Abstract
Travis Thompson’s lengthy review of Staddon’s The New Behaviorism requires several corrections and extensions. This response discusses Staddon’s analysis of Herrnstein’s matching law and concludes that Thompson misinterprets a gentle critique as a paean. The response goes on to defend the utility of models and “internal states” (i.e., postulated processes that are not directly measurable) as “formal representation[s] of the data reduced to a minimal number of terms,” a position similar to one of B. F. Skinner’s statements. The response ends with a defense of Skinner’s empirical brilliance, but a critique of his sweeping societal prescriptions.