Cornell University 
 Library 
 
 The original of this book is in 
 the Cornell University Library. 
 
 There are no known copyright restrictions in 
 the United States on the use of the text. 
 
 http://www.archive.org/details/cu31924029039083 
 
AN EXPERIMENTAL STUDY OF SENSORY 
 SUGGESTION 
 
 A THESIS 
 
 Presented to the Faculty of the Graduate School 
 
 OF Cornell University for the degree of 
 
 DOCTOR OF PHILOSOPHY 
 
 BY 
 
 AUSTIN SOUTHWICK EDWARDS 
 
 Reprinted from American Journal of Psychology, Vol. XXVI, 
 January, 1915, pp. 99-129. 
 E.V. 
 
AN EXPERIMENTAL STUDY OF SENSORY 
 SUGGESTION 
 
 By A. S. Edwards^ 
 
 TABLE OF CONTENTS 
 
 A. Introduction 99 
 
 B. Problem of this Study : Method, Apparatus, Observers loi 
 
 C. Quantitative Results 104 
 
 D. Qualitative Analysis of Consciousness under the Influence of 
 
 Suggestion. 
 
 Exp. I. Color no 
 
 Exp. 2. Colorless Discs 113 
 
 Exp. 3. Colorless Light 113 
 
 Exp. 4. Visual Extent 117 
 
 Exp. 5. Intensity of Noise 118 
 
 Exp. 6. Tonal Pitch 119 
 
 Exp. 7. Liminal Noise 120 
 
 Exp. 8. Smell 122 
 
 Exp. 9. Taste and Smell 123 
 
 Exp. 10. Pressure 124 
 
 Exp. II. Temperature 125 
 
 E. Discussion and Conclusions 127 
 
 A. Introduction 
 
 Many definitions have been offered of the term Suggestion, 
 and many experiments have been made in whose results the 
 effects of suggestion have found quahtative or quantitative 
 expression. But no one, we believe, has attempted by means 
 of careful introspection to study the suggested consciousness 
 itself. It is matter of common experience that suggestion may 
 arouse, deflect or inhibit movement; it is equally clear that 
 suggestion may change the meaning of a perception or idea 
 by shifting its context, that is, by giving it a new associative 
 setting. But if we ask whether or to what extent suggestion 
 may bring about intrinsic changes in conscious contents, we 
 receive only partial and incidental answer. A continuously 
 rising tone, for example, is presented under the suggestion 
 that the pitch is falling. The observer reports a drop in pitch. 
 Has he, then, heard what he reports, or is he subject to an 
 illusion of judgment? 
 
 In connection with sensation, especially, do we find equiv- 
 ocal statements and a lack of experimental study. We find 
 statements that seem to indicate a direct sensory result of sug- 
 
 1 From the Psychological Laboratory of Cornell University. 
 
100 EDWARDS 
 
 gestion. Baldwin, in the Dictionary of Philosophy and Psychol- 
 ogy, mentions a ' sensory ' suggestion ; and the illustration 
 given is that of a person who ' sees ' a red light as green.'' 
 But the question remains whether, in such an instance, the 
 observer actually sees green when red is presented ; or whether 
 he ' imagines ' that he sees green ; or whether, after all, the 
 effect is motor, and he makes his reply not as the result of a 
 sensory experience, but simply as a motor discharge which was 
 predetermined by the suggestion. In experiments with sug- 
 gestion, Scott finds ' illusions ' of heat and light f Gilbert re- 
 ports ' illusions ' of weight and sight ;* Katz writes that a pre- 
 sented blue appeared dark-brown ;° Seashore maintains that 
 suggested illusions are surprisingly realistic." But in no case 
 do we find a thoroughgoing analysis and methodical control of 
 the processes actually present in the minds of the observers. 
 Kiilpe writes : " The inference is, perhaps, allowable that in 
 certain circumstances they [centrally aroused sensations] are 
 qualitatively similar to peripherally excited contents." But 
 he concludes that the existing evidence from experiment is 
 unsatisfactory; the experiments that he refers to in this con- 
 nection, he thinks. " cannot be considered as wholly free from 
 objection."' 
 
 It is true that hypnotism has been appealed to, in proof of 
 the possibility of producing sensations by suggestion. The 
 hypnotized subject appears to suffer pain, to experience feel- 
 ings, pleasant or unpleasant, to realize any sensation suggested 
 to him ; indeed, Kiilpe affirms that to suggestion " every or- 
 ganic function, motor and sensory alike, is subject . . . The 
 wildest illusions of sense and the most absurd actions are 
 evoked with equal ease."* But sensory results from hypnotized 
 subjects, even if they were fully proved, — as they seem not 
 to be, — could point only to the likelihood, not to the certainty, 
 of similar results in the normal states of consciousness. 
 
 Our appeal, then, is to the laboratory and to introspection; 
 and certain factors in the situation must especially be taken 
 
 2 J. M. Baldwin and G. F. Stout, Dictionary of Philosophy and Psy- 
 chology, ii, 1902, 619. 
 
 3 W. D. Scott, Personal Differences in Suggestibility, Psychol. Bull, 
 7, 1910, 5 ff- 
 
 * J. A. Gilbert, Mental and Physical Development of School Children, 
 Yale Stud., 2, 1894, 44. 
 
 ° O. Katz, Versuche uber den Einfluss der " Gedachtnisf arben " auf 
 die Warnehmungen des Gesichtssinnes, Zentralh. f. Physiol., 20 (16), 
 1908, SI 7. 
 
 " C. E. Seashore, Measurement of Illusions and Hallucinations in 
 Normal Life, Yale Stud., 3, 1895, 62 ff. 
 
 ■^ O. Kulpe, Outlines of Psychology, 1909, 183. 
 
 8 Ihid., 453. 
 
EXPERIMENTAL STUDY OF SENSORY SUGGESTION 101 
 
 into account. Since there is a natural disposition to give posi- 
 tive answers, it is necessary to prove that the results do not 
 express the mere ripening of motor tendencies. It seems not 
 at all unlikely that a suggestion may set up a definite tendency 
 to report in a certain way, and that the appropriate situation 
 (the critical term in a series of stimuli) is then adequate to 
 touch off the motor response. Again, it is possible that the 
 supposed sensation is in reality an image ; the modifying effect 
 of expectant attention must be taken into account. Abnormal 
 conditions, such as exhaustion, morbid emotivity, etc., must, 
 of course, be eliminated. 
 
 From this brief introduction we proceed at once to a state- 
 ment of our special problem, methods and results. A history 
 of suggestion would be blind, and a critical sifting of the out- 
 come of previous experiments on suggestion is impossible, 
 until we gain a positive insight into the nature of the suggested 
 consciousness ; these, then, are topics that we reserve for a 
 future paper. Our present results are themselves hardly more 
 than preliminary; the experimental work requires much time, 
 and is often inconclusive ; but we hope that other investigators 
 may be attracted to the field, and that data may presently be 
 available that shall justify a physiological theory. 
 
 B. Problem of this Study ; Method, Apparatus, Observers 
 
 Our general problem is that of the nature and extent of the 
 intrinsic changes produced in consciousness by the influence of 
 verbal suggestion. We are especially interested in sensory re- 
 sults, such results as are normally aroused by an adequate sen- 
 sory stimulus. The experiments are, therefore, planned to favor 
 sensory results. The questions for which we seek an answer 
 are the following: Can verbal suggestion do the work of an 
 adequate sensory stimulus? What percentage of judgments 
 can be inverted, when supraliminal charges of sensory stimuli 
 are used? How far above the limen can we get these in- 
 versions of judgment? What are the conscious processes pres- 
 ent when such inversions are made?' 
 
 The experiments fall into three groups : ( i ) the suggestion 
 group, (2) the control group, and (3) the liminal group. 
 
 In the first group, the observers were called upon to discriminate 
 intensities, qualities, etc., under the guidance of a misleading sug- 
 gestion ; for example, a tone rose- in pitch when they were told that 
 it would fall. The first trial in every series was made with a minimal 
 difference of stimuli, and the difference was increased in succeeding 
 trials. E continued to increase the stimulus-difference until he feared 
 
 » In regard to the affective processes, we merely raise the question 
 whether the effect of suggestion, is not secondary, following upon the 
 arousal of organic sensations. We have no experimental data to offer. 
 
102 EDWARDS 
 
 that the observers would detect the deception. Suspicions on the part 
 of the observers were unavoidable; they had to be assured again and 
 again that E was giving the stimuli in accordance with his original 
 statement. Whenever there appeared results of suggestion strikingly 
 like the results ordinarily obtained from adequate sensory stimuli, 
 introspections were taken, and the experience was compared with 
 the normal: details will be given later. In the second group, the 
 observers made the same discriminations under the regular method- 
 ical instructions. They now made them, however, with full knowl- 
 edge of the suggestions employed in the first group, and with full 
 assurance that no misleading suggestion would henceforth be used. 
 Further, they were required to give full introspections of the fore 
 and midperiods, with constant reference to the conscious effect of 
 suggestion in the parallel experiments of the earlier group. The 
 third group consisted of determinations of difference limens, by the 
 method of limits, taken in order that the results of suggestion in 
 the first or suggestion group might be subdivided, the trials in which 
 the stimulus-differences were subliminal being set apart from those 
 which gave inversions of judgment when the stimulus-differences 
 were supraliminal. It is this latter set, in which we have what may 
 be called supraliminal inversions of judgment, that interests us in 
 the present paper. E took as the difference-limen the first difference 
 of stimuli, correctly judged, which had two correct judgments beyond 
 it. This is a slightly larger limen than is obtained by other forms 
 of procedure; but the larger liraen was deemed desirable, since com- 
 parison with the results of the suggestion group demanded a margin 
 of safety. In order to make the results of the suggestion and of 
 the liminal groups comparable, the steps in the corresponding series 
 were kept the same. The liminal series served also to assure E of 
 the normal acuity of the observers for the sense departments investi- 
 gated. 
 
 Experiments 7, 8, and 9 were complete without the control group. 
 In the remaining experiments, the control followed the suggestion 
 group in inverse order of series. If this arrangement made the burden 
 of memory greater for some, it greatly relieved it for other experi- 
 ments; and we considered it as, on the whole, the best plan to adopt. 
 The control for experiment 3 came immediately after the parallel 
 suggestion experiment; for 11, there was an interval of a week; 
 for 2 and 3, four and five weeks (3 months for one observer) ; for 
 10, two months; for i, three months; and for 6, one, four and five 
 months for the various observers. 
 
