Movement and Mental Imagery
Chapter 7: Simultaneous Movement Systems
Margaret Floy Washburn
A SIMULTANEOUS movement system is one where the movements, instead of being dependent each on the one which preceded it in time, are mutually interdependent and occur together. The linking by associative dispositions takes place in both directions. or in all directions if the system is composed of more than two movements. Not only are the kinaesthetic excitations produced by the movement A the necessary stimulus to movement B, but the excitations resulting from B are the stimulus to movement A. It is evident that in such systems the component movements must be compatible. They must be of such a nature that they can be performed at the same time. Thus two articulatory movements could not enter into a simultaneous movement system: you cannot pronounce t and b at the same moment. But the movements involved in pronouncing a word and those involved in looking at the word printed, or calling up a mental image of its appearance, are compatible, and could enter into a simultaneous system. In our perceptions of objects, some of the movements are compatible and enter into simultaneous systems, while others do not.
An originally simultaneous system is formed by the actual occurrence of the movements together, and strengthened by each repetition of their synchronous occurrence. Meyer (83), in 1910, following a method suggested by Ach, tried to bring about the formation of simultaneous systems under experimental conditions. The observer was shown a series of cards, each card carrying two groups of simple figures, and each shown for the very short instant of 135 thousandths of a second: the interval between each card and the next was the same very brief time. In the test, the observer was shown one half of the card and required to draw from memory the figures that were
( 129) on the missing half. The brief exposure was intended to keep the observer from attending successively to the two halves of the card, and force him to attend to them simultaneously. The short interval between exposures would, it was thought, in a similar way keep him from attending successively to the two parts of the memory after-image of the card. Learning and correct recall proved to be possible by this method. It is evident that only those parts of the card could really be simultaneously attended to which did not involve incompatible movements. It is also evident that most of the learning which we do in ordinary life is performed under conditions very different from these.
Where the movements which enter into the system are very simple, they may form themselves into simultaneous systems, probably, by actually occurring together. But in all the concrete examples one can think of where simultaneous systems are formed, one comes to do the two things together, or attend to them together, through a preliminary process of attending to them alternately. The leg and arm movements of a practiced swimmer form a simultaneous system, but in learning to swim they are performed alternately, and even after one really begins to swim one has difficulty in not attending to them alternately. The static movement system of holding the head perfectly still and the phasic system of the arm swing are formed into a simultaneous system when one learns to make a golf stroke, but their simultaneous performance is possible only by having attended to them alternately as a beginner. When one has perfectly learned a language, the sight of a word and its meaning are simultaneous; but in the beginning one attended alternately to the printed word and the idea of its meaning. When associative dispositions leading in both directions are formed between two movements, so that either one can excite the other successively, if they are compatible movements the tendency is apparently always for them to form simultaneous systems. The greatest aid to the formation of simultaneous systems is the association of each of the two movements with
(130) a common third. If C and B are compatible movements, and associative dispositions have been formed between A and C and between A and B, then when A occurs, there is a tendency for B and C to be simultaneously excited. Thus Miller and Pilzecker (90 found that if a syllable such as 'bez' were learned at one time in connection with the syllable 'gaf,' and at another time with the syllable 'jip,' when 'bez' was later given the observer might respond with a kind of hybrid like 'gap.' Both of the former associates of 'bez' were reproduced, but the incompatible movements were forced out and a compromise was reached in the combination of elements from both syllables. This method of forming simultaneous systems by linking each of several movements to a common motor response is of the utmost importance for our experience. It is precisely thus that those simultaneous movement systems are formed on which are based what we call perceptions of objects, as well as ideas of concrete objects and of abstractions.
