||Comments by Holger Ursin
Institute of Psychology, University of Bergen, 5000 Bergen, Norway
Aggression and the brain: Reflex chains or network? Adams bases his classification of intraspecific aggressive behavior on a postulated neural circuitry of social behavior. His system is interesting and challenging. However, acceptance of the model rests on acceptance on the neural circuitry. The flow diagrams are meant not only as intervening variables, representing a hypothetical flow of information; they are also assumed to represent structures, and therefore they assume the power of hypothetical constructs (MaCorquodale and Meehl 1948). Adams suggests that the information flow is really from the stimulus through several structures, ending up in responses.
My first difficulty is with this basic assumption of the brain functioning as a reflex chain, since we know that the brain is built as a network. There are complex feedback controls, gates, and a substantial source of variance intervening between stimuli and responses. In his discussion Adams mentions the fact that many of his arrows of postulated information flow may be reciprocal. I would suggest that all arrows on his flow diagrams are reciprocal, and that there is really no good reason, except tradition, to stick to one of these directions, If this is true, simple feedforward models for brain functions become hard to accept, in particular when stimulus "control" of behavior is assumed. Adams states that he finds ethological stimulus concepts like "releasing" and "motivating" stimuli helpful and meaningful.
At least my own neuroethological work on cats with frontal, cingulate, and septal lesions fails to agree with his statements. Flight and defense behavior in feral cats (= defense in Adam's definition) are not affected by these lesions (Ursin 1969; Ursin and Divac 1975). Also, I do not know what is lost due to a lesion affecting 90% of the mesencephalic central gray, but I doubt that the resulting behavioral deficits are explained satisfactorily as a loss of defense. The direction of information flow is particularly doubtful for this area. Ascending systems from very small areas within the central gray and in neighboring areas have a profound influence on a variety of behaviors: for instance, pain (see Liebeskind 1976), sexual behavior (Sodersten, Berge, and Hole (1978), and also defense behavior in a more strict sense (Hole, Johnson, and Berge 1977). For amygdala there is also a quite specific disagreement between Adams and other authors. This is due at least in part, to different behavioral terms.
The author postulates three main motivational systems defense, offense, and submission . There is a growing consensus that it is necessary to discriminate between offense and defense. However, Adams's defense concept differs markedly from the ordinary ethological one, as well as trom the general use of terms in psychology. Adams defines defense as the behavior or "motivational system" in "wild" animals, regardless of whether this is flight, defensive postures, or threats. When the identical behavioral pattern occurs in intraspecific behavior, it represent "submission." If it occurs as a response to an object or an unidentified stimulus source, like uncertainty itselt in open-field behavior, or a shock prod, it is impossible to classify the motivational state following this system. If the behavior is elicited by brain stimulation, it is likewise impossible to classify, since the stimulus condition is now bypassed.
The classification differs from conventional ethological definitions. Leyhausen's [q.v.] descriptions of cat defensive behavior (Abwehr - "defense sensustriction" - SS) is from a cat-cat situation, but identical behavior is observed in cats confronted with dogs or humans (Ursin 1964). Adams's classification is not based on neural circuitry, as was his original goal. The differentiation he suggests is not supported by such data. He assumes "parallel pathways" to explain why he postulates more than one system where only one has been demonstrated. The system also fails to account for the differentiation that has been found for flight and defense (SS) (see Kaada 1967). I believe that a cautious, nontheoretical approach based on ethological and neuroethological observations leaves us with three aggressive categories (defense, attack, and prey-killing) and one type of fear behavior (flight). In addition, freezing is a fear category that seems to be generally accepted In neuropsychological lesion work, freezing and flight have proven to be fruitful concepts for explaining learning deficits (see Kaada 1967; Ursin 1969).
The terms "aggression" and "fear," as used here, follow conventional definitions in psychology. Behavior is aggressive it an object or other individual is damaged or is threatened with damage. Fear is also related to threats, but the individual is now avoiding contact with the stimulus, either "passively" (freezing - passive avoidance) or "actively" (flight - active avoidance). This conservative set of definitions has the advantage of relating to experimental psychology and learning theory as well as neuropsychological data from limbic structures.
Before we conclude that a structure plays a crucial role for a certain behavior, several criteria should be satisfied;
1. The particular behavior should be readily identified in the naturally occurring behavior of the animal, and the terms used should be as close to ethological terms as possible.
2. The behavior should be elicited by electrical or chemical stimulation of that particular structure.
3. Units in that structure should change their activity during execution of that particular behavior.
4. The behavior should be reduced by a lesion to this particular structure.
5. Lesions should produce a handicap in learning problems where this particular action pattern is important for the execution of the instrumental behavior.
6. Electrical or chemical stimulation should "jam" the ordinary stimulus control of that behavior.
7. Pharmacological manipulations should elicit, eliminate, or jam this particular type of behavior.
8. At least some phylogenic homology should exist across species.
9. The behavioral changes produced should be fairly specific for !his particular response, and not be a general effect on all kinds of motivational systems.
The amygdala control of flight and defense behavior seems fairly well established, even with these criteria (Ursin 1965; 1972). For the other areas in Adams's flow diagrams. this strict criteria-set is not met. However, it should be admitted that it is much more difficult to fulfill these strict criteria for brain stem structures, where an anatomical localization is difficult due to the many ascending and descending systems. Adams's work is extremely interesting and challenging, but it does not resolve or eliminate controversies in this field. This was probably not the intention of the author, nor is it achieved by my comments