Book /Li: SMITHSONIAN DEPOSIT. Digitized by the Internet Archive in 2011 with funding from The Library of Congress http://www.archive.org/details/processofabstracOOmoor UNIVERSITY OF CALIFORNIA PUBLICATIONS IN PSYCHOLOGY Vol. 1, No. 2, pp. 73-197, 6 text- figures November 12, 1910 THE PROCESS OF ABSTRACTION AN EXPERIMENTAL STUDY Eg THOMAS VERNER MOORE BERKELEY THE UNIVERSITY PRESS UNIVERSITY OF CALIFORNIA PUBLICATIONS Note.— The University of California Publications are offered in exchange for the publi- cations of learned societies and institutions, universities and libraries. Complete lists of all the publications of the University mil be sent upon request. For sample copies, lists of publications or other information, address the Manager of the University Press, Berkeley, California, U. S. A. All matter sent in exchange should be addressed to The Exchange Department, University Library, Berkeley, California, U. S. A. OTTO HARRASSOWITZ R. FRIEDLAENDER & SOHN LEIPZIG BERLIN Agent for the series in American Arch- Agent for the series in American Arch- aeology and Ethnology, Classical Philology, aeology and Ethnology, Botany, Geology, Education, Modern Philology, Philosophy, Mathematics, Pathology, Physiology, Zool- Psychology. ogy, and Memoirs. Cited as Univ. Calif. PuM. Psychol. PSYCHOLOGY. —George M. Stratton, Editor. Vol. 1. 1. The Judgment of Difference, ■with Special Reference to the Doctrine of the Threshold, in the Case of Lifted Weights, by Warner Brown. Pp. 1-71, 4 text figures. September 24, 1910 50 2. The Process of Abstraction, an Experimental Study, by Thomas Verner Moore. Pp. 73-197, 6 text figures. November 12, 1910 1.00 PHILQSOPKY-^George H. Howison, Editor. Price per volume, $2.00. Cited as Univ. Calif. Publ. Philos. The first volume of the University of California Publications in Philosophy appeared in November, 1904, and was prepared in commemoration of the seventieth birthday of Pro- fessor George Holmes HoWison, under the direction of a committee Of his pupils composed of Evander Bradley McGilyary, Charles Henry Rieber, Harry Allen Overstreet and Charles Montague Bakewell, The price of the volume is $2.00. It may be had bound, or the papers may be obtained separately. Its contents are: Vol. 1. 1. The Summum Bonum, by Evander Bradley McGilvary. Pp. 1-27 $0.25 2. The Essentials of Human Faculty, by Sidney Edward Mezes. Pp. 28-55 .25 3. Some Scientific Apologies for Evil, by George Malcolm Stratton. Pp. 56-71 ... 15 4. Pragmatism and the a priori, by Charles Henry Rieber. Pp. 72-91 .20- 5. Latter-Day Flowing-Philosophy, by Charles Montague Bakewell. Pp. 92-114 .................. 20 6. Some Problems in Evolution and Education, by Ernest Norton Hender- son. Pp. 115-124 „ 10 7. Philosophy and Science in the Study of Education, by Jesse Dismukes Burks. Pp. 125-140 .. ...... .._ 15 8. The Dialectic of Bruno and Spinoza, by Arthur Oncken Lovejoy. Pp. 141-174 .............. 35 9. The Logic of Self -Realization, by Henry Waldgrave Stuart. Pp. 175- 205 30 10. Utility and the Accepted Type, by Theodore de Lopez de Laguna. Pp. 206-226 „ 20 11. A Theory of the Syllogism, by Knight Dunlap. Pp. 227-235 10 12. The Basal Principle of Truth-Evaluation, by Harry Allen Overstreet. Pp. 236-262 .: 25 UNIVERSITY OF CALIFORNIA PUBLICATIONS IN PSYCHOLOGY Vol. 1, No. 2, pp. 73-197, 6 text-figures November 12, 1910 THE PROCESS OF ABSTRACTION AN EXPERIMENTAL STUDY BY THOMAS VEENEE MOOEE. CONTENTS. PAGE Introduction 74 I. Literature of the Problem 76 II. The Method of Eesearch 116 1. The Problem and the Experiments 116 2. The Apparatus 119 3. Instructions to the Subject 119 4. Classification of the Experiments 120 III. Experimental Eesults 122 1. The Analysis of the Groups 122 (a) Isolation of the Common Element 122 (6) The Disappearance of the Surrounding Elements 124 2. The Process of Perception 127 3. The Factor of Memory in the Process of Abstraction 139 (a) The Method of Memorizing 139 (6) Memory as Belated to the Sequence of the Sur- rounding Figures 153 (c) Memory as Eelated to the Focality of Perception.... 158 4. The Process of Eecognition 160 (a) Analysis of the Experiments 160 (&) Interpretation of the Eesults 172 (i) The Immediate Experimental Conclusions 172 (ii) The Basis of Judgment in Eecognition 176 IV. The Product of the Process of Abstraction 180 Summary 190 A List of Eeferences 192 Appendix I: The Influence of Association on Perception 194 Appendix II: Generic Images 196 74 University of California Publications in Psychology. [To 1 - 1 INTRODUCTION. The decade that is just now drawing to a close has witnessed a notable extension of the field of psychological research. In the beginnings of modern psychology the field of experiment was seldom extended beyond the domain of sensation. Progress in physics and physiology made it possible to subject our sensa- tions to experiment, and for some time the sensory processes received the chief share of the attention of psychologists. It was not long, however, before the emotions began to receive their due amount of consideration and the invention of the plethys- mograph opened the way to a new line of research. But only within the last ten years has the experimental study of such higher processes of thought, as abstraction, commenced to de- velop. The impetus to this new development has come mainly from Professor Oswald Kiilpe at the University of "Wiirzburg. The present research, although its origins are not to be traced to the school of Wiirzburg, belongs to the field which Professor Kiilpe and his students have so admirably developed. Our problem was to study the mental processes involved in the formation of our abstract ideas. It is indeed true that the very existence of such ideas has been called in question. Still we may at present assume, for the purpose of stating our problem, that it is possible for the mind to perceive a series of objects, and to recognize some one quality or group of qualities as recurring constantly in every member of the series. The botanist exam- ining a set of specimens will classify them according to certain characteristics which mark off the genera and species. The group of characteristics constitutes what may be termed his concept of the genus or species that he has segregated. Of each species he has a more or less definite "concept," by which he can represent to himself a number of specimens, no two of which are precisely the same. Such "concepts," whatever may be their real nature, are facts of conscious experience; we form 1910] Moore: The Process of Abstraction. 75 them and use them incessantly. But what after all is the "concept"? What is the process of its formation? This is the problem of the present research. The history of the problem dates back to the days of the Greek philosophy; but only within the last few years has it been subjected to an experimental investigation. The more recent literature is of immediate interest for our present prob- lem. The metaphysical discussions, valuable as they are within their own sphere, bear only indirectly on the experimental question. Consequently, only the experimental literature bear- ing in some manner on the process of abstraction has been analyzed. Not every allusion in the extensive psychological literature of the day could be picked out, but a general account of the important pieces of experimental work from Galton to the present time has been given. The individual studies have been analyzed with some completeness because the history of the literature is an integral part of the evidence on one point of the present study, viz.: Is there or is there not a distinction between thought and imagery ; and if so, in what sense is thought to be interpreted? 76 University of California Publications in Psychology. t Vo1 - 1 LITERATURE OF THE PROBLEM. The first experiments which in any way approached the domain of our abstract ideas were made by Francis Galton in 1878. In the Proceedings of the Royal Institution of Great Britain for 1879 1 was published his memoir on Generic Images. In this article he refers to an earlier one in the Journal of the Anthropological Institute for 1878. 2 The bearing of these ex- periments on abstraction is suggested rather than direct. But they have become the basis of the now famous composite-image theory of ideas, and are therefore deserving of mention. Galton described in this article the composite photographs which he had just succeeded in obtaining. These he compared to "our general impressions." Just what he meant by "our general impressions" is not clear; but he congratulates himself that his explanation coincides with that of Professor Huxley in his work on Hume. "I am rejoiced," he says, "to find that from a strictly physiological side this explanation is considered to be the true one by so high an authority, and that he has, quite independently of myself, adopted a view which I also enter- tained, and had hinted at in my first description of composite portraiture, though there was no occasion at that time to write more explicitly about it." 3 Huxley's meaning is clearer, and to him we may turn for an outline of the theory. In the above-mentioned work on Hume, the following quotation gives a clear idea of the generic image theory of general concepts. i Pp. 161-171. 2 This article is mainly concerned with the technique of composite photographs. 3 Proceedings of the Royal Institution of Great Britain, 9, 1879-1881, p. 166. 1910] Moore: The Process of Abstraction. 11 "Now when several complex impressions which are more or less different from one another — let us say that out of ten impressions in each, six are the same in all, and four are dif- ferent from all the rest — are successively presented to the mind, it is easy to see what must be the nature of the result. The repetition of the six similar impressions will strengthen the six corresponding elements of the complex idea, which will therefore acquire greater vividness; while the four differing impressions of each will not only acquire no greater strength than they had at first, but in accordance with the law of asso- ciation, they will appear at once, and will thus neutralize one another. "This mental operation may be rendered comprehensible by considering what takes place in the formation of compound photographs — where the images of the faces of six sitters, for example, are each received on the same photographic plate for a sixth of the time requisite to take one portrait. The final result is that all those points in which the six faces agree are brought out strongly, while all those in which they differ are left vague; and thus what may be termed a generic portrait of the six, in contradiction to a specific portrait of any one, is produced. . . . "The generic ideas which are formed from several similar, but not identical, complex experiences are what are commonly called abstract or general ideas; and Berkeley endeavored to prove that all general ideas are nothing but particular ideas annexed to a certain term which gives them a more extensive signification, and makes them recall, upon occasion, other indi- viduals which are similar to them. Hume says that he regards this as 'one of the greatest and the most valuable discoveries that has been made of late years in the republic of letters,' and endeavors to confirm it in such a manner that it shall be 'put beyond all doubt and controversy.' ' ' I may venture to express a doubt whether he has succeeded in his object; but the subject is an abstruse one; and I must content myself with the remark, that though Berkeley's view 78 University of California Publications in Psychology. [Vol. 1 appears to be largely applicable to such general ideas as are formed after language has been acquired, and to all the more abstract sort of conceptions, yet that general ideas of sensible objects may nevertheless be produced in the way indicated, and may exist independently of language." 4 It would thus seem that Huxley's theory — and probably Galton's also — is that only our abstract ideas of sensible objects are to be compared with the composite photographs. Galton points out that the mind in forming its generic images is much less perfect in its mechanism than the camera. "Our mental generic composites are rarely defined; they have that blur in excess which photographic composites have in a small degree and their background is crowded with faint and incongruous im- agery. The exceptional effects are not overmastered, as they are in the photographic composites, by the large bulk of ordinary effects." 5 The experiments on composite photographs were not experi- ments on abstract ideas — they merely suggested a theory of general concepts. Nor did Galton's later experiments on mental imagery 6 approach very much nearer the problem. They called general attention to mental imagery and perhaps stimulated the next investigation of any importance 7 which was made by Ribot. In October, 1891, M. Ribot published in the Revue Philoso- phique 8 his "Enquete sur les idees generates. " This he after- wards amplified in the fourth chapter of his book, L'E 'volution des idees generales. 9 His problem was this : At the moment of thinking or reading or hearing a general term, what is there in consciousness — immediately and without reflection? On the basis of the imagery which his subjects reported he classified 4 Huxley, David Hume. New York, 1879, pp. 92-94. s Loc. cit., p. 169. e Cf. Inquiries into Human Faculty, 1883, Section on Mental Imagery. 7 The article entitled ' ' Observations on General Terms, " by S. E. Wiltse, in the American Journal of Psychology, 3, 1890, pp. 144-148, was only tentative and contained no definite results. s Vol. 32, pp. 376-388. 9 Paris, 1897. 1910] Moore: The Process of Abstraction. 79 them into: (1) The concrete type (visual or muscular imagery of an object). (2) Typographic visual type (visual image of the printed word). (3) Auditory type. A great many of his subjects said that they had nothing in mind. For example, fifty per cent, of the answers to the imagery of the word 'cause' said that the subjects had represented to themselves nothing at all. M. Ribot then asked himself the question, "What is this "nothing"? The word alone? No. Otherwise there would be no difference between a general term and a word of a language that one did not understand. The word is a sign of some object. We have learned the mental habit of designating many objects that have some point of agreement by this symbol. The objects designated lie hidden and are unconsciously represented by this general term. "General ideas are habits in the intellectual order. ' ' Our higher concepts consist of two elements — one clear and conscious, and this is always the word which may at times be accompanied by some shred of imagery. The other element is obscure and unconscious. M. Ribot refrains from saying precisely what this obscure and unconscious element is. From the context it would seem that he means the unconscious trace left by the habitual use of the word to designate various objects. A word of criticism may be said in passing. M. Ribot 's interpretation of this "nothing," which accompanies the per- ception of a general term, is purely theoretical and is not based on any published data given by his subjects. Furthermore, he has not followed out his sign theory to its logical conse- quences. For every sign, we have on the one hand the object signified and on the other, the signification. Smoke has on the one hand fire, of which it is a sign, and on the other a signifi- cation in the mind of the observer. If, then, general terms are signs, they have on the one hand the objects that they signify, and on the other a signification. This signification, as M. Ribot admits, is not the image and not the word itself. It certainly is not the unconscious factor he speaks of — for he would scarcely maintain that his subjects were not conscious of the meaning of the word. It is therefore a clearly conscious mental process distinct from both the image and the word. 80 University of California Publications in Psychology. t Vo1 - 1 M. Ribot's work comes closer to being an experimental study of the problem than Galton's experiments on composite photo- graphs. Had he examined the state of mind of his subjects when they lacked imagery and not trusted to theory on that point, he would have carried his investigation into the heart of the problem of abstract ideas. After Galton and Ribot the study of mental imagery took up no small portion of the time and labors of psychologists. But the extensive literature on imagery is not directly concerned with the problem of abstraction. An excellent piece of work in this field of research is that of William Chandler Bagley of Cornell University. 