The Importance of Information
The affordances of an animal's environment exert selection pressure both developmentally and evolutionarily on the course of an animal's activities. But affordances are merely facts of the environment, and even if these facts are considered in relation to an animal, the obvious question arises: what information is available to the animal about those affordances? We have seen that jumping spiders learn to guide their pounces by starting at an appropriate distance from their prey. Worms pull leaves and other objects into their burrows by the appropriate edge, but what is the information that lets spiders or worms appreciate what counts as appropriate? Psychologists have typically assumed that "experience"--especially trial and error--provides this information. But even in cases where trial and error demonstrably plays a role in nature (and these are fewer than one might think), such trial and error cannot possibly be explanatory. Even granting the effects of trial and error, there is still the residual question as to how trial and error is evaluated as to its meaning by the animal--does this action get me closer to my goal or not? What information does a worm have that distinguishes the edges of leaves in terms of their meaning--that pulling the leaf by this edge (and not by that one) is more efficacious for keeping the burrow snug? James Gibson's concept of ecological information specifying its environmental sources provides a general framework for answering this kind of fundamental question about animal behavior and experience.
One of the most difficult questions about the evolution of behavior concerns the relationship between ultimate evolutionary effects and proximate behavioral activities ( Tinbergen, 1951; Curio, 1994). An animal cannot simply know, as if by magic, that behaving in a specific way will tend to increase its fitness in the future. All sorts of theories have emerged around this problem, from notions of quasi-intelligent, albeit "selfish" genes ( Dawkins, 1976; Williams, 1993), to speculations that animals monitor their physiological processes like cost-conscious economists ( Real, 1992).
There is a much simpler and more plausible hypothesis than those usually put forth to explain the relationship between current behaviors and ultimate evolutionary meaning. However, this simpler hypothesis is not well known and has not been discussed seriously in the literature on animal behavior. This is James Gibson's ( Gibson et al., 1982; Gibson, 1979/ 1986) hypothesis concerning ecological information. Gibson's idea is that, because some affordances of the environment are in fact very persistent, even with respect to phylogenetic time, there may exist in the environment information speci-