Category-related components in visual evoked potentials: photographs of faces, persons, flowers and tools as stimuli.

One hundred and sixty black-and-white photographs of faces, persons, tools and flowers constituted the four stimulus categories in the first experimental paradigm. In the second paradigm, 160 black-and-white photographs of persons in a neutral (side view), fighting, greeting and "depressed" postures were used. The same nine female and nine male students volunteered as subjects in both paradigms. EEG responses were recorded through the electrodes F3, F4, Cz, Oz, T5 and T6 and referenced to "linked" mastoid electrodes. Individually averaged EPs and grand averages of the EPs for all 18 subjects were computed as well as bipolar responses for selected pairs of electrodes. Category-related components in the evoked potentials (EPs) were found mainly in the recordings through the midline and the frontal electrodes. These category-specific differences were more prominent, however, in the bipolar recordings between these electrodes and the respective ipsilateral electrodes (T5 or T6) than in the monopolar recordings. Monopolar and bipolar EPs could be divided into those evoked by person stimuli and by non-person stimuli. Category-related differences were demonstrated by computing EP-difference curves related to pairs of stimulus categories. In addition to the statistical analysis of peak amplitudes and peak latencies of the individual curves, a "running" Wilcoxon test was applied and cross-correlation functions between the EPs evoked by different stimulus categories were computed. The "face-related" components of the EPs described in earlier studies could be confirmed. They were prominent in the waves peaking between 130 and 150 ms and between 200 and 230 ms and were stronger when faces rather than full-figure "neutral persons" were used as stimuli. No significant hemispheric differences could be found between any of the EPs. In the second paradigm, EEG responses evoked by person stimuli in a neutral, fighting, greeting or depressed posture exhibited only slight differences, if at all. The "person-responsive" components in the EPs evoked by the photographs of a neutral standing person (side view) were smaller between 120 and 200 ms than in the other three categories in which "expressive" persons were the stimuli.