A credible link between Yeniseian and Na-Dené languages

[Preno]

This is actually last month's news, but in case you missed it like I did:

A genetic link has been demonstrated and acknowledged by the relevant specialists between the Na-Dené (http://en.wikipedia.org/wiki/Na-Den%C3%A9_languages) languages spoken here:

http://upload.wikimedia.org/wikipedia/commons/thumb/9/9c/Na-Dene_langs.png/200px-Na-Dene_langs.png

and the small Paleosiberian Ket language (http://lingsib.unesco.ru/en/languages/ket.shtml.htm) spoken here:

http://www.eki.ee/books/redbook/maps/141.gif

Ket is the last surviving representative of the small Yeniseian (http://en.wikipedia.org/wiki/Yeniseian_languages) family, several other related languages having died out in the last few centuries (Yugh probably at the end of the 20th). Na-Dené includes the Athabaskan languages, Eyak (last speaker of which died this January) and Tlingit. The relevant papers are available here (http://www.uaf.edu/anlc/dy2008.html). The key one is A Siberian Link to Na-Dené Languages (http://www.uaf.edu/anlc/docs/vajda-2008.pdf), there are also some on anthropology and comparative mythology (which is interested but blurred and inconclusive). Despite the linguistic link, however, the two groups are genetically unrelated. If true, this is so far the only established link between the New and Old World languages, with the obvious exception of Eskimo-Aleut (according to one of the linguists, this is "the longest pre-transport language spread" on Earth). That would certainly be big news.

And it's apparently not just another Whatever-Burushaski-North Caucasian-Basque mass comparison type of theory, the author actually systematically compares the strikingly similar verb morphologies of the two families, shows how Na-Dené evidence can help to clarify tonogenesis in Yeniseian and establishes sound correspondences.

[dancer_rnb]

Do you mean a genetic link between the speakers of the language,
or a link between the languages themselves?

[Preno]

Do you mean a genetic link between the speakers of the language,
or a link between the languages themselves?I mean that they share a common ancestor. "Genetically related" in linguistics means just that. Although I can see how this:
A genetic link has been demonstrated and acknowledged by the relevant specialists between the [languages]

[...]

Despite the linguistic link, however, the two groups are genetically unrelated.might have been a bit confusing.

[llanitedave]

That's a fascinating discovery.

[Gooch's dad]

Thanks for the excellent post, Preno! I enjoyed reading about some of those languages, and I listened to a sound clip of the Ket language on one of the websites you listed. I wonder what the Ket native speakers will think of this discovery.

[Ray Moscow]

Well, cool.

I have a very small fraction of Athabascan ancestry.

[Per Ahlberg]

Nice one! Am I correct in thinking that the Na-Dené speakers are regarded as relative latecomers in the Americas - later than the Amerind speakers anyhow? I seem to remember hearing this somewhere.

[Ray Moscow]

My understanding is that the Athabaskans arrived much later than the earlier immigrants.

The Navajo, for example, arrived relatively recently in an area that had already been settled by the pueblo peoples some thousands of years earlier.

[Preno]

Tbh, I don't know much about the anthropological side of the issue, and I'm not sure what the accepted opinion is (if there is one). As far as I can tell, there are several points of contention and the evidence seems to be conflicting.

(What there is an accepted opinion on, however, is that the affinity of the so-called Amerind languages has not been demonstrated - mass comparison is not considered to be a valid method of determining genetic relationships between languages.)
If anything, a Dene-Yeniseic language link only deepens a number of already deep mysteries surrounding the peopling of the New World. Archeological research has so far not yielded compelling evidence for the arrival of new colonists from North Asia into Alaska in the millennia between the foundering of the Bering land bridge at the end of the Pleistocene 10,000 years ago and the arrival across Bering Strait by the ancestral Eskimo-Aleut about 4,500 years ago. The intricate linguistic homologies described in my article would seem to suggest an arrival by ancestral speakers of Na-Dene during this intervening period – perhaps millennia after the first Paleo-Indian group(s) entered Alaska yet before the Eskimo-Aleut – a hypothesis earlier proposed for the Na-Dene by Joseph Greenberg (1987). The evidence presented by Ben Potter (this volume) on the temporal succession of prehistoric tool assemblages in Siberia and Alaska, however, reveals no clear evidence for new migrations into Alaska between 10,000 and 4,500 years ago. So far, there is no archeological evidence that ancestors of the Na-Dene peoples (Athabaskans, Eyak, Tlingit) entered the New World during the interval between 10,000 and 4,500BP, despite the fact that the historically attested locations of these peoples also suggest their arrival after other Paleo-Indian groups. If one accepts Dene-Yeniseic, the incongruence between the archeological and linguistic facts only widens, since the lexical and grammatical homologies between Na-Dene and Yeniseic suggest a separation younger than 10,000 years, given what is generally assumed about rates of language change over time. At present, no compelling reason exists to assume languages changed more slowly in prehistory than in the recent past, though I will return to this possibility below.Cognates tree names are limited to species shared across Late Pleistocene North Asia and Alaska: birch, alder, willow, as well as (probably) generic words for conifer needles and resin. Specific terms for conifer genera such as larch, fir, spruce, pine, etc. – species that remained in South Siberia through the Ice Age but which re-colonized interior Alaska only after 6,000 years ago (William Workman, personal communication) do not match up between Ket and Athabaskan.
Judging from the parallel findings of archeologists, it is conceivable that Dene-Yeniseic could prove to be a language family older than 10,000 years. If so, then the linguistic evidence from Dene-Yeniseic would suggest that the apparent incongruity between the Dene-Yeniseic linguistic facts and currently assumed dates for prehistoric trans-Bering migrations could be due to an over-estimation of how quickly languages change on average over time. Unless hard evidence of migrations into Alaska from North Asia between 10,000 and 4,500 BP can be found, my personal preference would be to at least entertain the ramifications of the latter possibility.