10 He under- took to study the effect of imperfectly formed words on the perception of spoken sentences, and parts of sentences. The words and sentences were first recorded by a phonograph, the initial, middle, or final consonants being unpronounced. The subject listened to the phonograph, and was called upon to repeat what he heard, and analyze the mental processes he experienced. The section of the work which bears upon our problem is that entitled "The Conscious Process Involved in the Apperception of Spoken Symbols." In comparison with the later German experiments, Dr. Bagley 's are remarkable for the very frequent occurrence of imagery of one kind or another. His subjects in perceiving the meaning of sentences report with surprising frequence the presence of visual, auditory or kin- esthetic imagery. This might be due to the fact that Dr. Bagley 's subjects were capable of sharper introspection than the German psychologists. Still this can hardly be the case. The German psychologists, among whom are such men as Pro- fessor Kiilpe, cannot be supposed to be lacking in the power of introspection. Another possible explanation is that Bagley laid special stress upon the report about imagery and in that way developed in his subjects a "task" to associate definite images with the given sentences. 10 "The Apperception of the Spoken Sentence: a Study in the Psychol- ogy of Language." American Journal of Psychology, 12, 1900-01, pp. 80- 134. 1910] Moore: The Process of Abstraction. 81 Though his subjects generally experienced imagery in the perception of the meaning of a sentence, still he finds cases in which this imagery is lacking. One factor which has to do with the lack of imagery is that "familiarity with the sentence sometimes militates against a clear and direct reference on the part of the observer." In this his results agree with those of Dr. Taylor reported below. 11 Bagley's general conclusion is that, "The consciousness concomitant with the apperception of auditory symbols is made up of sensational and affective ele- ments — some peripherally, some centrally aroused — in connec- tions which vary in character with different individuals and under different conditions. These connections are arranged in patterns which change rapidly into one another, and are in general transitory and fleeting. When the attention is directed to the peripherally excited elements exclusively — when the ex- ternal stimuli occupy the burning point of apperception — the meaning which they as symbols should convey is not clearly apperceived. When the attention is directed upon the centrally aroused ideas which the symbols suggest, the 'meaning' is ap- perceived, but errors and lapses in the stimuli are apt to pass unnoticed." (p. 125.) He thinks that Stout goes too far in suggesting the existence of representative mental contents different from "visual, audi- tory, tactual, and other experiences." He thinks that his ex- periments lead him to no such conclusion. "From the series of observations which were made in the course of our experi- ments, no conscious 'stuff' was found which could not be classed as sensation or affection, when reduced to its ultimates by a rigid analysis. Neither do our experiments show that there is in the apperception of spoken sentences such a thing as 'image- less apprehension.' " (p. 126.) Still Dr. Bagley finds something which he does not feel justified in putting down as either imagery or feeling. To this something which is not imagery or feeling and still has to ii See p. 87. 82 University of California Publications in Psychology. [Vol. 1 do with the understanding of the sentence, he applies the name 'mood.' "We may say with Stout that the new is referred to a mental 'system/ in so far as such a system is a mood, an attitude, a tendency, an adaptation. The mind adjusts itself uniformly to uniform conditions: this seems to be the essence of the apperceptive 'mood.' When C, in the sentence "The play was bad," interpreted 'play' as a drama, her mind adapted itself in a degree to the drama environment. This was not necessarily a focal reference to a given play, but the mind was in the dramatic 'mood.' Should particular parts of a typical play-environment have been ideally reproduced, the situation would only have been reinforced. Should certain verbal ideas such as 'drama,' 'theaters/ 'Shakespeare,' etc., have been reproduced in consciousness, either visually, auditorily or kinesthetically, these ideas would have been constituents of the dramatic 'mood,' but not necessarily the fundamental con- stituents. The fundamental constituents may and do vary from time to time. Only very seldom can they be called constant, and the 'constant supplements' which we have noticed are in- stances of such occasions. The fact that the focal constituents of the apperceptive consciousness are not necessarily consistent with the situation represented bears testimony to this point of view. "There was not room for a stove in the corner"; with this sentence one observer imaged distinctly a stove in the corner of a small, otherwise bare room. His own surprise at the inconsistency of this imagery was shown by his exclamation upon recording the introspection: "But there was a stove there!" (p. 127.) A 'mood/ therefore, is something that has to do with the past experience of the subject in regard to the words of the sentence that is understood. Just what it is, as a present psy- chical state, Dr. Bagley does not say. It is the revival of past experience. It is not mental imagery, although mental imagery enters in as a partial element in the complex termed 'mood.' If this is so, what is that present psychical state in the mood, which is neither imagery nor feeling ? It is not past experience, for the past is not present. It is not revived imagery and 1910] Moore: The Process of Abstraction. 83 feeling, for Dr. Bagley admits that the mood contains something besides imagery and feeling. It therefore seems that Dr. Bagley has found something more than he is willing fully to recognize. His experiments, like those of the later German writers, reveal the existence of an imageless mental content. Just what we call this is not of prime importance. But its existence should be recognized. In 1901 appeared Marbe's 12 experimental study of judgment. His main problem is not that of the present work, but it is an early attempt to apply the experimental method to supra- sensuous mental processes. He also mentions, as a side issue, his attempt to develop the experiments of Ribot. The work was done in Professor Kiilpe's laboratory at the University of Wurzburg, where Dr. Marbe was at the time Privatdozent of Philosophy. The method is essentially the same as that of the later Wurzburg experiments which are described below. His final conclusion in regard to the judgment is that any special mental process — a word or gesture or image — may become a judgment. Taken literally, this conclusion is not borne out by the experiments. What they seem to prove is that a judgment may be signified by a variety of different processes. And this may be the author's meaning (cf. Chapter III). The perception of a judgment, however, is not a sensation or an image or a feeling, or anything that can be pointed out in consciousness. The perception of a judgment is a knowing — a "Wissen" (cf. p. 17). In his concluding remarks on experimentation in the domain of logic Dr. Marbe mentions Ribot 's work, L' Evolution des idees generates, and refers to some similar experiments of his own on ideas and imagery. His meaning is not clear to me, so I quote entire the brief account of his work in this field (pp. 99- 101). "Seit den Zeiten des Sokrates hat man angenommen, da8 den Begriffen im Bewusstsein ausser den zugehorigen Worten irgend etwas direkt entspreche, d. h., daB es neben diesen Worten 12 K. Marbe, Experimentell-psychologische Untersuclmngen iiber das Urteil. Leipzig, 1901. 84 University of California Publications in Psychology. [ Vo1 - 1 psychische Gebilde gabe, welche der Gesamtheit der Gegen- stande, auf welche sich die Worte beziehen, korrespondieren sollen. Diese Gebilde wurden urspriinglich, wie gelegentlich noch heute im Gegensatz zu den Worten, die nur Zeiehen ihrer Bedeutungen sind, als Abbilder derselben aufgefasst, indem sie die gemeinschaftlichen Merkmale ihrer Gegenstande im Bilde enthalten sollten. Solche psychischen Gebilde hat man spater je nach dem Grade der Abstraktheit, den man ihnen zuschrieb, bald als Gemeinbilder, bald als allgemeine Vorstellungen, bald als Begriffsvorstellungen bezeichnet. Obgleich, wie bekannt, ihre Existenz schon im Altertum nnd Mittelalter bestritten und in der Neuzeit hauptsachlich dureh Berkeley bekampft wurde, so halt man doch auch heute noch vielfach in der einen oder anderen Form an derselben fest. Auch die Frage, ob es solche psychologische Aquivalente der Begriffe giebt, ist eine rein psy- chologische, und ihre Behandlung sollte nicht, wie es in der Regel geschieht, mit logischen Untersuchungen vermiseht wer- den. Die Aussagen unserer Versuchspersonen liber die Bewusst- seinsvorgange, welche sie nach dem Erleben von Urteilsworten und Urteilssatzen zu Protokoll gaben, enthalten iibrigens nichts von solchen Parallelerscheinungen der Begriffe, ebensowenig, wie die wertvollen Untersuchungen von Ribot, 13 in welchen dieser Forscher einer Reihe von Beobachtern Substantiva zurief, um sich dann von ihnen sagen zu lassen, was die gehorten Worte fur Erlebnisse auslosten. Ich selbst habe mehreren Beobachtern ca. 20 Substantiva zugerufen und mir dann von ihnen berichten lassen, was fur Erlebnisse die zugerufenen Worte erzeugten. Dann gab ich denselben Beobachtern der Reihe nach verschiedene Karten in die Hand, auf welchen jeweils ein Substantivum auf- gedruckt war, wahrend sie nach einigen Augenblicken die Erleb- nisse zu Protokoll geben mussten, die dureh den Anblick der gedruckten Worte in ihnen ausgelost wurden. Endlich stellte ich ihnen die Auf gabe, die Begriffe: Baum, Volk, Gesellschaft, Zeit u. a. zu denken und mir dann die Resultate ihrer Bemii- hungen mitzuteilen. In alien diesen Fallen zeigten die Proto- 13 L'Evolution des idees generates. Paris, 1897, p. 127ff. 1910] Moore: The Process of Abstraction. 85 kolle nichts von Begriffsvorstellungen u. dergl. Die Erlebnisse der Beobachter bestanden vielmehr aussehliesslich in Wahrneh- mungen, Vorstellungen und Bewusstseinslagen, die teilweise gefiihlsbetont, teilweise o-hne jeden Gefiihlston verliefen. Man wird also wohl sagen diirfen, daB es keine psychologischen Aquivalente der Begriffe im Sinne der Begriffsvorstellungen giebt. Jedenfalls aber sehen wir leicht ein, daB auch diese Frage experimentell behandelt werden kann und muB und daB sie die Logik weiter nicht tangiert." The obscurity arises from the fact that it is not perfectly clear whether Marbe merely denies the existence of the general images which Bishop Berkeley 14 termed abstract ideas, or that he claims that there are neither general images nor universal concepts. It would seem, however, that Marbe found no evidence for the existence of a general image in the understanding of the words given to his subjects. He did find, however, Wahrneh- mungen, Vorstellungen, and Bewusstseinslagen. This latter is a term introduced by Mayer and Orth 15 to represent certain "states of mind" which are more or less refractory toward all 14 <■ ' Whether others have this wonderful faculty of abstracting their ideas, they can best tell; for myself I dare be confident I have it not. I find indeed I have indeed a faculty of imagining, or representing to myself, the ideas of those particular things I have perceived, and of variously compounding and dividing them. I can imagine a man with two heads, or the upper parts of a man joined to the body of a horse. I can consider the hand, the eye, the nose, each by itself or separated from the rest of the body. But then whatever hand or eye I imagine, it must have some particular shape and colour. Likewise, the idea of man that I frame to myself must be either of a white, or a black, or a tawny, a straight or a crooked, a tall or a low, or a middle-sized man. I can not by any effort of thought conceive the abstract idea described [in his previous account of the abstract ideas of the traditional logic]. And it is equally impossible for me to form the abstract idea of motion distinct from the moving, and which is neither swift nor slow, curvilinear nor rectilinear; and the like may be said of all other abstract general ideas whatsoever." A Treatise concerning the Principles of Human Knowledge. Introduction, 10, pp. 141-142, vol. I of Fraser's Oxford (1871) Edition of his Works. It is evident from the context that Bishop Berkeley does not dis- tinguish between the mental image and the abstract concept — between what is termed by the later Wiirzburg School the Vorstellung and the Begriff. is < ' Zur qualitativen Untersuchung der Association. ' ' Zeitschrift fur Psychologie und Physiologie, 26, 1901, p. 6. 86 University of California Publications in Psychology. [Vol. 1 analysis — and in which images (Vorstellungen) are not to be found. In a later study 10 Dr. Orth attempted to show that the Bewusstseinslage was not a state of feeling. That part of Dr. Narziss Ach's work Uber die Willenstdtig- Jceit und das Denken, 17 which refers to thought has a direct relation to this line of work, which we may consider as originat- ing in Ribot's "Enquete sur les idees generales. " The Bewusst- seinslage of Marbe appears under the name of Bewtisstheit. This author too has recognized the existence of mental states in which there "could not be detected any such phenomenal elements as visual, auditory, or kinesthetic sensations or memory pictures of such sensations which qualitatively determined the mental content reported as knowledge" (p. 210). There are often present along with such states of consciousness words or fragments of words. "Such a presence of kinesthetic or audi- tory kinesthetic images may well be the cause of the widely dis- seminated hypothesis that our thought continually takes place in an inner speech or adequate visual, acoustic, or similar kinds of memory images. Against such a view one must point to the fact that there are very complex contents in which, as already mentioned, the partial contents are consciously represented in their manifold opposing relations and still these individual con- tents are not expressed by any adequate vocal designations and the like — and indeed, it is absolutely impossible that they should be" (p. 215). The question then arises, what are these imageless states of consciousness ? This Dr. Ach explains by an example : ' ' Every idea which is given in consciousness, for example, the word 'bell' puts, as is well known, a number of ideas in readiness, with which it stands in associative connections. This putting of ideas in readiness, or stimulation of tendencies to reproduc- tion, suffices for the conscious representation of what we call is Dr. Johannes Orth, Gefuhl und Bewusstseinslage. Sammlung von Abhandlungen aus dem Gebiete der Padagogischen Psychologie und Physi- ologie, edited by Ziegler and Ziehen. Vol. 6, No. 4. Berlin, 1903, cf. especially pp. 69-75. it Gottingen, 1905. 1910] Moore: The Process of Abstraction. 87 'meaning' without its being necessary that the ideas should act- ually become conscious" (p. 217). A nonsense syllable or a word in an unknown language does not place in readiness any such set of tendencies to reproduction and consequently has no meaning. Every signification and every idea is an associative abstraction because it picks out some of a vast number of pos- sible associations. And no signification is identical with any other but only more or less analogous. 18 The problem of the understanding of words and sentences was taken up by Dr. Clifton 0. Taylor in 1906. 