* I \ THE PLACE OF ACCESSORY MEANING IN ASSOCIATION By HARRIETT MAYBELLE ANDERSON A.B. University of Illinois, 1920 THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF ARTS IN PSYCHOLOGY IN THE GRADUATE SCHOOL OF THE UNIVERSITY OF ILLINOIS, 1922 URBANA, ILLINOIS Digitized by the Internet Archive in 2016 https://archive.org/details/placeofaccessoryOOande THE GRADUATE SCHOOL Mayi_25., - 192 - 2 - i HEREBY RECOMMEND THAT THE THESIS PREPARED UNDER MY SUPERVISION BY Harr&e It Ya yb e 1 1 e An Is r s on ENTITLED The Place of Accessory Meaning in Association BE ACCEPTED AS FULFILLING THIS PART OF THE REQUIREMENTS FOR Recommendation concurred in* Committee on Final Examination* Required for doctor’s degree but not for master’s ' ' ' TABLE OF CONTENTS Page I Historical Summary 1 II The Experiments S III Conclusions 38 Bibliography 42 1 THE PLACE OF ACCESSORY MEANING IN ASSOCIATION HISTORICAL SUMMARY Since the first extension of experimental methods to the study of memory and thought, a number of serious attempts have been made to reformulate the traditional " laws" of association, to give an analysis of what has been called the "associative conscious- ness", and to measure the capacity of association and memory. We may take us typical examples the work of Ebbinghaus, Meurnann, the G&ttingen School, Arnold, Calkins, and Titchener. Ebbinghaus was the pioneer. His method of nonsense series has been very important in the development of quantitative research in this field. For the past generation this method has been under development by G. E. Mttller and his pupils. By the term association Ebbinghaus means nothing more nor less than reproduction, which he conceives as a process or as activity. "While attention means limitation, memory means expansion We refer to this ability of expansion by the term memory, to the actual process of expansion by reproduction or association . " x Meurnann, on the other hand, distinguishes between associ- ation and reproduction, reserving the former term for the linking 1. Ebbinghaus, H. , Psychology; an elementary text-book. (Tr. by M. Meyer) 1908, pp. 93-93. 2 or consoiidatin of impressions. "Association is the initial formation of a connection between ideas which are present in consciousness simultaneously or in immediate succession, or which occur at least as links in a chain of ideas which constitutes some sort of a unitary whole for us. Reproduction, on the other hand, is the appearance of ideas in consciousness, - a phenomenon ’which is rendered possible in most cases by a previously established asso- _ ciation between these ideas and other ideas." 1 2 3 Arnold says that there are two aspects of association. In the one it is to be considered as something; already existing, depending upon nervous processes; in the other, it is simply the process of recall. In the former, the emphasis is placed upon the first member, while in the latter, it is placed upon the second member in its relation to the former. In the first place, then, association is looked upon as a cause, much as Meumann re- ports it; while in the second it becomes effect. "Considered as a cause, therefore, asso- ciation is the functional development of a psychological disposition; as an effect, it is the realization in serial order of the con- tinuity implicit in the present moment."' In her study. As so ciat i on, Calkins starts out with the following provisional definition: "Association may be defined as the observable connection between successive objects or partial objects of consciousness, of which the second is not an object of per- ception. ,,iJ 1. Meumann, E., The Psychology of Learning, 1913, p.3. 2. Arnold, F. , The Psychology of Association. 1906, p. 72. 3. Calkins, M. W. , Association. ( Psychol . Rev . Mono? . Suppl . . No. 2, Feb. 1S96. p. l) . 5 At the end of her study she gives as a complete definition: "Association is the connection between objects or elements of consciousness (of which the second is not perceptual), assumed to be respectively identical with preceding objects, or elements of consciousness which have stood to each other in a relation of similarity or succession. " 1 Apparently she does not confine association to ideas. The first element in the complex may be perceptual. Furthermore, she brings in a ’self’ to explain 'identical objects'. "Truth to tell there is no recourse here save in the recognition of that 'inexpugnable assumption' of the permanent self beneath the changing phenomena. One can never have the same states of consciousness, in successive hours or moments, but one may be conscious in the same way at different times, and in that sense only the successive objects of pnes con- sciousness may be called 'the same'."^ In hi3 Textbook of Psychology (1910), Titchener refers to association as a "group of ideas", or as "such and so many ideas found uniforraally together in such-and-such ways" (p. 3??) ; but in his later Beginner's Psychology (1916) he confirms the theory of James that meanings, not mental processes, are associated (p. 149). The definition of Ebbinghaus is obviously uncritical. Ebbinghaus is not especially interested in defining or analysing association, but rather, in discovering aids and hindrances to mem- ory. It is natural therefore that he should identify association with reproduction. Meumann has distinguished between association and reproduction; but his definition is practical rather than scientific. He does not enter into the analysis of the associative Ibid . , p. 10. Ibid . . p. 1. 1 . 2 . 