§ 3. Manifestations of Mental Process in Others.—No one can directly observe what is passing in the mind of another. He can only interpret external signs on the analogy of his own experience. These external signs always consist in some kind of bodily action or attitude. Thus when a man clenches his fist, stamps, etc., we infer that he is angry. When a dog wags its tail, we infer that it is pleased. The knowledge acquired in this way must be carefully distinguished from that which is obtained through intercommunication by means of language. When a man tells us that he is or was angry, he is not directly expressing his anger, but his knowledge of his anger. He is conveying to us the result of his own introspection. This source of information is in no way peculiar to psychology. It does not differ from any other communication of observed facts by means of words. The peculiarly psychological inference rests on signs which may or may not be noticed or understood by the subject who displays them. On the other hand, communication by means of language necessarily presupposes that the person communicating the information is himself aware of the meaning of the words which he uses. He must first understand himself in order to make others understand him. It may happen that the inference from the direct expression of the mental state may contradict the subject's own assertion about it. He may show most unambiguous symptoms of anger, and at the same time declare vehemently that he is not angry.

In the case of the lower animals and young children, it is impossible, and in the case of savages it is difficult, to obtain verbal descriptions of their own mental states and processes. This is partly because they either do not use language, or use a language inadequate for the purpose, and partly because they are not introspective. Under such conditions our only course is to rely on the interpretation of the appropriate external manifestations of the processes themselves. Interpretation becomes more difficult in proportion to the difference between the mind of the psychologist and the mind which he is investigating. The interpretation must rest on some analogy between the two. But if the analogy is only partial and accompanied by great diversity, a constructive process is necessary. It is in his own mind alone that the psychologist has the constituent elements from which an interpretation can be framed. "All depends on accurate resolution of his own complex consciousness into its constituents, and on recompounding these in such a way and in such proportions as to explain the nature and order of the signs which indicate to him the mental processes of others."* For instance, he finds among savages a widespread belief in the power of all kinds of odds and ends to influence the fortunes of the person possessing them. This is a prevailing tendency of savage thought; if the psychologist looks for analogies in his own mental life, he will find them few and far between. But they are not likely to be wholly absent. There are moments in which he either has been influenced or has felt strongly inclined to be influenced, by considerations in themselves as meaningless as those on which the savage relies. The fall of a picture, or the spilling of salt, or the presence of thirteen at table, may make him uneasy in spite of reason. If he has ever been carried away by the gambling impulse, he must have been almost irresistibly prompted to regard quite irrelevant details as having an essential bearing on his winning or losing. In order to construct the mental state of a savage, he must carefully observe and analyse these transient and occasional mental attitudes in which he approximates to savagery. He must then attempt to represent a mind in which tendencies, that, in him, are so overborne by other conditions as to be transient and occasional, are unchecked by opposing forces, and for that reason prominent and permanent. It sometimes happens that a man is so destitute of a certain kind of mental tendency himself, that he is unable to understand its presence in others. Thus, Charles Lamb tells us that his friend, George Dyer, could never be brought to say anything in condemnation of the most atrocious crimes, except that the criminal must have been very eccentric.

* Analytic Psychology, vol. i., p. 15.

The besetting snare of the psychologist is the tendency to assume that an act or attitude which in himself would be the natural manifestation of a certain mental process must, therefore, have the same meaning in the case of another. The fallacy lies in taking this or that isolated action apart from the totality of the conditions under which it appears. It is particularly seductive when the animal mind is the object of inquiry. The economy of a beehive displays such adaptation of means to ends, as to suggest strongly farreaching prevision and political faculty of a human kind in the bees. But it would be very rash to trust this first impression. We must first consider all the other actions of bees and similar insects; we must also examine in detail how the individuals concerned severally perform the separate acts which in their combination constitute the orderly scheme of organization of bee society.

We shall then find that the most essential modes of behaviour, especially on the part of the queenbee, are due to congenital tendencies, which operate independently of previous experience. We must further take into account the physical organisation of the bees. Their nervous system differs so widely and in such a manner from the human, as to make us hesitate before ascribing to them so very large a share of processes especially characteristic of human beings. Finally, we find that the division of labour which makes the bee community possible, is directly determined by congenital differences of physical organisation. The queenbee, the worker, and the drone, differ not only in their actual behaviour, but in their bodily constitution. The bodily constitution is so prearranged by nature as to be adapted for certain special functions. Here all analogy with the political organisation of human beings breaks down. This is a typical instance. The lesson to be learnt from it is that in investigating the mental conditions of persons or animals widely removed in their general circumstances and conditions from our own, we must assume an attitude of critical suspense until we have taken into account everything which can have a bearing on the problem.

This warning is the more important because human language is especially constructed to describe the mental states of human beings, and this means that it is especially constructed so as to mislead us when we attempt to describe the workings of minds that differ in any great degree from the human. The very implications of the words we are almost compelled to use in describing what we suppose to go on in the mind of a dog or a cat surreptitiously introduce interpretations which may be quite false, and often are so. It is, therefore, above all things necessary in these cases to criticise our language, avoiding popular phraseology, and substituting technical terms with fixed meanings carefully defined. A horse, having had a feed at a certain place on one day, stops of his own accord at that place on the second journey. People say that it remembers being fed there before, and infers that it will be fed there again. In all probability these words with their human implications are quite misleading. Suppose that the driver of the horse is a bibulous person, who takes a drink as a matter of course whenever he comes to a publichouse on the road. In order to do this he need not go through the process of remembering that he has had a drink at a publichouse before, or of inferring that he can have a drink at a publichouse again. He simply has a bias to stop at a publichouse whenever he comes to one. Probably the horse's act implies just as little of remembering or inferring.