Common pitfalls in human genomics and bioinformatics: ADMIXTURE, PCA, and the ‘Yamnaya’ ancestral component

invasion-from-the-steppe-yamnaya

Good timing for the publication of two interesting papers, that a lot of people should read very carefully:

ADMIXTURE

Open access A tutorial on how not to over-interpret STRUCTURE and ADMIXTURE bar plots, by Daniel J. Lawson, Lucy van Dorp & Daniel Falush, Nature Communications (2018).

Interesting excerpts (emphasis mine):

Experienced researchers, particularly those interested in population structure and historical inference, typically present STRUCTURE results alongside other methods that make different modelling assumptions. These include TreeMix, ADMIXTUREGRAPH, fineSTRUCTURE, GLOBETROTTER, f3 and D statistics, amongst many others. These models can be used both to probe whether assumptions of the model are likely to hold and to validate specific features of the results. Each also comes with its own pitfalls and difficulties of interpretation. It is not obvious that any single approach represents a direct replacement as a data summary tool. Here we build more directly on the results of STRUCTURE/ADMIXTURE by developing a new approach, badMIXTURE, to examine which features of the data are poorly fit by the model. Rather than intending to replace more specific or sophisticated analyses, we hope to encourage their use by making the limitations of the initial analysis clearer.

The default interpretation protocol

Most researchers are cautious but literal in their interpretation of STRUCTURE and ADMIXTURE results, as caricatured in Fig. 1, as it is difficult to interpret the results at all without making several of these assumptions. Here we use simulated and real data to illustrate how following this protocol can lead to inference of false histories, and how badMIXTURE can be used to examine model fit and avoid common pitfalls.

admixture-protocol
A protocol for interpreting admixture estimates, based on the assumption that the model underlying the inference is correct. If these assumptions are not validated, there is substantial danger of over-interpretation. The “Core protocol” describes the assumptions that are made by the admixture model itself (Protocol 1, 3, 4), and inference for estimating K (Protocol 2). The “Algorithm input” protocol describes choices that can further bias results, while the “Interpretation” protocol describes assumptions that can be made in interpreting the output that are not directly supported by model inference

Discussion

STRUCTURE and ADMIXTURE are popular because they give the user a broad-brush view of variation in genetic data, while allowing the possibility of zooming down on details about specific individuals or labelled groups. Unfortunately it is rarely the case that sampled data follows a simple history comprising a differentiation phase followed by a mixture phase, as assumed in an ADMIXTURE model and highlighted by case study 1. Naïve inferences based on this model (the Protocol of Fig. 1) can be misleading if sampling strategy or the inferred value of the number of populations K is inappropriate, or if recent bottlenecks or unobserved ancient structure appear in the data. It is therefore useful when interpreting the results obtained from real data to think of STRUCTURE and ADMIXTURE as algorithms that parsimoniously explain variation between individuals rather than as parametric models of divergence and admixture.

For example, if admixture events or genetic drift affect all members of the sample equally, then there is no variation between individuals for the model to explain. Non-African humans have a few percent Neanderthal ancestry, but this is invisible to STRUCTURE or ADMIXTURE since it does not result in differences in ancestry profiles between individuals. The same reasoning helps to explain why for most data sets—even in species such as humans where mixing is commonplace—each of the K populations is inferred by STRUCTURE/ADMIXTURE to have non-admixed representatives in the sample. If every individual in a group is in fact admixed, then (with some exceptions) the model simply shifts the allele frequencies of the inferred ancestral population to reflect the fraction of admixture that is shared by all individuals.

Several methods have been developed to estimate K, but for real data, the assumption that there is a true value is always incorrect; the question rather being whether the model is a good enough approximation to be practically useful. First, there may be close relatives in the sample which violates model assumptions. Second, there might be “isolation by distance”, meaning that there are no discrete populations at all. Third, population structure may be hierarchical, with subtle subdivisions nested within diverged groups. This kind of structure can be hard for the algorithms to detect and can lead to underestimation of K. Fourth, population structure may be fluid between historical epochs, with multiple events and structures leaving signals in the data. Many users examine the results of multiple K simultaneously but this makes interpretation more complex, especially because it makes it easier for users to find support for preconceptions about the data somewhere in the results.

In practice, the best that can be expected is that the algorithms choose the smallest number of ancestral populations that can explain the most salient variation in the data. Unless the demographic history of the sample is particularly simple, the value of K inferred according to any statistically sensible criterion is likely to be smaller than the number of distinct drift events that have practically impacted the sample. The algorithm uses variation in admixture proportions between individuals to approximately mimic the effect of more than K distinct drift events without estimating ancestral populations corresponding to each one. In other words, an admixture model is almost always “wrong” (Assumption 2 of the Core protocol, Fig. 1) and should not be interpreted without examining whether this lack of fit matters for a given question.

admixture-pitfalls
Three scenarios that give indistinguishable ADMIXTURE results. a Simplified schematic of each simulation scenario. b Inferred ADMIXTURE plots at K= 11. c CHROMOPAINTER inferred painting palettes.

Because STRUCTURE/ADMIXTURE accounts for the most salient variation, results are greatly affected by sample size in common with other methods. Specifically, groups that contain fewer samples or have undergone little population-specific drift of their own are likely to be fit as mixes of multiple drifted groups, rather than assigned to their own ancestral population. Indeed, if an ancient sample is put into a data set of modern individuals, the ancient sample is typically represented as an admixture of the modern populations (e.g., ref. 28,29), which can happen even if the individual sample is older than the split date of the modern populations and thus cannot be admixed.

This paper was already available as a preprint in bioRxiv (first published in 2016) and it is incredible that it needed to wait all this time to be published. I found it weird how reviewers focused on the “tone” of the paper. I think it is great to see files from the peer review process published, but we need to know who these reviewers were, to understand their whiny remarks… A lot of geneticists out there need to develop a thick skin, or else we are going to see more and more delays based on a perceived incorrect tone towards the field, which seems a rather subjective reason to force researchers to correct a paper.

PCA of SNP data

Open access Effective principal components analysis of SNP data, by Gauch, Qian, Piepho, Zhou, & Chen, bioRxiv (2018).

Interesting excerpts:

A potential hindrance to our advice to upgrade from PCA graphs to PCA biplots is that the SNPs are often so numerous that they would obscure the Items if both were graphed together. One way to reduce clutter, which is used in several figures in this article, is to present a biplot in two side-by-side panels, one for Items and one for SNPs. Another stratagem is to focus on a manageable subset of SNPs of particular interest and show only them in a biplot in order to avoid obscuring the Items. A later section on causal exploration by current methods mentions several procedures for identifying particularly relevant SNPs.

One of several data transformations is ordinarily applied to SNP data prior to PCA computations, such as centering by SNPs. These transformations make a huge difference in the appearance of PCA graphs or biplots. A SNPs-by-Items data matrix constitutes a two-way factorial design, so analysis of variance (ANOVA) recognizes three sources of variation: SNP main effects, Item main effects, and SNP-by-Item (S×I) interaction effects. Double-Centered PCA (DC-PCA) removes both main effects in order to focus on the remaining S×I interaction effects. The resulting PCs are called interaction principal components (IPCs), and are denoted by IPC1, IPC2, and so on. By way of preview, a later section on PCA variants argues that DC-PCA is best for SNP data. Surprisingly, our literature survey did not encounter even a single analysis identified as DC-PCA.

The axes in PCA graphs or biplots are often scaled to obtain a convenient shape, but actually the axes should have the same scale for many reasons emphasized recently by Malik and Piepho [3]. However, our literature survey found a correct ratio of 1 in only 10% of the articles, a slightly faulty ratio of the larger scale over the shorter scale within 1.1 in 12%, and a substantially faulty ratio above 2 in 16% with the worst cases being ratios of 31 and 44. Especially when the scale along one PCA axis is stretched by a factor of 2 or more relative to the other axis, the relationships among various points or clusters of points are distorted and easily misinterpreted. Also, 7% of the articles failed to show the scale on one or both PCA axes, which leaves readers with an impressionistic graph that cannot be reproduced without effort. The contemporary literature on PCA of SNP data mostly violates the prohibition against stretching axes.

pca-how-to
DC-PCA biplot for oat data. The gradient in the CA-arranged matrix in Fig 13 is shown here for both lines and SNPs by the color scheme red, pink, black, light green, dark green.