 Experiments were made — not all, as we shall see, with the same 
 thoroughness or success — upon sight (quality, intensity, extent), hear- 
 ing (noise intensity, tonal quality and intensity), taste (quality), 
 smell (quality, intensity and duration), pressure (intensity), and 
 temperature (intensity of warmth and cold). There were four ex- 
 periments upon sight; three upon hearing; one upon smell; one 
 upon taste and smell; and one each upon pressure and temperature. 
 Most of the experiments were planned for inversions of judgment; in 
 two cases, however, — one of the audition experiments and the smell 
 experiment, — the attempt was made to get, by means of verbal sug- 
 gestion, the process ordinarily aroused by adequate stimulus, and to 
 get it in the absence of any adequate sensory stimulation whatever. 
 In three cases, owing to the nature of the experiment, to lack of time, 
 or to difficulty with apparatus, we. failed to obtain liminal determina- 
 tions. 
 
EXPERIMENTAL STUDY OF SENSORY SUGGESTION 103 
 
 For vision, colored, black and white discs were mounted 
 on the Marbe mixer i^" recourse was also had to Whipple's 
 apparatus for the discrimination of brightness. ^^ For visual 
 extent, we used Wundt's space-limen apparatus. ^^ The Stern 
 variator, Fechner pendulum and modified Lechmann acoume- 
 ter^* were employed for sound experiments. Two paper fun- 
 nels were used in the smell experiments ; small candies for 
 taste and smell ; the Wundt pressure apparatus^* for pressure, 
 and brass cylinders for temperature. Continuous changes were 
 possible with the Marbe mixer, the Whipple box, and the 
 Stern variator. 
 
 In experiments of the first and second groups (with exceptions 
 noted later), the observers were told, either that a stimulus would be 
 given and that shortly afterwards a second would follow, or that a 
 continuously changing stimulus would be given; they were to judge 
 the direction of the change and, when asked, to give full introspec- 
 tions; the suggestion as to direction of change was made just before 
 each trial. The suggestion always ran counter to the facts; so that 
 if E said the change would be toward brighter, he changed the 
 stimulus toward the darker. The observers were informed that the 
 first changes would be small, and that the series would be con- 
 tinued until they could clearly perceive the change suggested. In the 
 experiment on smell, the observers were told that they would be 
 given an odor and that, as soon as they indicated their perception 
 of it, a second odor would be given which should neutralize the first. 
 No odors were used, and no experiment was begun until the observer 
 reported the room free from all odors. In the experiment on taste 
 and smell, the candies were given in pairs, the one to be eaten shortly 
 after the other; the two differed in shape, and one was so wrapped 
 as to appear of better grade than the other. The ingredients were 
 the same in both cases. The suggestion lay in the appearance of the 
 candies and in the verbal instructions ; the observers were to write 
 full introspective comparisons of the sensations of taste and smell 
 aroused by the two kinds. In one of the audition experiments, a 
 lead and a cotton ball were used, the sight of the falling cotton ball 
 being the suggestion for the sound. The object of these experiments 
 was, then, either to arouse through suggestion the process commonly 
 aroused by an adequate stimulus, or by means of suggestion to invert 
 judgment. The omission of several steps at the end of a series 
 permitted the experimenter to obtain, in a few instances, very large 
 supraliminal results, and at the same time to avoid the danger of 
 detection, which was always present, but was especially to be feared 
 in the longer series. 
 
 The experiments were carried on with five men in the Cor- 
 nell laboratory during the greater part of twoi years, and later 
 
 10 K. Marbe, Physiol. Centralb., 1894, No. 25, 811. 
 "G. M. Whipple, A Manual of Mental and Physical Tests, 1910, 
 163. 
 
 12 -y^r Wundt, Physiol. Psychologic, iii, 1911, 466. 
 
 13 E. B. Titchener, Exp. Psych., II., i, 190S, 24. 
 1* W. Wundt, op. cit., i, 1908, 674. 
 
104 EDWARDS 
 
 with five women at the University of Minnesota during one 
 year. The men will be referred to as A, B, C, D and E; the 
 women as F, G, H, I and /. Observers A, B and C had all 
 had thorough training in introspection, A and B in the Cornell 
 laboratory and C in the Harvard laboratory. One was an in- 
 structor in the department, one an assistant, and the third an 
 advanced student engaged in research. D and E were respec- 
 tively a fellow and a scholar in psychology. F, G, H, I and / 
 were advanced students in psychology, one being an assistant 
 in the department with some training in introspection; all 
 needed further training. 
 
 A number of trials were made with other than the regular observ- 
 ers. These did not call for introspective reports; they were under- 
 taken merely for comparative purposes. They showed that results 
 similar to those obtained from the regular observers could with little 
 difficulty be obtained from persons unconnected with the psycholog- 
 ical laboratory and unversed in psychology. Indeed, in some instances, 
 even more extreme results of suggestion appeared with these un- 
 trained observers. 
 
 In reporting our experiments, we shall first give the quantitative 
 results which show the effects of suggestion in relation to the dif- 
 ference-limen. We shall then turn to an analytical and qualitative 
 study of the conscious processes under the influence of suggestion. 
 
 C. Quantitative Results 
 
 In calling these results ' quantitative,' we do not mean that 
 we have measured the degree of suggestion, or that we have 
 discovered the greatest amount of supraliminal change of stim- 
 ulus that can be introduced while the judgment of the observer 
 under the influence of suggestion is still inverted. We are only 
 reporting, in numerical terms, the results of particular series 
 of experiments made under particular conditions. These con- 
 ditions were, in fact, distinctly unfavorable to suggestion. The 
 instructions demanded a critical attitude; the observers were 
 asked to make judgments in series which, so far as they knew, 
 were part of a study of liminal differences, and which de- 
 manded their most careful attention. Seven of the ten ob- 
 servers reported suspicions as to the nature of the experiment, 
 and all showed, at times, by action and tone of voice, that they 
 thought the investigation was not what it purported to be. 
 Finally, all such factors as fatigue, excitement, and the like, 
 which are favorable to suggestion, were absent. 
 
 The difficulties of the experiments, the suspicions of the 
 observers, and lack of time, made it possible for E to use 
 supraliminal charges of stimuli only in 900 of the 1621 trials 
 comprised in the first or suggestion group. In 253 of these 
 
EXPERIMENTAL STUDY OF SENSORY SUGGESTION 105 
 
 900 trials, the suggestion succeeded in inverting judgment. 
 In other words, suggestion was successful in 28 per cent, of 
 all cases in which supraliminal changes of stimuli were used. 
 This number does not include the judgments ' doubtful ' and 
 ' equal.' Of the 254 series in the group suggestion was suc- 
 cessful in no, or in 43 per cent. The following Tables i, 2 
 and 3, show ( i ) the percentage of successes of suggestion with 
 supraliminal changes of stimuli for the various series in the 
 suggestion group, and (2) for the individual experiments and 
 observers, and (3) the number of steps above the difference- 
 limen where suggestion was successful, with the number of the 
 successful cases for every step. To repeat and emphasize, all 
 the results noted in these Tables are results obtained with 
 supraliminal differences of stimuli. 
 
 Table i 
 Number and percentage of series from the suggestion group con- 
 taining inversions of judgment. 
 
 Part 1 Part 2 
 
 According to Experiments According to Observers 
 
 Expt. Suggestion No. % Observer Series % 
 
 1 Blue 17 71 A 10 25 
 
 3 Brighter 21 72 B 15 34 
 
 Darker 20 74 
 
 C 35 42 
 
 4 Larger 
 
 Smaller 2 20 D 3 100 
 
 5 Louder 4 80 E 4 100 
 
 Fainter 4 80 
 
 F 14 86 
 
 6 Higher 9 45 
 
 Lower G 6 35 
 
 7 Falling shot... 14 16 H 3 50 
 
 10 Heavier 7 100 I 10 52 
 
 Lighter 
 
 J 10 48 
 
 11 Warmer 3 36 
 
 Total series containing inversions = 110 or 43%. 
 
 In Part I, the first column indicates the number of the experi- 
 ment; the words Blue, Brighter, etc., the suggestion used. The num- 
 bers following show first the number (no.) and secondly the per- 
 centage (%) of series in which inversions of judgment followed the 
 use of suggestion. E. G., " 17 71 " means that in Experiment I 
 inverted judgments were found in 17 series, or in 71% of all series 
 in the suggestion group. 
 
 Part 2 indicates the number and percentage of series containing 
 inversions of judgment for each observer. 
 
106 
 
 EDWARDS 
 
 
 ■g 0«D 
 
 ^ S o 
 
 coco 
 
 a 
 
 o 
 
 CSl 
 
 #> 
 
 § 
 
 
 
 
 1 
 
 ?5 
 
 0010 
 
 
 i 
 
 
 uno 
 
 o°" 
 
 3h 
 
 a 
 
 CD 00 
 
 o-^ 
 
 e 
 
 
 
 g 
 
 
 ffl 
 
 00 
 
 pa 
 
 z 
 
 50 
 
 co^~ 
 
 00 
 
 
 0? 
 
 < 
 
 3° 
 
 < 
 
 
 £2- 
 
 ■>*o 
 
 0-* 
 
 
 3 0) 
 
 11 
 
 il 
 
 
 0., 1 
 
 CO 
 
 ■* 
 
 w 
 
 
 
 
 
 
 
 
 
 he sug- 
 ersions. 
 rors. 
 ere not 
 experi- 
 
 sg 
 
 ^^ 
 
 2 
 
 fe 
 
 ^g 
 
 
 
 
 
 
 
 
 
 
 fer to t 
 nal inv 
 sible er: 
 mens w 
 in this 
 
 523 
 
 COC75 
 
 3 
 
 00 
 
 i-H 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 aJ"-< OT.Sh— » 
 
 
 
 
 
 
 ►-S?3 
 
 
 , etc., r 
 
 upralim 
 
 the po 
 
 since 1 
 
 I and , 
 
 
 
 
 
 
 _.HO 
 
 
 
 
 
 
 
 
 '-'■*«3 
 
 
 
 
 
 
 
 
 
 to 0^0,0 
 
 
 
 
 
 
 ffiSiS 
 
 
 Blue, Bri 
 rcentage 
 I were IS 
 C in Ex 
 n for F, 
 
 
 
 
 
 
 
 (jt>oo 
 
 
 words 
 ind pe 
 , whicli 
 B and 
 :st lime 
 
 
 
 ffi 
 
 
 
 
 
 
 
 
 
 
 
 f^SS 
 
 
 The 
 ber i 
 rrors 
 r A, 
 highe 
 
 
 
 
 
 
 
 
 
 
 
 
 
 )-i 
 
 E ■" „, 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 i 
 
 nts I, 3. 4. etc. 
 ; show the nu 
 ; these mean 3 
 iclude results i 
 higher than thi 
 
 
 
 ,.8 
 
 
 
 OS?5 
 
 1 
 
 
 
 pj 
 
 
 
 mg52 
 
 
 
 experime 
 
 of the tabl( 
 
 mbers 3 IS 
 
 do not in 
 
 C are not 
 
 
 u^^ 
 
 CO 
 
 0" 
 
 8 
 
 S8 
 -• 
 
 
 s 
 
 010 
 
 ■*Q0 
 
 in 00 
 
 2 
 
 
 
 
 <S 
 
 S 
 
 i 
 
 n refer t 
 he body ^ 
 d the nui 
 rcentages 
 A, B and 
 correct. 
 
