been found that after exposure to aggressive erotica male aggression against females is at a higher level than after exposure to similarly arousing but nonaggressive erotica. No such difference was observed for intermale aggression, however.
In order to determine whether or not aggressive cues were significantly involved in the hedonic-valence effect obtained in the present investigation, a follow-up experiment was conducted ( Zillmann, Bryant, and Carveth, in press). An "aggressive" erotic film featuring sadomasochistic behavior was successfully matched on both excitatory potential and negative hedonic valence with a film featuring bestiality (with cooperative animals). Compared against a no-exposure control, both films increased motivated aggression substantially, but to virtually identical degrees. Consistent with Donnerstein's observations, then, aggressiveness in highly arousing erotic films seems of little consequence for intermale aggression. [This assessment excludes rapists. There is reason to expect that "aggressive" erotica might further aggressive action in this subpopulation (cf. Abel, Barlow, Blanchard, and Guild, 1977).] Independent of general implications, however, the results of the follow-up experiment give no support to the aggressive-cue hypothesis and thus rule it out as an alternative explanation to the present findings.
The follow-up experiment also included a stimulus condition low in excitatory potential and very high in positive hedonic valence. This communication was composed of the photographs that had to be dismissed as too pleasant (i.e., beautiful female nudes in sexually provocative poses). There was a tendency for exposure to these "hyperpleasant" stimuli to reduce intermale aggression (relative to the no-exposure control), but this tendency again failed to be reliable statistically. It appears, then, that the earlier reported failure to observe a reduction of retaliatory behavior after exposure to pleasant, nonarousing erotica was not due to the stimuli's insufficient pleasantness, but may indeed have resulted from the severity of the provocation treatment. Taken together, these failures to replicate can be interpreted as suggesting that the aggression-reducing effect of the erotic fare in question is limited to minor and moderate provocations. This suggestion, needless to say, constitutes a post-hoc account of the findings, and it remains to be determined, through experimentation, whether or not the effect of these erotica is, as projected, a function of the magnitude of provocation and annoyance.
*This investigation was supported by Grant BNS77-07194 from the National Science Foundation to Dolf Zillmann. Paul Cominsky was at the University of Massachusetts and Norman Medoff at Indiana University when the study was completed.