74 thoughts on “The beginnings of the Bronze Age in Europe”

1. @Robert

   Thank you for this guest post on the blog. Hopefully the first of many others to come. I’m really glad to host this original article of yours and I’m sure the readers are very glad to have you writing here too.

   Will start commenting on the content itself later. For now just wanted to express my gratitude and congratulate you for the very good start.

2. Although there are some interesting phenomena, I don’t really have a set conclusion, instead the overview hopes to begin an attempt toward analyzing elite mobility/ elite emulation, the making of identities, and emergence of ‘States’.

   Yes, this is something that will have to be done carefully in order to understand the processes that led to the socio-cultural (and linguistic) identities that we start to know only in the Iron Age and later (talking about Europe).

   Given that adornments, weapons and chariots can be traded & gifted, the only clear way to discern the process of identity making in groups such as Myceneans and Balkan Tumuli chiefs is via Bio-archeo-cultural modelling.

   I guess that the “Bio” part refers to what we can gather from ancient DNA. We already know that after the EBA there are no major population-wide changes in most parts of Europe, but we’re still waiting to see if the details will be informative enough. For example, looking at Great Britain we know that there was a huge genetic shift c. 2300 BCE, which in all likelihood was associated to a linguistic change. The next thing we know is that there was linguistic change in the mid 1st mill. BC (to Celtic) and another one in the mid 1st mill. AD (to Germanic). We have no idea how many shift (if any) there might have been between 2300 BC and 500 BC. The kind of subtle genetic changes that can be discerned in the two known linguistic shifts (not obvious at all), are equally seen in the preceding period mentioned above. But how do we know if they brought any linguistic change?

   Or looking at NE Europe, how many language shifts happened between c. 2800 BC and 500 AD (or even today, depending on where one places the Slavic homeland)? The assumption seems to be “none” because there’s no population-wide genetic shift. But how realistic is that (especially when we have seen different people moving around during all that period)?

   In Iberia we have (still waiting for the details to be published) populations that are genetically the same speaking different language families.

   So this is going to be a difficult task. And I welcome that you start to point out the relevance of certain aspects that we’ll have to know well and keep in mind when trying to figure out the making of identities you talk about (and their associated languages – specially the IE question that is a central part of many studies).

   Any thoughts about how this will unfold? Not asking about predictions of IE origins or the like. I refer to how do you think we’ll be able to make progress in figuring out these things. Maybe by working backwards -in time- from what we do know to what we don’t (assuming aDNA provides enough clues in the former cases that can be applied to the latter ones)? Or even that won’t work due to it being still rather speculative?

3. Hi Alberto
   Thanks, I hope the post was not too long winded.
   I absolutely think we will be able to reconstruct more realistic and fine grained scenarios behind the appearance of language / tribal/ cultural groups. I don’t think it would be very difficult with adequate sampling. Both anterograde and retrograde approaches would help in analysis.

   What we should like to see is aDNA utilised in a less “passive” manner or solely for “hypothesis testing”.
   I think the future lies in understanding detailed diachronic sequence of regions and individual key sites, to fine tune the great but often dispersed data sets we currently have.
   And I imagine newer generations of scholars involved might ultimately be able to balance all lines of evidence, including critically appraising them for their validity & constraints.

4. It would be very helpful if Chalcholithic R1b-L51 was detected somewhere, be it in Western Yamnaya, the Balkans or in any other place. R1b-L51 is so frequent in modern Europeans that its origin is relevant and Bell Beakers were almost exclusively R1b-L51. If Bell Beakers did not speak an IE language and even if they were early IE speakers, the detection of Chalcholithic R1b-L51 would possibly help to identify one of the most important substrate languages of modern IE varieties.

   At the moment, it is even nostalgic to read old books such as ”Encyclopedia of Indo-European Culture” by Mallory and Adams. As for ”Baden”, they speculate on the IE identity of Baden because of the fortified settlement of Vučedol and the presence of apsidal houses. They note that the apsidal houses that appear in Western Anatolia and Greece have been used to trace the movements of IE-speaking groups. However, they warn against linking a culture wth a language on the basis of a house form.

   As for Globular Amphorae, they mention that it has been linked with the Germanic-Baltic-Slavic continuum. Among the points they make are the ceramics that have been associated with the Maykop ceramics and that of the Lower Mikhaylovka group, evidence of suttee, the central role of oxen in burials and the presence of sun-discs. However, they note that such attributes as horse, animal burials and axes were all present in TRB and Lengyel.

   I do not have strong opinions about the languages spoken by archaeological cultures because of the inherent incertainties of such an approach and the inadequate amount of ancient yDNA in particular, but the ancient DNA has rather made me a supporter of late than early linguistic sweeps.

5. @Robert

   What we should like to see is aDNA utilised in a less “passive” manner or solely for “hypothesis testing”.

   Yes, I agree. I do understand that for geneticists getting into this field it was useful to use a well defined hypothesis as a starting point, or a framework to work with. But I’m afraid that it’s ultimately been detrimental as time has passed and they didn’t move on to keep a scientific attitude towards the data they were bringing out. This has been most obvious in the linguistic remarks in several studies.

   I’m eagerly waiting for a good sampling of Italy, which has been neglected so far. Now that larger scale studies are starting to come, it would be great to see one that covers Italy, especially since the Bronze Age to early Roman period.

6. @ Kristiina

   Yes as I’ve mentioned I incline to agree with similar ”recentist” views of socio-linguistics.

   About Baden (& similar groups of that horizon), you write ‘they speculate on the IE identity of Baden because of the fortified settlement of Vučedol and the presence of apsidal houses. They note that the apsidal houses that appear in Western Anatolia and Greece have been used to trace the movements of IE-speaking groups. However, they warn against linking a culture wth a language on the basis of a house form”

   To elaborate on this – it’s in fact more than hourse forms, but ceramics (”fluted wares”), stelae (which pre-date Yamnaya), etc. The above elements could neatly explain many things, including the dearth of steppe ancestry in Anatolia, but calls for a modification of models to form more along the lines of the Shennans and Garrett.

   Your other point ”It would be very helpful if Chalcholithic R1b-L51 was detected somewhere, be it in Western Yamnaya, ”
   Yes whether in Ukraine, or further east in Khvalynsk. We had other lineags in the steppe, only 2 or 3 of which succeeded to expand during the proto-Bronze Age. If these groups had otherwise been separated, living in their riverine niches since the Late Glacial, it’s doubtful they all spoke Pre-PIE (withing the steppe fraemwork). That’s why I feel that treating the entire steppe are a homogeneous putative homeland is problematic, and it deprives opportunities to comtemplate about what other languages could have expanded from there.

7. the overview hopes to begin an attempt toward analyzing elite mobility/ elite emulation, the making of identities, and emergence of ‘States’.

   A main factor in this respect is long-distance trade. Such trade, of course, didn’t just start in the BA. We already know, e.g., of Epi-Paleolithic Obsidian trade from E. Anatolia to the Levante, Mesolithic amber trade from the Curian Lagoon to the Vaidai hills, Piemontese Jadeite axes distributed as far as Denmark and (salt-exporting) Brittany, the E. Pannonian Bodrogkeresztur Culture supplying copper to Cucuteni in exchange for Moldavian salt, or Indian Carneol reaching Majkop. Controlling such trade allows for accumulating surplus that can be used to purchase food from satellite villages, enabling the development of larger, (peri-)urban settlements with specialised crafts, and for constructing exquisite burials. A prime early example is Varna, now as then a major seaport, that during the CA processed and exported commodities from the Bulgarian hinterland (salt, copper, gold) towards the Circum-Pontic area.

   The Bronze Age took trade towards a new dimension. Previously, it was fine to own, e.g., Rijkholt axes or shiny and sharp Sardinian Obsidian blades, but, if neccessary, knapping your local second-grade stone ressources would still have done the job. However, bronze-making requires alloying copper with tin, and while copper ressources are comparatively widespread, exploitable tin is scarce. There seem to have been a handful of smaller tin mines around the E. Mediterranean, e.g. in the Taurus mountains and the E. Balkans, but most of the tin for Egyptian / Levantine / Aegean bronze making needed to be imported from far away, namely either BMAC, or the Atlantic Facade (Cornwall / Brittany /Galicia). [Speculatively, the Erzgebirge (“Ore Mountains”) between Saxony and Bohemia, and Somaliland, could in addition have provided alluvial tin oxide (Casssiterite) that can be directly used for bronze-making].

   Against this background, I think it is anything but accidental that the two most outstanding European EBA cultures outside the E. Mediterranean both combine copper ressources with control of tin routes:
   – El Argar sat next to the S. Iberian copper (and gold/ silver) mining area (e.g. Rio Tinto), and controlled maritime tin trade from the Atlantic Facade towards the E. Mediterranean;
   – The Unetice core extended between the Nitra and Harz copper mining areas (the former well evidenced already for the BA, the latter Europe’s largest medieval copper producing area), and controlled trade of Cornish tin along the Elbe-Danube axis to the E. Mediterranean, and the Erzgebirge alluvial tin oxide deposits.
   Both were economically, to some extent also culturally, dependent upon, and ultimately peripheral of the E. Mediterranean. That dependence became obvious when the Thera eruption ca. 1,600 BC sent shock (and Tsunami) waves through the Aegean, eventually not only causing the demise of the Minoan civilisation, but also fundamental transformations in Iberia (El Argar I -> II shift), NW Europe (spread of bronze-making to Wessex, Elp & Tumulus cultures, plus Nordic BA) and of course the Balkans.

   The WC Asian situation was somewhat different. Bronze-making and BMAC tin exploration appear to have occurred slightly earlier than in Europe, and Mesopotamia and the IVC provided alternative markets. Still, one shouldn’t underestimate the relevance of BMAC tin (bronze) exports to the E. Mediterranean as evidenced a/o from Akkadian, Elamite and Sumerian records. An evidence of BMAC trade towards the E. Mediterranean is Lapis Lazuli (e.g. Tutankhamun burial mask), but in all likelyhood Lapis Lazuli was just piggy-backed onto tin trade in much larger volumes.
   I suppose that Kura-Araxes with its strong metallurgical focus played a main role in that trade. KA expansion to the Northern Levante is archeologically attested, and there is some indication that it didn’t stop there but reached at least Cyprus. Less clear is the role that KA played in Central and Western Anatolia, beyond Arslantepe, and ultimately maybe even in Thrace and Greece. I mean, Lazarides e.a. 2017 a/o came up with a feasible model of Mycenaeans as 37% Armenia MLBA (Kura-Araxes) – 63% ANF.
   Of course, Caucasian ancestry was already present in Barcin, NW Anatolia around 3800 BC, i.e. before the KA expansion, and, as mentionned by Alberto, we are getting to the point here where genetic changes introduced by migrants may be so subtle that they can hardly be discerned anymore [Still, the Laz. 2017 BA samples from Central Anatolian Harmanören and Minoans have been shown to slightly favour Iran_N over CHG, while the Damgaard e.a. 2018 “Hittite” samples clearly prefer CHG, so there seems to have been substructure, relating to different migrations, in CA/BA Anatolia and beyond].
   What is becoming apparent in any case is that not only the European, but also the West Asian BA experienced a shift of yDNA. G2a, once dominating ANF and EEF, seems to have virtually disappeared, while “Caucasian” J, especially J2a, is found all over the place from Namazga to Mycenae, the Levante, and Crete.

   The 2200 BC crisis around the E. Mediterranean was in all likelyhood climate-related – geological records e.g. from the Dead Sea indicate massive drought. To which extent the contemporary demise of BMAC had also climatic reasons, or was related to the loss of Mediterranean markets has so far remained unclear to me. In any case, it is probably no coincidence that around the same time also Kura-Araxes collapsed (and that bronze-making, apparently based on Central Asian tin that required alternative markets, made its way into NW China).
   Similarly, the 1600 BC Thera eruption coincided with turmoil in West Asia, e.g. the Hittite conquest of Babylon.

