The Early Iron Age Revival of the State, XXX: Quoi?

 

This one originally qualified as a response to something I read online: not here, of course: Over at Quora, where the best of the resident historical geneticists, Ygor Coelho, accepts the final collapse of the "Yamnaya expansion" thesis as far as it concerns ancient Anatolia --and then reconstructs it. 

Some time ago I wrote that my understanding about the origins of the Indo-European language family and the early Indo-European migrations, after reading many scientific papers about the archaeogenetic findings in connection with the archaeological ones, had been evolving to favor the Pontic-Caspian Steppe hypothesis, but not in its classic “Yamnaya hypothesis” (too late to be really representative of Proto-Indo-European, as opposed to some Indo-European branches, possibly those ancestral to Greek, Armenian and perhaps Albanian), and also in complete disagreement with those population geneticists that were interpreting the data as an evidence of an origin of Proto-Indo-European south of the Caucasus, probably close to Armenia.

It could also qualify as a book review, in that I took the decision to spend a lazy Saturday working this material over as an invitation to read Eric Cline's  After 1177: The Survival of Civilisations. Also, Narendra Modi is going to win re-election in India on his "Sure would be a shame if an ethnic cleansing were to just happen around here" platform, and if I can't do anything about that, at least I can direct some impotent aggression towards his Hindutva loons. 

So, first, Professor Cline. I read  1177 BC: The Year Civilisation Collapsed as a somewhat less than passionately felt book, and that is even more true of  Survival. That's not to say that it is a bad read, much less bad scholarship. I see omissions and neglect in the literature, But Cline has a professional expertise in the field so that is much more likely to be my misperception than reality, except insofar as I was hyperfocussed on his treatment of the brilliant Saro Wallace, and found his use of her work shallow. To be fair,  Travellers in Time came out too late to be considered; but Cline's whole monograph is permeated with the idea of a "successful collapse," and Wallace offers a mechanism for it that would explain what Cline finds so mysterious about the Phoenician anti-collapse and which might have come into his treatment of Israel if he had taken Finkelstein more to heart. (A redistribution of everyday economic activity across elevations enriches the "Phoenician" city states and makes the Kingdom of Judah possible). 

Oh, well, maybe I'm just white knighting it But, you know, Cline only catches fire when it ambles off the reservation to talk about climate change. I'm totally on board with worrying abot climate change, but the presumed "mega-drought" plays an important, if not quite starring role in Collapsed, and is central to Survival, is rooted in archaeobotanical studies, and drawing universal conclusions from localised archaeobotanical sites is a fraught activity, as witness repeated revisions of claims about forest cover changes based on revised understandings of the environmental history of specific sites. I get that Cline would like to use the enormous amount of money he has made for his publisher and turn into public intellectual clout in the service of something more important, but there are fine young scholars out there failing to get tenure-track jobs, and I'd like Cline to back off them if he can.  

Back to Ygor, who, as an Internet warrior still has his bones to make, and can get down into it. Ancient Anatolians do not have Steppe ancestry, and that's that. The Indo-European language family was not spread into Anatolia by a wave of demic advance. "Migration." So then he fixed it by finding a mutual ancestral group in the southern Caucasus in the right timeframe for Proto-Indo-European (4000BC, according to him. 

No disrespect to Ygor, but this is crazy. It's like, "I read some historical linguistics stuff on the Internet, and now I'm going to do a genetic study of the recovered DNA of more than 200 Neolithic individuals and unleash enough statistical analysis software on them to take a Lunar lander to the Sea of Tranquility and back." 

See? This is why linguistics is secretly the hardest historical science. We all take it for granted that we're not going to understand what the historical linguists are talking about, so we just nod along. It's like the Grand Unified Theory. Or it would be if we were using the contradictions between General Relativity and quantum mechanics to justify some light genocide. 

https://en.wikipedia.org/wiki/Ma%C5%9Fat_H%C3%B6y%C3%BCk

So let's talk about (proto-)Indo-European. Not in terms of "glottochronology might not be wrong," but in terms of, you know, actual evidence. That evidence takes the form of an ever-increasing mass monumental texts in Bronze Age/Early Iron Age monumental texts in "Heiroglyphic Luwian," and collections of documents in Hittite texts that are usually called archives, but might actually be working libraries. The best known of these is the Bogazkoy (ancient Hattusa) archive, which was assembled under Supiluliuma II and might have been abandoned with the climactic destruction of that city during the Late Bronze Age Collapse. Whatever we make of that theory, they were lost in a reasonably intact state, which is why we've been reading them for a century or so. In contrast, perhaps-similar collections associated with Tudhaliya I (1360--1344BC) and probably II at Ortakoy (Sapinuwa), Maşat Höyü (Tappika), and Samuha are in a fragmentary state and are being published much too slowly for an impatient scholarship. They do, however, give us a picture of the kind of texts (administrative documents and letters aside) felt necessary for the practice of literate kingship in 1300BC. And they are, perhaps not surprisingly given our Classical tradition, overwhelmingly linguistic in nature. Royalty seems to have spent a great deal of its time performing public cult, and much of that cult was borrowed from adjacent, non-Hittite-speaking cultures, and getting it right required knowing about Hattic, Hurrian, Luwian, Akkadian, and, as is now emerging, other languages as well, including variant forms of Hurro-Urartian and Anatolian. These were not people to commit unconsidered speech acts. 

