§ 16. ASSOCIATIONS.



1. The concept association has undergone, in the modern development of psychology, a necessary and very radical change in meaning. To be sure, this change has not been accepted everywhere, and the original meaning is still retained, especially by those psychologists who support, even to-day, the fundamental positions on which the association-psychology grew up (§ 2, p. 14 sq.). Association-psychology which is predominantly intellectualistic, pays attention to nothing but the ideational contents of consciousness and, accordingly, limits the concept of association to the combinations of ideas. hartley and hume, the two founders of association-psychology, spoke of "association of ideas" in this limited sense 1). Ideas were regarded as objects, or at least as processes which could be repeated in consciousness with exactly the same character as that in which they were present at first (p. 14, 8). This led to the view that association was a principle for the explanation of the so-called "reproduction" of ideas. Furthermore, it was not considered necessary to account for the rise of composite ideas through psychological analysis, since it was assumed that the physical union of impressions in sense perception was sufficient to explain their psychological combination, and so the concept of association was limited to those forms of reproduction in which the associated ideas succeed one another in time. For the discrimination of the chief forms of successive associations, aristotle's logical scheme for the memory processes was accepted. In accordance with the principle of classification by opposites the following forms were discriminated: association by similarity and contrast, and association by simultaneity and succession. These class-concepts gained by a logical dichotomic process were dignified with the name "laws of associations". Modern associationism has generally sought to reduce the number of these laws. Contrast is regarded as a special form of similarity, for only those contrasted concepts are associated which belong to the same general class; and associations by simultaneity and succession are both included under contiguity. Contiguity is then regarded as outer association and contrasted with inner association by similarity. Some psychologists believe it possible to reduce these two forms to a single, still more fundamental, "law of association" by making association by contiguity a special form of similarity, or, what is still more common, by explaining similarity as a result of association by contiguity. In both cases association is generally brought under the more general principle of practice or habituation.
 
 

[1) The author remarks that the English word idea as here used corresponds to the German Vorstellung. Tr.]
 
 

2. The whole foundation for this kind of theorizing is destroyed by two facts which force themselves irresistibly upon us as soon as we begin to study the matter experimentally. The first of these facts is the general result of the psychological analysis of sense perceptions, namely the fact that composite ideas, which association-psychology regards as irreducible psychical units, are in fact the results of synthetic processes which are obviously closely related to the complex processes commonly called associations. The second fact comes from the experimental investigation of memory processes. It is found that the reproduction of an idea in the strict sense of a renewal in its unchanged form of an earlier idea never takes place at all. What really does happen in an act of memory is the rise of a new idea in consciousness;

this new idea always differs from the earlier idea to which it is referred, and usually derives its elements from a number of preceding ideas.

It follows from the first fact that there are elementary processes of association which unite the components of ideas and are earlier in their appearance than the associations of composite ideas with one another, although it is this later group of processes to which the name associations is generally limited. The second fact proves that ordinary associations can be nothing but complex products of the earlier elementary associations. These conclusions show the utter lack of justification for excluding from the concept association the elementary processes the products of which are simultaneous ideas rather than successive ideas. Then, too, there is no reason for limiting the concept even to ideational processes. The existence of composite feelings, emotions, etc., shows, on the contrary, that affective elements also enter into regular combinations, which may in turn unite with associations of sensation elements to form complex products, as we saw in the rise of temporal ideas (§ 11, p. 159 sq.).

3. It follows from what has been said that the concept of association can gain a fixed, and in any particular case unequivocal significance, only when each association is reduced to its elementary processes, which elementary processes never appear in actual psychical processes except in so far as they give rise to more or less complex products. Therefore the only way to find out the character of elementary association is to subject complex associated products to a psychological analysis. The ordinarily so-called associations (the successive associations) are only one form, and the loosest at that, of all the forms of combination. In contrast with these we have the closer combinations from which the different kinds of psychical compounds arise. For these processes we have already adopted the general name fusions, because of the closeness of the union (p. 102 sq.). The next stage of combination is found in the simultaneous associations which arise when a given psychical compound is changed through the influence of the elements of other compounds acting upon it. We designate these processes, because of the way in which the elements interact, assimilations. In addition to these assimilations we have another group of associations which are also generally simultaneous in character, namely the processes which herbart called complications, and which consist in simultaneous associations of psychical compounds derived from different spheres of sensation. Finally, there are associations which unite psychical compounds into temporal successions of ideas. These are the forms of association which are most easily observed. They were, therefore, the only forms recognized at first. We call these successive associations.

A. FUSIONS.

4. The various forms of fusion of psychical elements which are possible, have been described in detail in the course of the discussion of psychical compounds. These compounds are, indeed, nothing more nor less than the products of such fusions. The various fusion processes require, therefore, at this point only a brief treatment with special reference to the definition of their relation to the other processes of association. With reference then to their special characteristics as association processes, the processes of fusion may be described as thoroughly fixed associations of psychical elements. An element of a fusion may, to be sure, appear in other combinations, but it can never appear alone. It is the processes of fusion, accordingly, through which all the real psychical compounds of our conscious experience arise, for there are no isolated elements in consciousness (p. 32). The existence of these simplest forms of association could have been inferred from the existence of more complex associations, even if there had been no direct evidence of the simple associations in the analysis of the various forms of psychical compounds. For it would hardly be comprehensible that combinations should arise between complex compounds if there were no predisposition toward these combinations in the elements. Indeed, it will appear as a fact in the later discussions, that the associations of complex compounds are always to be traced back to associations between the elements of these compounds (§ 16, 10).

