Argonauts of the West Balkans? Origins, spread and distribution of Haplogroup E-V13 v 1.2 Raf Ceustermans 2017-04-29
Contents 1 General introduction to Y-DNA research
1
2 V13, part of Haplogroup E
1
2.1 2.2 2.3 2.4
Introduction . . . . . . . . . . . . . . . . . . . The discovery of V13 and early theories . . . Further refining of the tree and new theories . Geographic spread of V13 by region . . . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
3 The geographic spread of different subgroups of V13
3.1 3.2 3.3 3.4 3.4 3.5 3.5 3.6 3.6
From L618 to V13 . . . . . . . . . . . . . . . . . . . . . . . . . . . . From V13 to CTS5856 . . . . . . . . . . . . . . . . . . . . . . . . . . V13 > Z1057 > CTS5856* . . . . . . . . . . . . . . . . . . . . . . . . CTS5 CTS585 8566 > A9723 (BY6203) . . . . . . . . . . . . . . . . . . . . . . CTS5 CTS585 8566 > S3003 . . . . . . . . . . . . . . . . . . . . . . . . . . . . CTS5 CTS585 8566 > S7461 . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.6. 3.6.11 CTS5 CTS585 8566 > S7461 > A8612 . . . . . . . . . . . . . . . . . . . 3.7 3.7 CTS5 CTS585 8566 > Y16729 . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.8 3.8 CTS5 CTS585 8566 > Y19509 . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.9 3.9 CTS5 CTS585 8566 > Z16663 . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.10 CTS5856 CTS5856 > BY6527 > Z38899 . . . . . . . . . . . . . . . . . . . . . 3.11 CTS5856 CTS5856 > Z5017 . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.11.1 3.11.1 CTS5856 CTS5856 > Z5017 > Z38334 . . . . . . . . . . . . . . . . . . 3.11.2 3.11.2 CTS5856 CTS5856 > Z5017 > Z19851 . . . . . . . . . . . . . . . . . . 3.11.3 3.11.3 CTS5856 CTS5856 > Z5017 > Z5016 > SK888 > CTS9320 . . . . . . 3.11.4 3.11.4 CTS5856 CTS5856 > Z5017 > Z5016 > SK888 > CTS9320 > S19928 . 3.11.5 3.11.5 CTS5856 CTS5856 > Z5017 > Z5016 > SK888 > CTS9320 > Z16988 . 3.11.6 3.11.6 CTS5856 CTS5856 > Z5017 > Z5016 > SK888 > CTS9320 > Z17107 . 3.11.7 3.11.7 CTS5856 CTS5856 > Z5017 > Z5016 > SK888 > CTS9320 > Z17264 . 3.11.8 3.11.8 CTS5856 CTS5856 > Z5017 > Z5016 > SK888 > CTS9320 > Z37530 . 3.12 CTS5856 CTS5856 > Z5018 . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.12.1 3.12.1 CTS5856 CTS5856 > Z5018 > A2192 . . . . . . . . . . . . . . . . . . . 3.12.2 3.12.2 CTS5856 CTS5856 > Z5018 > L17 . . . . . . . . . . . . . . . . . . . . 3.12.3 3.12.3 CTS5856 CTS5856 > Z5018 > Z16242 . . . . . . . . . . . . . . . . . . 3.12.4 3.12.4 CTS5856 CTS5856 > Z5018 > Z17293 . . . . . . . . . . . . . . . . . . 3.12.5 3.12.5 CTS5856 CTS5856 > Z5018 > S2979 . . . . . . . . . . . . . . . . . . . 3.12.6 3.12.6 CTS5856 CTS5856 > Z5018 > S2979 > FGC33621 . . . . . . . . . . . 3.12.7 3.12.7 CTS5856 CTS5856 > Z5018 > S2979 > Z16659 . . . . . . . . . . . . . 3.12.8 3.12.8 CTS5856 CTS5856 > Z5018 > S2979 > Z16659> L241 . . . . . . . . . 3.12.9 3.12.9 CTS5856 CTS5856 > Z5018 > S2979 > Z16659 > Y3183-S2972 . . . . 3.12.10 3.12.10 CTS5856 CTS5856 > Z5018 > S2979 > FGC114 C11457 57-F -FG GC1145 11451 1 . . . . .
1 2 2 3 3
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
4 5 6 6 7 7 7 7 8 8 8 8 8 8 9 10 10 10 11 11 11 11 11 11 11 12 12 12 12 13 14
4 Possible clues about migrations by V13-descendants
15
5 Genes, language and archeology
17
6 Linguistics: V13 and Indo-European
17
7 Archeological insights
7.1
7.2
The Eneolithic (c. 30 3000 - c. 23 2300 BC) . . . . . 7.1.1 Vucedol and the Early Helladic I to IIa IIa 7.1.2 The Post Vucedol per period - EHIIb IIb . . . . The early Bronze age (23 (2300-2100 BC) . . . . .
19
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
19 19 20 20
8 Balkan genetics
22
9 Puzzling it all together
24
10 Further research
25
1
General introduction to Y-DNA research
All males in the world descend in a straight paternal line from one male who lived roughly 260.000 years ago in Africa (designated Y-Adam). All current living males have inherited his Y-chromosome since this is passed on from father to son. However in all these thousands of years mutations (SNP’s and STR’s, see further) happened on the Y-chromosome. These mutations are passed on from father to son, and they allow us to identify different branches of the Y-chromosome. Such a branch is called a haplogroup. The term haplogroup is generally reserved for larger groups, underneath those terms like subbranch or subclade are used. In Y chromosome research two types of mutations are examined: STR’s (Short Tandem Repeats) and SNP’s (Single Nucleotide Polymorphisms). STR’s are small pieces of DNA that are repeated multiple times in the chromosome. When DNA is copied errors might happen whereby extra repeats are added, or some repeats are dropped. An STR-test tests a number of STR’s (generally 12, 25, 37, 67 or 111) and reports back the number of repeats for each of them. These STR-tests are useful for the very distant past and the very recent past, but not for the period in between. In the very distant past STR are correlated to the largest haplogroups, and such an STR-test will result in a prediction for a main haplogroup. There are several free predictors available on the internet, a very good one can be found at www.nevgen.org. The other mutation examined is the SNP. An SNP is a mutation of 1 letter of the Ychromosome (which is a chain of millions of such letters). SNP’s define different haplogroups, and since they are more reliable for the last thousands of years SNP-testing is generally required to pinpoint your exact haplogroup. For those who are unfamiliar with SNP testing there is a very good lecture by John Cleary that is freely viewable online. 1
2
V13, part of Haplogroup E
2.1
Introduction
V13 is one of the many branches of haplogroup E. In the early days of phylogenetical research letters were used as names for what researchers thought to be the principal haplogroups around the world. This classification was not entirely unbiased as a path downwards towards typical European (or maybe more accurate, typical white Anglo-Saxon) haplogroups was followed. Every side branch from this path got designated a letter, so that large parts of the world were lumped together under one letter, while very recent ”European” branches got their own letter. The letters were given out alphabetically and so it’s no surprise E split of fairly early, possibly around 50 to 60000 years ago. Different branches underneath these letters are nowadays designated by a defining mutation (like V13). As such we can write the path to V13 as follows: Y Adam > A0T > A1 > A1b > BT > C T > DE > E > P 147 > P 177 > P 2 > M 215 > M 35 > L539 > M 78 > Z 1919 > L618 > V 13
Branches of E dominate sub-Saharan Africa today, but most are rare outside of this continent, the exception being M35.2 The common ancestor of all M35 probably lived about 25000 years ago, in the Horn of Africa. Some descendants migrated northwards through the Nile Valley and would eventually be part of Neolithic migrations that brought farming to South-East Europe. Z1919, whose common ancestor would have lived about 12000 years ago has sub branches in Europe, the Levant (Middle east) and the Arabian peninsula. One subclade of Z1919 however, L618, is limited to a European distribution. It likely mutated somewhere in Southeast Europe about 8000-9000 years ago.
