In this, I'm with the mainstream of workers in the field, who do not see either the Anatolian nor the Armenian hypothesis as particularly serious, as both have very challenging problems that their theories do not even attempt to account for. The Armenian hypothesis is largely based on linguistics, and is—at best—handwavey about archaeology, while the Anatolian hypothesis does the opposite; proposes an archaeological solution that ignores linguistics. The only model that has the potential to be taken seriously is the Kurgan model. I'm kinda sorta reviewing the book (not that it's new, but I just read it just now) The Horse, The Wheel, and Language: How Bronze-Age Riders From the Eurasian Steppes Shaped the Modern World by David Anthony. He adds a bit to the discussion by drawing on some archaeological data that has been largely unavailable to the West in the past because of the Iron Curtain, as well as more work that he was personally involved with in detailing new Pontic-Caspian steppe fieldwork. And although it post-dates this book, genetic evidence, particularly the R1B1 Haplogroup, but even moreso this really quite new work further corroborates the Kurgan Theory. But it's a mistake to think, as the copy on the cover of the book I'm talking about says, that this is really new. The Kurgan Theory was first put forward by Lithuanian archaeologist Marija Gimbutas in 1956, further reinforced/reiterated by J. P. Mallory in the 80s, and now with the 2007 publishing of Anthony's book, it's even more solidly mainstream than it already was. Otto Schrader proposed the same homeland, using at least some of the same arguments, back at the very beginning of the 20th century, and even by Theodor Benfey some decades before that. The theory has been augmented by more information, but remains little changed since at least the 50s, when Gimbutas first laid it out.
In spite of the fact that ever more facts are brought to bear on question, it remains, of course, an unproven and unproveable hypothesis, a just-so story, if you will. In Anthony's book, the just-so stories tend to be on a more micro-level, and I'll mention a few specifically in a bit. I will point out that, of course, in this discipline, a just-so story isn't necessarily untrue just because it is unproven. Barring the completely shocking discovery of clay tablets or other decipherable texts in languages and from cultures believed to be illiterate, we will never be able to prove that Proto-Indo-European is associated with the steppe cultures, or how the languages spread. Anthony's work is clearly influenced by a number of ideas that are somewhat faddish these days—some of which the genetic information listed in the link above—are already partly refuted, or at least its implied that they are not correct. Among these faddish ideas that Anthony buys into are the following:
- Although he makes an attempt to step outside of the western box and accept the old Russian paradigm of folk migration, he still is very reluctant to do so, and fills such talk with caveats, virtue signaling, skepticism, social linguistic spread models that don't require much in the way of actual migrations of actual people, etc. Of course, the link above clearly indicates that a lot of Yamnaya people actually moved into areas that shortly thereafter emerged as speaking a differentiated Indo-European language.
- He believes that the original "Old European" languages of, for example, the large and long-lasting Cucuteni-Tripollye culture probably had a very early branch of the Afro-Asiatic language. This is based on a single possible loan-word (which Mallory mentions in his book decades earlier); the word *tawr- for bull. Given that their is no other evidence whatsoever for an Afro-Asiatic language in Europe, this seems far-fetched. Granted; the linguistic picture of Old Europe is very poorly documented, and therefore anything said about it is by necessity at least somewhat speculative, but if I were a betting man, I'd put money on it being related to the pre-Greek substrate language, such as Pelasgian, which might be related to either the Minoan language of Crete prior to the arrival of primitive Greek on Crete, or to the somewhat hastily proposed Tyrsenian language family which includes Etruscan. Both of which, I'll add, are geographically much closer to the Balkans, and which are observable not Afro-Asiatic, being rather either language isolates, or more likely members of families that were already in decline with the advent of writing, and therefore remained poorly documented and somewhat anonymous.
- The wheel had to have come from the Near East. Actually, there are three competing theories for where wheeled first developed. As it turns out, the Pontic-Caspian steppes and the Proto-Indo-Europeans of the Yamnaya Horizon are one of those three candidates.
- In general, Anthony seemed to be very reluctant to give too much credit to the Indo-Europeans about all kinds of things, but as we chip away at the biases and prejudices in the science through new information, we're coming around more and more to reiterating the ideas that were popular in the first part of the 20th century after all. Heck, I expect very shortly to see a re-embrace of the word aryan as a descriptor of the Proto-Indo-Europeans after all, since reflexes of it appear all over the place, even if they're now largely discounted. The "evidence" used to discredit this once wide-spread linguistic belief appears to be little more than, "the Nazis liked to use it, so therefore it's racist, and therefore its out of favor." While this may well do for a political idea, anyone who knows anything about how science works should know that this is no good.