4 mind. We are, no doubt, justified, however, in assuming that his 'ideas' are the 'ideas' of the associationists; that is, that they are meanings rather thtn mental processes. Meumann, like Ebbing- haus, has made an attempt to discover the conditions which aid or hinder recall. Ebbinghaus' p reate st contribution to the field of associa- tion v/as the introduction of a new kind of material and a new method of research. By using meaningless material - nonsense syllables - it seemed to him possible to study the strength and duration of new associative bonds. Since his material could be readily varied on the quantitative side, it was practically unlimited in amount. From his own experiments he sought to derive laws expressing the dependence of reproduction upon such conditions as the length of the series, the number and distribution of repetitions, and the elapsed interval of time. Meumann, Arnold, Calkins, and Titchener all attack the old historical laws of association, which were first formulated by Aristotle, - the laws of similarity, contrast, contiguity in space and time. Meumann would substitute for them the special conditions under which association and reproduction take place. He divides these conditions into three groups, impression, reproduction, and forgetting. Under the conditions of impression he would include time (duration of process, number of appearances, rhythm, etc.), attention, and emotion; under conditions of reproduction, the num- ber of associated processes at recall, and the present state of consciousness . Arnold points out that the old traditional laws of 5 association are not 'laws', but are simply logical forms of associa- tion. After casting them aside, he attempts to formulate a real law which must, as he thinks, take meaning into account. The law - a law of "redintegration" - he states as follows: "Any element tends to reinstate the entire moment of which it previously constituted a part, which moment tends (l) to diffuse itself along some one of the paths which have been already formed; (2) to leave a trace of itself as a whole for future revival and development."* 1 2 - Eecause of the great complexity of any moment, it is impossible to predict at any time what the second member will be. The various factors which enter in may be named under the 'secondary 1 laws of association, ’.which he defined (after James and Calkins) as five; repetition, vividness, recency, primacy, and emotional congruity. Arnold reviews the existing classifications. Although they are based in part, upon experiment, they are, as he finds, by no means free from logic. liss Calkins, likewise, rejects the old laws of similar- ity and contrast. The essential point for her is the distinction between "concrete association of things and the association of elements as qualities."- Her own experimental study bears upon the "secondary" laws of frequency, vividness, recency, and primacy. Titchener emphasizes the integrative side of association, contending that it is the "situation" which conjoins the ■''actors. The situation serves, apparently, by setting up a nervous disposi- tion which accounts for subsequent revivals. 1. Arnold, F. , op., cut., p. 62. 2. Calkins, M. W. , op. cit . , p. 17. / 6 "Let us," he says, "cull the brain pro- cesses that are correlated with mental pro- cesses ' psychoneural ' processes. Then we may say: IPhen a number of psychoneural pro- cesses, all of which are reinforced and all of which stand alike under the directive influence of a nervous disposition, occur together under certain favorable conditions, then associative tendencies are established among them, such that the recurrence of any one tends to involve, according to circum- stances, the recurrence of the others." 1 * 3 Turning then to mind, he gives the following law: "If a number of vivid perceptions or ideas, whose situational context is the same, occur together under favorable conditions, then the later appearance in the same situa- tional context o* any one will tend to be accompanied, according to circumstances, by r . the reappearance (as ideas) of the others. Furthermore, as this does not take into account the fact that we may pass from one situational context to another, he adis: "If certain of these reapnearing ideas belong also to a different situational con- text, they will tend to be accompanied, again according to circumstances, by ideas which formerly occured together (as percep- tions or ideas) within that context." 0 Titchener has gone further in the direction of analysis than others. His description, also, of the underlying neural conditions does explain, though it is largely speculative. There seems to be little value in any forms of classifi- cation given. The nearest Titchener comes to classification is a distinction of the different forms of organization, or modes of connection, of mental processes. The types of organization are 1. Titohener, E. P., A Pecrinner's r 'sycholomy, 1916, p. 164. 3. Ibid., p. 166. 3. Ibid . . p. 168. I 7 fusion and conjunction, the same as in perception. However, he finds no satisfactory description of the connection of the complexes as perception and idea. Meumann's enumeration of the conditions upon which asso- ciation depends is fruitful. It organizes the material and problems in an adequate manner. Using this as a suggestion, we may sum up these conditions in such an outline as the following: I. Conditions operative at the rime of presentation 1. Amount of material 2. Imaginal type of the observer 3. Repetitions Number Distribution 4. Manner of learning By whole By part 5. Type of material Meaningful Nonsense 6. Rhythm 7. Intent to learn 8. Attention II. Conditions operative during the interval between presentation and recall 1. Length of time 2. Retroactive influences 3. Character of filling III. Conditions operative during the period of recall 1. Attention 2 . Fat i gue 3. Method used A. Methods of reproduction (1) Method of retained members (The observer is asked to repro- duce as much as possible without