1
https://www.youtube.com/watch?v=hlxvdayxZiI Off course the slave trade of the last 500 years brought a lot of these Sub-Saharan E-branches accross the Atlantic ocean where they are now present in descendants of the former slaves. 2
1
2.2
The discovery of V13 and early theories
For a long period no sub branches under M78 were known. As a result genetic studies in the early 2000’s could only test up to this SNP and found it over all of Europe, North-Africa, the Levant and in Arabia. This distribution seemed to defy the regional logic that was observed for other haplogroups. If STR-profiles were taken into account the picture changed, STR-profiles were clearly clusterd in geographic groups of which the European was dubbed α. Only in 2006 was V13 finally discovered by a scientific team (together with five other sub M78 SNP’s). Testing of M78+3 persons revealed that the European α-cluster of STR-profiles corresponded to the V13 SNP.4 Subsequent research in 2007 showed that V13 was present in most Western European populations in 1-5%, but notably higher in most of Southeast Europe and in Southern Italy and on Sicily. It reached its peak in Albania at 32%. Based on STR-diversity a bronze age expansion of V13 was theorized.5 In 2008 Battaglia calculated a much older coalescence time 6 for V13 (about 10000 years ago) and proposed that V13 was a Mesolithic Balkan line that had adopted farming and spread it further into Europe. 7 This theory would be prevalent over the next years.
2.3
Further refining of the tree and new theories
In the years following the discovery of V13 sometimes SNPs underneath V13 were discovered, but their significance and relationship to each other was not always clear. Most of these were discovered at the company FTDNA, and because the large majority of their customers are Americans of British or German descent there was some bias towards SNP’s that were especially present in these populations, like L17, L143 and L540. In the past years the focus in Y-chromosome testing has shifted to ”next-generation-sequencing” where large parts of the Y-Chromosome are sequenced to discover new SNP’s. This has caused a massive increase in sub V13-SNP’s. This next generation sequencing also allows a better estimation of the age of SNP’s than STR distributions.8 Nowadays calculations based on SNP’s place the most recent ancestor of V13 in or right before the Bronze age, confirming Cruciani’s original theory. By 2007 it was already suggested that V13 in the British Isles was due to Roman troops from the Balkans rather than from Neolithic dispersal.9 In 2009 Di Gaetano had relied heavily on V13 to show that Sicilians had Greek heritage.10 So the combination of a Bronze age origin for V13 with the pan-European distribution and the higher levels on Sicily and in Southern Italy strongly suggest that V13 was spread from the Southern Balkans by both Greek colonists as well as Roman troops (from the Balkans, or possibly from Southern Italy and Sicily). 3
as a convention in this text + after a marker means positive for this marker, -means negative, and * means positive for the marker but negative for all known subclades. 4 Cruciani et al., Molecular Dissection of the Y-Chromosome Haplogroup M-78 (E3b1a): A Posteriori Evaluation of a MicroSatellite-Network-Based approach through six new Biallelic Markers, 2006. 5 Cruciani et al., Tracing Human Male Movements in Northern/Eastern Africa and Western Eurasia: New clues from Y-Chromosome Haplogroups E-M78 and J-M12, 2007. 6 Coalscence time: time to a common ancestor calculated by calculating the variation in a lot of STR-profiles and calculating how much time is needed to reach that variation. 7 Battaglia et al, Y-chromosomal evidence of the cultural diffusion of agriculture in southeast Europe, 2008. 8 In this document the calculated ages from the V13-draft tree by Steve Fix are generally followed. Other dates are available from Yfull (www.Yfull.com/Tree/E-V13/) and are generally younger. 9 Bird, Haplogroup E3b1a2 as a Possible Indicator of Settlement in Roman Britain by Soldiers of Balkan Origin, 2007. 10 Di Gaetano et al., Differential Greek and northern African migrations to Sicily are supported by genetic evidence from the Y chromosome, 2009.
2
2.4
Geographic spread of V13 by region
The above map is taken from www.eupedia.com. It shows percentages of V13 over Europe, based on scientific studies that try to remove any bias. It shows that nowadays the highest percentages of V13 are in the Balkans with a peak in Kosovo.
3
The geographic spread of different subgroups of V13
The next part gives an overview of the different subbranches of V13 with info on their geographical distribution. Since most scientific studies haven’t tested below V13 this is largely based on data from public DNA projects at FTDNA 11 , and it is important to realize that results there are heavily skewed towards white Anglo-Saxon Americans. A second source is the The National Geographic Societys Genographic Project 12 , which generally has a bias comparable to that of the FTDNA projects. A few scientific projects did test beneath V13 and are used here. These are principally a Whole Genome Sequencing project of 1200 Sardinians 13 , and the 1k Genomes project14 that did Whole Genome Sequencing on worldwide people. 11
https://www.familytreedna.com/ https://genographic.nationalgeographic.com/, with many thanks for giving me the opportunity to study their results 13 Francalacci et al., Low-pass DNA sequencing of 1200 Sardinians reconstructs European Y-chromosome phylogeny, 2013. With many thanks to the team for letting me study the raw data. Most of the samples are now also available on the Yfull-tree (www.yfull.com/sardinians/) 14 http://www.internationalgenome.org/ 12
3
The tree above is a simplified view of V13 and the major subclades that will be mentioned in the next paragraphs.
3.1
From L618 to V13
Before turning to the different subgroups of V13 it’s interesting to look at its direct parent, L618. L618 is currently defined by 57 different SNP’s. The common ancestor of all L618+ persons lived about 8000 to 9000 years ago, probably in an early Neolithic society in Southeast Europe. Today L618 is dominated by V13, which appears to form more than 99% of the L618 population, those who are L618+, V13- are denoted as L618*. There are seven proven L618* samples known at this moment, they originate from Sardinia, Bari (Southern Italy), Latvia, Denmark, England, Brittany and Germany. As most of these come from commercial testing companies there is no doubt a large selection bias at work. There are also two ancient DNA samples that merit attention. Ancient DNA is a very powerful tool as it confirms the presence of a group at a certain place and time. Unfortunately ancient DNA studies have not uncovered a lot of V13 up to now. However a Spanish Neolithic skeleton found in a cave turned out to be V13+, how is this reconcilable with a Bronze age
4
origin in the Balkans?15 V13 is only one of 33 SNP’s defining the V13 group. As everyone who is V13+ is also positive for the 32 other SNP’s we have no way of knowing which of these 33 mutations happened first, and which one last. The Spanish sample most likely already had the V13 mutation, but not all of the 32 V13-equivalent SNP’s (they were not tested so we can’t be sure either way). This Spanish sample would represent a side branch of V13, of which there might be no more descendants today. Another ancient DNA sample from the Neolithic Lengyel culture in Hungary was found L618+, it’s not unlikely it’s also somewhere between L618 and V13 (there is no test result known for V13). 16
3.2
From V13 to CTS5856
Just as L618 is dominated by its main clade V13, the latter is also made up almost entirely of one daughter clade: CTS5856. Those V13 that are not CTS5856+ are designated V13*. As explained above V13 has 32 equivalent SNP’s and might have consisted of multiple branches spread out over Europe that have since died out. At the start of the Neolithic period the population in Europe boomed rapidly due to the fact that farming can sustain a much larger population in a same territory than the previous hunting and gathering lifestyle. But after this initial boom populations started to dwindle again all over Europe, intensified by a very bad climate around 5900 years ago. By the end of the Neolithic, about 5500 to 5000 years ago, one V13 lived who once again saw his descendants boom. The location where we find V13* today might give some clues about where this common V13 ancestor lived.
