Imageless Thought

James Rowland Angell
University of Chicago


The problem of imageless thought has taken two main forms in current discussion. One of these has to do with the analysis of the reflective consciousness and with the exhaustive description of the contents of the mind distinguishable in such thinking. It is alleged, for example, by one considerable party of psychologists that a good deal of our thinking goes on without the presence of any discernible images at all. There is, in the second place, as a subordinate form of the problem, the doctrine that voluntary action requires for its initiation the presence of neither sensations nor images, and that it may be carried on as the consequence of a 'pure thought.' Clearly, this second form of the problem is essentially a corollary to the first form, inasmuch as the existence of such imageless thought is, in the nature of the case, a question prior to the use of such thought to control conduct.

Partly owing to a discussion carried on in the pages of the Philosophical Review with Professor Stout, the editor of Mind, and partly as the result of certain statements in the writer's 'Psychology,' his name has been involved in the controversies which have raged on this subject in contemporary literature. Up to the present time the author has held his peace in the face of the attacks to which he has been subjected by the defenders of imageless thinking. This occasion seems an appro-

(296) -priate one, however, upon which to initiate a counter attack. This course is particularly justified in view of the rapidly increasing literature on the experimental psychology of the thought processes, in connection with which the issue involved in the imageless thought doctrine is peculiarly important.

We shall address ourselves first to the question of the sheer existence of imageless thought and devote such time as may remain to a briefer consideration of the utilization of thought material of this kind in the control of voluntary action. It will perhaps serve to clear the ground in a helpful way if we turn back to the first explicit discussions of this question in our English literature of the subject, and see how the case stands. This is especially desirable in view of the persistent errors of citation, as they appear to the writer, rife among the defenders of imageless thought and even among its critics. Stout, for example, has been repeatedly cited as an advocate of imageless thought, despite his explicit disclaimer of entertaining under that rubric a doctrine such as certain of its advocates attribute to him.

The present author, in reviewing Stout's 'Analytic Psychology,' had called attention to the fact that imageless thought was put forward in that treatise as not only intelligible, but actual. Following this review, the author wrote a brief paper on the subject at issue, to which Stout himself replied in a subsequent number of the same journal.[2]

Stout's original statement had been as follows:

"An imageless thought is no absurdity. . . . There is no absurdity in supposing a mode of presentational consciousness which is not composed of visual, auditory, tactual, and other experiences, derived from, and in some degree resembling in quality, the sensation of the special sense. . . ." (‘Analytic Psychology,' Vol. I., p. 85.)

This seemed to be a sufficiently unequivocal statement and the present writer entered upon his critique, assuming that Stout really meant what he seemed to say, namely, that we have particular moments of thought in which no imagery and no sensational material is to be discovered. In the reply which Stout wrote to this article he made it quite clear that the doctrine which he advocated was in no sense

( 297) interpretable as the author had done. The ambiguity seems to attach to the word ‘mode' in the passage quoted, which Stout apparently used as equivalent to the word ‘phase,' whereas the reviewer had regarded it as equivalent to 'form,' or 'moment.' Stout said, for example:

"It will be said that in all such cases the presence of some specific item of sensation is necessary, and that similarly, in higher processes, the presence of some specific image is necessary. Now I do not feel sure that this is absolutely and always true. But I have no desire to contest the point. Certainly if imageless apprehension occurs otherwise than as apprehension of the meaning of a given sensation or image, it must be admitted that it plays no important part in our mental life. What I am concerned to maintain is that apprehension of meaning is a specific mode of consciousness, and that it cannot be resolved into the presence of sensation or image to which the meaning attaches. . . . I am inclined to believe that Professor Angell's criticism is largely due to his supposition that when I speak of imageless apprehension, .I have in view a total state of consciousness, rather than a partial constituent of a total state, which contains as another constituent some sensation or image. What I am really concerned to deny is that when A means B for consciousness, it does so by actually recalling an image of B; and that when no image is recalled, it can mean B only in the sense of being substituted for it."

This statement would seem to be sufficiently explicit, and yet the advocates of imageless thought continue to quote Stout as if he were in reality contending for precisely that which in the passage just quoted he denies advocating. Moreover, in a passage of the ‘Manual of Psychology' published a few years later he makes assurance doubly sure by the following statement which surely is not open to misinterpretation.

"An idea can no more exist without an image, than perception can exist without sensation. But the image is no more identical with the idea than sensation is identical with perception. The image is only one constituent of the idea; the other and more important constituent is the meaning which the image conveys" (p. 394)

At another point we read (p. 396)

"... Even the highest development of conception still involves imagery, though the imagery may be, and often is, purely verbal."

In an article in Mind, for 1907 [p.70 ff.], Mr. R. F. A. Hoernlé presents an admirable analysis of the connections of image, idea, and meaning, which has often been used, as have the writings of Stout, to support the imageless thought doctrine. Mr. Hoernlé, however, as I read him, holds a view which is far removed from such a position as that of Professor Woodworth, for instance, and can hardly be put into the ranks of

(298) the true defenders of imageless thinking without the exercise of logical and ethical violence. Mr. Hoernle is in no sense concerned to deny the presence of sensations and images in our ordinary conscious processes, but rather to point out that it is the meaning, rather than the gross substance of images or sensory materials which constitutes the important part of conscious processes. He says, for example:

"But the fact that a significant psychic whole contains as a subordinate factor sensational elements and that such a whole apparently cannot be present in our consciousness without the help of such elements, by no means entitles us to treat them as the only solid and substantial factor, the only objective content, as it were."

(The italics are not in the original.)

He devotes considerable attention to Professor James' account of the focus and the fringe of consciousness, in connection with which Professor James had asserted that the image or the sensation is actually the focal factor, whereas the meaning is ordinarily found in the fringe. Professor James is not on trial at the present time, and we pass over the justice or otherwise of Hoernlé's critique on this particular point. It is, however, interesting to remark that Hoernle appears to advocate the doctrine that either image or meaning maybe respectively focal or marginal. He says, nevertheless, that sign and meaning are inseparable. The sign, normally made up of sensational elements, is also normally subordinate, and the distinction of sign from meaning, is, in his judgment, a product of reflection. In his opinon meaning, and not sign, is ordinarily the focal thing. We regret that space does not permit us to consider more fully this particular doctrine of Hoernlé’s, for there is certainly some reason to feel that meaning and image are hardly to be distinguished as being in any given case the one focal, and the other marginal. It is at least a plausible view that they come and go together, and that when we say one is marginal and the other focal, we are describing a situation in which one kind of meaning is focal and one kind of image marginal, in disregard of the fact that another type of meaning is simultaneously marginal, and another matrix of imagery is focal. At all events, the facts are not so obvious concerning this issue, as Mr. Hoernlé's account would indicate. Moreover,

(299) there is ground for suspicion that Mr. Hoernlé shifts in his discussion between an account of logical meaning as such, and the experiencing individual's consciousness of meaning, which is often quite another thing.

