparatively minor points of the study, but also to the central assumptions on the diachronic syntactic and stylistic properties of the construction. 4.1.1. The ‘rise’ of the phrasal verbs As regards syntax, Kennedy describes some of the typical characteristics of the present-day English phrasal verb, including the post-verbal position of the particle. Kennedy’s description is not quite correct, though, since he states that non-pronominal objects always follow the particle except when they are very short while pronominal objects may never follow the particle. Consequently his examples for ‘exceptions’ from these rules in earlier periods are misleading, since the supposedly more irregular historical usage would be perfectly conceivable in present-day English as well. Also, although Kennedy’s study is clearly devoted to phrasal verbs (in the sense the term is used in the present study), he does not distinguish systematically between phrasal and prepositional verbs, and a few of his examples belong to the latter group. On the whole Kennedy observes a long-term increase in the frequency (both type and token) of the construction. Relying on Curme’s (1914)


119

discussion of the origins of ‘verbal compounds’ in Germanic, he claims that “the tendency to stress an adverbial particle following the verb more or less closely, worked, during the Old English and early Middle English periods, toward the elimination of the verb with unstressed, inseparable particle and the gradual increase of the verb-adverb combination” (Kennedy 1920: 11). In a small number of sample texts, then, Kennedy observes a marked increase in separable combinations at the end of the Middle English period, at the expense of inseparable formations from native material. In Old English he counts very few separable combinations as opposed to a high number of inseparable ones, and in Middle English the overall number of separable combinations (i.e. phrasal verbs) seems to remain rather low, but an assessment of the situation is difficult, he states, because of the French lexical influence and because of the differences in text types. Nevertheless, Kennedy speculates that the development of the phrasal verbs was slowed down by the influx of Romance ‘compounds’ (i.e. prefixed verbs). By the fifteenth century, “the combination begins to show real strength, altho [sic] it is evidently a part of the language of the common man, even as it has been ever since” (Kennedy 1920: 13). Kennedy’s account of the rise of the phrasal verbs is seriously flawed by the fact that all constructions with a particle or a prefix in preverbal position are counted as ‘compounds’ while only those instances where the particle is in postposition are counted as ‘verb-adverb combinations’; he concludes that in Old English “occurrences of the verb-adverb combination are practically nil” (Kennedy 1920: 12). That is to say, the post-verbal position of the particle is quite ahistorically taken as an unchanging characteristic of the construction. As a result, Kennedy’s discussion creates the impression that the phrasal verb is a new phenomenon in English and something specific to this language alone, i.e. that there is a ‘rise’ of the phrasal verb from the Middle English period onwards. To take just one prominent example (among very many others) of the popular afterlife of such notions: Burchfield (Fowler 1996, s.v. phrasal verb) states that the “earliest example known to me is to give up ‘to surrender’, which is recorded in 1154 … [t]he type thrived in the centuries that followed”. Moreover, the frequencies are based on a quite eclectic selection of texts, and Kennedy counts only those particles which occur in Modern English. But of course these observations do in no way reflect actual Old English usage (cf. the discussion in Chapter 3 above and see Kennedy 1920: 11–14 for his exact figures; a chart with these figures and with percentages is provided in Konishi 1958: 120). The inventory of particles is also subject to diachronic change, and whilst some of the Old English particles have disappeared from this inventory, others have entered it. Counting only the

120


particles from the Modern English inventory in a diachronic sample will thus quite predictably result in an observed increase of the frequencies, but this observation will be completely meaningless in regard to the quantitative development of the construction type. One example is the exclusion of phrasal verbs with the particle forth from his diachronic frequencies because of its loss of productivity in contemporary English, which seems acceptable in a study devoted to ‘Modern English’ until one realizes that this exclusion must have considerably slanted his historical figures for the relative frequencies of the construction type as a whole (which also used to be referred to in the literature for a long time, cf. Section 4.1.3 below; for an analysis of the decline of forth, see Akimoto 2006). Meaningful statements about the overall quantitative development of the verb-particle construction should be based on counts of all particle verbs in the respective samples (i.e. with all types of particles used, and with the particles preceding and following the verb) in mutually comparable text types. Kennedy’s counts meet none of these requirements. Since there is also no later study of the developments from Old English to Modern English which would meet all of them, it can only be asserted that so far longterm changes in the frequency of occurrence of the English particle verbs are not known. Yet, ever since Kennedy’s study changes in the frequencies of phrasal verbs within different periods have often been posited. This is not to say that there is definitely no increase. Indeed, considering the loss of many Old English (inseparable) prefixes which were in part functionally equivalent to some of the particles, this is even quite likely, cf. Chapter 5 below. But as many other claims regarding the development of the construction, this has not been subject to a thorough examination. For a more detailed discussion of the problems the quantitative analyses of the development of phrasal verbs have had to face, see Chapter 6 below. 4.1.2. Colloquiality, informality, nativeness Kennedy is clearly of the opinion that the phrasal verbs are by and large ‘colloquial’: The development of these combinations is essentially a process of the common, relatively uneducated mind; the verbs and the combining particles … are mainly of native, Anglo-Saxon stock and are in daily use by all classes of English-speaking people. On the other hand, the words that are being displaced, such as collapse (break down), comprehend (catch on), eliminate (cut out), issue (give out) were introduced into the language, for the most part, through scholarly influence. Many of these words are directly chargeable to the


121

artificial increase of our vocabulary by that school of ‘embellishers’ who flourished in the sixteenth and seventeenth centuries. Perhaps the common man is not to be blamed for avoiding the use of a vocabulary which has never really been his, and for utilizing in the expression of certain ideas his own familiar word stock. It is, in other words, a movement from the ground up. (Kennedy 1920: 40)

Despite the reference to the borrowing of hard words in Early Modern English, the discussion very much creates the impression that this has always been the case in the history of English, or at least since the putative ‘rise’ of the construction in Middle English. Characterisations of the construction as stylistically ‘low’, ‘colloquial’, etc. are all-pervasive in his discussion, and even though Kennedy concedes that this need not always be the case, the essentially colloquial character of the construction remains a recurrent theme, e.g.: “colloquial combinations” (5), “many are frankly colloquial or even slang” (10), “[w]hen the fifteenth century is reached … it is evidently part of the language of the common man, as it has been ever since” (13), “plebeian verb-adverb combination” (13), “seems to be an important element in that ‘shop-talk’ or ‘English of the common man’” (17), “likely to be familiar to the speaker possessed of a limited vocabulary” (29), and so forth. On the whole Kennedy wavers between tolerance and condemnation: The history of the development, and the strength of it in modern colloquial usage, render it entirely impossible to brush aside the matter as current slang and colloquialisms. On the other hand one hesitates to encourage too freely the use of many of these combinations because of the responsibility of the trained and far-sighted student of English for the maintenance of certain standards in usage. (Kennedy 1920: 42)

It seems that Kennedy takes for granted that phrasal verbs must be stylistically marked, although his examination of the linguistic evidence does not really lend any unbiased support to such a notion. This is particularly striking with respect to the relationship of phrasal verbs and synonymous Latinate verbs, which until today remains one of the most popular misconceptions in regard to the lexical status of the phrasal verbs (cf. e.g. the characterisation by McArthur 1992 discussed in Chapter 2 above): Much of the difference, indeed, between the speech of the majority of those who pretend to be fairly educated and of the few who take a real pride in the precision and dignity of their English lies in the use, on the one hand, of the verb-adverb combination and, on the other, of the less common but more exact synonym. (Kennedy 1920: 34)

122


It is clear from the very phrasing that what is actually described here is the avoidance of phrasal verbs by pedantic speakers; hence the deliberate use of Latinate verbs should be regarded as the marked case, not the use of phrasal verbs. But in the context of a discussion which invokes “mere linguistic laziness” of the “average speaker” (Kennedy 1920: 34) as an explanatory factor for the use of phrasal verbs, this is easily overlooked. Kennedy continues immediately with a list of phrasal verbs which are said to be characteristic of the “indifferent user of English” as opposed to the simple verbs used by “the average man of fairly good education”. But such lists – which have remained a stereotypical feature of the discussion of phrasal verbs – are extremely problematic for a number of reasons already discussed in Bolinger (1971): Kennedy’s examples contain non-compositional uses like e.g. cough up (vs. pay) or dig in (vs. apply oneself) while the compositional ones typically do not even have a Latinate counterpart. Many of such pairs can only be regarded as ‘synonyms’ by a very wide application of the term (e.g. hang out vs. reside or chip in vs. contribute, where the contexts for paradigmatic substitution are extremely limited). Another problematic aspect which seems to be consistently ignored in the discussion is the fact that the existence of etymologically distinct synonyms is a characteristic property throughout the English vocabulary (cf. e.g. the overview in Hughes 2000) and not just of phrasal vs. simple Latinate verbs, and that therefore such suggestive pairings are in themselves quite useless. For example, another item on Kennedy’s list is blow in vs. spend; but the OED s.v. blow v.1 shows that the sense ‘spend, squander’ is first and foremost an American slang expression found with the simple verb blow, from which a synonymous phrasal verb is derived. It would be unjustified to posit on the basis of this or similar pairs that all simple non-Latinate verbs in English are ‘colloquial’ etc., and it seems likewise unjustified to claim that the synonymous phrasal verb provides evidence for the colloquiality of phrasal verbs as a whole. Most importantly, Kennedy (1920: 34) states that he has found synonymous simple verbs which could be used “to advantage” for only 110 out of the 826 phrasal verbs he examined. In other words, probably no more than one seventh of the phrasal verbs in Kennedy’s sample could be replaced by stylistically more elevated simple verbs, and presumably only in certain senses. The observations on the typically Germanic etymologies of the phrasal verbs are closely connected to their stylistic classification; this again is connected to the observation that their verb elements are typically monosyllabic:


123

The reasons for the predominance of native, monosyllabic verbs in combination are clear, I am sure. In the first place, they are not only less cumbersome in the combining, but for the sake of a certain speech rhythm seem to call for weaker syllables to accompany them. Then again the native Teutonic verbs (and I include those of Scandinavian origin) are used more commonly, as a rule, and are more likely to be familiar to the speaker possessed of a limited vocabulary. (Kennedy 1920: 29)

As for the first part of this explanation, it is hard to tell what is meant by the characterisation of native (viz. Germanic) monosyllabic verbs as “less cumbersome in the combining”, while the speech rhythm can only be invoked in those cases where there seems to be no apparent difference in meaning between the simple verb and the simple verb plus particle (e.g. meet vs. meet up). What seems more relevant in the present context, though, is the reference to the “speaker possessed of a limited vocabulary”, which of course is again a mere invocation of the ‘simplicity’ and ‘colloquiality’ of phrasal verbs rather than an explanation; it is typically the noncompositional phrasal verbs which have Latinate synonyms, and these phrasal verbs are by definition lexicalized (e.g. knowing give and up does not provide access to the meaning of give up, and this phrasal verb is therefore just as lexicalized as the synonymous surrender). Thus the statement is once again one that can only make sense if it is taken to mean that the speaker with a limited vocabulary knows fewer Latinate words – which is first and foremost a statement about the status of the Latinate parts of the English vocabulary (cf. the collection of Early Modern English opinions on the issue in Jones 1953). But even this statement is a circular one, which detects limitations in a speaker’s vocabulary only where these limitations affect the Latinate parts of the English lexicon, which are traditionally the ‘hard’ ones, while the ‘native’ elements of the lexicon are ‘plain’ or ‘simple’. Hence the paradoxical situation that the phrasal verbs are at the same time ‘simple’ (when used by native speakers) and ‘difficult’ (when acquired by learners of English as a foreign language), which seems inextricably connected to the long-established notion of the particular ‘Englishness’ of the phrasal verbs, as the discussion in Chapter 6 below will show. 4.1.3. The impact of Kennedy’s study Despite the critical observations made so far, Kennedy’s study is impressive as the pioneering work which marks the beginnings of a linguistic research tradition in the history of the phrasal verbs. Kennedy manages to

124


provide a by and large convincing first overview, especially in view of the lack of more specialized historical studies. What is problematic, though, is the uncritical acceptance of many of Kennedy’s assumptions in the later literature on phrasal verbs, and this is also the reason why Kennedy’s study has been discussed here in more detail than would otherwise have been justified for a study published more than 90 years ago. Often the continued impact of Kennedy’s study will only be apparent at closer scrutiny; although some recent studies still rely explicitly on his results, others will do so only indirectly by referring the reader to more recent studies which in their turn, directly or indirectly, rely on Kennedy. 4.2.

Some textbooks and language histories

A short representative review of the treatment of phrasal verbs in the general literature shows that Kennedy’s assumptions have continued to prove a recurrent theme in a rather confused discussion. 4.2.1. The pitfalls of history Let us begin with a look at how a well-known older textbook presents the history of phrasal verbs in Old and Middle English. Quoting extensively from Mustanoja (1960: 348–349), Strang (1970: § 153), in the context of her discussion of the development of the Middle English prepositions, stresses the importance of influence from Old Norse and Old French and of “experimental exuberance” (Strang 1970: 275) for their development (the reference to Mustanoja resurfaces in Fischer’s 1992 discussion of syntactic developments in Middle English, cf. Section 4.2.2 below). The reasons for the development of the prepositions are unknown, Strang says, but it somehow must have had something to do with the phrasal verbs: It is not easy to understand the complex forces which led to these shifts of balance, and perhaps we shall never have the knowledge to penetrate them completely. But one factor clearly is of the highest importance, though itself of complex and not wholly understood provenance. This is the development of the verb-particle combination (phrasal verb), in which the particle may be a preposition or an adverb. Such combinations were virtually unknown in OE [viz. Old English], which used particles with verbs in separable prefix form … The separable use gave rise to many patterns in which the particle followed the verb, and for some reason this arrangement came increasingly to be preferred. (Strang 1970: 275)


125

But Strang does not offer any evidence for the close connection between changes in the inventory of prepositions and the rise of the phrasal verbs. The assumptions that the phrasal verbs somehow took the place of the Old English separable prefix verbs (i.e. that they are a new development) while at same time the phrasal verbs are, basically, separable prefix verbs with postverbal particles (i.e. they are a positionally restricted version of them) is another problematic aspect – only one of these two assumptions can be true. Strang (1970: 275) argues that the occurrence of give up ‘surrender’ as early as 1154 suggests that phrasal verb must have been “deeply entrenched” before that time; note that Burchfield (Fowler 1996) adduces the same example from the Peterborough Chronicle to suggest the opposite (cf. the quote in Section 4.1.1 above). Following Kennedy (1920), Strang goes on to propose that the development may at first have been restrained by the influx of new loan compounds from French and by stylistic reasons, since the verb-particle combinations “seem always to have had the air of colloquiality that still often clings about them” (Strang 1970: 276). It is not entirely clear from this statement whether it is also meant to apply to the Old English particle verbs. In this case it would be quite unfounded, if only because of the lack of lexical alternatives. Even in regard of the Middle English situation the assumption of colloquiality is based on very scanty evidence such as Kennedy’s somewhat anecdotal statements and can hardly be used as an explanation for the slow growth of lexicalized combinations, especially since colloquiality seems a somewhat problematic concept with respect to Old and early Middle English (cf. Kastovsky 1992: §§ 5.1.2 and 5.3.2.1 and Tristram 2004). Strang’s discussion provides a typical example of the treatment of phrasal verbs in the historical handbooks and in the textbooks: there appears to be a common conviction that the phrasal verbs are somehow particularly ‘English’ (which is bolstered by the observation that they typically contain verbs of Germanic descent and by the presumptive difficulties they pose for learners of English as a foreign language) and that their development may somehow have been influenced by language contact. Although those convictions have hardly ever been put to the test, they persist in the literature, as e.g. the following characterisations in a more recent introduction to Middle English show: An interesting case, demonstrating the impact of Norse and also (indirectly) French, is to do with the development of the phrasal verb. Phrasal verbs are a characteristic English formation that developed during the ME [viz. Middle English] period; they consist of a verb-particle formation of the model GIVE UP, SIT DOWN and so on. These verbs seem to derive from OE [viz. Old English] verbs such as bistandan STAND BY, but their increase in use during

126


the ME period probably derives from interaction with Norse. Strang (1970: 276) notes that there is a stylistic restriction on the use of the phrasal verbs even now: ‘The verb-particle combinations seem always to have had the air of colloquiality that still often clings to them’. (Horobin & Smith 2002: 75) The main syntactic innovation in the verb phrase during the ME period was the rise of two kinds of construction: the impersonal verb and the phrasal verb … The latter construction, still common in PDE [viz. present-day English], consists of a verb followed by another element which seems closely tied to it semantically, for example GET UP, WAKE UP, LOOK UP. Typically, as mentioned on p. 75 above, phrasal verbs in PDE are rather colloquial in register; typically also, they tend to have formal-register near-synonyms, cf. ARISE, AWAKE, CONSULT. (Horobin & Smith 2002: 99)

These two quotes from Horobin & Smith (2002) contain a representative selection of the most common preconceptions regarding the phrasal verbs: their Englishness, the putative influence of one or more contact languages, their stylistic restriction to colloquial registers and their rise as an innovation in Middle English. Taken together, the two quotes also provide a typical example of the widespread anachronistic and circular reasoning with regard to the issue of colloquiality: although the second quote refers to colloquiality in present-day English, in the first quote it is not only claimed that this has always been the case, but it is also implied (with reference to Strang 1970, who also merely claims this) that the colloquiality of the phrasal verbs was more marked in Middle English. Similar remarks can be found in Horobin (2007: 119), where the presumed colloquiality in Middle English is illustrated by two putatively colloquial examples from Chaucer. But in both passages the phrasal verbs are surrounded by French borrowings, which Horobin surely would not wish to label as ‘colloquial’; again, the procedure is circular. And as the comparative discussion in the preceding chapter has demonstrated, there is little reason prima facie to characterize the phrasal verb either as a particularly English construction or as a Middle English innovation. This also goes both for the assumption of language contact as a relevant factor in their development (see Chapter 5 below) and for the anachronistic assumption of colloquiality (which will be discussed in more detail in Chapter 6). 4.2.2. Coverage in the Cambridge History of the English Language (CHEL I and II) In CHEL I, Kastovsky (1992) in the chapter on the lexicon discusses lexical aspects of the development of the phrasal verbs on the basis of earlier


127

studies. Kastovsky (1992: § 5.1.3.1) presents the Old English particle verbs as part of the ‘associative’ vocabulary of Old English, in contrast to the phrasal verbs as typical of the ‘dissociated’ vocabulary of Modern English (see Leisi 1985: Section 12, but cf. Sanchez 2008, where it is shown that Leisi’s claims are to be taken with a considerable degree of caution). But one must doubt whether this distinction is applicable to the phrasal verbs, where new lexical units are formed on the basis of integrated material. In this respect phrasal verbs are not essentially different from the verbal prefixations Kastovsky (1992: 295–296, examples 5a–r) quotes. Kastovsky (1992: 298) points to the “complete collapse of the OE morphophonemic system because of its rapidly growing opacity … and the ensuing phonological, morphophonemic and morphological restructuring at the end of the OE and the beginning of the ME period, whose details still await a systematic investigation” and assumes that the growing semantic opacity of many Old English elements of word formation must be seen as the ultimate cause of the loss of the Old English prefixes. I.e., in this account the phrasal verbs came to replace the Old English prefix verbs because the prefixes had lost their semantic transparency: As Horgan (1980) and Hiltunen (1983[a]) have shown, the system of prefixes, in particular those occurring with verbs, was already at the end of the tenth century in a state of advanced decay, because many prefix-verb combinations were no longer transparent. With many verbal prefixes, e.g. a-, ge-, oþ-, it is impossible to establish consistent meanings, and frequently there does not seem to be any meaning difference at all between the simplex and the prefixed form. This points to a considerable weakening of the meaning of these prefixes, especially of a-, be-, ge-, and the prepositions/adverbs for and of. It is not surprising, therefore, that the OE prefix and preparticle system was an easy victim for both the Romance invasion of the lexicon and the rise of the postparticle (phrasal) verbs in ME. (Kastovsky 1992: 377)

This assumption rests at least in part on problematic conclusions drawn from earlier studies, e.g. by Horgan (1980).2 In this context Kastovsky also discusses the status of verbal loan formations (§ 5.2.1.5.3). Loan formations may result in knotty problems to a morphemic analysis of prefix 2

Horgan had shown that among different copies of the Old English Cura Pastoralis (Bodleian MS Hatton 20, Junius’ transcription of British Museum MS Cotton Tiberius B xi, Cambridge, Trinity College MS R 5.22 (717) and Cambridge University Library MS Ii 2.4) some prefixes are interchangeable with others and concluded that “the variation on the part of the copyist was deliberate, and presumably stylistic, since no actual change in meaning seems to be involved” (Horgan 1980: 130).

128


verbs where they are based on lexicalized Latin formations like instruere ‘instruct’ or praeferre ‘prefer’, which are rendered in Old English as ontimbran and foreberan respectively. However, Kastovsky claims, in such cases the prefixes are redundant or without specific meaning and play the role of mechanically imitating the Latin model (for a critical discussion of such a view of the role of calquing in Old English, see Johnston 2011 and the references given there). Kastovsky (1992: 314) also raises the question whether this process was facilitated by the semantic bleaching of the Old English prefixes at the time the loan translations were formed, or whether the resulting word formations themselves destabilized the prefix system, “by adding more and more instances of morphosemantically opaque formations to the numerous already existing ones” and he speculates that the two factors may well have operated together in the “radical loss of OE verbal prefixes in the subsequent period”. With regard to Scandinavian influence, Kastovsky remains somewhat vague and undecided, in keeping with the usually vague and contradictory statements in the literature: the phrasal verb type come on, make up – including muck up, muck about, where the verbs themselves are of Scandinavian origin (Poussa 1982: 73) – seems to be due to Scandinavian influence (Logeman 1906), or was at least strengthened by a parallel Scandinavian pattern (Hiltunen 1983[a]: 42–4), while its ultimate origin has been attributed to a Celtic substratum (de la Cruz 1972[d]: 171ff.). (Kastovsky 1992: 320)

Cf. Chapter 5 below for further discussion of the issues raised here. On the whole, verb-particle constructions clearly belong to the lexicon in CHEL I. In the chapter on syntax (Traugott 1992) the subject is not treated at all; in the light of the prominent place of the topic in the discussion of the development of English word order it is remarkable that particle verbs are excluded from the explicit syntactic discussion in that volume, although the topic is alluded to in some places, e.g. in the discussion of preposition stranding (230–232, cf. her examples 156 and 157). In contrast, the syntax chapter in the more recent companion textbook to CHEL (Fischer & van der Wurff 2006) discusses the development in some detail (cf. the references in Chapter 3 above). The chapter on syntax in CHEL II (Fischer 1992) includes the development of phrasal verbs (§ 4.9). The phrasal verbs, Fischer states, “almost completely replace the Old English prefixed verbs” (386). Mostly relying on Kennedy (1920), Mustanoja (1960), Strang (1970), Hiltunen (1983a and 1983c) and the contribution by Burnley (1992) in the same volume, she presents the development as a conflux of lexical and syntactic factors, and she states that the inseparable prefixes were replaced as a type by the


129

phrasal verbs. If I interpret it correctly, Fischer implies that the phrasal verbs replaced the inseparable prefix verbs after they had disappeared: “Since most Old English prefixes disappeared (cf. ch. 5 [viz. Burnley 1992], Section 5.2.11), this type was presumably simply replaced by a new verb-particle combination” (Fischer 1992: 386). According to her, the phrasal verbs may have been partly due to influence from Old French and Old Norse models, but they may also be seen as the native continuation of the Old English ‘separable prefix verbs’. Fischer does not elaborate on the details of the presumed combination of native and foreign factors and the assumption of language contact as playing a role in the development remains speculative. She attributes this assumption erroneously to Mustanoja (1960: 362–363), where the topic is not discussed. I assume the intended reference is to Mustanoja (1960: 348–349), which is also quoted in this connection by others (e.g. by Strang 1970; cf. the discussion above). But that passage from Mustanoja (1960) is not at all about foreign influence on the development of the phrasal verbs but about the influx of foreign prepositions and their influence on the native prepositional system. There is only one reference to prefixes (Mustanoja 1960: 352). But again, Mustanoja does not make any statement regarding the development of the phrasal verbs and his remarks on the mutual interchangeability of prepositions (Mustanoja: 350–352) are superseded by the more recent literature (cf. Dietz 2005 and the references given there). Fischer also does not comment on further details in her sketch of the syntactic development, which however seems to rest on the assumption that the elements subject to the positional changes are not semantically opaque, as is suggested by Kastovsky’s account in the preceding volume. In the chapter on the Middle English lexicon in CHEL II, Burnley (1992), like Kastovsky in CHEL I, mentions the rise of the phrasal verbs in connection to language contact, namely among what he regards as the effects of incomplete bilingualism, which are “probably due to the influence of Scandinavian” (Burnley 1992: 422). He also claims that “Scandinavian influence contributed to the development of particled verbs in Middle English” (Burnley 1992: 444), again without offering any evidence. Kastovsky’s phrasing, on the other hand, is remarkably guarded (cf. the quote from Kastovsky 1992: 320 above). But despite using similar hedges, Burnley’s presentation leaves little doubt that he assumes a connection with Scandinavian influence. Note the suggestive argumentation: “The earliest of the extensive use of verb+preposition/adverb colligations as phrasal verbs on the model of Old Norse is in the Peterborough Chronicle: gyven up (probably with Scandinavian initial /g/), faren mid, leten up and tacen to. The Ormulum contains numerous examples” (Burnley 1992: 422–

130


423). But none of this is real evidence of Scandinavian influence, and it remains quite unclear whether the Scandinavian influence caused the development or whether it only contributed to it. Compounding, Burnley states, is less productive in Middle English than it used to be in Old English (1992: 441). Burnley (1992: § 5.2.8) treats all separable verb-particle combinations in Old English as compounds and gives no criteria for distinguishing between separable and inseparable formations, and after a general characterisation of ‘verbal compounds’, he states that the types of Old English were redistributed into the two types ‘inseparable particle+verb compounds’ and phrasal verbs and that nominal compounds were derived of these two types (Burnley 1992: § 5.2.9), without going any further into the details of the development.3 He concludes: This emphasis upon the particled verb as the focus of derivation is symptomatic of the change which took place during the fifteenth century by which the formation of verbs became concentrated on the production of particled verbs, and compound verbs ceased to be productive as a type of word formation. (Burnley 1992: 445)

Unlike Kastovsky (1992), Burnley does not present a list of the relevant prefixes or particles, and most of the details of change mentioned by him remain somewhat vague. No connection is established between the history of the phrasal verbs and the fate of the Old English (inseparable) prefixes. In Middle English, Burnley states, “prefixation as a means of word formation was in retreat” (1992: 446): on the one hand, many Old English prefixes became unproductive and, on the other, new prefixes through borrowing from Latin and French did not become productive until the end of the Middle English period (Burnley 1992: §§ 5.2.11–13). As a result, “Middle English until the fifteenth century was somewhat depleted in its range of productive prefixes” (Burnley 1992: 447). Thus, with respect to the causes of the development Burnley’s account seems at odds with the

3

In the older literature the term compound verb may be applied to all kinds of particle verbs, including phrasal verbs (cf. e.g. the use in Curme 1914), but independent of this terminological issue the use in the dictionary de facto excludes the postpositional variants. Altogether it seems that the use of the term compound in connection with the English particle verbs tends to coincide with considerable confusion as regards their structural properties; cf. the characteristic statement by Nielsen (2005: 116), who claims that the “OE compound verbs with an adverb or preposition preceding the verbal base were partly continued and partly (co-existing with and) replaced by phrasal (particled) verbs in Middle English”.


131

preceding chapter in CHEL II on syntax, where Fischer points out the continuity with Old English and the ultimately syntactic reasons behind the emergence of verbs with normally postposed particles. In sum, there seems to be little disagreement as to the factors which may have played a role in the history of the phrasal verbs, although their respective evaluation may differ considerably. But taken together, consulting CHEL I and II leaves one with a somewhat hazy view of the development of the phrasal verbs: What are the mechanisms of change? And what exactly changes? But before we try to untangle the details of the development in the next chapter, it seems appropriate to offer a brief overview of the lexicographic coverage of particle verbs in the relevant historical dictionaries, which are quite likely to have contributed to the considerable confusion found elsewhere in the literature. Conversely, the unsatisfactory lexicographic treatment may also be assumed to be the result of the incomplete accounts in the earlier literature. 4.3.

Lexicographic coverage: a characteristic example

The present section, then, discusses the treatment of particle verbs in Old and Middle English in the standard historical dictionaries of English, viz. most notably for Old English Bosworth–Toller (published 1882–1898; Supplement from 1921), for Middle English the MED (1952–2001), and across the individual periods also the OED (substantially still 1884–1928). Until comparatively recently, with the publication of the first fascicles of the DOE, these were the main lexicographic aids available to scholars working on the subject (for a general overview of the English historical dictionaries, see Thim 2011b and the references given there). In none of these historical dictionaries the existence of particle verbs is satisfactorily acknowledged and there are striking parallels between the deficiencies of the lexical accounts discussed above and of those of the lexicographic coverage to be discussed in the present section. The present section does not aim at providing a full critical analysis of the coverage of particle verbs in the historical dictionaries but rather a very selective discussion of a typical example, with the aim to demonstrate the almost predictable limitations imposed by the dictionaries on the older research. The Old English verb forþfēran already introduced in Chapter 1 above may serve as an example of the lexicographic practice typical of the older historical dictionaries (the selection of a particle verb with forth is

132


due in part to the inclusion of the DOE in the discussion, which so far does not provide coverage beyond the letters A–G).4 4.3.1. Bosworth–Toller and other older dictionaries The treatment in Bosworth–Toller and its Supplement is quite characteristic. The coverage of the verb is spread over two entries (s.vv. forþferan and forþ-geferan):5 s.v. forþferan the senses ‘go forth, depart, die; decedere, defungi, mori, expirare’ are given. In the eight sample quotations, the particle precedes the verb immediately and the consistent spelling as one word adds to the impression of an inseparable prefix verb. Thus Bosworth– Toller s.v. feran contains only a cross reference to forþ-[feran] and is rather consistent in its de facto treatment of the combination as an inseparable prefix verb. Consequently there is also no cross reference between this entry and the entry for forþ-geferan (i.e. forþ-gefēran), which in fact would provide good evidence for the analysis of the construction as a particle verb: forþ-geferan … go forth, depart, die; decedere, mori: –Ðara monige forþgeferdon on Drihten many of whom died in the Lord.

If forþfēran was an inseparable prefix verb, one would not expect the prefix GE- to intervene between the prefix and the verb. It would, of course, be conceivable in principle that forþgefēran is a distinct prefixation (of a verb gefēran), but this would be a rather unsatisfactory explanation in the light of the separability of forþ in both cases, the synonymy of both 4

5

Dietz (2004: 596–600) points out several of the limitations of older research on Old English prefixes, which may be due to the authors’ reliance on the dictionaries, but also to their use of small or unrepresentative corpora. With respect to the status of Old English forþ Dietz (2004: 608–609) points out that it occurs predominantly in glosses of Latin prefixed verbs and he suggests that it is therefore problematic to label forþ as a native Old English prefix and that it had better be classified as a particle. Although this observation on the predominant occurrence in calques is certainly true (cf. DOE s.v. forþ-), it has no direct bearing on the present discussion where the focus is on the lexicographic coverage of such formations as verb-particle constructions rather than on their limited productivity because of their status as calques. The stem vowel of the verb is erroneously presented as short in Bosworth– Toller but emended to a long vowel (“é”) in the Supplement. In the examples from Bosworth–Toller the dictionary’s use of acute accents to indicate vowel length will be followed.


133

fēran/gefēran and forþfēran/forþgefēran, the general characteristics of Old English ge- (cf. e.g. the discussion in Kastovsky 1992: 377–380 and the comparative evidence discussed in Chapters 2 and 3 above; see also the discussions of related lexicographic problems by Butler & Mitchell and by Sauer in Bammesberger 1985). The Supplement to Bosworth–Toller, however, acknowledges Old English evidence for the verb with postposed particle by referring to féran forþ ‘die’ s.vv. féran and forþ. The supplementary entry for forþ, which has almost tripled in size compared to Bosworth–Toller, contains the following sub-entry (6a): of death and decay: – Gif ceorl and his wíf bearn hæbben gemǽne and fére se ceorl forð (and the husband die).

Moreover, the Supplement also adds s.v. féran (I.2b): of going from this world, to depart this life: … Férdon forð Tatwíne and Biéda … Gif ceorl and his wíf bearn hæbben gemǽne and fére se ceorl forð.

Crucially, there is no cross reference to the entry for forþféran and no substantial change s.v. forþféran. Also note in the two preceding examples that part of the entry for fēran reduplicates the supplementary entry for forþ, again without cross reference. Thus, while the meaning ‘die’ of the whole combination is haphazardly connected to either the simple verb or the particle in postposition, the postposed variant is in no way marked as related to the preposed one. Since this example shows the common practice in Bosworth–Toller and its Supplement, it is clear that the existence of the particle verb in Old English is far from being obvious in the dictionary, let alone systematically recorded, and it is easy to see how this procedure may have resulted in the impression first voiced by Kennedy (1920: 12) that in Old English “occurrences of the verb-adverb combination are practically nil”. A satisfactory solution would be a single entry for forþ(ge)fēran which points out the possibility of pre- and postposition of the particle. The example is typical. In many instances the Modern English glosses of the Old English headwords in Bosworth–Toller are highly suggestive of remarkable diachronic continuities. With respect to verbs listed as beginning with forþ- alone one finds no less than 65 verbs with Modern English glosses involving forth, including the considerable number of cases where the selection or the form of the headword in the dictionary seems problematic and also including the glosses with particles other than forth if they follow a gloss involving that particle (but not including the one-word glosses which often follow the ones involving particles and which are typically connected to non-literal and lexicalized senses).

134


As with forþfēran discussed here in greater detail, many of these apparent prefix verbs correspond throughout quite systematically to verbparticle sequences. How lexicographers of Old English who must have been familiar with the conventions of German dictionaries could present such forms as distinct from identical particle verbs with the particle in postverbal position is hard to explain, and the overall impression is very much comparable to a hypothetical dictionary of Modern German covering a particle verb fortfahren s.vv. fortfahren and fortgefahren for the preverbal position of the particle and s.vv. fahren and fort for the postverbal position. But nevertheless this procedure is expressly defended by Campbell (1972: v) in the preface to the Enlarged Addenda and Corrigenda to the Supplement, with no apparent reason except for a reference to “the tradition of the Dictionary”. These findings seem to confirm Kornexl’s (1994: 447, fn. 68) suspicion that “one might surmise that the relative scarcity of compound verbs in dictionaries of Old English to some extent at least results from Modern English structures unconsciously shaping the criteria applied to this previous stage of the language by present native speakers”. Remarkably, Grein’s earlier dictionary of the vocabulary of the AngloSaxon poets (1861–1864) offers a more satisfactory solution since it lists forð faran s.v. forð (5) and gives as one example (among others): “fêrdon ~ þonon”; i.e. Grein treats the particles as adverbs rather than prefixes and consequently his presentation avoids the inconsistencies characteristic of the later period dictionaries (although his treatment of other particles is sometimes a bit confusing, though; cf. e.g. the classification s.v. up). It comes as no surprise that Grein’s dictionary is not listed in Kennedy’s (1920) bibliography (just as Mätzner’s, 1878–1896, cf. below). Conversely, Borden’s much later Old English dictionary (1982, compiled on the basis of several earlier dictionaries and glossaries) is very unsatisfactory with regard to particle verbs since in this dictionary they are completely indistinguishable from inseparable prefix verbs, even after perusal of the entries for the particle verb, the simple verb, and the particle (cf. e.g. again s.vv. forðfēran, fēran, forð). The same criticism applies to the widely-used Concise Anglo-Saxon Dictionary by Clark Hall (first published in 1894, most recent edition 41960), which, moreover, lists the particle verb s.v. ±forðfēran, implying (“±”) that it is a prefix verb which may be preceded by ge-.


135

4.3.2. The Middle English Dictionary For Middle English, coverage of particle verbs in the Middle English Dictionary (MED) is in principle very similar to the older dictionaries of Old English, and hence the same criticism applies in those cases where the particles may still both precede and follow the verb (i.e. typically, but by no means exclusively, in early Middle English). Again there is no cross reference between the entries for forth-fāren and fāren, although it seems quite clear from the two entries taken together that in early Middle English forth-fāren with pre- and postposed particle is a single verb. In Middle English, the Old English verbs forþfaran and forþfēran seem to have almost completely merged into a mixed paradigm and are therefore both treated in the MED in one entry, just as the Middle English continuations of the simple verbs. It is therefore admissible to use the MED entries for (forth-)fāren to continue the discussion of the lexicographic treatment of Old English (forþ-)fēran (cf. the spelling variants and the morphological explanations given in the MED s.vv. fāren and forth-fāren). Thus s.v. forthfāren the senses ‘pass away, die; travel, go out, fare forth, issue out, advance (in age), pass (of a period of time)’ are given, i.e. the Old English literal and figurative senses of the verb are continued, e.g.:6 (1)

On þes ilces geares forð ferde se eadig biscop Ernulf of Roueceastre (a1126 Peterb. Chron. an.1124)

(2)

Go we þane narewe pað … [þ]ar forð fareð […] wel litel folc (a1225 (?c1175) PMor (Trin-C) 344)

(3)

I forthferde To walke, as I yow telle may ((a1393) Gower CA 1.98)

Characteristically, in this entry none of the sample quotations contains the particle in postposed position and the separability of the particle is marked only implicitly, although it might be more obvious to the casual user than in the older dictionaries of Old English, due to the spelling of some of the examples in two separate orthographical words, but also due to the morphological information provided at the beginning of the entry, where the past participle variants -fāre(n, -ivāre(n and -iværed are given, with the GE- (i-) prefix as a strong indication of separability of particle and verb. But this sign is likely to be interpreted correctly only by users who are aware of 6

It has not been possible to reproduce all of the diacritics used in the MED; in particular combinations of two diacritics, e.g. macron plus dot below, had to be omitted here. The abbreviations of the sources have been retained as used in the MED.

136


the comparative Germanic evidence anyway and who consult the dictionary also s.v. fāren and s.v. forth-; cf. e.g. the more confusing treatment with a separate and unrelated entry for the GE- variant in Bosworth–Toller discussed above (for a discussion of the position of GE-prefixes between preposed particle and verb in other Germanic languages, see Chapter 5 below). S.v. fāren, users will find examples of the verb with postposed particles (or, more precisely, only with the postposed particle), e.g.: (4)

Ic ne mihte na faren forð on þin ærende (a1175 (?OE) Bod. Hom. 18/1) [s.v. fāren (6a)]

(5)

Esau ferde forð ðeden to Seyr (a1325 (c1250) Gen. & Ex. 1836) [s.v. fāren (2a)]

(6)

Fare forthe … and fech as þou seggez (c1400 (?c1380) Cleanness 621) [s.v. fāren (6c)]

S.v. forth-, the particle is characterised as the first element of ‘compound verbs’, but, again, without reference to the possibility of postposition, and consequently no cross reference to the synonymous postposed uses, e.g. s.v. forth adv.: forth-, stressed prior member of cpd. verbs and action nouns [OE forþ-]. Examples: forth-bringen bring forth, bring up [cp. L prōdūcere]; -comen come forth; -cume departure; -gōn go forth, advance; -gang progress; -lēden lead out, carry out; -passen proceed; -sīth departure.

Once again the example is typical of the practice in the dictionary, cf. e.g. MED s.vv. forth-bringen (vs. bringen forth s.vv. bringen and forth, cf. Old English forþ-bringan), forth-callen (“[m]odeled on L prō-vōcare”), forthcasten, forth-clēpen, forth-comen, forth-dōn, forth-drauen, forth-fillen, forth-gangen, forth-gōn, forth-lēden, forth-nimen, forth-passen, forthsheuen, forth-wīsen, forth-wīten. A much earlier dictionary of Middle English (Mätzner 1885) treats the particle in a considerably more satisfactory way: Seit frühester Zeit finden wir zahlreiche lockere oder uneigentliche Zusammensetzungen mit der Partikel forð, von denen die mit Zeitwörtern häufig auch eine Trennung mit Umstellung der Partikel zulassen, während die Partikel anderweitig, namentlich an Substantiven, ihre Stellung behauptet. (Mätzner 1885: s.v. forð)


137

[‘Since the earliest time there have been numerous loose compoundings with the particle forð. Among these the ones with verbs often allow the particle to be separated and moved, while otherwise the particle, in particular next to nouns, retains its position.’]

This is the best explanation to be found in a dictionary to date (but cf. the strikingly similar though somewhat hazier statement in the OED quoted below). What renders treatment in the MED perhaps even more confusing than the Old English dictionaries discussed so far is the fact that sometimes, without apparent reason, there is a cross reference to the simple verb (e.g. s.vv. forth-nimen or forth-passen), sometimes odd bits of etymological information (e.g. s.v. forth-wenden: “[f]rom wenden go”) and twice (with respect to particle verbs involving forth) even an explicit cross reference to the postposed variant: s.vv. forth-bēren (“[a]lso bēren forth”) and forth-tēn (“[a]lso tēn forth”). As a result of these practices, it is very difficult to draw conclusions about the morphological, syntactic and semantic status and development of one or more particle verbs on the basis of the MED. It is also usually very hard to tell whether at some point the preposed particle became an inseparable prefix and also what the semantic relationship of such a prefix verb and a possibly still existing particle verb (presumably at that time one with postposed particle only, i.e. a phrasal verb) might have been; cf. the observation made by Burnley (1992) reported in Section 4.2.2 above that in the Middle English period there is a redistribution of the Old English particle and prefix verbs and the more detailed discussion in Chapter 5. Indeed, in the case of forth-fāren it seems on the basis of the dictionary evidence that in Middle English the metaphorical meaning ‘die’ becomes a property of the verb with preposed particle, while the spatial meaning continues to be found with the particle in pre- and in postposition (as witnessed by the absence of examples with postposed particles in the sense ‘die’ in the dictionary, cf. the representative examples quoted above and see also OED s.v. fare v.1 12, where there is no mention of fare forth ‘die’ after early Middle English), a finding which is strongly reminiscent of the typical distribution in some present-day Scandinavian languages discussed in Chapter 2 above. But this impression would certainly need a considerably more thorough analysis of the Middle English sources than can be provided here. The crucial point to be made in the present context is that the structure of the MED renders this kind of information extremely difficult to unearth, and almost completely invisible to anyone who is not actively looking for it. Another case in point is forth-fillen, where neither the MED nor the OED offer any evidence of a postposed particle. Moreover, all

138


quotations s.vv. in both dictionaries are from the late 14th century or later, and it seems that forth- in this formation had better be analysed as an (inseparable) prefix. But on the basis of the explicit information given in the MED, it is indistinguishable from the many particle verbs with forth. 4.3.3. The Oxford English Dictionary The Oxford English Dictionary (OED) offers a comparatively satisfactory treatment of particle verbs. It is obvious that the likely reasons for the positional characteristics of the construction were not yet known to the editors of the OED (although at the time of publication of the entries for the letter F the reasons for the positional properties could have been known to them due to the publication of Harrison’s study in 1892). Nevertheless the relationship between the preposed and the postposed variant tends to be pointed out in its entries, not only implicitly (via cross references), but also explicitly.7 Thus, to return to the example of forþfēran etc., the relevant entry in the OED (NED vol. IV.1, published in 1901) reads as follows: 1

† Forthfa·re, v. Obs. [OE. forðfaran, f. FORTH adv. + faran to go: see FARE v. ] 1. intr. To go forth, go away, depart, journey. c888 K. ÆLFRED Boeth. xxxiii. § 4 Swa ðætte hi æþer e forþ faraþ e eftcumaþ. a1200 Moral Ode 340 Go we .. þene wei grene þer forþ-fareþ lutel folc. a1300 E. E. Psalter x[i]. I, I sal forth fare, ife I wil, als a sparwe into þe hil. 13.. K. Alis. 6936 Sorwe and care That day thei letten forth fare. 2. To decease, die. O. E. Chron. an. 571, On þam ilcan eare he forþfor. a1175 Cott. Hom. 225. Noe lefede .. nion hund eare and fifti, and he þa forðferde. c1205 LAY. 11458 Þenne þu beost forð faren. c1320 Cast. Love 218 Atte laste he moste dyen and forþ-fare. c1350 Will. Palerne 5266 Þemperour was forþ-fare faire to crist.

And s.v. fare there is the following sub-entry: 1

12. Fare forth (analytical form of OE. forðfaran). See FARE v. and FORTH. a. To go forth, depart, start. c1200 Trin. Coll. Hom. 225 To heueriche hie sulle fare forð mid ure drihte. 1375 BARBOUR Bruce III. 345 All hyr cumpany, Lap on thar horss, and furth thai far. c1400 Melayne 206 Rowlande .. Fares forthe with Baners brade. 1647 H. 7

Apart from the comparatively good coverage of the particle verbs in medieval English, the OED is also quite strong on Modern English phrasal verbs, which are recorded with remarkable reliability (cf. e.g. its coverage of Early Modern English phrasal verbs established in Thim 2006a).


139

MORE Song of Soul I. I. xxvi, Like Doves so forth they fore. 1727–38 GAY Fables I.I. xiv. 5 Forth he fares, all toil defies. 1853 KINGSLEY Hyapatia xxi. 258 Before sunrise .. Raphael was faring forth gallantly. † b. To go on, advance, with respect to either space or time. In the latter sense also quasi-impers. 1340–70 Alex. & Dind. 939 Whan he is fare so forþ fer in his age. c1350 Will. Palerne 3260 It was forþ [to] nit faren bi þat time. † c. To go by, pass by. Obs. a1225 Leg. Kath. 1629 Pinen, þe fare forþ in an hondhwile.

But although the OED relates the preposed and the postposed variants to each other, it is also clear that this is again not done in an entirely systematic fashion, especially since, as the sample entries show, this is not done in a consistent and reciprocal way. Thus the cross references s.v. forthfare are quite satisfactory, especially when one bears in mind that the OED is not a dictionary of Old English, but a historical dictionary of the post-Anglo-Saxon language (cf. Stanley’s 1987 discussion of the treatment of Old English in the OED). These cross references relate the verb both to Old English forþfaran and to the entries for fare and forth; s.v. fare (12. fare forth), and the user is even given an explicit remark on the nature of the Modern English phrasal verb as the “analytical form” of the Old English verb and again a cross reference to forth, where in its turn (s.v. forth D. Forth- in composition) medieval English particle verbs are characterised as follows: “In OE. and ME. the combinations of forth adv. with vbs. are hardly to be considered compound verbs; whether the adv. precedes or follows the vb. depends on euphonic or other conditions which do not affect the sense”. On the basis of the combined evidence presented in the OED it is indeed possible to establish a more coherent historical picture than with any of the period dictionaries. But then again the information is altogether provided rather obliquely. Even leaving aside the somewhat dubious explanation of the status and the positional properties of forth provided s.v., it is not easy to deduce from these entries that both the obsolete prefix verb forthfare and the phrasal verb fare forth are ultimately derived from one Old English particle verb, especially since the references to Old English forþfaran in the entries for both verbs do not make it explicit at all that it is a particle verb. Rather, they create the impression of a prefix verb, an impression which seems to be played up even further by the characterisation of the phrasal verb as the ‘analytical form’ of that Old English verb (since that characterisation can only make sense if the variant with the particle in preposition is regarded as synthetic: i.e., an analysis which could be sound for Modern English is inappropriately applied to Old English).

140


The claim in The Oxford Companion to the English Language, “[c]itations [of phrasal verbs] date from Middle English” (McArthur 1992: s.v. phrasal verb), is a typical consequence and it shows clearly the connection between the characteristic lexicographic weaknesses and the wide-spread and misleading textbook accounts of the development of the phrasal verbs.

4.3.4. The Dictionary of Old English Initially the treatment of particle verbs in the Dictionary of Old English (DOE, as yet incomplete) had been quite unsatisfactory as well, largely in line with the various older historical dictionaries discussed so far. However, after Kornexl (1994: 447) argued in favour of a different treatment of particle verbs in the dictionary, namely as a “well-established morphological pattern in Old English”, the editors decided on a change in the treatment of such verbs. Among the weaknesses pointed out by Kornexl, the haphazard lemmatization of particle verbs with postposed particle is of particular interest in the present context. Thus in the following example (which is also mentioned by Kornexl 1994: 448, fn. 70) the DOE records the first particle verb s.v. beforan as an adverbial use of the particle (no cross reference to the entry for beforan-cuman glossing ‘praevenire’), while the second particle is recorded s.v. æfter as a preposition (there is no entry for æftercuman): (7)

gað ge beforon; ic eow cume æfter ‘go ahead; I will follow you’ [literally: ‘go you before; I you come after’] (GD 1 (C) [12.88.20])

Where the particle precedes the verb, there may be a separate entry for a ‘compound verb’; e.g. for the structurally completely equivalent verbs beforan-gān and æfter-gān, there are separate entries. However, apparently still in the tradition of Bosworth–Toller, these entries cover only the postpositional variants. Moreover, the treatment of these verbs is again quite unsystematic; cf. e.g. the following example, which is cited as evidence for æfter-gān (s.v.), but not included in the count for occurrences of beforan-gān: (8)

þa mengu þonne þa þe beforan eodan & þa þe æfter eodun cleopadun cwæþende ge-hǽl sunu dauiðes …


141

‘… those who went ahead and those who followed …’ [literally: ‘before-went … after-went’] (MtGl (Ru) 21.9) This is another example of the sometimes rather dubious handling of particle verbs in the DOE: the dictionary prints beforaneodan and æftereodun, but the edition (Skeat 1887, expressly referred to in the citation) prints a space between the particles and the verbs, while it remains quite unclear on the basis of both what the manuscript evidence is. Normally, however, the DOE would follow the entirely inconsistent conventions of the editions (cf. Gneuss 1973: 18, Jenkyns 1991: 388, Wetzel 1991: 223–225, Kornexl 1994: 446 and Dietz 2004: 597). That is to say, like the older dictionaries the DOE in the early fascicles does not recognize a connection between pre- and postverbal particles, let alone a separate category of particle verbs. Although the editors of the DOE have decided to follow Kornexl’s (1994: 447) plea for “a more generous inclusion of adverb plus verb formations as headwords” in the subsequent fascicles, the coverage of particle verbs in the DOE is still not entirely satisfactory. On the whole the main improvement in comparison to the older practices lies in the consistent insertion of cross references intended to establish explicit connections between the preposed and the postposed variants, e.g.: forþ Adv. … Some of the adverb and verb collocations cited here have elsewhere been taken as compounds. forþ- in verbal quasi-compounds, mainly element-by-element glosses of Latin pro-, where forþ may be taken as either adverb or prefix. forþ-fēran Vb. … Some instance of the compound have elsewhere been taken as adverb and verb.

