Just So Stories Are (Bad) Explanations. Functions Are Much Better ExplanationsPublished 24 September, 2012
A number of people have asked me about my reaction to Anthony Gottlieb’s September 17th piece in the New Yorker reviewing (unfavorably) David Barash’s recent book, Homo Mysterious: Evolutionary Puzzles of Human Nature.
I decided against writing a point by point rebuttal of Gottlieb’s remarks, first because I haven’t read Barash’s book, and second because the arguments Gottlieb advances have been addressed (repeatedly) before, here in the pages of this blog as well as elsewhere. (Barash has also written a reply.)
Instead I wanted to consider just one passage in the piece, Gottlieb’s favored explanation why human females don’t show the sort of swellings female chimpanzees show during the fertile phase of their cycle.
The simplest theory is that these swellings dwindled to nothing after our ancestors began to walk upright, because the costs of advertising ovulation in this way came to outweigh any benefits. Swellings could have made it harder to walk for several days each month, could have required more energy and a greater intake of water, and would be of less use as a signal when you were no longer clambering up trees with your bottom in males’ faces
This explanation for concealed ovulation – or, perhaps as Gottlieb would have it, the explanation for a lack of advertisement of ovulation – should be seen in juxtaposition with the theme of the article, entitled It ain’t necessarily so, with the tagline running beneath it: “How much do evolutionary stories reveal about the mind?” Gottlieb, following in the Gouldian tradition, is worried about evolutionary Just So storytelling.
The irony is that Gottlieb’s explanation is itself a Just So story: an explanation for a (lack of a) trait based on a theory of the relevant costs and benefits of a proposed function, in this case, efficient bipedal locomotion. As he himself might say, it ain’t necessarily so.
That is, of course, just fine. There’s nothing wrong with offering an explanation that might be wrong. Just because they’re false doesn’t make the stories any less explanations. Just So stories, in the sense of Kipling’s tales, happen to be very bad explanations, in no small part because they are inconsistent with what is known about the causal process that gives rise to traits like elephant trunks and camel humps. (The fact that they are bad explanations for the bits of animals can be forgiven. When Kipling was writing, Darwin’s ideas were quite new, and not universally accepted. Also, he was writing for children, so…) Kipling’s stories are especially bad scientific explanations because in addition to being in conflict with many known facts, they rarely if ever give rise to predictions that one might try to show to be false.
Not that it really matters, but how does Gottlieb’s idea about why human females don’t show genital swelling as chimpanzees fare as a scientific explanation? I haven’t carefully followed the literature on concealed ovulation, but it does seem to me that there’s something a bit funny about the comparative data. Orangs similarly don’t have genital swelling; do they spend more time walking upright than chimps do? (This isn’t a rhetorical question. My sense is that the answer is no, but I really don’t know the answer to the question.) That wouldn’t necessarily cement the issue, of course, but it hints at one way to go after the proposal. (It also occurred to me – and again, I’m not an expert in this area – that changes in color wouldn’t pose the mechanical problem Gottlieb is alluding to. So if, as Gottlieb seems to believe, there would be benefits to signaling, it is odd that there isn’t some sort of signal, a sort of vaginal blushing, that posed no impediment to walking?)
But whether Gottlieb is right or not is more or less beside the point. The point is – with all due respect to the importance of keeping these discussions on the highest possible plane – the deep hypocrisy that surrounds charges of storytelling. These charges often come from people –lay and scholarly alike – who themselves advance explanations that are no less susceptible to the charge. In the back and forth on this topic, for instance, I recall one proposal about how to distinguish stories from hypotheses, and in that case, the author’s own proposal failed the proposed test. Gottlieb’s explanation is no less, and perhaps no more, of a story than any other account.
I think, but am not sure, that the intuition underlying the “story” charge is the issue of evidence. In my interactions, the story charge is often followed by the claim that behavior doesn’t fossilize, and that we can’t know about history. This misunderstands the nature of claims in evolutionary psychology and the nature of what evidence is relevant to such claims. The argument about how evidence of design speaks to claims of function has been made in many places by many people, and I won’t repeat it here, but it seems possible that this one confusion lies at the heart of much of the endless discussion of this issue.
Another possibility is that there’s a more basic problem: that people find it hard, for whatever reason, to understand what an explanation is. (I thought David Deutsch’s The Beginning of Infinity was a nice recent treatment of the notion of explanation.) In my interactions with psychologists, I frequently find that people propose a word (“learning,” “culture,” and “salience” are favorites) as a potential explanation for a given phenomenon, even though most have come across the logical pitfalls of the “nominal fallacy.” I’m not suggesting that this confusion is limited to either journalists or psychologists. I actually think that there is something genuinely difficult about mastering the nature of explanation, and this difficulty is particularly visible in cases in which people like Gottlieb want to attack a book or a field for some entirely other reason.
That last point is important. One research area I have been thinking about recently is the literature on the willpower-as-resource model of self-control. This model is much more of a Just-So story than the proposals surrounding concealed ovulation, being unfalsifiable, and yet the response of the media (and scholarly community) has been adulation and praise rather than pointing out – let alone ranting about – how the model can’t be shown to be false (with one exception from Steve Pinker in the New York Times, who buried the charge of unfalsifiability in the penultimate paragraph amid an otherwise glowing book review).
For Gottlieb, and I think many like him, his own explanations are hypotheses (but not stories) while David Barash’s explanations are stories (but not hypotheses). In truth, they are both. The goal should not to expel stories from science, but rather to identify the stories that are also good explanations.