In a similar way, Binet has been repeatedly cited as an unequivocal advocate of this imageless thought as a real form of conscious experience, which may be present without either sensational or imaginal thought constituents. I am disposed to believe that Binet perhaps has held this view from time to time. But the passages which were originally quoted in support of this interpretation, are altogether equivocal. He said, for example, in a treatise on the experimental study of intelligence [3] that consciousness is chiefly made up of images and words, but that thought may be an unconscious act of the spirit, which in order to become definitely conscious requires words and images. In the same treatise he said that he came upon many cases of thought in which it was not possible for his subjects to detect any imagery. It appears, however, from these citations, that at that time he was not advocating a doctrine of imageless consciouness, but rather a doctrine of sub-conscious or non-conscious intellection. In a later article which appeared conjointly with Simon in the Année Psychologique for 19o8, p. 333 ff., and especially 339, he takes the ground that thought without image and without words is an actual phenomenon of what he calls intellectual sentiment, whatever this may mean. It appears, in any event, to connect the imageless thought process with affective aspects of mental life, as over against the purely cognitive ones. This would seem to give no comfort to a doctrine of purely intellectualistic thought content, disconnected with the ordinarily recognized elemental aspects of conscious experience.

Professor Woodworth has undoubtedly been the most persistent and unmitigated advocate of the imageless thought doctrine among American writers. Indeed, he has perhaps considered the matter from more angles of experimental study than any other one person. His statements lack nothing in

(300) radical character, however convincing they may be found. His views are so well known, and particularly to this audience, that extensive quotation is hardly essential. He has from the first advocated the most extreme form of the doctrine, in maintaining that we have moments 0f consciousness, from which both sensory and imaginal elements are entirely excluded. Moreover, he espouses the most extreme possible view about the independence of image and meaning. He says, for example:

"An image may call up a meaning, and a meaning may equally well call up an image. The two classes of mental content differ in quality as red differs from cold, and anger from middle C. They may also differ in importance for the purposes of a given thought. Otherwise, it is hard to see any essential psychological difference between them." (Journal of Psychology and Philosophy, Vol. III., 1906, p. 707.)

He renders it impossible to explain away his alleged phenomena by any reduction 0f them t0 physiological or sub-conscious activities, because he asserts with great vigor that the mental states which he has in mind, involve the clearest and highest lights of conscious experience.

In his recent little monograph on the 'Experimental Psychology of the Thought Processes,' and still more recently in his 'Textbook of Psychology,' Professor Titchener has entered upon a somewhat elaborate critical evaluation of the recent germanistic movement, dealing with the analysis of the thought process, as represented in such writers as Watt, Ach, Messer, Buhler, and others. This is altogether the most systematic effort which we have thus far had, to bring together the somewhat discrepant and wholly undigested literature of the subject. Mr. Titchener occupies a relatively conservative attitude 0n many of the essential points at stake. Nevertheless, he is quite outspoken in his unwillingness t0 accept the extreme interpretations 0f the alleged facts such as are advocated by Buhler, and Woodworth. He says, for example, that he remains entirely unconvinced that there is any such independent thought element as Buhler [and in a somewhat different manner Woodworth] advocates, and that neither as a result of his own experimentation, nor as a result 0f his careful and non-partisan reading of the experimental literature, is he able to feel at all convinced that

(301) the accounts 0f a knowledge process in terms of sensation and image are inadequate to account for all of the authenticated experiences.[4]

We have had a considerable number of briefer discussions in the columns of our technical journals, and one or two independent monographs, like Miller's 'Psychology of Thinking,' Dewey's 'How we Think' and Pillsbury's 'Psychology of Reasoning,' in none of which is the imageless thought doctrine openly espoused and in the first of which explicit ground is taken on the orthodox (?) side of this controversy. Pillsbury, in the Psychological Review for 1908 (Vol. XV., p. 158) has argued that meaning is primary and image secondary—in fact only another kind of meaning. The genetic primacy of meaning had already been advocated by Stout. Gore, in an interesting paper on meaning,[5] has advocated the doctrine suggested by certain of Dewey's positions, that the image represents the stimulus in consciousness, and meaning the response. This is to say that the image is correlative to the structure, whereas meaning is correlative t0 the use we make of our materials. Professor Mead and Dr. Kate Gordon have advocated similar views in occasional papers, and the present speaker, in his textbook of Psychology, has also an account in some ways closely resembling this. Mr. Thorndike has entered into the general circle of controversy with several papers, and with passages in his 'Elements of Psychology.' He, however, is mainly concerned with the question of volition, and will be considered at a later point. Under the influence of Woodworth and Thorndike, Betts has written a monograph on the Distribution and Function of Imagery [Columbia Cont. to Education, 1909, No. 26] in which he ventures to enlarge upon reports of imageless thinking, after having frankly eliminated verbal imagery from the discussion.

When one brings together all the literature of the subject in English, of which the above citations are illustrative, several distinct questions present themselves as containing the pith of the controversy. These may be formulated in a provisional way, at least, as follows:

( 302)

First, is there an aspect of cognitive experience, which is immediately given in consciousness, distinct from sensation and image?

Second, are there independent moments of cognitive consciousness, in contradistinction to mere aspects of such moments, in which neither sensation nor image is to be detected?

Third, supposing them to exist, are these alleged moments of consciousness which are devoid of sensational and imaginal components, in reality automatized forms of perception or of ratiocination, which in an earlier stage would have disclosed distinct evidences of imagery or sensation? Are they, in other words, processes well along towards automatism, having originally been characterized by a much higher degree of consciousness?

