same way, and that the machinery can be laid bare by experiment. All association, he says, is connection of elementary processes; the phrase ' association of ideas,' if it means that the ideas connect as such, is a misnomer. And the connection of elements is itself a twofold process. When a perception or idea is effective for association, its elements first arouse images that are like themselves; there is a fusion of like with like. But the terms of this fusion have been together, in former conscious presents, with other, unlike elements; the fusion is, therefore, at once supplemented; the homogeneous nucleus is surrounded by all sorts of connected processes. If the association is simultaneous, this is a full account of the matter. If it is successive, then some element in the cluster of associates about the original nucleus arouses its like; a new fusion is formed; and so the process is repeated.
Wundt's first proposition, that only the elementary components of perception and idea can enter into associative connection, is based partly upon the nature of the idea, and partly upon facts of observation. The idea is fluid, variable, instable; it does not stand still to be connected; it has no solidity, so to say, that should allow it to be coupled with another, equally solid idea. And if we observe associations under experimental conditions, we find that they hinge, in reality, not upon the ideas as wholes, but upon some simple constituent of the ideas.
This point established, we may go on to the mode of elementary connection. The first stage, the fusion of like with like, be- comes clear if we translate it into physiological terms. When a complex stimulus is presented to the organism, it arouses a com- plex excitation in the brain. But some at least of the component stimuli have impressed the brain in the past. These stimuli, then, running into the paths of previous impression, reexcite a previous brain-activity; the other, new stimuli have to make their im- pression for themselves. Now, if we retranslate, we have the 394 Association 394 Association conscious fact of fusion. The incoming sensation or image, if it has been in consciousness before, — and otherwise it could not be effective for association, — blends with its own image, with its rearoused self. Since the terms of this fusion are qualitatively alike, the conscious resultant is merely the element itself, given at increased intensity and with a high degree of clearness; we know, Wundt says, that the familiar elements in a complex situation stand out strongly and clearly, while the unfamiliar elements are weaker and more obscure; here, then, is evidence of the nuclear fusion. The second stage presents no difficulty. The reinforced central element, just because it has been in con- sciousness before, and is therefore fitted to arouse the nuclear fusion, must also stand in connection with many other elementary processes; and it is merely a matter of circumstances which of these shall actually be evoked.
What is to be said of this analysis? First, that it is by no means to be confused with the traditional doctrine of association. VVundt's fusion of like with like is not association by similarity; and his cluster of associates is not • association by contiguity. Every case of associates, whether 'by similarity' or 'by conti- guity,' involves, according to Wundt, both of the elementary con- nections: the rearousal of the like and its supplementing by the unlike. Secondly, that the analysis is, as it professes to be, an analysis of observed occurrence; it differs from the older ' laws of association,' not only in form and content, but also in derivation; it is not a product of logic, of reflection. The author offers only two criticisms. The one is, that Wundt has placed the whole mechanism of association in the realm of consciousness, whereas there seems to be no question that, in many instances, the mech- anism or a good part of it is purely physiological, and finds no conscious representation at all. And the other is, that the fusion of the incoming process with its imaginal twin, while as an hypoth- esis it is adequate to the facts, is nevertheless not directly attested by introspection. We might, perhaps, speak of a reinstatement of the like, rather than of a fusion of like with like; the concur- rence of the present excitation with the preexisting impressional tendency would then be a physiological process, to which the References for FurtJicr Reading 395 emergence of the single element, strong and clear, would directly correspond; there would be no rearousal of the mental double.
The law of association (§ 107) will now operate as follows. Two nonsense syllables, let us say, are given within the same con- scious present. Later, the one of them is presented alone. The single syllable, or that aspect of it (visual, auditory, kinaesthetic) which was prominent in consciousness at the time of its earlier presentation, reinstates itself; the previous impression of the nervous system makes the path of excitation easy, and the per- ception is clear and intensive. Along with reinstatement comes associative supplementing: the other syllable appears in imaginal form. So we have what appears to be a typical case of ' associa- tion by contiguity '; but we also see the danger (p. 380) of identi- fying the traditional law of contiguity with our own general law of association.
