Hemispheric superiority for recognizing faces depends upon how they are learned

Abstract

To investigate whether the type of information available influences encoding strategy, and subsequent hemispheric advantage, for face recognition, Experiment I presented photographs of faces for initial viewing centrally, under two conditions: verbal presentation of (a) molar personality or (b) physical feature information about each target face. The recognition task presented targets and distractors tachistoscopically to each visual field. Recognition accuracy scores displayed a highly significant pattern of contrasting hemispheric superiorities under the two informational conditions, but the direction of this contrast was not consistent for all subjects. The first part of Experiment II replicated the procedures of Experiment I, except that all verbal encoding information was omitted. Under these control conditions, the "crossover" pattern of hemispheric superiorities did not emerge, and subjects instead displayed consistent LH or RH superiorities. A second experimental session demonstrated the reliability and generality of the crossover phenomenon: When the same subjects were recalled and the original encoding manipulation applied to a new set of faces, the pattern of contrasting hemispheric superiorities again emerged. No aspect of subjects' performance in the control condition was found to predict the direction of crossover in the subsequent experimental session. We infer that intra-subject contrasts in hemispheric superiority reflect the use of alternative encoding or processing strategies under the two informational conditions, but the relation of such strategies to specific hemispheric advantages remains unclear. A general statement of the relation of cognitive processing, task demand characteristics, and observed hemispheric advantage is proposed.