Chapter IX - Stimulus Combination

Abstract

This chapter discusses stimulus combination in nonhuman primates. Experiments with animals have been concerned either with learning or with selective attention, and these studies have generally been presented in a few simple paradigms. The three stimulus combination experiments described in this chapter add to the data that are difficult to explain with Sutherland and Mackintosh's selective attention model. Most other data with monkeys likewise do not confirm the attention model predictions, so one possibility is that there are species differences in the model and another is that the selective attention model needs some modification before it can be applied to nonhuman primates. One of the experiments supporting the selective attention position is stimulus-compounding experiment. It was reasoned that the negative correlation between learning to discriminate orientation and brightness after confounded training was evidence of learning of switch-in analyzers. If a particular animal had a response bias and the model was sampled until finding an element corresponding to his expectancy, the response probability would be negatively correlated for the two sets of stimuli. Although there was no provision in sampling model for more than one pair of training stimuli and no mechanism for combining stimuli, it might be a bit more parsimonious to begin with a simple performance model than to postulate complex analyzers, which are switched in and out.