A preliminary attempt is made to analyze the intraspecific aggressive behavior of mammals in terms of specific neural circuitry. The results of stimulation, lesion, and recording studies of aggressive behavior in cats and rats are reviewed and analyzed in terms of three hypothetical motivational systems: offense, defense, and submission. A critical distinction, derived from ethological theory, is made between motivating stimuli that simultaneously activate functional groupings of motor patterning mechanisms, and releasing and directing stimuli that are necessary for the activation of discrete motor patterning mechanisms. It is suggested that motivating stimuli activate pathways that converge upon sets of homogeneous neurons, called motivational mechanisms, whose activity determines the motivational state of the animal.

A defense motivational mechanism is hypothesized to be located in the midbrain central gray. In addition to tactile, auditory, and visual inputs from the paleospinothalamic tract, lateral lemniscus, and (perhaps) from the pretectum, it may receive inputs from a major forebrain pathway whose functional significance is not yet understood.

A submission motivational mechanism is also thought to be located in the central gray. In addition to inputs for defense, it is thought to receive a necessary input from a “consociate (social familiarity cue) modulator” located in the ventromedial hypothalamus, which can switch behavior from defense to submission. The location of the hypothetical offense motivational mechanism is not known, although the pathways by which it is activated are traced in some detail.

Brain mechanisms of aggression in primitive mammals and in primates are apparently similar to those in rats and cats.

This study involved a partial replication of an experiment performed by Held & Rekosh (1963) on the relation between prism-induced changes in visuomotor feedback and shifts in the perception of straightness. The Held & Rekosh study is important because its methodology makes it possible to clearly separate changes in perception due to visuomotor feedback from those due to other perceptual factors. Thus it bears unambiguously on the hypothesis that the neurological processing of sensory input is dependent on the organization of motor activity. The present paper consists of a brief review of theory and experiment on sensorimotor phenomena in perception, followed by a detailed consideration of the Held & Rekosh study and the present replication. For greater effectiveness, Held & Rekosh should have used base-up and base-down prism orientations rather than the base-left and base-right orientations that they employed, but in any case our replication yielded negative results for both base-up and base-right. It is concluded that, given the minimally effective prism orientation and the attentional loads characterizing the Held & Rekosh study, there is need for further experimentation with these two variables. A more drastic suggestion is that there may be a need to confront anew the evidence for sensorimotor processes in perception.