- Although it seems possible—maybe even probable—Anthony's identification of the Maikop culture as non-Indo-European and probably proto-Kartvelian in nature (although the Pontic or the Caspian language families, also indigenous to the Caucasus and appearing to be isolate-families unrelated to each other or anyone else could do so as well) it remains speculative. I will add that although there are clear ties between Maikop and some of the steppe cultures, they do seem to be economically unrelated, at least. I suspect that Anthony is right in decoupling them from the PIE family, but I'll also note that this is a unique addition of his own.
- Anthony's description of how the languages may have corresponded with specific cultures, especially the spread of IE languages into Europe, and which languages later developed from specific archaeological material cultures is, by its very nature, a just-so story. Even if he's right, it's all speculative. He even acknowledges this tacitly in the text by sprinkling these discussions with a lot of "probably" and "possibly" and "perhaps" caveats.
That said, his model is quite nice, and describes the development and earliest spread of the Indo-European languages quite neatly. Lacking another model that does so any better, there's no reason not to accept Anthony's description of how it all worked. A few places where it differs from Gimbutas' own description:
- He believes that the Bug-Dniester culture may have been linguistically part of the same complex from which PIE derived, but that it probably was absorbed linguistically into the language spoken by the Criş farmers (which, see above, he equates with Afro-Asiatic, I presume it's more likely that it was a native Aegean/Balkan language instead.) In fact, much of the cultures associated with IE archaeology, he posits are too early to be anything other than pre-PIE cultures that were on their way to developing into PIE.
- The Suvorovo-Novodanilovka culture, which seems to have replaced the lower Danube Old European cultures like Cernavoda he associates with the very first group to split off of archaic PIE—the ancestors of the Anatolian branch. He also posits a minor climatic change (not unlike the Little Ice Age) as causing environmental shifts as the cause for this change, more than any cultural effect. Major changes to the Tripolye towns and on the steppes themselves seem to have happened concurrently.
- He has five "final Eneolithic" cultures that he posits as being on the very eve of classic PIE: Mikhailovka I, Post-Mariupol (which he admits is an awkward name for a culture), late Sredni Stog, late Khvalynsk and Repin. It's from the latter that he derives the Afanasievo migration and founding of the Tocharian branch, although he admits that continued contact with subsequent Yamnaya peoples flowed eastward for some time. Also on this eastern frontier of the Caspian steppe region (still on the west of the Urals, though) is where he places the domestication of the horse, first for hunting, and later for riding, earlier in the Neolithic. It's also here on the eastern frontier that true pastoral nomadism, with mobile dwellings, is established, which brings us into the Yamnaya horizon.
- He believes that the material cultural same-ness of the Yamnaya horizon (which is partially due to terminology; the Yamnaya horizon certainly had significant regional variants all along) probably disguise the spread of a "prestige dialect" from the east, which homogenized the PIE-speaking area. This actually isn't necessary, since we have no idea to what degree mutual intelligibility existed among the dialects already extant in the area.
- Although he replaces Mallory's rather vague story of steppe intrusions into the Balkans with specific Yamnaya sites in the Balkans, he still maintains an air of vagueness about it when he asserts that they "probably" are the roots of the Celtic and Italic branches, and maybe the German ones too after they moved up into the Corded Ware horizon. He later makes a vague case for Greek and Armenian coming out of the steppes a little later, but this is much more vague; and doesn't address at all where the Thracians, Dacians, Illyrians and Phrygians may have come from, since they have to be derived out of the Balkans as well. For that matter, the Baltic and Slavic families get barely a mention.
- He also tells an interesting story of the development of Proto-Indo-Iranian and its eventual split into specific Indic and Iranian language groups, but by this time, he's clearly just making assignments. Much of what he said was proposed, and then criticized by others as being too vague, years earlier; i.e., the identification of the Sintashta culture with Indo-Iranian, Andronovo with Iranian specifically, and some southern Andronovo variants as well as the filter through the BMAC as the vector for Indic. There isn't really any new archaeological evidence that he brings to bear, although he is a bit bolder than Mallory was willing to be in assigning ethno-linguistic identities to various material cultures.
All in all, a good book. An interesting one. I may have buy a copy to be a companion piece to Mallory's book, which I own and pull out and read every few years or so.