   A fascinating question is whether cultures are expansive “by nature”, or become so only in reaction to ecologic crisis and/or a collapsing trade system. E.g., Geoffrey Summers sees the Kura-Araxes expansion into the Central Zagros as having been motivated by the Late Uruk collapse. Similarly, while the emergence of the (Hellenic-speaking) Mycenaean civilization somewhat predates the Thera eruption, it only rose to full bloom afterwards, when it took over control over maritime trade from Tsunami-struck Minoan Crete.

   While I have mused above on Mycenaeans being potentially modeled as ANF-KA mix, I am actually quite certain that they represent some immigration from Europe. Point in case is the massive upswing in Mycenaean finds of Baltic Amber, a commodity that was also traded onwards e.g. to Crete and Syria, and found on the Uluburun shipwreck. Amber would have been unknown and of little interest to Anatolians, but has traditionally played an important role in North European cultures such as Narva, TRB and GAC, and in the Iberian Mesolithic and Neolithic. Most importantly, based on the a/m tradition, NW European Bell Beakers introduced conical amber beads pulled on strings, a form that has survived to date as Christian and Muslim prayer chains. Intriguingly, while Christian prayer beads may also be made from other materials such as (Indian) Carnelian, the Quran stipulates that prayer beads must be of amber.
   As such, I question the narrative of Unetice reflecting (solely) a movement from the southeast, i.e. the Balkans. We have a “Nordic” commodity (amber), turned by NW BB into a standardized and currency-like form (conical beads on a string) that ultimately became a symbol of Near Eastern (Muslim, Christian) faith – and the path clearly leads through Mycenae. Whether it also lead through the Balkans is another question, though. Classical Greek authors tended to locate the origin of amber at the caput adria (the well evidenced MLBA / IA “amber roads” ended in Aquilea), and the controversial MBA find of Baltic amber insigned in Linear B script from Bernstorf, Bavaria (40 km N of Munich, Tumulus Culture, largest BA fortified settlement north of the Alps) is also rather favoring an Adriatic connection.

8. @ Frank

   ”Of course, Caucasian ancestry was already present in Barcin, NW Anatolia around 3800 BC, i.e. before the KA expansion, and, as mentionned by Alberto, we are getting to the point here where genetic changes introduced by migrants may be so subtle that they can hardly be discerned anymore”

   Yes you’re right. The ”eastern shift” appears c. 4000 BC.
   The terminus ante quem, in the way of the Late Neolithic Tepecik samples (c. 5000 BC), are not dissimilar to West Anatolian Neolithic, whilst by 3800 BC, even western Anatolia is considerably shifted.

   This suggests most population flux occurs between 5000 and 4000 BC. This is hardly coincidental, as we know there are significant social ruptures in Anatolia at this time, although we are still attempting to understand them.

   So yes, this precedes the K-A phenomenon, which is said to have expanded from the south Caucasus. However, red-black burnished ware is seen already in east-central Anatolia c. 5000 BC. We are looking at pastoralist networks in the highland periphery north of the Uruk world.

   The 2500 BC trade -network, on the other hand, has the hallmarks of a Mesopotamiam (pre-Assyrian) phenomenon. Perhaps it did not alter the genetic fabric very much, and indeed appears to have itself collapsed with the 4.2 ky event, however what is more interesting are the resulting series of regional transformations & shifts.

9. By the way, here is a qpAdm model for Bulgarian EMBA

   left pops:
   Balkans_EBA
   Balkans_ChL
   Eneolithic_Ukraine
   Anatolia_BA

   right pops:
   Mbuti_DG
   Ust_Ishim
   Karitiana
   Onge
   Kostenki14
   Levant_N
   Ganj_Dareh_N
   Anatolia_N
   Iron_Gates
   West_Siberia_N
   MA1
   EHG
   CHG

   numsnps used: 321385

   best coefficients: 0.475 0.367 0.157
   std. errors: 0.052 0.019 0.053

   fixed pat wt dof chisq tail prob
   000 0 10 7.677 0.660356 0.475 0.367 0.157

   Adding Yamnaya Samara makes fit unfeasible
   As per earlier nMonte runs, Ukraine eneolithic seems enough to account for the “steppe” shift in the Balkans, which makes sense given that I2a2a1b hovered between Hungary, Bulgaria and Ukraine

   We have also established that by 3000 BC, Anatolian type ancestry had already made its way to the Balkans. Chad and I might have a look at a few more later

10. @Robert

    No better place to ask; how reliable are the datings of the Vinca Culture’s copper artifacts? If I remember correctly, the Policnik axes were found at a shallow depth.

    What’s your opinion on this?

11. This is from ” The Making of Bronze Age Eurasia” by P. Kohl:
    ”… If one is going to attribute the collapse of the Varna-related cultures to an invasion from the east, one also has the problem of circumventing the giant Tripol’ye-culture sites, which are beginning to develop at the time of the first postulated migration. The environmental-crisis model has the virtue of proceeding in the right direction: the observed sequential archaeological collapse from the southwest to the northeast corresponds to different latitudinal zones being affected at different times owing to this progressive onset of more arid conditions and changes in sea level. Some of the settlements described in this chapter were fortified, and many, including all the giant Tripol’ye settlements, show considerable evidence for burning and for having been destroyed by fire. Videjko (1996: 74) attributes the emergence of the gigantic settlements with their enclosures to internal competition and fights among different Tripol’ye groups, an interpretation that also is consistent with the environmental crisis model; times get tough – for whatever reason – and people get nasty. The direct archaeological evidence for such conflicts, however, is rather limited; the burning of the settlements may have been directly associated with their sequential abandonment and may have been conducted for ritualistic purposes, possibly associated with efforts to rejuvenate the soil. The available pollen analyses from the gigantic settlements are consistent with a pattern of deforestation and an overall reduction in biodiversity.

    Thus, the environmental-crisis model championed by Todorova seems consistent with the archaeological record. She (1995: 89) summarizes this model as follows: The brilliant development of the late Eneolithic cultural block was terminated at the end of the fifth millennium and the beginning of the fourth millennium B.C. by a colossal, global and multi-causal environmental catastrophe: the final stage of the climatic optimum, when the mean annual temperatures reached their post-glacial maximum. . . . The rising sea levels caused the water table to rise resulting in the swamping of the plains. The final blow to the Eneolithic economy was delivered by prolonged droughts which deprived the people of their means of existence and forest fires and erosion put paid to any chance of survival. . . . This phase of sea ingression . . . reached its culmination around 3,500 B.C. . . . These movements can be seen very clearly along the Bulgarian Black Sea coast. The steppes dried up, becoming deserts or semi-deserts in the south and spread beyond their Neolithic boundaries.

    What is indisputable is that two major shifts in settlement occurred: first, the abandonment of the Karanova VI Varna-related sites of the northern Balkans at the end of the fifth millennium and the consequent spread to the northeast and growth in size of the Cucuteni-Tripol’ye sites, culminating in the emergence of The Chalcolithic Prelude the gigantic settlements; and second, the subsequent collapse of these gigantic, agriculturally based Tripol’ye settlements around the end of the second quarter of the fourth millennium. If the demographic calculations for the giant sites that were presented above are at all accurate, tens of thousands of people were engaged in adopting an even more extensive, more mobile economy, relying principally on animal husbandry, both cattle-raising and increasingly the herding of sheep and goats, an economy which has been characterized as seminomadic. As Parzinger notes, the changes in settlement patterns, burial rites, and basic economy were fundamental, justifying the term postTripol’ye to refer to the new regional groups that emerge at this time. New areas of the steppe are occupied, such as the Lower Dniester region by the Usatovo post-Tripol’ye group, and the house remains of their settlements, such as at Usatovo-Bolshoi Kuyalnik and at Mayaki, are so insubstantial to nonexistent that it is reasonable to identify them as seasonal, possibly summer encampments, despite their size and surrounding ditches. In any case, the change from the time of the giant agricultural settlements is striking, and clearly what is being recorded is a shift from a more sedentary to a more mobile way of life, or, if you will, groups that are predominantly cultivators are transforming themselves to become groups that are principally herders.

    A major shift in intercultural relations also begins around the middle of the 4th millennium BC or slightly earlier and greatly affects these changes in the Northern Pontic and East European steppes. The advent of the well-known Early Bronze cultures of the Caucasus – the Maikop and later Novosvobodnaya cultures of the northern Caucasus and the Kura-Araxes (or Early Transcaucasian) “cultural community” of the southern Caucasus, eastern Anatolia, and northwestern Iran – marks a radical change in the production and exchange of metals throughout the entire interconnected area. Arsenical copper/bronzes, most of which originate in the Caucasus, replace the copper artifacts, which had been procured originally from the Balkans. Chernykh’s Circumpontic Metallurgical Province emerges to replace the no-longer functioning Carpatho-Balkan Metallurgical Province. Its emergence and the advent of these Early Bronze Caucasian cultures must somehow be related also to roughly simultaneous developments occurring farther south that involve the greater integration of northern Mesopotamia, including the Upper Euphrates drainage on the Anatolian plateau, into a larger Mesopotamian world, involving ultimately the movement of colonists and traders from southern Mesopotamia, or what now is referred to in the literature as the Uruk expansion.”

    Summary:
    – Balkans is one of the most advanced areas of Europe and the Near East during the Chalcolithic;
    – climatic crisis c. 4200-3000 BC and the collapse of the flourishing Chalcolithic cultures;
    – spread to the northeast and growth in size of the Cucuteni-Tripol’ye sites;
    – collapse of the gigantic, agriculturally based Tripol’ye settlements;
    – a shift from a more sedentary to a more mobile way of life of the post-Tripol’ye groups;
    – the rise of the Early Bronze cultures of the Caucasus – the Maikop and later Novosvobodnaya cultures of the northern Caucasus and the Kura-Araxes;
    – Chernykh’s Circumpontic Metallurgical Province emerges to replace the no-longer functioning Carpatho-Balkan Metallurgical Province;
    – integration of northern Mesopotamia with the Uruk expansion.

    This is the background for the emergence of the Yamnaya culture.

12. @ Vara
    Hi. I had not looked specifically into that because I did not realise there had been any doubt . i.e. that Copper axes etc were being used already in 6th millenium in Balkans. Is there any specific paper you might be referring to ?

    Here is one paper about the Policnkik axes etc
    http://www.doiserbia.nb.rs/img/doi/1450-5339/2009/1450-53390902165A.pdf

13. @ Kristiina

    Yes that’s a very good summary.
    Worth mentioning /correcting is that
    – not all ”Balkan” centres collapsed – eg in the northwest / Carpathian basin continued toward a more dispersed pattern (pre-Baden etc)
    – on the other hand, the collapse also affected the Suvorovo -Skelya culture, which also seems to have disappeared after 42/4000 BC
    – Anatolia remains something of a black box, and at the moment all we have is non-evidence based speculation on certain blogs. This will gap be filled in over the next year or so, including with more aDNA. IMO, we are close to understanding the exigious proto-Anatolians, & it ties in with the system of collapses, but ultimately, it is becomign increasingly untenable to place the steppe as the major constitutive element behind the ”proto-Indo-Hittites”.
    – I don’t think C-T derives from Karanovo refugees, but derives from earlier LBK -East groups.

14. I copy this from Anthrogenica:
    Someone posts some informations on the French forum about the thesis of Samantha Brunel from Institut Jacques Monod, Paris. Samantha defended her thesis last week: Paléogénomique des dynamiques des populations humaines sur le territoire Français entre 7000 et 2000 and Jool was here. He wrote the summary of the Brunel thesis results on the French forum and this is the Google translation!:
    “They have a good hundred samples from the North, Alsace and the Mediterranean coast, from the Mesolithic to the Iron Age.

