I have tried running supervised ADMIXTURE models by selecting distant populations based on PCAs and qpAdm results. The most accurate approximations to what the software should offer appear with a small K number, between K=5 and K=7, whether supervised or unsupervised, and adding more ancestral populations gives some weird results the more distant (in time) populations are from these selected samples.
Labels for ancestral components are used following those commonly referred to in the literature, although supervised ADMIXTURE using corresponding available samples (viz. Anatolia Neolithic for AHG, Iran Hotu and/or CHG for IHG, AG2, AG3 and Mal’ta for ANE, etc.) offer slightly different, less smooth outputs for some periods, especially among more recent populations.
Outputs depend on many different factors, and these files are intended as an overview of the evolution of these simplistic components. The number of available samples per period, the potential ancestry changes within each conventionally selected period, or whether or not each available sample is representative of the territory they were recovered from, among many other factors, influence the outputs and the maps.
NOTE. In summary, ADMIXTURE results like these below might be used to develop new ideas, to be then formally tested; they cannot be used to support anything. Don’t be like the Copenhagen group, randomly selecting “Steppe ancestry” with K=4, identifying this component as “Indo-Europeans”, and correlating its evolution with changes in vegetation composition in yet another obvious correlation = causation argument among many confounding factors left unaccounted for…
Static ADMIXTURE + culture maps
Colours correspond to the components as labelled in the video and in the files below.
The following maps offer natural neighbour interpolations of ancestral components in ancient DNA samples grouped by periods (conventionally selected following the same pattern as in the Prehistory Atlas).
Extrapolation (inferred ancestry beyond the frame created by available samples per map) is obtained by adding distant external locations (such as Greenland, Arctic, Alaska…) with a value of 0.
Videos offer a dynamic timeline.
Click on the images to see a version with higher resolution.
This ancestry peaks among Baikal HG, Ust’Belaya, Nganasans, or Ulchi, hence the different labels used.
Iran HG ancestry
ADMIXTURE maps by period
Click on each image for a higher resolution version.
Early Bronze Age
Middle Bronze Age
Late Bronze Age
Early Iron Age
Late Iron Age
These are the samples used for interpolations in each period (except for modern populations, which are those included in the Reich Lab curated dataset):
Many scholars in the Soviet Union and later the Russian Federation also have researched this topic over the last five decades. Notably the eminent Eugene Helimski dealt with this topic in several articles: his 1992 article (republished in Helimski 2000) on the emergence of Uralic consonantal stems used Indo-Iranian and other Indo-European loans as key evidence, and it was one of the first serious attempts to stratify the loanwords, paying attention to the non-initial syllables as well. Helimski (1997b) discusses Indo-Iranian loanwords more generally, but it is especially notable for the introduction of the “Andronovo Aryan” idea: Helimski argues that some loanwords in Ob-Ugric and Permic are derived from an unattested, third branch of Indo-Iranian. Helimski’s idea has been supported by at least Mikhail Zhivlov in a 2013 article, but otherwise it has not received wide acceptance. Helimski was also known for his criticism (see especially Helimski 2001) of Jorma Koivulehto’s etymological work: although the main targets of Helimski’s criticism were Koivulehto’s writings on Proto-Indo-European and Germanic borrowings (which fitted poorly with Helimski’s ideas of the Nostratic roots of Proto-Uralic and his other theories on Uralic linguistic prehistory), also some of his Indo-Iranian ideas received unnecessarily sharp criticism in Helimski (2001).
Vladimir Napol’skikh is another important Russian scholar who has written on several occasions about Indo-Iranian–Uralic contacts. His 2014 article is notable for its criticism on Helimski’s Andronovo Aryan theory and his arguments in favour of Indo-Aryan loanwords. Napol’skikh also considered some of the traditional Indo-Iranian loanwords to be borrowings from Tocharian (see below) in some of his earlier works, an idea which has been criticized by Kallio (2004) and Widmer (2002) and which Napol’skikh himself has since dropped in later publications (2010, 2014), where many of these alleged Tocharian loans are again considered Indo-Iranian.
Some of the main characteristics of Russian research is that the earliest Indo-European loanwords are usually considered to represent an inheritance from the Nostratic proto-language (Helimski ; Kassian, Zhivlov & Starostin ), an idea which is not widely accepted by scholars of Uralic in the West. Although this often does not concern the Indo-Iranian loanwords at all, or it concerns only a part of them, the works of Jorma Koivulehto, who dealt with both earlier Indo-European and Indo-Iranian loans, receive so much criticism from the Russian scholars that his important ideas are often totally rejected or left unmentioned in Russian research.
This kind of rejection of central etymological research literature can be considered one of the most pressing problems in Uralic loanword studies, and it leaves a regrettable gap between Russian and Western European scholars in this perspective.
Among the Indo-Iranian loanwords in Uralic, one can easily mention examples that follow the classification of semantic change as described above. For widening or generalization, vasara ‘hammer’ is a good example: the Indo-Iranian original denotes ‘the weapon of the god Indra’ in Indic and ‘the weapon of the god Mithra’ in Avestan, whereas Finnish ‘hammer’ (and the Mordvin meaning ‘axe’) are more general meanings of tools. Fi huhta is a good example of narrowing: Iranian *tsuxta- means simply ‘burned’, whereas in Finnic huhta means specifically ‘a burned patch used in slash-and-burn agriculture’. Metonomy has taken place in Mordvin, where čuvto denotes simply ‘tree’; this probably developed through the meaning ‘wood burned for agriculture’. Khanty (South) wǟrəs denotes ‘horse’s mane’, but its Iranian original probably had a more general meaning of hair (cf. Avestan varəsa- ‘hair of human and animal, mostly hair of the head’).
An interesting example of degeneration is the etymology of Finnic orja ‘slave’, probably borrowed from the Indo-Iranian ethnonym *(H)ārya- ‘Aryan’ (for the original semantics of this word, see the entry *orja in Chapter 2). A similar development is seen in English slave which is etymologically connected to the ethnonym Slav.
Distribution as a criterion in the dating of loanwords
(…) some of the Indo-Iranian loans seem to have a wide distribution, but upon a closer look it becomes clear that they include phonological irregularities, which can only be explained by assuming that they are parallel loans. The ability to recognize parallel borrowings is extremely important in Uralic loanword studies, and it has been developed with success in the research of Germanic and Baltic loanwords (see Junttila 2015).
Interestingly, K. Häkkinen (1983: 207) argues that although words disappear from languages, the most basic words often remain stable and are maintained for longer periods. Although this is probably true, here the notion of “basicness” is something that is open to different interpretations. Many central concepts in culture and livelihoods are often described with prestige words that are borrowed, and these central words can be very easily replaced. In determining the age of the loanwords one has to always keep in mind that a reflex of a very early cultural borrowing from Indo-Iranian to Proto-Uralic/Proto- West Uralic etc. can easily have been lost in some daughter language, if a later prestige loan for the same concept has been borrowed from some later contact language (such as from some form of Germanic or Baltic into Finnic or from some Turkic language into Udmurt, Mari or Mordvin).
In Uralic linguistics the common loanword layers shared by some intermediary proto-language have often been seen as giving support to the reconstruction of these stages, but K. Häkkinen (100–108) considers this problematic. It should also be noted that the distribution of Indo-Iranian loanwords very rarely matches the assumed taxonomic divisions: there are some loanwords confined to the Finno-Permic, Finno-Volgaic or Ugric languages, but very few loanwords that would be Finno-Permic, Finno-Volgaic or Ugric in the way that the word is found in all the languages that belong to the branch.
There are only very few possible examples of a consonantal substitution of the word-initial laryngeal. It seems probable that the word-initial laryngeal, if it was retained, was not substituted in any way in Uralic. *karšV (> Fi karhu), an uncertain etymology, is the only possible example.
(…) Even if *k was a result of laryngeal hardening, the development would probably be earlier than Proto-Indo-Iranian, meaning that by the time the word was borrowed, the Indo-Iranian word simply had the stop *k that was regularly substituted by Uralic *k.
Evidence for Andronovo Aryan and Indo-Aryan loanwords?
None of the loanwords have to be considered as Andronovo Aryan or Proto-Indo-Aryan based on the criteria that were presented in the Introduction. The Uralic palatal affricate *ć or sibilant *ś can in all cases be explained from Proto-Indo-Iranian *ć, and there is no need to assume that it should reflect Andronovo Aryan *ć or PIA *ś. In the etymological material of this study, no further positive evidence was found for the distinction of PU *ś and *ć as substitutions of the Proto-Indo-Iranian affricates. This means that at least in word-initial position there probably was no difference between *ć and *ś, and even though we do not know what this sound was phonetically, it is safe to assume that Uralic words showing *ś reflect a sound substitution of Indo-Iranian *ć and *Ʒ́.
Regarding the distribution of the etymologies within Indo-Iranian, all the loanwords which cannot be from Iranian because of the lack of attested Iranian cognates have a more or less secure Proto-Indo-Iranian etymology, and nothing prevents us from assuming that these words reflect Proto-Indo-Iranian borrowings. It is also possible that some words with solid Proto-Indo-Iranian etymologies were present in Iranian but were lost before the first Old Iranian texts were composed.
List of Indo-European and Indo-Iranian Etymologies
*asma ‘voracious’, *jama ‘twin’, *kekrä ‘wheel’, *mača ‘insect’ Of these, *mača from Proto-Iranian and *jama is ambiguous. As the -sm- in asma does not point to Proto-Indo-Iranian *ć, this is probably an Iranian loan too. It is possible that these words were borrowed into Proto-West Uralic, as there is no general support for a Finno-Saamic proto-language today. As the cognates within Finnic and Saami are regular, there is no need to assume parallel borrowings. *kekrä has to be from Proto-Indo-Iranian.
NOTE. Based on the discussion of stages of borrowing from Indo-Iranian, and of the distribution of *kekrä among Uralic dialects in particular, Holopainen probably means Pre-Indo-Iranian for this example.
Only in Mordvin and/or Finnic and/or Saami (can point to a borrowing into Proto-West Uralic):
Based on phonological criteria, these loans do not form a chronologically coherent layer, but probably their modern distribution is accidental (their original distribution can have been wider). *kečrä ‘spindle’ and *rećmä ‘rope’ are from Pre-II, *śaŋka, *waćara and *woraći from PII, *warsa and *wasa from later Iranian (Alanic). *ji̮ni is ambiguous. Also the loans confined to Finnic and Saami mentioned above probably were borrowed into Proto-West Uralic, as it is a more convincing taxonomic entity than Proto-Finno-Saamic.
Only in Mordvin, Finnic and/or Saami and Mari
This loan can be either from PII or PI. As it is obvious that these four branches do not form any taxonomical entity (Salminen 2002; J. Häkkinen 2009), it is only logical that there are no other loanwords with a “Finno-Volgaic” distribution.
Only in Mari:
*kVrtnV ‘metal’ (← PII, PI or later), Pre-Ma *pänti- ‘to bind’, PMa *pärća ‘ear of corn’, *si̮rńa ‘gold’ (← Old Iranian)
Only very few early Indo-Iranian loans can be found in Mari and in no other Uralic language. It is unclear what the reason for this is. It is, of course, possible that some uncertain loanwords like marij ‘man; Mari’ turn out to be correct after all, but even that does not make the number of loans in Mari very high. The situation has to be explained either with loss of vocabulary and replacement by later loans (from Turkic, and also perhaps from Permic) or with Mari’s location on the periphery at the time of the later contacts with the Iranian languages. Agyagási (2019: 254–258) argues that the current area where Mari is spoken was formed only relatively late, after the Mongol invasion in the High Middle Ages. If this is indeed correct, and Mari was spoken in more northern areas before that, it can be assumed that Pre-Mari had only sporadic contacts with the Iranian languages after it split off from Proto-Uralic.
Only in Permic (early loans; for later loans confined to Permic)
The number of loanwords in Permic is relatively high, and many of these can be considered to be Iranian loanwords. Technically many loans are ambiguous, but as some of the words were borrowed late due to historical reasons (‘iron’), and some were borrowed into a Pre-Permic which already had a phonological system that was different from Proto-Uralic (*šud- has d which cannot reflect PU *δ).
It is probable that the Permic languages were in continuous contact with the Indo-Iranian languages from the time they split from Proto-Uralic until the early mediaeval era.
There are so few convincing loanwords with a “Ugric” distribution that they provide very little evidence. Either of these loans could be from Proto-Indo-Iranian or Proto-Iranian, if we assume that *s > *h was a common Iranian sound change. Both loans were acquired
Of these, only *wåtå- has to be a very early loan because of *s > *t. *jäwi (> PS *jäə̑) and PS *wə̑rkə̑ were possibly acquired before the Proto-Samoyed vowel developments, making them probably early loanwords too. Formally all of them could be either from PII or PI. *pulə̑ ~ *pi̮lə̑ could have been borrowed into Proto-Samoyed (with Iranian *u corresponding to Samoyed *u), and because of the *l the word is probably from a relatively late, Middle Iranian language.
The following loanwords have a distribution with a cognate in both Samoyed and some other branch:
*śaδa- ‘to rain’, *tora- ‘to fight’ (also *itä-, which is more uncertain, belongs here)
The following loanwords have a distribution with regular cognates with at least one Ugric branch and some other branch, which points to early borrowing. Although formally *kana- and *kara- are ambiguous, they are probably from Proto-Indo-Iranian because of their distribution. The rest of the loans are from Pre-II or PII.
The following loanwords are found in at least two non-adjacent branches of Uralic (the ones listed in the above categories are not counted). As there are no widely accepted criteria for a word to be considered “Uralic”, all of these could be considered loanwords into Proto-Uralic, in this case probably from Proto-Indo-Iranian or Pre-Indo-Iranian.
Discussion about the distribution and its impact on Uralic taxonomy
(…) there are Proto-Iranian loanwords which were borrowed simultaneously into several early branches of Uralic, making it likely that Uralic had split into several branches by the time of these contacts.
Also the fact that many of the Proto-Indo-Iranian loanwords either show a restricted distribution (such as West Uralic *waćara, *woraći) or irregular correspondences (*asVra, *śasra, *śi̮ta) can point to the conclusion that Proto-Uralic was fragmenting by the time when contacts with Proto-Indo-Iranian took place.
The earlier, Pre-Indo-Iranian loanwords usually show a wider distribution and regular sound correspondences. Although the number of these earliest loans is quite small, based on their distribution and regular correspondences it can be assumed that the Pre-Indo-Iranian stage (after RUKI, *l > *r and the merger of velars and labiovelars but before the merger of non-high vowels) was concurrent with Proto-Uralic, with the changes leading to Proto-Indo-Iranian happening after the dispersal of Proto-Uralic.