19 His first experiment was based upon a similar one made by Marbe in the "Philosophischen Gesellschaf t " during the winter semester of 1904, and may be considered as a continuation of Marbe 's line of work. He read to his subjects the following sentence : "Imagine that in a rectangular space a plane is laid passing through the upper and lower edges of two opposing sides. The plane then must stretch obliquely through the space. How many such planes can you imagine in this space?" Then fol- lowed seven subordinate tasks based upon this fundamental problem. From the protocol obtained from his subjects it was evident, that for the understanding of sentences expressed in concrete terms the development of mental images can be useful, but that they are not indispensable. These auxiliary images be- come less frequent the more familiar the subject is with the text. A second experiment was carried out in which the subject read a text from Gegenbauer's "Anatomic" He had to take care that he understood the text perfectly, and while reading marked the places where he experienced any mental imagery. Visual imagery aided materially in understanding the text. But on writing out the text and then rereading the written copy, the imagery was reduced from fourteen pictures to but one in the third reading. On the other hand, from experiments with is For a criticism of Dr. Ach's view see below, p. 181 ff. 19 ' - Ueber das Verstehen von Worten und Satzen. ' ' Zeitschrift fiir Psychologie, 40, 1906, pp. 225-251. 88 University of California Publications in Psychology. ITol. 1 sentences composed of abstract terms, it seemed that the ap- pearance of images hindered rather than helped the under- standing of the text. With Taylor, as with other members of this school, the Beivusstseinslage comes into prominence, and it is found that these attitudes of consciousness are the more frequent as the subject is less familiar with his text. "We must now go back a few years to an independent line of research. Two years after the appearance of Marbe's ex- perimental study of judgment, .Binet published his brilliant L' Etude experimental de V intelligence.- His experiments were commenced before the appearance of Marbe's work, for we read on page 76 of a series made in November, 1900. Binet may therefore be considered as a real pioneer in this field. He must too have exercised no little influence on later German authors, for he showed how the method of controlled introspection could give very valuable assistance in the study of our higher mental processes. He used in his experiments a variety of subjects, but most of all his two daughters, Marguerite fourteen years old, and Armande thirteen. His first experiment was to give them the task of writing twenty words — any that they might wish. By questioning he then found out in what sense the words had been used — and how they came to be thought of. By class- ification of the words used in repeated experiments M. Binet found very characteristic differences in the vocabularies of his two children. Their environment having always been the same, this difference, he concluded, was due to their temperaments. Temperament therefore has its influence on our choice of words. From the fact that the words written formed well-defined groups, M. Binet concluded that association alone does not entirely account for our train of thoughts. Association ac- counts for word after word in any group — but it does not account for the origin of a new group. In another series of experiments the two little girls were given a word and instructed to tell their father of what they had thought. Binet was able to analyze this experiment into 20 Paris, 1903. 1910] Moore: The Process of Abstraction. 89 the following stages: (1) The hearing of the word. (2) The perception of its sense. (3) An effort to call up an image or determine a thought. (4) The appearance of the image. One of the observations of Armande shows very clearly the distinc- tion between stages two and three. ■ ' As yet there are no images (at the moment of choice) and I know why there are none: When there are many things such — for example, a house, there are many houses — it is necessary to choose. Just then I think about it without representing anything to myself as an image" (p. 75). Sometimes, however, says Binet, the image comes with- out being sought. Binet gives a special chapter to the problem of thought without images. The conclusion at which he arrives is that neither visual imagery nor internal words, either alone or to- gether, account for that complex mental process which we term thought. The grounds for this conclusion are the many in- stances in which his subjects had not and could not find any visual imagery for their thoughts. And again there were times when he thought that he could determine that word-imagery was also entirely lacking. Binet also attempted, and with success, to have his little girls give a rating for the clearness of their images. These ratings ranged from for very weak images up to 20 for images as clear and well defined as actual sensations of sight. In Marguerite there were three well-defined groups: I. A group in which the rating was usually 20, or a little below. This group contained memory images of well known objects or things recently seen. II. A group in which the rating ranged from 10 to 15. This group contained memory images of objects not recently seen. III. A group in which the rating ranged from 3 to 6. This contained memory images of things read or heard about and fictitious images of imagination. With his other subject, Armande, the differences were not so clear. M. Binet gave up hope of finding any regularity. He published, however, the ratings for the three classes. The 90 University of California Publications in Psychology. [ Vo1 - 1 averages, which he did not give, are for Class I, 6.2; Class II, 4.9; Class III, 2.8. Considering that Armande's ratings ranged from 0-12, and Marguerite's from 0-20, one would not expect the classes to be so well defined. The ratings of Armande, however, show the same tendency as those of Marguerite — only not so marked. The small number of cases, however, leaves the result uncertain. In his discussion of the theory of abstract thought and images M. Binet says that his data would support any theory. He makes this claim on the basis of a strange assumption that the discussions between nominalists and realists and conceptualists have always concerned images and not thoughts. He does not exactly state this assumption but it is evident from the text. He then comes to his intentional theory of the image. The image may be used by the person to represent a particular or general signification. It represents whatever the subject intends that it should. He would place, therefore, intentionism as a new theory alongside of realism, nominalism, and conceptualism. "With Binet then there is thought, image, and object. The image is an arbitrary sign to which the subject gives at will a particular or general significance. In our mental life there are those distinct classes of phenomena — thought, image, and interior language. Association alone does not account for the mechanism of thought. It is more complex and supposes con- stantly such operations as choice and direction. The stream of thought is far wider and deeper than that of our imagery. The last sentence of the book is this: "Finally — and this is the main fact, fruitful in consequences for the philosophers — the entire logic of thought escapes our imagery." The next experimental work on abstraction was that of Professor Kiilpe. A report of this was read at the German Psychological Congress, which met at G-iessen in the summer of 1904. This was the beginning of a series of experimental studies by several of Kiilpe 's students in his laboratory at Wiirzburg. The first experiments of Professor Kiilpe were made in the 1910] Moore: The Process of Abstraction. 91 summer of 1900, with Professor Bryan of Indiana. 21 Kiilpe was not satisfied with these and decided to take up the problem again with improved methods. By means of a stereopticon lantern he projected upon a screen in a dark room the figures that were to be observed by the subject. The objects pro- jected were nonsense syllables, four in number, which were grouped at equal distances around a given point of fixation. Each nonsense syllable consisted of a vowel and two consonants. The syllables might be in four different colors — red, green, purple, or black. In the different experiments also the four syllables were grouped so as to form various figures. A group of syllables forming with their different colors some kind of a figure was termed by Kiilpe an object. The subject could be instructed to observe the object from some definite point of view, or he could be left to observe the object without any prescribed task. There were four points of view given to his subjects : 1. The determination of the entire number of letters visible. 2. The determination of the colors and their approximate positions in the field of consciousness. 3. The determination of the figure which the grouping of the syllables formed. 4. The determination of as many letters as possible, with their positions in the field of vision. The number of statements possible to any subject about the individual letters could be classified as follows: (a) The entire number of statements made; (6) the correct statements; (c) the incorrect; (d) the indeterminate statements; and (e) those that could have been made but were not. Each division could then be rated by its proper percentage of the entire number of state- ments. Where task and statement come together (i.e., in the statements about the task) the percentage of correct (&) state- ments is a maximum and that of unmade (e), indeterminate (d), and in general also false (c) statements is a minimum. This 2iO. Kiilpe, "Versuehe iiber Abstraktion. " Bericht uber den I. Kon- gress fur experiment elle Psychologie in Giessen, 1904. Leipzig, 1904. 92 University of California Publications in Psychology. [Vol. 1 proves that "Abstraction in the sense of an accentuation of certain portions of a mental content, i.e., positive abstraction, succeeds best when a preoccupation of consciousness — a predis- position for the partial content, is given or provided for." (op. cit., p. 61.) Negative abstraction — the tendency to neglect or forget all but the one thing abstracted — is the more complete the greater the difficulty of the task. In explaining these results Kiilpe asks what is the reason for the effect of the task? "Were," he writes, "the elements or the colors seen differently under the influence of correspond- ing or heterogeneous tasks, or were they apprehended (aufge- fasst) differently? . . . According to our protocol and the entire conditions of the experiment, to that question one can only answer that the difference lies merely, or at least chiefly, in the apprehension and not in the sensations" (p. 66). The task does not affect sensation but it does affect apperception. If that is so, then there must be a distinction between sensations and our perception of them. "That this distinction must be made in much the same sense in which we distinguish between physical phenomena and our consciousness of them; that, in other words, the old doctrine of an inner sense with the in- volved idea of a distinction between the reality of consciousness and objectivity must now have its opportune renewal in the domain of psychology — this is the principal result that I would draw from my experiments." (p. 67.) Henry J. Watt, a student of Professor Kiilpe, published 22 in 1905 his ingenious attempt to approach the experimental treatment of the supra-sensuous mental processes by a study of reactions of association. As is well known, the reaction-time of association was originally measured by experiments in which the subject was instructed to respond to a given word with the first that occurred to his mind. Watt modified this form of procedure by limiting the freedom of the subject, setting 22 < ' Experimentelle Beitrage zu einer Theorie des Denkens. ' ' Archiv fiir die ges. Psychol., 4, 1905, pp. 289-436. 1910] Moore: The Process of Abstraction. 93 before him a more definite "task." He was to respond, not with any word at all, but with a word that bore a certain kind of relation to the word given as a stimulus. The subject had six of these tasks, constituting six separate sets of experiments. They were : 1. Seek a word under whose meaning the given word is included. 2. Seek one which is included under the meaning of the given word. 3. Seek the corresponding whole. 4. Seek a part. 5. Seek a coordinate idea. 6. Seek another part of the common whole. All the words given were familiar nouns, nearly always consisting of only two syllables, and never evidently compound words. Five hundred such words were found and printed for use in the various "tasks." One of the principal objects of research in this study was the influence of the "task" on the whole course of events in a given experiment. In analyzing the results it appears that there are two general classes into which the experiments may be divided: (1) That in which the association is found by a simple and direct process which suffers no disturbance in its course. Verbal and visual images may be present but they help, or at least do not hinder, the finding of the required association. (2) The second class is that in which the development is com- plex. The subject tries two or more paths before he hits upon the one that gives the desired result. The first class of associations is subdivided according as (a) visual images give rise to the association, or (b) a verbal image or a group of verbal images, or a condition of recollection, etc., or (c) no kind of imagery or media of association can be determined to show how the word spoken was found. In reproduction of complicated development one can point to two subclasses: (a) The subject sought for something else, or some other idea hung in his mind without his being able to 94 University of California Publications in Psychology. ITol. 1 determine just what it was. (b) The subject sought after some more or less definitely determined idea, but could not find it; or he had something in mind, but for one reason or another rejected it. One way in which the influence of the task manifested itself was the mode in which it determined the means of the associ- ation. Task 3 (whole), 4 (part), and 6 (part of the common whole) tend to increase the use of visual images. Task 2 (species) tends to increase the use of verbal images, and Task 5 (coordinate ideas) tends to do away with the use of both verbal and visual imagery. Under the head of visual images Watt brings forward some interesting facts that bear in the main upon two important problems. One is the problem with which Berkeley found so much difficulty: Are all images definite and concrete, or is there any such thing as a general image? The introspection of his subjects seems to point to the existence of what is at least a very indefinite image. We quote some examples that he has given as typical: "Hide: Image of an animal torso thickly covered with hair (very unclear). To what animal it belonged I do not know. Grain : Fleeting image of a rye or wheat field — the species was not clear. Mouth (Maul) : Beast. Dark image of an utterly undefinable animal. It could have been an ox, or a horse, or a dog with stronger definition of the head and mouth region." 23 Watt calls attention to the fact that in this last case the image did function as if it were universal. One can, he says, maintain that it was in reality concrete and definite, but he can not prove his contention. Still, scarcely any one would wish to make such a contention. Vague, indis- tinct images are often like a child's drawing — they need inter- pretation. When we label them we know what they are, but to the uninstructed observer they may stand for a number of things. After calling attention to the existence of such "gen- eral" images, Watt then points out how illogical it would be 23 Op. cit., p. 364. 1910] Moore: The Process of Abstraction. 