Enduring problems for molecular accounts of operant behavior.

Although the Associative Learner (AL) theory advocated by Shimp, Childers, and Hightower (1990) can simulate several features of human and nonhuman performance in free-operant situations, it fails to account for many of the findings that are theoretically most powerful. It underestimates the sensitivity of behavior to relative reinforcement rate and totally omits any role for incentive variables such as amount and delay of reinforcement. At present, it is not sufficiently comprehensive to serve as an adequate alternative to Herrnstein's (1961, 1970) matching theory as a general account of behavior.