||Comments by Pierre Karli
Laboatoire de Neurophysiologie, Centre de Neurochimie du CNRS, Strasbourg, France
Emotional responsiveness and relevant history of reinforcement are important determinants of social behavior. In keeping with the basic concepts of ethology, Adams's three hypothetical motivational systems are characterized by narrowly specific, rather rigid (with little change over time), linear, and unidirectional relations between sensory input and behavioral output. Since our data concerning the rat's mouse-killing behavior point to emotional responsiveness and prior social experience as essential determinants, we are led to imagine brain manipulations and to interpret the results obtained in a somewhat different perspective. More concretely, each time we elicit or facilitate, suppress or abolish a given behavior (attack, defense, submission, avoidance, escape, flight ) by means of some brain manipulation, the behavioral effect thus induced is not interpreted straight off in terms of altered functioning of a specific motivational mechanism. We rather try to take into account all the possible effects of the brain manipulation on the organism's level of overall responsiveness, on its capacity to integrate an affective significance with the sensory input so as to make the behavioral response coherent with previously shaped social-emotional adaptations, on the functioning and interactions of the systems of positive and negative reinforcement, on the bringing into play of these systems by some of the consequences of behavior that are anticipated or actually derived from it.
With regard to the existence of a rather rigid relation between specific motor-patterning mechanisms and a specific motivational mechanism, it might be pointed out that the mouse-killing response does not seem to be based on the same motivational state in the experienced killer-rat as when the rat is being presented with a mouse for the very first time (Karli et al 1974). The rather stereotyped and "cold-blooded" killing response displayed by the experienced killer-rat can be regarded as an appetitively motivated attack-behavior. When presented with a mouse for the first time(s), the rat mostly displays a killing- behavior that is rather of the "affective" kind - a behavior that can easily be elicited in the natural nonkiller by electric stimulation of medial hypothalamic and periaqueductal sites. The effective stimulation sites are in every case "switch-off" sites. Once the rat has learned to stop the brain stimulation, either by fleeing or by pressing a lever, it is much more difficult to induce it to kill a nearby mouse, This "affective" kind of mouse-killing might be considered a defense-behavior - a kind of active-avoidance behavior - killing being a way of putting an end to aversive experience. If the motivational state underlying the rat's mouse-killing behavior changes over time, some misinterpretations of experimental results are bound to occur, since initiation of mouse-killing in the natural nonkiller and abolition of killing behavior in the experienced killer-rat should no longer be regarded as two mirror-image processes.
It is both tempting and hazardous to try to specify precise and delimited locations for motivational mechanisms assumed to underlie behavior. A tempting project, since we are fully aware of the almost insuperable difficulties we would run against it we were to specify such mechanisms in terms of rather widely distributed and diffusely imbricated neuronal networks with complex reciprocal interactions. But also a hazardous project, since we are almost inevitably led to bestow upon a delimited brain structure or substructure an unduly global role with regard to the generation of a specific motivation. More concretely, let's consider the following two facts (1) an experimentally produced hyperreactivity (eg following a septal lesion) facilitates initiation of mouse-killing if and only if the rat did not previously develop a stable inhibition of interspecific aggression on the basis of repeated contacts with mice; (2) destruction of the corticomedial amygdala or interruption of the stria terminalis interferes with the development of such an inhibition on the basis of prior experience with mice (Karli et al 1977). In other words, whether or not a rat kills the mouse with which he is presented depends in an essential way on mechanisms in which the septum and the corticomedial amygdala are deeply implicated. Could it then be meaningful to search tor a "motivational mechanism" that would, for instance, be limited to the central gray and would thus comprise neither the septum nor the amygdala?
The search for narrowly specific mechanisms, in which the central gray or the medial hypothalamus may be deeply implicated, should not lead to overlooking the more general functional role played by the periventricular system in the generation of aversive experience, and of one of the two basic attitudes of the living organism towards its environment (i.e. retreat or escape, as opposed to approach) as well as in the negative reinforcement of approach behavior. As a matter of fact, the stimulation of a great many sites located in both the central gray and the medial hypothalamus induces one common effect: i.e, the rat readily learns to switch off such a stimulation by means of a variety of behavioral sequences. Furthermore, when two "switch-off" sites are simultaneously stimulated, the resulting escape speed corresponds most often to the sum of the escape speeds induced by stimulating either site alone (Schmitt and Karli 1979). With regard to this more general role of the periventricular system in motivational processes, there are complex interactions between central gray and medial hypothalamus on the one hand, and between these structures and the reward- approach system on the other hand. Central gray lesions depress hypothalamically-induced escape behavior (Schmitt, Paunovic, and Karli 1979), and one records in each of the two periventricular structures unit activities that are correlated with the escape speed induced by stimulating sites located in the same structure or in the other one (Sandner, Schmitt, and Karli 1979). On the other hand, central gray lesions provoke not only a decreased responsiveness to fear-inducing aversive stimulations and situations, but at the same time a general facilitation of various appetitively motivated behaviors (Chaurand, Vergnes, and Karli 1972; Karli et al 1974). It is clear that much further research is needed in order to better understand the part taken by these periventricular structures in general motivational processes and in more specific ones, respectively.