The perception of an object consists of a number of sensations, some of which are peripherally excited, that is, caused by the activity of our sense organs at the moment, while others are centrally excited, that is, the revival of former sense experiences. A piece of ice looks smooth, white, hard, and cold: we realize that only the whiteness of it is the result of present sense stimulation, and that the hardness, smoothness, and coldness are the effects of former experiences with the senses of touch and temperature. Now the way in which we form such combinations of sensations into perceptions is evidently not so much by adding bit to bit to form a mosaic, as by digging one bit after another out of an original whole. In first making acquaintance with an object we respond to it as an undifferentiated whole: later we come to make specialized responses to various parts and aspects of it; but it is the fact that it can be still responded to as a whole that keeps these specialized movements together in a single system, and thus gives the object its unity. An orange, or a chair, or a tree, is a single object, and not a mere aggregate of qualities and parts, because it can be
( 131) reacted to as a whole, and because every one of the movements involved in attending to its parts is associated with the movement of reacting to the whole object. Some of these movements which are associated with various parts or aspects of an object are compatible: the fact that they lead to the common outlet of a movement made to the whole object may thus transform them into simultaneous systems. Others are incompatible, and must be united in successive rather than simultaneous systems; thus, for instance, one cannot attend simultaneously to both ends of a pen, but the whole object is a unity because of the possibility of responding to the whole of it by a single movement or movement system. Of the motor responses thus linked together by their common outlet, the compatible ones become simultaneous systems, relating to those parts of the object that can be attended to together; while the incompatible ones become successive systems, either reversible or irreversible, relating to those parts of the object which must be attended to in succession.
On a non-motor theory of association, which would make the formation of an associative disposition result from a lowering of resistances at the synapses on a pathway directly connecting two sensory centres, and resulting from their simultaneous activity, it is commonly held that all associations are based on simultaneous rather than on successive experience. Thus Offner (98) says that when two successively occurring impressions are associated, it is because the second impression is simultaneous with the memory after-image or perseverative process of the first. Wohlgemuth (154) has recently maintained on the basis of experimental results that all association is between simultaneously occurring experiences. He required his observers to form an association between a color and a form (a) when the form was colored; (b) when a black shape was shown on a colored ground; (c) when a colored field was shown alongside of a black shape on a white ground; (d) when a colored field was shown followed by a black shape on a white ground, or vice versa. It was found that "the more the members of a group
( 132) are apperceived as a whole, the stronger their association with each other," and from this it is concluded that " all associations are due to simultaneity." It is evident that the superiority of the associations in impressions that were apperceived as a whole, for instance, of the association between color and shape when the shape is itself colored, is due to the fact that besides the simultaneous responses made to the two factors of color and shape, there is a single response to the whole impression; thus the simultaneous impression has the great aid of a single unifying motor reaction.
Just as the unity of the perception of an object depends on the possibility of making a single movement or movement system in response to the object as a whole, so of course the possibility of reviving a memory image of an object as a whole depends on such a single response. Upon the strength of the simultaneous and successive systems thus linked together by their common outlet will depend the completeness and accuracy with which a memory image can be analyzed into details corresponding with the original. It is not necessary to emphasize the importance of a word or name as furnishing a convenient unifying response to the whole object. One cause of the low stage of intellectual development of animals is the very limited extent to which they can hold together the parts of their experience by making reactions to a whole group of such parts as a single group. Beasts that, like the monkey, the elephant, and the raccoon, have grasping organs with which to move things about have a great advantage so far as the formation of perceptions of objects is concerned. But the unrivalled instrument of unifying motor responses is of course language. By its help we can not only hold together into a system our responses to the various aspects of a single concrete object, but we can in a similar way form systems out of the aspects or features which a number of objects have in common. We can form those systems which are the bases of general ideas or concepts, such as the concept dog or animal. There is almost no limit to the complexity of the system combinations which
( 133) can be formed through having a single motor outlet for an entire combination.
The process of learning practically always involves both simultaneous and successive movement systems. In trying to form a new successive system, such as a memorized series of nonsense syllables, one is of course really working not with single movements, but with systems of movements. Each syllable is not merely pronounced, which is a process involving comparatively simple simultaneous systems for each vowel and consonant, but there is very likely present a visual image of the syllable, betraying, if our theory is true, the presence of tentative eye movements; and there may be also other 'aids' to the learning process, each one involving its own movement systems.
When a number of stimuli, each calling for its own motor response, act together upon the organism, as of course they are constantly doing, one set of responses is prepotent, and other incompatible responses are completely inhibited. It goes without saying that the inhibited responses can form no associative connections. That which is wholly unattended to does not form associations. It is true that Scripture (123), working in the Leipzig laboratory more than twenty years ago, found that unattended-to parts of a picture, letters or small colored squares in the corner, for example, did when later shown alone recall the picture in a certain number of cases too large to be due to chance. But there is always a doubt as to whether, in these few cases, the elements which did the recalling were really unattended to, and Howe (53), repeating Scripture's experiments in the Cornell laboratory, failed to confirm his results. Ordahl (101) tested the question in 1911 by a better method: she had her observers learn the middle one of three ten-syllable series arranged in parallel vertical columns. After a short interval another set of three was presented whose middle series was one of the two side series in the preceding set. Would this middle series be better learned now because it was present, though unattended to, at the side of the series previously learned? The
( 134) results showed that it was not any better learned; its presentation without attention had formed no associative dispositions.