95 to infer from the existence of the general image the non-exist- ence of the universal idea. The second problem on which he touches under this heading- is the position of the mental image in our mental mechanism. The mere mention of the theory of types suffices to remind us that some authors write as if certain people made use of visual images in their mental operations to the exclusion of all others because they belong to what is termed the visual type. Watt points out that the kind of image used depends upon the "task" which the subject performs. 24 By changing the task the subject passes from the visual to the verbal type. Another point that he makes is this : The mental image is not always a merely secondary phenomenon like the illustration in a novel. It may seem at times merely to accompany the word used as a stimulus. On other occasions it is clearly the starting point for the solution of the task. In all probability the mental image never comes into the field of consciousness without exerting some influence on the development of associ- ations. Whether by inhibition or furtherance or direct sug- gestion of new ideas, it has its influence on the way in which the task is performed. In conclusion Watt sketches the outline of his theory of thought. It is an attempt to account for the flow of conscious- ness. He first calls attention to the fact that consciousness is not discrete but continuous. He then asks what determines the entrance of an idea into consciousness? The chief factor is the "task" that the mind is attempting to accomplish. The ten- dency of one idea to reproduce another is determined in 9 merely mechanical way by the number of times that the two ideas were perceived together in the past. But the many possi- bilities, the many tendencies to reproduction, are limited by the "task." In the much discussed problem of the relation between image, word, and concept, Watt admits the existence of all three and 24 Op. tit., p. 367. 96 University of California Publications in Psychology. ITol. 1 does not attempt to explain away the concept in terms of im- agery or words. From the statements of his subjects it was clear 4hat there was a distinction between the word and the understanding of the word. One could exist without the other, therefore they must be distinct. But is the understanding of the word the crowding into consciousness of a number of dark associations? One hears nothing of such associations in the understanding of the word used as a stimulus ; though in seeking for the word of response such associations do occur. The burden of evidence in his experiments rather favors the view that the understanding of a word is something other than crowding in of obscure associations. But for the final determination of this point he deems that further experiment is necessary. The following year August Messer 25 published the next study of the Wiirzburg School. Dr. Watt was one of his subjects. The general purpose of the study was expressed by the author as an attempt to investigate the conscious processes that are found in simple acts of thought. The method of the experi- ment was based on that of Watt's work, which has just been mentioned. There were fourteen series of experiments, some of which were taken from the "tasks" invented by Watt. 1. In the first series the subject was shown a word, and his task was to speak out as quickly as possible the first word that came to his mind. 2. In the second series the task was more restricted; the word of response had to be a word representing a coordinate object — that is, one that belonged to one whole along with the object represented by the given word. 3. In the third, the subject was to mention a coordinate concept — that is, one belonging to the same genus as the given word. 4. The response was to be any adjective. 5. A characteristic of the idea designated by the given word — but not its genus. 25 ' ' Experimentell-psyehologiscke Untersuchimgen iiber das Denken. ' ' Archiv fior die ges. Psychol., 8, 1906, pp. 1-224. 1910] Moore: The Process of Abstraction. 97 6. Remember an object belonging under the concept of the given word and make a statement concerning it. From the seventh to the eleventh experiment two words were shown one above the other. The upper was to be read first. The subject's tasks were: 7. Express the relation between the ideas designated by the given words. 8. Express the relation between the objects designated by the given words. 9. In the ninth series the two words were the names of celebrated men and the subject was to pass on their relative value, expressing a judgment which had real claim to objective validity. 10. In the tenth series the persons, things, or conditions represented by the given words were to be compared and a judgment expressed; but the judgment was to be one of merely subjective value and express what would be the subject's preference. 11. In the eleventh, a noun and an adjective were shown to the subjects. He was instructed to regard the two words either as a question or an assertion, and where possible to pass a judgment about them. 12. In this series the subject was shown sentences or groups of sentences and his task was to understand them and take up a position in regard to them. The groups of sentences repre- sented logical premises and conclusions formally correct. In the last two series of experiments the subject was shown real objects or pictures. 13. He was to speak the first word that came into his mind. 14. He was to make a statement about the object or picture. In the first series, though no special task was given, the subject made one for himself. He involuntarily sought a word that bore some relation to the given word. In other series also the tendency was noticed to specialize still further the task assigned. In the visual imagery of the subjects there is again found 98 University of California Publications in Psychology. ITol. 1 the "general image" mentioned by Watt. This proves to be an image so imperfect that the subject can designate it only by some such word as an animal, a bird, etc. Such an image may be spoken of as general because it can stand in consciousness for an entire class. The author also gives some account of the motor imagery that his subjects experienced during the experiments. Then after a discussion of the process of association he passes on to a problem more closely allied to our own, the understanding of the word — the concept as distinct from word and imagery. Generally the meaning seems to come with reading the word. But even in such cases the meaning is not a constant factor. It may exist in all degrees of perfection. The word may be scarcely understood at all. It may be perceived, but merely as a sound without meaning. Or the understanding may come partially with reading and take some time to grow. This latter form leads up to the case in which there is an actual separation between the perception of the word and the apprehension of its meaning. The conditions for the separation of the word from the apprehension of its meaning are as follows : 1. The strangeness of the word. 2. Incorrect reading of the word. 3. Equivocal character of the word. 4. Imperfect knowledge of the language. 5. Number and length of the words. 6. The occurrence of a purely automatic reaction on the basis of verbal association, e.g., Laut-Schall, Haustier-Maus. 7. Fatigue. 8. Excitement. The "meaning" of the word was often something that the subjects found it difficult to explain. It was frequently ex- pressed by such an expression as "I knew what was meant." The subjects were sometimes enabled to analyze this abstract "meaning" a little further. "The understanding of the word existed in the consciousness of that general sphere to which the word belonged" (p. 77). One of the subjects expressed it 1910] Moore : The Process of Abstraction. 99 as "the consciousness that something appropriate could be asso- ciated." Sometimes the " sphere "-consciousness is identified with the generic idea to which the object belongs; again, with the entire domain in which an object belongs. For example, Subject 2 with the word of stimulus, "Hegel," said: "It seemed to me at first as if the word were 'Hagel. ' As soon as the auditory image of ' e' sprang into consciousness, there came a direction toward the History of Philosophy." 26 At times the "sphere" of consciousness was an emotional element or word, or something similar. Again in the process of understanding, there was a consciousness of synonymous words or related objects, or some prominent characteristic of the thing represented by the word of stimulus. Sometimes the word instead of being understood in a general sense was taken in a special one, as where the word "garden" aroused the idea of a garden around a former home of the subject's family (p. 82). From all this it seems to the author extremely prob- able that in the process of understanding a word we have to do with phenomena of association and reproduction. What part, if any, has the subject's imagery in his under- standing of a word? The more perfect the imagery the less does it seem to cover what is meant by the general significance of the word. But the more schematic and faded the imagery, the less does it differ from the "meaning." 27 More important than the relation between the clarity of the image and the meaning, is the question: To what extent is imagery necessary to the signification? And here he says there is not one single example from which it is clearly evident that the understanding of the word was dependent on the awakening of a visual image. The most that can be said is that in a few solitary instances it was recorded that with the help of a visual image the meaning became clearer or more precise. But in the further progress 26 Page 79. 27 From what follows it is evident that the author does not mean to suggest that the meaning is nothing but faded imagery. The imagery fades into nothing, long before it gets anywhere near the ' ' meaning, ' ' which may at times be clear without imagery. 100 University of California Publications in Psychology. IT o1 - * of the experiment the subject's imagery plays an important part in the solution of the task. As to the understanding of the word of response, it often takes place before the subject can express his meaning, and when the word is found it does not always express fully the subject's mind. Sometimes too, the word of response is uttered before its meaning is understood. The further sections of this work on the psychology of judgment, etc., are more remotely connected with our problem. The more kindred section on "Begrifflichen und gegenstand- lichen Denken" confirms still further the distinctions between word, image and concept. 28 In immediate connection with the work of Watt and Messer is that of Dr. Schultze. 29 The foundation for his analysis is daily observation confirmed by his own experiments and those of others. His own experiments at the time of this article were to appear shortly under the title of "Beitrag zur Psychologie des Zeitbewusstseins. " 30 He was subject in Messer 's experi- ments and among his own subjects he numbered Kiilpe, Watt, and Messer. His own work claims to be in the domain of descriptive psychology. His first problem, and the one with which we are concerned, is this : In the classification of mental processes is it justifiable to make a distinction between the sensible appear- ances of things and thoughts (Erscheinungen und Gedanken) ? Originally he answered this question in the negative, but he was forced to give up this position on approaching the problem from the experimental point of view. The relinquishment of the old position seems to have required some effort, for he 28 For a criticism of the technique in Messer 's experiments, cf. E. Meu- mann, "Ueber Associationsexperimente mit Beeinflussung der Keproduk- tionszeit." Archiv fiir die ges. Psychol., 9, 1907, pp. 117-150. Messer re- plied in his article, "Bemerkungen zu meinen Experimentell-psychologischen Untersuchungen iiber das Denken. ' ' Archiv fiir die ges. Psychol., 10, 1907, pp. 409-428. 29 E. E. O. Schultze, ' ' Einige Hauptgesichtspunkte der Beschreibung in der Elementar-psychologie. I. Erscheinungen und Gedanken. ' ' Archiv fiir die ges. Psychol., 8, 1906, pp. 241-338. so Cf. Archiv fiir die ges. Psychol., 13, 1908, pp. 275-351. See especially Sec. 11, pp. 329-333. 1910] Moore: The Process of Abstraction. 101 writes : "It cost me a great resolution to say, that on the basis of immediate experiment, appearances and sensible apprehen- sions (Erscheinungen und Anschaulichkeiten) are not the only things that can be experienced. But finally I had to resign myself to my fate" (p. 277). His reason for doing so was that the data of appearances did not exhaust the content of experience. There are marked differences between appearances and thoughts. Appearances are apprehensible by the senses (anschaulich) but not so thoughts. Appearances are more or less localized. When there comes a pause in any series of appearances, during that pause we are conscious indeed of various sensations from the organs of the body — but is the consciousness of the pause the perception of such sensations? When there comes a blank over the mind, what is it that is lacking — sensation or thought? Thought. Thoughts are as much a matter of immediate experience as our sensations. Thoughts are not to be explained in terms of imagery. Thought can be perfectly clear and certain but the accompanying imagery is of various degrees of clarity or is altogether lacking. Thoughts are not feelings. (1) Because we can pass judgment upon matters of feeling without actually experiencing the slightest tremor of an emotional state. (2) We can experience feelings without any intellectual state con- nected with them, as for example, in certain unwarranted and inexplicable emotional states. (3) There is the same independ- ence between the clearness and importance of thought and feeling in our mental states as there is between thoughts and images. What then is our act of thought? Not the sensations that were active in the process of its acquisition. For we make frequent use of abstract concepts but seldom in connection with these concepts do we use the definitions and sensations necessary to their original formation. No sensation can con- ceivably exhaust all the characteristics of the concept. Con- cepts then are not sensations, not mental images, not feelings. They stand apart by themselves as special factors of our mental life. 102 University of California Publications in Psychology. [Vol. 1 The work of Watt, Messer, and Schultze was continued by Karl Biihler. 31 He thought it advisable to study the process of thought with materials which offered far more difficulty than the comparatively simple tasks of Watt and Messer. Accord- ingly such questions as the following were proposed to his subjects : "When Eucken speaks of a world-historical apperception do you know what he meant thereby?" The subject had to answer with a simple yes or no, and then give an account of all the mental processes he experienced in arriving at his answer. In a section on the Elements of our Mental Life of Thought he propounded the question — what are these elements, and which among them is the real bearer of the process of thinking ? From the protocol of his subjects there is one group of mental pro- cesses that may be easily characterized — the sense imagery, whether visual, or auditory, or sensomotor. To this may be added the consciousness of space. There are also feelings and such states as doubt, astonishment, etc. But this is not all. The most important phenomena do not fall in any of the above categories. There is something else that possesses neither the qualitative nor quantitative characteristics of the senses. These elements of our mental life are what the subject characterized as "the consciousness that," etc., or more properly and fre- quently as his 'concepts' (Gedanken) . Do we think by means of imagery or by concepts? The answer, based upon the subjective analyses given by his subjects, is that "what enters into consciousness so fragmen- tary, so sporadically, so very accidentally as our mental im- ages can not be looked upon as the well-knitted, continuous content of our thinking" (p. 317). Concepts then, not images, are the essential elements of our thinking. What then is the concept? Not an image nor a series of images, nor the relation to a series of images. The concept is a unit, a mental element, the ultimate result of the analysis si ' ' Tatsaehen und Probleme zu einer Psyckologie der Denkvorgange. ' ' Archiv fur die ges. Psychol, 9, 1907, pp. 297-365; 12, 1908, pp. 1-92. 