    At the bronze age, they have 5 samples with autosomal DNA, all in Bell Beaker archaeological context, which are very spread on the PCA. A very high sample close to the Yamnaya, a little above the Corded Ware, two samples right in the Central European Bell Beakers, a fairly low just above the Neolithic package, and one last full in the package. The most salient point was that the Y chromosomes of their 12 Bronze Age samples (all bell beaker) are all R1b, whereas there was no R1b in the Neolithic samples. ”

    This is Bernard’s comment:
    – French Bell Beakersare derived from a genetic mixture between a steppe population (sample located with Yamnaya in the PCA) and a local Neolithic population (sample located with Neolithic farmers)
    – The steppe population comes directly from the steppes (and not from Central Europe) otherwise we would not have a Bell Beaker sample that is located with the Yamnaya. So the Bell Beaker culture is not an emanation of the Corded Ware culture.

    IMO, this conclusion makes sense.

    Do we have any autosomal samples from western Yamnaya? I know that we have samples from Caucasus and Volga Yamnaya and mtDNA from Western Yamnaya.

15. @Kristiina

    Thanks for the news. It’s interesting to see a Bell Beaker from France that clusters with the Yamnaya samples. Not sure what it will mean exactly (we’ll have to see the details). We do have 3 samples from Yamnaya Ukraine, but all 3 females, and one of them is an outlier with something akin to Armenia_EBA admixture. The other 2 are similar to other Yamnaya samples (IIRC), maybe with slightly more admixture from Neolithic European populations.

    It will be interesting also to see if the samples that clusters with Neolithic farmers is a male carrying R1b-L51 or if it’s a female.

16. @Robert

    “Both authors emphasize that the metal objects were found at rather shallow depth of about 30 and 80 cm, respectively, directly below the humus layer and on top of the 1 – 3 m thick cultural layer. Therefore they should date close to 4000 B.C. ”

    https://www.researchgate.net/profile/Ernst_Pernicka/publication/249932266_Eneolithic_and_Early_Bronze_Age_copper_artefacts_from_the_Balkans_and_their_relation_to_Serbian_copper_ores/links/00b4952c51ec55501c000000/Eneolithic-and-Early-Bronze-Age-copper-artefacts-from-the-Balkans-and-their-relation-to-Serbian-copper-ores.pdf

17. @ Vara

    From what I understood with a quick reading, the Pernicka article suggests that the primacy of the Rudna Glava mine is not supported by the evidence. Instead, the isotopic signatures suggest that several mines were being exploited.
    But I don’t think this changes the fact that by the 5th millenium, copper was being used and entered a golden age in the Balkans.
    What is however hinted at, and also by other authors (eg F Klimscha. see Academia.edu) is that this zenith corresponds to a similar one in the Levant, making them doubt the autochthonist or at least isolated developed on Copper technology in S.E.E. as originally proposed by Renfrew.
    In other words, how are these parallel developments to be linked, when there is no clear line connecting Balkans , and Israel, for example. They hint at other centres, eg Anatolia or Iran as a possible spark for Copper technology, although it was experimented with much earlier, as far as PPNB).
    Given the changes to the ‘Old Neolithic’ world i touched upon above, there might be some tangible links to speak of as early as the 5th millenium.

18. Regarding the France samples. Indeed we need to know the details. But irrespective of that, they say we have 5 that yield good aDna…of those 2 are very close autosomally to Neolithic farmer and 2 are full Neolithic farmer and everyone focus on the one that sits closer to yamnaya?
    Isn’t this indicative that we are in fact under a skewed frame of mind regarding this issue?

19. Robert.
    CT is the culmination of the population movement that started as Boian and engaging locals such as Hamangia to at that point became Gulmenita Culture which in turn is seen moving northward where in there engage LBK to become Pre_Cucuteni. This the moviment of people in the second half of the 5th millennium BC in South balkans. This is where you find copper, carnelian beads, etc.

    Is there something you know of that disproves this? To me its a given.

    Note: it’s not a secret that I see this movement of population , taking as reference the kum6 girl, as the movement of South Caucasus to balkans after 4900BC. But that is not a secret where I come from, right?

20. @FrankN
    Yeah, the incursion of Caucasian/Iranian ancestry in Anatolia began in the AAF period of the Neolithic transition, well before 3800 BC.

21. @Robert

    It seems all roads lead to Iran. Thanks!

22. @ Vara

    I’m not sure exactly where it points to, or what it means. There were probably several flows around the regions in discussion. The main phase would be in the post-Ubaid period (also referred to as Late Chalcolithic , CFW, etc ). Could relate to metallurgy, seeking ores,

    @ OM
    The shift in Kumtepe 6 coul;d be different to that seen in Late Chalcolithic – Bronze Age Anatolia.

23. Hi Robert.
    can you expand a bit on your comment?
    “The shift in Kumtepe 6 coul;d be different to that seen in Late Chalcolithic – Bronze Age Anatolia.”

24. @ OM
    It seems at present that Kumtepe is not too different to early Barcin, but perhaps more CHG admixture, probably representing structure within Anatolia (perhaps this individual being representative of northeast turkey) .
    On the other hand, after 4000 BC, the samples are more profoundly shifted and of a different character. ? Post-Ubaid movements. We would need more samples to in turn compare those cf the Bronze Agers, which by account of the Y haplogroups (G2b2, J2a1) could point to Iran (?)

25. @Rob
    Good post 😉

    Given all recent and past findings, I-E expansion was either very early, or very late. And both have very drastic implications. Or of course, not I-E expansions.

    Also, I’m beginning to be suspicious of aDNA. aDNA is highly volatile, and within 3-4 generations, the composition can be very dissimilar from previous generations. For this reason, I lean towards Y-haplogroups as more useful genetic information (without discounting aDNA) in order to establish a new paradigm.

    @Kristiina: “… it is even nostalgic to read old books such as […] by Mallory and Adams” Yeah, I agree. It’s going to get worse and more nostalgic when more people accept what’s going on.

    @Vara
    Depends what you mean by Iran. That’s a floating definition throughout the ages, with very different original groups.

26. Sorry late question about – ” it’s in fact more than hourse forms, but ceramics (”fluted wares”), stelae (which pre-date Yamnaya), etc. The above elements could neatly explain many things, including the dearth of steppe ancestry in Anatolia ”

    Do you mean that Indo-Hittie might have emerged from a, EEF group, but not Anatolia itself ?

    What about the EEF collapse – was this partial ?
    http://eurogenes.blogspot.com/2018/12/europes-ancient-proto-cities-may-have.html

27. @ Atrior

    Both autosomes and uniparentals are required for the complete picture, i think most people would agree on that.
    I’d tend to agree with Kristiina that most langauges represent relatively recent expansions. Trying to connnect the various languages through the emerging evidence is the fun part. The evidence so far is becoming robust, but a few details are still missing.

28. @ Toby

    It’s certainly a possibility that I-H comes from an earlier, pre-steppe stratum. For example if further sampling from Anatolia also lacks steppe admixture.

    As per your second question – not all EEF collapsed, and those that did occurred at various times. E.g. the Varna-Karanovo VI horizon collapsed c. 4000 BC.
    C-T, cited in that paper, ended c. 3000 BC. Yamnaya ”collapsed” in the Balkans c. 2500 BC.
    Vucedol, Baden, Ezero are all predominantly EEF groups, albeit with increasing admixture from steppe and Anatolia.
    Cultures transform and admix continually.

    About C-T, i think in most of the western areas it is actually GAC which succedes it, which is an ”EEF” culture. Well actually, it’s not EEF, but ”MNE”. 😉
    The old Neolithic world came to an end, certainly. However, the details differ in each sub-region, there isn’t a singular explanation.

29. @Atri∂r

    I am very well aware of what the concept of Iran is since most of my personal research (Zoroastrianism) is related to it in one way or another. However, the Iran in question here is the modern country of Iran, specifically the region of greater Media, which shows a clear development in Tepe Sialk from native copper to experimental metallurgy in the neolithic all the way to making weapons 6kya.

    I also think that IE languages reached Europe later than expected and most likely early steppe cultures had nothing to do with PIE, but that might take years for some to accept.

30. A great summary that agrees with my general knowledge about this period too but I’ll have to disagree with this in particular (though I might be misunderstanding it, particularly in conjuction with “That’s why I feel that treating the entire steppe are a homogeneous putative homeland is problematic, and it deprives opportunities to comtemplate about what other languages could have expanded from there.” which I agree with, below):

    “but ultimately, it is becomign increasingly untenable to place the steppe as the major constitutive element behind the ”proto-Indo-Hittites”

    I think that despite various areas of cultural and/or genetic interactions on all sides of the Black Sea being responsible for the entities we see emerging later, the constitutive element of “Proto-Indo-Hittite” in at least the linguistic sense was (among the Balkan, Dnieper-Don, Don-Volga and Caucasus) the Don-Volga steppe one that’s so far represented by Khvalynsk and (Caucasus_)Steppe_Eneolithic (and the likely late PIE Yamna later on, i.e. everything but Anatolian and likely Tocharian) in our aDNA record.

    This also fits in with your comment about the Ukraine_Eneolithic-like element in the BA Balkans. It’s a cluster that seems to be made up of individuals of rather varying amounts of Ukraine_Neolithic (intermediate HGs), Anatolia_N (Ukraine_Neolithic outlier like, neighboring Balkan “Old European” cultures) and Khvalynsk/Steppe_Eneolithic/Yamna with the [coincidentally(?) R1a] Alexandria Sredni Stog II ascribed individual having most of the (newly arriving?) eastern steppe kind of ancestry. Of the three basic streams, it makes most sense to me to correlate PIE/PIH with the eastern steppe one when all recent papers are taken in together, though your points of caution are well-taken. Of course, one possible alternative is that Ukraine_Eneolithic is instead the result of an CHG-rich population arriving in the Ukraine “separately” from the eastern steppe around that same time and that together with the preceding Ukraine_N (plus some Anatolia_N) gives a result that looks very similar to an eastern steppe infiltration in its combination of EHG and CHG ancestry but I personally find that scenario less plausible right now. The lack of Western European R1b so far on the (eastern) steppe also complicates things a bit. And to what population is the Alexandrian R1a related, the western or the eastern steppe one? Some important questions still left even in a relatively sampled area…though the Wang samples might help clarify things.

    In related, though I’m sure people here have read about it, an upcoming paper apparently shows the separate CHG “Mediterranean” kind of ancestry also appearing in Sardinia. This leaves eastern Iberia as the last place to check for this Anatolian/Aegean impulse mentioned here (the choice to use Anatolia_BA for even more northern Balkan samples was very appropriate apparently! though I wonder if some steppe migrations also could have brought in more CHG than average…) and described in literature, by Heyd in conjunction with the more discussed online Beaker complex in the rightfully much brought up “Europe 2500 to 2200 BC”.

31. @ Egg

    Thanks for the comments bud.
    I will try to answer your questions/ comments in turn-

    a) ” in at least the linguistic sense was … the Don-Volga steppe one that’s so far represented by Khvalynsk and (Caucasus_)Steppe_Eneolithic (and the likely late PIE Yamna later on, i.e. everything but Anatolian and likely Tocharian) in our aDNA record.”

    How do we know that formative Indo-Hittite is most likely to have developed in the Caspian steppe, specifically ? What’s the evidence you’d propose.

    b) Ukraine Eneolithic R1a – certainly seems to represent an eastern shift, although I think we are still missing some samples to obtain an optimal fit. Of course, there is differences between where R1a might have trickled in from c.f. the autosomal idosyncracies we see in individuals and broad shifts we see at a population level. The Alexandrian male’s personal line could go back to the forest steppe, but ancestry muddled due to the interaction and admixture which connected the Sredni-Stog horizon from 45000 BC, increasingly connecting disparate groups which were initially confined to their riverine niches. But he certainly appears very eastern derived:

    Ukraine_Eneolithic:I6561 (male, R1a; 4000 BC)
    Samara_Eneolithic:I0122 54.9%
    Protoboleraz_LCA 28.7%
    Armenia_ChL 11.05%

    Whilst the R1b-Z2013 guy from (?) late Dereivka looks to hail from the Ukraine-Baltic axis.