A different problem, however, is furnished by the more ordinary cases of distraction. A thing to which we give no attention cannot be said to distract us at all: distraction occurs when our attention is forced to alternate between the movement system we are acquiring and another movement system quite unconnected with it. The effect of distraction seems to be greater, the more the two systems are, not merely disconnected with one another, but actually incompatible with one another. When they are really incompatible, if attention happens to be given to the distracting process the one to be learned is wholly shut out. Von Sybel (136a) says that distraction diminishes visual learning in favor of auditory-motor learning. Thus its effect in general seems to be that of simplifying the movement system that is formed. It prevents the stirring up of the more complex systems involved in visual imagery and in meaning connections. There are two ways in which one might conceivably explain this effect. One would be by supposing that only a limited amount of nervous energy is available in the cortex, and that if a part of it is occupied in the movements of the distracting system, there will not be enough left for other complex movement systems. The other explanation would rest on the possibility that the distracting system might contain movements incompatible with those of the system to be learned; and of course the more complex the system to be learned, the greater the likelihood that some of its parts would be incompatible with those of the distracting system. Both of these suppositions are very likely true.
The most concrete problems connected with the mutual relations of movement systems concern the way in which large systems are secure from distraction. The ability of any process that is going on in the mind to keep itself from being disturbed by distractions depends mainly on two factors. The first is the amount of effort or resolution that is put forth to 'keep one's mind on' the process to be attended to. This factor will be
(135) discussed in the chapter on "The Problem or Purpose." When the task is one of forming a new movement system, of learning, for instance, a series of nonsense syllables, nothing but effort will enable the material that is in process of being learned to hold attention against distraction. It will be suggested in the chapter just referred to that it is the naturally persistent character of the attitude of activity or effort which enables the task associated with it to be held to, despite distracting influences. But often, of course, it is the interest of the material that holds distraction aloof. Now, interest always means that some large, already formed movement system is back of the material attended to, and that its momentum, so to speak, is such that the associative dispositions not involved in it will be inhibited. The advantage is always with the older and complexer systems. Thus Toll (139) had his observers learn lists of words in which names of mammals alternated with names of American cities. The results showed that the tendency of one word was almost never to recall the word that followed it in the series, but rather to recall, if it were an animal name, the name of another animal; if it were a city name, the name of another city. The older and complexer systems had the right of way.
Levy-Suhl (70) used the ability of a well-established system of associative dispositions to resist distraction as a test of the normality of the mind. Insane and normal persons were allowed to start a train of ideas suggested by themselves, and when it was well under way, they were interrupted by pronouncing to them an irrelevant word. Only a hopelessly abnormal mind meekly accepted the distraction and followed the new line of thought without reference to the old. A curious instance of the effect of the problem or purpose involved in an experiment is shown by the fact that Baldwin (8), trying experiments by practically the same method, found that his observers, who were all normal, usually accepted the new train of thought almost at once. In his experiments the first train of thought, on which the interruption broke, was suggested by the experimenter instead of being self-suggested. The whole
( 136) attitude of the observers was therefore that of attending to anything that might come from the experimenter, instead of really allowing an associative system to get possession.
An experiment of Poppelreuter's (11e) well illustrates the independence of complex systems. He suggests that we take sentences from different stories and intermingle them, and that the result be read straight through from beginning to end. It is a fact that one will have at the end of the reading the two narratives side by side and each almost as clear as if it had been read without the other. Each system of movements has appropriated that which belonged to it, and each has in turn yielded place to the other. Thus we can carry on a conversation with a fair degree of intelligence while we read a story, and the story ideas do not mix with the conversation ideas except very occasionally.