1910] Moore: The Process of Abstraction. 103 of thought. As seeing is related to a sensation of sight as "sensing" to our sensations, so is knowing related to our thoughts. "Knowing" is distinct from "sensing." It may be accompanied by sensations but cannot be supplanted by them. Word imagery does not give us the signification of words. "A meaning can never be imaged but only known (Eine Bedeutung kann man uberhaupt nicht vorstellen, sondern nur wissen)." The solution of the task is not accomplished by a single series of concepts. Between concepts there goes on a great deal of thinking — the consciousness of the task to be performed — the relation of the given concepts to others and to the task. The general consciousness of the task and the consciousness of mani- fold relationships constitute a kind of setting or background in which special concepts appear. The understanding of words and sentences "is nothing less than a conscious logical relation, which brings into consciousness the connection between the thought to be understood and one already known" (12, p. 13). In many cases understanding took place by the entrance into consciousness of a more general concept, and thereupon the subject knew how and why the idea before him belonged under that concept. The mere entrance into consciousness of the more general concept does not seem to suffice, but it must be perceived as bearing a relationship to the problem before the subject. Sometimes the thought that the given idea suggests is not a more general one, but one which the subject perceives to be identical with the given thought. Sometimes the given sentence is understood by its suggesting a thought that would prove it. The analogy, between the process of understanding a sentence and the process of perceiving a geometrical figure, will be seen at once by comparing the above analysis of Biihler's work with our own section on the process of perception. 32 The division of Biihler's work entitled "Ueber Gedankener- rinerungen" is of great interest and value in the study of memory, but bears less directly on the general problem before 32 Below, pp. 127-139. 104 University of California Publications in Psychology. \yOfiflA<8> is OAZ*?*? v^^c Q InltdQW© H 12 ©05G W <^zs<^V(? s "H , 'B , T©¥QSQ io ff>*. § £ ®\*.P &&.*& Q o © 6 Wv P ■ 9?W ^ i" & * & BfltlPSRft® «?a ^s 1 %* % 8^© S3® HdP^&O'&t? & 12 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 Fig. 1. — The originals used were each about 1% times as large as in this reproduction (see Fig. 2, p. 122 for actual size). The numerals at the side and bottom are merely for convenience of reference in the text. In these references, the first numeral (in lighter face) indicates position along the axis of abscissas; the second numeral (in heavier face) indicates position along the axis of ordinates. Not all of the figures were actually used. On account of their very evident associations the following were ex- cluded: (1, 2)-(2, 2)-(5, 2)-(6, 2) -(11, 2)-(14, 5). The figures possess one great advantage. On account of their strangeness, the process of perceiving them goes through a longer course of development and thereby one is enabled to detect points which it would otherwise be impossible to notice, or at least could be obtained with difficulty and uncertainty. 1910] Moore: The Process of Abstraction. 119 The groups of figures in the actual experiments (see cuts on pages 122 and 123) contained five figures instead of three. This drew out to some length the process of isolating and perceiving the common elements, thereby allowing a better opportunity to observe the development of the mental processes involved. 2. The Apparatus. For the experiments in Leipzig, Wirth's memory apparatus with rotating disk was used. At the University of California I used Ranschburg's memory apparatus. Each performs the same function and the same disks may be used in either ap- paratus. Each rotates a disk and exposes suddenly a small surface and as suddenly removes it from view. In this ex- periment a group of five figures was exposed for a quarter of a second and then a blank space for a quarter of a second and so on till the series of twenty-five exposures came to an end or as much of the series was used as necessary for the experi- ment. It was at Professor Wundt's suggestion that I used this short time of exposure and interval between exposures. It tends to reduce the experiment to simpler and therefore more constant conditions by cutting out to a large extent such variable factors as reflection on what was seen, comparison, and voluntary association. To beat time I have used both the metronome and the time sense apparatus, but generally the former, which is sufficiently accurate for the purpose. Care was taken to keep both these pieces of apparatus out of the room in which the observer was seated. 3. Instructions to the Subject. The subject was instructed to look for the repetition of some figure and to turn a switch, which stopped the rotation of the disk, as soon as he was certain that he had seen some figure repeated. It was not required of him to see this figure in each group as it passed by, but merely to be sure that he had seen some figure twice. He was told not to wait until he knew all 120 University of California Publications in Psychology. t Vo1 - 1 about the figure but only to make sure that one and the same figure had occurred more than once. He was required at the end of the experiment to describe his state of mind during the experiment, and especially to tell what it was that he first noticed. 4. Classification of the Experiments. It soon became apparent that the method offered exceptional advantages for a genetic study of the process of abstraction. In handling the results and attempting to reduce them to some kind of order, the complex nature of abstraction became evident. And at the same time its analysis was greatly facilitated. Our five figures were found to constitute something of a unit which underwent a real process of breaking up. This was evidenced by the fact that the elements of a group have a different mental value after the perception of a common element than before. 2 Before the common element is noticed, the figures of any group have a tendency to persevere in memory, which varies with the foeality of their perception 3 and with their own inherent at- tractiveness. 4 After the common element has been perceived the tendency of the other figures to persevere in memory is greatly reduced. The group is no longer what it was before it was broken up. This breaking up of the group is one of the several processes which form the mental complex that we call abstraction. The breaking up of the group is intimately bound up with the perception of the common element. Perception then is another factor in abstraction. The figure perceived is remembered and recognized again upon its recurrence. We have then four points in our preliminary analysis of abstraction : (1) The breaking up of the group; (2) The process of per- ception; (3) The process of memory; (4) The process of recog- nition. Each one of these has been made the object of experi- ment and form the four main headings in our experimental data. 2 See below, pp. 124-127. s See below, pp. 158-159. * See below, pp. 122-124. 1910] Moore: The Process of Abstraction. 121 These experiments were commenced in Wundt's laboratory at the University of Leipzig. They were afterwards continued at the University of California. My thanks are due Professor Wundt for his kindly and valuable suggestions as to the method of experiment and also to Dr. Felix Krueger for his constant interest and assistance while I was working at Leipzig. I wish also to express my indebtedness to Professor Stratton and Dr. "Wrinch, with whose valuable cooperation the experiments were conducted at the University of California. The subjects who took part in the experiments were Miss Ball (B), Dr. Bessmer (Be), Dr. Brown (Br), Herr Blosfeldt (Bl), Miss Deamer (D), Herr Griinbaum (G), Dr. Krueger (K), Dr. Moore (Mo), Miss Mower (Mw), Miss Ross (R), Professor Stratton (S), Professor Eustachius von Ugarte (U), Mr. Wa- beke (W), Dr. Wrinch (Wr), and Herr Ziembinski (Z). 122 University of California Publications in Psychology. t Vo1 - 1 III. EXPERIMENTAL RESULTS. 1. The Analysis of the Groups. (a) Isolation of the Common Element. In abstraction some element or characteristic is always picked out from a group and is recognized as identical with that which was found in another group. In our experiments this element was the repeated figure. We may ask what is it in any element that accelerates the process of its isolation and perception? The answer as one might expect is — whatever attracts attention to the element. This may be the pure accident of its focal £ £ *p v ■ b o® R S> O V®^ A ®^-*D Fig. 2. — Showing grouping of 'elements' for actual display upon the disk, the common element following in the order 1, 3, 5, 3, 1, 3, 5. 1910] Moore : The Process of Abstraction. 123 perception. It may be the fact that it is rather larger than the other figures or blacker or more open. Small but symmetrical figures, e.g., 4, 12, and 5, 12, in Fig. 1, seem to pass by easily without being noticed. Another drawback is apparently the complication of the figure. This, however, is probably only apparent. The subject involuntarily waits to be informed about the complicated figures. Compli- cation is, in itself, an advantage because it attracts attention. But the subject waits to know just what is repeated. In spite of instructions, he cannot stop the apparatus as soon as he is sure of the bare fact that a figure of some kind has been repeated. The attempt was made to find out whether the sequence of position had any influence in the perception of the common element. If we number each of the five positions in a group 00 41 had a "feeling that it was there before he saw it." Reaction of the subject to disks with no common element: (1) Sees nothing. (2) Thinks none absolutely alike. (3) Negative doubt. Subject Z. I. An intimation of a common element, without any knowledge of its form. (8, 12)*-(13, ll)-(8, 13)-(4, 14)f-(2, 6)}-(16, 6)|-(3, 6)-(2, 13). || Confirmed by (15, 14) in two different experiments. * At first it seemed to the subject as if a common element was present. Then he looked here and there to find it. •f At first there was an abstract feeling of something common. % First noted something common and new. || At first there was an indefinite consciousness of something repeating itself. 170 University of California Publications in Psychology. [Vol. 1 II. Certainty that a common element is present, without any knowledge of its form. (3, 11). III. Probability that a common element is present, but an imperfect idea of its form. This stage is not to be found in the records of this subject. IY. Probability that a common element is present, and a true idea of its form. (2, 13)-(5, 12)-(10, 10). Confirmed by (16, 13)-(10, 12). V. Certainty that a figure is being repeated, but an imperfect idea of its form. (7, 12)-(13, 11)*-(17, 14). * Subject knew that the common element had something round in the middle. The following observations of this subject are interesting: (a) "There is no time to compare one figure with another, or one impression with a previous impression." (&) When a figure was used as a common element which the subject had not seen before, it generally happened that at first he noticed something new and then a special figure. (c) The perception of the common element has a tendency to obliterate the images of the other figures. Before perceiving the common element as common, the images of several figures that have just passed by float about in the mind. When the common element is perceived as such, they vanish at once. Reaction of the subject to disks with no common element : (1) Stops apparatus after nine exposures, and says he is perfectly certain that no figure is repeated in each group. (2) After nine exposures the subject was certain that no common element was present. (3) Subject thinks that a figure (4, 14) might possibly have been repeated. He drew it correctly along with another figure; which two figures were drawn when he was requested to reproduce everything he could remember as having been seen. This also happened with (17, 15). (4) No intimation of any figure having been repeated. (5) No intimation of any figure having been repeated. (6) Thinks that there was no common element. 1910] Moore: The Process of Abstraction. 171 (7) State of negative doubt. (8) Almost certain that no common element was present. (9) Negative doubt. (10) Thinks that no common element was present. (11) Subject thought several times that a common element was present. Then there came an ever increasing certainty that none was present, and at the end of the series he was certain that there was no common element. Another interesting stage with this subject is that in which the figure on being first noticed is recognized as familiar. By the word familiar, it is not meant that he had seen it before on other disks but that it comes into focal consciousness with a peculiar nuance which tells the subject that this is the common element. It seems that this tone of familiarity (Belcanntheits- qualit'dt) arises from the figure's being seen before but not analyzed out from the other figures. The subject whose results are about to be recounted could give by introspection at the end of the experiment no infor- mation at all about the development of the mental process he had just experienced. When later on in the semester I commenced to confirm the results of self -observation, I tried the same method with this subject. I cut the experiment short after he had seen fifteen groups of figures and then asked him simply: ''What do you think? Is there a common element present or not? Are you certain or merely inclined more or less to think that you see a common element? Draw what you remember!" In this way was obtained what the subject's introspective memory failed to reveal. Cross-sections were obtained in the course of development and fixed before they could fade from memory. Subject U. I. An intimation of a common element, without any knowledge of its form. (5, 12)-(7, 13)-(10, 9)-(10, 12). II. Probability that a common element is present, but an imperfect idea of its form. (2, 6)-(17, 15)-(16, 13)-(4, 9)-(2, 12). 172 University of California Publications in Psychology. IT o1 - 1 III. Probability that a common element is present, and a true idea of its form. This stage was not found with this subject. IV. Certainty that a common element is present, but an imperfect idea of its form. (10, 7)-(4, 14)-(8, 10)-(10, 6)*-(5, 9). * Certain only that lie had seen the "two eyes" recur. Reaction of the subject to disks with no common element : (1) No idea of any common element at the end of the experiment. (2) No idea of any common element at the end of the experiment. (3) Complete uncertainty at end of experiment. (4) No idea of any common element at the end of the experiment. (5) No common element noted. ( b) Interpretation of the Results. (i) The Immediate Experimental Conclusions. When we look at these results it becomes at once apparent that an element of certainty and uncertainty is involved in the process of recognition. If we ask ourselves what this means we must say that whenever the mind is certain of anything, it assents; and whenever we have an assent we have an act of judgment. One of the immediate empirical conclusions of our results may be stated thus : The process of recognition involves an element of certainty or uncertainty. From this we may conclude : That the process of recognition involves a judgment or a suspended judgment. For whenever I am certain I assent; and whenever my mind is in a state of uncertainty, assent is suspended. In the one case there is a judgment; in the other, judgment is suspended. It is not necessary that this judgment should be formulated in so many words. In fact, one may venture to say that in most cases of perfect recognition there is no verbal formulation of the judg- ment at all; but the psychological act of judging is nevertheless really and truly present. The presence of a judgment in the act of recognition proves that the act of perception which does not involve a judgment is 1910] Moore : The Process of Abstraction. 