5. We may distinguish as the chief forms of psychical fusion, intensive fusion and extensive fusion. This agrees with the results of our earlier discussions of psychical compounds. The intensive fusions subdivide into sensation fusions and affective fusions. The chief examples of sensation fusions are those which appear in clang compounds (p. 104), and the chief examples of affective fusions are composite feelings (p. 177). If we neglect for the moment those differences between various forms of intensive fusion which result from the nature and relations of the specific elements which in each case enter into the fusions, there are two distinguishing characteristics common to all intensive fusions. In the first place, such fusions result from the combination of sensation components, or affective components belonging to a single system. For example, the elements of a clang fusion belong to the sphere of tone sensations, the elements of a common feeling belong to the sphere of touch. In the second place, in every intensive fusion one element of the combination stands out as the predominant factor. For example, in a clang there is a chief tone, in a total feeling there is a chief feeling. Extensive fusions include spatial and temporal ideas, emotions, and volitional processes. They are more complex than the intensive fusions because they always include combinations of disparate elements. But even here there are certain predominating elements which give to the fusion products their unitary character. As predominating elements in the case of spatial ideas, we find outer tactual sensations and visual sensations. In the case of temporal ideas the feelings of tension and relief are such predominating factors. In the case of emotions and volitions the predominating factors are the partial feelings which result from the above mentioned feelings of tension and relief, and from exciting and quieting feelings (p. 178, 206). In point of complexity the various extensive fusions may be arranged in a series beginning with the least complex. The first members of such a series are the spatial ideas which are pure sensation ideas. They are, as compared with the other extensive fusions, relatively simple, while they are, as compared with intensive compounds, more complicated in character. Following the spatial ideas in the series, come temporal ideas. These contain both sensational and affective elements, but certain sensations are so closely fused with the dominating feelings that even the feelings are more or less ideational in character, that is, are directly referred to sensory impressions. The last members of the series are the emotional and volitional processes. These processes differ only in their closing phase, and all belong, therefore, together. They constitute, furthermore, the transitional stage between fusions and complex associations, because in them, complex compounds, such as spatial and temporal ideas and compound feelings, always enter as accessories to the main process. The extensive fusions, including the spatial ideas as their simplest form, and volitional processes as their most complex form, may be said to have the same characteristics in regard to the kinds of elements which they contain as have complications. They also show certain of the essential characteristics of successive associations. In this way it may be said that there are in the various forms of fusions, anticipations of each of the complex forms of association which are to be described. Assimilations are anticipated in intensive fusions; complications are anticipated in extensive spatial fusions; and, finally, successive associations are anticipated in temporal fusions and in emotional and volitional processes, which latter appear as the more highly developed complications arising from temporal ideas. Intensive fusions and spatial fusions may also be classified, together with assimilations and complications, as simultaneous processes. Temporal ideas, emotions and volitions belong, together with the memory processes to be described later and the related processes, under the general head of successive associations.

B. ASSIMILATIONS.

6. Assimilations are forms of association which constantly appear during the formation of intensive ideas and spatial ideas and thus serve to supplement the process of fusion. Assimilation is most clearly demonstrable when certain single components of the product of an assimilation are given through external sense impressions, while others belong to earlier ideas. In such a case the assimilation may be demonstrated by the fact that certain components of the idea which are wanting in the objective impression or are there represented by components other than those actually present in the idea itself, can be shown to arise from earlier ideas. Experience shows that of these reproduced components, those are most favored which are very frequently present. Certain single elements of the impression are, however, after the analogy of the dominating elements in fusion, usually of more importance in determining the association than are the others, so that when these dominating elements are altered, as may be the case especially with assimilations of the visual sense, the product of the assimilation undergoes a corresponding change.

7. Among intensive compounds it is the auditory ideas which are most frequently the results of assimilation. They also furnish the most striking examples of the influence on present processes of earlier combinations which have become familiar through repetition. Of all the auditory ideas, the most familiar are as a rule the readily available ideas of words, for these usually receive more attention than other sound impressions. As a result the hearing of words is continually accompanied by assimilations; the sound impression is incomplete, but it is entirely filled out by earlier impressions, so that we do not notice the incompleteness. So it comes that not the correct hearing of words, but the misunderstanding of them, that is, the erroneous filling out of incomplete impressions through incorrect assimilations, is what generally leads us to notice the process. We may find an expression of the same fact in the ease with which any sound whatever, as for example the cry of an animal, the noise of water, wind, machinery, etc., can be made to sound like words almost at will.

8. In the case of intensive feelings we note the presence of assimilations in the fact that impressions which are accompanied by sense-feelings and elementary aesthetic feelings, very often exercise a second direct affective influence for which we can account only when we recall certain ideas of which we are reminded by the impressions. In such cases the association is usually at first only a form of affective association, and only so long as this is true is the assimilation simultaneous. The related ideational association which explains the effect is, on the contrary, usually a later process which must be classified as a form of successive association. For this reason it is often hardly possible, when we have clang impressions or color impressions accompanied by particular feelings, or when we have simple spatial ideas, to decide what is the immediate affective influence of the impression itself, and what is the influence of the association. As a rule, in such cases the affective process is to be looked upon as the resultant of an immediate factor and an associative factor which unite to form a single, unitary total feeling in accordance with the general laws of affective fusion (p. 177).

9. Association in the case of spatial ideas is of the most comprehensive character. It is somewhat less noticeable in the sphere of touch when vision is present, on account of the small importance of tactual ideas in general and especially on account of the small importance of touch for memory. For the blind, on the other hand, touch is the essential means of rapid orientation in space, as for example in the rapid reading of the blind-alphabet. The effects of assimilation are most strikingly evident when several tactual surfaces are concerned, because in such cases assimilation is easily betrayed by the illusions which may arise in consequence of some disturbance in the usual interrelation of the sensations. Thus, for example, when we touch a small ball with the index and middle fingers crossed, we have the idea of two balls. The explanation is obvious. In the ordinary position of the fingers the external impression here given actually corresponds to two balls, and the many perceptions of this kind which have been perceived before, exercise an assimilative action on the new impression.