A final word of caution! We have spoken of impression, and of associative tendency, as if these things were real physiological characters. So, in one sense, they are: the nervous system behaves in certain definite ways which we are in duty bound to rec- ognise and to name. But it must be remembered that our know- ledge is altogether indirect, drawn from the results of psychological experiments. What the impression and what the associative ten- dency are, in themselves, — what goes on in the nervous system when a stimulus is impressed and an associative tendency estab- lished, — of all this we know nothing. The physiological explana- tion of association is, therefore, a problem for the future.
References for Further Reading §§ 105-m. H. Ebbinghaus, Ueber das Gedachtnis, 1885; Psychol- ogic, i., 1905, 633 ff.; E. Claparede, V association des idees, 1903; W. Wundt, Physiol. Psychol., iii., 1903, 518 ff. For experimental methods, C. S. Myers, A Text-book of Experimental Psychology, 1909, 144 ff. Discussions of association which have permanent value, but which the authors would probably modify if they were writing to-day, will be found in W. James, Principles of Psychol., i., 1890, 5506°.; O. Kulpe, Outlines of Psychol., [1893] 1909, 169 ff.
MEMORY AND IMAGINATION § 112. Retention: the Course of the Image. — An impres- sion made upon a plastic substance persists, for some length of time, after the removal of the impressing object; the substance retains the impression. Suppose, then, that a stimulus has impressed the brain: the nervous substance will retain this impression after the stimulus has ceased to act. The sensation or perception will be followed by an image or idea which — if nothing interferes with it — will remain in consciousness so long as the impression retains a certain depth. What happens to it in the meanwhile, during the gradual obliteration of the impression?
The question is not easy to answer. We saw in § 60 that every stimulus of moderate intensity arouses a wide- spread reaction; and it is equally true that every image, auditory or visual or what not, appears in a complex mental setting. When we are dealing with sensation, our organic attitude is determined and maintained by the stim- ulus, which further serves to guide and correct our asso- ciations; consciousness is in relatively stable equilibrium. But, when we are dealing with image, the organic attitude is likely to vary, and the associations, being of the same mental stuff as the image, are likely to influence it in vari- ous ways. Besides, the course of an image can hardly be followed, even under the most favourable experimental conditions, for any length of time. Some new impression is sure to stamp out the old, or some new stimulus to re- arouse the preexisting tendencies of the part of the brain im- pressed, and the image is thus cut across and interfered with.
§ 112. Retention: The Course of the Image 397 However, if we put together the results of investigations so far made, we may say that an image or image-complex is subject to three distinguishable modes of change: it may die away, it may approach a type, or it may be incorpo- rated, whole or part, in new imaginal formations.
We often read of the fading and decaying of images, though we do not so often find a description of the process. If we consider the attributes of the image, it seems that those which suffer directly by lapse of time are intensity and duration; loud images become faint, bright images (p. 204) become dull, and all images flit through consciousness the more hurriedly, the farther back the original experiences lie. These are intrinsic changes, due to the weakening of the nervous impression.
Other factors, however, are at work. Thus the quality of the visual image is definitely affected by the nature of the objective illumination to which the retina is subjected: images of colours and greys tend to lighten in the light and to darken in the dark. Tonal images tend to flat, possibly because the accompanying kin- aesthetic image of laryngeal adjustment tends to weaken. More important is the gradual shift from individual quality to regional or type quality. We saw in § 107 that an association may be aroused, not only by the recurrence of a familiar idea, but also by the occurrence of a novel but similar idea; the nervous conditions of like ideas are partially the same. It appears, now, that the nervous conditions of neighbouring sensory qualities are in part the same, or at least that the impression of any one subarouses the impressions of the others. For, if a particular colour or tone is impressed, the observer soon grows doubtful of its identity; when he seeks to call it up, he may image a quality that lies at some little distance from it on the scale of colours or tones; when a slightly different colour or tone is presented, he may be misled by his image into pronouncing it the same. The image is, of course, always an individual quality, but its quality is no longer strictly correlated with that of the primary sensation.
