No vindication of Joseph Greenberg?

By Razib Khan | July 25, 2012 3:06 am

A reader pointed me to this critique of Nick Wades’ telling in The New York Times Reports that the recent Reich et al. paper on Native Americans is a vindication of Joseph Greenberg’s ideas on the languages of the Americas. 90-Year-Old Consensus:

Nicholas Wade’s reported the Reich et al. research in the New York Times (July 11, 2012). Wade treats it as a vindication of a three-way genetic (historical linguistic) distinction among languages of the Americas proposed in Joseph Greenberg’s (1987) book of the same name, although Reich et al. do not cite it in their paper in Nature. (The only reference to Greenberg by Reich et al. is to a paper coauthored with Turner and Zegura and published in 1986 as one of the proponents of the three-way split.) The “vindication” here is entirely Wade’s. The bottom line is that this three-way distinction was known linguistically since the 1920s (for example, Sapir 1921). Basically, it’s a division among the Eskimo-Aleut languages, which straddle the Bering Straits even today, the Athabaskan languages (which were discovered to be related to a small Siberian language family only within the last few years, not by Greenberg as Wade suggested), and everything else. That’s essentially the three-way distinction that is constantly credited to Greenberg. We know of many major linguistic families among the “everything else”, worked out painstakingly through well-established methods, but don’t know how the “everything else” language families are connected to each other on a large-scale level.

Let me add that I am skeptical when someone says that a biological genetic grouping corroborates a historical linguistic grouping or vice versa for a simple reason: genetic material and language are transmitted by different mechanisms (I’ll skip my usual joke about this), so in principle a one-to-one correspondence should be surprising.

First, when I first heard about Greenberg’s system in the late 1990s I chatted up a few people who were involved in the linguistics of Native Americans on the issue. They thought he was full of it, but, they were also pretty much opposed to imposing any real coarse structure on Native American languages, including the model which is claimed to be pretty well known above. My sample could be very unrepresentative. I’ll let readers who know more weigh in (note that I’m skeptical of classifying all the “First American” languages into one group after 15,000 years, even assuming they derive from a common ancestor).

On the other hand, the author caricatures the paper which he discusses by suggesting a “one-to-one correspondence.” In fact, the two non-First American groups are genetic, but not linguistic, hybrids. Additionally, the argument that cultural and genetic transmission differ is specious. Certainly it is true, but the two are also correlated due to the effect of culture and language on marriage networks. This seems common sense, but many social scientists in my experience seem to recoil from any disciplinary integration with genetics, and so totally distinguish the character of cultural heritability and genetic heritability.

MORE ABOUT: Linguistics
  • Charles Nydorf

    That the Eskimo-Aleut languages constitute a single family descended from one original language is generally accepted. The same is true for the Athabaskan languages. In both cases, it is possible to get a rough idea of what the hypothetical ancestral language would be like. The languages that Greenberg grouped as Amerind constitute a different case because they are far more diverse. If they do go back to a single ancestral language it would be far older and far more difficult to reconstruct. Reich et als conclusion that speakers of Eskimo-Aleut and Athabaskan arrived later than speakers of languages in the Amerind grouping does not imply that Greenberg’s postulation of a giant Amerind family is valid or invalid.

  • gcochran

    Right. Even though the Amerinds ( as opposed to the Eskimo and Na-Dene) stem from a single, small, ancient colonization, one that almost certainly spoke a single language, maybe some of their existing languages (or the ancestors of those existing languages) were made up from scratch, like Klingon.

  • Karl Zimmerman

    2 –

    How would that work in a pre-literary era? One group of Paleo-Indians having a Pentacostal-like belief in “speaking in tongues” and the younger generation thinking they hear structure in this jabber, and developing it into a second language among themselves?

  • gcochran

    From Wikipedia, the free encyclopedia

    Sarcasm is “a sharp, bitter, or cutting expression or remark; a bitter jibe or taunt”, usually conveyed through irony or understatement.