In the sphere of visual sense perceptions, assimilative processes play a large part. They aid especially in the formation of ideas of the magnitude, of the distance, and of the three-dimensional character of visual objects. In this last respect they are essential supplements of immediate binocular motives for projection into depth. In this way we explain the correlation that exists between the ideas of the distance and ideas of magnitude of objects, as for example the apparent difference in the size of the sun or moon on the horizon and at the zenith. The perspective of drawing and painting also depends on processes of assimilation. A picture drawn or painted on a plane surface can appear three-dimensional only on condition that the impression arouses elements of earlier percepts which are assimilated with the new impression. This is most evident in the case of unshaded drawings that can be seen either in relief or in intaglio. Observation shows that these differences in appearance are by no means accidental or dependent on the so-called "power of imagination", but that there are always elements in the immediate impression which determine definitely the assimilative process. The elements which are thus operative are, above all, the sensations arising from the position and movements of the eye. Thus, for example, a linear design of a prism which is looked at with one eye only so as to eliminate the binocular data for the perception of depth, will be seen alternately in relief and in intaglio according as we fixate in the two cases the parts of the drawing which correspond ordinarily to a solid or to a hollow object. A solid angle represented by three lines in the same plane appears in relief when the fixation-point is moved along one of the lines starting from the apex; it appears in intaglio when the movement is in the opposite direction, that is, from the end of the line toward the apex. In these and all like cases the assimilation is determined by the rule that in its movement over the fixation-lines of objects the eye usually passes from nearer to more distant points, and when it fixates any point for a longer period of fixation, it generally turns toward those parts of the object which lie near at hand. Effects of assimilation are also noticeable in cases of misreading of words . These facts of misreading correspond fully with the facts of incorrect hearing described above (p. 258). In reading we overlook the misprints in a book. This is due, not so much to the fact that we fail to notice the wrong letter which is present, as to the fact that we see the right letter for the wrong one 2).

2) Assimilation processes which take place during reading may be studied most advantageously by means of the tachistoscope mentioned on page 236. This apparatus allows the words to be seen only for a short interval.
 
 

In other cases the geometrical optical illusions § 10 (19 and 20) which are due to the laws of ocular movements, produce as secondary effects certain ideas of depth which eliminate the contradictions between the retinal images from these figures and the illusions of length and directions which arise from the perceptions of the impressions. Thus, to illustrate, an interrupted straight line appears longer than an equal uninterrupted line (p. 140); as a result we tend to project the first to a greater depth than the latter. Here both lines cover just the same distances on the retina in spite of the fact that because of the different motor energy connected with their estimation, their lengths are perceived as different. An .elimination of the contradiction which thus arises is effected by the formation of different ideas of distance, for when one of two lines, the retinal images of which are alike, appears longer than the other, this longer line must, under the ordinary conditions of vision, belong to a more distant object. Again, to take another illustration, when one straight line is intersected at an acute angle by another line, the result is an overestimation of the acute angle, which overestimation sometimes gives rise, when the line is long, to an apparent bending of the line near the point of intersection (p. 140). Here, too, the contradiction between the true course of the line and the increase in the angle of intersection, is eliminated by the apparent projection of the line into the third dimension. In all these cases the perspective can be explained only as the assimilative effect of the elements of earlier ideas.

10. In none of the assimilations discussed is it possible to show that any former idea has acted as a whole on the new impression. Generally such action of a whole idea is impossible because we must attribute the assimilative influence to a large number of ideas, differing in many respects from one another. Thus, for example, a straight line which intersects a vertical at an acute angle, corresponds to innumerable cases in which an inclination of the line with its accompanying increase of the angle appeared as a component of a three-dimensional idea. But all these cases may have been very different in regard to the size of the angle, the length of the lines, and other attending circumstances. We must, accordingly, think of the assimilative process as a process in which not a single definite idea is operative, nor even a definite combination of elements from earlier ideas; but rather, as a rule, we must think of it as a process in which a great number of such combinations are operative. These many antecedents need agree only approximately with the new impression in order to affect consciousness.

We may gain some notion of the way in which this effect is produced from the important part which certain elements connected with the impression play in the production of the process, as for example the inner tactual sensations in visual ideas. Obviously it is these immediate sensation elements which serve to pick out from the great mass of ideational elements reacting on the impression, certain particular elements which correspond to themselves. The present sensations then bring these selected factors into a form agreeing with the form of the rest of the components of the immediate impression. At the same time it appears that not merely are our memory images relatively indefinite and therefore variable, but even the perception of an immediate impression may, under special conditions, vary within fairly wide limits. In this way the assimilative process starts primarily from elements of the immediate impression, chiefly from such as are of preeminent importance for the formation of the idea, as for example in visual ideas the sensations of ocular position and movement. These elements call up certain particular memory elements corresponding to themselves. These memories then exercise an assimilative effect on the immediate impression, and the impression in turn reacts in the same way on the reproduced elements. These separate acts are, like the whole process, simultaneous. For this reason the product of the assimilation is apperceived as an immediate, unitary idea. The two distinguishing characteristics of assimilation are, accordingly, 1) that it is made up of a series of elementary processes of combination, that is, processes which have to do with the components of ideas, not with the whole ideas themselves; and 2) that the united components modify one another through reciprocal assimilations.

11. On this basis we can explain without difficulty the main differences between complex assimilative processes, by the very different parts which the different factors necessary to such processes play in the various concrete cases. In ordinary sense perceptions the direct elements are so predominant that the reproduced elements are as a rule entirely overlooked, although in reality they are never absent and are often very important for the perception of the objects. These reproduced elements are much more noticeable when the assimilative effect of the direct elements is hindered through external or internal influences, such as indistinctness of the impressions or affective and emotional excitement. In all cases where the difference between the impression and the idea becomes, in this way, so great that it is apparent at once on closer examination, we call the product of the assimilation an illusion.