The same result may be brought about, indirectly, by verbal 39§ Memory and Imagination association. We may remember a colour as red, a light as dark grey, a tone as high. If we then try, at some later time, to image the colour or the light or the tone, we may image that special quality which, under the conditions of the moment, is the nearest representative of the class red, dark grey, high j that is, the qual- ity whose image is ' readiest,' most easily evoked by the associa- tive tendencies of the verbal idea. In such cases we may travel very far from the proper image of the primary sensation. Or, again, we may remember the original quality by absolute impres- sion (p. 313). As our experience grows, we form in all depart- ments ideas like the composite and standardised idea of space to which we referred in § 89; and these ideas may, like that, lapse into unconsciousness and be replaced by a cortical set, or may show only as total conscious attitudes, or may be represented on the particular occasion in some partial and fragmentary way. On its spatial side, e.g., the image is mainly determined by absolute impression, so that small extensions become still smaller, and large still larger, in the corresponding images. The same influence may be traced in the case of intensity 1 and duration. So with quality: if the colour strikes us as a beautiful red, the light as an unusually dark grey, the tone as excruciatingly high, we shall be likely, later on, to confuse the imaged qualities with those of other, similarly impressive stimuli.
1 The attribute of intensity has often been denied to the image. "The idea of the brightest radiance does not shine, that of the intensest noise does not sound" (H. Lotze, Outlines of Psychology, tr. 1886, 28); "the ideas of the slightest rustling and of the loudest thunder exhibit no difference in intensity whatever " (T. Ziehen, Inirod. to Physiol. Psychol., tr. 1895, r54)- rne author believes that such statements betray a form of the stimulus-error (p. 218). At all events, there is no doubt — since the experimental evidence is positive — that images have intensity. If now the image of the thunder is of long standing, and the image of rustling is recent; and if the underlying nervous impressions are allowed to fade out undisturbed; then there may come a time when the intensities are equally weak. Only under these exceptional condi- tions, however, can Ziehen's remark be true. And, as a rule, the image of thunder will always, whatever its age, be stronger than the image of rustling, because thunder makes on us the absolute impression of a typically loud sound, and rustling that of a typically faint, stealthy sound, § ii2. Retention: The Course of the Image 399 We have spoken of ' remembering ' the original colour or tone by means of the class-name and the absolute impression. A dis- cussion of the memory consciousness is not yet in place. Notice, however, that in this process of remembering the image may have disappeared altogether; it is only by accident that it can recur; in the great majority of cases it has been ousted by another image. Most images, indeed, are not allowed to live out their lives; the incoming stimuli and the preexisting tendencies of the nervous system are too much for them.
Instructive observations on the career of imaginal complexes may be made as follows. The observer draws on paper, from a copy or an object, some fairly simple figure: a fleur-de-lys, an heraldic animal. A week later, he is asked to repeat his drawing from the image of the previous drawing; a week later the same request is made, and so on. It is found that certain features of the image may disappear entirely, and that oftentimes the re- peated figure tends to approach a schematic type; these results are already familiar to us. It is also found, however, that the fig- ure may be transformed: certain principal lines of the original drop out, while certain secondary lines form associations of their own, and become dominant; so that, in course of time, the fleur- de-lys has changed, e.g., into a Greek cross. The method does not permit of detailed interpretation, but it shows that an image may persist, unsuspected, in consciousness, through incorporation in a number of successive ideas.
The Memory After-image. — Most observers find no difficulty in passing directly — that is, after a very brief interval (p. 29S) — from sensation to image; even in the case of vision, conditions may be arranged which prevent interference by after-images (pp.68, 72). Sometimes, however, the sensation is followed by a process, lasting perhaps from 5 to 10 sec, which Fechner called the memory after-image. This is not an after-image proper: for it depends, as the image does but the after-image does not, upon the clearness of the primary sensation; it appears only if, like an image, it is sought for, called up; it is stronger and clearer after a brief observation, while the after-image is better when stimulation is pro- longed; and it repeats the lights and colours of the original at a 400 Memory and Imagination time when the after-image, were it present, would be complement- ary. The memory after-image is, in fact, a sort of instantaneous photograph of the sensation or perception. It is reported by observers who have a very poor general furniture of visual imagery, and doubtless plays a larger part in the imageless than in the imaginal mind (§ 141); but many minds of the latter type possess it. We shall attempt an explanation later (§ 118).
