March 27, 2003

Walking Into the Past

Rich Baker walks into the human past with Luca Cavalli-Sforza's Genes, Peoples and Languages as his guide.


Sharp Blue: Genes, Peoples and Languages: There is a desert in humanity's past. Here, on the shores of the future, there are cities buzzing with commerce and industry, linked by highways and airliners and sea-lanes plied by gigantic cargo ships. Satellites race overhead, data streaks across global networks. A dizzying array of factories process raw materials into progressively more complex arrays of products. Start walking in the direction of the past and things slowly change. For the first few centuries of our trek, the cities are much the same but they become smaller and dirtier and less interconnected. Then suddenly, we reach the pre-industrial hinterland. For the next thirty or forty centuries, we're in a rolling countryside of small towns and scattered farms and the occasional sea full of sailing ships. Occasionally empires rise and fall. Clothing and the products of industry becomes simpler. Here and there are picturesque pyramids. Iron and steel disappear, then writing. For another seventy centuries, there are farming villages and guys riding around on horses with bows and stone axes, and the occasional small town of stone or timber. Then, even the farms vanish and we reach the desert. Hunter-gatherers live here. They are few in number, but look like us, only dirtier, uglier, and not wearing jeans and t-shirts. This desert stretches off to the horizon, spreading itself over hundreds of centuries of uncharted wastelands. The Palaeolithic Desert and its Neolithic borderland areso far from the shores of the future that they're strange and exotic and mysterious places. We know little of what happens there, of its peoples' histories and cultures and customs. The trackless desert covers most of the time for which anatomically modern humans have inhabited the Earth.

The traditional way of looking into prehistory is through archaeology. This studies the artifacts left behind by the people of the past: tools, paintings, campsites, buildings. Unfortunately, archaeology can't answer all the questions we'd like to ask. Archaelogical artifacts are the products of culture, and it's not clear if the spread of distinctive classes of items is a sign of the diffusion of culture or the migration and expansion of peoples. The population geneticist Luigi Luca Cavalli-Sforza has contributed as much as anyone to the understanding of these issues, and in his popular science book Genes, Peoples and Languages he explains his findings. The central idea is that the ancient migrations and diversifications and interbreedings of peoples have left traces in modern genetic diversity, and by studying these traces we can reconstruct the largest scale structure of the history.

After an introductory section on genetics, genetic dating, mutation and genetic drift, Cavalli-Sforza describes how genes from populations across the world can be used to construct a tree showing progressive divergences in the expansion out of Africa, through the Middle East, along the south cost of Asia, into Europe, from south Asia northwards, and from north-east Asia into the Americas. There follows a tour-de-force in which he outlines his work on the principal component analysis of genetic diversity in Europe. Principal component analysis separates out independent contributions to modern diversity. Each principal component corresponds to an expansion or contraction of an ancient population. When plotted on a map of Europe, the contours of the first principal component almost exactly coincide with the contours of the spread of agriculture from the Middle East. It thus seems that populations of farmers carried agricultural society across the continent rather than the idea of farming spreading by cultural diffusion. The second principal component contours trace the repopulation of Europe after the last ice age. The third component displays an expansion from the steppe to the north of the Black Sea, and supports the hypothesis that Indo-European nomads originating from that region brought their languages and culture to Europe. The fourth principal component shows the expansion of the Greeks along the shores of the Mediterranean. The fifth component shows the contraction of the Basques. That such features of history and prehistory can be read so clearly from genes seems to me remarkable.

The following chapters describe the reconstruction of linguistic trees and their comparison with genetic trees, and the mechanisms of the spread of culture. The section on linguistics is interesting and lucid but much more familiar than Cavalli-Sforza's own work. Unfortunately, the last section, on culture, is neither lucid nor interesting. Indeed, it feels tacked on, vague and incoherent. Nevertheless, Genes, Peoples and Languages is well worth reading, and it's made me want to eventually read his technical work The History and Geography of Human Genes.

Posted by DeLong at March 27, 2003 01:49 PM | TrackBack

Comments

Can't wait to compare with Jared Diamond's Guns, Germs and Steel thoughts. Thanks for the pointer.

