Material-specific perceptual interference for visual words and faces: implications for models of capacity...

Four experiments were conducted to test the hypothesis that perceptual interference effects would be greatest among items that are processed by the same hemisphere than among items processed by different hemispheres. Recognition of two peripheral faces, flanked on either side of a central item that also had to be identified, was worse when the item was a face than when the item was a word. The opposite results were obtained for peripheral word identification by naming (Experiment 1). Identical results were obtained when a nonsense form was substituted for the central face (Experiment 2). Since unfamiliar faces and nonsense forms are processed primarily by the right hemisphere and words are processed by the left, the results are consistent with the hypothesis. The interference effect was the same even when the central nonsense shape was identified by naming, suggesting that the locus of the effect is not at response selection (Experiment 3). Ignoring the central item only reduces the interference effects but does not eliminate it (Experiment 4). Although in general words were perceived better in the right visual field, and faces were perceived better in the left, small fluctuations in the magnitude but not in the direction of perceptual asymmetries were sometimes noted with changes in the central item. The fluctuations were usually opposite to those predicted by attentional models of perceptual asymmetries.