 PQ 
 
 pa 
 
 pa 
 
 P 
 
 pa 
 
 
 
 s::§s.k£ 
 
 COC<l 
 
 O-H 
 
 CO 
 
 rj" 
 
 .-ICO 
 
 
 
 
 3 .S s 
 
 •0 MH UMH " 
 
 int- 
 
 00 
 
 ;o 
 
 
 !S12 
 
 
 
 
 in the first a 
 The' figures 
 under A are 
 s and averag 
 If the limens 
 
 cketed figures 
 
 rHi-l 
 
 .-HO 
 
 < 
 c^ 
 
 
 tot* 
 < 
 c^co 
 
 
 
 
 II 
 
 SI 
 
 +j 
 
 •i 
 1 
 
 la 
 
 5 " 
 
 > 
 
 i 
 
 Numbers 
 gestiohs used. 
 For example. 
 
 The total 
 determined, 
 ment. the bra 
 
 m 
 
 to 
 
 c~ 
 
 
 
 
 
 5 J 
 
 3 
 
 H 
 
EXPERIMENTAL STUDY OF SENSORY SUGGESTION 107 
 
 
 
 
 
 .—1 
 
 
 
 
 
 
 
 
 
 
 
 O 
 
 fi 
 
 
 
 
 
 
 
 
 
 
 
 ■* 
 
 
 
 
 
 
 
 
 
 
 
 
 00 
 
 
 
 
 
 
 
 
 
 
 
 
 CO 
 
 
 
 
 
 
 
 
 
 
 
 
 CD 
 
 T-i 
 
 
 
 
 
 
 
 
 
 
 
 CO 
 
 
 
 
 
 
 
 
 
 
 
 
 -* 
 
 i-Ht-H 
 
 *-(i— 1 
 
 
 
 
 
 
 
 
 
 
 CO 
 
 
 
 
 
 
 
 
 
 
 
 
 cq 
 
 T-4fH 
 
 »H 
 
 
 
 
 
 
 
 
 
 
 CO 
 
 
 
 
 
 
 
 
 
 
 
 
 o 
 
 rHCSl 
 
 C<1 
 
 
 
 
 
 
 
 
 
 
 CO 
 
 
 
 
 
 
 
 
 
 
 a, 
 
 
 83 
 
 C^i-H 
 
 1-f 
 
 
 
 
 
 
 
 
 
 m 
 
 CO 
 
 
 pa 
 
 CVlrt 
 
 .-ICO 
 
 CO ra 
 "S 
 
 iS 00 
 
 CVI 
 
 C^ 
 
 .-H CO 1-1 
 
 i 
 
 Q 
 1 
 
 18 20 22 24 
 
 1 
 
 CMCOrH 
 
 CM-* 
 
 HH 
 
 f^ 
 
 
 00 ^^ca 
 
 •-ICO 
 
 l^to 
 
 CO 
 
 OOrt,-! 
 
 to 
 
 CM CM pa 
 
 >-H 
 
 o 
 
 
 tH 
 
 
 
 t— I 
 
 
 
 
 r-i 
 
 
 t-^ 
 
 ■< 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 i 
 
 o 
 
 1 
 
 to .-1 c^q cq .-H .-H i-i CXI 
 
 ^ >-ICOC<l COCSli-l 
 
 S C*a T-H .-H rH CX] 
 
 -MCSIrt 
 CM <-l 
 
 1 
 
 ^ CMrHNOO 
 
 c^j rt .H cjco 
 
 rH 
 
 
 
 OJ <-ICOCSl 
 
 caeo 
 
 Cfl O .-1 .-1 i-H ■* rt C<I CO 
 
 tn 
 
 O CMrt 
 
 PCrHrH 
 
 
 
 
 r-^ 
 
 
 
 1— I 
 
 
 
 
 rH 
 
 
 
 z 
 
 i-H 
 
 
 
 
 
 
 
 
 
 
 
 -U 
 
 1— ( 
 
 e>] y-l 
 
 CO* 00 
 
 THrHCO 
 
 pqcsji-i 
 
 CO 
 
 OOrHrHCOrH 
 
 pq .-1 
 
 
 
 1 
 
 00 rtrHPa 
 
 N 
 
 1 <o 
 
 c^cq 
 
 
 tOi-l *H 
 
 
 
 
 p. 
 
 
 
 
 •c 
 
 
 
 ■§ 
 
 
 
 
 M 
 
 (^ 
 
 to 
 
 C<1 
 
 
 
 rq.-i 
 
 
 & 
 
 Tf iHCM 
 
 
 
 2 
 
 
 'i- 
 
 
 
 
 
 w 
 
 
 
 
 
 
 i 
 
 
 
 05 
 
 
 
 
 
 ^ ■ • ■ ' • 
 
 
 
 
 
 g) ; 
 
 
 
 
 ^ 
 
 
 
 
 QJ 
 
 
 
 
 
 
 
 in c^ 00 --H N 
 
 Q 
 
 
 
 
 Q : : : : : 
 
 
 
 -i lo d csi ■*■ 
 
 ;r-HCJ; 
 
 I> p CM 00 ^' 
 
 •t-^CM" 
 
 
 
 
 
 
 ^i-H 
 
 1-H.-H.-H 
 
 C0.-Hf-H»HO 
 
 '. rH T— ) 
 
 
 l-^rH rH rH r-i 
 
 ir^CM 
 
 
 
 
 
 
 > > 
 
 > > > 
 
 
 > >> > > 
 
 • > > 
 
 
 > > > > > 
 
 : > > 
 
 
 
 
 
 
 assas 
 
 
 3 E S S d 
 
 ;s6 
 
 
 ssasB 
 
 [BB 
 
 
 
 
 
 
 '-H'oi'o'ino' 
 
 oq-t-T-^'to-oo" 
 
 :[>.-T 
 
 
 !> CO in 00 of 
 
 ItOtD 
 
 
 
 
 
 
 Lrit>t>oi> 
 
 Cjicocji>LO 
 
 • t-^td 
 
 
 couicMtoc-: 
 
 ■00 to 
 
 
 
 
 
 
 
 -^ 
 
 JhJiJhJJhJhJhJ 
 
 
 JJJJJJJJ 
 
 
 
 
 s^ri 
 
 JJJJJJ 1 
 
 QOQOQQQQ 
 
 
 QQQQQQQQ 
 
 
 
 
 QQ 
 
 Q 
 
 QQPPQ 
 
 jjjjhjUjj 
 
 
 ^iD^^PlD^t^ 
 
 ■<fflOfi.Offii-.» 
 
 fljmoteOKi— 
 
 <t:CQObOffii-i- 
 
108 
 
 EDWARDS 
 
 § goo 
 
 '-' tain 
 
 % 
 
 C3 
 
 ►?'■ 
 
 T3 
 
 I 
 
 1 
 
 hi 
 
 1 
 
 I 
 
 g(0 
 
 
 odd 
 > > > 
 
 cfd'cf 
 
 NPJCO 
 
 PJhJhJ 
 
 PQQ 
 
 ■*COrH 
 
 
 
 1 
 
 
 
 rH goo rH 
 
 oiinoj 
 dd'-i 
 
 > > > 
 
 JlJI-1 
 
 QPQ 
 <fflU 
 
 StO'-l'-HC 
 
 I 
 
 
 gsO'-' 
 
 
 ■2" 
 
 
 M 
 
 i 
 
 :s 
 
 ^ >>> 
 ess 
 
 S22 
 
 <!CQO 
 
 ,£rHi-HrH 
 
 ESS 
 
 OJNCO 
 
 ffi 
 
 I 
 
 Is 
 
 baO) 
 bo 
 
 I 
 
 S<o 
 
 •g 
 
 <U 
 
 I 
 
 So rH 
 N 00 N 
 
 M 
 
 § 
 > 
 
 § 
 I 
 
 t-;COt-j 
 
 > > > 
 
 SEE 
 
 hJJhJ 
 QQO 
 
 <!CQU 
 
 f-4d*^ 
 >>> 
 
 SEE 
 cacoin 
 
 QQQ 
 
 <UPQO 
 
EXPERIMENTAL STUDY OF SENSORY SUGGESTION 109 
 
 
 a 
 
 is 
 
 w. 
 
 o 
 
 1 
 
 I 
 
 a 
 
 o 
 
 .-HO C 
 
 
 O 
 
 •*05 
 CQrH 
 
 > > 
 
 r-iO 
 *-tf-( 
 
 hJnJ 
 
 ss 
 
 66 
 
 00 ■* 
 ■^00 
 
 QW 
 
 
 
 •r 
 
 t>coco 
 odd 
 
 > > > 
 
 666 
 
 cvidd 
 
 QQQ 
 
 <:CQO 
 
 I 
 
 8 
 
 
 s 
 
 g.i-irt CO 
 
 (3 
 
 fc, : : 
 
 0) . . 
 
 Q : 
 
 mooco 
 
 > > > 
 666 
 d(M"t> 
 Noid 
 h-lh-li-) 
 QQQ 
 
 <PQO 
 
 S g " ' 
 o63i 
 
 ^ s <u V 
 
 J) g <u cS •« 
 +5^ > n " 
 
 - . is 6 6 
 
 Jj ^ 0) o fa 
 
 -gto o.S 
 
 > 0«iH o 
 
 * So 
 
 ■1-' s 
 
110 EDWARDS 
 
 D. Qualitative Analysis of Consciousness under the 
 Influence of Suggestion 
 
 Experiment I. Color 
 The Marbe Mixer was used for the introduction of yellow by 
 continuous change into red. The change in the first exposure of the 
 series was from o° to 2°, in the second from 0° to 4°, and so on, 
 the increase in yellow proceeding by steps of 2°. Each series was 
 as long as E could make it without running an obvious risk of 
 detection. A large gray cardboard served as a screen. The observers 
 seated themselves at such a distance from the discs as made observa- 
 tion easiest for them; the distance was about 3m. The experiments 
 were made with only one observer at a time, for periods of 50 
 minutes. Exposures were of 6 sees, duration, increased slightly for 
 the trials where the greatest amounts of yellow were introduced, 
 in order that the rate of introduction might be roughly the same. In- 
 structions were as follows : " 1 am going to run a blue into a red on 
 the color mixer. The change will be continuous. I want you to tell 
 me whatever changes you notice in the color. The first changes will 
 be very small, the changes m succeeding exposures increasing until 
 you are sure of the change." The suggestion was generally, but not 
 always, repeated just before each exposure: "Towards the blue," 
 or " Still towards the blue." 
 
