This article was downloaded by: [Memorial University of Newfoundland] On: 13 November 2013, At: 02:29 Publisher: Routledge Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK

International Journal of Clinical and Experimental Hypnosis Publication details, including instructions for authors and subscription information: http://www.tandfonline.com/loi/nhyp20

The nature of task interference in hypnotic dissociation: An investigation of hypnotic behavior a

a

V. Jane Knox , Lila Crutchfield & Ernest R. Hilgard

a

a

Stanford University , Published online: 31 Jan 2008.

To cite this article: V. Jane Knox , Lila Crutchfield & Ernest R. Hilgard (1975) The nature of task interference in hypnotic dissociation: An investigation of hypnotic behavior, International Journal of Clinical and Experimental Hypnosis, 23:4, 305-323 To link to this article: http://dx.doi.org/10.1080/00207147508415953

PLEASE SCROLL DOWN FOR ARTICLE Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) contained in the publications on our platform. However, Taylor & Francis, our agents, and our licensors make no representations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of the Content. Any opinions and views expressed in this publication are the opinions and views of the authors, and are not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon and should be independently verified with primary sources of information. Taylor and Francis shall not be liable for any losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoever or howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use of the Content. This article may be used for research, teaching, and private study purposes. Any substantial or systematic reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http:// www.tandfonline.com/page/terms-and-conditions

The International Journal of Clanicd and Ezpcrimcn(d H y p n 1976, Vol. XXIII, No. 4, 305-323

Downloaded by [Memorial University of Newfoundland] at 02:29 13 November 2013

THE NATURE OF TASK INTERFERENCE IN HYPNOTIC DISSOCIATION: AN INVESTIGATION OF HYPNOTIC BEHAVIOR’, 2 V. JANE KNOXa, LILA CRUTCHFIELD, A m ERNEST R. HILGARD~.5 Stanford University

Abstract: Ss were 15 highly hypnotizable college students capable of performing a manual task (key-pressing according to a simple pattern) following a posthypnotic signal, with the task out of awareness (‘‘subconscious”). The “subconscious” task was sometimes performed alone, while S was presumably resting, and sometimes simultaneously with color-naming, always “conscious.” Control conditions included “conscious” key-pressing, both alone and simultaneous with colornaming. Most of the effects were in errors in key-pressing beyond those in the “conscious” condition alone: when “conscious” pressing was attempted simultaneously with color-naming; equally when “subconsciousJ’key-pressing w s performed alone; by a greater amount when performed “subconsciously” with color-naming. Maintaining dkociation apparently requires cognitive d o r t , hence adding to task interference. Individual differences were noted in departures from the average findings, with no evidence for performances of any Ss improved substantially by hypnotic dissociation.

The theory of dissociation has been out of style for many years, although the term dissociation has been used descriptively in discussing such familiar hypnotic phenomena as amnesia and analgesia. The claasical theory arose largely through studies of automatic writing and of multiple personality. Dynamic psychology, especially in its psychoanalytic version, showed little interest in these phenomena, and, except Manuscript submitted October 3, 1974; final revision received February

14,

1975. 1 This study was supported by Grant MH-3859 from NIMH to E. R. Hilgard and an Ontario Mental Health Foundation Traveling Fellowship to V. Jane Knox. * An earlier version of this paper was presented at the annual meeting of the Society for Clinical and Experimental Hypnosis, Newport Beach, California, December 1973. * Now at Queen’s University, Kingston, Ontario. 4 Reprint requests should be addressed to E. R. Hilgard, Hypnosis Research Program, Department of Psychology, Stanford University, Stanford, Califonia

94305. 5 The authors are grateful to Arlene H. Morgan for her assistance in the data analysis and the preparation of this report. 305