Quite as important as the distracting effects of one movement system on another are the favorable effects of one system on another. One of the simplest ways in which we can make associative dispositions help each other is to attach some distinguishing mark to each member in a series of movements that are to be formed into a successive movement system. Thus each member of the series becomes a simultaneous system of some complexity, and the various members get more individuality. A few examples will make this clear. Gordon (44) found that nonsense syllables could be better learned if each syllable was printed on a ground of different color; Peterson (107) that figures were better learned if they differed both in form and in color. The introduction of these variations increases the complexity of the material to be learned, and thus increases its dissimilarity: now, the less alike two movement systems are, the less the danger that their associates will get confused. When the differentiating marks are themselves connected into a system, the advantage they furnish is very much increased. Very simple illustrations of this state of things are furnished by the effect of rhythm in learning and the effect of associating a syllable with its place in the series.
The influence of rhythm in learning has been many times noted in experimental studies and in everyday life: perhaps it was a little exaggerated in the days when schoolrooms full of children used to chant the multiplication table at the top of their voices, but it has its uses as a mnemonic factor. Ebert and Meumann (30) found that without rhythm, it took twenty-three repetitions to learn a ten-syllable series; with rhythm only fourteen were needed for a twelve-syllable series. Indeed, so strong is the innate tendency to make rhythmic all our motor processes whenever we possibly can, that material to be learned cannot be read over repeatedly without falling into rhythm. The attempt not to make it rhythmic operates as a strong distraction of attention: thus M. K. Smith (129) showed that when nonsense syllables are presented without rhythm, at irregular intervals of time, some sort of rhythm has to be read into them or they simply will not be learned. One reason, then, for the value of rhythm in the presentation of material to be learned is simply that we can't help making it rhythmic, and the attempt to do so distracts the attention. But another reason, and the one which concerns us at this point, lies in the fact that when material is presented rhythmically, the words or syllables are differentiated by being associated with differences of accent, just as in Gordon's experiment they were differentiated by having differently colored backgrounds. Thus Müller and Schumann (91) proved that it took more repetitions to relearn a series of syllables in which the accented syllables were those which had been unaccented in the original series, than to relearn the original series with its accents unaltered. The association of a particular syllable with a particular stress helped to individualize it. A third reason for the advantage of rhythm is that it forms especially strong associative dispositions between the members of a single rhythmic foot; thus dividing up the whole system to be formed into a series of smaller systems. That a particularly strong disposition connects the two syllables which belong to the same rhythmic foot was demonstrated by Müller and Schumann. After a series had been
(138) learned in trochaic rhythm, new series were presented, in some of which the 'feet' of the original were preserved although their order was altered, while in others new 'feet' were made by putting together a syllable that had ended one foot and the syllable that began the next one. The learning of the first kind was helped by the associations already formed between the syllables belonging to a single foot; that of the second kind was helped by the associations already formed between syllables which were not parts of the same foot. The results showed that series of the first kind were learned more quickly than those of the second kind; therefore, presumably, the associative dispositions between syllables in the same foot are stronger than those between neighboring syllables in different feet. The exact reason for this is not so easy to make out.
The association of each member of a successive movement system with its place in the series is an aid to the formation of the system. What constitutes the association of a particular part of the material to be learned with a particular place in the series? How do we recognize that a certain syllable came near the beginning of a series formerly learned, or near the middle, or near the end? The syllables at the beginning and end of a series are apt to have distinguishing marks: thus, if the syllables were presented on a rotating cylinder, as is often the case in experiments where syllables are to be learned, the first and last syllables may be associated with the blank spaces at the beginning and end of the series.
