New Discoveries About Human History

Logic Bomb writes "The New York Times has an extensive article (free reg req, of course) about how scientists are finally able to take advantage of genetic data to really trace human history. Tracing mutations in genes is allowing anthropologists to map humanity's migrations going back 50,000 years...far longer than the 3,500 or so that we've been keeping historical records for. It's quite an interesting read."

Comments

Paper from Science

[Mad Hughagi]

Here's the actual paper found in the Nov. 10th issue of Science that prompted the media run-offs.

Re:Paper from Science

[Black Parrot]

FWIW, the paper by Colin Renfew that the authors cite (reference #3) is almost universally rejected, nay, despised, among people with the faintest clue about the prehistory and interrelationship of the various languages in the Indo-European family.

Re:Paper from Science

[Black Parrot]

> Maybe it's a problem to the linguists that Renfew's an anthropologist?

Interesting question.

First, let me point out that I'm not an authority on this. (I am at best an advanced amateur on the linguistic side.) My comment merely stems from observation of what actual authorities say, apparently very nearly unanimously.

Second, though I know much more about linguistics than I do about archaeology/anthropology, I am also a "fan" of those fields, and would be extremely reluctant to make any kind of blanket assumption that linguistic evidence trumps their evidence, or vice versa.

Third, I was vaguely aware that Renfrew was renowned in his field, and so perhaps I should have said more about that than I did. Also, I remember reading about Renfrew's IE hypothesis in Scientific American shortly after he first published it, and at the time I found it both sensible and appealing. It's just that now that I know more about it, I find it impossible to reconcile with the linguistic evidence. (Or, more accurately, I find that others better informed than myself find the reconciliation impossible.)

For people interested in that particular debate, I would recommend Mallory's In Search of the Indo-Europeans. Mallory is an archaeologist. The book is one step removed from a popularizing account, in the same sense that Scientific American is a step removed from a popularizing account of science; if you are moderately intelligent, moderately well-read, moderately patient, and fairly interested in the subject matter, you should be able to master the book.

Mallory apparently wrote the book partly, or perhaps even primarily, as a response to the controversy stirred up by Renfrew. However, the book is not written as a polemic. IIRC he mentions Renfrew explicitly only a handful of times (though he consistently disagrees with him in those mentions).

One very likeable thing about Mallory is that he does not feel compelled to produce a Grand Unified Theory That Explains Everything. He simply surveys the evidence and does the best he can with it. Metaphorically speaking, he proves some interesting lemmata, but finds himself unable to combine them into a theorem. More on Mallory near the end of this post.

[The following is for the benefit of any lurkers; I don't want you to think I'm trying to lecture you on topics that you probably understand better than I do.]

There is an interesting tension between linguistics and archaeology. I do not think of it as animosity, because I honestly think workers in both disciplines would really like to be able to use each other's work as orthogonal evidence to support a good theory. (Notice the current article's citation of both archaeologists {Renfrew,Gimbutas} and lingiusts {Ruhlen}.)

Unfortunately, this collaboration does not always work in practice.

There is a very appealing, albeit very naive, temptation to assume that a population of humans can be associated with a culture (what archaeologists study) and a language (what linguists study) in a 1:1:1 fashion. However, this is not necessarily the case.

The disconnect between population and language can be illustrated easily by looking at who speaks English in the modern world. Any notion of "the English speakers" as a meaningful genetic designation goes right out the window, since you can find native speakers of English who have biological ancestors hailing from every corner of the world. (To put it bluntly, look at the correlation between skin color and speaking English. But any other designation of a population on the basis of speaking English is not going to be very informative either, beyond the fact that they all do speak English.)

The disconnect between population and culture was well illustrated by an archaeology professor I once had. About a generation ago there was a rapid replacement, throughout most of the industrial world, of beverage containers made from tin-plated iron, by beverage containers made from aluminum. Unfortunately for any naive theories, this cultural horizon was not associated with any sort of population replacement, invasion, intrusion, or even fresh contact with another culture. It was simply an in situ development. The difference is hard, perhaps impossible, to detect rigorously simply on the basis of archaeological analysis.

So regarding the naive 1:1:1 relationship, you cannot say with certainty that either half holds absolutely, let alone the whole thing. Any attempt to map archaeology onto language, or vice versa, is going to have to rely on a chain of inferences only as strong as its weakest link.