The percentage of variation captured by each PC is often included in the axis labels of PCA graphs or biplots. In general this information is worth including, but there are two qualifications. First, these percentages need to be interpreted relative to the size of the data matrix because large datasets can capture a small percentage and yet still be effective. For example, for a large dataset with over 107,000 SNPs for over 6,000 persons, the first two components capture only 0.3693% and 0.117% of the variation, and yet the PCA graph shows clear structure (Fig 1A in [4]). Contrariwise, a PCA graph could capture a large percentage of the total variation, even 50% or more, but that would not guarantee that it will show evident structure in the data. Second, the interpretation of these percentages depends on exactly how the PCA analysis was conducted, as explained in a later section on PCA variants. Readers cannot meaningfully interpret the percentages of variation captured by PCA axes when authors fail to communicate which variant of PCA was used.

Conclusion

Five simple recommendations for effective PCA analysis of SNP data emerge from this investigation.

  1. Use the SNP coding 1 for the rare or minor allele and 0 for the common or major allele.
  2. Use DC-PCA; for any other PCA variant, examine its augmented ANOVA table.
  3. Report which SNP coding and PCA variant were selected, as required by contemporary standards in science for transparency and reproducibility, so that readers can interpret PCA results properly and reproduce PCA analyses reliably.
  4. Produce PCA biplots of both Items and SNPs, rather than merely PCA graphs of only Items, in order to display the joint structure of Items and SNPs and thereby to facilitate causal explanations. Be aware of the arch distortion when interpreting PCA graphs or biplots.
  5. Produce PCA biplots and graphs that have the same scale on every axis.

I read the referenced paper Biplots: Do Not Stretch Them!, by Malik and Piepho (2018), and even though it is not directly applicable to the most commonly available PCA graphs out there, it is a good reminder of the distorting effects of stretching. So for example quite recently in Krause-Kyora et al. (2018), where you can see Corded Ware and BBC samples from Central Europe clustering with samples from Yamna:

NOTE. This is related to a vertical distorsion (i.e. horizontal stretching), but possibly also to the addition of some distant outlier sample/s.

pca-cwc-yamna-bbc
Principal Component Analysis (PCA) of the human Karsdorf and Sorsum samples together with previously published ancient populations projected on 27 modern day West Eurasian populations (not shown) based on a set of 1.23 million SNPs (Mathieson et al., 2015). https://doi.org/10.7554/eLife.36666.006

The so-called ‘Yamnaya’ ancestry

Every time I read papers like these, I remember commenters who kept swearing that genetics was the ultimate science that would solve anthropological problems, where unscientific archaeology and linguistics could not. Well, it seems that, like radiocarbon analysis, these promising developing methods need still a lot of refinement to achieve something meaningful, and that they mean nothing without traditional linguistics and archaeology… But we already knew that.

Also, if this is happening in most peer-reviewed publications, made by professional geneticists, in journals of high impact factor, you can only wonder how many more errors and misinterpretations can be found in the obscure market of so many amateur geneticists out there. Because amateur geneticist is a commonly used misnomer for people who are not geneticists (since they don’t have the most basic education in genetics), and some of them are not even ‘amateurs’ (because they are selling the outputs of bioinformatic tools)… It’s like calling healers ‘amateur doctors’.

NOTE. While everyone involved in population genetics is interested in knowing the truth, and we all have our confirmation (and other kinds of) biases, for those who get paid to tell people what they want to hear, and who have sold lots of wrong interpretations already, the incentives of ‘being right’ – and thus getting involved in crooked and paranoid behaviour regarding different interpretations – are as strong as the money they can win or loose by promoting themselves and selling more ‘product’.

As a reminder of how badly these wrong interpretations of genetic results – and the influence of the so-called ‘amateurs’ – can reflect on research groups, yet another turn of the screw by the Copenhagen group, in the oral presentations at Languages and migrations in pre-historic Europe (7-12 Aug 2018), organized by the Copenhagen University. The common theme seems to be that Bell Beaker and thus R1b-L23 subclades do represent a direct expansion from Yamna now, as opposed to being derived from Corded Ware migrants, as they supported before.

NOTE. Yes, the “Yamna → Corded Ware → Únětice / Bell Beaker” migration model is still commonplace in the Copenhagen workgroup. Yes, in 2018. Guus Kroonen had already admitted they were wrong, and it was already changed in the graphic representation accompanying a recent interview to Willerslev. However, since there is still no official retraction by anyone, it seems that each member has to reject the previous model in their own way, and at their own pace. I don’t think we can expect anyone at this point to accept responsibility for their wrong statements.

So their lead archaeologist, Kristian Kristiansen, in The Indo-Europeanization of Europé (sic):

kristiansen-migrations
Kristiansen’s (2018) map of Indo-European migrations

I love the newly invented arrows of migration from Yamna to the north to distinguish among dialects attributed by them to CWC groups, and the intensive use of materials from Heyd’s publications in the presentation, which means they understand he was right – except for the fact that they are used to support a completely different theory, radically opposed to those defended in Heyd’s model

Now added to the Copenhagen’s unending proposals of language expansions, some pearls from the oral presentation:

  • Corded Ware north of the Carpathians of R1a lineages developed Germanic;
  • R1b borugh [?] Italo-Celtic;
  • the increase in steppe ancestry on north European Bell Beakers mean that they “were a continuation of the Yamnaya/Corded Ware expansion”;
  • Corded Ware groups [] stopped their expansion and took over the Bell Beaker package before migrating to England” [yep, it literally says that];
  • Italo-Celtic expanded to the UK and Iberia with Bell Beakers [I guess that included Lusitanian in Iberia, but not Messapian in Italy; or the opposite; or nothing like that, who knows];
  • 2nd millennium BC Bronze Age Atlantic trade systems expanded Proto-Celtic [yep, trade systems expanded the language]
  • 1st millennium BC expanded Gaulish with La Tène, including a “Gaulish version of Celtic to Ireland/UK” [hmmm, dat British Gaulish indeed].

You know, because, why the hell not? A logical, stable, consequential, no-nonsense approach to Indo-European migrations, as always.

Also, compare still more invented arrows of migrations, from Mikkel Nørtoft’s Introducing the Homeland Timeline Map, going against Kristiansen’s multiple arrows, and even against the own recent fantasy map series in showing Bell Beakers stem from Yamna instead of CWC (or not, you never truly know what arrows actually mean):

corded-ware-migrations
Nørtoft’s (2018) maps of Indo-European migrations.

I really, really loved that perennial arrow of migration from Volosovo, ca. 4000-800 BC (3000+ years, no less!), representing Uralic?, like that, without specifics – which is like saying, “somebody from the eastern forest zone, somehow, at some time, expanded something that was not Indo-European to Finland, and we couldn’t care less, except for the fact that they were certainly not R1a“.

This and Kristiansen’s arrows are the most comical invented migration routes of 2018; and that is saying something, given the dozens of similar maps that people publish in forums and blogs each week.

NOTE. You can read a more reasonable account of how haplogroup R1b-L51 and how R1-Z645 subclades expanded, and which dialects most likely expanded with them.

We don’t know where these scholars of the Danish workgroup stand at this moment, or if they ever had (or intended to have) a common position – beyond their persistent ideas of Yamnaya™ ancestral component = Indo-European and R1a must be Indo-European – , because each new publication changes some essential aspects without expressly stating so, and makes thus everything still messier.

It’s hard to accept that this is a series of presentations made by professional linguists, archaeologists, and geneticists, as stated by the official website, and still harder to imagine that they collaborate within the same professional workgroup, which includes experienced geneticists and academics.

I propose the following video to close future presentations introducing innovative ideas like those above, to help the audience find the appropriate mood:

Related

The origins of the Tumulus culture: Proto-Lusatian and potential Proto-Balto-Slavic origins

Interesting chapter The birth of a new world. Barrows, warriors, and metallurgists, by Przemyslaw Makarowicz. In: Urbańczyk P. (Ed.) THE PAST SOCIETIES. Polish lands from the first evidence of human presence to the Early Middle Ages, Warszawa 2017, vol. 3, U. Bugaj (Ed.) (2000 – 500 BC), Warszawa, pp. 127-186.

Some interesting excerpts from the introduction (emphasis mine):

In the 17th century BC the northern reaches of the Únětice culture oecumene experienced a structural crisis and a settlement hiatus; no such interruption in development occurred in the southern or western regions, or further west in the circle of the Blechkreiskulturen (Innerhofer 2000; Müller 2012, 257f.). In light of the most recent research, the decline of Únětice structures in the north was associated with a growing social and ecological crisis that resulted e.g., in the well-documented regression in the development of the fortified settlement in Bruszczewo in Greater Poland/Wielkopolska, which occurred ca. 1650/1600 BC (Kneisel 2012; Kneisel 2013, 101f.; Müller 2012). The settlement structure in that region only stabilized after several decades, with the emergence of Tumulus culture (Schurbein 2009; Cwaliński 2012, 16). In some parts of Central Europe (e.g., Bohemia, Bavaria, Hesse, Thuringia) a relatively gradual and smooth transition in the form of bronze items and pottery was observed between the periods of BA2 and BB1, diagnostic for the Early and Middle Bronze Age respectively (Rittershofer 1984; Innerhofer 2000). The term ‘pre-Tumulus’ horizon (BA3) was introduced to denote the stage that followed the disappearance of Early Bronze Age cultural structures and preceded the formation of Tumulus culture at the foothills of the Alps (Innerhofer 2000, 241f.)