Vaschide (142), as long ago as 1896, made a special investigation of the process of associating an impression with its place in a series. He used series of eight, ten, twelve, or twenty words, sometimes visually and sometimes auditorily presented. A few seconds afterwards, the observer having first tried to recall the series, the words were given to him in altered order, and he had to assign to each its original place in the series. The localizing was done in various ways. Usually only the first and last words could be immediately placed, no doubt by some association with stimuli which occurred at the beginning or end of the
( 139) series, although Vaschide implies that the placing was done directly and not through the aid of consciously realized associations. Other words were localized in groups, or by association with a number, or by some associative scheme such as a story, or forming the initials into a word; in still other cases a 'sentiment' was the localizing mark, such as that of effort or difficulty connected with words in the middle, or of relief at nearing the end. Sometimes the marks which determined the place of a word were too indistinct to be analyzed; sometimes it was localized negatively, so to speak, as belonging, for instance, neither to the beginning nor the end and therefore necessarily to the middle. Thus in many ways an impression in a series can enter into simultaneous movement systems which stand for its place in that series. That these marks of absolute position form associative dispositions which aid learning a series of nonsense syllables was proved by Miller and Schumann. After having learned, on one day, four series of twelve syllables each, they learned next day a series in which the feet were those of two of the series previously learned, six feet being taken from each of the two, and each foot being in the position it had held in its original series. That is, the first foot was the first foot of Series I of the preceding day; the second foot was the second foot of Series II, the third foot the third foot of Series I, and so on. They also learned another series made up in like manner of feet from Series I and II of the day before, but with these feet not in the absolute position which they had occupied previously: thus, the first foot was the second of Series I, and so on. The learning of the series in which the feet preserved the same absolute position which they had had on being previously learned in a different series showed a saving of 15% of the repetitions necessary to learn the series with the absolute position changed. It was not the order of the feet that was maintained, since the new series had feet selected from two different old series, but their absolute position, as at the beginning, middle, or end. Each syllable must in the first learning have entered into certain simultaneous movement systems,
( 140) perhaps with the varying degrees of effort, fatigue, and relief characteristic of different positions in the series; and these must have aided the relearning with absolute position preserved. Müller and Pilzecker (90) found them so influential that sometimes instead of the right syllable another would be substituted which had nothing in common with it except the fact that each was. say, the fifth syllable in its series. Such associations with absolute position are naturally, as Nagel (97) showed. more influential with nonsense material than with sense material, where the movement systems formed are so much more complex.
The association of a visual object with a definite position in space nearly everybody will attest from personal experience is a help to remembering it. One recalled a certain rule in one's Latin grammar by thinking of its position on the lower part of the left-hand page. Jacobs (56) conceived the idea of trying the effect of a purely mental localization of visual images on learning: he provided a series of circles in which the observer was to imagine placed the nonsense syllables that were read to him. Such a method involves too much distraction of attention, and it is not surprising that it offered no advantage over simply listening to the syllables. Gordon (44) actually presented the syllables in particularly striking spatial arrangements. They were shown either in a straight line with equal intervals, or in a straight line with unequal intervals, or around a circle with equal intervals, or around a circle with unequal intervals. The circle with equal intervals gave the best results: the spatial positions of the syllables were varied enough to distinguish them, and not so varied as to be confusing. There is a limit to the effectiveness of distinguishing marks added to the material to be learned. If the marks are too complicated, the formation of the movement systems which they involve may interfere with the formation of the new successive system which is the main object: too many sidewise dispositions may detract from the straight-ahead dispositions. This was clearly shown in Peterson's (107) experiments, where the observers
( 141) remembered colors better if the colored objects had different forms, but were not helped when a size variation also was introduced.
The best way to make simultaneous systems aid in the formation of successive systems is of course to have the auxiliary simultaneous systems themselves linked into a succession by already formed associative dispositions. This constitutes the enormous superiority of material that makes sense, over nonsense material. 'Sense' always means already formed associative dispositions. No matter how new an idea may be to us, if it has any meaning at all it evidently appeals to something out of our past experience. There is an overwhelming tendency to bring old and already formed movement systems into activity whenever new ones are to be formed. If you open the pages of a dictionary at random, and select any two disconnected words between which to form a new associative disposition, you have only to attend to them for a few seconds and by some hook or crook, some byway of past experience, you will find that they are already connected. Even nonsense syllables, selected, because of their freedom from old associations, for use in experiments to discover the laws under which new systems are formed, constantly stir up old systems. Take the syllables 'bap' and 'dif'; almost immediately an old system springs into activity and one thinks, 'The Baptists differ from other sects.' It is with the greatest difficulty that students in a psychological laboratory can be induced to learn nonsense syllables without 'making sense' of them.
These already formed systems in some cases help the formation of a new system by breaking it up into units, each unit forming part of an old movement system. This process of using aids to the learning of nonsense material has recently been exhaustively studied by Miller (89). When the aids are simultaneous systems, "apprehended by a single act of attention," Müller calls them complexes; when they are successive systems he calls them associative groups. All kinds of links are used in the formation of these subsidiary groups. Several syllables
(142) may be held together in a group because the first and last ones rhyme, or because the first and last ones have tall consonants, or by means of any sort of meaning association. In recall, a whole group will present itself as a simultaneous unity, and can then be turned into a successive group by having its several members successively attended to. But aids too have their disadvantages: the old systems on which they rest may sometimes transform and falsify the material presented. Müller says they may cause certain elements to be neglected by attention; moreover having too many aids at one's disposal may occasion hesitation, and one may sometimes reject a right syllable because one does not remember using it as an aid and thinks one would have done so had it been there. All this simply means that if one sets old systems into activity, instead of depending on patient repetition to form the new one, the old systems may become so very active and so numerous that they interfere with one another and with the formation of the new system.
