Annoyance, type and duration of postannoyance activity, and aggression: the "cathartic effect"

A series of experiments was conducted to elucidate the conditions conductive to a decrease in aggression following annoyance. The potential capacity of expression of aggression to bring about a reduction in the amount of subsequent aggression was of particular interest. This empirical concern was supplemented by tests of several influential and competing theoretical concepts dealing with the cathartic aspects of human aggressive behavior. Given the failure of such concepts to account for major portions of the data, an integrative theoretical model was proposed. experiment 1 evaluated the usefulness of the hydraulic, self-arousal, and dissipation of anger concepts in accounting for the earlier demonstrations of the cathartic effect. In a 2 x 3 x 2 design, half of the subjects were annoyed by a confederate, while the other half were treated neutrally. During the next stage (the interpolated period), a third of all subjects gave "shocks" to the confederate, another third simply waited, while the remaining third worked on mathematical problems. Orthogonal to the first two facotrs was the duration of the interpolated period (7 to 13 min). The main dependent measure was the number of shocks administered to the confederate in the final stage of the experiment. It was found that annoyed subjects gave more shocks than nonannoyed ones did, and that only the former were substantially affected by other manipulations. In the case of the annoyed wait and annoyed math subjects, the anger dissipation hypothesis correctly predicted that the mere passage of time would decrease the amount of subsequent aggression, presumably due to the action of homeostatic processes. The self-arousal hypothesis correctly predicted that the annoyed math subjects would give fewer shocks than the annoyed wait ones would. Since the subjects were engaged in an absorbing activity, the likelihood of their arousing themselves by ruminations about the preceding annoying incident was minimized, and the amount of subsequent aggression reduced. Yet, when annoyed subjects had given the confederate a moderate number of shocks in the interpolated period, they subsequently gave him fewer shocks than the 7-min annoyed wait and annoyed math subjects; this was the only outcome predicted correctly by the hydraulic model. In contrast, when a large number of shocks had been administered in the interpolated period, the amount of subsequent aggression was relatively high. The interpretation of the latter result in terms of an "adaption effect" was tested by further experiments.