Fourth, going still further than the previous question, are these alleged moments of consciousness, in reality instances of unconscious cerebration, to use a phrase more familiar to our forbears? Are they instances in which cerebral processes, for the time being excluded from direct contribution to the focus of consciousness, produce consequences which then appear in the conscious stream as the results of reflective thought? In this connection, of course, the advocates of 'split-off consciousness,' of double consciousness, are in position to allege that such phenomena have to do with dissociated foci of consciousness?

Fifth, are these alleged moments of rational thinking cases in which consciousness is monopolized by the awareness of attitudes, in which, originally at least, kinaesthetic sensory elements were conspicuous, but which, in the developed mind become, as attitudes, so habitual as to render introspection extremely difficult?

Sixth, in part equivalent to the second question and appearing as a subordinate feature of each of these problems, is the question whether meaning is a conscious element separated out and existing apart from both image and sensation.

( 303)


With reference to the first point, I doubt whether any psychologist of repute would call in question the affirmative reply to this issue.[6] Modern writers have varied widely in the extent to which they have seen fit to emphasize this aspect, i. e., the meaning aspect, of our cognitive experience, but it would be difficult to bring any satisfactory evidence to show that they disbelieve in its reality, and in the case of not a few writers, from Thos. Brown down, of whom Stout in his 'Analytic Psychology' may serve as a conspicuous example, there has been abundant emphasis on the reality of this feature of cognition.

I find it difficult, therefore, to explain the necessity which Mr. Woodworth has felt to defend so energetically the presence in perception of non-sensory elements of consciousness.[7] When he points out, for example, the peculiar manner in which we may interpret equivocal drawings as now one thing and now another, he seems to me to be stating merely the familiar doctrine of perception, without adding anything which seriously needs defense, and above all, a doctrine whose truth should not be exploited in the interests

( 304) of imageless thought, a bird of very different plumage I do not know where among contemporary psychologists of standing he will find the doctrine seriously defended, that perception involves the framing in the mind of an image prior to the apprehending of the perceived object. Certainly it is not the common view. Yet it is this conception apparently, in one or other of its forms, against which he is arrayed. When it comes to the question of terminology, to determine what names we shall assign to the various apperceptive factors of experience, it is more difficult to come to agreement, and the actual literature of our subject discloses considerable variance in usage. This difficulty appears at the very threshold in the fact that a writer like Mr. Woodworth, in his discussion of the non-sensory components of perception, is evidently assigning to the category of sensation a highly abstract and artificial significance. It must be admitted that a good deal of our structuralistic psychology has been guilty of such an abstraction. Indeed, it has rather gloried in it and defended its own excesses. But, whether such abstractions be judicious or injudicious, to use the term sensation as applicable to the sheer sensuous qualities of a sense experience, and then upon taking that step to plead earnestly for the introduction of so-called non-sensory 'elements,' is a precarious procedure, provided it can be shown, with any reasonable persuasiveness, that the elements thus called in, are quite as much sensory, as those which they are enrolled to assist: that is, that they appear quite as genuinely in the response of consciousness to peripheral stimulation of the organism. And this they certainly do. This, however, is a relatively minor issue and may be allowed for the present, to rest. We repeat, however, that such a position as Stout's, in which he alleges that every instance of cognitional consciousness contains in addition to such imagery and sensation as may be present, elements of another kind, represented by the meanings involved, this doctrine has, so far as the writer is aware, no serious opponents. Mr. Woodworth, however, while recognizing such elements would apparently maintain that they do not always accompany imagery.[8]


We come, in connection with our second point to what is really the crux of the whole matter, and to the most baffling of all issues. Indeed, after going through the literature, one is left with the feeling that the case is largely reduced to mere assertion and denial, occasionally to vituperative recrimination. It seems to be largely a matter of "It is!" or "It isn't!", adorned with such adjectives as taste may dictate and capacity afford.

Competent introspectionists are arrayed on the two sides of the question, and the results which they bring in are equally unequivocal and equally dogmatic. Under such conditions the burden of evidence would seem to be with those who take the negative attitude and deny the presence of that which the other party alleges to be discernible in its own person. This is logically, no doubt, the case, but scientifically it is a safe tradition which leads to the principle that any radically new scientific element shall justify its existence by unambiguous and convincing proof. However, this is no occasion to stand upon the niceties of etiquette in the matter of proof and disproof. The practical question is, after all, whether these parties, who are divided on this issue, can find some line of approach which will bring them together in reasonable agreement. It is unthinkable that either party to the controversy is prepared to rest short of a friendly conversion of its adversaries, and our immediate problem is first to find some adequate explanation of our differences of opinion, and then, if possible, to remove the source of these differ ences.

The most peaceable, if not the most natural, explanation for the extreme divergence of views which we meet is that we are confronted by two radically disparate types of mental organization, to one of which a form of thought is native which to the other is substantially unknown. The other obvious hypothesis involves the inference that the introspection of one or the other of the protagonists is essentially defective. The latter alternative seems the more plausible, for it surely seems more reasonable to believe that we differ in our capacity to identify a particular kind of conscious

( 306) material, than it does to believe that we are separated from one another by the possession or lack of possession of a fundamental thought quality.

In studying the introspective literature of the subject, one feels that the differences which divide certain of the writers are largely those of mutual misunderstanding as to the precise phenomena under discussion, a misunderstanding fostered in part by wide divergences in thought processes, such as our investigations of imagery types have made familiar, but not a misunderstanding based on the presence in one individual of a form of consciousness wholly lacking in the other. There are a few writers, however, of whom Woodworth and Biihler are fair examples, for whom such an explanation seems relatively inapplicable. I do not mention the other Wurzburg contributors, because I find them rather less unequivocal in expression. I judge, however, that Watt might well be added.

