European Journal of Archaeology 21 (1) 2018, 3–17

Discussion: Are the Origins of

Indo-European Languages Explained by
the Migration of the Yamnaya Culture to
the West?

LEO S. KLEJN1, WOLFGANG HAAK2, IOSIF LAZARIDIS3, NICK PATTERSON3,

DAVID REICH3, KRISTIAN KRISTIANSEN4, KARL-GÖRAN SJÖGREN4,
MORTEN ALLENTOFT5, MARTIN SIKORA5 AND ESKE WILLERSLEV5
1
Saint Petersburg University, Russia
2
Max Planck Institute for the Science of Human History, Iena, Germany
3
Department of Genetics, Harvard Medical School, USA
4
Department of Historical Studies, University of Gothenburg, Sweden
5
Centre for Geogenetics, Natural History Museum of Denmark, Copenhagen, Denmark

Two co-authored articles in Nature (Haak et al., 2015; Allentoft et al., 2015) caused a sensation.
They revealed genetically the mass migration of steppe Yamnaya culture people in the Early Bronze Age
to central and northern Europe. The authors considered this event as the basis of the spread of Indo-
European languages. In response, the Russian archaeologist, Leo S. Klejn, expresses critical remarks on
the genetic inference, and in particular its implications for the problem of the origins of Indo-European
languages. These remarks were shown to the authors and they present their objections. Klejn, however,
has come to the conclusion that the authors’ objections do not assuage his doubts. He analyses these objec-
tions in a further response.

Keywords: Indo-European languages, migration, Yamnaya culture, Bronze Age

DO THE INDO-EUROPEAN LANGUAGES OF ancient DNA from archaeological sites

EUROPE STEM FROM THE STEPPE PEOPLE and inferred that massive migration at the
OF THE YAMNAYA CULTURE?1 transition of the Neolithic to the Bronze
L.S. Klejn Age brought Indo-European languages
from the eastern European steppes to the
Attainments and doubt rest of Europe. The steppe Yamnaya (Pit-
Grave) culture appeared to be associated
Two studies published in Nature (Haak with the Proto-Indo-European language,
et al., 2015; Allentoft et al., 2015) analysed while the origin of the derivative language
groups (Greek, Germanic, Slavic, Celtic,
1 among others) was ascribed to the Corded
Originally my text was much longer, but to make it
publishable, it was substantially abridged. I have not
wanted to restore the original text since it was presented
to my opponents in the abridged form and their objec- my argument, which may facilitate the reading. On the
tions were made in response to this abridged version. recommendation of a peer reviewer, I also added a new
Yet I have seen fit to reinstate some minor portions of section on the characteristics of the Yamnaya and
the original text and some references that do not change Corded Ware cultures.

© European Association of Archaeologists 2017 doi:10.1017/eaa.2017.35

Manuscript received 23 October 2016,
accepted 24 April 2017, revised 19 March 2017
Downloaded from https://www.cambridge.org/core. University of New England, on 03 Feb 2018 at 20:33:03, subject to the Cambridge Core terms of use, available
at https://www.cambridge.org/core/terms. https://doi.org/10.1017/eaa.2017.35
 4 European Journal of Archaeology 21 (1) 2018

Ware cultures. This inference coincides since I derived from them the Donets
with the steppe hypothesis of the origins of Catacomb culture (more recent than
the Indo-Europeans. There are, however, Yamnaya). These are cultures of a quite
several rival hypotheses on this subject. different kind. Some Corded Ware cul-
The authors of these articles are aware of tures have barrows, some not. Men are
possible criticisms and often qualify their laid to rest in positions that differ from
inferences, e.g. perhaps not all Indo- those of women. The pottery is completely
European peoples stem from the Yamnaya different—the main types are amphorae
but only some of them. But if it is only and beakers—and the most popular
some of them, then it is not the cradle of weapon is the stone battle hammer-axe.
Proto-Indo-European, only one of its sub- These cultures are widely distributed in
families. If this were the case, the steppe the forest and meadow zone, but do not
hypothesis of the origin of Indo-Europeans extend onto the steppe.
is transformed into the steppe origin of, I doubt that the Yamnaya people spoke
say, Indo-Iranians. That scenario is quite the Proto-Indo-European language. In
realistic, but it would be strange to suppose what follows, I discuss the problem with
their spread over northern Europe; further- reference to the latest archaeological data
more, Indo-Iranian has nothing to do with and suggest how we can try to resolve the
most European languages! controversy using both genetic and arch-
In common with many other archaeolo- aeological findings.
gists, I doubt that the discoveries in ques-
tion reflect a direct migration from the
Yamnaya to the Corded Ware cultures. Breakup of the Indo-European proto-
Let me explain some of their differences. language
I have excavated on several occasions
steppe barrows containing burials of the On the basis of glottochronology and cladis-
Yamnaya culture. I know these burials well: tics, all the acknowledged dates for the
they are usually primary in the barrow, breakup of the Proto-Indo-European lan-
sometimes secondary, with strong skeletons guages are located within the seventh to fifth
laid on their backs with raised knees, and millennia BC (Figure 1): the median date
they are densely covered with red ochre. according to Gray and Atkinson (2003) is
Men and women are buried in the same 6700 BC; for Bouckaert and colleagues (2012/
way. The graves usually contain handmade 2013) it is 5500 BC; for Swadesh (1955) it
ceramics without handles, i.e. small, round- takes place before 5000 BC; for Dybo and
bottomed, egg-shaped vessels, sometimes Kassian (Kushniarevich et al., 2015) it is
with a corded decoration. They also often 5000 BC; for Starostin (2007) it lies around
contain hammer-headed pins made of 4670 BC; and for Chang and colleagues
horn, and occasionally stone shaft-hole (2015) it is 4500 BC. (Even if we expand
hammer-axes, and bronze lance-shaped the range by adding the confidence
knives and awls. This culture is widely dis- intervals, this cannot rejuvenate dates
tributed on the steppe. On the western significantly. The millennial gap remains.)
steppe, the Yamnaya is of mixed character,
retaining its own distinctive way of inter-
ment in barrows but adopting different, Disappearance of the Yamnaya culture
local ceramics.
I was also much involved in studying The Yamnaya culture is now well dated by
the Corded Ware and Battle-Axe cultures calibrated radiocarbon chronology (Chernykh

Downloaded from https://www.cambridge.org/core. University of New England, on 03 Feb 2018 at 20:33:03, subject to the Cambridge Core terms of use, available
at https://www.cambridge.org/core/terms. https://doi.org/10.1017/eaa.2017.35
 Klejn et al. – Are Origins of Indo-European Languages Explained by Migration of the Yamnaya Culture? 5

Figure 1. The breakup of the Proto-Indo-European language (in various datings) and the branching
off of the Yamnaya culture.