The processes behind the development of this new cultural phenomenon may become clearer if one considers the origins of the new ideology of warriorhood apparent in the most progressive formations of the late stages of the early Bronze Age in the Carpathian Basin (Vandkilde 2007, 129; 2014; the beginnings of the Middle Bronze Age in Hungarian chronology; Hänsel 1968; Bóna 1992; Harding 2000, Fig. 1.3).This factor is particularly relevant in the case of the centralized communities of the Otomani-Füzesabony culture. Its members built impressive fortified settlements, knew advanced methods of bronze casting, and maintained a vast network of contacts that connected the north of Europe with the eastern reaches of the Mediterranean world (e.g., Bouzek 1985; Furmánek, Veliačik, Vladár 1991; Kristiansen, Larrson 2005; David 2007)
(…)
The composition of some spectacular hoards and the presence of military items in some of the graves associated with such communities may suggest that a new type of individualized elite (military aristocracy) emerged in this very culture (Kristiansen 1998, 376f.; 1999; Kristiansen, Larrson 2005). The attractive ideology would then have spread to the west and north-west and be adapted by the ‘post-Early-Bronze’, de-centralized and mobile communities (most likely based on kinship) of animal farmers inhabiting the upper Danube basin and the upper Rhine basin, as well as by the peoples of the Nordic regions (Vandkilde 2014, Fig. 5). This process went hand in hand with the dissemination of the custom of tumulus-building and the associated religious concepts, funerary practices, and territorial behaviour. The mechanism behind the adoption of this custom remains unknown. It may have been the result of imitating the barrows of Corded Ware culture, already present in the landscape of Central Europe – a similar process took place in the communities of the Trzciniec circle (Makarowicz 2009; 2010; 2011). It is also possible that the tumuli were based on the few existing Únětice barrows, though in this case the similarities are more apparent in the stone elements beneath the barrows’ mound. In both cases there was no direct contact between the earlier cultural formation and the emerging group.

proto-lusatian-tumulus
Spatial range of the Silesian-Greater Polish Tumulus Culture (‘Vorlauzitzer Kultur’) after M. Gedl 1992, amended

The new lifestyle became a pan-European phenomenon, but involved a considerable degree of regional diversity that stemmed primarily from contact with local tradition (Bóna 1975; Gedl 1989; Jockenhövel, Kubach [eds.] 1994; David 2002; Jockenhövel 2013). But how did this model spread? It appears that analogies for this development may be found in the social processes and interactions that took place at the beginning of the 3rd millennium BC and led to the emergence of the Bell Beaker phenomenon (Burgess 1986; Nicolis 2001 [ed.]; Czebreszuk 2001; 2004 [ed.]; Heyd 2013; Van der Linden 2013, further literature therein). The most important elements of the ‘Tumulus set of cultural patterns’ included: warriorhood (conveyed through the presence of individual weaponry as grave goods), characteristic types of territorial behaviour (methods of familiarizing space that largely relied on constructing tumuli – monumental graves with a unique external form and internal architecture that was singular, spectacular, and immensely symbolic), and a specific array of valuables made of bronze or, less frequently, of amber or glass (such items indicated the status, gender, and sometimes also the social role of the deceased with whom they were buried). Local cultural milieux transmitted and adapted a set of ideological, social, and political principles that gave the emerging formation coherence and a new ‘quality’. The symbolism of the stone barrow construction (rings, kerbs, cores, rays, etc.), the high value of bronze and amber, and the emergence of the custom of cremation suggests that ‘Tumulus’ communities had a large part to play in the dissemination of the solar cult during the Middle Bronze Age (cf. Kristiansen, Larsson 2005; Czebreszuk 2011, 164-171).

The decline of the Central European early Bronze Age civilization and the birth of a new, pan-European formation was a complex process that lasted at least several decades. It may be surmised that the downfall of Únětice structures and the Otomani-Füzesabony-Gyulavarsánd complex in the Carpathian Basin was brought about primarily by internal structural crises, yet the reasons for the emergence of Tumulus culture lay in the attractive, almost ‘Dionysian’ ideology of warriorhood. Its solidification coincided with the decline of the ‘old’ Early Bronze Age elites that ruled over centralized structures that were territorial in character (fortified settlements with proto-urban characteristics) and were buried in magnificent, richly furnished graves covered with mounds (Fürstengräber). It was also concurrent with the emergence of active kinship-based and de-centralized groups led by the ‘new’ elite class of warriors (the beginnings of military aristocracy?). The significance of such groups continued to grow during the pivotal period – and the decline of the Únětice world and the final turbulent phase of the development of centres in the Carpathian Basin may well be thus described. The process was facilitated by the escalation of military conflicts that occurred in the Bronze Age (Harding 1999; 2007; Kristiansen 1999; Osgood, Monks, with Thoms 2000; Kristiansen, Larsson 2005; Hårde 2006; Vandkilde 2011; 2014). War became an inherent part of social life, as indicated by the increasing presence of weaponry in male graves, rock carvings and steles depicting warriors and their equipment, as well as arrowheads and spearheads embedded in the bones (soft tissues) of the deceased, and plentiful evidence of injuries caused by melee weapons (e.g., Osgood 2006). New types of weaponry (swords, spears) started to be used in the first half of the 2nd millennium BC, leading to more efficient methods of combat (e.g., Harding 2006; Thrane 2006). This must have resulted in the emergence of new types of units, combat styles, and military strategies. It may also be surmised that ‘Tumulus’ communities adopted a hitherto unknown, institutionalized model of warriorhood based on groups of men who dealt with warfare professionally (cf. Sarauw 2007, 66).

The origin of the Tumulus culture meant therefore a pan-European ideological socio-political and ideological change, that may be associated with the last true North-West Indo-European dialect continuum in Europe, as evidenced in Archaeology by long-distance cultural contacts, in Linguistics potentially by late layers of shared vocabulary, and in Ancient Genomics by the different origins of combatants studied from the the Tollense valley.

proto-lusatian-settlements
Settlement points of the Silesian-Greater Polish Tumulus Culture in the Prosna-Odra interfluve (‘close zone’) superimposed on a hypsometric map. By Jakub Niebieszczański

The origins of Tumulus culture in what is now Polish territory most likely resulted from a combination of different factors. In the hitherto prevailing narrative its arrival in the Odra-middle Vistula interfluve was associated with an invasion (aggressive migration) of the Tumulus peoples from enclaves in the middle Danube basin, the destruction of Únětice centres and the Nowa Cerekwia Group, and the subsequent conquest of the western territories inhabited by members of the Trzciniec culture (Gedl 1975, 81; 1989; 1992; Gediga 1978). There is, however, much evidence to suggest that the provenance of this cultural group is more complex.

Recent archaeological research and environmental analyses indicate that the decline of the Únětice culture in the northern reaches of its scope (e.g., the economic and settlement crisis of the Kościan agglomeration with its centre in Bruszczewo and the princely barrow graves in Łęki Małe) was mainly the result of excessive human activity and overly intense exploitation of natural resources (Kneisel 2012; 2013; Müller 2012). Palynological data from the period of1700-1500 BC collected in this part of the North European Plain indicates a decline of human activity. It coincides with the devolution of settlement centres (hamlets and necropolises) dated to the end of the Early Bronze Age and the beginning of the Middle Bronze Age (depopulation?). The decline of Early Bronze Age settlements occurred between 1700 and 1600 BC, whereas the beginning of the Silesian-Greater Polish Tumulus culture may be dated to 1600-1500 BC. A renewed increase in human activity, indicated e.g., by the ‘opening’ of the landscape, did not occur until ca. 1500-1400 BC, in the classic period of the development of ‘Tumulus’ cultural structures (Kneisel 2012, 221).

The whole paper is interesting from the point of view of the potential formation of a Proto-Balto-Slavic community in the Proto-Lusatian or Silesian-Greater Polish Tumulus culture, before its expansion to the east.

After O&M 2018, the only plausible alternative to this model of Balto-Slavic homeland is that Proto-Lusatian represents a Temematic community instead, and an Indo-Slavonic community formed in East Yamna, whereby Balto-Slavic would have possibly expanded with Srubna, and only much later over Temematic territory, absorbing its language as a North-West Indo-European substratum.