The universality of assimilation makes it certain that such processes occur also between reproduced elements, in such a way that any memory idea which arises in the mind is immediately modified by its interaction with other memory elements. Still, in such cases we have, of course, no means of demonstration. All that can be established as probable is that even in the case of so-called "pure memory processes", direct elements in the form of sensations and sense-feelings aroused by peripheral stimuli, are never entirely absent. In reproduced visual images, for example, such elements are present in the form of inner tactual sensations of the eye.

C. COMPLICATIONS.

12. Complications, or the combinations between unlike psychical compounds, are no less regular components of consciousness than are assimilations. Just as there is hardly an intensive or extensive idea or composite feeling which is not modified in some way through the processes of reciprocal assimilation between direct and reproduced elements, so almost every one of these compounds is at the same time connected with other, dissimilar compounds, with which it has some constant relations. In all cases, however, complications are different from assimilations in the fact that the unlikeness of the compounds makes the connection looser, however regular it may be, so that when one component is direct and the other reproduced, the latter can be readily distinguished at once. There is, however, another reason which makes the product of a complication appear unitary in spite of the easily recognized difference between its components. This is the predominance of one of the compounds, which pushes the other components into the obscurer field of consciousness.

If the complication unites a direct impression with memory elements of disparate character, the direct impression with its assimilations is regularly the predominant component, while the reproduced elements sometimes have an influence noticeable only through their affective tone. Thus, when we speak, the auditory word ideas are the predominant components, and in addition we have as obscure factors, direct motor sensations and reproductions of the visual images of the words. In reading, on the other hand, the visual images come to the front while the others become weaker. In general it may be said that the existence of a complication is frequently noticeable only through the peculiar coloring of the total feeling which accompanies the predominant idea. This is due to the power of obscure ideas to have a relatively intense effect through their affective tones on the attention (p. 244). Thus, for example, the characteristic impression of a rough surface, a dagger-point, or a gun, arises from a complication of visual and tactual impressions, and in the last case, of auditory impressions as well; but as a rule such complications are noticeable only through the feelings they excite.

D. SUCCESSIVE ASSOCIATIONS.

13. Successive association is by no means a process which differs essentially from the two forms of simultaneous association, assimilation and complication. It is, on the contrary, due to the same general causes as these, and differs only in the secondary characteristic that the process of combination, which in the former cases consisted, so far as immediate introspection was concerned, of a single instantaneous act, is here protracted and may therefore be readily divided into two acts. The first of these acts corresponds to the appearance of the reproducing elements, the second to the appearance of the reproduced elements. Here too, the first act is often introduced by an external sense impression, which is as a rule immediately united with an assimilation. Other reproduced elements which might enter into an assimilation or complication are held back through some inhibitory influence or other — as for example through other assimilations that force themselves earlier on apperception — and do not begin to exercise an influence until later. In this way we have a second act of apperception clearly distinct from the first, and differing from it in psychical content. The difference is the more essential, the more numerous the new elements which are added through the retarded assimilation and complication, and the more these new elements displace the earlier elements because of their differences in character.

14. In the great majority of cases the association thus formed is limited to two successive ideational or affective processes which are connected, in the manner described, through assimilations or complications. New sense impressions or some apperceptive combinations (§ 17) may then connect themselves with the second member of the association. Less frequently it happens that the same processes which led to the first division of an assimilation or complication into a successive process, may be repeated with the second or even with the third member, so that in this way we have an associational series. Generally, however, such a series is formed only under exceptional conditions. Such conditions arise when the normal course of apperception has been disturbed, as for example, in the so-called "flight of ideas" of the insane. In normal cases and under the ordinary conditions of life, serial associations hardly ever appear.

14a. Such serial associations may be produced most easily under the artificial conditions of experimentation, when the effort is purposely made to suppress new sense impressions and apperceptive combinations. But the process resulting in such cases differs from that described above in that the successive members of the series do not connect, each with its immediate predecessor, but all go back to the first, until a new sense impression or an idea with an especially strong affective tone furnishes a new starting point for the succeeding associations. The associations in the "flight of ideas" of the insane generally show the same typical tendency to return to certain predominant centers.

a. Sensible Recognition and Cognition.

15. The cases in which the ordinary form of association which is made up of two partial processes, may be most clearly observed arising out of simultaneous assimilations and complications, are the cases designated by the special names, sensible recognition and cognition. The qualification "sensible" is added when referring to these associative processes, to indicate, on the one hand, that the first member of the process is always a sense impression, and, on the other, to distinguish these associations from the logical processes of cognition.

The case of recognition which from the psychological point of view is the simplest, is that in which an object has been perceived only once and is recognized as the same when met a second time. If this second perception follows very soon after the first, or if the first was especially emphatic and exciting, the association usually takes place immediately, as a simultaneous assimilation. This process differs from other assimilations, which take place in connection with every sense perception, only in the characteristic accompanying feeling, the feeling of familiarity. Such a feeling is never present except when there is some degree of "consciousness" that the given impression has been received before. It is, therefore, evidently one of those feelings which come from the ideas obscurely present in consciousness. The psychological difference between this and an ordinary simultaneous assimilation must be looked for in the fact that at the moment when, in the apperception of the impression, the assimilation takes place, there arise in the obscure regions of consciousness some components of the original idea which do not enter into the assimilation. The relation of these obscure components to the elements of the idea which is apperceived, finds expression in the feeling of familiarity. The unassimilated components may be elements of an earlier impression which were so different from certain elements of the new, that they could not be assimilated, or, and this is usually the case, they may be complications which were clear before, but now remain unobserved. This influence of complication explains how it is that the name of a visual object, for example the proper names of persons, and often other auditory qualities, such as the tone of voice, are very great helps in recognition. To serve as such helps, however, they need not necessarily be clear ideas in consciousness. When we have heard a man's name, the recognition of the man the next time we meet him may be aided by the name without our calling it clearly to mind.

