Abstract
Theories of generalization distinguish between elemental and configural stimulus processing, depending on whether stimulus in a compound are processed independently or as distinct entities. Evidence for elemental processing comes from findings of summation in animals, whereas configural processing is supported by experiments that fail to find this effect when similar stimuli are employed. In humans, by contrast, summation is robust and independent of similarity. We show how these results are best explained by an alternative view in which generalization comes about from a visual search process in which subjects process the most predictive or salient stimulus in a compound. We offer empirical support for this theory in three human experiments on causal learning and formalize a new elemental visual search model based on reinforcement learning principles which can capture the present and previous data on generalization, bridging two different research areas in psychology into a unitary framework.