Joseph Greenberg and the Amerindians

Joseph Greenberg believed that the native languages of the Americas fell into three macro-families: Eskimos-Aleut, Na-Dene (mainly the Athabaskan languages, spoken in Western Canada and the American Southwest), and Amerind (all of the rest).  Often this theory was paired with the notion of three separate  migrations: first the Amerinds, then the Na-Dene, then the Eskimo.

It does indeed look as if there were three separate migrations – another gem from Reich and Patterson.  The study was marred by complete non-cooperation from Amerindians in the US, but they managed.  The Canadian Na-Dene tribe they looked at was mostly Amerind but had about 10% of the second wave.  The Eskimos were about half Amerind and half the third wave.  I would guess that Na-Dene groups such as the Navaho and Apache have a higher fraction of that second wave, because they simply look more like modern East Asians, as do the Eskimo.

Both of the later waves seem to have started out settling  areas in the far North, regions that became habitable late and where settlment required specialized techniques. Interestingly, the genetic sample from the old Saqqaq culture in Greenland looks to be almost entirely second wave ( 85%) – so some people linked to the Na-Dene likely occupied all of the Arctic coast of North America before the Eskimo.  Curiously, the Eskimos don’t show a noticeable fraction of second-wave ancestry.  They must have replaced them without much admixture.  Perhaps more detailed studies will show a bit. Lesson: when one human population expands into another, there is almost always some admixture (enough to transmit adaptive alleles, certainly) but there isn’t always much.  Laurent Excoffier thinks that such events are always similar to a ‘range expansion’ – diffusion and random mating – but that happens not to be the case. Not when AMH met Neanderthals and Denisovans, not in Neolithic Europe, not in the Philippines, not in the Americas.

Fairly recently, Na-Dene has been linked to a particular family of Siberian languages, Yeniseian, using standard linguistic techniques rather than  Greenberg’s multiple comparisons approach.  Of course, Merrit Ruhlen, a student of Greenberg, had made the same connection earlier using multiple comparisons, while Alfredo Trombetti suggested the same thing 90 years ago.

Most linguists studying Amerindian languages disagreed with Greenberg.  They thought (and still think) that it is impossible to identify a language macro-family that originated as much as 15,000 years ago.  They also to tend to dismiss evidence from genetics, not least because they don’t know any.

It is conceivable that Greenberg’s method is invalid, and that it repeatedly gave the right answer  (on the three-wave settlement of the Americas, on Dené–Yeniseian, and in his generally accepted classification of the African languages) by chance. Maybe Joe Greenberg was just lucky.

Certainly his critics aren’t.









This entry was posted in Uncategorized. Bookmark the permalink.

24 Responses to Joseph Greenberg and the Amerindians

  1. dearieme says:

    Maybe it’s like Sir James Black being lucky. Twice.

  2. SwamOwl says:

    Greg you may want to check the statement “Curiously, the Eskimos don’t show a noticeable fraction of second-wave ancestry.”
    From the abstract in Nature: “However, speakers of Eskimo–Aleut languages from the Arctic inherit almost half their ancestry from a second stream of Asian gene flow…”
    I was a bit confused with the example that Dr. Andres Ruiz-Linares used to explain that the Chipewyans, having mostly Amerind (1st wave) DNA, speak a Na-Dene (2nd wave) language. He used the example of the prevalence of Spanish in Latin America, probably implying a conquest scenario where the conqueror’s DNA got diluted into the greater population. However it is not clear to me that the bulk of Latin American DNA is non-European. Are there any sources to find out this percentages?

    • ghazi-less says:

      I’m not an expert, but my impression, too, is that Europe dominates Latin American ancestry. See some support for that here:

    • gcochran9 says:

      What I said was correct: the abstract put things in a slightly confusing way.

      As for ancestry in Latin America, things vary. In Mexico, average ancestry is 55% Spanish, 40% Amerindian, 5% African. In Bolivia or Paraguay, I think the Amerindian fraction is well over half. In Cuba, the African fraction is much higher than it is in Mexico. Etc.

  3. tommy says:

    Greenberg’s approach was invalid (he used a good deal of bad data), but not as invalid in some respects as in others. Sorting out the relationships between various Amerind language families is difficult and drawing any conclusions about sub-groupings of families based on Greenberg’s approach is questionable–more careful methodologies since Greenberg have suggested different relations in a number of cases. On the other hand, recognizing that the Na-Dene languages are very different from other Indian languages is much easier. Greenberg got the big picture right, it’s the details where he fell short.