Downloaded by [Memorial University of Newfoundland] at 02:29 13 November 2013

306

KNOX, CRUTCHFIELD, AND HILGARD

for occasional reports of multiple personality, the evidence has remained largely that of the 19th century. Until recently, the major theoretical &cuasion of the problem (30years ago) ended with the conclusion that dissociation was not a useful concept by which to explain hypnotic phenomena (White & Shevach, 1944). Lately, however, somenewlaboratory hdings have led to a reexamination of the concept, and to the propo&g of a neodieaociation theory to explain not only hypnotic phenomena but some related aspects of cognitive activity as well (Hilgard, 1973a, 1973b, 1974; &OX, Morgan, & mgard, 1974). Hull’s students and collaborators performed a few experiments on interfering tasks a number of years ago, with one task out of awaxeness, but they interpreted any evidence of task interference as denying dissociation (Messerschmidt, 1927; Mitchell, 1932). Although Sears (1936), a f b r reviewing the evidence, felt that a modernized approach to disaociation might prove productive, he did not follow up his own suggestion. Stevenson (1972), working in our laboratory, 888e888d task interference in a dissociation paradigm, extending some unpublished earlier work of Cass (1942). The tasks consisted of color-naming, in which colored squarea mounted on a chart were named one after another, always with full awareneas of what was being done, and automatic writing in the form of arithmetical computation, such 88 successive addition of sevens to an assigned beginning number. The writing hand was concealed from view in a box, and S, while engaging in the task automatically, was unaware of what he was doing. The automatic writing was suggested while S was hypnotized, but was carried out as a posthypnotic suggestion to a signal for which S was amnesic. Stevenson demonstrated that two kinds of interference occur when a “conscious” task is performed simultaneously with a “subconscious” task. The first kind of interference he interpreted as owing to the cognitive cost involved in keeping the task “subconscious” over “conscious” performance of that same task.b The second cost he interpreted as due to doing two things a t once, so that the tasks are further impaired when they are done simultaneously with one of them out of awareness. He found the amount of impairment to be a function of the difficulty of the tasks; the more difficult computation task produced more interference whether it was performed consciously or subconsciously. Performances of paired tasks by Ss simulating hypnotic dissociation were less interfered with than by the truly hypnotized, except in the easiest conditions, when there was no ditference between them. The quotation marks around “subcowciow” and “conscious” will hereafter be dropped because these terms refer here to experimental conditions and not to any special theory about alterations in the states of consciousness.

Downloaded by [Memorial University of Newfoundland] at 02:29 13 November 2013

TASK INTERFERENCE AND DISSOCIATION

307

The present experiment was designed to reveal the h e r aspects of this interference. A voice record of the conscious task (color-naming) was obtained on a polygraph that also recorded performance on the subconscious task, in this CW, pressing two keys in alternation (three to the left, three to the right, and so on). Through the obtained records, the nature of the interference could be examined both in terms of latency relationships and of errom made.

METHOD

Subjects The Ss were 15 highly hypnotizable Stanford undergraduates selected from those scoring 9 or 10 on a modified version of the Harvard Group Scale of Hypnotic Susceptibility, Form A of Shor and E. Orne (1962) which had 10 items,’ and maintahhg their high scores on a more cognitive scale, the Stanford Hypnotic Susceptibility Scale, Form C of Weitzenhoffer and Hilgard (1962). They repreaent approximately the upper 5 % of hypnotic responsiveness in this student population. Prior to participation in the experiment proper, they had an opportunity to practice posthypnotic automatic key-pressing with amnesia for the cue which led to the key-pressing and to the activity of key-pressing. The Ss were not retained unless they could meet the requirement for amnesia regarding automatic key-pressing; of 20 Ss invited to participate in the experiment, 4 were dropped because they continued to be aware of keypressing when it should have been automatic, and 1 waa dropped because of an uneasiness evoked by the dissociated experiences which made him prefer not to continue.

Practice Session On the practice day, S became familiar with the two tasks, keypressing and color-naming, to be described below. He first practiced both tasks singly and then simultaneously with both conscious. Following hypnotic induction, in which posthypnotic key-pressing out of awareness was suggested, S was dehypnotized, and the posthypnotic performance tested according to the implanted signal. An inquiry was conducted to ascertain whether or not the task had been performed without awareness. When there was any uncertainty, further practice was carried out. If success was not achieved, S was dismissed. Once successful, the required simultaneous task was also attempted, conscious color-naming 7 The two items omitted of the 12-item 1IGSHS:A were: item 1. Head falling and item 10. Eye catalepsy.

