Psychology as a distinct discipline began with psychophysics, which was once understood as relating events of the physical world to a presumably distinct mental one constituted of sensations and perceptions. In behavior analysis, of course, sensation and perception are replaced by sensitivity and discrimination. Despite this fundamental difference, the evolving concerns whereby psychophysics has become sensory psychology have methodological and conceptual relevance for behavior analysis.

As characterized in Blough's review of Stebbins' edited book, Animal Psychophysics: The Design and Conduct of Sensory Experiments,animal psychophysics originated as a relatively straightforward application of behavioral techniques to the concerns of another field. The shared strategy of intensively studying individual subjects, in contrast to the more common emphasis on statistical group designs in psychology, simplified the merger. As Blough shows, however, the procedural focus on analyzing sensory capacities inevitably shades over into conceptual issues. Sensory effects can be assessed only if control by discriminative stimuli is adequately separated from control by the reinforcement contingencies, and especially from control by discriminative properties of the contingencies themselves. Thus, successful animal psychophysics involves not only technical sophistication but also conceptual engagement with the subtle contingencies embedded in the experimental procedures.

The problem of separating discriminative relations in behavior from effects of the consequences of the behavior is easy to overlook in experiments that involve instructed human behavior. The potential confounding is always there, however, and went unnoticed for scores of years in sensory psychology. The key insight of signal detection theorists was to recognize that sensory thresholds are potentially affected by the consequences of reporting the presence or absence of a stimulus. Signal detection theory offered a solution to the problem, construing it in terms of distributions of presumed internal events (signal plus noise versus noise alone) judged according to the observer's criteria for reporting or not reporting them.

Nevin's review of Green and Swets' Signal Detection Theory and Psychophysicspoints out that the events treated in signal detection theory can be directly measurable distributions of environmental events rather than inferred internal events. Nevin's observation not only enables a more tractable handling of the distribution of discriminated events; it also provides for detailed exploration of the consequences of detection, in the context of dynamic contingencies rather than the constrained discrete-trial payoff matrix of classic signal detection theory. His review offers counterintuitive parallels between reinforcement variables in the schedule control of behavior and more standard psychophysical relations. This seminal essay led to Davison and Tustin's (1978) integration of signal detection theory with the generalized matching law (Baum, 1974), which had been derived from studies of choice. The result is a thoroughgoing behavioral detection theory synthesizing quantitative relations that originated in disparate scientific domains.

The reviews by Killeen and by Staddon link behavior analysis with more traditional issues in sensory psychology. In reviewing Stevens' Philosophy and Psychophysics,Killeen addresses the work of a classic psychophysicist who embraced the measurement of sensation at face value. On the one hand, as Killeen notes, Stevens' philosophy illustrates the methodological behaviorism that has been rejected by most behavior analysts. On the other, Stevens' emphasis on data as structuring theory, with attendant importance given to invariances in data, has close affinity to behavior analytic research strategies (e.g. Sidman, 1960).

Staddon's review of both von Békésy's Sensory Inhibitionand Ratliff's Mach Bands: Quantitative Studies on Neural Networks in the Retina focuses on the controversial concept of inhibition (cf. Donahoe & Palmer, Part III). Theorists who emphasize parsimony, as behavior analysts do, are chary of a concept whose main function is to account for non-occurrence. Such is inhibition, initially imported from neurophysiology and often invoked to account for the absence of behavior. As Staddon argues, however, several consistently verified properties of operant behavior justify the term, especially in the domain of stimulus discrimination. Appealing to potential similarities between steady-state receptor systems and steady-state operant behavior, Staddon postulates active tendencies opposing behavior corresponding to each term of the three-term contingency. The robustness of the tripolar relation between antecedents, behavior and consequences has become increasingly apparent; it has withstood attempts at fractionation into component pairwise relations (Rescorla, 1994), and it has been incorporated into extended networks of higher-order conditionality (e.g. Sidman, 1986). Staddon's characterization takes the operant beyond the implicit passivity of primitive S-R formulations, but also suggests a Newtonian, ballistic character for the operant (e.g., Killeen, 1992) that contrasts with the Darwinian, selectionist conception (cf. Palmer & Donahoe, 1992). Thus, the status of inhibition continues as an unsettled issue in behavior theory.

As discussed in Part IV and elsewhere, one issue of contention between behavior analysis and many other psychological viewpoints arises from the broadly held assumption that adequate explanations of behavior must appeal to mediating events underlying that behavior. Behavior analysis is not alone in its minority position on this issue, however, for even the occasional cognitivist has challenged the pervasive assumption (Watkins, 1990). The review by Costall of Gibson's The Ecological Approach to Visual Perceptionoutlines the same controversy as it has occurred among perception theorists. Costall illustrates how pervasive the mediational assumption is and makes a case that the notion of internal representation is a vulnerable manifestation of that assumption. Representations are posited as obviously necessary not only by most perception theorists but also by most theorists of human thinking and remembering. Costall identifies J. J. Gibson as a powerful ally in the struggle to unmask and challenge the troublesome assumption, for Gibson mounted an explicit frontal attack on representation as an explanatory concept, thus battering a keystone of the edifice that shelters the most explicit adherents of mediational theory.

To be sure, Gibsonian and behavior analytic approaches differ in significant ways. A hint of autonomous agency lingers within Gibson's characterization of the perceiver, whereas for behavior analysts the perceiver is a more neutral locus of interaction. Also, Gibson's theory more heavily emphasizes the match between the perceiving organism and its environmental niche. Nonetheless, Gibson is not the ally of behavior analysis merely in sharing a common opponent; rather, Gibsonian theory and behavior analysis have fundamental points in common. Costall sketches the complementarities of the positions (see also the concise introduction to Gibsonian theory in Michaels & Carello, 1981), placing both in the context of the controversies that have surrounded them.