But it is altogether unclear – in particular in the light of the cross references – why s.v. forþ-fēran there are only quotations with the preposed particles, while s.v. forþ (A.6.a.i) there are only quotations with the postposed particle. Thus despite the commendable cross references, the particle verbs are once again separated into ‘compounds’ and ‘phrases’ (or ‘collocations’) depending on the position of the particle. There are 41 cross references to ‘quasi-compound verbs’ s.v. forþ-, but none to forþ-fēran; that is, the verb can also not be regarded as sufficiently covered by the reference to a Latin model (and such references to Latin prefix verbs would not account for the positional properties of calqued particle verbs anyway). It must, of course, be conceded to the editors that the satisfactory classification of such

142


elements is not an easy task, as pointed out by Mitchell (1978) long before the appearance of the first fascicle: “How best to deal with these problems in future dictionaries and glossaries remains a difficulty … [t]he editors of the new Dictionary of Old English have a problem here!” (Mitchell 1978: § 51). The limitations of the DOE and the DOEC for syntactic research have been discussed in more detail by Koopman (1992), who is particularly concerned about the absence of variant manuscript readings and who shows that therefore important syntactic information is likely to be overlooked. Quite obviously, this applies also to variants involving pre- and postposition of the particle, or the use of particle verbs vis-à-vis simple verbs. But there is little doubt that at least the treatment of forþ-fēran could be considerably improved if it were not distributed across various entries, and that the overall quality of the DOE would likewise benefit greatly from the consistent adoption of such an approach throughout the dictionary. Again it is especially the lack of consistency and explicitness which is problematic. Thus the coverage of other particle verbs with forþ is much better, cf. e.g. s.v. forþ A.5.a.ii (forþ (be)cuman ‘to come forth/out; to emerge appear’), which gives quotations with both preposed and postposed particles. But then again, it is not clear why the construction should be treated s.vv. forþ, forþ-, forþ-cuman and cuman. Of course in this way the editors of the DOE can avoid taking a clear stance on the question how they actually classify the Old English particle verbs, but the users of the dictionary would be well served with a less cautious policy which does not result in the distribution of connected information across several different entries. Similar observations have been made by Wetzel (1991) in his review of the first fascicle of the DOE: Bei allen Schwierigkeiten der Abgrenzung im einzelnen sollte hier darauf geachtet werden, daß gleichartig gelagerte Fälle tatsächlich gleichartig behandelt werden und Doppelaufnahmen vermieden werden. S.v. dragan wird S. 555 im Zitat LawGer 9 me mæig myxendincgan ut dragan … in Übereinstimmung mit der Edition ut dragan geführt, bei den Querverweisen (S. 556) daneben auch ūt-dragan. Nach Konsultation von Hall (ūtdragan – LL 454,9 [=LawGer 9] ist immerhin vorstellbar, daß obige Stelle im Band des Buchstabens U unter ūt-dragan ein weiteres Mal verbucht werden wird. (Wetzel 1991: 224–225) [‘Despite all difficulties of drawing individual distinctions, care should be taken that equivalent cases are also treated equivalently and that double inclusions are avoided. S.v. dragan on p. 555 in the quotation LawGer 9 me mæig myxendincgan ut dragan … corresponds to the edition and has ut dragan, while the cross references (p. 556) also have ūt-dragan. Consulting Hall (ūtdragan – LL 454,9 [=LawGer 9], one can at least imagine that the quote


143

from above will be included once more in the volume for the letter U under ūtdragan.’]

Cf. also Kornexl’s (1994: 425) more general criticism: “In complex cases, what seems to be most important is consistency in treatment and clarity of presentation … one would at times wish for a more detailed and precise description of editorial policies and practices”. I am, however, somewhat sceptical whether Kornexl’s plea for a more liberal inclusion of particle verbs as headwords in the dictionary is sufficient. The appropriate treatment of such constructions depends crucially on the recognition that both the preposed and the postposed variants need to be included in one entry, where it should be made explicit that the preposed particles are distinct from inseparable prefixes. Similarly, although it is an interesting and lexicographically relevant piece of information whether a verb is a calque, the lemmatization of such verbs should in my view depend on the separability of the prefix/particle rather than on its etymology. 4.3.5. Concluding remarks on the historical dictionaries To sum up, it should have become evident that the coverage of Old and Middle English particle verbs in the main historical dictionaries is far from satisfactory, and it seems plausible that the perception of the long-term lexical development of the English phrasal verbs was considerably influenced by the inconsistencies and deficiencies which can be found in all of them, including the most recent ones. Although the typical users of such dictionaries can be assumed to be fairly knowledgeable about the history of English, it is unlikely that on the basis of the English historical dictionaries these users will be able to adequately perceive the lexical status of the Old and Middle English particle verbs – especially since the standard textbooks accounts of their history are similarly unsystematic and inadequate. 4.4.

Conclusion

In the preceding Chapter 3, which dealt with the development of the positional characteristics of the English verb-particle construction, with a strong focus on cross-linguistic parallels, it had been argued that the development of the English construction is rather unremarkable compared to the development of preverbs in other languages. But as has been shown in the present chapter, ever since Kennedy (1920), the loss of preverbal particle position has been referred to as the ‘rise of the phrasal verbs’, which are

144


portrayed as syntactically, semantically and lexically exceptional. Moreover, the analysis of the major historical dictionaries revealed that they fail to document the Old and Middle English particle verbs sufficiently and transparently, and thus contribute to the wrong impression that phrasal verbs are an essentially new structure which fully emerges only towards the Modern English period.

Chapter 5 Word formation, sound change and semantics As has been shown in the preceding chapter, the assumption of specifically ‘English’ properties of the verb-particle construction in English has traditionally played a prominent role in the research literature, and the notion is widespread that there must have been a ‘rise’ of the phrasal verb, which may be only explainable as the result of language contact – or at least that the development of some characteristics of the English phrasal verb cannot be explained language-internally. In the light of the comparative data from other Germanic languages, of the cross-linguistic parallels from genetically unrelated languages and of the diachronic syntactic development discussed in Chapters 2 and 3, this is a rather surprising state of affairs. In the present chapter it will be shown that such a view is in fact quite unfounded, not only with regard to the syntactic properties of the phrasal verb as a construction type in English, but also with regard to its lexical and semantic properties. The bulk of this chapter, then, will be devoted to identifying the main factors in the development of the Modern English construction type systematically. 5.1.

Changing prefix inventories in English

Since the relationship between the Old English prefix verbs and the particle verbs has been a constant concern in the literature it is relevant in the present context to attempt to establish criteria for distinguishing between prefixes and particles in the historical corpus, and to establish an inventory of Old English preverbs. Perhaps more importantly, though, the connections between the loss of many native prefixes and the development of the particle verbs ought to be delineated, especially since the Modern English phrasal verbs are in many ways functionally equivalent to the older prefix verbs. De la Cruz (1975: 77–78) provides the following list of prefixed verbs in the Ancrene Riwle (early 13th century) and its translation into Modern English by Morton (1853): gefor-

i clumben i streihte urswoluweð uoruret

climbed up outstretched up swalloweth up frets away

146

be-

to-

a-

Word formation, sound change and semantics vorworpen uorkeoruen bilepped binimen bileauen to blowen tospret tofleoted towarpled adruweð a cwenchen avleieð

cast away cut off wrapped up take away leave off blown up stretches out flits away shaken off drieth up put out driveth away

Examples of this kind suggest that the Modern English particles are in many ways functionally equivalent to older native prefixes, in particular with regard to their aspectualizing function, which the prefixes have in most of the above examples and which in the Modern English translation is consistently expressed by the particles (see e.g. Brinton 1988: 202–214 for further examples of Old and Middle English prefix verbs and their Modern English equivalents). One important question in regard to the development of the particle verbs concerns the way in which older native prefixes came to be replaced by particles. It has often been claimed that the prefixes were somehow ousted by the particles. But as will be argued in this section, the traditional views of the relationship between the development of the prefixes and of the phrasal verbs in English are in many respects quite unsatisfactory, and alternative explanations will be examined. Although a full-scale exploration of the topic is beyond the scope of this chapter, the characteristic semantic paths of development of preverbs will be considered, and the emergence of non-literal verb-particle constructions in the history of English will be sketched out. 5.1.1. The Old English prefixes General criteria for distinguishing between prefixes and particles are wellknown from the morphological and syntactic literature (see also Chapter 3 above). They can all be reduced to the simple observation that the former are inseparable and fused to the verbal stem, while the latter may occur in positions which do not immediately precede the verb. This goes along with prosodic and semantic differences which are quite similar to those found in the present-day Continental West Germanic languages (cf. the more


147

detailed discussion of the relevant changes in Section 5.2 below). But although there is solid evidence in principle in the Old English corpus for these differences, they are often impossible to establish for individual instances, not least because reliable test criteria are often difficult to apply to historical language stages. It must be assumed that the changing inventory of Old English prefixes and the concomitant loss of invariably bound forms reflects the general patterns of development of preverbs discussed in the Chapter 3, where the emergence of prefixes was shown to characteristically result in synchronic layering of cognate affixes and free forms. On the basis of the observation that all invariably bound prefixes are unstressed in Old English and of the comparative evidence from other Germanic languages, where verbal prefixes are unstressed even in those cases where there are homonymous free forms, it seems justified to use lack of stress as a criterion for prefix status in Old English as well. This approach has been criticized by Minkova (2008: 25) as circular, but this criticism may be admissible from a purely phonological point of view only. Minkova is right in pointing out that the mere reference to the stress patterns in Dutch and German is an insufficient means to establish the stress patterns of Old English. If however the comparative evidence, the diachronic development and the syntactic criteria for free forms are also taken into account, the plausibility of the identification of unstressed and inseparable prefixes is greatly enhanced by evidence from other Germanic languages on the one hand and from the distinct and systematic phonological and morphological differences in the further development of stressed and unstressed syllables on the other. There remains little ground for the possibility that productive verbal prefixes were stressed in Old English, despite Minkova’s (2008: 25–26) contention, “[t]he syntactic criteria used to establish separability of the particles in the modern languages … do not work well for Old English”, for which she refers to Fischer et al. (2000), whose views on this matter appear to be exactly the opposite, though. Positive evidence for stress in Old English is only to be gathered from the poetry, as in the following passage from Beowulf (for the use of metrical evidence for the establishment of the inventory of prefixes, see Dietz 2004): (1)

/ sie sio bær gearo | ædre be

the byre

ready

geæfned / þonne we ut

at once done

when

cymen |

we out come

‘Let the byre be ready, speedily prepared, when we come out.’ (Beowulf 3105–3106)

148

Word formation, sound change and semantics

As the underlining in l. 3106 indicates, alliteration provides clear evidence for stress on the particle (ut cymen) while the prefix in the preceding halfline precedes a stressed verbal stem (geæfned) and is thus quite unlikely to be stressed. On the basis of such evidence, an inventory of invariably unstressed verbal prefixes can be established, and the available evidence attests to the inseparability of these prefixes beyond doubt; cf. the general discussions by Harrison (1896: 58) and Campbell (1959: 32), and see also the discussion of Old English and Germanic stress in prefix and particle verbs by Bülbring (1902: 26–30) and the comparative discussion by Leinen (1891). Elenbaas (2007: 112–131) provides an overview of some Old English prefixes, but her etymological and semantic statements are almost entirely based on various rather dated reference tools (mostly Clark Hall 1960, but also Hiltunen 1983a) and are not always wholly reliable (cf. e.g. the conflation of etymologically and functionally distinct for- and fore-). Kastovsky (1992: 375) in his discussion of Old English word formation does not attempt to distinguish between inseparable prefixes and particles but only identifies those elements as ‘genuine prefixes’ for which no homonymous free forms exist. This schematic approach is unsatisfactory, though, since it results in a classification of invariably inseparable and unstressed æt- or be- together with separable and stressed adun and up rather than with inseparable and unstressed a- and ge-, etc., which in the light of the present discussion turns out to be undesirable on phonological, morphological and syntactic grounds. The inventories provided below – in (2) for the prefixes and in (3) for the particles – are based on the discussion in Dietz (2004) and supplemented by a critical survey of the sources I have used for Tables 5-1 and 5-2 below. Thus the following invariably inseparable and unstressed verbal prefixes can be established for Old English (for their etymologies, see Table 5-1):1 (2)

a-, æt-, an-/on-, be-, ed-, fer-, for-, fræ-, full-, ge-, mis-, oþ-, ot-, samod-, te-, tŏ-, un-/on-

I will draw no distinction between strong and weak forms of the prefixes (cf. e.g. Campbell 1959: 30–31), since there is little evidence in Old

1

Vowel length is marked only in tŏ- in order to distinguish it from the particle tō, cf. (3). The prefix an-/on- ‘towards, away’ given in (2) derives from ProtoGermanic *anda-, the particle an-/on- ‘on’ given in (3) from Proto-Germanic *ana.


149

English for synchronically productive strong-weak allomorphy (cf. Minkova 2008: 26–27 and the references given there). All the prefixes listed in (2) are of common Germanic descent, and about half of them are found as prefixes in all Germanic languages. For these there are typically (though not necessarily) no corresponding free forms in Old English, cf. Tables 5-1 and 5-2 below (for a detailed comparative case study see also Dietz 2005). Almost one half of the prefixes in (2) have ceased to be productive by late Old English, the prefixes fræ- and tebecame unproductive already in pre-Old English. However, other forms (e.g. of-, ofer-, on-, under-, tō-, þurh-) may be either stressed or unstressed in preverbal position, and in these cases the inventory of prefixes overlaps considerably with particles and also with prepositions and adverbs (cf. the overview in Hiltunen 1983a: 192–216, and the discussion in Chapter 3 above and the references there). In (3), I provide a core inventory of Old English adverbial particles: (3)

an/on, abutan, (a)dun, æfter, (be)foran, behindan, fore, forð, fram, geond, in(n), mid, niðer, of, ofer, ongean, onweg/aweg, under, tō, þurh, up(p), ut, wiþ, wiþer, ymb(e)

This list is not meant to be exhaustive, not least because the status of some forms is not entirely clear (cf. Kastovsky 1992 and Dietz 2004). To some extent the particles overlap with inseparable prefixes and with prepositions with respect to both form and meaning. All the Old English particles listed in (3) are spatial. Many of them continue being used in verbparticle constructions in Modern English; forþ becomes increasingly rare in the course of the Modern English period, while to is quite rare in verbparticle constructions already in Old English. For a discussion of the development of forþ, see Akimoto (2006) and the references given there (and also the discussion of the lexicographic coverage of forth in Chapter 4 above), while the marginal status of to is discussed by Elenbaas (2007: 144–145), who states that to is not found in verb-particle constructions in present-day English (it is certainly no longer productive, but it is found in idiomatic constructions like e.g. come to etc., cf. OED s.v. to). Some elements like æfter-, fore- or forþ- occur in (inseparable) loan translations of Latin prefix verbs, but rarely (or sometimes never) in independent native prefix formations (cf. DOE s.vv.). Minkova (2008) criticizes Dietz’ assumption that this is evidence of morphological opacity, and she points out justly that “calquing would in itself be evidence that these are perceived as independent morphological entities” (Minkova 2008: 43). Nevertheless the value of such formations as evidence for Old English

150


prefixation remains open to debate (cf. also the insightful discussion of Old English calques by Johnston 2011). Tables 5-1 and 5-2 below provide an overview of the Old English verb prefixes, based on my critical scrutiny of Bosworth–Toller, Dietz (2004), DOE, the DOE Corpus, EWA, Kastovsky (1992), Kluge–Seebold, Koziol (1972), W.P. Lehmann (1986), and on the MED. The most salient observation in the present context is that the inventories of the older prefixes are etymologically largely synonymous with the more recent particles, and this is of course exactly what is to be expected in the light of the discussion of the cross-linguistic properties of preverbs in Chapter 3. Thus the Old English ‘pure’ prefixes listed in (2) above mostly have spatial etymologies, as shown in Table 5-1, and with few exceptions these have developed also into aspectualizing or intensifying meanings in Old English, as we shall see below. (The term ‘pure’ prefixes is used differently by de la Cruz 1975, who only includes a-, be-, ge-, for-, of-, on- and toamong them. Following this practice, Hiltunen 1983a includes only these prefixes in his discussion and consequently his discussion of the Old English prefixes is somewhat limited. Since most later studies of the Old English particle verbs rely to some extent on Hiltunen’s database, this tends to apply to their coverage as well.) Table 5-2 offers a simple summary of the semantic development of the prefixes. But their individual histories are in part highly intricate and resist straightforward listing; cf. the detailed (though at times somewhat dense) discussion in Dietz (2004) and the references given there. Perhaps the most remarkable fact about the inventory of Old English verb prefixes provided in Tables 5-1 and 5-2 is its radical difference from the Modern English one. With the exception of five prefixes (be-, mis-, over-, un-, under-, marked * in the penultimate column of Table 5-2), all Old English prefixes are no longer fully productive in Modern English. But the loss of productivity of prefixes begins even before early Old English (the prefixes marked †††), and already at the end of the Old English period many other prefixes have ceased being productive (the prefixes marked ††), while others lose their productivity in the course of the Middle English period (the prefixes marked †). For a general discussion of the problems connected to establishing productivity in a historical corpus, see e.g. Kastovsky (1992: 355–361) and the more detailed discussion by Cowie & Dalton-Puffer (2002). The loss of productivity in Middle English proceeds, as in Old English, with different speed in different dialects (cf. e.g. Koziol 1972: §§ 197–289).


151

Table 5-1. Old English verb prefixes and their etymologies Old English prefix

aætan-/onan-/onbeedfer-/forforefræfullgegeondmisofoferonge(a)n oþsamodtetŏtōþurhun-/onunderwiþ(er)ymb(e)-

Proto-Germanic (**), West Germanic or pre-Old English (*) **uz **at **ana **anda **βi **iþ, **ið *fur **fur *fra **fulla **γa Gothic jaind **missa **aβa **uβer(i) Old High German ingagan **unþa *samuð *te *tuz *tō *þurh, *þerh **und/anda **under **wiþra **um-βi

etymology

‘out’ ‘at, to’ ‘to, in, on’ ‘away (towards)’ ‘by’ ‘back (again)’ ‘away’ ‘in front, before’ ‘forward, away’ ‘full’ ‘near, at, with’ ‘there’ ‘wrong’ ‘off’ ‘over’ ‘towards, against’ ‘away’ ‘together’ ‘apart’ ‘apart’ ‘towards’ ‘through’ ‘away (towards)’ ‘below’ ‘against’ ‘around’

Note that the term ‘bleaching’ is used in Table 5-2 as a general cover term for the loss of the literal, mostly spatial, meaning and the often concomitant emergence of new, more abstract meanings (on the complex semantic interplay of æt- and oþ-, see Dietz 2004: VII and cf. DOE s.v. æt); cf. also the discussion in Sections 5.2.1 and 5.2.4 below.

152


††† †† † *

– + + – – + – + – + – + + + + + + + – + + + – + + +

+ + + + + + + – + + + – – + – – + – + + – + + + – –

†† † †† †(†) * †† †† † ††† † †(†) †† * † * † †† †† ††† † † † * * † †

not productive after pre-Old English not productive after Old English not productive after Middle English productive in Modern English

further development if productive after OE

productivity

+ + + + + + + – + – + – + – – – + – + + – – + – – –

‘bleaching’

– + + – – – – + – – – + – + + + + + – – + + – + + +

literal meaning in OE

common Germanic prefix

aætan-/onan-/onbeedfer-/forforefræfullgegeondmisofoferonge(a)n oþsamodtetŏtōþurhun-/onunderwiþ(er)ymb(e)-

homonymous particle

Table 5-2. Old English verb prefixes: overview (OE = Old English, ME = Middle English, ModE = Modern English)

– ME at– – ModE be– – ME fore– ME fulME i– ModE misME ofModE overME again– – – ME toME toME thurghModeE unModE underME withME umb(e)-


153

5.1.2. The prefixes in Middle English and beyond Although many Old English prefixes have been lost, the prefix inventory of Modern English is larger than the Old English one. This is mainly due to the considerable influx of Romance loans in the centuries after the Norman Conquest, which provided the basis for the increasing productivity of borrowed affixes from late Middle English onwards (cf. Dietz 2002, Schaefer 2006 and Lutz 2002b and 2008). In Table 5-3, I provide an overview of the Modern English verb prefixes and their history which is based on my critical survey of the handbooks on English word formation. Although often in need of corrections in detail, they can be regarded as sufficiently reliable for the general contrastive overview intended here. The chief sources are Marchand (1969) and Koziol (1972) (which, however, are at times completely at odds), and I have crosschecked them against the more recent but less detailed discussions in Quirk et al. (1985), Bauer (1983) and Schmid (2005) for present-day English and Nevalainen (1999) for Early Modern English and also in part for earlier and later periods (cf. the overview in Nevalainen 1999 and the references given there). Unfortunately, the articles in the CHEL volumes on Middle English and on Late Modern English (Burnley 1992 and Algeo 1998) do not provide very much relevant information. Consequently Table 5-3 is considerably less exhaustive than might appear at first sight. Rather, it is the result of an evaluation and a critical comparison of the available reference works, where many problems remain undiscussed, which would deserve (and, in view of the extant literature, in fact would seem to be in need of) more detailed separate studies. At first sight the most remarkable observation is the considerable increase in prefix types from Middle English onwards. This observation, however, is somewhat misleading since the productivity of most of these prefixes is subject to significant restrictions. The new prefix formations are typically restricted to formal and technical registers, while the more common verbs with borrowed prefixes are typically not English prefix formations, but rather were already borrowed as prefix verbs (and consequently provided the model for the analogical rise of the borrowed prefixes). Moreover, when they are spatial, the borrowed prefixes only restrictedly combine with native verbs (cf. e.g. Marchand 1969, Koziol 1972 and Schmid 2005).

154


+ + + – – + + – + + + – + + + – + – – + – + + + + – + + + – – – + +

– – – – – – – + – – – – – – – – – – + – – – – – – – – – + – – – – +

+ – – ± + – – – ± ± – ± + – – + (–) + – – + – ± – – + (–) – – + + + – –

productivity in Modern English

post-Middle English borrowing

+ – – + + – – – – – – + – – – + – + – – + – – – – + – – – + + + – –

productivity in Middle English

Middle English from French or Latin

aabadbackbecircumcocontracounterdedes-/disdownenenterexfor-1 for-2 forehyperhypoin-1 in-2 interintromalmis-1 mis-2 nonoboffoutoverparaper-

Germanic

Table 5-3. Modern English verb prefixes (prefixes whose productivity is limited or doubtful are marked ±; prefixes of Germanic descent which are productive in Middle English are in bold type)

– – ± + since late Middle English + but increasingly marginal ± + ± since 17th c. + since Early Modern English + since 18th c. + since Early Modern English ± since Middle English + since late Middle English (?) ± in 15th–16th c. only + since 16th c. – since Early Modern English (merged with for-1) + Early Modern English only ± since Early Modern English ± since Early Modern English + + (~ in-1) + since late Middle English ± since Early Modern English ± since 17th c. + (merged with mis-1) ± since Early Modern English ± since Late Modern English very rare after Middle English + since late Middle English + ± ±

Changing prefix inventories in English postprepro-/porresubsupersurtransununderupwith-

– – – – – – – – + + + +

+ + + + + + + + – – – –

– – – – – – – – – – – –

– ± – ± – – – – + + ± +

155

± since Early Modern English + since Early Modern English – + since 15th c. + since Early Modern English + after Middle English ± after Middle English + after Middle English + + ± since Middle English ± decreasing since Old English

It is remarkable how ready many authors are to offer simple causal explanations for the connection between the developments of prefixes and particles which at closer inspection turn out to be rather unsatisfactory. E.g., referring to a claim made by Hiltunen (1983a: 92), Claridge (2000: 86) states in her overview of the literature on the ‘rise’ of the phrasal verb that “the rise of the phrasal verb is also connected with the fall of another construction, OE prefix verbs. The latter’s very sudden decline to almost zero productivity and use by early ME left a slot to be filled by the newer analytic constructions”. Already Samuels (1972: 163) had discussed the development as “a major example of the combined working of push- and drag-chain processes of replacement”, glossing rather broadly over diachronic inconsistencies and morphological and semantic details; note however that Samuels sees the process as gradual rather than sudden. But although such claims are not entirely unfounded as far as there is indeed a connection between the loss of preverbal affixes and the development of the phrasal verbs, their premises are quite mistaken, as a closer look at the productivity of the prefix inventories in Middle English, especially in comparison to the development in Old English, shows (cf. Tables 5-1 and 5-2 above). One very remarkable and perhaps rather unexpected finding is worth being pointed out: the number of productive prefixes of Germanic origin in Middle English is actually higher than the number of productive prefixes of Romance origin, and there are even new Germanic prefixes in Middle English, while the Romance prefixes very typically only become productive in late Middle English or even later, in Early Modern English (see also Dalton-Puffer 1996 and Dietz 2002). This is not to say that Burnley’s (1992) claim mentioned in the preceding chapter that Middle English became depleted of productive prefixes in comparison to Old English would be fully justified. Since many of the Old

156


English prefixes had already become (or were about to become) unproductive in the course of the Old English period (cf. Table 5-2 above) the pattern we can observe is rather one of continued loss and renewal. Although this topic appears to be in need of further study, it can be safely concluded that prefixation as a native word formation type remained productive in Middle English. Moreover, as Schröder (2008) has shown, the remaining native prefixes continue to be used as a productive means of verbal word formation to a considerably greater degree than had traditionally been assumed. The Middle English borrowings, then, cannot have ousted the Old English prefixes for a number of reasons, and each of these reasons would in itself be sufficient to preclude that possibility. With regard to the native prefixes, it emerges that some were already lost before the influx of French borrowings began, while others were in the process of being lost. As is indicated in Table 5-3, this applies not only to the productivity of these prefixes, but also to their semantics. Many of the originally spatial preverbs had more or less lost their spatial meaning already in Old English. As we shall see below (Section 5.2.4), the Old English prefix verbs were typically semantically equivalent to the non-spatial phrasal verbs in Modern English, while spatial meaning was already typically expressed by the adverbial particles. Although this is identified correctly by Kastovsky (1992: 377), his metaphorical description of the development is misleading (cf. the discussion in Chapter 4 above). For a critical overview of the metaphorical characterisations of the loss of Old English prefixes and the often inherently contradictory explanations for the development, see Lutz (1997: 258– 261), where it is also criticized that the traditional treatments of Old English verb formation fail to include satisfactory discussions of the particle verbs. The post-medieval development of the adverbial particles in English word formation shows the relationship of word order and morphologisation of affixes quite clearly and thus underlines the validity of the general paths of change discussed in Chapter 3 above. The syntactically conditioned disappearance of preverbal particles coincides with the loss of the emergence of new native prefixes, cf. Marchand (1969: 109): “Those particles which by the 15th century had not acquired the character of inseparable prefixes could no longer precede verb forms, except the nominal ones (verbal substantives and participles)”. Although he nevertheless treats those prefixes in his chapter on compounding, Marchand correctly notes their phonological and semantic characteristics, which unambiguously identify them as prefixes (cf. also Hackstein 2011 on the emergence of unstressed prefixes out of stressed preverbs in Indo-European).


157

The analogical formation of borrowed prefixes, on the other hand, is unaffected by either the presence or the absence of native ones, especially since the borrowed prefixes have a different, complementary role in the English lexicon (cf. also the discussion in Chapter 6 below). Schaefer (2006: 272) stresses that these borrowings ought to be seen in the context of “the intensive elaboration of the vernacular by enhancing the vocabulary with a ‘more appropriate’ word”, that is, appropriate for new types of discourse in the vernacular. Supportive evidence for this claim can also be found in the other Germanic languages, cf. in particular the loss of native prefixes in the North Germanic languages, where later new prefixes were formed on the basis of German loans, while in the Continental West Germanic languages almost identical sets of Latin- or Greek-derived prefixes were adopted despite the continued productivity of native prefixes. This is not to say, however, that the degree of integration of borrowed items in these languages were not different from each other and from English (see in particular the discussion in Lutz 2002b and 2008 and the references given there). These conclusions have been obscured so far by the mode of presentation of borrowings and prefixes in the handbooks, where Germanic-based prefixes are usually treated separately from Romance-based prefixes, which are usually ascribed to the Middle English period because the borrowed verbs containing these prefixes are mostly first attested in Middle English. In the English research tradition, Germanic prefixes are treated as a particularly Old English phenomenon, Romance prefixes in the context of Middle English borrowing, and ‘phrasal verbs’ as a new and particularly English phenomenon. The changes in word order are usually almost entirely neglected by studies of word formation, although these changes play a central role in the development of verbal prefixation and in the emergence of phrasal verbs. Thus, what is missing is a treatment of the different types of word formation in a coherent and systematically related way. Such a discussion would also have to describe the historical development of new particle types, including those with a Romance etymology (e.g. apart), the persistence of some Old English particles and the emergence of new, Germanic-based prefixes in Middle English, a topic which so far has been almost entirely neglected in the literature. A truly thorough account of the development will also have to consider the relationship to the loss of the native verbal suffixes, as has been exemplarily demonstrated by Peters (2006) with regard to Old English -ettan; cf. Bauer’s (2003b) discussion of the possible typological implications of the loss of Old English prefixes (where, however, the connection to particle verbs is not explored). Similarly, the connection between the loss of verbal suffixes and the

158


development of derivational prefixes in the history of German is discussed in Erben (2006: 136–137). Broad metaphorical generalizations which point to the loss of vitality of the inherited prefixes and the adoption of new, borrowed prefixes as both the cause and the gap-filling result of the demise of the old prefixes will predictably be of little help for a deeper understanding of the development. 5.2.

Preverbs and particles in medieval English

This leaves us with the task of actually accounting for the loss of the majority of the Germanic prefixes towards Modern English. As has been shown in the preceding section, the loss of the native prefixes is by no means a sudden phenomenon at the beginning of the Middle English period. Moreover, the common explanations of the disappearance of those prefixes are either rather vague and metaphorical or too general to be of explanatory value. This is true in particular of those accounts which see the loss of prefix verbs and their substitution by phrasal verbs as due to the ‘drift’ of English towards greater analyticity, as e.g. suggested by van der Gaaf (1930), Konishi (1958) or Hiltunen (1983a). Claridge (2000: 87) justly observes that the suggestion that “the general trend towards analytic constructions disfavoured the synthetic prefixal constructions … may be true, [but] it contains a certain amount of circularity, which disqualifies it as a ‘reason’”. Unless one were to identify this drift as an independently operative force in language change (for which there is no evidence at all), one would still have to explain more precisely what the driving forces behind the loss of the prefixes are. Moreover, one would also have to explain why new (native and borrowed) prefixes (i.e. synthetic forms) occur in this ‘drift towards analyticity’. 5.2.1. On sound change and word formation In most descriptions of the loss of native prefixes reference is made to the loss of vitality of the native prefixes and to the greater vitality of the new Romance prefixes, e.g. by Baugh & Cable (1993: §§ 137–138) and Koziol (1972: § 192); see the critical discussion in Lutz (1997: 259), who observes that the older studies create the impression that the native prefixes were both ousted by the Romance borrowings and “doomed anyway”. How can such a ‘loss of vitality’ be described more precisely? Traditionally it had been seen in connection to the fading of semantic content of the prefixes (cf. Kastovsky 1992: 377 and the references given there), but although the


159

abandonment of prefixes is not unlikely to be connected to their semantic development, this cannot have been the only causal factor in the development in English, as a closer look at the Old English prefixes on the following scale of semantic independence (based on Minkova 2008: 30) serves to show: be-, ge-, for-, of-, on-, to-, and-, ed-, mis-, un-, ... geond-, ofer-, þurh-, wiðer-, ymbleast independent

most independent

What is represented here (apparently on the basis of Horgan 1980, Hiltunen 1983a: 55–91 and Wedel 2001) as semantic independence is, in fact, a synchronic and diachronic cline from the etymological, typically spatial, meaning (‘most independent’) to no apparent meaning at all, up to the point where simple and prefixed verb may be synonymous, or where etymologically different, even antonymous prefixes may become interchangeable. The cline thus shows what Minkova identifies as the desemantization of the particles. In fact, although the process of desemantization of many prefixes was clearly well under way in Old English and continued in Middle English, the scale shows clearly that this process cannot be correlated to the loss of prefixes in a straightforward manner: the five prefixes on the left side of the scale were “largely desemanticized” (Minkova 2008: 30) already in Old English. But Minkova’s conclusion that there is a clear correlation between the semantic and prosodic weakness of the prefixes is misleading: as a comparison with Tables 5-2 and 5-3 above shows, some of the semantically least independent prefixes continue to exist in English well beyond the Old and even the Middle English period, while conversely some of the semantically most independent ones are abandoned rather early. A case in point are the two leftmost (i.e. the least independent) prefixes shown above, be- and ge-: while be- continues in English as a productive prefix until today, ge- was lost in Middle English, and this process is clearly very much under way already in Old English (on Wedel’s 1997 and 2001 view that ge- was used systematically as a marker of perfectivity in Old English, see the discussion in Section 5.2.3 below). Lutz (1997) presents an alternative explanation with directed sound change in unaccented syllables as the ultimate cause of the demise of many Old English verbal prefixes. Thus the major factor leading to the loss of native prefixes is to be found in the phonotactically determined destabilization of the prefixes, and she shows that those prefixes were lost whose phonotactic structure rendered them prone to weakening and loss of

160


pretonic consonants. This is in principle not dissimilar to an earlier suggestion by Marchand (1969: 130), where, however, just as in the other traditional accounts of the development, the phonological details of the loss of some prefixes are not explored systematically and in greater detail. Depending on their phonotactic stability, different Old English prefixes were abandoned at different times between Old English and Early Modern English, and the surviving native prefixes and the new borrowed prefixes are, characteristically, quite similar with regard to their stable phonotactic structure. Lutz (1997: 260) therefore argues that “it is not possible to explain the decline of the Old English means of prefixing on the basis of the assumption that prefixation as a means of deriving complex verbs from simplex verbs has been given up altogether” (see there for a short summary of the underlying theoretical assumptions, and cf. also Lutz 1991 and 1992). The effects of sound change on the unaccented syllables of the prefixes can be summarized as follows: phonetic attrition in unaccented syllables → morphological decay → loss of prefixes → lexical gaps → new verbs (with phonotactically more stable structures)

Thus the development of the Old English preverbs can be seen as ultimately caused by changes in other areas of linguistic structure: phonology (affecting the development of the Old English inseparable prefixes) and syntax (since the changes in English word order towards strict SVO resulted in the general loss of preverbal particles; cf. the discussion in Chapter 3 above). This model is in accord with the cross-linguistically attested model of the development of preverbs proposed in Chapter 3 above, where the long-term development of preverbs along a diachronic cline ‘discourse > syntax > morphology > morphophonemics > zero’ was shown; cf. also Givón’s (1979: 209) classic model of change from discourse to zero as a unidirectional cline (but see e.g. Traugott 2003b: 630– 631 for a critique of this as a model of grammaticalization). A clear case of loss due to phonetic attrition is provided by the prefix ge-, which has already been touched upon briefly. It was shown already by Pilch (1955) that its loss was ultimately caused by phonetic attrition (with subsequent paradigmatic levelling and complete loss):


161

Jedoch schafft der Übergang ʒe->i- die Voraussetzungen für die phonetische Abnutzung unseres Präfixes und leitet damit indirekt seinen Untergang ein. I- traf nämlich häufig mit gleichem oder ähnlichen Vokal im Hiatus zusammen. Mit diesem wurde es kontrahiert und fiel aus, z.B. hi iseoþ > hiseoþ > hi seoþ. Damit entstanden auf rein phonetischem Wege aus alten Kompositis neue, unpräfigierte Formen, und zwar vorzüglich nach vokalisch auslautenden Proklitika. (Pilch 1955: 39–40) [‘The transition from ʒe->i-, however, creates the conditions for the phonetic attrition of our prefix and thus leads indirectly to its demise, since i- frequently was in hiatus with an identical or similar vowel with which it coalesced and it was then elided, e.g. hi iseoþ > hiseoþ > hi seoþ. In this purely phonetic manner new, unprefixed forms developed out of old compounds, mainly after old proclitics that ended in a vowel.’]

Pilch (1955: 38–48) provides a detailed analysis of the successive stages of development (for a thorough discussion of the older literature on the semantics of Old English ge-, see Lindeman 1965 and 1970 and cf. also the discussion by Wischer 2004 and Wischer & Habermann 2004). But in the literature the loss of ge- is widely ascribed to semantic weakening, which, as we have seen, is rather unlikely considering the retained but similarly desemanticized prefix be-, while the distribution and spread of the phonetic reduction of ge- (> Middle English i- > ø) ties in with the general phonotactic developments observable in English. Cf. e.g. the discussion of Kastovsky’s (1992) account in Chapter 4 above and the more detailed discussion in Lutz (1997: 277–278) and the references given there. Hiltunen (1983a: 65) states that although there seems to be a decline in the use of gein the course of Old English (though his data tend to resist a more straightforward claim), little changes in the semantics of the prefix. But nevertheless Minkova still assumes that “functional loss is a precondition for its phonological demise” (2008: 30). In this context a comparison with the development of the two prefixes ge- and be- in German is instructive (cf. Habermann 1994; for the more recent literature, see Erben 2006). In present-day German, ge- has entirely ceased being productive as a derivational affix, and it is fully grammaticalized as a past participle marker, while in Early New High German it still has a marginal derivational function (Habermann 1994: 372–388). The prefix be- has retained a considerable number of derivational meanings, although Habermann observes a tendency in Modern German for this prefix to become more grammatical as well (see Habermann 1994: 255–293; for an overview of the derivational uses of be- in present-day German, see Fleischer & Barz 1995: 320–321). English be-, on the other hand, loses most of its

162


derivational meanings in the course of the Early Modern English period. Marchand (1969: 148) states that “[i]t is only with the shade of overloadedness, disparagement, or ridicule, that be- is a productive morpheme in Present-day English”. Erben (2006: 138) points out that in German the functional growth of native prefixation is the result of the complex historical interplay of various written discourse traditions from the Middle Ages onwards (see there for further literature). In contrast, the corresponding discourse traditions in English favoured the use of Romance loans; cf. the discussion of the role of discourse traditions in Schaefer (2006) and the comparative assessment of the role of word formation and borrowing in English and German in Lutz (2002b and 2008). The approach suggested here accounts well for the loss of prefixes containing phonetically weak consonants such as oþ-, where the inherently weak consonant /-θ-/ occurs in a positionally weak final position in an unaccented syllable. The consequent replacement of the prefix in later Old English by the inherently stronger prefixes of- and æt- ties in nicely with other phonotactically conditioned replacements of dental fricatives from late Old English onwards, just as the later abandonment of the positionally equally weak but inherently stronger labiodental fricative in of-. Another phonotactic factor must be taken into account for the loss of æt-, namely that some formations with this prefix would have resulted in unacceptable consonant clusters after resyllabification of Old English /æt$’…/ to /ə$’t…/ in Middle English, e.g. atbear ‘carry off’ (*/tb-/), atfall ‘fall away’ (*/tf-/), atglide ‘slip away’ (*/tg-/), atcome ‘get away’ (*/tk-/), etc. In those cases where the resultant consonant cluster was phonotactically acceptable the morphological structure was bound to become obscure, e.g. atren ‘run away’ (?a$’tren), atshoot (?a$’tshoot), etc. Thus, the plosive would either be lost altogether – just like ed- with the phonetically weaker lenis plosive /-d-/ already in Old English (“due to the phonotactic vulnerability of plosives in pretonic position”, Lutz 1997: 273) – or be subject to resyllabification. In any case, the prefix was eventually reduced to word-initial pretonic /ə-/, which was phonetically indistinct from other unstressed elements, in particular the formerly distinct but similarly weakened prefixes a-, on- and of-. See Lutz (1991: 94–116 and 1997: 267–270) for more examples. Dietz (2004: 573–579) argues that prefix substitution was due to semantic factors at least in the case of æt- substituting oþ-. But his argument rests on the assumption that the substitution of oþ- by the etymologically converse prefix æt- (and sometimes vice versa) must have been due to mutual influence. He does not elaborate on the details of this mutual influence, but his discussion of ætberan vs. oþberan in Beowulf shows that the prefixes become synonymous as soon as the origin or the goal of the verbal action is


163

specified; the semantic convergence does not explain why it is oþ- (and not æt-) which is abandoned first. The developments of wið- and to- call for more detailed comment, since the former is abandoned considerably later than would otherwise be expected with regard to its phonotactic structure (and it still exists in present-day English in a small number of lexicalized formations, e.g. withdraw), while for the disappearance of the latter no phonotactic explanation at all is at hand. Lutz (1997: 274–275) suggests that wið- resisted destabilization for longer than expected due to a combination of various factors: First, from late Old English onwards the dental plosive in the preposition mid was phonotactically weakened (> mið), and the preposition was gradually replaced by wiþ (while the Old English preposition wiþ was entirely abandoned and replaced by against); consequently, the “dental fricative of the Middle English prefix with- was formally identified with that of the preposition with (< OE mid), whose fricative must have been stronger than that of OE with-” (Lutz 1997: 275). Second, the onset /w/ and the nucleus /i/ contributed to the greater phonotactic stability of the prefix. And third, it is conceivable that the existence of the homonymous preposition provided formal support for the prefix. With regard to to-, Lutz (1997: 278–279) assumes that this prefix was abandoned due to the rise of the toinfinitive and the avoidance of repetition (on the rise and spread of the toinfinitive, see e.g. Fischer 1992: § 4.6.2, Fischer et al. 2000: Ch. 7 and Los 2005). But although both hypotheses are not implausible, they clearly show that just like in the cases where attrition was to result in homonymic clashes a wider functional perspective is often necessary to account for the ultimate loss of the prefixes in a satisfactory way. Thus, although it is clear that nearly all of the Old prefixes which were eventually lost were subject to phonetic attrition, this in itself does not always provide a sufficient explanation for their complete disappearance. It is notable that, quite typically, phonotactically conditioned phonetic attrition led to homonymic clashes, as in the case of ymb- (> Middle English um-/em-), fore- (> Middle English for-), which provides an explanation for the abandonment of one or more of these homonyms. But although homonymy has been used extensively as an explanation of linguistic loss in the older structuralist tradition, it should be stressed that homonymy in itself is never sufficient as an explanation of loss, unless the homonymic elements occur in identical contexts (see e.g. Samuels 1972: 67–75 et passim, de la Cruz Cabanillas 1999 and Hopper & Traugott 2003: 102–103). In the cases at hand homonymy destroys the former paradigmatic distinctions between the prefixes and thus greatly impairs their communicative value by rendering them dysfunctionally ambiguous. Consequently, their replacement by less

164


ambiguous elements can be ultimately ascribed to the phonological changes which rendered them homonymous. Altogether it seems clear, though, that the semantic development of the prefixes is unlikely to be the sole, let alone the primary driving force behind the abandonment of many of the native prefixes in the history of English, although the loss of semantic content is likely to have acted as a complementary factor to the abandonment of destabilized prefixes and also as an independent factor, as the abandonment of phonotactically stable but semantically extremely diffuse /ə-/ from various sources in Middle English shows. Since verbal prefixation remains productive as a word formation type in English, the loss of older prefixes must be regarded as the sum of individual losses, each of which must be explained individually. There is no evidence for the particles ousting the prefixes in the sense that they were a causal factor in their disappearance. Rather, as soon as the formation of new prefixes was blocked by the loss of the necessary preverbal input, this pattern of attrition and replacement could be continued only by using the now postverbal particles or, alternatively, by analogical formation of borrowed prefixes. Several details of the loss of the Old English prefixes are still in need of further exploration, although the major contributing factors have already been identified in the literature. Nevertheless, as we have seen there appears to be a recurrent pattern of phonetic attrition and desemantization, up to the point of complete loss of morphological and semantic substance. Moreover, this process sets in as early as pre-Old English and is therefore quite unlikely to be essentially connected to external factors entering the scene in the course of the history of English. In the context of the present section it is worth stressing this observation again, since it is inextricably connected to the development of the verbparticle construction. The relationship between the prefix verbs and the phrasal verbs is thus clearly much less vexed than has often been assumed in the literature, and at closer inspection even the presumed loss of vitality of the prefixes turns out to be rather doubtful, since the particles turn out to be their functional and semantic equivalents in almost every respect, with two significant differences, namely first with regard to their positional properties, which, as has been shown in Chapter 3 above, must be assumed to prevent their development into preverbally bound affixes, and second with the fact that they are stressed. These two factors work together not only to prevent phonetic attrition of the particles, but also to effect a higher degree of prosodic and semantic salience.


165

5.2.2. Some comparative evidence A more-than-cursory, in-depth study of the lexical developments of at least a representative number of particle verbs is conspicuously absent from the literature and would be highly desirable. Apart from the information available in the OED, very little and practically nothing systematic has been undertaken in this direction so far (for an overview of the changing semantic properties of the particles in medieval English, see Hiltunen 1983a: Part IV). The present study as well is not primarily concerned with the development of individual lexical items. With the morphological and phonological discussion of the preceding sections in mind, we will now turn to a brief general overview of the major semantic paths in the development of the English verb-particle construction. As has been argued in the discussion in Chapter 3, these paths are quite similar for all kinds of preverbs, with a high degree of cross-linguistic consistency, and in English just as in the other Germanic languages the long-term changes observable in separable prefix verbs and in particle verbs are largely identical. To start with, let us therefore return briefly to the typical path of development of preverbs sketched in Chapter 3 above. The viability of the model sketched there is provided by the Germanic preverb GE- (ProtoGermanic *γa). The particle is not attested in its probable etymological sense ‘near, at, with, together’ in any Germanic verb (cf. W.P. Lehmann 1986: s.v. ga), but evidence for the etymological meaning of GE- is found outside verbal prefixation in nominal formations (cf. e.g. Kluge–Seebold s.v. ge-). The development from free form to bound affix via preverbal particle is nevertheless evident from a comparative perspective. In all modern Germanic languages where GE- still exists it has been a bound affix throughout recorded history, as e.g. in present-day Dutch and German, where the prefix ge- has undergone grammaticalization as a past participle marker. But the situation is different in morphologically more archaic Gothic, where the cognate preverb ga has not yet been subject to complete morphological fusion with the verbal stem. Gothic ga is synchronically both a bound derivational affix, which may occur in fully lexicalized formations, and a residually free particle (Braune–Heidermanns 2004: §217a Anm.2; cf. there for further references). Moreover, it seems that already in Gothic bound ga- has a marked tendency to become largely desemanticized, while the aspectualizing meaning tends to be associated with the freer particle. Although the exact function of ga is difficult to establish, it has been known for long that ga can be an aspectualizer in Gothic, but it can also be used as a transitivizing derivational affix. However, in other cases ga seems

166


to be semantically almost empty, as the synonymous use of some verbs with and without the preverb shows (cf. e.g. Streitberg 1920: 194–200). The classic example of aspectualizing ga- is durative bairan ‘bear’ vs. perfective ga-bairan ‘give birth’. But although relatively straightforward examples such as this do by no means form the bulk of the Gothic evidence, Streitberg insisted that all Gothic preverbs have always an aspectualizing function comparable to the one found in Slavic (cf. Streitberg 1891). His proposal sparked one of the most furious debates in the history of linguistics, in the course of which it became clear that his views cannot be reconciled with the Gothic data (see Lloyd 1979 for a critical discussion of the older literature, and Lindemann 1970 and Brinton 1988: 199–202 et passim with particular reference to English; cf. also the bibliography in W.P. Lehmann 1986 and Braune–Heidermanns 2004: § 217a Anm.3). Leiss (1992: Section 2.5 and 2002a: 11–14), who is quite determined to present the development of GE- as the loss of an old aspect system, simply claims that this is the case and glosses over an enormous amount of difficulties in analysing the Gothic and other Germanic data, and seems not to consider the possibility that the Germanic aspectualizers have never formed a fully coherent system of aspect marking, as e.g. the typological observations by Bybee, Perkins & Pagliuca (1994: 87–88) would suggest. To my mind, her examples show that the assumption of systematic aspectual verb pairs in Gothic (and corresponding pairs involving verbonominal combinations in Modern German) cannot be upheld. The semantic interpretation of her examples seems arbitrary, just like the selection of examples, while the semantic motivations for the common Germanic development she suggests is similarly based on little persuasive evidence. This applies especially to the geographical distribution of the disappearance of aspectualizing particles, which Leiss notes but for which she cannot offer an explanation. In contrast, the phonological, morphological and syntactic reasons suggested here serve to explain this development very well (see also the discussion below); for a critique of similar arguments with respect to Old High German put forward by Schrodt (2004), see Fleischer (2006). In the present context the details of the analysis of Gothic ga- must be ignored, and it suffices to state the largely uncontested observation that the Gothic preverbs may have aspectualizing values, while no reassessment of the discussions of the issue will be undertaken (which Lloyd 1979: 8 summarizes by stating that “Streitberg’s critics have been far more successful, however, in demolishing – or at least badly denting – his theory than in developing a more satisfactory one”). In Gothic, then, ga is regularly separated from the verb by clitic particles in Wackernagel position, as e.g.


167

ga-saiƕan ‘see’ in the following example from Wulfila’s Bible translation (after van Kemenade & Los 2003: 98; the example as quoted here is slightly altered following Streitberg 2000 [1919] and Fortson 2004: 313): (4)

ga- u-

ƕa-

seƕi

ga- INT- anything- see:PRET.3SG.SUBJ

‘whether he saw anything’ (Wulfila, Mark 8:23) Van Kemenade & Los (2003) seem to take the intervention of a clitic particle as evidence against independent morphological status of the preverb, and they argue that this may have been the driving force behind the development towards bound affix status of ga- (and if I work out their argument correctly, other Gothic preverbs as well), where the particle was supposedly reanalysed as a bound morpheme if learners had otherwise no evidence for independent ga- (van Kemenade & Los 2003: 99). But this interpretation is unnecessarily complicated, since it is much more economical to analyse the possibility of tmesis with ga- as a relic from an earlier independent usage (which is acknowledged by van Kemenade & Los), while the preverb had otherwise already become a prefix. Moreover, they do not take the morphonological evidence against their analysis seriously (although they refer to it [113, n. 12]) and they ignore the fact that clitic particles in Gothic normally follow full words only (cf. e.g. Streitberg 2000 [1919]: s.v. u- and Fortson 2004: 313–314). In fact, van Kemenade & Los do not comment at all on the fact that the intervening particles occur in Wackernagel position, which affords a serious obstacle to a reanalysis of the clause-initial particle as a prefix, and their parallels from non-IndoEuropean languages are somewhat doubtful and of little help in a synchronic analysis of the Gothic data. Moreover, Gothic also offers evidence for layering of older and younger preverbs, which is particularly noticeable in the co-occurrence of synonymous doublets, e.g.: (5)

iþ

atta

us-gaggands ut

but father out-going

bad

out asked

ina him

‘But his father, coming out, pleaded with him.’ (Wulfila, Luke 15:28) This kind of construction may be interpreted as evidence that the prefix is about to lose its original spatial meaning, or at least that semantic bleaching has already affected the prefix to such an extent that reinforcement through doubling occurs; cf. Götti (1974: 35–42), and see also Dietz’ (2004: 584–

168


586) discussion of doubled prefixes in Old English and the discussion below, where it is argued that at least in Old English the particle does not reinforce the prefix. In the translation of the Greek Bible text usgaggan and utgaggan may also be used interchangeably (cf. Streitberg 2000 [1919]: s.vv.). Evidence like (5) for the relative order of the two preverbs shows the greater degree of fusion of us- with the verb: us- is always closer to the verb than ut-; also, if both preverbs precede the verb, only the sequence utus-verb is attested, but never *us-ut-verb. That is to say, some Gothic preverbs, like us, are close to becoming fully bound affixes (with concomitant semantic changes), while others, like ut, are positionally freer adverbial particles (cf. van Kemenade & Los 2003: 102); cf. the more detailed discussion of usgaggan vs. utgaggan by Götti (1974: 41) which despite some corrections in detail supports Kemenade & Los’ general argument: “Ich meine, daß utgaggan ein Versuch ist, die Bewegungsrichtung im Präfix deutlich auszudrücken, wie das etwa bei ingaggan der Fall ist. Es ist denkbar, daß wir hier den Beginn einer Entwicklung fassen können, an deren Endpunkt eine deutliche Scheidung der beiden Verben usgaggan und utgaggan steht” [‘I think that utgaggan is an attempt at clearly expressing the spatial direction in the prefix, as is the case with ingaggan. It is conceivable that we can perceive here the beginning of a development which ultimately leads to a clear distinction between the two verbs usgaggan and utgaggan.’]. A particularly remarkable instance of identical doubling is found with ga, which invariably precedes the verbal stem: (6)

ga ga-leikon ga

sik

ga-liken:INF oneself

‘change oneself’ Cf. also in-galeikon, miþ-galeikon, þairh-galeikon (Streitberg 2000 [1919]: s.vv.), where ga- appears to have fused with the stem completely – there is no simple verb *leikon attested in Gothic (but of course the corpus of Gothic texts is rather small) – while the particles retain their meanings as free forms. This also explains the sequence of elements in clauses like the following:


(7)

ga-h

þan miþ

ga-and then with

169

[ga-]sandedidum imma broþar [ga-]send:PRES.1PL him

brother

‘and then we are sending with him a brother’2 (Wulfila, II Cor. 8:18) A comparison between the two mss. containing the Gothic text shows (cf. Streitberg 2000 [1919]: 311) that only one of them (ms. A) has gasandedidum, while the other (ms. B) has the simple sandedidum. Desemantization can likewise be observed with other Gothic prefixes, e.g. fra ‘away’ and us ‘out’. In contrast to ga, however, these two prefixes have retained their etymological spatial meaning to a larger extent. The more archaic nature of the Gothic corpus shows also in the fact that Gothic only has nine prefixes, while the higher numbers of prefixes in the more recent old Germanic dialects reflect the secondary, post-Proto-Germanic development of these prefixes (cf. Braune–Heidermanns 2004: §217a). Old English fræ- and a- are cognate to Gothic fra- ‘away’ and us- ‘out’ (cf. W.P. Lehmann 1986: s.vv., EWA s.vv., and DOE s.vv.) and occur only as bound affixes. Like ge-, they are abandoned at some point in the history of English (cf. Table 5-2 above). Moreover, the prefixes in Old English share practically all characteristics with their Gothic cognates, except that the Old English prefix system is clearly representative of a more advanced stage of the Germanic preverb development. In contrast to the situation in Gothic, the Old English prefixes are invariably bound, and fræ- and a- have lost their etymological spatial meaning almost completely; fræ- has even ceased being productive already in early Old English, while a- continues being used throughout the Old English period, possibly as an intensifier (this is at least the common interpretation of this largely desemanticized prefix, cf. e.g. de la Cruz 1975 and Hiltunen 1983a; but cf. the discussion below). As in Gothic, GE- is semantically perhaps the most elusive of all productive prefixes in Old English, and it seems that it is often added to a simple verb without any change of meaning at all. However, throughout the early Germanic dialects it can be used as an aspectualizer, as the following parallel passages from three entirely independent Bible translations (Wedel 1997: 327) indicate; the correspondences are even more remarkable if it is considered that the three Germanic translations are not based on the same text, not even the same language (the Gothic text is a translation from Greek, the Old English and Old High German ones from Latin): 2

On gah- (with -h < Proto-Indo-European *kwe ‘and’) in this example, see Braune–Heidermanns (2004: § 24 Anm.2 and § 218).