Another one to throw into the mix would be Richard Nisbett's The Geography of Thought.

http://radio.weblogs.com/0117032/2003/03/24.html#a129

Posted by: Zack Lynch on March 27, 2003 02:58 PM

Can't wait to compare with Jared Diamond's Guns, Germs and Steel thoughts. Thanks for the pointer.

Another one to throw into the mix would be Richard Nisbett's The Geography of Thought.

http://radio.weblogs.com/0117032/2003/03/24.html#a129

Posted by: Zack Lynch on March 27, 2003 03:02 PM

Two quibbles: horseback-riding is relatively recent -- the earliest date I remmebr is about 4000 BC, though it could be a couple of millenia further back.

And the dirtiness of peoples correlates mostly with the convenience of bathing (for example, in rivers, stream, and pools), not with level of development.

Posted by: zizka on March 27, 2003 03:44 PM

Quick comment, for what it's worth. There was an article by this author in Atlantic Monthly (~2 years ago) perhaps an exerpt from this book. Amazing to me was his description of the peopling of the Americas about 15,000 yrs. ago, from Siberia. He says that 95%(almost all, don't remember precisely) of the indigenous inhabitants in the Americas could trace their ancestry to ONE man, who had a distinctive genetic mutation on his Y chromosome, so his progeny could be tracked. So Apaches, Hopi, Maya, Aztecs, Incas, all had one common ancestor. And either there were very few crossings from Asia, or almost all ended in extinction for those groups who tried. And that the migration was relatively rapid, down the west coast to Chile - the advantage to this route being in transportation (coast-hugging canoes) and in having one, consistent ecological niche to exploit. Then they spread east across the continents.

Posted by: andrew b. on March 27, 2003 10:51 PM

Cavalli-Sforza is one of the scientists on whose work Diamond's book builds. The first two books cited in the "Further Reading" section of Guns, Germs and Steel are Cavalli-Sforza's The History and Geography of Human Genes and The Great Human Diaspora. Genes, Peoples and Languages doesn't cover technological, cultural and social development in detail as the locus of his interests is migrations and expansions of populations. I've just started reading another book which covers all of that stuff: Rudgley's Lost Civilisations of the Stone Age. It's good so far, and I'm sure a review will follow eventually.

Posted by: Richard Baker on March 28, 2003 06:49 AM

According to linguists the Americas were settled in three waves: the original wave (95% acc. to C-S) which was apparently one man and his family, or one clan; the Athabaskans, mostly in N. Canada but also Apaches and Navahos; and Eskimos and Aleuts.

The Eskimo way of life is not primordial but was developed in historical times (though we don't have a history of it of course). Eskimos are said to be technological whizzes who survived by dint of a lot of ingenuity.

Posted by: zizka on March 28, 2003 08:30 AM

"Two quibbles: horseback-riding is relatively recent -- the earliest date I remmebr is about 4000 BC, though it could be a couple of millenia further back."

Actually, it's no quibble. It turns out that horses were domesticated at the same time and in the same place as the origin of the Indo-European nomads (more or less Ukraine today), and that likely the technological edge of having horses was one of the factors which lead to the rapid spread and success of languages in europe, the middle east, and asia.

Posted by: X-patriot on March 28, 2003 10:04 AM

Diamond gives 4000BC as the approximate date of the domestication of horses in the Ukraine. Cavalli-Sforza says Indo-European originated about 9000-10000 years ago and the Indo-European expansion began 3-4000 years later; but he also reports that the archaeologist David Anthony thinks that the Kurgan culture domesticated horses around 3000BC. This all suggests to me that the date of domestication isn't yet very well determined, so I think that although I was a little sloppy with that part of the review I wasn't too sloppy. And, of course, there were not just towns of timber or stone, but lots of towns of mud-brick too.

Posted by: Richard Baker on March 28, 2003 11:12 AM

From what I've seen, the Indo Europeans used oxcarts and fought as infantry. The big cavalry armies from the steppe appeared in the West around 700 BC and in China around 300 BC. MacNeill suggests that there might have been a symbiosis between the steppe peoples and the Assyrians. In any case, cavalry are not primordial, but much younger than civilization.

