I think I will try another post on sort-of-anthropology, again with webcomics. Everyone loves anthropology posts! Right? Right? I can see your reaction now!
So in my post on inequality, I alluded to the fact that some of the evidence given in the talks I discussed seemed a little sketchy to me. This wasn’t meant as a dig on the researchers involved. Instead, if anything, it reflects the difficulty of the study. In fact, you’d have to be a little odd to study it. Steve Shennan, for example, is a professor of theoretical archaeology. Peter Turchin is a professor of mathematical history (n.b.: that isn’t history of mathematics!). Do these professions strike you as normal? Probably not.
The reason these professions are not normal is indeed that it is incredibly difficult to study history and anthropology empirically. Take Peter Turchin’s research on war as a motivator of supersocial behavior. His hypothesis is that, based on Price’s Equation, places at natural geographic boundaries are much more likely to lead to empire creation. To test his hypothesis, he collected information on geographical location of empires. But how much data was there? There just weren’t that many empires. And that brings in the choice of dataset. The empires of nomads in Central Asia were usually called different things. The Roman Empire was always called the Roman Empire, but existed for a very long time. Does that mean that they should be weighted differently? How? These questions are hard, and the only way to answer them, really, is with way more data. But I can’t think of a way to get this data from the real world without trying to make new empires, which isn’t a very sunny prospect.
It’s a little funny: anthropology and psychology are both social sciences, and, on certain topics at least, pretty closely border one another. And they both have rather suspect reputations among people who prefer “hard” science. And yet, the reason for this is opposite. Psychology has so much data readily available, the perception is that they unscrupulously use it to prove whatever the hell they want. In Anthropology, the data is so sparse, there is just very little that can be said with scientific rigour.
Think about it this way, and it starts making a lot of sense to try to simulate history. Suddenly all the world-as-simulation stuff starts sounding less and less crazy and you get to wondering when everyone will finally start mooning the aliens.