308

KNOX, CRUTCHFIELD, AND HILGARD

Downloaded by [Memorial University of Newfoundland] at 02:29 13 November 2013

with subconscious key-pressing. If all was succeBsful, S was accepted for the regular experimental session to follow on a later day. E x p e r i d Session The experiment proper waa conducted in a single d o n following the practice pend. The order of events is given in Table 1. Because the measurement periods were uniformly of 1-minute length, the whole could be completed comfortably within 1 hour, including time for a post-experimental inquiry. For each S a mean could be computed of 4 periods of color-naming alone, equally spaced at beginning and end, of 4 periods of conscioue key-pressing alone, and of 4 of subconscious keyp r d g , together with 4 periods of simultaneous tasks, half with keyp d g conscious and half subconscious. Although it was intended that there be an equal number of 8 s in TABLE 1

ORDEROF CONDITIONS. ON EXPERIMENTAL DAY Order B

Order A

Single T a s b

Simultaneous Tasks

Single Tasks

Key-pressing (Cs) Color-naming (Cs) Key-pressing (Scs) Color-naming (Cs) Key-pressing (Sca) 6. Key-pressing (Cs)

1. 2. 3. 4. 5.

7. Color-naming (Cs) ; Key-pressing (Cs) 8. Color-naming (Cs); Key-pressing (Scs) 9. Color-naming (Cs) ; Key-pressing (Scs) 10. Color-naming (Cs) ; Key-pressing (Cs) 11. Key-pressing (Cs) 12. Key-pressing (Scs) 13. Color-naming (Cs) 14. Key-pressing (Scs) 15. Color-naming (Cs) 16. Key-pressing (Cs)

Key-pressing (Cs) Color-naming (Cs) Key-pressing (Scs) Color-naming (Cs) Key-pressing (Scs) 6. Key-preseing (Cs) 1. 2. 3. 4. 5.

7. Color-naming (Cs) ; Key-preming (Scs) 8 . Color-naming (Cs) ; Key-pressing (Cs) 9. Color-naming (Cs) ; Key-pressing (Cs) 10. Color-naming (Cs) ; Key-pressing (Scs) Key-pressing (Cs) Key-pressing (Scs) Color-naming (Cs) Key-pressing (Scs) Color-naming (Cs) 16. Key-pressing (Cs) 11. 12. 13. 14. 15.

Note.-The 8 had been hypnotized and dehypnotized prior to the initial single tasks in order t o reinstate the posthypnotic instructions, and also prior to the simultaneous tasks, and prior to the final set of single tasks. * Cs = Conscious, Scs = Subconscious.

TASK INTERFERENCE AND DISSOCIATION

309

Downloaded by [Memorial University of Newfoundland] at 02:29 13 November 2013

order A and in order B, it turned out because of the dismissals that of the 15 Ss in the finalsample, 9 were tested in order A m d 6 in order B. However, no significant differences were noted aa a consequence of order, so that the two orders are not separately treated in the malyses to follow.

