On Perception and Imitation
Mr. Hobhouse, in speaking of the possibility of imitation in animal behavior, remarks: "What has first to be settled is the possibility of a still simpler mental act—learning by the perception of an event and its consequence—when that consequence directly affects the learner."[1] Giving the results of his own experiments, he concludes that, "Whether by the perception of what was done by another, or by noting the results of their own actions, it seems fair to say that what the animal learnt to do was, in some instances, though not in all, to effect a certain perceptual change as a step to procuring food."[2] There is no question, then, that Hobhouse in his somewhat liberal interpretation of animal intelligence recognizes learning by a perception of an event and its consequence in animal behavior, and therefore that the basis for possible imitation is also there. Indeed, in his account of the monkey upon which he experimented he definitely ascribes certain achievements to the influence of imitation. I am not interested in criticizing Hobhouse's results, though one cannot but feel that he failed to recognize the importance of the past training which these apes had had, and that he took the data given him by the keepers in the Bel Vue Gardens in Manchester with too little criticism.
It is rather the bearing of the doctrine of perception upon imitation to which I wish to call attention in the few words which I have to say. Mr. Hobhouse follows the growth of perception from mere stimulation with purely instinctive response, to stimulation which has assimilated a certain character out of the active content of the experience, on to the perception of objects in what he calls concrete experience and practical judgment. In this latter stage the "character," which for example we find in our own immediate perceptions of distance, has developed into a consciousness of relation. There is not as yet the simultaneous presence of both terms in the relation as parts of a whole and yet separate from each other which characterizes a more reflective consciousness, but the experience is consciously controlled by a content in which both elements are felt.
The problem in animal psychology is to find out in what way consciousness of the consequences of an act may get into the stimulation that leads to the act. The steps in Hobhouse's treatment are unconscious assimilation of the affective nature of the consequence—its agreeableness or disagreeableness—the assimilation of a certain character of the consequence, a character which has a felt intellectual content, not simply an affective one; and finally such a consciousness of the consequence that it can conceivably control immediate reactions, that is a working consciousness of relations—a working, not a discurie [sic] consciousness.
This type of animal psychology, which in its fundamental technique is not different from that of Lloyd Morgan's, is after all at the mercy of a logical analysis. The consciousness of relations, which is the essence of reason, is identified, by our own introspection, with the feeling of characters and affective states where these have in them the implicit relation. The behavior of animals seems to justify us in ascribing such feelings to their consciousness, as effective contents.
It is refreshing to turn to Mr. Thorndike's analysis, where one finds intellectual states of consciousness used as little as possible in the explanation of animal behavior. Mr. Thorndike suggests that in the place of the association of ideas, or of any states of consciousness as such, we may explain most instances of the learning of animals, by the association of an impulse with a stimulus. The cement for this association the author still finds in pleasure and pain, but the association is one that lies not between watered down logical elements, but between the stimulation and the whole act of the animal. By a system of trial and error the different impulses of the animal are provided with appropriate stimuli.
Of course, Mr. Thorndike does not imply that this is the only method of explaining all animal behavior, but he insists that this method of explanation of animals' learning tricks should be used where possible.
The association of a complete impulse with a sense-stimulus leaves no room for imitation in the sense in which Hobhouse uses the term.
There could be mimicry and automatic imitation, because both of these imply simply that impulses to instinctive reaction are mediated by certain stimuli that mediate like impulses in other forms. Mr. Thorndike leaves us somewhat in the dark as to the exact nature of the impulse. My understanding of his use of the term is that of an activity which goes off under stimulation without cognitive control, that is without conscious analysis of the stimulus. The means of stamping in this relation between an impulse and a particular stimulus is, as has been said, the pleasure or pain that accompanies the experience.
Where Mr. Thorndike can explain animal conduct by this formula, he finds no evidence of imitation on the part of the animals, that is, of imitation which, to use his own phrase would [be] an "associative transference of the self."
So far as this formula can be used to explain animal behavior it would be out of place, as well, to use the term perception for the consciousness of the stimulation and its relation to the impulse which it mediates, but the formula gives us a basis, on the other hand, for a psychological in the place of a logical analysis of perception.The other type of theory imply that there are certain states of consciousness there, due to stimulation of sense organs and organic tracts, which have in themselves possible cognitive values, that the animal learns by associating these contents with the execution of the act and its consequences. (The term "association" is used in the general sense, including assimilation and fusion.)