& Orlovskaya, 2004a; Nikolova, 2012). It The Poltavka culture begins in c. 2600
begins at the very earliest around the thirtieth BC (Morgunova, 2013), the Catacomb cul-
century BC, but most of its dates fall within tures also begin around 2600 BC, but, if
the second third of the third millennium BC. we broaden the confidence interval to two
Likewise, Heyd (2011) dates Yamnaya in sigmas, then the earliest limit will be
the west (in the area of Hungary) to the first around 2900 BC (Chernykh & Orlovskaya,
half of the third millennium BC. 2004b). This is still very far even from the
One must also take into account that nearest possible date of the language
the derivative cultures (and their derivative breakup around 4500 BC.
languages) can branch off from their Thus, there is gap of about 2.5 millen-
matrix no earlier than this time and nia (1.6 millennia at the very least)
perhaps much later. Furthermore, we must between the breakup of the language and
correlate the breakup of the required that of the culture (Figure 1). We must
proto-language not with the beginning of a add that the breakup of the language will
culture (say, the Yamnaya) but with its not immediately follow the fragmentation
disintegration, the end of its existence, its of a society. So the gap will be even wider.
replacement by new cultures formed on The inference is clear. The language
the basis of Yamnaya—those that might spoken by the Yamnaya people cannot have
derive from it. By this criterion, the been Proto-Indo-European: the temporal
Poltavka culture, Catacomb grave cultures, gap between the breakup of the Proto-
and others might be derivative. Indo-European language and the

Downloaded from https://www.cambridge.org/core. University of New England, on 03 Feb 2018 at 20:33:03, subject to the Cambridge Core terms of use, available
at https://www.cambridge.org/core/terms. https://doi.org/10.1017/eaa.2017.35
 6 European Journal of Archaeology 21 (1) 2018

Figure 2. Distribution of the ‘Yamnaya’ genetic component in the populations of Europe (data taken
from Haak et al., 2015). The intensity of the colour corresponds to the contribution of this component
in various modern populations. The scale of intervals is to the right. The purple line represents the
borders of the Yamnaya area. The brown arrow shows the direction of migration postulated by the
proponents of a Yamnaya origin for the Indo-Europeans of Europe. The red arrows show the direction
of the movement of the ‘Yamnaya’ component in accordance with the gradient shown on this distribu-
tion. The map shows that the ‘Yamnaya’ genetic component is hardly Yamnaya in origin; rather it is a
more ancient component originating in the populations of northern Europe from whence it spread both
to the steppes and to the cultures of central Europe and elsewhere. Map by O.P. Balanovsky.

disintegration of the Yamnaya culture is similar cultures must be significantly

too wide. earlier than the other. The geneticists in
question and the archaeologists on their
teams argue that Yamnaya is much earlier
The beginning of Corded Ware cultures than the spread of the Corded Ware cul-
tures across central, northern, and eastern
The idea of migration from the steppes to Europe (including eastern European
the rest of Europe implies that one of the Russia). This was the situation imagined

Downloaded from https://www.cambridge.org/core. University of New England, on 03 Feb 2018 at 20:33:03, subject to the Cambridge Core terms of use, available
at https://www.cambridge.org/core/terms. https://doi.org/10.1017/eaa.2017.35
 Klejn et al. – Are Origins of Indo-European Languages Explained by Migration of the Yamnaya Culture? 7

New ideas about the beginning of the

Yamnaya culture

Over the last decades, suggestions have been

made that the beginning of the Yamnaya
culture (Ivanova, 2004; Morgunova, 2013,
among others) is earlier than previously
thought. Ivanova (2004), a specialist of the
Yamnaya culture, accepts all these sugges-
tions, but has observed a remarkable thing:
the main bulk of the Yamnaya culture (in
Figure 3. Ethnocultural situation in central and the Ukraine) has no such early dates and the
eastern Europe in the Late Mesolithic and Early earlier dates are concentrated at the very
Neolithic (sixth to fifth millennia BC). 1:
Maglemose culture area in the seventh millen-
western and the very eastern edges of this
nium BC (after G. Clark). 2–7: Mesolithic area. Yet, the origin of the culture cannot
cultures of the sixth millennium BC of the post- have spread from both edges to the centre!
Maglemose cultural tradition (after Kozlovsky One of the two concentrations must be
and Zaliznyak). 2: de Leijen-Wartena. 3: false; but if one is false, the other also
Oldensloe-Gudenaa. 4: Chojnice-Pieńki. 5:
becomes doubtful.
Janicsl avice. 6: finds of Janisl avice artefacts
=

beyond its main distribution area. 7: Donets

culture. 8: directions of the Janisl avice culture
=

settlement (after Kozlovsky and Zaliznyak). 9: A source in the north?