With scientists of their training and experience, to say nothing of some of their reagents, the charge of inaccurate introspection is likely to seem to most persons highly improbable, if not preposterous. One may hope that the time has passed when the charge of inaccuracy in such a matter need indicate any personal animus, much less any disposition to reflect upon the general scientific competency of the person charged. Nowhere is the difficulty of observation so great as in the field of controversy now under consideration, and no man has ever worked in psychology with serious intent, who has not at one time or another found himself convicted of serious error in his observational reports. So long therefore as the number of persons remains small, who have been thoroughly trained in introspective methods, and who report affirmatively upon the presence of imageless thought, it will always seem a reasonable interpretation that errors of observation are responsible for the discrepancy between their reactions and those of other trained observers. But the number of individuals who are now to be counted among those making affirmative allegations is suspiciously large to justify this interpretation, although the author is at present obliged, as

(307) the lesser of two evils, to believe that this line of explanation is essentially correct.[9]

There are a number of conscious activities, with which it is perhaps easy to confuse the supposed imageless thought process, if one may judge by the accounts offered of this process. For example, the literature is full of evidence thdt in framing their conclusions at least, not a few persons have failed altogether to distinguish between the presence of objective imagery, especially of a visual character, and the presence of word imagery, whether of purely auditory character, or as more often happens, auditory-motor; or as frequently happens (and when it does happen, it creates the maximum of introspective difficulty) the presence of the suppressed enunciatory movements themselves. Again, the descriptions offered strongly suggest in some cases the presence of highly schematic and extremely evanescent imagery, whether of verbal or other character. In other cases the descriptions clearly indicate that sensory materials, including the attitudes reported by kinaesthetic sensation, are employed as the carriers of meaning, in which case the search for imagery in the proper sense is necessarily futile and predestined to failure. Finally, there are not a few instances in which the descriptive accounts make it all but certain that the actual occurrences involve essentially subconscious activities, which emerge with rational results, and which are then attributed to imageless thought. I am not at all confident that there are not other conditions in which confusion is equally explicable.

These alternative hypotheses, are, however, well known to such a psychologist as Mr.Woodworth, and he energetically rejects all as entirely inapplicable to his own case. He insists that the process at stake is no subconscious affair, and quite the contrary, is characterized by the keenest sort of consciousness. Except by inference I do not find him disclaiming in quite so convincing a manner the presence of sensory and attitudinal factors, although whenever he refers to these as in any way present, he takes particular pains to speak of them

( 308) as irrelevant. Whether they are so unimportant as he imagines, I have no method of determining, but their presence I understand him in many instances to admit. I shall later call attention to a fact suggesting that they are not so irrelevant as they appear. The only types of imagery whose denial his statements do not seem to me wholly convincing about, are those which Mr. Colvin has recently termed mimetic, and which the present writer had previously spoken of as schematic, or symbolic. Mr. Woodworth, however, must be entirely conversant with descriptions of this type of thinking, and presumably he would have identified it, if it were present in his consciousness. Certainly it is far from my intention surreptitiously to inject any such heterodox psychological virus into his system, although I frankly doubt whether his thought is as pure as he supposes. He sometimes refers to it as 'naked' and in view of its lack of descriptive raiment, this seems to be a good term.

A naturally trustful disposition is somewhat disturbed by the absence from Mr. Woodworth's reports of detailed and accurate descriptions of his imageless equipment. He surrounds himself with a cloud of negatives, denying that it is visual or auditory or tactual, denying that it has any sensational or imaginal composition, and leaving us with a mere apotheosis of the void, so far as concerns any positive assertions. It appears to be closely related to the smile of the Cheshire cat which remains after the cat has disappeared. Of course, it may be that those of us who are organized on the simpler, old-fashioned plan of sensations and images with their meaningful aspects, may be incapable of appreciating a more accurate description of these sublimated experiences. But at least we should like to have a try at it, and we seem to get but little assistance on our way, by listening to this mere chorus of negatives, telling us what the imageless thoughts are not like. Jesting aside, it certainly seems odd that it should be impossible to do more than sit back and allege the presence of these elements, if they are so important in the structure of the thinking of these individuals. Certainly the products of their thought are not markedly less

( 309) valuable, and, when reduced to words, are not notably less lucid than those of persons frankly indebted to the use of imagery for their thought processes. It seems odd, therefore, that so little has been done, and that therefore presumably so little can be done to make clear to those of us less richly organized what kind, of possession an imageless thought really is. We may be abnormally stupid.

Buhler [10] makes a serious effort of this kind, but after he has gotten through denying.sensation, little is left but functional terms such as I myself should use in describing attitudes or meaning or relation.

Mr. Woodworth has, so far as I know, laid rather more stress upon the presumptive support given to his docrtine by the facts of cerebral anatomy and physiology, than any of the other writers on this subject. He has repeatedly referred to the fact, which he regards as highly significant, that so small a portion of the cerebral cortex is sensory in character. From this he draws the inference that presumably the other areas of the cortex subserve functions which are distinct from sensation, and which consequently may be expected to find counterparts in consciousness which are not of a sensory character. It is, of course true, that anatomically there are very considerable regions of the cortex which do not receive impulses directly from sense organs, but on the other hand, it is not to be forgotten that the evidence is very strongly indicative of the fact that whereas one large group of regions receives impulses from the sense organs, and other large regions are concerned with the discharge of impulses into the muscles of the body, the remaining regions serve anatomically, and presumably physiologically, to connect these sensory and motor regions with one another. Ordinarily, as regards the great mass of these interconnective, or 'association regions' as Flechsig calls them, stimulation is transmitted from sensory regions either into other sensory regions, or, as probably occurs ultimately, out into motor regions. From this point of view the system is essentially unipolar in character, and it is difficult to suppose that any cortical

(310) portion can be wholly devoid of some sensory component in view of this fact that the sensory components are prior in the series of normal innervations to the stimulations of other regions 0f the cortex. However, all this type of speculation must be regarded as of extremely problematic value, whether t0 defend, or t0 attack, the view of thorough-going imageless thought—at least thought considered as strictly conscious.

Mr. Woodworth commits himself also to a view which he has expressed in more extreme language than the author recalls to have seen elsewhere, in his separation, already referred to, of meaning and image. [This concerns our sixth question.] He speaks in the most confident way of the appearance of a meaning entirely devoid of an image, and conversely of an image entirely devoid of a meaning. For example, he says: "Meaning is not felt as a relation between the image and an object, but as the thought of the object . . . . It (i. e., meaning = thought of the object—J. R. A.) is as substantial an element of thought as the image, and there is no absurdity in the notion that it may be present alone. (Jour. Psy. Phil. Sc. Meth., 1906, p. 707.) We have already quoted another passage in which he makes the radical announcement which follows. It seems altogether improbable that Mr.Woodworth can have in mind by 'meaning' just the sort of thing that some of the rest of us have, when he makes this statement:

. . . The two classes of mental contents differ, in quality . . . they may also differ in importance . . . otherwise it is hard to see any essential psychological difference between them." [Loc. cit.]