    I wasn’t surprised that further genetic work on the problem would reveal multiple migrations. I recall a study a few years ago which suggested that there was only a single migration and I knew this couldn’t be right. I expected this would be another case, like early research that indicated the Finns had not even slight Asiatic admixture, that would be overturned by later findings. I’m sure further studies will clarify even more about these migrations. Perhaps even the first Amerind migration will turn out to be a lot more interesting than we currently imagine.

    • billswift says:

      Using lots and lots of data, especially from many sources, is a good thing, even if some of it is bad. Unless the data is biased somehow, and using many sources tends to reduce that risk, bad data tends to cancel out to some extent; leaving a usually useful remnant. Think of “wisdom of crowds” but with raw information rather than the people presenting and interpreting the information.

  4. Sid says:

    Steven Pinker states in “The Language Instinct” that after 10,000 years or so, a language will have so thoroughly evolved from what it was in the past that it will bear no resemblance to its past incarnation.

    Linguists who have studied the Mayan languages, Nahuatl, Quechua and Cherokee can’t find any familial connections between them, not the way you can find clear connections between Ancient Greek, Latin and Sanskrit. In all probability, the aforementioned Native American languages have a common origin, but it goes back some 15,000 years, which is a surplus of time for those languages to change their forms so utterly that they bear no resemblance to each other.

    • billswift says:

      I suspect that is largely a result of Sanskrit, Greek, and Latin having written forms, which tends to preserve the languages more.

      • Sid says:

        That’s probably true for vocabulary, maybe for grammar, but probably not so much for phonology.

      • Sid says:

        For what it’s worth, the earliest records we have of Greek are from 1400 BC (Linear B), Latin from 500 BC, and Sanskrit is unusual, since the first written records are from 1100 AD but the earliest compositions we have from the Rig Veda are estimated to be from 1500 BC. In short, these had about 2500-3500 years to evolve on their own without writing, since Proto-Indo-European began to individuate around 4000 BC.

  5. dearieme says:

    How could Mr Pinker know, Sid? No-one was writing languages down 10000 years ago, nor were their sound recording devices.

    • Sid says:

      Try reading reconstructions of Proto-Germanic (over 2,000 years ago), and see how similar that is to English. Now try Proto-Indo-European reconstructions (around 6,000 years ago). If English persists and one of its descendants is still around some 4,000 years from now, it will be dramatically different from what is spoken now. So over the course of 10,000 years, Proto-Indo-European will have evolved into something else altogether, with maybe the word “me” remaining the same.

      • dearieme says:

        In other words he doesn’t know, he’s guessing about the future, or using inventions about the past.

      • Sid says:

        There’s no doubt that the 10,000 year mark is an estimate. When you compare that estimate with languages as they’ve evolved over the written records, then it’s one borne out with experience.

  6. tommy says:

    Language evolution is complex and you have to keep in mind that the 10,000 year mark or so is going to be an average. Besides, nobody is claiming all currently isolated Amerind families split apart from one another at the 15,000 year mark. I suspect we’ll be able to unite more of the Amerind families in due time, but how far we’ll get is anybody’s guess. I agree that uniting them all looks extremely unlikely, but you have to flesh out the proto-languages in good detail before you find out how much more progress you might make and, even for established families, this seems like a slow and uneven process in Amerind linguistics. This is probably because the number of linguists familiar with any given family is usually tiny and not all of them are interested in attempting reconstructions.

    • Sid says:

      The problem is that Native American languages tend to be hard to learn. Damn hard. Nahuatl is a polysynthetic language, which apparently has no easy connections with the polysynthetic Mayan languages, which has no known connections with the polysynthetic Quechua. Learning just one of those languages is a headache, and learning one of them won’t help you jump to another one easily.

      Native American languages are also often moribund, so it’s hard to find speakers to converse with, and if you’re still speaking a Native American language instead of English or Spanish, you’re probably not going to be the most cosmopolitan person interested in expanding Western science.

    • ohwilleke says:

      “nobody is claiming all currently isolated Amerind families split apart from one another at the 15,000 year mark.”