Map with localisable V13* samples. Red dots are from public FTDNA-projects, orange from The National Geographic Societys Genographic Project. The samples in North-Western Europe are concentrated mostly along the border regions of the Roman empire. The large number of samples here is also connected to a large population, and large emigration to the USA in the past. V13* is relatively scarce but is present in almost all of Europe. Since it’s not tested in scientific studies with random selection it suffers from severe selection bias towards Anglo-Saxon 17 15
Lacan et al., Ancient DNA reveals male diffusion through the Neolithic Mediterranean route, 2011. Szcsnyi-Nagy, Molecular genetic investigation of the Neolithic population history in the western Carpathian Basin, PhD thesis Johannes Gutenberg-Universitt in Mainz, 2015. And deeper classification in: Lipson et al., Parallel ancient genomic transects reveal complex population history of early European farmers, 2017. 17 In this text Anglo-Saxon denotes people with typical English surnames, generally living in the USA, but can occassionaly also cover people living in countries with a large British heritage like Australia, or even Ireland and 16
5
samples. That makes it all the more surprising that on the British Isles it’s only present in one Scottish, one Irish and three English samples. The Irish sample is closely connected to a Macedonian sample, probably sharing a common ancestor about 2200 years ago, which strongly suggests that the Irish sample descends from a Thracian soldier stationed in Britain. Apart from these five British there are some other Western European V13*’s: 1 Danish, 1 Spanish, 1 Belgian, 1 Luxembourgish, 1 Dutch, 9 German, 8 Italian (from near Rome, Genoa (2), Mafalda (Molise), Milan, Naples (2) and Salerno) and 2 Swiss. For the Italian samples it’s remarkable how many seem to come from middle and northern-Italy rather than from Southern Italy (including Sicily), as there is both a higher percentage of V13 in Southern Italy, as well as a larger historical emigration towards the USA leading to more testers from this area. Outside of Western-Europe we find 1 Hungarian, 2 Serbian-Croatians, 1 Serbian, 1 Greek, 2 Turkish, 1 Iranian, 1 Belarus (Ashkenazi), 1 Polish and two Serbo-Montenegrin clans. The first Serbian-Montenegrin clan are the Rajovici, a large family concentrated in the village Zunjevice, close to the Serbian-Kosovar border. Their own oral history claims that in the 17th century they moved here from the territory of the Kuci-tribe in Montenegro, led by a certain Rajo. In this Kuci-region a small hamlet by the name of Rajovici (village of Rajo) exists, which might lend some credibility to the story.18 The Kuci tribe lived in eastern Montenegro, close to the border with Albania. It was mostly Greek-orthodox and most members had Serbian names, but Albanian names were also common. The Kuci tribe (like a lot of other Greek-orthodox tribes) claimed descendance from a medieval Serbian noble. This is possible, but another possibility is that Albanian catholics converted to the Greek orthodox religion and took on a Serbian identity.19 The Kuci also seem V13, but they don’t appear closely related to the Rajovici. The other clan are the Vasojevici, they are a much larger group, living in an area northeast of the Kuci. According to 15th century sources they originated more to the south, not very far from where the hamlet Rajovici is situated. Like the Rajovici the Vasojevici identify as Serbian and Greek-Orthodox, in the recent Montenegrin independence referendum they voted to stay part of Serbia in overwhelming numbers. Also like the Rajovici they claim descendance from a Serbian noble family from the Middle Ages. 20 Recent research has also shown a new level between V13 and CTS5856, Z1057 (called CTS8814 at FTDNA), from those tested the Scot was negative, and the Swiss, the Pole, the Macedonian the Irishman and the Vasojevici positive. All these, except the Scot off course, were also positive for PH1246 under Z1057. As stated above the Macedonian and the Irishman form a branch together, as do the Swiss and the Pole.
3.3
V13
>
Z1057
>
CTS5856*
A number of people who tested superpacks or Big Y were found negative for all subgroups. Due to the large number of sub CTS5856 groups (that is still expanding) few people have been tested for all possibilities. Tested for all important SNP’s and most minor SNP’s are two Englishmen, two Germans, a Pole, a Swiss, two Italians (one from near Milan) a Greek from Arcadia and a person from Kosovo/Macedonia. Tested for most major SNP’s but not for Z5017/Z5018 are two Germans, as Swiss and two Englishmen.
3.4
CTS5856
>
A9723 (BY6203)
For now this group has been found in a Bulgarian and a French sample. This group is relatively small, but seems to confirm to a pattern of Southern Balkan origin and spread by Greek Colonisation/Roman soldiers. the UK. 18 communicated by a Rajovici family member 19 Robert Elsie, The tribes of Albania, History, society and culture, 2015, p.13. 20 https://en.wikipedia.org/wiki/Vasojevi%C4%87i
6
3.5
CTS5856
>
S3003
This rather large group is somewhat exceptional within CTS5856. Almost all S3003+ persons are descended from one common ancestor that lived about 1500 years ago and had the L540 mutation. These L540+ persons live mostly in Eastern Germany and Poland, with occasional members from neighboring countries to the north and east. This seems to imply a medieval expansion that can either be linked to the Slavic expansion in the early middle ages (500-1000AD), or colonization (Ostsiedlung) by Germans in the high middle ages (1000-1500AD). Only one non L540 S3003 sample is known, a Hungarian from a scientific study. So S3003 is completely absent from the Balkans in our current knowledge. Hopefully additional sampling in the future will shed light on how S3003 fits in the CTS5856 puzzle. One of the L540’s has done extensive research into this group and set up a website to document it: http://www.gwozdz.org/L540.html.
3.6
CTS5856
>
S7461
This group again conforms to a basic pattern of Balkan origins and a spread through Greek/Roman colonization. Although a late discovery it seems this group is the most numerous subclade behind Z5018 and Z5017. Most of S7461 is part of A8612, those who are not are classified as S7461*. There is one confirmed S7461* from Switzerland. We can identify a number of other probable S7461* based on two criteria: first there is a possible defining marker YSC0000078, this marker was also found in a The National Geographic Societys Genographic Project member from Puglia (the heel of Italy). The second criteria we can apply is a very disctinct marker at DYS19=9, this is present in the known Swiss S7461* and some relatives. Apart from that it is found in one Irish and four Italians (one from Anzi in Basilicata, one from the shores of Lake Como, very close to Switzerland, and a family from either Sicily or Campania). Although no S7461*has been found in the Balkans we can suspect a Balkan origin. One Greek from the Pelopennese was recently tested S7461 (but not tested deeper). 3.6.1
CTS5856 > S7461 > A8612
The same pattern can be observed for A8612+ samples. They mostly come from Southeast England and from places close to the Rhine, where in antiquity a lot of Roman soldiers were stationed. In the Balkans we have two Bulgarian samples, but none from the western Balkans or Greece. One of the most remarkable aspects of A8612 is its presence in two Druze samples. From older scientific studies it is known that Druze, a distinct religious group living mainly in Lebanon, Syria and Kurdistan, had moderate levels of V13. Because of the lack of testing of Druze it is hard to understand why they have V13. The two A8612 are the only known Druze that have tested, and its not clear whether this is a coincidence or A8612 really is dominant within the V13-Druze. In any case this is not due to a recent founder effect, as the two Druze are on very separated lines, one of them is in fact closer to an English sample than to the other Druze. The English and Druze both belong to the group Z40644 where they form the subgroup BY5169. Also in Z40644 but BY5169- is another Englishman. Another A8612 subgroup is Z40645 to which an Englishman, a German and an Anglo-Saxon belong. A third, tentative group is FGC44177, made up of a Belgian, a Swede and a third person of unknown origins. The Swede and the unknown form a subclade. Within FGC44177 there is also a possible marker we can use to identify The National Geographic Societys Genographic Project samples. It seems this marker is quite a bit younger than A8612: CTS1489. It is shared by the Belgian, an Irishman, a German (Bavaria) and two Sicilians (Caccamo and agrigento). However this marker is also positive in other Haplogroups, so its relevance is still under question.
3.7
CTS5856
>
Y16729
Y16729 is a very remarkable group as it splits in two big groups: a British and an Arabian. No Y16729+ have so far been found outside of these regions. Within Britain the samples are mostly concentrated in Scotland and Northern Ireland. These British samples all descend from a quite recent common ancestor who lived about a thousand years ago. The split with the Arabian 7
branch is much earlier. Y16729 probably is a very small group, and maybe future research will find some in the Balkans.
3.8
CTS5856
>
Y19509
This group is as remarkable as Y16729 in having only been found North-West Europe. There is a German/Dutch family in this group (descending from a common ancestor in the late middle ages), but most members come from Wales or the rest of the British Isles (descending from a common ancestor around 400BC). This probably connects it to a scientific study that was conducted in 2011 and found that one unexpected haplogroup made up 30% of a study of 500 people in North-Eastern Wales. Although never officially confirmed it is clear this group was V13. They might be linked to a founder effect of an Y19509, although it’s not known at this moment what subgroups are in this 30%. Did this group wander of early and end up in Britain? Or are they also descended from Roman soldiers but did we not yet find relatives in Italy or the Balkans? Telling is that the region where the study was performed is right next to Chester, the original Roman capital of Britain. It is even possible we will never be able to have an answer, as maybe all Y19509 is extinct outside of North-West Europe.
3.9
CTS5856
>
Z16663
Until recently the Z16663 group was only known from scientific studies (found in two Sardinians and one Tuscan), but recently several people were found positive through SNP-testing. It was found in a Pontic Greek (Greeks that lived around Trabzon in Turkey on the Black sea coast, either descending from Classical age Greeks, or medieval Byzantine colonists), a Slovakian, a Dutchmen, an Anglo-Saxon and a Swede. The Slovakian forms a subbranch with the Sardinian samples (Z21731). The three other members from a different branch.