173 an essentially different and less highly developed mental state. Recognition is indeed a perception, and over and above this a judgment is passed upon the perception. This judgment in- volves the statement that what is now perceived has been per- ceived before. If recognition is incomplete the judgment hangs in suspense and cannot be definitely passed. In the further study of recognition we have only to ask our- selves, what is the basis of this judgment? Do the experiments help us out? If we run through the results we will find that any degree of certainty may be accompanied by any degree of the perfection of perception. A person can be certain that a figure was re- peated and have a perfect image of the figure, or an imperfect image, or no image at all. A second empirical conclusion may be stated thus: Assured recognition is not dependent upon per- fect perception. And why this statement? Simply because it is an empirical fact that assured recognition can exist with a very imperfect perception, — a perception that is so imperfect that it involves no mental image whatsoever. 17 "While indeed we have not found out, as yet, on what the judgment of recognition depends, we have at least discovered something on which it does not depend. And that is the mental image. This suffices finally to dispose of one theory of recog- nition, now generally rejected by psychologists — the theory, namely, that recognition is brought about by the comparison of the present sensation with a revived mental image. Identity being perceived, the object seen is then recognized. That such a comparison of images is unnecessary appears from the ex- periments. Why? Because recognition takes place not only when there is no revived mental image of the past perception, but when the present perception itself is too imperfect to leave any trace of mental imagery in the mind. Recognition, how- ever, may take place by a comparison of mental images. In general the rapidity of succeeding impressions made this an impossible, or at least a very awkward, process. It once hap- " Cf. also above, pp. 134-136. 174 University of California Publications in Psychology. [Vol. 1 pened, however, that a subject reported that she had used just this metho'd in arriving at certainty of recognition. On thinking that she had seen a certain figure twice, she tried to call up the previous image that she had in mind as identical with the figure just seen, and institute a comparison between the two images. However we must note that recognition was already in the probable stage when this was done. And the comparison that was attempted was after all only an auxiliary method. The comparison of images, therefore, may come in as an aid, but it is not necessary to recognition nor is it the normal method. One might object to the use of the word 'normal' here as carrying us beyond the limits of legitimate deduction. Was not the rapidity with which the exposures succeeded one another expressly chosen to exclude the possibility of comparing mental images? That is true, and our experiments prove only that the comparison of mental images is not necessary in the process of recognition. As to its being the normal method, we can from our own experiments only conjecture. But there are other experiments along this line. I refer to those on the recognition of the identity of time intervals, tones, etc. When a subject listens to two raps separated by a short interval, and then, after a period of waiting, hears two more raps, how is it that he recognizes that the second two raps mark off an interval of time equal to that of the first? Does he really compare some kind of mental images of the two time intervals? It would seem from the experimental research on this point, that he does not. 18 Professor Frank Angell has made it abundantly clear that the recognition of tones does not depend on a comparison of mental images. In his study of the "Discrimination of Clangs for Different Intervals of Time," he arrived at the following results : "The main conclusion to be drawn from the distraction experiments is that judgments of tone discriminations can take place, and in the majority of our experiments did take place, is Cf. Wundt, Physiologische Psychologic, III, 5, 476-517. 1910 1 Moore: The Process of Abstraction. 175 without conscious comparison between the present sensation and a memory-image of a past sensation. When, for example, a reagent, after a long time-interval filled with interesting reading, from which he had to be practically aroused by a sharp signal in order to prepare himself for the apprehension of the second tone, nevertheless delivered a judgment with a feeling of con- siderable security, it is idle to speak of "memory-images" or indeed of comparison in the ordinary meaning of the word. Or when a reagent, after having accurately discriminated six pairs of tones, decided with ease that a tone just given is like or unlike a tone 4 vibrations higher or lower sounded 60 seconds before, and is correct in these decisions 63 times in 100, it is evident that the ordinary theories of tone-comparisons need readjustment. "No more is it explicable on the theory of memory compar- ison that there should not have been a great increase in doubtful judgments in passing from undistracted to distracted discrimi- nation, or indeed in failures to judge at all, or that the several forms of distraction should not have shown a far greater dif- ference in effect than was actually the case. ' ' 19 In the light, then, of our experiments, and also those on the recognition of various sensory stimuli, it is not too much to say that the comparison of mental images is not the normal method of recognition. Summing up, then, the conclusions that we may regard as established by the experiments of this section we may state : A. The process of recognition involves an element of cer- tainty or uncertainty. From this follows: The process of recognition involves a judgment or a suspended judgment. B. Certain recognition is not dependent on perfect percep- tion. From this it was seen to follow that: A comparison of mental images is not necessary to the process of recognition. is Amer. Journ. Psychol, 12, 1900-1901, p. 69. 176 University of California Publications in Psychology. [yo\. ! An empirical fact rather than a conclusion from these exper- iments is stated in the following proposition : Certain recognition can take place without the formation of any mental image of the thing that is recognized. 20 (ii) The Basis of Judgment in Eecognition. When we are asked to give an account of the real basis of judgment in recognition we naturally ask, how is the object that is recognized remembered? The factors of memory, one might suppose, are active to a large extent in the process of recognition. We are naturally concerned with the factors which enabled the subjects to memorize the figures used in our experiments which represented rather complex conditions. It cannot be taken for granted that the basis of recognition is the same for simple sensations and complex perceptions. In fact it is rather likely that what serves as our cue in one case does not meet the de- mands of another, that what is the chief basis of recognition of a simple tone may become a very minor factor in the recog- nition of a time-interval. And what is prominent in recog- nizing a time-interval may become subordinate in the recognition of a street or a house as places where one has been before. On this account it is desirable to take for experiment such complex material as our figures, in order to see if any factors enter into the process of recognition that have not been noticed in the usual experiments on time-intervals, colors, tones, etc. The process by which the figures are remembered should give us some clue to the method used in their recognition. In the section entitled, "The Factor of Memory in the Process of Abstraction" (p. 139), we compared memory by visualization and by motor imagery, with memory by association and analysis. A marked advantage was found in favor of the latter. Memory by association consisted in relating the figure to known objects, or analyzing it and thus relating it to certain mental categories. These mental categories are the 20 This conclusion is based in great part upon the experiments given on pp. 134 ff. 1910] Moore: The Process of Abstraction. 177 bonds which hold the figure in place and make possible its recall when it has left the field of consciousness. In fact, it seems that if all the conceptual ties could be cut, or be lacking from the beginning, the figure would fade away completely and recall would be impossible. We find also that, in the process of per- ception, the essential element is not the formation of a visual image but the relating of the object perceived to one or more mental categories. Nor must we regard this relating of the impression of an object to its categories as a manipulation of separate and distinct psychical entities. It is rather what Wundt would call an assimilation. The sensation and the gen- eral concepts form a psychical compound which differs from its elements and is a new mental product. What are the elements of this compound? Wundt speaks of the feelings involved, especially that of familiarity, the sensation and the images to which it is assimilated. But we may question the completeness of this analysis. There seems to be something that is not included therein and that something lies in the mental categories that couple the perception of the object to the train of memory. These, the essential elements of assimilation in perception, are also the elements par excellence of recognition. An assimilation does take place, and on the basis of our experiments on memory and the analysis of perception we may venture to say that the chief elements of assimilation are the concepts to which the sensations are assimilated in the process of perception. When the figure is seen it is at once assimilated to certain mental categories ; it is regarded as made of straight or curved lines ; it has elements that curve ; it is an open or a dark figure ; it is symmetrical and regular, or just the reverse. These phrases do not stand for images that are present; this the cases of recognition without imagery prove. But suppose they do so stand ; suppose we have in recognition an assimilation of a present sensation to a number of revived images — of lines, curves, points, etc. Certainly the new psychical product should be an image, a product of the sensation and the imagery of past experience. But there were 178 University of California Publications in Psychology. ITol. 1 eases in which recognition took place without the trace of an image. Consequently the assimilation would appear to be of elements that are not images. These elements we may speak of as mental categories or concepts. The sensation of a figure never stands alone. Perhaps no sensation ever does. It is related to an appropriate series of concepts. These are not all in focal consciousness. Perhaps all remain unanalyzed in the background of consciousness until by reflection we consider what kind of a figure we have seen. But the sensation plus the con- cepts with which it is associated, — these are assimilated and constitute a new psychical product. This psychical product is what is known as our 'idea' of the figure. My subjects have sometimes said: "I have an idea of what the figure is, but I cannot draw it." And then after some thought they would give a very inadequate description which would relate the figure to some concept. On being allowed to look for the figure they would find it among the entire lot of figures that made up our material. Our 'idea' of the figure is whatever image may be present plus the concepts to which it is assimilated. That which is the chief factor in perception, that by which we recall figures, is also that by which we recognise them. And this is the figure's series of associated concepts. When a figure is seen once, some kind of an 'idea' of the figure is formed — it is fitted in to one or more mental categories. When it is seen again the new percept is assimilated to the old. The old series of associated concepts falls in with the new. And in this way, perhaps, is produced the tone of familiarity. In the process of assimilation there is nothing that jars; on the contrary there is a reinforce- ment at least of some members of the associated train of con- cepts. New concepts may be brought out, but they fit in with the old. Merely similar figures, however, might on a later per- ception bring out new concepts which would contradict the old and thereby destroy the feeling of familiarity and give rise to doubt as to the identity of the figures. One who is not disposed to give such individuality to the concept as distinct from sensation and mental imagery might 1910] Moore: The Process of Abstraction. 179 have recourse, as Wundt does, to the feelings. When we find no image in the process of recognition we must not jump to a conclusion that a concept distinct from our mental imagery is present. There are the feelings to be taken into consideration. Perhaps these mental categories are groups of feelings and not a class of mental states by themselves. The examination of this point leads us to our next chapter, in which we analyze the product of abstraction. 180 University of California Publications in Psychology. t Vo1 - 1 IV. THE PRODUCT OF THE PROCESS OF ABSTRACTION. In our analysis of the process by which an abstraction is formed, we have necessarily learned something about the final product. "We have watched the growth of a complex mental state and must necessarily know something about that mental state in its final stage. Are there any evident elements in the final product of abstraction that we may regard as facts of experience? Yes. Our experiments have revealed some to which we called attention in our section on the process of per- ception. From the results of that section, confirmed as they are by the succeeding chapters, there are two important facts that were abundantly evident. (a) There exist imageless mental contents representative of a visible object. Our own experiments are not the only evidence on this point. A reference to the history of the problem 1 will show that a number of psychologists have determined the ex- istence of various kinds of imageless mental contents. The consensus of evidence is such that 'thoughts' without imagery must be looked upon as established mental facts. And when we take perception to mean the result of the process of per- ception, our experiments show conclusively that we can have a perception of a visible object in which there is no visual imagery. Our idea of that visual object is therefore not a mental picture, although under such conditions as obtained in our experiments we should expect, if at all, to find visual im- agery constantly developed. Without, however, making any assumption as to the nature of these imageless mental contents we may regard their existence as an established fact. They are the essential elements in the product of perception and abstraction. The existence of any i Cf . pp. 76 ff. 1910] Moore: The Process of Abstraction. 