Abstract:
The present experiment attempted to address the local and global contingencies which control responding in a choice procedure. An alternation procedure (similar to Williams’, 1971a; 1971b experiments) was implemented to observe how six pigeons responded, given the predictions made by the formal definition of ‘reinforcement’. These strictly alternating conditions were compared with randomly alternating conditions, which acted as a control in order to answer the question of whether the results found by Williams could be at least partially explained by the effect of the overall distribution of reinforcers. The incompatible predictions made by theories of local control and global control made it possible to directly ascertain what variables were affecting the behaviour of the subjects. For example, if the subjects were responding to local contingencies, then at points in the experiment where the local probability of a reinforcer being found at a given location was 0, the responses should be zero – however, if they were responding to global contingencies, then we would expect the distribution of responses to reflect the overall probability of a reinforcer being found at that location. These effects were measured using preference pulses and although there was some evidence of local contingencies having an influence on behaviour, it was concluded that the longer-term aggregations of reinforcers was the factor that largely controlled the distribution of behaviour. The research also extends the work of Krägeloh et al. (2005) by examining the ramifications of the local effects of sequences of reinforcers on behaviour using a steady-state procedure instead of a frequently-changing procedure (Davison & Baum, 2002). This was done to investigate whether the findings were consistent across different experimental paradigms.