3.10
CTS5856
>
BY6527
>
Z38899
Another very rare subbranch is BY6527. Found in a Sardinian sample from a scientific study, a Sicilian and an Anglo-Saxon. The Sardinian and the Anglo-Saxon form a branch together, Z38899. Again no samples have been found from the Balkans yet, but it might conform to a Balkan origin spread by Greek colonists to Sicily and Sardinia. It’s important to remark that none of the three members of this group has been tested for Z5017, and so this might well turn out to be a subbranch of Z5017.
3.11
CTS5856
>
Z5017
There are two Z5017 samples that do not belong to the three known subgroups of Z5017: a Greek (from Epirus) and an Italian. All others belong to Z38334, Z19851 or Z5016. 3.11.1
CTS5856 > Z5017 > Z38334
Remarkably this group contains a Swede, a Dane, an Anglo-Saxon and a Tatar. They appear to have had a common ancestor around 1000BC. Generally Scandinavian samples point to a link with Bulgaria, but no Bulgarian Z38334 has been found so far. 21 3.11.2
CTS5856 > Z5017 > Z19851
This group so far has been found in a Swede, a Serbian, a Greek (from Thessaly but likely with Bulgarian roots) an Austrian, a Bulgarian, a Scot, an Englishman and an Italian (from near the border of Emilia Romagna and Lombardy). Based on STR clusters this SNP is thought be 21 In multiple subbranches of V13 it seems like Scandinavian and Bulgarian samples are close to eachother. Some checks in other Haplogroups seem to confirm this pattern. It seems likely that this is a phenomenon from the early Middle ages, probably linked to Slave trade from Bulgaria to Viking age Scandinavia. That would also explain why we find V13 in this region where no Romans set foot.
8
widespread in the western Balkans, it has been found in several members of the Serbian Orlovici tribe.22 3.11.3
CTS5856 > Z5017 > Z5016 > SK888 > CTS9320
Almost all of Z5016 belongs to SK888, the exceptions are a Tuscan from a scientific study and a Spanish sample. Likewise almost all SK888 is also CTS9320+, with the exception of a German from Hesse. This represents a sort of bottleneck as there is a period of about 1000 years between Z5016 and CTS9320, and apparently sidebranches from this period only rarely survived. CTS9320 is by far the most dominant group within Z5017. It is relatively recent, with a most common ancestor that would have lived about 1500BC or slightly later. CTS9320 that cannot be classified in one of the subgroups is found in a German, a Pole and a Greek (presumably from Izmir/Smyrna).
The above map shows the members of the E-M35 project at FTDNA belonging to CTS9320 (as far as a location for their oldest ancestor is known). Black dots represent CTS9320 that hasn’t been tested deeper. The orange dots are CTS9320*. Red is Z17264, green Z1707, purple Z16988 and blue S19928. It’s clear that CTS9320 is most prominent in the Balkans. The small cluster around Strassburg is probably related to the Roman military camp there. 22
communication by mail from a specialist in Balkan Genetics. This large cluster is also visible in Serbian DNA results: http://dnk.poreklo.rs/tabela-po jedinacne-grupe/?grp-filter=E.
9
The same graph but with samples from the Genographic project. Again there is a concentration in the Western-Balkans and a cluster around Strassburg, but there are also more Greek samples visible here. 3.11.4
CTS5856 > Z5017 > Z5016 > SK888 > CTS9320 > S19928
Found in a Sardinian from a scientific study, a Frenchman from the Alsace, a Slovakian, an Italian (from Trentino), two British and an Anglo-Saxon. 3.11.5
CTS5856 > Z5017 > Z5016 > SK888 > CTS9320 > Z16988
In the scientific study on Sardinian samples Z16988 was dominant within CTS9320 with five samples. Apart from that there is a German group (mostly from Baden), a few British/Anglosaxon, a Frenchman, a Hispanic, a Pole (maybe from Slovakia), a Norwegian group, a Greek (maybe with Albanian roots) and a Serb (from Southeast Serbia). The Pole and the Norwegians form the Z27131 branch (common ancestor about 500BC), while the Hispanic, one German and two Anglosaxons form the Z38664 branch. The French and Serbian belonged two neither of these two groups, while the others weren’t tested underneath Z16988. In the Albania DNA project Z16988 was found, and seems typical in the Kelmendi tribe.23 In Serbia this subclade also seems well represented.24 3.11.6
CTS5856 > Z5017 > Z5016 > SK888 > CTS9320 > Z17107
A small subgroup of CTS9320, found in two Albanians (one from Western FYROM, one from northern Albania), a Croatian, a Swede and a Russian. Apart from the Russian they all belong to the subgroup Z38456 (with a common ancestor who probably lived after 1000BC). In the Albania DNA project Z17107 was found, and seems typical in the Dibrri tribe. 25 It seems very absent from Serbian results, for the moment. 26 23
https://www.familytreedna.com/groups/albanian-bloodlines/about/results http://dnk.poreklo.rs/tabela-pojedinacne-grupe/?grp-filter=E 25 https://www.familytreedna.com/groups/albanian-bloodlines/about/results 26 http://dnk.poreklo.rs/tabela-pojedinacne-grupe/?grp-filter=E 24
10
3.11.7
CTS5856 > Z5017 > Z5016 > SK888 > CTS9320 > Z17264
Within this group there is a Bulgarian (from Northern Bulgaria), an Englishman, an Italian (Naples area), a Russian, a Slovak and a Spaniard that are not tested further. Two Norwegians form a group together. Another group is formed by a Sardinian from a scientific study, an Anglo-Saxon and some other samples. 3.11.8
CTS5856 > Z5017 > Z5016 > SK888 > CTS9320 > Z37530
In this group one Swede wasn’t tested deeper than Z37530. Another Swede and a Bulgarian form a separate group (A10953), a Belarussian was found Z37530+, A10953-.
3.12
CTS5856
>
Z5018
Next to Z5017 this is the other main group under CTS5856. As a general rule Z5018 tends to become much more dominant compared to Z5017 as one gets further away from the Balkans (with the exception of Scandinavia). This seems to be connected to the more recent boom of CTS9320, and the exception for Scandinavia is probably due to the clear genetic links there are between Bulgaria and Scandinavia that probably date to the early Middle Ages. Z5018 falls mainly into five large groups. Samples that are Z5018+ but not part of any subgroup are two separate French groups, a Czech, a Hungarian and a Swede. Based on a marker that appears equivalent to Z5018 there is one Romanian who seems to be between CTS5856 and Z5018. 3.12.1
CTS5856 > Z5018 > A2192
This cluster is dominated by L143, but there might be a selection bias here. L143 was one of the first sub V13 SNP’s that was discovered, and as such was tested in a lot of FTDNA-members looking for a group under V13. The L143 group is limited to English samples. Two samples are A2192+, L143-: a Czech and a Bulgarian. 3.12.2
CTS5856 > Z5018 > L17
This is one of the rather large clusters that have not yet been found in the Balkans. This is especially remarkable as L17 has been known for quite some years and has been extensively tested. Furthermore it was also tested in The National Geographic Societys Genographic Project that covered a lot of Balkan samples. L17 is mostly found in the Iberian Peninsula (and consequently in Latin America) and the British Isles. Apart from these groups samples at FTDNA are also known from Czechia, Austria, Italy, Poland and Ukraine. At least one of the Iberian samples at FTDNA is Sephardic (Spanish Jew). In The National Geographic Societys Genographic Project there were also 3 Italians (one from Sicily) and 2 Germans, the 19 others with a listed origin are from the British isles (10) or Iberian (9). How can this distribution be explained? Before V13* was found in the Balkans it seemed that the distribution closely matched that of some R1b branches and that it was spread alongside them. While this might still be true it has since become clear from NGS tests that the Iberian samples seem younger than the British. The Italian samples suggest that L17 might have been a rare branch in the Balkans, maybe not existant anymore, and that it boomed on Sicily from where it was spread by Roman colonists to both Spain and England. 3.12.3
CTS5856 > Z5018 > Z16242
This group is comparable to L17 in having a primarily Iberian-British distribution. 3.12.4
CTS5856 > Z5018 > Z17293
A rare group that has been found in a Tuscan from a scientific study, a Brit from a scientific study and an Irishman.
