Many psychologists, like the philosophers before them, have looked inside themselves for explanations of their behavior. They have felt feelings and observed mental processes through introspection. Introspection has never been very satisfactory, however. Philosophers have acknowledged its inadequacies while insisting that it is nevertheless the only means of self-knowledge. Psychologists once tried to improve it by using trained observers and the brass instruments of which William James had such a low opinion. Introspection is no longer much used. Cognitive psychologists may see representations and may even argue that they are the only things that can be seen, but they do not claim to see themselves processing them. Instead, like psychoanalysts, who face the same problem with processes that cannot be seen because they are unconscious, they have turned to theory. Theories need confirmation, however, and for that many have turned to brain science, where processes may be said to be inspected rather than introspected. If the mind is “what the brain does,” the brain can be studied as any other organ is studied. Eventually, then, brain science should tell us what it means to construct a representation of reality, store a representation in memory, convert an intention into action, feel joy or sorrow, draw a logical conclusion, and so on.

But does the brain initiate behavior as the mind or self is said to do? The brain is part of the body, and what it does is part of what the body does. What the brain does is part of what must be explained. Where has the body-cum-brain come from, and why does it change in subtle ways from moment to moment? We cannot find answers to questions of that sort in the body-cum-brain itself, observed either introspectively or with the instruments and methods of physiology.

The behavior of the organism as a whole is the product of three types of variation and selection. The first, natural selection, is responsible for the evolution of the species and hence for species behavior. All types of variation and selection have certain faults, and one of them is especially critical for natural selection: It prepares a species only for a future that resembles the selecting past. Species behavior is effective only in a world that fairly closely resembles...