 Suggestion Group 
 
 Observer A reports as follows with the following degrees of 
 yellow present : 10°, " Barely possible it is darker ;" 20°, " Perhaps 
 slightly darker;" 22°, "Not at all certain whether the disc is a little 
 bluish or not now. Quite certain of change but not sure of direction 
 of change. Is it toward blue ?" ; 30°, " I suspect all of a sudden 
 that maybe there is blue in it and I've missed it before ;'' 36°, " Slight 
 edge of color of rather greenish gray." A reported the following 
 " prepossession " : " E is working on suggestion, may give me no color 
 at all, may give me wrong colors, may give me no changes, or wrong 
 changes. May make movements intended to deceive me, also make 
 time intervals wrong for deceit ;" " Tendency to be careful ;" tendency 
 to make statements with " qualification, ' perhaps.' " 
 
 Observer B reports gray film, and says that it " may be due to 
 the running in of gray." In the second of the two series B says, 
 with 8° of yellow, " I guess there is a little blue in it ;" and calls 
 6, 7, and 8 (12°, 14°, and 16°) the same as 4 (8° of yellow). 
 
 Among the reports of blue by observer C occur the following: 8° 
 of yellow, " Blue ring perceived and decided it was no illusion ;" 12°, 
 "Blue ring definitely present;" 28°, "Blue ring definitely present;" 
 32°, " Definite blue ring not so vivid as 14 [28° of yellow] but more 
 
EXPERIMENTAL STUDY OF SENSORY SUGGESTION 111 
 
 than IS [30°] ;" 32°, "Blue ring as faint as 15:" From another series: 
 12° of yellow, "Obviously bluer;" 22°, "Much bluer than i;" 28°, 
 " Color has blue in it." After this series C was asked to indicate 
 with discs how much blue he thought had been used, and he set them, 
 with 35° of blue exposed. 
 
 Observer F gave the following reports : with 18° of yellow, " Slightly 
 more blue than 5 and 6;" 10°, "Noticeably more blue;" 12°, "Same 
 as 5; perception of blue was less marked in appearance but lasted 
 for a longer time ;" 28°, " Perception of red changed distinctly towards 
 blue; . . . surprised to see the blue so distinctly. . . . The 
 blue present was very definite ;" 34°, " Blue was very distinct, so 
 much so that the color was quite blue ;" 36°, " Blue lasted longer 
 than in 15 ;" 42°, " The red seemed to be slightly darker, with a 
 gray more than a blue." 
 
 Observer G was frankly suspicious after a few trials; her brother, 
 she said, had ' fooled ' her so often that she was generally on her 
 guard. Some results were obtained : 8° of yellow, " Doubtful ; first 
 judgment ' darker,' second judgment ' lighter ' ;" 14°, " Two judg- 
 ments; one that the hue grew darker; second judgment 'no change ';" 
 16°, " Changes from darker to lighter and darker again." G reported 
 ' no change ' or ' doubtful ' for changes of 8°, 10°, 12°, 16°, 22°, 24°, 
 26°, 28° and 30°. At 28° she said: "Might have been a change 
 towards blue." 
 
 Observer H _ reported the following : 10° of yellow, " Very faint 
 blotches of slightly darker hue ;" 10°, " Darker hue seemed to be 
 in narrow bands;" 10°, "Slightly darker;" 10°, "Darker hued;" 18°, 
 "A bluish-pink tinge in a filmy form, seeming to stand out from the 
 background." Many reports of " No change " were made with the 
 yellow shown at 10° to 40°. 
 
 In a few cases, E reversed the direction of stimulus-change at the 
 end of a series, in order to avert a suspicion which had become evi- 
 dent in the observer's voice, manner or replies. In one series with H 
 the results show that within a few trials a large increase of yellow 
 and a large decrease of yellow can both be judged as changes in 
 the same direction. In another series H reported " No change " 
 when the stimulus-change was an increase of 34° and 40° yellow 
 (14 and 17 steps beyond the DL). 
 
 Observer / frequently reported blue : 14° of yellow, " Medium blue 
 discs . . blue shadows ;" 16°, "Single grayish blue disc rises 
 
 over red f rpm below and remains, not quite covering it ;" 20°, " Red 
 seems to blend with blue-gray;" 14°, "Bluish shadow appeared about 
 a second after disc was exposed, . . . moved back and forth ;" 18°, 
 "Distinctly a violet-blue film over surface;" 20°, "Change to a 
 purplish red with a great deal of gray in it, and this darker red, not 
 a purple but a purplish red, . . . seemed to fuse this time except 
 at edges and right in center of disc ;" 22°, " Yes, grayish blue, . 
 a little denser and a little darker and more mixed with the red than 
 it had been before;" 22°, "Change to blue. It was unquestionably 
 blue. . . . The blue was like a veil over the red, but it was un- 
 questionably blue. It did not fuse with the red." Other reports of 
 like nature were made. 
 
 Observer / reported the following : 10° of yellow, " Part . . . 
 
 seemed a little darker . . . ;" 28°, " Columns of shadow 
 
 . . . ;" 22°, "Columns of darkness;" 14°, "the orange-red 
 
 changed to a brownish-red ;" 30°, " Perception of a bluish-lavender, 
 
 . perception of a more brownish-red background; the bluish- 
 
112 EDWARDS 
 
 lavender gradually changed to a more distinctly dark blue perception. 
 The dark blue was not very definite, although I was certain that it 
 was blue and no other color;" 32°, "A bluish-lavender appeared, 
 . . . I am certain that the perception was that of bluish-lavender ;" 
 36°, "No change." 
 
 Further results of the same kind might be quoted from the reports 
 of all observers. Needless to say, pauses were often made in the 
 longer series, to avoid fatigue and eye-strain; the observers were con- 
 tinually cautioned beforehand not to make an observation unless 
 entirely free from after-images; and the experimenter frequently 
 enquired, during an experiment, about the presence of after-images. 
 
 Control Group 
 
 Introspections were secured from observers F, G, H, I and /. 
 We can give only brief summaries. 
 
 Observer F rarely reports imaginal processes; judgment comes " fol- 
 lowing directly the perception of the blue," " with the perception of 
 blue as the clearest thing in consciousness." Once she reports a 
 " readiness to say blue." There is apparently no change of attitude 
 from the one group to the other; and assurance is no greater in the 
 control than, in the suggestion series. 
 
 Observer G expresses her judgments in "Yes," or "Positive," but 
 a good deal of doubt is still reported. There is sometimes much 
 imagery, expectant and memorial, and sometimes again this imagery 
 precedes judgment; but ordinarily the judgment is reported as direct. 
 There is change of attitude, but no marked increase of assurance. 
 
 Observer H has a good deal of imagery, and sometimes reports 
 verbal processes, though both alike are usually " very dim," '" merely 
 an undercurrent of little importance." In recalling earlier series H 
 says : " The changes to blue ... in the suggestion series were 
 clear perceptions, though not of so great change. . . . They 
 were perceptions from the fact that what I saw was different from 
 what I expected to see, i. e., a dark sector on red." 
 
 The shift from a suggestion series to a control series was made 
 with observer / during one of the hour-periods. It was not, how- 
 ever, until 18° of blue were introduced that she said " Blue was 
 not so blue in the other series." Her introspections show two kinds 
 of judgment: in the one, there were many imaginal processes, in 
 the other no imaginal processes at all were reported, and the judg- 
 ment was touched off directly by the stimulus. Her assurance is the 
 same in both groups. _ She used, in the first group, such words as 
 " unquestionably," " distinctly," " clearly," in referring to changes 
 toward blue, and showed no trace of doubt in tone of voice or in 
 word. "As to the earlier group there was no question of the clear- 
 ness of change and my sureness of the change. It was distinct, 
 and a change, so far as I can tell, actually of sensation . . . ; 
 the changes had all the characteristics of sensory experiences, and, 
 at the time, I should not have thought of calling them anything else 
 than sensory." 
 
 J's judgments occasionally come with a '"yes," and kinaesthetic- 
 verbal elements are characteristic of her rejwrts. In comparing the 
 two groups she says : " No more clear or distinct than in the cases 
 where I was sure of _ the blue ;" " Experiences exactly like those in 
 other experiments; things seem to go just the same." 
 
EXPERIMENTAL STUDY OF SENSORY SUGGESTION 113 
 
 Summary of Results from Experiment i 
 
 (1) The records of the suggestion group show results that are 
 very similar to (in some cases indistinguishable from) results re- 
 ported in the control group. There are a few of these for A, B and 
 G; many for C, F, H, I and J. Even where the observers think 
 that they have found specific differences in the experiences of the 
 two groups (as happened in certain instances with / and 7), refer- 
 ence to the record shows that exactly the same things are reported 
 in both sets of experiments. 
 
 (2) The assurance of F, H, I and 7 is the same for both groups; 
 that of G is not markedly different. _ 
 
 (3) The introspections show (a) judgments with a good deal of 
 imagery; (6) judgments which seem to_ indicate the touching-off 
 of a tendency to agree with the suggestion, to say " Yes " or the 
 like; and (c) direct judgments without imagery and with no apparent 
 intervention of the motor impulse to accept the suggestion; here the 
 observation itself appears to be modified. Judgments of the third 
 or sensory type were largely predominant in this experiment with 
 observers F and H, and were in the majority with G and /. 
 
 (4) The reversal of direction of large stimulus-changes, for two 
 observers, is in the one case barely noticed and in the other, is 
 accepted as a continuation of the series. Definitely supraliminal 
 changes in opposite directions are, by both observers, judged as changes 
 in the same direction. 
 
 Experiment z. Colorless Discs 
 This experiment, made with black and white discs on the Marbe 
 mixer, seemed fatally to arouse (or confirm) the suspicions of the 
 observers. Hence E took only one series with A and five with C. 
 In the series with A, the suggestion was " Towards brighter." With 
 6°, 8°, and 10° decrease of white A reported no change; with 12° 
 he said : " Might have increased that time ; can't be sure at all." The 
 obvious suspicion of A made it necessary to discontinue the experi- 
 ment. Observer C gave the judgment " equal," under the same sug- 
 gestion, with decreases of 8° and 10°. 
 
 Experiment 3. Colorless Light 
 The Whipple box, set up in the dark room, was arranged to show 
 light only in one window. The bar connected with the reflector was 
 lengthened, so that small changes of stimulus were possible. Steps 
 of 2° on either side of the neutral position of the reflector were used. 
 The instructions were : " I shall give you a continuously changing 
 light, and I want you to report as soon as you notice the change in 
 brightness. Please indicate also the direction of the change. The 
 change in the first trial will be very small, and will be increased 
 in each succeeding trial until you notice the change." The suggestion 
 
114 EDWARDS 
 
 was repeated at the beginning of every series, and generally before 
 every trial: "Towards the brighter," or "Towards the darker." 
 Fifteen minutes were allowed for dark adaptation, and care was taken 
 throughout the experiments to avoid after-images. The exposure was 
 of 6 sees, duration, and was slightly increased for the larger changes. 
 
