Who’s Afraid of the Big, Bad Theory of Evolution?Published 26 November, 2010
I think that Brian Scholl is one of the best experimental psychologists working in the field today. There, I said it.
I’ve known Brian – I’ll refer to him by first name here, and with apologies for the informality – for a number of years now, and he is not only among the best experimental psychologists working in the field today – there, I said it again – and he’s also a Really Nice Guy.
Ok, with that in mind, in today’s post, I’m going to briefly discuss, and then make some small, vaguely critical remarks about, a piece that recently came out of his lab, just published in Psychological Science. (Tao Gao was the first author on the paper.)
Brian does research on vision, and he and his collaborators have demonstrated a number of very interesting visual effects. This work investigates how people perceive, basically, pack hunting. They had subjects watch a display with shapes moving around and, in some of them, the shapes move randomly but “point toward” a disc that is also moving around. In one study, the subject controls a little green disc, and the task is to avoid the “wolf,” one of the other shapes on the screen. When there was a wolf pack in the stimulus – those shapes moving around, pointing at the subject’s sheep – performance wasn’t as good as when there wasn’t a wolf pack. In other words, the presence of the wolf pack stimuli, though irrelevant to the evasion task, impaired performance. (You have to check out the demos; I think this one is among the most compelling.)
The authors present a number of similar findings in a series of experiments, with a number of interesting control conditions. When there are good cues that multiple individuals are oriented toward a target, it really does give rise to a compelling sense of animacy.
Ok, now my quibble. Readers of this blog would probably have some ideas about why the “wolfpack effect” exists. It’s not crazy to think that humans – even humans living in New Haven, with little experience of being encircled by canid predators – have reliably developing mechanisms that are designed to deliver inferences and predictions about the social behavior of multiple individuals whose orientations toward a target are coordinated in a way that would facilitate predation.
It’s actually really hard to imagine that this research group didn’t have a function along these lines in mind when they designed the stimuli; and, of course, they call it the “wolfpack effect” because they do have such a function in mind. My question is, why is there no explicit reference to the evolved function of this system? The work is presented as a demonstration of an effect, but no explanation for its existence is offered for it other than, again, the references to wolfpacks. It’s a little like the authors are winking at the reader, saying, “look, we all know that we have an evolutionary explanation in mind – talking about wolfpacks makes this pretty clear – but we also all know that it’s unfashionable to say so. Therefore, we’ll just leave the animating idea, as it were, implicit.”
What is with the reluctance? Would the hypothesis that the human visual system is sensitive to these cues of coordinated activity become less falsifiable if a motivation for it in terms of evolved function were provided? What is the harm in being up front about the explanation for why these particular psychophysical cues draw the viewers’ attention in this way? In this context, I might note that Brian isn’t allergic to such functional explanations, in general. His work on motion-induced blindness (with Josh New) includes an idea about a particular function. (I’ll omit the details here; it’s a very cool idea though.) But it’s clear that you wouldn’t have checked to see if this very particular pattern of stimuli – coordinated pointing toward a target – without the idea that this very particular pattern was useful for people to detect.
It seems to me that this is a symptom of a broader phenomenon. It’s fine to talk about evolved function for low level stuff, like the motion sensing and object recognition systems. (There, indeed, systems are named in terms of their function.) When the content is social, it becomes taboo to refer to evolved function.
I suppose Brian would say – feel free to chime in, my friend – that calling it the wolfpack effect is more or less naming an evolved function. Maybe. But if so, then there would have been little harm in saying that in the piece. Right?