aid) (2) Method of learning by heart (Learning is measured by the number of repetitions necessary for com- plete retention) (3) The saving method (Ersparnismethode) (The observer is instructed to re- learn the series) 8 (4) The method of sample associates (Tref i erme thode ) ( P art of the material is given and conjoining parts are to be sup- plied) (5) The method of aids (Also a completion method, but in this case the observer is aided when reproduction fails) E. Methods of recognition (Usually all of the material is given together with some additional material, the problem being to recognize the old; Our own problem, a study in certain accessory factors in association, falls in the first group, for it concerns conditions operative during the period of presentation. It arises out of the common appearance of adventitious associates in recall. In re- calling the declension of a certain French lord, we frequently re- member its exact place on the page in our first French grammar. The recall of a motor trip may bring with it the memory that a companion wore a peach colored scarf which was blown by the breeze. It is commonly assumed, but without experimental verification, that such associates not only anpear in revival but that they are actual / aids to recall. To put the matter to the proof, we have selected certain outside factors; i_. e_. , factors which lie outside of the central meaning of the situation, and we have attempted to discover whether or not they do actually serve as integrating factors. In order to discover just the place they occupy in mind, we have taken pains to note where they stand as regards clearness and as regards organization, during the periods of impression and of recall. Fefore describing our own experiments, we should mention in review a number of related problems; 'mediate' association, the effect of 'subliminal* factors on perception, the possibility of 8 spontaneous survival through an alleged ' perseverative 1 tendency, and the problem of incidental memory.* No one of these problems is exactly the same as our own. Scripture, in his study of •mediate' association, and the investigators of the subliminal pro- cesses were dealing with unconscious or subconscious processes which were supposed to be connected in some unknown way with the observa- ble processes of mind. The assumption of a perseverative tendency is based upon an assumed rhythmical return of processes to mind. There seems to be no final evidence for the existence of such a ten- dency; and if it existed it would not account for our observations. In the problem of incidental memory we find something more closely related to our own, the chief difference being that in the former an effort is made to determine to what extent these outside factors may be recalled, while we are ourselves endeavoring to determine to what extent, if at all, obscure processes under impression are con- cerned in the associative complex. THE EXPERIMENTS In work done in our own laboratory antecedent to the present study, the learning of nonsense syllables was the main task and color was used as the adventitious factor. In order to obtain stronger associative bonds than this material afforded, German nouns were later employed with the articles 'ier', 'lie', and 'das'. As this material was found to lack uniformity, nonsense words were 1. See Potter, M. L. , The Role of Adventitious Associ- ates in Recall, 1921. 10 formed by combining two nonsense syllables and dropping one of the middle consonates (e_.g_. , sudow, fibux), and, in place of the three articles, the symbols +, =, and # were used. In making up the exposure series, two independent factors bearing upon our problem had to be considered, namely, color and segregation. The factor of color was brought in by presenting the 'words and symbols on colored cards. In the segregated series words with like symbols were grouped together upon the colors, while, in the white series, they were presented in a mixed arrangement. In order to exploit an entirely different kind of setting for the associative complex, three bodily postures, standing, sitting, and kneeling, were assumed by the observers, - thus throw- ing in, as advantitious factors, organic and kinae3thetic processes. From the results of both these procedures, nonsense words with colored backgrounds and with postures, the conclusion was reached that "segregation is a strong incentive to recall and advantitious associates aid recall only when they are given in con- J junction with segregated groups." 1 ’Border * Associates . It is at this point that we take up the problem for further experimental research. Our first step was toward further refinements of method. In the first place, white cards with colored borders were used in place of colored cards, thus removing differences in legibility, reducing after-ima?;es, and 1. Ibid., p. 33. 11 removing the color from the field of direct vision. Secondly, each word with its symbol was printed on a separate card both for the bordered and the borderless series, thus making these condi- tions the same. Thirdly, the interval was doubled in length at the end of each five words to break up the mixed as well as the segregated series into three groups of five. Fourthly, the timing device was improved with a soundless bob swung by a silk thread at a one-second length. Finally, the former exposure screen was replaced by a reading desk adjusted to the proper reading distance. The desk was covered with a large gray cardboard sheet having a window 4x6 inches. The exposure was effected by a gray cardboard shutter which opened and closed the window by means of rubber bands and a cord running over a system of pulleys. As before, a complete set consisted of four series of fifteen words each; five of the