11
3.12.5
CTS5856 > Z5018 > S2979
S2979 covers by far the biggest part of Z5018. S2979 itself falls mostly into Z16659 and FGC11457, while there is also the smaller group FGC33621. One British man was found S2979*. 3.12.6
CTS5856 > Z5018 > S2979 > FGC33621
This marker was found in a Brit, a German, an Italian and a Bulgarian, and also in an Albanian cluster of very distinctive STR-profiles. The Albanians clearly form a recent founder effect and are mostly from the Albanian-Kosovarian borderlands, where FGC33621 seems to be the disitinctive marker of the Berisha-clan. The Bulgarian forms a group together with the Albanians. 3.12.7
CTS5856 > Z5018 > S2979 > Z16659
No Z16659* is known, the group seems about equally divided in L241 and Y3183. 3.12.8
CTS5856 > Z5018 > S2979 > Z16659> L241
There is not a lot of resolution under L241 despite quite a few NGS-tests. This seems to point to a fast, starlike expansion. Interesting is that this expansion happened after a bottleneck of a few hundred years. In this respect it closely mirrors CTS9320 and just like that marker it must have mutated quite late, between 1500 and 1000 BC, but is nevertheless a big part of V13. In the Balkans it has been found in two Kosovars (one Serbian?), a Bulgarian, and two Greeks (one from Messenia). Just like CTS9320 it was very present in the Sardinian samples from a scientific study, turning up in four of them. There are also quite some L241+ samples in Central and Eastern Europe, this seems partly connected to a group of Ashkenazi Jews that are L241. Further it was also found in an Italian from Bari, a Lybian, two Frenchmen, three Germans, a Swiss, and a large number of British samples. A Sephardic (Spanish) Jew is also L241+. The Lybian might be connected to the Greek city of Cyrene. This city reached its peak in the fifth century BC. In the late Roman period it was destroyed by several earthquakes, and abandoned. Cyrene was known for having an important Jewish population which might fit the Ashkenazi and Sephardic Jews. Off course lots of big Greek cities had Jewish communities so this might be a coincidence. The Greek colony nearest to Bari was Taranto, established by Spartans, but given the proximity to Albania the L241 there might not come from Greece at all. Nevertheless the pattern of distribution seems a good fit for Greek colonization. Very tentatively a connection to Dorian speaking Greeks could be suggested.
12
On this graph the red dots show the samples from the E-M35 project that are tested L241+.
On this graph the red dots show the samples that are L241+ from The National Geographic Societys Genographic Project.
3.12.9
CTS5856 > Z5018 > S2979 > Z16659 > Y3183-S2972
Most of Y3183 is part of S2972, but an Austrian, a German, a Scot, an Irishman and a Bulgarian were found Y3183*. Within S2972 an Englishman and an Anglo-American were found negative
13
for its two main groups: A7136 and Z16661. Within A7136 two Hispanics and an Irishman form a group (A8458) while a Greek (Southern Peloponnese), a Macedonian, a Russian and an Italian (Malta/Southern Italy) were found negative for this marker. Two Germans and an Anglo-Saxon were not tested underneath A7136. Within Z16661 most samples belong to S2978: two Sardinian samples from a scientific study, and three British samples. S2978- are a Bulgarian (Northern Bulgaria) and an Italian, while a Pole was not tested beneath Z16661. 3.12.10
CTS5856 > Z5018 > S2979 > FGC11457-FGC11451
No FGC11457* is known. One small subgroup consists of two British samples (A4979), all the rest belongs to the main subrgroup FGC11451. The subgroup FGC11451 has a special property. All V13’s that have a value of 11 at DYS607 and that have been SNP-tested turned out FGC11451+. Vice versa every FGC11451-tested sample who has tested at least 67 STRmarkers has the value of 11 at DYS607. This means a much larger dataset is available to assess the distribution of FGC11451. Unfortunately this does not make things a lot clearer. Most of FGC11451 belongs to FGC11458, four samples do not, an Est, a German, a Bulgarian/Greek and a Ukranian (or Turkish?). At least some of these four might be Ashkenazi. Two FGC11458 are negative for the only subgroup FGC11450: a Sicilian and a Mallorcan. Within FGC11450 a German, a Scot, a Pole, two Norwegians, a Swede and an Anglo-American belong to BY4793. The Norwegians form a group apart (Z24130) while the others together form FGC11447. Deeper down the Scot, the Anglo American and the Swede form a group (Z38485), and below that the Scot and the Anglo American (Z38519). A sister clade to BY4793 contains a Brit while an Albanian, a Bulgarian, a Swiss, a German and a Portuguese are negative for both these clades. A Greek and two Sardinians from a scientific study were not tested below FGC11450.
On this graph the red dots show the samples from the E-M35 project that are predicted to be FGC11451+, or are confirmed FGC11451. The orange dots are FGC11451 from The National Geographic Societys Genographic Project.
14
4
Possible clues about migrations by V13-descendants
The question many people ask is where V13 came from, and how and when it came to where they live now. In the following paragraphs some clues from the spread and distribution of V13 and its daughter clades will be analyzed. Many V13-people in Northwestern European countries can reliably trace back their ancestors to the Middle Ages. Generally there are no indications they were part of a recent migration. Looking back at population movements during the Middle Ages and the distribution of V13 it seems likely that V13 arrived at the latest in Roman times, whatever the exact origins and path of migrations might have been. That means that V13 would have arrived in Northwestern Europe before 200 AD. There is equally evidence to place at least part the influx of V13 in exactly this period. Above the connection was mentioned between an Irish V13* and a Macedonian (of possible Bulgarian descent) V13*. They seem to share a common ancestor about 2300 years ago. The most logical connection between the two would be a Roman soldier from Thracia/Macedonia being stationed in Northern England. While some V13 might have another source (which would explain the deviating distribution of some groups like L17.) a Roman source for the majority of North-Western V13 seems likely. A second question is what the source of Roman V13 was. As shown above in at least one case a link to the Balkans could be established, showing Thracian/Macedonian troops certainly formed a source. So V13 must have been present there before 200AD also. However there are also indications they might not have been the only source. V13 in the Balkans (with the exception of Greece) is dominated by the quite recent marker CTS9320. Preliminary results show it forms about 50% of all V13 in most Balkan countries. This is not the case in North-Western Europe where it is about 10%. If Roman soldiers from the Balkans were the sole source of North-Western European V13 we might expect CTS9320 to be fairly close to 50% there also. Two possible explanations are that either some V13 in North-West Europe has a pre-Roman source (for example L17), or that there were other sources for Roman migrants outside of the Balkans that also carried V13, but with a lesser percentage of V13. It was remarked above that CTS9320 is lower in Greece but there are no sources showing Greek migration towards North-Western Europe in Roman times (or before). A more credible source might be formed by populations from Southern Italy and Sicily. Here V13 is present in higher percentages than in other parts of Europe outside of the Balkans. The obvious source is Greek colonisation during the last millenium BC. As Greece is already lower in CTS9320 this would explain the lower levels found in North Western European V13’s. If we assume V13 was brought to Southern Italy and Sicily by Greek colonists that provides an additional date. Given the fact that Greek colonisation was mostly an affair of the 8th to 6th century BC, it’s clear V13 must have been present sizable percentages in Greece by then. The period between about 1200 and 800BC is called the Greek ”Dark Ages” as it is a period where the advanced society of the Mycenaean age had crumbled, and population and economy declined. Is it possible to derive whether V13 was already present in Greece during the Mycenaean age or not? As shown above Greece, certainly the southern parts of it, has a lot less CTS9320 than more northern parts of the Balkans. If V13 had arrived in Greece after the CTS9320 boom we would expect that CTS9320 would be as dominant in Greece as in the rest of the Balkans. As this is not the case we can conclude that V13 in Greece must have been present before CTS9320 boomed, so we need to date this event. CTS9320 has a common ancestor that lived about 1500BC or shortly afterwards. Nevertheless the CTS9320 expansion is likely an Iron age phenomenon connected to Illyrian/Thracian tribes, so it does not allow to conclude that V13 should have been in Greece before the dark ages. Yet if the fact that V13 seems to have been present in all parts of Greece is taken into account, it does seem likelier that it was there before the Dark Ages than not. Ancient DNA might be able to provide more solid conclusions in the future. A different clue might come from FGC11450. As explained above this marker has the advantage of being perfectly correlated to an STR-marker, and a lot more people have been STR-tested than SNP-tested. FGC11450 has a common ancestor about 4000 years ago. Its distribution seems to correlate very well to a Greek colonisation pattern, with for example a lot of Italian samples. 15
At first glance it seems more rare north of Greece. One Albanian sample is known from a Scientific study (also positive for FGC11450, but not for its daughters), while one Bulgarian Sample has been SNP-tested and found FGC11450+, but negative for its subgroups. The problem is of course that it cannot be excluded that the Albanian and Bulgarian samples are the result of later Greek Colonisation of the coastal regions of these countries (like for example the Greek colony of Apollonia in Albania), or even contacts in the Mycenaean period. Additionally, people from the Balkans tend to be underrepresented in commercial tests compared to for example Italians. They are also more likely to opt for a 12, 25 or 37 marker test than a 67 or 111 marker test, and only the last two contain the marker that is significant for FGC11450. Nevertheless with caution it can deduced that if migration to Greece had taken place long after the origin of FGC111451 it should be visible in more exclusive Balkan lineages. The fact that this is not the case but there is some FGC11451 in the Balkans outside of Greece suggests it’s origins must be not a lot sooner or later than the flow towards Greece. FGC11451 is a daughter of FGC11457 whose sister clade FGC33621 seems to be more limited to the Balkans. This seems to suggest that a few hundreds of years before 2000 BC the common ancestor of both (S2979) would have lived north of Greece, and his descendants would partially stay there, and partially migrate southwards. This also confirms that the migration took place from north to south, as we have branches in the north that we don’t find in Greece, but not the other way around. That is hard to explain if one would place the origin of V13 in Greece itself. The dates given above should be interpreted with a margin, as there is always a variation on the number of mutations per generation. Recent finds of V13* point to the Montenegrin-Serbian-Albanian border regions as a probable candidate for the emergence of V13 and CTS5856. Nevertheless, it should be kept in mind that generally the diversity in V13-groups is larger in the Eastern Balkans (Bulgaria) and that V13* is also found in Macedonia.