16. The observations described show what are the conditions under which a recognition may pass from a simultaneous into a successive association. If a certain interval elapses before the elements of the earlier idea which gradually rise in consciousness, can produce a distinct feeling of familiarity, the whole process divides into two acts, into the act of perception and the act of recognition. Perception depends on the ordinary simultaneous assimilations only, while in recognition, the obscure unassimilated elements of the earlier idea show their influence. The line of division between these partial processes is, accordingly, more distinct the greater the difference between the earlier impressions and the new one. In a case of marked difference not only is there usually a long period of noticeable inhibition between perception and recognition, but certain additional apperceptive processes, namely the processes of attention which take part in the act of recollection, also come to the aid of the association. As a special form of this kind of process we have the phenomenon called "mediate recognition". This consists in the recognition of an object, not through its own attributes, but through some accompanying mark, which stands in a chance connection with it, as for example when a person is recognized because of his companion. Between such a case and a case of immediate recognition there is no essential psychological difference. For even those characteristics which do not belong to the recognized object in itself, still belong to the whole complex of ideational elements that help in the preparation and final carrying out of the association. And yet, as we should naturally expect, the retardation which divides the whole recognition into two ideational processes, and often leads to the cooperation of voluntary recollection, generally appears in its most evident form in mediate recognitions.

17. This simple process of recognition which takes place when we meet again an object that has been perceived once before, is a starting point for the development of various other associative processes, for processes which like recognition stand on the boundary between simultaneous and successive associations, and for processes in which there is a more marked degree of that retardation in the formation of associations and complications which leads to a successive rather than simultaneous occurrence of the processes. Thus, the recognition of an object which has often been perceived is easier and, therefore, as a rule it is an instantaneous process. It is also more like the ordinary assimilation because the feeling of familiarity is much less intense. Sensible cognition differs generally but little from the recognition of single familiar objects. The logical distinction between the two concepts consists in the fact that recognition means the establishment of the individual identity of the newly perceived object with a formerly perceived object, while cognition is the subsumption of an object under a familiar concept. Still, there is no real logical subsumption in a process of sensible cognition any more than there is a fully developed class-concept under which the subsumption could be made. The psychological equivalent of such a subsumption is to be found in this case in the mere process of associating the impression in question with an indefinitely large number of objects. This presupposes an earlier perception of various objects which agree only in certain particular properties, so that the process of cognition approaches more nearly to the ordinary assimilation in its psychological character, the more familiar the class to which the perceived object belongs, and the more the object agrees with the most common objects of this class. In equal measure the feelings peculiar to the processes of cognition and recognition decrease and finally disappear entirely, so that when we meet very familiar objects we do not speak of a cognition at all. The process of cognition is noticeable only when the assimilation is hindered in some way, either because the perception of the class of objects in question has become uncommon or because the single object shows some unique characteristics. In such a case the simultaneous association may become successive by the separation of perception and cognition into two successive processes. Just in proportion as this happens, we have a specific feeling of cognition which is indeed related to the feeling of familiarity, but, as a result of the different conditions for the rise of the two, differs from the feeling of familiarity especially in its temporal course.

b. Memory processes.

18. There is another direction, essentially different from that just described, along which the process of recognition may develop. This shows itself when the hindrances to immediate assimilation which give rise to the transition from simultaneous to successive associations, are so great that the ideational elements which do not agree with the new perception, unite — either after the recognition has taken place or even when there is no such recognition whatever — to form a special idea referred directly to an earlier impression. The process which arises under such circumstances is a memory process and the idea which appears is a memory idea, or memory image.

18a. Memory processes were the ones to which association psychology generally limited the application of the concept association. But, as has been shown, these are associations which take place under especially complicated conditions. The erroneous view of association psychology rendered an understanding of the genesis of an association impossible from the first, and it is easy to see that the doctrine accepted by the associationists is limited essentially to a logical rather than a psychological classification of the association products which are to be observed in memory processes. An insight into the character of the more complex processes is possible, however, only through a study starting with the simpler associative processes, for the ordinary simultaneous assimilations and simultaneous and successive recognitions present themselves very naturally as the antecedents of memory associations. But even simultaneous recognition itself is nothing but an assimilation accompanied by a feeling which comes from the unassimilated ideational elements obscurely present in consciousness. In the second process these unassimilated elements serve to retard the process, so that the recognition develops into the primitive form of successive association. The impression is at first assimilated in the ordinary way, and then again in a second act with an accompanying feeling of recognition which feeling serves to indicate the greater influence of certain reproduced elements. In this simple form of successive association the two successive ideas are referred to one and the same object, the only difference being that each time some different ideational and affective elements are apperceived. With memory associations the case is essentially different. Here the elements of the earlier impressions which are different from those in the present impression predominate, and the first assimilation of the impression is followed by the formation of an additional idea. This idea is made up of elements of the present impression and of elements belonging to certain earlier impressions, which earlier impressions are suitable for the assimilation because of certain of their components. The more the elements of the earlier impression which differ from the elements of the present impression, predominate, the more the second idea differs from the first, or, on the other hand, the more the like elements predominate, the more the two ideas will be alike. In any case the second idea is always a reproduced idea and distinct from the new impression as an independent compound.

19. The general conditions for the rise of memory images may also exhibit shades and differences which correspond to the differences which appear in the processes of recognition and cognition (15, 17). Thus the recognition of an object perceived once and the recognition of an object familiar through frequent perceptions, and finally, the cognition of an object that is familiar in its general class-characteristics may all become sources of various modifications in memory processes.