When material is learned with the help of any kind of aid, it belongs to two or more movement systems when the learning is complete: one, the new system which has been established, the others, the old systems which functioned in the learning. Its recall when the point of correct recitation has been reached is due to the combined effect of all the systems, the newly formed and the older ones. When a number of associative dispositions thus combine to produce the same result, we may use the term 'constellation,' by Ziehen's (161) suggestion, to designate the process. Many times, when a disposition would be too weak to bring about recall unaided, it may be made the dominant one by the coöperation of others: the word 'plant' might call up an image of something with stem and leaves, and set going other vegetable reminiscences if it did not occur in the phrase 'manufacturing plant,' in which case we may get instead an image of machinery and smokestacks. Constellation is really the most important influence in determining recall. Even the action of a subliminal disposition may so aid another disposition
(143) as to secure its victory in a contest. This was shown by some curious experiments of Miller and Schumann's. Those patient learners committed to memory four twelve-syllable series in trochaic rhythm. Later, enough later for these series to have been partly forgotten, other series were learned which consisted of the accented syllables of the former series. Immediately afterwards, a third set of series composed of the unaccented syllables of the first set was learned, and the number of repetitions required to learn this third series indicated that it was actually helped by the previous learning of the accented syllables; apparently the unaccented ones had been 'set in readiness' by the presentation of their accented companions.
When the material to be learned has meaning, as in the case of a series of words, the tendency to link the words with a connected and ready-made system through their meanings is irresistible unless a great effort is made to counteract it. The words, however randomly chosen, are made into a story. And when the material has connected meaning, forms a coherent narrative or exposition, the influence of these previously formed simultaneous and successive systems is so great that a single reading is often enough to establish the new system. The experimental evidence that meaningful material is learned more readily than meaningless material is plentiful. Ogden (99) found sense material learned about ten times as fast as nonsense material, and Balaban (7) got a corresponding result; the latter and Radossawljewitsch (113) report that it is forgotten much less rapidly. The speed with which different individuals learn sense material is more uniform than the speed with which they learn nonsense material: Michotte (84) says that while four observers had very different capacities for the mechanical learning of pairs of words, they all four accomplished about the same amount of learning in a given time when they were instructed to think of relations between the meanings of the words; that is, to make use of old associative dispositions.
An important effect of meaning associations is that they
( 144) unify successive movement systems. We noted in the chapter on The "Memory After-Image that weak backward associative dispositions are formed, through the influence of the tendency of a movement just performed to be re-excited, between the second member of a successive movement system and the first. Now a much wider opportunity for the functioning of associative dispositions in such an apparently reversed direction is given when the successive movement system is associated as a whole with a simultaneous system as its meaning. Thus the latter part o' a familiar quotation can easily suggest its beginning; the words 'falling fast' may instantly call up the whole line, 'The shades of night were falling fast.' When a successive movement system, each of whose members normally could produce only the next following one in the series, aside from a much weaker tendency to produce the preceding one through the memory after-image process, is associated as a whole with a meaning, that is, a simultaneous system, the latter part of the successive system may suggest the meaning and the meaning may serve to revive the whole successive system beginning with its first members. This 'initial tendency in recall,' as Arnold (6) has called it, is due not to any reversal of the action of associative dispositions, but to the fact that the latter part of the successive series calls up the first part through the mediation of a simultaneous system with which the successive system as a whole is associated.