southern border of Mesolithic and Early Neolithic
cultures of post-Swidrian and post-Arensburgian I have already stressed (Klejn, 2015a,
traditions. 10: northern border of the settlement
area of Balkan-Danubian farmers (late sixth to 2015b, 2015c) a remarkable fact: the
early fifth millennium BC). 11: Bug-Dnestr strange distribution of the genetic ‘steppe’
culture. 12: Neolithic cultures formed on a post- contributions to the Corded Ware cultures
Maglemose ethnocultural basis. E: Ertebølle- and their descendants revealed in Haak
Ellerbeck. D: Dnieper-Donets culture. M: et al. (2015): very rich in the north of
Mariupol culture (western variants). N: Niemen
culture. After Koncha, 2004 (from ideas of L. Europe and increasingly weaker towards
L. Zalisnyak), redesigned by P. Deyneka. the south, in Hungary, just where the
western edge of the Yamnaya culture itself
is located. This distribution is at odds
with the suggestion that the source of the
contribution to the Corded Ware cultures
in the twentieth century (by advocates of is the Yamnaya culture in the south-east;
the steppe hypothesis, including but the same distribution appears quite
Gimbutas, Mallory, and Anthony). Now, natural if one suggests that the common
the many radiocarbon dates at our disposal source (of both cultural units) is located in
have changed this notion. The Yamnaya the north of Europe—hence the common
dates have been discussed above. Let us cause of the genetic similarity (Figure 2).
now look at the dates of the Corded Ware The respected Ukrainian archaeologist
cultures: all begin in the early third millen- L.L. Zaliznyak (2005) has proposed to
nium BC—immediately after 3000 BC shift the date of the Proto-Indo-European
(Stöckli, 2002; Furholt, 2003a, 2003b). language to the Late Mesolithic–Early
They are practically contemporary with the Neolithic (sixth–fifth millennia BC) so as to
Yamnaya culture. better correspond to the glotto-

Downloaded from https://www.cambridge.org/core. University of New England, on 03 Feb 2018 at 20:33:03, subject to the Cambridge Core terms of use, available
at https://www.cambridge.org/core/terms. https://doi.org/10.1017/eaa.2017.35
 8 European Journal of Archaeology 21 (1) 2018

chronological datings mentioned above. the dates for Proto-Indo-European based

On archaeological and physical anthropo- on rates of language change. As Klejn
logical grounds, he reconstructs the move- points out, all estimates of the divergence
ment of Maglemose and post-Maglemose of Anatolian (the most deeply divergent of
(Ertebølle–Ellerbeck) populations of Indo-European languages) and other
Jutland and the south-eastern Baltic region Indo-European languages based on rates
to the Dnieper and Donets areas of language change are at least ∼1500
(Figure 3). Consequently, a large area from years older than the archaeologically
the Baltic to the Dnieper was settled by attested date of the Yamnaya, although
people with similar archaeological, physical such dates are known to be highly
anthropological, and supposedly genetic dependent on the modelling assumptions
characteristics—the cradle of Indo- that go into them (Chang et al., 2015).
European peoples, according to Zaliznyak. The main problem with Klejn’s critique,
My notes have no intention of detract- however, is that when we refer to the lan-
ing from the brilliant success of the guage spoken by the Yamnaya, we are
geneticists. I merely try to correct their explicitly not referring to Proto-Indo-
interpretation of it. European, but instead to a continuum of
languages that is likely to have existed in
the western Eurasian steppes well after the
ACKNOWLEDGEMENTS splitting off of the Anatolian branch. As
we write in our paper, ‘the location of the
This article was read in manuscript form proto-Indo-European homeland that also
by my colleagues E.V. Balanovskaya, O.P. gave rise to the Indo-European languages
Balanovsky, A.V. Dybo, A.G. Kozintsev, of Asia, as well as the Indo-European lan-
and A.A. Kovalev, and they have given me guages of southeastern Europe, cannot be
valuable advice. For corrections to my determined from the data reported here’
English I am thankful to S.D. Leach. (Haak et al., 2015: 211).
Klejn mischaracterizes our paper as
claiming that practitioners of the Corded
Ware culture spoke a language ancestral to
RESPONSE L.S. KLEJN’S CRITIQUE OF
TO all European Indo-European languages,
OUR ARTICLE including Greek and Celtic. This is incor-
Wolfganf Haak, Iosif Lazaridis, rect: we never claim that the ancestor of
Nick Patterson, and David Reich Greek is the language spoken by people of
the Corded Ware culture. In fact, we expli-
L.S. Klejn’s critique of our article, ‘Massive citly state that the expansion of steppe
migration from the steppe was a source for ancestry might account for only a subset of
Indo-European languages in Europe’ (Haak Indo-European languages in Europe.
et al., 2015), is based on misunderstanding Klejn asserts that ‘a source in the north’
of our claims. Klejn mischaracterizes our is a better candidate for the new ancestry
article, suggesting we were stating that the manifested in the Corded Ware than the
practitioners of the Yamnaya culture were Yamnaya. While it is indeed the case that
speakers of ‘Proto-Indo-European’: the lan- the present-day people with the greatest
guage spoken prior to the breakup of affinity to the Corded Ware are distribu-
known Indo-European languages. ted in north-eastern Europe, a major part
If this was our claim, there would of the new ancestry of the Corded Ware
indeed be some tension with estimates of derives from a population most closely

Downloaded from https://www.cambridge.org/core. University of New England, on 03 Feb 2018 at 20:33:03, subject to the Cambridge Core terms of use, available
at https://www.cambridge.org/core/terms. https://doi.org/10.1017/eaa.2017.35
 Klejn et al. – Are Origins of Indo-European Languages Explained by Migration of the Yamnaya Culture? 9

