All the after-images are sensory in character. So too are the memory colours that we habitually lay over familiar objects (p. 63), and that make us see snow as white and gold as yellow and coal as black, just because they are ordinarily or typically white and yellow and black. So also are the recurrent images, those troublesome and haunting images to which most of us are subject at times: the tunes that run in our head and that we cannot get rid of, the rows of figures that obsess us after a long morning of calculation, the bright disc that keeps cropping up after we have spent several hours at the microscope. So, again, are the images that serve to complete and round out an imperfect perception. A favourite device of modern advertising is to outline the human figure only in part and to leave the remainder to the imagination; and you will perhaps notice, if you look attentively at such a figure, that the outline, so far as the suggestion of the neighbouring lines is unambiguous, is indeed completed in image, black on white or colour on colour; only where the completion is uncertain do the images fail. These tied images, so called because they are unequivocally bound up with the sensory portion of the perception, occur also in the sphere of sound; a missing orchestral part, if it is familiar, may be clearly heard by the conductor.

Not everyone has recurrent images; and perhaps only a large minority have tied images. The image—even if we decide that it is only a secondary sensation, psychologically indistinguishable from sensation—nevertheless represents a later stage of biological development than the sensation proper, and our equipment of images is correspondingly variable; your own experience may be richly imaginal, while your friend, under the same conditions, has hardly a trace of imagery. Those who do possess recurrent and tied images agree that they are distinguished from sensations rather by their context, by the presence or absence of certain other processes, than by any difference of nature. The same thing holds of those abnormal phenomena to which Hume referred. Hallucinatory images are by no means uncommon in the drowsy period that precedes sleep; we hear the telephone bell, or we hear our name called; some of us—there are, again, great differences in individuals—have hallucinations of sight. Dream images also differ markedly from individual to individual; but the dream is nearly always accepted as a real event. One of the most puzzling facts in this connection is the occurrence of concomitant or synæsthetic images. In the commonest case, that of coloured hearing, any auditory stimulus arouses, along with the appropriate sensation of hearing, whether tone or noise, a visual image of light or colour. The sound of the word Tuesday, for instance, may be seen as a light grey-green followed by a yellow! We might suppose, at first thought, that coloured hearing is due to association, to a connection between sight and hearing set up in childhood and continued into adult life; but the evidence points to some inborn connection in the nervous system; coloured hearing tends strongly to run in families. Moreover, we know of no natural or normal association of colours with tones, although the attempt has often been made to illustrate music by colours; the recent colour-scoring of the Russian composer Scriabin is, for instance, nothing more than an idiosyncrasy, and will make no general or permanent appeal to the musical public. There are many other kinds of synæsthesia, besides this connection of sight and sound; and we have no reason to think that every instance is to be explained in just the same way; in all cases, however, we have a particular sensation uniformly accompanied by another, which we may call either a secondary sensation or an image of sensory character.

Coming back to the normal life, we have next to note the part played in certain minds by habitual images. Just as, in Wagner’s operas, the performer comes upon the stage to the accompaniment of some characteristic musical phrase, some ‘motive,’ as it is called, which recurs again and again as he enters and reenters to take his share of the action, so in minds of the imaginal type such general notions as ‘virtue’ and ‘commerce’ and ‘summer’ may regularly call up mental pictures, little groups of images, which illustrate or characterise the notions: thus, virtue may be pictured mentally by the flash of a human figure, standing very upright. These pictures are usually incomplete, mere impressionist sketches; but they may remain unchanged for years.

Finally, we come to the images which enter into our ideas of memory and of imagination. We discuss these ideas later; here we need only say that the psychological distinction between sensation and image, if it is to be drawn at all, must be drawn between sensation and the free images of memory and imagination, and cannot be drawn earlier. Some psychologists believe that a memory-image can always be distinguished from a sensation, that the two processes differ in their intrinsic nature. It is difficult to put the question to the test of experiment; but what evidence we have seems to look the other way. We shall do best to suspend judgement.

The word ‘image’ is unfortunately used, as the foregoing paragraphs have shown, both for the simple image and for groups or clusters of images; thus, the recurrent image and the habitual image are always complex. Summing up our results, with this warning in mind, we may say that positive and negative after-images, memory colours, and synæsthetic images are definitely sensory in character; that the simple images which make up memory after-images, recurrent and tied and habitual images, hallucinations and dreams, appear to be of the same kind; and that the simple images which compose our ideas of memory and imagination may or may not be intrinsically different from sensations. The simple image may therefore be defined as an elementary mental process, akin to sensation and perhaps indistinguishable from it, which persists when the sensory stimulus is withdrawn or appears when the sensory stimulus is absent. We may say further that, while every normal person has very much the same equipment of sensations, there are great individual differences in the matter of secondary sensations or images; in some cases they are interwoven into the whole tissue of experience, in others they are infrequent or even lacking; we shall see presently how they may be replaced. In general, images of sight and sound are common; then come images of touch and temperature, and then again images of taste and smell, which are uncommon; organic images are very rare. Kinæsthetic images undoubtedly occur, and probably occur frequently; but they are likely to blend with kinæsthetic sensations, and so to escape notice.

[§ 18]. Simple Feelings and Sense-Feelings.—Many of our experiences are indifferent; but many of them, again, are pleasant or unpleasant. These two words, pleasant and unpleasant, denote elementary mental processes of a different sort from sensations and images; they are known as simple feelings. The term ‘feeling’ is itself even more ambiguous than the term ‘image’; it is natural to speak of ‘feeling’ a strain or effort, a warmth or cold; but we shall henceforth use it only in its technical meaning, to indicate the way in which stimuli affect us, pleasantly or unpleasantly. We must discard altogether the words pleasure and pain, although they have long been current as the names of the simple feelings, and although they are much less clumsy than pleasant and unpleasant. We discard them because pain is a sensation (p. 43); and pains, while usually unpleasant, may at times be pleasant; the scratching that relieves an itch and the nip of the wind on a brisk winter’s day are both pains, but they are also both pleasant.

The main difference between sensation and simple feeling is that a feeling cannot be made the object of direct attention. Try to attend to the pleasantness or unpleasantness of an experience, and the feeling evaporates, eludes you; it is like clutching a ghost; you find yourself beyond the feeling, so to speak, and face to face with some obtrusive sensation or image that you had no wish to meet. This peculiarity of feeling must, of course, be taken account of in our conduct of the psychological method of observation. The formula of observation (p. 19) was:

psychological (vivid experience → full report).

In the case of sensation, the observer is set or disposed, beforehand, to attend to sensation and to report upon sensation; the sensation comes, and is attended to; and the report which follows is determined, under the influence of the preliminary set or disposition, by the nature of the sensation. In the case of feeling, the observer is set to attend to sensation, but to report upon the feeling which accompanies the sensation; the sensation comes and is attended to; and the report then describes, under the influence of the preliminary set, the feeling which accompanied the sensation. That sounds a little paradoxical; but the method is not difficult in practice; and it has the advantage that we can use all manner of sensory stimuli (colours, tones, everything) in our study of feeling.

We find, first of all, that pleasant and unpleasant are really opposites; the colour or tone that is most often reported as pleasant is least often reported as unpleasant, and conversely. An obvious result? Not at all; for what is obvious to common sense demands very careful consideration at the hands of science; and the fact that, in this instance, common sense turns out to be right does not at all mean that we should have been justified in taking it for granted. We find, secondly, that intensity of feeling behaves like intensity of sensation (p. 67); the more pleasant or unpleasant an experience is, the more must the stimulus be changed if we are to feel a difference; and the less pleasant or unpleasant it is, the less change need be made to produce a change of feeling.