See also:

Y-DNA haplogroup R1b-Z2103 in Proto-Indo-Iranians?

chalcolithic_early-asia

We already know that the Sintashta -> Andronovo migrants will probably be dominated by Y-DNA R1a-Z93 lineages. However, I doubt it will be the only Y-DNA haplogroup found.

I said in my predictions for this year that there could not be much new genetic data to ascertain how Pre-Indo-Iranian survived the invasion, gradual replacement and founder effects that happened in terms of male haplogroups after the arrival of late Corded Ware migrants, and that we should probably have to rely on anthropological explanations for language continuity despite genetic replacement, as in the Basque case.

Nevertheless, since we have very few samples, I think we could still see a clear genetic contribution from Yamna to Corded Ware immigrants in the North Caspian region (from Abashevo, in turn a mix of Fatyanovo/Balanovo and Catacomb/Poltavka cultures) in terms of:

  • Ancestral components and PCA in new Sintashta-Petrovka, Andronovo, and/or later samples – similar the ‘steppe’ drift seen in Potapovka relative to Sintashta samples, both formed by incoming Corded Ware migrants – ; and
  • R1b-L23 subclades, either appearing scattered during the Sintashta melting pot (of Abashevo/R1a-Z645 and East Yamna-Poltavka/R1b-Z2103 peoples), or resurging after this period, as we have seen in Pre-Balto-Slavic territory.

This contribution could better explain the obvious language continuity in the region, beautifully complementing the complex anthropological model we have now of archaeological continuity of Sintashta and Potapovka with the previous Poltavka, seen in a similar material and symbolic culture that survived the arrival of newcomers.

A lot of people seem to be looking like crazy since O&M 2018 for some sort of connection between Corded Ware and Yamna migrants in Eastern and Central Europe (wheter in SNP calls of samples published, or among almost forgotten academic papers), either to support the ideas of the 2015 papers – for those who relied on their conclusions and built (even if only mentally) far-fetched migration models around it – , or just because of some sort of absurd continuity theory involving modern R1a-Z645 subclades:

NOTE. The situation we have seen with the hundreds of samples from O&M 2018, and with the recent additional Eastern European samples, depict an unexpected absolutely clear-cut distinction in Y-DNA haplogroups between Corded Ware and Yamna/Bell Beaker: I really can’t see how the situation could be more obvious for everyone, so I doubt any further samples will make certain people change their minds. Their hope is, I guess, that just one sample may give some more oxygen to infinite pet theories, as we are still surprisingly seeing even with reactionary R1b autochthonous continuists in Western Europe…

However, looking into the most likely future for the field, what we should be expecting right now is continuity of Yamna ancestry and lineages in early Proto-Indo-Iranian territory. Since we only have a few samples from Sintashta-Petrovka, Potapovka, and Andronovo, I think there might be a sizeable number of R1b-Z2103 subclades in the territory inhabited by those who – no doubt – spread the language into Central Asia.

Haplogroup_R1b_(Y-DNA)
Modern Y-DNA haplogroup R1b distribution, by Maulucioni at Wikipedia

While full population replacement by R1a-Z93 lineages in the North Caspian region ca. 2000 BC is not impossible, I don’t think it is very likely, since we already know that there are R1b-Z2103 lineages widely distributed in Indo-Iranian-speaking territory, and Z93 is now known to be an older subclade than YFull’s mean formation date suggested (due to the Ukraine_Eneolithic I6561 sample‘s SNP call), so what we can infer now that actually happened in Sintashta -> Andronovo is not exactly the spread of haplogroup Z93 during its formation, but rather a regional reduction in its variability coupled with the expansion of some of its subclades.

The main question, after the South Asia paper is finally published, will then be:

  1. Given that Yamna peoples were an elite group of patrilineally-related families mainly of R1b-L23 subclades:
  2. Accepting that PCA, ADMIXTURE, and other statistical methods are not relevant (alone) for ethnolinguistic identification: e.g. Yamna ‘outliers’ and East Bell Beaker migrants of R1b-L23 lineages without steppe ancestry; N1c1a1a-L392 lineages and Siberian ancestry unrelated to Uralic speakers; R1a-Z645 and steppe ancestry in North-East Europe related to Uralic-speaking cultures
  3. If we find now, as I expect, genetic continuity of east Yamna in Sintashta -> Andronovo (relative to other late Corded Ware peoples), probably including haplogroup R1b-Z2103 mixed with R1a-Z93 before its further reduction of subclades (e.g. to L657) and expansion during its subsequent spread southward…

bronze_age_early_Asia-andronovo
Diachronic map of migrations in Asia ca. 2250-1750 BC

Why exactly do we need Corded Ware to explain migrations of Late Indo-European speakers?

In other words: if we had the data we have today in 2015, would we have a need for Corded Ware to explain Indo-European migrations from the steppe? Are some people so blinded by their will to (appear to) be right in their past interpretations that they can’t just let go?

NOTE. On a side note, wouldn’t it be nice for this paper to publish some other R1b-L23 (x2103) sample – maybe even R1b-L51 – in Yamna, Andronovo, or Afanasevo territory, to end both autochthonous continuity theories (of North-Eastern and Western Europe) at the same time?

I really hope someone in David Reich’s team understands this matter, or else they will still identify Corded Ware as the (now probably ‘a’ instead) vector of expansion of Indo-European languages, and some of us will still have fun for another 2 or 3 years with such conclusions, until someone in the lab realizes that ancestry ≠ population ≠ ethnic identification ≠ language.

NOTE. It seems rather dull to read how people are discussing in the Twitterverse conventional constructs like ‘human race‘ as found in Reich’s op-ed in The New York Times, as if such grandiose semantic discussions had any practical meaning, when basic anthropological questions actually relevant for Genomics, like the essential ancestral component ≠ people tenet seem not to be of interest for anyone in the field….

Since our Indo-European demic difusion model (and its consequences for our reconstruction of North-West Indo-European) and this blog are becoming more and more popular each day – judging by the constant growth in visits in the past 6 months or so – , I guess the simplemindedness and predictability of certain geneticists is benefitting traditional anthropology directly, driving more and more amateur geneticists to look for sound academic models to answer the growing inconsistencies of genetic research.

NOTE. I am not saying the rejection of Corded Ware as spreading Indo-European is definitive. Maybe more samples within some years will depict a clear ancient expansion of Early or Middle Proto-Indo-Europeans from Khvalynsk to the forest-steppe and forest zone, and later with certain Corded Ware migrants into Central Europe, over whose territory a Late Indo-European dialect from Bell Beakers became the superstrate, as some have proposed in the past – e.g. to explain Krahe’s Old European hydronymy. I really doubt you could demonstrate such an old ethnolinguistic identification with a clear, unbroken archaeological trail, though, and we know now that this old hydronymy is probably of Late Indo-European nature (possibly even more recent).

What I am saying is: with the data we have now, it does not make any sense to keep the anthropological models invented by geneticists ex nihilo in 2015, and the hundred different alternative Late Indo-European migration models that arebornwitheachnewpaper.

These Yamna -> Corded Ware migration models didn’t have any sense for me since early 2016, but now after O&M 2017, and especially O&M 2018, I don’t think any geneticist with a little knowledge in Linguistics or Archaeology (if they are decent about their quest for truth in describing ancient European migrations) would buy them, if not for some sort of created ‘tradition’. So let’s ditch Corded Ware as Late Indo-European-speaking, let’s accept that late Corded Ware migrants should most likely be identified as early Uralic speakers, and then future data will tell if we are – again – wrong.

Please, don’t let Genomics become another pseudoscience based solely on Bioinformatics like glottochronology: let anthropologists (preferably mainstream archaeologists, but also the true Indo-Europeanists, linguists) help you interpret your raw data. Don’t deceive yourselves thinking that you have read enough about the Indo-European question, or that you know enough Indo-Europeanists (say what?) to derive your own conclusions.

Use the South Asia paper to begin expressly retracting the Corded Ware mess.

Please pretty please with sugar on top?

Related:

For commenters: this post concerns an anthropological question, and deals with the expansion of Late Proto-Indo-European speakers from Yamna, and the controversy surrounding the role of Corded Ware migrants that a handful of academics propose spread from it, based on a renewed model of Gimbutas’ outdated Kurgan theory and on the so-called ‘Yamnaya’ ancestry.

It happens so that the discussion has turned lately mainly to ancient Y-DNA haplogroups, because they help confirm previous mainstream anthropological models of cultural diffusion and migration. It is obviously not reasonable to judge prehistoric ethnolinguistic migrations from ca. 5,000 years ago based on historical nation-states and ethnic or religious concepts invented since the Middle Ages, coupled with “your” people’s main modern (or your own) paternal lineage.

