Tongue ties: a language bridge across the Bering Strait

A Western Washington University professor has compared native languages in North America to those in Asia and found ties that suggest they come from the same ancestors.
Crosscut archive image.

Members of the Ket people of Central Siberia, photographed in 1906.

A Western Washington University professor has compared native languages in North America to those in Asia and found ties that suggest they come from the same ancestors.

The bones, arrowheads, and DNA (most recently found in fossilized poop) all agree: Our North American continent was first peopled by immigrants from north Asia over 10,000 years ago. And now linguist Edward Vajda has found remnants of this ancient heritage in words spoken today.

Vajda, a professor at Western Washington University in Bellingham, recently demonstrated a convincing kinship between a Siberian language family called Yeniseic and a Native American family called Na-Dene, which includes languages spoken in the Pacific Northwest and the Southwest.

Compared to the hard evidence of archaeology, language is more like quicksand, which is why the new link is surprising. "It's been assumed that the rate of language change is so rapid that all evidence of linguistic relationships would have disappeared by (this) time," said Vajda.

Despite languages' tendency to morph, Vajda found enough similarities between the Yeniseic and Na-Dene families to convince fellow linguists that both are derived from a common ancient tongue at a recent symposium held at the University of Alaska in Fairbanks. To do this, he drew upon his expertise in Ket, an endangered language spoken in central Siberia. "In this one little obscure language with fewer than 200 (speakers), we have this evidence of a link across the oceans," he said.

To learn this endangered language that hadn't been completely documented, Vajda relied on Russian scholars, rare books, and his own field work in Tomsk, a city in Siberia, where he worked with native Ket speakers. Ket people were hunter-gatherers until they were forced to settle during Soviet collectivization campaigns, said Vajda.

While learning Ket, Vajda originally wondered about a connection with Native American languages. "I just saw features in the Ket verb that just really sort of reminded me of Navajo," he said. After the 15 years it took him to master Ket, he could rigorously establish similarities between Yeneseic languages, which included Ket and extinct Siberian languages like Pumpokol and Yugh, and Na-Dene languages like Navajo and Athabaskan, which is spoken in Alaska and Washington.

A link between the Yeniseic and Na-Dene families had been supposed for some time, since they share look-alike words that have similar meanings. But such "look-alikes" alone do not prove languages are related because they can occur by chance. Vajda's work moves beyond these superficial similarities by finding true "cognates," words that have a common origin, and by showing similar word structures and consistent sound correspondences.

For instance, take the Ket word for "hair": It's made by adding the word for "head" to the word for "fur" to make "head-fur." It turns out that this very same "head-fur" construction is used to make the word "hair" throughout the Na-Dene languages, and not only is it a cognate, but the individual words for "head" and "fur" are also cognates. "That type of thing is really strong evidence that languages are related," Vajda said.

Notably, the cognates he found occur in vocabulary relevant to a hunter-gatherer lifestyle, as expected for words derived from the same ancient language. "You really have to look at the basic vocabulary because these are the words that are probably in the language five or 10 thousand years ago," he explained.

These cognates might also be unwitting informants about North American prehistory, Vajda suggested. "This link (might) help us understand what was original in the lifestyle of these people — what kind of terrain, and what kind of trees and animals (they had) in their original homeland," he said.

Vajda also found similarities in verbs, which are particularly complicated in these languages. Verbs in both families are modified by a complex series of prefixes, and Vajda showed that the order, form, and meanings of these prefixes were the same in both language families.

For example, Ket uses a "shape" prefix, which describes the shape of the object being acted upon by the verb. It's a very specific construction: For hammering a flat plank, the verb "to hammer" is preceded by a word meaning "flat"; for hammering something round, like a stone, there is a prefix meaning "round"; and for hammering something long, there is a "long" prefix.

This complicated shape prefix system also occurs in Na-Dene languages. "We have the same three prefixes, with the same three meanings, and (in) the same slot in the verb," he said.

Finally, Vajda identified enough cognate words between Yeniseic and Na-Dene that he could make a glossary of the sound correspondences between them, which shows how sounds in one language have morphed into a different sound in another. For example, a /d/ sound in German corresponds to a /th/ sound in the English cognate. Think "Dick" and thick, or "Leder" and leather.

Vajda found that an /s/ sound in Ket corresponds to a /ts/ sound in Athabaskan. So the word for "mosquito" is pronounced "soo-ee" in Ket and "tsoo-ee" in Athabaskan. "These basic sound correspondences tell you that this isn't coincidental, that these similarities come from a common origin," said Vajda.

Part of what made Vajda's comparisons so convincing is that he relied on the research of other linguists who had independently worked out the structures of the Na-Dene languages, without any interest in finding similarities between them and Asian languages. "I could systematically connect Ket with the system they had established," he said.

Many of these linguists agreed with Vajda's analysis. "This is more material than for many language families that are recognized (as) deeply linked in time," said Jim Kari, an expert in Athabaskan languages at the University of Alaska in Fairbanks. Getting these linguists together for the symposium also generated more evidence. "Even at breakfast one morning, we apparently found a cognate for 'merganser,'" said Kari.

The work also underscores the importance of documenting obscure languages before they die out. Vajda closed his symposium paper with this thought: "Who could have guessed that the ancient words Native American and Native Siberian boarding-school children were punished for speaking aloud just a few short decades ago would prove to wield a power vast enough to reunite entire continents?"


Please support independent local news for all.

We rely on donations from readers like you to sustain Crosscut's in-depth reporting on issues critical to the PNW.


About the Authors & Contributors