  • Chris Irwin Davis

    It would have been impossible in a pre-literary era for someone to develop “speaking in tongues” into a second language. Even modern recording technology and computational analysis has revealed glossolalia (as it’s known to linguists) to elude identification of any lexical or grammatical structure.

    Perhaps I am arguing semantics. If someone claimed that the numerical substring “999999” was significant because it recurred frequently in a stream of random numbers, could it actually be said to be “based on” that stream if the claim were not, in fact, true?

  • magscanner

    I like gcochran’s idea. Imagine you have a stable society, and the upper class develops a private language based on a sarcastic view of the lower classes. They make up words, grammar, and all of that stuff, and start using it amongst themselves, but keep it secret from the toiling classes.

    Eventually the lower classes revolt, and throw the erstwhile bosses out of the tribal group, leaving them to survive as best they can while using a jabbering language that was made up and is probably inefficient.

    I’d like to see some alternate-history fantasy writer try this on for size.

  • Jess Tauber

    I’ve got evidence (lexical, morphological) that Yahgan, from the southern tip of South America, has some sort of relatively recent connection to Salishan (from the North American Pacific Northwest). What the nature of this relation is (I lean towards contact) remains open.

    Yahgan was one of the languages Greenberg mucked up, splitting it into two different ‘Yahganan’ languages because he didn’t understand the different nomenclatural conventions used in two different research eras, and the poor renderings from the more recent one.

    He and his research tradition predecessors (Swadesh, Sapir) thought they saw Salishan related to local neighbors Chemakuan and Wakashan, and later connected through Kutenay and Algonkian-Ritwan. There may be some evidence for Salish-Kutenay linkage (some of the Kutenay forms resemble Yahgan more than Salishan), perhaps Ritwan to Yahgan as well. Chemakuan and Wakashan are no-go.

    Greenberg also connects Yahgan to other ‘Andean’ language families, but the evidence is relatively sparse (very distant if true). Yahgan is very different from its neighbors such as Qawesqar and various Chon languages (except for borrowings). I have some evidence that makes a not-so-solid case for a link between Salishan and Mapudungun.

    It will be interesting to see whether long-distance genetic relationship is as important on the Pacific coasts of the Americas as it seems to have been in select cases elsewhere. Can Greenberg’s ‘Amerind’-internal splits be trusted?

  • Mike Keesey


    “one that almost certainly spoke a single language”

    How could we possibly know that?

  • John Emerson

    This has been of only slightly more than casual interest for me, but whatever I’ve read on this topic has always been consistent with a.) a well-defined Eskimo-Aleut language group (relatively recently arrived), b.) a well-defined Athabaskan group (arrived less recently) and c.) an undefined group which arrived before them. I’ve never seen much of an argument for a single arrival or a single original language.

  • gcochran

    How could we know it, or at least strongly suspect it? Well, genetics: we know from genetics that the original settlers were few in number. Against, from genetics, it looks as if there was a single ancient arrival. A priori, that makes a certain amount of sense, since the first settlers have a huge dual continent to expand into. After even a couple of thousand years, the same number of people in a later migration would be a drop in the bucket, hardly noticeable. The Eskimo-Aleuts and the Na-Dene seem, for the most part, to have originally settled the far north, which became habitable later and required specialized survival skills.

    And of course, according to Greenberg and Ruhlen, there are shared characteristics of those Amerind languages.


Discover's Newsletter

Sign up to get the latest science news delivered weekly right to your inbox!

Gene Expression

This blog is about evolution, genetics, genomics and their interstices. Please beware that comments are aggressively moderated. Uncivil or churlish comments will likely get you banned immediately, so make any contribution count!

About Razib Khan

I have degrees in biology and biochemistry, a passion for genetics, history, and philosophy, and shrimp is my favorite food. In relation to nationality I'm a American Northwesterner, in politics I'm a reactionary, and as for religion I have none (I'm an atheist). If you want to know more, see the links at


See More


RSS Razib’s Pinboard

Edifying books

Collapse bottom bar