EDIT (27 MAR 2018): Minor corrections and post made shorter.

Population substructure in Iberia, highest in the north-west territory (to appear in Nature)

A manuscript co-authored by Angel Carracedo, from the University of Santiago de Compostela, and (always according to him) pre-accepted in Nature, will offer more insight into the population substructure of Spain, based on autosomal DNA.

Carracedo’s lecture about DNA (in Galician), including his summary of the paper (from december 2017):

Some of the points made in the video:

  • The study shows a situation parallelling – as expected – the expansion of Spanish Medieval kingdoms during the Reconquista (and subsequent repopulation).
  • In it, the biggest surprise seems to be the greater substructure found in Galicia, the north-western Spanish territory – greater even than expected by the authors.
  • As a side note, Galicia shows a great influence from Moorish” ancestral components, due mainly to the influx from Portugal, which shows more.

It is difficult to judge only from the image and his words, but one could say that there are:

  • Certain quite old ancestral Galician groups;
    • then two – also quite old – ancestral Basque groups;
      • then more recent Galician groups;
        • and then a common, central Spanish group – including
          • a wider Asturian-Catalan group, with a western Asturian-Leonese, and an eastern Catalan subgroup;
          • and a central Castillian-Aragonese group, also with a western Castillian, and an eastern Aragonese subgroup.
spain-autosomal
Spain’s population substructure, from the video.

We thought that certain parts of the British Isles could show ancestral components related to the old population, although this has not proven exactly right, due to more recent population expansions.

However, this paper might shed light to the controversy surrounding Lusitanian (possibly Gallaico-Lusitanian) as a Pre-Celtic Indo-European group of Iberia, either slightly older as an Italo-Celtic dialect, or potentially from the Bell Beaker expansion, whose genetic imprint might have survived the Roman conquest, which apparently didn’t replace its ancestral population.

Given the presence of a central Spanish group opposed to the other minor groups – and knowing that (at least part of) the Medieval kingdoms should be related to the Occitan region – due to the Celtic expansion, and also potentially later during the Visigothic Kingdom, and the Carolingian Empire – , we can only guess that the other (north-western and Basque) groups are potentially quite old, and reflect prehistoric population structures.

Just speculating here, of course. Another interesting genetic paper to await…

Seen first in the Facebook group Iberia ADN.

Related:

Olalde et al. and Mathieson et al. (Nature 2018): R1b-L23 dominates Bell Beaker and Yamna, R1a-M417 resurges in East-Central Europe during the Bronze Age

The official papers Olalde et al. (Nature 2018) and Mathieson et al. (Nature 2018) have appeared. They are based on the 2017 preprints at BioRxiv The Beaker Phenomenon And The Genomic Transformation Of Northwest Europe and The Genomic History Of Southeastern Europe respectively, but with a sizeable number of new samples.

Papers are behind a paywall, but here are the authors’ shareable links to read the papers and supplementary materials: Olalde et al. (2018), Mathieson et al. (2018).

NOTE: The corresponding datasets have been added to the Reich Lab website. Remember you can use my drafts on DIY Human Ancestry analysis (viz. Plink/Eigensoft, PCA, or ADMIXTURE) to investigate the data further in your own computer.

olalde_pca
Image modified by me, from Olalde et al (2018). PCA of 999 Eurasian individuals. Marked is the late CWC outlier sample from Esperstedt, showing how early East Bell Beaker samples are the closest to Yamna samples.

I don’t have time to analyze the samples in detail right now, but in short they seem to convey the same information as before: in Olalde et al. (2018) the pattern of Y-DNA haplogroup and steppe ancestry distribution is overwhelming, with an all-R1b-L23 Bell Beaker people accompanying steppe ancestry into western Europe.

EDIT: In Mathieson et al. (2018), a sample classified as of Ukraine_Eneolithic from Dereivka ca. 2890-2696 BC is of R1b1a1a2a2-Z2103 subclade, so Western Yamna during the migrations also of R1b-L23 subclades, in contrast with the previous R1a lineages in Ukraine. In Olalde et al. (2018), it is clearly stated that of the four BB individuals with higher steppe ancestry, the two with higher coverage could be classified as of R1b-S116/P312 subclades.

This is compatible with the expansion of Indo-European-speaking Yamna migrants (also mainly of R1b-L23 subclades) into the East Bell Beaker group, as described with detail in Archaeology (and with the population movement we are seeing having been predicted) first by Volker Heyd in 2007.

yamna-bell-beaker
Yamna – East Bell Beaker migration 3000-2300 BC. Adapted from Harrison and Heyd (2007), Heyd (2007)

Also, the resurge of R1a-Z645 subclades in Czech and Polish lands (from previous Corded Ware migrants) accompanying other lineages indigenous to the region – seems to have happened only after the Bell Beaker expansion into these territories, during the Bronze Age, probably leading to the formation of the Balto-Slavic community, as I predicted based on previous papers. The fact that a sample of R1b-U106 subclade pops up in this territory is interesting from the point of view of a shared substrate with Germanic, as is the earlier BB sample of R1b-Z2103 for its connection with Graeco-Aryan dialects.

All this suggests that a North-West Indo-European dialect – ancestor of Italo-Celtic, Germanic, and Balto-Slavic -, supported in Linguistics by most modern Indo-European schools of thought, expanded roughly along the Danube, and later to northern, eastern, and western Europe with the Bell Beaker expansion, as supported in Anthropology by Mallory (in Celtic from the West 2, 2013), and by Prescott for the development of a Nordic or Pre-Germanic language in Scandinavia since 1995.

copper-age-late-bell-beaker
Diachronic map of Late Copper Age migrations including Classical Bell Beaker (east group) expansion from central Europe ca. 2600-2250 BC

Maybe more importantly, the fact that only Indo-Iranian-speaking Sintashta-Petrovka (and later Andronovo) cultures were clearly associated with R1a-Z645 subclades, and rather late – after mixing with early Chalcolithic North Caspian steppe groups (mainly East Yamna and Poltavka herders of R1b-L23 subclades) – gives support to the theory that Corded Ware (and probably the earlier Sredni Stog) groups did not speak or spread Indo-European languages with their migration, but most likely Uralic – as seen in recent papers on the much later arrival of haplogroup N1c – (compatible with the Corded Ware substrate hypothesis), adopting Indo-Iranian by way of cultural diffusion or founder effect events.

As Sheldon Cooper would say,

Under normal circumstances I’d say I told you so. But, as I have told you so with such vehemence and frequency already the phrase has lost all meaning. Therefore, I will be replacing it with the phrase, I informed you thusly

I informed you thusly:

The Indo-European demic diffusion model, and the “R1b – Indo-European” association

yamna_bell_beaker_cut

Beginning with the new year, I wanted to commit myself to some predictions, as I did last year, even though they constantly change with new data.

I recently read Proto-Indo-European homelands – ancient genetic clues at last?, by Edward Pegler, which is a good summary of the current state of the art in the Indo-European question for many geneticists – and thus a great example of how well Genetics can influence Indo-European studies, and how badly it can be used to interpret actual cultural events – although more time is necessary for some to realize it. Notice for example the distribution of ‘Yamnaya’ in 3000 BC, all the way to Latvia (based on the initial findings of Mathieson et al. 2017), and the map of 2000 BC with ‘Corded Ware’, both suggesting communities linked by admixture and unrelated to actual cultures.

Some people – especially those interested in keeping a simplistic picture of Europe, either divided into admixture groups or simplistic R1b-Vasconic / R1a-Indo-European / N1c-Uralic (or any combination thereof) – want (others) to believe that I am linking ‘Indo-Europeans’ with haplogroup R1b. That is simply not true. In fact, my model dismisses such simplistic identifications of the reconstructible proto-languages with any modern peoples, admixtures, or haplogroups.

vasconic-uralic
Simplistic Vasconic/R1b-Uralic/N1c distribution, and intruding Indo-European/R1a, according to Wiik.

The beauty of the model lies, therefore, precisely in that if you take any modern group speaking Indo-European languages, none can trace back their combination of language, admixture, and/or haplogroup to a common Indo-European-speaking people. All our ancestral lines have no doubt changed language families (and indeed cultures), they have admixed, and our European regions’ paternal lines have changed, so that any dreams of ‘purity’ or linguistic/cultural/regional continuity become absurd.

That conclusion, which should be obvious to all, has been denied for a long time in blogs and forums alike, and is behind the effort of many of those involved in amateur genetics.

Main linguistic aim

The main consequence of the model, as the title of the paper suggests, is that reconstructible Indo-European proto-languages expanded with people, i.e. with actual communities, which is what we can assert with the help of Genomics. From a personal (or ethnic, or political) point of view genomics is useless, but from an anthropological (and thus linguistic) point of view, genomics can be a very useful tool to decide between alternative models of language diffusion, which has given lots of headaches to those of us involved in Indo-European studies.