The influence of meanings, that is, of old and ready formed movement systems, on recall will account for a difference which McDougall (73) regards as fundamental between two kinds of memory, and which he takes as a convincing proof that a motor theory of association is impossible, and hence that psycho-physical parallelism in general must be abandoned. Parallelism, he says, will be discredited "if it can be shown that habit and memory do not obey the same laws." As an argument in support of the belief that they obey different laws, 'McDougall supposes himself to be set the task of learning a series of twelve nonsense syllables in a certain number of repeti-
( 145) -tions. After the learning is complete, he says, "I can throw my mind back and remember any one of the twelve readings more or less clearly as a unique event in my past history. I can remember perhaps that during the fifth reading I began to despair of ever learning the series, that I made a new effort, that someone spoke in the adjoining room and disturbed me disagreeably; I may perhaps remember what he said." "If the repetition by heart of the nonsense syllables and the remembering of any one of the readings of the series are both to be called evidences of memory, it must be admitted that two very different functions, two very different modes of retention, are denoted by the same word."
The chief differences between them he states to be the following: —
(1) The syllable learning involves the formation of a habit to which each repetition contributes a little; the recall of the events which characterized a single repetition "depends wholly on a single act of apprehension." The reason for this, we should say, is simply that the syllable learning involved the formation of a wholly new movement system, while the events of a single repetition are already related to many old movement systems; in other words, they have meaning. We have often before felt weary and discouraged when half way through a task; we have been interrupted and have felt annoyed on many previous occasions. These happenings are so readily recalled because they are relevant, and to be relevant means to be already imbedded in the same movement systems.
(2) Reproducing the syllables involves a forward-looking attitude; recalling a particular repetition involves a backward-looking attitude. To this we should agree; it is immediately related to the next point of difference.
(3) Recalling the syllables is not helped by any effort to cast back thought to the moment of apprehension; recalling a particular repetition is aided by "voluntary rummaging in the past." Naturally, we should answer, since the syllables form a wholly new movement system, and the memory of a partic-
( 146) -ular repetition is as we have seen largely interwoven with older systems.
(4) The syllable learning involves the connecting of eight simple impressions only, yet it requires twelve or more repetitions: the remembrance of a particular event involves a very complex set of impressions, yet it depends only on a single act of apprehension. Naturally, because it is aided by the old movement systems of meanings.
(5) The series is quickly forgotten; the particular event may be recalled for a long time. Naturally, for the same reason that meaningful material may be longer retained.
Another instance of the effect of auxiliary movement systems is furnished by the location that is assigned to visual imagery. There seem to be two sources from which the systems are derived on which such localization is based. One is the actual surroundings of the individual at the moment; the other is what he recalls of the actual location of the real object that is being imaged. Milhaud (85) reports that some observers localize the mental image of an object with reference to their own position at the time, while others feel themselves transported to the real position of the object represented: the difference, he suggests, is due to the fact that observers of the former type are more interested in their own kinaesthetic sensations and cannot lose the sense of their actual position. Martin (75) enumerates no fewer than twelve different ways in which a mental image may be located, and a great variety of conditions which determine the localization, all of which, however, seem to fall into two general classes, namely, those due to actually present objects and the position of the observer's body, and those connected with the memory image itself, such as the recollection of having last seen the original object in a particular situation.
The distinction drawn by Perky (106) between a memory image and an image of imagination is based on a difference in the complexity of the movement systems involved. Perky caused the persons she experimented on, sitting in a dark room
( 147) with one eye closed, to call up what she called a memory image and an imaginary image of the same object. By a memory image she meant the image, say, of a particular horse recently seen; by an imaginary image she meant the image of a horse not connected with any particular occasion of past experience. This use of the terms memory and imagination may have historical justification, but it is surely not the ordinary one. Imagination forms new combinations out of past experiences: one may imagine a centaur, but one has to remember a horse, and one remembers, not imagines, it, whether it is thought of as with or without a context. Baldwin and Stout, in the former's "Dictionary of Philosophy and Psychology," say of the term 'imagination': "It seems better to adopt the current usage of popular language, and to restrict the term to that forming of new combinations which is made possible by the absence of objective limitations confining the flow of ideas." Perky found that the 'memory images' of her observers required more eye-movement and more kinaesthesis generally, than their 'imagination images'; also that "in imagination consciousness is narrowed and there is inhibition of irrelevant associations, while in memory attention wanders and the image is unstable." This is what we should expect: if an observer is required to call up, besides what we should call the memory image of an object, memory images of its surroundings and attendant circumstances on a particular occasion, the motor systems will naturally be more complicated than when the object is recalled alone.