Simple recognition becomes a memory process when the immediate assimilation of the impression is hindered by elements that belong, not to the object itself, but to circumstances which attended its earlier perception. Just because the former perception occurred only once, or at least only once so far as the reproduction is concerned, these accompanying elements may be relatively clear and distinct and sharply distinguished from the surroundings of the new impression. In this way we have transitional forms between recognition and remembering; the object is recognized, and at the same time referred to a particular earlier sense perception the accompanying circumstances of which add a definite spatial and temporal relation to the memory image. The memory process is especially predominant in those cases in which the elements of the new impression which gave rise to the assimilation are entirely suppressed by the other components of the image, so that the associative relation between the memory idea and the impression may remain entirely unnoticed.

19a. Such cases have been spoken of as "mediate memories", or "mediate associations". Still, just as in the case of "mediate recognitions", so here, we are dealing with processes which are fundamentally the same as ordinary associations. Take, for example, the case of a person who, sitting in his room at evening, suddenly remembers without any apparent reason a landscape through which he passed many years before; examination shows that there happened to be in the room a fragrant flower which he saw for the first time in that landscape. The difference between this and an ordinary memory process in which the connection of the new impression with an earlier experience is clearly recognized, obviously consists in the fact that here the elements which recall the idea are pushed into the obscure background of consciousness by other ideational elements. The not infrequent experience, commonly known as the "spontaneous rise" of ideas, in which a memory image suddenly appears in our mind without any assignable cause, is in all probability reducible in every case to such latent associations.

20. Memory processes that develop from recognitions which have been often repeated and from cognitions, are, in consequence of the greater complexity of their conditions, different from those connected with the recognition of objects perceived but once. When we perceive an object which is familiar either in its own individual characteristics or in the characteristics of its class, the range of possible associations is incomparably greater, and the way in which the memory processes shall arise from a particular impression depends less on the single experiences which give rise to the association, than it does on the general disposition and momentary mood of consciousness, and on the interference of certain active apperceptive processes together with the intellectual feelings and emotions which are connected with these processes. Word ideas are important aids to association. These ideas are in many cases connected with individual objects (proper names), but they are especially important when they refer to class characteristics of ideas (class names). With conditions which are so varied, it is easy to see that as a general thing it is impossible to calculate beforehand what the association in any given case will be. As soon as the act of memory is ended, however, the traces of its associative origin seldom escape careful examination, so that we are justified in regarding association as the universal and only cause of memory processes under all circumstances.

In thus deriving memory from association, it is never to be forgotten that every concrete memory process is by no means a simple process, but is made up of a large number of elementary processes, as is apparent from the fact that every such process is produced by a psychological development of its simple antecedents, namely, the simultaneous assimilations. The most important of these elementary processes is the assimilative interaction between some external impression and the elements of an earlier psychical compound, or between a memory image already present and such elements. Connected with this there are two other processes which are characteristic of memory processes; one is the hindrance of the assimilation by unlike elements, the other is to be found in the assimilations and complications

20a. It is obvious that the usual classification, which makes all memory processes associations by either similarity or contiguity, is entirely unsuitable if we attempt to apply it to the modes of psychological genesis which these processes manifest. On the other hand, it is too general and indefinite if we try to classify the processes logically according to their products, without reference to their genesis. In the latter case the various relations of subordination, superordination, and coordination, of cause and end, of temporal succession and existence, and the various kinds of spatial connection, find only inadequate expression in the very general concepts "similarity" and "contiguity". When, on the other hand, the origin is studied, every memory process is found to be made up of elementary processes which may be called partly associations by similarity, partly associations by contiguity. The assimilations which serve to introduce the process and also those which serve to bring about the reference to a particular earlier experience at its close, may be called associations by similarity. But the term "similarity" is not exactly suitable even here, because it is identical elementary processes which give rise to the assimilation, and when an identity of elements does not exist, such identity is always produced by reciprocal assimilation. In fact, the concept of "association by similarity" is based on the presupposition that composite ideas are permanent psychical objects and that associations take place between these finished ideas. The concept of association by similarity must be rejected when once this presupposition is given up as entirely contradictory to psychical experience and fatal to a proper understanding of experience. When certain products of association, as for example two successive memory images, are similar, this likeness is always reducible to processes of assimilation made up of elementary combinations resulting from identity or contiguity. The association through identity may take place either between components which were originally the same, or between those which have gained this character through assimilation. Association by contiguity is the form of combination between those elements which hinder the, assimilation, thus dividing the whole process into a succession of two processes, and also contributing to the memory image those components which give it the character of an independent compound, different from that of the impression which gave rise to it. The joint action of associations of identity and contiguity is very clearly the most natural explanation, especially in the case of the simplest forms of memory association, namely in those forms which are made up of simple sensory impressions. Indeed, it is only by means of this joint action of the two forms of association that we can give any natural explanation of the facts in question. Thus, when a yellow color impression calls up in the mind the similar color orange, the explanation offered by the pure theory of association by similarity is that the close similarity between the two colors produced the association. On the other hand, the pure theory of contiguity explains the same fact on the ground that the two impressions have been seen next to each other an indefinite number of times, in the rainbow, in the spectrum, and in the shadings of painted surfaces. In reality the facts are not as stated in either of these explanations. It is true, rather, that colors, like tones, form a continuous sensation series within which the impressions standing nearest to each other are always most closely associated on account of the conditions of their natural rise and variation. There are always brought into consciousness with any given color-impression, other associated colors, especially those which lie nearest to the given color. This is possible only because the present color calls up first the color which is identical with itself in some memory complex, and then calls up through this identical color the one next to it in the memory complex. Yellow, for example, can call up the yellow which has been seen before in the spectrum (association by identity) and then through this first process, may further call up the neighboring orange (association by contiguity). It is especially obvious in this case that there must be a combination of the two forms of association, because the two stages in the complete association are much more distinctly separate than in the case of complex ideas where the two stages unite at once into a single composite process.