related to Armenians (Haak et al., 2015) archaeological, linguistic, and genetic data
and hunter-gatherers from the Caucasus speak against such a hypothesis.
(Jones et al., 2015). This ancestry has not
been detected in any European hunter- Absolute dating of the Yamnaya and
gatherers analysed to date (Lazaridis et al., Corded Ware cultures
2014; Skoglund et al., 2014; Haak et al.,
2015; Fu et al., 2016), but made up some The Yamnaya culture starts around 3000
fifty per cent of the ancestry of the BC (online archaeological text to Allentoft
Yamnaya. The fact that the Corded Ware et al., 2015), and Corded Ware no earlier
traced some of its ancestry to the southern than 2800 BC. The majority of dates
Caucasus makes a source in the north less cluster around 2600 BC, when populations
parsimonious. had consolidated. Klejn’s arguments on
In our study, we did not speculate about this issue are incorrect.
the date of Proto-Indo-European and the
locations of its speakers, as these questions
are unresolved by our data, although we Proto-Indo-European split
do think the genetic data impose con-
straints on what occurred. We are enthusi- The dates produced by the promising, yet
astic about the potential of genetics to still highly experimental, phylogenetic
contribute to a resolution of this long- methods are so variable, ranging from
standing issue, but this is likely to require 6700 BC to 4500 BC, that at this point
DNA from multiple, as yet unsampled, they cannot be considered reliable.
ancient populations. The reconstructed Proto-Indo-European
vocabulary concerning weaving, wool pro-
duction, horse breeding, and wagon tech-
nology is incompatible with dates earlier
ARCHAEOLOGY, LANGUAGE, AND than the fourth millennium BC.
GENETICS: YAMNAYA AND CORDED
WARE. A RESPONSE TO LEO KLEJN
Kristian Kristiansen, Karl- Scandinavia as a source for Yamnaya
Göran Sjögren, Morten Allentoft, genetic dominance
Martin Sikora, and Eske Willerslev
Klejn’s argument is based on contemporary
Leo Klejn proposes an alternative scenario patterns of genetic variation, which are often
for the archaeological and genetic formation poor predictors of their distributions in
of the Corded Ware culture in northern the past. Available aDNA data demonstrate
Europe, and subsequently the spread of that neither Mesolithic nor Neolithic
Indo-European languages. He wishes to see Scandinavians have Yamnaya affinity
an early origin in Late Mesolithic cultures (Lazaridis et al., 2014; Skoglund et al., 2014).
of Scandinavia, who migrated south to the
Dnieper and Donets areas, where they later
turned into Yamnaya. To promote such a The Yamnaya migration and its
hypothesis, he attempts in points 1–4 to archaeological and genetic impact in
undermine the existing chronology of the northern and central Europe
Yamnaya and Corded Ware cultures, as well
as the date for the breakup of the Indo- It has recently been demonstrated that two
European proto-language. However, male burials of the earliest Corded Ware

Downloaded from https://www.cambridge.org/core. University of New England, on 03 Feb 2018 at 20:33:03, subject to the Cambridge Core terms of use, available
at https://www.cambridge.org/core/terms. https://doi.org/10.1017/eaa.2017.35
 10 European Journal of Archaeology 21 (1) 2018

culture in Poland and southern Germany divergence in approaches by geneticists

were non-locals and had a hammer-headed and several archaeologists and linguists.
pin of steppe type as grave goods
(Pospieszny et al., 2015; Sjögren et al., The origins, spread, and divergence of
2016). In another article (Rasmussen et al., Indo-European languages
2015) it could be demonstrated that an
early form of plague occurred in Yamnaya First of all, both responses prefer to reduce
and Corded Ware populations, which could the main controversy to the secondary
account in part for the strong decline of the question of whether all Indo-European
so-called Neolithic DNA. This, however, languages have stemmed from the language
needs to be tested on Neolithic samples. of the migrated Yamnaya culture, or not
all of them but only the Indo-European
languages of Europe, or even not all of
those. Allegedly, I had accused the authors
Conclusion
of tracing all Indo-European languages
back to Yamnaya, whereas they did not
We do not exclude the possibility that
trace all of them but only a portion!
migrations preceding the Yamnaya expan-
Well, I shall not reproach the authors
sion into northern temperate Europe could
for their ambiguous language: it remains
have brought steppe DNA, since a good
the case that (beginning with the title of
genetic coverage of Late Neolithic cultures
the first article) their qualifications are lost
in this region, such as the Globular
and their readers have understood them as
Amphora culture, is not yet available.
presenting the solution to the whole ques-
However, we contend that in our present
tion of the origins of Indo-European
state of knowledge the models presented in
languages.
the two articles (Allentoft et al., 2015;
Another factor is more important. The
Haak et al., 2015) are the more likely scen-
authors of the first response specify they
arios for what happened during a few dra-
had in view not the Proto-Indo-
matic centuries in the first half of the third
European before the separation of the
millennium BC in western Eurasia. We do
Hittites, but the language that was left
not claim to have found a definite origin of
after the separation. Yet, this was still the
Indo-European languages. But we have
language ancestral to all the remaining
gathered genetic evidence to document a
Indo-European languages, and the fol-
substantial human migration occurring
lowers of Sturtevan and Kluckhorst call
shortly after 3000 BC showing remarkable
only this language Proto-Indo-European
similarities with one of the models proposed
(while they call the initial one Indo-
for the spread of Indo-European languages.
Hittite). The majority of linguists (specia-
lists in Indo-European languages) is now
IS YAMNAYA MIGRATION TO CENTRAL
A inclined to this view. True, the breakup
AND NORTHERN EUROPE GENETICALLY of this younger language is several
CONFIRMED AND DOES THIS EXPLAIN hundred years more recent (nearly a thou-
THE ORIGINS OF INDO-EUROPEAN sand years later according to some glotto-
LANGUAGES? chronologies) than the separation of
L.S. Klejn Anatolian languages, but it is still around
a thousand years earlier than the birth of
The two responses have much in common cultures derived from Yamnaya (see my
but they also reveal the extent of the Figure 1).

Downloaded from https://www.cambridge.org/core. University of New England, on 03 Feb 2018 at 20:33:03, subject to the Cambridge Core terms of use, available
at https://www.cambridge.org/core/terms. https://doi.org/10.1017/eaa.2017.35
 Klejn et al. – Are Origins of Indo-European Languages Explained by Migration of the Yamnaya Culture? 11