Finally, a profound influence is exerted by old movement systems upon new ones in the processes which underlie the alteration of memory images with time. It may be laid down as a law that whenever in a movement system it is possible for an old associative disposition, based on much repetition, to take the place of a new one whose strength lies rather in recency than in repetition, the substitution occurs. This is of course precisely the fundamental law of perception. If the third and fourth fingers are crossed, and a pencil is laid be-
( 148) -tween them, the patient's eyes being closed, he will have excited the tentative movements that belong to two pencils rather than those belonging to one, even though he knows there is only one pencil: the associative dispositions recently set up by the sight of the single pencil are not able to withstand the old dispositions connected for life with the stimulation of two points on the shin that are not normally reached by a single object at the same time. On this tendency of old dispositions to supplant new ones is based the tendency of memory images to alter in the direction of the ordinary and normal experience: to lose their peculiar and unusual features. Thus Warren and Shaw (143) found indications that the memory image of a large square tends to groom- smaller and that of a small square to grow larger, both being as it were attracted towards a mean. Leuba (69) and Lewis (71 ), working by entirely different methods, noted a tendency in the memory image of a bright light to grow dimmer and that of a dim light to grow brighter, or rather to be judged as the image of a brighter light than that which had actually produced it. Phillippe (109), who had his observers look at a collection of five small objects and then draw them from memory after various intervals of time, reports that the images "seem to tend towards a type preexisting in the mind" which exerts an attraction: thus the features on a Japanese mask tended in the memory image to resemble the European type. Kuhlmann (64) says that the memory images of the picture of an object tend towards representing the object itself, which of course involves older movement systems than those of the picture. Some time ago I saw at an exhibition of the National Academy of Design Waugh's painting called "The Knight of the Holy Grail." I remember that the parts of the picture which impressed me most were the sky, with a single faint star and the line of mountains underneath, the red glow of the Grail, and the very faint halo around the Grail. A stanza from Tennyson's poem was inscribed on the frame, and the last line, "And starlike mingled with the stars," ran in my mind, 'perseverated,' for several days after,
( 149) which probably accounts for the persistence of the memory image of the sky and the Grail. For this is the part of the picture which is now clear in my memory; next come the figures of the angels, while the figure of Galahad I can scarcely get at ail. When I attend to the angel figures, I see them grouped around the Grail in a way that I know must be incorrect, because it is inconsistent with my image of the Grail seen against the sky. I think this wrong grouping of the angels has been borrowed from another picture in which the angels are carrying a little child: I have a vague tendency to complete the angel group in this way. As I let my attention dwell on the angels, suddenly the image of a mast at the bow of the boat appears, at first with an air of authority about it; but presently I reject it. I do not believe it was in the picture: some other boat picture has 'contaminated,' as the philologists say, the Galahad one. I now fix my attention on the figure of Galahad. I see him in profile, kneeling, with hands pressed together in the conventional attitude of prayer. His face is like that of the Galahad in the Abbey pictures. Now, in order to test the accuracy of the memory image which I have thus called up and developed, I can fortunately appeal to an illustrated catalogue of the exhibition. I find that the mountains are much higher and wilder in the real picture than they were in my memory image; that instead of there being group of two or three angels at the same level, there are two pairs, one above the other, an angel of the upper pair holding the Grail against the sky; that the figure of Galahad is not kneeling but sitting, though with the hands folded in prayer as I had remembered them; that my memory image had placed him near the centre of the picture instead of at the extreme left where he really is, and finally that the lower pair of angels carry in their hands tall wax candles in still taller candlesticks. It occurs to me that these candlesticks are responsible for my picture of the mast at the bow of the boat. There is no mast. My rejection of the mast image when it presented itself was accompanied by the incipient, undeveloped thought that a mast in that position would spoil the
(150) picture. Was it perhaps not the long, straight, nearly vertical lines of the candlesticks in the bow that suggested a mast? Now, almost all the errors in my memory image are in the direction of making it more commonplace, more generic. The rugged wildness of the mountains was softened; the unusual distribution of the angel figures became the more familiar image of two or three figures at the same level, and tended to blend with other pictures of the same general kind, as is shown by the intrusion of the child's figure. The figure of Galahad was shifted towards the centre to make the composition more commonplacely symmetrical; the correct image of the hands pressed together suggested the ordinary accompaniment of a kneeling rather than a sitting figure. Finally, the candlesticks were so unusual a feature of the bow of a boat that they transformed themselves into a mast.