Last week I alluded to a view about evolution that motivates some to sympathize with sensorimotor theory, and more generally, with the view that perception is for bodily action, and not (except incidentally) for the detection of states of affairs independently of their relevance to bodily action.
One version of the argument goes something like this.
Selection cares only about (what Patricia Churchland has raffishly called) the “4 Fs”—Feeding, Fighting, Fleeing, and Reproduction. Since 4F success does not depend on apprehending the truth, selection cannot explain a truth-seeking organism, much less an abstract-truth seeking one.
Arguments like this have considerable currency. Alvin Plantinga takes them to show that God is needed to explain the reliability of human thought. From Stephen Gould’s perspective, they show that non-4F thought is a “spandrel”—a non-adaptive by-product of natural selection. And Tyler Burge writes that false perception “is not a failure of biological function”, since it is response not detection that is biologically functional.
Today, I continue my discussion of actionist theories of perception and thought by taking issue with this anti-intellectual view of human evolution. The 4F view is not mandated by evolutionary theory.
Some say that abstract truth-seeking is at best an add-on, more likely a wasteful luxury, in a world where only the 4Fs matter. But what if early hominins lived in groups that, for whatever reason, valued abstract thinking? In such groups, access to mates and to shared food may have differentially depended on a capacity to think abstractly. Assortative mating would then have brought about the unfitness of non-thinkers, regardless of their 4F capacities. Indeed no single individual would need to be a pure 4Fer if the group managed to share such duties. This is not a fanciful scenario: surely we need something like it to explain the extremely rapid evolution of human cognitive ability. In this scenario, we are the 4FK species, where ‘K’ stands for ‘Knowing’.
Of course, such reasoning does ultimately get back to the 4Fs—the capacity to think abstractly is needed to get food and reproductive opportunities in groups of a certain sort. But the point is that the evolutionary value of truth-seeking is not limited to sussing out situations for fit bodily response. Rather, the value of truth-seeking cognition is located in its contribution to group and sexual selection.
This scenario is just the start of the argument. A second important component is the observation that very primitive organisms are capable of association. By association, I mean the Humean/Pavlovian process by which a feature F can begin to trigger behaviour specific to feature G in organism O because F and G are repeated sensed as occurring together. In a very simple organism, the sensorimotor profile of F and G may merge. Yet, F and G would not become the same, for the organism can become deconditioned: the association of F and G can be erased.
The point I want to make quite generally is this: there are learning mechanisms that creates records of contingent associations without merging the sensory identity of the features it brings together. Christopher Stephens has shown that when these learning mechanisms ramify, truth-seeking becomes the best strategy. This lays the foundation for a type of organism that values truth independently of bodily response, and which congregates in bands that protect and valorize cognitive excellence.