      Actually, that is precisely the problem to a great extent. The time that elapsed from a first wave settlement to having the entire continent populated was probably less than 4,000 years. In much of the New World, especially the part not experiencing new waves of Asian immigrants, like South America, where linguistic diversity is greatest in the Americas, the next 10,500 years or more were eras where there is no genetic or archaeological evidence of much in the way of subsequent mass migration or conquest for most of the continent. You have rapid dispersal and then very long term isolation in situ where each little tiny region can evolve linguistically with only slight diffusion influence from their immediate neighbors.

      Eurasia and Africa probably looked similar linguistically to pre-Columbian South Amerrica apart from the Inca Empire, before they adopted herding and farming. It was probably highly fractionated politically and lingustically diverse. Modern language families owe their coherence to a great extent to the expansion of food producing populations (and at least as importantly, later waves of copper age to iron age conquests) in the last few thousand years and to the fact that as time as moved on the world has gotten smaller due to improving transporation and communication technologies. Today, you can go anywhere in the global on an upper middle class budget in two or three days. In the 1930s it took weeks and you had to be rich to do it. In pre-Columbian South America (which lacked even horses or camels) a three day expedition might have gotten you 75 miles if you traveled light and were swift, the distance, for example, from North Denver to South Suburban Colorado Springs – and the trip would have been dangerous and over sparsely populated land.

  7. Tschafer says:

    I never really understood all the invective directed at Greenberg by Americanist linguists. Even if you thought that the guy was totally wrong (and it doesn’t look like he was), there was really no excuse for the level of vituperation he experienced. Does anyone have an explanation for this? Admittedly, academics tend towards this sort of sorority-style hissy-fit politics, but the Greenberg case seemed to transcend this.

  8. RS-prime says:

    Good day for him, only it’s too bad Greenberg is as I recall gone.

    For some years I knew an Eskimo — racially unmixed to my knowledge — which I guess most people haven’t. This was in a mainstream Lower-48 context. Great guy, not totally dissimilar in character to a contempo E Asian. Not super similar to Mex & Central Mestizos, mostly foreign-born, that I used to know. (I realize of course that the latter are quite admixed with Euro.)

    To me there is an obvious affinity between Pac NW visual art and NEAsian art. I also see affinities between Pac NW taste and the (more interesting and better developed) Mesoamerican taste, whereas it’s pretty hard to pick up on any real Mesoamerican-NEAsian art nexus. If Pac NW and Mesoamerican people are both Amerind, it could be that the Meso people diverged more esthetically mainly because their presumably much greater population size conduced to greater rates of positive selection. I don’t know the archeology of their pop sizes, I just know that the sun’s energy makes potatoes pretty efficiently and salmon pretty inefficiently.

    I guess another, very contrasting hypothesis would be that form of drift which we call allele surfing. I’m aware of the debate but have no real opinion on it — would be hard work, no doubt, to attain to a worthwhile opinion.

    • RS-prime says:

      > it could be that the Meso people diverged more esthetically

      I might clarify that with the word ‘esthetically’, I was referring to their art. Not their own appearance — their faces, bodies or what have you.

  9. bob sykes says:

    Linguistics has shown itself to be another pseudoscience if it can’t accept Greenberg’s results.

    • Sid says:

      That’s a silly, silly proclamation. The basic reasoning for having an “Amerind” language family is that there was one wave of migration from Beringia some 15,000 years, and subsequent waves have related language families. If that’s the criterion for having a linguistic superfamily, then where do we stop? Since Caucasoids and Mongoloids both left Africa around the same time, they spoke a common language, from which so many Old World families must emerge. But are there are any familial linguistic ties between the Dravidian and Semitic languages? There are none yet discerned, grammatically or phonologically, even though they logically must share a common ancestor.

  10. a very knowing American says:

    Greenberg’s successes with African language macro-families, and probably with New World languages, suggest that we should take pretty seriously another proposal of his, the “Eurasiatic” family, including Indo-European, Uralic-Yukaghir, Altaic, Japanese, Korean, Ainu, several more Siberian families, and Eskimo-Aleut. Any theories about the prehistory here?

    • Sid says:

      There is the Nostratic superfamily hypothesis, which was popular in the former Soviet Union but was regarded rather skeptically in the West.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s