170


(8)

Gothic (Wulfila, John 17:12 and Luke 2:19): (i) Þan was miþ im in þamma fairau, ik fastaida ins in namin þeinama. (ii) iþ Maria alla gafastaido þo waurda, Þagkjandei in hairtin seinamma.

(9)

Old English (Gospels, John 17:12 and Luke 2:19): (i) Þa ic was mid him ic heold hi on þinum naman. (ii) Maria geheold ealle þas word on hyre heortan smeagende.

(10) Old High German (Tatian ζ 178,4 and α 6,6): (i) mit diu ih uuas mit in, ih hielt sie in thinemo namen. (ii) Maria uuarliho gihielt allu thisu uuort ahtonti in ira herzen. (i)

‘While I was with them in the world, I kept them in thy Name.’ (AV John 17:12) (ii) ‘But Mary kept all these things, and pondered them in her heart.’ (AV Luke 2:19) The parallel use of GE- in these three examples appears to offer evidence for a consistent and systematic use of GE- in Gothic, Old English and Old High German, possibly to express perfective aspectual meaning (cf. also the comparative discussion of verbal prefixation and aspectuality in Old English and Old High German by Wischer & Habermann 2004). Nevertheless this example is less cogent than it might appear at first sight, since the Gothic translation renders Greek σύν-, and the Old English and the Old High German ones render Latin con-, and each of these prefixes is the literal translation of Gothic ga-, at least in its etymological spatial sense. Wedel’s examples of GE- formations which do not translate prefixed verbs (2001: 328) are more persuasive, although, as in similar postulations of aspect in Gothic, there is little discussion of a representative amount of individual examples which could serve to demonstrate the actual existence of such an aspect distinction. Leiss (1992: 64) concludes (with reference to Rice 1932) from the observation that there is no consistent one-to-one correspondence between the prefixes in Greek and those in the Gothic translations that the Gothic prefixes cannot be translations of the Greek ones and takes this as an argument in favour of regarding them as aspectualizers. But this argumentation does not appear to be entirely justified, since she does not take into account the fact that the prefix inventories of the two languages are quite different to start with, nor does she discuss the possibility of lexicalized prefix formations in Gothic which need not be parallel to the Greek verbs, or the possibility that the higher number of


171

prefix verbs in the Gothic translation might be a typical translation phenomenon (cf. e.g. the occurrence of loan formations in Old English translations from Latin). Leiss also does not mention Rice’s (1931: 122– 123) observation that ga- very consistently appears to translate desemanticized prefixes of Greek whose perfectivizing value is far from indisputable. Thus, in default of more systematic investigations, the issue cannot be regarded as settled, while the arguments in favour of systematically aspectualizing preverbs in Gothic run a strong risk of circularity. Following Lloyd (1979), Wedel (2001) claims that Old English ge- always expresses such aspectual contrasts, and indeed his analysis of gein Cynewulf’s Elene provides evidence that the prefix can be used to express aspectual contrasts (more specifically, what he calls ‘complexive aspect’; see also Wedel 1997). But his very examples show that this is in fact not systematically the case, just as he does not show how the use of gein examples (6)– (8) above is representative of systematic expression of aspectual contrasts in the respective languages. Wedel’s interpretations of the text in favour of this hypothesis are imaginative but not convincing. Cf. e.g. his analysis of hyrdon vs. gehyrdon (Wedel 2001: 208–209), where an aspectual difference is very hard to make out in two otherwise identical clauses Hwæt we ðæt (ge)hyrdon þurh halige bec except for an ‘aspectual’ contrast construed on the basis of the presumptive aspectual meaning of the prefix: the approach is circular. 5.2.3. Prefix variation in Old English With respect to Old English, it becomes clear through material collected by Hiltunen (1983a) that the assumption of systematically aspectualizing prefixes is untenable (see also e.g. Pilch 1955: 45–46, Sprockel 1973: 38 and Horgan 1980: 128; for an overview of the often contradictory views on the semantics of ge-, see Lindemann 1965). Hiltunen had carried out a comparison of the use of gein identical contexts in different manuscripts and had discovered considerable variation in the occurrence of the prefix, e.g.:3

3

Quoted after Hiltunen (1983a: 56). Hiltunen explores the use of a few prefixes in a small number of texts only; his results are therefore far from being exhaustive, but they appear to be sufficiently reliable and representative to be used here as an extended example of the semantic and functional overlaps observable in Old English. For more examples, see also Ogura (2002: Ch. V and App. II).

172


(11) (i) … þæt he him sylfum geagnað þa þenunge GD H (35.6) (ii) … þæt he agnað him sylfum þa þenunga GD C and O (35.5) (translating officium apostolici nostri domini sibimet usurpare indoctus praesumpsit) (12) (i) … he gespræc to him Mk L (4:33) (ii) … hiæ spreocað to him Mk R (4:33) (iii) … he spræc to him Mk C and H (4:33) (translating talibus multis parabolis loquebatur eis) (13) (i) & oft gehergode on Peohtas (ii) & oft hergode on Peohtas ‘and often made raids on the Picts’

ChronE 875 (75.1) ChronA 875 (74.1)

Hiltunen finds very little systematic variation among the manuscripts and none that can be shown to be connected to diachronic, dialectal or stylistic factors; there is hardly any evidence for a consistent preference or dispreference for the use of gein individual manuscripts. Hiltunen concludes: it is not often possible to determine the exact value of ge-. In many contexts there is no apparent reason for its presence or absence … In terms of verb forms, ge- is most commonly found with past tense forms, in particular the simple past. But, as we have seen, it is difficult to assess the significance of this in more precise terms. Beyond these tendencies the use of ge- seems haphazard. Some stylistic values may have been involved, but it is not possible to pin them down to anything very concrete either. (Hiltunen 1983a: 65)

Thus Horgan’s (1980) suggestion that prefix substitution in parallel texts may be due to stylistic factors cannot be confirmed either, see also the discussion above. Hiltunen (1983a: 62) also finds it difficult to confirm Samuels’ (1949) claim that already in Old English ge- is a clear past marker (although, it should be added, Samuels is concerned with the Lindisfarne glosses and what he assumes to be parallels to Old Norse there). That claim had already been disproved by Lindemann (1970), whose claims as to the lexical derivational function of the prefix cannot be confirmed with Hiltunen’s data either, though. The findings thus seem to support the model of the diachrony of gein Old and Middle English suggested by Pilch (1955), whom Hiltunen seems not to have consulted. The observation that the loss or substitution of native prefixes occurs first in the Northern dialects must be seen as a strong (and probably conclusive) argument against the possibility of French influence (via the borrowing of French verbs). More importantly, however, there is little evidence for an increase of the variable


173

use of the prefix within the Old English period. Together with the observation that towards Middle English GE- is increasingly used as a past participle marker (the function which is to become fully grammaticalized in Dutch and German), this can be regarded as conclusive proof that semantic factors cannot have caused the disappearance of this native prefix, as is often claimed in the literature (cf. Chapter 4 and Section 5.2.1 above). The other Old English prefixes can likewise be subject to prefix variation, although to lesser degrees (see below). The prefix which is most similar to gein this respect is a-, but other prefixes also vary in their occurrence among the parallel texts, e.g.: (14) (i) … þe heo fedde GD C and O (70.8) (ii) … þe heo afedde GD H (70.9) (translating ecce enim gallinas, quas nutrit, vulpis comedit) (15) (i) & bebead þæm l him (ii) & he him bead (translating et praecipebat eis) (16) (i) & Ælmar abb hi lætan aweg (ii) & Ælmar abbod hi farleton aweg ‘and they led away Abbod Ælmar’

Mk L and R (8:18) Mk C and H (8:18) ChronE 1011 (8.18) T-ChronF 1011 (267.16)4

Moreover, prefix variation is not confined to the presence versus absence of individual prefixes; as is to be expected from the examples above, there is a certain amount of interchangeability among the semantically weaker prefixes. Again this occurs mostly with ge- but also with the other semantically bleached prefixes, e.g.: (17) (i)

… het Eadweard cy[ni]ng atimbran þa norðran burg … ChronA 913 (96.19) (ii) … het Eadweard cing getimbran þa norðran burh … T-ChronB 913 (816.12) ‘… King E. had the fortress in the north built …’

4

Hiltunen takes the variant reading of MS F of the Anglo-Saxon Chronicle from Thorpe’s edition (i.e. T-ChronF in the list of short titles); Plummer’s reading is confirmed by the facsimile edition (ChronF in the list of short titles) now available (at 63v).

174


(18) (i) & swiðe bebead him (ii) & he him swyðe forbead (translating et vehementer cominabatur eis)

Mk L and R (3:12) Mk C and H (3:12)

However, this kind of variation is rare and restricted, in line with the general semantic tendencies of prefix variation: As a rule – we may assume – one prefix can be substituted for another only if the contents of the two are not contradictory; i.e. there must be at least a partial overlap between the items. Ge- poses few problems in this respect … its use did not change much from eOE [viz. early Old English] to lOE [viz. late Old English]. The same applies to a- … However, some signs of the incipient decline of ge- and a- may be seen even in the present material. One such sign is the tendency in GD (H) and Mk (CH) to use be-, for-, on-, or to-, instead of ge-, most probably because of their greater expressiveness. This does not mean that ge- and a- would always have been semantically empty, only weaker and less distinctive than the other prefixes. Thus, because semantic fading is more a feature of ge- and a- than of the other items, other explanations must be sought for their decline … Gradually, as the non-prefixed variants take over, the coexistence grows less and less permissible. (Hiltunen 1983a: 84)

Lindemann’s (1970) hypothesis that genuine prefixes are by definition interchangeable in Old English because of their abstract meaning was highly implausible even before the consideration of further empirical evidence. The available evidence, then, fully serves to disprove it (see also Hiltunen 1983a: 55). The scale of semantic independence of Old English prefixes suggested by Minkova (2008: 30; cf. the discussion above) is clearly reflected by the possibility of omission or alternation of the prefixes. At the same time, Hiltunen’s findings also provide further support for the viability of the reasons for the long-term weakening discussed in Section 5.2.1 above. But the (semantically) more independent prefixes tend to retain their etymological spatial meaning (cf. e.g. Bosworth–Toller s.vv. geond-, ofer-, þurh-, wiðer-, ymb-) and, moreover, these typically secondary (i.e. not common Germanic and thus younger) prefixes show considerably less developed aspectualizing functions, if at all. Brinton & Traugott (2005: 125) claim that the Old English particles are already aspectual while “the verbal prefixes … more consistently express concrete spatial meanings”. But this is historically unfounded (cf. also Table 5-2 above), and the scenario is chronologically unlikely and supported by dubious examples (e.g. be- glossed ‘round’ and to- glossed ‘motion towards’), which are not supported by Brinton’s (1988: 204–212) earlier discussion of prefix meanings in Old English (see also the discussion in


175

Section 5.2.4 below). Just like in the Gothic examples discussed at the beginning of the present section, in Old English the weakening process is marked with the older prefixes and fully ties in with the cross-linguistic properties of spatial preverbs (cf. Chapter 3 above). Wedel’s (2001) hypothesis that some Old English prefixes (and in particular ge-) systematically function as aspectualizers is, as has been shown so far, demonstrably wrong. But Wedel’s observations also point in a more appropriate direction: The prefix [ge-] was, thus, used to achieve aspectual contrast. And, indeed, in a narration not all events are of equal importance. By alternating – wherever possible – between the prefixed and the unprefixed verbal forms, Cynewulf was able to make his narration more vivid. He used the unprefixed forms for the background information and reserved the prefixed verbal forms for the major events. (Wedel 2001: 210)

Again, the assumption of aspect distinctions made here is circular (for a discussion of and further literature on the discourse-pragmatic approach to aspect which is here implicitly taken by Wedel, see Binnick 2001: 562). In the light of the Old English data investigated by Hiltunen (1983a), it had better be argued that the evidence discussed by Wedel may support a path of development running from spatial meaning to telicity to semantic emptiness, possibly – as the reference to ‘major events’ indicates – via boosting. This suggestion would tie in with Lenker’s (2008) discussion of Old English booster prefixes (for a full discussion of boosters, see Peters 1993; for more recent literature, see the references provided by Lenker 2008). Although Lenker is concerned with adjective boosters, her analysis can also be applied to the development of the English preverbs, especially since there are considerable etymological and functional parallels. Etymologically, prototypical boosters are spatial adverbs, and in fact the Old English inventories of verbal prefixes discussed so far and of adjective booster prefixes overlap considerably. As we shall see, these stages of development are also to be found with the particle verbs, with the difference that the younger preverbs there involved continue to be found in Modern English in all their successively developed meanings. The diachronically conditioned synchronic semantic layering of the particles can thus be taken as further evidence that the above sketched development of the older prefixes, of which in English only the later stages are attested, is essentially correct, which, of course, is the expected result in view of the cross-linguistic evidence.

176


5.2.4. Particle semantics in medieval English As has been shown so far, on the whole there is little reason to assume that any Germanic language has ever developed systematic aktionsart marking comparable to the one found in the Slavic languages; at least this cannot be established on the basis of the historical evidence (but see Haverling 2003 on the loss of aktionsart marking in Romance; on the analysis of Slavic ‘aspect’ as lexical derivational aktionsart marking, see Dahl 1985: 27 et passim). However, it is also remarkable that in all Germanic languages prefixes and particles are used to some extent as aspectualizers, which is a consequence of the specific spatial semantics of these prefixes and particles. In the light of the cross-linguistically widely attested considerable variation in the expression of aspectual distinctions, it is not entirely surprising that no Germanic language has ever developed systematic aktionsart marking. And despite well-documented tendencies towards the derivational expression of aspectual contrasts, there is certainly no need for such a system anyway, as a cross-linguistic comparison shows. Sasse (1991: 43) observes: “It is clear that there are aspectual types of states of affairs which are universal, but not universally lexicalized. Moreover, it is clear that there is a universal tendency to systematically express the conceptual distinction of situation and situation change, at least for certain types of states of affairs” (his emphasis); see also the more recent discussion in Sasse (2006). Moreover, systematic aktionsart marking may in fact break down once it has been established, as has been shown e.g. by Haverling (2003) with respect to Early and Classical Latin as opposed to the Romance daughter languages. However, this is not to say that synchronically the Old English prefixes and particles are functionally and semantically equivalent, as suggested by Brinton (1988). Despite their identical etymologies the Old English prefixes and particles are clearly at different stages in the development of preverbs, not just with respect to morphological fusion, but also, and perhaps more importantly, with respect to their meaning. Van Kemenade & Los (2003) justly criticize Brinton (1988) for arguing in favour of such a view, and they point out that in particular the frequent occurrence of prefix–particle doublings shows that “it looks as though the inseparable prefixes, phonologically weak as they are, are in the process of losing their distinctive meaning and cease to encode W in the R-LCS [viz. a secondary predicate in the resultative lexical conceptual structure]” (van Kemenade & Los 2003: 105). Moreover, there is little support for Brinton’s (1988: 185) claim based on the older literature, “[t]hough many of the modern postverbal particles are the etymological counterparts of the verbal prefixes, it seems clear that the system of post-verbal particles represents a new


177

development …, not a continuation in any direct way of the older system of prefixation”, which neglects the comparative evidence for the emergence of preverbs and the syntactic factors relevant in the English development discussed in Chapter 3 above, while the claims concerning the sudden loss of productivity of native prefixes by early Middle English are similarly unfounded (see Section 5.1.1 above). Although the greater degree of prosodic and semantic salience characteristic of the particles diachronically underlies the co-occurrence of prefixes and particles which is common in Old English (cf. also the data from Gothic discussed above), this co-occurrence synchronically only rarely results in semantic doubling (i.e. in pleonasm), although this has been frequently stated (or implied) in the literature (cf. e.g. Samuels 1972: 164, Hiltunen 1983a: 98–99, Denison 1985: 44–48, Brinton 1988: 215–216 et passim, Claridge 2000: 87, van Kemenade & Los 2003: 101–103 and Elenbaas 2007: 114–115 and 146). Similarly, Leiss (1992: 71 and 257–258) has argued that the doubling of the aspectualizing function of verbal prefixes and verbo-nominal combinations (‘Funktionsverbgefüge’) in Modern German correlates with the long-term weakening of the aspectualizing function of the prefixes (see also Leiss 2002a and b; but see van Pottelberge 2001: 223–230 for a critique of Leiss’ argumentation and for further literature). It is remarkable that the recurrent stock examples in the literature are restricted to a small set of prefixes and particles. What is more, quite a few of them only exemplify the possibility of co-occurrence of prefixes and particles, but hardly the intensification of the prefix meaning through the particle. Some of these doubtful examples may look conclusive, but at closer examination such certainties tend to evaporate. To take one example, Brinton (1988: 217) claims that the particle off “often reinforces a weakened ā- verbal prefix”. At first glance the evidence she provides seems valid: although Old English a- is etymologically not quite identical to off, both the considerable semantic overlap and the meaning of at least some a-prefixations do not preclude that possibility. This seems to receive further support from the example she gives: (19) Gif man cealf of adrife … ‘If someone drives off a calf …’ (Ancient Laws (Thorpe) i, 72, 1) However, verbs with the prefix a- also co-occur with other particles, e.g. (examples from the DOE):

178


(20) Aaron ahæfde his hand upp on gebedum ‘A. raised his hand up in prayers.’ (ÆLS (Pr Moses) 26) (21) ða clypodon hi ealle, kyrrieleyson, up ahafenum handum wið heofonas weard ‘Then they all cried, Kyrie Eleison, with their hands raised up towards heaven.’ (ÆLS (Basil) 450) (22) He adune astah ‘He went down.’ (PsGlG (Rosier) 71.6) (23) & ðæt gode mod ðe sio hælo ful oft aweg adriefð ðæt gemynd ðære medtrymnesse geedniwað ‘and the good mood, which the health very often drives away, the memory of sickness restores’ (CP 36.255.16) (24) & he him anweald sealde untrumnessa to hælanne, & deofolseocnessa ut to adrifanne ‘and he gave him the power to heal sickness, and to drive out the devil-sickness’ (Mk 3.15) Conversely, these particles also co-occur with other prefixes, e.g.: (25) & Bryttas him wið gefuhton. & hæfdon sige. & hi bedrifon ut on ane ea. & manige adrencton ‘and the Britons fought with them, and won, and they drove them out on a water, and many drowned’ (ChronE 890.5) Although Brinton (1988: 217–223) also uses some of these examples, she does not remark on the contradictory nature of her evidence. Surprisingly, the extent, spread and finer details of this use have never been investigated in detail, despite the fact that it is mentioned in most studies of the Old English verb-particle construction. Such examples show that the cooccurrence of prefixes and particles can hardly be explained satisfactorily by pointing out that the ‘weak’ prefixes are semantically ‘reinforced’ (cf. Brinton 1988: 221) by the particles, unless one argues that a- means ‘up’, ‘off’, ‘away’, ‘out’, ‘down’, etc. simultaneously, while mutatis mutandis


179

the same is true of be- and ge-, etc. But all the studies mentioned above in fact postulate several or even all of these senses for a- (and similarly for the other desemanticized prefixes). Plainly, if all these prefixes can mean anything, there is nothing to reinforce to start with. Only those prefixes seem to participate in this pattern of prefix/particle doubling which are largely desemanticized anyway and which not only tend to be mutually interchangeable, but also completely omissible (cf. Bosworth–Toller or, where applicable, the DOE, s.vv. the respective simple vs. prefixed verbs; on the little investigated topic of double prefixation cf. also Dietz 2004: 584–586). Kastovsky (1992) remarks about the meaning of a-: In view of the vagueness of the meaning of the prefix, which only occurs with verbs or deverbal adjectives, it is difficult to give precise semantic patterns, and in many instances it does not seem to have added anything to the meaning of the stem, cf. abacan/bacan ‘bake’, abarian/barian ‘lay bare’, aberan/beran ‘bear’. In some instances it seems to denote ‘out’, e.g. aberstan ‘burst out’, abrædan ‘spread out’, acleopian ‘call out’; in others, it seems to add an intensifying or completive element, e.g. abeatan ‘beat to pieces’, acalan ‘become frost-bitten’, adrygan ‘dry up’. But in the overwhelming majority of instances, its meaning is no longer transparent. (Kastovsky 1992: 378)

One must therefore conclude that the reinforcement of earlier but semantically weakened prefixes by particles is unlikely to have been a factor of major relevance in the development of the verb-particle construction, nor in the co-occurrence of both prefixes and particles. However, what both the older prefixes and the younger particles share is their propensity for developing into aspectualizers, which is ultimately based on their shared spatial etymologies. This topic has been discussed in some detail by Brinton (1988), who argues at length that the development is essentially based on metonymical processes: The prefixes and particles occur first in contexts in which spatial meanings are primary, with verbs of motion or of physical action. In many of these contexts, though, both spatial and aspectual meanings are possible. This is the ‘locus for change’ in focus from spatial to non-spatial meaning … Once such a change in meaning from spatial directionality to situational boundedness has taken place, the prefixes and particles can then occur freely in combinations in which spatial meanings are impossible … [T]he shift described here is metonymic, not metaphoric. The particles do not assume figurative value, nor does the combination of verb and particle effect some figurative shift. There seems to be no indication that the prefixed or phrasal verb is used metaphorically any more often than simple verbs. (Brinton 1988: 197–198)

180


For an overview and critique of the traditional views of the process as ‘fading’, figurative extension and metaphorization, see Brinton (1988: 191– 193); for a useful survey of the older scholarly opinions concerning the aspectual nature of the particles, see also Appendix B in Brinton (1988). But it seems somewhat unfortunate that Brinton (1988: 58) argues that the aspectualizing function of the particles is “fully established in their earliest attestations”, and it has been justly pointed out by Traugott (1991: 223) that this claim is “close to circularity without further analysis”. In fact, there is very little support in favour of Brinton’s panchronic conflation of spatial and aspectual meaning in prefixes and particles, and the particles are very clearly first and foremost spatial in Old English (see Diensberg 1990: 192– 193). Traugott (1991: 225), moreover, argues that metonymy is only a secondary factor in the emergence of aspectualizing particles. Rather, Traugott states, “the metonymy results from … linguistic inference arising out of the use of spatial terms rather than from the abstract linguistic organization”; for a fuller discussion of the role of pragmatic inferences in semantic change, see Traugott & Dasher (2002). Likewise, Brinton’s (1988: 225) contention that postverbal position of the particle is a prerequisite for the emergence of aktionsart meaning is unfounded; cf. Allen’s (1990: 248) critique of this assumption: “It is difficult to understand why post-verbal position should be seen as a syntactic peculiarity of aktionsart markers, and Brinton has not established that the post-object position as a possibility for the aktionsart particles post-dates their first usage with such meanings”. In fact, Brinton does not take into account the syntactic factors involved in the loss of preverbal particle position, while her hypothesis concerning the connection between serialization and semantics rather undermines her contention that the prefixes and the particles are functionally equivalent. The path of change from spatial to aspectual meaning outlined here has been confirmed in numerous cross-linguistic studies, cf. e.g. by Bybee & Dahl (1989): To be located spatially in an activity is also to be located temporally in an activity, so that from the beginning the meaning of such constructions has temporal implications. Gradually the locative meaning weakens while the temporal implications stabilize, giving rise to the aspectual meaning. (Bybee & Dahl 1989: 81)

In fact, the historical development from spatial to more abstract meaning is probably the single most consistent finding of typologically oriented research into the emergence of semantic generalizations; see e.g. Bybee & Pagliuca (1985) on the role of semantic generalizations in grammaticaliza-


181

tion processes, and Bybee, Perkins & Pagliuca (1994: 6) on alternative terminologies, including ‘bleaching’, a term which had perhaps better be avoided if the emergence of new albeit more abstract meaning is focused on. For discussions of the universal path from spatial to temporal, aspectual and other types of more abstract meaning, see e.g. Traugott (1978), Heine, Claudi & Hünnemeyer (1991), Traugott & König (1991) and the overview in Hopper & Traugott (2003). A short overview of the relevant cognitive processes is provided in Diewald (1997: Ch. 3); for cross-linguistic attestation see in particular Heine & Kuteva (2002). In contrast to the native prefixes, where much of the syntactic and semantic development is only accessible indirectly through the comparative method, the history of the particles can be observed in the historical corpus; see e.g. Brinton (1988: 204–214) for examples of the less desemanticized prefixes used both with literal and aspectual meaning. The examples containing particles adduced so far easily serve to show that the typical meaning of the Old English particles is spatial, and the typical meaning of the Old English verb-particle construction is transparent. Brinton & Traugott’s (2005: 124–125) claim that already in Old English the particles must have been grammaticalized as markers of aspect is therefore somewhat misleading, considering that the vast majority of particle tokens in Old English will have been spatial (while the associated claim that the prefixes express spatial meanings, is clearly wrong, see Section 5.2.3 above). But the critical contexts for the metonymical shift from spatial to aspectual meaning are certainly present in the Old English verb-particle constructions, cf. the following examples, which are arranged in a sequence from fully literal to fully aspectual:5 (26) Hi eodon þa ut to þam inran gate. ‘They then went out to the inner gate.’ (ÆLS (Martin) 1158) (27) þa wearp se broðor þæt glæsene fæt ut. ‘then the brother cast out the glass barrel’ (ÆCHom II, 11 104.25)

5

Quoted from Brinton (1988: 222–223), with slightly different translations supplied (cf. also Diensberg 1990 and Reichl 1995); see Brinton (1988: 216– 233) for a more detailed discussion and more examples with out and other particles.

182


(28) deofolseocnessa ut to adrifanne ‘to drive out the devil-sickness’ (Mk 3.15) (29) & lett agan ut hu fela hundred hyda wæron innon þære scire ‘and let them find out how many hundreds of hides were in the shire’ (ChronE 1085.26) This development of out is further continued in Middle English, where the particle assumes a mostly aspectualizing function already noted by Denison (1985: 56–57): “the semantic development of out is a straightforward matter either of metonymy from a resultative spatial sense or the metaphorical use of a particular collocation”; cf. MED s.v. out(e), e.g.: (30) Marrch wass þa Neh all gan ut till ende. (?c1200 Orm.(Jun 1) 1892) (31) Evere comeþ out uvel sponnen wolle. (?a1300(c1250) Prov.Hend.(Dgb 86) st.37) (32) I wol breke out fram þat baret. (a1375 WPal.(KC 13) 486) (33) Now, lorde, lach out my lyf, hit lastes to longe. (c1400(?c1380) Patience (Nero A.10) 425) (34) Þe deuyll putyth owth þe fyre of charyte. (a1500 GRom.(Glo 22) 779/13) Cf. also De Smet’s (2010) discussion of the history of out and the consequences of its collocational and semantic overlap with forth, where it is argued that inter-particle interference may also play a role in the development of individual particles. Los (2004) regards forth and on as grammaticalized secondary predicates and argues that their path of development is from resultative to durative to event-modifying meaning. For a discussion of the history of complex predicates in Dutch, see Blom & Booij (2003). The threefold semantic categorization of verb-particle constructions in present-day English discussed in Chapter 2 above can thus be graphically represented as in Figure 5-1 as the result of two historical processes: first, the metonymical evolution of aspectualizing meaning observable in the particles, and second, the further lexicalization of individual complex words (cf. C. Lehmann 1989 and Blank 2001), with accompanying increase in semantic non-compositionality.


183

verb-particle combination

compositional

directional particle

non-compositional

aspectual particle

Figure 5-1. Semantic development of verb-particle constructions

The second of these two processes has not been subject to a detailed historical investigation, despite the discussions of the lexicalization of some Old and Middle English particle verbs in de la Cruz (1972b) and in Hiltunen (1983a: 192–220). For a general account of the development of phrasal verbs and other types of complex predicates in the history of English, see Traugott (1999). But there can be little doubt that the generalizations based on a critical survey of the literature which have been offered in the present section would benefit from more detailed analyses of a large number of verbs and particles. Like the present study, earlier studies have been content with providing discussions of individual particles and verbs; the OED offers good histories of individual phrasal verbs (cf. Chapter 4 above), and it seems a promising project to use the information contained there and in the other historical dictionaries for a much more comprehensive lexicalhistorical survey (cf. the discussion in Matsumoto 2008: 125–141). Moreover, a synchronic-diachronic study of the type undertaken for German particle verbs by Habermann (1994) could serve to shed more light on the overall development of different subtypes of aspectual meaning. But it seems that a prerequisite for studies of this kind in English would be the abandonment of the preoccupation in English studies with the mutual relationship of phrasal verbs and Romance loans (see Section 5.3 below). 5.2.5. Some conclusions In the light of the foregoing discussion, it can be explained in principle why the Old English prefix verbs are very often best translated with a Modern English phrasal verb. As we have seen, this cannot simply be taken as an

184


indication that the phrasal verbs ousted the prefix verbs at some point in the history of English; the native prefixes successively ceased being productive, while the spatial particles developed aspectual functions. But the relationship between these two developments is clearly a contiguous one, rather than a causal one. Talmy (e.g. 1985, 1991 and 2000) has classified English, like the other Germanic languages, as a satellite-framed language, which prototypically uses directional particles to encode the motion-path (as opposed to e.g. the modern Romance languages). It is conceivable that the enormous productivity of the verb-particle construction type in the Germanic languages (and the cross-linguistically well-attested emergence of aspectualizing functions out of path-expressing satellites) is connected to this typological trait of the language. More recently, however, Beavers, Levin & Tham (2010: 369) have argued that “Talmy’s typology results from numerous converging factors, including the overlap of path/manner (or rather manner/result) encoding in the verb’s obligatoriness, and the independent availability of various means of encoding manners and paths, combined with certain preferences for certain non-verbal encoding possibilities” – in other words, that the phenomenon is ultimately epiphenomenal (cf. also Slobin 2006). Nevertheless the approach may yield interesting results for the historical analysis of English, especially with respect to the borrowing and integration of Romance verbs after the Norman Conquest and the pleonastic particles discussed below. With regard to the possibility of Scandinavian influence, there is no reason for the assumption that language contact may have had any impact on the development of the verb-particle construction in English. It has also been suggested that the development may have been influenced by language contact with Celtic or French, but these suggestions are generally (and justly) regarded as unfounded (cf. the discussions in Diensberg 1983 and by Veselinović 2006). In his methodological discussion of language contact as an explanation of language change, Lass (1997) demonstrates that “any attempt to show extensive contact influence must be bolstered by examples that could not have happened any other way” (Lass 1997: 207). This requirement is not fulfilled by either the peculiarities of the contact situation between speakers of Old Norse and of Anglo-Saxon, nor by the mere existence of particle verbs in Old Norse (Thim 2008a). The suggestive references to North Germanic parallels which are so typically encountered in the literature (cf. e.g. the speculations in Kastovsky 1992, Burnley 1992 and Horobin & Smith 2002, which were mentioned in Chapter 4 above; for full bibliographical references, see Thim 2008a) only serve to show that verb-particle constructions are a common phenomenon in all Germanic languages, while there is nothing in that development in English


185

which cannot be explained as language-internal. Lutz (1997: 262, fn. 8) argues that Scandinavian influence has been “effective in a more global manner, namely by way of accelerated reduction of unaccented syllables (including inflectional endings) in the Scandinavianized Northern dialects of English, which must have resulted in a likewise accelerated fixing of word order in those dialects”. As far as the reduction of unaccented syllables goes, this account ties in nicely with the dialectal evidence and it serves to explain the consistently earlier phonotactic attrition of the unstressed prefixes in the Northern dialects (cf. the discussion in Section 5.2.1 above). With regard to the fixing of word order, however, the link between the loss of inflectional morphology and fixed word order, although plausible at first sight, has so far not been demonstrated conclusively, and there is considerable comparative evidence that the two developments are not necessarily connected (cf. e.g. the discussion in Allen 2006 and the references there). On the whole it is not clear which of the properties of the English particle verbs should actually derive from language contact, since each of them can be explained as the result of long-term language-internal developments characteristic of all Germanic languages, as in very many, perhaps even all, other languages with comparable preverbs (cf. the discussion in Chapter 3 above). 5.3.

An outlook to Modern English

The essential characteristics of the Modern English ‘phrasal verb’ can be regarded as fully developed by late Middle English. But as is well-known, the Middle English period is also characterized by an unprecedented influx of Romance loans, which has traditionally always been seen in connection to the development of the phrasal verbs, both in regard to their internal etymological structure and in regard to their overall stylistic role in the lexicon. 5.3.1. Etymology and integration As has already been noted, descriptions of the Modern English phrasal verb rarely fail to make the observation that one of its characteristic features is its typically Germanic (West Germanic or Old Norse) etymology. But as with many other statements in this context, attempts at actually substantiating it have been rare.

186


Thus, what can we actually say about the etymologies of Modern English phrasal verbs? Claridge (2000: 116) counts 326 different simple verb elements (types) among the phrasal verbs in the Lampeter Corpus (1640– 1740). In 62 per cent of these the simple verbs are of Germanic descent and in 32 per cent of ‘Romance’ (i.e. Latin or French) descent (see also the list of phrasal verbs supplied in Claridge 2000: App. I). The remaining verbs (18 types) are of unclear etymology according to the OED, which Claridge uses to establish the etymological origins (although the etymologies in the OED are now rather dated and certainly not reliable in all details, for the distinction between native and Latin- or French-derived verbs the information provided there can be regarded as sufficiently reliable). Claridge (2000: 116) concludes that “there is hardly anything unexpected to be found in that area” in her corpus, but the very observation that about one third of all phrasal verb types in the corpus do not conform to the stereotypical characterizations of phrasal verbs as etymologically ‘Germanic’ deserves some attention. Conversely, one might also wonder why there are more phrasal verbs with Germanic etymologies than phrasal verbs with Romance etymologies. But this question is virtually absent from the literature, so that both the reasons for the prevalence of Germanic verb types and those for the existence of the non-Germanic ones (in this case, as many as one third of the verb types) are left unexplained. But in the light of the discussion both in this and in the preceding chapters, it would seem odd if borrowed simple verbs did not participate in the formation of phrasal verbs, since the formation of verb-particle constructions has always been a productive process in English. Although there are some restrictions on morphological productivity according to etymological provenance in English (see e.g. Marchand 1969, Bauer 1983, Nevalainen 1999 and Schmid 2005), such restrictions are unlikely to be fully operative in the case at hand, since borrowed verbs participate rather frequently in the formation of particle verbs. This is still true in present-day English. In a spot-check of phrasal verbs in Cowie & Mackin (1975), Lutz (1997: 286– 287) presents clear evidence for the etymologically mixed character of Modern English phrasal verbs (bold type marks non-Germanic items, the provenance of non-Anglo-Saxon items is given in brackets, with F = French, L = Latin, ON = Old Norse): by down off out

come ~, drop ~, get ~ (ON), lay ~, pass ~ (F) break ~, close ~ (F), come ~, get ~ (ON), go ~, grind ~, jot ~ (L/G), keep ~, pass ~ (F), put ~, set ~, shut ~ feed ~, freeze ~, go ~, keep ~, lay ~, pay ~ (F), play ~, tail ~, take ~ (ON), marry ~ (F), pass ~ (F) beat ~, carry ~ (F), check ~ (F), clean ~, fade ~ (F), leave ~, move


up about across (F) apart (F) away forward over

187

~ (F), pass ~ (F), point ~ (F), rule ~ (F), work ~ act (L), bolster ~ , book ~, clear ~ (F), dress ~ (F), dry ~, foul ~, get ~ (ON), keep ~, open ~, pass ~ (F), set ~, bring ~, carry ~ (F), go ~, play ~, push ~ (F), set ~ come ~ , file ~ (F), get ~ (ON), put ~ , press ~ (F) fall ~, pull ~ (F), put ~, set ~, take ~ (ON), tell ~ back ~, carry ~ (F), die ~ (ON), drain ~, fade ~ (F), keep ~, pass ~ (F), tear ~, wear ~ bring ~, carry (F), look ~, push ~ (F), put ~, step ~ boil ~ (F), gloss ~ (L), leave ~, move ~ (F), pass ~ (F), take ~ (ON)

Clearly, non-Germanic simple verbs may combine easily with the Germanic particles, but also vice versa, as the particles across and apart of French descent show. Incidentally, this is also true to a large extent of present-day English prefixes and verbs. Although there is a certain tendency for etymologically consistent prefixes and bases, it appears more likely, Lutz argues, that combinability is ultimately driven by stylistic considerations which subsume the remaining native prefixes like be- and mis- together with the larger number of borrowed ones (cf. Table 5-3 above) as marked for formality (cf. Lutz 1997: 285–286). Moreover, the non-Germanic elements appear to be typically borrowed from French. As de la Cruz (1972c: 22–27) shows, such hybrid formations occur already in early Middle English, and Diensberg (1983: 255) takes this as evidence for the lexical integration of such borrowings and the general productivity of the construction. Lutz therefore concludes: It is true that the lexical influence of French (and of Latin) was enormous, not only on the more formal registers but also upon colloquial English … As regards the verbs, this lexical influence makes itself felt in an impressive number of straightforward Romance loans, Romance formations and etymologically mixed verbs in all registers of English. (Lutz 1997: 287)

It thus seems clear that as a construction type the phrasal verbs are in principle insensitive to the etymological provenance of their elements as long as these elements are either inherited from (West) Germanic or borrowed from Old Norse and in particular from French. The fact that the occurrence of verbs borrowed from Old Norse in verb-particle constructions is by no means particularly high furnishes another, if only marginal, argument against the assumption of Scandinavian influence on the development of the construction (see also Thim 2008a for a fuller discussion of fallacious examples containing Norse-derived simple verbs in support of that assumption).

188


But what about verbs borrowed from Latin? As will be discussed in more detail in the next chapter, the Latin borrowings in English can to a large extent be described as abstract and ideational. In other words, verbs borrowed from Latin are hardly ever among the “verbs of motion or of physical action” identified by Brinton (1988: 197) as the typical input to the formation of verb-particle constructions (cf. Section 5.2.4 above). Hence the failure of so many Latinate verbs to combine with spatial particles is only to be expected, just like the suitability of semantically rather untypical Latin borrowings to occur in verb-particle constructions nevertheless, cf. e.g. act in the list illustrating the etymologically mixed character of the Modern English phrasal verbs provided above. With regard to the relationship of Germanic- vs. French-derived simple verbs as part of verb-particle constructions, it would be worthwhile to explore whether their proportion corresponds to the overall proportion of native verbs vs. French borrowings in the English lexicon. In his assessment of the etymological origins of the lexicon of Modern Standard English, Scheler (1977) finds a consistently higher proportion of French borrowings for the overall lexicon than the one third noted by Claridge (2000) and Thim (2006b; cf. the discussion in Section 5.3.2 below) for late Middle English to Late Modern English phrasal verbs. But Scheler does not distinguish between different word classes, and in fact in comparison to other word classes the number of verbs which were borrowed from French is actually disproportionally low (see Dekeyser 1986). The major factor determining the appearance of a simple verb in a verb-particle construction is its meaning, and only indirectly its etymological provenance. Moreover, very many of these borrowed verbs are prefix verbs, with the spatial or aspectual meaning associated with the English particles already expressed by the prefix. Cf. e.g. Serjeantson’s (1935) examples of Latin verbs borrowed in Middle English: accede, adjure, admit, combine, commend, commit, conclude, confide, discuss, dissent, distend, exclude, expend, immix, import, infect, interest (Serjeantson 1935: 260–261)

And cf. also the list of verbs borrowed from Latin in Early Modern English which were to become permanent additions to the English lexicon in Baugh & Cable (1993): adapt, alienate, assassinate, benefit, consolidate, disregard, emancipate, eradicate, erupt, excavate, exert, exhilarate, exist, extinguish, harass, meditate (Baugh & Cable 1993: § 161)


189

As these examples show, adding a particle would in most instances be likely to result in some kind of pleonasm. This is most obvious in the case of some of the borrowed prefix verbs, e.g. combine, exclude, eradicate, etc., in the lists above, where verb-particle constructions are possible but tend to be avoided (e.g. ?combine together, ?exclude out, ?eradicate out), while with others spatial or aspectual semantics typically added by the particles appear to be semantically incongruent with the verb meaning. These last points become particularly obvious in a contrastive analysis of English verbs borrowed from (late medieval) Latin and their German equivalents. Lutz (2008: 10) has drawn up a list of such verbs in order to illustrate the different degrees of integration of Latin borrowings in the two languages (borrowings are in italics, German prefixes and particles are underlined): distract frustrate include interrupt magnify prevent promote prosecute reject remit reprehend subdivide subjugate submit subscribe substitute supplicate suppress testify

ablenken vereiteln, durchkreuzen, frustrieren einschließen, umfassen unterbrechen vergrößern verhindern fördern, unterstützen verfolgen ablehnen, verwerfen vergeben, erlassen, aufschieben tadeln, kritisieren unterteilen unterjochen, unterwerfen unterwerfen, unterbreiten, vortragen (unter)zeichnen, beitragen, subskribieren einsetzen, ersetzen, substituieren anflehen unterdrücken, niederschlagen, abstellen (be)zeugen, attestieren

As this list shows, in German very systematically native prefix and particle verbs are used where English has borrowed from Latin. The fact that some of these verbs are, in fact, loan formations (on Latin models) is not only relevant to the evaluation of the role of borrowing in English and German, as Lutz (2008: 11) points out, but it also shows that transfer and integration of the loans depend on bilingual individuals who can analyse the Latin verbs as semantically equivalent to complex Germanic ones (for the typological backgrounds to this difference between Germanic and Romance languages, cf. the brief discussion in Section 5.2.5 above).

190


It is likely that a number of factors contribute both to the general avoidance of pleonastic particles and to their occasional occurrence. The general explanation sketched so far presumably accounts for the vast majority of instances, or perhaps rather for the fact why most verbs borrowed from Latin are indeed unlikely candidates for verb-particle constructions, unless pleonastic use of the particle is tolerated. But why pleonastic particles were sometimes used with these verbs is not easy to tell, and explanations so far, although not necessarily wrong, fail to be fully satisfactory. C. Lehmann (2005: 130) hypothesizes: “Yet other examples evince a desire to equip a naked noun, verb or adjective with a companion so that it need not stand alone. The word alone seems too weak”. But although one half of his examples (past experience, resulting effect, unexpected surprise, return back, sink down, fall down, repeat again, fly through the air) contain phrasal verbs, the explanation C. Lehmann (2005: 131) offers is not completely sufficient: “In some cases, this horror vacui may be motivated purely phonologically, by reasons of rhythmic euphony … This variety [of pleonasm] may be called phatic pleonasm. The modifier is not stressed and in most cases cannot even be stressed because there is no possible contrast”. But this does not apply to the phrasal verbs he adduces as examples, and, more importantly, it does not explain why this kind of pleonasm is clearly much more of an exception than the rule (while the factors conditioning the exceptions remain rather vague). Cf. C. Lehmann (2005: 119– 123) for a terminological discussion of various types of linguistic redundancy, where it is argued that pleonasm should not be confused with superfluity or uselessness, since it fulfils a number of diverse communicative functions; cf. also Traugott & König (1991: 191): “The fundamental process we see at work is a principle of informativeness or relevance, essentially the principle: Be as informative as possible, given the needs of the situation”. Lehmann’s observation that the phenomenon ought to be seen as a particular type of linguistic redundancy, which is characteristically associated with borrowing, seems to point in an appropriate direction in the present context: Borrowing an item presupposes some degree of knowledge of the donor language, and the item is borrowed precisely for its properties. The most one can say is that the speaker wants to make sure, within the means of the code he is currently using, that the item has the properties needed in the discourse. The most transparent way of guaranteeing this is the application of a productive operation whose operator confers just the desired property … Safety pleonasm evinces a basic insecurity in the control of the code. Since none of us is the master of the norm, we do not have full certainty of the meaning of a word and the service it can do in our speech. Therefore we prefer to play it safe and to


191

combine it with another sign which should also contribute the desired meaning and of which we feel a little more sure. (C. Lehmann 2005: 148–149)

This general characterization, then, helps to shed some light on the pleonastic use of particles, and in particular the use of particles with borrowed verbs, whose exact meaning is more likely to have been connected to the “basic insecurity in the control of the code” Lehmann refers to. In Margaret Paston’s Letters (written between 1441 and 1478), five per cent of all phrasal verbs are pleonastic (see Thim 2011a for details). Already MüllerSchotte (1955: 365–366) had discussed the use of pleonastic particles in texts from throughout the Modern English period and also pointed out remarkable stylistic parallels to the German Chancery style. In a corpus of 18th-century letters, Denison (2007: 124) notes the following examples of pleonastic phrasal verbs: adjourn out, declare off, inquire out, repair up, return back. These are all impossible (or at least highly doubtful) in present-day English, and Denison speculates that this is to do with a phonological constraint which is operative in present-day English but weaker in 18th-century English; namely that the typical simple verbs in verb-particle constructions are stressed on the first syllable. But what is visible in present-day Standard English is the result of the normative dislike of pleonasm (cf. the discussion in the next chapter). This is much more likely to be heeded by speakers using Latinate verbs, which, for the reasons outlined above, are also much more likely to give rise to pleonastic combinations. The occurrence of such pleonastic particles in informal 18thcentury letters is likely to be the result of the communicative strategies outlined here. Each of Denison’s examples contains a borrowed verb with a prefix-like element at the beginning, and his argumentation is not entirely convincing. He adduces piss off / cheese off / put off vs. annoy off as arguments in favour of the constraint, but does not discuss the fact that his annoy and annoy off are synonymous and therefore stylistically incongruent, whilst piss and piss off are not. The precise nature of the factors at work in the selection of pleonastic particles remain to be explored. I agree with Lehmann’s conclusion: Older textbooks teach that language changes because we adapt it to new needs. Younger textbooks teach that it changes because the language acquisition device comes up with an original analysis of the input. Pleonasm and hypercharacterization confirm what Coseriu … said long ago …: Language changes because we create it every day. We have to do so to the extent there is no ready-made language that we could rely upon. (C. Lehmann 2005: 149–150)

192


Nevertheless a more precise analysis of the factors governing the occurrence of pleonastic forms in historical texts seems desirable. For presentday English, Hampe (2002) has demonstrated convincingly that apparently ‘redundant’ particles may play an important role from a cognitive point of view, which tended to be overlooked by earlier discussions: redundancy in verb-particle constructions must be regarded as a form of conceptual overlap. In contrast to the truth-based notion of redundancy, conceptual overlap is a gradable concept. More specifically, the particles refer to elements which are already parts of the conceptual bases of the respective simple verbs … Rather than inserting completely new conceptual material, the particles profile material already (to some degree) activated. (Hampe 2002: 246; her italics)

The history of this type, which Hampe calls ‘superlative verbs’, remains to be studied in detail, though. 5.3.2. Phrasal verbs in 15th- and 16th-century English By way of concluding this chapter let us have a brief look at the use of phrasal verbs in a sample of 15th- and 16th-century English texts, i.e. texts written at the beginning of the Modern English period (based on Thim 2006b). An electronic full-text search for phrasal verbs with the particles back, down, forth, out and up of the 15th- and 16th-century letters contained in the Corpus of Early English Correspondence (1,026,247 words altogether; for a description of the corpus, see Nevalainen & RaumolinBrunberg 1996 and 2003 and Nurmi 1998) can serve to confirm the general observations on the percentage of French borrowings in verb-particle constructions. Cf. Figure 5-2, which refers to the type count of the simple verbs combining with the five particles. With regard to token count, however, the situation changes dramatically, and a mere five per cent of French-derived simple verbs remain, while the native simple verbs make up almost 95 per cent of all occurrences. For a more detailed discussion of the data, see Thim (2006b) and cf. also the discussion of some of the quantitative results obtained in that study in Chapter 6 below. Likewise, in the smaller corpus of Margaret Paston’s letters of the 15th century, there are 81 per cent Germanic and 19 per cent French types, while the French tokens make up 9 per cent of the corpus (cf. Thim 2011a: 362). Such figures are a consequence of the fact that a very small number of simple verb types is involved in the majority of phrasal verb tokens in the corpus (cf. Thim 2006b: 216).


193

28% French

1% Germanic

Latin 71%

Figure 5-2. Etymologies of phrasal verbs (types) in English letters, 1450–1600 (according to Thim 2006b)

All these simple verbs (viz. come, send, set, fall, take, bring, go, lay, make, put, give and break) in Margaret Paston’s letters clearly belong to the core inventory of the English lexicon and are of Germanic descent. Hence it becomes clear that the statements concerning the typical etymological structure of phrasal verbs need to be expressed more precisely than usual. With regard to the token frequency of occurrence, it is certainly true that the vast majority of phrasal verbs contain simple verbs of Germanic descent, while with regard to the types, French-derived simple verbs make up almost one third of all phrasal verbs (cf. also the almost identical results obtained by Claridge 2000 for a later corpus discussed in Section 5.3.1 above). In this context, it is also remarkable that all of the 159 simple verbs occurring as parts of these phrasal verbs are still current in present-day English, while each of them appears to be well-integrated in Early Modern English. This was confirmed by consultation of the OED s.vv. the simple verbs (for a full list of these verbs, cf. Thim 2006b: Appendix). Hence it must be concluded that Germanic simple verbs represent the core inventory of phrasal verbs in the corpus, and while not all Germanic simple verbs are highly productive, all the highly productive ones are Germanic. Moreover, pleonastic particles, although altogether quite rare, were typically found with the borrowed verbs, although not exclusively, e.g. (the abbreviations used for the identification of the sources are those used in the Corpus of Early English Correspondence; cf. the references given above): (35) neuer ceasing to lift up my handes and hart with deuout [prayers] for your most prosperous safe and sure succes in this voiage (RO1 1594 ELIZABETH1 110)

194


(36) Wherfore his Grace thought hit best that my Lord Steward shold advaunce forth and bryng his hole army as nere to gether as he myght (A 1522 T TMORE 260) (37) he showlde rather retorne back againe then to continue theare (CA 1594 T WCECIL 155) While (35) contains the native verb lift in combination with a pleonastic directional up, (36) and (37) provide evidence for the more typical case of borrowed simple verbs in combination with a pleonastic particle expressing a directional orientation already present in the semantics of the simple verb. Perhaps more importantly, though, the syntactic and semantic characteristics of the phrasal verbs in present-day English as described in Chapter 2 above are fully developed in this corpus of late Middle English/Early Modern English texts. It thus turns out that from late Middle English onwards, phrasal verbs as a construction type have remained almost entirely unchanged. This is also why the data from Thim (2006b) need not be discussed here further, since it would be rather pointless to amass examples which are identical to present-day verb-particle combinations, while there simply are no significant differences (cf. also Thim 2011a, where all phrasal verbs found in the 15th-century corpus of Margaret Paston’s letters are quoted in context). This, however, does not apply to the development of individual phrasal verbs; thus Denison (1998: 223) quotes the following examples of phrasal verbs from 19th-century literature which are now obsolete: (38) What’s going forward?6 (39) You have been bred up in the country. (40) … and shrugging up his shoulders … (41) And, oh, have you mended up all the open pens in the study? (42) … hastily checking herself up … Cf. also the 18th-century examples discussed in Section 5.3.1 above, and cf. Nevalainen (1999: 424) for examples of Early Modern English phrasal verbs which have ceased to exist in present-day English. Thus evidence for 6

This does exist in present-day English, though not in the sense ‘What’s going on?’ as in this example (and cf. also adverbial going forward ‘in the future’).