More than that, I analysed in my criti- Dating the breakup of the Indo-
cism both possibilities—the case for all European proto-language
Indo-European languages spreading from
Yamnaya and the case for only some of My opponents’ first objection appears to
them spreading from Yamnaya. In the clearly contradict their second objection.
latter case, it is argued that only the lan- The latter, made by both sets of authors,
guages of the steppes, the Aryan (Indo- attempts to address my point that there is
Iranian) are descended from Yamnaya, not a gap of more than a thousand years
the languages of northern Europe. between the breakup of the Indo-European
Together with many scholars, I am in proto-language and the birth of cultures
agreement with the last possibility. But, derived from Yamnaya. In both responses,
then, what sense can the proposed migra- the authors draw attention to the state of
tion of the Yamnaya culture to the Baltic linguistic conclusions: they point out that
region have? It would bring the Indo- the dates of the breakup of the Indo-
Iranian proto-language to that region! Yet, European proto-language are not facts but
there are no traces of this language on the hypotheses based on a certain model—
coasts of the Baltic! here the shaky foundations of glottochron-
However, the Yamnaya culture, or ology are referred to. As if the method of
rather its western variety, has many con- admixture used by geneticists is not itself
tacts in the Middle Danube basin (not in based on some model!
the north of Europe), where mixed assem- Yes, glottochronology is not exact—
blages are found with Yamnaya burial rites neither is radiocarbon dating—but, within
and local ceramics (probably representing certain parameters, it is reliable enough,
marriages of steppe men with local and these bounds of tolerance are deter-
women). This may be reflected in the mined not by the arbitrary will of some
long-recognized distribution of Indo- scholars (which may be infinite) but by
Iranian terms relating to power, religion, some commonly recognized factors of
and war in the Italo-Celtic languages: raj/ uncertainty.
rex, reg (king), upasti/foss (servant), asi/ One of the two responses mentions the
ensis (sword), etc. (Vendryès, 1917; connection of the later date (fourth mil-
Koncha, 2005). lennium BC) with the presence in the
I am afraid that in my critical article my Proto-Indo-European glossary of general
main concern about geneticists’ treatment terms for weaving, horse breeding,
of languages was not made plain. My main wagons, etc. Yet, this connection is not
concern is that, to my mind, one should absolute, for borrowed terms might by
not directly apply conclusions from genet- analogy take forms developed in this lan-
ics to events in the development of lan- guage (i.e. a word spread by borrowing
guage because there is no direct and could look outwardly similar to forms
inevitable dependence between events in derived from the proto-language). The
the life of languages, culture, and physical whole sixth chapter of the second part of
structure (both anthropological and Koncha’s (2017) work Indoevropeytsi is
genetic). They can coincide, but often they devoted to the substantiation of this
all follow divergent paths. In each case the aspect. However, even if we take the most
supposed coincidence should be proved recent date for the breakup of the Indo-
separately. European proto-language, the thousand-
year gap between this breakup (even if

Downloaded from https://www.cambridge.org/core. University of New England, on 03 Feb 2018 at 20:33:03, subject to the Cambridge Core terms of use, available
at https://www.cambridge.org/core/terms. https://doi.org/10.1017/eaa.2017.35
 12 European Journal of Archaeology 21 (1) 2018

without the Hittites) and the birth of cul- with language continuity is very difficult
tures derived from Yamnaya remains. to ascertain.
The new work of Reich’s team
(Lazaridis et al., 2016) has retrieved data
Northern European roots that adds weight to this objection. Large-
scale sample analysis has shown that a
The authors’ third objection concerns the mighty fifty-seven per cent contribution of
increase of the genetic similarity of eastern hunter-gatherers had entered the
European population with that of the genetic pool of the Early Bronze Age
Yamnaya culture. This increases in the steppe; a similar contribution (43 per cent)
north of Europe and is weak in the south, had entered the genetic pool of the
in the places adjacent to the Yamnaya Scandinavian hunter-gatherers. This sug-
area, i.e. in Hungary. This gradient is gests that the steppe population (including
clearly expressed in the modern popula- Yamnaya) had a component akin to
tion, but was present already in the that of the Scandinavian Mesolithic popu-
Bronze Age, and hence cannot be lation (something I had suspected from
explained by shifts that occurred in the archaeological and anthropological consid-
Early Iron Age and in medieval times. erations). The steppe cultures of the Early
However, the supposed migration of the Bronze Age have partly the same roots as
Yamnaya culture to the west and north Scandinavian cultures (although not only
should imply a gradient in just the oppos- Scandinavian): the impact of eastern
ite direction! European hunter-gatherers is evident in
It is precisely this paradox that has led many other cultures of eastern Europe. As
to my suggestion to search for the source for the Armenian data, they remain com-
of the supposed impact in the north of pletely outside the net of connections
Europe. My suggestion coincides with the drawn by Reich’s team.
long-noted similarity of a Cromagnon-like
population in the northern European
Mesolithic with the steppe population of Further issues
the Early Neolithic and Bronze Age
steppes, in particular with the Yamnaya. To this set of objections, Kristiansen and
Researchers of the Ukrainian and his colleagues add two more counterweights
Belarussian Mesolithic also advance arch- to my arguments: 1) in two early burials of
aeological hypotheses of this kind. the Corded Ware culture (one in Germany,
the other in Poland) some single attributes
of Yamnaya origin have been found; 2) in
Causasian roots? the Yamnaya and Corded Ware popula-
tions an early form of plague (pestis) was
The authors’ fourth objection is a continu- found (Rasmussen et al., 2015) and both
ation of the third. Both of my opponents’ bacteria stem from one source. As to the first
responses stress the discovery of the point, if this is the full extent of Yamnaya
(Trans) Caucasian (Armenian) roots of the infiltration into central Europe—two
steppe population and hold them as a burials (one for each country) from several
counterweight to the possibility of nor- thousands (and from several hundreds of
thern roots. Yet, usually every archaeo- early burials)—then it hardly amounts to
logical culture has several roots in different large-scale migration. As to the second
directions. Which of them is connected point, the plague could have spread without

Downloaded from https://www.cambridge.org/core. University of New England, on 03 Feb 2018 at 20:33:03, subject to the Cambridge Core terms of use, available
at https://www.cambridge.org/core/terms. https://doi.org/10.1017/eaa.2017.35
 Klejn et al. – Are Origins of Indo-European Languages Explained by Migration of the Yamnaya Culture? 13