    Ukraine_Eneolithic:I5884 (Male, R1b-Z23013, 2800 BC)

    Ukraine_Mesolithic 45.25%
    Narva_Lithuania 26.9%
    Protoboleraz_LCA 22.6%
    Armenia_ChL 5.1%

    (perhaps the R1b-pre-297 in the Baltic_HGs might be a further clue)

    Then, by 3300 BC, a couple of lineages came to be dominant in the steppe propper, e.g. R1b-Z2013 for the Don-Volga region.
    It will be very interesting to see what western Yamnaya shows, not to mention the forgotten Cernavoda & Usatavo cultures.

    c) ”In related, though I’m sure people here have read about it, an upcoming paper apparently shows the separate CHG “Mediterranean” kind of ancestry also appearing in Sardinia. This leaves eastern Iberia as the last place to check for this Anatolian/Aegean impulse mentioned here (the choice to use Anatolia_BA for even more northern Balkan samples was very appropriate apparently! ”

    Yes this must related to the ”Bronze Age package” amply described in literature which reached the Aegean & Italy by 2000 BC.
    I am curious to see if there is any genetic manifestation of this ”CHG-Med” in El Agar, as some of its citadels (esp. La Bastida) are said to show Mediterranean inspiration. Judjing from the abstract made available by the Reich team on Iberia, i couldn’t recall them mentioning it.

32. @Egg

    It is tempting to propose a linear development of IE on the steppe. However, what falsification criteria could be accepted for this hypothesis?

    Keep in mind that most basal phylum of IE covered the entire southern half of Anatolia and more. We are talking about not only Hittite, but numerous linguistic groups within the Anatolian branch. I would find it extremely difficult to believe that this could have happened without major demic impact considering that Anatolia was much more developed than the steppe at the time. If Haak and colleagues can’t find steppe admixture in Anatolia, I’d think it more realistic to dismiss the steppe entirely as a formative stage of IE. The European model would still work, but it would necessarily have to be EEF groups carrying IE to Anatolia.

    Anyway, I think it all hinges on the upcoming samples from the critical regions. If Haak et al. with their hyperfocus on the steppe will be unable to tease out Yamnaya admixture in early Anatolian samples, I would say that Indo-Hittite from the steppe should be considered definitively falsified.

33. As for haplogroups, I think they really are muddying the waters when it comes to the dispersal of languages. While R1b-L51 had a signficant expansion in Bronze Age Europe, so did for example E-V13 and I-M253. It would require a leap of faith to correlate the latter two to the expansion of any known language group, but the association of R1b and IE has become gospel despite the fact that one of the most R1b-rich groups doesn’t speak IE.

34. @Marko

    Yes, I agree that Anatolian languages don’t have any reasonable source population in the steppe. As you say, it’s not just a small elite of Hittite rulers. Most of Anatolia was IE speaking, with quite diverse languages. We’d need a large and very early migration from the steppe to explain this.

    About R1b-L51 I agree too. For now no connection to IE, but strong connection to non-IE. It’s not modern Basques, it’s large areas of Western Europe, and 100% of the population who didn’t switch quite late to Celtic or para-Celtic (Lusitanian as only survivor of possible pre-Celtic migrations bringing para-Celtic languages). It becomes difficult to explain this large population turnover with no known linguistic impact (all Bell Beaker languages in Western Europe went extinct by the Iron Age?).

35. @Alberto

    You are absolutely right – a careful look at the ethnic-linguistic situation in Western Europe during the metal ages will be crucial. Even more interesting than Iberia perhaps could be Britain & Ireland. The latest research on the Insular Celtic languages by Schrijver, Matasović and others indicates that the Iron Age Celts encountered distinctly non-Indo-European groups upon entering the islands. It is very unlikely that late Neolithic languages survived the overwhelming replacement by the Bell Beakers in Britain and Ireland – the non-IE substrates in Brythonic and Gaelic are remnants of the Bell Beaker languages I have no doubts! It is now only a matter of collating the linguistic evidence with the emerging evidence from population genetics.

36. @ Alberto

    “”About R1b-L51 I agree too. For now no connection to IE, but strong connection to non-IE. ”

    I wouldn’t discount it completely
    As I outlined, it could be certain subsections (eg the Bavarian group) served as Indo-Europeanisers of their kin further west
    Of course, if Mycenaean shaft graves come out R1b-L51, then further discussion is moot

37. @ Marko

    “Keep in mind that most basal phylum of IE covered the entire southern half of Anatolia and more. We are talking about not only Hittite, but numerous linguistic groups within the Anatolian branch. I would find it extremely difficult to believe that this could have happened without major demic impact considering that Anatolia was much more developed than the steppe at the time. If Haak and colleagues can’t find steppe admixture in Anatolia, I’d think it more realistic to dismiss the steppe entirely as a formative stage of IE. The European model would still work, but it would necessarily have to be EEF groups carrying IE to Anatolia.”

    This is a fundamental fact which needs to be digested.
    Some of the discussions one recalls amidst the more speculation-oriented fora includes :

    A) Anatolia was too densely settled, so steppe admixture became rapidly washed away .
    This is not actually the case. There is a documented population decline between 5000 and 4000 BC. These are prime conditions for any migration from the steppe to be palpable. Indeed, the Damgaard study detected an nomad from Anatolia in the Iron Age, when demography was several fold higher

    B) IE chiefs created state – sanctioned linguistic change. We still need aDNA and archaeological evidence for steppe conquest, elite migration etc. How is it that the same process happened to have occurred in all of central & western Anatolia at the same time? As per above/ there should be plenty of evidence for this. We should see kurgans and barrows throughout Anatolia.

    C) This just for laughs- Anatolians adopted the IE language from entrepid and mysterious travellers from the steppe out of awe.

    Some more aDNA from western Anatolia will be instructive, if steppe admixture turns up, it’ll be there.

38. @ Rob
    Thanks for explanation.
    your point about Anatolia is interesting, i remember the article someone posted on French thread of Anthrogenica – ”Climatic changes and social transformations in the Near East and North Africa during the ‘long’ 4th millennium BC: A comparative study of environmental and archaeological evidence”

    With regard to movements – what about the Suvorovo culture – that seems to be conclusive evidence of steppe migrations into the Balkans. It probably continued on to Anatolia as well. Some samples from Mathieson could be of Suvorovo origin.

39. @Rob

    “How do we know that formative Indo-Hittite is most likely to have developed in the Caspian steppe, specifically ? What’s the evidence you’d propose.”

    Leaving archaeological arguments that are more ambiguous, especially within the steppe, and carry even more disagreement aside, that likely IEzation in our aDNA record so far correlates with the appearance of the eastern EHG/CHG steppe element [and the relevant Y-DNA in some areas so far; though the latter is mostly Z2103 on the steppe itself unfortunately but its R1b sibling can’t be far off and the relevant R1a clades aren’t implausibly also derived from that area (R1a in Khvalynsk too after all) but more sampling will prove this more conclusively right or wrong]. That’s the strongest argument we have so far that doesn’t leave as much room for ambiguity in my view. But as I mentioned, I wouldn’t totally discount the western (Dnieper-Don) steppe either just yet since an influx of CHG ancestry in it could potentially give a similar pseudo-eastern steppe effect due to the HG ancestry there in Ukraine_Neolithic (in Global25 for example this kind of model for Ukraine_Eneolithic works relatively well too), it just appears much much less plausible to me and right now I’m considering a scenario of eastern to west migration within the steppe as more likely which would put the early IH/IE locus on the eastern Don-Volga steppe, with mixing and levelling between the two areas occuring later. The fact that specifically the Sredni Stog II ascribed individual I6561 is also the one that seems to be carrying that element at its highest among the sampled and is harder to try and model with CHG instead of something from the eastern steppe (and that it has the relevant R1a at that) is even more interesting, then. I’m looking forward to the published Wang data at any rate.

    If you were referring specifically to the Anatolian aspect of IH like Marko did and keeping solely to genetics, that’s certainly muddier at the moment though some of the sampled Anatolians seem to show subtle steppe-related ancestry to me too. Similarly I don’t see any later movements from the Transcaucasus onto the steppe that could be responsible per my response to Marko below.

    @Marko

    I admit that, like everyone else, I have certain preconceptions in mind about the relevant movements and general linguistic associations so your question is appropriate. For example, what if pre-Phrygian/Greek/etc. Anatolia turns out to have overwhelmingly local Y-DNA (non Z2103 or I2a etc.) and ancestry, to the extent that the steppe associations become minuscule to the point of being almost invisible? You can turn to the more ambiguous archaeological arguments and admit that most people wouldn’t have expected steppe ancestry to be that prominent in (most) Anatolians anyway, which is true, but it’d be a bit of a double standard considering the broader arguments (though, as a sidenote, I also don’t see “falsificationism” as a be-all end-all, we all follow more anarchic procedures when examining lines of evidence and some end up being more fruitful than others in the long-term). So far in the Balkans we see some early steppe-infused outliers with Y-DNA that’s biased towards the farmer aside if anything. Some of those could be evidence for some of the early migrations also giving rise to Anatolian but some might very well be dead-ends that were incorporated in Balkan societies (I remember even Gimbutas arguing for this, early peaceful incorporation of steppe individuals in neighboring Balkan societies; those plausibly didn’t give rise to anything, just adopted settled down and adopted the local languages).

    For post-Anatolian IE, I think it’s harder to disagree with a steppe initial locus. The Steppe_MLBA element for Indo-Iranian is entirely likely as the catalyst per the recent papers too (just less prominent in autosomal DNA as theorized sometimes due to admixture with related ANE-heavy groups along the way) and the difference between Mycenaeans and Minoans is precisely a steppe-carrying element. Considering Crete_Armenoi (and maybe even the Anatolia_IA MA2197, named Anatolia_IA_low_res in Global25? it’s tempting to see it as a Greek colonist of Hellenistic times or a related Balkan group considering its apparent makeup and modern and ancient distances, though its exact associations are uncertain), this difference is also more likely related to European movements than West Asian ones.

    I don’t find this argument of yours plausible though: “I would find it extremely difficult to believe that this could have happened without major demic impact considering that Anatolia was much more developed than the steppe at the time.”

    “Development” can be tricky sometimes and I think that historically this has also often happened. Hungarians don’t have seem to have much (if any) Uralic ancestry, large swathes of North Africa don’t seem to have much Arabian ancestry and Pontic Turks or Turkish-speaking Pontic and Cappadocian Greeks don’t seem to have any Turkic ancestry. Even if we end up finding, say, some 10% pre-Balkan IE Steppe_EBA ancestry in Anatolia, mediated via admixed groups, I’d consider it important enough not to discount.

    “despite the fact that one of the most R1b-rich groups doesn’t speak IE.”

    I’ve brought up the Basques myself but I think this is why time-series and specific contexts are important too. In the particular example, it’s very plausible that Beaker (or even later movements like Urnfield) didn’t affect all of Iberia in similar ways in the long run and didn’t affect language change everywhere, while still being the most likely carriers of IE (one thing you could do is think of those large horizons as multilingual, a bit like later steppe confederacies that could incorporate various groups but it’s essentially a difference of opinion whether you see it one or the other way and the kind of subtlety with regard to aDNA that might never be fully clarified to everyone’s satisfaction). Alberto brought up a similar example with Northeastern Europe and the Trzciniec horizon earlier. It’s plausible that some of the varieties spoken there were dead-ends or their (geographic) history a bit more complex, considering the cultural successions in the rough area of Poland, for example. LBA-IA Europe is barely sampled though so aDNA can’t offer much there yet. With specific reference to Celtic I’d consider the case of Ireland more problematic in a way considering that its R1b doesn’t seem to show great impact of later Celtic-associated migrations and, to my knowledge, the archaeological record doesn’t show great impact of Central European horizons like Urnfield, Hallstatt and La-Tene either. This makes the case of all of “Celtic” as an LBA-IA development a bit less obvious and in need of further disentangling.

