Genes and the double standard

Between 2008 and 2010, Edward Vajda presented a series of papers and lectures arguing in favor of a genetic relationship between the Yeniseian and Na-Dené language families. (For non-linguists, the phrase "genetic relationship" strictly means a relatedness between languages and says nothing about the DNA of their speakers). The Yeniseian languages are spoken in central Siberia, fairly distant from the eastern coastline. The Na-Dené languages are spoken throughout North America and were probably the last major family to enter North America prior to the Colombian Age starting in 1492.

Unlike most recent advocates of long-distance relationships between language families, Vajda is no fringe linguist or loon. His arguments are conservative, openly acknowledges their problems, and link to deep within the grammatical structures of the two families. His critics, however, have been less convincing; most of them revolve around the "It just couldn't happen argument:"

  1. Nichols (2010) claims it's geographically implausible. I find this to be the worst counter-argument of all, given the extreme time depths. We are talking about thousands upon thousands of years of time to migrate away from each other.
  2. Ives (2010) and others criticize the dearth of archaeological support. Fair enough, but that's a problem to be hashed out in time. We could also point to the under-funding of archaeological digs in Siberia as well.
  3. Several linguists criticize the actual linguistics; the real "meat and potatoes" of Vajda's arguments. The pronouns don't match (although this can be easily explained cross-linguistically) and the phonological correspondences are problematic (this is actually a real problem that must be acknowledged). For a fuller discussion of the linguistic difficulties, read here (PDF warning).

But what really makes me sigh are people who point to Scott & O’Rouke's (2010) biostudy that showed little-to-no genetic relationship (and now I do mean this in a medical sense) between both Yeniseian and Na-Dené peoples. Every so often, an article is published that shows high genetic relatedness (the DNA kind of relatedness) between two groups of people that speak two different languages. What is implied by the studies is that perhaps the two different languages are linguistically related because they are genetically related. For example, the unrelated Khoisean languages in southern Africa.  

The problem with this implication is that language replacement makes this mean diddly-squat. Hungarian is a Uralic language but the Hungarian people are genetically more like Balto-Slavs and Balkan folk. My friend Xiao speaks only English but I can tell you right now, Xiao is not Anglo-Saxon. So, why am I talking about this? Because people who are quick to point out the problems with using DNA as a measure of language relationships are the same people who point to Scott & O'Rouke. Can we just agree to stop using genes and focus on the linguistics, people?


Ives, J. W. "Dene-Yeniseian, Migration and Prehistory". Anthropological Papers of the University of Alaska. Special Issue: The Dene-Yeniseian Hypothesis. Ed. by J. Kari & B. Potter. Vol. 5, Is. 102. 2010.

Nichols, J. "Proving Dene-Yeniseian Genealogical Relatedness". Anthropological Papers of the University of Alaska. Special Issue: The Dene-Yeniseian Hypothesis. Ed. by J. Kari & B. Potter. Vol. 5, Is. 102. 2010.

Scott, G. R. & D. H. O'Rourke. "Genes across Beringia: a Physical Anthropological perspective on the Dene-Yeniseian hypothesis". Anthropological Papers of the University of Alaska. Special Issue: The Dene-Yeniseian Hypothesis. Ed. by J. Kari & B. Potter. Vol. 5, Is. 102. 2010.