Spatial overshadowing in pigeons: Evidence for an acquisition deficit

Abstract

We investigated theoretical accounts of spatial overshadowing using a landmark-based spatialsearch task in a touchscreen preparation with pigeons. Pigeons first learned to find a hidden target on a screen using a compound of two visual cues as landmarks. Landmark A was proximal to the target while landmark X was distal to the target. Experiment 1 replicated our prior spatial overshadowing effect whereby landmark A overshadowed the development of spatial control by X. Spatial control by X was also poorer than by landmark Y which had been paired with the target alone but with the same absolute distance to the target as X had. Thus, the poor spatial control by X was not merely due to the greater X-target distance (relative to the A-target distance). Experiments 2a and 2b failed to find recovery from spatial overshadowing of X through either post-training extinction or counterconditioning of overshadowing landmark A, respectively. We interpret our results as being consistent with acquisition-focused models of elementary associative learning, but not with performance-focused models.