||Comments by P. R. Wiepkema
Department of Zoology, University of Groningen, Kerklaan 30, 9751 NN Haren, The Netherlands
On the specification ot motivational systems. In brain/behaviour studies an ultimate question is how the brain organizes and regulates entities like reproductive behavior, food intake, and the like as they occur under natural conditions. In spite of much research, the answers to this question are still very preliminary. It is the merit of Adams's paper to survey the relationship between one such entity- intraspecific aggression - and a number of brain structures, and to speculate about the functional significance of these neural circuits.
Although Adams's paper raises many questions, I want to comment mainly on the concept of motivational systems and the way the author has translated this into defense, submission, and offense. We have to be as clear as possible about what we mean by such concepts, when these are used to elucidate brain. behaviour relationships.
Although not stated explicitly by the author, a motivational system comprises not only a set of specific stimuli and internal motivating and motor mechanisms, but also a specific overt behavioural output. Otherwise, Adams could not have made the distinction between three behavioural entities such as defense, submission, and offense. An overriding question is, then, what precisely the characteristics of such motivational systems are, and whether these can be distinguished from each other. Since Adams's motivational system is very similar to behavioural systems in ethology, it is worthwhile to mention briefly a present-day view on the characteristics of such behavioural systems (cf Baerends 1976).
Overt behaviour consists of stereotyped elements (movements, postures, etc.) that enable a quantification of that behaviour within a species These elements do not occur at random, and this can be demonstrated by different techniques (cf Colgan 1978: Hazlett 1977).
Associations of behavioural elements in time are often similar for many conspecifics under comparable conditions and are called behavioural systems Such systems have often been represented as hierarchical models in which a number of behavioural elements show some common causal factor. Although there is a regular warning against a unitary interpretation of this common factor, it has often been interpreted as a recognizable internal unity such as aggression, flight, or hunger. Although it is not quite clear what Adams means by homogeneous neurons underlying defense, submission, or offense, he strongly suggests that these neurons partially represent the internal hierarchy that ethologists have been looking for. Such an idea is strengthened by the hierarchical models that Adams presented in a recent paper (Lehman and Adams 1977).
However, although hierarchical models are helpful in understanding and describing the organisation of overt behaviour, they may facilitate sham explanations, First, most if not all behavioural observations used to define behavioural systems have been made in a limited set of well circumscribed situations. The observations cited by Adams are no exception to this rule - for instance, the behaviour of a dominant male meeting an unknown intruder male in its home cage, or the behaviour of two males next to each other while receiving unexpected pain shocks. At least after some experience all animals react to such a situation in a more or less stereotyped manner; say offense or defense.
However, such a behavioural stereotype does not mean that system A (e.g. offense) and system B (e.g. defense) must also have their own specific underlying internal mechanisms. Such a conclusion would be tenable only if it could be demonstrated that under varying conditions the animal uses either A or B but never mixtures of both. If, however, the latter would be found, then a more likely model is that under different conditions an animal may use different behavioural patterns, composed out of elements of one basic system. The separate behavioural patterns are then the average responses of an animal adapted to a specific situation. If such reasoning is correct, one would expect different patterns among individuals dealing with a similar behavioural situation, or that elements of, say, defense, submission, and offense can be intermingled in different ways by conspecifics in different situations. The last expectation is supported by the fact that lactating females may show such a mixture of offense and defense.
Therefore, before we can categorize aggressive behaviour of a species into three motivational systems, suggesting the existence of three hierarchies (defense, submission, and offense), we have to investigate how robust these systems are under varying and more complex conditions than hitherto investigated.
A second drawback of hierarchical models is that they do not specify whether or not the behavioural elements belonging to a particular system show specific sequences or patterns in time. The statistical procedures used all emphasize the presence of a common factor and are less explicit about a specific patterning of behaviour in time and space. This may lead to a loose handling of such systems. For instance, if it is stated that as a result of a given brain lesion defense does not disappear, one is inclined to conclude - as Adams does - that for the occurrence of defense that particular brain structure is not necessary. The real and often unanswered question, however, is whether the entire original pattern of the defense response is still intact (normal latencies, durations, frequencies, sequences, goal directedness). Only very sensitive behavioral measurements can inform us about possible changes in behaviour in the limited number of behavioural situations used.
I really doubt, for instance, that forebrain structures fail to form an essential part of a defense motivating mechanism (as Adams suggests). The answer seems to be given when, in simple experimental conditions, no changes are recorded. What really has to be investigated is how such lesioned animals behave in much more complex situations; for instance, in natural colony conditions.
In my opinion the existing literature is either far too imprecise to support, say, a yes/no relationship between certain brain structures and defense or submission, or else the definitions of the latter two concepts are too vague to be useful.
In this context it is somewhat surprising that the author does not refer to newer approaches to the problem of the organisational structure of behavioural systems. These approaches emphasize the specific characteristics of behavioural programs. The point is that behavioural systems like feeding. drinking. sexual behaviour, different forms of aggressive behaviour, and so on, are all conceived as regulatory systems (Archer 1976: Baerends 1976; Toates and Archer 1978; Wiepkerna 1978). All these systems are behavioural programs directed at the homeostasis of specific aspects of the internal and/or external environment.
To realize this, each organism has at its disposal a limited set of behavioural elements that can to a certain extent be arranged and rearranged according to individual experience with a particular situation. Such an approach is attractive not only in that It assumes a similar basic organisational structure of all gross behavioural systems, but also because it accentuates the need for precision in describing time and space patterns of the behavioural programs involved.
In order to know what is regulated, one has to know the differences between and within individuals in terms of aggressive behavioural programs. This brings us to one of the most intriguing fields of present behavioural research; the question of the flexibility of behavioural programs or systems (cf. Hinde and Stevenson-Hinde 1973). At the moment we hardly know anything about the flexibility of "systems" like defense, submission, or offense.