Posted by: zizka on March 28, 2003 04:06 PM

I've actually read both of the Cavalli-Sforza books mentioned, and in fact I own a copy of "The History and Geography of Human Genes." For those who want a to follow human population genetics more closely, I'd recommend a Google search for the American Journal of Human Genetics' website. They have an online archive of pretty much everything they've published in the last 8 years, and many of the most exciting papers in this field have been published in the AMJHG. Access to all but the most recent 6 issues is completely free. Another good source is the Proceedings of the National Academy of Science (PNAS), which also has a 6-month restriction, and finally, PubMED is useful for looking up references.

One thing worth mentioning is that there is really nothing special about the fact that 95% of Native Americans can trace their Y chromosomes back to a single shared ancestor within the last 26,000 years - given any given set of human males, one can always find a Most Recent Common Ancestor (MRCA) along the Y line. This is a simple consequence of Coalescent Theory.

To simplify massively, if we model gene inheritance within a population of fixed size using a random variables approach, given the fact that any gene has a less than 100% chance of propagating into the next generation, we see that most lines are destined to die out, and eventually the entire population will carry copies descended from a single one of the individuals in the initial population. It is a fact that the effective human population size for most of history hasn't been constant, but the model can be extended to accomodate this fact.

This is why it is common to speak of a "Mitochondrial Eve" or of an "Adam" in population genetics - there is nothing special about the mitochondrial genes or the Y-chromosome, other than the fact that they do not recombine, and in consequence are easier to study. Generally speaking, given any two individuals of Old World ancestry (by which I mean of European, Asian or African ancestry), it is pretty much a certainty that said individuals share an ancestor somewhere within the last 2000 years, and pretty much everyone in Europe and the Middle East shares an ancestor within the last 1000 years (this isn't just theoretical speculation - the presence of certain tell-tale alleles at a high frequency in even relatively "pure" [i.e, isolated] European and East Asian populations shows the contributions of Sub-Saharan Africans to their gene pools in the relatively recent past).

From the viewpoint of evolutionary biology, what I find noteworthy is the tendency of science writers, and even "real" scientists such as anthropologists, to assume that the MRCA estimate of ~150,000 years given by the mitochondrial data is a proper estimate of the origin of the human species: in fact, it is no more than an upper bound, and not even the lowest one, at that. Looking at, say, the ABO blood group, would give an estimate in the millions of years, predating the split between us, gorillas and chimpanzees, which is obviously ridiculous, but the same scepticism isn't applied to the estimates from mitochondrial DNA.

Looking at the Y-chromosome data, a more likely estimate is that our species originated no more than ~60,000 years ago, and certainly no more than 80,000 years ago, with the MRCA for all of us alive today existing no more than 50,000 years b.p. The early migration of the Australian Aborigines helps to set a firm, fossil-dated lower bound which no genetic evidence can contradict.

Posted by: Abiola Lapite on March 28, 2003 04:18 PM

I have to admit that Cavalli-Sforza does occupy a somewhat special place in my own private gallery of intellectual 'heroes' (probably in the space between Zvi Griliches and Wittgenstein).

He is tremendously interesting for all sorts of reasons, but here two will have to suffice.

First, he is the sort of scientist who got out of his office to have a look at what was actually going on in the world. The most interesting case in point being his investigation of the pygmies of Africa's tropical rain forest which began in 1966. Why would a geneticist want to do anthropological field work? Because, and in his own words, he wanted to 'understand'.

This need to understand lead him to investigate how social complexes evolve, and how there can be, again in his own words 'societies without a future'. How did the pygmies get stuck? Answering this question may help us understand what may be happening on a goat track not too far from Nasyria right now.

Another of his questions, and this one is still hotly contested is: do people or technology move? Gregg Clark has sctrached his head asking something similar about textile machinery in 19th century India. The argument in terms of the agricultural 'revolution' seems to boil down to whether there actually was one (accompanied by dramatic population migrations, Cavalli-Sforza's view) or rather was there a steady slow-drip diffusion.

