||Comments by Paul Leyhausen
Max-Planck-Insitut fur Verhaltensphysiologie, 5600 Wuppertal 1, Federal Republic of Germany
Are neurophysiological techniques adequate to account for agonistic behavior? David B. Adams's attempt to review and organize the available data concerning the brain mechanisms of agonistic behavior is admirable for its consistency and its mastery of methods and material. Since the data are still far from complete and the gaps in our knowledge are all too apparent, no one could at this stage succeed in creating an overall picture to everyone's satisfaction. Hence, I will simply state my general agreement with most of Adams's efforts and briefly outline my main reservations.
These concern three points (1) an apparent bias, (2) an apparent overconfidence in the adequacy of neurophysiological methods and techniques, and (3) a misapprehension of some ethological views and behavioral phenomena.
1. Adams appears to be one of the last believers in the idea that a rigid stimulus-response hypothesis provides an adequate explanation for all behavior. Thus, since there is motivation, there must be motivating stimuli, and there cannot be stimulus-independent motivation. This leads to the demand for unitary motivational mechanisms consisting of "sets of homogeneous neurons." For reasons given later, I doubt the existence of such homogeneous motivational mechanisms. Moreover, while no one doubts that stimulation can bring about changes in the motivational state of animals, there is equally no doubt that internal, stimulus-independent changes of that state are continuously going on, and that the internal, autonomous generation of motivation, variously named "drive," "instinct," or "propensity," is not only equally important but constantly "setting" the peripheral and central afferent mechanisms concerning which kinds of stimuli are to be reacted to at any given time, and how, and which stimuli are to be disregarded. A stimulus is not a stimulus unless the internal motivational mechanisms make it one.
Consequently, I cannot agree with the distinction made between the motivating and the releasing function of stimuli. Nor does it make any difference whether a stimulus releases a motor pattern or its directional mechanisms (taxes), which are also motor mechanisms. The function of the stimulus is merely to touch off and sustain this motor apparatus, which in turn does all the orienting. Therefore, strictly speaking, there is no such thing as a directing or orienting stimulus. In conclusion. and in accordance with N Tinbergen, who first made this plain, there are releasing (or eliciting), supporting, and consummatory (switch-off) stimuli, and that is all.
2. Adams states: "Despite the fact that electrical stimulation of the hypothalamic ventromedial nucleus produced affective defense, the neurons of that nucleus are not active during affective defense elicited in a seminatural situation. Puzzled, I came to the conclusion that our behavioral control was not adequate to the demands of of neurophysiological techniques" (commentator's emphasis). Apart from finding the second sentence somewhat cryptic, I should like to reverse what I take it to mean: Neurophysiological techniques certainly are still far from adequate to the control of behavior as exercised by the organism itself, notwithstanding all modern refinements. Even a micro-electrode is too clumsy a tool, compared with the cellular mechanisms with which it rather brutally interferes. And however tiny the electrodes used may be, they are not even micro-electrodes in most electrostimulation experiments, and certainly not in the type of experiment Adams is referring to here.
Furthermore, all of the more complex behavioral patterns that may be elicited by electrical brain stimulation can be elicited from a wide rage of loci, usually extending from the amygdala down to the midbrain and even the medulla. This is in keeping with the wide variety of internal and external factors that contribute, separately or jointly, toward the manifestation of the behavior in question. To assume that in any given case all should always be active seems to me quite out of the question. It would be most interesting to provide all the loci from which, say, attack can be elicited with electrodes in the same experimental animal and then vicariously use one for stimulation and all others for recording. To my knowledge, this has never been done so far; but my prediction would be that the number and kind of loci from which recordings could be obtained would vary greatly with the site chosen for stimulation, stimulation intensity and duration, and the interval between stimulations.
The systems we try to study are of a rather complex nature and certainly farther from being unitary or homogeneous than Adams seems to allow. Much as the stimulation experiment is by far the most useful tool at present available for such investigations, we must not hope that a point-by-point stimulation procedure, as originally followed by W.R. Hess and more or less by all his successors, can provide us with a true picture of all the organisational complexities involved in behavior such as defense or rival-fighting. Even less may we expect this from the results of lesion experiments, which are notoriously misleading when used to interpret the neuronal basis of complex behavioral patterns. Here. I feel, it would have been more valuable if Adams had attempted to trace clearly the still extant gaps rather than to depict a fairly complete system.
3. When a cut-off tail end of an earthworm is attached to its former front end with a piece of thread, it will follow the front with well-coordinated creeping movements. Thus was it demonstrated that the locomotion of an earthworm is a chain reflex. Later, when E. von Holst (1932; 1933) treated the middle of an earthworm with an acid that ate away all the tissue except the nerve fibers, that preparation also showed well-coordinated locomotion in both front and tail, although the nerve fibers were too tender to exert any pull on the tail end; thus it was demonstrated that earthworm locomotion is controlled by a central nervous automatism The whole case highlights the danger of thinking in alternatives where biological systems are concerned.
Likewise, agonistic behavior is produced by both endogenous and exogenous factors in parallel, the endogenous producing the generalized or "ideal" pattern, the exogenous modulating and adapting it to suit the irregularities of the here-and-now conditions of the actual situation. This is one reason why it is not possible to subdivide agonistic behavior into rigidIy closed subsystems of offense, defense, submission and flight. The other is that these subsystems have a varying number of behavioral elements in common. In fact, the subsystems of offense (including .'aggression") and defense have most of them in common, regardless of whether they are being directed toward a conspecific or heterospecific adversary or toward a prey animal. The difference lies not so much in the few elements specific to each of them but in the relative order and intensity in which the elements are performed. Even within each subsystem, that order varies according to the situation and the individual and actual motivational state of the individual before the releasing situation arises. It is this hierarchically organized, endogenously coordinated subsystem, created anew each time, which is the "motivational mechanism," and it quite certainly is not unitary, nor is it permanent: it is a disposition, a potentiality, rather than a fixed structure, and for all we know it may use a fairly large number of brain structures and mechanisms vicariously. The ethological analysis of these complexities and my interpretation of the results I have given in detail elsewhere (Leyhausen 1965: 1979 a. b).
E. von Holst and U. von St Paul (1960) have demonstrated that even such complex problems of behavioral organization can be attacked successfully using electrical brain stimulation. It is to be deplored that no other investigator has followed their example.
Reviews such as Adams's are both useful and necessary, not in spite of but because of their at least partly controversial nature. Apart from this. I am inclined to think that, in that part of neurophysiology which attempts to deal with complex behavior, the rule of parsimony leads to oversimplification more often than not, and proves itself a hindrance rather than a help to progress.