words in each series being pre- sented with =, five with +, and five with #. The four series were as follows: (A) A white mixed series (all plain white cards with the three symbols arranged in chance order) ; (B) a colored mixed series (the cards identical with those used in A except that they had borders of colored lines on the white sheets, five of them having red, five green, and five violet borders, the colors being arranged in chance order in reference both to position and to symbols); (C) a colored segregated series (cards identical with those used in B except that the same symbol appeared on cards having the 3ame colored borders in each case, thus the five #'s on cards with red borders, the five =’s on cards with green borders, and the five + *s on cards with violet borders, the symbols and • - ' ■ 12 colors in this case being arranged in segregated order); and (D) a white segregated series (the cards identical with those in A but arranged in segregated order) . White , Unsegregated Colored Segregated \. A t » B 4 > C ^ D Figures I, II, III, and IV illustrate Series A, E, C, and D respectively. For the first, part of the experiment, six complete series of the above description were made up, making 360 nonsense words in all. For reproduction, the words of each series were typed in chance order on a single sheet of paper. A series of cards was placed at the window with the shutter closed. The observer sat before the apparatus in a com- fortable position. The experimenter was out of view behind the apparatus. Each card was exposed for two seconds, followed by an interval of two seconds, except that the interval after the fifth and tenth cards was four seconds. The interval between repeti- tions was kept as uniform as possible, about ten seconds. An interval of thrity seconds, during which the observer was asked to do indicated sums, followed the sixth repetition. As has already been mentioned, each series was divided into three grourjs of fives. The groups of five were numbered 1, 2, and 3. In the six repetitions the three fives were pre- sented in every possible order; viz . , 13 + yazon = mupet + xoyej r f offuj = suhid = tovaq # dapoy + hi gem # nirr.ic •: sejap + refil Fi~ure I. Series A 14 = rax oh = be nek # payov = giwup 4- le gaw 4 - fijas I yeluf Figure II. Series E. r f umox 1 + qapuh = miqop f" i' hazeq j ahul = puzek # siyos + vuhag = yeduv 1 4i veyij - # boh oh 4- bigad = vivel + pcsen = cahas Fi ure III. Series C. 15 = buzib + turef # xehow = lafot + codiw * # zeran = yufop ■ 4 - hizim • # zusox = j ocad + bavec Z kiqet = majig + cowiq # j ituv Figure IV. Series D. 1 ? Presentation I - order i, 2,3 tt tl II - " 2, 3,1 n tt III - " 3, 1,3 n n IV - " 1, 3,2 tt tt V - " 2, 1,3 It H VI - " 3, 2,1 A short practice series was given each observer before the regular experiment was begun. At the beginning the observer worked, under the following instructions: "Two seconds after a 'ready' signal you will be presented with a series of nonsense words, each of which is written after a sym- bol. The series will be divided into three parts by means of short intervals. There will be six repetitions of each complete series. During the interval which follows the sixth repetition you will do indicated sums in addition and subtraction. At the end of this interval a new sheet, the repro- duction sheet, will be laid before you con- taining the nonsense words of the foregoing series. To these words you will affix as many of the correct symbols as you can. After each reproduction you will be called upon for introspections and comments." After the experiments had begun, it was found that two of the observers, Y and K, were using various devices for learning. This is shown by the following observational comments: "In some cases I read a meaning into the svllable and connect the symbol with a mean- ing." Obs. K: C(2) "The following shows meanings used to connect symbols and words: matoy = my toy, dicej -•-+ dice etc." Obs. K: A (2) "In this case, I was able to organize the words of + and = into a group. This aided much in recall." Obs.* K: D(4) . "A great many associations learned during the series acted as cues. For example, in gelug, g = g "Vamut" suggests "vomit" and "vomit" has a 'sharp' pain, therefore 'f'." Obs. Y . ■ le "There is an intellectual play which attempts to put meaning upon the combina-- tions and to find relations between the form and symbol. This play continues un- til the next exposure." Obs. Y The results already obtained from Obs. Y and K were then thrown out, as they were obviously obtained under widely warying and inconstant conditions. It was obviously necessary to rule out — if possible — •these artificial aids. For this purpose the instructions were modified as follows: "Two seconds after a ’ready’ signal you will be presented with a series of non- sense words, each of which is written after a symbol. Each complete series will be divided into three parts by means of short intervals. There will be six repetitions of each complete series. After each repetition, AT THE SIGNAL ’HEST' YOU WILL DROP THE SERIES AND PREPARE FOR THE NEXT REPETITION. DO NOT REVIEW THE PREVIOUS SERIES. AVOID LOGICAL ASSOCIATIONS IN LEARNING. During the interval which follows the sixth repetition you will do indicated sums in addition and subtraction. At the end of this interval, a new sheet, the reproduc- tion sheet, will be placed before you con- taining the nonsense words of the foregoing series. To these words you will affix as many of the correct symbols as you can, tak- ing them in the order given on the sheet. You will indicate the degree of assur- ance by placing the number 1, 2, or 3 before each sumbol. (l = very certain, 2 = fairly certain, a.nd 3 = no certainty) After each reproduction you