181 kind of mental imagery in the complex product is not essential. Imageless mental content and not imagery is therefore the true product of abstraction. The second fact of experiment is this: (&) Perception is a process of assimilating the data of sense experience to their appropriate mental categories. By this assimilation the object is perceived. The word category is not here taken in any pre- conceived sense. It is a fact that in perceiving a figure the earlier stages were designated as a knowledge that the figure was "pointed" or "open" or "round" or "had the top lines crossed," etc. These expressions are examples of what I mean here by categories. It is a fact, too, that these expressions were not descriptions of mental images. The figures, however, had been seen with the eyes, and in perceiving them they were inter- preted in terms of the previous knowledge of the subject. This I have expressed by saying that the figures were assimilated to appropriate mental categories. So far this is all that I mean by the word category. Let us now ask, what are these mental categories in terms of our modern psychological terminology? A current psycho- logical division of our mental states leaves room for nothing but (a) sensations and their images, (5) feelings, and (c) will, which by some psychologists is explained in terms of feeling. To these states and combinations of them many psychologists have attempted to reduce our mental processes. We may now ask ourselves to which of these classes do the mental categories of perception belong? (1) Do they belong to the class of sensations and images? The 'mental categories' are not, of course, sensations, and we have already shown that they can not be directly interpreted as images, because they exist without imagery. Dr. Ach, how- ever, has a theory 2 by which they might be the combined effect of many images. They are not images but the tendencies of a whole host of images to reproduce themselves. This theory was excogitated to explain the meaning of words. A word is 2 Cf . above, p. 86. 182 University of California Publications in Psychology. [Vol. 1 understood because it sets a number of images in readiness, all of which have a tendency to reproduce themselves. This tendency of the images to reproduce themselves is the meaning of the word. Against this as a theory of the meaning of words one may object : (a) If a single image can not constitute a meaning it is hard to see how the tendency of a whole host of meaningless images to come into consciousness would constitute a meaning. (6) If we refer to the section in the experiments of Buhler 3 entitled "Ueber das Auffassen von G-edanken" we will see that the 'mental categories' which were used by his subjects in the understanding of sentences cannot be analyzed into any known form of mental imagery. (c) Furthermore, words express objects for which we can have no adequate imagery. How then can the mere tendency of this inadequate imagery to reproduce itself constitute the meaning of the word? The same objections which prevent our acceptance of Ach's theory as an explanation of the meaning of words preclude its application to the 'mental categories' of our own experiments. It would explain meaning by the tendency of meaningless mental contents to reproduce themselves; for pure sensation independent of its associations has no meaning; neither has an image. It must be associated with other mental states to be understood. If these mental states are themselves but a host of images, each one of which has no significance in itself, from their combinations we can not bring about meaning. Nor can this tendency to appear in consciousness be said to constitute meaning. For the mere tendency of meaningless mental states to appear in consciousness would give no meaning that was not in these states themselves. One might challenge the statement that pure sensations or mental images independent of their associations have no meaning. Let us therefore develop this point a little further. 3 Archiv fur die ges. Psychol., 12, pp. 12 ff. 1910] Moore : The Process of Abstraction. 183 Whatever may be our theory, it is a fact that a complex of sensations on being received into the mind is interpreted. This is evident from our section on perception. The interpretation takes place by means of the something that we may term 'mental categories,' to which the sensation is associated. These give it a meaning. But suppose the sensation is not assimilated to these mental categories? Is this not merely to say that it is not understood and has no meaning? What is left to meaning when you deprive it of every possible association and every mental category into which it might be resolved? It dwindles to nothing and ceases to be meaning. These mental categories possess meaning by their own right and are qualitatively distinct from sensations and images. One might bring in at this point Kibot's 'intentional' theory of the mental image. 4 Sensations and mental images are signs of their objects. But as we said in our passing criticism of Eibot, if the mental image is a sign of the object that it repre- sents, it must be understood. On one side of the sign is the object signified, on the other is the meaning of the sign. If the mental image is a sign it must not only have an object but also a meaning. Consequently, to say that the image is a sign does not help us to get along without any kind of an idea or concept which functions as a meaning. If, therefore, by acting as a sign sensations and mental images cannot account for meaning, if they themselves are not the meaning, we must seek for meaning elsewhere than in sensations and their mental images. However, if we could take Dr. Ach's "Vorstellung" in the sense of a mental 'concept with meaning' we have in the theory a good analysis of a number of those states which Marbe and his followers have termed " Bewusstseinslagen." 5 They are mental states in which several concepts tend to appear in con- sciousness — but no one succeeds in doing so. As a result, you have a more or less unanalyzable mental state without definite * Cf . above, pp. 78 ff. s Cf . above, p. 85. 184 University of California Publications in Psychology. t Vo1 - 1 characteristic. The tendency of the many 'concepts with meaning' to appear in consciousness results in an imageless mental content, which is hard to characterize, simply because many characteristics tend to come before the mind but no one succeeds in doing so. (2) Are the 'mental categories' feelings? Those who hold to the opinion that feelings of pleasure and pain constitute the sole elements of our emotional life will not be disposed to seek in these affective states an interpretation of our 'mental categories.' These 'mental categories' express knowledge; and knowledge is not pleasure and pain, though it may be pleasurable or painful. Nor does it make any difference how we may extend the idea of feeling; if we still mean by it something that is not knowledge, then thoughts and 'mental categories' can never be explained in terms of feeling. For if the word 'feelings' remains an exact scientific term to desig- nate those very mental processes which do not give us knowledge, if feeling is opposed to sensation and to all our cognitive mental processes, then the 'mental categories' we have defined above are not states of feeling. Such considerations as these could hardly have escaped Wundt. Yet he would interpret our 'thoughts,' and I suppose what I have termed 'mental categories,' as a complex of images and the "adequate" feelings which are involved. Our 'mental categories,' he claims, are not feelings alone and not images alone, but a complex of both. But if imagery is in itself mean- ingless, if we can have 'thoughts' which are not images, then the representative function of our thoughts and 'mental cate- gories' must be performed by the 'feelings.' No single indi- vidual can place a limit to the meaning of a term. Thorndike calls every single one of our mental processes a 'feeling.' To this even Wundt would object. Still, if he were to insist on embracing under the term 'feelings' the representative con- tent of our thoughts as well as their affective tone, he should at least admit that there are two very distinct classes of feeling — one which gives the affective tone and another which repre- 1910] Moore: The Process of Abstraction. 185 sents the object. Wundt has nowhere made this admission. In fact, from his writings it would seem that the representative function is ascribed by him to the imagery in the complex mental content termed a 'thought.' But on being accused of this by Buhler he strenuously objected that Buhler had not read his works 6 and maintained that in his analysis of thought there was also the concept of the feelings. Consequently, the question arises: Do these feelings represent the object or not? If not, they can never account for the representative function of 'thought.' If they do, then surely we must classify our feelings into those that represent an object and those that do not; for it is certainly clear that there exists a large class of feelings which are not representative of objects. If then there are 'feelings' which can represent an object, how, we may ask, does this come about? How in the absence of imagery, and independent of it, can any combination of Wundt 's entire tri-dimensional system of feelings account for the meaning of words and phrases or the mental categories formed in the perception of our figures? Pleasure and pain, tension and relaxation, excitement and repose, might conceiv- ably combine to form complex emotions, moods, and a variety of non-representative mental states which accompany our pro- cesses of recognition, abstraction, analysis, etc. But that they should take over in their combination a function which is qualitatively distinct from any that is inherent in them as elements, is an unwarrantable assumption. One might bring forward at this point the following objec- tion : Tour contention that there exist imageless mental contents is based in great measure on the experiments in which a common element was certainly perceived, although the subject did not at all know precisely what kind of a figure was present. But to conclude from such experiments that imageless ideas exist is not warranted, because the experiments may be explained without such an assumption. These experiments represent those cases in which the common element was never seen in the focal e Psychol. Studien, 3, pp. 347-348 (note). 186 University of California Publications in Psychology. IT o1 - 1 point of consciousness. But wherever it was perceived, however far in the background it might be, it gave rise to certain feelings of relaxation and restfulness, perhaps even of pleasure or dis- pleasure. The peculiar combination of these feelings gave rise to the feeling of certainty that a figure was being repeated. This feeling had connected with it no visual imagery that the subject could recall. From such an analysis it is evident that from the lack of mental imagery you can not jump to the conclusion that there are imageless concepts. Such an objection would not be based upon a complete analysis of the evidence. The existence of imageless concepts is not founded solely upon these rare cases — but also upon cases in which the subject was certain of parts of the figure that could easily have been drawn had any visual imagery been present. The subject described things such as points or curves or angles, which certainly could be pictured, but claimed to have no picture and could draw nothing that would represent his state of mind. Now, points and angles and curves are not mere feelings. And if they are present to the mind without imagery they are not images. Furthermore, in the cases of recognition of figures without any knowledge whatsoever of their special nature, it is perfectly true that the basis of the subject's judgment to a large extent was some combination of feelings such as was mentioned in the supposed objection. But we must not forget that the basis of a judgment is not the judgment itself. And we must also remember that in all these cases there is in the subject's mind the abstraction, 'some kind of a figure' plus the knowledge that 'the figure was repeated.' The knowledge expressed by these two terms constitutes the judgment, 'some kind of a figure was repeated.' This judgment is not constituted by the feelings which evidenced the presence of a figure. It is based upon them, but it does not consist in them. It is therefore something over and above them. The elements into which this judgment can be analyzed are the abstraction 'some kind of a figure' and the knowledge that it ' was repeated. ' 1910] Moore: The Process of Abstraction. 187 Since these elements are not feelings and are not mental images, there is nothing left in the current division of elemen- tary mental processes into which they can be relegated except the acts of will. But certainly we can not place them there. We must therefore recognize the existence of another division of mental processes to which our thoughts and mental categories must be relegated. Consequently in the final product of ab- straction there is an element distinct from imagery and feelings. This element, since it is the bearer of the meaning, is the kernel of the product and it may truly be termed the 'thought' or the 'concept.' Imagery and feeling may cluster about this concept; but as far as the imagery is concerned it is certainly lacking at times, as our experiments have shown. As to feeling, we can not say for certain on the basis of our experiments whether or not it is necessarily present. The concept of the figures in our experiments, though dis- tinct from imagery and feeling, was not itself an elementary process. It was manifestly compound in a number of instances. For one and the same figure was assimilated to several mental categories. It was a concept made up of several more elementary concepts. Between the concepts of which it was constituted there was a conscious bond. The sensation in being assimilated picked out its categories by the necessary process of its assimi- lation and these united to form a concept of the figure which the subject was afterwards enabled to analyze with more or less completeness. If such is the case one may ask how was the first concept formed? Does man come into the world equipped with a whole system of mental categories by which he is enabled to perceive and understand the things about him? This question leads us on into a problem far beyond the limits of the present research. Our problem has been the analysis of the process of abstraction in the adult. The process of perception which is the initial stage of abstraction was found to be one of assimilating the sensation to previously formed mental categories. Whence originated these mental categories, is another problem. These 188 University of California Publications in Psychology. ITol. 1 mental categories and their function in perception are facts. The origin of the mental categories, and the process of perception and abstraction in the child, are very different problems from our own. But ignorance of child psychology does not destroy the facts of adult mental life. However, it may not be out of place to suggest a theory as to the origin of our mental categories. And this I would do as follows : As Kiilpe suggests, 7 the data of sense are perceived. There exists something of the nature of an 'inner sense' — a central consciousness which perceives the phenomena of the external senses. When consciousness first dawns the data of external sensations are perceived. Perhaps at first in the auto- matic life of the child the sensations that are perceived are more or less intermittent and vary in their nature. But every time a sensation arouses consciousness the child is aware of a change in its mental life. At first this change is not interpreted because there are as yet no mental categories. Every change is just an awareness. The child simply realizes that something has happened. And this realization develops into his first mental category. As time goes on, experiences multiply and the several different kinds of experiences make the child not only aware that 'something has happened' but that something of a more partic- ular nature has happened. Something painful, something pleas- ant — something hot or something cold, etc. In this way he forms still further sets of mental categories into which his future experience is received. Out of these develop the categories of identity and diversity, — when, we do not know; nor is it neces- sary for us to settle this point here. But by a gradual deter- mination of the most general of his mental categories — 'some- thing' — his experience grows and is assimilated. The first determinations are of very particular experiences. The most varied things are given one and the same name, simply because he has but a few general concepts and his sensations are assim- ilated by necessity to whatever categories may have been devel- oped. The child's experience — his inner perception of a train 7 Bericht ii. d. I. Kongress f. exp. PsycJiol. in Giessen, 1904, p. 67. 