195

a constant turnover of the inventory is easily come by, as opposed to the generally accepted view of a constant increase in phrasal verb types from late Middle English onwards (cf. the following Chapter 6 for a more detailed and decidedly sceptical discussion of any strong version of that view), although a more detailed investigation of this topic is to be desired for the future. This also applies to the development of phrasal-prepositional verbs, for which hardly any systematic historical information has been gathered so far, except for the general assumption of a diachronic increase, cf. e.g. Denison (1998: 223). Interestingly, Denison (1998: 224) also discusses the overall increase of prepositional verbs (cf. there for further literature on the topic), but his approach to ‘group-verbs’ seems to prevent him from establishing a connection between the two developments. In the approach taken here, where phrasal verbs are regarded as periphrastic word formations, the issue, of course, tends to disappear altogether, at least as far as the history of the verb-particle construction is concerned: there is no reason why the increasing use of prepositions to satisfy the argument frame of the verb should not affect simple and complex verbs alike, while that increase, worthy of attention as it is, has no direct connection to the history of the verb-particle construction. 5.4.

Conclusion

In this chapter the emergence of the Modern English verb-particle construction has been examined from a morphological, lexical and semantic point of view. I have argued that a language-internal account of the history of the English verb-particle construction is possible and, what is more, decidedly more plausible than assumptions of external influences. In this context an evaluation of the long-term loss of most of the native (Germanic) prefixes has shown that the loss of the native prefixes was in principle independent of the development of the particles (and also of the establishment of borrowed prefixes), although the particles, with their similar spatial etymologies, were bound to take similar paths of semantic development. As a consequence, preverbs which came to be fused with the verbal base have been largely replaced by preverbs which, due to syntactic changes, by now occur separated from the verb, in postverbal position. As a construction type, particle verbs turn out to be rather insensitive to the etymological origins of their components, as long as the meaning of a simple verb does not preclude its combination with spatial or aspectual particles. This is the case with many borrowings from Latin, which clearly

196


prefer combining with Latinate prefixes. However, the preference of the Latinate verbs for etymologically congruent prefixes is also, and perhaps more importantly, connected to their special status in the post-medieval English lexicon, a topic we will return to in the next chapter.

Chapter 6 Frequency, style and attitudes A number of studies have focused on the quantitative development of phrasal verbs from the late Middle Ages onwards. Some of these studies are mainly concerned with an overall diachronic quantification of phrasal verbs in the English lexicon, while others have tried to take into account considerations of variationist aspects connected to their occurrence in different text types, registers, or national varieties. Once again it seems that many of such investigations take for granted the notion that there is something like a rise and subsequent spread of the phrasal verb in English, and quite clearly all of them are confident that the phrasal verbs have always been stylistically marked as ‘low’, ‘informal’, or ‘colloquial’. In the following it will be argued that on closer scrutiny many of the results turn out to be rather inconclusive and that so far a satisfactory quantitative account of the history of the phrasal verbs remains to be written. In particular, the common stylistic evaluation of phrasal verbs throughout the history of English must be regarded with a considerable degree of caution and the quantitative observations adduced in its favour are far from persuasive. This will become clear through a representative discussion of the normative tradition and of the linguistic practice of a number of style-conscious authors from between the end of the 17th and the beginning of the 20th century. 6.1.

Counting phrasal verbs

Let us begin with a discussion of a number of studies which provide figures for the overall frequencies of phrasal verbs in various texts since the late Middle Ages. 6.1.1. Quantitative long-term developments from Middle English to the 20th century As so often Kennedy (1920) sets the pace for much of 20th-century research (cf. the discussion in Chapter 4 above). Thus, his influence is also noticeable in Konishi’s (1958) study of the long-term developments in English. For medieval English, Konishi relies entirely on the figures given

198

Frequency, style and attitudes

by Kennedy, and he explicitly connects the development of the phrasal verbs to the analytic tendency in English, where the development of the exclusively postpositional pattern is seen to engage in an interplay with the loss of inseparable verbal prefixes during and after the Old English period and the influx of Romance loanwords in Middle English, which he describes as the “long history of struggle between the native compound verb and the verb-adverb combination” and the “sudden inrush of a multitude of Romanic compound verbs since the fourteenth century” (Konishi 1958: 119). Konishi also follows Kennedy in connecting the phrasal verbs to what he calls “the ordinary conversation of the masses”, which is taken as the primary explanation for the quantitative developments from Middle English onwards: Kennedy wishes to emphasize in view of certain conclusions that the development of the verb-adverb combination would have been much more rapid had it not been weakened for some generations or even centuries, by the adoption into the English language of numerous Romantic [sic] verbs. Certainly this is true of written English or the formal speech of the educated people, but, I venture to say, it is not true of the ordinary conversation of the masses to whom these combinations mainly belong. The inrush of Romanic verbs had been strong enough to check the growth of the phrasal verbs, indeed, but at the same time I must add that it had been strong enough to drive out the native compounds which stood firm in the way of the combination. In other words, I would rather say, the verb-adverb combination had succeeded in conquering its original enemy by the hands of the newcomer, so that the obstacle had been got out of the way for its startling progress at the early period of Modern English. (Konishi 1958: 119–121)

In the present context it is significant that Konishi extends the somewhat adhoc-ish line of stylistic reasoning in his rather metaphorical narrative to the later periods of the language. In the OED entries for ten frequent simple verbs, he observes a marked and steady increase in first quotations for phrasal verbs between the 13th and the 17th century, which is followed by a slump towards the 18th century; only in the 19th century is there again an increase. The long-term development, then, is represented as in Figure 6-1 below. The wave in Figure 6-1 suggests a continuous development, which Konishi (1958: 121–123) explains with references to the respective linguistic spirits of the age and contemporary stylistic norms and attitudes. Since his figures are based on the first quotations of phrasal verbs or new senses of existing ones every 50 years in the OED quotations, their validity depends crucially on the dictionary’s coverage.


199

Figure 6-1. Development of phrasal verbs (Konishi 1958: 125)

But although the treatment of verb-particle constructions in the OED is superior to that in other historical dictionaries (cf. Chapter 4 above), statements concerning the development of the English lexicon in different periods need to be taken with a fair degree of caution if they are based on the OED (i.e. a considerable portion of such statements made in the course of the 20th century). Even though the OED database, with its strong literary bias, can be assumed to be relatively homogeneous, the coverage of the dictionary is remarkably uneven, as Schäfer (1980 and 1989) has shown. This applies to the numbers of sources used for the compilation of the dictionary, which vary enormously over time. According to Schäfer (1980); the 16th century is relatively poorly covered and there is a considerable increase in the 17th century, while for the 18th century again fewer sources were used. The increase in 19th-century sources reflects the origins of the OED in the second half of that century. Schäfer has also demonstrated that some sources were read much more closely than others. Hoffmann (2004) has shown that not only the number of quotations in the OED is unevenly distributed across time, but also the average length of the quotations: there is a slow but steady increase, and the Modern English part of the quotation corpus is altogether considerably larger. As a result of these and some other peculiarities of the OED, the history of the English lexicon has often been distorted, due to a confusion of the substantial vagaries of the

200


lexicographical coverage and the ‘real’ history of the language (cf. Brewer 2000, Mugglestone 2005 and Thim 2011b). In other words, it is very problematic to draw conclusions about long-term changes in the productivity of late medieval and Modern English phrasal verbs on the basis of the OED entries, unless the peculiarities of the dictionary’s documentation are taken into account. Konishi’s figures really reflect the quantitative diachronic makeup of the OED, rather than the development of the phrasal verbs in the language. A later study, however, seems to confirm Konishi’s findings. Using a very different (but still a literary) database, Spasov (1966) observes a longterm increase in the use of phrasal verbs between late medieval and 20thcentury English. Counting the incidence of phrasal verbs in samples of 300 verbs in each of 46 plays, Spasov concludes that there is a “steady increase in the use of phrasal verbs since Renaissance times” (Spasov 1966: 23), which he illustrates as shown in Figure 6-2. But just like Konishi’s figures, Spasov’s look decidedly less cogent on closer inspection. His samples are rather small and it is not quite clear what they are representative of. On the whole, the graph creates an impression of growth. But there is nothing in Spasov’s data which would justify the rise towards the group of ‘Mysteries and Miracle Plays’, since his counts start with them. There is also no reason for the two following groups, ‘Shakespeare’s Contemporaries’ and ‘Shakespeare’ to be ordered successively, when in fact the plays in the former group were, on average, written later than the ones in the latter (the average date of composition for the eight plays by Shakespeare is 1599, while the average date of composition for the eight plays by his contemporaries is 1609). There seems to be indeed an increase towards the ‘Restoration’ group, which is taken as evidence for a steady overall increase; the decrease in the two subgroups for the 18th century (in particular in the three plays used as samples for the second half of the century), though, is explained away by a wholesale reference to stylistic predilections. Spasov’s explanations for the frequencies are, once again, in line with the traditional narrative: The only considerable drop, as seen in the above scheme, is to be observed in the 18th century, and that particularly in the second half when writers seem to fight shy of phrasal verbs. But then, it should be remembered that it is not improbable that in certain periods writers may not have reflected the real state of affairs in the language, following the fashion of the day or personal likes and dislikes. On the whole, authors whose language is based on popular speech are apt to use a larger number of phrasal verbs than those who introduce refined or sophisticated speech in their works. (Spasov 1966: 22)


201

Figure 6-2. Phrasal verbs in English plays (Spasov 1966: 21)

In other words, if the number of phrasal verbs increases, this is taken to reflect the general trend in the ‘real’ language, and if it decreases, it is taken to be merely a writer’s personal predilection, or the result of the linguistic taste of the time. After the 18th century, the evidence for an overall increase towards the 20th century appears to be somewhat more robust, and seems to be corroborated by more extensive data (cf. the discussion below). But the apparently crucial developments in the 19th century are left unaccounted for, since there is a gap of more than a hundred years in the coverage, and most of the 19th century is not covered at all (the last play from the 18th century is from 1777, while the first play from the ‘Pre-IWorld-War’ group is from 1885). With regard to a small additional sample of 20th-century novels the reason for one character having 17% phrasal verbs in all his verbs in the first three chapters is, predictably, explained with the author’s attempt at making the figure “more true to life”, while the complete absence of phrasal verbs in the speech of another character in these chapters makes her language “more individual” (Spasov 1966: 23). Despite its weaknesses, Spasov’s analysis, just like the long-term account offered by Kennedy (1920), has been perpetuated by Bolinger (1971: xi, fn. 1) in his very brief historical reference to Spasov’s findings.1 Nonetheless 1

In the past the book must have been rather difficult to procure. E.g. both Pelli (1976: 43) and Claridge (2000: 97) state that they could not consult it, and both (Pelli 1976: 126 and Claridge 2000: 98) reproduce the graph from Bolinger (1971: xi, fn. 1). Although the decline in frequency observed by

202


one cannot but agree with Martin’s (1990: 32) critique, “Spasov’s study tantalizes the curiosity but does not satisfy it”. But although the older quantitative studies are unsatisfactory and cannot be accepted as a reliable basis for any kind of observation on the development of the phrasal verbs, Kennedy (1920), Konishi (1956) and Spasov (1966) continue being quoted extensively in the literature, where they tend to be regarded as reliable sources (cf. e.g. the references in Hiltunen 1994, Claridge 2000 or Elenbaas 2007 and the uncritical use in the handbooks), and in several instances the chain of quotes seems to turn into a game of Chinese whispers (cf. also the discussion of Dryden’s presumptive replacements of phrasal verbs in Section 6.2.4 below). Among the long-term studies, only Martin (1990) seems to provide a more dependable overview. Being an unpublished dissertation, the study appears to have had virtually no impact at all on the subsequent historical research on phrasal verbs. Martin’s corpus consists of private letters written between the 15th and the 20th century, and her quantitative analyses are based on the first one hundred phrasal verbs (tokens) found in the subcorpora for each century. This is, of course, a problematic procedure, not least since it is hard to see why the large number of correspondents announced in the descriptions of the sub-corpora should be needed for such a small number of phrasal verbs and what the criteria for the selection of individual informants/letters within these sub-corpora were; here again representativeness is an issue. But with regard to absolute numbers of phrasal verbs investigated over time, Martin’s study is as yet unsurpassed. Her diachronic corpus consists of six sub-corpora for British English (15th– 20th century, one for each century) and three for American English (18th– 20th century). The choice of letters as the basis of the corpus is motivated by the assumption that “their language is closer to the vernacular of ages past than any other available linguistic record” (Martin 1990: 47), although she is aware that the functional domains of letter writing may have changed considerably over time (Martin 1990: 97–98); see e.g. Schaefer 1996 for a critical examination of such claims with regard to the 15th-century Paston Letters. On the whole, the corpus attempts to include informal letters, ideally holographs written by writers of comparable socio-economical status and to avoid the overrepresentation of idiolectal features; for a more Spasov for the 18th century is one of the central concerns of their studies, both do not cover the whole 18th century because of their corpora, and they must rely on Bolinger’s somewhat casual description of Spasov’s database and his results; cf. also the discussion below.


203

detailed account, see Martin (1990: 57–73), where, however, no information is provided as to which letters exactly were included in the corpus (e.g. for the Paston Letters there is just a list of the members of the Paston family whose letters were used rather than a list of the individual letters). Of interest here are Martin’s results for the ‘overall density’ in texts, i.e. the “ratio of number of phrasal verbs to number of overall words” (Martin 1990: 79) where she finds a considerable long-term increase, but on the whole little difference between British and American English: Phrasal verbs density in informal letters has increased over the centuries, as expected. This is to say that the letters of today, at least those studied, feature almost five times as many phrasal verbs as those of five centuries ago. Unexpected was the similarity of British and American trends, particularly in the 20th century, when density is virtually identical. This finding does not support the common perception of Americans as the heavy users of the phrasal verb form. (Martin 1990: 100)

The percentages presented by Martin are mostly based on estimated figures for the overall number of words in the texts which make up the corpus (for the details of the quantitative procedure, see Martin 1990: 78–79). It appears, then, that there is little evidence for the universal assumption that phrasal verbs are more typical of American English than of British English. This is a topic decidedly in need of more thorough investigation, as the somewhat different results obtained by Rohdenburg (2009) for at least some phrasal verbs attest. But to date even the genesis of the assumption of a difference between the two varieties (which must lie somewhere in the 19th century) is unclear, and statements like the one in Fowler s.v. phrasal verbs (1996, ed. Burchfield), “[i]t is clear that the use of phrasal verbs began to increase in a noticeable manner in America from the early 19c. onward. From there, many have made their way to Britain during the 20c.”, are, although not necessarily wrong, as yet not supported by persuasive evidence. Interestingly, though, the figures obtained by Martin seem to confirm some of the general trends posited by Konishi (1956) and Spasov (1966) on the basis of their unrepresentative and generally rather problematic corpora. Overall, the incidence of phrasal verbs rises from 0.2% in the 15th century (corroborated by the figures in Thim 2011a) to 1.1% in the 20th century. There is only a very slow but steady increase between the 15th and the 17th century; i.e. the dramatic rise shown by Konishi’s figures is quite predictably not confirmed by Martin’s more balanced corpus, but there is support for the general trend towards greater frequency observed by Spasov. In the 18th century, then, the figures remain steady, and in the

204


British sub-corpus there is even a slight decrease at the end of the century, which however is considerably less marked than in Spasov’s study, while the figures for the American sub-corpus remain quite stable throughout the 18th century. This would seem to confirm my assessment that the 18thcentury dip claimed by Konishi merely reflects the coverage in the OED. Again this may not be taken as evidence for the claim that phrasal verbs are more typical of American English in general, since in both varieties there is an increase in the 19th century which Martin (1990: 102) characterizes as a ‘skyrocketing’ one (cf. also Smitterberg’s 2008 figures for the 19th century). That increase, moreover, appears to be decidedly steeper for British English, to the point that the absolute figures for the 19th century are considerably higher in British English than in American English, while the absolute figures for the 20th century appear to be identical in both varieties. Martin does not give any absolute figures for the overall token incidence of the phrasal verbs in her sub-corpora, so that one has to rely on her Figure 1 (Martin 1990: 103) for these observations – especially since she does not comment on the 19th-century divergence between British and American English in her data at all. Moreover, since there is a gap of 70 years between the ‘18th-century’ and the ‘19th-century’ sub-corpus, it is impossible to make any more detailed statements about the developments in the two varieties on the basis of these data. One problematic aspect of Martin’s study is that the respective sub-corpora for the 15th to 20th centuries are restricted to the following periods: 1440–1479, 1533–1573, 1637– 1661, 1760–1790, 1860–1890, 1975–1987 (Martin 1990: 56), although the chart with the results creates the impression of a continuous development, as in the earlier studies discussed above. Thus the results can be used only for the establishment of rough diachronic tendencies, and the exact figures are not provided. But nevertheless on the whole the observation that a rather dramatic increase occurs in the course of the 19th century appears to be a very robust, yet so far totally unexplained, finding (but cf. Section 6.1.4 below). These findings are also borne out by Pelli’s (1976) study of phrasal verbs in 68 American dramas published between 1765 and 1972. Pelli’s corpus of approximately 1.25 million words is divided into six sub-periods (Pelli 1976: 1–7) and it shows the beginnings of a considerable increase around the middle of the 19th century, while his overall percentages tally nicely with those of Martin. Albeit not at odds with any of the other studies discussed here, his data would seem to offer new challenges of interpretation, which shall not be addressed here in greater detail, e.g.: why is there a decrease in his last period compared to the preceding one? And, more intriguingly in the light of the previous observations on the late 18th-century


205

dip: why is there a slight decrease after the first period – is the late 18thcentury dip a phenomenon of the long 18th century? If so, there would seem to be some grounds for the assumption that the development of phrasal verbs in the late 18th century may indeed be seen in connection to extra-linguistic factors. Martin’s assumption that the American development simply lags behind the British one because Americans used to carry “a cultural chip on their shoulder” (Martin 1990: 108) seems rather problematic, not least because she does not offer any comparable British data which could be used to justify that claim. Pelli’s explanations for his figures follow the predictable traditional pattern of referring to informality, common people and colloquiality. We will return to this in Section 6.2 below. 6.1.2. Early Modern English frequencies Compared to the little attention Late Modern English – in particular the 19th century – has received, the Early Modern English period has been given a fair deal of attention, and there are a number of studies which are based on extensive corpora, most notably those by Hiltunen (1994) and Claridge (2000). But the figures for Early Modern English vary considerably among the different corpora, and also within them, and often no clear correlation between date of composition, text type, and social variables can be established. With all the studies discussed here, the mutual comparability is an issue. In brief, comparisons between these studies yield at best vague approximations since they are based on corpora containing widely different texts, but also because there are differences in the authors’ categorizations of elements as phrasal verbs. Claridge (2000) is based on the Lampeter Corpus of Early Modern English Tracts, which contains 1,172,102 words in complete texts from the Tract Collection in the Founder’s Library of the University of Wales, Lampeter. The tracts are non-literary prose texts covering a variety of topics, which serve as the basis for the sub-division of the corpus in the domains ‘Science’, ‘Economy’, ‘Religion’, ‘Law’, ‘Politics’ and ‘Miscellaneous’. These groups are of roughly even size for the ten decades covered by the corpus and information on the socio-economic status of the authors is provided as far as possible, with the aim to facilitate socio-historical research (for a more detailed description of the corpus, see Claridge 2000: Ch. 2). The Lampeter Corpus covers the period 1640–1740, i.e. the time when Konishi’s and Spasov’s data suggest a drop in frequency.

206


Figure 6-3. Development of phrasal verbs per 1,000 words in the Lampeter Corpus, 1640–1740, per decade (according to Claridge 2000: 178)

Figure 6-3 shows the distribution of phrasal verbs over time Claridge finds per 1,000 words of text. With respect to quantitative changes within the period covered by the corpus, Claridge’s findings are quite inconclusive, so that on the basis of her data no safe statements on the development of phrasal verbs between 1640 and 1740 can be made: Another point regarding frequencies concerns the question of linguistic change …, in particular Konishi’s (1958) and Spasov’s (1966) claim of the decline of phrasal verbs in the period in question. This is connected with the question of suitability of the Lampeter Corpus for this investigation. While nothing can be said about phrasal-prepositional verbs, because their overall frequency is too low … phrasal verbs, on the other hand, produce a rather erratic graph, which failed to point to either stability or any clear development. Thus, the above authors’ suggestion can neither be repudiated nor substantiated on the basis of the present data. However, the pres??ent [sic] result may be taken as an indication that their data base was too small to really warrant such a claim. (Claridge 2000: 277; her italics)

Thus, the lines of development assumed by Konishi (1958) and Spasov (1966), which, as we have seen above, are open to doubt anyway, cannot be confirmed by a more detailed study of a considerably larger corpus, and both the increase thought to be observed in the Early Modern period and the drop towards the 18th century are possibly quite fortuitous (but cf. Section 6.3 below). Moreover, the erratic graph produced by the data in


207

Claridge’s Lampeter Corpus is a clear indication that even in a relatively homogeneous corpus any diachronic snapshot of the frequencies of phrasal verbs may be completely unrepresentative. In the case at hand, for a comparison per decade the figures are not statistically significant anyway (Claridge 2000: 176–177), and for the time being it must be concluded that there is no reliable evidence at all for quantitative diachronic changes between the middle of the 17th and the middle of the 18th century. But it also seems that there is no other factor which can be convincingly connected to the use of phrasal verbs in the texts of the corpus, and this is also true of their occurrence in other corpora, as will be argued in Section 6.2 below. Hiltunen (1994) discusses phrasal verbs in the Early Modern English part of the Helsinki Corpus, which contains 551,000 words in three subcorpora (1500–1570, 1570–1640, 1640–1700) of roughly even size, with texts drawn from various text types.2 Hiltunen’s discussion is restricted to phrasal verbs containing the particles away, back, forth, off, out and up, of which there are altogether 851 instances in the corpus. His brief comparison of the relative frequencies of occurrence of phrasal verbs in the Early Modern English part of the Helsinki Corpus and the present-day English London-Oslo/Bergen Corpus (LOB)3 suggests that by and large the relative order of frequencies of combination is similar in both corpora and that “the most common collocations tend to be made up of the same lexical items in both periods” (Hiltunen 1994: 134). Since the two corpora differ in size and with respect to the kinds of texts included, more far-reaching 2

3

The Helsinki Corpus has been used for a very large number of studies and is certainly among the best-described corpora; its merits and limitations are quite well known and it is therefore not necessary to discuss it here in greater detail. For further information, see e.g. Kytö, Ihalainen & Rissanen (1988), Rissanen (1992 and 1994), Kytö (1993), Rissanen, Kytö & Palander-Collin (1993), and also the VARIENG bibliography online (). LOB contains roughly one million words in 500 texts of approximately even length; together with the Brown University Corpus of American English (the ‘Brown Corpus’), it marks the beginning of electronic corpus research tradition in English linguistics; both corpora were published in 1961 and were compiled on the same principles. For a description of LOB, see in particular Johansson (1978); for an overview of corpus linguistics and the use of LOB and the Brown Corpus, cf. e.g. McEnery & Wilson (2001), Lindquist (2009) or Mukherjee (2009). Hiltunen does not provide any figures for LOB except for the comparison of a number of frequent individual phrasal verbs.

208


conclusions are not drawn, although Hiltunen thinks it possible that most of the different frequencies of occurrence may be due to the different subject matters of the texts included in the two corpora rather than represent diachronic developments. Nevertheless the material suggests a general tendency for combinations with forth to have decreased and for combinations with out to have increased in frequency since the Early Modern English period. Without discussing diachronic developments within the Early Modern English period, Hiltunen establishes three groups of phrasal verbs according to particle frequencies in his corpus (<100: back, off, forth; >100: down, away; >200: out, up). Claridge (2000: 172) provides normalised figures for phrasal verbs with the six particles studied by Hiltunen on the basis of his figures for the Helsinki Corpus and compares them to those for the Lampeter Corpus used by herself. This comparison is summarized in Table 6-1 below. Again, a comparison of these corpora shows that diachronic frequencies cannot be taken to offer more than very rough approximations. Since the Lampeter Corpus is almost exactly twice as big as the Early Modern English part of the Helsinki Corpus, one might perhaps expect that the absolute numbers of phrasal verbs found in the two corpora would behave correspondingly. But on the basis of these figures it appears that most phrasal verbs are, in fact, about twice as frequent in the Lampeter Corpus (i.e. the absolute figures are four times higher) than in the Helsinki Corpus, with those containing the particle forth being the only noticeable exception. Since there is no increase to be observed within the period covered by the Lampeter Corpus, one conceivable explanation would be that these figures reflect a dramatic increase which takes place before 1640 but considerably after 1500. But this would be in contradiction to Konishi’s, Spasov’s and Martin’s findings (even if the figures provided by Konishi are normalized with respect to the coverage of the OED) and it would hardly be borne out by the diachronic developments to be observed in the Helsinki Corpus either. Also, in the small Early Modern English corpus covering the period 1500–1700 used in Thim (2006a), the phrasal verbs tend to be even more frequent than in the Lampeter Corpus (cf. Section 6.1.3 below). A comparison with 15th- and 16th-century letters contained in the Corpus of Early English Correspondence (CEEC) confirms the impression that the figures from the Lampeter Corpus seem to be out of line and, perhaps more remarkably, that there is again little reason to assume an increase on the basis of a comparative evaluation of these figures and those provided by the other corpora.


209

up

down

in

forth

off

away

back

home

on

Helsinki Corpus 1500–1700 Lampeter Corpus 1640–1740

out

Table 6-1. Phrasal verbs in two corpora

209

228

125

–

84

57

127

21

–

–

932

992

427

323

196

356

335

84

1

254

Table 6-2 below summarizes the occurrence of phrasal verbs with one of the particles out, up, down, forth and back in a sample from the Corpus of Early English Correspondence (originally discussed in Thim 2006b); cf. also the etymological and semantic discussion of phrasal verbs in the corpus in Chapter 5 above. For a description of the corpus, see e.g. Nevalainen & Raumolin-Brunberg (2003), who also discuss a number of Early Modern English changes in progress on the basis of the corpus, and cf. also the pilot studies using the corpus in Nevalainen & Raumolin-Brunberg (1996). The sample from the corpus used here contains the full 15th- and 16th-century material with the exception of the Paston Letters: 246,055 words from the Corpus of Early English Correspondence Sampler and 780,192 words from the copyrighted material contained in the non-public version of the corpus, i.e. 1,026,247 words altogether. This part of CEEC is thus about one tenth smaller than the Lampeter Corpus; the texts it contains are on average 200 years older. The figures in Table 6-2 are based on a WordCruncher analysis of the corpus, where I have searched for all spelling variants of the particles; on this basis the phrasal verbs were thus retrieved. The five particles discussed here were selected since they are of high, medium and low frequency and can be assumed to be fairly representative of the use of phrasal verbs in the corpus (cf. the more detailed discussion of these data in Thim 2006b). Table 6-3 provides a comparison between CEEC, the Helsinki Corpus and the Lampeter Corpus. Taking into account their different sizes, it is remarkable that there is no significant difference between the occurrences of phrasal verbs in CEEC and the Helsinki Corpus, despite the differences in diachronic coverage and text types, while some phrasal verbs in the Lampeter corpus are approximately twice as frequent. Stylistic differences between the texts in the corpora might offer an explanation for these findings, but we shall see in Section 6.2 below that this is not the case.

210


Table 6-2. Phrasal verbs in the Corpus of Early English Correspondence CEEC 15th–16th century

out

up

down

forth

back

429

499

226

127

41

Table 6-3. Phrasal verbs per 1,000 words in three corpora CEEC 15th–16th century Helsinki Corpus 16th–17th century Lampeter Corpus 1640–1740

out

up

down

forth

back

0.48

0.49

0.22

0.12

0.39

0.37

0.41

0.23

0.15

0.38

0.79

0.85

0.36

0.16

0.72

6.1.3. Relative frequencies of particles The relative frequencies of the different particles in the historical corpora are an indication that the figures discussed so far are not completely random. The results of a comparison of the seven particles covered in each of the studies by Hiltunen (1994), Thim (2006a) and Claridge (2000) are summarized in Table 6-4. As the table shows, the relative frequencies – in particular those of the more frequent particles in the two large corpora – are almost identical. Indeed, the most notable exception is forth, whose frequency is considerably lower in the Lampeter Corpus than in the other two corpora (but cf. also the figures for away). Since forth is subject to a longterm decrease, which can be assumed to be well under way in the 18th century (cf. OED s.v. and Akimoto 2006), the lower frequency in a corpus where the earliest texts are 140 years younger than the earliest texts in the other two corpora is likely to be a result of this trend. Hiltunen (1994) and Claridge (2000) do not provide information on the developments of individual particles within the periods covered by their corpora, so that a more precise statement on the basis of their data is not possible. The Everyday English corpus (EvE) first used in Thim (2006a), on the other hand, is too small for any observations on developments taking place within the period it covers. It is rather small (c. 36,000 tokens from Cusack 1998), and although the overall correspondences to the other corpora as regards the relative frequencies of the particles may be taken as an indication that the use of phrasal verbs in EvE can be regarded as representative, the size of the corpus renders it problematic with respect to the rarer particles (cf. e.g. the figures for back).


211

Table 6-4. Relative frequencies of particles in three corpora Helsinki Corpus 1500–1700 Everyday English 1500–1700 Lampeter Corpus 1640–1740

out

up

down

forth

off

away

back

total

0.25

0.27

0.14

0.09

0.07

0.15

0.02

851

0.23

0.34

0.16

0.09

0.06

0.06

0.04

170

0.25

0.27

0.12

0.05

0.09

0.09

0.02

3645

While the texts in EvE were never intended for being printed and tend to be quite remote from the norms of the emerging standard, the tracts in the Lampeter Corpus are all by definition printed, and were in fact typically written for print publication. The texts in the Helsinki Corpus, on the other hand, include diverse text types such as letters, handbooks, bible translations, etc. On the whole the use of phrasal verbs in these three rather different corpora matches the long term trends in the relative frequencies of the individual particles and is, moreover, characterised by a high degree of consistency. But the absolute numbers of phrasal verbs (or rather: their frequencies relative to the size of the corpus) generally do not seem to offer any significant insights. This conclusion receives support both from quantitative long-term comparisons and from stylistic considerations, two topics we will turn to in the following sections. 6.1.4. Quantitative long-term comparisons A diachronic comparison of frequencies of phrasal verbs since the 15th century can be attempted on the basis of a number of studies; an overview is provided in Table 6-5 below. A meaningful interpretation of the figures given in Table 6-5 is extremely difficult. It is, of course, possible that some or even all of these studies cannot be relied on. But with the exception of Biber et al. (1999), in each of these studies the criteria for classification as phrasal verb are quite similar, and there is little reason to doubt the general reliability of the counts. Biber et al. (1999) use the Longman Spoken and Written English Corpus (LSWE) as the basis of their quantitative analyses. But the figure for phrasal verbs quoted here from Biber et al. (1999: 408) is problematic for a number of reasons, in particular since it is not quite clear from their discussion what is actually classified as a phrasal verb. It seems that for intransitive phrasal verbs the defining criterion is idiomaticity, while for transitive phrasal verbs it is the positional properties of the particle and the object (Biber et al. 1999: 404–409). This is a somewhat

212


unfortunate definition (cf. also the discussion in Chapter 2 above), and the examples given in the grammar are also not entirely consistent: e.g. go in is adduced as an example of a ‘free combination’ (i.e. not a phrasal verb), while walk in and move in are treated as examples of ‘activity intransitive’ phrasal verbs. Since some transitive compositional phrasal verbs will have been excluded from the count, it seems that the figure quoted here should be slightly higher; the other studies quoted in this table have included in their counts all verb-particle constructions independent of their semantics. But although it is claimed that the corpus findings for the phrasal verbs are based on an analysis of the whole corpus (Biber et al. 1999: 1134), this does not seem to imply the truly Herculean task of a thorough and dependable count and analysis of all phrasal verbs in a corpus of that size. Even if one were to leave aside the extremely low figure given for LSWE in Biber et al. (1999), one would have to account in some way or another for the fact that the frequency of phrasal verbs in Shakespeare’s English (Castillo 1994) is apparently higher than in the English of present-day university students (Waibel 2007). And even if that observation could be explained somehow by referring to some peculiarity of Shakespeare’s language (cf. the discussion below), one would have to account for the fact that mid17th-century authors of tracts (Claridge 2000) use almost as many phrasal verbs as those students, while the 20th-century texts in LOB (Johansson & Hofland 1989) and in Martin’s letter corpus (1990) contain a considerably higher number of phrasal verbs. But it seems very dubious to identify diachronic frequencies which look like the one given in Table 6-5 as an overall increase. On the basis of these figures alone, there is only one conclusion to be drawn. If we are to ignore the gross differences between these corpora, we cannot use any of their results for statements on the diachronic developments in English, unless the counts in some of them could be shown to be wrong. If, on the other hand, we are to take them seriously, we must admit that the figures are completely inconclusive. Plag, Dalton-Puffer & Baayen (1999), in a study of the productivity of suffixes in different present-day English registers, show that morphological productivity may be dependent on register and that the productivity of individual morphological elements may vary considerably across registers, for which it seems difficult to find simple explanations. With respect to phrasal verbs, there is of course the possibility that the figures are distorted by stylistic factors which account for the different frequencies in different text types and hence in the different corpora. A brief look at the arguments put forward in favour of such an explanation shows that this has not been demonstrated convincingly, cf. Section 6.2.1 below.


213

Table 6-5. Frequencies of phrasal verbs in nine corpora (tokens per 1,000 words)4 1440–1479 1441–1478 c1600 1500–1700 1640–1650 1640–1740 1800–1900 1960 1975–1987 present day present day

[some] Paston Letters (Martin 1990) Margaret Paston’s letters (Thim 2011a) Shakespeare (Castillo 1994) EvE (Thim 2006a) Lampeter Corpus (Claridge 2000), one decade Lampeter Corpus (Claridge 2000), complete CONCE (Smitterberg 2008) LOB (Johansson & Hofland 1989) private letters (Martin 1990) LSWE Corpus (Biber et al. 1999) LOCNESS (Waibel 2007)

2.0 1.9 6.5 5.8 4.5 3.6 5.9 8.6 10.9 1.4 5.2

As matters stand, it seems that many statements made so far about the quantitative diachronic development of the phrasal verb as a construction type throughout the Modern English period are meaningless. One could of course try to arrive at new long-term comparisons by analysing the development of phrasal verbs in a very large diachronic corpus. The work would be enormous, though, especially since it would be problematic to look at only a selection of particles or phrasal verbs (typically those which are easily retrievable with a concordance program, as has been done in a number of studies). Since the inventories of particles and of verbs are subject to change, the frequency of the construction type can only be reasonably established by an analysis of all of them; note that the homonymy of some of the particles to some highly frequent prepositions would render an electronic access inadvisable in an untagged corpus. One could only hope for significant insights, I would suggest, if such a long-term quantification also included a transparent semantic classification and thus 4

CONCE: Corpus of Nineteenth-century English (seven genres), period 1 (1800–1830, 346,176 words) and period 3 (1870–1900, 298,796 words). – LOB: based on the figures given in Johansson & Hofland (1989) for the tagged Corpus. – LOCNESS: the Louvain Corpus of Native English Essays contains 324,304 words in essays written by British pupils and British and American university students (the figures in Waibel 2007 exclude the essays written by pupils, and there remain 90 essays by British students and 207 essays by American students; the corpus as used by Waibel then contains 263,974 words. – LSWE: the Longman Spoken and Written English Corpus used in Biber et al. (1999) contains over 40 million words in 37,244 texts, which are subdivided into four major register categories (‘conversation’, ‘fiction’, ‘news’ and ‘academic prose’).

214


managed to trace the development of aspectual and idiomatic constructions over time. This would also be necessary for the establishment of changes in their productivity (cf. e.g. Baayen & Renouf 1996, Plag 1999, Barðdal 2008 and Bybee 2010); none of this has been done in any of the studies of the history of phrasal verbs available so far. It will be argued in the following section that counting phrasal verbs in order to find out about their history may to some extent be beside the point anyway (cf. also Adamson’s 1998: 592 critique of similar quantifying approaches). Looking at the changing frequencies of phrasal verbs in the history of English has always been connected to the idea that there is something like a rise of the phrasal verb. And even if there is a diachronic increase in the use of phrasal verbs, this is hardly an instance of colloquialization of Late Modern English (pace Mair 2006). Rather, it had better be regarded as an indication of an ongoing de-formalisation of the language, as the discussion of some stylistic considerations in the following sections will show, where I will argue that particle verbs are a stable property of common English, and that it is quite likely that by and large the frequencies described in the literature so far cannot be explained simply with reference to general patterns of growth, and only indirectly with reference to diastratic and diaphasic factors (cf. Thim 2006a). This may have changed to some extent in the course of the 19th century, when it seems that phrasal verbs are increasingly regarded as symbolic of colloquial styles; see also Smitterberg (2008), whose figures suggest changing frequencies in individual text types rather than an overall increase of the construction type. Moreover, although an overall increase between the late Middle Ages and today is quite possible, this had better not be seen in connection to the particular ‘Englishness’ of the construction, as the comparison with German shows, where considerable patterns of growth have been reported (cf. Habermann 1994 for Early New High German and Herbers 2002 for Middle High German), though not fully explained. 6.2.

Style and attitudes

The general opinion on the stylistic evaluation of the phrasal verbs in the history of English has been summed up nicely by Konishi (1958): With the gradual establishment of Modern English analytic language this wordformation has made remarkable progress through the propitious process of the colloquial speech of the uneducated. (Konishi 1958: 119)

Style and attitudes

215

This of course echoes Kennedy’s views (1920), and similar statements will be found in practically any other publication on the history of the phrasal verb. It is therefore unnecessary to give extensive references to the literature here: this view is more or less taken for granted in every single title on the subject. Moreover, this view is also commonly encountered in grammars and popular style guides for present-day English (often with unreliable historical background information; cf. e.g. the discussion of the treatment in McArthur 1992 and in Biber et al. 1999 in Chapter 2 above). In the preceding section several attempts at explaining the frequencies of phrasal verbs by referring to their presumptive stylistic properties have already been discussed. Typically such connections are drawn on an ad hoc basis, in order to explain diachronic patterns of growth or decline in corpora which on closer inspection turn out to be rather speculative anyway. Interestingly, phrasal verbs appear never to be mentioned in general discussions of colloquial Middle or Early Modern English (at least I am not aware of phrasal verbs being mentioned in such studies; cf. e.g. the complete absence of phrasal verbs in the comprehensive discussions by Görlach 1999 and Adamson 1999). This in itself could, of course, simply be a weakness of the extant research, but it could also be an indication that looking at phrasal verbs with a determination to see them as colloquial might be misguided. The present section has a number of related concerns. First, it will be shown that the evaluations of phrasal verbs in different historical text types attempted so far are not convincing and that here again the explanatory value of such attempts has been quite low. There is also very little reason to assume that phrasal verbs in general were regarded as being stylistically marked by speakers of English before the middle of the 18th century, and it is quite likely that they were perceived as ‘colloquial’ only from the 19th century onwards. 6.2.1. Text type and frequency As has been shown in the preceding section, the older historical studies of the phrasal verbs tend to introduce stylistic considerations whenever diachronic changes need to be accounted for. An increase is typically attributed to a long-term tendency, while a decrease is attributed to stylistic peculiarities of the texts at the point in time in question. For example, if the percentage of phrasal verbs in the 20th-century LOB is higher than in the 17th-/18th-century Lampeter Corpus, this is taken to reflect an overall increase, but if the percentage in Shakespeare is higher than in the

216


Lampeter Corpus, this is taken to reflect Shakespeare’s “extraordinary linguistic inventiveness” (Claridge 2000: 125); cf. also the discussion by Hiltunen (1999: 158–160). This observation, however, does not apply to studies which focus on Old English alone, where the status of the particle verbs as a completely unmarked feature of the language seems never to have been seriously questioned (except implicitly in statements on the status of phrasal verbs throughout the history of English, which, however, typically assume a rise of phrasal verbs sometime in the Middle English period; cf. the discussion in Chapter 4 above). Any other view would be decidedly in need of explanation in the light of the syntactic, lexical and comparative evidence, and indeed such a view has, to my knowledge, never been seriously voiced. Since there is no reason to assume a ‘rise of the phrasal verb’ in the Middle English period (except in the sense that the English particle verbs are subject to syntactic and semantic changes, which is clearly not what is implied by the traditional metaphorical characterisations of the development), the traditional attitude regarding the phrasal verbs as colloquial is, then, confronted with the question of when the phrasal verbs did become colloquial (Thim 2007). But this question has hardly ever been asked, let alone been answered. Rather, it seems that the colloquiality in historical texts is simply taken for granted, even if the evidence does not support that assumption. This is particularly true for the analyses of Early Modern English corpora by Hiltunen (1994), Claridge (2000) and Blake (2002); for a more detailed discussion of these three analyses, see Thim (2006a: 298–300). Hiltunen (1994) in his study of phrasal verbs in the Helsinki Corpus (Early Modern English part, 1500–1700) finds a rather unexpected distribution of phrasal verbs in the eight different text types of the corpus and concludes that the phrasal verb “can be employed for a variety of purposes beside that of conveying overt informality” (Hiltunen 1994: 139). But Hiltunen does not provide a single example of the use of a phrasal verb as a marker of overt informality, nor does the distribution of phrasal verbs across the text types necessarily imply such a connection. The distribution is as follows (100% = all phrasal verbs in the corpus): official correspondence 2.9%, statutes 7.2%, sermons 9.7%, private letters 11.5%, trials 12%, Bible 17.2%, fiction 18%, handbooks 21.5% (Hiltunen 1994: 135–139). Although the sermons, private letters and trials are characterized as “basically very interactive in character” (Hiltunen 1994: 137), there are considerably more phrasal verbs in the Bible, in fiction and in handbooks. Perhaps the frequency in the Bible may be attributed to Tyndale’s (and subsequent translators’) preference for the “native idiom” (Hiltunen 1994: 137); but cf. Lutz (2002b), where it is shown that, in particular in com-

Style and attitudes

217

parison to Luther’s German translation of the Bible, Tyndale actually uses a high number of Latinate loans, which can be regarded as evidence for the higher degree of integration of non-Germanic lexical items in English. But no ready stylistic explanation for the high frequencies in fiction and in handbooks can be offered, nor for the relatively low frequencies in private letters and in sermons. Similar observations apply to Claridge’s (2000) data. Although she mentions the possibility that the use of phrasal verbs in the Lampeter Corpus may not be connected to informality at all (cf. e.g. Claridge 2000: 103– 104 or 172, fn. 4) and although her discussion of the corpus findings does not make such a connection very likely, she does not conclude that there is simply no evidence in the corpus for a particular informality of the phrasal verbs. The frequencies are not connected to a stylistic evaluation of individual texts, but to ad hoc explanations as to why texts with higher frequencies might be more colloquial (cf. Thim 2006a: 299): the lowest number of phrasal verbs is found in the texts labelled ‘Politics’, the highest number in the texts labelled ‘Economy’. Six Royal Society authors writing on ‘Science’ show widely divergent frequencies, from very low to extremely high. In sermons the proportion of phrasal verbs is higher than in other religious texts, but lower than in ‘Economy’, ‘Law’, ‘Miscellaneous’, and about as high as in ‘Science’. In two texts containing ‘real’ dialogue (printed tracts containing transcripts of court proceedings), it is higher than in ‘fictitious’ dialogue. Women seem to use fewer phrasal verbs than men, but no clear correlation between the authors’ educational/social background and the use of phrasal verbs can be established. Claridge’s conclusion (2000: 197) that phrasal verbs in the corpus are typical of colloquial language and of texts directed at a wider audience is therefore hard to accept, and the same would seem to go for Smitterberg’s (2008: 276–281) suggestions for the remarkably uneven distribution of phrasal verbs in various 19th-century genres. Blake (2002) in his discussion of phrasal verbs in Shakespeare’s English notes the use of phrasal verbs in “elaborately formal poetry” (2002: 32) and labels them as informal in the very same sentence. A comparison between the use of phrasal verbs in the Quarto and First Folio editions of Richard III yields a number of instances where one edition has a simple verb while the other has a phrasal verb or where both editions have different phrasal verbs. Some of the examples Blake (2002: 34–35) quotes are problematic as instances of phrasal verbs, since he uses the term in a rather loosely-defined way. Nevertheless it is clear that the general impression holds true that there are some significant changes with respect to phrasal verbs between the two editions. But while he ascribes changes from simple verb to phrasal

218


verb between Quarto and Folio to the informality of phrasal verbs, he takes the considerable number of instances where the verb element is changed but the particle remains as evidence that the particle can carry more ‘semantic weight’ than the verb. With regard to the question of foreign influence on the genesis of the phrasal verbs, Blake argues that what he sees as their colloquial origins may speak against the possibility of foreign influence. But this argument is based on the assumption that phrasal verbs in Early Modern English must have been colloquial and it extrapolates from that assumed colloquiality in Early Modern English to the situation 500 years earlier, in early Middle English. To sum up, I suggest that in previous studies little conclusive evidence has been offered that phrasal verbs have always been colloquial. It must now be asked whether any actual contemporary evidence for the colloquiality of the phrasal verbs in the older periods of the language can be found. 6.2.2. Pre-1800 evidence for colloquiality? None of the studies on phrasal verbs in Early Modern English discussed so far provides any evidence for explicit contemporary views on the phrasal verbs as being informal or colloquial, despite the fact that from the 16th century onwards there is a considerable number of contemporary discussions of the status of the English language. All through the 16th and the 17th century phrasal verbs play practically no overt role at all in stylistic considerations, as their complete absence in the 16th- and 17th-century sources available in Jones (1953) attests; cf. also (especially for the 17th to the 18th century) the discussions by Söderlind (1964), McIntosh (1986), Adamson (1998 and 1999) and Görlach (1999), and see Cohen (1977) and Michael (1970) for an overview of English grammars before 1800. Historical attitudes towards phrasal verbs have been investigated in particular in Hiltunen (1983b), Brinton (1996) and Claridge (2000: Ch. 8). The present discussion of the grammatical tradition draws on these studies and on the material available in the Dictionary of English Normative Grammar 1700–1800 (Sundby, Bjørge & Haugland 1991) and in Knappe (2004); on the study of English grammars and usage between the late 18th and the 20th century, see Finnegan (1998). Attitudes towards phrasal verbs in the grammatical and the lexicographic tradition of the late 18th and the 19th century were almost entirely neglected until recently (Thim 2006c and 2008b and Wild 2008).