Tasks and InstructiOns The colors to be named were presented on a cardboard display panel with small square patches of eight distinctive colors mounted in 16 rows of 16 patches each. To avoid losing a place by too long a line of colors, there waa a space after the eighth color in a row, and a larger space between the eighth and ninth rows. The whole had the appearance of four larger square8 composed of.color patches with a white crosa centered among them. The center of the crom provided a fixation point during those triah in which colors were not to be named. The chart wiw symmetrical but the individual colors randomized, so that by rotating the chart 90" between trials four orders were used for the naming. The S identified the colors by reading from left to right one full row a t a time, as in the reading of a book. Each trial was 60 seconds long. The S waa instructed to find a congenial rate, to move along steadily, but not to hurry because accuracy was desired and not speed. A voice record on a polygraph indicated the timing of each color name. The E kept a record of accuracy. Under these circumstances, SS were remarkably consistent in their repeatsd performances over time, with slight practice effects. For example, the first two trials of color naming yielded an average of 72 colors named per minute, the last two an average of 74, with a range of 45 for the slowest S and 104 for the most rapid S, with a correlation of .89 between the two rates. Because of the high correlation, this small gain with practice proved to be statistically significant (p = .01). For the key-pressing task, S placed his hand in a box which was out of his line of e o n and in which two keys were mounted. The S rested the first two fingers of his right hand on the two keys. His task waa to press three times on the left key, then three times on the right key, and 80 on,three to the left, three to the right. Again, each trial was 60 seconds long and S was instructed to strive for accuracy and not speed. As with color-naming, Ss tended to adopt a congenial rate. The keys were connected to a polygraph so that a record of each individual key-press was provided. For purposea of tabulating results, the key-pressings were counted as units, 1 unit being two triplets of pressing, that is, 3L-3R. A unit was in error if there waa an extra or omitted press, for example, 2L-3R, 3L-4R, etc. The rate of key-pressing, like that of color-naming,

Downloaded by [Memorial University of Newfoundland] at 02:29 13 November 2013

310

KNOX, CRUTCHFIELD, AND HILGARD

wa8 q i t e consistent. The first two trials of key-pressing yielded a mean of 26 wts of 3L-3R in 1 minute, and the last two trials a mean of 28 sets, with a range from 17 for the slowest S t o 44 for the most rapid S , with a correlation of .94between the two rates. The slight gain was significant ( p = -05). Because the automatic key-pressing was to be done to a signal t h a t would produce the response posthypnotically, and out of awareness, either singly or in combination with color-naming, the specific instructions were carefully prepared and presented verbatim. T h e instruction for the initial posthypnotic single taak (subconscious key-pressing) was as follows, Shaving first been hypnotized:

Please place your hand in the box, with your fingers on the keys. After you awake, 1 shall say: “You may rest, but remain alert.” I want you then to look straight ahead at the center of the cross that is formed by the colon. You will feel as though you are having a little rest between trials, but you will remain alert, and when you hear me say BEGIN 3’s your fingers will begin to tap out alternately 3 to the left and 3 to the right, and will continue to do this until you hear the timer shut off a t the end of a minute. You will have no awareness of what your hand is doing. Your hand will tap accurately according to the instructions, and will not hurry. You will pay no attention to what your hand is doing, and will not even know that it is tapping out 3’s. Later on, when you are aroused from hypnosis, you will not remember that you have received these instructions, or even that you were hypnotized, until I say, “Now you can remember everything.” Please tell me now what you are to do after I arouse you from hypnosis and say: “You may rest but remain alert, BEGIN 3’s.”[If S’s replies are incomplete, repeat instructions as necessary.] That’s fine. Note that the word BEGIN always means key-pressing without awareness. Now I will arouse you by counting backwards from 10 to 1 and you will be fully aroused but you will have forgotten what we have done or even that you were just hypnotized, and you will carry out the instructions quite automatically. 10, 9, 8, . . . 1. How do you feel? Because of t h e several conditions, the instructions were necessarily somewhat complex; fortunately, the necessary discriminations turned out t o b e acquired readily by these highly responsive hypnotic Ss.8 The five different tasks, with the verbal cues which introduced each of them, are listed in brief in Table 2. In addition to some contextual statements, the main differentiations mere between GO, for color-naming alone, START 3’8, whenever the 3L-3R key-pressing was t o be done consciously, and BEGIN 3’5, whenever the key-pressing was t o be done subconsciously, whether alone or simultaneously with color-naming.

* Complete protocols

may be obtained from E. R. Hilgard.