The distribution of loanwords reinforces the old idea that Samoyed is a lexical outlier, as only few convincing Indo-Iranian etymologies for Proto-Uralic words (*saδa- ‘to rain’, *tora- ‘to fight’) have a convincing reflex in Samoyed. However, the fact that such etymologies exist means rather that the situation is due to lexical loss in Samoyed, and that the earliest contact occurred before Samoyed split off from Proto-Uralic.
There are very few loanwords that have a Ugric distribution (being found in at least one Ob-Ugric branch and Hungarian), and likewise rather few in Ob-Ugric. The few loans that have a distribution confined to Ugric were borrowed before the change *s > *θ took place. This means that the Ugric distribution does not mean much from the point of view of chronology or taxonomy, as the words were borrowed into a language that was still identical to Proto-Uralic. Even some loans borrowed into Khanty and Mansi have to be so early.
Impacts on dating and the location of the contact zones
Because of the very limited number of convincing etymologies found only in Finnic or Saami, it is probable that there were not (extensive) contacts with Pre-Finnic or Pre-Saami after the split of Proto-West Uralic.
The great number of loanwords of varying ages in Permic inevitably points to the conclusion that the pre-form of the Permic branch had been constantly spoken in an area that was adjacent to the Iranian languages. The different layers of loanwords in Permic clearly point to chronological differences in the donor languages, but it also seems that Permic was in contact with various forms of Iranian and not with different diachronic stages of the same language.
In general, the words that have been borrowed are typical cultural words, and the contacts between Indo-Iranian and Uralic seems to have been a typical contact situation in which a culturally less-advanced language group borrows various cultural terms from a more “advanced” group. The words in various loanword layers related to horse and cattle breeding show obvious cultural influence in the field of domesticated animals, and the borrowing of some names of grains points to agricultural influence from the Indo-Iranians on the speakers of Uralic.
Needless to say, many of the borrowings I listed in A Song of Sheep and Horses suffer from the same ailment attributed to Indo-Europeanists in general:
With slight exaggeration one can agree with the remark by Koivulehto (1999a: 209–210) that the Indo-Europeanists often use outdated sources or are simply uninterested in the topic. The problem is further complicated by the various and often obsolete views expressed in even relatively modern Uralicist works, such as those of Rédei (1986c; 1988) or Katz (2003); (…) Mallory & Adams (2006) adequately refer to the importance of the early loanwords, but they use mostly Rédei’s outdated reconstructions and stratigraphy in support of their theories.
I need to review all related texts with this thesis and the works recently published by Kümmel, as well as the recent book of the Leiden school on Indo-Uralic.
Also, does anyone know the (traditional?) why of the resistance to the Indo-Uralic concept among Uralicists? Maybe it’s a reaction against the Nostraticist and Siberian views of Uralic espoused by the Soviets?
Two new interesting papers concerning Corded Ware and Bell Beaker peoples appeared last week, supporting yet again what is already well-known since 2015 about West Uralic and North-West Indo-European speakers and their expansion.
Below are relevant excerpts (emphasis mine) and comments.
The discovery of the Alexandria outlier represented a clear support for a long-lasting genomic difference between the two distinct cultural groups, Yamnaya and Corded Ware, already visible in an opposition Khvalynsk vs. late Sredni Stog ca. 4000 BC, i.e. well before the formation of both Late Eneolithic/Early Bronze Age groups.
However, the realization that it may not have been an Eneolithic individual, but rather a (Middle?) Bronze Age one, suggests that Sredni Stog was possibly not directly related to Corded Ware, and a potential direct connection with Yamnaya might have to be reevaluated, e.g. through the Carpathian Basin, as Anthony (2017) proposed.
This new paper shows two early Corded Ware individuals from Obłaczkowo, Poland (ca. 2900-2600 BC) – hence close to the supposed original Proto-Corded Ware community – with an apparently (almost) full “Steppe-like” ancestry, clustering (almost) with Yamnaya individuals:
Similar to the BAC individuals, the newly sequenced individuals from the present-day Karlova in Estonia and Obłaczkowo in Poland appear to have strong genetic affinities to other individuals from BAC and CWC contexts across the Baltic Sea region. Some individuals from CWC contexts, including the two from Obłaczkowo, cluster closely with the potential source population of steppe-related ancestry, the Yamnaya herders. Notably, these individuals appear to be those with the earliest radiocarbon dates among all genetically investigated individuals from CWC contexts. Overall, for CWC-associated individuals, there is a clear trend of decreasing affinity to Yamnaya herders with time.
NOTE. Interestingly, this sample is almost certainly attributed to the skeleton E8-A, which had been supposedly already investigated by the Copenhagen group as the RISE1 sample:
We note that RISE1 is also described as the individual from Obłaczkowo feature E8-A. However, their genetic results differ from ours. They present this individual as a molecularly determined male that belongs to Y-chromosomal haplogroup (hg) R1b and to mtDNA hg K1b1a1 while our results show this individual to be female, carrying a mtDNA hg U3a’c profile
NOTE. They might show contributions from Pre-Yamnaya-influenced Sredni Stog, though, but if they show a contribution of Yamnaya, then they are probably outliers, related to Yamnaya vanguard groups (see image below). And for them to show it, then both sources, Yamnaya and Corded Ware, should be clearly distinguishable from each other and their relative contribution quantifiable in formal stats, something difficult (if not impossible) to ascertain today.
Their position in the published PCA – a plot apparently affected by projection bias – suggests a cluster in common with early Baltic samples, which are known to show contributions from East European sub-Neolithic populations (see qpAdm values for Baltic CWC samples).
NOTE. Results for previous samples labelled as Poland CWC are unreliable due to their low coverage.
The most interesting aspect about the ancestry shown by these early samples is their further support for an origin of the culture different than Sredni Stog, and for a rejection of the Alexandria outlier as ancestral to them, hence for a Volhynian-Podolian homeland of Proto-Corded Ware peoples, with an ancestry probably more closely related to the late Maykop Steppe- and Trypillian/GAC groups admixed with sub-Neolithic populations of the Eastern European Late Eneolithic.
NOTE. That is, unless there is a reason for the apparent increase in so-called “Steppe-ancestry” during the northward and westward migration of CWC peoples that represents another thing entirely…
#UPDATE (27 OCT 2019): Apparently, the PCA was actually not affected by projection bias:
Sample poz44 clusters ‘to the south’, with other early German ones, but also close to Yamnaya. Its poor coverage makes qpAdm results unreliable, but its common cluster close to central European and eastern CWC groups – despite belonging to the same Obłaczkowo site – supports that it is more representative of the Proto-CWC population than poz81.
Sample poz81 clusters with Yamnaya samples – or at least with the wider, Steppe-related cluster. Nevertheless, analyses with qpAdm – in combination with values obtained for other early Baltic samples – support that the ancestry of poz81 is more closely related to a core Corded Ware population admixed with sub-Neolithic peoples (similar to Samara LN).
NOTE. I have selected Czech CWC as a potential source closer to the Proto-CWC population, similar to models with Baltic samples. Since Czech CWC samples are later than these from Obłaczkowo, I have also checked the reverse model, with Poz81 and GAC Poland as a source for Czech CWC, and the fits are slightly worse. Anyway, ‘better’ or ‘worse’ p-values can’t determine the direction of migration…
I.2. CWC expansion under R1a bottlenecks
The two males in our dataset (ber1 and poz81) belonged to Y-chromosome R1a haplogroups, as do the majority of males (16/24) from the previously published CWC contexts, while a smaller fraction belonged to R1b [3/24] or I2a [3/24] lineages. The R1a haplogroup has not been found among Neolithic farmer populations nor in hunter–gatherer groups in central and western Europe, but it has been reported from eastern European hunter–gatherers and Eneolithic groups. Individuals from the Pontic–Caspian steppe, associated with the Yamnaya Culture, carry mostly R1b and not R1a haplotypes.
Sample poz81 is of basal hg. R1a-CTS4385*, an R1a-M417 subclade, supporting once again that most Corded Ware individuals from western and central European groups expanded under R1a-M417 (xZ645) lineages. The Battle Axe sample from Bergsgraven (ca. 2620-2470 BC) shows a basal hg. R1a-Y2395*, a R1a-Z283 subclade leading to the typically Fennoscandian R1a-Z284.
Both findings further support that typical lineages of West CWC groups, including R1a-M417 (xZ645) subclades, were fully replaced by incoming East Bell Beakers, and that the limited expansion of R1a-Z284 and I1 (the latter found in one newly reported Late Neolithic sample from Sweden) was the outcome of later regional bottlenecks within Scandinavia, after the creation of a maritime dominion by the Bell Beaker elites during the Dagger Period.
I.3. CWC and lactase persistence
(…) one of these individuals (kar1) carried at least one allele (-13910 C->T) associated with lactose tolerance, while the other two individuals (ber1 and poz81) carried at least one ancestral variant each, consistent with previous observations of low levels of lactose tolerance variants in the Neolithic and a slight increase among individuals from CWC contexts.
The fact that two early CWC individuals carry ancestral variants could be said to support the improbability of the individual from Alexandria representing a community ancestral to the Corded Ware community. On the other hand, the late CWC individual from Estonia carries one allele, but it still seems that only Bell Beakers and Steppe-related groups show the necessary two alleles during the Early Bronze Age, which is in line with a late Repin/early Yamnaya-related origin of the successful selection of the trait, consistent with the expansion of their specialized semi-nomadic cattle-breeding economy through the steppe biome during the Late Eneolithic.
I.4. West Uralic spread from the East
The BAC groups fit as a sister group to the CWC-associated group from Estonia but not as a sister group to the CWC groups from Poland or Lithuania (|Z| > 3), indicating some differences in ancestry between these CWC groups and BAC. Supervised admixture modelling suggests that BAC may be the CWC-related group with the lowest YAM-related ancestry and with more ancestry from European Neolithic groups.
While the results of the paper are compatible with a migration from either the Eastern or the Western Baltic into Scandinavia, phylogeography and archaeology support that Battle Axe peoples emerged as a Baltic Corded Ware group close to the Vistula that expanded first to the north-east, and then to the west from Finland, continuing mostly unscathed during the whole Bronze Age mostly in eastern Fennoscandia with the development of Balto-Finnic- and Samic-speaking communities.
Radiocarbon dating showed that the three individuals from the Öllsjö megalithic tomb derived from later burials, where oll007 (2860–2500 cal BCE) overlaps with the time interval of the BAC, and oll009 and oll010 (1930–1650 cal BCE) fall within the Scandinavian Late Neolithic and Early Bronze Age
In my last post, I showed how the ancestry of Corded Ware from Esperstedt is consistent with influence by incoming Yamnaya vanguard settlers or early Bell Beakers, stemming ultimately from the Carpathian Basin, something that could be inferred from the position of the Esperstedt outlier in the PCA, and by the knowledge of Yamnaya archaeological influences up to Saxony-Anhalt.
Yamnaya settlers are strongly suspected to have migrated in small so-called vanguard groups to the west and north of the Carpathians in the first half of the 3rd millennium BC, well before the eventual adoption of the Proto-Beaker package and their expansion ca. 2500 BC as East Bell Beakers.
Tauber Valley infiltration
As I mentioned in the books, one of the known – among the many more unknown – sites displaying Yamnaya-related traits and suggesting the expansion of Yamnaya settlers into Central Europe is Lauda-Königshofen, in the Tauber Valley.
A series of CW cemeteries have been excavated in the Tauber valley. There are three large cemeteries known and some 30 smaller sites. The larger ones are Tauberbischofsheim-Dittingheim with 62 individuals, Tauberbischofsheim-Impfingen with 40 individuals, and Lauda-Königshofen with 91 individuals. The cemeteries are dispersed rather regularly along the Tauber valley, on both sides of the river, suggesting a quite densely settled landscape.
The Lauda-Königshofen graves consisted mostly of single inhumations in contracted position, usually oriented E-W or NE-SW. A total of 91 individuals were buried in 69 graves. At least 9 double graves and three graves with 3–4 individuals were present. In contrast to the common CW pattern, sexes were not distinguished by body position, only by grave goods. This trait is common in the Tauber valley and suggests a local burial tradition in this area. Stone axes were restricted to males, pottery to females, while other artifacts were common to both sexes. About a third of the graves were surrounded by ring ditches, suggesting palisade enclosures and possibly over-plowed barrows.
In particular, Frînculeasa, Preda, & Heyd (2015) used Lauda-Königshofen as representative of the mobility of horse-riding Yamnaya nomadic herders migrating into southern Germany, referring to the findings in Trautmann (2012) about the nomadic herders from the Tauber Valley, and their already known differences with other Corded Ware groups.
The likely influence of Yamnaya in the region has been reported at least since the 2000s, repeatedly mentioned by Jozef Bátora (2002, 2003, 2006), who compiled Yamnaya influences in a map that has been copied ever since, with little improvement over time. Heyd believes that there are potentially many Yamnaya remains along the Middle and Lower Danube and tributaries not yet found, though.
NOTE. Looking for this specific site, I realized that Bátora (and possibly many after him who, like me, copied his map) located Lauda-Königshofen in a more south-western position within Baden-Württemberg than its actual location. I have now corrected it in the maps of Chalcolithic migrations.
Althäuser Hockergrab…Bell Beakers
Unfortunately, though, it is very difficult to attribute the reported R1b-L51 sample from the Tauber valley to a population preceding the arrival of East Bell Beakers in the region, so there is no uncontroversial smoking gun of Yamnaya vanguard settlers – yet. Reasons to doubt a Pre-Beaker origin are as follows:
1. This family of the Tauber valley shows a late radiocarbon date (ca. 2500 BC), i.e. from a time where East Bell Beakers are known to have been already expanding in all directions from the Middle and Upper Danube and its tributaries.
2.Archaeological information is scarce. Remains of these four individuals were discovered in 1939 and officially reported together with other findings in 1950, without any meaningful data that could distinguish between Bell Beakers and Corded Ware individuals.
This site is located in the Tauber valley, ca. 100 km to the northwest of the Lech valley. The site was discovered during the construction of a sports field in 1939 and was subsequently excavated by G. Müller and O. Paret. Four individuals in crouched position were found in the burial pit of a flat grave. The burial did not contain any grave goods, but due to the type of grave and positioning of the bodies (with heads pointing towards southwest) the site was attributed to the Corded Ware complex.
The classification of this burial as of CWC and not BBC seems to have been based entirely on the numerous CWC findings in the Tauber valley, rather than on its particular burial orientation following a regional custom (foreign to the described standard of both cultures), and on its grave type that was also found among Bell Beaker groups. Like many human remains recovered in dubious circumstances in the 20th century, these samples should have probably been labelled (at least in the genetic paper) more properly as Tauber_LN or Tauber_EBA.