 Suggestion Group 
 
 Observer A, with the suggestion, "Towards brighter" and the 
 stimulus-change actually from o° to 12° darker, reports: "Believe 
 it did get lighter right near end." Again, with 12° change towards 
 darker: "A wee mite lighter and then a wee mite darker, very small 
 but very definite." With 6° and 8° darker A reported " lighter, doubt- 
 ful ;" and with 10° darker, " Wee mite lighter." It is to be noted 
 that, just before this series, A had asked E to give a few practice 
 trials in each direction, and that E had granted the request. 
 
 B's reports frequently raise the question whether the change is 
 as suggested. With 8°, 10° and 12° change in both directions, "no 
 change " and " doubtful " are reported. B was evidently suspicious ; 
 and in the last series, where the suggestion was " Towards darker," 
 E reversed his direction. The change was from 8° in the one direction 
 to 12° in the opposite direction, a change of 20° ; it was, however, 
 accepted simply as a continuation of the series. 
 
 Observer C gives many " doubtful " and " no change " judgments, 
 even where the stimulus-changes are clearly supraliminal. But he 
 also gives inversions of judgment with full assurance. With the 
 stimuli 8° and 10° lighter he reports " darker " emphatically. Again, 
 with the stimulus 12° darker he says : " Brighter, sure." Once, when 
 the stimulus-change was 12° darker, he reports : " Three periodic 
 changes constituting increase of brightness, each followed by a dark- 
 ening. That actually occurred subjectively." 
 
 Observer F, with suggestion "Towards brighter" and stimulus- 
 change 20° darker, says : " Got brighter, distinctly brighter, then 
 darker. . . . Change from lighter to darker was very slight." In 
 another series, with suggestion " Towards brighter " and stimuli 10°, 
 12°, 14°, 16°, 18° and 20° darker, all judgments are " brighter." With 
 18°, e. g., F says : " There was a change from darker to brighter, quite 
 distinct, short duration ;" with 20° : " Much brighter than last, grew 
 brighter for a time, and then a dark streak through center." Just 
 before the second series of this experiment P asked : " Isn't there 
 a good deal of suggestion here? " and added, " My natural tendency 
 is to do the opposite." Immediately afterwards, with the suggestion 
 " Towards darker " and actual changes of 8°, 12° and 14° towards 
 the lighter, F reported " darker," and said in every case that she 
 was sure of the change. 
 
 Observer G was still on her guard. When the judgments were in- 
 verted, she was generally doubtful. With a stimulus-change 22° 
 brighter, she reports: "That seemed darker; quite surprised at judg- 
 ment; not very sure of it; seemed to get lighter at end; then more 
 sure of judgment after I looked at it again; seemed an evident 
 change." In another series, with a stimulus-change 18° brighter, she 
 says : " Seemed as though getting a little darker ; not sure ; two or 
 three judgments; first that it was darker; then that it was not darker; 
 then that it was darker. Think it was imagination more than any- 
 thing else." 
 
EXPERIMENTAL STUDY OF SENSORY SUGGESTION 115 
 
 Observer H gets definite results with full assurance. With sug- 
 gestion "Towards darker" and stimulus-changes of io°, 12°, 14°, 
 16°, 18°, 20° and 22° towards lighter, all in one series, H reports that 
 the changes are towards darker. In another series, " Brighter " is 
 reported when the stimulus-changes are 6°, 8°, 10°, 12°, 14°, 16° 
 (18°, "no change"), 20°, 22°, 24° and 26° darker. 
 
 The reports of / also show definite results. Characteristic is the 
 following : 18° darker, " Change half way through ; seemed to light 
 up and a difference in depth; seemed to have third dimension, depth, 
 as if I could put my finger into it; then stayed at this brighter 
 intensity the rest of the time." Once when the stimulus-change is 
 18° brighter / reports changes in both directions, but says she is 
 surer of the change towards darker. It is to be noted that this 
 observer often reports twofold changes of this kind, and that she 
 is always surer of the suggested changes than of those due to the 
 actual change of stimulus. / _ even uses the word " brilliant " in 
 describing a change with the stimulus 20° darker. 
 
 Observer J, with suggestion " Towards darker " and stimulus- 
 change 10° lighter, reports : " Noticeable change first towards darker ; 
 constant for a while; then slightly brighter; change towards darker 
 came suddenly, was distinct but only a slight change; change towards 
 lighter was not quite as distinct; I was not as sure of that." With 
 a change of 22° towards lighter she reports : " Distinctly towards 
 darker; very distinctly noticeable; greater change than in any of 
 the other experiments." 
 
 Control Group 
 
 Observer A reports an "'agreement " which he " seems to have ahead 
 of time," and he often replies with a " Yes ;" " Judgment shot itself 
 off, ' Yes.' " Sometimes A also reports a good deal of imagery. 
 He is, however, an observer of a rather extremely 'objective' type, 
 with a persistent tendency to hold fast to the stimulus and to accept 
 the sense-impression as it comes. His judgments, so far as we can 
 tell, are nearly always of the sensory type. There seems to be no 
 greater assurance in this group than in the suggestion group. 
 
 Observer B often replies by " yes," and reports much kinaesthesis 
 (eye-strain, tendency to nod head, etc.). Once he reports: "A vague 
 memory of previous experiments when I had to be on my guard." 
 His attitude and assurance were changed in the control group. 
 
 Observer C reports changes no greater or more definite in this 
 than in the suggestion group. In fact we find an emphasis placed 
 on a few, at least, of the inverted judgments which we do not find 
 in the control group. C gives many imaginal and mixed judgments; 
 but the sensory are the more numerous. 
 
 Observer F reports judgments with and without imagery. When 
 asked to compare the experiences of this group with those of the sug- 
 gestion group she said : "At the time they seemed to be actually sensa- 
 tions. There did not seem to be anything imaginal about them. I should 
 probably have declared under oath that there was a change ' towards 
 brighter' or whatever you suggested." After another trial she said: 
 " Same effect though not as distinct as last time. This was about 
 the same as in the first group. I don't think that if I had the two 
 side byi side I could have told any difference, unless this latter might 
 be a little clearer and I can't be sure of that. They were very much 
 alike, I am sure, of that." And again : " In; trying to decide whether 
 my experiences were sensory or imaginal, there seemed to be both 
 sensations and images, but I seemed to get things I did not expect 
 
116 EDWARDS 
 
 and had not imagined. Sometimes the light seemed to be brighter, 
 so much so that I felt surprised. . . . Some of the results were 
 different from that imagined, i. e., expected. It seems to me now that 
 suggestion affected both imagery and sensations, but I can't tell which 
 was affected most." 
 
 Observer G expresses much doubt in this group, but less than in 
 the suggestion group. With 20° change towards brighter she is 
 doubtful, and with 24° she says : " Guess that was brighter, not 
 sure." After an observation with 20° darker, she remarks : " This 
 seemed about the same as results in the suggestion group. I don't 
 seem to notice any difference between this last experiment and the 
 suggested experiences." Later, in support of her statement that her 
 former judgments were due to 'imagination,' she says: "In sensa- 
 tion I see the change towards darker as a cloud. If it were imagina- 
 tion, it would be confined more to the window itself. When I shut 
 my eyes I can get a clear-cut image of the window itself." But 
 several reports in the suggestion-group refer to just this "clou4"- 
 effect, which the observer maintains would malce the experience 
 sensory. 
 
 Observer H can find no difference in the experiences of the two 
 groups. " Can't now pick out any differences. Changes seemed 
 definite then and they are definite now." "The changes in the first 
 group seemed just as much out in front of me; these in the second 
 group no more clear or definite. ... I think the change was 
 sensorjr because the change came differently than I expected it would; 
 I had images of what I expected, and what I saw was contradictory 
 to the images." It should be noted that in this case the control group 
 followed the suggestion group at an interval of only a few days. 
 
 Observer / agrees with H. "At the time I looked at them [the 
 exposures of the first group] they were perfectly real; they looked 
 just like a moving picture or panorama that moved in front of me; 
 it was different from imagination and I did not doubt its objectivity." 
 ' They seemed like sensory experiences in that they seemed to be 
 presented to the eye with visible changes. . . . There was also 
 an . . . idea that the changes were not quite what I expected 
 them to be; it was a little feeling of dissatisfaction, although I had 
 no definite images of what I should get. ... If you had said 
 to me that it was just my imagination, I should have taken exception 
 to it, because I really saw something there." 
 
 Observer / has slightly less assurance here than she had in the 
 suggestion group. She reports that there is no difference that she 
 can name between the experiences of the suggestion and of the 
 control groups ; " The effect of suggestion in earlier group was cer- 
 tainly, I think, on the sensations." 
 
 Summary of Results from Experiment 3 
 
 (i) Again, similar (and in part indistinguishable) results are 
 obtained from the experiments of the two groups. 
 
 (2) The assurance of A, C, F, H, I, and / is approximately the 
 same in both groups. I expresses greater confidence in the sug- 
 gested than in the actual change, when on certain occasions she 
 reports change in both directions. 
 
 (3) The introspections again show three types of judgment, the 
 imaginal, the motor, and the direct sensory judgment. The sensory 
 tjrpe is the most frequent in this experiment: the judgments of ob- 
 
EXPERIMENTAL STUDY OF SENSORY SUGGESTION 117 
 
 servers A, H, I and / are predominantly direct; those of C, F and G 
 are sensory or imaginal or mixed; those of B are largely motor. 
 
 (4) A finds a " very definite " change (judgment being inverted) 
 immediately after B has given, at A's request, special practice trials 
 with knowledge. B, in the suggestion group, accepts a very large 
 change of direction as simply a continuation of the series. 
 
 Experiment 4. Visual Extent 
 The Wundt apparatus was arranged to show a white rectangle, 14 
 by 60 mm., upon a black background. The rectangle could be length- 
 ened or shortened at will; the unit of change was i mm. The 
 observer saw the standard strip for i sec; then came a pause of 
 S sees.; and then the comparison strip for i sec. The instructions 
 were : " I shall expose a rectangle, first, at the standard size ; then 
 I shall expose it at a length either greater or less than the standard. 
 You are to note any change in length, and to indicate whether the 
 second rectangle is longer or shorter than the standard. The change 
 in the first pair will be very small, and will be increased in suc- 
 ceeding pairs till you notice the change." The suggestion was re- 
 peated before every trial : " Towards the longer," or " Towards the 
 shorter." A head-rest brought the eyes of the observer 90 cm. from 
 the frame. A gray card, placed just before the stimulus, aided 
 fixation and also helped to eliminate after-images. Great care was 
 taken to rule out secondary criteria. A, B and C took part in the 
 
 experiment. 
 
 Suggestion Group 
 
 Observer A gave comparatively few inversions. With the variable 
 4 mm. larger, he reported, with no doubt at all, " shorter." With a 
 change of S mm. he said " larger, doubtful ;" with 6 mm. he was 
 sure of " longer." In another trial a decrease of 3 mm. was judged 
 " longer " with no doubt. There were many " doubtful " and " equal " 
 judgments. A declared his suspiicion of the use of suggestion 
 almost at the beginning of the experiment, and said : " I seem to be 
 more suggestible in this experiment than in the others." 
 