will be called upon for introsx^ect ive comments." Furthermore, the observers were asked to mark all pairs in -which ’logical’ associations appeared. IS The numerical results obtained are tabulated in Tables I, II, and III. In every case the totals for A are highest, the difference being greatest for Obs. Y. When the results are thrown into graphical form, a dip appears at B for all three Obs. (See graphs P.22) In the cases of Obs. Y and 0, there is a rise again at D. The differences for Obs. K were so slight throughout as to have little significance. We find wide variations in the result of the individual observers, which are not accounted for in the introspections. While the general tendency is clear in the totals, it- is so slight as to be masked in the single sets. For Obs. 0 the total per cent, correct in A is 85.7, (M.V. 11.6); in B, 81.2 (M.V. 9); in C, 83.7 (M.V. 8.6); and in D, 83.5 (M.V. 10.6). In no case do we find an adequate explanation in this observer's commentary for these serial differences. We do find explanations, however, for some of the differences in accomplish- ment in the results of the other observers. Obs. K notes in A and D(s) distraction coming in through an effort to inhibit logical meanings. In (3) where B stands relatively high, he notes that there was particular attention to colors during the fifth repeti- tion. Obs. Y mentions an inhibited tendency to name the colors in C(2); in B of the same set, a difficulty to maintain uniform atten- tion because of extraneous associations; and in A, a lack of inter- est. As shown by the tables the percentage correct in these cases was decidedly low. For C(2), where the number (percent.) correct was only 46.6, Y reports a drop in intent to learn. We are led to believe that these wide variations in the i ! ‘ 20 P * 1 1 O rH * CO CJ CO CT> m CO - O 1 — 1 rH rH 1 — 1 rH 1 — 1 n £h S P • o • to vo in c- i o H"* o 00 .S3 to S3 • • • • • • • P o in c 00 rH p o o o O 1 c- O © O P CO in o o CD o S-. £ rH rH © -H - O o CO 03 o o o © si CO o m 03 CO i n o o Sh p rH t — i bib 5-« " o O o a O o © © rH o o c o o c Q O rH rH rH iH rH 1 — 1 p 1 o I— 1 • in CJ CJ * — 1 CJ CO - - o i 1 rH rH rH 1 — I rH O Eh S P 3 O • • o 03 o 03 CO m in 43 to to • • • • • • « P o S3 o CJ o O CO CO CO V° •rl G » o cr> 03 C~ D- c- CO O *3 ^ 1 — 1 P <*H >> , o CO CO o O P CO 1 1 CO © c Jh © -H o ID c o c o J-. © aS Cl o CO in o o CO O p< p 1 — I rH . - jH o to u — a a O o o o © © T 1 o O o o o o Q o 1 — 1 1 — 1 1 — 1 rH rH 1 — 1 p + o rH * CO in CO CJ CO cr> C\J » - O 1 — 1 t — 1 rH rH rH o £h 55 p 3 O • CO o C- in CO CJ 43 • b£> • • • • • • • p to S3 CO o £> 00 CO c- V 5 . •H O CO o 03 c- CO m 00 & g £Sl rH p «W >1 o o o c o o o O P CO o o o o a> S3 © -H o o iH o CJ CO i © © CJ o o o in 00 CO o Sh p o to {-. & o o o o o © © rH o o o o o c Q O 1 — 1 1 — 1 1 — 1 iH rH 1 — 1 p J o rH * CO m CO cr> CO CO - - O 1 — 1 rH 1 — 1 rH o Eh p pi o • • c- in o c— o CF> c- 43 M to • • • • • • • P o S3 in CJ o in o o CO •H _3 — CO o CO in c n CO gs rH V » o C- O o o t- o p CO in ID o p £ o © -H o o o m o © © CO CO CO o o c- o Sh t -4 + 1 H Sh to S^ o O o O o o o © © — 1 o o o C o o i o O o ■ — ! rH i — l rH J — 1 1 — 1 CQ © • • •H > ttl s- rH CJ CO in lO <3 1 -a © L° CO G P o © u u o o rH • 9 OoH^COrl^OOD rH rH i — 1 i — ! i l rH i — 1 CO V. c- O CO O C- CO o t~ ID O t— 00 CO 0— CO *D p • H OOOOrHrHOCpCOCO o i — \ i — ! rH iH iH j — 1 00 © Eh O o Sh S-. COC'C-COCOC-C'C- o ©©rjlt-t-©©© o p rH • O CO 01 i — ! CO O CO CT> 'si* o - o s — 1 rH rH j — 1 C> © Eh S Q O O CO C- !> C- O u ^ CJ © t~ © © © © O VL o p rH * ai^WHinmno^ o - O £h JS <3 S-. o on o cj o c- c- lOC-COt-OCOCOlO o V?. rH o © a) •H rH C! CO vf pH CO CO ^ f-i P o O 03 Eh 31 M H H w § E-i o Cor- rect Tot- al No . -II SI SI 9 II 01 01 Without Log. M'ng L 64.2 54.5 63 .6 35.7 36.6 80.0 64.4 Correct ( %] Degree of certainty CO o co o o » • CO O cC5 02 iC) 02 O O r—tO- C- C0 'O 02 t- CD 001 001 001 001 001 0 o Cor- rect Tot- al No. 13 5 12 13 12 7 11 3 O » bO to s -P o - * k T ►— i ' 86.6 25.0 70.0 85.7 80.0 46.6 66.6 Correct (‘ Degree of certainty 001 0 02 67 27 57 82 40 39 fH © o o o o o O O o O cr> O 1 — I 1 — 1 < — I I — t 1 — 1 Cor- rect Tot- al No . 10 3 12 9 13 12 11- Correct (%) Without Log. M'ng. 66.6 46.1 78.5 60 .0 86.6 61.5 65.8 Degree of ■certainty co 0 33 LOO LOO 60 02 64 50 67 60 75 50 <-4 100 100 100 100 100 < Cor- rect Tot- al No . 15 8 14 13 14 13 13- Correct ( %) Without Log. M'ng 100 50 90 84.6 92.8 36.6 83.9 Degree of certainty CO 100 0 LOO LOO 100 I 02 100 64 86 83 89 80 100 100 100 100 100 100 Series rH 02 CO 02 • k" < 23 results represent some variable factor in the self -instruction or on the side of extraneous processes which have not appeared in the introspections or comments. The numerical results indicate that the colored borders did not actually serve as mediating factors; but - at least in the case of Obs. G and Y - as positive distractions. Furthermore, there is insufficient evidence that segregation served as an aid to re call . The next problem seemed to be to get a more detailed analysis of mind at the time of presentation, the object being to discover by what processes these extraneous factors were carried and -where they stood in the total organization of mind. To this end a number of series were completed under the following instruc- tions : "Two seconds after a 'ready' signal, you will be presented with a series of non- sense words, each of which is written after a symbol. Each complete series will be divided into three parts by means of short intervals. After the first repetition, AT THE SIGNAL 'BEST' ’"ILL DPO? THE SEPIES AND PREPARE FOR THE NEXT REPITITION. DO NOT REVIEW THE PREVIOUS SERIES. AVOID LOGICAL ASSOCIATIONS IN LEARNING. Following the second repetition you will either