3. In terms of ancestry, there seem to be no gross differences between the Lech Valley BBC individuals and previously reported South German Beakers, originally Yamnaya-like settlers admixing through exogamy with locals, including Corded Ware peoples, as the sex bias of the Lech Valley Beakers proves (see PCA plot below). In other words, northern and eastern Beakers admixed with regional (Epi-)Corded Ware females during their respective expansions, similar to how southern and western Beakers admixed with regional EEF-related females.
The two available Tauber Valley samples (“Tauber_CWC”) show the same pattern: a quite recent Steppe-related male bias and Anatolia_Neolithic-related female bias. Nevertheless, the male sample clusters ‘to the south’ in the PCA relative to all sampled Corded Ware individuals (see PCA plot below), and shows less Yamnaya-like ancestry than what is reported (or can be inferred) for Yamnaya from Hungary or early Bell Beakers of elevated Steppe-related ancestry.
The ancestry and position of the Althäuser male in the PCA is thus fully compatible with recently incoming East Bell Beakers admixing with local peoples (including Corded Ware) through exogamy, but not so much with a sample that would be expected from Yamanaya vanguard + Corded Ware-related ancestry (more like the Esperstedt outlier or the early France Beaker). Compared to the more ‘northern’ (fully Corded Ware-like) position ancestry of his female counterpart, there is little to support that both are part of the same native Tauber valley community after generations of ancestry levelling…
#UPDATE (27 OCT 2019): The PCA shows that the Althäuser male clusters, in fact, ‘to the north’ of the female one, almost on the same spot as a Bell Beaker sample from the Lech Valley.
Despite their reported damage and poor coverage, there seems to be a trend for qpAdm values to prefer a source population for the male (Alt_4) close to Germany Beakers, whereas the female sample (Alt_3) shows ‘better’ fits when a Corded Ware source is selected.
Also relevant is the Corded Ware ancestry of the male – closer to a Czech rather than German CWC source – compatible with an eastern origin, hence supporting a recent arrival via the Danube, in contrast to the local source of the CWC admixture of the female. The poorer coverage of the female sample makes these results questionable, though.
4. The haplogroup inference is also unrevealing: whereas the paper reports that it is R1b-P310* (xU106, xP312), there is no data to support a xP312 call, so it may well be even within the P312 branch, like most sampled Bell Beaker males. Similarly, the paper also reports that HUGO_180Sk1 (ca. 2340 BC) shows a positive SNP for the U106 trunk, which would make it the earliest known U106 sample and originally from Central Europe, but there is no clear support for this SNP call, either. At least not in their downloadable BAM files, as far as I can tell. Even if both were true, they would merely confirm the path of expansion of Yamnaya / East Bell Beakers through the Danube, already visible in confirmed genomic data:
II.2. Proto-Celts and the Tumulus culture
The most interesting data from Mittnik et al. (2019) – overshadowed by the (at first sight) striking “CWC” label of the Althäuser male – is the finding that the most likely (Pre-)Proto-Celtic community of Southern Germany shows, as expected, major genetic continuity over time with Yamnaya/East Bell Beaker-derived patrilineal families, which suggests an almost full replacement of other Y-chromosome haplogroups in Southern German Bronze Age communities, too.
This Central European Bronze Age continuity is particularly visible in many generations of different patrilocal families practising female exogamy, showing patrilineal inheritance mainly under R1b-P312 (mostly U152+) lineages proper of Central European bottlenecks, all of them apparently following a similar sociopolitical system spanning roughly a thousand years, since the arrival of East Bell Beakers in the region (ca. 2500 BC) until – at least – the end of the Middle Bronze Age (ca. 1300 BC):
Here, we show a different kind of social inequality in prehistory, i.e., complex households that consisted of i) a higher-status core family, passing on wealth and status to descendants, ii) unrelated, wealthy and high-status non-local women and iii) local, low-status individuals. Based on comparisons of grave goods, several of the high-status non-local females could have come from areas inhabited by the Unetice culture, i.e., from a distance of at least 350 km. As the EBA evidence from most of Southern Germany is very similar to the Lech valley, we suggest that social structures comparable to our microregion existed in a much broader area. The EBA households in the Lech valley, however, seem similar to the later historically known oikos, the household sphere of classic Greece, as well as the Roman familia, both comprising the kin-related family and their slaves.
The gradual increase in local EEF-like ancestry among South Germany EBA and MBA communities over the previous BBC period offers a reasonable explanation as to how Italic and Celtic communities remained in loose contact (enough to share certain innovations) despite their physical separation by the Alps during the Early Bronze Age, and probably why sampled Bell Beakers from France were found to be the closest source of Celts arriving in Iberia during the Urnfield period.
Furthermore, continued contacts with Únětice-related peoples through exogamy also show how Celtic-speaking communities closer to the Danube might have influenced (and might have been influenced by) Germanic-speaking communities of the Nordic Late Neolithic and Bronze Age, helping explain their potentially long-lasting linguistic exchange.
Like other previous Neolithic or Chalcolithic groups that Yamnaya and Bell Beakers encountered in Europe, ancestry related to the Corded Ware culture became part of Bell Beaker groups during their expansion and later during the ancestry levelling in the European Early Bronze Age, which helps us distinguish the evolution of Indo-European-speaking communities in Europe, and suggests likely contacts between different cultural groups separated hundreds of km. from each other.
All in all, there is nothing to support that (epi-)Corded Ware groups might have survived in any way in Central or Western Europe: whether through their culture, their Y-chromosome haplogroups, or their ancestry, they followed the fate of other rapidly expanding groups before them, viz. Funnelbeaker, Baden, or Globular Amphorae cultural groups. This is very much unlike the West Uralic-speaking territory in the Eastern Baltic and the Russian forests, where Corded Ware-related cultures thrived during the Bronze Age.
It was about time that geneticists caught up with the relevance of Y-DNA bottlenecks when assessing migrations and cultural developments.
From Malmström et al. (2019):
The paternal lineages found in the BAC/CWC individuals remain enigmatic. The majority of individuals from CWC contexts that have been genetically investigated this far for the Y-chromosome belong to Y-haplogroup R1a, while the majority of sequenced individuals of the presumed source population of Yamnaya steppe herders belong to R1b. R1a has been found in Mesolithic and Neolithic Ukraine. This opens the possibility that the Yamnaya and CWC complexes may have been structured in terms of paternal lineages—possibly due to patrilineal inheritance systems in the societies — and that genetic studies have not yet targeted the direct sources of the expansions into central and northern Europe.
Some of the early farmers studied were part of the Neolithic Bell Beaker culture, named for the shape of their pots. Later generations of Bronze Age men who retained Bell Beaker DNA were high-ranking, buried with bronze and copper daggers, axes, and chisels. Those men carried a Y chromosome variant that is still common today in Europe. In contrast, low-ranking men without grave goods had different Y chromosomes, showing a different ancestry on their fathers’ side, and suggesting that men with Bell Beaker ancestry were richer and had more sons, whose genes persist to the present.
There was no sign of these women’s daughters in the burials, suggesting they, too, were sent away for marriage, in a pattern that persisted for 700 years. The only local women were girls from high-status families who died before ages 15 to 17, and poor, unrelated women without grave goods, probably servants, Mittnik says. Strontium levels from three men, in contrast, showed that although they had left the valley as teens, they returned as adults.
(…) it has long been assumed that prior to the Athenian and Roman empires,—which arose nearly 2,500 and more than 2,000 years ago, respectively—human social structure was relatively straightforward: you had those who were in power and those who were not. A study published Thursday in Science suggests it was not that simple. As far back as 4,000 years ago, at the beginning of the Bronze Age and long before Julius Caesar presided over the Forum, human families of varying status levels had quite intimate relationships. Elites lived together with those of lower social classes and women who migrated in from outside communities. It appears early human societies operated in a complex, class-based system that propagated through generations.
A long-lasting and pervasive social system of Bronze Age elites under Yamnaya lineages strikingly similar to this Southern German region can be easily assumed for the British Isles and Iberia, and it is likely to be also found in the Low Countries, Northern Germany, Denmark, Italy, France, Bohemia and Moravia, etc., but also (with some nuances) in Southern Scandinavia and Central-East Europe during the Bronze Age.
Therefore, only the modern genetic pool of some border North-West Indo-European-speaking communities of Europe need further information to describe a precise chain of events before their eventual expansion in more recent times:
the relative geographical isolation causing the visible regional founder effects in Scandinavia, proper of the maritime dominion of the Nordic Late Neolithic (related thus to the Island Biogeography Theory); and
NOTE. Rumour has it that R1b-L23 lineages have already been found among Mycenaeans, while they haven’t been found among sampled early West European Corded Ware groups, so the westward expansion of Indo-European-speaking Yamnaya-derived peoples mainly with R1b-L23 lineages through the Danube Basin merely lacks official confirmation.
I have recently written about the spread of Pre-Yamnaya or Yamnaya ancestry and Corded Ware-related ancestry throughout Eurasia, using exclusively analyses published by professional geneticists, and filling in the gaps and contradictory data with the most reasonable interpretations. I did so consciously, to avoid any suspicion that I was interspersing my own data or cherry picking results.
Now I’m finished recapitulating the known public data, and the only way forward is the assessment of these populations using the available datasets and free tools.
Understanding the complexities of qpAdm is fairly difficult without a proper genetic and statistical background, which I won’t pretend to have, so its tweaking to get strictly correct results would require an unending game of trial and error. I have sadly little time for this, even taking my tendency to procrastination into account… so I have used a simple model akin to those published before – in particular, the outgroup selection by Ning, Wang et al. (2019), who seem to be part of the only group interested in distinguishing Yamnaya-related from Corded Ware-related ancestry, probably the most relevant question discussed today in population genomics regarding the Proto-Indo-European and Proto-Uralic homelands.
I try to prepare in advance a bunch of relevant files with left pops and right pops for each model:
It seems a priori more reasonable to use geographically and chronologically closer proxy populations (say, Trypillia or GAC for Steppe-related peoples) than hypothetic combinations of ancestral ones (viz. Anatolian farmer, WHG, and EHG).
This also means using subgroups closer to the most likely source population, such as (Don-Volga interfluve) Yamnaya_Kalmykia rather than (Middle Volga) Yamnaya_Samara for the western expansion of late Repin/early Yamnaya, or the early Germany_Corded_Ware.SG or Czech_Corded Ware for the group closest to the Proto-Corded Ware population (see below), likely neighbouring the Upper Vistula region.
I usually test two source populations for different targets, which seems like a much more efficient way of using computer resources, whenever I know what I want to test, since I need my PC back for its normal use; whenever I don’t know exactly what to test, I use three-way admixture models and look for subsets to try and improve the results.
I have probably left out some more complex models by individualizing the most relevant groups, but for the time being this would have to do. Also, no other formal stats have been used in any case, which is an evident shortcoming, ruling out an interpretation drawn directly and only from the results below.
Full qpAdm results for each batch of samples are presented in a Google Spreadsheet, with each tab (bottom of the page) showing a different combination of sources, usually in order of formally ‘best’ (first to the left) to ‘worst’ (last to the right) fits, although the order is difficult to select in highly heterogeneous target groups, as will be readily visible.
Corded Ware origins
The latest publications on the Yampil barrow complex have not improved much our understanding of the complexity of Corded Ware origins from an archaeological point of view, involving multiple cultural (hence likely population) influences. This bit is from Ivanova et al., Baltic-Pontic Studies (2015) 20:1, and most hypotheses of the paper remain unanswered (except maybe for the relevance of the Złota group):
In the light of the above outline therefore one should argue that the ‘architecture of barrows’ associated in the ‘Yampil landscape’ of the Middle Dniester Area with the Eneolithic (specifically, mainly with the TC), precedes the development of a similar phenomenon that can be observed from 2900/2800 BC in the Upper Dniester Area and drainage basin of the Upper Vistula, associated with the CWC [Goslar et al. 2015; Włodarczak 2006; 2007; 2008; Jarosz, Włodarczak 2007]. The most consuming research question therefore is whether ritual customs making use of Eneolithic (Tripolye) ‘barrow architecture’ could have penetrated northwards along the Dniester route, where GAC communities functioned. One could also ask what role the rituals played among the autochthons [Kośko 2000; Włodarczak 2008; 2014: 335; Ivanova, Toshchev 2015b].
This issue has already been discussed with a resulting tentative systemic taxonomy in the studies of Włodarczak, arguing for the Złota culture (ZC) in the Vistula region as an illustration of one of the (Małopolska) reception centres of civilization inspirations from the oldest Pontic ‘barrow culture’ circle associated with the Eneolithic and Early Bronze Age [Włodarczak 2008]. Notably, it is in the ZC that one can notice a set of cultural traits (catacomb grave construction, burial details, forms and decoration of vessels) analogous to those shared by the north-western Black Sea Coast groups of the forest-steppe Eneolithic (chiefly Zhyvotilovka-Volchansk) and the Late Tripolye circle (chiefly Usatovo-Gordinești-Horodiștea-Kasperovtsy).
Taking into account that I6561 might be wrongly dated, we cannot include the Corded Ware-like sample of the end-5th millennium BC in the analysis of Corded Ware origins. That uncertainty in the chronology of the appearance of “Steppe ancestry” in Proto-Corded Ware peoples complicates the selection of any potential source population from the CHG cline.
Nevertheless, the lack of hg. R1a-M417 and sizeable Pre-Yamnaya-related ancestry in the sampled Pontic forest-steppe Eneolithic populations (represented exclusively by two samples from Dereivka ca. 3600-3400 BC) would leave open the interesting possibility that a similar ancestry got to the forest-steppe region between modern Poland and Ukraine during the known complex population movements of the Late Eneolithic.
It is known that Corded Ware-derived groups and Steppe Maykop show bad fits for Pre-Yamnaya/Yamnaya ancestry, and also that Steppe Maykop is a potential source of “Steppe-related ancestry” within the Eneolithic CHG mating network of the Pontic-Caspian steppes and forest-steppes. Testing Corded Ware for recent Trypillia and Maykop influences, proper of Late Trypillia and Late Maykop groups in the North Pontic area (such as Zhyvotylivka–Vovchans’k and Gordineşti) side by side with potential Pre-Yamnaya and Yamnaya sources makes thus sense:
Akin to how Yamnaya patrilineal descendants hijacked regional EEF (±CWC) ancestry components mainly through exogamy, dragging them into the different expanding Bell Beaker groups (see below), but kept their Indo-European languages, these hunter-gatherers that admixed with peoples of “Steppe ancestry” were the most likely vector of expansion of Uralic languages in Eastern Europe.
Baltic Corded Ware
One of the most interesting aspects of the results above is the surprising heterogeneity of the different regional groups, which is also reflected in the Y-DNA variability of early Corded Ware samples.
Seeing how Baltic CWC groups, especially the early Latvia_LN sample, show particularly bad fits with the models above, it seems necessary to test how this population might have come to be. My first impression in 2017 was that they could represent early Corded Ware groups admixed with Yamnaya settlers through their interactions along the Dnieper-Dniester corridor.