Style and attitudes

219

From the 18th century onwards phrasal verbs get noticed by grammarians and lexicographers, but there is little reason to assume that they were subject to negative prescriptivist attitudes much before the end of the 18th century. This is for example shown in Hiltunen (1983b), an overview of the treatment of phrasal verbs in a representative selection of 43 grammars published between 1586 and 1839. Hiltunen has surveyed two grammars from the end of the 16th century, eleven from the 17th century and 26 from the 18th century (cf. the full list in Hiltunen 1983b: Appendix). Claridge (2000: Ch. 8) has surveyed a similar selection of grammars and a number of dictionaries and textbooks, and her results by and large confirm Hiltunen’s. Interestingly, both have found a considerable number of instances where phrasal verbs are used in these works, e.g. in definitions in the dictionaries or in sample sentences in the grammars, but never in the context of prescriptive verdicts against the construction. Especially for the 17th century the evidence turns out to be largely negative, and Hiltunen concludes: The Latin tradition must have impeded many of them from taking up constructions so alien to Latin in their accounts of English. At best one may speak of a latent awareness of the existence of the English phrasal constructions in the works of the sixteenth and seventeenth-century writers. In some cases (e.g. Wallis [1653] and Miège [1688]), this awareness materializes when English and Latin are contrasted. Miège’s grammar is the earliest of those examined which specifically draws attention to the structural differences between the two languages in the use of ‘prepositions’, recognizing the analytical constructions as single units. (Hiltunen 1983b: 378–379)

It is only in the 18th century that phrasal verbs begin to be treated more regularly as syntactically and semantically distinct from free combinations of verbs and adverbs, as e.g. the grammars by Mattaire (1712), Lowth (1762) and Wiseman (1767) testify. But still the discussion of phrasal verbs (in whatever context) is not the norm, as many other grammars from throughout the 18th century show, where they are neglected (cf. Hiltunen 1983b: 379). Mattaire appears particularly advanced in his insightful comments on the lexical and syntactic characteristics of phrasal verbs, as the following passage from his grammar shows (see Section 6.3 below for a discussion of the linguistic terminology employed by Mattaire and Lowth in this and the following example and its implications for their linguistic analyses): The English prepositions may compound words by being put after, without governing a word; as to go on, to go out, to run in … The Particle, which com-

220


pounds the verb by following it, does not always go next to the verb; but the Noun, which is governed by the verb, is often placed between … The pronoun ever goes between. (Mattaire 1712: 110–111)

Indeed the description of the syntactic and semantic properties of the phrasal verb offered by Mattaire remains unsurpassed for more than a century to come. But neither Mattaire nor any other grammarian of the 18th century makes any explicit statement on the stylistic value of the phrasal verb. For example, fifty years after Mattaire, Robert Lowth, the most influential English grammarian of the 18th century, describes the phrasal verb as follows: Verbs are often compounded of a Verb and a Preposition … and this composition sometimes gives a new sense to the Verb … But in English the Preposition is more frequently placed after the Verb, and separate from it like an Adverb; in which situation it is no less apt to affect the sense of it, and to give it a new meaning; and may still be considered as belonging to the Verb, and as a part of it. As, to cast, is to throw: but to cast up, or to compute, an account, is quite a different thing … So that the meaning of the Verb, and the propriety of the phrase, depend on the Preposition subjoyned. (Lowth 1762: 128–129)

As is well known, Lowth served as the prescriptive bogeyman of modern descriptive linguistics for a very long time; for a critical discussion of this practice, see in particular Tieken-Boon van Ostade (2006a) and her discussion of Aitchison’s (1981: 23–24) stereotypical account of Lowth as a complacent eccentric. Such an account shows little familiarity with the history of normative grammar in the 18th century; this is typical of the “persistent prejudice and preconceptions concerning the aims and intentions of the normative grammarians” (Tieken-Boon van Ostade 2006a: 552). Cohen (1977: 84–85) had already pointed out that “Lowth is prescriptive only with regard to his insistence on following what he determines to be established (and approved) usage, and he repeatedly acknowledges different customs among existing languages” (see also Tieken-Boon van Ostade 2006b for a brief overview of the beginnings of the English normative tradition in the 18th century). But clearly Lowth does not proscribe the use of phrasal verbs; in fact, he does not even comment on their stylistic value at all. Tieken-Boon van Ostade (2006a: 552) shows that “Lowth’s grammar was so popular because it gave its readers exactly what they were looking for, a respected norm to imitate if they wished to improve their position in society”. The absence of any kind of proscription of the construction in Lowth’s grammar can hence be taken as an indication that around the middle of the 18th century phrasal verbs were not yet associated

Style and attitudes

221

with colloquiality or informality. Clearly, phrasal verbs were not evaluated at all. The examples from various other grammars quoted by Hiltunen (1983b: 379–383), where phrasal verbs are described with varying degrees of linguistic precision, but never with the slightest hint of stylistic censure, show that Lowth’s treatment is quite typical. Thus Hiltunen’s conclusion that the 18th-century grammarians’ normative approach also entailed the stigmatization of phrasal verbs because of their very difference from the Latin model (1983b: 384–389) is surprising; at least it cannot be substantiated with actual evidence from 18th-century grammars. The same is true of Brinton’s (1996: 189) claim, based on the discussion in Hiltunen (1983b), that phrasal verbs “seem to have evoked the objections of traditionalists as early as the eighteenth century”. What Brinton means by referring to traditionalists is not clear, and there appear to be no explicit objections from anyone at that time (but cf. Section 6.3 below for a discussion of possible indirect consequences of normative verdicts on the phrasal verbs). 6.2.3. An example: Samuel Johnson In the lexicographical tradition it is also in the 18th century that phrasal verbs are first systematically noticed, and again they are described, but by no means proscribed. The best-known lexicographic comment from the 18th century is from the Preface to Samuel Johnson’s dictionary (1755; cf. the quote at the beginning of Chapter 1) and Johnson has sometimes been brought forward as a witness for negative attitudes towards phrasal verbs in the 18th century, e.g.: There are many examples of a negative attitude towards the use of phrasal verbs, see e.g. Tucker [1961], p. 83, where James Harris’ Philological Inquiries in Three Parts (1781, pp. 105–6) is quoted. The great lexicographers, Dr Johnson and Noah Webster, were also strongly against phrasal verbs, although both of them listed many such combinations in their Dictionaries. But it has to be kept in mind that phrasal verbs were more of a novelty in the fifteenth, sixteenth and seventeenth centuries than they are today. Still, the negative attitude towards them has not altogether vanished. (Hiltunen 1983b: 384–385) [Phrasal verbs are] an area which, for many writers, marks an expert knowledge of English, and which has therefore attracted considerable interest down the centuries. According to Samuel Johnson in the 18th century (Rambler No. 203), for example, it is an area where even native speakers produce ‘colloquial barbarisms’, ‘licentious idioms’ and ‘irregular combinations’ of which he has ‘laboured to clean the language’. (Ahulu 1995: 29)

222


We shall see shortly that there is little evidence at all for claims like the ones given in these two quotations. But unfounded beliefs about Johnson’s negative attitude towards phrasal verbs appear to have been fairly widespread for some time. Thus we find Jowett (1951) in a short popular article on phrasal verbs in present-day English state as follows: [Phrasal verbs] are the speech of the man in the street rather than of the pedant – the language of ‘every Tom, Dick and Harry’ rather than that of Thomas, Richard and Henry. As one might expect, such collocations were not liked by Samuel Johnson, who thought them vulgar. The great lexicographer hated such verbs as bind up, bring in, look on, and went so far as to declare to come by (to obtain) to be an irregular and improper use, though finding this phrase in Hooker, Shakespeare, Bacon, and Dryden. Johnson, of course, had scant respect for the linguistic standards of Tom, Dick, and Harry. An educated speaker of English is, however, constantly required to choose the level at which he shall speak, and to suit language to context and occasion. It is much more natural to say ‘He came into the room, picked up a book, looked at it casually, put it down, and went out’ than ‘He entered the room, seized a book, examined it, discarded it and departed’ – indeed, in the latter sentence two of the verbs used, ‘seized’ and ‘discarded’, are inappropriate but are yet more suitable than any other single verb which I have been able to substitute for them. (Jowett 1951: 154)

For one thing, this passage is yet another example how 18th-century authors can be made responsible at will for any kind of prescriptivist attitude, but it is quite instructive for a number of reasons; incidentally, although Johnson does quote examples from the authors mentioned by Jowett s.v. come by, the negative verdict on the construction is mysteriously absent from the entry (cf. also the discussion of the entry below). In particular, the passage is symptomatic of the author’s dilemma to decide whether the phrasal verbs are proof of “the virility and versatility of the English verb” and “one of the glories of our tongue”, or whether they have “yet to be subjected to a severe testing process before being finally accepted or rejected by the language” (Jowett 1951: 153 and 156). In the context of the present section, such statements seem even more relevant because they imply that the 20th-century perception of phrasal verbs as colloquial (or stylistically marked in a similar way) has a considerable historical background, and that there is evidence for the colloquial status of verb-particle constructions after the mid-18th century and for the disapproval they met from an author as eminent and influential as Johnson, but also from other commentators. None of this is supported by actual contemporary evidence, though. In the verdict alluded to by Hiltunen (1983b: 384–385) in the quote above,

Style and attitudes

223

Harris (1781) disapproves of monosyllables, not of phrasal verbs (cf. the discussion in Section 6.3 below). The passage from Johnson which Ahulu (1995: 29; cf. the quotation at the beginning of the present section) refers to reads as follows: Whatever shall be the final sentence of mankind, I have at least endeavoured to deserve their kindness. I have laboured to refine our language to grammatical purity, and to clear it from colloquial barbarisms, licentious idioms, and irregular combinations. Something, perhaps, I have added to the harmony of its cadence. When common words were less pleasing to the ear, or less distinct in their signification, I have familiarized the terms of philosophy by applying them to popular ideas, but have rarely admitted any word not authorized by former writers; for I believe that whoever knows the English tongue in its present extent, will be able to express his thoughts without further help from other nations. (Samuel Johnson, The Rambler, no. 208, 1752 [Bate & Strauss 1969: 318–319])

It is quite impossible to find anything here that would permit the establishment of an explicit connection to phrasal verbs; neither is there anything in the rest of the essay that would suggest such a connection. Note also that the quote in question is from number 208 of the Rambler, not from number 203, and that Ahulu’s statement in its turn appears to be based on a misquotation of Johnson by Konishi (1956: 128). Johnson’s linguistic practice does not allow us to draw such conclusions either. No. 203 of the Rambler is not at all concerned with linguistic issues. It is a short text of slightly less than 1,300 words, which is concerned with happiness and the transience of life. It contains the following phrasal verbs (quoted after the critical edition by Bate & Strauss 1969): (1)

…when time has abated the confidence with which youth rushes out to take possession of the world, we endeavour, or wish, to find entertainment in the review of life …

(2)

No man past the middle point of life can sit down to feast upon the pleasures of youth without finding the banquet embittered by the cup of sorrow …

(3)

… he may enjoy the nobler pleasure of looking back upon distress firmly supported …

(4)

… with whom he had planned out amusements for his later years …

(5)

… we must soon lie down in the grave with the forgotten multitudes of former ages …

224


(6)

… who, like us, shall be driven away …

(7)

They whose souls are so chained down to coffers and tenements …

(8)

When this vacuity is filled up …

Thus it is quite clear that Johnson’s linguistic practice does not preclude the use of phrasal verbs. Of course there can be a difference between an author’s practice and his normative advice. But again, Johnson’s attitude towards phrasal verbs seems to be essentially descriptive, as the quote from the Preface to his dictionary shows (cf. Chapter 1 above), and there is no indication of a particular stylistic evaluation of phrasal verbs. An analysis of a characteristic dictionary entry shows that they are in fact treated accordingly. S.v. come there are the following sub-entries for phrasal verbs in Johnson’s dictionary: 16. To COME about. To come to pass; to fall out; to come into being. Probably from the French venir a bout … 17. To COME about. To change; to come round … 22. To COME in. To comply; to yield; to hold out no longer … 24. To COME in. To arrive at a port, or place of rendezvous … 25. To COME in. To become modish; to be brought into use … 26. To COME in. To be an ingredient; to make part of a composition … 27. To COME in for. To be early enough to obtain: taken from hunting, where the dogs that are slow get nothing … 28. To COME in to. To join with; to bring help … 29. To COME in to. To comply with; to agree to … 33. To COME off. To deviate; to depart from a rule or direction … 34. To COME off. To escape … 35. To COME off. To end an affair; to be dismissed with our lot … 36. To COME of from. To leave; to forbear … 37. To COME on. To advance; to make progress … 38. To COME on. To advance to combat … 39. To COME on. To thrive; to grow big … 40. To COME over. To repeat an act … 41. To COME over. To revolt … 42. To COME over. To rise in distillation … 43. To COME out. To be made publick … 44. To COME out. To appear upon trial; to be discovered … 45. To COME out with. To give a vent to; to let fly … 46. To COME to. To consent or yield … 50. To COME up. To grow out of the ground … 51. To COME up. To make appearance … 52. To COME up. To come into use, as a fashion comes up. 53. To COME up. To amount to … 54. To COME up to. To rise to … 55. To COME up with. To overtake.

Although this is a long entry (the relevant sub-entries are quoted here in full except that the sample quotations are left out), it is a representative one with respect to phrasal verbs: in Johnson’s dictionary, phrasal verbs are hardly ever marked as stylistically ‘low’ or ‘informal’ and in this respect Johnson’s dictionary is similar to earlier monolingual English dictionaries. Among the more than 1,000 ‘branded words’ in English dictionaries before

Style and attitudes

225

Johnson which Osselton (1958) discusses, there are eight phrasal verbs ‘branded’ as ‘low’, ‘colloquial’, etc., but there is a considerably greater number of simple words where this applies. Likewise, in her analysis of Johnson’s dictionary, Knappe (2004) counts 1,663 ‘phraseological units’, of which only 63 have a usage label at all. This corresponds almost exactly to the percentage of usage labels for non-phraseological entries, and there is only a very small number of individual phrasal verbs that are criticized, while the use of phrasal verbs in the definitions is also a typical feature of the whole dictionary. See e.g. Knappe (2004: 429–450) for a discussion of the copious literature on Johnson’s dictionary; Knappe also provides a helpful overview of other pre-1800 lexicographic approaches to phraseology including, in the classification of her study, phrasal verbs. For the exceptionally full coverage of phrasal verbs and other types of word formation in Johnson’s dictionary compared to earlier dictionaries, cf. Stein (1984) and Osselton (1995); a brief account of the development of usage labels in English dictionaries is provided by Cassidy (1997). In fact, Osselton (1986) has shown that Johnson’s dictionary is the first monolingual English dictionary which treats phrasal verbs systematically and that for the selection of phrasal verbs Johnson relies on a range of earlier bilingual dictionaries, where phrasal verbs were treated much more extensively than in contemporary monolingual dictionaries. The claim that Johnson disapproved of phrasal verbs cannot be upheld. On the whole, Samuel Johnson’s attitude, his usage, and his lexicographic practice do not justify the claim that he was “strongly against phrasal verbs”, as suggested by Hiltunen (1983b) in the quote above. On the contrary, there are strong indications that Johnson uses and treats them as unmarked elements of the English lexicon; the same is true of Webster. This has been discussed in more detail by Osselton (1986), Claridge (2000: 208–210), Knappe (2004) and Thim (2007). Wild (2008) provided an extensive analysis of phrasal verbs labelled ‘colloquial, familiar, popular’ and phrasal verbs with proscriptive labels in Johnson (1755), Webster (1828) and OED1. Moreover, the case of Johnson is representative, both of the 18th century practice and of its misrepresentation in the linguistic literature in regard to the phrasal verbs. There is no contemporary opposition to the phrasal verb as a construction type in the 18th century, neither in the dictionaries nor in the grammars or in other accounts of the language. As a look through the material available in Sundby, Bjørge & Haugland (1991) proves, there is no opposition to the phrasal verbs in principle, but only against individual phrasal verbs. Indeed, where 18th century grammarians describe phrasal verbs, they tend to do so with a completely neutral attitude.

226


6.2.4. James Cook rewritten and John Dryden revised Why should the examples from literary language discussed here be relevant with regard to the development in Standard English in general, one might ask. From the Early Modern period onwards, both are inextricably connected: The rise of a national Standard language in the period 1476–1776 … had its literary counterpart in the formation of a national literature, embodied in the works of those whom influential opinion identified as the nation’s ‘best authors’. Indeed, the codifying of language and the canonising of literature were not merely simultaneous but symbiotic processes, with the ‘best authors’ being quarried for instructive examples as much by grammarians and language teachers as by rhetoricians and literary critics. (Adamson 1999: 539)

A discussion of what kind of changes were introduced with respect to phrasal verbs by two style-conscious authors will cast further light on contemporary attitudes and practices. One text is from the second half of the 17th century, the other from the second half of the 18th century, and I will discuss the two texts in reverse chronological order. McIntosh (1986: 105–114) compares the manuscript of James Cook’s journal of the voyage of the Endeavour to Tahiti and New Zealand (1770), which was probably never intended for publication, to John Hawkesworth’s published version of Cook’s journal from 1773 (quoted in McIntosh 1986 after Beaglehole 1955). Hawkesworth, a friend of Johnson’s, had received a considerably better education than Cook, who had risen from lowly origins (cf. McIntosh 1986: 105–106). Fanny Burney observed about Hawkesworth’s conversation, “He has an amazing flow of choice of words & expressions … All he says is just,–proper, & better express’d than most written language” (quoted in Tieken-Boon van Ostade 2006b: 258). McIntosh (1986: 106) finds that Hawkesworth “never meets lower-class language in Cook but he removes it”; the following two parallel passages are among the examples she provides: a. Cook’s manuscript: one need hardly wish for a better was the Access to it from the Eastward less dangerous, but this difficulty will remain untill some better way is found out than the one we came, which no doubt may be done was it ever to become an object to be look’d for b. Hawkesworth’s version: better would not need to be desired, if the access to it, from the eastward, were less dangerous: that a less dangerous access may be discovered, I think there is

Style and attitudes

227

little reason to doubt, and to find it little more seems to be necessary, than to determine

It is obvious that Hawkesworth rewrites Cook’s manuscript version extensively, and a comprehensive discussion of his rewrite would go far beyond the scope of the present discussion (cf. e.g. the use of passives in the manuscript). But one of the changes Hawkesworth makes affects the phrasal verb find out, which he replaces by discover. How can this replacement of a phrasal verb be explained in the light of my claim that there is no evidence for colloquiality of the phrasal verbs, and in view of the absence of prescriptive censure directed against them? Claridge (2000), who also discusses this example from McIntosh (1986), notes that the replacement favours Latinate over Germanic (‘native’) vocabulary, and interprets this as evidence that phrasal verbs “tended to be identified with more colloquial and less prestigious styles” (Claridge 2000: 219). But McIntosh, in her discussion of this passage, had already offered a different analysis of the rewritten version: Gone is the old-fashioned conditional, in its place an if clause with the correct form of the verb, the subjunctive. Gone is the unspecific and possibly inaccurate abstraction “difficulty”, and the common verb found gives way to the slightly more genteel discovered. Gone is “the [way] we came” – to come a way is a bourgeois idiom at best and could be accused of redundancy … Gone is the ambiguous which referring to an entire event, not to a single substantive; gone is the awkward repetition of “be done … become … be look’d for”. (McIntosh 1986: 106; her square brackets)

That is to say, Hawkesworth has produced a formal text for publication, which meets the contemporary stylistic requirements on every linguistic level from the lexicon to textual structure, and McIntosh thus concludes: Without exception, so far as I know, the genteel John Hawkesworth corrects or omits all solecisms and archaisms and colloquialisms he finds in journals by the lower-middle-class James Cook. (McIntosh 1986: 114)

To some extent this seems to involve the preference of Latinate over native vocabulary. McIntosh points out that what is affected here is the simple verb find because of its status as a ‘native’ word, i.e. what is favoured is not the simple over the phrasal verb, but rather the Latinate over the native verb. It must be added, though, that McIntosh’s explanation is not entirely satisfactory, since Hawkesworth re-introduces the simple verb find almost immediately afterwards (“… and to find it little more seems to be necessary…”). But still the passage cannot be taken as evidence for

228


Hawkesworth avoiding phrasal verbs, as another passage shows (quoted from McIntosh 1986: 107): a. Cook’s manuscript: Another custom they have that is disagreeable to Europeans which is eating lice a pretty good stock of it which they generally carry about them. b. Hawkesworth’s version: as they live in a hot country, and have no such thing as a comb, they are not able to keep their heads free from lice, which the children and common people sometimes pick out and eat: a hateful custom, wholly different from their manners in every particular.

Here Hawkesworth even inserts phrasal verbs in the course of his extensive rewriting of the manuscript (pick out; and cf. similarly Hawkesworth using take away not in Cook’s manuscript in another passage quoted by McIntosh 1986: 109). Evidently he is not averse to phrasal verbs in general. It seems more plausible to suggest that the introduction of discover (in combination with the Latinate noun access replacing way) may indeed be part of a deliberate Latinisation of the text, as the other replacements indicate.5 But this does not necessarily go along with the avoidance of Germanic vocabulary, as the use of the synonymous find immediately afterwards shows, while the omission of the particle – unless fortuitous – may in this case be explained as the avoidance of redundancy (cf. the more detailed discussion of this factor in Section 6.3 below), but certainly not as the avoidance of phrasal verbs, as the use of other phrasal verbs in the revised text show. The characterization of phrasal verbs as ‘colloquial’ or ‘informal’ on the basis of this or similar examples is thus rather problematic, particularly in view of the negative evidence from prescriptive writings and the inconclusive evidence from the quantitative analysis of larger corpora. It seems that such characterizations address the issue from the wrong perspective. What is at stake here is stylistic value of the Latinate borrow5

As Andrew James Johnston points out to me (p.c.), it is conceivable that the choice of discover is ideologically motivated and part of Hawkesworth’s goal of celebrating Cook as a deliberate discoverer. Seen from this perspective, the reason for the substitution of find out is to be found in the different prestige value that is connotatively associated with discover. I find this suggestion very appealing. In addition, the use of discover also establishes a link to the title of the book (The journals of Captain James Cook on his voyages of discovery), which can easily be imagined to be a strong reason for Hawkesworth to introduce that verb.

Style and attitudes

229

ings in English as opposed to the ‘native’ vocabulary, whose “different and complementary expressive properties” have been described by Adamson (1999) as follows: What Cheke and other mid-sixteenth-century purists perhaps did achieve by resisting the latinate invasion and defending the dignity of native ‘Saxon’ English was to develop a general awareness of the etymological origin of words and an appreciation that the Saxon and latinate elements in the word-stock had different and complementary expressive properties … these properties relate directly to the conditions in which the two layers of the lexicon are learned. Saxon words are typically learned early, learned through speech and in the context of physical experience. Hence no one needs to be told the meaning of light or strong; they consult their memories of all the experiences with which the world is connected. Words like illumine or energial, by contrast, are learned late, learned through education and interpreted by reference to explicit definition. They are therefore associated not only with a formal, public style but also with a range of meaning that is primarily abstract and ideational, whereas Saxon words are associated with private and intimate discourse and their semantic range is characteristically experiential. (Adamson 1999: 573)

For the tendency to prefer Latinate vocabulary in certain contexts, for which there is ample evidence throughout the Modern English period, see e.g. the discussions in Nevalainen (1999), Görlach (1999) and Adamson (1999), or the overview of contemporary attitudes and practices in Görlach (1994) and Barber (1997). Crucially, the phrasal verb is affected by this development only indirectly, since the majority of phrasal verbs contain simple verbs of Germanic origin, while the compositional combinations of verbs of motion and directional particles are characteristically not abstract and ideational. Interestingly, this had already been tentatively suggested by Hiltunen (1994) in his discussion of the distribution of phrasal verbs in the Early Modern English part of the Helsinki Corpus, when he proposed that “[t]he low frequency [in statutes and official letters] may indicate a preference for Latin-based verbs in the official documents generally (although this is not shown by the present data)” (Hiltunen 1994: 136), while he observed a “tendency to employ phrasal verbs when the action denoted by the combination is itself in the foreground” (Hiltunen 1994: 139). This ties in with my detailed analysis of the use of phrasal verbs in a small corpus of everyday English texts from 1500 to 1700 (Thim 2006a, cf. Section 6.1.3 above), where it is shown in detail that the choice of phrasal verbs depends to a considerable degree on the topic (and not on the style) of the texts and the semantics of the phrasal verbs: phrasal verbs tend to be particularly frequent in descriptions of spatial events (where there is hardly ever a lexical

230


alternative), while there is no observable tendency to avoid phrasal verbs in more formal texts, although they tend to be rarer in more abstract types of discourse. But even in heavily Latinized texts there is an enduring use of nonLatinate vocabulary. Adamson (1999) explains the context of this as follows: Perhaps because the grand style was so clearly defined in functional rather than formal terms and because its function was so clearly understood to be persuasion or moving, most renaissance writers ground the magniloquent latinate in the homely Saxon. In a trivial sense, they have no choice: since the closed class words of English … have remained almost exclusively Germanic, even the most ardent Latiniser is bound to produce a hybrid text. It is only in the open class … that significant choice can be made and at this level, from the midsixteenth century onwards, the norm for the grand style is to interweave latinate and Saxon … Styles which, by contrast, collocate latinate with latinate … tend to appear in parodies rather than instances of the grand style, as the marker of an ‘affectate’ discourse or a speaker out of touch with reality. (Adamson 1999: 574)

In the light of this analysis, we can now address the changes made by John Dryden in the subsequent editions of his Essay of Dramatick Poesie (1st edn. ‘Q1’ from 1668, 2nd edn. ‘Q2’ from 1684; cf. Monk et al. 1971). It has been claimed repeatedly that Dryden avoided phrasal verbs, and the changes in the Essay have been taken as the main evidence for that claim. Indeed, for the second edition (Q2, 1684), Dryden himself undertook a thorough revision of the text, which Ker (1900: I.xxxv) characterizes as Dryden’s “most elaborate piece of criticism, and the most careful of his prose works” (cf. the Textual Notes in Monk et al. 1971 for further details). According to Bately (1964: 268), “[t]hese corrections are of considerable interest to the linguist and emphasize Dryden’s consciousness of language, more particularly ‘correct’ language”; cf. also Simon (1963). An analysis of the changes is therefore particularly instructive with regard to late 17thcentury norms of correctness. McIntosh characterizes the sociolinguistic implications of the revision of the Essay as follows: It was written … for an upper-class readership, to defend English literary culture. Dryden composed and revised it with fastidious care. The effect of about 80 percent of his revisions for the second edition of 1684 is to make his language less colloquial, more correct, or more up-to-date … it shows us a writer actively distancing himself from lower-class norms, and so provides an opposite end to what then becomes a sociolinguistic spectrum of language. (McIntosh 1986: 12)

Style and attitudes

231

As we will see, the differences in phrasal verbs between the first and the second edition of the Essay can be seen as part of his conscious attempt at producing a text in a ‘heightened’ style rather than as evidence for avoidance of phrasal verbs altogether. And indeed, the claim that Dryden avoids phrasal verbs does not become true through continuous repetition in the linguistic literature. This has become a bit of a chimaera in the literature on phrasal verbs. E.g. Claridge (2000: 1999) quotes Von Schon (1977: 37) as evidence for Dryden replacing phrasal verbs by Latinate synonyms in his manuscripts, while this claim made by Von Schon (“Dryden repeatedly crossed out a phrasal verb in his manuscripts in order to replace it with a Latinate synonym”) in its turn rests on Konishi (1958: 122), where, however, it is merely observed that Dryden replaced two phrasal verbs by Latinate synonyms in the reprint of one essay. A look through the subsequent editions of the Essay reveals that there are four instances where Dryden changed a phrasal verb. To render the comparisons easier, the passages in question are here in italics. Consequently Dryden’s italicisations, mostly proper nouns, are printed here in roman type like the rest of the examples: (9)

But what will Lisideius say if they themselves acknowledge they are too strictly ti’d up by those lawes [Q1] / bounded by those laws [Q2], for breaking which he has blamed the English? (Dryden 1668/1684 [Monk et al. 1971: 51])

(10) … when by experience they had known how much we are bound up [Q1] / limited [Q2] and constrained by them … (Dryden 1668/1684 [Monk et al. 1971: 51]) (11) … but when e’re they endeavour to rise up [Q1] / rise [Q2] to any quick turns and counterturns of Plot … (Dryden 1668/1684 [Monk et al. 1971: 53]) (12) But when, by the inundation of the Goths and Vandals into Italy new Languages were brought in [Q1] / introduced [Q2], and barbarously mingled with the Latine … (Dryden 1668/1684 [Monk et al. 1971: 70]) Dryden replaces two phrasal verbs by simple Latinate verbs in (10) and (12), and he replaces one phrasal verb with a more precise (and less metaphorical) simple verb in (9). Although the verb bound is of Romance origin (cf. OED s.v.) this can hardly have been clear to Dryden – the verb could just as well be Germanic. In the fourth instance he omits the particle but leaves the verb as it stands. As will be argued below, the two latter

232


substitutions are likely to be connected to Dryden’s pursuit of perspicuity (cf. the discussion in the next section). But Dryden also kept phrasal verbs even in sentences which he otherwise changed. Thus we find substitutions like the following, where the sentence is rendered more explicit, while the phrasal verb is kept: (13) … which in reason might render him more wary another time, and make him punish himself with harder fare and courser cloaths to get it up again [Q1] / to get up again what he had lost … [Q2] (Dryden 1668/1684 [Monk et al. 1971: 43]) In the 105 lines of the dedicatory epistle alone, there are four phrasal verbs, which Dryden kept in all editions (cf. Monk et al. 1971: 3–6; this line count includes only Dryden’s text and not the long quotations from other authors in the text): (14) Seeing then our Theaters shut up, I was engag’d in these kind of thoughts with he same delight with which men think upon their absent Mistresses (Dryden 1668–84 [Monk et al. 1971: 3]) (15) It was an honour which seem’d to wait for you, to lead out a new Colony of Writers from the Mother Nation … (Dryden 1668–84 [Monk et al. 1971: 4]) (16) Sometimes, like a Schollaer in A Fencing-Scool I put forth my self … (Dryden 1668–84 [Monk et al. 1971: 5]) (17) And yet, my Lord, this war of opinions, you well know, has fallen out among the Writers of all Ages … (Dryden 1668–84 [Monk et al. 1971: 5]) Similarly, in the Essay proper there is no indication for Dryden avoiding phrasal verbs, cf. the following two examples from the beginning (again in all editions): (18) after which, having disingag’d themselves from many Vessels which rode at Anchor in the Thames, and almost blockt up the passage towards Greenwich, they order’d the Watermen to let fall their Oares more gently (Dryden 1668–84 [Monk et al. 1971: 8]) (19) Eugenius lifting up his head, and taking notice of it, was the first who congratulated to the rest that happy Omen of our Nations Victory (Dryden 1668–84 [Monk et al. 1971: 9])

Style and attitudes

233

It is particularly noteworthy that Dryden leaves lift up in (19) unchanged, although he did change the similarly pleonastic rise up in (11). And Dryden continues using phrasal verbs throughout the Essay (e.g. cry out, grow up, run out, set out, come in, etc.). Thus, once again, there is but little evidence for the avoidance of phrasal verbs in Dryden’s Essay and, by implication, no evidence for the ‘colloquiality’ of the phrasal verbs. Rather, the development of the phrasal verbs in Modern English should be seen in the context of other developments, mainly such as are connected to prescriptive attitudes and stylistic ideals, and that any discussion which looks at the phrasal verbs in isolation is doomed to failure. 6.3.

The colloquialization conspiracy: a first suggestion

I should now like to suggest how the stylistic views of phrasal verbs current in the 20th century may have come about. This suggestion will allow us to avoid the inconsistencies and paradoxes we have encountered in the discussion of so many of the traditional accounts of the history of the phrasal verb in Modern English. Clearly, different factors have played a role in the development, which can therefore be characterized as a colloquialization conspiracy. Not least because the prescriptive attitudes of the 19th century are as yet fairly unchartered ground and far from being easily accessible, parts of the present sketch will remain to be completed in the future, though. Attitudes towards the vernacular in the Early Modern period have long been extensively documented and the prescriptive tradition of the 18th century has been subject to intensive research over the last years, but the 19th-century sources are as yet difficult to access (cf. Görlach 1998), and there is nothing comparable to e.g. Leonard (1929), Jones (1953), Sundby, Bjørge & Haugland (1991) or Knappe (2004); cf. also the bibliographical survey provided by Gneuss (1996). A concise overview of attitudes towards English in the 19th century is given in Mugglestone (2006). In the revised edition of Dryden’s Essay of Dramatick Poesie, the avoidance of stranded prepositions and the changes in prepositions, conjunctions and relative pronouns are much more notable than the omission of phrasal verbs. Bately (1964), in her discussion of Dryden’s usage and the changes in the second edition of the Essay, places these revisions in a wider context within the chronology of his works: After 1672 Dryden ceases to use ‘prepositions’ at the end of a sentence in prose (they are in fact not found after the Essay of Dramatic Poesy), but continues to use them at the end of a clause, exceptionally before a colon and in other

234


related positions. However, this usage too shows a sharp decrease after 1672. (Bately 1964: 271)

This observation ties in nicely with Dryden’s well-known attitude towards stranded prepositions, made explicit in his criticism of Ben Jonson as follows, in the “Defence of the Epilogue: Or, An Essay on the Dramatique Poetry of the last Age”, printed as an appendix to The Conquest of Granada (1672) (Loftis et al. 1978; cf. Knorrek 1938): The preposition in the end of the sentence; a common fault with him, and which I have but lately observ’d in my own writings. (Dryden 1672 [Loftis et al. 1978: 208])

In Section 6.2.2 it has already been noted that the particles of the phrasal verbs tend to be treated as prepositions in contemporary grammars. Even Mattaire (1712: 110–111), in his very precise description of the positional properties of the phrasal verbs, refers to the particles also as ‘prepositions’ (cf. the quotation in Section 6.2.2), and this is also done by Lowth in his discussion (cf. the quotation from Lowth 1762: 128–129 in Section 6.2.2 above). This is entirely in line with the usage in other contemporary grammars and remains current in the English grammatical tradition up until the 20th century; it is due to the absence of an established terminology for such elements in the Latin-based grammatical tradition (on that tradition, cf. the studies quoted above). In the Latin tradition praepositio is used as a cover term for both prepositions and prefixes, and thus even where the phrasal verbs are treated among the compounds – as by Mattaire and many others – the particles will consequently be classified as ‘prepositions’. In his discussion of the topic, Sundby (1995: 31–35 and 90–92) points out that those particles which are synonymous to adverbs only appear to have been subject to the same classification due to analogy: “By analogy with (e.g.) on, over, which are also used as preparticles but are still ‘prepositions’, that name is transferred to prefixes … For similar reasons, ‘particles’ (= prepositions and adverbs mainly) include prefixes” (Sundby 1995: 91); see also Michael (1970) and Vorlat (1975), where the history of the terminological tradition is treated in greater detail. The first critical observation involving phrasal verbs found so far needs to be seen in this context. Knappe (2004: 473 fn. 80) mentions a remark by Obadiah Walker from 1659 on “the infelicity of the English, by reason of Prepositions disjoyned from the Verb”; Walker suggests using Latin loans instead, e.g. import and intervene rather than bring in and come between. The remark is therefore not, as Knappe claims, one of the few “overt stylistic criticism[s] of phrasal verbs” (Knappe 2004: 473 fn. 80)


235

since it is neither a stylistic criticism, nor one of phrasal verbs. What Walker actually criticises is the fact that English has a particle and a preposition where the corresponding Latin verbs have prefixes; but since all three would be termed praepositiones in classical Latin grammar, it is inferred that the English ones are in the ‘wrong’ position. With respect to the phrasal verbs, this excessive reliance on the model of Latin appears to have had little impact on subsequent grammarians and is, as far as one can see, by and large without parallels. With respect to Dryden, this could offer another part of the explanation of his avoidance of some phrasal verbs in the revised/rewritten versions while other phrasal verbs are left unchanged or even introduced. Dryden does not seem to change a phrasal verb if the object follows the particle, i.e. in those cases where the sequence is identical to the ordinary sequence of ‘verb + preposition + noun’, as examples (15) to (19) in Section 6.2.4 above show. Cf. also Dryden (1668–84 [Monk et al. 11, 13, 23]): and helps out his numbers, we shall take up more time this Evening; but with more swiftness then it brought them on, etc. But Dryden does change the order with a pronominal object in (13), where he replaces the pronoun with a full noun phrase. He also tends to change the phrasal verb where it is not followed by an object, as in (9) to (12). That this is not done in an entirely consistent way, as seen in (14), ties in with Bately’s (1964) analysis of Dryden’s usage with respect to real stranded prepositions, which is not entirely consistent either. The avoidance of placing particles in sentence-final position, as a hypercorrect application of the normative rules against sentence-final prepositions, may well have played a part in the avoidance of phrasal verbs altogether, and it may also have had some indirect impact on the later association of phrasal verbs with ‘colloquiality’ (for a general discussion of the influence of normative notions on 17th- to 19th-century English see Auer 2006). For the time being this must remain a mere (if, to my mind, plausible) hypothesis to be tested on the basis of a considerably larger corpus. There is no denying that stranded prepositions continued to be a stylistic concern well into the 20th century, as a look into popular style guides from throughout the 20th century confirms. In a usage guide as recent as The New Fowler’s Modern English Usage (ed. Burchfield 1996), it is stated in the section on ‘history of attitudes’ that “the myth of the illegitimacy of deferring prepositions had clearly been destroyed by the end of the 19c.”, while in a ‘final verdict’ it is concluded that “[i]n most circumstances, esp. in formal writing, it is desirable to avoid placing a preposition at the end of a clause or sentence, where it has the appearance of being stranded” (Fowler 1996 s.v. prepositions). But the end of the 19th century is

236


presumably the heyday of that ‘myth’, and Burchfield’s advice is in essence identical to the 18th- and 19th-century stance on the issue. The advice given earlier by Gowers (1954: 139), “[d]o not hesitate to end a sentence with a preposition if your ear tells you that that is where the preposition goes best”, is clearly more progressive, but the very treatment of the matter is evidence of the stylistic uneasiness associated with stranded prepositions (cf. also the discussions by Fowler 1931: 92 or Partridge 1954: s.v. preposition at end). Gowers’ (1954: 139) warning, “[s]ometimes, when the final word is really a verbal particle, and the verb’s meaning depends on it, they form together a phrasal verb … and to separate them makes nonsense”, seems to offer additional evidence that the stylistic uneasiness concerning stranded prepositions may indeed have had some influence on the use of phrasal verbs. As Bately’s study shows, although Dryden appears to have tried to avoid stranded prepositions after 1672, his avoidance of preposition stranding correlates directly with the kind of texts he produces: Dryden, then, seems to have come to avoid ‘prepositions’ at the end of the sentence, or rather, after their expected regimen, in all those constructions where they could occur in normal English, the extent of this avoidance being determined by such considerations as the closeness to conversation of the medium used and the type of subject-matter. (Bately 1964: 273–274)

Thus, some stranded prepositions can still be found in his later comedies, but in his non-dramatic verse they are virtually absent after 1672, and Bately goes on to suggest that the avoidance strategy even results in a tendency for the complete omission of the prepositions in passive clauses. From the late 17th century onwards preposition stranding clearly had become a growing concern for style-conscious writers, and from the second half of the 18th century, it developed into one of the targets of prescriptivism. Lowth (1762) makes the following observation: This is an idiom, which our language is strongly inclined to; it prevails in common conversation, and suits very well with the familiar style in writing; but the placing of a preposition before the relative, is more graceful, as well as more perspicuous, and agrees much better with the solemn and elevated style. (Lowth 1762: 127–128)

Until recently 17th- and 18th-century usage of and attitudes to preposition stranding have received surprisingly little attention, despite being one of the major issues in the prescriptive tradition. Yáñez-Bouza (2006 and 2008a and b) has shown that there is a decrease in preposition stranding in the


237

second half of the 18th century, and this is very likely due to the influence of normative grammars: The analysis of the occurrence of stranded prepositions in six formal and informal genres has shown a drastic decrease in frequency in the course of the eighteenth century, the reasons being, evidently, the stigmatisation of the vernacular idiom; as such, the development is part of the stylistic drift towards a more literate prose style during the eighteenth century … this shift was not so much an expression of a genuine grammatical change internal to the language, but rather the result of the chief prescriptive ideals of correctness and politeness. (Yáñez-Bouza 2006: Conclusion)

Another factor which may have played a role is connected to the typical monosyllabic structure of the verbs and the particles. Harris (1781) advises against placing monosyllables at the end of a sentence as follows: It has been called a fault in our Language, that it abounds in Monosyllables. … care should be had that a sentence end not with a crowd of them, those especially of the vulgar, untunable sort, such as, to set it up, to get by and by at it, etc. for these disgrace a Sentence that may be otherwise laudable, and are like the Rabble at the close of some pompous Cavalcade. (Harris 1781, quoted after Tucker 1961: 83)

That is to say, although Harris does not proscribe the use of phrasal verbs, his attack involves a sequence of monosyllabic words typical of phrasal verbs, as his examples show. Ever since the Renaissance the English monosyllables had been subject to very different stylistic judgments (cf. e.g. the discussion of the sources in Jones 1953) and it is not altogether clear whether monosyllabicity is really likely to have had a direct impact on the perception of phrasal verbs; cf. in particular Oldireva Gustafsson’s (2008) discussion of the impact of different rhetorical parameters on attitudes towards and use of monosyllables at the end of a clause. However, as Tony Fairman has suggested (p.c.), the social implications of Harris’ choice of words (the ‘rabble’ vs. the ‘cavalcade’) may be rather strong and indicate an association of the knowledge of ‘pompous’ and polysyllabic ‘high words’ with a high position on the socio-economic scale, which may be demonstrated by stylistic choices unavailable to speakers at the lower end of the socio-economic scale (cf. also Yáñez-Bouza’s reference to politeness as a factor influencing the use of stranded prepositions quoted above). But the situation is likely to be considerably more complex, as Adamson’s (1998) discussion of the changing status of the Latinate vocabulary and the emergence of a ‘middle’ style as a stylistic ideal shows:

238


Whereas Mrs Malaprop [Sheridan, 1775] … is ridiculous because she cannot manage vocabulary derived from the classical languages, Mr Micawber [Dickens, 1850] … is ridiculous precisely because he can. Dickens’s satire of him echoes Maculey’s criticism of Johnson for avoiding ‘strong plain words’ in favour of unnatural bombast, and both form part of a strenuous campaign in favour of ‘Saxon-English’ which, by the end of the nineteenth century, had largely succeeded in driving latinate vocabulary out of the literary lexicon. By the mid-twentieth century, it was being evicted from its refuge in academic and administrative discourse … The consequence for literature has been a narrowing of vocabulary range and a loss of expressive contrast between ‘plain’ and ‘elevated’ language. (Adamson 1998: 609)

For the time being, this issue will therefore have to remain unsettled, but it is noteworthy that in Pelli’s study of phrasal verbs in American plays there is a remarkable increase in types and tokens after 1815 (see Pelli 1976: Chapter 9). If this can be taken to be parallel to the overall increase in phrasal verbs in the 19th century observed in other studies, in the light of Adamson’s analysis it is quite likely that that increase is due to changing literary norms rather than to a sudden growth of the construction type. The Appendix to Lindley Murray’s successful English Grammar (1st edn. 1795; 26th edn. 1815) aimed at “assisting young persons to write with perspicuity and accuracy” (Murray 1815: 274). Murray writes: The fifth rule for the strength of sentences is, to avoid concluding them with an adverb, a preposition, or any inconsiderable word. Agreeably to this rule, we should not conclude with any of the particles, of, to, from, with, by. For instance, it is a great deal better to say, “Avarice is a crime of which wise men are often guilty,” than to say, “Avarice is a crime which wise men are often guilty of.” This is a phraseology which all correct writers shun; and with reason. For as the mind cannot help resting a little, on the import of the word which closes the sentence, it must be disagreeable to be left pausing on a word, which does not, by itself, produce any idea. For the same reason, verbs which are used in a compound sense, with some of these prepositions, are, though not so bad, yet still not proper conclusions of a period: such as, bring about, lay hold of, come over to, clear up, and many other of this kind; instead of which, if we can employ a simple verb, it always terminates the sentence with more strength. Even the pronoun it should, if possible, be avoided in the conclusion: especially when it is joyned with some of the prepositions; as with it, in it, to it … Besides particles and pronouns, any phrase, which expresses a circumstance only, always appears badly in the rear of a sentence. (Murray 1815: 306–307; his italics)

Adamson (1999) thoroughly discusses the central role of perspicuity as a post-Restoration stylistic ideal and explains the rise of that negative attitude


239

towards stranded prepositions in the context of the neo-classical stylistic quest for perspicuity (cf. also the invocation of the principle in the quotation from Lowth 1762 given above): increasingly the ‘Natural Order’ was equated with the English order … But these preferences were justified by an appeal not only to norms of English usage but to universal reason, and where the ‘Natural Order’ of conversational practice turned out to be at odds with the ‘Natural Order’ of rational grammar, the latter was often preferred. Hence Dryden’s revision of his own style to reduce the practice of preposition-stranding … Although very common in spoken English, preposition-stranding was regarded by some as a violation of the logic by which a preposition was so called because it was pre-posed, its ‘natural place’ being in front of the word it governs. (Adamson 1999: 603–604)

The modern usage guides mentioned above unanimously create the impression that the verdict was invented by Dryden and never really heeded by anyone afterwards – a good example how both the reasons for and the impact of post-Restoration attitudes to linguistic issues were misrepresented throughout much of the 20th century. Adamson’s suggestion ties in nicely with Bately’s (1964) analysis of the reasons for the avoidance, who had convincingly refuted the notion that the avoidance of stranded prepositions may have been influenced by the application of rules of Latin grammar to English, for which there is indeed no contemporary evidence at all in the case at hand (cf. Bately 1964: 274–276), as the comment on the matter by Miège in his grammar of English shows: “Ainsi la Preposition devient (s’il m’est permis de le dire) une Postposition. Le Discours en est plus coulant, je l’avouë, & c’est a l’Imitation de l’Ore terms des Latin” (Miège 1685: 114, quoted after Bately 1964: 275). The enduring impact of the principle of perspicuity in the 19th and 20th century can, moreover, be seen in continued negative attitudes towards pleonastic particles. These have come in for a considerable amount of criticism ever since the 18th century, and again it seems to be Dryden who, although he does not appear to have taken an explicit stance against them, is among the first to reduce this kind of redundancy (albeit not consistently, cf. example (19) above), as the correction of rise up in the first edition to a simple rise in the subsequent editions of the Essay shows, cf. (11) above. In 18th-century grammars, then, such particles are frequently criticized and sometimes labelled ‘improper’, ‘vulgar’, ‘absurd’, etc., as the examples in Sundby, Bjørge & Haugland (1991: 375–376) show: advance forward, approach near labelled ‘absurd’; combine together, cover over labelled ‘improper’, etc. (for further examples see Sundby, Bjørge & Haugland 1991: 375–376). There is a continued concern about the abuse of such

240


‘superfluous’ elements in 20th century usage guides and in popular opinion (cf. also Müller-Schotte 1955). Again, the 19th-century attitudes are as yet difficult to access, but there is clearly a tradition which runs from the 18th century up to the present day. Moreover, in this context it should be remembered that many of the 18th-century grammars, dictionaries and handbooks continued to be used (and published) throughout the 19th century; cf. e.g. Gneuss (1996: 40), who points out that “as far as works intended to describe or teach the living language are concerned, it would make little sense to draw a clearly defined line between the eighteenth and nineteenth centuries”. In the 20th century, warnings against ‘superfluous’ particles are voiced for example in the influential usage guides by Gowers (1954 and 1973 ed. Fraser) and Fowler (1965 ed. Gowers and 1996 ed. Burchfield); cf. also the contemporary voices collected by Foltinek (1964). The English grammatical tradition of the Renaissance appears to be silent in this regard, perhaps not quite surprisingly considering the central stylistic position of copia (cf. Adamson 1999). But from the 18th century onwards ‘superfluous’ elements are criticised as ‘expletives’, ‘pleonasms’, ‘tautologies’, ‘redundancies’ (cf. Sundby, Bjørge & Haugland 1991: 347), and still in the mid-20th-century Gowers (1954) treats such ‘superfluous’ elements as instances of ‘padding’ (and thus continues the Early Modern metaphorisation of style as clothing). Interestingly, 20th-century British commentators (including linguists, e.g. Barber 1964: 140–141) are unanimous in their condemnation of the phenomenon as a recent development, for which American influence is to be blamed; see Mittins et al. (1970: 45– 47) for an overview of the earlier 20th-century usage guides and for a discussion of the low acceptability of the stereotypically mentioned meet up with among British English informants (cf. also Brinton 1996 and Mair 2006: 16). Considering the abundant evidence for such uses in pre-1800 British English, that view is open to some doubt, though. In the 18th century phrasal verbs in general tend to be regarded as idiomatic in the full range of meaning of the term as found in the contemporary literature. We have already encountered the use of the term in Johnson’s derogatory phrase “licentious idioms” in the Preface to his dictionary, and also the mistaken claim that Johnson had used that phrase with regard to phrasal verbs. Earlier, Dryden in his “Dedication to Troilus and Cressida” had been troubled by the dangers of ‘idiomatic’ writing: But how barbarously we yet write and speake, your Lordship knows, and I am sufficiently sensible in my own English. For I am often put to a stand, in considering whether what I write be the Idiom of the Tongue, or false Grammar, and nonsence couch’d beneath that specious Name of Anglicisme; and have no


241

other way to clear my doubts, but by translating my English into Latine, and thereby trying what sence the words will bear in a more stable language. (Dryden 1679 [Novak & Guffey 1984: 222])

But, as Knappe (2004: 451–478) has shown in her thorough analysis of the use of idiom(atic) in the pre-1800 linguistic tradition, the “term ‘idiom’ … is not automatically equated with an ‘anomaly’ that must be corrected” (Knappe 2004: 456); rather, despite frequent occurrence in pejorative contexts, idiom(atic) is used primarily in the definition given by Johnson as follows: A mode of speaking peculiar to a language or dialect; the particular cast of a tongue; a phrase; phraseology. (Johnson 1755: s.v. idiom)

Obviously, in the light of this definition phrasal verbs can be regarded as idiomatic due to their specific syntactic properties, and also due to their potential for semantic non-compositionality. The fact that phrasal verbs are cross-linguistically quite unremarkable is irrelevant here, since the crosslinguistic parallels, which continue to be unknown to many current English linguists, cannot be expected to have been noted (let alone systematically explained) by 18th-century observers. Sundby, Bjørge & Haugland (1991: 18) point out: “idiom in the sense ‘figurative expression’ is almost a pejorative term with the grammarians, who were uneasy about the unitary meaning of multi-word lexemes and the syntactical constraints on them” (cf. also Voitl’s 1969: 197–205 discussion of the history of the term idiom). Moreover, the procedures of syntactic analysis available to the grammarians were almost bound to reinforce a view of phrasal verbs as oddities, as Knappe (2004: 467) concludes in her discussion of Lowth’s syntactic analyses: “In the grammars, then, the functional members of the word classes as seen in their potential phraseological complexity is persistently undermined by the procedure prevalent in the parsing exercises”. But, as Knappe (2004: 472) also points out, there is a different (and, as I would argue, complementary) strand of tradition to be found in comparative studies which had long before the 18th century treated the noncompositional phrasal verbs as a problem to be covered in pedagogical grammars and dictionaries (see Knappe 2004: 455 for a list of positive connotations of the term, in particular in contexts where idiomatic use is opposed to violations committed by foreigners, or, more generally, where violations of established usage are criticized). Echoing Johnson’s observations, the notion that phrasal verbs are ‘idiomatic’ is expressed clearly by Anselm Bayly in his Plain and Complete Grammar of the English Language (1772):

242


Besides this composition by prefixing the preposition, there is another, peculiar to the English, by subjoining it, as, cast off, up, down; set out, go in, go in unto, come out, come out from among: This method not only modifies the signification of words, oftentimes with a wilderness and equivocation that may be diverting to the natives, though perplexing to foreigners; but also renders the construction difficult and disputable even to our best writers, who do not agree in the use of the preposition before the noun. (Bayly 1772: II.76, quoted after Sundby, Bjørge & Haugland 1991: 94)

Here we have the central themes of the common perception of the phrasal verbs up until today: Englishness (as witnessed by the reference to foreigners) and irregularity, which are both connected to idiomaticity. The Germanic analogues and the history of the construction are, of course, likely to have been quite unknown to the contemporary commentators, who were mostly familiar with Latin and French (cf. also the extensive discussion of bilingual dictionaries and foreign language grammars in Knappe 2004). The idiomatic, moreover, had already been connected to the colloquial at the beginning of the 18th century, as Joseph Addison’s (1712) caveat to the poet shows: If Clearness and Perspicuity were only to be consulted, the Poet would have nothing else to do but to cloath his Thoughts in the most plain and natural Expressions. But, since it often happens, that the most obvious Phrases, and those which are used in ordinary Conversation, become familiar to the Ear, and contract a kind of Meanness by passing through the Mouths of the Vulgar, a Poet should take particular care to guard himself against Idiomatick ways of speaking. (The Spectator 285, 26 January 1712 [Bond 1965: 10])

The key terms here are perspicuity and idiomaticity. Phrasal verbs may violate the principle of perspicuity in several ways, while at the same time they are ‘idiomatic’. Cf. also the definition given in the OED s.vv. idiom and idiomatic, where the chain of associations which leads from ‘typical of a specific language’ to ‘illogical’ or ‘colloquial’ is clearly visible: [s.v. idiom] The form of speech peculiar or proper to a people or country; own language or tongue … In a narrower sense: That variety of a language which is peculiar to a limited district or class of people; dialect … A form of expression, grammatical construction, phrase, etc., peculiar to a language; a peculiarity of phraseology approved by the usage of a language, and often having a signification other than its grammatical or logical one. [s.v. idiomatic] Peculiar to or characteristic of a particular language; pertaining to or exhibiting the expressions, constructions, or phraseology approved by the


243

peculiar usage of a language, esp. as differing from a strictly grammatical or logical use of words; vernacular; colloquial.

It seems plausible that phrasal verbs may thus have become stereotypically associated with “[t]he uttered speech of private life”, to borrow Wyld’s definition of colloquiality (1936: 359). I have avoided in this chapter entering on a discussion of the term colloquiality itself; Wyld’s use of the term appears to be among the more sensible ones. No doubt the major problem with the term is its extremely vague use throughout most of the linguistic literature, where it is typically associated with ‘low’ and ‘informal’ styles – if not explicitly, then, more often, by collocation and context. Compare, for example, Murray’s famous diagram in the “General Explanations” of the OED (xxiv), where ‘colloquial’ is placed below the ‘common’ core of the language right on the way to ‘slang’ below (and note that the first paragraph of that text, which could by no stretch of the term be labelled ‘colloquial’, contains as many as three phrasal verbs: shade off, fade away, widen out). But Wyld’s application of the term in his book is itself is so vague as to render the term altogether useless. Thus he includes “formulas used in beginning and ending letters” (Wyld 1936: 361) in his discussion, although even in Wyld’s definition a less colloquial textual element is hardly imaginable, but also “dialogue”, “complimentary banter”, “oaths”, “euphemisms” – in other words, the width and vagueness of its application renders the term well nigh vacuous and virtually useless for linguistic discussion. Cf. also Moser’s (1969) discussion of the German equivalent Umgangssprache in the context of German dialectology, where the distance from the linguistic norms of the standard and, in a hierarchizing metaphor, a position below ‘standard’ and above ‘dialect’ is a defining property of the term. But such views can hardly be reconciled with more recent approaches to language standardization or with the insights of variationist sociolinguistics of the last decades (cf. e.g. Petyt 1980, Hudson 1996 or Wardhaugh 2006 for general introductions to these topics and for further references). On closer inspection, the term colloquial appears to create more problems than it solves. This is also true of its German equivalent umgangssprachlich, which is identical in meaning, and in claims to the contrary it is overlooked that the presumptive differences between the two terms in the German and in the English traditions are simply the result of different dialect situations (cf. e.g. the different definitions in Bußmann 1990 s.v. and 2008 s.v.). Moser acknowledged the problems in his conclusion (1969: 228), but he nevertheless tried to integrate the term in his model of a hierarchically ordered diastratic, diatopic and diaphasic linguistic space (Moser 1969: 230–231). In fact, the

244


metaphorical placing of colloquiality on a cultural and linguistic level from ‘top’ to ‘bottom’ remains an essential characteristic of the term in most contexts where it is applied (see Kenyon 1948 for a critical discussion of this practice). Moreover, as Kenyon’s (1948: 28–29) discussion of definitions and examples of ‘colloquiality’ in the literature shows, even his careful application of the term as a “functional variety of standard English”, which is “used chiefly in conversation” (Kenyon 1948: 28) can hardly be applied to the phrasal verbs as a construction type in general, since that would completely disagree with the use of phrasal verbs in all kinds of texts. Rather, the connection of phrasal verbs to colloquiality appears similar to mistaken claims with regard to the reduction of vowels in unstressed syllables in “colloquial pronunciation”, while, as Kenyon (1948: 30) justly points out, that reduction is typical of all natural English speech styles, be they ‘colloquial’ or not. As we have seen in the present chapter, there is good reason to assume that throughout the history of English phrasal verbs have always had a place in the ‘common core’ of the language, as defined in the introduction to the OED: The English vocabulary contains a nucleus or central mass of many thousand words whose ‘Anglicity’ is unquestioned; some of them only literary, some of them only colloquial, the great majority at once literary and colloquial, – they are the Common Words of the language. (OED, “General Explanations”, xvii)

And this may indeed account for an increase of phrasal verbs in 20thcentury English which, if by anything, is characterized by an impetus towards common speech style in all genres; cf. Bailey’s (1992) description of the long-term trends in the development of English since the Renaissance: The ideal English thus became centered firmly in a conceptual midpoint between the florid and the inarticulate, the aristocracy and the folk, the overly learned and the uneducated. Subsequent disputes turned to the issue of just what sort of English lay in the middle ground and just which expressions from intellectual and geographical voyagers would be admitted to it. (Bailey 1992: 57)

But the assignment of the phrasal verbs to the “colloquial speech of the uneducated” (cf. Konishi 1958 quoted at the beginning of Section 6.2) seems hardly justified.