16
5
Genes, language and archeology
The previous chapter dealt with the question of the migration routes and regions of origins and their timing. Another question a lot of people have is in which ”culture” V13 originated and boomed, and which language the early V13-people may have spoken. For this, genetics have to be compared to archeological evidence and linguistics, but this has proven very difficult. Any group in the past would have formed a mix of different haplogroups, and only ancient DNA can give us more certainty on the subject. Unfortunately, very little has b een done in the field of ancient DNA in the Balkans, certainly for the Bronze age period that is of most interest to V13. Nevertheless, it’s clear that the common ancestor of all V13 (and CTS5856) must have belonged to a certain group, and it’s reasonable to assume that the rapid expansion and spread of V13 should be somewhat correlated to archaeological phenomena. First the link between language and archaeology will be examined to provide a solid starting point to try and fit in the genetics.
6
Linguistics: V13 and Indo-European
The Indo-European languages are a family of languages that are spoken all around the world today. But most of this distribution is due to the European empires of the past few centuries. If we would look before that, say in 1500 AD, Indo-European languages would be spoken in most of Europe, Iran and India, but not outside of that, hence the name Indo European languages. Just like a human family is descended from one ancestor, a language family also descends from one language. For the Indo-European languages this is the proto-Indo-European language (PIE). Ever since scientists realized the grouping of Indo-European languages they have been searching the answer to two big questions: where was PIE spoken, and how and when did its daughter languages come to be spread out over such an enormous area? The first question has been hotly debated, often by people who tried to prove that PIE originated in their own country. Over the past fifty years two competing theories were in vogue: the Anatolian hypothesis that claimed PIE was spoken in Anatolia (Turkey) at the start of the Neolithic (about 8000BC). It was then brought along by the first farmers from Anatolia and spread by them over all of Europe, Iran and India. This theory was favored by archaeologists, especially those who were studying the Greek prehistory, as they disliked the idea of recent migrations into Greece to explain how an Indo-European language arrived there. The other theory placed PIE in the Pontic Steppe (in Ukraine and Southern Russia) about 4000BC. This theory was mostly favored by linguists as they could show that words that were shared by all or most Indo-European languages described inventions only made about 4000 BC, like for example the cart. Furthermore, other words common in all or most Indo-European languages were coherent with the fauna and flora and the type of society that could be expected in this region. Over the last ten years genetics has been used to answer the question. Large studies have confidently shown that the advent of farming introduced a lot of new genes into Europe, coherent with a large migration. However, it also showed that most of these farmer-genes are not so abundant in todays population. This was because a second gene flow could be identified at the start of the Bronze age, this time from the Russian steppe. As such now even archaeologists agree that the Anatolian hypothesis is no longer defendable, and the PIE-homeland is firmly fixed on the Pontic steppe ca. 4000BC.27 As for the second question when and how these Steppe people succeeded in spreading both languages and genes over such a wide area is still very much an open question. One possibility that seems favored recently is that the plague was brought to Europe by steppe people and caused large scale death within native populations. What is now the link to V13? In the area where we suspect V13 to have originated (whether the Western or the Eastern Balkans), and in all the areas where it is present in high levels today, Indo-European languages were spoken in the last millennium BC. Most of those have since disappeared and have in large part been replaced by Slavic languages. (Serbo-Croatian, Bulgarian, Macedonian) that arrived in the middle ages with Slavic settlers. The two surviv27
Kristiansen et al., Re-theorising mobility and the formation of culture and language among the Corded Ware Culture in Europe, 2017.
17
ing languages are Greek, spoken mostly in the Greek peninsula, and Albanian, mostly confined to Albania and neighboring Kosovo. The disappeared languages are very scarcely attested, and their relationship to each other and to Greek and Albanian is not clear. Generally, a difference is made between Illyrian (spoken by Illyrian tribes near the Dalmatian coast), Thracian (Bulgaria) and Dacian (Romania). Greek is one of the earliest attested Indo-European languages. It was written down from about 1400BC in the Mycenaean palaces all over Greece. Soon after it also replaced the non-Indo-European language that was previously used on Crete. This Mycenaean Greek was not necessarily spoken all over Greece, it is known that even in the classical age some areas spoke a non-Greek language, which might well have been older than Greek. The Greek language is not the only Indo-European aspect we can find with the Mycenaeans. From their writings we know which gods they worshipped and at least part of them seem to come from the Indo-European pantheon. The Mycenaeans already worshipped Zeus who is identical to Zeus Pater, the principal Indo-European god. For some of the early myths and literature linguists have proposed parallels with Iranic and Indian works, and suggested they go back on a shared Indo-European heritage. 28 All of this suggests that by 1400BC Mycenaean palatial society had a very Indo-European flavor. When it arrived remains an open question. According to Anthony Indo-european migrations marked by Kurgans (tumuli or burial mounds) arrived at the mounds of the Danuba about 3100BC.29 They belonged to the so-called Yamnaya-culture that existed on the Pontic Steppe and that is generally seen as the source for most, if not all, Indo-European languages. In the following 300 years they spread upstream through the Danube valley, settling in Romania, Serbia and Hungary. The settlement in Hungary was by far the largest. So broadly we are looking at a period between 3000 and 1500BC where Indo-Europeans might have arrived in Greece. As pointed out above, V13 likely arrived in Greece before the dark ages. So it’s possible that the same processes that brought V13 to Greece also brought the Proto-Greek language and an Indo-European culture to Greece. But even though we do not know where exactly V13 originated and boomed, it seems unlikely that it was in the Pontic Steppe that was the home of Indo-European languages. So, at the very least, V13 should have originated in a region that was already Indo-Europeanised, or at least was before V13 moved to Greece. On the other hand, maybe the influx of V13 in Greece was completely independent of the process that IndoEuropeanised Greece.
28
Mallory and Adams, The Oxford introduction to Proto-Indo-European and the Proro-Indo-European world, p.423 29 D. Anthony, The horse, the wheel, the language, 2007, p. 345.
18
7
Archeological insights
As a next step, the archaeological cultures of the Balkans and Greece in the Third and second millennia BC will be examined. As far as Greece is concerned, its prehistory has been examined thoroughly and a lot of publications exist. For the Balkans the biggest problem is the lack of research. What exists is often not published in English and reliable chronologies are mostly lacking. Nevertheless, what follows is an attempt at a short summary by a layman.