TASK INTERFERENCE AND DISSOCIATION

311

TABLE 2 TASKS TO

BE PERFORMED Ah‘D VERBAL CUES TO INITIATE

Downloaded by [Memorial University of Newfoundland] at 02:29 13 November 2013

Tssks

Conscious color-naming alont Conscious key-pressing alone Subconscious key-pressing alone Conscious color-naming with conscious key-pressing Conscioua color-naming with subconscious key-pressing

THEM

Verbal Cues

“Please name the colors. GO.” “No color naming this time. START 3’s.” “You may rest, but remain alert. BEGIN 3’s.’’ “Please name the colors. START 3’s.” “Please name the colors. BEGIN 3’s.”

These cues were acquired readily during the practice session, and operated with little difficulty in the experiment proper.

RESULTS When simultaneous tasks are employed, as in this investigation, the tasks are mutually interfering, hence in presenting the results, attention will be given first to the interference of key-pressing upon color-naming, followed by some interferences within key-pressing alone, and finally by interference of color-naming upon key-pressing.

Color-Naming as Interfered with by Key-Pressing The chief etTect of interference by simultaneous key-pressing was on speed of color-naming rather than on accuracy. Because so few errors were made in any condition (a mean of 0.2 errors per minute with colornaming alone, and a mean of only 0.5 errors per minute in the most interfering condition) it is possible to neglect the errors without distorting the results. The rate of color-naming decreased from a mean of 73 f 15 colors per minute to 53 f 18 colors per minute when conscious key-pressing interfered, and decreased to 59 f 21 colors per minute when subconscious key-pressing interfered. An analysis of variance for the three conditions (color-naming alone, and the two interfering conditions) yielded a significant F ratio of 14.13, df 2,30; p = .001,justifying t tests of the differences. The decrease in rate of color-naming when conscious key-pressing interfered was significant at p = .001, and when subconscious keypressing interfered was significant at p = .01 (all t tests two-tailed). Although there wm a slight tendency for color-naming to be more rapid when interfered with by subconscious key-pressing than when interfered with by conscious key-pressing, the dif€erencedid not reach the .05 level

312

KNOX, CRUTCHFIELD, AND HILGARD

Downloaded by [Memorial University of Newfoundland] at 02:29 13 November 2013

of significance. The results may be summarized as indicating that simultaneous key-pressing produced a decrease in the rate of color naming of about 2Q or 30% without any substantial increase in errors, and not significantly m e r e n t for the two conditions of key-pressing. Key-Pressing Rates and Errors Under Four Conditions

The four conditions under which key-pressing was performed were: alone, conscious; alone, subconscious; simultaneous with conscious color-naming, key-pressing conscious; and simultaneous with conscious color-naming, key-pressing subconscious. Two analyses of variance were conducted of the same type as for color-naming, one for rate of keypressing, one for the percent of errors. By expressing the errors in a percent there is a correction for differences in the number of key-pressings attempted in the 1-minute trials. Treatments proved significant in both cases (p = .Ol), so that comparisons of the conditions by t tests are permissible. The modification in rate and errors will be presented according to the conditions which are to be compared. Changes between mciolls and submciuua key-pressing. The change from conscious to subconscious key-pressing modified the performance, independent of any secondary task. The number of key-pressing units attempted in the conscious condition averaged 27 f 6 per minute, and in the subconscious condition averaged 29 f 11 units,a nonsignificant mean increase of 2 units attempted. This small average increase was contributed to by a few Sa who greatly speeded up their key-pressing in the subconscious condition, a t the same time making many errors. E R O between ~ the conscious and subconscious condition increased from 1.7 f 2.4 percent-errors in the conscious condition to 13.1 f 18.8 percent-errors in the subconscious condition, a significant increase of 11.4 % ( p = -05). The rather large individual differences require a further analysis to be considered after the remaining major findings have been presented. Key-pressing as interfered with by color-naming. It has already been noted that when key-pressing was combined with color-naming the rate of color-naming was reduced, with little effect upon errors. This occurred whether the key-pressing was conscious or subconscious. The reciprocal effect, for conscious key-pressing interfered with by color-naming, showed as both a decrease in rate of responding and an increase in errors. From the normal rate of key-pressing alone of 27 f 6 units per minute the rate declined to 21 units, a significant decrease (p = .001). The percent of errors increased from 1.7 f 2.4 to 13.1 f 20.1 ( p = .05). It may be noted that the increase in errors due to conscious task interference is just equal to the increase in errors for key-pressing alone when tlhe change was from conscious to subconscious pressing.