any massive migration (a few isolated different: pit graves are not rectangular but
contact points may have been sufficient). oval, skeletons are not on their backs with
I believe I have answered to all the bent legs but curled up on their sides or
authors’ objections. As can be seen, they supine, ochre is scanty, and ceramics are
have not caused me to suspend my doubts. not round-bottomed (as on the Dnieper or
Don) but are of Balkan type.
By contemporary Russian standards,
New genetic results, new interpretations this must be characterized as another, sep-
arate culture, not Yamnaya. Among
Quite recently we have witnessed the Eneolithic steppe cultures classified by
success of a group of geneticists from Rassamakin and Kotova in the Ukraine,
Stanford University and elsewhere (Poznik some are very similar to the Lower
et al., 2016). They succeeded in revealing Danube cultures (Lower Mikhailovka,
varieties of Y-chromosome connected with Kvityana): barrows, oval pit graves, skele-
demographic expansions in the Bronze tons supine or lying curled on their sides,
Age. Such expansion can give rise to and cromlechs. The culture in question
migration. Among the variants connected was one of these, and its possible connec-
with this expansion is R1b, and this hap- tion with Yamnaya (kurgan, what else?) is
logroup is typical for the Yamnaya culture. too limited.
But what bad luck! This haplogroup con- Furthermore, with regard to the barrow
nected with expansion is indicated by the burials of the third millennium BC in the
clade L11, while the Yamnaya burials are basin of the Danube, although they have
associated with a different clade, Z2103, been assigned to the Yamnaya culture, I
that is not marked by expansion. would consider them as also belonging to
It is now time to think about how else another, separate culture, perhaps a mixed
the remarkable results reached by both culture: its burial custom is typical of
teams of experienced and bright geneticists the Yamnaya, but its pottery is absolutely
may be interpreted. not Yamnaya, but local Balkan with
imports of distinctive corded beakers
(Schnurbecher). I would not be surprised if
The cultural affinities of barrow burials Y-chromosome haplogroups of this popu-
in the Danube basin lation were somewhat similar to those of
the Yamnaya, while mitochondrial groups
I have read the outstanding archaeological were indigenous.
works by Volker Heyd with great interest. As yet, geneticists deal with great
In a recent work in co-authorship with blocks of populations and prefer to match
Frı̑nculeasa and Preda (Frı̑nculeasa et al., them to very large and generalized cultural
2015) he summarizes the results of many blocks, while archaeology now analyses
years’ excavations of barrows in the basin more concrete and smaller cultures, each
of the Danube. The results are extremely of which had its own fate.
important. The early date of some burials
on the steppe (last third of the fourth mil-
lennium BC) is substantiated. Yet Heyd
and his co-authors have shown that these REFERENCES
early barrow burials (as distinct from those
of the third millennium, similar in some Allentoft, M.E., Sikora, M., Sjögren, K.-G.,
ways to Yamnaya burials) are significantly et al. 2015. Population Genomics of

Downloaded from https://www.cambridge.org/core. University of New England, on 03 Feb 2018 at 20:33:03, subject to the Cambridge Core terms of use, available
at https://www.cambridge.org/core/terms. https://doi.org/10.1017/eaa.2017.35
 14 European Journal of Archaeology 21 (1) 2018

Bronze Age Eurasia. Nature, 522: 167–72. Landscapes: Burial Mounds in the Copper
doi: 10.1038/nature14507 and Bronze Ages (Central and Eastern
Bouckaert, R., Lemey, P., Dunn, M., et al. Europe – Balkans – Adriatic – Aegean, 4th–
2012. Mapping the Origins and Expansion 2nd millennium BC). Lyon: Maison de
of the Indo-European Language Family. l’Orient et de la Méditerranée, pp.
Science, 337: 957–60. doi: 10.1126/ 236–55.
science.1219669. Corrections and Ivanova, S.V. 2004. Istoricheskaya rekonstrukt-
Clarifications. Science, 342 (2013): 1446. siya i arkheologicheskie realii (Yamnaya
doi: 10.1126/science.342.6165.1446-a kulturno-istoricheskaya oblast) [Historical
Chang, W., Cathcart, C., Hall, D. & Reconstruction and Archaeological Reality
Garrett, A. 2015. Ancestry-constrained (Yamnaya Cultural and Historical Area)].
Phylogenetic Analysis Supports the Indo- Naukovi praci istorichnogo fakultetu
European Steppe Hypothesis. Language, Zaporiskogo universitetu, 18: 330–59.
91: 194–244. Jones, E.R., Gonzales-Fortes, G., Connell, S.,
Chernykh, E.N. & Orlovskaya, L.B. 2004a. et al. 2015. Upper Palaeolithic Genomes
Radiouglerodnaya khronologiya drev- Reveal Deep Roots of Modern Eurasians.
neyamnoy obshchnosti i istoki kurgannykh Nature Communications, 6: article no.
kultur. Rossiyskaya Arkheologiya, 1: 84–99. 8912. doi: 10.1038/ncomms9912
Chernykh, E.N. & Orlovskaya, L B. 2004b. Klejn, L.S. 2015a. Arkheologicheskaya osnova
Radiouglerodnaya khronologiya katakomb- stepnoy gipotezy proiskhozhdeniya indoevro-
noy kulturno-istoricheskoy obshchnosti peytsev: kriticheskiy vzglyad [Archaeological
(sredniy bronzovyj vek). Rossiyskaya Basis of the Steppe Hypothesis of Indo-
Arkheologiya, 2: 15–29. European Origin: Critical View] [à propos of
Frı̑nculeasa., A., Preda, B. & Heyd, V. 2015. K. Kristiansen’s article]. Genofond, August
Pit-graves, Yamnaya and Kurgans along 2015 [online] [accessed 5 April 2017].
the Lower Danube: Disentangling IVth Available at: <http://xn–c1acc6aafa1c.xn–
and IIIrd Millennium BC Burial p1ai/?page_id=4424>
Customs, Equipment and Chronology. Klejn, L.S. 2015b. Archkheolog v odnoy
Praehistorische Zeitschrift, 90: 45–113. komande s genetikami [Archaeologist in
Fu, Q., Posth, C. Hajdinjak, M., et al. 2016. One Team with Geneticicts]. Interview of
The Genetic History of Ice Age Europe. Kristian Kristiansen by Leo Klejn in
Nature, 534: 200–05. doi: 10.1038/ August 2015. Genofond, September 2015
nature17993 [online] [accessed 5 April 2017]. Available
Furholt, M. 2003a. Absolutchronologie und at: <http://xn–c1acc6aafa1c.xn–p1ai/?page_
die Entstehung der Schnurkeramik. id=4535>
Online 16 December 2003 <http://www. Klejn, L.S. 2015c. Otkrytie drevnego stepnogo
jungsteinsite.uni-kiel.de/pdf/2003_furholt. vklada v genetiku evropeyskogo naseleniya –
pdf> [accessed 16 May 2017]. uspekhi i slozhnosti [The Discovery of
Furholt, M. 2003b. Die absolutchronologische the Ancient Steppe Contribution to
Datierung der Schnurkeramik in Mitteleur- the Genomic Pool of the European
opa und Südskandinavien. Bonn: Rudolf Population – Successes and Problems].
Habelt. Conversation with Wolfgang Haak, Part
Gray, R.D. & Atkinson, Q.D. 2003. I. Genofond, November 2015 [online]
Language-tree Divergence Times Support [accessed 16 May 2017]. Available at:
the Anatolian Theory of Indo-European <http://www.genofond.ru>.
Origins. Nature, 426: 435–38. doi: Koncha, S.V. 2004. Perspektivi etnogenetich-
10.1038/nature02029 nich rekonstruktsiy za kam’yanoy dobi
Haak, W., Lazaridis, I., Patterson, N., et al. (materiali indoevropeistiki) [Perspectives
2015. Massive Migration from the Steppe on the Ethnogenetic Reconstruction of
Was a Source for Indo-European Lan- the Stone Age (Indo-European
guages in Europe. Nature, 522: 207–11. Materials)]. Kam’yana doba Ukraini
doi: 10.1038/nature14317 [Ukrainian Stone Age], vol. 5. Kiev:
Heyd, V. 2011. Yamnaya Groups and Tumuli Shlyakh, pp. 191–293.
West of the Black Sea. In: S. Müller- Koncha, S.V. 2005. Ario-kelto-italski zv’yazky
Celka & E. Borgna, eds. Ancestral (za leksichnymy danymy) [Aryan-Celtic-