21. All the results of memory associations, so far as they are related to earlier impressions, are commonly grouped together under the name memory. This concept memory originated in popular psychology and was then carried over into the now abandoned faculty-psychology. Memory must, of course, in every particular case be subjected to a special analysis to show what are the elementary association processes involved in the special phenomena under consideration, and what are the particular effects of these association processes. Such analysis finds the simplest conditions for its application in those cases in which the memory associations take place between simple impressions, or at least between impressions which arise under relatively simple and uniform conditions. Thus, one may investigate the memory for tone sensations, or for simple visual objects, by measuring the accuracy of such memory in terms of the clearness with which an earlier impression is recognized after the lapse of a given interval. As a result of such measurements it appears that immediately after an impression is given, its reproduction is relatively accurate. Very soon (in the case of tones after two seconds, in the case of simple visual objects after an interval somewhat, but not very much, longer) reproduction reaches its maximum of accuracy and then begins to decrease with gradually lessening rapidity until, finally, (after about 60 seconds) it reaches a point at which it remains approximately constant for a long time. In the course of this general decrease in accuracy of reproduction, there appear successive periods of fluctuating accuracy which probably are related to the fluctuations of attention already mentioned (p. 236).

Of special interest for the investigation of the relation of intervals of time to memory processes are the facts of time memory. By time memory we mean the memory for temporal intervals. This form of memory can be investigated with the highest degree of exactness, just as can the attributes of time ideas in general, by using so-called empty intervals marked off by auditory impressions. Through investigations of this kind it appears that the relation between the memory image of an interval and the objective length of this interval depends, in the first place, on the length of the interval in question, and, in the second place, on the amount of time that elapses between the giving of the impression and the formation of the memory image. The length of the given interval affects the process in accordance with the general rule, that short intervals are overestimated in memory and long intervals are underestimated. Between these two forms of false estimation lies an indifference-interval for which the remembered interval is, on the average, equal to the given interval. When the reproduction follows the impression very quickly this indifference-interval is 0.5—0.6 sec. If the interval is increased in length there appears here also a kind of periodic recurrence of exact estimations, for which the regular rule is that all the multiples of the indifference-interval are more accurately estimated than are the intervals lying between these multiples. This periodic recurrence of exact estimations is probably due to the fact that longer intervals have to be broken up into groups of short intervals in order to be grasped in consciousness as single wholes. In such division and grouping the indifference-interval presents itself as the standard simple unit. The fact of periodic accuracy in estimation is also doubtless connected with the above described processes of involuntary rhythmical subdivision of long time intervals (p. 169). When the period between impression and reproduction is longer, the exactness of estimation suffers a general decrease, just as in the case of the reproduction of qualitative tone sensations and light sensations. It finally reaches a minimum at which it continues for a relatively long period of time at an approximately constant level. With a lengthening of the period between impression and reproduction the reproduced interval becomes more and more clearly shortened in comparison with the original interval. No very exact determinations have been made of these last described facts. They are, however, familiar from every-day experience.

21a. The experimental analysis of the facts of memory has engaged a great deal of attention in modern psychology, both because of the theoretical interests which attach to the problem and also on account of the practical pedagogical interests which are involved, especially in the application of the results of such experiments to complex ideas and series of ideas. The methods which have been developed for such experiments may be divided into two classes; the learning methods and the methods of recognition. Each of these methods exhibits certain sub-forms which may be distinguished as follows. The learning method in the narrower sense of the word, or the memorizing method as we may call it, consists in requiring an observer to repeat nonsense syllables, words, or sentences, under definite and voluntarily variable conditions until he has them learned by heart. A second looser form of the learning method may be designated as the guessing method. It consists in presenting a certain combination of syllables once or several times and repeating later only a part of the combination, allowing the part which has been omitted to be reproduced in memory. In the recognition methods, on the other hand, an impression is given to the observer and after a given period, either the same impression or one differing from the first is presented. If the observer is able to recognize correctly the second impression as like the first, or different from it, or if he fails to give a correct answer, the degree of his memory can be determined. Two types of this method can be distinguished. One may be designated as the method of modified repetitions and the other as the method of identical series. The method of modified repetition is especially suited to the testing of memory for simple sensations and for relatively simple and temporal ideas. It consists in presenting at first an impression which is known as the normal impression, and then after a time presenting a second impression which is known as the comparison impression. The comparison impression is gradually modified in successive cases until the point is reached at which the difference can be just recognized, that is, until the so-called difference-threshold (§ 17, 10a) is reached. The method of identical series is especially appropriate for the study of complex groups of experiences such as series of words, numbers, etc. It consists in presenting a series of impressions which are separated by short intervals from each other and after an appreciable interval of time allowing the same series to appear again in order that the number of correctly recognized impressions may be determined. Of the methods which have been described, the "learning methods", especially in the form of the memorizing methods, are the most commonly used. This method is open, however, to the criticism that the psychological conditions under which the experiments are performed are very much more complex than are those which are necessary for the recognition methods. As a result, the measure of the work done by memory, which has been utilized in this method, has not infrequently been fundamentally wrong. When, for example, the measure for memory work has been sought in the time or number of repetitions necessary for complete memorizing, the presupposition has been that mental work increases in proportion to the time. This supposition is, however, as a matter of fact, not correct, because mental work, as is shown by a great number of experiences, decreases rather than increases with an increase in the time.