 Observer B showed in tone of voice and in manner that he ques- 
 tioned the nature of the experiment; and failure in four series led 
 E to discontinue it. There were no inversions, but the judgments were 
 mostly " equal " and " doubtful." 
 
 Observer C gave no inversions when the change was actually 
 towards the smaller. There were, however, many " equal " and 
 " doubtful " judgments with changes as_ great as 5 and 6 mm. C 
 reports that, when E says the change will be in one direction, " the 
 suggestion " seems to make it just the opposite. 
 
 Control Group 
 
 Observer A finds the " readiness " or " agreement " that he has 
 reported before. The ts^pical introspection shows little or no imagery. 
 
118 EDWARDS 
 
 He says, e. g., "Attention strong on second stimulus ; ' shorter ' came 
 out immediately upon clear visual and kinaesthetic perception of the 
 stimulus." 
 
 Observer B gives his judgment characteristically by "yes.'' "A 
 peculiar auditory after-image immediately succeeded B's voice, a 
 sort of echo accompanied by verbal idea ' yes, smaller.' I not only 
 say to myself ' smaller ' but say to myself ' yes, smaller.' Sometimes 
 I think it quite irrelevant whether I tell you my judgment or not. 
 The fact that I say ' yes ' to myself seems to show the working of 
 the suggestion." " Strong repetition of instructions seemed to con- 
 stitute the suggested consciousness. When I actually do see the 
 second stimulus it is surprisingly more different than I expect. That 
 surprise is mainly kinaesthetic and organic, partially in verbal com- 
 ment as if setting of mouth to say 'why?'" 
 
 Despite the verbal set, B reports that there is no tendency to give 
 a judgment before the second stimulus is perceived. 
 
 Observer C reports a good deal of imagery which appears to be 
 largely responsible for the ' equals ' and ' doubtfuls ' of the suggestion 
 group. " The perception of the white is not constant ; the apperceptive 
 process undergoes development and changes." 
 
 Results of Experiment 4 
 
 (i) This experiment was not successful. Only in one or two 
 cases do we find inverted judgments given with an assurance as 
 great as (or greater than) that of the correct judgments. 
 
 (2) The types of judgment — motor, imaginal, sensory and mixed — 
 are fairly evenly represented. A and B tend here toward the motor, 
 C is definitely of the imaginal type. 
 
 Experiment 5. Noise Intensity 
 The point marked 45 on the graduated arc of the Fechner pendulum 
 was chosen as that from which to give the standard stimulus. The 
 steps used for the variable were the divisions of the scale. The ob- 
 servers were instructed as in the former series. The interval between 
 standard and variable was 8 sec. The suggestion was given before 
 every trial ; " Towards the louder,'' or " Towards the fainter." A, B 
 and C took part in the experiment. 
 
 Suggestion Group 
 
 Observer A stated near the beginning of this experiment that he 
 gave himself the counter-suggestion of equality of the stimuli; and E 
 did not venture to give many supraliminal changes. With the sug- 
 gestion " Towards louder " A reported : 38, " Weaker, very doubtful ; " 
 42, "weaker, doubtful, just a trifle weaker;" 41, "equal or a trifle 
 louder ; second seemed to have more volume ; " 40, " equal or a trifle 
 louder." With the suggestion " Towards fainter " A responded : 47, 
 "equal, doubtful;" 49, "louder, doubtful;" 50, "trifle louder;" si> 
 "weaker, not very much weaker, but my judgment was fairly sure." 
 A reversal of stimulus-change, made at the end of one series, was 
 accepted without comment, as if a continuation of the series. AH of 
 these changes were supraliminal. 
 
EXPERIMENTAL STUDY OF SENSORY SUGGESTION 119 
 
 Observer B reported : " Sometimes it seems to go opposite to what 
 you say it will. I sometimes say ' doubtful ' and wait till another 
 step." With the suggestion " Towards louder " B replied : 42, " doubt- 
 ful, a difference, not very sure whether louder or not ; " 40, " equal ; " 
 38, " equal ; " 37, " certainly sounds less ; " 36, " less." E then gave a 
 large supraliminal change in the opposite direction, and B merely said 
 " more," accepting the trial as a continuation of the series. With the 
 suggestion " Towards fainter " B answers : 51, " about equal ; " 50, 
 " equal ; " another 50, " That was less ; that was quite surprisingly 
 less ; " 52, " That wasn't less, equal if anything ; " S3, " equal ; " SS, 
 " equal ; " and in this series it is only at 56 that B says " louder." 
 E then altered his direction, and gave 40. B said : " Less, not as 
 sure about it as other, but no doubt." This change, too, was accepted 
 as a continuation of the series ; although, just before it was begun, 
 some practice trials had been made in both directions. 
 
 Observer C reports suspicion of the experiment. Yet with the sug- 
 gestion "Towards louder" he responds: 42, "louder, sure judgment;" 
 36, "very slightly louder or equal;" 41, "louder;" 39, "louder" (C 
 adds here that he could "judge confidently; my attention has been 
 good, hence the confident tone of voice"); 38, "louder," again with 
 confident voice. With the suggestion " Towards fainter," C gives at 
 51, " fainter, fairly confident." 
 
 Control group 
 
 Observer A gives his judgments nearly always by " yes " or " plainly " 
 or some like word. In nearly every case the analysis of judgment 
 shows the intervention of anticipatory auditory _ images. The situation 
 differs from the preceding, in that now there is no self-suggestion of 
 the equality of the two stimuli. 
 
 Observer B often refers to the automatic release of the response. 
 " The word ' yes ' was almost forced upon me, as if I could not say 
 anything else." 
 
 Observer C also reports several cases in which the judgment was 
 touched off automatically. He has, however, a good deal of imagery, 
 visual and kinaesthetic, and once mentions an anticipatory image of 
 sound. 
 
 Results of Experiment 5 
 
 (i) All three observers give inverted judgments with as much assur- 
 ance as the judgments of the control group. 
 
 (2) Reversal of direction of the stimulus-change is accepted by A 
 and B without surprise or comment. With both observers, the changes 
 involved were large. 
 
 (3) The judgments in this experiment were characteristically of the 
 motor type. With B these motor judgments are largely preponderant; 
 and with A (who inclines strongly to the sensory) and C (who in- 
 clines to the imaginal) they are in the majority. A reports the almost 
 constant appearance of auditory images. 
 
 Experiment 6. Pitch 
 
 The standard stimulus was a tone of 870 v. d. from the Stem 
 
 variator; the unit of change, up and down, was 2 v. d. The duration 
 
 of the stimulus was 6 sec, increased slightly for the largest changes. 
 
 The observers sat about 5 m. from the apparatus. The instructions 
 
120 EDWARDS 
 
 were: "I am going to give you a continuously changing tone, and 
 I want you to report as soon as you notice the change in pitch. 
 Please indicate also the direction of the change. The change in the 
 first trial will be very small, and will be increased in each succeeding 
 trial until you notice the change." The suggestion was repeated 
 before each trial: "Towards the higher," or "Towards the lower." 
 Observers A, B and C took part in the experiment. 
 
 Suggestion Group 
 
 Observer A was so suspicious of the experiment that but few series 
 could be taken. There was, however, one inversion of judgment 
 with supraliminal stimulation. 
 
 Observer B was also very suspicious, and the series were therefore 
 curtailed. There were, however, clear supraliminal inversions: e. g., 
 with the suggestion " Towards higher " and the stimulus-change to 
 860 V. d., B judged "higher." 
 
 Observer C declared, before the end of this experiment, that he 
 thought the experiments were on suggestion. E told him emphatically 
 that he was making a study of limens, but that he did not wish to 
 explain further. The few results obtained are the more striking as 
 E gave C on request a number of trials for practice, both up and 
 down, at the beginning of the series. Most of the inversions occurred 
 at 6 v. d. ; but once, with the change of 8 v. d. higher, C repUed 
 " down " with no comment ; and once, with the change of 8 v. d. 
 lower, he replied " higher " with no comment. 
 
 Control Group 
 
 Observer A finds nothing new in introspection except the prevalence 
 of tonal imagery. 
 
 Observer B reports characteristically by " yes " or some like word. 
 Once he says : " My readiness to make the judgment was present 
 before the tone came." 
 
 Observer C still reports much imagery. 
 
 Results of Experiment 6 
 
 (i) This experiment was not successful. A few inverted judg- 
 ments, especially with observer C, show an assurance equal to that 
 of the control group; but the results are very scanty. 
 
 (2) B gives judgments chiefly of the motor, C of the imaginal or 
 mixed type. A, as in the preceding experiments, is of the sensory type. 
 
 Experiment 7. Noise 
 Series of 10 stimuli were given by the modified acoumeter; 8 with 
 the larger lead ball commonly used; 2 (at varying places in the 
 series) with a cotton ball made compact by a few stitches of thread 
 and blackened with ink. Preliminary observations were taken with 
 the lead ball, in order that this sound might, in the experiments 
 proper, be made definitely supraliminal. The observers were seated 
 
EXPERIMENTAL STUDY OF SENSORY SUGGESTION 121 
 
 at a distance of 3.5 to 4.5 m., from which they could easily see the 
 ball di-op. The intervals between trials were of 8 sec. duration. The 
 instructions were : " I want to make a few preliminary trials to 
 find the limen roughly and see if I can make use of the experiment. 
 Teil me whether or not you hear the sound from the ball dropping. 
 You are to watch the ball, and to report after you see it drop." The 
 suggestion lay in the fall of the cotton ball. Observers A, B, C, G, 
 I and / took part in the experiment. 
 
 Suggestion Group 
 
 Observer A expressed surprise that he saw the ball drop but did 
 not hear it. Once he "heard it hit the side after it dropped;" and 
 once he replied confidently with " Yes." A little later he said he 
 believed E was using two balls and trying to ' fool ' him. 
 
 Observer B reported hearing the cotton ball three times; once he 
 gave the judgment in a " yes " that indicated conviction. During 
 one series he remarked: "When I don't hear it I think it is a lapse 
 of attention; it takes about all the attention you have." 
 
 Observer C often reported that he heard the ball drop. His reports 
 were given with full assurance, notwithstanding the fact that he 
 several times failed to hear the ball and complained of lapses of 
 attention. 
 
 Observer G reported 'hearing' the ball drop once; I, once; and 
 / not at all. Failure was due, in part at least, to suspicions on the 
 part of the observers. G expressed some doubt about her judgment, 
 although she called the experience a "perception of sound;" and her 
 final judgment was "positive." 
 
 Control Group 
 
 No introspections were taken from G and /. The two following 
 reports from I are typical. 
 
 Lead ball used : " Visual perception of falling shot ; fell in straight 
 line. Auditory perception: sound of falling shot; quality clear; 
 auditory image of ticking of clock. Duration very short; sound 
 faint and far away." 
 
