The proposal in brief
The Dene-Yeniseian hypothesis is a proposal by historical linguists that the Na-Dene languages of North America and the Yeniseian languages of Siberia descend from a common ancestor. It was put forward in the late 2000s, most prominently by the linguist Edward Vajda, and it attracted attention because few proposed language links between the Old World and the Americas have met with serious professional interest. It remains a hypothesis, not an accepted fact, and the discussion is still active.
For readers of this site the question is a narrow one. Apache languages are part of the Athabaskan family, and Athabaskan is part of Na-Dene. If the hypothesis holds, Apache languages would be a distant branch of a larger family with roots in northern Eurasia.
The language families involved
Na-Dene is a family in northwestern North America. Its best-known branch is Athabaskan, which spreads from Alaska and western Canada through the Pacific coast to the Southwest, where it includes Navajo and the Apache languages. Eyak, in southern Alaska, and Tlingit, on the Pacific coast of Alaska and Canada, are generally treated as related to Athabaskan within Na-Dene, although the Tlingit link has been argued about for a long time.
Yeniseian is a small family named for the Yenisei River in central Siberia. Its only surviving language is Ket, spoken by very few people today. Several other Yeniseian languages are recorded in older sources and are no longer spoken. Ket has a particularly complex verb, which is one reason it interested specialists in Athabaskan.
The evidence offered
Supporters point to several kinds of similarity.
- Verb structure. Both families build long verbs with a template of slots in a fixed order, and some of the details of how those slots work are said to match in unusual ways.
- Sound correspondences. Proposals have been made that certain consonants in Ket regularly match certain consonants in Athabaskan words of similar meaning.
- Shared vocabulary. Lists of words with similar meaning and form have been compiled, mostly for body parts and basic terms.
- Tone. Some argue that tone systems in both families can be compared, although tone can arise independently.
Matches in grammar are valuable to historians of language because they are harder to borrow than words. That is the main reason the proposal has been taken seriously.
The doubts
The hypothesis has not won universal agreement, and doubts come from several directions.
- Languages that are far apart in time and space can share similarities by chance, and the more words one has to choose from, the easier it is to find matches.
- The time depth would be very large, probably many thousands of years, and the usual methods of comparing languages become less reliable at such distances.
- Some historians of language want stricter and more complete regular sound correspondences than have yet been shown.
- Others have questioned whether specific comparisons, especially in vocabulary, hold up.
Some linguists find the case persuasive or at least promising, while others consider it unproven. Both positions are reasonable, and scholars are still examining the material.
What a language link does and does not show
A relationship between languages is not the same as a relationship between populations. People can adopt a language without descending from its earlier speakers, and groups with a shared ancestry can end up speaking unrelated languages. A proposed family tree for languages therefore cannot be read as a family tree for people.
It also says nothing in itself about when or how any ancestors of Na-Dene speakers reached North America, or about what happened afterward. Archaeological evidence, genetic research and linguistic reconstruction each address different parts of the question, and they do not always agree. Genetic studies of Native American ancestry are a developing field, and consumer DNA tests, including Apache DNA ancestry tests marketed to the public, cannot tell anyone which nation they belong to.
Nor does the hypothesis suggest that Apache peoples arrived in the Southwest from Siberia in any recent period. The Apachean migration southward from the north is a separate and later question, and the timing of the arrival of Athabaskan speakers in the Southwest is itself debated by archaeologists. Apache oral tradition accounts give their own understanding of origins, and those accounts stand independently of linguistic theory.
Why the idea gets attention
The proposal draws interest because it would be, if confirmed, the first widely accepted demonstration of a genetic link between a Native American language family and a family in the Old World. Earlier suggestions of such links have usually been rejected. That makes careful testing important, and it is why many linguists have asked for clear presentations of the sound correspondences.
It also draws interest outside linguistics, which is a reason for caution. Popular articles can present a proposal as a discovery, and online summaries sometimes turn it into a claim about the ancestry of every Athabaskan-speaking people. The accurate summary is more modest: a serious hypothesis, supported by some linguists, questioned by others, and not yet settled.
Where it fits among other questions
Within the wider debate over Apache origins, Dene-Yeniseian is a distant background topic. The nearer questions are how the Athabaskan language family spread across North America, what archaeological evidence exists for Athabaskan-speaking arrivals in the Southwest, and what Apache nations themselves say about their history. Those are covered where they apply elsewhere on this site. Readers who want the latest state of the debate should consult recent linguistic publications, because the discussion is still developing.