Thursday, October 16, 2008

Hunter-gatherers and farmers, the continuing saga   posted by Razib @ 10/16/2008 04:18:00 PM

Sandy made another response to my assertions about HG's vs. farmers and quality of life, Agriculture Reduced The Periodicity & Amplitude Of Nutritional Stress. He's done a lot of research to support his specific contentions, and certainly everything he reports is generally true. But, I don't think it's necessarily relevant or representative of the issues at hand.

First, he says:
I've done some reviewing of the literature and I still think that the probability that a child of an agriculturalist will reach reproductive maturity is higher than that of a hunter gatherer. Hell, that's why there's been a population boom ever since the Neolithic revolution.

No, not necessarily. In a previous post he alludes to r vs. K strategies. Agriculture could simply be a r strategy. Children never born are not children ever had. So agriculturalists could have a higher growth rate by simply increasing the total number of infants produced over a reproductive lifetime. To make it explicit:

Woman 1 = 6 live births and 3 survive to reproduce
Women 2 = 12 live births and 4 survive to reproduce

Obviously the second woman's offspring have a higher mortality rate before reproductive age, but that doesn't matter, there are more copies of her genes floating around. Population growth or decline is a dynamic which can be the outcome of many combinations of the parameters (e.g., fertility, mortality and generation time). If you know that in population A the mean number of children per woman is 3, and in B it is 4, you can not conclude that B has a higher population growth than A without knowing the death rates.

Over the long term agriculturalists have had on average somewhat higher growth rates, otherwise hunter-gathering wouldn't have been totally marginalized as a way of life. But, that doesn't mean that populations grew very fast; the logic of compounding growth means that very small differences in average growth rates can result in great divergences over long periods of time.

But, as I said in my previous post the issue isn't really about looking over the long term. When the Europeans arrived in North America and Australia they perceived the land to be "empty." Now, obviously there were peoples settled in these territories, but they're average density was simply far lower than was typical in Europe. Why? Because these peoples did not have the cultural toolkit to extract as many calories per unit out of land. Europeans, with their more productive agricultural toolkit could naturally support far larger populations than the natives from the same land because of culturally contingent factors.

I think the same sort of dynamic can be projected back into the past. Imagine a group of agriculturalists who landed on the Breton coast 8,000 years ago, coming via sea from Portugal. The locals are hunter-gatherers who need enormous areas of range territory to support a small population. A group of farmers might be able to win or negotiate a "small" tract of land from the natives to farm. The natives might see this concession as minor; after all, they might not comprehend that a "small" territory from the perspective of farmers is actually going to be a "large" plot when conditioned on the agricultural techniques farmers had. During the early years farmers would experience plentitude as there is a surplus of land. Soon though they would need to expand, and here their ability to organize and project force because of their numbers would come to the fore. Not only would they be able to push back the hunter-gatherers, but likely powerful infectious diseases would sweep in front of their demographic wave. Soon, the only hope that hunter-gatherer tribes would have of defending against this would be take up farming themselves and become sedentarists. As this occurred the remaining hunter-gatherers would be pushed into marginal and less desirable land, further making the lifestyle relatively less attractive (any quality of life advantaged would be swamped out). There are plenty of data points which support these points from the last 500 years, from the European settlement of North America down to the relationship between hunter-gatherers and farmers in Botswana.

Most of the rest of Sandy's post is an analysis of correct, but irrelevant or unrepresentative data. Or at least I think it is irrelevant, and suspect it is unrepresentative. To a great extent much of the human race today lives outside of the Malthusian trap, and the conditions of the past do not necessarily apply to today. Additionally, though over the long term populations grew rather slowly in the pre-modern era, there were local fluctuations up and & down. We know this for a fact from cultures where we have demographic data (e.g., England and China), so a "snapshot" of a given period does not allow us to generalize well across time. Comparing contemporary hunter-gatherers today with contemporary agriculturists is probably not representative of past conditions. For me the biggest problem is that hunter-gatherers today are selection biased toward regions where agriculture is not at much of a comparative advantage. This is why I tend to be more interested in comparisons of remains within a particular region which went through the lifestyle transition in the past. My general impression is gross physiological health declined. In other words, farmers are, on average, subject to greater morbidity, even if hunter-gatherers have more mortality.

In any case, one of the problem here is bigger meta-issue in the debate between Sandman and myself:
Anyways, in my previous post I outlined some of the ethnographic reasons why I think so. In this post, I'll share some of the demographic data that supports my hypothesis.

I know less cultural anthropology than he does, and I'm not interested getting to know the whole literature. I could go to google scholar and look up some research articles which support my view because there are a lot of articles out there, and I'm sure I could find some trash which would support any contention I make. That's why I'm leaning back on the theory about Malthusian assumptions that I can be rather certain of, even if I'm wrong as an empirical matter I can argue the logic with some concreteness and consistency. I'm not convinced by Sandman for all the reasons above, but, at the end of the day I wouldn't be convinced by a few surveys of the literature because I'm sure I could selectively troll through the literature and find authorities to support me too. If I actually knew the literature with any fluency I could ascertain the value of any given citation I dig up, but since I don't know the literature I'm really concerned about just selection biasing what I get based on the way I formulate a google scholar query. This is also why I'm pretty down on the citation wars which occasionally crop up in the comments, I strongly have the suspicion that the interlocutors aren't really digging into each other's citations, but rather just going to search for more papers to support their own argument. What's really the point of all that? So I have decided to ask some people who really are in the "know" (that is, anthropologists who aren't CAFRs) what they think.

Oh, and also:
Razib equates hunter-gatherers to lions, and farmers to antelopes. Lion cub mortality rates are higher than antelopes, upwards of 80% of cubs do not make it to adulthood. Whereas antelope calves, if they survive predation, have much more greater chances at seeing sexual maturity. It is much more precarious to count on mom to provide milk and hunt at the same time, than it is to graze. So, this issue is less about absolute availability and more about reducing the periodicity and amplitude of nutritional stress. And this is one of the reasons why the agricultural diet has 'won' over many humans, even despite the many shortcomings it has.

To the first approximation I somewhat accept this, but I'm not totally sure, and the characterization might elide some of the details. Imagine a scenario like the chart below, where Y = nutritional quality and X = generation. The average for a hunter-gatherers might be higher, but the far higher "floor" of farmers might make all the demographic difference over the long term.

Addendum: I don't know cultural anthropology well, aside from the fact that CAFRism is endemic, but I do know something like the society of the Roman late antique period relatively well for a lay person, so I have seen people make arguments based on obviously retarded citations. Now, since these people are frankly retarded, dishonest or ignorant (or any combination thereof) any well informed response is really totally futile; people who can't evaluate the plausibility of their own argument based on the quality of the authorities which they cite aren't going to be able to evaluate the plausibility of someone else's authority. This explains why I'm responding to Sandman in broad theoretical generalities and methodological objections; if I "dug in" by spending time looking for literature support my case I'd turn this into a lawerly dialogue, and what's the point of that? Granted, I know Sandman better than if he was just a "random" from the internet, but still, life is short.