If we start from he point of view represented by Mr. Thorndike's formula we can ask what are the conditions under which a cognitive, that is a control, element would arise within the act and so get the conditions for the appearance and growth of perception. We would look for these elements in so far as not only the stimulation but also the response in its relation to the stimulus could be conceived of being present in the consciousness of the form. In other words, we would take the functionalist's point of view and recognize no contents in animal consciousness whose function was not the ground of their existence.To give a concrete illustration: if a monkey is offered a stick with which to push a banana within his reach, we imply that he perceives the stick as a possible tool, and this perception implies that the sensing of the stick awakens the tendency to respond, that the tendency to grasp and move the stick reacts back upon the sensing of it. These two processes control each other. The visual process awakens the impulse to seize and manipulate, the tendency to seize and manipulate keeps the sensing to those features of the stimulation which mediate the impulse. Those who have watched animals or who are familiar with recent observations of animal conduct will recognize that evidence of this sort of relation between the stimulation and response is just what is lacking. The stimulation calls forth, it may be, well-directed and even complicated responses, but there is little evidence that the tendency to respond is there to hold the eye to the object or what is done with it.
This amounts to saying that the ground for the selective character of perception lies in the interrelation of the response and the stimulus, and when we say that it is impossible to get an animal's attention we imply that the inhibited tendency to respond is not there to hold the eye to what we are presenting to his vision.If we ask now what type of behavior will indicate such possible perception in the conduct of lower animals, the reply will be, where we find intermediate acts which must be adapted to final results. In such cases, if the response and its stimulus can be both present in terms of mutual control these intermediate acts can serve to mediate the final act, and would represent in a possible consciousness just the contents which are called for in perception. On the other hand, many complicated acts take place in the experience of animal forms where each step simply provides the stimulus for a succeeding step, and in which there would be no necessity of the preceding act being present in its inhibited tendencies. The first step need not even lead up to a unique stimulus. After the completion of the first step the organism may be ready to respond to a number of different stimuli, e.g., a cat seeking to reach its feeding place and finding a door shut may respond to the stimuli which take it about by another path without necessarily having a perception of the house as a whole.
If, however, the ape whose behavior Mr. Hobhouse describes adapted its throwing of its blanket so as to bring in the banana without training and not by a purely accidental throw, its consciousness of the blanket would certainly have the same character as our own perceptions, that is, there would arise certain sensuous contents which would be sensed with the unity of the act which it mediates, while the kinesthetic contents, which the inhibition of the tendencies to react supplies, would provide the material out of which the consciousness of a physical object arises, and the adaptation of the process to the later act would carry with it the crude relational content out of which the meaning of the percept springs.
There is a further consideration that deserves attention here. What type
of physical reaction lends itself to this mediating experience out of
which perception may arise? If one passes in review the tricks and test
acts which the comparative psychologist uses in his observation of
animals, he finds two general types: those of the maze, and those which
require some sort of manipulation that must precede the final act, e.g.,
the opening of a door by various devices, the use of a stick, etc.
The first type has served to indicate the comparative case with which a
series of acts may be built up, each providing the stimulus for the
succeeding step, without the necessity of implying any perceptual process.
The second type is of interest because it suggests the fact that our own
perceptions consist so largely in the interpretation of what comes
through the eye, the ear, and other distance sensations, through the
suggested kinesthetic experience of possible contact. This fact, which
lies at the basis of the older distinction between the primary and
secondary sensations, and finds further expression in the inevitable
presentation of the outer world in terms of solid matter, i.e., in the
imagery of actual manipulations, this fact suggests that a rich
kinesthetic experience in manipulation may be almost a precondition of
perception, that is, that the sort of mediate experience in which stimulus
and response would mutually control each other in the adaptation of one
act to another could hardly arise before the primate with his highly
sensitive flexible hand.
If we accept this or an analogous definition of perception, it would follow that imitation could not arise as a conscious phenomenon before such mediate acts appeared. One cannot have imitation in this sense without perception, and given perception it is hard to see how imitation can lag far behind.
A crucial test for the presence of imitation would be something like the following: Make sure that the animal has a series of acts well coordinated which lead up to a final act. If, then, it will vary this final act while keeping control over the mediate processes, after seeing the variation carried out by the observer, it would be difficult to deny the animal the power to perceive and to imitate. If, for example, to open a door a latch must be lifted and a plug pulled out, and the observer should push in the plug instead of pulling it out, and the animal should lift the latch and then push in the plug, there would probably exist no doubt in the observer's mind that his subject was imitating.