    As for your mention EV13 and I1, the association of L51 to IE isn’t a huge leap of faith considering it’s the close sibling of the clearly steppe-related Z2103 (from a phylogenetic perspective, R1a would be harder to associate with IE in that sense, to bring up another online aDNA war regarding Corded Ware lol). Someting similar happened with I2 in farming communities in many areas of Europe where it relegated G, which was the early farming-associated haplogroup, to minority status. For my part, I didn’t generally even focus as much on Y-DNA much here, but more the autosomal side which clearly shows a correlation with the eastern steppe EHG/CHG element in my view.

    But I don’t see these specifics as discounting the general trends. They are still in need of explanation rather than handwaving obviously and we also need more data to clarify them to everyone’s satisfaction.

    Still, can you outline a dispersal of Indo-European that makes more sense with e.g. reference to the Transcaucasus and with no recourse to the steppe for later branches at least? To me, it’s more likely that movements with steppe roots that greatly impacted Northern Europe and also affected Central and South Asia and Southern Europe to a good extent than movements that affected mostly Western Asia and Southeastern Europe and had only very secondary impacts in the rest of Europe are associated with IE. At least the steppe scenario creates potential problems mostly for Anatolia, the Transcaucasus scenario creates problems for most of the range. Unless you do think that the steppe was a secondary stage, in which case the question becomes what kind of movement you think gave rise to IE on the steppe. I think that the Wang paper was an extra blow to Transcaucasian scenaria though as I mentioned previously you could always envision an CHG-heavy group arriving to the Ukraine and giving rise to IE there. But that would have to argue for a hypothetical R1b-heavy CHG group that doesn’t exist in our record and discount the R1b-heavy steppe groups that do exist in our record (nevermind the relevant R1a that also seems to be associated with some IE movements and also appears on the steppe but not the Transcaucasus). Or alternatively the coming R1b-less group bringing IE but having its Y-DNA lineages displaced by the local ones. This is not impossible of course but I think it’s less likely.

40. @ Egg

    You’re correct in pointing out that the most expansive element, genetically is the classic steppe motif (EHG/CHG) which we understand to have formed in the Caucaso-Caspian region. Of course, it depends if this early steppe CCR groups spoke Indo-Hittite, and / or how we treat Sintashta groups for ex,, who are thousand years older and and genetically differentiated to EBA steppe groups. Further – have we figured out what happened to the putatively Tocharian Afansievans ? Where did they go after the Siberian Okunevan’s took over their territory ?

    To elaborate on further points, which you addressed to Marko – we must be careful of false analogies, and treat every case on its own merit. In otherwords, there is little point in drawing parallels to the Arab expansion or Roman Empire as they represent different situations, different periods (much more recent, in State /Empire forming period). All the more if we already have a pretty decent idea of what was occuring in the situation in question. The most commony accepted language models have proto-Anatolian splitting between 5000 and 4000 BC. Whilst the Khvalynsk culture certainly had some local chiefs and heirarchies present, Im not aware of any model as to how we can expect them to have >acculturated< most of Anatolia. In fact, no real model has been offered by Mallory, Anthony, Parpola. etc, apart from a few stray comments about Ezero, Troy and sceptre-chiefs. This problem is all the more acute given that the Khvalynsk culture seems to have ended as a result of the Varna collapse, just like the Suvorovo system did.

    Instead of negative data, we might focus on the data we do have, and I take a somewhat different view – archaeology isn't ambiguous – not if it is understood correctly, and a reasonable body of analysis exists (Eg archaeologists long drew parallels between CWC and steppe-like cultures. It was only post-modernist influences which drew away from that. Jeneusse predicted that BB is a ''steppe-culture'', and not native Iberian.).

    Similarly, if we understand Balkan & Anatolian arcaheology, couple it with the aDNA data points which exist, then we can already formulate some decent ideas.
    So as per your suggestion – ''Even if we end up finding, say, some 10% pre-Balkan IE Steppe_EBA ancestry in Anatolia, mediated via admixed groups, I’d consider it important enough not to discount.''
    And I'd agree. It might suggest that Proto-Anatolians arrived with EEF -rich groups from the northern Balkans or Carpathian, ones that had been in contact steppe groups (the 10%), and then back-migrated to Anatolia in groups, singly and/or in ideology. And it's nice to know that's the archaeology data suggests, someof which i mentioned in the main post above.
    Later, the main bulk of IE could have expanded with the expansive steppe-rich groups who had acquired the language along with the package of ideological and cultural adaptations (kurganism, pastoralism, metallurgy) they have acquired over the centuries of contact with their neighbours.
    As a final point, I would simply highlight we are missing data points for large parts of the western steppe, esp. the important pre-Yamnaya cultures (Cernavoda, Usatavo) which will shed further light on the Yamnaya autosomal genetics which seems to take the focus currently.
    Of course, we have to explain away the large scale migration into Anatolia, Greece and Italy by bearers of genetic signal which is similar to northern Iran, BMAC and the Indus valley, precisely during the Bronze Age. Whislt we might be content with calling this non-IE, what texts did BB preserve which might allow us to confidently asccribe a lnaguage ?

    NB your point on MInoans vsd Myceneans. It is true a major difference is that Minoans have no steppe ancestry, perhaps the horses couldn;t swim yet 🙂 But a more prominent difference is the degree of local Neolithic ancestry the Minoans have c.f. Myceneans.

    NB 2: the above hypotheses are within the occidental framework. I'd like to see further clarification for South Asia, esp. with regard to the temporal contraints required for the Vedas and splitting of I-A. I understand that the elaboration of sampling might attempt to do so in the very near future.

41. @ Toby

    ”your point about Anatolia is interesting, i remember the article someone posted on French thread of Anthrogenica – ”Climatic changes and social transformations in the Near East and North Africa during the ‘long’ 4th millennium BC: A comparative study of environmental and archaeological evidence”

    Yes that s a great paper. Im aware of it

    ”With regard to movements – what about the Suvorovo culture – that seems to be conclusive evidence of steppe migrations into the Balkans. It probably continued on to Anatolia as well. Some samples from Mathieson could be of Suvorovo origin.”

    Yes there are early steppe -admixed individuals in Bulgaria – Varna outlier, Smyadovo. However, this are within local contexts, and appear assimilated individuals. It does not appear that Suvorovo were raiders, but instead traders who gained prestige via their contacts with centres like Varna.
    When the East Balkan centres collapsed, this sent repercussions onto the steppe. In fact, the successor steppe culture – Cernavoda, cannot be dated before 3700 BC, so there is the well-known steppe hiatus in the western Steppe, as there is a hiatus in much of Bulgaria. So there is no invasion horizon to speak of, and the early period of steppe peoples drifting southward came to a precoccious end.
    Successor groups in Bulgarian EBA are EEF and western steppe derived, but not from the Volga – which is where the Indo-European narrative had them placed.

42. @Egg

    Your points are well taken regarding the general trend seen in the Bronze Age, with steppe ancestry defined as CHG/EHG having an overwhelming expansion. I also strongly agree with your point regarding the purported southern influence in the steppe and that the findings of the Wang paper make it exceedingly unlikely that this type of ancestry arrived there any later than, say, the Middle Neolithic.

    What I have a hard time accepting is your analogy regarding the acculturation effected by the Magyars in the Middle Ages. The developmental differences in the Chalcolithic are significant because it would be difficult to bring about such a change without rather sophisticated methods of warfare and tribal organisation, both of which I do not see in place in the Chalcolithic steppe. Note how the expansions of BB & CWC always seem to follow sources of wealth, making it look as though those populations were racketeers or traders whose material wealth led to differential reproductive success vis-a-vis the farmers. They did not install themselves as rulers of local populations by means of warfare. That kind of dynamic I associate more with the later Bronze Age and the Iron Age.

    As for an alternative model, I do not have any strong opinions yet seeing how little data we have from the key locations. I will say however that I find the focus on Europe a bit misguided – the linguistic landscape of Europe seems to have come about mostly through late language sweeps dating to the later metal ages. For all we know Europe outside of the Aegean could have become IE-speaking only very late and with rather negligible demic impact. Anatolian/Greek/Indian samples will be instructive in that regard.

43. @Rob

    “Further – have we figured out what happened to the putatively Tocharian Afansievans ? Where did they go after the Siberian Okunevan’s took over their territory ?”

    That particular aspect is a mystery for me too currently. We have R1b Afanasievo, the Okunevans who were often taken to displace the Afanasievans towards the Tarim Basin in archaeology and apparently did displace them or mix with them to an extent while partially retaining their Y-DNA lineages (some R1b in Okunevo). Then the supposedly Tocharian-speaking mummies come out as R1a-Z93. There are various possibilies there (various IE layers in the Tarim, including Indo-Iranian after all) but we definitely need stronger (and intermediate? Qawrighul has sometimes been taken to be a potentially intermediate culture in the Tocharian trek from memory) connections in this case.

    “And I’d agree. It might suggest that Proto-Anatolians arrived with EEF -rich groups from the northern Balkans or Carpathian, ones that had been in contact steppe groups (the 10%), and then back-migrated to Anatolia in groups, singly and/or in ideology”

    I think that already doesn’t seem implausible considering what the early Balkan outliers look like, whatever their exact fate before we see a clear trend appear in the record. They look half steppe/half farmer and in a generic, genetic sense “quasi-Corded Ware but even more farmer”. In general, the argument in this case is more about the original-original IE locus, in my view, rather than the complex processess than followed.

    “NB your point on MInoans vsd Myceneans. It is true a major difference is that Minoans have no steppe ancestry, perhaps the horses couldn;t swim yet But a more prominent difference is the degree of local Neolithic ancestry the Minoans have c.f. Myceneans.”

    Here I’d have one objection that something like Global25 can’t confidently and definitively differentiate between two already very similar farmer groups in that kind of modelling and might prefer one or the other just due to subtle differences in (W)HG-related ancestry etc. Similarly it’s possible that the extra steppe they have inflates their CHG kind of ancestry so they grab onto some extra CHG-heavy Anatolia_BA ancestry as well, though it’s possible that it also reflects the ongoing merging of local and new, Anatolian-derived groups with the steppe ancestry also arriving around that time. But if your point is more about the Baden_LCA kind of ancestry (Balkan-related farmer) in your graph, whatever we suppose it exactly stands for, I don’t disagree since it’s very possible the (pre-)proto-Greek groups were already mixed steppe/Balkan farmer rather than straight-up Catacomb or whatever culture one prefers.

    @Marko

    “For all we know Europe outside of the Aegean could have become IE-speaking only very late and with rather negligible demic impact.”

    Echoes of the Drews discussion under another post but I can agree with that to an extent or a certain interpretation of it. The Basque case that has been mentioned might be related to what happened in parts of Europe that might have been definitively Indo-Europeanized (or even differentially Indo-Europeanized per your “language sweeps” comment) only with later movements. We don’t have Etruscan or Rhaetic related data either, to be sure, and we know those guys were most likely (just not to fully dismiss the the very few linguists who adhere to some early IE dialect interpretation, here) non-IE, all the way to Central Europe. If they end up having some steppe ancestry too, it’ll be clear that things weren’t that simple, even in the more steppe-influenced Northern Europe where attestation was much worse. A related phenomenon in Northern Europe is that during the LBA – early IA, steppe ancestry steadily goes down while the local very WHG-heavy farmer ancestry goes up which points to ongoing admixture between two somewhat different groups and might point to some ongoing multilingualism along class lines down to a certain period even in specific locales. So far we have only strong hints of this (Nordic_IA, Levanhluhta_IA outliers, Tollense/Welzin_LBA, Baltic_BA-IA) but there’s a Bavarian paper coming out that will provide local time-series. Similarly the Iberian IA paper might help expose some consistent differences between the definitively attested IE and non IE groups we know historically.