Whether he has gained any comfort from Messrs. Pillsbury and Gore, who have spoken as though image and meaning were in some sense co-terminous with one another, the author does not  know, but he is disposed to believe that these gentlemen have a different doctrine in mind. To say that two psychological qualities differ from one another as 'anger from middle C' and then to polish off the statement with the assertion that it is hard to see any other essential difference between them, is a good deal like saying that black is not

(311) white, but that otherwise they are highly similar. If it proves anything for Mr. Woodworth's doctrine, it would appear t0 prove too much, for it would indicate that meaning, which is for him, as it is for others, one of the essential relational features of consciousness, is itself an image in disguise. However, the present writer is too much in doubt as to Mr. Woodworth's intent in this particular of his doctrine to discuss it intelligently. In the article from the 1906 Jour. Phil. Psy. and Sci. Meth. (p. 708), he says, however, that probably all thinking originates in terms of meaning, but that some persons have more excitable sensorial processes and so get images. This conception of course makes the image a mere annex to the essential process. He calls it 'by-play.' Buhler, be it said (op. cit.) holds that his 'thoughts' are quite distinct in every way from the 'ideas' with which imagery is found connected.

If now we return for a moment to our first assumption to explain the fundamental discrepancies between our controversialists, we are obliged to accept the existence of two radically distinct types of organization for thought. There is perhaps no more a priori reason for refusing to make this assumption than there is for questioning the reality of race distinction on the physical side, but all the analogies which it is easy to lay hands upon in the way of organic divergences within the same species, suggest variation in the line of special development, or lack 0f development, in tissues which are common to all the members of the species. The case now before us, if we accept this analogy, is of a different kind. It is comparable in character to the discovery of a race of men having six fingers instead of five on each hand, 0r having two noses or three eyes. Mythology presents us with characters of this sort. The anatomical museums also present us with such materials, but they are recognized at once as abnormal in the latter case and imaginary in the former. They do not represent types which are perpetuated and constitute any considerable portion of the race.

Now with the best will in the world, the author does not find it possible to discover any middle ground between these

( 312) two hypotheses, provided one excludes the assumption that we are really dealing with a subconscious process or with the unconscious cerebration of the earlier writers. Woodworth and Buhler, at least, are certainly not talking about subconscious processes. Binet, on the other hand, in some of his writings, certainly has had this in mind, and has given his assent to the reality of such subconscious thought. But this alternative being put aside, and all fallacy of introspection being ruled out, there appears to be no other adequate explanation except that of the existence of two different types. As between these two necessary alternatives, the writer does not hesitate to take the former as scientifically the more conservative, and up to the present time the more justifiable. It is more reasonable to assume that the introspection of a few men, however competent, may have gone astray (perhaps as several critics have urged, because of their method), than to introduce an element which many men, presumably equally competent, are wholly unable to verify. Undoubtedly it is the part of an open mind and of scientific honesty to be hospitable to all such new comers in the field of scientific observation. But whenever such a new-comer arrives without a properly certified passport, it is not unfair to insist that he be detained at the frontier until he can be carefully examined, and his fitness for ultimate citizenship convincingly demonstrated. This appears, in the author's view, to be the present situation in regard to the hypothesis of imageless thought.

It is a far cry from our catalogue of points to be discussed, for we have tarried unduly long perhaps on the second of the issues raised. But this was really the central issue and the rest are of secondary significance. Our third point concerned the possibility that the alleged imageless thoughts were in point of fact automatized mental acts which in an earlier stage would have revealed distinct evidences of imagery.

Speaking for himself, the author has no question that in his own thought processes there are many occasions where thinking is carried forward by verbal imagery so highly schematized and so automatized as all but wholly to escape

( 313) identification. Moreover this schematism in the imagery is often an incident of the process of 'telescoping' to which many of the writers on this topic have called attention. In view of this fact, the author cannot refrain from strong suspicion that in the case of many of the less experienced observers, imagery of this variety may well have been overlooked. Nor is it the contention of the more thoughtful advocates of imageless thought that such schematic and compressed imagery may not serve a purpose in the conveying of the meanings of thought. On the other hand, however, we have a number of instances of observers whose reports cannot be called in question on the ground of lack of training and general professional competency. When these authorities insist that there is not a vestige of such schematic imagery, one must either accept their statement as fact and make the best of it, or resort as do the writers previously cited to a charge of introspective fallacy. Wundt, for example, alleges that the imageless thought is a consequence both of defective observation and false presupposition.[11] Titchener attributes the fallacy 'to the stimulus error,' by which he means the failure to distinguish the attributes which belong to the stimulus from those which belong to the consciousness of it.[12] An issue of this character is hardly likely to be settled merely by dogmatic affirmation and denial, which is the level at present characterizing the controversy. Unless some more crucial experiment can be discovered than has hitherto been employed, so that one party or the other can be brought to a conviction of the correctness of the view of the adversary, we must simply wait for the amassing of evidence in the hope that the accumulation of experimental results will slowly create a presumptive proof for the position of one or other of the contestants.


Our fourth consideration touching the explanation in terms of unconscious cerebration is also thrown out of court as irrevelant by most of the defenders of imageless thought. As we have seen, they allege in the most explicit way that the experiences to which they refer, so far from being nonconscious, are among the most vivid and distinct which they ever meet. This, of course, does not prevent their entertaining the hypothesis that non-conscious cerebralistic activities may produce consequences which may emerge in the field of consciousness at some later period. Binet, for example, speaks of the fact that thought itself is unconscious, but that unconscious processes exercise a directive influence over the flow of conscious states. I am not aware that any of the cohorts of imageless thought has appealed to the facts of 'split-off consciousness' to account for the phenomena involved. They seem rather to have in mind a highly explicit form of normal awareness, which however, baffles description in any ordinary terms.

