6. A consociate modulator that switches the animal from defense to submission; organized perinatally by corticosteroids; activated by corticosteroids and inhibited by prolactin in adulthood.
I have suggested in previous work that muroid rodents have two related motivational mechanisms, defense and submission, and that there is a neural switch, called the consociate modulator, that suppresses defense and activates submission when the opponent is a familiar conspecific or consociate (Lehman and Adams, 1977; Adams, BBS 2(4) 1979). In most respects, including most of their motivating stimulus inputs and motor pattern outputs, defense and submission are similar. But defense is appropriate against a potential predator; it includes a damaging attack and may be seen in wild animals or in laboratory animals with £orebrain lesions which abolish submission. Submission, on the other hand, is appropriate to familiar consociate opponents; it includes submissive acts and postures that would not be respected by a potential predator, but that can inhibit offense by a conspecific opponent. The following motivating stimulus inputs are shared by both mechanisms: sudden visual movement and auditory stimuli, pain, restraint, dorsal tactile stimuli, defense pheromones, and the stimuli which activate neophobia. The following motor pattern outputs are also shared: sideways and upright posture, escape leaps and flight locomotion, urination and defecation, squeal or chit vocalization, and freezing crouch. The following motor patterns are not shared: lunge-and-bite attack, that is activated only by defense; full submissive posture and ultrasonic piping vocalization that are activated only by submission. The effects of hormones are not uniform across all of the motor patterns of defense and submission. For this reason, I have suggested that hormones do not act upon the motivational mechanisms themselves, but upon the consociate modulator that determines which of the two motivational mechanisms are operative in any given situation (figure 1, site 6). A facilitation of the consociate modulator would be expected to facilitate submission, including its motor patterns of full submissive posture and ultrasonic vocalization, and inhibit defense, including its motor pattern of lunge-and-bite attack. Inhibition of the consociate modulator would have the opposite effect; it would facilitate the lunge-and-bite attack and inhibit the full submissive posture and ultrasonic vocalization. Other .-ambivalent" motor patterns such as upright posture, flight, squealing, and crouching would be expected to appear under both conditions, although there might be some differences in the degree of their prepotency.
The effects of adrenal corticosteroids upon aggressive behavior may be understood as an activation of the consoc1ate modulator. In mice, administration of corticosteroids may facilitate passive avoidance of a previously victorious opponent (Moyer and Leshner, 1976) and increase the rate of submissive upright posture in response to attack (Nook and Leshner, 1976). The effect is not due to ACTH secretion, since there is no effect if ACTH levels are altered and corticosteroid levels are kept constant (Moyer and Leshner, 1976). And in a. situation that does not involve a conspecific opponent, the novel open field test situation, the suppression of pituitarv-adrenal activity by hypothalamic implants of hydrocortisone does not affect freezing behavior in rats (Stern et al, 1973).
The effect of prolactin upon aggressive behavior is the opposite of the effect of adrenal corticosteroids. Females are more aggressive when lactating, and especially likely to use the lunge-and-bite attack that indicates defense; this suggests that prolactin inhibits the consociate modulator. Whether confronted with a conspecific or a potential predator the lactating female shows not only offense behaviors, as mentioned earlier, but also the lunge-and-bite attack in deermice (Eisenberg, 1962), laboratory mice (Noirot et aI, 1975) and rats (Erskine et aI, 1978). Both males and non-lactating females may also show a lunge-and-bite attack against a potential predator such as a human handler (Kom and Moyer, 1968), but only a lactating female is likely to use such an attack against a familiar conspecific opponent. Although one experiment has reported decreased rates of upright posture and boxing by lactating females in response to footshock (Thoman et al, 1970), we have found in our laboratory that in such a. situation the lactating female is more likely to show a lunge-and-bite attack that produces freezing and escape in the opponent, and these are incompatible with upright posture and boxing. It is possible that androgen may have some effect upon the consociate modulator. When a male tat is subjected to electric shock and confined with a conspecific in a small cage, it is more likely than a female to show the upright posture and boxing (Conner and Levine, 1969b; Hutzell and Knutson, 1972; Thor et a1, 1974), an effect that is abolished by castration (Hutchinson et a1, 1965) and dependent upon neonatal androgen (Conner and Levine, 1969b; Baenninger, 1974). Since females are, if a~hing, more sensitive to shock than males (Beatty and Beatty, 1970), one cannot attribute the difference in behavior to differences in sensory thresholds or lack of motivating stimulus input. There are at least two possible explanations in terms of.the present model. First, androgen may facilitate the motor patterning mechanism of.the upright posture and boxing. Second, if one assumes that the upright posture is more strongly facilitated by defense than by submission, androgen might act to inhibit the consociate modulator. In situations that do not involve conspecific opponents, such as handling by the experimenter (Korn and Moyer, 1968) and many open field test situations (Archer, 1975) there are no measurable sex differences in defense; these data are consistent with both possible explanations. There are other data, however, that cast doubt on the role of androgen as an inhibitor of the consociate modulator. In mice, castration and androgen injections do not affect the submissive postures of animals defeated in fighting (Nock and Leshner, 1976). And in rats tested against familiar conspecific partners in a competitive fighting paradigm, we have found no difference s between males and females in their responses to a bite-and-kick attack by the opponent. Both males and females respond with upright posture about 40% of the time and with full submissive posture the other 60%. Before reaching any conclusion on the effects of androgen upon defense and submission, we need to have positive data from tests besides those of the highly artificial shock-induced fighting situation.
I have suggested elsewhere that the consociate modulator may be located in the ventromedial hypothalamus (Adams, BBS 2(4) 1979). Consistent with this is the great increase in the lunge-and-bite attack which is seen following ventromedial lesions, to be expected from the removal of an inhibitory influence upon the defense motivational mechanism. Hormone uptake data are also consistent with this. Androgen is taken up by the ventromedial hypothalamus, and corticosteroids are taken up in the septum which provides a. major input to the ventromedial. hypothalamus (Stumpf and Sar, 1976). As mentioned above, it is not yet known where prolactin acts upon the central nervous system.
Corticosteroids may have an organizing influence upon the neural substrate of the consociate modulator around the time of birth. Higher rates of submissive posture have been found in adult mice that received neonatal corticosteroid injections (Schwartz and Leshner, 1977). .