Sample and comparison location as factors in matching acquisition, transfer, and acquired equivalence.

Pigeons learned symbolic matching with samples appearing equally often on left and right keys. For a location-relevant group, the reinforced comparison choice for each sample reversed across sample locations; for a location-irrelevant group, the reinforced choices were the same. Consistent with the hypothesis that samples at different locations are functionally different for pigeons, Experiment 1 showed that matching acquisition was comparable in these two groups. Nevertheless, the location-irrelevant group eventually ignored sample location, given that their performances subsequently transferred to a novel (center-key) sample location. This transfer was not simply due to sample familiarity at different training locations; rather, it required that left- and right-key samples occasion the same reinforced choices in training. Acquired equivalence between those samples was then assessed in Experiment 2. The location-irrelevant group showed the predicted equivalence effects, but the location-relevant group did not--in fact, its results were the opposite of those predicted by equivalence. Their results indicate that the functional comparison stimuli are also defined in terms of their locations.