Monday, April 23, 2018

David Reich, "Who We Are and How We Got Here: Ancient DNA and the New Science of the Human Past"

Friends have been asking me for years if I can recommend a book on the new science of ancient DNA. I have always said, "No, the field is changing too fast." But now there is finally something I can recommend, David Reich's very impressive Who We Are and How We Got Here.

There is not much original in this book, which summarizes the results of ancient DNA studies over the past decade. Reich has been in the thick of those discoveries, first as part of the team that showed Neanderthals and modern humans had interbred, then as the leader of a lab that has read and analyzed a huge amount of DNA. His writing is not great but it is good. You may find some of his technical explanations hard to follow; I did, and have been reading about this stuff for years. My advice is that if you get bogged down in a particular passage, just skip it and go on. If you do that, you should be able to get a huge amount of information out of this book however minimal your background in genetics or statistics.

I have been writing here all along about the big discoveries, which I would say are these:
  • Modern humans interbred with Neanderthals, and outside of Africa human genomes are on average 1 to 2 percent Neanderthal.
  • Modern humans in East Asia also interbred with another lineage of archaic humans we call Denisovans, although as Reich shows in this book the identity of Denisoavans is a complex story.
  • Our modern races are not ancient branches of the human tree, but the result of mixing that has taken place since the origin of agriculture; 20,000 years ago Eurasia was home to a different assortment of races, and the first farmers of Syria were as different, genetically, from those of Iran as Chinese and Welsh are today.
  • There are in the Eurasian family tree groups with no clear obvious descendants today, such as "ancestral West Eurasians"; we call these "ghost populations".
  • Modern Europeans are descended from three quite different ancient groups: the Paleolithic hunter-gathers of old Europe, the farmers who migrated from the Middle East beginning around 9,000 years ago; and invaders who came in from the Steppes during the Bronze Age.
  • The people of India are a mix of an ancient population ("Ancestral South Asians"), farmers who migrated from Iran during the Neolithic, and West Eurasians; the West Eurasian invaders seem to have come in multiple waves beginning in the Bronze Age. As you would expect by looking at people in India, the mixture of West Eurasian ancestry is much greater in the north, but there is some admixture even in the far south. This mixture skews male, indicating that the invaders were majority male, or else that they did their most successful breeding with local women.
  • East Asians are a mixture of an ancient Yangtze Valley ghost population, which spread after the invention of agriculture, with numerous other ghost populations, in particular one in the Yellow River Valley; modern Han Chinese originated within the past 5,000 years from a mixture of those two and possibly other groups.
Reich concludes by recapitulating the essay he published in the Times on whether this genetic data confirms or will feed racism. I have wondered why he went out on a limb over these issues, and some of my friends have as well. I can only imagine that he knows or expects that genetic data is going to come out that somebody will think has racist implications, and he wants to get ahead of the curve.

One of the anti-racist points Reich makes in both his book and his Op-Ed is that modern races are all mixtures of ancient races. There are no pure racial types; miscegenation is our legacy and our origin. It is true that some groups have remained relatively pure for the past 3,000 years or so, but compared to the vast sweep of human evolutionary history that just isn't very long.

In other words, the vast movements and upheavals of modern times – migrations, conquests, the rise and fall of empires, the disappearance of old peoples and the appearance of new ones – are not an aberration. The medieval world seems to have experienced a few thousand years or so of comparative stasis and endogamy, leaving us a legacy of people who think they belong to a pure race (Koreans, Japanese, Han Chinese, Germans, Jews, Celts). But that only goes back a certain distance, and we now have the tools to peer back much farther. In that long view, migration and mixing are the human norm.


Shadow said...

I have a question that may very well be ignorant given my lack of knowledge about this, but it has to do with this:

"Modern humans interbred with Neanderthals, and outside of Africa human genomes are on average 1 to 2 percent Neanderthal."

What does 1 to 2% really mean?

We hear things like a chimp's and human's genome are 99% the same. I think I've heard that 80 or 90% of our genome is the same as a watermelon's. And while this makes sense on one level because we are all cut from the same cloth, it makes little sense in comparison. We are obviously very different than chimps and watermelons. So how different has that 1 to 2% made us?

John said...

It just means that of your 20,000 or so genes, 200 to 400 come from Neanderthals rather than modern humans as they evolved in Africa. Nobody knows how much difference it makes, since everything important about us seems to involve the interaction of hundreds of genes. Reich says one of the genes is related to Keratin, the protein in hair and nails, so maybe the hairiness of some Eurasians derives from Neanderthals. Light skin seems another obvious candidate for something we picked up from Neanderthals, but I don't think there's any evidence for that yet.

I think Reich wrote his racism Op-Ed to get out in front of the possibility that some gene clearly related to brain function might eventually be shown to be very different one racial group or another. You can just imagine all the racists going on about variation in the HEX99 gene or whatever it is. But so far you can learn more about what genes do to us by looking at people than you can from their genomes.

Shadow said...

I asked this because we hear about genome similarities and percentages all the time, but almost no one ever says what they mean (or don't mean). We are not 95% like chimps. As I understand it, it's pretty much how you described it: the interaction between genes, between transcription RNA and DNA, and between other DNA components matters. This is why finding a single gene that might have something to do with intelligence may not matter as much as the initial finding indicates. Change at the gene level is ongoing (variability). Most of these changes don't matter, until and unless an environmental event occurs that triggers an advantage for certain gene combinations, and then things can start changing fast.