However, I recently predicted that the most likely admixture leading to their ancestry and PCA cluster would involve a Corded Ware-like group and a group related to sub-Neolithic cultures of eastern Europe, whose best proxy to date are EHG-like Khvalynsk samples (i.e. excluding the outlier with Pre-Yamnaya ancestry, I0434):
Relevant are also the mixtures of Corded Ware from Esperstedt, and particularly those of the sample I0104, which I have repeated many times in this blog I suspected to be influenced by vanguard Yamnaya settlers:
The infeasible models of CWC + Yamnaya_Kalmykia ± Hungary_Baden (see below for Bell Beakers) and the potential cluster formed with other samples from the Baltic suggest that it could represent a more complex set of mixtures with sub-Neolithic populations. On the other hand, its location in Germany, late date (ca. 2500 BC or later), and position in the PCA, together with the good fits obtained for Germany_Beaker as a source, suggest that the increase in Steppe-related ancestry + EEF makes it impossible for the model (as I set it) to directly include Yamnaya_Kalmykia, despite this excess Steppe-related ancestry actually coming from Yamnaya vanguard groups.
These results confirm at least the need to distrust the common interpretation of mixtures including late Corded Ware samples from Esperstedt (giving rise to the “up to 75% Yamnaya ancestry of CWC” in the 2015 papers) as representative of the Corded Ware culture as a whole, and to keep always in mind that an admixture of European BA groups including Corded Ware Esperstedt as a source also includes East BBC-like ancestry, unless proven otherwise.
Bell Beaker expansion
A hotly (re)debated topic in the past 6 months or so, and for all the wrong reasons, is the origin of the Bell Beaker folk. Archaeology, linguistics, and different Y-chromosome bottlenecks clearly indicate that Bell Beakers were at the origin of the North-West Indo-European expansion in Europe, while the survival of Corded Ware-related groups in north-eastern Europe is clearly related to the expansion of Uralic languages.
Nevertheless, every single discarded theory out there seems to keep coming back to life from time to time, and a new wave of interest in “Bell Beaker from the Single Grave culture” somehow got revived in the process, too, because this obsession – unlike the “Bell Beakers from Iberia Chalcolithic” – is apparently acceptable in certain circles, for some reason.
This success in ascertaining a closer Beaker source is probably due to the physical isolation of the specific groups (related to Germany_Beaker, Netherlands_Beaker, and NE_Mediterranean_Beaker samples, respectively) after their migration into regions dominated by peoples without Steppe-related ancestry. Furthermore, Celtic-speaking populations expanding with Urnfield south of the Pyrenees also show a good fit with a source close to France_Beaker.
So I decided to test sampled Bell Beaker populations, to see if it could shed light to the most likely source population of individual Beaker groups and the direction of migration within Central Europe, i.e. roughly eastwards or westwards. As it was to be expected for closely related populations (see the relevant discussion here), an attempt to offer a simplistic analysis of direction based on formal stats does not make any sense, because most of the alternative hypotheses cannot be rejected:
Not only because of the similar values obtained, but because it is absurd to take p-values as a measure of anything, especially when most of these conflicting groups with slightly ‘better’ or ‘worse’ p-values represent multiple different mixtures of the type (Yamnaya + EEF) + (Corded Ware + EEF ± Yamnaya), impossible to distinguish without selecting proper, direct ancestral populations…
A further example of how explosive the Bell Beaker expansion was into different territories, and of their extensive local admixture, is shown by the unsuccessful attempt by Olalde et al. (2018) to obtain an origin of the EEF source for all Beaker groups (excluding Iberian Beakers):
Now, there is a simpler way to understand what kind of Steppe-related ancestry is proper of Bell Beakers. I tested two simple models for some Beaker groups: Yamnaya + Hungary Baden vs. Corded Ware + GAC Poland. After all, the Bell Beaker folk should prefer a source more closely related to either Yamnaya Hungary or Central European Corded Ware:
The admixed Yamnaya samples from Hungary that will hopefully be published soon by the Jena Lab will most likely further improve these fits, especially in combination with intermediate Chalcolithic populations of the Middle and Upper Danube and its tributaries, to a point where there will be an absolute chronological and geographical genomic trail from the fully Yamnaya-likeYamnaya settlers from Hungary to all North-West Indo-European-speaking groups of the Early Bronze Age.
The only difference between groups will be the gradual admixture events of their source Beaker group with local populations on their expansion paths, including peoples of mainly EEF, CWC+EEF, or CWC+EEF+Yamnaya related ancestry. There is ample evidence beyond ancestry models to support this, in particular continued Y-DNA bottlenecks under typical Yamnaya paternal lineages, mainly represented by R1b-L51 subclades.
European Early Bronze Age
European EBA groups that might show conflicting results due to multiple admixture events with Corded Ware-related populations are the Únětice culture and the Nordic Late Neolithic.
The results for Únětice groups seem to be in line with what is expected of a Central European EBA population derived from Bell Beakers admixed with surrounding poulations of East Bell Beaker and/or late (Epi-)Corded Ware descent.
Potential models of mixture for Nordic Late Neolithic samples – despite the bad fits due to the lack of direct ancestral CWC and BBC groups from Denmark – seem to be impossible to justify as derived exclusively from Single Grave or (even less) from Battle Axe peoples, supporting immigration waves of Bell Beakers from the south and further admixture events with local groups through maritime domination.
Balkans Bronze Age
The potential origin of the typical Corded Ware Steppe-related ancestry in the social upheaval and population movements of the Dnieper-Dniester forest-steppe corridor during the 4th millennium BC raises the question: how much do Balkan Bronze Age groups owe their ancestry to a population different than the spread of Pre-Yamnaya-like Suvorovo-Novodanilovka chieftains? Furthermore, which Bronze Age groups seem to be more likely derived exclusively from Pre-Yamnaya groups, and which are more likely to be derived from a mixture of Yamnaya and Pre-Yamnaya? Do the formal stats obtained correspond to the expected results for each group?
Since the expansion of hg. I2a-L699 (TMRCA ca. 5500 BC) need not be associated with Yamnaya, some of these values – together with the assessment of each individual archaeological culture – may question their origin in a Yamnaya-related expansion rather than in a Khvalynsk-related one.
NOTE. These are the last ones I was able to test yesterday, and I have not thought these models through, so feel free to propose other source and target groups. In particular, complex movements through the North Pontic area during the Late Eneolithic would suggest that there might have been different Steppe-ancestry-related vs. EEF-related interactions in the north-west and west Pontic area before and during the expansion of Yamnaya.
One of the key Indo-European populations that should be derived from Yamnaya to confirm the Steppe hypothesis, together with North-West Indo-Europeans, are Proto-Greeks, who will in turn improve our understanding of the preceding Palaeo-Balkan community. Unfortunately, we only have Mycenaean samples from the Aegean, with slight contributions of Steppe-related ancestry.
Still, analyses with potential source populations for this Steppe ancestry show that the Yamnaya outlier from Bulgaria is a good fit:
The comparison of all results makes it quite evident the why of the good fits from (Srubnaya-related) Bulgaria_MLBA I2163 or of Sintashta_MLBA relative to the only a priori reasonable Yamnaya and Catacomb sources: it is not about some hypothetical shared ancestor in Graeco-Aryan-speaking East Yamnaya– or even Catacomb-Poltavka-related groups, because all available Yamnaya-related peoples are almost indistinguishable from each other (at least with the sampling available today). These results reflect a sizeable contribution of similar EEF-related populations from around the Carpathians in both Steppe-related groups: Corded Ware and Yamnaya settlers from the Balkans.
In hobby ancestry magic, as in magic in general, it is not about getting dubious results out of thin air: misdirection is the key. A magician needs to draw the audience attention to ‘remarkable’ ancestry percentages coupled with ‘great’ (?) p-values that purportedly “prove” what the audience expects to see, distracting everyone from the true interesting aspects, like statistical design, the data used (and its shortcomings), other opposing models, a comparison of values, a proper interpretation…you name it.
I reckon – based on the examples above – that the following problems lie at the core of bad uses of qpAdm:
In the formal aspect, the poor understanding of what p-values and other formal stats obtained actually mean, and – more importantly – what they don’t mean. The simplistic trend to accept results of a few analyses at face value is necessarily wrong, in so far as there is often no proper reasoning of what is being assessed and how, and there is never a previous opinion about what could be expected if the alternative hypotheses were true.
In the interpretation aspect, the poor judgement of accompanying any results with simplistic, superficial, irrelevant, and often plainly wrong archaeological or linguistic data selected a posteriori; the inclusion of some racial or sociopolitical overtones in the mixture to set a propitious mood in the target audience; and a sort of ritualistic theatrics with the main theme of ‘winning’, that is best completed with ad hominems.
If you get rid of all this, the most reasonable interpretation of the output of a model proposed and tested should be similar to Nick Patterson’s words in his explanation of qpWave and qpAdm use:
Here we see that, at least in this analysis there are reasonable models with CordedWareNeolithic is a mix of either WHG or LBKNeolithic and YamnayaEBA. (…) The point of this note is not to give a serious phylogenetic analysis but the results here certainly support a major Steppe contribution to the Corded Ware population, which is entirely concordant with the archaeology [?].
Very far, as you can see, from the childish “Eureka! I proved the source!”-kind of thinking common among hobbyists.
The Mycenaean case is an illustrative example: if the Yamnaya outlier from Bulgaria were not available, and if one were not careful when designing and assessing those mixture models, the interpretation would range from erroneous (viz. a Graeco-Aryan substrate, as I initially thought) to impossible (say, inventing migration waves of Sintashta or Srubnaya peoples into Crete). The models presented above show that a contribution of Yamnaya to Mycenaeans couldn’t be rejected, and this alone should have been enough to accept Yamnaya as the most likely source population of “Steppe ancestry” in Proto-Greeks, pending intermediate samples from the Balkans. In other words, one could actually find that ‘the best’ p-values for source populations of Mycenaeans is a combination of modern Poles + Turks, despite the impracticality of such a model…
I haven’t been able to reproduce results which supposedly showed that Corded Ware is more likely to be derived from (Pre-)Yamnaya than other source population, or that Corded Ware is better suited as the ancestral population of Bell Beakers. The analyses above show values in line with what has been published in recent scientific papers, and what should be expected based on linguistics and archaeology. So I’ll go out on a limb here and say that it’s only through a careful selection of outgroups and samples tested, and of as few compared models as possible, that you could eventually get this kind of results and interpretation, if at all.
Whether that kind of special care for outgroups and samples is about (a) an acceptable fine-tuning of the analyses, (b) a simplistic selection dragged from the first papers published and applied indiscriminately to all models, or (c) cherry picking analyses until results fit the expected outcome, is a question that will become mostly irrelevant when future publications continue to support an origin of the expansion of ancient Indo-European languages in Khvalynsk- and Yamnaya-related migrations.
Feel free to suggest (reasonable) modifications to correct some of these models in the comments. Also, be sure to check out other values such as proportions, SD or SNPs of the different results that I might have not taken into account when assessing ‘good’ or ‘bad’ fits.
The current state of research reveals no firm evidence of crop cultivation in the region before the LBA (Piličiauskas et al. 2017b; Grikpėdis and Motuzaitė-Matuzevičiūtė 2018). Current archaeobotanical data firmly suggest the adoption of farming during the EBA to LBA transition. (…) By comparison, in other parts of N Europe subsistence economy of CWC groups was characterized by strong emphasis on animal husbandry, however crop cultivation was also used (Kirleis 2019; Vanhanen et al. 2019). CWC sites from the Netherlands, Denmark, Sweden and Germany reveal evidence of the cultivation of H. vulgare var. nudum, T. dicoccum, Linum usitatissimum (flax) (Oudemans and Kubiak-Martens 2014; Beckerman 2015; Kubiak- Martens et al. 2015).
It is (…) striking that earliest evidence of farming in the SE Baltic only appears in the deposits dating over 4,000 years later.
The environmental conditions of the SE Baltic presented a significant barrier and numerous genetic adaptations were required before farming could successfully spread into the region (Motuzaitė-Matuzevičiūtė 2018). Adaptations through seasonality changes usually play a major role in adapting to new environments (Sherratt 1980). These include establishing genetic controls on seasonality, especially flowering times and length of growing season (Fuller and Lucas 2017). Therefore, it could be argued that farming was only firmly established in the region around the LBA after several crop species, primarily barley, became adapted to the local environment and the risk of crop failure was reduced (Motuzaitė-Matuzevičiūtė 2018). The transition to farming was further aided by the climate warming which started around 750 cal bc (Gaigalas 2004; Sillasoo et al. 2009). In such a case the fragmented evidence from earlier periods is a likely illustration of the early attempts that have failed.
The LBA agrarian intensification of the SE Baltic was most likely not an isolated case but rather a part of broader social, economic and technological developments sweeping across northern Europe.
Evidence from sites across the Baltic Sea shows that the end of the EBA (ca. 1200 bc onward, after Gustafsson 1998) was marked by intensification of agriculture and changes in landscape management. This coincides with the agricultural developments observed on the SE fringes of the Baltic Sea and provides a context for the eventual arrival of farming, followed shortly by the rapid agrarian intensification of the region. Looking just south from the study region, we see that data from northern Poland reveal a sharp increase in both scale and intensity of agricultural activities during the EBA to LBA transition. Pollen records show significant environmental changes starting around 1400/1300 bc (Wacnik 2005, 2009; Wacnik et al. 2012). These were mostly a result of development of a production economy based on plant cultivation and animal raising. Even more significant changes during this period are visible in southern Scandinavia. Pollen records from S Sweden present evidence for an opening up of the forested landscape and the creation of extensive grasslands (Berglund 1991; Gustafsson 1998). Major changes are also apparent in archaeobotanical assemblages.
In general, during the end of the EBA northern Europe underwent a massive transformation of the farming system moving towards a more intensified agriculture aimed at surplus production. However, this should not be regarded as an isolated occurrence, but rather as a radical change of the whole society which took place throughout Europe (Gustafsson 1998). Intensification of contacts across northern Europe have integrated previously isolated regions into a wider network (Kristiansen and Larsson 2005; Wehlin 2013; Earle et al. 2015). It is therefore likely that farming spread into the SE fringes of the Baltic Sea alongside other innovations including malleable technologies and developments of social structure.
The presence and scale of intensifying connections is well illustrated by SE Baltic archaeological material.
Firstly, the appearance of stone ship graves has served as a basis for locating the Nordic communication zones. Construction of such graves was limited to the coastal regions of Kurzeme, Saaremaa Island and the Northern Estonian coast near Tallinn and Kaliningrad (Graudonis 1967; Okulicz 1976; Lang 2007) and is generally regarded as a foreign burial custom which was common in Gotland and along the Scandinavian coast. This is also supported by the Staldzene and Tehumardi hoards (Vasks and Vijups 2004; Sperling 2013), which contained artefacts typical of Nordic culture.