Through the application of these methods the various factors which are recognized in ordinary experience as determining the ability to remember, can be varied in a more or less exact way. Such factors are the number and duration of presentations of the impressions which are to be remembered, the interval between these presentations, and, finally, the length of each of the series which is to be remembered. Negative factors may also be measured, such as the gradual disappearance of memory traces and the effect of the interval between presentation and reproduction. The results of measurements on the influence of these different factors show 1) that the amount which can be retained in memory increases at first rapidly and then slowly with the increase in the number and duration of the single impressions. 2) The amount which can be remembered increases in the absolute with every increase in the length of the series and decreases relatively, that is, decreases in the number of retained impressions, as compared with the total number of impressions which are presented. 3) Those factors in a series which are accompanied by a strong feeling, as for example the members of a series of sounds which are accented, are more easily retained than those which are unaccented. The consequence of this relation is the familiar fact that rhythmically divided series are in general easier to remember than others. 4) The length of the interval between the various presentations may be of a certain length which constitutes the optimum or interval of greatest advantage for memory. Above and below this optimum the efficiency of memory decreases. This can easily be explained when we consider that there is a certain opposition between memory and the tendencies of fatigue and recovery on the one hand and the process of forgetting on the other. The process of forgetting consists in the gradual disappearance of the disposition to reproduce the impression. The most favorable interval for memory will, accordingly, lie at the point where recuperation and fatigue balance each other, and where at the same time the disposition toward reproduction is not too much reduced. 5) As the interval between the presentation of the impressions and their reproduction increases, the amount which can be retained diminishes at first very rapidly and then more gradually. All of these results, especially those which show the influence of feeling and those which emphasize a relation between fatigue and voluntary effort, show that memory and all processes of recollection are complex results made up of elementary associations and of those apperceptive processes which are to be considered in later paragraphs. In this respect, memory is analogous to a large number of other complex forms of mental work, such, for example, as reading, writing, counting, and using numbers for complex processes of calculation.

22. The character of memory ideas is intimately connected with the complex nature of the memory processes. The description of these ideas as weaker, but otherwise faithful, copies of the direct sense perceptions is as far out of the way as it could possibly be. Memory images and sense perceptions differ, not only in quality and intensity, but most emphatically in their elementary composition. We may diminish the intensity of a sensible impression as much as we like, but so long as it is perceptible at all it is an essentially different compound from a memory idea. The incompleteness of the memory idea is much more characteristic than the small intensity of its elements. For example, when I remember an acquaintance, the images I have of his face and figure are not mere obscure reproductions of what I have in consciousness when I look directly at him, but most of the features do not exist at all in the reproduced ideas. Connected with the few ideational elements which are really present and which can be but little increased in number even when the attention is voluntarily concentrated upon the task, are certain factors added through contiguity and certain complications, such as the environments in which I saw my acquaintance, his name, and finally, and more especially, certain affective elements which were present at the meeting. These accompanying components are what make the image a memory image.

23. There are great differences in the effectiveness of these accompanying elements and in the distinctness of the sensation elements of the memory image in the cases of different individuals. Some persons locate their memory images in space and time much more precisely than do others; the ability to remember colors and tones is also very markedly different. Very few persons seem to have distinct memories of odors and tastes; in place of these most of us have, as substitute complications, accompanying motor sensations of the nose and taste-organs.

These differences between different individuals are all referred to as differences in "memory". The concept memory is, thus, a supplementary concept which is very useful in giving clear expression to these individual differences in the memory processes. It must, however, never be forgotten that the term always refers to what is in reality a series of processes, and that in each particular case a special explanation of the facts is required. We speak of a faithful, comprehensive, and easy memory, or of a good spatial, temporal, and verbal memory, etc. These expressions serve to point out the different directions in which, according to the original disposition or habit of the person, the elementary assimilations and complications occur.

One important phenomenon among the various differences referred to, is the gradual weakening of memory with old age. The disturbances resulting from diseases of the brain agree in general with the results of this weakening of memory through age. Both are of special importance to psychology because they exhibit very clearly the influence of complications on memory processes. One of the most striking symptoms of failing memory, in both normal and pathological cases, is the weakening of verbal memory. It generally appears as a lack of ability to remember, first proper names, then names of concrete objects in the ordinary environment, still later abstract words, and finally, particles that are entirely abstract in character. This succession corresponds exactly to the possibility of substituting in consciousness for single classes of words other ideas that are regularly connected with them through complication. This possibility is obviously greatest for proper names, and least for abstract particles, which can be retained only through their verbal signs.
 
 

References. Bain, The Senses and the Intellect, 4th ed., p. 335 seq. (The part on Intellect is a thoroughgoing exposition of associationism). Wundt, Bemerkungen zur Assoziationslehre Phil. Stud., vol. 7, and Grundz. 5th ed., vol. III Chap. 19; Lectures, lectures 19 and 20. On Recognition, Discussion of Association by Similarity and Contiguity: Hoeffding, Vierteljahrsschr. f. wiss. Philos., vols. 13 and 14, and Phil. Stud., vol. 5. Lehmann, Phil. Stud., vols. 7 and 8. On Forms of Association, and Association Time: Trautscholdt, Phil. Stud., vol. 1. Aschaffenburg, Kraepelin's Psychol. Arbeiten, vols. 1, 2, 3. On Mediate Association: scripture, Phil. Stud., vol. 7. cordes, Phil. Stud., vol. 17. On Memory: Wolfe, (on memory for tones), Phil. Stud., vol. 3. Radoslawow (on memory for simple visual objects), Phil. Stud., vol. 15. Ebbinghaus, Das Gedächtnis, 1885. Müller and Schumann, Zeitschr. f. Psychol., vol. 6. Müller and Pilzecker, Supplement to Zeitschr. f. Psychol., 1900. Binet and Henri, Année psychol., vol. I, 1894. Bolton, franz, Houston, W. G. Smith, Psych. Review, vol. III, 1896. Meumann, Oekonomie und Technik des Auswendiglernens, 1901. Reuther, Psychol. Stud., vol. I (with complete bibliography). On Memory for Time: Vierordt, Der Zeitsinn, 1868. Kollert, Estel, Glass, Phil. Stud., vols. 1, 2, 4. Meumann, Phil. Stud., vols. 8—12. On Complex Phenomena of Memory: Kraepelin, Psycholog. Arbeiten, vols. 1—4. Wundt, Grundz., 5th ed., vol. III pages 581 seq. On Diseases of the Memory: Ribot, Diseases of the Memory. Wundt, Völkerpsychologie, vol. I, Pt. 1, chap. 5 (on word memory).