Downloaded from https://www.cambridge.org/core. University of New England, on 03 Feb 2018 at 20:33:03, subject to the Cambridge Core terms of use, available
at https://www.cambridge.org/core/terms. https://doi.org/10.1017/eaa.2017.35
 Klejn et al. – Are Origins of Indo-European Languages Explained by Migration of the Yamnaya Culture? 15

Italian Communication]. Magisterium, (5): e0155083. doi: 10.1371/journal.

Arkheologichni Studii, 20: 47–52. pone.0155083
Koncha, S.V. 2017. Indoevropeytsi: Piznannya Skoglund, P., Malmström, H., Omrak, A.,
doistorii [Indo-Europeans: Lessons from et al. 2014. Genomic Diversity and
History]. Kiev. Admixture Differs for Stone-Age
Kushniarevich, A., Utevska, O., Chuhryaeva, M., Scandinavian Foragers and Farmers.
et al. 2015. Genetic Heritage of the Science, 344: 747–50. doi: 10.1126/
Balto-Slavic Speaking Populations: A science.1253448
Synthesis of Autosomal, Mitochondrial Starostin, S. 2007. Indo-European among
and Y-Chromosomal Data. 2015/9/2. Other Language Families: Problems of
PLoS ONE, 10(9): e0135820. doi: Dating, Contacts and Genetic
10.1371/journal.pone.0135820 Relationships. In: S.A. Starostin, ed.
Lazaridis, I., Nadel, D., Rollefson, G., et al. Trudy po yazykoznaniyu [Works on
2016. Genomic Insights into the Origin of Linguistics]. Moskva: Nauka, pp. 806–20.
Farming in the Ancient Near East. Stöckli, W.E. 2002. Absolute und relative Chron-
Nature, 536: 419–24. ologie des Früh- und Mittelneolithikums in
Lazaridis, I., Patterson, N., Mittnik, A., et al. Westdeutschland (Rheinland und Rhein-
2014. Ancient Human Genomes Suggest Main-Gebiet). Basel: Archäologie-Verlag.
Three Ancestral Populations for Present- Swadesh, M. 1955. Towards Greater Accuracy in
day Europeans. Nature, 513: 409–13. doi: Lexicostatistic Dating. International Journal of
10.1038/nature13673 American Linguistics, 21: 121–37.
Morgunova, N.L. 2013. Radikarbonnaya khro- Vendryès, J. 1917. Les correspondences de
nologiya yamnoy kultury Volgo-Uralskogo vocabulaire entre l’indo-iranien et l’italo-
mezhdurechya [Radiocarbon Chronology of celtique. Mémoires de la Société de
the Yamnaya Culture at the Volga-Ural Linguistique de Paris, 20: 265–85.
interface]. Kratkie soobshcheniya Instituta Zaliznyak, L.L. 2005. Ukraine and the
arkheologii (Moskva), 230: 5–22. Problem of the Proto-Indo-European
Nikolova, A.V. 2012. Absolyutna khronologiya Original Homeland. In: L.L. Zaliznyak &
yamnoy kultury Pivnichnogo J.C. Carter, eds. Archaeology at Kiev-
Nadchornomor’’ ya v svitli dendrodate Mohyla Academy. Kiev: Stilos, pp. 12–37.
[Absolute Chronology of the Yamnaya
Culture in North Nadchornomor in the
Light of Dendrodates]. Archeologiya
(Kyiv), 4: 14–31.
Pospieszny, L, Sobkowiak-Tabaka, I., Price, T. BIOGRAPHICAL NOTES
D., et al. 2015. Remains of a Late Neolithic
Barrow at Kruszyn. A Glimpse of Ritual and Leo S. Klejn is Emeritus Professor of
Everyday Life in Early Corded Ware Archaeology at St Petersburg University,
Societies of the Polish Lowland.
Praehistorische Zeitschrift, 90: 185–213. his interests in archaeology are theory and
Poznik, G.D., Xue, Y., Mendez, F.L., et al. the Bronze Age, in philology Homeric
2016. Punctuated Bursts in Human Male studies, and the origin of Indo-Europeans
Demography Inferred from 1244 in anthropology Slavic pagan religion.
Worldwide Y-Chromosome Sequences.
Nature Genetics, 48: 593–99. doi: 10.1038/
ng.355
Address: Zheleznovodskaya 27, apt. 27,
Rasmussen, S., Allentoft, M.E., Nielsen, K., 199155, St Petersburg, Russia. [email:
et al. 2015. Early Divergent Strains of [email protected]]
Yersinia pestis in Eurasia 5000 Years Ago.
Cell, 163: 571–82. doi: <http://dx.doi.org/
10.1016/j.cell.2015.10.009> [accessed 16
Wolfgang Haak is leader of the molecular
May 2017].
Sjögren, K.-G., Price, T.D. & Kristiansen, K. anthropology group at the Max Planck
2016. Diet and Mobility in the Corded Institute for the Science of Human
Ware of Central Europe. PloS ONE, 11 History. He works at the interface of