The demic diffusion theory for the three main stages of the proto-language expansion was originally, therefore, a dismissal of impossible-to-prove cultural diffusion models for the proto-language – e.g. the adoption of Late Proto-Indo-European by Corded Ware groups due to a patron-client relationship (as proposed by Anthony), or a long-lasting connection between cultures (as proposed by Kristiansen, and favoured by “constellation analogy” proponents like Clackson, who negated the existence of common proto-languages). It also means the acceptance of the easiest anthropological model for language change: migration and – consequently – replacement.

By the time of the famous 2015 papers, I had been dealing for some time with the idea that the shared features between Indo-Iranian and Balto-Slavic may have been due to a common substrate, and must have therefore had some reflection in genomic finds. The data on these papers, and the addition of a weak connection between Pre-Germanic and Balto-Slavic communities, together with their clearest genetic link – R1a-M417 subclades (especially European Z283) – made it still easier to propose a Corded Ware substrate, partially common to the three.

Allentoft Corded Ware
Allentoft et al. “Arrows indicate migrations — those from the Corded Ware reflect the evidence that people of this archaeological culture (or their relatives) were responsible for the spreading of Indo-European languages. All coloured boundaries are approximate.”

Before the famous 2015 papers (and even after them, if we followed their interpretation), we were left to wonder why the supposed vector of expansion of Indo-European languages, Corded Ware migrants – represented by R1a-Z645 subclades, and supposedly continued unchanged into modern populations in its ‘original’ ancestral territories, Balto-Slavic and Indo-Iranian – , were precisely the (phonetically) most divergent Indo-European languages – relative to the parent Late Indo-European proto-language.

My paper implied therefore the dismissal of an unlikely Indo-Slavonic group, as proposed by Kortlandt, and of a still less factible Germano-Slavonic, or Germano-Indo-Slavonic (?) group, as loosely implied by some in the past, and maybe supported in certain archaeological models (viz. Kristiansen or partially Anthony), and presently by some geneticists since their simplistic 2015 papers on “massive migrations from the steppe“, and amateur genetic fans with infinite pet theories, indeed.

A common Corded Ware substrate to Balto-Slavic and Indo-Iranian, and common also partially between Balto-Slavic and Germanic (as supported by Kortlandt, too, albeit with different linguistic connotations), would explain their common features. The Corded Ware culture (and Uralic, tentatively proposed by me as the group’s main language family) is a strong potential connection between them, further supported by phylogeography, too.

Other consequences

Interpretations in my paper help thus dismiss the simplistic Yamna -> Corded Ware -> Bell Beaker migration model implied with phylogeography in the 2000s, and revived again by geneticists and Kristiansen’s workgroup based on the famous 2015 papers, whereby – due to the “Yamnaya ancestral component” – the Yamna culture would have been composed of communities of R1a-M417 and R1b-M269 lineages which remained against all odds ‘related but separated’ for more than two thousand years, sharing a common unitary language (why? and how?), and which expanded from Yamna (mainly R1b-L23) into Corded Ware (mainly R1a-M417) and then into Bell Beaker (mainly R1b-L51), in imaginary migration waves whose traces Archaeology has not found, or Anthropology described, before.

While phylogeography (especially the distribution of ancient samples of certain R1b and R1a subclades) was the main genetic aspect I used in combination with Archaeology and Anthropology to challenge the reliability of the “Yamnaya ancestral component” in assessing migrations – and thus Kristiansen’s now-popular-again modified Kurgan model – , my main aim was to prove a recent expansion of Late Proto-Indo-European from the steppe, and a still more recent expansion of a common group of speakers of North-West Indo-European, the language ancestral to Italo-Celtic, Germanic, and probably Balto-Slavic (or ‘Temematic’, the NWIE substrate of Balto-Slavic, according to some linguists).

My arguments serve for this purpose, and modern distributions of haplogroups or admixture are fully irrelevant: I am ready to change my view at any time, regarding the role of any haplogroup, or ancestral component, archaeological data, or anthropological migration model, to the extent that it supports the soundest linguistic model.

proto-indo-european-stages
Stages of Proto-Indo-European evolution. IU: Indo-Uralic; PU: Proto-Uralic; PAn: Pre-Anatolian; PToch: Pre-Tocharian; Fin-Ugr: Finno-Ugric. The period between Balkan IE and Proto-Greek could be divided in two periods: an older one, called Proto-Greek (close to the time when NWIE was spoken), probably including Macedonian, and spoken somewhere in the Balkans; and a more recent one, called Mello-Greek, coinciding with the classically reconstructed Proto-Greek, already spoken in the Greek peninsula (West 2007). Similarly, the period between Northern Indo-European and North-West Indo-European could be divided, after the split of Pre-Tocharian, into a North-West Indo-European proper, during the expansion of Yamna to the west, and an Old European period, coinciding with the formation and expansion of the East Bell Beaker group.

Gimbutas’ old theory of sudden and recent expansion served well to support a real community of Proto-Indo-European speakers, as did later the Yamna -> Corded Ware -> Bell Beaker theory that circulated in the 2000s based on modern phylogeography, and as did later partially Anthony’s updated steppe theory (2007). On the other hand, Kristiansen’s long-lasting connections among north-west Pontic steppe cultures and Globular Amphorae and Trypillian cultures, did not fit well with a close community expanding rapidly – although recent genetic data on Trypillia and Globular Amphorae might be compelling him to improve his migration theory.

So, if data turns out to be not as I expect now, I will reflect that in future versions of the paper. I have no problem saying I am wrong. I have been wrong many times before, and something I am certain is that I am wrong now in many details, and I am going to be in the future.

If, for example, R1b-L23(xZ2105) is demonstrated to come from Hungary and not the steppe (as supported by Balanovsky) or R1a-M417 samples are proved to have expanded with West Yamna settlers (as recently proposed by Anthony, see below the Balto-Slavic question), I would support the same model from a linguistic point of view, but modified to reflect these facts. Or if a direct migration link is found in Archaeology from Yamna to Corded Ware, and from Corded Ware to Bell Beaker (as proposed in the 2015 papers), I will revise that too (again, see the image below). Or, if – as Lazaridis et al. (2017) paper on Minoans and Mycenaeans suggested – the Anatolian hypothesis (that is, one of the multiple ones proposed) turns out to be somehow right, I will support it.

calcolithic-expansion
My map of Late Proto-Indo-European expansion (A Grammar of Modern Indo-European, 2006), following Gimbutas and Mallory.

Haplogroups are the least important aspect of the whole model, they are just another data that has to be taken into account for a throrough explanation of migrations. It has become essential today because of the apparent lack of vision on the part of geneticists, who failed to use them to adjust their findings of admixture with findings of haplogroup expansions, favouring thus a marginal theory of long-lasting steppe expansion instead of the mainstream anthropological models.

Since many of these alternative scenarios seem less and less likely with each new paper, it is probably more efficient to talk about which developments are most likely to challenge my model.

Main points

My main predictions – based mostly on language guesstimates, archaeological cultures, and anthropological models of migration -, even with the scarce genomic data we had, have been proven right until know with new samples from Mathieson et al. (2017) and Olalde et al. (2017), among other papers of this past year. These were my original assumptions:

(1) A Middle Proto-Indo-European expansion defined by the appearance of steppe ancestry + reduction in haplogroup diversity and expansion of (mainly) R1b-M269 and R1b-L23 lineages;

(2) A Late Proto-Indo-European expansion defined by steppe ancestry + reduction in haplogroup diversity and expansion of (mainly) R1b-L23 subclades; and

(3) A North-West Indo-European expansion defined by steppe ancestry + reduction in haplogroup diversity and expansion of (mainly) R1b-L51 subclades.

The expansion of Corded Ware peoples, associated with steppe ancestry + reduction in haplogroup diversity and expansion of (mainly) R1a-Z645 subclades, represents thus a different migration, which is compatible with the different nature of the Corded Ware culture, unrelated to Yamna and without migration waves from one to the other (although there were certainly contacts in neighbouring regions).

As you can see, neither of the 3+1 expansion models imply that no other haplogroup can be found in the culture or regions involved (others have in fact been found, and still the models remain valid): these migrations imply a reduction of haplogroup diversity, and the expansion of certain subclades as is common in population expansions throughout history. While we all accept this general idea, some people have difficulties accepting just those cases not compatible with their dreams of autochthonous continuity.

Nevertheless, there are still voids in genetic investigation.

