Catania's Concept of the Operant Revisited: Empirical Analysis of Response Variation and Controlling Contingencies

Abstract

Six Shaver‐Starcross hens with no prior experimental experience served as participants in two experiments which were designed to empirically test Catania's concept of the operant. The stimulus consisted of a cream coloured rectangle against a black background shown on a standard LCD computer monitor. An infrared touch screen mounted to the front of the LCD monitor recorded the location of all responses made by the participants. Experiment one consisted of autoshaping the rectangle pecking response. This was done using an automated computer program designed to eliminate latencies between responses and reinforcement, as well as positional biases which may have been introduced via manual autoshaping. The program successfully eliminated latencies and positional biases introduced by the experimenter, but took longer than anticipated to autoshape the desired response in the participants. It is suggested that procedural differences account for the unusual length of time taken to autoshape the participants in this experiment. Preliminary inquiry investigating procedural differences shows that it may be possible to model speed of acquisition more accurately than done so at present, without retrospective analysis of the acquisition data itself. In experiment two the active, reinforced zone of the rectangle was reduced. The participants had no visible cue demarcating active and inactive zones of the rectangle, yet in accordance with Catania's operant, responding across all participants came to fall within the active, reinforced zone of the rectangle. The results; shifts in response distributions in relation to the changes in contingencies, offer empirical support for Catania's operant in terms of positive reinforcement across a single parameter of responding. The implications of these findings imply that when variability in responding is important; for example in learning, exploring, creating, and problem solving, Catania's operant may be favourable over Skinner's operant as a vehicle for identifying and controlling variables associated with behavioural outcomes due to its greater topographical inclusivity. Contemporary behavioural analysis favours a behavioural systems approach where the respondent and operant class distinction merges; the environment affects the organism as much as the organism affects the environment. Catania's operant is complementary to behavioural analysis in this vein.