The monumental evidence is in some ways even more enigmatic. In some ways, they are the earliest ancient texts known. Writing about the presumptive universal empire of the Egyptian Middle Kingdom, one of Herodotus' strongest arguments that it existed at all was that monuments with hieroglyphic illustrations could be found in Anatolia. We  have found some of the moments that Herodotus saw, and we know that they are actually in Hieroglyphic Luwian, which we know well from considerable monumental building at Carchemish by a neo-Hittite state that we can trace (from all those inscriptions) from its founding by a cadet of the Hittite royal line before 1000BC down to its annexation by the Assyrians under Sargon II or perhaps definitively by Sennacherib (705--681). Herodotus, a Greek of Miletus with extensive Carian informants, if not himself bilingual in this Anatolian language, was surely well-travelled enough to have seen Carchemish, so it is mildly amazing that he had no intimation that he was seeing something indigenously Anatolian. 

On the other hand, there is always something intriguing about the early Greek historical sense. They, unlike us, lived in a world saturated with Late Bronze Age survivals. Their understanding of the past is rooted in a Late Bronze Age era that we reconstruct as 'Mycenaean," they as "Homeric." But how much did they actually know? They were clearly prone to deeply misunderstanding the archaeological record; but how much of Homeric tradition is authentic? What texts did they have? Mycenaeans had a written Greek, Linear B, which was lost in the Collapse, or which we cannot exclude was perpetuated in perishable materials until replaced by alphabetic writing. 

Which raises an interesting question: What do we know about Mycenaean Greek? Proto-Indo-European isn't a totem to wave around discussions of more congenial subjects. It is a  hypothesised and much reconstructed entity. Linguists do three main things with this reconstruction process:

i) They reconstruct vocabulary. This is an interesting and worthy endeavour, but apart from arguments about whether the existence of "circle" in Proto-Indo-European proves that wheels and equestrianship were invented somewhere in the deep steppe in the Eneolithic, which is more controversial than is sometimes allowed;

ii) They reconstruct sound changes. This might be the core of the enterprise;

iii) They reconstruct syntax and attempt to understand its changes. This is the most perplexing part of the historical linguistic enterprise since it requires understanding grammar at a highly theoretical level. No-one, thank Heavens, speaks a language with three genders, seven persons, and fourteen cases! And it is an open question as to whether syntax changes to conform to sound changes! Thus perhaps the feminine gender enters the daughter languages of proto-Indo-European for no better reason than that feminine endings "sound right." 

Proto-Indo-European differed from all its daughter languages, and shared with Anatolian, two features: It lacked the masculine/feminine gender present in, for example, Spanish, German and French (and to preserve an explanation of its presence in so many daughter languages, "late Proto-Indo-European"; and it had (probably) three "laryngeal phonemes." I'm not going to pursue the technical account at Wikipedia here, just point out that if it is correct to say that their subsequent disappearance is illustrated by examples like the Germanic "wh" and Latin/Roman "qu-" "question words," it seems that, as a logical conclusion, proto-Germanic and proto-Latin are the direct daughter languages of proto-Indo-European, making the splitting up of proto-Indo-European an early Iron Age phenomena, which, as long-time readers will know, is the idee fixe around here. This converges with explanations of the emergence of grammatical gender separately in the different daughter languages, and makes the splitting up of proto-Indo-European an early Iron Age phenomena, which, as long-time readers will know, is the idee fixe around here

This raises the question of whether Mycenaean Greek has laryngeals. And it does! On the basis of the disappearance of laryngeals, proto-Greek is "first to split," or, alternatively, closest to the original in a paradigm that imagines language change within the Indo-European continuum geographically rather than chronologically. ("Synchronic" versus "diachronic.") So does Mycenaean Greek have grammatical gender?  It is, apparently, weird and not for the faint of heart. So I will do what I do when engineers try to explain free-turbine turboprops with heat exchangers, and bail on the conversation. 

I'll just say here that proto-Indo-European is Luwian, not Hittite, since if it were Hittite we'd see lots of Hurrian and Hattic words. The explanation for its charisma and dominance is the nexus with the military use of the horse, and I would go into that except, oh, look, it's 4pm and I'm working tomorrow. 

 

You knew I would have an explanation, and that it would be a doozy.