Now back to Mallory. He reports on a very nice archaeological analysis by Gimbutas (also cited in the Science article, and well regarded by historical linguists) that makes a plausible chain of inferences between the Kurgan culture of the steppes and the speakers of an early IE dialect, probably ancestral to the Indic, Iranian, Baltic, and Slavic sub-families of IE (represented respectively by modern Hindi, Farsi, Lithuanian, and Russian, plus many other languages both living and dead).

Some would like to derive the entire IE language/population/culture from that Kurgan culture, but Mallory balks at that. (And IMO with good reason.)

The problem arises when you step back and look at western Europe. There, you can make an equally plausible chain of inferences to derive the speakers of the language(s) ancestral to the Celtic family (Irish, Welsh, etc.) from the Hallstadt culture, and arguably even back to the Urnfield culture.

So far, so good. The problem arises because nobody, whether linguist or archaeologist, seems willing to derive the Hallstadt/Urnfield culture from the Kurgan, or vice versa. So to preserve a unified early IE culture, which would seem to be a well-justified desire in principle, you would have to derive both the western European and the steppe culture from a single earlier culture. I'm not aware of anyone who would even posit a suggestion, except perhaps Renfrew. (And I say "perhaps" because I'm not sure even he would derive the Kurgan culture from an Anatolian antecedent. No one else that I know of seems interested in doing so, and at any rate Renfrew's posited common ancestor would be so far back in time that it stretches belief that the languages of western Europe would have remained so transparently related to the languages of western Asia well into the historical period.)

Despite many efforts to pin down "the IE homeland" over the last couple of centuries, the problem seems to be beyond solution with the knowlege we currently possess. At any rate that's my take on it.

I don't pretend that the above in any way disproves Renfrew, because frankly, I don't even remember the details of his model. I do stand by my assertion that, so far as I have observed, historical linguists almost universally reject his model. However, as suggested by tesserae, it is not hard to imagine that archaeoligists would subscribe to Renfrew and universally reject conflicting claims from linguists (though Mallory does not seem to fit that pattern).

caveat lector: the the sake of brevity, I have included some gross simplifications in the "facts" above. If not already so far off topic, interesting and sometimes critical comments could be made about Gimbutas and Ruhlen as well. Also, as the saying goes, "Your spelling may vary."

Side note re inbreeding.

[Black Parrot]

Notice that you have 2 biological parents, 4 biological grandparents, 8 biological great-grandparents, and so forth, for 2^N ancestors at the Nth preceding generation. This number grows exponentially.

Meanwhile, population has been growing geometrically (exponentially?) as well, but in the opposite direction in time.

So at some point, the number of your posited (great*)grandparents must exceed the human population of the earth at that time.

For example, take 30 generations ago, a nominal 25 years per generation, and assume no non-human input, then you can calculate that you needed about a billion (28*great)-grandparents a mere 750 years ago, when the world's human population was surely much less than one billion. Another ten generations back and you would have needed a trillion (38*great-grandparents), still a mere 1,000 years ago.

That's right, folks. We're all inbred like the veriest Ozark Hillbillies.

Humor aside, I wonder whether any biologists here can tell us something quantitative about the rate of the, erm, call it "biofeedback", in a modern human's genes, and how that rate compares to other species, such as chimps.

Re:Side note re inbreeding.

[spankfish]

This is an interesting point, but you seem to be forgetting that for the vast majority of human existence, parents have generally had more than one child. Usually at least two or three. Refactor that into the calculation, it ought to bring the numbers down significatly. Say for the 30 generation example, assuming only that each pair of parents produced 3 offspring. 3 children with two distinct parents. Let's call that 1 child with 2/3 of a distinct parent. Going back over time with the same rate compouding, you get the picture.

Although, yes, it's probably reasonable to assume that we are all related in some way. Although I wonder about couples who cannot have children - perhaps in some cases their DNA is just different enough to have a chance of not being able to create a human embryo.

Do we have species differentiation going on within the homo sapiens clan? We must, surely, or at least the very beginnings of it. It's ludicrous to think that we couldn't be evolving, after all, we're still shagging and mutating, and what else do you need for evolution to occur?

Also, you can't really leave the non-human element out of this. There is evidence around that homo sapiens and homo neanderthalis mated on numerous occasions.

Imagine the soap opera potential in that. (chuckle)

--