(source) (http://www.uaf.edu/anlc/docs/dy_vajda_perspective.pdf)

[lpetrich]

And it's apparently not just another Whatever-Burushaski-North Caucasian-Basque mass comparison type of theory, the author actually systematically compares the strikingly similar verb morphologies of the two families, shows how Na-Dené evidence can help to clarify tonogenesis in Yeniseian and establishes sound correspondences.
I think that verb morphology is a good clue. For an example close to home, although English has an enormous amount of borrowed vocabulary, its verb morphology is still recognizably Germanic.

The next stop, I think, is the North Caucasian languages. They include Chechen, the native language of Chechnya. However, they are rather difficult to compare, and Sergei Starostin's reconstruction of their ancestor remains controversial.

Sino-Tibetan is also rather difficult to compare, because many members are rather short on word morphology, and because only Chinese and Tibetan have written histories past 1000 years.

But if Dené-Yeniseian can successfully be linked to North Caucasian or even Sino-Tibetan, then we can move on to Basque, Burushaski, and perhaps also Hurro-Urartian, Hattic, and Etruscan.


Wikipedia's contributors are very fast: Dené-Yeniseian languages, and they have stuff on broader comparisons, like Dené-Caucasian languages.

Nice one! Am I correct in thinking that the Na-Dené speakers are regarded as relative latecomers in the Americas - later than the Amerind speakers anyhow? I seem to remember hearing this somewhere.
That's pretty much it. The Siberian Origins of Native Americans (http://pandora.cii.wwu.edu/vajda/ea210/SiberianOriginsNA.htm)

The First (Amerind) Migration (earlier than 10,000 BC)

The Second (Na-Dene) Migration (c. 8,000 BC)

The Third (Eskimo-Aleut) Migration (c. 4,000 BC)

[RAFH]

And it's apparently not just another Whatever-Burushaski-North Caucasian-Basque mass comparison type of theory, the author actually systematically compares the strikingly similar verb morphologies of the two families, shows how Na-Dené evidence can help to clarify tonogenesis in Yeniseian and establishes sound correspondences.
I think that verb morphology is a good clue. For an example close to home, although English has an enormous amount of borrowed vocabulary, its verb morphology is still recognizably Germanic.

The next stop, I think, is the North Caucasian languages. They include Chechen, the native language of Chechnya. However, they are rather difficult to compare, and Sergei Starostin's reconstruction of their ancestor remains controversial.

Sino-Tibetan is also rather difficult to compare, because many members are rather short on word morphology, and because only Chinese and Tibetan have written histories past 1000 years.

But if Dené-Yeniseian can successfully be linked to North Caucasian or even Sino-Tibetan, then we can move on to Basque, Burushaski, and perhaps also Hurro-Urartian, Hattic, and Etruscan.


Wikipedia's contributors are very fast: Dené-Yeniseian languages, and they have stuff on broader comparisons, like Dené-Caucasian languages.

Nice one! Am I correct in thinking that the Na-Dené speakers are regarded as relative latecomers in the Americas - later than the Amerind speakers anyhow? I seem to remember hearing this somewhere.
That's pretty much it. The Siberian Origins of Native Americans (http://pandora.cii.wwu.edu/vajda/ea210/SiberianOriginsNA.htm)

The First (Amerind) Migration (earlier than 10,000 BC)

The Second (Na-Dene) Migration (c. 8,000 BC)but apparently no archaeological finds to substantiate the 8000BCE migration. Is it possible they used another route? Is anything known of where the Ket came from, their previous migrations? The language root fades in at 500BCE, splitting out at 1 CE. Apparently no tracing back beyond that to much of anything.

The Third (Eskimo-Aleut) Migration (c. 4,000 BC)

Yep, interesting. It will be even more interesting if some archaeological finds support this someday. Is there anything on the Dene migrations in North America before they arrived at their current locations, any traces in western Alaska that could suggest a time or place of passage?

[Preno]

I think that verb morphology is a good clue. For an example close to home, although English has an enormous amount of borrowed vocabulary, its verb morphology is still recognizably Germanic.I agree, but I think this situation was rather unusual, in that the verb morphology of was very complex, apparently more so than with other Siberian languages, and hence provided more evidence to compare, and the historical system of conjugation stands out typologically against the rest of the surrounding languages. The rest of the Siberian languages largely conjugate by suffixing, unlike Yeniseian, which was historically prefixing. (Interestingly enough, due to areal influence, this was completely reversed when Ket developed a new lexical slot at the beginning of the verb, the historical prefixes becoming suffixes everywhere except for the historically oldest verbs.)

Another massive advantage was that areal influence could clearly be dismissed out of hand, which is hardly so with most other Siberian languages, where both options are entirely possible (as with Uralic-Yukaghir, for example). Ket itself apparently borrowed from Selkup (a Samoyedic language).