1910] Moore: The Process of Abstraction. 189 of similar mental events constitute a mental category which is his idea of those events. The first mental category, the child's awareness of something, enters though not consciously and ex- plicitly into all his later concepts. Some of these later ones group together, and so on, until under the influence of language and education the events of the external world receive their interpretation. 190 University of California Publications in Psychology. IT o1 - 1 SUMMARY. We are now in a position to summarize the process of ab- straction as revealed in our experiments. 1. The process of abstraction is initiated by the breaking up of the group presented for perception. In this breaking up of the group the common element becomes accentuated at the expense of the surrounding elements. These are not merely neglected, but are positively cast aside and swept more or less completely from the field of consciousness. 2. This breaking up of the group initiates the process of perceiving the common element. This is accomplished by assim- ilating to known mental categories the sensations perceived. Perception proceeds from that which is more general to that which is particular. The formation of a reproducible image represents a later and unessential stage of perception. 3. The retention in memory of the figure perceived depends in great measure on the method of memory. Memory by analysis and association has a very decided advantage over memory by imagery. The memory of the figure depends, fur- thermore, upon the focality of perception. The accuracy of memory decreases rapidly with the distance of the figure from the focal point in the act of vision by which it was perceived. The perception of new groups after a figure has been perceived has a tendency to obliterate it from memory. 4. The recognition of a figure once seen involves an element of certainty or uncertainty. Consequently there is implied in recognition assent or doubt, and therefore a judgment or a suspended judgment. In recognizing a figure any degree of certainty of recognition can accompany any degree of perfection in the perception of a figure, so that a subject may be certain of the repetition of a figure and still may have no knowledge as to what manner of figure it was — or the subject may know 1910] Moore: The Process of Abstraction. 191 all about a given figure and simply draw it as remembered, or as very doubtfully the common element. 5. The final product of abstraction, that which is perceived as common to many groups, is essentially a concept distinct from imagery and feeling. It is not an elementary concept, but represents the assimilation of that which is perceived by the senses to a more or less complex mental category, or perhaps to several such categories. These mental categories may be re- garded as the results of past experience. 192 University of California Publications in Psychology. ITol. 1 REFERENCES.* Ach, Narziss: Ueber Willenstdtigheit und das BenTcen. Gottingen, 1905. Angell and Harwood: "Discrimination of Clamgs for Different Inter- vals of Time." Part I, American Journal of Psychology, 1899-1900, Vol. XI, pp. 67-79. Part II, by Prof. Angell alone. Op. cit., Vol. XII, pp. 58-79. Aster, E. von: "Die psychologische Beobachtung und experimentelle Untersuchung von Denkvorgangen. " Zeitschrift fur Psychologie, 1908, Vol. XLIX, pp. 56-107. Baglet, William Chandler: "The Apperception of the Spoken Sen- tence." American Journal of Psychology, 1900-01, Vol. XII, pp. 80-134. Berkeley, George: A Treatise concerning the Principles of PLuman Knowledge. Vol. I, Fraser's, Oxford (1871) Edition of his works. Bigham, John: "Memory." Psychological Review, 1894, Vol. I, pp. 453-461. Binet, Alfred: L 'Etude experimental de I'intelligence. Paris, 1903. BfJHLER, Karl : ' ' Tatsachen und Probleme zu einer Psychologie der Denkvorgange. " Archiv fur die ges. Psychologie, 1907, Vol. IX, pp. 297- 365; 1908, Vol. XII, pp. 1-92. : ' ' Antwort auf die von W. Wundt erhobenen Einwande. ' ' Archiv fur die ges. Psychologie, 1908, Vol. XII, pp. 93-123. : "Zur Kritik der Denkexperimente. " Zeitschrift fur Psy- chologie, 1909, Vol. LI, pp. 108-118. Davies, Arthur Ernest: "An Analysis of Psychic Process." Psy- chological Review, 1905, Vol. XII, pp. 166-206. DtJRR, E.: "Ueber die experimentelle Untersuchung der Denkvor- gange" Zeitschrift fur Psychologie, 1908, Vol. XLIX, pp. 313-340. Galton, Francis: "Composite portraits made by combining those of many different persons into a single resultant figure. ' ' The Journal of the Anthropological Institute of Great Britain and Ireland, 1879, Vol. VIII, pp. 132-144. : "Generic Images." Proceedings of the Royal Institute of Great Britain, 1879, pp. 161-171. : Inquiries into Human Faculty, New York, 1883. Grunbaum, A. A.: "Ueber die Abstraktion der Gleichheit." Archiv fur die ges. Psychologie, 1908, Vol. XII, pp. 340-478. Huxley, Thomas H.: David Hume, New York, 1879. * The list is not a bibliography of the subject but contains merely those works referred to in the present study. 1910] Moore : The Process of Abstraction. 193 KtJXPE, O.: "Versuche iiber Abstraktion. " Bericht uber den I Kon- gress fiir experiment elle Psychologie in Giessen, 1904, pp. 56-68. Marbe, K. : Experimentell-psychologische Untersuchungen uber das Urteil. Leipzig, 1901. Mayer und Orth: "Zur qualitativen Untersuchungen der Associa- tion." Zeitschrift fiir Psychologie, 1901, Vol. XXVI, pp. 1-13. Messer, August: " Experimentell-psychologische Untersuchungen iiber das Denken. " Archiv fiir die ges. Psychologie, 1906, Vol. VIII, pp. 1-224. : "Bemerkungen zu meiner ' Experimentell-psychologischen Untersuchungen iiber das Denken." Archiv fiir die ges. Psychologie, 1907, Vol. X, pp. 409-428. Meumann, E.: "Ueber Associationsexperimente mit Beeinflussung der Eeproduktionszeit. " Archiv fiir die ges. Psychologie, 1907, Vol. IX, pp. 117-150. Mittenzwei, Kuno: " Ueber abstrahierende Apperzeption. " Psychol. Studien, 1906-07, Vol. II, pp. 358-492. Moore, T. V.: "The Process of [Recognition. " Atti del V. Congresso intemazionale di Psicologia, tenuto in Soma dal 26 al 30 Aprile, 1905, Roma, 1906, pp. 286-287. Orth, Johannes: "Gefiihl und Bewusstseinslage. " Sammlung von Abhundlungen aus dem Gebiete der Padagogischen Psychol, und Physiologic Edited by Ziegler and Ziehen, Vol. VI, No. 4, Berlin, 1903. Bibot, Th. : "Enquete sur les idees generales." "Revue philosophique, 1891, Vol. XXXII, pp. 376, 388. : L'Evohition des idees generales. Paris, 1897. Schultze, F. E. Otto: "Einige Hauptgesichtspunkte der Beschreibung in der Elementarpsychologie. I. Erscheinungen und Gedanken. " Archiv fiir die ges. Psychologie, 1906, Vol. VIII, pp. 241-338. : "Beitrag zur Psychologie des Zeitbeweisstseins. " Archiv fiir die ges. Psychologie, 1908, Vol. XIII, pp. 275-351. Spearmann, C. : " The Proof and Measurement of Association between Two Things. ' ' American Journal of Psychology, 1904, Vol. XV, pp. 72-101. Taylor, Clifton O.: "Ueber das Verstehen von Worten und Satzen." Zeitschrift fiir Psychologie, 1906, Vol. XL, pp. 225-251. Watt, Henry J.: " Experimentelle Beitrage zu einer Theorie des Den- kens." Archiv fiir die ges. Psychologie, 1905, Vol. IV, pp. 289-436. Wiltse, S. E.: "Observations on General Terms." Am. Journal of Psychology, 1890-91, III, pp. 144-148. Wundt, Wilhelm: Grundziige der physiologischen Psychologie. 5th Edition, Leipzig, 1902-03. : "Ueber Ausfrageexperimente und iiber die Methoden zur Psychologie des Denkens." Psychologische Studien, 1907, Vol. Ill, pp. 301-360. : "Kritische Nachlese zur Ausf ragemethode. " Archiv fiir die ges. Psychologie, 1908, Vol. XI, pp. 444-459. 194 University of California Publications in Psychology. [Vol. 1 APPENDIX I. THE INFLUENCE OF ASSOCIATION ON PERCEPTION. In the course of the experiments a number of eases occurred in which the subject's drawing of the common element differed from the actual figure in such a way that the error was evidently due to the association that was reported. Some of these cases are given below. The drawings given under the heading ''sub- ject's drawing" reproduce the essential characters of those made by the observer. They are not however exact reproductions of his drawings. Common Element. Subject's Drawing. Association. a CO "Wurst." Q Q Omega. B The subject drew the figure correctly at first. He then changed his mind and drew a see- ond figure with a n double curve, say- lx ing that this was more correct. He said the figure looked like a (Laufer). His second figure does in fact re- semble the bishop in some forms of chess- men. O ft Mushroom. X Two half-moons. A Open scissors. 6 Apple. 1910] Moore: The Process of Abstraction. 195 These errors lend additional evidence to the theory of per- ception advanced in the body of this work. The actual imagery arising from the figure itself is not the first thing noticed. It fits into and is interpreted in the light of the subject's past experience. The association comes into the subject's mind first. He sees that, and interprets the data of vision in its light before the true image is perceived. Had the series in which these errors occurred been sufficiently long there can be no doubt but that the error of assimilation would have been corrected. The true image which was constantly being impressed upon the retina would have eventually been noticed as it was in itself. But because perception does not consist in merely seeing with one's eyes but in interpreting the data of the senses, such errors as the above are not only possible but natural. 196 University of California Publications in Psychology. IT o1 - 1 APPENDIX II. GENEEIC IMAGES. "What looked like a fusion of mental images occurred twice in the course of the experiments. Such fusions are interesting in view of the generic image theory of ideas. The first case was less evidently one of fusion. The disk was inaccurately made. The accompanying figure occurred as the common element, now in one, now in another of the two positions as given. ® ® The result was that the two circles were drawn correctly. Just what was in the inner one remained doubtful. The second case seemed evidently a fusion of mental images. It occurred on a disk with no common element. The accom- A A panying two figures occurred several expositions apart. The subject drew the common element. The outline he said was subject drew /\ as the common element. The outline he said was certain, the dotted inner line was doubtful. Since the two figures appeared several expositions apart there can be no question at all of a retinal fusion. The phenomenon must be due to some central cause. Just such cases as these resemble very closely those postulated by the Huxley-Galton theory of general ideas. The common features are deeply impressed and therefore retained; the vari- able, but faintly, and are neglected. There is however a very important difference between the universal idea and such "generic images," as were found in the entire course of the experiments. In the formation of a general idea there is a kernel picked out as constantly recurring and therefore essential, while that 1910] Moore: The Process of Abstraction. 197 which is variable is neglected and forgotten, or recognized as unessential. If in our experiments there was any fusion of images at all, that which was common was indeed clearly impressed, — but that which was variable was neither neglected nor forgotten but remained obscure and doubtful. At all events the extreme rarity of the phenomenon postu- lated by the Huxley-Galton theory shows that it cannot be the usual way in which we form our concepts, — not even those of sensible objects. The analysis of abstraction made possible by the experiments points to a process that has little to do with composite photography. Transmitted December, 1909. LBJa'13 UNIVERSITY OF CALIFORNIA PUBLICATIONS-(CONTINUED) Vol.2. 1. The Dialectic of Plotinus, by Harry Allen Overstreet. Pp. 1-29. May, 1909 - 25 2. Two Extensions of the Use of Graphs in Elementary Logic, by William Ernest Hocking. Pp. 31-44. May, 1909 15 3. On the Law of History, by William Ernest Hocking. Pp. 45-65. Sep- tember, 1909 - — - 20 4. The Mystical Element in Hegel's Early Theological Writings, by George Plimpton Adams. Pp. 67-102. September 24, 1910 35 AMERICAN AECHAEOLOGY AND ETHNOLOGY.— Alfred L. Kroeber, Editor. Price per volume $3.50 (Volume 1, $4.25). Volumes 1-7 completed. Volumes 8 and 9 in progress. BOTANY.— W. A. Setchell, Editor. Price per volume $3.50. Volumes I (pp. 418), II (pp. 354), III (pp. 400) completed. Volume IV in progress. CLASSICAL PHILOLOGY.— Edward B. Clapp, William "A. Merrill, Herbert C. Nutting, Editors. Price per volume $2.50. Volume I (pp. 270) completed. Volume II in progress. ECONOMIGS.— Adolph C. Miller, Editor. Vol. 1. Gold, Prices and Wages under the Greenback Standard, by Wesley C. Mitchell. 632 pages, with 12 charts. March, 1908 $5.00 Vol. 2. A History of California Labor Legislation, with a Sketch of the San Fran- cisco Labor Movement, by Lucile Eaves. 461 pages. August 23, 1910 4.00 EDUCATION. Vol. 4. The Development of the Senses in the First Three Years of Childhood, by Milicent Washburn Shinn. 235 pages and Index. July, 1908 ....$2.00 A continuation of the author's Notes on the Development of a Child (Volume I of this series, 423 pages, 1893-1899, reprinted March, 1909, $3.50). Vol. 5. 1. Superstition and Education, by Fletcher Bascom Dresslar. 239 pages. July, 1907 : • • • •'■--- — 200 GEOLOGY— Bulletin of the Department of Geology. Andrew C. Lawson, Editor. Price per volume $3.50. Volumes I (pp. 428), II (pp. 450), in (pp. 475), and IV (pp. 462) completed. Volume V in progress. MATHEMATICS.— Mellen W. Haskell, Editor. Vol. 1. 1. On Numbers having no Factors of the Form p (kp + 1). »y Henry W. Stager. (In press.) MODERN PHILOLOGY.— Charles M. Gayley, Lucien Foulet, and Hugo K. Schilling, Edi- tors. Price per volume $2.50. Vol 1. 1. Der Junge Goethe und das Publikum, by W. B. R. Pinger. Pp. 1-67. May, 1909 2. Studies in the Marvellous, by Benjamin P. Kurtz. Pp. 69-244. March, 1910 2.00 3. Introduction to the Philosophy of Art, by Arthur Weiss. Pp. 245- 302. January, 1910 «...- - • - - - 50 4. The Old English Christian Epic, by George A, Smithson. Pp. 303-400. September 30, 1910 • -- l- 00 Vol. 2. 1. Wilhelm Busch als Dichter, Ktinstler, Psychology und Philosoph, von Fritz Winther. Pp. 1-79. September 26, 1910 75 UNIVERSITY OF CALIFORNIA PUBLICATIONS- (CONTINUED) „ AnT nrv w E fitter and Charles A. Kof oid, Editors. Price per volume $3.50. ZOGLOGY.-W. B . Rittei : ana ^| ries ^ £ } ^ ( 393) ^ v (pp. 440) com' Sted VoiuS'v?aM VII ?ta pToSs. Commencing with Volume II, this seriei SnSL the^trlhuSs f rom the Laboratory of the Marine Biological Associate of San Diego. MEMOIRS OF THE UNIVERSITY OF CALIFORNIA (Quarto). Vol 1 No. 1. Triassic Ichtbyosauria, with special reference to the American Vol. l. JNo. i.^^ By John o. Merriam. Pages 1-196, plates 1-18, 150 text , figures. Septemher, 1908 ...».._ - • ■ - * a, T Vol 2. The Silva of California, by Willis Linn Jepson. (In press.) Other series in Botany, Economics, Engineering, Entomology, Geology, Lick Observatoj Bulfetins, Sck Observatory Publications, Mathematics, Physiology, Publications of t Academy of Pacific Coast History, and Zoology. TTWTVTTiSlTV OF CALIFORNIA CHRONICLE.-An official record of ^versity life UNIVERSITY %^f^^ edited by a committee of the faculty. Price, fl.00 P* year. Current volume No. XII. ATHvmrr^TEATlVE BULLETINS OF THE UNIVERSITY OF CALIFORNIA.-Edited b; ADMINISTRATIVE gg^tor" «» Faculties. Includes the Register, the President' Report, the Secretary's Report, and other official announcements. Address all orders, or requests for information concerning the above publications K The University Press, Berkeley, California. Deacidified using the Bookkeeper process. Neutralizing agent: Magnesium Oxide Treatment Date: Nov. 2004 PreservationTechnologies A WORLD LEADER IN PAPER PRESERVATION 111 Thomson Park Drive Cranberry Township, PA 16066 (724)779-2111