David Sloan Wilson has just posted a five-part series on the importance of the evolutionary toolkit in the social sciences. I’ve found the series hard work, but in the fifth post Wilson has pointed to an interesting exchange in Edge between his cousin Timothy Wilson and Steven Pinker. Timothy Wilson starts with an examination of the state of social psychology, and then turns to the role that evolutionary psychology can play:
There are some striking parallels between psychoanalytic theory and evolutionary theory. Both theories, at some general level are true. Evolutionary theory, of course, shows how the forces of natural selection operated on human beings. Psychoanalytic theory argues that our childhood experiences mold us in certain ways and give us outlooks on the world. ... But both theories led to a lot of absurd conclusions, and both are very hard to test rigorously. ... Evolutionary theory ... can explain virtually anything. It can be a useful heuristic, as I mentioned. But at the same time, I think it is way too broad.
To make his point, Wilson creates an adaptive explanation of why blood is red. Pinker swats it away like the fly it is - drawing on the chemistry and physics, the non-adaptive explanations for red blood are known. Pinker also suggests a range of empirical tests of Wilson’s faux claim.
In further defence of evolutionary psychology, Pinker argues that evolutionary psychology has been successful and, in particular, uses empirical evidence:
In a 2003 _Psychological Bulletin_ article, David Buss listed fifty novel predictions about social behavior derived from evolutionary theory, most of which had been supported at the time by empirical tests. Entire fields of social-psychological research—on violence, love, beauty, motherhood, religion, sexual desire, parent-offspring conflict, dominance, status, self-conscious emotions, and yes, sex differences (which everyone in the world but Wilson thinks is an important phenomenon)—have been driven by tests of evolutionary hypotheses. Many other evolutionary hypotheses—the nepotism theory of homosexuality, for example, and the Trivers-Willard hypothesis applied to female infanticide—have been empirically falsified as well, leaving the phenomena in question unexplained. It's simply not true _that evolutionary hypotheses that make correct empirical predictions_ can "explain anything."
Having defended his turf, Pinker then lines up social psychology:
Why doesn't social psychology get more respect? I readily agree that social psychology, not least Wilson's own research, has made profound discoveries, which deserve a greater place in policy and personal recommendations. But the field has been self-handicapped with a relentless insistence on theoretical shallowness: on endless demonstrations that People are Really Bad at X, which are then "explained" by an ever-lengthening list of Biases, Fallacies, Illusions, Neglects, Blindnesses, and Fundamental Errors, each of which restates the finding that people are really bad at X.
When you make broad statements such as this to experts in the field, you are almost assured of getting counter-examples - and Wilson and some other participants in the discussion give just that. However, Pinker keeps asking why, why, why. There is a limit to the generality of an explanation if it ignores, in the case of social psychology, the biological underpinnings. Pinker describes this issue nicely:
A satisfying explanation invokes principles that are fewer in number, more general, earlier in the causal chain, and closer to irreducible physical and mathematical laws than the ones that immediately fit the data in question. And that will almost always take one outside the boundaries of one's academic specialty. In the case of social psychology, any explanation must ultimately invoke a conception of what our social emotions and reasoning processes are _for._
Reading this, I kept thinking how similar Pinker’s description of social psychology is to my perception of behavioural and experimental economics. Behavioural economists have generated a mass of biases, illusions and heuristics but lack a framework to put them together. It is reflected in wikipedia pages like this. Claiming that there is no framework at all is overreach, in the same way that Pinker’s generalisation may have missed some specific examples, but as a general critique it holds. Each time a new paper comes out that finds a bias or heuristic, I generally don’t read beyond the abstract. There are so many of them that it is hard to know if it matters. Further, if there was a framework, we would probably discover that many of the biases are versions of the same feature. We’d end up with a much smaller list.
Pinker closes by making an obvious point:
Tim asks who would be best equipped to solving a social problem, a social psychologist, an evolutionary psychologist, or an economist, but this strikes me as the wrong question. The right question is, who is better equipped, a social psychologist who uses relevant ideas from evolutionary biology and economics (and other fields), or a social psychologist who doesn't?
To me, the point is so obvious that it is trite, but it seems to need plenty of repeating. In his commentary, David Sloan Wilson noted the lack of cross-referencing between evolutionary and social psychology - and that lack applies both ways. In his words, they inhabit “parallel universes”.
And again, this point applies equally to economics. Who is better equipped: an economist who uses relevant ideas from evolutionary biology, or an economist who doesn’t? Over the next decade, the evidence is going to strongly favour the economist trained in evolutionary biology.