 Cotton ball used: "Visual perception of falling shot. Seemed to 
 fall faster than before. Auditory perception, sound of shot falling, 
 clear; duration longer than before. Sound quite loud and unmistak- 
 able; visual and auditory image of dried pea falling on glass counter." 
 
 Results of Experiment 7 
 
 We have found cases — a considerable number with one observer — 
 in which a cotton ball is 'heard' to drop just after a lead ball has 
 given a definitely supraliminal stimulus. The judgments are passed 
 with assurance; and they are not corrected later, although the true 
 stimulus is immediately repeated. There is nothing to show that 
 the suggested experience could be distinguished from that aroused 
 by adequate stimulus. 
 
122 EDWARDS 
 
 Experiment 8. Smell 
 Two paper funnels projected from a small closet in which E sat 
 during this experiment. E often made various noises to suggest 
 the presence of apparatus; no odors were used. The observer sat 
 about 3 m. away, directly in front of the funnels. The instructions 
 were : " I shall give you eiither a pleasant or an unpleasant odor, 
 which will come from one of the funnels, and I want you to tell 
 me as soon as you notice it. I shall keep the time with a stop-watch. 
 As soon as you report the first odor, I shall give you another which 
 will neutralize the first, and I want you to tell me as soon as the 
 first odor has disappeared." The suggestion was made before each 
 trial : " I shall now give you a pleasant [or unpleasant] odor." The 
 observer was asked to give as full an introspection as possible after 
 each successful trial. Observers D and E took part in the experiment. 
 
 Observer D; suggestion of " unpleasant odor." Duration, 31 sec. : 
 "Got just one whiff of unpleasant odor; it smelled like hydrogen 
 dasulfid. I had a decidedly unpleasant odor." Duration, 106 sec. : 
 " Before getting the smell there was a watering of the eyes, then 
 rapidly followed a tickling in the nose and in about the same interval 
 the smell came. I tried hard to get smell all through period; con- 
 centrated attention to get it. It does not seem to stay or grow 
 more intense but comes and goes. The odor was very clear; I do 
 not believe it could be any clearer. It burst out qirickly; got no 
 after-image. My eyes smart a little now." Duration, 97 sec. : " Same 
 characteristic tightening in forehead and sensations in throat. The 
 smell is the same I have spoken of only more intense. It was dis- 
 agreeable and I smelled it a long time before it was disagreeable; 
 not very clear at any time, but got the odor definitely; what I re- 
 ported I got." Only two positive results came with the suggestion 
 of " pleasant odor." Once D said : " I can't really smell it, but think 
 there is a smell there because of the change in blood pressure;" 
 and again : " Slight odor ; no pleasantness or unpleasantness." 
 
 Observer E: Suggestion of "unpleasant odor." — Duration, 96 sec: 
 " It was rather irritating; seemed to irritate lining of nose and throat; 
 something like ammonia. It wasn't intense enough to be really un- 
 pleasant. I could only get it by keeping my attention on it. Fairly 
 clear." Duration, 97 sec. : " The unpleasant odor seemed fainter 
 than before, and I was rather conscious of smelling something than 
 of a particular smell. It was irritating. It was not clear."' — Sug- 
 gestion of " pleasant odor." Duration, 56 sec. : " Very faint and of 
 a pleasant soothing nature. I got it as an odor only, that is there 
 was no irritating effect, and I got it in the fore part of my nose 
 and it was fairly clear." Duration, 134 sec: "It was pleasant, of 
 a soothing, sweet nature; I got it in the fore part of my nose. It 
 was rather clear and lasted about three minutes." 
 
EXPERIMENTAL STUDY OF SENSORY SUGGESTION 123 
 
 Results of Experiment 8 
 
 (i) The reports of odors are given with full assurance. 
 (2) Although the reports mention various modes of tactual experi- 
 ence, yet they distinguish these from the experience of smell proper. 
 
 Experiment 9. Taste and Smell 
 Specially prepared small candies were used. Half of them were 
 wrapped at the time of the experiment in tin foil; these 
 were also of different shape from the unwrapped. All 
 were made at the same time, and were of the same material 
 and the same size. The foil-covered candies were called A, the 
 others B. The observer tasted A (or B) for 15 sec; then rinsed 
 the mouth (about 30 sec.) ; then tested B (or A) for 15 sec; and then 
 gave his judgment. The instructions were: "I am going to give you 
 two stimuli, and I want you to make a careful comparison between 
 the two as regards gustatory and olfactory sensations." A, B and C 
 took part in the experiment. Each observer made 10 comparisons. 
 Differences were reported as follows : 
 
 Observer A: A sweeter 6; B sweeter 2; A more bitter 4; B more 
 bitter i ; A more olfactory 4 ; B more olfactory i ; A smoother 2. 
 
 Observer B : A sweeter i ; B sweeter 6 ; A more bitter 4 ; B more 
 bitter i ; A more sticky 6. 
 
 Observer C: A sweeter 2; B sweeter 3; A more bitter 4; A more 
 olfactory 2. 
 
 The following are typical introspections. Observer A : " In A the 
 sweet tasted first quite intense, then bitter, with smell quite intense. 
 Intensive changes or alternations. A also is harder than B. In B 
 the sweet is less intense than in A. Bitter is a long time coming as 
 also is the smell, neither so strong as in A. No such alternation of 
 large-small intensity as in A." Observer B: "A called forth first 
 tactual and vague kinaesthetic sensations, sweet, not strong, some 
 slight cold sensations in middle of tongue. Finally a bitter almond 
 taste, not quite as strong as the sweet. Touch outlasted them all. 
 B was very much sweeter, sticky, and the last burnt smell came only 
 immediately before rinsing the mouth. The kinaesthetic sensations 
 from chewing were somewhat clearer than in A, but not more in- 
 tense." Observer C: " B held at the tip of tongue. At first a very 
 bitter sensation there. Then the stimulus was broken up. Now 
 sweet sensations for a considerable period mild, then becoming very 
 strong with a very faint admixture of bitter sensations. A held at 
 tip of tongue. Stronger and more widely distributed bitter sensations 
 than with B. Then the stimulus was broken up. Now sweet sensa- 
 tions for a time mild, then becoming seemingly as strong as those 
 of B. But, perhaps, not maintaining this high intensity for so long a 
 time. Admixture of stronger bitter sensations than with A." 
 
 At the close of the experiments the observers were asked if they had 
 noticed any constant differences between A and B. Observer A de- 
 cided that A was generally smoother, softer and sweeter; quite often 
 more bitter and having a stronger smell than B; taste and smell sensa- 
 
124 EDWARDS 
 
 tions rise more quickly in A than in B. (His reports show that A is 
 generally reported as sweeter, more bitter and more olfactory than B; 
 B also is reported as harder than A, and A as smoother than B.) 
 
 Observer B decided that B was sweeter and more sticky than A. 
 (This agrees with the reports, which show that B was sweeter and 
 more sticky than A; B is also reported, however, as more bitter.) 
 
 Observer C decided that A and B were the same, except that A was 
 the more bitter. (This excess of bitter is the difference most promi- 
 nent in his reports.) 
 
 Results of Experiment g 
 
 (i) The reports of difference are given with full assurance, and 
 differences are often reported as large, although the observers, espe- 
 cially A, had their suspicions that the stimuli were identical. 
 
 (2) The judgments are all (except in one instance with observer 
 C) of the direct sensory type. 
 
 Experiment 10. Pressure 
 Stimuli were applied to the back of the hand by the ivory button of 
 the Wundt apparatus; 50 gr. was the standard; change was made by 
 steps of 2 gr. Application was for i sec, and the interval was 5 sec. 
 Care was taken to prevent fatigue. Instructions were as in other 
 experiments. The suggestion was repeated before each trial : " To- 
 wards the lighter," or " Towards the heavier." Observers D and E 
 took part in the experiment. 
 
 Suggestion Group 
 
 D gives many "equal" judgments. With suggestion "Towards 
 heavier," and lighter stimuli 6 (once), 12 (once), and 8 (several 
 times), he says "heavier." Twice E reversed the direction of stimulus- 
 change, to avoid suspicion : in the one case, in which the second stimu- 
 lus was lighter, D said " heavier " at 6, and " equal " at 8 to 16 ; 
 change was then made from 16 lighter to 8 heavier, and D accepted 
 the stimulus as a continuation of the series, and still judged " equal." — 
 With suggestion "Towards lighter" there were no inversions, but 
 many judgments of " equal." 
 
 Observer E reported " heavier '' when the stimulus was 8 lighter 
 (twice) and 10 lighter (twice). There were no inversions with the 
 suggestion " Towards lighter," though there were several " equal " 
 judgments. 
 
 Control Group 
 
 Observer D reports a great deal of imagery, visual, verbal-auditory, 
 and anticipatory tactual. Thus : " Sensation from first stimulus ; 
 images left consciousness entirely, tactual sensation very clear. Then 
 tactual image came, and verbal-auditory image of word ' strong.' " 
 
 Observer E also describes much imagery, chiefly visual. 
 
 Results of Experiment 10 
 
 This experiment was inconclusive. Several inversions, however, 
 occur with full assurance. Reversal of direction of stimulus-change 
 was possible without comment from the observer. Judgments were 
 predominantly of the imaginal type. 
 
EXPERIMENTAL STUDY OF SENSORY SUGGESTION 125 
 
 Practice experiments made with observer F and another whom we 
 shall call K (a man taking advanced work in psychology) led to 
 results that may be mentioned here. Discrete stimuli were given with 
 a small spring pressure-apparatus, and no misleading suggestion was 
 employed. The observers were told only that two pressures would 
 be given, the second heavier than the first, and were asked to make 
 as full introspections as possible. 
 
 Observer F reported : " First noticed the instrument before it 
 touched my hand; then felt pressure and tried to carry it over in 
 memory; held breath till second came. Second more intense; slight 
 pain; seemed to feel it coming by degrees till instrument touched 
 hand." E asked F to say more about the fore-period, before the 
 stimulus was applied. " First feeling was less intense. It seemed to 
 last longer than the pressure itself, quite a little longer. It was clear 
 and definite, but much less intense than the pressure; seemed to get 
 nearer by jerks. Just as clear as actual pressure, only less intense. 
 Remember surprise at feeling it before stimulus touched my hand." 
 And again : " First felt pressure near third finger ; then felt touch 
 from first stimulus, much less intense than pressures in other experi- 
 ments and not lasting so long. . . . The first [pressure] before the 
 stimulus touched the hand . . . was like a big surface ; seemed 
 heavier than the two following [from stimuli], not so clear cut; other 
 two real ones seemed less intense than the first. . . . All three 
 seemed about the same in clearness. The first, i. e., the one before 
 the stimulus, seemed to be of longest duration. The two from the 
 stimuli seemed to have smaller area." 
 