7.1 7.1.1
The Eneolithic (c. 3000 - c. 2300 BC) Vucedol and the Early Helladic I to IIa
The Eneolithic, or copper age, is a period in between the Neolithic (Stone age) and the Bronze age. In Southeast Europe this period appears to last from about 3000 BC to 2300 BC. In Montenegro a number of large Tumuli have been found containing very rich graves. The burials date from the Eneolithic period (they are dated to about 3100-2500BC). They are considered princely graves, erected for elites. No comparable founds have been made in the Adriatic in such concentration. The burial in tumuli seems to point towards a Yamnaya related culture, possibly another group of the ones that settled in central Serbia and eastern Hungary that are discussed by David Anthony.30 Apart from the use of tumuli itself the deceased were buried in a flexed position facing east-west, as is typical for Yamnaya burials. It’s hard not to see a connection between these tumuli and the first arrival of R1b-Z2103 in this region. It’s also interesting that these tumuli are also located on the highland plains where the two V13* clans that were mentioned earlier would have their origins. Of course, these tumuli were erected long before these clans were first mentioned in the 15th century CE, and even their origin on these highlands cannot be firmly proven. The tumuli are at the south end of the area are thought to belong to the so called Vucedol culture. This culture was named after a Croatian site that seems to have been very important during the Eneolithic, and was right adjacent to the region in Serbia were Yamnaya Tumuli turned up in this period. So it’s no surprise that Vucedol is seen as a fusion between Yamnaya elements and local eneolithic cultures. In fact, one ancient DNA sample from a Vucedol site in Hungary turned up R1b-M269+, and it’s likely he would have belonged to the Z2103 strand. The Vucedol culture is roughly contemporary with the Early Helladic I (3100-2700BC) and IIa (2700-2400) phases in Greece. The Early Helladic I is mostly a continuation of the previous Final Neolithic phase. The Early Helladic IIa presents more of a break with the past as a stratified society starts to develop. 30
D. Anthony, The horse, the wheel, the language, 2007, p. 345.
19
Map of the Vucedol-culture, source: Wikipedia.org 7.1.2
The Post Vucedol period - EHIIb
About 2500BC the common Vucedol culture fell apart in a number of regional groups. This roughly coincides with the start of EHIIB in Greece. The stratification started in EHIIa continued with little chiefdoms ruled from so called corridor-houses. Seals and sealings prove some form of administrative organization. In this period Greece became part of larger trade links within the Mediterranean. One typical example of this is the island of Lefkas where around this time a number of tumuli appeared that cannot be clearly linked to any group before. Links to the western Balkans, Anatolia and Sardinia have all been proposed. What most scholars agree on is that Lefkas was some sort of Emporium, a stop in a long range trade network. North of Greece regional groups were scattered about. Near the heart of the previous Vucedol culture the Vinkovci group flourished, while in Serbia Bubanj Hum II I is known. In Dalmatia and its hinterland the so called Cetina group existed. The earliest Cetina pottery is dated to about 2500 BC. We are now about the time that CTS5856 would have been rapidly expanding. If this happened in the Balkans none of this presents a clue to as where this might have happened.
7.2
The early Bronze age (2300-2100 BC)
The period between about 2300 and 2100 BC was one of great changes and dynamic in the Aegean and the Balkans. The transition from EHIIb to EHIII in Greece was often catastrophic. Archeological remains show clear traces of destruction which in the past were explained as the result of some type of invasion. In the past decades, under the impulse of processual archeology, 20
this explanation has been abandonded and internal evolutions have gotten more attention. The proposed solutions are climatological catastrophes, land degradation and erosion. 31 At the same time the Balkans and the Western Aegean saw the short-lived Cetina phenomenon. In the previous phase Cetina was limited to the Dalmatian coast and its hinterland, but around 2300 the Cetina pottery and other artefacts associated with this culture start rapidly expanding and are found all along the Adriatic and Ionian coasts, from Istria to Malta and the Peloponnese.32 Also associated with the Cetina culture are burials in Tumuli. In many places settlements or burials associated with this Cetina material is seen as alien and intrusive, so there seems to have been not only a rapidly booming trade network, but also some movement of people. The spread of this seafaring population inspired Joseph Maran to dub them Argonauts of the West Balkans.33 The network was also present in Greece, where tumuli were found in Olympia and Loutraki. Olympia was also a main source of Cetina pottery finds, probably making it something of a trade post. Typologically these tumuli closely resemble the widespread tumuli of the following Middle Helladic period (Ayios Ioannis-Papoulia,Vodokilia and Kastroulia in Messenia, Aphidna and Vrana I in Attica, Argos I and possibly Asine in Argolis, and Drachmani in central Greece). 34 Remarkable in Greece in this period is the use of so called Ritual mounds in these periods, for example in Lerna. In Thebes a group of people were buried on a destruction level and covered by a large mound.35 Although this would seem to suggest external influences specialists of the Aegean Bronze Age are adamant that they were raised by locals. 36 Another very prominent location of the Cetina phenomenon is Albania. Either during the Cetina phenomenon or immediately following it (the chronology isn’t very clear) a large number of Cetina related material and tumuli appear all over Albania. 37 This phenomenon would still expand throughout most of the second millennium BC. Apulia (Puglia) is another region featuring Cetina related material. 38 For Friuli the presence of Cetina has been proven, its link to the big fortress of Monkodonja, and its presence in Veneto have been proposed but may be more speculative.39 About 2150BC the Cetina phenomenon seems to abruptly stop again, at the start of the so called Middle Bronze age. This middle bronze age presented a period of a few centuries with depopulation, little trade and no organized society. Recently a connection between the onset of this period, the disruption of the Cetina tradenetwork and the bad climatic conditions around 2200-2100BC (the 4.2kya event) have been suggested. 40 More or less contemporary with the Cetina phenomenon two small, probably related PostVucedol groups were present on the Central Balkans: Bubanj-Hum III in eastern Serbia and the West of Bulgaria, and Armenochori in Macedonia. 31
S. Voutsaki, Mainland Greece, The Oxford handbook of Aegean Bronze age, p. 137. J. Maran, Seaborne contacts between the Aegean, the Balkans and the central Mediterrenean in the 3d millenium BC, 2007. 33 J. Maran, Seaborne contacts between the Aegean, the Balkans and the central Mediterrenean in the 3d millenium BC, 2007. 34 S. Mller-Celka, Burial Mounds and Ritual Tumuli of the Aegean Early Bronze, 2011. 35 S. Mller-Celka, Burial Mounds and Ritual Tumuli of the Aegean Early Bronze, 2011. 36 J. Forsn, Mainland Greece, The Oxford handbook of Aegean Bronze age, p. 97. 37 S. Oikonomidis et al., The emergence and the architectural development of the Tumulus Burial Custom in NW Greece (Epirus and the Ionian Islands) and Albania and its connections to settlement organization, 2011. 38 G. Recchia , Burial mounds and ”Specchie” in Apulia during the Bronze Age, 2011. 39 Elisabetta Borgna and Paola Cssola Guida, Seafarers and Land-Travellers in the Bronze Age of the Northern Adriatic, 2009. 40 Reinhard Jung and Bernhard Weninger, Archaeological and environmental impact of the 4.2 ka cal BP event in the central and eastern Mediterranean, 2014. D. Pullen, The Early Bronze Age in Greece, 2009, p. 36. 32
21
This map represents Early Bronze age cultures in the Balkans. The purple shaded area is the core area of the Cetina-culture, while red dots represent places that are associated with the Cetina-phenomenon (this might just imply trading contacts). The green shaded area is the Armenochori culture, and the orange shaded is Bubanj Hum III.
8
Balkan genetics
Previously the distribution of V13 was discussed but here there will be some broader info about other haplogroups that are common in the Balkans, as they will be necessary to understand the whole picture. If we look at the distribution of some common groups in the Balkans we see that V13 and R1b are amongst the highest. Important is to remark that we know from more detailed tests that practically all R1b here (maybe with the exception of Croatia) is R1b-Z2103 and not the R1b-L51 that is very common in Western-Europe.