Conclusion

6.4.

245

Conclusion

In this chapter some space has been devoted to a discussion of John Dryden’s Essay of Dramatick Poesy, no doubt one of the most style-conscious and elaborate prose texts ever written in English. In the literature on phrasal verbs, that essay has been repeatedly adduced as evidence for Dryden’s presumptive avoidance of phrasal verbs. The implication, of course, is that Dryden avoided phrasal verbs because of their ‘informality’, or ‘colloquiality’, a notion which has traditionally been used to describe the stylistic value of phrasal verbs throughout their history. But it turned out on closer inspection that there is very little evidence for this claim. In those cases where Dryden actually removed phrasal verbs for the second edition of the Essay, it is highly likely that the removal was the result of a more general conscious effort to produce a text which was logical, abstract, ideational, and perspicuous. But this did not preclude the use of phrasal verbs as such, as the goodly number of phrasal verbs which remained unchanged in the second edition shows. More importantly, the example of Dryden appears to be representative: there is no evidence to be found for the presumptive ‘colloquiality’ or ‘informality’ of the phrasal verbs – neither in linguistic practice nor in metalinguistic comments. It is likely that the traditional view of the phrasal verbs as a particularly ‘English’ and ‘colloquial’ construction is beginning to take off in the 18th century as the indirect result of a number of metalinguistic and stylistic factors; I have tentatively characterized the operation of these factors as a colloquialization conspiracy, which is inextricably connected to the widespread explicit classification of phrasal verbs as ‘idiomatic’ from the 18th century onwards. But, as has been argued throughout this study, this is a misconception deriving from an imperfect recognition of the diachronic and cross-linguistic regularity of the construction, which was bound to be overlooked by 18th-century observers. In the light of this suggestion, it is also possible to offer an explanation why the various quantitative studies discussed in the first part of this chapter are so hopelessly inconclusive, especially where they attempt to correlate the use of phrasal verbs to stylistic factors. Primarily, the use of phrasal verbs as a specific construction type has never really been a stylistic issue at all. However, it is likely that the use of phrasal verbs has been influenced indirectly by the choice of Latinate verbs, which well into the 19th century were deemed the appropriate choice for the ‘high’ registers of English. It remains to be explored in further studies whether this has indeed been the case, and whether this may offer an explanation for the apparent

246


increase in the use of phrasal verbs in the course of 19th century, which coincides with the extension of the ‘middle’ style in English writing.

Chapter 7 Conclusion Ever since the 18th century, phrasal verbs have been regarded as specifically ‘idiomatic’ and ‘English’. But even a cursory look at other Germanic languages indicates that this view is in need of some explanation, since there are corresponding constructions in all these languages. Thus, one main concern of this study has been to examine the actual degree of Englishness of the phrasal verb. More space, however, has been devoted to the discussion of the long-term diachronic development of particle verbs in English, with a strong focus on comparative and cross-linguistic considerations. From this discussion it has emerged that there is in fact very little reason to see anything specific in the English development, which essentially runs parallel to that in the other Germanic languages. Without exception, the syntactic, semantic and stylistic characteristics of the English verb-particle construction are merely epiphenomenal to other characteristic features of the language. What is particularly remarkable in English, I therefore suggest, is the attitude towards the construction, rather than the construction itself. 7.1.

Summary

Chapter 2 provided an analysis of the verb-particle construction in presentday English and other Germanic languages. It turned out that from a contrastive perspective the syntactic and semantic characteristics of the construction are remarkably similar in all present-day Germanic languages, including English. In all these genetically related languages, the relative order of verb and particle depends fundamentally on the word order of the clause. Thus, since word order in Modern English is normally strictly V-3 (SVO), the postverbal position of the particle is entirely in line with the position of the particle in the other Germanic languages, where the particle may precede the verb, but only in clauses with a different word order. Likewise, the semantic types of English particle verbs are also found in all other present-day Germanic languages. Some previous classifications of the English phrasal verbs have been rejected as unsatisfactory in this study. In particular, the viability of ‘group verbs’ as a distinct class in English has proved doubtful, since the various types of verbal constructions subsumed under that heading have only little in common. With regard to phrasal

248

Conclusion

verbs, at any rate, the criteria proposed in favour of their inclusion in such a class can be shown to be inadequate, since they do not sufficiently capture the syntactic, semantic and lexical characteristics of the construction. I have suggested that this is also true of arguments presented more recently in favour of analysing the particles as intransitive prepositions. From a functional point of view the English particle verbs had better be analysed as periphrastic word formations, contrary to the traditional practice in studies of English word formation, where phrasal verbs have always been excluded from the discussion. Thus, particle verbs and prefix verbs can be regarded as cross-linguistically widely attested subtypes of complex predicates. Moreover, the analysis of phrasal verbs in terms of word formation helps explain some of their properties which otherwise remain unaccounted for, like their derivational force, and which again are also found with the particles in other Germanic languages. In line with recent work in Construction Grammar, I have put forward arguments in favour of a constructional approach to these periphrastic word formations and in favour of treating the specific positional properties of the English construction as instances of allostructional variation. In Chapter 3 the development of the postverbal particle position was explored. Like the particles in the present-day Continental West Germanic languages discussed in the preceding chapter, the Old English particles may precede or follow the verb depending on clause type. In syntactic discussions which focus on Old English alone, the classification of these particles presents formidable problems, due to the partial homonymy of prepositions, postpositions, adverbs and prefixes in Old English. To some extent this belongs with a well-known set of problems in analysing dead languages, where typically neither speaker intuitions nor syntactic tests are available. Crucially, however, this partial categorial indeterminacy is also an instance of layering, where a synchronic analysis provides a snapshot of a systematic diachronic development. A discussion of cross-linguistic and comparative evidence from several non-Indo-European and Indo-European languages has revealed that the development in English can be regarded as an instance of the widely attested evolution of preverbs, which accounts for the history both of the inseparable prefix verbs and of the particle verbs. There is evidence for this in a wide array of genetically unrelated languages, where the decategorialization of adverbs (typically derived from formerly independent relational nouns) can be observed in syntagmatic strings in which adverbs are preceded by oblique nouns and followed by simple verbs. Such adverbs may develop either into adpositions or into verb particles which tend to fuse with the verbal base and consequently to develop into prefixes. In the Indo-European languages there is ample

Summary

249

attestation for such developments, which depend crucially on word order in the clause: wherever the position of the verb changes from clause-final to a position earlier in the clause, the preverb comes to occur in postverbal position. Consequently, the positional properties of the particles must be regarded as epiphenomenal to other syntactic developments. Evidence from Germanic to medieval English serves to show that this is indeed the case in the history of English, where the loss of preverbal particles is ultimately caused by the typologically significant long-term loss of basic OV order between Proto-Germanic and Modern English. The systematic correspondences found in the present-day Germanic languages have thus received a coherent historical explanation. This also explains the decreasing formation of new native prefixes in the history of English, since the necessary input of particles immediately preceding the verb is lost as well. On the whole, both developments are therefore entirely in line with the other Germanic languages. In Chapter 4 a critical evaluation of the research tradition showed that many of the traditional assumptions have rarely been subject to scrutiny and that they are often unsatisfactory; although there is general agreement on the factors which may have played a role in the development, their discussion has often been rather confused. In many ways Kennedy’s (1920) pioneering monograph has had a sustained influence on later studies, where many of Kennedy’s claims have continued to be accepted as evidence. As a result, up until today the relevant handbooks speculate about the reasons for what they present as the rise of the phrasal verb in Middle English, although there is hardly any evidence which would suggest anything but a continuous development from pre-Old English to present-day English. To a considerable extent this situation is due to the widespread neglect of insights into the syntactic causes of the loss of preverbal particles in the literature on historical word formation and lexicology. In the syntactic literature, on the other hand, there is but little interest in the lexical evaluation of the development. Also, the coverage of Old and Middle English particle verbs in the relevant historical dictionaries is far from satisfactory. It is likely that the insufficient coverage in the dictionaries is the result of (but presumably also a contributing factor to) the failure of older studies to present a sufficiently coherent account of the emergence of the phrasal verb. Chapter 5 was devoted to sketching such a coherent account, examining the history of the English verb-particle construction from a morphological, lexical and semantic point of view. Clearly the development of the particle verbs is inextricably connected to the history of the prefix verbs, both the native, Germanic ones and the borrowed, Romance ones. This relationship has been portrayed in much of the literature in a more or less vaguely

250

Conclusion

metaphorical fashion, and it has often been claimed that the older native prefixes were somehow ousted by the borrowed prefixes and replaced by the particles. But such accounts cannot be accepted as valid, since the longterm loss of the majority of the native prefixes in the history of English is ultimately caused by morphonological factors which are quite independent of the adoption of borrowed prefixes, as the chronology of the development shows. From pre-Old English onwards native prefixes are subject to progressive phonetic attrition and a recurrent pattern of desemantization, but up until the Middle English period new Germanic prefixes continue to emerge, following the path of development described in Chapter 3. These younger Old English prefixes tend to be homonymous with free particles. When in the course of the Middle English period the particles stop occurring in preverbal position, their further development into bound affixes is prevented. In all other respects, however, the particles have turned out to be fully equivalent to the prefixes, as the semantic developments of native prefixes and particles has shown. Thus, the evolution of the typical semantic properties of verb-particle constructions corresponds to that of the native prefixes. This, of course, was only to be expected in the light of the comparative evidence (cf. Chapter 3). Further support has come from a discussion of preverbs in Gothic, which represent an earlier Germanic stage in the development of the oldest bound prefixes in Old English. In the light of this discussion it has also become clear why there are functional and semantic parallels between the Old English preverbs and the particles of the Modern English phrasal verbs, while the assumption of language contact as a directly relevant factor in the development has turned out to be entirely unnecessary and implausible. The adoption of borrowed prefixes, on the other hand, must be seen in the context of the elaboration of the vernacular for written discourse, which is heavily dependent on Latin and French models and which, due to the long-term political and cultural consequences of the Norman Conquest, strongly favours the use of Latinate items. This explains also the often noted etymological and prosodic characteristics of the Modern English phrasal verbs. Thus the frequently mentioned etymological and phonological constraints on phrasal verbs are most likely to be epiphenomenal, too. Chapter 6 connected to the central concerns of the study in several ways. In it the emergence of the specifically English attitudes towards the phrasal verb was discussed and various topics which had already been introduced in earlier chapters were taken up more systematically, in particular the wide-spread stylistic characterizations of phrasal verbs as colloquial and informal. Following Kennedy’s (1920) ahistorical reasoning examined in Chapter 4, quantitative studies of the development have

Summary

251

unanimously taken such characterizations for granted, and they have been very much concerned with showing a diachronic increase assumed to be linked both to the presumptive rise of the construction in Middle English and to its stylistic peculiarities. At closer inspection, however, these studies have turned out to be inconclusive in almost every respect, and even the figures they present in support of an increase in phrasal verbs from the late Middle Ages onwards can be shown to be based on methodologically dubious procedures. Studies based on evidence from the OED merely reflect the peculiarities of the dictionary, rather than any quantitative developments in the language. The connections the quantitative studies try to establish between the development of phrasal verbs and stylistic and/or social variables also lack persuasive force: although there is some evidence which actually suggests an overall long-term increase in the use phrasal verbs (in particular in the 19th century), so far no viable explanation for this has been put forward. Following up on the discussion in Chapter 5, I have therefore presented an alternative explanation and suggested that the use of phrasal verbs has essentially never been a stylistic issue. On the basis of a critical analysis of attitudes towards the construction, I have argued that there is no pre-1800 evidence for its presumed colloquiality at all. What is significant is the use of Latinate vocabulary and its connection to changing stylistic ideals rather than the use of phrasal verbs. Moreover, a representative examination of the use of phrasal verbs by a number of styleconscious 17th- and 18th-century authors has shown that these authors use phrasal verbs quite freely in highly elaborate texts. Likewise, there is no evidence for the negative attitudes towards phrasal verbs which have been widely ascribed to these authors in the literature. Rather, the traditional view of phrasal verbs as a typically English and particularly colloquial construction has its roots in the 18th century as the indirect result of a number of metalinguistic and stylistic factors, which I have described as a colloquialization conspiracy. Relevant factors include the normative verdicts against preposition stranding, monosyllables and pleonasm. Since the diachronic and cross-linguistic regularity of the construction was bound to be overlooked by the early grammarians and lexicographers, phrasal verbs have been regarded as idiomatic from the 18th century onwards, and I have suggested that connotations of idiomaticity have played a crucial role in their subsequent characterization as English, irregular and colloquial.

252

Conclusion

7.2.

Outlook

Phrasal verbs are far from being “wildly irregular”, as Samuel Johnson had characterized them in the quote from the Preface to his Dictionary (1755) which was given at the very beginning of this study. But of course, engaging in a battle against a linguistic assessment made more than 250 years ago would in itself hardly justify a study of any length, were it not for the enduring afterlife of such views in the modern linguistic discussion. Contrary to such wide-spread claims, this study has shown that phrasal verbs are neither specifically English, nor the result of language contact, nor a Middle English innovation; they are also not irregular, colloquial, or informal. Note that this characterization refers to phrasal verbs as a type of construction. That any of these points may apply to any individual phrasal verb – as to any other English verb – is of course completely indisputable. But many English verb-particle constructions simply belong to the common core of the English lexicon, and in essence no special explanation is necessary for their development, which in fact can be regarded as highly regular, in particular when it is viewed from a cross-linguistic and comparative perspective. More importantly, perhaps, this clears the way for more detailed studies of many topics which have not been discussed here, or which were touched upon only in passing. By way of concluding this study, let me mention a few such topics. First, it should be worthwhile to include the particle verbs in a largerscale historical study of verbal word formation, which looks into the changing structural and discursive properties of different types of verbal affixation in English. Some initial proposals in this direction have been made in the preceding Chapters 5 and 6, where it was argued that many of the borrowed prefixes and many of the particles belong to distinct and complementary functional domains. But so far there have been only rudimentary explorations of the historical emergence of this Modern English situation. Rather than relying on quantitative factors of the kind so popular in historical corpus linguistics, such a description would be likely to benefit greatly from approaching the subject in a more quality-oriented way, which takes into account the development of discourse traditions and their historical implementation in English. The semantic development of the verb-particle construction has been sketched here only in principle. So far we know only little about the exact ways in which individual particles acquire their respective aspectualizing, derivational or valency-changing properties, and even less about historical patterns in the emergence of new idiomatic phrasal verbs. To me it seems

Outlook

253

that all these issues provide prime examples of constructionalization and can be discussed most fruitfully in the context of surface based exemplar models of language which do not believe any longer in the reality of boundaries between syntax, morphology and the lexicon. At any rate, a consistent semantic and functional classification of the construction would also be a necessary condition for quantitative diachronic and variationist studies of the productivity and of changing type and token frequencies. The lack of such a systematic classification is one reason why the figures discussed in Chapter 6 do not help to establish significant insights on the development of the phrasal verb. Diachronic counts of individual particles without distinction between their compositional, aspectual, idiomatic, derivational uses etc. are not likely to produce useful data. Other topics which would deserve a detailed examination include the history of nominalized phrasal verbs, 19th-century usage and attitudes, or the development of variety-specific uses. The use of pleonastic particles has been discussed in Chapters 5 and 6. As we have seen, this has attracted a good deal of prescriptive censure in Late Modern English. From the 19th century onwards this has been attributed to American influence. But since there is evidence for pleonastic uses from throughout the history of the phrasal verbs, this view would seem to be in need of further explanation. The combination of borrowed verbs with pleonastic particles, on the other hand, has sometimes been seen as a means to ease their integration. That view appears to be more plausible, but I would argue that it is the cognitive profiling created by such particles which ultimately underlies their use in combinations with all kinds of verbs, and that the combination with borrowed verbs is just a subset of that use. A thorough historical study of this phenomenon would therefore ideally cover the whole history of the language. The relevant issues have been discussed in various places in the present study. Most prominently, they include the relation of prefix and particle semantics in Old and Middle English (for Old English, the use of pleonastic particles with simple verbs has so far been investigated nowhere in the literature), the combination of particles with borrowed verbs mainly from French and Latin in Middle English and Early Modern English, and the impact of the normative dislike of pleonasm on usage in the Standard varieties of the language. Of course, this would also deserve scrutiny with respect to non-Standard varieties and conceivable differences to the Standard. Clearly, all work in this area would be facilitated if the lexicographic deficits criticized in Chapter 4 were to be satisfactorily addressed. None of these suggestions for further research is of course likely to surprise anyone familiar with the relevant historical research on German, where particle verbs have traditionally been treated in the context of word

254

Conclusion

formation and where these developments have been studied in considerably more detail. Thus, as a last point, I should like to suggest an extension of comparative research, which has tended to concentrate on the Germanic languages in their earliest attestations, to the more recent periods of their histories. It has been one major aim of this study to argue that such a reorientation of historical research on the English verb-particle construction is indeed likely to yield fruitful new insights.

References Aarts, Bas. 1989. Verb-preposition constructions and small clauses in English. Journal of Linguistics 25. 277–290. Aarts, Bas. 1992. Small clauses in English: The non-verbal types. Berlin: Mouton de Gruyter. Aarts, Bas. 1997. English syntax and argumentation. Basingstoke: Macmillan. Ackerman, Farrell. 2003. Aspectual contrasts and lexeme derivation in Estonian: A realization-based morphological perspective. In Yearbook of Morphology 2003, ed. Geert Booij & Jaap van Marle. Dordrecht: Kluwer. 13–31. Ackerman, Farrell & Gert Webelhuth. 1998. A theory of predicates. Stanford: CSLI Publications. Adams, Valerie. 1973. An introduction to Modern English word-formation. London: Longman. Adams, Valerie. 2001. Complex words in English. Harlow: Longman. Adamson, Sylvia. 1998. Literary language. In The Cambridge History of the English Language, vol. 4: 1776–1997, ed. Suzanne Romaine. Cambridge: Cambridge University Press. 589–692. Adamson, Sylvia. 1999. Literary language. In The Cambridge History of the English Language, vol. 3: 1476–1776, ed. Roger Lass. Cambridge: Cambridge University Press. 539–653. Ahulu, Samuel. 1995. Variation in the use of complex verbs in international English. English Today 11. 28–34. Aitchison, Jean. 1981. Language change: Progress or decay? London: Fontana. Akimoto, Minoji. 2006. On the decline of after and forth in verb phrases. In Syntax, style and grammatical norms: English from 1500–2000, ed. Christiane Dalton-Puffer, Nikolaus Ritt, Herbert Schendl & Dieter Kastovsky. Bern: Lang. 11–31. Algeo, John. 1998. Vocabulary. In The Cambridge History of the English Language, vol. 4: 1776–1997, ed. Suzanne Romaine. Cambridge: Cambridge University Press. 57–91. Allen, Cynthia L. 1990. Review of Laurel J. Brinton, The development of English aspectual systems: Aspectualizers and post-verbal particles (Cambridge: Cambridge University Press, 1988). Journal of Linguistics 26. 245–250. Allen, Cynthia L. 2006. Case syncretism and word order change. In The handbook of the history of English, ed. Ans van Kemenade & Bettelou Los. Oxford: Blackwell. 201–223. Allerton, David. 2004. The analysis of fixed expressions in a text. In Phraseological units: Basic concepts and their application, ed. David Allerton, Nadja Nesselhauf & Paul Skandera. Basel: Schwabe. 87–105. Andrew, S.O. 1940. Syntax and style in Old English. Cambridge: Cambridge University Press. (Reissued New York: Russell & Russell, 1966.)

256

References

Antonsen, Elmer H. 1975. A Concise Grammar of the Older Runic Inscriptions. Tübingen: Niemeyer. Antonsen, Elmer H. 2002. Runes and Germanic linguistics. Berlin: Mouton de Gruyter. Aronoff, Mark. 2000. [History of morphological research II: Research traditions in the 20th century:] Generative grammar. In Morphologie: Ein internationales Handbuch zur Flexion und Wortbildung (Morphology: An International Handbook on Inflection and Word-formation), vol. 1, ed. Geert Booij, Christian Lehmann & Joachim Mugdan, in collaboration with Wolfgang Kesselheim & Stavros Skopeteas. Berlin: de Gruyter. 194–209. Askedal, John Ole. 1994. Norwegian. In The Germanic languages, ed. Ekkehard König & Johan van der Auwera. London: Routledge. 219–270. Askedal, John Ole. 1995. Geographical and typological description of verbal constructions in the modern Germanic languages. In Drei Studien zum Germanischen in alter und neuer Zeit, ed. John Ole Askedal & Harald Bjorvand. Odense: Odense University Press. Askedal, John Ole. 2001. Conceptions of typological change. In Language Typology and Language Universals: An International Handbook, vol. 2, ed. Martin Haspelmath, Ekkehard König, Wulf Oesterreicher & Wolfgang Raible. Berlin: de Gruyter. 1624–1640. Askedal, John Ole. 2005. [The typological development of the Nordic languages II:] Morphology and syntax. In The Nordic Languages: An International Handbook of the History of the North Germanic Lnaguages, vol. 2, ed. Oskar Bandle et al. Berlin: de Gruyter. 1872–1886. Auer, Anita. 2006. Precept and practice: The influence of prescriptivism on the English subjunctive. In Syntax, style and grammatical norms: English from 1500–2000, ed. Christiane Dalton-Puffer, Nikolaus Ritt, Herbert Schendl & Dieter Kastovsky. Bern: Lang. 33–53. Auwera, Johan van der. 1995. Les préverbes du Néerlandais: Une comparaison avec l’Allemand. In Les préverbes dans les langues d’Europe: Introduction à l’étude de la préverbation, ed. André Rousseau. Lille: Presses Universitaires du Septentrion. 77–94. Auwera, Johan van der. 1999. Dutch verbal prefixes: Meaning and form, grammaticalization and lexicalization. In Boundaries of morphology and syntax, ed. Lunella Mereu. Amsterdam: Benjamins. 121–136. Baayen, R. Harald & Antoinette Renouf. 1996. Chronicling the Times: Productive lexical innovations in an English newspaper. Language 72. 69–96. Bacchielli, Rolando. 1999. An essential bibliographical guide to the synchronic and diachronic study of verb-particle combinations (‘phrasal verbs’) in English. In SLIN Newsletter 20, 4–21 and 21, 14–23 (also on the internet: ). Bækken, Bjørg. 1998. Word order patterns in Early Modern English: With special reference to the position of the subject and the finite verb. Oslo: Novus. Bailey, Richard W. 1992. Images of English: A cultural history of the English language. Cambridge: Cambridge University Press.

References

257

Baldi, Philip. 1979. Typology and the Indo-European prepositions. Indogermanische Forschungen 84. 49–61. Bammesberger, Alfred (ed.). 1985. Problems of Old English lexicography: Studies in memory of Angus Cameron. Regensburg: Pustet. Barber, Charles. 1964. Linguistic change in present-day English. Edinburgh: Oliver. Barber, Charles. 1997. Early Modern English. 2nd edn. Edinburgh: Edinburgh University Press. Barðdal, Jóhanna. 2008. Productivity: Evidence from case and argument structure in Icelandic. Amsterdam: Benjamins. Barnes, Michael P. & Eivind Weyhe. Faroese. In The Germanic languages, ed. Ekkehard König & Johan van der Auwera. London: Routledge. 190–218. Barz, Irmhild. 2007. Wortbildung und Phraseologie. In Phraseologie: Ein internationales Handbuch der zeitgenössischen Forschung (Phraseology: An International Handbook of Contemporary Research), vol. 1, ed. Harald Burger, Dmitrij Dobrovol’skij, Peter Kühn & Neal R. Norrick. Berlin: de Gruyter. 27– 36. Bate, W.J. & Albrecht B. Strauss (eds.). 1969. The Yale Edition of the Works of Samuel Johnson, vol. 5: The Rambler (part 3). New Haven: Yale University Press. Bately, Janet M. 1964. Dryden’s revisions in the Essay of Dramatic Poesy: The preposition at the end of the sentence and the expression of the relative. The Review of English Studies 15. 268–282. Bately, Janet M. (ed.). 1980. The Old English Orosius. EETS SS 6. Bately, Janet M. (ed.). 1986. The Anglo-Saxon Chronicle. A collaborative edition, general eds. David Dumville & Simon Keynes, vol. 3: MS A (A semidiplomatic edition with introduction and indices). Cambridge: Brewer. Bauer, Laurie. 1983. English word-formation. Cambridge: Cambridge University Press. Bauer, Laurie. 2003a. Introducing linguistic morphology. 2nd edn. Edinburgh: Edinburgh University Press. Bauer, Laurie. 2003b. English prefixation: A typological shift? Acta Linguistica Hungarica 50. 33–40. Baugh, Albert C. & Thomas Cable. 1993. A history of the English language. 4th edn. London: Routledge. Baugh, Albert C. & Thomas Cable. 2002. A history of the English language. 5th edn. London: Routledge. Bayley, Anselm. 1772. A plain and complete grammar with the English accidence. London: Ridley. (Repr. Menston: Scolar Press, 1974.) Beaglehole, J.C. (ed.). 1955. The journals of Captain James Cook on his voyages of discovery. 8 vols. and a portfolio, ed. from the original mss. Vol. 1,1: The voyage of the Endeavour: 1768–1771. (Repr. 1999.) Woodbridge: Boydell Press. Beal, Joan C. 2004. English in modern times 1700–1945. London: Arnold.

258

References

Bean, Marian C. 1983. The development of word order patterns in Old English. London: Croom Helm. Beavers, John, Beth Levin & Shiao Wei Tham. 2010. The typology of motion expressions revisited. Journal of Linguistics 46. 331–377. Beekes, Robert S.P. 1995. Comparative Indo-European linguistics: An introduction. Amsterdam: Benjamins. Behaghel, Otto. 1909. Beziehungen zwischen Umfang und Reihenfolge von Satzgliedern. Indogermanische Forschungen 25. 110–142. Behaghel, Otto. 1924. Deutsche Syntax: Eine geschichtliche Darstellung, vol. 2: Die Wortklassen und ihre Wortformen (B. Adverbium – C. Verbum). Heidelberg: Winter. Behaghel, Otto. 1930. Von deutscher Wortstellung. Zeitschrift für Deutschkunde 44. 81–89. Behaghel, Otto. 1932. Deutsche Syntax: Eine geschichtliche Darstellung, vol. 4: Wortstellung – Periodenbau. Heidelberg: Winter. Benson, Larry D. (general ed.). 1988. The Riverside Chaucer. 3rd edn. Oxford: Oxford University Press. Berg, Helma van den. 2003. Spatial prefixes in Dargi (East Caucasian). Acta Linguistica Hungarica 50. 201–225. Bergs, Alexander & Gabriele Diewald (eds.). 2008. Constructions and language change. Berlin: Mouton de Gruyter. Besch, Werner, Anne Betten, Oskar Reichmann & Stefan Sonderegger (eds.). 1998–2004. Sprachgeschichte: Ein Handbuch zur Geschichte der deutschen Sprache und ihrer Erforschung. 4 vols. 2nd edn. Berlin: de Gruyter. Besten, Hans den & Corretje Moed-van Walraven. 1986. The syntax of verbs in Yiddish. In Verb second phenomena in Germanic languages, ed. Hubert Haider & Martin Prinzhorn. Dordrecht: Foris. 111–135. Bethurum, Dorothy (ed.). 1957. The Homilies of Wulfstan. Oxford: Clarendon Press. Biber, Douglas, Stig Johansson, Geoffrey Leech, Susan Conrad & Edward Finegan. 1999. Longman Grammar of Spoken and Written English. London: Longman. Binnick, Robert I. 2001. Temporality and aspectuality. In Language Typology and Language Universals: An International Handbook, vol. 2, ed. Martin Haspelmath, Ekkehard König, Wulf Oesterreicher & Wolfgang Raible. Berlin: de Gruyter. 557–567. Blake, Norman F. 2002. Phrasal verbs and associated forms in Shakespeare. Atlantis 24. 25–39. Blank, Andreas. 2001. Pathways of lexicalization. In Language Typology and Language Universals: An International Handbook, vol. 1, ed. Martin Haspelmath, Ekkehard König, Wulf Oesterreicher & Wolfgang Raible. Berlin: de Gruyter. 1596–1608. Blockley, Mary. 2001. Aspects of Old English poetic syntax: Where clauses begin. Urbana: University of Illinois Press.

References

259

Blom, Corrien & Gert Booij. 2003. The diachrony of complex predicates in Dutch: A case study in grammaticalization. Acta Linguistica Hungarica 50. 61–91. Bolinger, Dwight 1971. The phrasal verb in English. Cambridge, MA: Harvard University Press. Bond, Donald F. (ed.). 1965. The Spectator, vol. 3. Oxford: Clarendon Press. Booij, Geert. 1990. The boundary between morphology and syntax: Separable complex verbs in Dutch. In Yearbook of Morphology 1990, ed. Geert Booji & Jaap van Marle. Dordrecht: Foris. 45–63. Booij, Geert. 2002a. The morphology of Dutch. Oxford: Oxford University Press. Booij, Geert. 2002b. Separable complex verbs in Dutch: A case of periphrastic word formation. In Verb-particle explorations, ed. Nicole Dehé, Ray Jackendoff, Andrew McIntyre & Silke Urban. Berlin: Mouton de Gruyter. 21–41. Booij, Geert. 2007. The grammar of words: An introduction to linguistic morphology. 2nd edn. Oxford: Oxford University Press. Booij, Geert. 2010. Construction Morphology. Oxford: Oxford University Press. Booij, Gert & Ans van Kemenade. 2003. Preverbs: An introduction. In Yearbook of Morphology 2003, ed. Geert Booij & Jaap van Marle. Dordrecht: Kluwer. 1– 11. Booij, Geert & Jaap van Marle (eds.). 2003. Yearbook of Morphology 2003. Dordrecht: Kluwer. Borden, Arthur R. 1982. A Comprehensive Old English Dictionary. Washington: University Press of America. Bosworth–Toller = Joseph Bosworth & T. Northcote Toller. An Anglo-Saxon Dictionary (Oxford: Oxford University Press, 1882–1898); Supplement by T. Northcote Toller (Oxford: Clarendon Press, 1908–1921); Enlarged addenda and corrigenda to the Supplement by Alistair Campbell (Oxford: Clarendon Press, 1972). Braune–Heidermanns 2004 = Braune, Wilhelm. 2004. Gotische Grammatik: Mit Lesestücken und Wörterverzeichnis. 20th edn. by Frank Heidermanns. Tübingen: Niemeyer. Braunmüller, Kurt. 1982. Syntaxtypologische Studien zum Germanischen. Tübingen: Narr. Braunmüller, Kurt. 1999. Die skandinavischen Sprachen im Überblick. 2nd edn. Tübingen: Francke. Brewer, Charlotte. 2000. ‘OED’ sources. In Lexicography and the OED: Pioneers in the untrodden forest, ed. Lynda Mugglestone. Oxford: Oxford University Press. 40–58. Brinton, Laurel J. 1985. Verb particles in English: Aspect or aktionsart? Studia Linguistica 39. 157–168. Brinton, Laurel J. 1988. The development of English aspectual systems: Aspectualizers and post-verbal particles. Cambridge: Cambridge University Press. Brinton, Laurel J. 1996. Attitudes toward increasing segmentalization: Complex and phrasal verbs in English. Journal of Linguistics 24. 186–205.

260

References

Brinton, Laurel J. & Minoji Akimoto. 1999. Introduction. In Collocational and idiomatic aspects of composite predicates in the history of English, ed. Brinton & Akimoto. Amsterdam: Benjamins. 1–20. Brinton, Laurel J. & Minoji Akimoto (eds.). 1999. Collocational and idiomatic aspects of composite predicates in the history of English. Amsterdam: Benjamins. Brinton, Laurel J. & Elizabeth Closs Traugott. 2005. Lexicalization and language change. Cambridge: Cambridge University Press. Bülbring, Karl D. 1902. Altenglisches Elementarbuch, 1. Teil: Lautlehre. Heidelberg: Winter. Burchfield 1996 (see Fowler 1996) Burger, Harald. 2007. Phraseologie: Eine Einführung am Beispiel des Deutschen. 3rd edn. Berlin: Schmidt. Burger, Harald, Dmitrij Dobrovol’skij, Peter Kühn & Neal R. Norrick (eds.). 2007. Phraseologie: Ein internationales Handbuch der zeitgenössischen Forschung (Phraseology: An International Handbook of Contemporary Research), vol. 1. Berlin: de Gruyter. Burnley, David. 1992. Lexis and semantics. In The Cambridge History of the English Language, vol. 2: 1066–1476, ed. Norman F. Blake. Cambridge: Cambridge University Press. 409–499. Bußmann, Hadumod (ed.). 1990. Lexikon der Sprachwissenschaft. 2nd edn. Stuttgart: Kröner. Bußmann, Hadumod (ed.). 2008. Lexikon der Sprachwissenschaft. 4th edn. Stuttgart: Kröner. Butler, Sharon & Bruce Mitchell. 1985. Some lexicographical problems posed by Old English grammar words. In Problems of Old English lexicography: Studies in memory of Angus Cameron, ed. Alfred Bammesberger. Regensburg: Pustet. 79–89. Bybee, Joan L. 1985. Morphology: A study of the relation between meaning and form. Amsterdam: Benjamins. Bybee, Joan L. 2010. Language, usage and cognition. Cambridge: Cambridge University Press. Bybee, Joan L. & Östen Dahl. 1989. The creation of tense and aspect systems in the languages of the world. Studies in Language 13. 51–103. Bybee, Joan L. & William Pagliuca. 1985. Cross-linguistic comparison and the development of grammatical meaning. In Historical semantics – historical word-formation, ed. Jacek Fisiak. Berlin: Mouton. 59–83. Bybee, Joan L., Revere Perkins & William Pagliuca. 1994. The evolution of grammar: Tense, aspect, and modality in the languages of the world. Chicago: The University of Chicago Press. Cambridge Phrasal Verbs Dictionary. 2006. Cambridge: Cambridge University Press. Campbell, Alistair. 1959. Old English Grammar. Oxford: Clarendon Press. Campbell, Alistair. 1972 (see Bosworth–Toller)

References

261

Campbell, George L. 1991. Compendium of the world’s languages. 2 vols. London: Routledge. Cappelle, Bert. 2004. The particularity of particles, or why they are not just ‘intransitive prepositions’. In Adpositions of movement, ed. Hubert Cuyckens, Walter de Mulder & Tanja Mortelmans. Amsterdam: Benjamins. 29–57. Cappelle, Bert. 2005. Particle patterns in english: A comprehensive coverage. K.U. Leuven: PhD dissertation. . Cappelle, Bert. 2006. Particle placement and the case for ‘allostructions’. Constructions all over: Case studies and theoretical implications Constructions, special vol. 1. . Cappelle, Bert. 2007. When ‘wee wretched words’ wield weight: The impact of verbal particles on transitivity. In Collocations and Idioms 1: Papers from the First Nordic Conference on Syntactic Freezes, Joensuu, Finland, 19–20 May 2006, ed. Marja Nenonen & Sinikka Niemi. Joensuu: University of Joensuu. 41–54. Cappelle, Bert. 2009. Can we factor out free choice? In Describing and modeling variation in grammar, ed. Andreas Dufter, Jürg Fleischer & Guido Seiler. Berlin: Mouton de Gruyter. 183–201. Carstensen, Broder. 1964. Zur Struktur des neuenglischen Wortverbandes. Die Neueren Sprachen N.F. 13. 305–328. Cassidy, Frederic G. 1997. The rise and development of modern labels in English dictionaries. Dictionaries: Journal of the Dictionary Society of North America 18. 97–112. (Repr. in Lexicography: Critical concepts, ed. R.R.K. Hartmann, Routledge: London, 2003, vol. 3, 255–269.) Castillo, Concha. 1994. Verb-particle combinations in Shakespearean English. Neuphilologische Mitteilungen 95. 439–451. Chapman, Don. 2008. Fixer-uppers and passers-by: Nominalization of verbparticle constructions. In Studies in the History of the English Language IV: Empirical and analytical advances in the study of English language change, ed. Susan M. Fitzmaurice & Donka Minkova. Berlin: Mouton de Gruyter. 265– 300. CHEL = The Cambridge History of the English Language. 1992–2001. 6 vols., general ed. Richard Hogg. Cambridge: Cambridge University Press. Cherchi, Marcello. 1994–1997. Verbal tmesis in Georgian. 1994 (part I) and 1997 (part II). ΑΙΩΝ [Annali del Dipartimento di Studi del Mondo Classico e del Mediterraneo Antico: Sezione Linguistica] 16 and 19. 33–115 and 63–137. Chomsky, Noam & Morris Halle. 1968. The sound pattern of English. New York: Harper & Row. Clackson, James. 2007. Indo-European linguistics: An introduction. Cambridge: Cambridge University Press. Claridge, Claudia. 2000. Multi-word verbs in Early Modern English: A corpusbased study. Amsterdam: Rodopi.

262

References

Clark, Cecily (ed.). 1958. The Peterborough Chronicle 1070–1154. Oxford: Oxford University Press. Clemoes, Peter (ed.). 1997. Ælfric’s Catholic Homilies: The first series. EETS SS 17. 1997. Cohen, Murray. 1977. Sensible words: Linguistic practice in England 1640–1785. Baltimore: The Johns Hopkins University Press. Colman, Fran. 1991. What positions fit in? In Historical English syntax, ed. Dieter Kastovsky. Berlin: Mouton de Gruyter. 51–102. Comrie, Bernard. 1976. Aspect: An introduction to the study of verbal aspect and related problems. Cambridge: Cambridge University Press. Cowie, Anthony Paul & Ronald Mackin. 1975 [2006]. Oxford Dictionary of Current Idiomatic English, vol. 1: Verbs with prepositions and particles [2nd edn. Oxford Phrasal Verbs Dictionary for Learners of English, 2006, ed. Colin McIntosh]. Oxford: Oxford University Press. Cowie, Clare & Christiane Dalton-Puffer. 2002. Diachronic word-formation and studying changes in productivity over time: Theoretical and methodological considerations. In A Changing world of words: Studies in English historical lexicography, lexicology and semantics, ed. Javier E. Díaz Vera. Amsterdam: Rodopi. 410–437. Croft, William. 2006. Evolutionary models and functional-typological theories of language change. In The handbook of the history of English, ed. Ans van Kemenade & Bettelou Los. Oxford: Blackwell. 68–91. Croft, William & David Alan Cruse. 2004. Cognitive linguistics. Cambridge: Cambridge University Press. Cruz, Juan M. de la. 1972a. The origins of the Germanic phrasal verb. Indogermanische Forschungen 77. 73–96. Cruz, Juan M. de la. 1972b. Transference and metaphor in Middle English verbs accompanied by a locative article. Orbis 21. 114–135. Cruz, Juan M. de la. 1972c. The Latin influence on the development of the English phrasal verb. English Philological Studies 13. 1–43. Cruz, Juan M. de la. 1972d. A syntactical complex of isoglosses in the northwestern end of Europe (English, North Germanic and Celtic). Indogermanische Forschungen 77. 171–180. Cruz, Juan M. de la. 1975. Old English pure prefixes: Structure and function. Linguistics 145. 47–81. Cruz, Juan M. de la. 1977. Synchronic-diachronic remarks on the nature of prefixation. Orbis 26. 262–292. Cruz Cabanillas, Isabel de la. 1999. The conflict of homonyms: Does it exist? Cuadernos de Investigación Filológica 25. 107–116. Curme, George O. 1914. The development of verbal compounds in Germanic. Beiträge zur Geschichte der deutschen Sprache und Literatur 39. 320–361. Cusack, Bridget (ed.). 1998. Everyday English 1500–1700: A reader. Edinburgh: Edinburgh University Press. Dahl, Östen. 1985. Tense and aspect sytems. Oxford: Blackwell.

References

263

Dalton-Puffer, Christiane. 1996. The French influence on Middle English morphology: A corpus-based study of derivation. Berlin: Mouton de Gruyter. Daneš, František. 1966 [1964]. A three-level approach to syntax. In L’école de Prague d’aujourd’hui, ed. František Daneš, Karel Horálek, Vladimir Skalička, Pavel Trost & Josef Vachek. Prague: Academia. [Repr. University of Alabama Press, 1966.] 225–240. Day, Mabel (ed.). 1952. The English text of the Ancrene Riwle. Ed. from Cotton MS Nero A.xiv, on the basis of a transcript by J.A. Herbert. EETS 225. De Smet, Hendrik. 2010. Grammatical interference: Subject marker ‘for’ and the phrasal verb particles ‘out’ and ‘forth’. In Gradience, gradualness and grammaticalization, ed. Elizabeth Closs Traugott and Graeme Trousdale. Amsterdam: Benjamins. 75–104. Deeters, Gerhard. 1930. Das karthwelische Verbum: Vergleichende Darstellung des Verbalbaus der südkaukasischen Sprachen. Leipzig: Markert & Petters. Dehé, Nicole. 2002. Particle verbs in English: Syntax, information structure and intonation. Amsterdam: Benjamins. Dehé, Nicole (ed.). 2003– . Verb-particle constructions: A bibliography. . Dekeyser, Xavier. 1986. Romance loans in Middle English: A re-assessment. In Linguistics across historical and geographical boundaries, ed. Dieter Kastovsky & Aleksander Szwedek. Berlin: Mouton de Gruyter. 253–266. Delbrück, Berthold. 1893–1900. Vergleichende Syntax der indogermanischen Sprachen in 3 parts (part 1: 1893; part 2: 1897; part 3: 1900). Vol. 3–5 of Grundriss der vergleichenden Grammatik der indogermanischen Sprachen: Kurzgefasste Darstellung der Geschichte des Altindischen, Altiranischen (Avestischen u. Altpersischen), Altarmenischen, Albanischen, Lateinischen, Oskisch-Umbrischen, Altirischen, Gotischen, Althochdeutschen, Litauischen und Altkirchenslavischen by Karl Brugmann & Berthold Delbrück. Strassburg: Trübner. Denison, David. 1981. Aspects of the history of English group-verbs: With particular attention to the syntax of the Ormulum. University of Oxford: unpublished PhD thesis. Denison, David. 1984. On get it over with. Neophilologus 68. 271–277. Denison, David. 1985. The origins of completive up in English. Neuphilologische Mitteilungen 86. 37–61. Denison, David. 1987. On word order in Old English. In One hundred years of English studies in Dutch universities: Seventeen papers read at the Centenary Conference, Groningen, 15–16 January 1986, ed. G.H.V. Blunt, E.S. Kooper, J.L. Mackenzie & D.R.M. Wilkinson. Amsterdam: Rodopi. 139–155. Denison, David. 1993. English historical syntax: Verbal constructions. London: Longman. Denison, David. 1998. Syntax. In The Cambridge History of the English Language, vol. 4: 1776–1997, ed. Suzanne Romaine. Cambridge: Cambridge University Press. 92–329.

264

References

Denison, David. 2007. Syntactic surprises in some English letters: The underlying progress of the language. In Germanic language histories ‘from below’ (1700– 2000), ed. Stephan Elspaß, Nils Langer, Joachim Scharloth & Wim Vandenbussche. Berlin: de Gruyter. 115–127. Denison, David. 2008. Patterns and productivity. In Studies in the History of the English Language IV: Empirical and analytical advances in the study of English language change, ed. Susan M. Fitzmaurice & Donka Minkova. Berlin: Mouton de Gruyter. 207–230. Diensberg, Bernhard. 1983. Zur Genese und Entwicklung des neuenglischen Phrasal Verbs. Folia Linguistica Historica 4. 247–270. Diensberg, Bernhard. 1990. English phrasal verbs expressing aspect and aktionsart? [Review article of Laurel J. Brinton, The development of English aspectual systems: Aspectualizers and post-verbal particles (Cambridge: Cambridge University Press, 1988)]. Folia Linguistica Historica 11. 187–197. Dietz, Klaus. 2002. Lexikalischer Transfer und Wortbildung am Beispiel des französischen Lehngutes im Mittelenglischen. In Historische Wortbildung des Deutschen, ed. Mechthild Habermann, Peter O. Müller & Horst Haider Munske. Tübingen: Niemeyer. 381–405. Dietz, Klaus. 2004. Die altenglischen Präfixbildungen und ihre Charakteristik. Anglia 122. 561–613. Dietz, Klaus. 2005. Die altenglischen Nominalpräfixe æ- und o-, das Verbalpräfix a- und ihre althochdeutschen Entsprechungen: Ein Beitrag zur vergleichenden Wortbildung der altgermanischen Sprachen. Sprachwissenschaft 30. 1–47. Diewald, Gabriele. 1997. Grammatikalisierung: Eine Einführung in Sein und Werden grammatischer Formen. Tübingen. Niemeyer. Diewald, Gabriele. 2007. Konstruktionen in der diachronen Sprachwissenschaft. In Konstruktionsgrammatik: Von der Anwendung zur Theorie, ed. Kerstin Fischer & Anatol Stefanowitsch. Rev. 2nd edn. [1st edn. 2006]. Tübingen: Stauffenburg. 79–103. Dik, Simon C. 1997. The theory of Functional Grammar, part 1: The structure of the clause. 2nd, rev. edn., ed. Kees Hengeveld. Berlin: Mouton de Gruyter. Dikken, Marcel den. 1995. Particles: On the syntax of verb-particle, triadic, and causative constructions. New York: Oxford University Press. Dirven, René. 2001. The metaphoric in recent cognitive approaches to English phrasal verbs. Metaphorik.de 1. 39–54. . Dixon, Robert M.W. 1991. A new approach to English grammar, on semantic principles. Oxford: Clarendon Press. Dobbie, Elliot Van Kirk (ed.). 1953. Beowulf. In The Anglo-Saxon Poetic Records, vol. 4: Beowulf and Judith. London: Routledge. DOE = Dictionary of Old English. 1986–. Ed. by Angus Cameron, Ashley C. Amos & Antonette diPaolo Healey. Toronto: Pontifical Institute of Mediaeval Studies. . (Dictionary of Old English in Electronic Form A–F. 2003. Center for Medieval Studies, University of Toronto.)

References

265

DOE Corpus = Antonette diPaolo Healey (ed.). 1997. Dictionary of Old English Corpus online. . Donaldson, Bruce C. 1993. A Grammar of Afrikaans. Berlin: Mouton de Gruyter. Donalies, Elke. 1999. Präfixverben, Halbpräfixverben, Partikelverben, Konstitutionsverben oder verbale Gefüge? – Ein Analyseproblem der deutschen Wortbildung. Studia Germanica Universitatis Vesprimiensis 3. 127–143. Dongen, W.A. van. 1919. ‘He put on his hat’ and ‘He put his hat on’. Neophilologus 4. 322–353. Dryer, Matthew S. 1995. Word order typology. In Syntax: Ein internationales Handbuch zeitgenössischer Forschung (An International Handbook of Contemporary Research), vol. 2, ed. Joachim Jacobs, Arnim von Stechow, Wolfgang Sternefeld & Theo Vennemann. Berlin: de Gruyter. 1050–1065. Duden 2005 = Der Duden, vol. 4: Die Grammatik. 7th edn. Mannheim: Bibliographisches Institut. Dufresne, Monique, Fernande Dupuis & Mireille Tremblay. 2006. Preverbs and particles in Old French. In Yearbook of Morphology 2003, ed. Geert Booij & Jaap van Marle. Dordrecht: Kluwer. 33–60. Dumville, David (ed.). 1995. The Anglo-Saxon Chronicle. A collaborative edition, general eds. David Dumville & Simon Keynes, vol. 1: Facsimile of MS. F: The Domitian Bilingual. Cambridge: Brewer. Dunkel, George E. 1992. Die Grammatik der Partikeln. In Rekonstruktion und relative Chronologie: Akten der der VIII. Fachtagung der Indogermanischen Gesellschaft, Leiden, 31. August – 4. September 1987, ed. Robert Beekes, Alexander Lubotsky & Jos Weitenberg. Innsbruck: Institut für Sprachwissenschaft der University Innsbruck. 153–181. Eisenberg, Peter. 2006. Grundriss der deutschen Grammatik. 2 vols. (vol. 1: Das Wort; vol. 2: Der Satz). 3rd edn. Stuttgart: Metzler. Eitrem, H. 1903. Stress in English verb+adverb groups. Englische Studien 32. 69– 77. Elenbaas, Marion. 2007. The synchronic and diachronic syntax of the English verb-particle combination. Utrecht: LOT. Ellinger, J. 1910. Über die Betonung der aus Verb+Adverb bestehenden englischen Wortgruppen. In Programm der k. k. Franz-Joseph-Realschule in Wien. [Quoted in Kennedy 1920 and in Bacchielli 1999; n.p., n.pag.; cf. the review by Albert Eichler in Zeitschrift für Österreichische Gymnasien 64, 1913, at 185.] Emonds, Joseph. 1993. Projecting indirect objects. The Linguistic Review 10. 211– 263. Engel, Ulrich. 1988. Deutsche Grammatik. Heidelberg: Groos. Erben, Johannes. 2006. Einführung in die deutsche Wortbildungslehre. 5th edn. Berlin: Schmidt. EWA = Etymologisches Wörterbuch des Althochdeutschen, ed. Albert L. Lloyd & Otto Springer [et al.]. 1988– . 10 vols. [published so far: vol. 1–3 (to fûstslag)]. Göttingen: Vandenhoeck & Ruprecht.