Our fifth point touches the question of the possible explanation of the alleged imageless thought phenomena in terms of awareness of attitudes, especially such as involve experiences marked by the presence of kinaesthetic sensations, due to the assumption of particular bodily poses. Here again, the present author would take a strong affirmative position so far as concerns his own thinking. In many instances the closest approach which he ever secured to a state in any way strongly suggestive of the descriptions of the imageless thought proprietors, is in connection with certain reflective activities, in which there is from moment to moment an almost entire absence of describable ideational material, but with a most vivid consciousness of the directional and attitudinal kind involving vivid kinaesthetic experiences. When these experiences are met with, it is often quite impossible to find anything describably present except the awareness of an attitude of expectant strain sensorially reported, together with which there is a keen apprehension of the direction in which the thought is about to move. This is often characterized by a very definite sense of the

( 315) multifold associations of a nascent kind connected with the thought, but not rising to the focus of attention. And again, the present author would be strongly disposed to interpret the introspective deliverances of many unpracticed subjects in accordance with an explanation of this type. But we are met with so positive and heated denials of the reality of this explanation on the part of the most sophisticated of the observers among our adversaries, that we are obliged to question the adequacy of such an explanation for some at least of the experiences at issue. Here again our only lines of advance would appear to be those suggested in connection with point four.

As regards point six, few of the imageless thinkers take so extreme a view as does Mr. Woodworth in his separation of the meaning factor from the image. We have already quoted him to the effect that either image or meaning may come to consciousness entirely independent of the other. We have also remarked that this view is flatly contradicted by a number of high authorities conspicuous among whom is Stout. As in all these cases, Mr. Woodworth will brook no appeal except to fact, and in his own case, he is apparently perfectly certain that meaning and image appear in the separate and distinct way above indicated. No other view could well be tolerated by him because of the extreme form in which he defends his imageless thought. The issue reverts again, therefore, to the old question of reliability in introspective report.

Thus far I have only remarked two writers, Buhler and Woodworth, who seem to adhere to this extreme formula and Buhler's conception of his 'Gedanken' is such as to render this formulation misleading as applied to him. Binet's first statements about the matter were certainly couched with a view to recognizing the fact of unconscious cerebral activities rather than the presence of a strictly conscious experience devoid of sensational content. His later accounts have, as already indicated, introduced the notion of the intellectual sentiment or feeling, a suggestion which the present writer can partially sympathize with in so far as it indicates the presence of

(316) thought processes in which the strictly cognitive elements are submerged under affective attributes. The difficulty of description might well connect itself with this affective, noncognitive character. Miss Calkins has expressed herself in a way which confirms my own reading. She says:[13]

"The writer of this paper frankly deprecates the tendency of certain psychologists —of Stout, Buhler and Woodworth, for example—to insist that the occurrence of imageless thought has been proved. . . . What is abundantly proved is that along with imagery, and often in the focus of attention, when one compares and reasons and recognizes, are elements neither sensational nor affective."[14]

The author, together with many other psychologists, after the most painstaking efforts to ascertain what actually occurs in his own experience, is wholly unable to confirm the appearance in consciousness of any such meaningless image, or of any such imageless or sensationless meaning. For reasons that no doubt have grown out of the similar observations of a long line of psychologists and logicians, the general theory has arisen which alleges the invariable interconnection of these two phases of mental life. To suppose that one can occur without the other is utterly to destroy the entire foundation on which rests the theory advocated by most of us concerning cognitive operations. Doubtless each of us would promptly abandon the theory provided we were confronted by convincing evidence of the fallacy of the data upon which it rests. But until we can secure something like unanimity of competent introspective opinion on the matter, we can hardly be asked to throw overboard our hard-won convictions, which equally with those of our opponents are based upon what we believe to be unimpeachable facts. A meaningless image is to the present writer not only a thing never experienced, but also a thing in the nature of the case ridiculous, a physical object free from the attribute of gravity, a light devoid of wave-length. In the same way and for the same reasons an imageless or sensationless meaning seems an impossible as well as an unexperienced event.

If we accept Woodworth's pure thoughts we commit ourselves to a belief in the prodigality of nature such as has not

( 317) generally been entertained in modern times. I do not find that he assigns in any convincing manner, a function to these pure thoughts, essentially different from that subserved by ideas of the ordinary variety, i. e., meaningful images of objects or words or acts. Evidently, unless this can be done, we are invited to suppose that we have two generically different thought materials to accomplish one and the same result. Clearly this may be true. But it flies in the face of well-grounded prejudice and tradition to admit it. Possibly Mr. Woodworth calls in question the doctrine that ordinary ideas operate as we have always supposed. Certain of his statements seem to lend themselves to such interpretation.

Relevant to this point it may be added that so far as I can discern, the appearance of imageless thoughts is not connected with any specifically assignable kind of situation. They may appear on one occasion only to have their places taken by ordinary ideas on another similar occasion. This seems at all events to be the experience of many of Mr. Woodworth's subjects. They appear, therefore, to be somewhat sporadic and irregular phenomena, which is certainly not what we might expect, if they play so important a part in thinking as their defenders would have us believe. So long as they disclose vagaries of this kind, we may justly demand and wait for their further analysis before accepting them.

Buhler takes rather different ground on this issue. He apparently regards his Gedanken as serving a different and higher function than ideas. Time fails me for a careful analysis of his contentions on this point, but it is perhaps not unfair to say that they have received very severe criticism from a number of competent psychologists. Wundt, as will be remembered, refers to the Gedanken as constituting a return to the actus gurus of the scholastics. Whether this be scholasticism or not, the view which regards ideas and 'pure thoughts' as subserving essentially different functions, appears to be much better entrenched logically, than one which more nearly identifies their properties. I do not find Buhler's citation of fact anywise conclusive as to his theory,

(318) but merely as theory, it is intelligible and defensible. It smacks, nevertheless of the Kantian Reason.


We turn next to a brief survey of the evidence concerning the part played by sensory and imaginal elements in the control of our movements. Here again, Mr. Woodworth has been among the most active critics of what he calls the orthodox view, and he has contributed not a little of the most interesting experimental evidence bearing on this issue. His colleague, Mr. Thorndike, has also been a stalwart advocate of essentially the same views as those advanced by Mr. Woodworth. On the fundamental issue as to whether or not we possess the imageless thought for which these authors stand sponsor, there is little to be added to what we have already said. But they bring forward in the case of voluntary control a number of considerations which do not figure in the discussion of reflective and constructive thinking. A few notes on these additional points may therefore be permitted.

