National Geographic News has just published an article about the recent symposium in Alaska regarding a possible connection between Yeniseic languages in Siberia and Na-Dene languages in the Americas. John Roach’s article, Siberian, Native American Languages Linked — A First, highlights the recent work of Edward Vajda, who defended his connection during the February symposium. Vajda goes deeper than cognate lists in his parallels, providing several corresponding grammatical systems, particularly verb prefix structure. Ket, his primary Siberian source, is the only living Yeniseic language (which remains highly endangered) and bears some striking grammatical similarities to Navajo. Yeniseic languages have a unique verb prefix system: unique enough that Vajda could not find a corresponding system throughout Northern Asia. Na-Dene was the closest family geographically with a similar system. Johanna Nichols, a groundbreaking Historical Linguist and Linguistic Anthropologist, attended the symposium and made comment. Roach quotes:
With the exception of the Eskimo-Aleut family that straddles the Bering Strait and Aleutian Islands, this is “the first successful demonstration of any connection between a New World language and an Old World language,” Nichols said.
Vajda has not yet published his findings, so the extent of his linguistic claims is not yet clear. However, based on Roach’s summary of his discussion, there are two major points of controversy. First, Roach states that Vajda found “several dozen” cognates. Whether or not the comparative method for linguistic reconstruction was used remains to be seen. Regardless, a cognate list under 50 seems a bit thin to solidify a connection at all, let alone begin reconstruction. Furthermore, the public at this point has no access to the words to assess their status as true cognates. Without a doubt, a consistent and corresponding element of grammatical structure is a strong argument for a common ancestor, but we must consider the systems of linguistic change, particularly sound change (which requires cognates), as a central factor.
A second point of controversy is the matter of depth: how long ago does the proposed connection date back? Vajda makes no direct claims, but states that this would be the oldest known language link if it corresponds to the late Pleistocene migrations evident in the archaeological record. Unfortunately, the field of linguistics currently has no reliable absolute dating techniques, and relative dating such as glottochronology, has been widely discredited. In this case, it seems the lack of cognates would help secure this relationship as an old one. If that were indeed the case, a volume of cognates would become evident in the reconstructions of Proto-Yeniseic and Proto-Na-Dene. Whether or not Vajda has taken this into consideration remains to be seen. At any rate, Nichols is not convinced of a 10,000 year-old connection:
“I don’t think there is any reason to assume the connection is [10,000 years] old … this must surely be one late episode in a much longer and more complicated history of settlement,” she said.
At this point it is very difficult to make any generalizations. Vajda has not yet published his findings, but merely opened the door to discussion on the topic. Until he does, the foundation of our support or criticism is unknown.