Secondly, studies of early metallurgy and its products, both imported and created in the SE Baltic, have concluded that metal consumption in the LBA had more than doubled compared to the EBA (Sidrys and Luchtanas 1999). The SE Baltic region lacks any metal artefact types exclusive to the region and metal objects are dominated by artefact types originating from Nordic and Lusatian cultures (Sidrys and Luchtanas 1999; Lang 2007; Čivilytė 2014). This indicates that even after metal crafting reached the region, the technology remained exclusively of foreign origin. Rarely identifiable negatives of clay casting moulds were also made for artefacts of Nordic influence, such as Mälar type axes or Härnevi type pins (Čivilytė 2014; Sperling 2014).
Lastly, emerging social diversification was accompanied by the establishment of the first identifiable settlement pattern. Settlement locations were strategically chosen alongside economically significant routes, primarily on the coast and near the Daugava River. Hilltop areas were prioritized over the lowlands, and excavations on these sites have often revealed several stages of enclosure construction (Graudonis 1989). This has also been explained as a reflection of intensifying communication networks between Nordic and Lusatian cultures, and the indigenous communities of the SE Baltic.
Kortlandt’s position regarding Balto-Slavic is that it is in fact simply ‘Proto-Baltic’, a language that would stem thus from an Indo-Baltic branch, which would be originally represented by Corded Ware, and which would have split suddenly in its three dialects without any common development between branches, including some intermediate hypothetic “Centum” Temematic substrate that would explain everything his model can’t…
The site of Turlojiškė in southern Lithuania (ca. 908-485 BC) – which Mittnik et al. (2018) classified as “Bronze Age, Trzciniec culture?” – can be more reasonably considered a settlement of incoming intensive agrarian communities under the influence of the Lusatian culture, like the Narkūnai hilltop settlement in eastern Lithuania (ca. 800–550 BC), or the enclosed hilltop settlement of Kukuliškiai in western Lithuania (ca. 887-506 BC), just 300 m east of the Baltic Sea, also referred to in the paper.
While the dates of sampled individuals include a huge span (ca. 2100-600 BC), those with confirmed radiocarbon dates are more precisely dated to the LBA-EIA transition. More specifically, the first clearly western influence is seen in the early outlier Turlojiškė1932 (ca. 1230-920 BC), while later samples and samples from Kivutkalns, in Latvia, show major genetic continuity with indigenous populations, compatible with the new chiefdom-based systems of the Baltic and the known lack of massive migrations to the region.
Contacts with western groups of the Nordic Bronze Age and Lusatian cultures intensified – based on existing archaeological and archaeobotanical evidence – in the LBA, especially from ca. 1100/1000 BC on, and Baltic languages seem to have thus little to do with the disappearing Trzciniec culture, and more with the incoming Lusatian influence.
Both facts – more simple dialectalization scheme, and more recent Indo-European expansion to the east – support the spread of Proto-Baltic into the south-east Baltic area precisely around this time, and is also compatible with an internal separation from Proto-Slavic during the expansion of the Lusatian culture.
Even though comparative grammar is traditionally known to be wary of resorting to language contamination or language contact, the truth is that – very much like population genomics – trying to draw a ‘pure’ phylogenetic tree for Balto-Slavic has never worked very well, and the most likely culprit is the Slavic expansion to the south-east into territories which underwent different and complex genetic and linguistic influences for centuries (see here and here).
The relative chronology of hydrotoponymy in the East Baltic shows that essentially all ancestral layers to the north of the Daugava must have been Uralic, while roughly south of the Daugava they seem to be mostly Indo-European. The question remains, though, when did this Indo-European layer start?
Interestingly, though, it is well known that some modern Baltic toponyms can’t be easily distinguished from the Old European layers – unlike those of Iberia or the British Isles, which show some attested language change in the proto-historical and historical period – which may imply both (a) continuity of Baltic languages since the EBA, but also that (b) the Baltic naming system is a confounding factor in assessing the ancestral expansion of Old European. The latter is becoming more and more likely with each new linguistic, archaeological, and genetic paper.
In summary, a survival of a hypothetical late Trzciniec language in Lithuania or as part of the expanding Lusatian community is not the most economic explanation for what is seen in genetics and archaeology. On the other hand, the cluster formed by the Tollense samples (a site corresponding to the Nordic Bronze Age), the Turlojiškė outlier, and the early Slavs from Bohemia all depict an eastward expansion of Balto-Slavic languages from Central Europe, at the same time as Celtic expanded to the west with the Urnfield culture.
NOTE. Another, more complicated question, though, is if this expanding Proto-Baltic language accompanying agriculture represents the extinct
early Proto-Baltic dialect from which Balto-Finnic borrowed words, hence Proto-Baltic proper expanded later, or if this early Baltic branch could have been part of the Trzciniec expansion. Again, the answer in archaeological and genetic terms seems to be the former. For a more detailed discussion of this and more, see European hydrotoponymy (IV): tug of war between Balto-Slavic and West Uralic.
As I said recently, the slight increase in Corded Ware-like ancestry among Iron Age Estonians, if it were statistically relevant and representative of an incoming population – and not just the product of “usual” admixture with immediate neighbours – need not be from south-eastern Corded Ware groups, because the Akozino-Malär cultural exchange seems to have happened as an interaction in both directions, and not just as an eastward migration imagined by Carpelan and Parpola.
Archaeology and genetics could actually suggest then (at least in part) an admixture with displaced indigenous West Uralic-speaking peoples from the south-west, to the south of the Daugava River, at the same time as the Indo-European – Uralic language frontier must have shifted to its traditional location, precisely during the LBA / EIA transition around 1000 BC.
The tight relationship of the three communities also accounts for the homogeneous distribution of expanding haplogroup N1c-VL29 (possibly associated with Akozino warrior-traders) in the whole Baltic Sea area, such as those appearing in the Estonian Iron Age samples, which have no clearly defined route(s) of expansion.
Fortunately, the current obsession with simplifying ancestry components into three or four general, atemporal groups, and the common use of the same ones across labs, make it very simple to merge data and map them.
Corded Ware ancestry
There is no doubt about the prevalent ancestry among Uralic-speaking peoples. A map isn’t needed to realize that, because ancient and modern data – like those recently summarized in Jeong et al. (2019) – prove it. But maps sure help visualize their intricate relationship better:
Edit (29/7/2019): Here is the full Steppe_MLBA ancestry map, including Steppe_MLBA (vs. Indus Periphery vs. Onge) in modern South Asian populations from Narasimhan et al. (2018), apart from the ‘Srubnaya component’ in North Eurasian populations. ‘Dummy’ variables (with 0% ancestry) have been included to the south and east of the map to avoid weird interpolations of Steppe_MLBA into Africa and East Asia.
Anatolia Neolithic ancestry
Also interesting are the patterns of non-CWC-related ancestry, in particular the apparent wedge created by expanding East Slavs, which seems to reflect the intrusion of central(-eastern) European ancestry into Finno-Permic territory.
The cline(s) between WHG, EHG, ANE, Nganasan, and Baikal HG are also simplified when some of them excluded, in this case EHG, represented thus in part by WHG, and in part by more eastern ancestries (see below).
Arctic, Tundra or Forest-steppe?
Data on Nganasan-related vs. ANE vs. Baikal HG/Ulchi-related ancestry is difficult to map properly, because both ancestry components are usually reported as mutually exclusive, when they are in fact clearly related in an ancestral cline formed by different ancient North Eurasian populations from Siberia.
When it comes to ascertaining the origin of the multiple CWC-related clines among Uralic-speaking peoples, the question is thus how to properly distinguish the proportions of WHG-, EHG-, Nganasan-, ANE or BaikalHG-related ancestral components in North Eurasia, i.e. how did each dialectal group admix with regional groups which formed part of these clines east and west of the Urals.
The truth is, one ought to test specific ancient samples for each “Siberian” ancestry found in the different Uralic dialectal groups, but the simplistic “Siberian” label somehow gets a pass in many papers (see a recent example).
Below qpAdm results with best fits for Ulchi ancestry, Afontova Gora 3 ancestry, and Nganasan ancestry, but some populations show good fits for both and with similar proportions, so selecting one necessarily simplifies the distribution of both.
A simplistic Iran Chalcolithic-related ancestry is also seen in the Altaic cline(s) which (like Corded Ware ancestry) expanded from Central Asia into Europe – apart from its historical distribution south of the Caucasus:
The first question I imagine some would like to know is: what about other models? Do they show the same results? Here is the simplistic combination of ancestry components published in Damgaard et al. (2018) for the same or similar populations:
NOTE. As you can see, their selection of EHG vs. WHG vs. Nganasan vs. Natufian vs. Clovis of is of little use, but corroborate the results from other papers, and show some interesting patterns in combination with those above.
Baikal HG ancestry
Ancient North Eurasians
Once the modern situation is clear, relevant questions are, for example, whether EHG-, WHG-, ANE, Nganasan-, and/or Baikal HG-related meta-populationsexpanded or became integrated into Uralic-speaking territories.
When did these admixture/migration events happen?
How did the ancient distribution or expansion of Palaeo-Arctic, Baikalic, and/or Altaic peoples affect the current distribution of the so-called “Siberian” ancestry, and of hg. N1a, in each specific population?
NOTE. A little excursus is necessary, because the calculated repetition of a hypothetic opposition “N1a vs. R1a” doesn’t make this dichotomy real:
There was not a single ethnolinguistic community represented by hg. R1a after the initial expansion of Eastern Corded Ware groups, or by hg. N1a-L392 after its initial expansion in Siberia:
Different subclades became incorporated in different ways into Bronze Age and Iron Age communities, most of which without an ethnolinguistic change. For example, N1a subclades became incorporated into North Eurasian populations of different languages, reaching Uralic- and Indo-European-speaking territories of north-eastern Europe during the late Iron Age, at a time when their ancestral origin or language in Siberia was impossible to ascertain. Just like the mix found among Proto-Germanic peoples (R1b, R1a, and I1)* or among Slavic peoples (I2a, E1b, R1a)*, the mix of many Uralic groups showing specific percentages of R1a, N1a, or Q subclades* reflect more or less recent admixture or acculturation events with little impact on their languages.
*other typically northern and eastern European haplogroups are also represented in early Germanic (N1a, I2, E1b, J, G2), Slavic (I1, G2, J) and Finno-Permic (I1, R1b, J) peoples.
The problem with mapping the ancestry of the available sampling of ancient populations is that we lack proper temporal and regional transects. The maps that follow include cultures roughly divided into either “Bronze Age” or “Iron Age” groups, although the difference between samples may span up to 2,000 years.
NOTE. Rough estimates for more external groups (viz. Sweden Battle Axe/Gotland_A for the NW, Srubna from the North Pontic area for the SW, Arctic/Nganasan for the NE, and Baikal EBA/”Ulchi-like” for the SE) have been included to offer a wider interpolated area using data already known.
Similar to modern populations, the selection of best fit “Siberian” ancestry between Baikal HG vs. Nganasan, both potentially ± ANE (AG3), is an oversimplification that needs to be addressed in future papers.
NOTE. The samples from Levänluhta are centuries older than those from Estonia (and Ingria), and those from Chalmny Varre are modern ones, so this region has to be read as a south-west to north-east distribution from the Iron Age to modern times.
Baikal HG-like ancestry
The fact that this Baltic N1a-VL29 branch belongs in a group together with typically Avar N1a-B197 supports the Altaic origin of the parent group, which is possibly related to the expansion of Baikalic ancestry and Iron Age nomads:
The dilution of Nganasan-like ancestry in an Arctic region featuring “Siberian” ancestry and hg. N1a-L392 at least since the Bronze Age supports the integration of hg. N1a-Z1934, sister clade of Ugric N1a-Z1936, into populations west and east of the Urals with the expansion of Uralic languages to the north into the Tundra region (see here).
The integration of N1a-Z1934 lineages into Finnic-speaking peoples after their migration to the north and east, and the displacement or acculturation of Saami from their ancestral homeland, coinciding with known genetic bottlenecks among Finns, is yet another proof of this evolution:
Similarly, WHG ancestry doesn’t seem to be related to important population movements throughout the Bronze Age, which excludes the multiple North Eurasian populations that will be found along the clines formed by WHG, EHG, ANE, Nganasan, Baikal HG ancestry as forming part of the Uralic ethnogenesis, although they may be relevant to follow later regional movements of specific populations.
It seems natural that people used to look at maps of haplogroup distribution from the 2000s, coupled with modern language distributions, and would try to interpret them in a certain way, reaching thus the wrong conclusions whose consequences are especially visible today when ancient DNA keeps contradicting them.
The evolution of each specific region and cultural group of North Eurasia is far from being clear. However, the general trend speaks clearly in favour of an ancient, Bronze Age distribution of North Eurasian ancestry and haplogroups that have decreased, diluted, or become incorporated into expanding Uralians of Corded Ware ancestry, occasionally spreading with inter-regional expansions of local groups.
Given the relatively recent push of Altaic and Indo-European languages into ancestral Uralic-speaking territories, only the ancient Corded Ware expansion remains compatible with the spread of Uralic languages into their historical distribution.
Interesting excerpts (emphasis mine, modified for clarity):
To understand the genetic structure and influence of the Viking expansion, we sequenced the genomes of 442 ancient humans from across Europe and Greenland ranging from the Bronze Age (c. 2400 BC) to the early Modern period (c. 1600 CE), with particular emphasis on the Viking Age. We find that the period preceding the Viking Age was accompanied by foreign gene flow into Scandinavia from the south and east: spreading from Denmark and eastern Sweden to the rest of Scandinavia. Despite the close linguistic similarities of modern Scandinavian languages, we observe genetic structure within Scandinavia, suggesting that regional population differences were already present 1,000 years ago.
Maps illustrating the following texts have been made based on data from this and other papers:
Maps showing ancestry include only data from this preprint (which also includes some samples from Sigtuna).
Maps showing haplogroups of ancient DNA samples based on their age include data from all published papers, but with slightly modified locations to avoid overcrowding (randomized distance approx. ± 0.1 long. and lat.).
We find that the transition from the BA to the IA is accompanied by a reduction in Neolithic farmer ancestry, with a corresponding increase in both Steppe-like ancestry and hunter-gatherer ancestry. While most groups show a slight recovery of farmer ancestry during the VA, there is considerable variation in ancestry across Scandinavia. In particular, we observe a wide range of ancestry compositions among individuals from Sweden, with some groups in southern Sweden showing some of the highest farmer ancestry proportions (40% or more in individuals from Malmö, Kärda or Öland).
Ancestry proportions in Norway and Denmark on the other hand appear more uniform. Finally we detect an influx of low levels of “eastern” ancestry starting in the early VA, mostly constrained among groups from eastern and central Sweden as well as some Norwegian groups. Testing of putative source groups for this “eastern” ancestry revealed differing patterns among the Viking Age target groups, with contributions of either East Asian- or Caucasus-related ancestry.