be ashed to reproduce the symbols or to give an introspective account of the presentation period." The observers reported visual, auditory, and kinaesthetic processes; but we got little information as to just what these pro- cesses stood for and as to their place in mind. Subsequently, an ' 24 analysis-sheet was used, upon which the observer was asked to desig- nate, not only the mental processes but also their clearness value. Clearness was designated by numbers ranging from 1 to 5, 1 being very clear, 2 fairly clear, 3 medium, 4 fairly obscure, and 5 very obscure , The introspective results obtained in this manner are shown in the following table: Processes Visual sensations Words Colors Auditory sensations Exposure Kinaesthetic sensations Verbal Visual imagery Auditory imagery Kinaesthetic imagery Obs. G Clearness values Obs. Y Obs. K 1 be com. 2 or 3 1 and 2 1 2 i 1 to 3 4 and 5 4 and 5 4 and 5 1 and 2 2 and 3 3 to 5 occasionally 4 and 5 (also after Im, ) (after images) (afterimages) 3 and 5 " 3 3 occasionally 3 or 4 These comments throw considerable light upon the organization of mind during the period of presentation and give strong evidence that the colored borders were distracting factors. 1 Since the introspections do not account for the fact that the totals for A (white unsegregated series) were higher than those for D (white segregated series), it was suspected that the reason must lie in a different self-instruction for the segregated and the unsegregated series. The observers were asked to report upon any 1. The clearness values of Y were not wholly reliable, as this Obs. did not sharply distinguish between process clearness and cognitive clearness. 25 difference of disposition or set which appeared in these two series. They were not, however, informed of the quantitative results. In the case of Obs. G, we found that in the D series, during the pre- sentation period, there was a tendency for the common symbol to be carried over by obscure processes from word to word throughout the part-series. Ye should expect this carrying over to improve repro- duction. It was also found that in the unsegregated series each pair was silently repeated two or three times during each exposure; while in the unsegregated series each pair was repeated only once. Obs. K reported the A series to be more difficult, so effecting the attentive set. Obs. Y also reported a like difference in attentive se t . 1 Angle * associates . To get rid of the distracting ele- ment of color and to secure a stronger bond, something more nearly parallel to position on the printed page, a new kind of material was used. The words and symbols of the B and C series, in place of being centered on the white part of the cards were placed in the different corners (five of each series in the upper right-hand cor- ner, five in the upper left-hand corner, and five in the lower left- hand corner). Lines were drawn perpendicular to the edges, enclos- ing the word and symbol in the center of the space and emphasizing the quadrant where the pair appeared. The white sheets were then pasted on the gray cards in such a position that the fixation point was kept constant. Samples 26 A new lot of words was made up by reversing the old ones and then marking out all which duplicated words already used. Four complete set 3 (240 words in all) were made up. The A and D series . P bG £ 0 CO CO O ^ — > •H ,_} %.-H GO CO cr> 0 •H 0 m ■cf in in V. Vi n >1 1 O 1 O > m 0 O p o p 00 p 0 H-> CO 03 in O c 0 © © •H cO c- O © © •H t- cO C~ u © ctf 03 CO cO in >- © cri 03 CO in u G p u IQ p o to G 00 O 0 0 to O • 1 o © © 1 1 00 O 0 0 © © 1 1 O Q © rH rH « 0 cH , -P 1 1 P 1 u o p • t. O p ft 1 o © o o CO ■cF in 0 O 0 rH O CT> in P cC rl o- CO 0 CO •© 1-4 r^i cn Cl CO Ti V •H 0 mrn m CO 00 o V fe rH 0 > •-4 ^2 c; CJ p o p O © p CfH >1 0 1 O © p >1 O 0 c~ G o O p CO O S 0 O P 00 O 03 CO © c rH © G rH p G © •H 0 0 O p u © •H in CO CO G © d 03 0 m O w u © oi 03 in 03 CO o g p rH rH G 0 Sh p •H o bQ G CvJ O 0 •r! 0 bG $-• 0 « © © i — ! CJ* O 0 © © 1 1 l 0 — ' c O rH 1 — 1 — - CQ O rH 1 -p i t P 1 > u O p • + > Cr O p ft I M o © O i — 1 o in O © 0 rH O 0 c~ CO o- o g EH 3 P 7 rH rH 1 — 1 w O U fcH © S w CQ p P < 3 • CO e- 3 • 03 CD 0 00 £-* o • bO • ft • O ft to ft ft ft ft c{ to r-» t — ! 0 in in QQ to c C31 03 0 03 •p o m cr> 0 cn cr> P 0 — cO 03 sF Vi •H G rH •H »-4 s: CQ p >, O 1 l CQ >1 c 0 0 o o p CO m P O P CO 0 o rj O ft rH G © •H 0 O 0 © © •H in CO IQ © 3 03 0 O CO s* © 3 03 C~- CO o g p fc. u p o to g O O O 0 bli 0 0 O © © rH 1 0 O 0 0 9 © rH in 0 0 g o o G H re 1 1 rH rH rH 0 E - 1 © S p • O CO 03 p 3 • to ft ft ft 3 ft 0 CO CO CO O bC c in 0 03 0 ft to • • ft ft ,c o to- CT» 0 CO -G to G Q CO CO cO p i-4 rH P O - CO 03 in ■cF i-4 rH V V < >> 0 0 1 < in >. O CO » p o P 00 0 in in p O P CO CO < CO CD ft © ft • •H 03 H* rH > ft •H 03 > CO u 03 H« rH < © IQ 03 rH *4 .n © ® O c n CO - - . 2S •c o X -p » o -p s O -H e> oJ J . -p to u a s CO i i i 02 0 50 rH ~C= lD t co e- c Cor- rect Tot- al No. CO CO rH CO Correct ( %) Wi thout Log. M'ng. 75 50 25 Degree of certainty CO 1 I o 02 1 o o lT> 1 ! 