44. @ Egg

    I agree that regional time-slices will help clarify the genesis of the know Late Iron Age tribal/ linguistic groups. Until then, it is hard for me to speculate, but agree with the general view that the steppe theory is the most economical one, at least for NPIE. Of course, given that language reflects human behaviour, it might be more complex than what appears to be the mathematically parsiminous model. T/f my approach is to explore all possibilities, as I’m sure most people do too.

    BTW, w.r.t to your discussion with Marko, I think evidence of conquest exists already in the BB period. Look at El Agar, the citadel at La Bastida. Somebody came to rule.

45. @ Egg
    About the G25, my understanding is that it was created to pick up recent drift. When statistically sound, G25 and qpADM more or less match, although it depends on decent number and coverage of samples.

    My tentative models for Myceneans, for example, match what has long been established by archaeologists in a relatively unsupervised manner. Similarly, observe a run for one of the Anatolian MBAs:

    Anatolia_MLBA:MA2203
    Anatolia_EBA 74.6%
    Balkans_ChL 19.9%
    Yamnaya_Bulgaria:Bul4 3.9%
    West_Siberia_N 1.6%
    Yamnaya_Samara 0%
    Armenia_EBA 0%
    Levant_BA_North 0%

    Distance 2.6629%

    If the occidental PIE model is defended*, then I am fairly confident my northeast Balkan / West Pontic Chalcolithic variant will be the path. Wait for the data from Ikiztepe..

    * although given the recent rise of alternative models, it might not ? Either way, for those of us who aren’t ”haplo-centric”, we look forward to any more data.

46. @Rob

    Though one of the things that caught my attention from D-stats comparing Anatolia_EBA to Anatolia_MLBA was that Minoans had a strong signal in the latter. G25 can confirm that signal. Whether it means something or it’s just a random statistical fit due to the few samples available is something that I couldn’t tell.

    Anatolia_MLBA
    Anatolia_EBA 45.3%
    Minoan_Lasithi 24.4%
    Levant_BA_North 14.1%
    Armenia_EBA 13.1%
    Yamnaya_Bulgaria 1.3%
    Hajji_Firuz_ChL 1.3%
    Seh_Gabi_ChL 0.5%
    Botai 0%
    Tisza_LN 0%
    Yamnaya_Samara 0%
    Anatolia_BA 0%
    Balkans_ChL 0%

    Distance 1.0513%

    I’d mostly think that this is rather anecdotal and that Anatolia EBA and MLBA are pretty much the same people, and probably both spoke mostly IE languages. It’s probably better to look for a population shift ca. 4000-3500 BC or thereabouts.

47. Alberto,
    Yes I agree it’s probably just shared ancestry
    4000 BC is the likely timeframe, however such signal could remain hidden in northern or western Anatolia until 2000s

48. @Rob

    On the archaeological front, the only plausible movement from Europe into Anatolia would be the apsidal houses & the megara from Thessaly-Thrace (Karanovo) to Troy, no? I’ve read about recently excavated megara as far south and east as Adana.

    Could also be a movement from Anatolia into Europe though.

49. @Rob

    I think it’s definitely very obviously the case in some areas, like e.g. picking up later Northwestern vs Northeastern European like drift (sources like Baltic_BA a very strong signal there and pop up in all likely or certainly Balto-Slavic related sources and even some pre-Slavic Southeast Europeans to some extent while not liking similarly HG-heavy “Northwestern” populations like e.g. the Levanluhta outliers at all) but I’m not sure if it can differentiate as well with overall very related farmer populations in the sense of picking the absolutely “right” source from closely related ones rather than just trying to balance things. At least I haven’t seen something just as obvious as the example above pop up empirically in that case so I’m skeptical either way of very specific models in the case of e.g. Mycenaeans before we have more data, beyond a generic Aegean-Anatolian like + steppe(+maybe Balkan)-like.

    As for Anatolia, Alberto brings up a good point with a potentially very early arrival into Anatolia itself too, even if it spread eastwards later. I’m not against either in theory personally and I wonder if some individuals like Anatolia_Chl I1584 simply have just extra CHG (or even some EHG coming from populations like Armenia_Chl) or some steppe ancestry too. Though one or two Anatolia_MLBA individuals (I forget which right now) seem to have subtle steppe ancestry compared to Anatolia_EBA ones and extra near eastern admixture might be masking their extra steppe ancestry. But it’s subtle and I’m not 100% sure that Global25 is capturing something real vs trying to balance pull-and-push sources, it’s definitely not a clear-cut case like the other discussed populations.

    As for your “occidental” model, I think it’s correct at least in the general sense that the IE process was a complex cultural one involving various populations, even if they did not all provide the core of the language itself and so weren’t the central, expanding element in a sense. A bit similarly to how the historical branches later came about in their own way via post BB/CW/Balkan BA events, even if their ultimate linguistic ancestor (for the most part and in my view, of course) was the eastern steppe element. It’s true that abstract modelling can obviously simplify those complex processes, as you said.

50. @ Marko
    Yep. There’s also the fluted pottery simiarities, resembling like those of Boleraz , Baden , Cernavoda.
    The houses themselves are hard to track which direction, most of the earlier research is dated, but currently more work is underway.

    Well yes, we know there is a colonization of parts of Europe from Anatolia, however most people perceive this to be non-IE.
    However, if we are to making anything of the -ss- & -nth- words around the Aegean, various interpretations have been made for them, from non-IE, to a pre-Greek (”Pelasgian”) IE, etc.

51. @ Egg

    ”At least I haven’t seen something just as obvious as the example above pop up empirically in that case so I’m skeptical either way of very specific models in the case of e.g. Mycenaeans before we have more data, beyond a generic Aegean-Anatolian like + steppe(+maybe Balkan)-like.”

    I think it can certainly distinguish between Greek Neolithic and north Balkan / Carpathian Middle Neolithic.

    2) The one Anatolia Chalc. does not have any real EHG ancestry. She has loads of extra south Caucasus ancestry.
    The other , Kumtepe 4, is 0.01 coverage, therefore of little use.

52. @Rob

    I think even more interesting than the Anatolian migration to the Aegean and the wider Mediterranean would be the Thracian region. Cernavoda III marks the beginning of the Bronze Age in Europe, but what happens after that remains unclear. It looks like Bulgarian elements spread towards the west, and their encounter with Vučedol in the Carpathian basin results in the formation of the Nagyrev culture. I think the sites in question aren’t described very well in the literature.

    Whether these movements involved Anatolians I do not know for sure, but considering their association with bronze and the distribution of Near Eastern paternal haplogroups in Europe I would think they did.

53. Though, reviewing the Copper and Early Bronze Age cultures in the Carpathian basin, I can’t help but notice that Drews might be right about the relatively unwarlike nature of the inhabitants including Vučedol and its successor Nagyrev-Hatvan well until the late Ottomani culture. This is a bit baffling seeing how the Carpathian basin appears to harbor the greatest number of weapon hoards in the Middle to Late Bronze Age.

    This is in contrast to the South Iberian and Helladic Bronze Age where weapons seem to have been given much more importance already at an early period.

54. @Vara
    Replying late, not sure if you are still reading this thread.

    I’ve heavily criticized the Pontic Steppe theory for a few years as Rob and others know.

    Too many disparate elements to discuss chronology, namely since genetics and linguistics aren’t matching in the least on any level, if most were honestly looking at it.

    I suspect though that I might disagree with you in regards to Iran and Zoroastrianism, but that’s another subject, though intricately connected to the mystery.

55. @Atri∂r

    I think you misunderstood my comment. My comment to Rob was about the origins copper technology which seems to have it’s root in Iran. I do not think PIE was spoken in Iran 10kya. If you ask me, I’d say Pre-PIE or very very Early PIE was perhaps spoken around the wine cultures of Greater Armenia, from Shulaveri to Haji Firuz. Though, I still think Maykop is the oldest culture that fits with the definition of PIE.

    On the subject of Zoroastrianism, I tackled that purely independent of any PIE hypothesis and I do have unique views on it so there is no problem with disagreeing with me. But yeah, Zoroastrianism and Indo-Iranians in general made me move away from a steppe homeland years ago.

56. @ Marko

    Sorry for late reply, I had other matters to attend.
    I am not sure about the direction of relationship between Boleraz and Cernavoda. which mark a large horizon in the south of east-central Europe. There is much new work going on in west Balkans, however ive not seen it for Vucedol as yet. I might have to contact some old friends..
    But about Nagyrev – see the reference on Kiss, Kulcsar et al

    But you’re correct, the ability to command large armies and subject contiguous stretched of territory was a development after 200o BC in Europe, based on models adapted from western Asia.

    About South Iberian BA (El Agar), we are awaiting the aDNA, though the snippets we’ve heard point to R1b, like Bell Beaker. Yet it represents a departure from the BB culture in many respects. So a bioarchaeological analysis will be worthwhile, given the teams already involevd in that area directly, Im sure theyll propose something fascinating.

57. @ Vara; Atri∂r

    When groups first make contact, there is an initial period of ”testing the waters”, so to speak, then a mix of migration, admixture and/or replacement. The initial period could be rather brief to lasting several hundred years.
    With the upcoming data, we can look into how it ocurred in South Central Asia. A guess would be that contact between the steppe/ IAMC & BMAC began c. 2100 BC, and this was followed by a settlement over the mid to late 2nd millenium.

58. @Rob

    “I think it can certainly distinguish between Greek Neolithic and north Balkan / Carpathian Middle Neolithic. ”

    On its own of course but I mean in the relative contributions when combined with other (and often relatively closely-related) sources too. I think it can be very dependent on the sources you pick in that case. To use another example than the OP models, if you use all the ancient sources currently included in Global25, the first one that pops up for the Mycenaean average seems to be Koros_N. But is that because there’s genuine ancestry from Central Europe reflected in that pick or because Koros_N is a relatively HG-free Neolithic source, close to Anatolian and Greek Neolithic sources, and just works well (maybe even better due to potential better coverage etc. I mean I’d doubt the Minoans have any Baden ancestry at all)?

    The model you picked in the OP for Mycenaeans makes sense from an archaeological perspective too for example but is it necessarily the most accurate one in its particular specifics? That’s hard to say, especially since you can come up with very similar but slightly different models that provide even lower distances. Here’s a very similar one for example, with similar distance (they also get some sub-Saharan due to damage but not using it, even if it improves the fits, to keep to the OP modelling):

    [1] “distance%=2.5669”

    Mycenaean

    Minoan_Lasithi,82.4
    Beaker_Hungary_o,17.6

    or alternatively with a slightly better fit

    [1] “distance%=2.1181”

    Mycenaean

    Anatolia_BA,40.2
    Minoan_Lasithi,31.2
    Baden_LCA,19
    Beaker_Hungary_o,9.6

    An Aegean-Anatolian-related source (one assumption here is that the pre-Greek substrate in southern Greece would have started to resemble the Minoans more over time compared to the less CHG Peloponnese_N average; but we already have the Peloponnese_N outlier that’s even more CHG than the Minoans so it’s not that outrageous either) and a Carpathian source (which would also accord with one of your preferences for the northern sources but this one is presumed not admixed before moving on) that resembles various steppe and steppe-derived MLBA groups. (adding Anatolia_BA and Baden_LCA improves the fit here but marginally so while potentially overfitting)

    Which of the two models corresponds more to what happened? I’d have no clue right now either way. All I can tell is that we have a combination of a heavily Anatolia_N-related source, a Bronze Age source with a decent amount of CHG and a source with at least some steppe and that it’s likely that those are shared so that the sources would be strictly speaking two when the mixing happened. That’s basically my point here, that it’s hard to distinguish from very related models in a case like this one until we have earlier or more data thought it’s funny that with just four samples things get good enough that there’s likely not huge room to disagree on the broad model and if we had the Armenoi_Crete sample, things might have been even more illuminated.