Overall, our findings suggest that the genetic makeup of VA Scandinavia derives from mixtures of three earlier sources: Mesolithic hunter-gatherers, Neolithic farmers, and Bronze Age pastoralists. Intriguingly, our results also indicate ongoing gene flow from the south and east into Iron Age Scandinavia. Thus, these observations are consistent with archaeological claims of wide-ranging demographic turmoil in the aftermath of the Roman Empire with consequences for the Scandinavian populations during the late Iron Age.
Genetic structure within Viking-Age Scandinavia
We find that VA Scandinavians on average cluster into three groups according to their geographic origin, shifted towards their respective present-day counterparts in Denmark, Sweden and Norway. Closer inspection of the distributions for the different groups reveals additional complexity in their genetic structure.
We find that the ‘Norwegian’ cluster includes Norwegian IA individuals, who are distinct from both Swedish and Danish IA individuals which cluster together with the majority of central and eastern Swedish VA individuals. Many individuals from southwestern Sweden (e.g. Skara) cluster with Danish present-day individuals from the eastern islands (Funen, Zealand), skewing towards the ‘Swedish’ cluster with respect to early and more western Danish VA individuals (Jutland).
Some individuals have strong affinity with Eastern Europeans, particularly those from the island of Gotland in eastern Sweden. The latter likely reflects individuals with Baltic ancestry, as clustering with Baltic BA individuals is evident in the IBS-UMAP analysis and through f4-statistics.
Genetic clustering using IBS-UMAP suggested genetic affinities of some Viking Age individuals with Bronze Age individuals from the Baltic. To further test these, we quantified excess allele sharing of Viking Age individuals with Baltic BA compared to early Viking Age individuals from Salme using f4 statistics. We find that many individuals from the island of Gotland share a significant excess of alleles with Baltic BA, consistent with other evidence of this site being a trading post with contacts across the Baltic Sea.
The earliest N1a-VL29 sample available comes from Iron Age Gotland (VK579) ca. AD 200-400 (see Iron Age Y-DNA maps), which also proves its presence in the western Baltic before the Viking expansion. The distribution of N1a-VL29 and R1a-Z280 (compared to R1a in general) among Vikings also supports a likely expansion of both lineages in succeeding waves from the east with Akozino warrior-traders, at the same time as they expanded into the Gulf of Finland.
Vikings in Estonia
(…) only one Viking raiding or diplomatic expedition has left direct archaeological traces, at Salme in Estonia, where 41 Swedish Vikings who died violently were buried in two boats accompanied by high-status weaponry. Importantly, the Salme boat-burial predates the first textually documented raid (in Lindisfarne in 793) by nearly half a century. Comparing the genomes of 34 individuals from the Salme burial using kinship analyses, we find that these elite warriors included four brothers buried side by side and a 3rd degree relative of one of the four brothers. In addition, members of the Salme group had very similar ancestry profiles, in comparison to the profiles of other Viking burials. This suggests that this raid was conducted by genetically homogeneous people of high status, including close kin. Isotope analyses indicate that the crew descended from the Mälaren area in Eastern Sweden thus confirming that the Baltic-Mid-Swedish interaction took place early in the VA.
N1a-VL29 lineages spread again later eastwards with Varangians, from Sweden into north-eastern Europe, most likely including the ancestors of the Rurikid dynasty. Unsurprisingly, the arrival of Vikings with Swedish ancestry into the East Baltic and their dispersal through the forest zone didn’t cause a language shift of Balto-Finnic, Mordvinic, or East Slavic speakers to Old Norse, either…
NOTE. For N1a-Y4339 – N1a-L550 subclade of Swedish origin – as main haplogroup of modern descendants of Rurikid princes, see Volkov & Seslavin (2019) – full text in comments below. Data from ancient samples show varied paternal lineages even among early rulers traditionally linked to Rurik’s line, which explains some of the discrepancies found among modern descendants:
A sample from Chernihiv (VK542) potentially belonging to Gleb Svyatoslavich, the 11th century prince of Tmutarakan/Novgorod, belongs to hg. I2a-Y3120 (a subclade of early Slavic I2a-CTS10228) and has 71% “Modern Polish” ancestry (see below).
Izyaslav Ingvarevych, the 13th century prince of Dorogobuzh, Principality of Volhynia/Galicia, is probably behind a sample from Lutsk (VK541), and belongs to hg. R1a-L1029 (a subclade of R1a-M458), showing ca. 95% of “Modern Polish” ancestry.
Firstly, modern Finnish individuals are not like ancient Finnish individuals, modern individuals have ancestry of a population not in the reference; most likely Steppe/Russian ancestry, as Chinese are in the reference and do not share this direction. Ancient Swedes and Norwegians are more extreme than modern individuals in PC2 and 4. Ancient UK individuals were more extreme than Modern UK individuals in PC3 and 4. Ancient Danish individuals look rather similar to modern individuals from all over Scandinavia. By using a supervised ancient panel, we have removed recent drift from the signal, which would have affected modern Scandinavians and Finnish populations especially. This is in general a desirable feature but it is important to check that it has not affected inference.
The story for Modern-vs-ancient Finnish ancestry is consistent, with ancient Finns looking much less extreme than the moderns. Conversely, ancient Norwegians look like less-drifted modern Norwegians; the Danish admixture seen through the use of ancient DNA is hard to detect because of the extreme drift within Norway that has occurred since the admixture event. PC4 vs PC5 is the most important plot for the ancient DNA story: Sweden and the UK (along with Poland, Italy and to an extent also Norway) are visibly extremes of a distribution the same “genes-mirror-geography” that was seen in the Ancient-palette analysis. PC1 vs PC2 tells the same story – and stronger, since this is a high variance-explained PC – for the UK, Poland and Italy.
Evidence for Pictish Genomes
The four ancient genomes of Orkney individuals with little Scandinavian ancestry may be the first ones of Pictish people published to date. Yet a similar (>80% “UK ancestry) individual was found in Ireland (VK545) and five in Scandinavia, implying that Pictish populations were integrated into Scandinavian culture by the Viking Age.
Our interpretation for the Orkney samples can be summarised as follows. Firstly, they represent “native British” ancestry, rather than an unusual type of Scandinavian ancestry. Secondly, that this “British” ancestry was found in Britain before the Anglo-Saxon migrations. Finally, that in Orkney, these individuals would have descended from Pictish populations.
(…) ‘UK’ represents a group from which modern British and Irish people all receive an ancestry component. This information together implies that within the sampling frame of our data, they are proxying the ‘Briton’ component in UK ancestry; that is, a pre-Roman genetic component present across the UK. Given they were found in Orkney, this makes it very likely that they were descended from a Pictish population.
Modern genetic variation within the UK sees variation between ‘native Briton’ populations Wales, Scotland, Cornwall and Ireland as large compared to that within the more ‘Anglo-Saxon’ English. This is despite subsequent gene flow into those populations from English-like populations. We have not attempted to disentangle modern genetic drift from historically distinct populations. Roman-era period people in England, Wales, Ireland and Scotland may not have been genetically close to these Orkney individuals, but our results show that they have a shared genetic component as they represent the same direction of variation.
As in the case of mitochondrial DNA, the overall distribution profile of the Y chromosomal haplogroups in the Viking Age samples was similar to that of the modern North European populations. The most frequently encountered male lineages were the haplogroups I1, R1b and R1a.
Haplogroup I (I1, I2)
The distribution of I1 in southern Scandinavia, including a sample from Sealand (VK532) ca. AD 100 (see Iron Age Y-DNA maps) proves that it had become integrated into the West Germanic population already before their expansions, something that we already suspected thanks to the sampling of Germanic tribes.
Haplogroup R1b (M269, U106, P312)
Especially interesting is the finding of R1b-L151 widely distributed in the historical Nordic Bronze Age region, which is in line with the estimated TMRCA for R1b-P312 subclades found in Scandinavia, despite the known bottleneck among Germanic peoples under U106. Particularly telling in this regard is the finding of rare haplogroups R1b-DF19, R1b-L238, or R1b-S1194. All of that points to the impact of Bell Beaker-derived peoples during the Dagger period, when Pre-Proto-Germanic expanded into Scandinavia.
Also interesting is the finding of hg. R1b-P297 in Troms, Norway (VK531) ca. 2400 BC. R1b-P297 subclades might have expanded to the north through Finland with post-Swiderian Mesolithic groups (read more about Scandinavian hunter-gatherers), and the ancestry of this sample points to that origin.
However, it is also known that ancestry might change within a few generations of admixture, and that the transformation brought about by Bell Beakers with the Dagger Period probably reached Troms, so this could also be a R1b-M269 subclade. In fact, the few available data from this sample show that it comes from the natural harbour Skarsvågen at the NW end of the island Senja, and that its archaeologist thought it was from the Viking period or slightly earlier, based on the grave form. From Prescott (2017):
In 1995, Prescott and Walderhaug tentatively argued that a dramatic transformation took place in Norway around the Late Neolithic (2350 BCE), and that the swift nature of this transition was tied to the initial Indo-Europeanization of southern and coastal Norway, at least to Trøndelag and perhaps as far north as Troms. (…)
The Bell Beaker/early Late Neolithic, however, represents a source and beginning of these institution and practices, exhibits continuity to the following metal age periods and integrated most of Northern Europe’s Nordic region into a set of interaction fields. This happened around 2400 BCE, at the MNB to LN transition.
NOTE. This particular sample is not included in the maps of Viking haplogroups.
Among the ancient samples, two individuals were derived haplogroups were identified as E1b1b1-M35.1, which are frequently encountered in modern southern Europe, Middle East and North Africa. Interestingly, the individuals carrying these haplogroups had much less Scandinavian ancestry compared to the most samples inferred from haplotype based analysis. A similar pattern was also observed for less frequent haplogroups in our ancient dataset, such as G (n=3), J (n=3) and T (n=2), indicating a possible non-Scandinavian male genetic component in the Viking Age Northern Europe. Interestingly, individuals carrying these haplogroups were from the later Viking Age (10th century and younger), which might indicate some male gene influx into the Viking population during the Viking period.
As the paper says, the small sample size of rare haplogroups cannot distinguish if these differences are statistically relevant. Nevertheless, both E1b samples have substantial Modern Polish-like ancestry: one sample from Gotland (VK474), of hg. E1b-L791, has ca. 99% “Polish” ancestry, while the other one from Denmark (VK362), of hg. E1b-V13, has ca. 35% “Polish”, ca. 35% “Italian”, as well as some “Danish” (14%) and minor “British” and “Finnish” ancestry.
Given the E1b-V13 samples of likely Central-East European origin among Lombards, Visigoths, and especially among Early Slavs, and the distribution of “Polish” ancestry among Viking samples, VK362 is probably a close description of the typical ancestry of early Slavs. The peak of Modern Polish-like ancestry around the Upper Pripyat during the (late) Viking Age suggests that Poles (like East Slavs) have probably mixed since the 10th century with more eastern peoples close to north-eastern Europeans, derived from ancient Finno-Ugrians:
Similarly, the finding of R1a-M458 among Vikings in Funen, Denmark (VK139), in Lutsk, Poland (VK541), and in Kurevanikha, Russia (VK160), apart from the early Slav from Usedom, may attest to the origin of the spread of this haplogroup in the western Baltic after the Bell Beaker expansion, once integrated in both Germanic and Balto-Slavic populations, as well as intermediate Bronze Age peoples that were eventually absorbed by their expansions. This contradicts, again, my simplistic initial assessment of R1a-M458 expansion as linked exclusively (or even mainly) to Balto-Slavs.
The nature of the prehistoric languages of the British Isles is particularly difficult to address: because of the lack of ancient data from certain territories; because of the traditional interpretation of Old European names simply as “Celtic”; and because Vennemann’s re-labelling of the Old European hydrotoponymy as non-Indo-European has helped distract the focus away from the real non-Indo-European substrate on the islands.
Alteuropäisch and Celtic
An interesting summary of hydronymy in the British Isles was already offered long ago, in British and European River-Names, by Kitson, Transactions of the Philological Society (1996) 94(2):73-118. In it, he discusses, among others:
Non-serial hydronyms: Drua-/Drav-/Dru-, from drew- sometimes reshaped as derw-; ab-; ag-; al-; alb-; alm-; am-; antjā-; arg-; aw-; dan-; eis-; el-/ol-; er-/or-; kar(r)a-, ker-; nebh-; ned-; n(e)id-; sal-; wig-; weis-/wis-; ur-, wer-; etc.
Serial elements: -went-, -m(e)no-, -nt-o-, -n-; -nā-, -tā-; -st-, -r-; etc.
Probably non-Celtic suffixes are found e.g. in Tamesis, paralelled in the Spey Tuesis, and also in Tweed (<*Twesetā?); or -no-/-nā- is also particularly frequent in Scottish river-names, but not in English ones. Another interesting case is the reverse suffix relative order into -r-st- instead of -st-r-.
Most if not all of them can be explained as of Old European nature. I will leave aside the discussion of particular formations – most of which may be found repeated, complemented, and updated in more modern texts.
Bell Beakers as Old Europeans
(…) Bell-beakers are in fact the only archaeological phenomenon of any period of prehistory with a comparably wide spread to that of river-names in the western half of Europe. The presumption must I think be that Beaker Folk were the vector of alteuropäisch river-names to most of western Europe. Rivers in the base Arg-, which we have seen there is cause to think was not already in use at the earliest stage of the river-naming system, and which therefore should be associated with such a vector if one existed, fit their distribution exceptionally well.
That they were a single-speech community can be asserted more confidently of the Beaker Folk than of most archaeologically identified groups for the very reasons that have caused archaeologists difficulty in interpreting them. As McEvedy (1967:28) put it, ‘the bell-beaker folk march convincingly in every prehistorian’s text, but they do so from Spain to Germany in some and from Germany to Spain in others, while lately there has been a tendency to make them go from Spain to Germany and back again (primary and reflux movements)’. One ‘firm datum seems to be that the British beaker folk came from the Rhine-Elbe region.’
This confirms what the long chronology now indicated for Common Indo-European would suggest anyway, and what to me, as remarked above, the rareness of non-Indo-European names in England suggests, that the old dissenting minority of Celticists were right to see the arrival in Britain of Indo-Europeans, as evinced in river-names whether or not in ethnic proto-Celts, as early as the third millennium. McEvedy’s map of Beaker Folk identifies them linguistically with Celto-Ligurians, but in that his admirably tidy mind was, typically, a degree too tidy. Considerations of phonology indicate that more than one linguistic group was involved.