 Observer K reported : " Auditory perception of E'% voice. Some 
 imagination, imagining what feeling would be like. Really, a tactual 
 sensation before the stimulus was applied ; rather diffuse, seemed 
 much lighter, smaller, to cover less space than perception with stimu- 
 lus. The real perception much clearer ; following the perception was 
 an after-image." K added that the ' imagination ' of sensation before 
 the stimulus was applied "would be as clear as a very light sensation 
 produced by the instrument." 
 
 In these reports we have a mention of anticipatory ' images ' which 
 seem closely to simulate sensation. 
 
 Experiment ii. Temperature 
 Cylinders i cm. in diameter were used; 35° C. was the standard 
 stimulus, and the variable increased or decreased by 1° in successive 
 trials. Instructions were as usual; the suggestion was repeated before 
 every trial : " Towards the warmer," or " Towards the colder." The 
 stimulus was applied for 2 sec. ; the interval between stimuli was 5 
 sec. The inner side of the fore-arm was stimulated. Observers A, 
 B and C took part in the experiment. 
 
126 EDWARDS 
 
 Suggestion Group 
 
 Observer A said after a few trials that he thought the stimuli of 
 each pair were probably the same. However, with the suggestion 
 " Towards colder " he judges 4° warmer as " colder ; " S° warmer as 
 " equal ; " and in another series, 3° warmer as " colder, doubtful." 
 There were no supraliminal inversions when the second stimulus was 
 colder. 
 
 Observer B with suggestion " Towards warmer " gives one inver- 
 sion, at 2° colder. With suggestion "Towards colder" he judges 3° 
 warmer as "colder" (twice), as "slightly colder" and as "equal" 
 
 Observer C with suggestion " Towards warmer " misjudges only 
 once ; 2° colder is called " warmer." With suggestion " Towards 
 colder " he calls 3° warmer, " very decidedly colder ; " 4° warmer, 
 "very slightly colder or equal;" 6°, "equal;" 7°, "colder." After 
 this judgment C said: "These judgments are more marked, the judg- 
 ments more confident, than in the case of the visual sensations." 
 
 Control Group 
 
 Observer A several times reports anticipatory images. " E's voice 
 ... in background; but was sufficient to arouse visual and tactual 
 image of coming stimulus." Again : " Tactual image of how the 
 stimulus would feel." Again : " E's ' ready ' and ' now ' though heard 
 inattentively caused increased strains and anticipatory tactual image." 
 Once he speaks of a "tactual and cold(?) image." 
 
 After one of these reports, E asked A why he used the words 
 " image '' and " sensation " as he did. A replied : " If I had used the 
 word ' sensation,' that would have meant that you touched me or that 
 I thought you did. A touch sensation is different from a tactual 
 image. I should say there is generally more of a sensation. I mean 
 that the image is abbreviated, does not contain all that the sensation 
 contains ; perhaps, is of less extent, less intensity, generally, and I 
 think (keeping this experiment in mind) that its extent is less clearly 
 defined." After another experiment he remarked : " The criterion 
 is probably the source of the stimulus, though why I answered as 1 
 did was probably unconscious." 
 
 Observer B reports : " Here I had a good anticipatory cold image, 
 localized on the fore-arm before the stimulus was given. Verbal com- 
 ment in imaginal terms : ' Why, here I am feeling cold.' Attitude of 
 surprise." Asked why he used the word " image," B replied : " Might 
 have been ' sensation ' if the tactual sensations had also been present." 
 In another trial B says : " Slight pressure and distinct warm sensa- 
 tions from the time E said 'now' up to the giving of the second 
 stimulus." Here he uses the word " sensation." 
 
 Observer C reports " a diffuse cutaneous image of warmth in a 
 certain cutaneous region of the forearm. Under the influence of the 
 Aufgabe this became more intense." Again :■ " In mid-period kin- 
 aesthetic-auditory images 'the next will be warmer;' then attitudinal 
 process meaning roughly ' It is really so ; ' then cutaneous image of 
 warmth, warmer than first [sensation] but not so warm as second 
 proved to be ; feeling of surprise." In differentiating between " sen- 
 sation " and " image " at the time of these accounts, C said that, in 
 the case of the sensation, there were other intermingled sensations. 
 There was also a different attentional accommodation; the sensation 
 was more intense and compact; the image was, as a rule, less intense. 
 
EXPERIMENTAL STUDY OF SENSORY SUGGESTION 127 
 
 Results of Experiment ii 
 
 (i) The suggestion group shows inversions of judgment for all 
 three observers; the judgments were made with full assurance, and 
 in the face of some suspicion as to the nature of the experiments. 
 
 (2) The various types of judgment — motor, imaginal, sensory and 
 mixed — are in this experiment fairly evenly represented. 
 
 (3) Anticipatory " images " of fair intensity play a part in the 
 judgments. 
 
 E. Discussion and Conclusions 
 
 We have relied, in this paper, upon the reports of trained 
 observers ; and we have been obliged, in consequence, to utilize 
 the observers A, B and C for experiments upon which they 
 entered with definite suspicion. The effects of the verbal sug- 
 gestions were thus, we may suppose, markedly lessened. Un- 
 less, however, we had the analyses given by these observers, 
 we could hardly place any high degree of confidence in the 
 positive statements of the less trained observers ; types of judg- 
 ment must be distinguished before interpretation is possible. 
 It is regrettable that the limits of space have prevented our 
 quoting in full some of the elaborate reports of the Control 
 Group; still, enough has probably been given to indicate the 
 general nature of the work. 
 
 Our first difficulty, then, lay in the necessity of recourse to 
 trained observers. These men, just because they were trained, 
 were less suggestible than the average and were likely to sus- 
 pect the character of the experiments. A second difficulty arose 
 from the necessity of prolonging the single observations. If 
 a continuously changing color is to be wrongly ' seen,' or if a 
 second noise is to be wrongly ' heard ' in relation to a preced- 
 ing noise, then the color must not change too quickly and the 
 sounds must not occur in too quick succession. Our durations 
 and intervals were chosen, after preliminary experiments, as 
 times that should be long enough for the working of the sug- 
 gestion and yet not so long as to preclude the possibility of a 
 direct sensory judgment. We cannot pretend to have made 
 them optimal; further work is desirable, with variation of 
 times, just as further work is desirable with more trained ob- 
 servers and with other modes of stimulation. 
 
 Our positive results may now be summed up as follows : 
 
 (i) We have found three types of suggested consciousness, 
 yielding three typically different kinds of judgment, (a) In 
 some cases we note a carriage of the verbal suggestion, as if 
 across the sensory contents, to the far end of the arc, i. e. to 
 the motor expression of judgment. If it is too much to say 
 that the sensory contents in such cases are irrelevant, we can 
 
128 EDWARDS 
 
 at least affirm that rather startling approximations to that 
 state of affairs have occurred in the course of the experiments. 
 The judgments of B, in particular, are largely of this type; 
 they seem to represent the ripening of a motor impulse ; they 
 ' go off in a Yes ' or ' say themselves.' It is hardly question- 
 able that many instances of ' sensory suggestion ' are, in real- 
 ity, of this motor sort, (b) In other cases we note a free play 
 of relevant imagery: of imagery that is largely due, in all 
 probability, to the long-drawn-out character of the observa- 
 tions. The imagery clearly serves, in many instances, to fur- 
 nish a new context or setting to the presented stimulus, and 
 thus influences or determines judgment; in other words, the 
 judgment expresses an imaginally modified situation, and does 
 not bear directly upon the stimuli. Fully half of C's judg- 
 ments, and rather more than half of E's, are of this imaginal 
 sort, (c) In a third or sensory type of judgment, imagery is 
 either lacking or fleeting and apparently irrelevant, while there 
 is no indication of any motor impulse to express the sugges- 
 tion. Judgments of this kind are the most frequent with the 
 remaining observers and form (so far as we can judge) just 
 over 40 per cent, of the total number of judgments passed in 
 the Control Group. They alone can furnish evidence of a 
 truly ' sensory ' suggestion, (d) Lastly, there are judgments 
 which cannot be placed outright under any one of these three 
 rubrics, and that must therefore be classed as mixed. The 
 combination imaginal-sensory is fairly common, and observer 
 A gives a good number of judgments in which the sensory and 
 the motor determinations are both involved. 
 
 (2) We have found sensory suggestion, vouched for by 
 sensory judgments, with color and light (continuous change), 
 with ' liminal ' noise, with smell, taste and temperature. Our 
 experiments with grey discs and tonal pitch (continuous 
 change) were unsuccessful, mainly, it appears, by reason of 
 the suspicion of the observers ; those with visual extents failed, 
 partly for the same reason and partly because the experiment 
 favored other than sensory judgments ; and those with noise 
 intensities and pressures were inconclusive, because the experi- 
 ments favored respectively motor and imaginal judgments. 
 Other observers and other experimental arrangements might, 
 of course, yield positive results. We have found, also, 
 ' images ' of touch, warmth and cold which simulate the corre- 
 sponding sensations. 
 
 We do not hesitate to draw the conclusion that, in certain 
 departments of sense, a verbal suggestion may arouse conscious 
 processes which are, phenomenologically, identical with those 
 
EXPERIMENTAL STUDY OF SENSORY SUGGESTION 129 
 
 ordinarily aroused by an adequate stimulus or change of stim- 
 ulus. The departments are those of sight, smell, taste and 
 temperature. We think it probable that hearing and touch may 
 be added to this list; but for those senses our evidence is 
 incomplete or inconclusive. We have already pointed out, on 
 the other hand, that many of the illusions ascribed to sensory 
 suggestion by previous writers are of doubtful character. No 
 observation may be put down as a case of sensory suggestion 
 unless (a) the type of judgment is known to be sensory, and 
 {b) changes of the sense-organ (adaptation, after-image, etc.) 
 have been prevented by a fitting control of the experimental 
 procedure. 
 
 We do not offer any physiological theory of our results, nor 
 do we enter here upon the question of the relation of ' sensa- 
 tion ' to ' image.' It would be premature to attempt a theory 
 (or theories) on the basis of the present observational mate- 
 rial; we still do not know, e. g., the qualitative range of color 
 suggestion, or whether Baldwin's " red for green " is possible. 
 
 (3) No proof was needed that observers are ' suggestible.' 
 We were, however, surprised at the potency of suggestion in 
 the case of observers who had frankly avowed suspicion. A 
 and C, the one unusually 'objective' and the other markedly 
 ' subjective ' in type, would both, again and again, ' accept ' a 
 suggestion against which they were, in general, on guard. We 
 have here, then, a striking confirmation of a principle that of 
 late years has won its way to recognition: the principle that 
 the experimenter must exercise great care in the phrasing of 
 his instructions, and must not be wearied of repeating them. 
 
 If we count the inverted judgments in the series which were 
 taken both by men and by women, we find that the women 
 have an excess of 3 per cent. In view of the lesser training 
 of the women, we regard this difference as negligible. 
 
Cornell University Library 
 BF395 .E26 1915 
 
 olin 
 
 3 1924 029 039 083 
 
:,■ T: ^ \