I clearly prefer the former view because it fits in with my understanding of the importance of 19th and 20th century migratory processes. But this could be just 'wishful thinking' because the evidence is contradictory. In any event there is a clear missmatch between how we perceive ourselves to be (stable, sedantary cultures, deep rooted nations etc) and how we all too often are (a race of invetarate immigrants). The truth, as always, is to be found somewhere in the mixing.

For those who are interested in taking this further, and not tired of banging their heads against the wall of intractable problems, this collection is a good State of the Art summary:

http://www.amazon.com/exec/obidos/ASIN/1902937082/qid%3D1049182622/sr%3D11-1/ref%3Dsr%5F11%5F1/002-5758523-1993634

Which brings us to two more names to drop: Colin Renfrew (more from the 'there'll always be a Britain school') the archeogentics 'guru', and William Calvin for his work on climate and evolution, which while it may not be proven right in the details, is certainly interesting and thought provoking.

Posted by: Edward Hugh on March 31, 2003 11:41 PM

I have to admit that Cavalli-Sforza does occupy a somewhat special place in my own private gallery of intellectual 'heroes' (probably in the space between Zvi Griliches and Wittgenstein).

He is tremendously interesting for all sorts of reasons, but here two will have to suffice.

First, he is the sort of scientist who got out of his office to have a look at what was actually going-on in the world. The most interesting case in point being his investigation of the pygmies of Africa's tropical rain forest which began in 1966. Why would a geneticist want to do anthropological field work? Because, and in his own words, he wanted to 'understand'.

This need to understand lead him to investigate how social complexes evolve, and how there can be, again in his own words, 'societies without a future'. How did the pygmies get stuck? Answering this question may help us understand what may be happening on a goat track not too far from Nasyria right now.

Another of his questions, and this one is still hotly contested is: do people or technology move? Gregg Clark has sctrached his head asking something similar about textile machinery in 19th century India. The argument in terms of the agricultural 'revolution' seems to boil down to whether there actually was one (accompanied by dramatic population migrations, Cavalli-Sforza's view) or rather was there a steady slow-drip diffusion.

I clearly prefer the former view because it fits in with my understanding of the importance of 19th and 20th century migratory processes. But this could be just 'wishful thinking' because the evidence is contradictory. In any event there is a clear missmatch between how we perceive ourselves to be (stable, sedantary cultures, deep rooted nations etc) and how we all too often are (a race of inveterate immigrants). The truth, as always, is to be found somewhere in the mixing.

For those who are interested in taking this further, and not tired of banging their heads against the wall of intractable problems, this collection is a good State of the Art summary:

http://www.amazon.com/exec/obidos/ASIN/1902937082/qid%3D1049182622/sr%3D11-1/ref%3Dsr%5F11%5F1/002-5758523-1993634

Which brings us to two more names to drop: Colin Renfrew (more from the 'there'll always be a Britain school') the archeogentics 'guru', and William Calvin for his work on climate and evolution, which while it may not be proven right in the details, is certainly interesting and thought provoking.

Posted by: Edward Hugh on March 31, 2003 11:44 PM

Those who weren't sent to sleep by my previous post will have noted that I threatened TWO points. The second one is this: Luigi Cavalli Sforza is important for what he teaches us about what science is. Using his own words:

'Can the history of humanmankind be reconstructed on the basis of today's genetic situation?

This was the question I posed myself over forty years ago. I made a personal bet that it could be done, because I believed the theory of evolution gives us the key. Human genetics was an infant science when I started working in research in 1941, and I spent twenty years obtaining the tools I needed to answer it.'

What I think this tells us is that 'good' science is all about being able to think outside the box, having what Kant would have called strong and clear intuitions, and then having the determination to follow them through, whatever the critics say to begin with. Important ideas are nearly always wrong in their details when they begin their career.So I'll close with my favourite quote from Arthur Eddington (another from that land of dark satanic mills that is forever.....):

'It is a good rule not to put too much confidence in the observational results that are put forward until they are confirmed by theory.'

Posted by: Edward Hugh on April 1, 2003 12:07 AM
Post a comment