It is normal in reconstructed Indo-European for groups of related words not all to have the same vowel in the root syllable. The commonest vowel gradation is between e, o, and zero; (…) Language-groups that level short a and o include Germanic and Baltic, Slavonic, Illyrian, Hittite and Indo-Iranian; but Celtic and Italic like Greek and Armenian preserve the original distinction. It follows that Celts speaking normal Celtic sounds cannot have been wholly responsible for bringing alteuropäisch river-names to any area. It would seem to follow, as Professor Nicolaisen has consistently urged, that in Spain, Gaul, Britain, and Italy, where the only historically known early Indo-Europeans were speakers of non-levelling languages, they were preceded by speakers of levelling languages not historically known. This hypothesis, pretty well required by the linguistic evidence, finds so good an archaeological correlate in the Beaker People that I think it would now be flying in the face of the evidence not to accept those as bearers of the river-names to these countries.
The funny note is the rejection of the steppe homeland by Kitson in favour of Central European Neolithic cultures, due in part to the ‘impossibility’ of proto-Finnic loans from East Indo-European, if Proto-Indo-European was spoken in the steppe. As I said recently, the lack of knowledge of Uralic languages and Indo-European – Uralic contacts has clearly conditioned the Urheimat question for both, Proto-Indo-European and Proto-Uralic researchers.
The question is, though, to what extent the reasoning of those researchers was as detailed so as to consider it a modern approach to the question, because Krahe in the 1940s seems to offer the first reliable data to make that assumption. In any case, Gimbutas’ idea of Kurgan warriors imposing Indo-European languages everywhere, so over-represented in Encyclopedia-like texts since the end of the 1990s, was not the only, and probably never the main hypothesis among many Indo-Europeanists.
Celts part of Bell Beakers?
Regarding Koch and Cunliffe’s revival of the autochthonous Celts idea, one can find a similar traditional view among British researchers of the early and mid-20th century – and a proper rejection based on hydrotoponymy. It seems that many fringe theories in Indo-European studies, from Nordic or Baltic homelands to autochthonous Celts to the Europa Vasconica, can be traced back to revivalist waves of romantic views of the 19th c.:
What the late Professor C. F. C. Hawkes called in British archaeology ‘cumulative Celticity’, built up by successions of comparatively small tribal migrations, will then have operated on the linguistic side as well. That the predecessors of the Celts proper for so long had in most of Britain been people of similar Indo-European speech explains why there is not a significant survival of recognizably non-Indo-European river-names, and why the few serious candidates for non-Indo-European among recorded place-names all seem to be in Scotland. That the river-names kept their north European non-Celtic phonology will be because the Celts proper took them over as names, with denotative not fully lexical meaning. (…)
(…) I think non-Celtic Indo-European-speakers are likely to have been involved in fact, whether or not they are the whole story, both because that it is the hypothesis which makes best sense of the archaeological evidence (…)
(…) because it is widely accepted that placenames in the Low Countries imply the existence of at least one group of not historically attested Indo-European-speakers, not the same as the ones we are concerned with. So do names in Spain, another country where the only historically attested early Indo-Europeans were Celtic. Comparing Spanish alteuropäisch names with British ones gives a glimpse of the dialectal range that must have characterized the Beaker phenomenon. Either group shares one feature with historical Celtic that the other lacks. The Spanish names like Celtic proper mostly keep Inda-European o. There the diagnostic feature is initial p (Schmoll 1959:93, 78-80; Rodriguez 1980), lost from Celtic and the alteuropäisch of Britain.
Interesting is also the early reaction against Vennemann’s much publicized interpretation of Krahe’s Old European as ‘Vasconic’. This is a useful comment which is still applicable to the same non-existent ‘problem’ found by some Indo-Europeanists, depending on their ideas about Indo-European dialectalization:
It is again naughty of Vennemann (1994:244) to call his laryngealist explanation ‘the only kind of explanation that I know’. At least he does not quite go so far in his laryngealism as to posit a proto-Indo-European in which the vowel a never existed, as Kuiper does.
NOTE. It is difficult to understand why the work of so many Indo-Europeanists is usually not known, while Vennemann’s far-fetched theory has been endlessly repeated. I reckon it must be the same phenomenon of personal and professional contacts, involvement in editorial decisions, and simplification in mass media which makes Kristiansen and his theories frequently published and cited nowadays.
Based on these data, I entertained the idea of arguing for a Pre-Celtic Indo-European language in A Storm of Words, called Pre-Pritenic, with a tentative fable based on the data described below for the Insular Celtic substrate, but eventually deleted the whole text, because (unlike other tentative fables, like the Lusitanian or Venetic ones) it was pure speculation with not even fragmentary data to rely on. Here is a fragment of the discussion:
Among the main reasons adduced to reject the non-Celtic nature of Pritenic is Orkney, a region where Pictish carved stones have been found (indicator of a centralised Pictish power and identity). The name was attested first as Gk. Orkas / Orkádos (secondary source, from Pytheas of Massilia, ca. 322-285 BC, or possibly much later) and Lat. Orchades / Orcades (by Latin sources in the 1st century AD), and it was used to describe the northernmost promontory in Scotland, commonly identified as Dunnet Head in Caithness. It is supposed to derive its name from Cel. *φorko- ‘pig’, because speakers of Old Irish interpreted the name for the island later as Insi Orc ‘island of the pigs’. Therefore, Pritenic would have undergone the prototypical Common Celtic evolution of NWIE *p- → Ø- (see above).
This argument is flawed, in so far as it could have happened (with the interpretation of the name from a Celtic point of view) what happened later with Norwegian settlers, who reinterpreted the name according to Old Norse orkn ‘seal’, to identify it as ‘island of the seals’. In fact, texts published in the 19th and 20th century looked for an even closer etymology to the interpreter, who usually saw it as ‘island of the orcas’.
The region name orc- could be speculatively linked to NWIE *ork-i- ‘cut off, divide’, cf. Ita. *erk-i- (vowel analogically changed), Hitt. ārk- (<*hork-ei-), in Latin found with the meaning ‘divide (an inheritance)’, hence noun Lat. erctum ‘inheritance, inherited part’.
Maybe more interesting is a connection to *or-, as found in British rivers or streams Arrow, Oare Water (Som), Ayre , Armet Water, Arnot Burn, Ernan Water etc. for which cognates Skt. arvan(t)- ‘running, swift’, árṇa- ‘surging’, Gmc. *arnia- ‘lively, energetic’ have been proposed (Forster 1941; Nicolaisen 1976; Kitson 1996). Similar to these derivatives in -n-, -m-, one could argue for a denominative suffixation in *-ko-, not uncommon in Old European toponyms (Villar Liébana 2007), which could be interpreted originally as ‘(region) pertaining to the Or (river, stream)’. The a-vocalism of Old European does not need further explanation, being fairly common in the British Isles (Kitson 1996).
I tried to look for rivers and streams in Caithness that fit a potential border for an ancestral tribe, but after reading many (and I really mean too many) texts on Scotland’s hydronymy, which is a quite well-researched area, I didn’t like the idea of plunging into such a speculative task; not when I have this blog for that… I deleted the text from the book, seeing how it doesn’t really add anything of value and may have distracted from its real aim. If any reader wants to post potential candidates for this delimiting river ‘Or’ in Caithness, feel free to post that below.
Weak (if any) support of a non-Celtic nature of the names might also be found in the late description of Ptolemy’s Geographia (originally ca. 150 AD), Tauroedoúnou tēs kai Orkádos kaloumenēs, translated in Latin as Tarved(r)um, quod et Orcas promontorium dicitur. The original name seems to be formed from *tau-r-, as is common in Indo-European *taur-o- (compare also river Taum), whereas the commonly used Latin translation seems to rely on a Celtic *tarw-o-.
As with other Pictish material, these questions are unlikely to be settled without unequivocal sources pointing to the original names and their meaning. The autochthonous trend is set lately by Guto Rhys, whose work is thorough and methodologically sound, although his reviews tend to dismiss all evidence of a non-Celtic (or even non-Brittonic) layer in Pictland as described in previous works, mostly because of the lack of direct sources or uncontroverted data:
Where a supposed divergence is found in certain names, a lack of proper reading or interpretation of materials (or lack of enough cases to generalize them), combined with similar names in other (neighbouring or distant) Celtic languages, is adduced.
However, the same arguments can indeed be used to reject his proposal of a Celtic nature of many names which cannot be simply explained with other clearly Celtic examples: namely, that all similarities are due to later influences, re-analysis and modifications of Old European terms according to Celtic phonemic (or etymological) patterns, or that the Brittonic nature of many names are due to convergence of the attested Pritenic naming conventions with neighbouring dialects.
In the end, the only conclusion is that there is a clear impasse in hydrotoponymic research in the British Isles, particularly in Scotland, with an impossibility of describing non-Celtic or non-Indo-European Pre-Pritenic layers, due in great part – in my opinion – to the trend among many British Celticists to consider Celtic as autochthonous to the Atlantic. This hinders the proper investigation of the question, just like the trend among Basque studies to consider the western Pyrenees as the eternal Vasconic homeland hinders a fair investigation of the actual Vasconic proto-history.
The syntactic parallels between Insular Celtic and Afro-Asiatic languages (which used to be called Hamito-Semitic) were noted more than a century ago by Morris-Jones (1899), and subsequently discussed by a number of scholars. These parallels include the following.
The VSO order, attested both in OIr. and in Brythonic from the earliest documents (…).
The existence of special relative forms of the verb, (…).
The existence of prepositions inflected for person (or prepositional pronouns), (…).
Prepositional progressive verbal forms, (…).
The existence of the opposition between the “absolute” and “conjunct” verbal forms. (…)
The aforementioned features of Old Irish and Insular Celtic syntax (and a few others) are all found in Afro-Asiatic languages, often in several branches of that family, but usually in Berber and Ancient Egyptian (see e.g. Isaac 2001, 2007a).
Orin Gensler, in his unpublished dissertation (1993) applied refined statistical methods showing that the syntactic parallels between Insular Celtic and Afro-Asiatic cannot be attributed to chance. The crucial point is that these parallels include features that are otherwise rare cross-linguistically, but co-occur precisely in those two groups of languages. This more or less amounts to a proof that there was some connection between Insular Celtic and Afro-Asiatic at some stage in prehistory, but the exact nature of that connection is still open to speculation.
Insular Celtic also shares a number of areal isoglosses with languages of Western Africa, sometimes also with Basque, which shows that the Insular Celtic — Afroasiatic parallels should be viewed in light of the larger framework of prehistoric areal convergences in Western Europe and NW Africa.
The text goes on with typologically rare features found in West Europe and West Africa, such as the inter-dental fricative /þ/ (also in English, Icelandic, Castillian Spanish); initial consonant mutations/regular alterations of initial consonants caused by the grammatical category of the preceding word; the common order demonstrative-noun (within the NP) reversed; the vigesimal counting system; or use of demonstrative articles.
(…) only 38 words shared by Brythonic and Goidelic without any plausible IE etymology. These words belong to the semantic fields that are usually prone to borrowing, including words referring to animals (…), plants (…), and elements of the physical world (…). Note that cognates of these words may be unattested in Gaulish and Celtiberian because these languages are poorly attested, so that the actual number of exclusive loanwords from substratum language(s) in Insular Celtic is probably even lower. In my opinion it is not higher than 1% of the vocabulary. The large majority of substratum words in Irish and Welsh (and, generally, in Goidelic and Brythonic) is not shared by these two languages, which probably means that the sources were different substrates of, respectively, Ireland and Britain; (…)
The thesis that Insular Celtic languages were subject to strong influences from an unknown, presumably non-Indo-European substratum, hardly needs to be argued for. However, the available evidence is consistent with several different hypotheses regarding the areal and genetic affiliation of this substratum, or, more probably, substrata. The syntactic parallels between the Insular Celtic and Afro-Asiatic languages are probably not accidental, but they should not be taken to mean that the pre-Celtic substratum of Britain and Ireland belonged to the Afro-Asiatic stock. It is also possible that it was a language, or a group of languages (not necessarily related), that belonged to the same macro-area as the Afro-Asiatic languages of North Africa. The parallels between Insular Celtic, Basque, and the Atlantic languages of the Niger-Congo family, presented in the second part of this paper, are consistent with the hypothesis that there was a large linguistic macro-area, encompassing parts of NW Africa, as well as large parts of Western Europe, before the arrival of the speakers of Indo-European, including Celtic.
Even more interesting than the discussion of potential non-Indo-Europeans still lingering in Ireland until well into the Common Era, is the discussion on his paper Lost Languages in Northern Europe (2001). Apart from other non-Indo-European borrowings in northern Europe, most of which must clearly be included within the European agricultural substrate, Schrijver tries to interpret the relative chronology of a substratum language of northern Europe, described by Kuiper (1995) as A2, and by Schrijver as “language of geminates“.
This substrate language is heavily present in Germanic (see e.g. Boutkan 1998), but also in Celtic and Balto-Slavic:
A highly characteristic feature of words deriving from this language is the variation of the final root consonant, which may be single or double, voiced or voiceless, and prenasalized. (…)
Incidentally, the language of geminates cannot be Uralic, as another of its characteristics is the frequent occurrence of word-initial *kn- and *kl-, and Uralic languages do not allow consonant clusters at the beginning of the word. On the other hand, and at the risk of explaining obscura per obscuriora, one might consider the possibility that the consonant gradation of Lappish and Baltic Finnic is somehow connected with the alternation of consonants at the end of the first syllable in the “language of geminates”.
The idea that the Northern European language of geminates could play an intermediary role in loan contacts between Northern and Western Indo-European on the one hand and Finno-Ugric on the other may also account for the fact that Finno-Ugric words could end up as far away as Celtic, which as far as we know was never in direct contact with a branch of Uralic.
Schrijver later changed his view about certain aspects of this substrate, from a “language of geminates” influencing Balto-Finnic which in turn influenced Germanic, to Pre-Balto-Finnic speakers being the substrate of Germanic, and both evolving at the same time in contact in Scandinavia. In fact, we know that Pre-Proto-Germanic evolved in southern Scandinavia, with a core in Jutland that shifted to the south, so the location must have been close to the North European Plain.
Also fitting this model is the substrate behind Balto-Slavic (spoken in the West Baltic), which must have also been (Para-)Balto-Finnic. However, the frequent word-initial *kn- and *kl- and the loanwords appearing in the Celtic homeland (also including Early Balto-Finnic) must place this Uralic(± non-Indo-European) language contact also well into Central European Corded Ware groups.
The only archaeological culture that could fit most of these data, in the currently known relative chronological time frame, would be the Megalithic expansion in Western Europe, or potentially (maybe in addition to this early layer) the expansion of the Proto-Beaker package, which could have spread a Basque-Iberian language (see e.g. my take on Basque-Iberians).
Whether the language behind the Insular Celtic substrate (or, rather, some of its dialects) had true Afroasiatic syntactic features or it was just a language with features which happened to be similar to Afroasiatic is irrelevant. It’s impossible to reconstruct with confidence a Pre-Proto-Basque language with the currently available information.