Mr. Thorndike, for example,[15] alleges that in willing not to make a movement we frequently find present the resident and remote sensations caused by the movement itself (or the images of these sensations), which, according to the orthodox theory, as he interprets it, ought to be used only for the willing to make the movement. Therefore he infers that these sensations or images must be irrelevant to the actual production of the movement.

This case when it arises, instead of affording a difficulty, would seem to constitute a peculiarly telling confirmation of the importance of the sensations and images concerned. To will not to make a movement is in so far to have that movement in mind. If the mere act of this negative willing itself reveals these sensations and images present, when all that is required is the identifying of the movement to be inhibited, it would seem to offer strong presumptive evidence of their essential significance for executing the act to which they properly belong. In the particular case cited, they are ir-

( 319) -relevant, except as reliable representatives of the movement to be estopped, and in any event I do not understand Mr. Thorndike to allege that they are always present on such occasions.

Again, Mr. Thorndike writes that we cannot well possess images, whose resident sensations are not obtainable. He cites those of willing to move the eye smoothly across a line of print, and urges that because we now know that the eye does not move smoothly, but moves by jerks, we could never really have willed this act in the manner alleged by the orthodox view.

It may certainly be admitted as truistic that we cannot obtain images of sensations which we cannot experience. But so far as concerns controlling eye movements this fact appears to be a trifle irrelevant. It may well be in the case of eyemovement, which has so many reflex factors involved, that the resident sensations and images play an insignificant part in motor control. At least, this may easily be the case, if resident be taken to mean kinaesthetic and be set over against retinal factors. It does not appear, however, just what bearing this consideration has on the fact that a voluntary movement proves under scientific scrutiny not to be precisely what common sense has supposed. It requires some kind of 'cue' under its new guise, as well as under its former one. Whether this cue be sensory or imaginal, or consists in one of the pure thoughts, is a matter for further determination. The argument appears to have force 'only against the theory that voluntary control always involves the employment of imagery which reinstates strictly kinaesthetic sensations, a theory whose proof ought not to be demanded of the believers in the invariable presence of sensory or imaginal material of some kind at the outset of voluntary movement.

Again, Mr. Thorndike writes that it requires too much time to get the image of each movement in a series of coordinated movements, such as those of speech, or writing, or the playing of a musical instrument.[16] I have not been able to discover any orthodox (!) psychologist, who has made

( 320) statements reasonably interpretable as this assertion requires in order to have point, but I am quite prepared to believe that such statements are in existence. Certainly the view which has commonly been held by writers like James, for example, has been that a single cue may be adequate to release a long series of coordinated movements, provided only that these acts have been built up in some habitual manner, so that they are more or less automatized. Relative to this point, the author may remark that a number of experiments carried on in his own laboratory have revealed cases in which it has been possible to secure imagery very much more rapidly than the movements could be executed expressive of the imagery.

In the experimentation which Mr. Woodworth and his students have carried out, or which Mr. Thorndike has undertaken, there has been a crucial defect in method in the writer's judgment, which he has more than once pointed out. It is utterly unsound as a matter of method to give subjects verbal instructions concerning the movements which they are to make, and then to draw a negative inference concerning the part played by sensory and imaginal material, because, under such conditions, a certain percentage of these persons detect no imagery in the execution of the act a little later on. In the verbal cue, in this case, we find all that is needed to give precisely that 'set' to the nervous system, which Mr. Woodworth has himself so tellingly described. But the 'set' itself is occasioned by the sensorially apprehended directions. The author has participated in a long series of experiments directed to the analysis of this movement consciousness. Neither in himself nor in the many students who have served as subjects for these experiments, has there ever been a scintilla of real evidence for the initiation or control of a voluntary movement entirely without sensory or imaginal supervision. This control has often been found emanating from factors which ordinary introspection would have regarded as entirely irrelevant. For example, the control of the writing process has in certain of our subjects shown itself as dependent to a very marked degree on the sound made by the pen as it moves over

(321) the surface of the paper.[17] Ordinarily this experience is regarded as constituting a mere annoyance and as in no sense an essential part of the control. Some of Mr. Woodworth's irrelevant conscious facts may have a similar unsuspected function.

In most of the experiments which these authors have relied upon to make their case, they have used relatively habitual forms of action, in which only a slight cue is in any case necessary to precipitate the act. In the other type of case which they cite, where a relatively new coordination is brought under control, the evidence seems to the writer perfectly convincing as regards the presence in an essential way of sensory or imaginal factors. For example, to look at the foot, or to attend to sensations from it, while trying to gain control over the muscles of the toes, is certainly in no sense to produce a situation in which sensory direction is entirely wanting.[18]

There is, here again, as in connection with the topic of reflective thought, a tendency to magnify unduly those features of other authors' writings, which they feel confirm their own views. This is no doubt an entirely unintentional over-emphasis on one side of the case. Bair, for example, whose work has sometimes been adduced in support of this view of the pure thought control over movement, makes it perfectly obvious as I read him, that in his experiments, sensory factors were repeatedly found indispensable to securing control over the muscles of the ear.[19] The author has already made quotations from Stout, which show how different his view is from that held by the gentlemen now under discussion. Of him Mr. Woodworth writes:

"Some authors, Stout, Binet, and recently Bühler, boldly assert the existence of imageless thought." (Journal of Philosophy, Psychology, and Scientific Methods, Vol. III., 1906, p. 701.)

One does not like to call attention in a spirit of carping criticism to matters of this kind, but if the appeal is in any

( 322) way to be made to the authority of opinion, it is fair that this opinion should be ranged where it belongs.

On the whole, the author is much more disposed to be dogmatic about the case of voluntary control, than about the case of reflective thinking. His own experience, and that of a considerable number of expert persons, who have worked on the problem in his own laboratory and elsewhere, leads him to utter skepticism of the control of voluntary movement, without sensational or imaginal factors. Moreover, there is good indirect reason why such control should not occur, although evidence of this type, if unsupported by more strictly empirical evidence, would have no especial significance. But if one thing stands out more clearly than another in the general contributions of comparative psychology, to our general theory, it is that the original reactions of animals are made in response to sensory stimulations, and that the primal significance of the reactions is to be found in the adjustment which they represent to the stimulating objects. Memory, when it makes its appearance, also finds its significance in the guidance of these movements of reaction, and it is memory of the sensory incentives and consequences—not memory of the purely extra-sensorial kind. If, at a later stage of development, thought processes are developed, which have no direct dependence upon, nor connection with, the world of sensory experience, such thought processes must have only a very secondary reference to the control of ordinary muscular activities. But, as said above, considerations of this kind are of altogether secondary consequence as compared with the direct introspective evidence gained under experimental conditions. For himself, the author is perfectly clear that these control conditions are always of a sensory or imaginal character. But it needs to be repeated that for him neither sensation nor image as such is to be regarded as isolated from the correlative fact of meaning, and meaning is not for him in any sense equivalent to the mere sensuous content of the mental state, whether it be dominantly conditioned by peripheral or central stimulation.

