Evolutionary psychology, demarcation, and grand unified theories
Last week evolutionary psychologist Satoshi Kanazawa published the highly controversial “Why Black Women are Less Physically Attractive Than Other Women.” I’m not interested in debunking his claims (that has been done quite ably elsewhere). Instead, I’ll add my two cents to a debate that cropped up on TNC’s blog. Should we consider evolutionary psychology a science? And why does it seem to repel people and is open to “hucksterism?”
I’ve discussed the science versus not-science distraction with respect to economics, and I think a similar analysis also holds here. What matters is whether a particular piece of ev-psych research is good scholarship, not whether the field as a whole can be classified as “science.”
The second question is far more interesting, and one that I’ve been meaning to blog about for a while. I think ev-psych, with its grand theorizing on how evolution explains everything we observe today, can sometimes display really bad physics envy. Outside of physics, it’s often not possible to explain all phenomena with a single set of laws. Some would argue it’s not even possible in physics.
At any rate, here is the comment I left over at TNC. Of course I quote Natural Reflections at the end!
Frankly, I don’t think we should even be trying to ascertain whether ev-psych is “a science.” That question is too vague to be meaningful. What we should be doing is determining whether a particular claim is backed up by sound evidence, strong analysis and general good scholarship. Being officially considered “science” should almost be besides the point. Some research is very weak and must be taken with a grain of salt even if it is lucky enough to be labeled science. I would put much biomedical science in this category. On the other hand, some historical scholarship (e.g. the role of blacks in the Confederacy) is bolstered by reams of supporting evidence. We rightly trust such work even though historians aren’t considered scientists.
It’s important to note that apart from the opening paragraphs, Ryan spends almost no time questioning the scientific status of evolutionary psych writ-large. Rather, he meticulously and carefully demolishes Kanazawa’s specific research. That’s the approach we should always try to use. In this case, Kanzawa’s work is crap whether or not we deem it science. And since, as Ace Rock shows above, certain aspects of ev-psych do have explanatory power, it makes no sense to offer either blanket praise or condemnation for the field.
As for ev-psych’s being open to hucksterism, I attribute part of the problem to the false seduction of grand unified theories. That is, we’re always trying to find the single theory that will explain everything about everything. Physicists have been very successful at this, and I think other scientists sometimes unfortunately take a similar approach. But once the question gets very hard (basically anything that can’t be described perfectly with math, which is basically everything humans care about), we’re left with the rather boring, mundane statement that we need many different factors working together to explain anything.
I think that many intuitively grasp this concept–that complicated phenomena don’t have single-cause explanations. And so when they see some evolutionary psychologists boldly claiming that they have found the reason for such and such phenomena, people get understandably annoyed.
Along these lines, I strongly recommend Barbara Herrnstein-Smith’s “Natural Reflections.” While the book is about science and religion, she does an excellent job eviscerating the strong interpretation of ev-pysch. From page 66:
In seeking to account for any complex behavioral, cultural, or social phenomenon, a good starting assumption would be that it was the emergent outcome of multiple factors of various kinds, operating at many scales and levels, interacting over time. The starting assumption of evolutionary psychology and “cognitive” approaches to religion, however, is that the best way to explain any behavioral, cultural, or social phenomenon is by demonstrating that it is the outward effect of the activation of some underlying mental mechanism. A methodological tradition of this sort puts a premium on ingenuity with respect to the hypothesizing of mental mechanisms and, by the same token, encourages negligence with respect to the investigation of possibly relevant environmental, experiential, and developmental factors.