Controversial aspects

In my humble opinion, these are potential conflict periods and the most likely areas of change for the future of the theory:

1. When and how did R1b-M269 lineages become “chiefs” in the steppe?

Based on scarce data from Khvalynsk, it seems that during the Neolithic there were many haplogroups in the North Pontic and North Caspian steppes. A reduction to R1b-M269 subclades must have happened either just before or (as I support) during (the migrations that caused) the Suvorovo-Novodanilovka expansion among Sredni Stog, probably coinciding also with the expansion (or one of the expansions) of CHG ancestry (and thus the appearance of ‘Steppe component’ in the steppe). My theory was based initially on Anthony’s account and TMRCA of haplogroups of modern populations (both ca. 4200-4000 BC), but recent samples of the Balkans (R1b-M269 and steppe ancestry) seem to trace the population expansion some centuries back.

If my assessment is correct, then modern populations of haplogroup R1b-M269* and R1b-L23* in the Balkans probably reflect that ancient expansion, and samples related to Proto-Anatolian cultures in the Balkans will most likely be of R1b-M269 subclades and R1b-L23*. After admixture in the Balkans, posterior migrations of Anatolian languages into Anatolia might be associated with a different admixture component and haplogroups, we don’t have enough data yet.

If the haplogroup reduction and expansion in Khvalynsk happened later than the Suvorovo-Novodanilovka expansion, then we might find the expansion of Pre- or Proto-Anatolian associated with many different haplogroups, such as R1b (xM269), R1a, I, J, or G2, and more or less associated with steppe ancestry in the Balkans.

Another reason for finding such variety of haplogroups in ancient samples from the Balkans would be that this Khvalynsk group of “chiefs” traversed – and mixed with – the Sredni Stog population. Nevertheless, if we suppose homogeneity in haplogroups in Khvalynsk during the expansion, a high proportion of different haplogroups explained by admixture with the local population of Sredni Stog would challenge the whole “chief domination” explanation by Anthony, and we would have to return to the “different culture” theory by Rassamakin and potentially an older migration from Khvalynsk. In any case, both researchers show clear links of the Suvorovo-Novodanilovka phenomenon to Khvalynsk, and a differentiation with the surrounding Sredni Stog culture.

A less likely model would support the identification of the whole Eneolithic Pontic-Caspian steppe as a loose Indo-Hittite-speaking community, which would be in my opinion too big a territory and too loose a cultural bond to justify such a long-lasting close linguistic connection. This will probably be the refuge of certain people looking desperately for R1a-IE connections. However, the nature of the western steppe will remain distinct from Late Proto-Indo-European, which must have developed in the Yamna culture, so autochthonous continuity is not on the table anymore, in any case…

suvorovo-novodanilovka-region
Coexistence of the Varna-Gumelniţa culture and the Suvorovo phase of the sceptre-bearer communities. 1 — Fălciu; 2 — Fundeni-Lungoţi; 3 — Novoselskaja; 4 — Suvorovo; 5 — Casimcea; 6 — Kjulevča; 7 — Reka Devnja; 8 — Drama; 9 — Gonova mogila; 10 — Reževo; 11 — geographically separate Decea variant of the sceptre bearer group (after Govedarica, Manzura 2011: Abb. 5, adapted).

2. How did R1a-M417 (and especially R1a-Z645) haplogroups came to dominate over the Corded Ware cultures?

If I am right (again, based on TMRCA of modern populations), then it is precisely at the time of the potential expansion of Proto-Corded Ware from the Dnieper-Dniester forest, forest-steppe, and steppe regions, ca 3300-3000. Furholt’s recent radiocarbon analysis and suggestions of a Lesser Poland origin of the third or A-horizon, on which disparate archaeologists such as Anthony or Klejn rely now, seem to suggest also that Corded Ware was a cultural complex rather than a compact culture reflecting a migration of peoples – similar thus to the Bell Beaker complex.

This cultural complex interpretation of Corded Ware contrasts with the quite homogeneous late samples we have, suggesting clear migration waves in northern Europe, at least at some point in time, so Genomics will be a great tool to ascertain when and from where approximately did Corded Ware peoples expand. Right now, it seems that Eneolithic Ukraine populations are the closest to its origin, so the traditional interpretation of its regional origin by Kristiansen or Anthony remains valid.

3. How was Indo-Iranian adopted by Corded Ware invaders?

This is rather an anthropological question. We need reasonable models of founder effect/cultural diffusion necessary for that to happen – similar to the ones necessary to explain the arrival of N1c subclades into north-east Europe, or the arrival of R1b subclades in Basque/Iberian-speaking regions in south-west Europe. My description of potential events in the eastern steppe – based partially on Anthony – is merely a short sketch. Genomic data is unlikely to offer more than it does today (replacement of haplogroups, and gradually of some steppe component, by late Corded Ware groups in the steppe), but let’s see what new samples can contribute.

As for what some Indians – and other people willing to confront them – are looking for, regarding R1a-M417 and/or Indo-European origins in India, I don’t see the point, we already know a) that the origin of the expansion is in the steppe and b) that Hindu nationalist biggots will not accept results from research that oppose their views. I don’t expect huge surprises there, just more fruitless discussions (fomented by those who live from trolling or conspiracies)…

4. Yamna settlers from Hungary

Anthony’s new theory – and the nature of Balto-Slavic – hinges on the presence of R1a-M417 subclades (associated with later Corded Ware samples) in Yamna settlers of Hungary, potentially originally from the North Pontic area, where the oldest sample has been found.

My ‘modified’ version of Anthony’s new model (the only I deem just remotely factible) includes the expansion of a Proto-Corded Ware from Lesser Poland, but (given the overwhelming R1b found in East Bell Beaker), with R1a-M417 being associated with the region. How to explain this language change with objective data? Well, we have Bell Beaker expanding to these areas at a later time, so we would need to find R1b-L23 settlers in Lesser Poland, and then a resurge of R1a-M417 haplogroup. If not, resorting yet again to cultural diffusion Yamna “patrons” to Corded Ware “clients” of Lesser Poland would bring us to square one, now with the ‘steppe ancestry’ controversy included…

Since some Eastern Europeans are (for no obvious reason whatsoever) putting their hopes on that IE-R1a-CWC association, let’s hope some samples of R1a-M417 in Yamna or Hungary give them a break, so that they can begin accepting something closer to mainstream anthropological models. We could then work from there a Yamna-> Bell Beaker / North-West Indo-European association truce, and from there keep accepting that no single haplogroup from Yamna settlers is linked with modern languages, cultures or ethnic groups.

yamna-region
localization of Central-European funerary monuments with elements of the Pit Grave culture (after Bátora 2006);

5. How and when was Balto-Slavic associated with haplogroup R1a?

If we accept the Southern or Graeco-Aryan nature of Balto-Slavic with influence from an absorbed North-West Indo-European dialect, “Temematic” (as Kortlandt does), then Indo-Slavonic adopted in the steppe from Potapovka by Sintashta and Poltavka populations divided ca. 2000 BC into Indo-Iranian (migrating to the east with Andronovo), and Balto-Slavic (migrating westward with the Srubna culture). History from there is not straightforward, and it should follow Srubna, Thraco-Cimmerian, or other late expansions from cultures of the steppe.

On the other hand, if it is a Northern dialect related closely to Germanic and Italo-Celtic (in a North-West Indo-European group), then its origin has to be found in the initial expansion of East Bell Beakers, and its development into either the Únětice culture (of Balkan and thus potentially “Southern IE” influence), or the Mierzanowice-Nitra culture (of Corded Ware and thus potentially Uralic influence), or maybe from both, given the intermediate substrate found in Germanic and Balto-Slavic.

It is my opinion that the association of Balto-Slavic with haplogroup R1a is quite early after the East Bell Beaker expansion, probably initially with the subclade typically associated with West Slavic, R1a-M458. I have not much data to support this (apart from the most common linguistic model), just modern haplogroup distribution maps and common TMRCA, and highly hypothetical archaeological-anthropological models. Genetics will hopefully bring more data.

Let’s see also what information on ancient haplogroups we can obtain from the Tollense valley (already showing a close cluster with modern West Slavic populations) and steppe regions.

6. How did Germanic, Celtic, and Italic expand?

Germanic is probably the most interesting one. Following the expansion of R1b-L51 subclades (especially R1b-U106) and steppe ancestry (a confounding factor, with the previous expansion of R1a-Z284 subclades) in Scandinavia is going to be fascinating. Anthropological models already point to a linguistic and archaeological expansion of Pre-Germanic with Bell Beaker peoples.

The expansion of Celtic seems to be associated with chiefdoms, untraceable today in terms of haplogroups, and it seems thus different from previous expansions. New studies might tell how that happened, if it was actually in successive ways, as proposed, or maybe we don’t have enough data yet to reach conclusions.