266

References

Faarlund, Jan Terje. 1995. De la préposition au préverbe en Nordique. In Les préverbes dans les langues d’Europe: Introduction à l’étude de la préverbation, ed. André Rousseau. Lille: Presses Universitaires du Septentrion. 61–75. Faarlund, Jan Terje. 2001. From ancient Germanic to modern Germanic Languages. In Language Typology and Language Universals: An International Handbook, vol. 2, ed. Martin Haspelmath, Ekkehard König, Wulf Oesterreicher & Wolfgang Raible. Berlin: de Gruyter. 1706–1719. Farrell, Patrick. 2005. English verb-preposition constructions: Constituency and order. Language 81. 96–137. Feilke, Helmut. 2007. [Syntaktische Aspekte der Phraseologie III:] Construction Grammar und verwandte Ansätze. In Phraseologie: Ein internationales Handbuch der zeitgenössischen Forschung (Phraseology: An International Handbook of Contemporary Research), vol. 1, ed. Harald Burger, Dmitrij Dobrovol’skij, Peter Kühn & Neal R. Norrick. Berlin: de Gruyter. 63–76. Finegan, Edward. 1998. English grammar and usage. In The Cambridge History of the English Language, vol. 4: 1776–1997, ed. Suzanne Romaine. Cambridge: Cambridge University Press. 536–588. Fischer, Kerstin & Anatol Stefanowitsch (eds.). 2007. Konstruktionsgrammatik: Von der Anwendung zur Theorie. Rev. 2nd edn. [1st edn. 2006]. Tübingen: Stauffenburg. Fischer, Olga. 1992. Syntax. In The Cambridge History of the English Language, vol. 2: 1066–1476, ed. Norman F. Blake. Cambridge: Cambridge University Press. 207–408. Fischer, Olga. 2004. Grammar change versus language change: Is there a difference? In New perspectives on English historical linguistics: Selected papers from 12 ICEHL, Glasgow, 21–26 August 2002, vol. 1: Syntax and morphology, ed. Christian Kay, Simon Horobin & Jeremy Smith. Amsterdam: Benjamins. 31–63. Fischer, Olga. 2007. Morphosyntactic change: Functional and formal perspectives. Oxford: Oxford University Press. Fischer, Olga. 2008. Is there life beyond generative syntax? Beiträge zur Geschichte der deutschen Sprache und Literatur 130. 199–235. Fischer, Olga, Ans van Kemenade, Willem Koopman & Wim van der Wurff (eds.). 2000. The syntax of early English. Cambridge: Cambridge University Press. Fischer, Olga & Wim van der Wurff. 2006. Syntax. In A History of the English Language, ed. Richard Hogg & David Denison. Cambridge: Cambridge University Press. 109–198. Fischer, Susann. 2010. Word order change as a source of grammaticalization. Amsterdam: Benjamins. Fleischer, Jürg. 2006. Zur Methodologie althochdeutscher Syntaxforschung. Beiträge zur Geschichte der deutschen Sprache und Literatur 128. 25–69. Fleischer, Jürg. 2009. Paleographic clues to prosody? – Accents, word separation, and other phenomena in Old High German manuscripts. In Information structure and language change: New approaches to word order variation and

References

267

change, ed. Roland Hinterhölzl & Svetlana Petrova. Berlin: Mouton de Gruyter. 160–189. Fleischer, Wolfgang. 2000. Die Klassifikation von Wortbildungsprozessen. In Morphologie: Ein internationales Handbuch zur Flexion und Wortbildung (Morphology: An International Handbook on Inflection and Word-formation), vol. 1, ed. Geert Booij, Christian Lehmann & Joachim Mugdan, in collaboration with Wolfgang Kesselheim & Stavros Skopeteas. Berlin: de Gruyter. 892– 897. Fleischer, Wolfgang & Irmhild Barz. 1995. Wortbildung der deutschen Gegenwartssprache. 2nd edn. Niemeyer: Tübingen. Foltinek, Herbert. 1964. Die Wortverbindung Verbum +up oder out im modernen Englisch. Moderne Sprachen 8. 94–104. Fortson, Benjamin W., IV. 2004. Indo-European language and culture: An introduction. Oxford: Blackwell. Fourquet, Jean. 1938. L’ordre des éléments de la phrase en germanique ancien: Études de syntaxe de position. Strasbourg: Publications de la Faculté des Lettres de l’Université de Strasbourg. Fowler, Henry Watson. 1931. The King’s English. 3rd edn. Oxford: Clarendon Press. Fowler, Henry Watson. 1965. A Dictionary of Modern English Usage. 2nd edn. rev. by Ernest Gowers. Oxford: Clarendon Press. Fowler, Henry Watson. 1996. The new Fowler’s Modern English Usage. 3rd edn. by Robert William Burchfield. Oxford: Clarendon Press. Fraser, Bruce. 1965. An examination of the verb-particle construction in English. Cambridge, MA: M.I.T PhD dissertation. Fraser, Bruce. 1966. Some remarks on the verb-particle construction in English. In Report of the Seventh Annual Round Table Meeting on Linguistics and Language Studies. Washington: Georgetown University Press. Fraser, Bruce. 1970. Idioms within a Transformational Grammar. Foundations of Language 6. 22–42. Fraser, Bruce. 1974. Review of Dwight Bolinger, The phrasal verb in English (Cambridge, MA: Harvard University Press, 1971). Language 50. 568–575. Fraser, Bruce. 1976. The verb-particle combination in English. New York: Academic Press. Friedrich, Paul. 1975. Proto-Indo-European syntax: The order of meaningful elements. (Journal of Indo-European Studies. Monograph Series 1.) Butte: Journal of Indo-European Studies. Fries, Charles C. 1940. On the development of the structural use of word-order in Modern English. Language 16. 199–208. Fritz, Matthias A. 2005. Die trikasuellen Lokalpartikeln bei Homer: Syntax und Semantik. Göttingen: Vandenhoeck & Ruprecht. Frost, William & Vinton A. Dearing (eds.). 1987. The Works of John Dryden, vol. 5: Poems: The works of Virgil in English, 1697. Berkeley: University of California Press.

268

References

Gaaf, W. van der. 1930. The passive of a verb accompanied by a preposition. English Studies 12. 1–24. Gamkrelidze, Thomas V. & Vjačeslav V. Ivanov. 1995. Indo-European and the Indo-Europeans: A reconstruction and historical analysis of a proto-language and a proto-culture, part 1: The text. (With a preface by Roman Jakobson, English version by Johanna Nichols, ed. by Werner Winter.) Berlin: Mouton de Gruyter. Gerritsen, Marinel. 1984. Divergent word-order developments in Germanic languages: A description and tentative explanation. In Historical syntax, ed. Jacek Fisiak. Berlin: Mouton. 107–135. Givón, Talmy. 1979. Understanding grammar. New York: Academic Press. Gneuss, Helmut. 1973. Guide to the editing and preparation of texts for the Dictionary of Old English. In A Plan for the Dictionary of Old English, ed. Roberta Frank & Angus Cameron. Toronto: University of Toronto Press. 9–24. Gneuss, Helmut. 1996. English language scholarship: a survey and bibliography from the beginning to the end of the nineteenth century. Birminghamton, NY: Center for Medieval and Early Renaissance Studies. Godden, Malcolm (ed.). 1979. Ælfric’s Catholic Homilies: The second series. EETS SS 5. Goldberg, Adele E. 1995. A Construction Grammar approach to argument structure. Chicago: Chicago University Press. Goldberg, Adele E. 2006. Constructions at work: The nature of generalization in language. Oxford: Oxford University Press. Görlach, Manfred. 1994. Einführung ins Frühneuenglische. 2nd edn. Heidelberg: Winter. Görlach, Manfred. 1998. An annotated bibliography of 19th-century grammars of English. With a foreword by Ian Michael. Amsterdam: Benjamins. Görlach, Manfred. 1999. Regional and social variation. In The Cambridge History of the English Language, vol. 3: 1476–1776, ed. Roger Lass. Cambridge: Cambridge University Press. 459–538. Görlach, Manfred. 2001. Eighteenth-century English. Heidelberg: Winter. Götti, Ernst. 1974. Die gotischen Bewegungsverben: Ein Beitrag zur Erforschung des gotischen Wortschatzes mit einem Ausblick auf Wulfilas Übersetzungstechnik. Berlin: de Gruyter. Goolden, Peter (ed.). 1958. The Old English Apollonius of Tyre. London: Oxford University Press. Gowers, Ernest. 1954. The complete Plain Words. London: Her Majesty’s Stationery Office. Gowers, Ernest. 1973. The complete Plain Words. 2nd rev. edn. by Bruce Fraser. London: Her Majesty’s Stationery Office. Greenbaum, Sidney. 2000. The Oxford Reference Grammar, ed. Edmund Weiner. Oxford: Oxford University Press. Greenberg, Joseph H. 1966 [1963]. Some universals of grammar with particular reference to the order of meaningful elements. Universals of Grammar, ed. Joseph H. Greenberg. 2nd edn. Cambridge: MIT Press. 73–113.

References

269

Grein, C.W.M. 1861–1864. Sprachschatz der angelsächsischen Dichter. Erster Band: A–G (1861); zweiter Band: H–Z. Eigennamen. Nachträge (1864). Kassel: Wigand. Gries, Stefan Thomas. 2003. Multifactorial analysis in corpus linguistics: A study of particle placement. New York: Continuum. Haase, Martin. 2001. Lokalkasus und Adpositionen. In Language Typology and Language Universals: An International Handbook, vol. 1, ed. Martin Haspelmath, Ekkehard König, Wulf Oesterreicher & Wolfgang Raible. Berlin: de Gruyter. 736–740. Habermann, Mechthild. 1994. Verbale Wortbildung um 1500: Eine historischsynchrone Untersuchung anhand von Texten Albrecht Dürers, Heinrich Deichslers und Veit Dietrichs. Berlin: de Gruyter. Hacken, Pius ten. 2000. Inflection and derivation. In Morphologie: Ein internationales Handbuch zur Flexion und Wortbildung (Morphology: An International Handbook on Inflection and Word-formation), vol. 1, ed. Geert Booij, Christian Lehmann & Joachim Mugdan, in collaboration with Wolfgang Kesselheim & Stavros Skopeteas. Berlin: de Gruyter. 349–360. Hackstein, Olav. 1997. Präverb, Post- und Präposition im Tocharischen: Ein Beitrag zur Rekonstruktion der urindogermanischen Syntax. Tocharian and IndoEuropean Studies 7. 35–60. Hackstein, Olav. 2003. Reflexivpronomen, Präverbien und Lokalpartikel in indogermanischen Sprachen. Tocharian and Indo-European Studies 10. 69–95. Hackstein, Olav. 2011. Proklise und Subordination im Indogermanischen. In Indogermanistik und Linguistik im Dialog: Akten der XIII. Fachtagung der Indogermanischen Gesellschaft vom 21. bis 27. September 2008 in Salzburg, ed. Thomas Krisch & Michael Lindner, with Michael Crombach & Stefan Niederreiter. Wiesbaden: Reichert. 192–202. Hale, Kenneth L., Mary Laughren & Jane Simpson. 1995. Warlpiri. In Syntax: Ein internationales Handbuch zeitgenössischer Forschung (An International Handbook of Contemporary Research), vol. 2, ed. Joachim Jacobs, Arnim von Stechow, Wolfgang Sternefeld & Theo Vennemann. Berlin: de Gruyter. 1430– 1451. Hall, Christopher J. 2000. Prefixation, suffixation and circumfixation. In Morphologie: Ein internationales Handbuch zur Flexion und Wortbildung (Morphology: An International Handbook on Inflection and Word-formation), vol. 1, ed. Geert Booij, Christian Lehmann & Joachim Mugdan, in collaboration with Wolfgang Kesselheim & Stavros Skopeteas. Berlin: de Gruyter. 535– 545. Hall, J.R. Clark. 1960. A Concise Anglo-Saxon Dictionary. 4th edn., with a supplement by Herbert D. Meritt. (1st edn. 1894.) Cambridge: Cambridge University Press. (Repr. Toronto: University of Toronto Press, 1996.) Hampe, Beate 2002. Superlative verbs: A corpus-based study of semantic redundancy in English verb-particle constructions. Tübingen: Narr. Harbert, Wayne. 2007. The Germanic languages. Cambridge: Cambridge University Press.

270

References

Harnisch, Karl-Rüdiger. 1982. ‘Doppelpartikelverben’ als Gegenstand der Wortbildungslehre und Richtungsadverbien als Präpositionen: Ein syntaktischer Versuch. In Tendenzen verbaler Wortbildung in der deutschen Gegenwartssprache, ed. Ludwig M. Eichinger. Hamburg: Buske. 107–133. Harris, Alice C. 1985. Diachronic syntax: The Kartvelian case. Orlando: Academic Press. Harris, Alice C. 2003. Preverbs and their origins in Georgian and Udi. In Yearbook of Morphology 2003, ed. Geert Booij & Jaap van Marle. Dordrecht: Kluwer. 61–78. Harris, Alice C. (ed.). 1991. The Kartvelian languages. (Vol. 1 of The indigenous languages of the Caucasus, general ed. John A.C. Greppin.) Delmar, NY: Caravan Books. Harris, Alice C. & Lyle Campbell. 1995. Historical syntax in cross-linguistic perspective. Cambridge: Cambridge University Press. Harris, James. 1781. Philological inquiries in three parts. (3 vols.) London: Nourse. (Repr. New York: AMS, 1975.) Harris, Martin & Nigel Vincent (eds.). 1988. The Romance languages. London: Croom Helm. Harrison, Thomas Perrin. 1892. The separable prefixes in Anglo-Saxon. (PhD thesis, Johns Hopkins University.) Baltimore: Murphy. (Repr. College Park: McGrath, 1970.) Haspelmath, Martin. 2000. Periphrasis. In Morphologie: Ein internationales Handbuch zur Flexion und Wortbildung (Morphology: An International Handbook on Inflection and Word-formation), vol. 1, ed. Geert Booij, Christian Lehmann & Joachim Mugdan, in collaboration with Wolfgang Kesselheim & Stavros Skopeteas. Berlin: de Gruyter. 654–664. Haspelmath, Martin. 2002. Understanding morphology. London: Arnold. Haspelmath, Martin. 2007. Pre-established categories don’t exist: Consequences for language description and typology. Linguistic Typology 11. 119–132. Haspelmath, Martin. 2008. Parametric versus functional explanations of syntactic universals. In The limits of syntactic variation, ed. Theresa Biberauer. Amsterdam: Benjamins. 75–107. Haspelmath, Martin. 2010. Framework-free grammatical theory. In The Oxford handbook of linguistic analysis, ed. Bernd Heine & Heiko Narrog. Oxford: Oxford University Press. 341–365. Haspelmath, Martin & Thomas Müller-Bardey. 2004. Valency change. In Morphologie: Ein internationales Handbuch zur Flexion und Wortbildung (Morphology: An International Handbook on Inflection and Word-formation), vol. 2, ed. Geert Booij, Christian Lehmann, Joachim Mugdan & Stavros Skopeteas, in collaboration with Wolfgang Kesselheim. Berlin: de Gruyter. 1130–1145. Hasselblatt, Cornelius. 1990. Das estnische Partikelverb als Lehnübersetzung aus dem Deutschen. Wiesbaden. Harrassowitz. Haverling, Gerd. 2003. On prefixes and actionality in Classical and Late Latin. Acta Linguistica Hungarica 50. 113–135.

References

271

Hecht, Hans (ed.). 1900–1907. Bischof Wærferths von Worcester Übersetzung der Dialoge Gregors des Grossen über das Leben und die Wundertaten italienischer Väter und die Unsterblichkeit der Seelen, aus dem Nachlasse von Julius Zupitza, nach einer Kopie von Henry Johnson. 2 vols. 1900 (1. Abt.), 1907 (2. Abt). Leipzig (vol. 1) and Hamburg (vol. 2): Bibliothek der Angelsächsischen Prosa 5. (Repr. Darmstadt: Wissenschaftliche Buchgesellschaft, 1965). Heine, Bernd, Ulrike Claudi & Friederike Hünnemeyer. 1991. Grammaticalization: A conceptual framework. Chicago: University of Chicago Press. Heine, Bernd & Tania Kuteva. 2002. World lexicon of grammaticalization. Cambridge: Cambridge University Press. Helbig, Gerhard & Joachim Buscha. 2001. Deutsche Grammatik: Ein Handbuch für den Ausländerunterricht. Berlin: Langenscheidt. Herbers, Birgit. 2002. Verbale Präfigierung im Mittelhochdeutschen: Eine semantisch-funktionale Korpusanalyse. Tübingen: Niemeyer. Herbst, Thomas & Michael Klotz. 2009. Syntagmatic and phraseological dictionaries. In The Oxford History of English Lexicography, ed. A.P. Cowie, vol. 2: Specialized dictionaries. Oxford: Clarendon Press. 219–244. Hill, Leslie A. 1968. Prepositions and adverbial particles: An interim classification, semantic, structural, and graded. London: Oxford University Press. Hiltunen, Risto. 1983a. The decline of the prefixes and the beginnings of the English phrasal verb: The evidence from some Old and early Middle English texts. Turku: Turun Yliopisto. Hiltunen, Risto. 1983b. Phrasal verbs in English grammar books before 1800. Neuphilologische Mitteilungen 84. 376–386. Hiltunen, Risto. 1994. On phrasal verbs in Early Modern English: Notes on lexis and style. In Studies in Early Modern English, ed. Dieter Kastovsky. Berlin: Mouton de Gruyter. 129–140. Hiltunen, Risto. 1999. Verbal phrases and phrasal verbs in Early Modern English. In Collocational and idiomatic aspects of composite predicates in the history of English, ed. Laurel J. Brinton & Minoji Akimoto. Amsterdam: Benjamins. 133–165. Hinderling, Robert. 1982. Konkurrenz und Opposition in der verbalen Wortbildung. In Tendenzen verbaler Wortbildung in der deutschen Gegenwartssprache, ed. Ludwig M. Eichinger. Hamburg: Buske. 81–106. Hinterhölzl, Roland & Svetlana Petrova (eds.). 2009. Information structure and language change: New approaches to word order variation and change. Berlin: Mouton de Gruyter. Hirt, Hermann. 1928. Indogermanische Grammatik, part 4: Doppelung – Zusammensetzung – Verbum. Heidelberg: Winter. Hock, Hans Henrich. 1991. Principles of historical linguistics. 2nd rev. edn. Berlin: Mouton de Gruyter. Hoekstra, Jarich F. 2001. Standard West Frisian. In Handbook of Frisian studies/Handbuch des Friesischen, ed. Horst Haider Munske. Tübingen: Niemeyer. 83–98.

272

References

Hoffmann, Sebastian. 2004. Using the OED quotations database as a corpus: A linguistic appraisal. ICAME Journal 28. 17–30. . Hoffner, Harry A. Jr. & H. Craig Melchert. 2008. A Grammar of the Hittite Language, part 1: Reference grammar. Winona Lake: Eisenbrauns. Holmberg, Anders & Jan Rijkhoff. 1998. Word order in the Germanic languages. In Constituent order in the languages of Europe, ed. Anna Siewierska. Berlin: Mouton de Gruyter. 75–104. Holmes, Philip & Ian Hinchliffe. 1994. Swedish: A Comprehensive Grammar. London: Routledge. Hopper, Paul J. 1975. The syntax of the simple sentence in Proto-Germanic. The Hague: Mouton. Hopper, Paul J. 1991. On some principles of grammaticization. In Approaches to grammaticalization, vol. 1: Focus on theoretical and methodological issues, ed. Elizabeth Closs Traugott & Bernd Heine. Amsterdam: Benjamins. 17–35. Hopper, Paul J. & Elizabeth Closs Traugott. 2003. Grammaticalization. 2nd edn. Cambridge: Cambridge University Press. Horgan, Dorothy. 1980. Patterns of variation and interchangeability in some Old English prefixes. Neuphilologische Mitteilungen 84. 127–130. Horobin, Simon & Jeremy Smith. 2002. An introduction to Middle English. Edinburgh: Edinburgh University Press. Horobin, Simon. 2007. Chaucer’s language. Basingstoke: Palgrave. Hückel, W. 1968. Einige Fragen des Wortverbandes im Englischen. Fremdsprachen 12. 252–260. Huddleston, Rodney & Geoffrey K. Pullum (eds.). 2002. The Cambridge Grammar of the English Language. Cambridge: Cambridge University Press. Hudson, Richard A. 1996. Sociolinguistics. 2nd edn. Cambridge: Cambridge University Press. Hughes, Geoffrey. 2000. A history of English words. Oxford: Blackwell. Iacobini, Claudio & Francesca Masini. 2006. The emergence of verb-particle constructions in Italian: Locative and actional meanings. Morphology 16. 155–188. Iglesias-Rábade, Luis. 2011. Semantic erosion of Middle English prepositions. Frankfurt am Main: Lang. Irvine, Susan. 2004. The Anglo-Saxon Chronicle. A collaborative edition, general eds. David Dumville & Simon Keynes, vol. 7: MS E (A semi-diplomatic edition with introduction and indices). Cambridge: Brewer. Jackendoff, Ray. 1997. Twistin’ the night away. Language 73. 534–595. Jackendoff, Ray. 2002. English particle constructions, the lexicon, and the autonomy of syntax. In Verb-particle explorations, ed. Nicole Dehé, Ray Jackendoff, Andrew McIntyre & Silke Urban. Berlin: Mouton de Gruyter. 67– 94. Jacobs, Neil. 2005. Yiddish: A linguistic introduction. Cambridge: Cambridge University Press.

References

273

Jacobs, Neil, Ellen Prince & Johan van der Auwera. 1994. Yiddish. In The Germanic languages, ed. Ekkehard König & Johan van der Auwera. London: Routledge. 388–419. Jacobsson, Bengt. 1951. Inversion in English: With special reference to the Early Modern English period. Uppsala: Almqvist & Wiksell. Jenkyns, Joy. 1991. The Toronto ‘Dictionary of Old English’ resources: A user’s view. The Review of English Studies n.s. 42. 380–416. Jespersen, Otto. 1905. Growth and structure of the English language. (2nd rev. edn. 1912.) Leipzig: Teubner. Jespersen, Otto. 1936. A Modern English Grammar on Historical Principles, part 3: Syntax (vol. 1). 4th edn. Heidelberg: Winter. Johansson, Stig. 1978. Manual of information to acompany the LancasterOslo/Bergen Corpus of British English, for use with digital computers (in collaboration with Geoffrey N. Leech & Helen Goodluck). Oslo: Department of English, University of Oslo. . Johansson, Stig & Knut Hofland. 1989. Frequency analysis of English vocabulary and grammar based on the LOB Corpus, vol. 2: Tag combinations and word combinations. Oxford: Clarendon. Johnson, Samuel. 1755. A Dictionary of the English Language. 2 vols. London. (Repr. New York: AMS, 1967.) Johnston, Andrew James. 2011. Calques and culture: Revisiting an issue in Old English lexical morphology. In More than words: English lexicography and lexicology past and present (Essays presented to Hans Sauer on the occasion of his 65th birthday: part I), ed. Renate Bauer & Ulrike Krischke. Frankfurt am Main: Lang. 67–79. Jones, Richard Foster. 1953. The triumph of the English language: A survey of opinions concerning the vernacular from the introduction of printing to the Restoration. Stanford: Stanford University Press. Jowett, W.P. 1951. On phrasal verbs. English Language Teaching 5. 152–157. Kastovsky, Dieter. 1982. Wortbildung und Semantik. Düsseldorf: Bagel. Kastovsky, Dieter. 1992. Semantics and vocabulary. In The Cambridge History of the English Language, vol. 1: The Beginnings to 1066, ed. Richard Hogg. Cambridge: Cambridge University Press. 290–408. Kemenade, Ans van. 1987. Syntactic case and morphological case in the history of English. Dordrecht: Foris. Kemenade, Ans van. 2003. Word order in Old English prose and poetry: The position of finite verbs and adverbs. In Studies in the history of the English language: A millennial perspective, ed. Donka Minkova & Robert Stockwell. Berlin: Mouton de Gruyter. 355–371. Kemenade, Ans van & Bettelou Los. 2003. Particles and prefixes in Dutch and English. In Yearbook of Morphology 2003, ed. Geert Booij & Jaap van Marle. Dordrecht: Kluwer. 79–117.

274

References

Kemenade, Ans van & Bettelou Los. 2006. Discourse adverbs and clausal syntax in Old and Middle English. In The handbook of the history of English, ed. Ans van Kemenade & Bettelou Los. Oxford: Blackwell. 224–248. Kennedy, Arthur Garfield. 1920. The Modern English verb-adverb combination. Palo Alto: Stanford University Press. Kenyon, John S. 1948. Levels of speech and colloquial English. The English Journal 37. 25–31. Ker, W.P. (ed.). 1900. Essays of John Dryden. 2 vols. Oxford: Clarendon Press. Kiefer, Ferenc & László Honti. 2003. Verbal ‘prefixation’ in the Uralic languages. Acta Linguistica Hungarica 50. 137–153. Kirkpatrick, John. 1912. Handbook of idiomatic English: As now written and spoken; containing idioms, phrases, and locutions; adapted for students and travellers of all nationalities. Edinburgh: Thin [also published elsewhere]. Klotz, Michael. 2000. Grammatik und Lexik: Studien zur Syntagmatik englischer Verben. Tübingen: Stauffenburg. Kluge–Seebold = Friedrich Kluge. 2002. Etymologisches Wörterbuch der deutschen Sprache. 24th edn. by Elmar Seebold. Berlin: de Gruyter. Knappe, Gabriele. 2004. Idioms and fixed expressions in English language study before 1800. Frankfurt am Main: Lang. Knorrek, Marianne. 1938. Der Einfluß des Rationalismus auf die englische Sprache: Beiträge zur Entwicklungsgeschichte der englischen Syntax im 17. und 18. Jahrhundert. Breslau: Priebatsch. Koch, C. Friedrich. 1891. Historische Grammatik der englischen Sprache, vol. 3: Die Wortbildung der englischen Sprache, zum Drucke besorgt von Richard Wülker. Kassel: Wigand. Koefoed, Geert & Jaap van Marle. 2000. Productivity. In Morphologie: Ein internationales Handbuch zur Flexion und Wortbildung (Morphology: An International Handbook on Inflection and Word-formation), vol. 1, ed. Geert Booij, Christian Lehmann & Joachim Mugdan, in collaboration with Wolfgang Kesselheim & Stavros Skopeteas. Berlin: de Gruyter. 303–311. Kohonen, Viljo. 1978. On the development of English word order in religious prose around 1000 and 1200 AD: A quantitative study of word order in context. Åbo: Åbo Akademi. Kolehmainen, Leena. 2006. Präfix- und Partikelverben im deutsch-finnischen Kontrast. Frankfurt am Main: Lang. König, Ekkehard. 1973. Englische Syntax, vol. 2: Struktur des einfachen Satzes. Frankfurt am Main: Athenäum Fischer. König, Ekkehard & Johan van der Auwera (eds.). 1994. The Germanic languages. London: Routledge. Konishi, Tomoshichi. 1958. The growth of the verb-adverb combination in English: A brief sketch. In Studies in English grammar and linguistics: A miscellany in honour of Takanobu Otsuka, ed. Kazuo Araki, Taiichiro Egawa, Toshiko Oyama & Minoru Yasui. Tokio: Kenyusha. 117–128.

References

275

Koopman, Willem F. 1985. The syntax of verb and particle combinations in Old English. In Linguistics in the Netherlands 1985, ed. Hans Bennis & Frits Beukema. Dordrecht: Foris. 91–99. Koopman, Willem F. 1992. The study of Old English syntax and the Toronto ‘Dictionary of Old English’. Neophilologus 76. 605–615. Kornexl, Lucia. 1994. Progress in historical lexicography: The ‘Dictionary of Old English’. Anglia 112. 421–453. Kornexl, Lucia. 2003. ‘Unnatural words’? Loan formation in Old English glosses. In Language contact in the history of English, ed. Dieter Kastovsky & Arthur Mettinger. 2nd edn. Frankfurt am Main: Lang. 195–216. Kortmann, Bernd. 1991. The triad ‘tense-aspect-aktionsart’. Perspectives on aspect and aktionsart (Belgian Journal of Linguistics 6), ed. Carl Vetters & Willy Vandeweghe. 9–30. Koziol, Herbert. 1972. Handbuch der englischen Wortbildungslehre. 2nd edn. Heidelberg: Winter. Krapp, George Philip (ed.). 1931. Exodus. In The Anglo-Saxon Poetic Records, vol. 1: The Junius Manuscript. New York: Columbia University Press. Kroch, Anthony & Ann Taylor. 2000. Verb-object order in early Middle English. In Diachronic syntax: Models and mechanisms, ed. Susan Pintzuk, George Tsoulas & Anthony Warner. Oxford: Oxford University Press. 132–163. Kroch, Anthony, Ann Taylor & Donald Ringe. 2000. The Middle English verbsecond constraint: A case study in language contact and language change. In Textual parameters in older languages, ed. Susan C. Herring, Pieter van Reenen & Lene Schøsler. Amsterdam: Benjamins. 353–391. Kruisinga, Etsko. 1925. A handbook of present-day English, part 2: English accidence and syntax. 4th edn. Groningen: Noordhoff. Kruisinga, Etsko. 1932. A handbook of present-day English, part 2: English accidence and syntax. 5th edn. Groningen: Noordhoff. Kuryłowicz, Jerzy. 1964. The inflectional categories of Indo-European. Heidelberg: Winter. Kytö, Merja. 1993. Manual to the diachronic part of the Helsinki Corpus of English Texts: Coding conventions and lists of source texts. 2nd edn. University of Helsinki: Department of English. Kytö, Merja, Ossi Ihalainen & Matti Rissanen (eds.). 1988. Corpus linguistics: Hard and soft. Amsterdam: Rodopi. Lakoff, George. 1993. The contemporary theory of metaphor. In Metaphor and thought, ed. Andrew Ortony. Cambridge: Cambridge University Press. 202– 251. Lambrecht, Knud. 1994. Information structure and sentence form: Topic, focus, and the mental representation of discourse referents. Cambridge: Cambridge University Press. Lang, Ewald & Gisela Zifonun (eds.). 1996. Deutsch: Typologisch. Berlin: de Gruyter. Lasch, Alexander & Alexander Ziem (eds.). 2011. Konstruktionsgrammatik III: Aktuelle Fragen und Lösungsansätze. Tübingen: Stauffenburg.

276

References

Lass, Roger. 1990. How to do things with junk: Exaptation in language evolution. Journal of Linguistics 26. 79–102. Lass, Roger. 1994. Old English: A historical linguistic companion. Cambridge: Cambridge University Press. Lass, Roger. 1997. Historical linguistics and language change. Cambridge: Cambridge University Press. Lass, Roger. 2000. Language periodization and the concept of ‘middle’. In Placing Middle English in context, ed. Irma Taavitsainen, Terttu Nevalainen, Päivi Pahta & Matti Rissanen. Berlin: Mouton de Gruyter. 7–41. Lass, Roger. 2004. ‘Ut custodiant litteras’: Editions, corpora and witnesshood. In Methods and data in English historical dialectology, ed. Marina Dossena, Roger Lass & Maurizio Gotti. Bern: Lang. 21–48. Lavotha, Ödön. 1960. Észt Nyelvkönyv. Budapest: Tankönyvkiadó. Lavotha, Ödön. 1973. Kurzgefaßte estnische Grammatik [German translation of Lavotha 1960, by Hans-Hermann Bartens]. Wiesbaden: Harrassowitz. Lehmann, Christian. 1983. Latin preverbs and cases. In Latin linguistics and linguistic theory: Proceedings of the 1st International Colloquium on Latin Linguistics, Amsterdam, April 1981. Amsterdam: Benjamins. 145–161. Lehmann, Christian. 1985. Grammaticalization: Synchronic variation and diachronic change. Lingua e Stile 20. 303–318. Lehmann, Christian. 1989. Grammatikalisierung und Lexikalisierung. Zeitschrift für Phonetik, Sprachwissenschaft und Kommunikationsforschung 42. 11–19. Lehmann, Christian. 1995. Thoughts on grammaticalization. Revised and expanded version; 1st published version [draft version from 1982]. München: LINCOM Europa. Lehmann, Christian. 2005. Pleonasm and hypercharacterisation. In Yearbook of Morphology 2005, ed. Geert Booij & Jaap van Marle. Dordrecht: Springer. 119–154. Lehmann, Winfried P. 1972. Proto-Germanic syntax. In Toward a Proto-Germanic syntax, ed. Frans van Koetsem & Herbert L. Kufner. Tübingen: Niemeyer. 239–268. Lehmann, Winfried P. 1973. A structural principle of language and its implications. Language 49. 47–66. Lehmann, Winfried P. 1986. A Gothic Etymological Dictionary. Based on the 3rd edn. of Vergleichendes Wörterbuch der gotischen Sprache by Sigmund Feist; with bibliography prepared under the direction of Helen-Jo J. Hewitt. Leiden: Brill. Lehmann, Winfried P. 1993. Theoretical bases of Indo-European linguistics. London: Routledge. Leinen, Rudolf. 1891. Über Wesen und Entstehung der trennbaren Zusammensetzung des deutschen Zeitwortes mit besonderer Berücksichtigung des Gotischen und Althochdeutschen. Strassburg: Heitz. Leisi, Ernst. 1985. Das heutige Englisch: Wesenszüge und Probleme. 7th edn. Heidelberg: Winter.

References

277

Leiss, Elisabeth. 1992. Die Verbalkategorien des Deutschen: Ein Beitrag zur Theorie der sprachlichen Kategorisierung. Berlin: de Gruyter. Leiss, Elisabeth. 2002a. Die Rolle der Kategorie des Aspekts im Sprachwandel im Deutschen: Ein Überblick. In Grammatische Kategorien aus sprachhistorischer und typologischer Sicht: Akten des 29. Linguisten-Seminars (Kyoto, 2001), hg. von der Japanischen Gesellschaft für Germanistik. München: Iudicium. 9–25. Leiss, Elisabeth. 2002b. Der Verlust der aspektuellen Verbpaare und seine Folgen im Bereich der Verbalkategorien des Deutschen. In Grammatische Kategorien aus sprachhistorischer und typologischer Sicht: Akten des 29. LinguistenSeminars (Kyoto, 2001), hg. von der Japanischen Gesellschaft für Germanistik. München: Iudicium. 26–41. Lenker, Ursula. 2008. Booster prefixes in Old English: An alternative view of the roots of ME forsooth. English Language and Linguistics 12. 245–265. Lenker, Ursula. 2010. Argument and rhetoric: Adverbial connectors in the history of English. Berlin: De Gruyter Mouton. Leonard, Sterling Andrus. 1929. The doctrine of correctness in English Usage 1700–1800. Madison: University of Wisconsin Studies in Language and Literature. Li, Charles N. & Sandra A. Thompson. 1974. Historical change of word order: A case study in Chinese and its implications. In Historical linguistics I: Syntax, morphology, internal and comparative reconstruction (Proceedings of the First International Conference on Historical Linguistics, Edinburgh 2nd–7th September 1973), ed. John M. Anderson & Charles Jones. Amsterdam: NorthHolland. 199–217. Liebermann, Felix (ed.). 1903. Das Gesetzbuch der Könige Alfred–Ine. In Die Gesetze der Angelsachsen, vol. 1: Text und Übersetzung. Halle: Niemeyer. Lindelöf, Uno. 1935. English agent-nouns with a suffixed adverb. Neuphilologische Mitteilungen 36. 257– 282. Lindelöf, Uno. 1937. English verb-adverb groups converted into nouns. Helsinki: Akademische Buchhandlung. Lindemann, J.W. Richard. 1965. Old English preverb ge-: A re-examination of some current doctrines. Journal of English and Germanic Philology 64. 65–83. Lindemann, J.W. Richard. 1970. Old English preverbal ge-: Its meaning. Charlottesville: University Press of Virginia. Lindquist, Hans. 2009. Corpus linguistics and the description of English. Edinburgh: Edinburgh University Press. Lipka, Leonhard. 1972. Semantic structure and word-formation: Verb-particle constructions in contemporary English. München: Fink. Liuzza, Roy M. (ed.). 1994. The Old English version of the Gospels, vol. 1: Text and introduction. EETS 304. Live, Anna H. 1965. The discontinuous verb in English. Word 21. 428–451. Lloyd, Albert L. 1979. Anatomy of the verb: The Gothic verb as a model for a unified theory of aspect, actional types, and verbal velocity. Amsterdam: Benjamins.

278

References

Loftis, John, David Stuart Rodes & Vinton A. Dearing [et al.] (eds.). 1978. The Works of John Dryden, vol. 11: Plays: The Conquest of Granada, Marriage ala-mode, The Assignation. Berkeley: University of California Press. Logeman, Henri. 1906. On some cases of Scandinavian influence in English. Archiv für das Studium der neueren Sprachen und Literaturen 117 (n.s. 17). 29–46 (I-III) and 268–286 (IV-V). Lohse, Barbara, John A. Hawkins & Thomas Wasow. 2004. Domain minimization in English verb-particle constructions. Language 30. 238–261. Longman Phrasal Verbs Dictionary. 2000. Ed. Adam Gadsby. Harlow: Longman. Los, Bettelou. 2004. From resultative predicate to event-modifier: The case of forth and on. In New perspectives on English historical linguistics: Selected papers from 12 ICEHL, Glasgow, 21–26 August 2002, vol. 1: Syntax and morphology, ed. Christian Kay, Simon Horobin & Jeremy Smith. Amsterdam: Benjamins. 83–102. Los, Bettelou. 2005. The rise of the to-infinitive. Oxford: Oxford University Press. Lowth, Robert. 1762. A Short Introduction to English Grammar. (Facs. repr. Menston: Scolar Press, 1969.) LPD = John C. Wells. 1990. Longman Pronunciation Dictionary. London: Longman. Lundskær-Nielsen, Tom. 1993. Prepositions in Old and Middle English: A study of prepositional syntax and the semantics of At, In and On in some Old and Middle English texts. Odense: Odense University Press. Lutz, Angelika. 1991. Phonotaktisch gesteuerte Konsonantenveränderungen in der Geschichte des Englischen. Tübingen: Niemeyer. Lutz, Angelika. 1992. Lexical and morphological consequences of phonotactic change in the history of English. In History of Englishes: New methods and interpretations in historical linguistics, ed. Matti Rissanen, Ossi Ihalainen, Terttu Nevalainen & Irma Taavitsainen. Berlin: Mouton de Gruyter. 156–166. Lutz, Angelika. 1997. Sound change, word formation and the lexicon: The history of the English prefix verbs. English Studies 78. 258–290. Lutz, Angelika. 2002a. When did English begin? In Sounds, words, texts and change: Selected papers from 11 ICEHL, Santiago de Compostela, 7–11 September 2000, ed. Teresa Fanego & Elena Seoane. Amsterdam: Benjamins. 145– 171. Lutz, Angelika. 2002b. Sprachmischung in der deutschen und englischen Wortbildung. In Historische Wortbildung des Deutschen, ed. Mechthild Habermann, Peter O. Müller & Horst Haider Munske. Tübingen: Niemeyer. 407–437. Lutz, Angelika. 2008. Types and degrees of mixing: A comparative assessment of Latin and French influences on English and German word formation. Interdisciplinary Journal for Germanic Linguistics and Semiotic Analysis 13. 131–165. Mack, Frances M. (ed.). 1934. Seinte Marherete: Þe meiden ant martyr. Re-ed. from MS. Bodley 34, Oxford, and MS. Royal 17A xxvii, British Museum. EETS 193.

References

279

MacLeish, Andrew. 1969. The Middle English subject-verb cluster. The Hague: Mouton. Macmillan Phrasal Verbs Plus. 2005. Oxford: Macmillan. Mätzner, Eduard. 1878–1896. Altenglische Sprachproben: Nebst einem Wörterbuche. Zweiter Band: Wörterbuch. Erste Abtheilung: A–D (1878); zweite Abtheilung: E–H (1885); dritte Abtheilung: I–misbileven (1896, with Hugo Bieling). Berlin: Weidmannsche Buchhandlung. Mahler, Andreas. 2002. ‘This is the kind of thing up with which I will not put’: Einige Beobachtungen zu Syntax und Morphologie komplementierter ‘multiword verbs’ im Englischen. Anglia 120. 508–533. Mair, Christian. 2006. Twentieth-century English: history, variation and standardization. Cambridge: Cambridge University Press. Makkai, Adam. 1972. Idiom structure in English. The Hague: Mouton. Malkiel, Yakov. 1978. Derivational categories. In Universals of human language, vol. 3: Word structure, ed. Joseph H. Greenberg. Stanford: Stanford University Press. 125–149. Marchand, Hans. 1969. The categories and types of present-day English wordformation: A synchronic-diachronic approach. 2nd edn., completely revised and enlarged. München: Beck. Marks, Jonathan. 2005a. Phrasaled out? Don’t worry! Help is at hand. In MED Magazine 32. . Marks, Jonathan. 2005b. Phrasal verbs international. In MED Magazine 33. . Marks, Jonathan. 2005c. The truth revealed: Phrasal verbs in writing and speech. In MED Magazine 34. . Martin, Pamela. 1990. The phrasal verb: Diachronic development in British and American English. Columbia University New York: PhD thesis. (Ann Arbor: University Microfilms International.) Masini, Francesca. 2006. Diacronia dei verbi sintagmatici in Italiano. Archivio Glottologico Italiano 91. 67–105. Matsumoto, Meiko. 2008. From simple verbs to periphrastic expressions: The history of composite predicates, phrasal verbs, and related constructions in English. Bern: Lang. Mattaire, Michael. 1712. The English Grammar. (Facs. repr. Menston: Scolar Press, 1967.) Matthews, Peter H. 2007. Syntactic relations: A critical survey. Cambridge: Cambridge University Press. McArthur, Tom (ed.). 1992. The Oxford companion to the English language. Oxford: Oxford University Press. McArthur, Tom & Beryl Atkins. 1992. Dictionary of English Phrasal Verbs and their Idioms. 14th edn. London: Collins.

280

References

McEnery, Tony & Andrew Wilson. 2001. Corpus linguistics. 2nd edn. Edinburgh: Edinburgh University Press. McIntosh, Carey. 1986. Common and courtly language: The stylistics of social class in 18th-century English literature. Philadelphia: University of Philadelphia Press. McLelland, Nicola. 2008. Approaches to the semantics and syntax of the adverb in German foreign language grammars. Beiträge zur Geschichte der Spachwissenschaft 18 (Themenheft: Das Adverb in der Grammatikographie, Teil II). 37–58. MED = Middle English Dictionary. 1952–2001. 13 vols. Ed. Hans Kurath & Sherman M. Kuhn. Ann Arbor: University of Michigan Press. (Electronic MED: .) Meier-Brügger, Michael. 2003. Indo-European linguistics. With contributions by Matthias Fritz and Manfred Mayrhofer. Translated by Charles Gertmenian. Berlin: de Gruyter. Meyer, George A. 1975. The Two-word Verb: A Dictionary of the VerbPreposition Phrases in American English. The Hague: Mouton. Michael, Ian. 1970. English grammatical categories and the tradition to 1800. Cambridge: Cambridge University Press. Miège, Guy. 1685. Nouvelle méthode pour apprendre l’anglois. London: Thomas Bassett. (Facs. repr. Menston: Scolar Press, 1970.) Miller, Thomas (ed.). 1890–1898. The Old English version of Bede’s Ecclesiastical History of the English People. With a translation and introduction. 4 vols. EETS 95, 96, 110, 111. Minkova, Donka. 2008. Prefixation and stress in Old English: In memoriam Richard Hogg (1944–2007). Word Structure 1. 21–52. Mitchell, Bruce. 1964. Syntax and word-order in The Peterborough Chronicle 1122–1154. Neuphilologische Mitteilungen 65. 113–144. Mitchell, Bruce. 1978. Prepositions, adverbs, prepositional adverbs, postpositions, separable prefixes, or inseparable prefixes, in Old English? Neuphilologische Mitteilungen 79. 240–257. Mitchell, Bruce. 1985. Old English syntax. 2 vols. Oxford: Oxford University Press. Mitchell, Bruce & Fred C. Robinson. 2001. A guide to Old English. 6th edn. Oxford: Blackwell. Mitchell, T.F. 1958. Syntagmatic relations in linguistic analysis. Transactions of the Philological Society. 101–118. Mittins, W.H., Mary Salu, Mary Edminson & Sheila Coyne. 1970. Attitudes to English usage: An enquiry by the University of Newcastle upon Tyne Institute of Education English Research Group. London: Oxford University Press. Monk, Samuel Holt, A.E. Wallace Maurer & Vinton A. Dearing [et al.] (eds.). 1971. The Works of John Dryden, vol. 17: Prose 1668–1691: An Essay of Dramatick Poesie and shorter works. Berkeley: University of California Press. Morris, Richard (ed.). 1874–1880 [1967]. The Blickling Homilies. 3 vols. [repr. 1967 in one vol.]. EETS 58, 63, 73.

References

281

Morton, James. 1853. The Ancren Riwle: A treatise on the rules and duties of monastic life. London: Camden Society. Moser, Hugo. 1969. Umgangssprache: Überlegungen zu ihren Formen und zu ihrer Stellung im Sprachganzen. Zeitschrift für Mundartforschung 27. 215–232. Müller, Stefan. 2002. Complex predicates: Verbal complexes, resultative constructions, and particle verbs in German. Stanford: CSLI Publications. Müller, Stefan. 2008. Head-Driven Phrase Structure Grammar: Eine Einführung. 2nd edn. Tübingen: Stauffenburg. Mugglestone, Lynda. 2005. Lost for words: The hidden history of the Oxford English Dictionary. New Haven: Yale University Press. Mugglestone, Lynda. 2006. English in the nineteenth Century. In The Oxford history of the English language, ed. Lynda Mugglestone. Oxford: Oxford University Press. 274–304. Mukherjee, Joybrato. 2009. Anglistische Korpuslinguistik. Berlin: Schmidt. Müller-Schotte, Hans. 1955. Pleonastische Verberweiterungen im Englischen. Die Neueren Sprachen N.F. 4. 363–367. Murray, Lindley. 1815. English Grammar: Adapted to the different classes of learners. With an appendix, containing rules and observations for assisting the more advanced students to write with perspicuity and accuracy. 26th edn. [1st edn. 1795]. York: Wilson and Sons. Mustanoja, Tauno F. 1960. A Middle English syntax, part 1: Parts of speech. Helsinki: Société Néophilologique. Napier, Arthur S. (ed.). 1894. Invention of the cross. In History of the holy roodtree: A twelfth century version of the cross legend. EETS 103. Nash, David. 1982. Warlpiri preverbs and verb roots. In Papers in Warlpiri grammar in memory of Lothar Jagst, ed. Stephen Swartz (work-papers of SIL-AAB, series A, vol. 6). Berrimah: Summer Institute of Linguistics. 165–216. Nash, David. 1986. Topics in Warlpiri grammar. New York: Garland. Nevalainen, Terttu. 1997. Recycling inversion: The case of initial adverbs and negators in Early Modern English. Studia Anglica Posnaniensia 31 (A festschrift for Roger Lass on his sixtieth birthday). 203–214. Nevalainen, Terttu. 1999. Early Modern English lexis and semantics. In The Cambridge History of the English Language, vol. 3: 1476–1776, ed. Roger Lass. Cambridge: Cambridge University Press. 332–458. Nevalainen, Terttu & Helena Raumolin-Brunberg. 2003. Historical sociolinguistics: Language change in Tudor and Stuart England. London: Longman. Nevalainen, Terttu & Helena Raumolin-Brunberg (eds.). 1996. Sociolinguistics and language history: Studies based on the Corpus of Early English Correspondence. Amsterdam: Rodopi. Nielsen, Hans Frede. 1998. The continental backgrounds of English and its insular development until 1154. Odense: Odense University Press. Nielsen, Hans Frede. 2005. From dialect to standard: English in England 1154– 1776. Odense: Odense University Press.

282

References

Novak, Maximillian E. & George R. Guffey (eds.). 1984. The Works of John Dryden, vol. 13: Plays: All for Love, Oedipus, Troilus and Cressida. Berkeley: University of California Press. Nurmi, Arja (ed.). 1998. Manual for the Corpus of Early English Correspondence Sampler CEECS. New ICAME Corpus Collection CD-ROM [1999]. Bergen: The HIT-Centre. OED = The Oxford English Dictionary. NED (= OED1) = A New English Dictionary on Historical Principles. 10 vols. Ed. James A.H. Murray, Henry Bradley, William A. Craigie & Charles T. Onions. Oxford: Clarendon Press, 1884–1928. Reedited [under the title The Oxford English Dictionary, being a corrected reissue, with supplement], with introduction, supplement and bibliography by William A. Craigie & Charles T. Onions. 13 vols. Oxford: Clarendon Press, 1933. Supplement ed. by Robert W. Burchfield. 4 vols. Oxford: Clarendon Press, 1972–86. OED2 = The Oxford English Dictionary. 2nd edn. 20 vols., reset with corrections, revisions and additional vocabulary, prepared by John A. Simpson & Edmund S.C. Weiner. Oxford: Clarendon Press, 1989 [CD-ROM Version 1.0, 1992; CD-ROM Version 2.0, 1999. Oxford: Oxford University Press]. OED3 = 3rd edn. Ed. John A. Simpson. OED Online, . Oxford: Oxford University Press, 2000– . Oesterreicher, Wulf. 2001. Historizität – Sprachvariation, Sprachverschiedenheit, Sprachwandel. In Language Typology and Language Universals: An International Handbook, vol. 2, ed. Martin Haspelmath, Ekkehard König, Wulf Oesterreicher & Wolfgang Raible. Berlin: de Gruyter. 1554–1595. Ogura, Michiko. 2002. Verbs of motion in medieval English. Cambridge: Brewer. Oldireva Gustafsson, Larisa. 2008. Phonaesthetic assessment of words in 18thcentury prescriptions and later. In Perspectives on prescriptivism, ed. Joan Beal, Carmen Nocera & Massimo Sturiale. Bern: Lang. 83–112. Olsen, Susan. 1997. Über den lexikalischen Status englischer Partikelverben. In Lexikalische Kategorien und Merkmale, ed. Elisabeth Löbel & Gisa Rauh. Tübingen: Niemeyer. 45–71. Olsen, Susan. 2000. Composition. In Morphologie: Ein internationales Handbuch zur Flexion und Wortbildung (Morphology: An International Handbook on Inflection and Word-formation), vol. 1, ed. Geert Booij, Christian Lehmann & Joachim Mugdan, in collaboration with Wolfgang Kesselheim & Stavros Skopeteas. Berlin: de Gruyter. 897–916. Osselton, N.E. 1958. Branded words in English dictionaries before Johnson. Groningen: Wolters. Osselton, N.E. 1986. Dr Johnson and the English phrasal verb. In Lexicography: An emerging international profession, ed. Robert Ilson. Manchester: Manchester University Press. 7–16. Osselton, N.E. 1995 [1990]. The character of the earliest English dictionaries. Chapter 1 in his Chosen words: Past and present problems for dictionary makers. Exeter: University of Exeter Press [originally published in 1990 as ‘English lexicography from the beginning up to and including Johnson’ in Wörterbücher/Dictionaries/Dictionnaires: Ein internationales Handbuch zur

References

283

Lexikographie, vol. 2, ed. Franz Josef Hausmann, Oskar Reichmann & Herbert Ernst Wiegand, Berlin: de Gruyter, 1943–1953]. Palmatier, Robert Allen. 1969. A descriptive syntax of the Ormulum. The Hague: Mouton. Partridge, Eric. 1954. The concise usage and abusage: A modern guide to good English. London: Hamilton. Paul, Hermann. 1995 [1920]. Prinzipien der Sprachgeschichte. 5th edn. (Repr. 1995.) Tübingen: Niemeyer. Pelli, Mario G. 1976. Verb-particle constructions in American English: A study based on American plays from the end of the 18th century to the present. Bern: Francke. Peters, Hans. 1993. Die englischen Gradadverbien der Kategorie booster. Tübingen: Narr. Peters, Hans. 2006. The Old English verbal suffix –ettan. In Language and text: Current perspectives on English and Germanic historical linguistics and philology, ed. Andrew James Johnston, Ferdinand von Mengden & Stefan Thim. Heidelberg: Winter. 241–254. Petyt, K.A. 1980. The study of dialect: An introduction to dialectology. London: Deutsch. Pilch, Herbert. 1955. Der Untergang des Präverbs ge- im Englischen. Anglia 73. 37–64. Pinault, Georges-Jean. 1995. Le problème du préverbe en Indo-Européen. In Les préverbes dans les langues d’Europe: Introduction à l’étude de la préverbation, ed. André Rousseau. Lille: Presses Universitaires du Septentrion. 35–59. Pintzuk, Susan. 1991. Phrase structures in competition: Variation and change in Old English word order. Philadelphia, PA: University of Pennsylvania PhD thesis. Pintzuk, Susan & Ann Taylor. 2006. The loss of OV order in the history of English. In The handbook of the history of English, ed. Ans van Kemenade & Bettelou Los. 249–278. Plag, Ingo. 1999. Morphological productivity: Structural constraints in English derivation. Berlin: Mouton de Gruyter. Plag, Ingo. 2003. Word-formation in English. Cambridge: Cambridge University Press. Plag, Ingo, Christiane Dalton-Puffer & Harald Baayen. 1999. Morphological productivity across speech and writing. English Language and Linguistics 3. 209–228. Plummer, Charles (ed.). 1892–99. Two of the Saxon Chronicles parallel. Oxford: Clarendon Press. Pope, John C. (ed.). 1967–68. Homilies of Ælfric: A supplementary collection. 2 vols. EETS 259, 260. Pottelberge, Jeroen van. 2001. Verbonominale Konstruktionen, Funktionsverbgefüge: Vom Sinn und Unsinn eines Untersuchungsgegenstandes. Heidelberg: Winter.