75 50 0 Series 4 — rH CO ^ Crf rH CO cj. -P O 32 "In this case there was a definite ten- dency to place each word in one of the three corners. The symbol was associated with the corner as much as the word itself. The cor- ner arrangement became an aid in this way in connecting symbol and word, though at the first there was some distraction of attention, due to the newness of the situation." Obs. K C(33) "In all cases of those marked 1, the 'words and symbols were recalled as belonging to the corner in which they were presented." Obs. K C(44) T’e find that, for Obs. G and Y, the A and D series have changed places. Whereas, in the first group the numerical results were considerably higher for A than for D, in this group we find A is slightly lower than D. This may be accounted for by the period of introspective training at the end of the first group. Due to this training there seems to have been less self-instruction, the observer taking a more uniform attitude toward the different series. As before we found wide variations in the results. Obs. G’s total (percent.) correct in A was 51.1 (M.V. 2.9); in B, 43.6 (M.V. 14.1); in C, 42.2 (M.V. 10.3); and in D, 53.3 (M.V. 4.4) For Obs. Y the average deviation runs as high as 20 in C where the total correct is only 43.3$. We have not enough results here to be determinative, especially in view of the wide variability. The results obtained by the Trsf fer method of reproduc- tion are few and without significant differences in the four series. It was thought that the small differences present might become more distinct if the observers ?/ere placed under heavier pressure. The number of repetitions was therefore reduced to . . ' three and the exposure and interval times were reduced to 1.33 seconds. The first method of reproduction was again used and the observers were instructed to write down the symbols as quickly as possible, no time being allowed for introspection or comment. Under these conditions six series each with K and Y were given at a sitting. The results obtained are tabulated in Tables X and XI. It is interesting to note that in this case, Obs. K's total-number correct for A falls, for the first time, materially below his totals for B, C, and D. This accelerated method had many advantages, and valuable results might have been obtained with fresh material and more data. One bad factor was the old material in which some combinations were mere familiar than others, having been better learned during previous presentations. From the observers* commentary we find differences as to the general attitude toward the experiment. Some cf the differ- ences may, to a certain extent, account for the above mentioned differences in the quantitative results. Apparently, Obs. K and Y set themselves the- task of "learning" and held themselves respon- sible for finding some means to accomplish the task. To this end they tended throughout to use different devices for holding the material together. In place of simply taking the word and symbol as they stood, given in juxtaposition, these observers attempted to find some means of binding them together; they tried to find some meaning relationship between the two. This self-instruction to discover and to dwell attentively upon meanings is extremely common in the task of "learning". On the other hand, the problem given was simply to maintain uniform attention aid to let mind and the ' TABLE X (Accelerated time) [ Cor- rect Tot- al 130. 13 15 12 14 14- « Correct (%) p 3 » O ♦ ho 43 bO $4 P O - te: 80.0 100.0 76.9 91.6 87.2 «P >» o p c co 100 50 © -H © 53 U P t>0 t, © © O O C\2 r 71 100 63 100 rH <5000 o o o o rH rH iH rH Cor- rect Tot- al No . 14 13 12 13 13 o VL Without Log. M'ng. 93.3 84.6 62.5 85.7 84.0 Correct l Degree of certainty CO ! O O 1 C3 O O O O O in io rH rH 90 100 100 100 PQ Cor- rect Tot- al No . 12 15 13 14 14- Correct ( %) Without Log. M'ng. 75.0 100.0 85.3 92.3 88.1 Degree of certainty CO LOO LOO w 67 100 50 80 1 — i 1 100 100 100 100 Cor- rect Tot- al No . 10 12 11 13 12- -P O © U u o o Without Log. M'ng. 66 .6 76.9 66.6 83.3 73.0 Degree of c ertainty CO 0 0 67 03 57 50 40 71 rH 100 100 100 100 1 Series 11 22 33 44 Av. 34 * - ' ’ - - 35 1 p t + 5-. O P • 00 o O C- CO CO O re To al No rH rH O p o CO CO rH CO cO o CO P • hO • • » • • * * p bo c CO cO C- 00 CO 02 HO- to CO n m n 5-° gs J SS o P o o o O O o P O P CO in in in o c: © © -H CO t- O o o o S-. © erf cc n CO in m CO H 4 v. 5-. P o M U o O O O o o o © © H o o o c o o O o rH rH rH iH rH 1 — 1 + I P » O t~ in cr> o CO 03 u © P • t 1 i — 1 o © O rH o J* £-> « S o p cO CO co O CO rH O 3 • • • • • • • 9 O • bfl CD 02 co o c- 4rf fcfi £ CO co co CO in in P O - 1 ^ ' •H J cr> < t. o P ■ rH rH rH o o ©oho H erf S P 3 • CO CO o- CO CO CO co &q O • bO • • • • • ♦ • ^ Me; CO CO o CO 02 t- CQ P o - tn CO CO CO n < V rH k-3 S E H m p *M >1 o i o o O o a O P CO in o © c rH f © -H CO 02 c- o o 02 fcl © erf 02 C" H 4 CO cr> CD 02 o fot o O ' o O o O O a s 1 1 o o o o o o s — 1 1 — 1 i — ! rH rH rH 1 p « 1 s~, o p • cO CO CO 02 CJ1 CT> o o 2 £ r erf a rH rH p 3 • o o CO in in o cr> O • bC • • • • • • SI 'ctf C O O co 00 CO o in P o - St 4 cf 00 t- 01 in ✓—> •H JS v° fcg c P >> o 1 CO 1 CO o p O P CO co CO o srf © © *rj 02 O o CO CO CO 5-i © erf 02 •eje c t- 02 t- u ^ P © u CO 02 CO < © O CC 67 nervous s\^stem do the rest, the observer simply recording faith- fully whatever came. The instructions against this sort of thing made it difficult to maintain a uniform and constant atti- tude toward the problem. This was especially true in the case of Obs. Y. A few observational comments will illustrate: "In this case I was able to organize the words of + and = into a group. This aided much in recall." Obs. K C(3) "In connecting the symbols with the nonsense 'words, I pronounced the words to myself and made movements with my fingers to correspond with the symbol. This seemed to help fix the visual impression." Obs. K E(3) "The learning is so mechanical that it is hard to maintain a uniform atten- tion. There was a tendency toward ex- traneous associations relative to matters of the day. These were somewhat inhibited." Obs. Y B(3) "It seems that there is a tendency to busy myself .with something, if not with logical associations then with extra mater- ial . It is very difficult to be very passive and say these syllables in a mechani- cal way and to think of nothing." Obs. Y C(2) "At the first of the series a feeling of discouragement - I can't ever learn these! It is hard to keep up interest in so mechan- ical a process," Obs. Y B(s) For G there was little evidence of this intent to "learn" and, apparently, a greater uniformity in general attitude. Another interesting group of differences appeared in the character of the processes in the foreground, during both the period of impression and that of recall. Obs. G, reported verbal kinaesthetic processes in the foreground during the period of pre- 38 sentation, while both Obs, K and Y reported these processes in the background and visual processes in the foreground. For Obs. G and Y rec all was carried in every case in terms of auditory-kin- aesthetic (verbal) and kinaesthetic processes, i_. e. , the word was chiefly audit or y-kinae sthetic and the symbol visual. Kinaesxkes- is (non-verbal) occasionally entered into the reproduction of the symbol. Obs. K had no name for the symbol; both G and Y had. Another interesting difference, though it has little bearing upon our present problem, concerned the reliability of .judgments of certainty. In the cases of Obs. K and Y, degree 1 of assurance almost always insured correctness. This was not true in the case of Obs. G, as only 61$ of those marked 1 (Table VI) were correct. The difference is, no doubt, due largely to the fact that G did not use logical meanings. The following comments bear this out: "No accessory processes upon which to base certainty. The only cue I would have is a little hesitancy in the appearance of the kinaesthesis and in a very general affective and ideational judgment of the •whole period." G. "1 usually means that there was some cue.” Y. CONCLUSIONS We set out to determine whether or not certain adventi- tious factors, factors lying outside the central meaning of the situation, were aids to recall. Furthermore, we .wished to dis- cover by what processes these extraneous meanings were carried * . ■ 39 and where these processes stood in the general organization of mind. We introduced a number of extraneous factors. Included in the visual group were colored papers, colored borders, and location of words in different sectors. The second group intro- duced organic and kinaesthetic factors by the various postures of standing, sitting, and kneeling. In general we have obtained but little positive evidence that our extraneous factors were actually integrative in associa- tion. ' On the contrary, they have frequently appeared as distract- ing or disintegrating moments. The result suggests that the incidental meanings which supply the setting for the main subject- matter and which frequently reappear in reproduction at a high level of clearness do not serve as incentives to recall but are sheer associative luxuries. Incidentally we have come upon the problem of segrega- tion as a factor. While the earlier results gave evidence of some advantage in segregation, the later results g:.ve very little. In fact, a part of the evidence is negative. As has been shown, this negative evidence may be due to bad self-instruction on the part of the observer. The problem is beginning to show promise and invites further investigation. A number of modifications might be made both by way of refining the method and by way of introducing new ’outside' factors. In the first place we suggest a new kind of symbol, a symbol which would be less likely to be named. Japanese charac- ters or non-sense geometric-:! figures might be used. Secondly, . . * . 40 an accelerated, time, each as .7 as used in the last part of the experiment, would be an improvement. Under these conditions attention would tend to greater uniformity with less temptation to learn by logical means. Thirdly, in order to obtain accurate results, it would seem necessary to give each observer a period of thorough training. During this period careful introspections and running comments by the observer should be taken in order to discover any devices for 'learning' and variations in self-instruction. Not until all of these things have been arranged should the regular experiment be started. Fourthly, a 'recognition' method might be used to supple- ment the procedure of 'reproduction'. Thus, a partial or com- plete list of words, with symbols already attached (some correct and some incorrect), would be given the observer, who would be instructed to identify the correct combinations. This method would have the advantage of similarity to the mode of presentation. There are several new adventitious factors which might be used. We have already explored the visual and kinaesthetic fields. We might turn next to audition, using, in place of colored borders, tones of unlike pitches or intensities. In the E and C series we should then employ high, medium, and low quali- ties or strong, medium, and weak tones of the same pitch. Again, affective processes might be brought in as an extraneous factor, in connection, say, with odors. The general arrangement, however, would have to be modified to si: it the material, the series being 41 divided into two parts, for pleasantness and unpleasantness, in place of three. Furthermore, there would he, in this case, not one factor hut two, affection and olfactory sensation. Much could he added on the side of qualitative analysis hy more careful and more detailed reports. Introspections from a long period of training would add materially to the body of information. 42 BIBLIOGRAPHY Arnold, F., The Psychology of Association. ISOo. Calkins, M. W. , Association ( Psychol . Rev . Monog . Suppl , , No. 2, Feb. 1896). Ebbinghaus, H. , Psychology; an Elementary Text-book. Xtx. by M. Meyer) 1908. Meumann, E. , The Psychology of Learning. 1913. Potter, M. L. , The RSle of Adventitious Associates in Recall. 1921. Titchener, E. B. , A Beginner’s Psychology. 1316.