Much contemporary research in the field of animal learning is characterized by a shift away from extremely behavioristic, response-centered approaches. If the material in this book provides any yardstick, the use of abstractions and inferences that go well beyond simple relationships among observables is more frequent and acceptable now than it was 10 or 20 years ago. This swing of the pendulum toward more flexible interpretations of animal learning seems to have its roots in the belief that highly descriptive or associationistic approaches, which scrupulously avoid the postulation of mechanisms intervening between environmental stimulus and observable response, have not provided us with many truly significant or novel insights and are unlikely to do so in the future.