Downloaded from https://www.cambridge.org/core. University of New England, on 03 Feb 2018 at 20:33:03, subject to the Cambridge Core terms of use, available
at https://www.cambridge.org/core/terms. https://doi.org/10.1017/eaa.2017.35
 16 European Journal of Archaeology 21 (1) 2018

human genetics, medical sciences, archae- 40530 Göteborg, Sweden. [email: k.kris-
ology, anthropology, and linguistics. [email protected]]

Address: Max Planck Institute for the

Science of Human History, Kahlaische Karl-Göran Sjögren is docent and senior
Strasse 10, 07745 Jena, Germany. [email: researcher at the University of Gothenburg
[email protected]] covering the Neolithic of northern
Eurasia, mobility, and diet based on stron-
tium isotopic tracing.
Iosif Lazaridis is a post-doctoral fellow at the Address: Department of Historical Studies,
Reich Laboratory, Harvard Medical School. University of Gothenburg, Box 200,
40530 Göteborg, Sweden. [email: karl-
Address: Department of Genetics, Harvard [email protected]]
Medical School, Boston, Massachusetts
02115, USA. [email: lazaridis@genetics.
med.harvard.edu] Morten Allentoft is associate professor at
the Centre for GeoGenetics at
Copenhagen University.
Nick Patterson is deputy lab head at the
Reich Laboratory, Harvard Medical School. Address: Centre for Geogenetics, Natural
History Museum of Denmark, Øster
Address: Department of Genetics, Harvard Voldgade 5–7, 1350 København K,
Medical School, Boston, Massachusetts Denmark. [email: [email protected]]
02115, USA. [email: nickp@broadinsti-
tute.org]
Martin Sikora is associate professor at the
Centre for GeoGenetics at Copenhagen
David Reich is leader of the Reich University.
Laboratory in medical and population
genetics at Harvard Medical School. Address: Centre for Geogenetics, Natural
History Museum of Denmark, Øster
Address: Department of Genetics, Harvard Voldgade 5–7, 1350 København K,
Medical School, Boston, Massachusetts Denmark. [email: [email protected]]
02115, USA. [email: [email protected].
harvard.edu]
Eske Willerslev is Lundbeck Foundation
professor and leader of the Centre for
Kristian Kristiansen is professor of archae- GeoGenetics at Copenhagen University,
ology at the University of Gothenburg. and Prince Phillip professor at Cambridge
His research covers the Bronze Age as his- University.
torical epoch in western Eurasia, archaeo-
logical theory, and cultural heritage. Address: Centre for Geogenetics, Natural
History Museum of Denmark, Øster
Address: Department of Historical Studies, Voldgade 5–7, 1350 København K,
University of Gothenburg, Box 200, Denmark. [email: [email protected]]

Downloaded from https://www.cambridge.org/core. University of New England, on 03 Feb 2018 at 20:33:03, subject to the Cambridge Core terms of use, available
at https://www.cambridge.org/core/terms. https://doi.org/10.1017/eaa.2017.35
 Klejn et al. – Are Origins of Indo-European Languages Explained by Migration of the Yamnaya Culture? 17

Débat : pout-on expliquer les origines des langues indo-européennes par la

migration de la culture Yamna vers l’ouest ?

Deux articles parus dans la revue Nature (Haak et al., 2015 ; Allentoft et al., 2015) firent sensation.
Ils révélaient, du point de vue génétique, qu’une migration de masse de peuples des steppes appartenant
à la culture Yamna affecta l’Europe du centre et du nord à l’âge du Bronze Ancien. Leurs auteurs
tiennent cet évènement comme formant la base de la diffusion des langues indo-européennes. En
réponse, Prof. L.S. Klejn, archéologue à Saint Pétersbourg (Russie), émit certaines critiques à l’égard des
déductions basées sur la génétique, en particulier ses répercussions sur la question des origines des langues
indo-européennes. Ses remarques furent soumises aux auteurs des deux articles, qui à leur tour
présentèrent leurs contre-arguments. Cependant Klejn en vint à conclure que les objections de ces auteurs
n’ont pas atténué ses doutes, ce qui l’amène à une seconde réponse. Translation by Madeleine
Hummler

Mots-clés: langues indo- européennes, migration, culture Yamna, âge du Bronze

Diskussion: Kann man den Ursprung der indoeuropäischen Sprachen mit der
Zuwanderung der Jamnaja-Kultur nach Westen erklären?

Zwei Artikel, welche die Zeitschrift Nature in 2015 veröffentlichte (Haak et al., 2015; Allentoft et al.,
2015), haben großes Aufsehen erregt. Diese lassen, aus genetischer Sicht, eine Massenmigration der
Steppenvölker der Jamnaja-Kultur nach Mittel- und Nordeuropa in der Bronzezeit erkennen. Nach
Auffassung der Verfasser bildet dieses Ereignis die Grundlage der Verbreitung der indoeuropäischen
Sprachen. Als Antwort darauf äußerte sich Prof. L.S. Klejn (Archäologe in Sankt Petersburg, Russland)
kritisch über die genetischen Rückschlüsse, besonders über die Auswirkungen auf die Frage des Ursprungs
der indoeuropäischen Sprachen. Diese kritischen Bemerkungen wurden den Verfassern der Artikel vorge-
legt und die Letzteren haben dann ihre Einwände dargelegt. Klejn ist aber zum Schluss gekommen,
dass die Einwände der Verfasser ihn nicht überzeugen, und untersucht diese Gegenargumente in einer
zweiten Antwort. Translation by Madeleine Hummler

Stichworte: indoeuropäische Sprachen, Migration, Jamnaja-Kultur, Bronzezeit

Downloaded from https://www.cambridge.org/core. University of New England, on 03 Feb 2018 at 20:33:03, subject to the Cambridge Core terms of use, available
at https://www.cambridge.org/core/terms. https://doi.org/10.1017/eaa.2017.35