    Btw, for fun I decided to make a 2/3 Baden_LCA, 1/3 Beaker_Hungary_o individual and (probably unsurprisingly considering they’re similar steppe + Balkan-Danubian mixes) it comes close to some of the more southwestern Beaker, Balkan_BA and Hungary_BA samples. This kind of individual might very well represent the proto-Greek speakers that came to the Aegean but the exact situation might be a bit different too with the relevant individuals being more steppe-shifted. To reiterate the previous, it’s just hard to pick a source from a steppe-to-Hungary/Balkan cline yet imo even with the general power of something like Global25 in detecting more recent drift (which power does seem to be lower in some cases than others per above) and that’s where I’m generally coming from in regard to this kind of point.

    @Atri∂r

    “namely since genetics and linguistics aren’t matching in the least on any level”

    Care to elaborate on that because I’m not fully understanding the objection? I think there are a lot of things to still unpack with regard to those correlations but the general distributions arising from a steppe model (with Corded, Beaker, Balkan Bronze Age, post-Yamna steppe cultures as vectors) don’t seem to be that incongruent either compared to some of the generally preferred trees/relationships?

    In my general view, you have a positionally (both geographically and linguistically) westernmost Italo-Celtic, including the other early branches of the western area that are hard to classify exactly within that categorization, a Balkan IE that’s positioned between Italo-Celtic to the West and Indo-Iranian, Indo-Iranian between Balkan IE and Balto-Slavic and Balto-Slavic being on the opposite geographic-linguistic end of Italo-Celtic. Germanic seems weird but might occupy a position between Balto-Slavic, Indo-Iranian and Balkan IE. In a broad cultural sense then Italo-Celtic might correspond to Beaker, Germanic between Beaker, Corded and Balkan Bronze Age, Balkan IE between Balkan Bronze Age and post-Yamna steppe, Indo-Iranian between post-Yamna steppe and Corded Ware and Balto-Slavic in Corded Ware. I have some theoretical problems with timing and some associations too but I don’t think that this kind of situation is unbelievable with regard to traditional models, considering those groupings were often in touch with each other and so might have slowed down linguistic separation too despite often hostile relations (Beaker replacing/mixing with Corded Ware in areas, Corded Ware-derived forest steppe cultures replacing/mixing with relict Balkan IE on the steppe).

    Of course everything in the above might end up being bullshit but the point is more that it’s not an unbelievable or outrageous model that requires much special pleading based on current data and corresponds to linguistic considerations of relationships and trees too. The association with specific Y-DNA (that you said you privilege) lineages doesn’t seem that out of place either with certain, different branches of R1b associated with Italo-Celtic and Balkan IE, certain, different branches of R1b and R1a associated with Germanic and Indo-Iranian and certain branches of R1a in Balto-Slavic.

59. @Vara
    I remember some of your comments on other sites and remember agreeing with most. As for its start, I hold that Zoroastrianism started in E. Iran. I’ve been to a few Zoroastrian temples and most are just carrying the vestiges of their past religion. I have found excellent 19th and 20th century books in their libraries though that offered interesting insights.

    @Rob
    Yes, my position since I first brought up with you hasn’t changed much. I don’t think the dna about to be released will change anything. It will only open Harvard’s eyes to history already documented in books and folklore. They’ll figure it out after they’ve finished poaching the comments sections of the internet.

    @Egg
    I exaggerate the point to illustrate it. Here’s how I think, if it helps. Y-haplo more important to tracking movements than aDNA. ADNA more important than mtDNA. And language evolves faster than all of these groups.
    Here’s some hard info than many ignore.

    – no r1a in W. Europe
    – r1a heavily present in Estonia, Hungary, Altai
    – hardly any r1b in India
    – most Basque r1b
    – language changes fast unless it is recorded. Predicting languages from 3-4Kya without records is a fool’s errand.
    – aDNA can change at any moment when a new ancient sample is found.
    – academics make up stories for samples. I.e. this sample is a merchant from a distant land.

    These eclectic bullet points are all to say one thing. The I-E tree is wrong. Berkeley has begun to catch on and I recently read a good paper on Italo-Celtic.

    Also, time for Indo-Uralic to make a comeback.

    I’m pretty sure where it ends, no one is going to be happy.

    But approaches like Rob’s are the best, because they layout the information with little bias, as it should be.

60. Robert,

    You mentioned that there was a R1b-U106 sample (I4178) in Chalcolithic/ BB Hungary. However, there is no confirmation that this specific sample is R1b-U106.

    Alex Williamson, the owner of Ytree.net, took a look at this sample. He concluded that there is not sufficient evidence that this sample is R1b-U106. He is not even sure this sample belongs to R1b-L151:

    ”
    I don’t have any confidence in calling I4178 as R-U106 > Z18. There’s only one read and it is a G > A mutation at the last position. This sort of mutation is typical of damaged aDNA. I can provide you a list of a dozen other calls for this kit that are downstream of L51 that are also supported by just one read and they’re all from different branches. As it is, I’m not even sure he belongs to R-L51.”

    https://anthrogenica.com/showthread.php?13555-R1b-U106-from-%DAn%26%23283%3Btice-Culture-2200%961700-BC&p=373568&viewfull=1#post373568

61. @Egg

    Quick follow-up to your last (was typing from phone earlier/not user-friendly).

    There is too much bias that has entered the discussion by associating genes with languages. Correlation is fine, but most of the recent genetic papers have high levels of bias and frequently associate genetic trees with linguistic trees.

    Also, it is a sounder approach to work language trees from the most recent to the most distant in my opinion. Right now, the reverse has been done, and I blame this for most of the errors.

    Lastly, there is no way to know what BB or CWC spoke until we find records. Educated conjecture is fine, but in my experience, these guesses become repeated so many times until in time, they erroneously become accepted as fact.

    We have I-E records in Greece and that is where I-E should begin in Europe, until further proofs are found.

62. @ Tesmos
    Thanks for that, i will check Alex’s site in detail. From an earlier glance, I did note it to be very useful & well presented.
    I shall update the post accordingly. I guess so far, then, we have U106 in Unetice and the transitional/ post-CWC burial from Scandinavia as the earliest appearance of U106. Perhaps more sampling form the southern Baltic will help.

63. @ Egg
    I take your points, but it isn’t about ”picking” models. A sensible model is easy enough to construct – choose sources from places which (literally) surround the target & whose dates are historically feasible. So to model 2nd mill Myceneans, one chooses 3000 BC source of Left pops (whether qpADM or nMonte) – Greece LN, a central Balkan group, central-west Anatolia EBA, etc.
    Anyhow, as you suggest ” a heavily Anatolia_N-related source, a Bronze Age source with a decent amount of CHG and a source with at least some steppe ancestry” is the crux of it. There might be little point in trying to discern the exact source group because with autosomal analysis, as we can see, the movements from the North and from the East occurred continuously rather than a single wave, with an accretion of such ancestry. Therefore Uniparentals might be more instructive as to how things went down at the local scale. It might seem like knit-picking, but it’s actually very important in formulating a final understanding.

64. @Atri∂r

    Thanks for the specifics. I’ll make a few points as I see things for your consideration. 😉

    “– no r1a in W. Europe”

    I’m not sure why we should expect it to be found in Western Europe (west of Germany or thereabouts that is and I assume most of the other major R1a-related clades were carried by Scandinavian migrations unless there are some minor clades in Western Europe I’m not familiar with though I don’t think they’d change your overall point here). It doesn’t seem to have been a particular prominent clade on the steppe and neither was L51 for that matter compared to Z2103, even though the last two are siblings. It’s possible they only became prominent after certain, lucky expansions out of the steppe which would have meant that they had the opportunity to multiply their Y-DNA lineages in Europe even beyond their representation on the steppe. In that sense Marko’s mention of EV13 and I1 are important too, they seem to have undergone something similar during the Bronze Age.

    ” – r1a heavily present in Estonia, Hungary, Altai”

    I assume some of those clades are Balto-Slavic related and some Indo-Iranian and entered Uralic and Turkic (etc.) speaking populations from post-Corded Ware movements? Though I’m not against certain branches of R1a being associated with the spread of Uralic etc. alongside other clades too if that’s what you were getting at here, even if I don’t know the phylogeny that well to see how possible it is.

    ” – hardly any r1b in India”

    There is some Z2103 though, right? Much like in Iran etc. I think Z2103 must be the only major clade that’s present throughout the IE and former IE range, even though it seems more associated with certain branches like Balkan IE and Indo-Iranian to a lesser extent, which might make sense considering its early prominence on the steppe. It was carried by pretty much all movements even if those were dominated by other clades too.

    “– most Basque r1b”

    Yes, for sure, a very strong founder effect but linguistic survival nonetheless. We had a previous discussion on that and what it might imply (in my view, communities affected by Beaker that nonetheless didn’t undergo long-term linguistic change but I can understand why people can easily disagree there; it’s just an assumption right now!). Iron Age Iberian data will help for sure, especially the comparison between contemporary IE and non-IE speakers from the same general area.

    ” – language changes fast unless it is recorded. Predicting languages from 3-4Kya without records is a fool’s errand.”

    I’m not sure if this is a hard law per se though. If a language is recorded in a region of intense contact with others, it might also change faster even if recorded, no? But yes, it’s more about general correlations than absolute certainties like in the case of Mycenaean Greek where Linear B sets an indisputable record (and even that has been disputed by some in the past even after the original decipherment, if we wanna go that deep!)

    ” – aDNA can change at any moment when a new ancient sample is found.”

    That’s true to an extent (as in South Asia for example) but I don’t think we should undermine ther stability that can exist at least in well-sampled eras/areas. For example would you agree we have a decent understanding of what Neolithic Europeans (at least west of a certain point) were like in general that’s not likely to change in radical ways in the future?

    ” – academics make up stories for samples. I.e. this sample is a merchant from a distant land. ”

    Sure, though I imagine that’s also why they work with archaeologists and linguists on the major papers, to try and make a general case for various things like that. I think you exaggerate a little bit with this point though.

    “The I-E tree is wrong. Berkeley has begun to catch on and I recently read a good paper on Italo-Celtic.”

    What are you referring to here? If you recall that paper, feel free to link to it btw.

    “I’m pretty sure where it ends, no one is going to be happy.”

    What do you mean by “where it ends”? The precursors of “Indo-Uralic” or what?

    “Also, it is a sounder approach to work language trees from the most recent to the most distant in my opinion. Right now, the reverse has been done, and I blame this for most of the errors. ”

    What do you mean by recent and distant? Are you referring to the year of tree calculation or to the chronology of branch attestation?

    “Lastly, there is no way to know what BB or CWC spoke until we find records”

    I don’t think we’ll ever find decipherable records from Northern Europe from that period in time so as much as we can disagree about things, for my part, I’m not sure I like that kind of complete and utter agnosticism when we can at least connect certain broad branches with certain broad horizons even if the specifics are uncertain. It’s a bit like saying “why bother” in that kind of case. It’s possible (very likely even) that e.g. all of Beaker wasn’t responsible for producing surviving branches but out of what general horizon could Italic and Celtic have come from other than it?

    “We have I-E records in Greece and that is where I-E should begin in Europe, until further proofs are found.”

    Sure, though it depends on what you mean by that since e.g. I know pretty much no prominent archaeologist or linguist who ever thought that Greece circa MBA would have been an entry point of IE into Europe or anything. And we know that the clear difference between Minoans and Mycenaeans is the EHG/CHG (eastern) steppe element right now, anyway so I don’t get your point here. Even if what you mean is that it might have entered Greece from Anatolia rather than the Balkans (which I currently doubt considering Crete_Armenoi to stick to genetics), it still likely had to come from somewhere around the steppe.

    Out of curiosity, what’s your very general scenario on the origins and spread of IE? Do you have one at all or just waiting to see how things go?

    @Rob

    “It might seem like knit-picking”

    That’s fine because I was essentially nit-picking too for the sake of discussing a few specifics if anything.

