American Experiments on the Paranormal Cognition of Drawings

American Experiments on the Paranormal Cognition of Drawings ERNEST TAVES, GARDNER MURPHY, and L. A. DALE Abstract: Upon publication in 1944 of Whately Carington’s paper, “Experiments on the Paranormal Cognition of Drawings, III: Steps in the Development of a Repeatable Technique,” it was decided to perform a similar series of experiments at the A.S.P.R. in order to test the hypothesis of repeatability of results as suggested in the Carington paper. Accordingly, during 1944 and 1945 four series of experiments on the paranormal cognition of drawings were per- formed at the Society. 272 percipients, scattered throughout the United States (a few in Canada), took part. The results were scored both by Carington’s catalogue and by an American catalogue de- scribed elsewhere in this issue. By neither method of assessment were the total results shown to be significantly positive, Significant decline effects, however, were shown to exist. Strictly speaking, our results cannot be considered to confirm Carington’s hypothesis; this may be duc, however, to the fact that in some respects our procedure differed from his. In another paper in this JourNat (1) is described the construction of an American catalogue for the evaluation of experimental results in the paranormal cognition of drawings. It is the purpose of this paper to describe the procedures and results of the experiments upon which the construction of that catalogue was based. During the period between May, 1944 and January 1945 we performed at the rooms of the Society a series of ex- periments on the paranormal cognition of drawings, modeled after the experiments of Whately Carington (2, 3, 4, 5). Four series of experiments were performed, designated as CA, CB, CC, and CD; of these the first two were composed of four experimental sessions, and the last two of three experimental sessions each. Thus there were fourteen eve- nings of experimentation. Upon each of these fourteen evenings ten drawings were used as target items. These were selected by means of random numbers from a list of suitable items from Caring- Paranormal Cognition of Drawings 145 ton’s catalogue. The procedure was such that no one knew until the actual moment of selection which target item was next to be employed. The agents, or experimenters, varied during the series from night to night, and included different combinations of the following: the three present writers, Dr. Lois B. Murphy, Dr. Alice Taves, Dr. Gertrude Schmeidler, and Miss Adele Wellman. From two to five of these persons were present at each experimental session. Each session lasted thirty minutes (8:00 P.M. to 8:30 P.M.), thus allowing a three-minute interval for the ex- posure of each drawing. The session began with the random selection of a target item. The person making the selection (by means of a previously prepared list of random num- bers) announced it to the group; one member, chosen sim- ply on the basis of his ability to make freehand drawings, rapidly made such a drawing and tacked it up on a board, where it remained until a total of three minutes had elapsed. The drawing was then taken down, and the procedure thus continued until ten drawings had been selected, drawn, and exposed. During the three-minute periods the other agents in some cases made théir own “non-official” drawings of the target items, or in other cases simply “concentrated” upon the items. Subjects, located throughout the United States (and a few in Canada), were instructed to record their impressions at the appropriate times, upon blank forms supplied by us for that purpose, and to mail these forms to us with all pos- sible dispatch. Subjects were cautioned that only simple, con- crete, “drawable” items would be used; instructions were also given to label each drawing appropriately, with a single word if possible. The numbers of subjects involved in the four experiments were, respectively, 56, 95, 90, and 149." From these sub- jects a total of 8723 responses were received. Of these, 538 1272 individuals took part in the entire series; certain of these individuals, how- ever, participated in two or more experiments.

—S— 146 Journal of the American Society for Psychical Research were “hits” of different types, to be described shortly, leaving 8185 “misses.” These “misses,” when properly tallied, comprise the American catalogue used in scoring the material. Hits were tabulated into three categories, alpha, beta, and gamma, as follows: Alpha hits were “directly on the nose,” so to speak. Within the three minute period during which an item was being drawn and exposed, the subject recorded upon his record sheet his impression as to that drawing. TABLE I Final results, beta hits, all series, scored by two catalogues Expt. Carington scoring ASPR scoring CA Score Variance Sigma CR P CB Score Variance Sigma CR P CC Score Variance Sigma CR P CD Score Variance Sigma CR Pp Total Score Variance Sigma CR P Paranormal Cognition of Drawings 147 Beta hits were hits scored upon a drawing used on that same evening. Gamma hits were hits scored upon drawings used during the appropriate series of evenings (rather than on the evenings of the other three series ). Following Caring- ton, all statistics derive from the beta hits obtained. (Alphas of course are contained within the beta category. ) All of our data were scored both by Carington’s catalogue, using the multiple-unit procedure described by him, and by our own catalogue, with results as shown in Table I. It will be seen that the two results are not identical, although generally more or less comparable; this finding is discussed in the other paper in this JourNAL previously referred to. The table shows that in neither case was a significant over- all result obtained. The results are not, however, completely devoid of inter- est. Comparison of the four successive CR’s obtained when scored by our catalogue provides a curve as shown in Figure 1. This indicates that the phenomenon of initial success followed by subsequent decline—what we call the “Midas touch in reverse’—has once again been demon- -h CA cB cc cD EXPERIMENT Fic. 1—CR’s obtained for four series of experiments, beta hits. ASPR catalogue.

