Submission differs from defense on a number of levels. In terms of its motor patterns, submission includes the full submissive posture and ultrasound (in rats), while defense includes a lunge-and-bite attack and striking (in cats). Other motor patterns are similar for defense and submission. In terms of motivating stimuli, submission is shown against an animal's consociates (i.e. familiar conspecifics or other animals with whom it has lived), and defense is shown against other animals. In terms of life history determinants, defense is shown primarily by wild animals, and submission is shown primarily by tame or laboratory animals. One may assume that the process of taming wild animals consists, in part, in shifting the dominant behavior from defense to submission. Functionally, defense is often damaging, while submission is not. The defensive animal is a dangerous one, likely to inflict a bite on the face of its opponent (or the hand of an experimenter), while the submissive animal is a vulnerable one, likely to show a full submissive posture that might inhibit conspecific offense but would not be expected to be effective against a predator. Phylogenetically, defense probably evolved first to deal with predators, and the submission system evolved later to modify defense behavior when the animal was confronted with a conspecific whose offense behavior could be inhibited by particular submissive postures. Finally, as will be reviewed below, defense and submission depend upon different, although parallel, neural substrates. In particular, lesions of the amygdala may enhance submission, while lesions of the septum and ventromedial hypothalamus enhance defense.
Most of the motor patterns and motivating inputs of defense and submission are similar. This suggests that their respective motivational mechanisms consist of sets of homogeneous neurons with very similar neural connections, and that they may have the same locus, submission having evolved as a subset of defense during the course of phylogeny. As in the case of defense, the motor patterns of submission appear to be organized in a graded hierarchical series depending upon intensity of activation of the submission motivational mechanism and differential thresholds of activation of the various motor patterning mechanisms. Since most of the motor patterning mechanisms of the two systems may be shared, their differential thresholds would be expected to be the same for both systems.
I suggest that there is a mechanism that switches the behavior of an animal from defense to submission in the presence of a consociate, or familiar individual. Such a mechanism may be called a "consociate modulator." In a previous publication (Adams 1977) I suggested that it should be called a "conspecific defense modulator," but upon further reflection I think that the term should be "consociate modulator" because there are old experiments that show that behavior towards individuals is modified by familiarity, even if they are not members of the same species; for example, cats can be made to live peacefully with rats rather than treating them as prey (Kuo 1930). And I have dropped the term "defense," because such a mechanism may ultimately turn out to modulate other behaviors besides defense, such as feeding and sex.
The consociate modulator, according to the present formulation, is a set of homogeneous neurons (similar to those of motivational mechanisms) that is activated by stimuli associated specifically with consociate animals, and that, in turn, facilitates the submission motivational mechanism and inhibits the defense motivational mechanism. For rats, the stimuli that activate the consociate modulator are probably olfactory; in cats the cues may come from other sensory modalities as well.
A number of lines of evidence, none of them compelling when considered alone, have suggested to me that the ventromedial nucleus of the hypothalamus may be the site of the hypothetical consociate modulator. These lines of evidence include: a reanalysis of the results of single-neuron recording from the ventromedial hypothalamus during affective defense in the cat (Adams 1968); consideration of data on the neuronal circuitry and the inputs of the ventromedial hypothalamus; an attempt to explain why fleeing and affective defense are obtained from different (although neighboring) zones with electrical stimulation of the forebrain; and speculations on the potential interactions of defense and submission in the forebrain, which might explain why forebrain lesions can release defense.
A reanalysis of single-unit recording results from the ventromedial nucleus of the cat during affective defense suggests that its neurons may be active during submission rather than during defense. In my original study I was quite surprised to find that neurons of this nucleus were not active during affective defense, despite the general belief expressed in the literature that cells in this region should be involved in aggressive behaviors (Adams 1968). One neuron recorded in the capsule of the ventromedial nucleus acted in a way that I could not interpret at the time. It fired at high rates if and only if the animal cowered or tried to escape from its opponent. This neuron was not reported in the publication, however, because I could not record from it during affective defense, which was, after all, the purpose of the study. In what at the time seemed an unfortunate occurrence, the cat refused to show affective defense during testing of the neuron, and instead simply cowered and tried to flee. This event reflected a bias I had built into the design of the experiment: it focused on defense and not submission. Those cats (about 50%) in the colony who showed cowering and flight (i.e. submission) were not used for the experiment; only those who showed consistent affective defense were implanted with the recording electrodes.
(Section continued on next page)