Backward masking of lateralized faces by noise, pattern, and spatial frequency

Experiments 1 and 2 measured the critical interstimulus interval at which a face presented to the right or left visual field escaped a trailing noise, pattern, or spatial-frequency mask. The function relating target duration to critical ISI was multiplicative in the noise and spatial-frequency condition, but additive at longer durations in the pattern mask condition. An advantage of about 8 msec for the left visual field and 2 msec for the right field was found in the pattern and spatial-frequency masking condition, respectively. No consistent visual field differences were found in the noise mask condition. Taken together, these results suggest that hemispheric difference in face recognition are either absent or inconsistent at early, peripheral, energy-sensitive stages of processing, but emerge strongly at higher order central stages. The results also suggest that the left and right hemispheres are not differentially sensitive to the output of high- and low-spatial-frequency channels, respectively. If it is assumed that the central face processor is functionally localized to the right hemisphere, one can infer from these results that interhemispheric transmission time is not greater than 8 msec, and the output of sensory analysis and/or relational features are transferred across the interhemispheric commissures.