Introduction and Motivational systems
(Section continued from previous page)
What are the criteria for a motivational mechanism? And do I give a convincing example? These questions are raised explicitly by Koolhaas and by Miczek. Although I imply the criteria in the first three paragraphs of the section on defense, I am not sufficiently explicit. In fact, the criteria are best expressed by Ursin in his commentary, although I am not sure that stimulation must "jam" the stimulus control of a behavior. The Ursin criteria define a motivational mechanism if one restricts them further to insist that a lesion of the structure would totally and irreversibly abolish all of the behaviors of the motivational system, that is, all of its motor patterns except those that are "ambivalent" when they are activated by other motivational systems as well. Although Ursin does not acknowledge it, I think that these criteria are met for the midbrain central gray and immediately adjacent tegmentum as the motivational mechanism of defense: (1) the behaviors of defense may be described in ethological terms; (2) defense may be elicited by both electrical and chemical stimulation here; (3) units here change their activity during defense - in this case I grant Koolhaas's objection that the unit data on the rat are weak, but I invite him to consider my unit data from the cat, which are much stronger; (4) the behaviors of defense are all permanently abolished by total lesions here; (5) an animal will no longer learn a response to escape from stimulation after lesions here; (6) the question of jamming is controversial as stated above; (7) pharmacological manipulations such as those involving morphine have their effects upon defense here; (8) there are complementary data from animals as widely removed phylogenetically as chickens, rats, cats, and monkeys; and (9) the behavioral effects of manipulations here, while not confined to defense, may be quite different for other behaviors, for example, facilitation of "appetitively motivated behaviors" as noted by Karli in his commentary. It should be noted that Ursin's criteria need not refer exclusively to a motivational mechanism. If one substitutes the term "motor pattern" for Ursin's "behavior," then the criteria would define a motor patterning mechanism. The criteria apply to a motivational mechanism only if the term "behavior" stands for all of the motor patterns of a particular motivational system, and if the behaviors are irreversibly abolished by a lesion. By these criteria, for example, the amygdala cannot contain a motivational mechanism for defense, because defense survives removal of the entire forebrain.
My assumption that a vertebrate neural mechanism would consist of a set of homogeneous neurons is criticized by Glantz from the standpoint of his work with invertebrate nervous systems. The question he raises is interesting, but I do not agree with his answer that neural systems must consist of circuits of small numbers of heterogeneous neurons. Glantz and many other invertebrate neurophysiologists bias their analysis by recording only from a few large neurons. For example, in the abdominal nerve cord of the crayfish studied by Glantz the giant escape command cells have the two largest axons, as shown in the figure from Krasne and Wine on page 276 in the Hoyle (1977) volume. What about the hundreds and perhaps thousands of smaller neurons in the figure that have not been categorized because they are too small for intracellular recording? Are they homogeneous or heterogeneous? It is true that invertebrate nervous systems have smaller numbers of neurons than those of vertebrate but Davis in the Fentress (1976b) volume still concedes that there are at least 10,000 neurons in the central nervous system of a snail and 100,000 in a crayfish. Most of these neurons are small and no easier to study than the neurons of a vertebrate. Vertebrates, in some cases, also have a few giant neurons like the Mauthner neuron of fish and the Muller cells of the lamprey. But, as in invertebrates, it is not appropriate to characterize their nervous systems by these few large neurons, Instead, these neurons are exceptions that have evolved to handle peculiar functions for which large size is useful. There is only one complex animal that I know in which all neurons are large, and that is the vertebrate Necturus. (This does argue, by the way, that there should be more intensive study of the behavior and nervous system of this organism.)
There are several arguments suggesting, but not demonstrating, that vertebrate neuronal aggregates consist of pools of homogeneous neurons. The best studied neural system, the vertebrate retina of Necturus, does consist of populations of homogeneous neurons of only five (or a few more) classes of neurons, each class homogeneous in terms of its anatomical location, inputs, outputs, and types of synapses. However, the retina, one may argue, is a sensory system and should not be taken to represent an integrative system such as that described here. Another simple consideration argues for a great degree of homogeneity in vertebrate neural populations, no matter what system; the absolute size of the vertebrate nervous system, and its total number of neurons, are determined less by function than by size and metabolism of the species (Blumenschine, Mink, and Adams 1978). For example, a rat brain has about ten times as many neurons as a mouse brain, yet surely no one would argue that a rat brain has ten times as many classes of neurons; the difference must be in the degree of redundancy in homogeneous pools of neurons. Finally, there is a point of view, to which I adhere, that the actual genetic instructions for neurons are both few in kind and quite limited in number. Strumwasser (1967) has suggested that there may be as few as eight kinds of information encoded in a neuron, and data from my own work in behavior genetics suggest that the genetic factors underlying differences in behavior among strains of rats may be quite small in number (Adams 1978).
(Section continued on next page)