Source: www.eupedia.com Almost all R1a, I1 and I2b in this table is the result of Slavic migrations during the early Middle Ages. We can remove these groups to get a very rough idea how te distribution was before the Slavic migrations:
22
Source: www.eupedia.com We see that V13 (as principal part of E1b1b) is very high over all of the Balkans, and that that the previous peak in Albania isn’t that big anymore, but more due to less impact of later Slavic migrations. Most of the other groups are very small, and are probably remnants of Neolithic populations (groups like C, F,G, J1, E outside of V13), which seems supported by the fact that these groups are generally more present in Greece than in the rest of the Balkans. Haplogroup G seems somewhat higher ( 5%) in Serbia than in most parts of the Balkans (Macedonians are also fairly high). G-L42 is most typical for Balkans apart from Greece and might represent a Celtic migration into the Balkans in the last millennium BC, as it seems to originate from Switzerland. The other two groups outside of V13 that have a high pre-Slavic presence are R1b and J2. J2 is made up mostly of Neolithic branches, but there is one exception: J2-M241. Although rare in most of the Balkans, it reaches about 15% in Albania. This might not just be a founder effect, as J2 appears about as high in the rest of the Balkans when the recent Slavic groups are removed. Samples from Albania and other parts of the Balkans all belong to the subclade Z2507 which is about 4400 years old according to Yfull. Older subclades of M241 are mostly limited to the Western-Mediterrenean, and absent from the Balkans. This seems to point to a mobility in the Western Mediterrenean that might be linked to the spread of Megalithic cultures like the Arzachena-culture on Sardinia. 41 From the fourth to the third millenium BC there is a long bottleneck in this group, followed by a renewed expansion underneath Z597. It appears that this expansion happened somewhere around the Thyrrenian sea, but groups quickly spread out to the Eastern-Mediterrenean, while keeping contact with their homeland. This expansion might be linked to the ’international spirit’ of about 2500-2300 BC, as earlier it was shown that contacts between Sardinia and the Ionian coast were suggested in the second half of the third millennium BC. So it’s conceivable that M241 arrived in the Balkans in this period. But just like with V13 it’s hard to pinpoint an exact timeframe and their arrival could also be in a later period. In the Mycenaean period there were certainly contacts between Greece and the western Mediterranean. It seems that the Mycenaeans had taken over the previous trading network of the Minoans after the decline of this Cretan civilization. By the end of the Mycenaeen period sources speak about raiders and one of these groups are called ’Sharden’, what some have explained as ’Sardinians’. If the Sardinians were competing with the Mycenaeans economically and military, the latter probably in the form of raids and piracy in a Viking-style, an operational base close to Greece but far enough so that Greek land troops or ships could not easily reach it would be ideal. When sailing along the coast of Italy the Sardinians would find the shortest sea route to the Greek coast by crossing the strait between the heel of Italy and Albania. From there they could control lucrative shipping routes to Italy and the Adriatic coast and sail along the Greek west-Coast to launch raids on Mycenaean settlements. Both their disruption of Mycenaean trade as well as their raids might have contributed to the rapid decline of Mycenaean civilization. Of course, once trade disappeared and societies declined the Sardinians would also have dwindled. R1b xM269 is rare in most of the Balkans. On the other hand M269 xL23 is more common, especially near Albania, this is R1b-PF7563, although quite removed from the R1b found in the Yamnaya the most likely explanation seems it was part of Steppe people invading in the Balkans. One possibility is that it’s the result of a later flow, for example through Cimmerians. R1b-L23 41
https://en.wikipedia.org/wiki/Arzachena culture
23
xL51 (or R1b-Z2103) is another group that is very important in the Balkans. As explained above this group was found in genetic tests of Yamnayab burials in the Pontic steppe. So it seems acceptable that the flow that brought R1B-Z2103 to the western Balkans is connected to the Yamnaya-like tumuli that pop up in Serbia and Montenegro at about 3000 BC. Most likely this flow introduced Indo-European languages and culture to the Western-Balkans.
9
Puzzling it all together
We have very little clues for the origin of V13. In contrast to a lot of other haplogroups we have little info on basal sidebranches, as there is a period of a few thousands of years between V13 and its most recent ancestor, L618. This ancestor was found in Lengyel, Hungary, while a V13 subbranch was found in the Mirabol cave in Spain. This already indicates that L618/V13branches, while no doubt a very minor clade, were present over a large area during the Neolithic. It’s anybody’s guess from where the V13 came that eventually would boom in the Southern Balkans. Nevertheless it seems likely that the first boom of V13 already happened in or near the Montenegrin-Albanian-Serbian border area. This probably happened between 3500 and 3000BC. The next, bigger boom underneath CTS5856 happened about 700-1000 years later, likely in the second half of the third millennium BC. When comparing to the archeological data the best fit for this second boom might be the Cetina phenomenon. If there really was an intensive trade network in the Adriatic and beyond this must have been a period of economic and demographic growth. People were needed to produce and consume these goods, but evidence also points to at least some migrations. If V13 was caught up in these migrations it provides an elegant solution for the sub V13 branches that were not found in the Balkans and sometimes have very peculiar distributions, like L17. They might have migrated towards Italy or the Northern Adriatic coast or even mutated in migrated communities there. That might also help understand why V13 is relatively high all over Italy, and very present on the Italian-Swiss border. They might have been what Maran called ”the argonauts of the West-Balkans”. The distribution of V13* as shown above seemed to have a Western Balkan focus that agreed well with the core area of the Cetina culture near the Dalmatic coast. The V13* in Italy also seemed to be a good fit for the distribution of Cetina-ware, althought it also may be from a much later date. Still, it must be kept in mind that this does not offer a good explanation for the high levels of V13 in Bulgaria, and the high diversity there. Moreover Bulgaria was closer to the Indo-European homeland and seems like a logic pathway for languages and cultures to flow to Greece and the western Balkans. However there does not seem to be a process like the Cetina phenomenon to explain such a flow. A possible explanation might be that CTS5856 boomed a little earlier and was present in different Post-Vucedol cultures, but the current distribution does not suggest such a split in V13-subclades. More testing data might clear up this point. Although specialists of the Greek Bronze age dismiss any possibility of migration into Greece at this period, it seems likely that there was also a movement of Cetina people to the south. It has been suggested that the Cetina-network competed with a Eastern-Aegean network, and that this competition was the reason for the destructions seen at the end of EHII. 42 It seems very likely that in the following EHIII period there was some presence of Cetina related people in Olympia and in places along the coast of the Ionian Sea. At about 2200-2150 BC the whole Mediterranean seems to have suffered a severe crisis, likely linked to very bad climatic conditions. Although society crumbled and population shrinked, the Cetina-related tumuli seem to multiply in Greece, spreading to all the areas were the later Mycenaean civilization would flourish. It seems likely that in this period V13 spread out all over Greece. Afterwards V13 was spread around the Mediterrenean by Greek colonists, and later on by Roman soldiers, both from Italy and the Balkans, leading to its current pan-European distribution. 42
James Clinton Wright, Early Mycenaean Greece, The Cambridge Companion to the Aegean Bronze Age, 2008, p.232.
24
It’s not illogical to connect the influx of Cetina-related people into Greece with the protoGreek language and Indo-European-culture. The Cetina-people probably had an Indo-European culture and language that was brought to the region by an earlier migration from the steppe, that also brought the R1b-PF7562 and R1b-Z2103 haplogroups to the Western Balkans. Maybe a third input to the Cetina cuture came from the Western Mediterennean, and the trading and seafaring aspect of the Cetina culture might have come from Sardinia through people that also brought haplogroup J-M241. As such the Western Balkans would have been a melting pot of different genetic and cultural inputs.
10
Further research
The biggest missing puzzle piece is without a doubt ancient DNA. While hundreds of samples have been analyzed from other parts of Europe, the Balkans and Greece have seen little results except for the Neolithic period. This is surprising since Ancient DNA might largely settle some of the big debates about the history of the regions, like the ”coming of the Greeks”. Ancient DNA would also allow to better pinpoint the region of origin of V13 and whether there is a link with the Cetina phenomenon. It is very well possible that it will lead to new insights that will totally disprove the theories put forward above. Apart from ancient DNA research where nothing seems to be forthcoming the research on current DNA continues. NGS-testing is steadily fleshing out the tree of subbranches underneath V13 while SNP-packs allow those with less financial means to profit of the new discoveries of NGS-testing. Every test, whether NGS or SNP-pack, adds to the knowledge about the distribution of the different subclades of V13. Without ancient DNA it’s risky to try and align subclades to different regions or cultures, but hopefully more and more results will allow the big picture to emerge.
25
