Many discussions of aggression emphasize only controlled actions in which the aggressors deliberately attempt to achieve some goal. Without minimizing the importance of these analyses, I will here present a sampling of my own research highlighting the determinants of impulsive, relatively uncontrolled, aggressive conduct. This research program was narrowly focused at the start, and was especially concerned with the effects of frustrations. But then, as time went by, my studies became much more general in nature and considered the effects of a variety of other aversive occurrences. The implicit theorizing governing these investigations was at first also fairly narrow, and was guided by a simplified version of the S-R associative learning concepts prevalent at the time. My analyses also broadened as the studies continued, making more use of cognitive notions. Then, as I came closer to retirement, I made increasing use of associative-network ideas.

Although my initial research publication appeared in 1953, I began my systematic attention to aggression in 1958, with an examination of the frustration–aggression hypothesis, principally as published in 1939 by Dollard, Doob, Miller, Mowrer, and Sears. These authors defined “frustration” as an “interference with the occurrence of an instigated goal response” (with this interference supposedly produced either externally or within the person). This blocking, according to Miller's later modification, produces an instigation to aggression along with other inclinations. If the perceived frustrating agent is injured, the Yale group held, the aggressive drive theoretically will be lessened, but will increase if the frustrater is not hurt for one reason or another.

In my early 1962 book on aggression, and in other discussions of this topic, I frequently spoke of the resulting aggressive drive as anger. I later went on to propose that the arousal of anger will not produce open aggression in the absence of aggressive cues – “stimuli associated with the present or previous anger instigators” in the external environment or in the mind. These cues presumably “pull” (evoke) aggressive responses from the angry person. However, this insistence on the necessity of aggression-associated stimuli is probably too strong a statement, and I now prefer to say that these cues only automatically intensify the ongoing aggressive drive.