We don’t know either how Italic expanded into the Italian Peninsula, or whether Latin expanded with peoples from Italy, if at all, or it was mostly a cultural diffusion event, as it seems.

Regarding Etruscan, while I think it is a controversy initiated based on fantastic accounts, and ignited with few finds of Middle Eastern ancestry (that seem logical from the point of view of regional contacts), it will be important for Italian linguists and archaeologists, also to accept the most likely scenario.

As for Palaeo-Hispanic languages, while steppe ancestry is found quite reduced in R1b-L51 subclades (after so many different expansions and admixture events since the departure from the steppe), their distribution from the Chalcolithic onwards and the resurgence of native haplogroups may serve to ascertain which Pre-Roman tribes were associated with the oldest regions where these subclades dominated. For that aim, a closer look at the developments in Aquitania and other pre-Roman Vasconic- and Iberian-speaking regions may shed some light on how founder effects might develop to leave the native language intact (in a case similar to the adoption of Indo-Iranian by post-Corded Ware Sinthastha and Potapovka in the eastern Pontic-Caspian steppe).

NOTE: Although mostly unrelated, linguistic questions may also be somehow altered with a change of migration models. For example, our current Corded Ware Substrate Hypothesis – strongly contested by Kortlandt and others – implies that Uralic was potentially the language spoken by Eneolithic Ukraine / Proto-Corded Ware peoples, therefore early Uralic languages were spoken by Corded Ware peoples, as a substrate for Germanic and Balto-Slavic, and Balto-Slavic and Indo-Iranian. If an Indo-Hittite branch different from Late PIE is accepted for Eneolithic Ukraine (thus suggesting a millennia-long cultural-historical community in the steppe), then the model still stands (e.g. Ger. and BSl. *-mos/-mus, as stated by Kortlandt, would correspond to the oldest morphological IE layer). As you can read in the different versions of our model, the different possibilities for the common substrate are stated, and the most likely one selected. But the most likely a priori option sometimes turns out to be wrong…

NOTE 2: You can comment whatever you want here, but I opened a specific thread in our forum if you want serious comments on the model to stuck and be further discussed.

Featured images: from the book Interactions, changes and meanings. Essays in honour of Igor Manzura on the occasion of his 60th birthday. Țerna S., Govedarica B. (eds.). 2016. Kishinev: Stratum Plus.

See also:

Our monograph on North-West Indo-European (first draft) is out

I wrote yesterday about the recently updated Indo-European demic diffusion model.

Fernando López-Menchero and I have published our first draft on the North-West Indo-European proto-language. Our contribution concerns mainly phonetics, and namely two of its most controversial aspects: a common process of laryngeal loss and two series of velars for PIE.

There is also an updated linguistic model for the Corded Ware substrate hypothesis, which seeks to explain certain similarities between Germanic and Balto-Slavic, and between Balto-Slavic and Indo-Iranian, and potential isoglosses between the three.

Available links:

As you probably know, our interest is (and has been for the past 15 years or so, even before our common project) the reconstruction of a North-West Indo-European proto-language, the ancestor of Italo-Celtic, Germanic, and Balto-Slavic. At least since Krahe’s proposal of an Alteuropäische substrate to European hydronymy, some 70 years ago, Indo-Europeanists have been supporting an Old European branch of Proto-Indo-European.

Old_European_hydronymic_map_for_the_root__Sal-,__Salm-
Root *sal-, *salm in European river names. Krahe (1949). From Wikipedia.

However, dialectal divisions were tentative. Since Oettinger, some 30 years ago, we have a clearer picture of a group of closely related dialects, namely Italo-Celtic, Germanic, and Balto-Slavic. Although the nature of Balto-Slavic is somehow contended (for the few scholars who support an Indo-Slavonic group), the minimalist view holds that at least the substrate language of Baltic and Slavic, Holzer‘s Temematic, was part of the North-West Indo-European group.

A North-West Indo-European (NWIE) proto-language not only solved the controversial question of Pan-European IE hydronymy (clearly of Late Indo-European nature), but also – and more elegantly – the question on the origin of the many fragmentary languages attested in Western Europe, usually attributed to a “Pre-Celtic” or “Pre-Italic” nature depending on their surrounding languages (Venetic has even said to be related to Germanic…).

proto-indo-european-stages
Stages of Proto-Indo-European evolution. IU: Indo-Uralic; PU: Proto-Uralic; PAn: Pre-Anatolian; PToch: Pre-Tocharian; Fin-Ugr: Finno-Ugric. The period between Balkan IE and Proto-Greek could be divided in two periods: an older one, called Proto-Greek (close to the time when NWIE was spoken), probably including Macedonian, and spoken somewhere in the Balkans; and a more recent one, called Mello-Greek, coinciding with the classically reconstructed Proto-Greek, already spoken in the Greek peninsula (West 2007). Similarly, the period between Northern Indo-European and North-West Indo-European could be divided, after the split of Pre-Tocharian, into a North-West Indo-European proper, during the expansion of Yamna to the west, and an Old European period, coinciding with the formation and expansion of the East Bell Beaker group.

Described first mainly in terms of lexical isoglosses, the concept of a NWIE language was then gradually and strongly founded in common grammatical features, contributed to mainly by the German, North American, and Spanish schools (as you know, the British or French schools are quite divided on the nature of Proto-Indo-European itself…). Recent archaeological models pioneered by Harrison and Heyd (2007) showed how this might have happened, with Yamna migrants that evolved as the East Bell Beaker group, and their subsequent expansion into most of Europe.

Genetics is now clearly supporting such a closely related group, too.

yamna-bell-beaker
Yamna – East Bell Beaker migration 3000-2300 BC according to Heyd in Harrison and Heyd (2007).

The work of Prescott and Walderhaug (1995) on the Pre-Germanic homeland, and the more precise archaeological migration model developed by Prescott clearly established the advent of Bell Beakers in Scandinavia as the key factor for the development of a unitary Pre-Germanic language in Scandinavia during the Dagger Period of the Nordic Late Neolithic.

The nature of Únětice and Mierzanowice/Nitra cultures as of Bell Beaker absorption of preceding Corded Ware cultures made the identification of the Balto-Slavic homeland in the Lusatian culture as quite likely – and this is now being confirmed with the study of Bronze Age samples, like those of the Tollense battlefield, which cluster closely to West Slavic and East German samples.

At the time of Marija Gimbutas’ breakthrough model of the “kurgan peoples” a common dialect from this Old European branch was deemed to be ‘Northern European‘ (or ‘Germano-Balto-Slavic’), which greatly influenced her work, supporting an identification of different burial types as stemming from the same source. This model, rejected already some years after Gimbutas’ proposal, has sadly survived to this day because of tradition (due e.g. to the work and influence of Kristiansen, and to some extent Anthony), and for some years (until the advent of ancient DNA) because of the modern distribution of haplogroup R1a in Europe and its relation to the ancient distribution of the Corded Ware culture.

This traditional model of a ‘Corded Ware -> Bell Beaker expansion of NWIE’ which we also followed until recently, never fit well with the known migrations paths from Yamna (into Balkan Early Bronze Age cultures), with the geographic distribution of Old European hydronymy, or with the guesstimates for Late Indo-European and North-West Indo-European. This compelled us to support a break-up of the proto-language further back in time than warranted by models of language change, and it needed certain unlikely cultural diffusion events over huge areas (because no such migration from Yamna to northern Europe has been attested): along the steppe/forest-steppe zone first, for a diffusion from Yamna into Corded Ware cultures, and along the Danube or the Rhine later, for a diffusion of Corded Ware into Bell Beaker. These models were also based on the wrong interpretation of the first radiocarbon dates of Beakers – placing an origin of the Bell Beaker people in Iberia (which has been rejected in Archaeology, and now also in Genetics).

Such a ‘Germano-Balto-Slavic’ group faded in Linguistics long ago, with most Indo-Europeanists preferring to talk about late contacts (viz. Celto-Germanic or Italo-Germanic contacts), and for some there is – if any subgroup at all – a core West Indo-European or Italo-Celto-Germanic group, which may be supported by recent genetic research on Bell Beaker peoples, with the Beaker group of the Netherlands being the key. Our research on the potential language spoken by Corded Ware peoples – most likely related to Uralic, from an Indo-Uralic community from the Pontic-Caspian steppe – can elegantly explain the isoglosses that both European dialects share.

copper-age-late-bell-beaker
Diachronic map of Late Copper Age migrations including Classical Bell Beaker (east group) expansion from central Europe ca. 2600-2250 BC

Read also: Schleicher’s Fable in Proto-Indo-European – pitch and stress accent

Related: