Occasion Setting, Disjunctive Problem Structures, and the Art of Rationalizing Mistakes
Abstract
I endorse the efforts proposed by Leising et al. (2025) to bridge terminological and conceptual gaps within and across disciplines. Occasion setting may indeed represent one of the most universally studied problems in human and nonhuman learning, occurring whenever a learned contingency between two variables depends on the status of a third (explicit or latent) variable. I argue that identifying the (partial) “disjunctive structure” and stimulus representations fundamental to occasion setting allows for recognizing a broader range of relevant tasks and phenomena of theoretical interest in human category learning, operant conditioning, and related fields. This perspective has potential implications for theoretical concepts of error-driven reinforcement learning and may inform investigations into how humans reason about occasions when learned stimulus–outcome contingencies are reinforced or nonreinforced. Such insights could enhance our understanding of behavioral adaptation on a broader scale (e.g., the cognitive processes underlying lying, or rationalization of errors).
Keywords: contextual modulation, contingency reversal, rule abstraction, category learning, Gamblers’ fallacy