International Journal of Comparative Psychology

Interference between cues is generally regarded as different from interference between outcomes in the (animal and human) predictive learning literature. In consequence, current theories of learning generally attempt to explain one or the other effect but not both. In general, cues are assumed to interfere with each other only if they are trained in compound as predictors of the same outcome, whereas outcomes are assumed to interfere with each other only if they have been individually paired to the same cue. In two experiments with humans, we examined the parallel between interference between elementally-trained cues and between elementally-trained outcomes, using a behavioral preparation. Experiment 1 showed that both interference effects are similarly affected by identical contextual manipulations. Experiment 2 showed that the two effects take place when the interfering association shares an element with the target association: When the shared element is the outcome, interference between cues takes places; when the shared element is the cue, interference between outcomes occurs. These results add to the growing body of evidence that calls for the integrative study of interference between cues and between outcomes in predictive learning situations.