There's a new paper out in Genome Research
which shows something rather fascinating that seems to have happened in human history, right after the agricultural revolution. The two graphs below show the number of men (left) and women (right) alive at various times in history whose genes are still around today. (We can measure this separately because Y-chromosome DNA is transmitted only through men, and mitochondrial DNA only through women)
There are lots of reasons that you would expect this curve to increase as you move forward in time. The simplest part of this is the "common ancestor" effect. If someone is your ancestor, then all of their
ancestors are your ancestors, too. This means that, as you go farther back in time, anyone who's along your family tree is an ancestor of a bigger and bigger chunk of people. In fact, once you go far enough back, you'll encounter a person who is a common ancestor for everyone in your population group, or even in the world -- and once you've encountered this first common ancestor, every one of their
ancestors is a common ancestor, too! This means that a bit further back in time, you suddenly pass a second threshold: at that point, everyone
who was alive then is either a common ancestor of everyone alive today, or of nobody alive today. (If you're curious about this, there are a few famous papers by +Douglas Rohde
and a few others on this subject, where with a combination of historical population data and computer simulations, they managed to show that the most recent common ancestor of all humanity probably lived only a few thousand years ago, and in either southeast or northeast Asia: http://tedlab.mit.edu/~dr/Papers/Rohde-MRCA-two.pdf
So because of this effect, you would expect that as you go far back in time, the number of people who are ancestors of people alive today would end up being a roughly fixed fraction of the population: everyone's either a common ancestor, or not an ancestor at all.
Now, the other important thing about the agricultural revolution is that it made the population boom: grain fields can support orders of magnitude more people than hunting and gathering or nomadic herding. (This is also why the agricultural revolution leads to the original rise of cities)
If you look at the curve on the right -- estimated number of women who are ancestors of living humans today, as a function of time -- you see exactly that. Right around 15,000 years ago (15kya), the number of women skyrockets, and starts to level off around 10,000 years ago. This is exactly what you would expect if nutrition suddenly improved by a lot,
and it suggest that it was the early agricultural revolution -- that first cultivation of crops, rather than the rise of effective mass agriculture and the rise of early cities -- that had the biggest effect.
But the plot for men is bizarrely different. At 15kya, the gauge for men doesn't move. And then at 10kya, when the "big" agricultural revolution hits and cities start to emerge, the number for men plummets,
only to recover and show the giant population-related spike around 5,000 years ago. At its most extreme, the ratio of female to male ancestors was 17:1!
What happened here? The authors suggest that this was most likely a cultural effect, rather than a mysterious plague which only affected men. My own quick summary of thoughts:
(1) The effects which created the initial surge in female long-term reproduction, around 15kya, don't seem to have affected men much at all. This suggests that we're seeing a huge nutritional effect on the success rate of pregnancies.
(2) The crash in male reproduction around 10kya suggests that most men were suddenly unable to reproduce, even as lots of women were doing so. This means that small numbers of men were having lots
of children, and most weren't having any at all, or at least none which appear to have survived. Since you would suspect that most men might object to this, that suggests rather extraordinary application of force: i.e., the rise of the agricultural state brought with it tremendous power asymmetries and the rise of very wide polygyny.
(3) Around 5kya, this effect seems to have vanished even more quickly than it appeared. If anything, that's more fascinating, because 5kya is already within visibility of the literary record. (The story of the marriage of Inanna, for example, contains some fairly clear allusions to the tension between nomadism and agriculture) A change this rapid, from extremely concentrated harems to some kind of more level marriage system, would seem to require a tremendous social event going with it, something big enough that I'm surprised that we don't see at least allusions to it in a wide range of early literary records.
In fact, this third point is enough to make me actively suspicious: this is a huge
effect, something which would have defined human society for hundreds of generations and the response to which would likely have had effects for hundreds of generations to come. Its uniformity across geographic regions (colors in the graph) is similarly surprising: cultural shifts affecting the entire world don't Just Happen.
So I'm going to take this result with a great deal of caution until there's further confirmation, but the questions which it poses are fascinating, and this is clearly a direction worth more research.