Finally, then, I find the doctrine of imageless thought open to suspicion on the following points:

1. The method of its experimental investigation is at least not wholly satisfactory in meeting the demands of ordinary experimental procedure. This is true, both as regards the problem of reflective consciousness, and that of voluntary muscular control.

2. Imageless thought seems with many observers to be a sporadic and occasional phenomenon. Its appearance is not in their cases invariably connected with any special kind of situation. It consequently lacks one usual characteristic of well established thought forms, and may therefore well be regarded a trifle askance.

3. Unless purely functional and logical terms be used, it seems almost impossible to describe it, save in negative terms. This suggests either that the analysis is not yet complete, or that the thing analyzed is not really a content of consciousness.

4. If Mr. Woodworth's variety be accepted as genuine, we must apparently recognize two generically different kinds of thought material to serve one general function. This will be at variance with our conceptions of the parsimony of nature. Mr. Woodworth's effort to exhibit the two forms as respectively primary and secondary to one another, will scarcely serve to avoid the factual difficulties even though it afford a logical solution. Meaning and imagery differ from one another in essential function not as does one form of imagery from another form, merely in quality. If we accept Biihler's conception, we have ideas limited in their function in a manner utterly at variance with ordinary opinion.

5. There are many well recognized conscious states which may obviously be readily confused with imageless thought. The consciousness of attitude, springing out of very primitive physiological attitudes, is an important case in point.

6. The presence of interpretative factors in perception gives no real comfort to belief in imageless thinking.

My own conclusion is that at present, the only demonstrable imageless thought is subconscious, and so primarily a matter of cerebralistic physiology. Even this would be imaginal, if it got above the limen. But I shall try to keep an open mind.


  1. An address delivered at Columbia University, February 3, 1911.
  2. Phil. Rev., VI., 1897, p. 532, p. 646; Vol. VII., 1898, p. 72.
  3. L'Etude experimentale de L'Intelligence,' p. 108, Paris, 1903.
  4. Op. Cit., pp. 180-182.
  5. 'Studies in Logical Theory,' John Dewey and others, p. 183.
  6. Inasmuch as the main contentions of this paper are definitely critical of the experiments and the theories of Mr. Woodworth and Mr. Thorndike among American writers, and of the Wurzburg school and its critics among the Germans, to whom more explicit reference will be made below, it is only fair to myself to say that I value very highly the influence of their work in stimulating psychologists to an analysis of the thought processes keener and more searching than any previous generation has known. That I think some features of their procedure open to objection, and that I demur to their conclusions in certain essential particulars, does not at all diminish my sense of obligation for arousing us all to a new interest in problems previously too much exempted from experimental analysis. I am indebted in a personal way for such criticisms on my own published views as have enabled me to clarify to myself, and I hope to others, the convictions which I hold and the grounds upon which they rest.
    The more important of the recent German monographs and papers touching the general topics at one point or another are: Ach, 'Ueber Willensthaetigkeit u. d. Denken,' 1905; Watt, 'Exp. Beitraege z. einer Theorie d. Denkens,' Arch. f. d. ges. Psy., IV., 1905, p. 289; Messer, 'Exp. psycho'. Untersuch. u. d. Denken,' Arch. f. d. ges. Psy., VIII., 1906, p. 1; Buhler, ‘Tatsachen u. Probl. z. einer Psycho'. d. Denkvorgange, Arch. f. d. g. Psy., IX., 1907, p. 297; Starring, 'Exp. Untersuch. fib. einf. Schlussprozesse,' Arch. f. d. ges. Psych., XI., 1908, p. 1; Durr, 'Ueber exp. Untersuch. d. Denkvorgange,' Zeitschr. f. Psy., XLIX., 1908, p. 313; van Aster, 'D. Psycho'. Beobachtung u. exp. Unters. d. Denkvorgange,' Zeitschr. f. Psy., LXIX., 1908, p. 97.
  7. Jour. Phil., Psy., Sc. Meth., IV., 1907, p. 170.
  8. Jour. Phil., Psy. & Sci. Meth., III., 1906, p. 707.
  9. In this connection should be mentioned the paper of Dürr who took part in many of the Wurzburg experiments and who writes critically of Bühler's more extreme positions. Zeit. f. Psychol., XLIX., 1908, p. 313.
  10. Op. cit.
  11. Wundt says: "The acts gurus of the thought experiments is no fact of observation, but simply a consequence of defective observation and false presupposition." (Psych. Studien., Vol. III., 1907, p. 347. 
    And again: "If one abstracts from the logical reflexions of the observers, the results show that the observers have observed nothing! The thought stood as a whole clearly before consciousness. But it was disembodied, incorporeal. Here at last we have come again upon the acts purus of the scholastics." (Ibid., 344-5.)
  12. Experimental Psychology of the Thought Processes, p. 145.
  13. Am. Jour. Psych., Vol. XX., 1909, p. 277.
  14. For reasons which the present writer has made clear in a previous paragraph, the inclusion of Stout in this list appears to be unwarranted.
  15. Journal of Philosophy, Psychology and Scientific Methods, Vol. IV., 1907, p. 40.
  16. Burnett enlarges upon this point in a paper reporting experimental observations. 'Studies in Phil. and Psycho'.'; Garman Commem. Volume, p. 393.
  17. Cf. Downey, 'Control Processes in Modified Handwriting,' Psycho'. Rev. Monog., Vol. IX., 1908, Whole No. 37.
  18. Cf. Woodworth, 'Le Mouvement,' p. 330
  19. PSY. REV., 1901, Vol. VIII., p. 476.

Valid HTML 4.01 Strict Valid CSS2