NOTE. I will not resort here to typologically-based arguments similar to the “Hamito-Semit(id)ic” and “Vasconic-Uralic” Europe that were commonly in use in the 1990s, because they are in great part based on the mere re-labelling of Old European layers as “Vasconic” and flawed mass lexical/grammatical comparisons. For linguists favourable to this kind of reasoning, the theory set forth here is probably easier, though, as will be for those supporting a Neolithic expansion of Indo-European from the Mediterranean. This, however, has its own set of problems, as I have already discussed.
Single Grave culture
The non-Indo-European substrate of Insular Celtic, in combination with the oldest hydrotoponymic layers – almost exclusively of Old European nature – of Britain and likely all of Ireland, can more easily be explained as a first layer of North-West Indo-European speakers heavily influenced by an Afroasiatic(-like) substrate reaching the British Isles, possibly with a slightly richer set of non-Indo-European loanwords at the time. Their language would have been later replaced by the closely related Celtic dialects imposed by elites in the Early Iron Age, which could have then easily absorbed this (mainly syntactic) substrate.
There is little space to argue for a hypothetic non-Indo-European expansion from another region, or for an in situ substrate, due to:
the presence of the same (mainly syntactic) substrate in both Goidelic and Brittonic; and
the minimal non-Indo-European lexical borrowings and hydrotoponymy, different in each island;
Based on archaeological and palaeogenomic data, the only reasonable direct connection of north-western Bell Beakers and this substrate language would be then the Corded Ware groups from north-western Europe – i.e. the traditionally named Single Grave culture from northern Germany and Denmark, and the Protruding Foot Beaker culture from the Netherlands.
The main reasons for this are as follows:
1. Early Corded Ware wave
The earliest Corded Ware burials from northern Europe (ca. 2900-2800 BC) show important differences, so no strict funerary norms existed at first (Furholt 2014):
In southern Sweden the prevailing orientation is north-east–south-west, and south–north; contrary to the supposed rule, male individuals are regularly deposite on their left and females on their right side
In the Danish Isles and north-eastern Germany, the Final Neolithic / Single Grave Period is characterized by a majority of megalithic graves, with only some single graves from typical barrows.
In south Germany, west–east and collective burials prevail, while in Switzerland no graves are found.
In Kuyavia (south-eastern Poland), Hesse (Germany), or the Baltic, west–east orientation and gender differentiation cannot be proven statistically.
In genetics, the area that would become the ‘core Corded Ware province’ only after ca. 2700 BC also shows a surprising variability in the oldest samples in terms of haplogroups (which may indicate a recent departure of migrants from a mixed homeland); in terms of admixture, at least one sample clusters close to EEF groups, while later ones from Esperstedt – of hg. R1a-M417 (possibly xZ645) – show a likely admixture with Yamna vanguard groups expanding from the Carpathian Basin.
The Corded Ware culture in Denmark was particularly weak in its human impact compared to previous farmers (see e.g. Feeser et al. 2019), and also in its cultural traits, adopting Funnel Beaker culture traits up to a point where even the Copenhagen group describes cultural continuity, likely entailing an important substrate language impact (see e.g. Iversen and Kroonen 2017).
As it appears from the analysis above, the situation in East Denmark during the 3rd millennium BC is culturally rather complex. The continued use of megalithic entombments and the almost total rejection of the Single Grave burial custom show a strong affiliation with old Funnel Beaker traditions even after the end of the Funnel Beaker culture. (…) With an almost total lack of the two defining elements of the Single Grave culture – interments in single graves and the prominent position of stone battle axes – one can hardly talk about a Single Grave culture in East Denmark. What we see is rather the adoption of various Single Grave, Battle Axe and Pitted Ware cultural traits into a setting that was basically a continuation of Funnel Beaker norms and traditions (Iversen 2015).
The reason why East Denmark so conservatively upheld the Funnel Beaker traditions must be found in the area’s old position as a ‘megalithic heartland’, which reaches back to the early 4th millennium BC when dolmens and passage graves were constructed in very large numbers. (…) The result was a cultural blend governed by old Funnel Beaker norms and the use of Pitted Ware, Single Grave and Battle Axe material culture. This situation continued until the beginning of the Late Neolithic (ca. 2350 BC) when cultural and social development took a new course and flint daggers and metal objects appeared/ re-appeared in South Scandinavia.
The Corded Ware culture in the Netherlands is particularly disconnected culturally from its eastern core areas, which is reflected in the likely survival of a non-Indo-European language around the Low Countries, in the so-called Nordwestblock area. From Kroon et al. (2019):
The connections between changes in ceramic production techniques and social changes (see Fig. 2) allow for the formulation of hypotheses about the technological impact of the scenarios that archaeologists have proposed for the introduction of the CWC. If migration (i.e. an influx of new communities that bring new material culture) causes the spread of the CWC, then CWC vessels should differ from the vessels of previous communities in all respects: resilient, group-related, and salient techniques. However, if the introduction of the CWC is the result of diffusion of stylistic traits and moving objects, both these imported objects (different raw materials and production sequences) and changes in salient techniques should be observed when comparing CWC vessels to VLC vessels. Network interactions should yield the same changes as diffusion, as the combined movement of people, objects and styles within existing networks leads to the introduction of CWC. However, network interactions should yield one additional characteristic. Given that new people are integrated into extant communities, the occurrence of vessels with different resilient techniques, but group-related techniques that are stable relative to previous communities, is to be expected.
The over-arching transitional process in the Western coastal area of the Netherlands is local continuity with diffusion and network interaction traits. Interestingly, the supra-regional networks of the VLC communities in this region, as well as some of the defining technological practices within these networks, remain intact throughout the CWC transition.
In the absence of detailed genetic and isotopic data from Late Neolithic individuals from the western coastal areas of the Netherlands, direct conclusions on the relations between the migrations demonstrated by genetic analyses in other regions and the outcomes of this study remain speculative. However, if a similar shift in the late Neolithic gene pool from this area can be detected, this raises questions on the impact of such migrations on knowledge transmission and local traditions. If such a change cannot be attested, questions should be raised about the nature of the CWC in this particular area. Questions that will ultimately boil down to what we define as CWC.
In other words, the introduction of Corded Ware in the Netherlands, which we can assume were driven by migrations – evidenced by the arrival of “Steppe ancestry” (see below) – would need to be interpreted in light of the adoption of a different set of cultural traits in this region. Combining linguistic and archaeological data, there is strong evidence that the Corded Ware ideology and its internal coherence might have been broken in the westernmost territories, hence the likely survival of the local culture and language(s).
Further reasons for this independence from the Uralic homeland, supporting the advantages of a cultural and linguistic integration among regional groups, include:
This predominant non-Indo-European language would later be the substrate language of Bell Beakers from the Lower Rhine and the British Isles.
Culturally, the same process as in the previous Single Grave culture period may have happened in the Low Countries, due to the culturally favorable situation there. This might be inferred from the continuity of Protruding Foot Beaker into All-Over Ornamented Beaker, most likely an imitation of the expanding Proto-Beaker package by locals of the Single Grave culture.
Arguably, though, the same situation should have happened in all other Proto-Beaker regions favourable to cultural change and witnessing admixture with locals, such as Iberia, and the social relevance of this imitation is far from being accepted by almost anyone except for archaeologists working around the Rhine… From Heise (2014):
While in 1955 the Maritime Beaker was considered to be intrusive, the 1976 work seemed to prove that in the Netherlands a continuous development from Protruding Foot Beaker (PFB) to All-Over Ornamented (AOO) Beaker to Maritime Beaker occurred. Nevertheless, the authors stressed that it was not possible to identify ‘the’ origin of the ‘Bell Beaker Culture’ in the Lower Rhine Area since typical artefacts (wristguards, daggers) were not known to be associated with the early AOO and Maritime pottery. Furthermore they argued against the “misleading simplification” of a single point of origin (Lanting & van der Waals 1976, 2). However, this last observation was not appreciated or was simply ignored by large parts of the research community and the theory was subsequently applied as a universal solution in many parts of Europe.
In fact, most archaeologists have unequivocally rejected a Single Grave – Classical Bell Beaker continuity, and Heyd’s model has been recently confirmed in paleogenomics, which shows an evident expansion of East Bell Beakers from Yamna settlers in the Carpathian Basin (see here). We may nevertheless still save the following assertion, as particularly relevant for the continuity of non-Indo-European languages among the Single Grave groups of the Lower Rhine:
Marc Vander Linden argued that the “local validity of the Dutch sequence cannot […] be questioned” (2012, 76).
Olalde et al. (2019) showed how British, Dutch, and French Beakers have excess “Steppe ancestry” relative to Central European Beakers from Germany, who are in turn closest to the origin of Old Europeans in Iberia (i.e. Galaico-Lusitanian, “Ligurian”), the Lower Danube (i.e. Celtic), Italy (i.e. Italic, Venetic, Messapic), Sicily, and even Denmark (i.e. Germanic). This excess “Steppe ancestry” probably implies admixture with local Single Grave populations of the Lower Rhine, which is further supported by the position of these Lower Rhine Beakers in the PCA (using British Beakers and Netherlands BA as proxies), clustering – among Bell Beakers – closest to Corded Ware samples.
Futhermore, the emergence of Bell Beakers in the British Isles represents a radical replacement, with a population turnover of ca. 90% of the local population, and Yamna lineages representing more than 90% of the haplogroups of individuals in Chalcolithic and Bronze Age Britain and Ireland, apart from an evident Y-chromosome bottleneck under hg. R1b-S461 (and its subclade R1b-L21), maintained during the whole Bronze Age. The scarce non-Indo-European hydrotoponymy attests to the lack of integration of local populations or their languages into the new society. All this suggests an initial swift and massive intrusion marking the linguistic evolution of the British Isles until the Iron Age.
The arrival of Insular Celtic in the British Isles will be likely defined by an increase in ancestry related to Central Europe (and probably haplogroups, too). Since the Afroasiatic-like substrate is unrelated to Common Celtic, the non-Indo-European substrate must be associated with preceding Bronze Age populations of western Europe, most likely with Bronze Age Britons, who are in turn derived from Bell Beakers from the Lower Rhine admixed with Single Grave peoples. The latter, therefore, must have passed on their Afroasiatic-like language as the substrate of Lower Rhine Beakers.
5. Vasconic from the north
Another indirect proof to the survival of non-Indo-Europeans in northern Europe is offered by Basques. Vasconic speakers came originally from some place beyond Aquitaine, and very recently before the Roman conquests, because place- and river-names show an overwhelming Old European substratum to the north of the Pyrenees, and exclusively Old European to the south.
Their origin is potentially quite far away, since Modern Basques show a similar cluster to that found in Iron Age Celtiberians of the Basque country. This could essentially mean that Basques were peoples of north/central European ancestry (see below fitting models of origin populations), because they must have arrived to Aquitaine after the arrival of Celtiberians, and with a similar ancestry.
(…) increases in Steppe ancestry were not always accompanied by switches to Indo-European languages. This is consistent with the genetic profile of present-day Basques who speak the only non-Indo-European language in Western Europe but overlap genetically with Iron Age populations showing substantial levels of Steppe ancestry.
The Tollense Valley near Rügen in the West Baltic shows LBA people clustering with Modern Basques (see here). This is compatible with the arrival (or displacement) of Vasconic-speaking Northern/Central Europeans close to the Rhine, possibly originally from northern France, very likely close to the Atlantic area during the Final Bronze Age / Early Iron Age based on cultural interactions.
Pre-Steppe languages in Europe?
An alternative to Old Europeans of the British Isles would be to support some kind of non-Indo-European/Vasconic continuity in the Atlantic façade close to the English Channel and the North Sea, given the current lack of palaeogenomic data on Bell Beakers and later groups in the area, and the potential Vasconic nature of Megalithic/Proto-Beaker groups that might have survived there.
The main problems with this approach are the lack of such an Afroasiatic-like substrate in Gaulish, which should have shown the same substrate as Insular Celtic, and the impossibility of associating this Afroasiatic-like substrate with Vasconic, both potentially representing completely different languages. A counterargument would be that we don’t have that much information on Gaulish and its dialects – or on the syntax of Vasconic, for that matter – to reject this hypothesis straight away…
In any case, the survival of pockets of non-Indo-European, non-Uralic speakers in northern Europe, even after Steppe-related expansions, should not shock anyone:
If the survival of non-Indo-European-speaking groups happened despite the swift expansion and radical population replacement brought about by the Bell Beaker folk – so called traditionally because of its unitary culture suggesting a unitary language community -, and non-Uralic-speaking groups in areas dominated by Corded Ware peoples, it could certainly have happened, and even more so, with Corded Ware and Bell Beaker groups at the western and northern edges of their expansions, due to the early loss of contact with their respective core cultural regions.
Even obscure components of place or river names, like those from northern Europe, the Nordwestblock area, and the British Isles, might be better explained as Old European exceptions than any other alternative, i.e. either as an Indo-European layer over a non-Indo-European one or vice versa, or both in different periods, before the eventual unifying Celtic, Roman, and (later) Germanic expansions.
All in all, one could say about substrates and hydrotoponymy in the British Isles, the Lower Rhine, and in northern Europe as a whole, that the potentially interesting non-Indo-European forms are precisely those which do not interest either scholarly ‘faction’:
those supporting a non-Indo-European Western Europe, because it doesn’t represent the whole substrate, and can’t be used to argue for a Europa Vasconica or Europa Afroasiatica;
those supporting a Palaeo-Indo-European Western Europe, because their limited presence concentrated in isolated pockets doesn’t deny the Indo-Europeanness of the Old European layer anywhere.
However, these are the details that should be studied and that could define what happened exactly after steppe-related migrations, e.g. in the Single Grave cultural area before and after North-West Indo-Europeans admixed with its population, and thus what happened in the British Isles, too.
Ignoring the (mostly useless) typological comparisons, my bet would be for an ancient Uralic layer heavily admixed with local non-Uralic peoples, especially intense in the Single Grave culture. This Proto-Uralic layer would be of a dialect or dialects (assuming succeeding CWC waves and later local expansions) different from the known Late Proto-Uralic – which expanded with eastern Corded Ware groups.
Describing the phonetic features of this layer could improve our knowledge of Early Proto-Uralic, as well as some specifics of the evolution of Germanic, Balto-Slavic, and potentially Celtic and Balto-Finnic.
This would be similar to the relevance of Aquitanian toponyms for Proto-Basque reconstruction, or of the alteuropäische substratum when it conflicts with the Proto-Indo-European dialectal reconstruction of some linguists (e.g. the laryngeal Pre-Indo-Slavonic of Kortlandt) which, like Kitson implies, should question the dialectal reconstruction of this minority of Indo-Europeanists, and not the Indo-European nature of the substratum.