284

References

Poussa, Patricia. 1982. The evolution of early standard English: The creolization hypothesis. Studia Anglica Posnaniensia 14. 69–85. Poutsma, Hendrik. 1926. A Grammar of late Modern English. Groningen: Noordhoff. Quirk, Randolph, Sidney Greenbaum, Geoffrey Leech & Jan Svartvik. 1985. A Comprehensive Grammar of the English Language. London: Longman. Reichl, Karl. 1995. Review of Laurel J. Brinton, The development of English aspectual systems: Aspectualizers and post-verbal particles (Cambridge: Cambridge University Press, 1988). Anglia 113. 80–84. Reszkiewicz, Alfred. 1962. Main sentence elements in The Book of Margery Kempe: A study in major syntax. Wrocław: Zakład Narodowy Imienia Ossolińskich. Reszkiewicz, Alfred. 1966. Ordering of elements in late Old English prose in terms of their size and structural complexity. Wrocław: Zakład Narodowy Imienia Ossolińskich. Rice, Allan Lake. 1932. Gothic prepositional compounds in relation to their Greek originals. (University of Pennsylvania dissertation.) Language Dissertations published by the Linguistic Society of America (Supplement to Language) 11. Philadelphia, PA: University of Pennsylvania. Rickard, T.A., W.H. Shockley & A.E. Pratt. 1909–1910. Standardization of English in technical literature. Transactions of the Institution of Mining and Metallurgy for 1909–10. [Quoted in Kennedy 1920; n.pag.] Rissanen, Matti. 1992. The diachronic corpus as a window to the history of English. In Directions in corpus linguistics, ed. Jan Svartvik. Berlin: de Gruyter. 185–205. Rissanen, Matti. 1994. The Helsinki Corpus of English Texts. In Corpora across the centuries, ed. Merja Kytö, Matti Rissanen & Susan Wright. Amsterdam: Rodopi. 73–79. Rissanen, Matti, Merja Kytö & Minna Palander-Collin (eds.). 1993. Early English in the computer age: Explorations through the Helsinki Corpus. Berlin: de Gruyter. Roberts, Ian. 2007. Diachronic syntax. Oxford: Oxford University Press. Roberts, Murat H. 1936. The antiquity of the Germanic verb-adverb locution. Journal of English and Germanic Philology 25. 466–481. Roelcke, Thorsten. 1998. Typologische Unterschiede in den Varietäten des Deutschen. In Sprachgeschichte: Ein Handbuch zur Geschichte der deutschen Sprache und ihrer Erforschung, vol. 1, ed. Werner Besch, Anne Betten, Oskar Reichmann & Stefan Sonderegger, 2nd edn. Berlin: de Gruyter. 1000–1013. Rohdenburg, Günter. 2009. Grammatical divergence between British and American English in the nineteenth and early twentieth centuries. In Current issues in Late Modern English (Papers presented at the 3rd International Conference on Late Modern English, held at the University of Leiden in 2007), ed. Ingrid Tieken-Boon van Ostade & Wim van der Wurff. Bern: Lang. 301–329. Ronneberger-Sibold, Elke. 2004. [Systeme morphologischer Struktur: Sprachskizzen:] Deutsch (Indogermanisch: Germanisch). In Morphologie: Ein

References

285

internationales Handbuch zur Flexion und Wortbildung (Morphology: An International Handbook on Inflection and Word-formation), vol. 2, ed. Geert Booij, Christian Lehmann & Joachim Mugdan, in collaboration with Wolfgang Kesselheim & Stavros Skopeteas. Berlin: de Gruyter. 1267–1285. Rosier, James L. (ed.). 1962. The Vitellius Psalter. Ed. from British Museum MS Cotton Vitellius E.xviii. Ithaca, NY: Cornell University Press. Rousseau, André (ed.). 1995. Les préverbes dans les langues d’Europe: Introduction à l’étude de la préverbation. Lille: Presses Universitaires du Septentrion. Samuels Michael L. 1949. The ge- prefix in the OE gloss to the Lindisfarne Gospels. Transactions of the Philological Society 47. 62–116. Samuels Michael L. 1972. Linguistic evolution: With special reference to English. Cambridge: Cambridge University Press. Sanchez, Christina. 2008. Consociation and dissociation: An empirical study of word-family integration in English and German. Tübingen: Narr. Sasse, Hans-Jürgen. 1991. Aspect and aktionsart: A reconciliation. Perspectives on aspect and aktionsart (Belgian Journal of Linguistics 6), ed. Carl Vetters & Willy Vandeweghe. 31–45. Sasse, Hans-Jürgen. 2006. Aspect and aktionsart. In Encyclopedia of language and linguistics, 2nd edn., editor-in-chief Keith Brown. Oxford: Elsevier. 535–538. Sauer, Hans. 1985. Die Darstellung der Komposita in altenglischen Wörterbüchern. In Problems of Old English lexicography: Studies in memory of Angus Cameron, ed. Alfred Bammesberger. Regensburg: Pustet. 267–315. Schäfer, Jürgen. 1980. Documentation in the O.E.D.: Shakespeare and Nashe as test cases. Oxford: Clarendon Press. Schäfer, Jürgen. 1989. Early Modern English lexicography. 2 vols. Oxford: Clarendon Press. Schaefer, Ursula. 1996. The late Middle English Paston Letters: A grammatical case in point for reconsidering philological methodologies. In Proceedings of the Anglistentag 1995, Greifswald, ed. Jürgen Klein & Dirk Vanderbeke. Tübingen: Niemeyer. 313–323. Schaefer, Ursula. 2006. Textualizing the vernacular in late medieval England: Suggestions for some heuristic reconsiderations. In Language and text: Current perspectives on English and Germanic historical linguistics and philology, ed. Andrew James Johnston, Ferdinand von Mengden & Stefan Thim. Heidelberg: Winter. 269–290. Scheible, Silke. 2005. ‘Upgrading’, ‘downsizing’ and co.: Revitalizing a moribund word-fomation pattern in present-day English? Arbeiten aus Anglistik und Amerikanistik 30. 177–200. Scheler, Manfred. 1977. Der englische Wortschatz. Berlin: Schmidt. Schmid, Hans-Jörg. 2005. Englische Morphologie und Wortbildung: Eine Einführung. Berlin: Schmidt. Schmidt, Günther Dietrich. 1987. Das Affixoid: Zur Notwendigkeit und Brauchbarkeit eines beliebten Zwischenbegriffs der Wortbildung. In Deutsche Lehnwortbildung, ed. Gabriele Hoppe, Alan Kirkness, Elisabeth Link, Isolde Nortmeyer, Wolfgang Rettig & Günther Schmidt. Tübingen: Narr. 53–101.

286

References

Schmidt, Karl Horst. 1969. Zur Tmesis in den Kartvelsprachen und ihren typologischen Parallelen in indogermanischen Sprachen. In To George Akhvlediani, ed. Shota Dzidziguri et al. Tbilisi: University. 96–105. Schmidt, Karl Horst. 1998. Versuch einer geschichtlichen Sprachtypologie des Deutschen. In Sprachgeschichte: Ein Handbuch zur Geschichte der deutschen Sprache und ihrer Erforschung, vol. 1, ed. Werner Besch, Anne Betten, Oskar Reichmann & Stefan Sonderegger, 2nd edn. Berlin: de Gruyter. 993–1000. Schneider, Edgar W. 2004. How to trace structural nativization: Particle verbs in World Englishes. World Englishes 23. 227–249. Schrodt, Richard. 2004. Althochdeutsche Grammatik II: Syntax. Tübingen: Niemeyer. Schröder, Anne. 2008. Investigating the morphological productivity of verbal prefixation in the history of English. Arbeiten aus Anglistik und Amerikanistik 33. 47–69. Schultze-Berndt, Eva. 2003. Preverbs as an open word class in northern Australian languages: Synchronic and diachronic correlates. In Yearbook of Morphology 2003, ed. Geert Booij & Jaap van Marle. Dordrecht: Kluwer. 145–177. Seoane, Elena. 2006. Information structure and word order change: The passive as an information-rearranging strategy in the history of English. In The handbook of the history of English, ed. Ans van Kemenade & Bettelou Los. Oxford: Blackwell. 360–391. Serjeantson, Mary S. 1935. A history of foreign words in English. London: Kegan Paul. Shluktenko, J.A. 1955. Über die sogenannten ‘zusammengesetzten Verben’ vom Typ stand up in der englischen Sprache der Gegenwart. Sowjetwissenschaft: Gesellschaftswissenschaftliche Abteilung, ed. by the Zentralvorstand der Gesellschaft für deutsch-sowjetische Freundschaft. Vol. for 1955 (2). 223–235. [Ю.А. Жлуктенко. 1954. О так называемых «сложных глаголах» типа stand up в совремнном англиӥском яэыке. Вопросы яэыкоэнаиия 5. 105– 113. German translation by G. Riegel, revised by H.H. Bielfeldt.] Sievers, Eduard (ed.). 1892. Tatian: Lateinisch und Altdeutsch mit ausführlichem Glossar. 2nd edn. Paderborn: Schöningh. (Repr. 1966.) Siewierska, Anna. 1993. On the interplay of factors in the determination of word order. In Syntax: Ein internationales Handbuch zeitgenössischer Forschung (An International Handbook of Contemporary Research), vol. 1, ed. Joachim Jacobs, Arnim von Stechow, Wolfgang Sternefeld & Theo Vennemann. Berlin: de Gruyter. 826–846. Simon, Irène. 1963. Dryden’s revision of the ‘Essay of Dramatic Poesy’. The Review of English Studies 14. 132–141. Simpson, Jane. 1991. Warlpiri morpho-syntax: A lexicalist approach. Dordrecht: Kluwer. Sinclair, John. 1991. Corpus, concordance, collocation. Oxford: Oxford University Press. Sitta, Horst. 1998. Der Satz. In Duden Grammatik der deutschen Gegenwartssprache. 6th edn. Mannheim: Dudenverlag. 609–858.

References

287

Skeat, Walter (ed.). 1871–1887. The Holy Gospels in Anglo-Saxon, Northumbrian, and Old Mercian versions. 4 vols. Cambridge: Cambridge University Press. Skeat, Walter W. (ed.). 1881–1900 [1966]. Ælfric’s Lives of the Saints. 4 vols. (Repr. 1966 in 2 vols.) EETS 76, 82, 94, 114. Slobin, Dan I. 2006. What makes manner of motion salient? Explorations in linguistic typology, discourse, and cognition. In Space in languages: Linguistic systems and cognitive categories, ed. Maya Hickmann & Stéphane Robert. Amsterdam: Benjamins. 5–81. Smirnova, Elena & Tanja Mortelmans. 2010. Funktionale Grammatik: Konzepte und Theorien. Berlin: de Gruyter Mouton. Smith, Logan Pearsall. 1925. Words and idioms: Studies in the English language. London: Constable. Smitterberg, Erik. 2008. The progressive and phrasal verbs: Evidence of colloquialization in nineteenth-century English? In The dynamics of linguistic variation: Corpus evidence on English past and present, ed. Terttu Nevalainen, Irma Taavitsainen, Päivi Pahta & Minna Korhonen. Amsterdam: Benjamins. 269–289. Söderlind, Johannes. 1964. The attitude to language expressed by or ascertainable from English writers of the 16th and 17th centuries. Studia Neophilologica 36. 111–126. Spasov, Dimiter. 1966. English phrasal verbs. Sofia: Naouka i Izkoustvo. Sprockel, Cornelis. 1973. The language of the Parker Chronicle, vol. 2. The Hague: Nijhof. Sroka, Kazimierz A. 1972. The syntax of English phrasal verbs. The Hague: Mouton. Standop, Ewald. 1999. Syntaxtheorie und englische Syntax – quantum vis. Anglia 117. 236–258. Stanley, Eric G. 1987. Old English in ‘The Oxford English Dictionary’. In Studies in lexicography, ed. Robert Burchfield. 19–35. Starke, Frank. 1977. Die Funktionen der dimensionalen Kasus und Adverbien im Althethitischen. Wiesbaden: Harrassowitz. Stefanowitsch, Anatol & Kerstin Fischer (eds.). 2008. Konstruktionsgrammatik II: Von der Konstruktion zur Grammatik. Tübingen: Stauffenburg. Stein, Gabriele. 1984. Word-formation in Dr Johnson’s Dictionary of the English Language. Dictionaries: Journal of the Dictionary Society of North America 6. 66–112. Stockwell, Robert P. 1977. Motivations for exbraciation in English. In Mechanisms of syntactic change, ed. Charles N. Li. (2nd printing 1980.) Austin: University of Texas Press. 291–314. Stockwell, Robert P. 1984. On the history of the verb-second rule in English. In Historical syntax, ed. Jacek Fisiak. The Hague: Mouton. 575–592. Stockwell, Robert P. 1990. Review of Ans van Kemenade, Syntactic case and morphological case in the history of English (Dordrecht: Foris, 1987). Lingua 81. 90–100.

288

References

Stockwell, Robert P. & Donka Minkova. 1991. Subordination and word order change in the history of English. In Historical English syntax, ed. Dieter Kastovsky. Berlin: Mouton de Gruyter. 337–348. Strang, Barbara M.H. 1962. Modern English structure. London: Arnold. Strang, Barbara M.H. 1970. A history of English. London: Methuen. Streitberg, Wilhelm. 1891. Perfective und imperfective actionsart im germanischen. Beiträge zur Geschichte der deutschen Sprache und Literatur 15. 70–177. Streitberg, Wilhelm. 1920. Gotisches Elementarbuch. 5th and 6th edn. Heidelberg: Winter. Streitberg, Wilhelm. 2000 [1919]. Die Gotische Bibel. 7th edn. [based on the 2nd edn. 1919] with a supplement by Piergiuseppe Scardigli. 2 vols. (vol. 1: Der gotische Text und seine griechische Vorlage: Mit Einleitung, Lesarten und Quellennachweisen sowie den kleineren Denkmälern als Anhang; vol. 2: Gotisch–Griechisch–Deutsches Wörterbuch). Heidelberg: Winter. Sundby, Bertil. 1995. English word-formation as described by English grammarians 1600–1800. Oslo: Novus. Sundby, Bertil, Anne Kari Bjørge & Kari E. Haugland. 1991. A dictionary of English normative grammar: 1700–1800. Amsterdam: Benjamins. Svenonius, Peter Arne. 1994. Dependent nexus: Subordinate predication in English and the Scandinavian languages. University of California, Santa Cruz: Ph.D. dissertation. Svenonius, Peter Arne. 1996. The verb-particle-alternation in the Scandinavian languages. Ms., University of Tromsø. 36pp. Sweet, Henry (ed.). 1871. King Alfred’s West Saxon version of Gregory’s Pastoral Care. 2 vols. EETS 45, 50. Talmy, Leonard. 1985. Lexicalization patterns: Semantic structure in lexical forms. In Language typology and syntactic description, vol. 3: Grammatical categories and the lexicon, ed. Timothy Shopen. Cambridge: Cambridge University Press. 57–149. Talmy, Leonard. 1991. Path to realization: A typology of event conflation. In Proceedings of the Seventeenth Annual Meeting of the Berkeley Linguistics Society, ed. Laurel A. Sutton, Christopher Johnson & Ruth Shields. Berkeley: Berkeley Linguistics Society. 480–519. Talmy, Leonard. 2000. Toward a cognitive semantics, vol. 2: Typology and process in concept structuring. Cambridge, MA: MIT Press. Tanabe, Harumi. 1999. Composite predicates and phrasal verbs in The Paston Letters. In Collocational and idiomatic aspects of composite predicates in the history of English, ed. Laurel J. Brinton & Minoji Akimoto. Amsterdam: Benjamins. 97–132. Taylor, Simon. 1983. The Anglo-Saxon Chronicle. A collaborative edition, general eds. David Dumville & Simon Keynes, vol. 4: MS B (A semi-diplomatic edition with introduction and indices). Cambridge: Brewer. Thim, Stefan. 2006a. Phrasal verbs in everyday English, 1500–1700. In Language and text: Current perspectives on English and Germanic historical linguistics

References

289

and philology, ed. Andrew James Johnston, Ferdinand von Mengden & Stefan Thim. Heidelberg: Winter. 291–306. Thim, Stefan. 2006b. Phrasal verbs in late Middle and Early Modern English: Combinations with back, down, forth, out, and up. In Syntax, style and grammatical norms: English from 1500–2000, ed. Christiane Dalton-Puffer, Nikolaus Ritt, Herbert Schendl & Dieter Kastovsky. Bern: Lang, 213–228. Thim, Stefan. 2006c. When did the phrasal verbs become colloquial? Paper read at the 14th International Conference on English Historical Linguistics, Bergamo, 21–25 August 2006. Thim, Stefan. 2007. Lexicalization, usage and attitude: The colloquialization of the verb-particle construction in Late Modern English. Paper read at the 3rd Late Modern English Conference, Leiden, 30 August–1 September 2007. Thim, Stefan. 2008a. The rise of the phrasal verb in English: A case of Scandinavian influence? In Anglistentag 2007 Münster: Proceedings of the Conference of the German Association of University Teachers of English 29, ed. Klaus Stierstorfer. Trier: Wissenschaftlicher Verlag Trier. 291–304. Thim, Stefan. 2008b. ‘There is another kind of composition more frequent in our language than perhaps in any other’: The ‘Englishness’ of the ‘phrasal verbs’. Paper read at the 15th International Conference on English Historical Linguistics, Munich, 24–30 August 2008. Thim, Stefan. 2011a. On the phrasal verbs in some Paston Letters. In More than words: English lexicography and lexicology past and present (Essays presented to Hans Sauer on the occasion of his 65th birthday: part I), ed. Renate Bauer & Ulrike Krischke. Frankfurt am Main: Lang. 355–385. Thim, Stefan. 2011b. Historical dictionaries of English. Lexicographica 27. 63–99. Thorpe, Benjamin (ed.). 1840. Ancient laws and institutes of England. 2 vols. London: Commission on the Public Records of the Kingdom. Thorpe, Benjamin (ed.). 1861. The Anglo-Saxon Chronicle according to several authorities. 2 vols. Rolls Series. London: Longman, Green, Longman & Roberts. Thráinsson, Höskuldur. 1994. Icelandic. In The Germanic languages, ed. Ekkehard König & Johan van der Auwera. London: Routledge. 142–189. Thráinsson, Höskuldur, Hjalmar P. Petersen, Jógvan í Lon Jacobsen & Zakaris Svabo Hansen. 2004. Faroese: An overview and reference grammar. Tórshavn: Føroya Fróðskaparfelag. Tieken-Boon van Ostade, Ingrid. 2006a. Eighteenth-century prescriptivism and the norm of correctness. In The handbook of the history of English, ed. Ans van Kemenade & Bettelou Los. Oxford: Blackwell. 539–557. Tieken-Boon van Ostade, Ingrid. 2006b. English at the onset of the normative tradition. In The Oxford history of the English language, ed. Lynda Mugglestone. Oxford: Oxford University Press. 240–273. Tiersma, Pieter M. 1985. Frisian Reference Grammar. Dordrecht: Foris. Tiisala, Seija. 2008. Adverbs and particles in Swedish grammars from the 17th and 18th century. Beiträge zur Geschichte der Sprachwissenschaft 18 (Themenheft: Das Adverb in der Grammatikographie, Teil II). 101–134.

290

References

Traugott, Elizabeth Closs. 1978. On the expression of spatio-temporal relations in language. In Universals of human language, vol. 3: Word structure, ed. Joseph H. Greenberg. Stanford: Stanford University Press. 369–400. Traugott, Elizabeth Closs. 1991. Review of Laurel J. Brinton, The development of English aspectual systems: Aspectualizers and post-verbal particles (Cambridge: Cambridge University Press, 1988). Studies in Language 15. 221–226. Traugott, Elizabeth Closs. 1992. Syntax. In The Cambridge History of the English Language, vol. 1: The Beginnings to 1066, ed. Richard Hogg. Cambridge: Cambridge University Press. 168–289. Traugott, Elizabeth Closs. 1999. A historical overview of complex predicate types. In Collocational and idiomatic aspects of composite predicates in the history of English, ed. Laurel J. Brinton & Minoji Akimoto. Amsterdam: Benjamins. 239–260. Traugott, Elizabeth Closs. 2003a. From etymology to historical pragmatics. In Studies in the history of the English language: A millennial perspective, ed. Donka Minkova & Robert Stockwell. Berlin: Mouton de Gruyter. 19–49. Traugott, Elizabeth Closs. 2003b. Constructions in grammaticalization. In The Handbook of historical linguistics, ed. Brian D. Joseph & Richard D. Janda. Oxford: Blackwell. 625–647. Traugott, Elizabeth Closs & Richard B. Dasher. 2002. Regularity in semantic change. Cambridge: Cambridge University Press. Traugott, Elizabeth Closs & Ekkehard König. 1991. The semantics-pragmatics of grammaticalization revisited. In Approaches to grammaticalization, vol. 1: Focus on theoretical and methodological issues, ed. Elizabeth Closs Traugott & Bernd Heine. Amsterdam: Benjamins. 189–218. Trips, Carola. 2002. From OV to VO in early Middle English. Amsterdam: Benjamins. Tristram, Hildegard L.C. 2004. Diglossia in Anglo-Saxon England, or what was spoken Old English like? Studia Anglica Posnaniensia 40. 87–110. Trousdale, Graeme & Nikolas Gisborne (eds.). 2008. Constructional approaches to English grammar. Berlin: Mouton de Gruyter. Tucker, Susie. 1961. English examined. Cambridge: Cambridge University Press. Vennemann, Theo. 1974a. Language type and word order. Acta Universitatis Carolinae, Philologica 5, Linguistica Generalica I, 219–229. Vennemann, Theo. 1974b. Topics, subjects, and word order: From SXV to SVX via TVX. In Historical Linguistics I: Syntax, morphology, internal and comparative reconstruction (Proceedings of the First International Conference on Historical Linguistics, Edinburgh 2nd–7th September 1973), ed. John M. Anderson & Charles Jones. Amsterdam: North-Holland. 339–376. Vennemann, Theo. 1984. Verb-second, verb-late, and the brace construction. In Historical syntax, ed. Jacek Fisiak. The Hague: Mouton. 627–636. Veselinović, Elvira. 2006. How to put up with cur suas le rud and the bidirectionality of contact. In The Celtic Englishes IV, ed. Hildegard L.C. Tristram. Potsdam: Potsdam University Press. 173–190.

References

291

Vestergaard, Torben. 1977. Prepositional phrases and phrasal verbs: A study in grammatical function. The Hague: Mouton. Vetters, Carl & Willy Vandeweghe (eds.). 1991. Perspectives on aspect and aktionsart (Belgian Journal of Linguistics 6) Amsterdam: Benjamins. Vincent, Nigel. 1999. The evolution of C-structure: Prepositions and PPs from Indo-European to Romance. Linguistics 37. 1111–1153. Voitl, Herbert. 1969. Probleme der englischen Idiomatik. Germanisch-Romanische Monatsschrift N.F. 19. 194–212. Von Schon, Catherine Virginia. 1977. The origin of phrasal verbs in English. New York State University: unpublished dissertation. (Ann Arbor: University Microfilms International.) Vorlat, Emma. 1975. The development of English grammatical theory 1568–1737: With special reference to the theory of parts of speech. Leuven: Leuven University Press. Vriend, Hubert Jan de (ed.). 1986. The Old English Herbarium and Medicina de Quadrupedibus. EETS 286. Waibel, Birgit. 2007. Phrasal verbs in learner English: A corpus-based study of German and Italian students. Freiburg University: doctoral dissertation. . Wälchli, Bernhard. 2001. Lexical evidence for the parallel development of the Latvian and Livonian verb particles. In The Circum-Baltic languages: typology and contact, vol. 2: Grammar and typology, ed. Östen Dahl & Maria Koptjevskaja-Tamm. Amsterdam: Benjamins. 413–441. Walker, Obadiah. 1659. Some instructions concerning the art of oratory: Collected for the use of a friend of a young student. London: J.G. for R. Royston. Wallis, John. 1653. Grammatica Linguæ Anglicanæ. (Repr. Menston: Scolar Press, 1969.) Wardhaugh, Ronald. 2006. An introduction to sociolinguistics. 5th edn. Oxford: Blackwell. Watkins, Calvert. 1963. Preliminaries to a historical and comparative analysis of the syntax of the Old Irish Verb. Celtica 6. 1–49. Watkins, Calvert. 1964. Preliminaries to the reconstruction of Indo-European sentence structure. In Proceedings of the Ninth International Congress of Linguists, Cambridge, Mass., August 27–31, 1962, ed. Horace G. Lunt. The Hague: Mouton. 1035–1042 [with a discussion, 1043–1045]. Webster, Noah. 1828. An American Dictionary of the English Language. (Facs. ed. on CD-ROM, San Francisco: Foundation of American Christian Education, 1995.) Webster, Noah. 1909. Webster’s New International Dictionary of the English Language. Ed. William Torrey Harris (editor-in-chief) & Frederic Sturges Allen (general editor). London: Merriam. Wedel, Alfred R. 1997. Verbal prefixation and the ‘complexive’ aspect in Germanic. Neuphilologische Mitteilungen 98. 321–332. Wedel, Alfred R. 2001. Alliteration and the prefix gein Cynewulf’s ‘Elene’. Journal of English and Germanic Philology 100. 200–210.

292

References

Wetzel, Claus-Dieter. 1991. Bemerkungen zur ersten Lieferung des ‘Dictionary of Old English’. Indogermanische Forschungen 96. 218–237. Wild, Kate. 2008. Phrasal verbs in Late Modern English. Paper read at the 15th International Conference on English Historical Linguistics, Munich, 24–30 August 2008. Wilhelm, Christopher. 2002. Old Hittite postpositions: A fuzzy problem. In New Insights in Germanic Linguistics III, ed. Irmengard Rauch & Gerald F. Carr. New York: Lang. 273–288. Williams, Edwin S. 1981. On the notions ‘lexically related’ and ‘head of a word’. Linguistic Inquiry 12. 245–274. Wischer, Ilse. 2004. Old English prefixed verbs and the question of aspect and aktionsart. In Anglistentag 2003 München: Proceedings of the Conference of the German Association of University Teachers of English 25, ed. Christoph Bode, Sebastian Domsch & Hans Sauer. Trier: Wissenschaftlicher Verlag Trier. 71–84. Wischer, Ilse & Mechthild Habermann. 2004. Der Gebrauch von Präfixverben zum Ausdruck von Aspekt/Aktionsart im Altenglischen und Althochdeutschen. Zeitschrift für Germanistische Linguistik 32. 262–285. Wiseman, Charles. 1764. A Complete English Grammar on a new plan: For the use of foreigners. London: Nicol. Wood, Frederick T. 1956. Verb-adverb combinations: The position of the adverb. English Language Teaching 10. 18–27. Wright, William Aldis (ed.). 1909. The Authorised Version of the English Bible 1611, vol. 5: The New Testament. Cambridge: Cambridge University Press. Wyld, Henry Cecil. 1936. A history of modern colloquial English. 3rd edn. Oxford: Blackwell. Yáñez-Bouza, Nuria. 2006. Prescriptivism and preposition stranding in eighteenthcentury prose. Historical Sociolinguistics and Sociohistorical Linguistics 6. . Yáñez-Bouza, Nuria. 2008a. Preposition stranding in the eighteenth century: Something to talk about. In Grammars, grammarians, and grammar-writing in eighteenth-century England, ed. Ingrid Tieken-Boon van Ostade. Berlin: Mouton de Gruyter. 251–278. Yáñez-Bouza, Nuria. 2008b. To end or not to end a sentence with a preposition: An eighteenth-century debate. In Perspectives on prescriptivism, ed. Joan Beal, Carmen Nocera & Massimo Sturiale. Bern: Lang. 237–264.

Index ad hoc formation 16, 25, 33, 57, 58 Addison, Joseph 242 adposition 75, 77, 82, 85–89, 248; see also adverb, postposition, preposition, preverb, serialization adverb(ial particle) 3, 10, 11, 13–14, 21, 26, 35–36, 56, 58, 62, 65, 66–67, 69, 75–79, 81–88, 99, 106, 115, 117, 119, 124, 127, 129, 130, 133, 134, 140, 141, 149, 156, 168, 175, 219, 220, 234, 238, 248 adverbial, clause-initial 95, 98, 99, 103 affix 32, 49, 50, 59, 62, 67, 78, 147, 153, 155, 156, 161, 164, 165, 167–169, 250, 252 affixation vs compounding 35, 58, 64, 67 affixoid 64 Afrikaans 47–49 ahistorical 118–119, 250; see also anachronistic aktionsart 16–17, 19, 176, 180; see also aspect(ualizer) allostruction 23, 71–73, 248 American English see varieties of Modern English American influence see varieties of Modern English (American English) anachronistic 103, 126; see also ahistorical analytic(ity) 36, 38, 39, 66, 138, 139, 155, 158, 198, 214, 219 vs synthetic(ity) 3, 39, 118, 139, 158; see also drift Anglo-Saxon 76, 110, 120, 139, 184, 186; see also English (Old English)

areal features 81 aspect(ualizer) 12–14, 16–21, 28, 38–41, 46, 58, 63, 65, 146, 150, 165–166, 169, 170–171, 174–184, 188–189, 195, 214, 252, 253 attitudes 2, 9, 198, 216, 218–219, 221–222, 224–225, 226, 229, 233, 234, 236, 237–240, 247, 250, 251, 253; see also normative tradition attrition, phonetic 87, 160–161, 163– 164, 185, 250; see also phonotactics Bayly, Anselm 241–242 Behaghel’s Laws 24–25, 85, 95, 99 Bible (translation) 167–169, 211, 216–217 bleaching 128, 151–152, 167, 173, 181 BNC see British National Corpus boosters 175; see also intensification borrowing 6–8, 41, 81, 88, 121, 126, 156, 157, 162, 172, 184, 186– 195 of particles 187 of prefixes 130, 153–154, 156– 158, 160, 164, 187, 195 of verbs compared to other word classes 188 see also calque, Celtic, contact, Englishness, French, hard word, Latin, Latinate vocabulary in English, loans, Norse, Romance Bosworth–Toller 131–134, 136, 140, 150 brace (construction) see clausal brace Bradley, Henry 2–3

294

Index

British English see varieties of Modern English British National Corpus (BNC) xiv Brown Corpus 207 calque 80–81, 128, 132, 141, 143, 149–150; see also borrowing, loans CEEC see Corpus of Early English Correspondence Celtic 85, 128, 184 clausal brace 47, 97–102, 104, 107, 114–115 clitic 79, 83–84, 95, 102, 161, 166– 167; see also Wackernagel’s Law cognitive profiling 253 colligation 76, 129 collocation 55, 76, 141, 182, 207, 222, 230, 243 colloquiality 9, 31–32, 42–45, 120– 122, 125, 126, 187, 197, 205, 214–218, 221–223, 225, 227– 228, 230, 233, 235, 242–245, 250–252 term 243–244 colloquialization 214 colloquialization conspiracy 233, 245, 251 common core of the English lexicon 193, 243–244, 252 compositionality 5, 11–15, 18–20, 24–25, 27, 33, 39, 41, 46, 56– 59, 65, 70, 81, 122–123, 182– 183, 212, 229, 241, 253 compound 3, 32, 35–36, 58–59, 62– 64, 68–69, 73, 76, 80, 88, 119, 125, 130, 136–137, 139–141, 156, 161, 198, 220, 234 compound stress see stress CONCE see Corpus of Nineteenthcentury English conceptual overlap 192; see also doubling, pleonasm, redundancy, superfluity

construction (Construction Grammar) 2, 8, 58, 60, 63, 69– 72, 248 constructionalization 56, 253 contact 5, 48, 80–81, 103, 126, 129, 145, 184–185, 250, 252; see also borrowing convergence 5, 81 conversion 36, 57–58, 68 Cook, James 226–228 copia 240 corpora 55, 100, 105–106, 110, 132, 191, 202–213, 228, 252; see also British National Corpus, Brown Corpus, Corpus of Early English Correspondence, Corpus of Nineteenth-century English, Dictionary of Old English Corpus, Everyday English (1500–1700), Helsinki Corpus, Lampeter Corpus, LondonOslo/Bergen Corpus, Longman Spoken and Written English Corpus, Louvain Corpus of Native English Essays, PennHelsinki Parsed Corpus of Middle English, York-TorontoHelsinki Parsed Corpus of Old English Prose Corpus of Early English Correspondence (CEEC) 192– 193, 208–210 Corpus of Nineteenth-century English (CONCE) 213 correctness 230, 237 cranberry verbs 30 Danish 45, 50, 52 Dargi 79 decategorialization 87, 248 Delbrück, Berthold 83–85 derivation 30, 32–33, 35, 49–50, 58– 60, 63–67, 73, 130, 158, 160– 162, 165, 172, 176, 248, 252

Index desemantization 159, 161, 164–165, 169, 171, 179, 181, 250; see also semantic weakening dictionaries bilingual 225, 242 historical 8, 76, 117, 131–144, 183, 199, 249; see also Bosworth–Toller, Dictionary of Old English, Oxford English Dictionary, Middle English Dictionary of phrasal verbs 54–55 phrasal verbs in pre-19th century dictionaries 218–219, 221, 225, 251; see also Johnson, lexicographer, lexicography, normative tradition Dictionary of Old English (DOE) 131–132, 140–143, 150–152 Dictionary of Old English Corpus 110 directional particle see spatial particle discourse tradition 162, 252 DOE see Dictionary of Old English doubling 167–168, 176–177, 179; see also pleonasm drama 204; see also plays, Shakespeare drift 39, 93, 158, 237; see also analyticity Dryden, John 202, 222, 226, 230– 236, 239–241, 245 Dutch 46–49, 51, 53, 62, 69, 74–76, 92, 96, 105, 147, 165, 173, 182 Early Modern English see English education (educated and uneducated speakers) 120–122, 198, 214, 217, 222, 226, 229, 244 elaboration 157, 250 English 14th-century 62, 102, 138 15th-century 154–156, 192, 194, 202–204, 208–211

295

16th-century 154, 192, 199, 208– 210, 218–219 17th-century 7, 9, 154, 197, 198– 199, 203, 207, 210, 212, 215, 218–219, 226, 230, 235–236, 251 18th-century 1, 9, 117, 154, 191, 194, 198–207, 210, 215, 218– 222, 225, 226, 233, 237, 239– 242, 245, 247, 251 19th-century 7, 36, 194, 198– 199, 201, 203–205, 214–215, 217, 218, 233, 235–236, 238, 239–240, 245–246, 251, 253 as a foreign language 1, 20, 54– 55, 123, 125, 241–242 Early Modern English 7, 103, 121, 123, 138, 153–155, 160, 162, 188, 193–194, 198, 205– 209, 215–216, 218, 226, 233, 240, 253 Late Modern English 7, 153–154, 188, 214, 253 Middle English 7, 41, 62, 74–76, 88–89, 100–103, 105, 111– 115, 119, 121, 124–126, 129– 130, 131, 135–137, 140, 143– 144, 146, 150, 152–159, 161– 164, 172–173, 177, 182–183, 185, 187–188, 194–195, 197– 198, 215–216, 218, 249–253; see also Middle English Dictionary Old English 3–8, 74–76, 80, 85, 89–115, 119–120, 124–125, 127–137, 139–144, 145–157, 159–164, 168–178, 180–181, 183, 198, 216, 248–250, 253; see also Anglo-Saxon, Bosworth–Toller, Dictionary of Old English periodization 7 present-day English 2, 5, 7, 8, 75, 79, 90, 99, 101, 103, 113–115, 118, 126, 149, 153, 162, 163,

296

Index

182, 186, 191–194, 207, 212, 215, 222, 249 Standard English 188, 191, 211, 226, 244 see also varieties of Modern English Englishness 56, 123, 126, 214, 242; see also colloquiality, English as a foreign language, idiom, irregularity Estonian see Finno-Ugric etymology 6–7, 9, 40, 64, 79, 82, 118, 122, 137, 143, 148, 150– 151, 159, 162, 165, 169–170, 174–176, 177, 179, 185–188, 193, 195–196, 209, 229, 250 EvE see Everyday English (1500– 1700) Everyday English (1500–1700) 210– 211, 213, 229 exbraciation 98–99, 104, 107, 110, 115; see also clausal brace exemplar models 253 Faroese 46, 52, 99 Festus 83 figurative use 5, 14, 50, 135, 139, 180, 241 Finnish see Finno-Ugric Finno-Ugric 39, 80–81 foreigner see English as a foreign language formality 9, 42–45, 120, 126, 153, 187, 191, 197, 198, 202–203, 205, 214, 216–218, 221, 224, 227–230, 237, 243, 245, 250, 252; see also colloquiality, style French influence on English 6, 40–41, 119, 124–126, 129–130, 154, 156, 172, 184, 186–188, 192– 193, 250, 253 language 82, 242

frequency 79, 85, 90–92, 98, 102– 102, 161, of phrasal verbs 1, 44, 63, 118– 120, 193, 197–201, 203, 205– 217, 220, 229, 237, 253 Frisian 46–49, 51 fronting 15, 17, 24–25, 51, 91, 98 functionalism 2, 49, 90 Funktionsverbgefüge 177 fusion 8, 19, 78, 83–85, 105, 146, 165, 168, 176, 195 generativism 23, 60, 68, 69, 70, 71, 90, 100 Georgian 51, 78–79, 81; see also Kartvelian German 64, 75–76, 80–82, 92, 100, 115, 117, 134, 147, 158, 161– 162, 165, 173, 189, 191, 214, 217, 243, 253 Early New High German 161, 214 Middle High German 214 Modern German 74, 166, 177 Old High German 110–111, 166, 169, 170 present-day German 3–4, 46–51, 53, 59–61, 105, 165 Germanic 3–8, 40, 45–56, 61, 63, 64, 72, 73, 74, 80, 81, 84–86, 89– 94, 97, 99, 104–105, 115, 117, 119, 122–123, 125, 136, 145, 147–149, 152, 154, 155, 157, 158, 165–166, 169, 174, 176, 184–189, 192–193, 195, 227– 231, 242, 247–250, 254; see also native North Germanic 51, 89, 91–93, 98, 157, 184 Proto-Germanic 6, 86, 89, 92, 148, 151, 169, 249; see also Runic West Germanic 4, 6, 47, 50, 53, 55–56, 74, 88, 91–93, 97, 105, 146, 151, 157, 185, 187, 248

Index

297

see also Afrikaans, Danish, Dutch, English, Faroese, Frisian, German, Gothic, Icelandic, Luxembourgeois, Norse, Norwegian, Pennsylvania German, Runic, Scandinavian, Swedish, Yiddish Gothic 151, 165–171, 175, 177, 250 grammar, normative see normative tradition grammaticalization 70, 79, 87, 91– 92, 99, 160–161, 165, 173, 180–182 Greek 39, 78, 82, 84, 86, 157, 168– 171 group-verb 36–39, 195, 247; see also multi-word verb

inflectional morphology 32, 50, 60 loss 99, 185 informality see formality inseparable prefix see prefix (inseparable) integration of borrowed items 8, 157, 184, 187, 189, 217, 253 intensification 17–18, 150, 169, 177; see also boosters interchangeability of prefixes 127, 159, 168, 173–174 intransitive preposition see preposition inversion 11, 15, 20; see also fronting irregularity 1, 221–223, 242, 251, 252 Italian 82

hard word 118, 121, 123 Harris, James 221, 223, 237 Harrison, Thomas P. 3, 117, 138 Hawkesworth, John 226–228 Helsinki Corpus 207–211, 216, 229 Hittite 82–83, 85–86 homonymic clash 163 Hungarian 80; see also Finno-Ugric

Johnson, Samuel 1–2, 221–226, 238, 240–241, 252 Jonson, Ben 234

Icelandic 45–46, 51–52, 92, 99 ideational 188, 229, 245 idiom 3, 28, 38–39, 54–56, 216, 221, 223, 227, 236–237, 240–242, 245, 247, 251 idiomatic (semantic type of phrasal verb) 2, 11–13, 15, 19–21, 25, 46, 56, 62, 81, 149, 211, 214, 252–253 idioms containing phrasal verbs 25, 72 Indo-European 6, 50, 64, 77, 78–79, 81–87, 89, 92, 156, 248 Proto-Indo-European 77, 82–83, 86, 169 inference 180

Kartvelian 39, 78–79; see also Georgian, Svan Kennedy, Arthur Garfield 3, 117– 125, 128, 133, 134, 143, 197– 198, 201–202, 215, 249, 250 Lampeter Corpus 186, 205–211, 213, 215–217 Late Modern English see English Latin 78, 82–84, 86, 176, 219, 221, 235, 241–242, 250 borrowed from 6, 41, 130, 154, 157, 186–190, 193, 195, 229, 234, 235 grammar 3, 219, 234–235, 239 translated from 6, 110–111, 114, 128, 132, 141, 149, 169–174 see also normative tradition Latinate vocabulary in English 40– 41, 44, 121–123, 188, 191, 196, 217, 227–231, 237–238, 245, 250–251

298

Index

Latvian 81 layering 85, 87, 104, 147, 167, 175, 248 letters 192–194, 202–203, 208, 211, 213, 216–217, 229, 243; see also Paston Letters lexicographer 1, 45, 134, 219, 221– 222, 251; see also dictionaries lexicography 26, 55, 131, 133, 135, 140, 143, 149, 218, 221, 225, 253; see also dictionaries lexicalization 6, 12, 19, 25, 28, 30, 36, 38–39, 42, 55, 56, 59, 60, 70, 123, 125, 128, 133, 163, 165, 170, 176, 182–183 light verb 37 literal use 6 literary language 199–200, 226, 230, 238, 244 Livonian 81; see also Finno-Ugric loans 6, 125, 127–128, 149, 153, 157, 162, 171, 183, 185, 187, 189, 198, 217, 234; see also borrowing, calque LOB see London-Oslo/Bergen Corpus London-Oslo/Bergen Corpus (LOB) 207, 212–213, 217 LOCNESS see Louvain Corpus of Native English Essays Longman Spoken and Written English Corpus (LSWE) 211– 213 Louvain Corpus of Native English Essays (LOCNESS) 213 Lowth, Robert 1, 219–221, 234, 236, 239, 241 LSWE see Longman Spoken and Written English Corpus Luther, Martin 217 Luxembourgeois 46 Mattaire, Michael 1, 219–220, 234 metaphor 5, 6, 14, 137, 156, 158, 179–180, 182

metonymy 6, 14, 179–182 Middle English see English Middle English Dictionary (MED) 135–138, 150 middle style 237, 244, 246 Miège, Guy 219, 239 monosyllabic verbs 7, 40, 122–123, 237 monosyllables 223, 237, 251; see also preposition stranding morphonological factors 78, 167, 250 motion 5, 15, 40, 114, 179, 188, 229 motion-path 184 multi-word verb 15–16, 36–39; see also group-verb multifactorial analysis 23, 71, 96 Murray, James 243 Murray, Lindley 238 native 8, 88, 119, 120, 123, 129, 132, 145–146, 149, 153, 156–160, 162, 164, 172–173, 181, 184, 186, 187–189, 192, 194, 195, 198, 216, 227 , 229, 249–251; see also Germanic, Englishness nominalisation 24, 30–33, 62, 67, 253 Norman Conquest 6–8, 153, 184, 250 normative (grammar/tradition) 9, 191, 197, 220–221, 224, 237, 251, 253; see also prescriptive tradition, proscription, style guides Norse (Old Norse) 103, 124–126, 129, 172, 184–187; see also Scandinavian Norwegian 45, 50–51 OED see Oxford English Dictionary Old English see English Old High German see German Old Norse see Norse

Index opacity morphological 149 semantic 30, 127, 129 order see serialization Ormulum 101, 113, 129 OV see serialization Oxford English Dictionary (OED) 2, 117, 131, 137, 138–140, 183, 186, 198–200, 204, 208, 210, 242–244, 251 palaeographical clues to prosody in Old High German 110; see also spelling and morphological boundaries particle see adverb, postposition (of particle), prefix (separable), preposed particle, preverb, serialization (of particle and object), verb-particle construction particle shift 103, 112 particle verb, term 3–4 Paston Letters 113, 191–194, 202– 203, 209, 213 path (of change) 6, 8, 65, 79, 82, 87– 88, 146, 156, 250 semantic 165, 175, 180–182, 195 syntactic 104–105 Penn-Helsinki Parsed Corpus of Middle English 111 Pennsylvania German 46 perfectivizer 171; see also aspectualizer periodization see English periphrasis 55, 63, 67 periphrastic word formation see word formation (periphrastic) persistence 102, 157 perspicuity 232, 236, 238–239, 242, 245 Peterborough Chronicle 100–101, 112–113, 125, 129 phonological constraints on verbparticle combinations 191, 250

299

phonotactics 159–160, 162–164, 185 phrasal verb in American English see varieties in British English see varieties ‘rise’ 62, 74, 90, 100, 118–119, 121, 125–127, 129, 143, 145, 155, 197, 200, 216, 249, 251 term 2–4 phrasal-prepositional verb 11, 26, 28–29, 37–38, 195, 206 phraseology 55–56, 59, 61, 70, 80, 225, 238, 241–242 plain language 238, 242 plays 200–201, 238; see also drama pleonasm 19, 177, 184, 189–194, 233, 239–240, 251, 253; see also doubling, redundancy, superfluity poetry 114, 147, 217, 234 politeness 237 polysyllabic verbs 7, 237 postposition (adposition) 4, 59, 75, 82, 85–88, 100, 248 postposition (of particle) 4, 5, 51, 59, 66, 103, 106, 114, 119, 130, 131, 133, 135–143, 198, 239 predicate (complex/composite) 8, 16, 36–37, 64–65, 77, 182–183, 248 prefix bound 49, 63–65, 77, 147, 164– 165, 167–169, 250 inseparable 6, 50–51, 60–61, 65, 75, 77–78, 81, 104–105, 109, 119–120, 128–130, 132, 134, 138, 143, 146–149, 156, 160, 176, 198, 248 inventory 65, 145, 147–157, 170, 175 loss of prefixes 8, 87, 120, 127– 128, 145, 147, 150, 155–164, 177, 180, 195, 198, 249–250; see also attrition (phonetic) separable 3, 4, 8, 46–47, 49–50, 53, 60–61, 64–65, 69, 75–77,

300

Index

79, 88, 119, 124–125, 129– 130, 148 prefix verb see prefix, preverb prefixoid see affixoid preposed particle 59, 111, 113, 133, 136–139, 141–143 preposition 10–11, 26, 67, 75–76, 82–86, 88, 92, 105, 124–125, 127, 129–130, 140, 149, 163, 195, 213, 219, 220, 242, 248 intransitive 21, 27–28, 38, 66, 248 Latin terminology (praepositio) 234–235 stranding 29, 105, 128, 233–239, 251; see also Dryden, monosyllables see also adposition, preverb prepositional adverb 75–76 prepositional passive 29 prepositional phrase 15, 85, 106 prepositional verb 2, 11, 26–29, 37– 39, 41–42, 54, 62–63, 66–67, 118, 195 present-day English see English prescriptive tradition 219–220, 222, 228, 233, 236–237, 253; see also normative tradition, proscription, style guides preverb 6, 8, 35, 39, 48, 50–51, 53, 63–65, 75–89, 104–105, 115– 116, 143, 145–147, 149–150, 156, 158, 160, 164–169, 171, 175–177, 185, 195, 248–250; see also Delbrück, FinnoUgric, Gothic, Greek, Hittite, Indo-European, Kartvelian, Latin, Latvian, tmesis, Tocharian, Wackernagel’s Law, Warlpiri preverbal particle 6, 8, 48, 50, 63– 64, 69, 75–79, 105, 107–109, 111, 113, 115, 119, 134, 143, 149, 156, 160, 164–165, 180, 249–250

productivity 11, 14, 30, 35–36, 41, 57–58, 60, 63–64, 81, 120, 130, 132, 147, 149, 150, 152– 157, 159, 161–162, 164, 169, 177, 184, 186–187, 190, 193, 200, 212, 214, 253 pronominal object 11, 95, 99, 101, 102, 114, 118, 235 proscription 220, 221, 237; see also normative (grammar/tradition), prescriptive tradition, style guides prosody 27, 50, 86–87, 110, 146, 159, 164, 177, 250; see also stress purism 229 quantitative approaches 9, 110, 111, 120, 192, 197–198, 200, 202– 203, 206–207, 211, 213–214, 228, 245, 250–253 redundancy 17, 128, 190, 192, 227– 228, 239–240; see also doubling, pleonasm, superfluity register 43–44, 102, 126, 153, 187, 197, 212–213, 245 Renaissance 200, 230, 237, 240; see also English (Early Modern English) replacement of phrasal verbs 34, 40–41, 43, 59, 122, 202, 227–228, 231 of prefix (verb) 8, 127–130, 146, 155, 162–164, 195, 250 Righthand Head Rule 68–69 ‘rise’ of the phrasal verb see phrasal verb (rise) Romance influence on English 119, 127, 153, 155, 157–158, 162, 185– 187, 198, 231, 249 languages 82, 84, 176, 183, 184, 189

Index see also French, Italian, Latin Runic 91–92; see also ProtoGermanic satellite-framed see motion-path Scandinavian 6, 47–48, 51, 99, 103, 123, 128–130, 137, 184–185, 187; see also Norse schematicity 56 semantic types of phrasal verbs 13– 20 semantic weakening 127, 161, 175, 177; see also desemantization sentence brace see clausal brace separable prefix see prefix, preverb, tmesis serialization and idiomaticity 15, 24–25, 81 (basic) word order 5–6, 47–48, 77, 81, 84, 85–87, 88 English word order 8, 59, 74, 90– 103, 128, 156–157, 160, 185, 235, 239, 247, 249 extraposition 91, 104, 108, 114 in subordinate clauses 47–48, 76, 92–97, 99, 101–102, 105–106, 108, 111–113 multifactorial analysis 23 V-2 47–48, 74, 90–99, 101–104, 107–109, 116 V-3 47–48, 247 V-F 47–48, 74, 90–94, 96, 98– 99, 101, 106–107, 110–111, 116 of particle and object 4–6, 11, 21–25, 47–54, 70–72, 103–116 of verb and particle 51, 88–89, 103–116 OV 6, 47, 56, 86–94, 96–98, 100–104, 249 SVO 48, 90, 99–103, 11, 112, 115, 116, 160 247 VO 48, 86–94, 96, 101, 103–104, 108, 110 see also clausal brace, fronting,

301

inversion, postposition (of particle), preposed particle Shakespeare, William 200, 212–213, 215–217, 222 Slavic 81, 84, 166, 176 small clause 21, 28, 60 spatial adverb/particle/prefix/preverb 5, 7, 11, 18, 65, 79, 137, 149– 151, 153, 156, 159, 167–170, 174–176, 179–182, 184, 188– 189, 195 spelling and morphological boundaries in Afrikaans, Dutch and German 49–51, 53 in Anglo-Saxon manuscripts 110–111; see also palaeographical clues to prosody in Old High German sprachbund 81 Standard English see English standardization 243 stereotype (Bolinger) 14 stranded preposition see preposition stranding stress 27, 35–36, 50, 68, 78, 81, 104, 119, 136, 147–149, 156, 162, 185, 190, 191, 244; see also prosody structural nativization see varieties of Modern English style 7, 10, 37, 40, 42–45, 54, 72, 94, 101, 111, 118, 121–122, 125– 126, 127, 172, 185, 187, 191, 197–198, 209, 211, 214–215, 217–218, 220–222, 224, 226– 240, 243–244, 245–246, 247, 250–251; see also formality, colloquiality, middle style style guides 1, 20, 42, 215, 235, 239–240; see also normative tradition, prescriptive tradition, proscription suffix 31–33, 50, 67, 69, 100, 157, 160, 212

302

Index

superfluity 190, 240; see also conceptual overlap, doubling, redundancy, pleonasm superlative verbs 192 Svan 78–79, 84; see also Kartvelian SVO see serialization Swedish 45, 50–53, 80, 117 synchrony vs. diachrony (dichotomy) 2 synonymy with a simple verb 6, 38– 41, 43, 118, 121–123, 126, 166, 191, 228; see also replacement of phrasal verbs synthetic(ity) 3, 39, 63, 64, 118, 139, 158; see also analytic(ity) vs synthetic(ity), drift telicity 17–18, 175; see also aspect(ualizer) text type 119, 197, 205, 207, 209, 211–212, 214–216 tmesis 49, 77–78, 81, 83–84, 167; see also preverb Tocharian 83, 85 tracts see Lampeter Corpus transitivity 11, 20–21, 23, 27–29, 37, 42, 44, 62, 72, 106, 165, 211– 212 Turkish 39 Tyndale, William 216–217 typology 2, 47, 55, 64, 77, 79, 85– 87, 90, 116, 118, 157, 180, 184, 189, 249 univerbation 84–85 usage guides see normative tradition, style guides

V-2 (etc) see serialization valency 38, 59–60, 67, 77, 252 varieties of Modern English American English 42, 122, 202– 205, 207, 213, 238, 240, 253 British English 42, 202–205, 213, 240 non-Standard 253 structural nativization 42 variational space 2 variationist approaches 2, 197, 243, 253 Vedic 82–83, 85 verb-particle construction, term 3–4 vernacular 157, 202, 233, 243, 250 VO see serialization vulgarity 222, 237, 239, 242 Wackernagel’s Law 78, 84, 166–167 Walker, Obadiah 234–235 Warlpiri 81 Webster, Noah 117, 221, 225 Wiseman, Charles 219 word formation (periphrastic) 8, 10, 38, 55–73; see also affix, compound, derivation, nominalisation, periphrasis etc word order see serialization Yiddish 46, 48, 53, 92 York-Toronto-Helsinki Parsed Corpus of Old English Prose 109

Phrasal Verbs Tesis

Recommend Documents