Association for Behavior Analysis International

Most of the concepts we use today to describe behavioral processes were the product of a mutually beneficial relationship in which Keller and Schoenfeld (Principles of Psychology) used Skinner's treatment of his research findings (The Behavior of Organisms) as the foundation for a systematic account of the science of psychology. In greater or lesser degree, this was the history of such concepts as operant conditioning, reinforcement and extinction, response induction and differentiation, successive approximation, shaping, stimulus generalization and discrimination, fading, conditioned reinforcement, chaining, escape, avoidance, and punishment.

Skinner not only provided us with a rich empirical and implied legacy, butalso a fertile conceptual one. One arena he shed his conceptual light upon is education. Through Skinner we've come to know that student learning is not accidental--it is the direct result of what has been designed, intentionally or unintentionally, by educators, instructional designers and society. Through Skinner we've learned that behavior is orderly, order emerges in single subjects, the "subject is always right," understanding behavior requires analyzing contingencies in fine detail, a complex repertoire can be synthesized from elementary units by the careful arrangement of contingencies of reinforcement, units of behavior should be determined empirically rather than defined in advance, and other important principles. One hundred years after Skinner's birth, by standing on his shoulders and those of other giants, Headsprout has updated the teaching machine to become a true, live contingency-matrix, one that has the potential to teach millions of children to read.

Of his many conceptual contributions, I will focus on Skinner's use of a small number of behavioral concepts and principles (respondent pairing, operant reinforcement and extinction, discriminative stimulus control, etc.) as a basis for a molecular interpretation of any conceivable event involving behavior. His analysis of the results of complex animal experiments in these terms illustrates the power of these few basic relations, and constitutes strong support for their extension to the interpretation of any form of human behavior. I will describe some outstanding examples of Skinner's interpretive approach and suggest its general importance for the field of behavior analysis.

Skinner partitioned the science of behavior into two inter-related enterprises-experimental analysis and interpretation. Experimental analysis is restricted to those situations in which all behaviorally relevant variables are manipulable, controllable, or measurable. Interpretation is the explanation of behavior that occurs in situations that do not meet the demands of experimental analysis. Such explanations must appeal exclusively to functional relations that have been identified in prior experimental analyses if they are to qualify as interpretations. The analysis-interpretation distinction is not peculiar to the science of behavior, but inheres in science generally. According to this distinction, many worthy behavioral studies do not satisfy the criteria of experimental analysis (e.g., the study of choice by means of concurrent schedules). Further, nearly all human behavior is the province of interpretation, not experimental analysis. The failure to honor the analysis-interpretation distinction has greatly impeded progress in the science of behavior, most especially in normative psychology.