The Perseverative Tendency. — Images themselves seem, at times, to crop up of their own accord; we are haunted by tunes, by tags of verse, by a picture, by the face of a drowned man; and ideas occur to us in the most incongruous way. Experiments on verbal association (pp. 274 f.) show the same phenomenon; the ob- server will repeat a word again and again, in his series of responses, without realising that he is obsessed by this particular associate. Grouping all these facts together, certain psychologists have con- cluded that the brain is the seat of what they term perseverative tendencies. The impression does not fade out steadily, but re- covers itself, so to speak, from time to time and under favourable conditions, so that the idea may surge back spontaneously into consciousness.
If this hypothesis is correct, we have to distinguish three sets of nervous tendencies. First, we have the impressional tendency, which represents the ' readiness ' of an idea to emerge, the dis- tance below the conscious limen at which its excitatory process is now going on. Secondly, we have the associative tendency, which represents the strength of the connection between one impression and another, or the degree of excitation that will accrue to the one when the other is reexcited. Thirdly, we have the perseverative tendency, which is a sort of rhythm imposed on the impressional tendency, such that the idea does, now and again, emerge with- out the aid of the associative tendencies.
Why, then, should we separate impressional and perseverative tendency? Why should we not say, at once, that the impressional tendency varies, oscillates, fluctuates? Because the status of the impressional tendency, as we have defined it, is fairly well assured, whereas the status of the perseverative tendency is doubtful. All these hauntings and recurrences may, in fact, be accounted for in § 113. Retention: The Process of Dissociation 401 one or other of the three ways outlined on p. 386, and accounted for more satisfactorily than on the assumption of perseverative ten- dency. Perseveration is strongest under two, somewhat contradic- tory conditions: immediately after the original perception, and during the onset of fatigue. In the former event, both the im- pressional and associative tendencies will be strong, so that there is little to choose between the alternative explanations. But in the second case, of fatigue, it is not easy to see how the perseverative tendency should be set in operation, whereas it is natural that the more ingrained, more permanent associative tendencies should alone remain active, and that the range of consciousness should thus be restricted.
There is, then, no harm in keeping the terms 'perseveration' and ' perseverative tendency ' to designate a certain mode of behaviour of images and a certain part-problem of nervous retention. But there is, in the author's opinion, no positive evidence that the behaviour is unique, or the problem insoluble by appeal to impres- sional and associative tendencies.
§ 113. Retention: The Process of Dissociation. — The as- sociation, like the image, is retained for a time; the associ- ative tendencies persist along with the impressions. But an association, if left to itself, soon begins to break up; the associative tendencies weaken, at first quickly, then more and more slowly, until finally they cease, so far as conscious- ness is concerned, to act at all. It is possible, by means of nonsense syllables, to trace out this process of dissociation, and it is also possible to determine, by variation of the experimental conditions, what are the principal influences that make for permanent retention.
The use of nonsense syllables permits us to follow the associa- tive tendencies from their first establishment to their final decay; there is practically no danger of reinforcement or of inhibition in the intervals of the experiment. It seems certain that these limited and clean-cut tendencies do, in time, disappear; they die 2D 402 Memory and Imagination of old age. The much more complicated tendencies established by meaningful material seem, on the other hand, to persist, below the limen of consciousness, for very long periods, possibly throughout the individual life. We learn poems, in childhood, which we may never think of again until we find our own chil- dren learning them twenty or thirty years later. We try our memory, and discover that, except for a tag here and a tag there, we have forgotten everything. Nevertheless, if we sit down to memorise one of these old poems and another, new poem of the same length, the same metrical form, and the same level of imagination, we regain the old with considerably fewer readings than must be given to the new; the associative tendencies were there, in subliminal degree, although the associations had long since vanished. Childhood, of course, is a plastic period; the original impressions were deep, and the original associations were little interfered with. But even if the experiment is transferred to adult life, the associative tendencies show an extraordinary per- sistence. Ebbinghaus learned some stanzas of Byron's " Don Juan " in his thirty-sixth year, and did not look at them again for twenty-two years. He had completely forgotten them, but he found evidence that the associative tendencies had not died out.
