Joined: Jan. 2006
|I basically agree--or don't have a reliable basis for disagreeing--with your assessment of the Greenberg-Ruhlen "Amerind" hypothesis (which, with the exception of Ruhlen's "stick to the wall" Ket-Na Dene speculative stab, is really the only "original" claim they've made in the field of "origins of American languages/peoples--since the grouping of the Na-Dene and Inuit-Aleut languages was, as you say, much older, |
No, I don't think anyone has ever grouped Na-Dene and Inuit-Aleut together. They were just the last two entries into the New World, but that doesn't mean they're related.
|and even the Na-Dene/Haida connection had been hypothesized much earlier--by Sapir? or one of those cats).|
I think by Sapir. I'd have to look it up. Enrico's articles would probably say.
|Not being a linguist, I do find the work of the proto-language family "lumpers" interesting, and it seems not entirely un-, anti-, or psuedo-scientific to me. Though obviously the farther back you try to reconstruct, the sketchier and more controversial things will get...|
In this regard, though, did Greenberg use an entirely different procedure to generate the similarities and connections which assertedly support his Amerind hypothesis than he did in mega-grouping the sub-Saharan African languages, which--if I understand correctly-- eventually was accepted?
What I gather is that Greenberg succeeded in Africa where he failed in the New World for a handful of different reasons; apparently, before Greenberg tackled Africa, the state of historical linguistic work on African languages was still in its infancy, and no one had seriously done the work on major groupings yet. So what he did was lay out the big families. However, by the time of Language in the Americas, the work on linguistic grouping in the New World was already very far advanced indeed -- it was nothing like the semi-blank slate Africa had been. The kind of work Greenberg had done for Africa had already been done for the New World.
Another problem is that the linguistic situations in Africa and the New World aren't all that similar. The New World is simply much more diverse, with a LOT more old language families that can't be linguistically related. Africa really does seem to be reduceable to a small handful of big language stocks, the existence of which is accepted by workers on those language families. However, in the New World, not only do no specialists in American Indian languages accept 'Amerind', none of the intermediate supergroupings Greenberg posited (like 'Almosan-Kere-Siouan', etc.) have been accepted, either. So even when Greenberg tried to posit larger stocks in the New World, his ideas haven't held up.
And finally, again, I am not an Africanist, but I am told that the degree to which Greenberg's ideas about African languages have been accepted is somewhat exaggerated -- that is, some of the groupings he came up with in the '50's haven't stood the test of time, tho I gather most have.
|And the work of the Russian linguistic reconstructivists may bear watching. The interplay of the "genetic" and linguistic hypotheses may eventually lend some added support--or ultimately falsify--this entire enterprise.|
Maybe, maybe not. I personally can't accept purely genetic evidence for linguistic hypotheses. I'd call it circumstantial evidence, at best.
|My understanding here, again, is that the Amerind-Asian "split" may well be much older than the Na Dene-Yeniesian "split." |
Right. It looks pretty likely that the proto-Na-Dene speakers were the second-to-last entrants into the New World, a couple thousand years after Clovis. And if semi-convincing cognates can still be found between a language family centered on Alaska and one centered in central (not even eastern) Siberia, that's a hint that the split can't have been THAT long ago...
|Which would make the reconstruction and connecting up of proto-Amerind (supposing, of course, that there even is such a thing) with some proto-whatever Asian language family even more difficult to recover and reconstruct, I presume...|
My feeling is that even if Proto-Amerind did exist once -- and there's no terribly compelling reason to think it did -- it's not reconstructible at all. The time depth is just too huge by now, and the language families just too diverse.
But your main point is true, tho, that trying to connect ANY New World family to any family in Asia is made extremely difficult by the time depth involved. If the pre-Clovis sites are real, and there were already people in the New World say, 22,000 years ago, then some of the languages might have come over so long ago as to make connecting them to Asia flat out impossible.
|In that regard, your point about the relative distance of Haida and Yeniesian (Ket) from Na-Dene is interesting in and of itself. In "cladistic" terms, one might envision a tree ("unrooted" at the moment, though I guess G'n'R and others might group the whole shebang with ?Caucasian? or some such proto-family--?Norstratic?)--I don't have the tools to draw this, much less link it to this blog, sorry, so the reader will have to bear with me--in which the Yeniesian group of languages (with Ket as the exemplar) and Haida are deeply split, but connected on one branch, and at the base of that branch is an even deeper split, the other arm of which runs off to the Na-Dene languages, which then splits again (to generate the Tlingit-Ayak-Athapaskan group of languages--with the Athapaskan subgroup being familiar to most readers through Navajo and Apache, though the family seems to "center" geographically in NW Canada, with outliers in Oregon and other places along the way to the U.S. Southwest).|
Yeah, basically core-Na Dene consists of two main branches: Athabaskan/Eyak on the one hand, and Tlingit on the other, as two co-equal branches (Eyak is closer to Athabaskan than it is to Tlingit.) And it's accepted that Athabaskan had to have originated in Alaska, since that's where the most diversity in the family is. It's clear that the Apachean and Pacific Coast branches of Athabaskan were much more recent. The entry of Apachean (including Navaho) into the southwest may have actually been post-contact.
Yeneseian and Haida would be somewhere much further up the tree. The jury is still out on both of them. If the similarities between Yeneseian and Na-Dene are extensive enough, then it means they're related, since the enormous geographic separation between them means the similarities between them can't be due to contact.
And contact is the whole problem with the similarities between Haida and Na-Dene. Before Enrico started his work, pretty much everyone had swung over to assuming that Haida was an isolate that had undergone some huge (and ancient) influence from Tlingit. However, Enrico was the first person to do new fieldwork on Haida for decades, and he's convinced that Haida is actually genetically related to Na-Dene. Considering his expertise on Haida -- no other linguist is really an expert on the language -- his views can't be dismissed, but the jury is still out on the issue. There seems to be no consensus on the subject yet, at least not that I've heard.
|Or do you visualize a different tree, Arden? Between Yeniesian-Na Dene, and then a very early split between Haida and the rest of Na-Dene? |
There's probably no other choice, given the geography -- provided Haida really is related.
|I would love to get a copy of Vajda's Yeniesian-Na Dene talk, Arden. Maybe we could work that out somehow (is it the one he gave at the 2000 SSILA meeting?)...|
No, it's from a talk he gave in Europe this summer. I wasn't there for it, but a friend of mine was, and he xeroxed me a copy of the handout. Send me a message, and I'll make some arrangement to send you a copy of it.
"Rich is just mad because he thought all titties had fur on them until last week when a shorn transvestite ruined his childhood dreams by jumping out of a spider man cake and man boobing him in the face lips." - Erasmus