148 Journal of the American Society for Psychical Research 3.0 i 2 3 + EXPERIMENT CA Fic. 2.—CR's obtained on four evenings, CA No. 1, 2, 3, and 4, beta hits. ASPR catalogue. strated. Moreover, when the scores of the first experi- ments (CA) are analyzed, a similar curve appears, as shown in Figure 2. It will be noted that the first evening’s work, as evaluated by the American catalogue, gave a CR of 2.88 (P = .002), which is significant. The results as plotted graphically in these two figures we believe to be the only significant finding in these experiments —the demonstration of another case of initial extrachance scor- ing, followed by a decline to subchance levels. The data were also examined in respect of alpha and gamma hits; the results were essentially without interest. The alpha hits are too scanty to permit adequate statistical oe and the gamma curves parallel those of the beta its. Before concluding this brief report, one further matter requires clarification. During the winter of 1941-42 we had attempted a series of long-distance telepathy experiments patterned in a general way after Carington’s procedure, obtaining null results. Carington was able to show, in terms Paranormal Cognition of Drawings 149 of his association theory of paranormal cognition (5), that this failure may have been due to the fact that not only the actual target items, but all potential target items had been associated in the mind of the agent with the idea of the experiment. Our procedure at that time was as follows: Two of the present writers (GM and LD) chose 208 words which were considered suitable as target items. These words were then illustrated by a City College student. LD, who was subsequently the agent in the experiment, placed the drawings in envelopes, numbered them, etc., and thus was entirely familiar with all 208 potential originals. Re- garding this, Carington wrote: “Mrs. Dale was evidently the ‘culprit,’ so to speak, for there was every opportunity for the 208 potential originals to be firmly associated in her mind with the idea of the experiment—and that is all that is necessary” (6). The possibility exists that the present experiments will be subjected to similar criticisms because ET had inten- tionally, at the end of experiment CA, prepared a modified list of target items based upon Carington’s catalogue; cer- tain items considered by Carington to be suitable target items were either rejected or grouped together with other items to form a single target item. Thus, in experiment CA target items were drawn from all of the items set in Roman type in Carington’s catalogue, whereas in CB, CC, and CD target items were drawn from the modified list, which con- tained 311 items. Experiment CA (which was not signifi- cant as a whole) is exempt from criticism since there we followed, we believe, essentially the procedure outlined by Carington. In all experiments subsequent to CA, however, the situation in respect to ET was comparable to that of LD in the 1941-42 series: that is, all potential target items may have been associated in ET’s mind with the idea of the experiment by virtue of his labors to prepare the modified list. Therefore, in terms of Carington’s association theory a null result might again have been predicted if ET had been the only agent in experiments CB, CC, and CD. But this in fact was never the case (see List of Agents at end

150 Journal of the American Society for Psychical Research of paper). The CR’s obtained on the two evenings when ET was not present (CC No. 1 and CD No. 1) were, respec- tively, — .29 and —.78. Now as to the other agents: GM and LD were agents in CB; they were familiar with the Carington catalogue, but not with the modified list. (In- deed, if a reading acquaintance with the catalogue dis- qualified an agent, its value as a research tool would be very limited.) In experiment CC Dr. Lois B. Murphy acted as one of the agents on two out of the three evenings. She was acquainted with neither catalogue nor modified list. Dr. Gertrude Schmeidler, who likewise had had no contact with the catalogue, with the modified list, nor with any part of the previous experiments, came to New York to act as an agent in CD. She took part on all three evenings. This use of “naive” agents throws further doubt upon the hypothesis that our null results may be attributed to lack of differentiation in the agents’ minds between target and non-target items. In conclusion, it may be stated that, although our results differ from his, we do believe that the Carington long- distance telepathy technique is a signal contribution and should be tried by other experimenters in other laboratories. We believe, moreover, that the phenomenon of initial suc- cess, which has been demonstrated over and over again, is worthy of further investigation, using entirely different types of experiments. LIST OF AGENTS CA 1: ET, AW; 2: ET, AT, LD; 3: ET, AW; 4: ET, LD. CB 1: GM, ET, LD; 2: ET, LD; 3: GM, ET, LD; 4: GM, ET, LD. CC 1: GM, LBM, LD; 2: GM, ET, LD; 3: GM, LBM, ET, LD. CD 1: GM, GS, LD; 2: GS, ET, LD. 3: GM, LBM, GS, ET, LD. REFERENCES . Taves, Ernest, “The Construction of an American Catalogue,” this issue, 1 pp. 151ff. 2. Carington, Whately, “Experiments on the Paranormal Cognition of Draw- ings, 1.” Proc. S.P.R. Vol. XLVI (1940-41), pp. 34ff. 3. , “Experiments on the Paranormal Cognition of Drawings, I,” Proc, S.P.R., Vol. XLVI (1940-41), pp. 277ff. 4. , “Experiments on the Paranormal Cognition of Drawings, III: Steps in the Development of a Repeatable Technique,” Proc. A.S.P.R., Vol. XXIV, January, 1944. . “Experiments on the Paranormal Cognition of Drawings, IV.” Proc. S.P.R., Vol. XLVII (1944), pp. 155ff. 6. “Correspondence,” Journat A.S.P.R., Vol. XXXVII, July, 1943, pp. 148ff. de