Attention shifts during matching-to-sample performance in pigeons

Two pigeons performed a matching-to-sample task in which a color, a line orientation, or both, might appear on the sample. Stimulus control by (attention to) the color and line components of a compound sample was measured when (a) either component might be relevant (i.e., appear as a comparison stimulus on that trial), and (b) when only one component was relevant. Degree of stimulus control by a given component of the compound sample was higher when only it was relevant than when either component might be relevant. Because the matching-to-sample task separated sample inspection from response choice, interpretation of the results in terms of response competition was ruled out. Because the training procedure led to shifts in stimulus control by the components of the compound sample, explanation of previous matching-to-sample results in terms of reduced perceptual clarity of compound samples was also ruled out. Thus, variations in stimulus control by components of the compound sample were interpreted as showing shifts in attention to the components of the compound sample.