    “A sensible model is easy enough to construct – choose sources from places which (literally) surround the target & whose dates are historically feasible”

    Sure but my point as you put it was more about “little point in trying to discern the exact source group”, or rather than little point (though I assume you meant “currently” anyway), an inability to really differentiate between different but very similar sources that can give similar results too in this case, rather than the broad model. Early (and/or different, if somehow Grave Circle samples end up being more illuminating) samples might help even more if the Y-DNA ends up being biased one way or another too and I don’t think we disagree on the general stuff in this case.

    Btw, does anyone know what’s happened to paper supplements on biorxiv? They seem to be gone in general?

65. @ Egg. Yeah I just noticed the Supp. is missing now from Dzudzuana paper..
    Not sure why

66. @Egg
    You give good scenarios for genetics, but not for language. But so many assumptions required; too many for my poor brain.

    Also, founder effect is over-used as an argument and a weak one imo; Founder effect is present sometimes, sure, but it needs to be demonstrated with proofs (i.e. early populations of Utah Whites are a solid example). It’s not a panacea for all unknown situations and it makes no sense in respect to the lack of r1a in Western Europe. Why is the presence of r1a in Europe (yes, I exclude recent Norse DNA) almost non-existent in W. Europe?

    I prefer Occam’s razor. And I suspect r1a was a movement that went E to W, and r1b moved in accordance. Such movements, as an aside, have been the pattern on the Steppe over and over again. Once that is understood, then certain things become very clear about the I-E split.

    If you really are interested, I’ll need to dig up that Berkeley paper. Key words: Berkeley School, Italo-Celtic, Indo-European tree.

    I do have a model. With it, predicted the IAMC before Harvard Paper, Mycenaean and Hittite J2 before the samples came out, the role of Turan aDNA in South Asia, and feel confident about my understanding of R1a, and its role in I-E. But I have a second model as well, as some key data is still missing and can alternate the vectors.

    It is a pitfall to use genetics as you mention in respect to the Mycenaean and Minoans. The situation is not as certain as I understood your statement to be. Need more samples. And older samples. Where are the generations of Mycenaean that have 50% Steppe?

    The subject is so heated and discussed because it is related to the cultural identity of hundreds of millions. Significant change to one’s sense of identity does not occur without backlash or resentment. And scientists and historians are not as immune to this as they might hope to be. Hence bias.

67. Supplementary material on all papers on biorxriv now seems to be inaccessible.

68. Huh checked again, it’s available for the “ Neolithic Britain” paper by Brace et al , at least

69. @Rob

    No problem! I agree, we need more samples from the southern Baltic area. We have a significant amount of Bell Beaker- and CWC samples and none of them belong to R1b-U106. Suddenly, U106 shows up before 2000 BC in Scandinavia (average date is around 2150 BC) and Unetice (date for Unetice sample is 2200-1700 BC to be precise, so average date for this sample is around 1950 BC). It’s still a mystery how U106 arrived in Central Europe and Scandinavia.

70. @ Atrior & Egg

    To add on some of your discussion

    – Atrior, the lack of R1a in western Europe reflects the fact that CWC – the original vector behind R1a expansion acrross Europe – did not expand beyond the Rhine (apart from a few, AOC-beaker, E-W oriented graves in Paris basin, etc).
    The Netherlands was the (natural) limit for CWC to the West, however, it was soon supplantd by Bell Beaker, and then MBA groups, and so forth.
    These, and subsequent expansions, eg Celts, Franks, do not seem to have been R1a-rich, hence a continued relative dearth of R1a.
    Still, there is some 10% R1a-L664 in Netherlands.

    – Egg, I agree with Atrior that there might be more to Minoans vs Myceneans than steppe and not steppe, but we’ve discussed that already amply.

    Anyhow, I think when more data is released from key regions which will lead to future discussions. Hopefully soon

71. @Atri∂r

    Depends on what you mean by E. Iran. I don’t think the Median or Bactrian Homelands are the right ones but rather the one in Drangiana. I believe West Iran had little to do with Zoroastrianism before the Medes and was most likely on the Indo-Aryan side.

    @Rob

    I don’t disagree. Though, I do not think that’s how the Indo-Iranian ethnogenesis began. IMO, the OIT crowd has done enough work criticizing the chronology but they lose credibility when most of them want to tie the Vedic homeland to their backyards.

    IMO, I think the approach should be comparative literature/religious studies> archaeology>genetics>>>linguistics. However, the problem with this approach is that any preconceived notion or incompetence when dealing with a ritual or burial within a certain archaeological complex can lead one down a rabbit hole, eg. Anthony incorrectly claiming that Srubnaya dog sacrifices were the same as those that aren’t actually found in the Rigveda . But still, there has to be a good explanation to why culture xyz is considered PIE or IE in the first place other than because ancestry x was abundant there.

    Merry Christmas!

72. Belated merry Christmas and happy holidays to everyone

    @Atri∂r

    “It’s not a panacea for all unknown situations and it makes no sense in respect to the lack of r1a in Western Europe. Why is the presence of r1a in Europe (yes, I exclude recent Norse DNA) almost non-existent in W. Europe?”

    Rob answered the specifics about R1a within Europe west of the steppe but the more pertinent and unanswered question you might have implied here is why it expanded with only via that movement and not any others from the steppe? I agree that we need more data to really clarify the locus of R1a within the steppe. We see some early steppe-heavy individuals with clearly assimilated farmer lineages for example so it’s not inconceivable that R1a was also local to parts of the Dnieper-Don area (and perhaps originally even further north of the steppe, on the forest steppe) too, rather than arriving together with the eastern steppe EHG/CHG kind of ancestry (and not to repeat here the other possible scenario of separate CHG arrival that I mentioned above). It’s only one individual so far with the relevant clade after all.

    “I prefer Occam’s razor.”

    We have both R1a and R1b on the steppe so far and the one branch of R1b that’s missing from the record right now is the sibling of the one that isn’t missing, so I’m personally not expecting to find it west of the Carpathian Basin and unconnected to steppe movements, whatever the specifics end up looking like. We can easily call that a simple solution with few assumptions too, even if it turns out to be mistaken in some way.

    “And I suspect r1a was a movement that went E to W, and r1b moved in accordance. Such movements, as an aside, have been the pattern on the Steppe over and over again. Once that is understood, then certain things become very clear about the I-E split. ”

    What do you mean by this? That R1b moved from west to east? Overall it doesn’t seem unlikely but for the relevant IE-associated subclades, it doesn’t seem too likely with Z2103 having a clear presence on the steppe but not west of it and L51 being its relatively young sibling? How do you think these opposite movements pertain to the IE split?

    “If you really are interested, I’ll need to dig up that Berkeley paper. Key words: Berkeley School, Italo-Celtic, Indo-European tree. ”

    I think I found what you meant. It’s a paper by Calvert Watkins where he apparently posits no intermediate Italo-Celtic stage after PIE. It’s an interesting position and, not being a linguist, I wonder if it’s possible that Celtic itself is simply just a grab bag of “relatively archaic (non-innovative) IE dialects”, alongside other non-easily categorized varieties like Lusitatian? This might explain e.g. the R1b in Ireland showing no secondary influences on the same level as the Beaker impact that might explain the arrival and spread of a new kind of “Celtic” (though even a subtler influence doesn’t necessarily mean no linguistic change either of course).

    I actually didn’t find the article itself with a quick look but it’s mentioned in “The Question of Italo-Celtic Unity” by Alan Yates which also mentions alongside it: “Based on a consideration of this evidence, Watkins 1966 firmly rejects the notion of Italo-Celtic unity, concluding: “The only common language from which both Italic and Celtic can be derived is Indo-European itself” (43-44). Yet the same evidence has led Cowgill 1970 not just to posit unity, but to forward a plausible migratory theory to account for it. His theoretical model assumes an extreme western Italo-Celtic group speaking a very late form of Proto-Indo-European, including only very basic developments (such as laryngeal loss) and its distinct set of shared innovations”

    This kind of scenario is what I envisioned too based on how I’ve generally understood Indo-European interrelations (as I mentioned above, with “Italo-Celtic” being the most extremely ‘western’ group, both geographically and how its linguistic relations look like) and the recent aDNA evidence which make an association of Italic and Celtic with post-Beaker cultures the most likely. Even if an intermediate Italo-Celtic stage didn’t really exist and much of the Beaker area didn’t leave linguistic descendants (due to assimilation by locals, secondary spreads etc. not impossible), I don’t personally see what other scenario makes as much sense as some surviving IE-speaking cultures in former Beaker territory being responsible for their spread.

    “I do have a model. With it, predicted the IAMC before Harvard Paper, Mycenaean and Hittite J2 before the samples came out, the role of Turan aDNA in South Asia, and feel confident about my understanding of R1a, and its role in I-E. But I have a second model as well, as some key data is still missing and can alternate the vectors.”

    What do you mean with the first three? You should elaborate a bit on what you mean by the role of R1a in IE and explain your primary and secondary models a bit. Nothing too specific, just the gist since you keep hinting at it but you’re giving nothing away. I’m really curious at this point!

    “It is a pitfall to use genetics as you mention in respect to the Mycenaean and Minoans. The situation is not as certain as I understood your statement to be. Need more samples. And older samples. Where are the generations of Mycenaean that have 50% Steppe? ”

    There might be no MBA Greek speaker with 50% steppe if the intrusion was during the Early Helladic. It really depends. Nonetheless, subtler than in Northern Europe or even Iberia as it currently looks like (and I’ll repeat “currently” because the Crete_Armenoi sample might have some 20% Yamnaya-like/25% Steppe_MLBA-like ancestry), it’s clear than an EHG-carrying element of some sort was added to the Mycenaeans if you compare them to the Minoans. A population carrying steppe ancestry seems like the simplest solution especially considering the likelihood that the Minoans were non-IE while the Mycenaeans certainly were. More and older samples will help clarify the situation but I don’t see how they’ll change this basic difference imo (whatever other kinds of ancestry came along with the EHG one that might be harder to disentangle currently, like Rob and I discussed above).

    “The subject is so heated and discussed because it is related to the cultural identity of hundreds of millions. Significant change to one’s sense of identity does not occur without backlash or resentment. And scientists and historians are not as immune to this as they might hope to be. Hence bias.”

    I don’t disagree with that [though on a personal note, I think my main bias in this specific area – as opposed to others 🙂 – likely comes from having been introduced to the PIE homeland question via Mallory who, politely agnostic as he’s generally been on it, generally always seem to have preferred a steppe homeland (and in the 80s, in a comment to a paper by Anthony he even focused on the importance of the eastern steppe as opposed to just the Dnieper-Don area that Anthony seems to have preferred at the time), rather than some personal investment which I assume would have leaned more in an EEF direction, anyway, unless I somehow liked the aesthetics of the steppe scenario more] but when you get down to it, in the specific case of a steppe scenario we have a population that can be called “European” in the HG/basal sense (but very differently so compared to moderns in the ANE/non-ANE sense) and in the geographic sense (though at the very margins, in south Russia, so still “Eastern European” if one wants it to be so), even though they didn’t exactly respect our modern continental constructs, but also one that’s not exactly more linguistically related to any of its descendants than another and that left varying and comparable amounts of ancestry in both European and non-European areas. Basically, from my perspective, it’d be weird to somehow “co-opt” that kind of ur-population that’s related to most of Eurasia (linguistically and genetically), anyway, and the on-going answer to the academic question of “where and when” is interesting enough.

    @Rob, Matt

    Yeah, strange stuff. I can understand with papers that have been published but some are unpublished so far to my knowledge, like the recent Wang et al. one.

73. @Vara, Rob, Egg

    Happy New Year 2019 to all!

    I had wrote a very long response to all a few days ago, but lost it before posting.

    At any rate, resolution for this year will be to start posting/vlogging more. TBD.

    Btw, looks like meltdowns have begun on other sites. Will 2019 be the year?

74. Happy new year
    I’m not sure what new papers are due. A good start might be final versions of last years pre-prints and the abstracts we’ve seen, to break the silence

Comments are closed.