It has been shown, by the experiments with nonsense syllables, that, if two associative tendencies are of the same strength but of different ages, a repetition of the association has the greater value for the older-established tendencies. Hence the advantage of distributing in time the readings of the material to be memorised (§ roo); the associations that are strengthened by the successive readings are older than they would be were the readings massed together. The explanation appears to be as follows. Remote and retroactive associations disappear more quickly than direct associations. The successive readings will therefore sustain and reinforce the direct associations, while they may have actually to reestablish the others; the benefit of the readings will fall mainly to the direct associations. Distribution in time thus plays directly into the hands of the law of initial inhibition. Contrariwise, the massing of the readings will keep the secondary associations alive, and in so far will delay the action of the law.
§ 114. Retention: Individual Differences. — The image is a later development than the sensation, and we may ex- pect, accordingly, that it will show a greater individual variation. The psychology of sensation is concerned primarily with uniformities; all those who possess normal sense-organs have the same general endowment of sensa- tions; and we refer striking peculiarities like colour-blind- ness, tone deafness, insensitivity to pitch differences, — we refer these peculiarities, when they appear, to some abnor- mality of the organ. The psychology of the image, on the other hand, is essentially an individual psychology. The normal brain is a much more variable thing than the normal sense-organ, and the ideas of different minds are constituted in very different ways.
Attempts have been made to reduce these differences to order, and to classify observers in terms of their imagi- nal type. Four principal modes of ideation have thus been distinguished: the visual, the auditory-kinaesthetic, the kinaesthetic, and the mixed. The visually minded ob- server, for instance, retains his experiences in terms of vis- ual imagery; his perceptions, of whatever kind, are trans- lated into visual ideas. An observer of the mixed type repeats in image what he has received in sensation, though he will probably have a certain leaning towards a particu- lar class of images. It seems that there is no pure audi- tory type, and no visual-kinaesthetic type; at any rate, these cases are exceptional. Words are retained in simi- larly characteristic forms: as visual, and as auditory-kinaes- thetic images. It is probable, again, that the auditory- kinaesthetic elements do not occur separately, although the emphasis may be preponderantly upon the one or the other.
4°4 Memory and Imagination While, however, these gross differences undoubtedly ex- ist, generalisation must not be pushed too far. Thus, we cannot argue from verbal to total type; a man may imao-e words as auditory-kinaesthetic, and yet be, on the whole, visually minded. Indeed, it may be questioned whether, apart from the mixed mode of imagery, a total type can be said to exist at all; the imagery of a given observer will vary both with the manner of presentation of the original material, and with the purpose or intention with which the material is approached. We might, perhaps, sum up the situation by saying that individuals are predisposed for different kinds of imagery; that, as a rule, the predispo- sition represents a line of naturally least resistance, but does not prevent the opening of other lines (by the nature of the stimulus, by special cortical set); and that, in certain cases, the predisposition is exclusive and supreme.
We made a brief reference to imaginal type on p. 199. The subject is, evidently, of great importance for education as well as for psychology; it has therefore received much attention, and many methods for the determination of type have been devised and applied. The principal result of the investigations is the proof that type is far more variable and more complex than had at first been supposed. Two points, in particular, may here be mentioned. The first is that the presence of imagery does not necessarily imply the use of imagery; my mind may be full, f.g., of visual images, and yet I may habitually mean and under- stand, think and remember, in other than visual terms. And the second is that a man's talent, or his choice of a profession, is no indication of his imaginal type. "I should have thought," re- marks Galton, "that the faculty [of visualisation] would be com- mon among geometricians, but many of the highest seem able somehow to get on without much of it." " I am myself a good draughtsman," says James, "and have a very lively interest in pictures, statues, architecture, and decoration. But I am an ex- § H4- Retention: Individual Differences 405