In this anti-behaviourist age, anyone who wishes to argue that cognitive processes might, in part, be constituted by physical manipulation of the environment must be prepared for the following sort of retort: And pigs might fly! This provides a rather contrived segue. Pigs do not fly, of course, because they do not have wings. But why is this? It is tempting to think that, from an evolutionary point of view, wings would be selectively advantageous, allowing one to escape one's most dangerous predators, to spot suitable food items, habitats, etc. So, one might be led to suppose that there would be a significant selection pressure towards the evolution of wings by pigs and other ground-based vertebrates. Why, then, have pigs not developed wings?

In the case of pigs, the answer is almost certainly that the necessary genetic variation has never been available (Lewontin 1979). And the reason for this, again almost certainly, has to do with the nature of evolution. There are pressing historical facts that constrain any straightforward development towards optimality. In evolution, there is no tabula rasa. To use an analogy developed by Dawkins, imagine the designers of the first jet engine being obliged to develop (to ‘evolve’) their product from an existing propeller engine, changing one component at a time, nut by nut, screw by screw, etc. Furthermore, imagine that not only the end product of the development, but also each successive intermediate stage, must provide a viable form of propulsion for an aircraft. In fact, not only must each successive intermediate stage provide a viable form of propulsion, it must provide a form of propulsion that is superior to the one that preceded it.