Downloaded by [Memorial University of Newfoundland] at 02:29 13 November 2013

TASK INTERFERENCE AND DISSOCIATION

313

The crucial question is whether subconscious key-pressing was any freer from interference than when it was conscious. If rate done were considered, this would seem to be the case, because the normal rate wm sustained with subconscious key-pressing while color-naming. The normal rate for key-pressing alone in the conscious state, 27 f 6 , dropped only to 25 f 10 Units in subconscious pressing when combined with color-naming, a change not significant, and the final rate wm si&cantly higher than when the simultaneous tasks had both been conscious ( p = .05). However, this advantage for the subconscious condition was gained at the expense of errors in key-pressing. When the simultaneous tasks had been performed with key-pressing conscious, there had been 13.1 f 20.1 percent-errors, but when key-pressing waa subconscious these increased to a mean of 27.2 f 32.3%. Because of the very high variability, this increase was only marginally significant ( p = .07). To summarize these somewhat complex findings, it may be indicated that the mean percent errors in key-pressing increased with each complication, first from conscious to subconscious key-pressing, when performed alone. The change from conscious to subconscious produced about the same increase in errors as WBS found when the change was from conscious key-pressing alone to such key-pressing combined with conscious color naming. Finally, the most complex condition waa that in which conscious color-naming was combined with subconscious keypressing, and this brought the largest number of errors, though with high variability.

Individual fiflerences in Tmk Interference The great variability of results led to marginal statistical significances in some instances, raising the question whether mean results may not have concealed some important individual differences. Because a major issue is whether or not interference can be reduced when a task is out of awareness, it is important to examine the data to see whether or not there were some cases in which the results were opposite from the generality. To permit an examination of the results for individuals, the keypressing responses under the several conditions are presented in Table 3. To facilitate comparative examination of the individual data, Ss have been arranged in order of the errors they made on the most complex task, and they have been subdivided into two groups, Ss of one group having made fewer errors (means from 0 to 4%) and the other more errors (means between 18 and 90%). Data for the first and second group are presented in Table 3. The mean results for the low-error group show practically no interference with the accuracy of key-pressing as a result of the usually disruptive conditions. Holding key-pressing subconscious did not change

314

KNOX, CRUTCHFIELD, AND HILGARD

Key-PressingCombined with Color-Naming

Key-Pressing Only

Downloaded by [Memorial University of Newfoundland] at 02:29 13 November 2013

Subjst

Conscious

Subconscious

C

~

O

U

subconscious

S

% E m a s Number I%Ermrr

10 11 6 4 7 9 15

29

30 21 33 24 25 25 26

21 36 25

29 16 27

1 2 0

14 11 22

1 0 2 4

23

14 10 30 22

20

28

24 26

21 25

Ss Making More Errors 2 13 3 5 1 12 14 8

30

8

29

24 19

0 0 2 3 0

24 23

x

27.6

36 42 25 17 28

0

4

2.1

I

17 0 2

30.9

67 17 31 26

I

3 8 6

19 27 31 24 12 22

28

41 57 28 32 13

16

20

68 49 24 0 21

-~ 23.5

1

21.4

I

22.4

I

17 22 26 34 54 27 30 18

18 20 23 52 54 59 85

28.5

50.1

90

Note.-The Ss have been arranged in order of increasing errors in the last column.

either rate or errors. Although there was a slight reduction in the rate of key-pressing as a consequence of simultaneous color-naming (from 26.3 to 20.0),there waa no added &ect as a consequence of the hypnotic subconscious key-pressing condition. Examination of the data for colornaming (not in the table) showed that there had been no trade-off between key-pressing and color-naming; that is, there was the usual reduction in the rate of color-naming 89 a consequence of pairing with keypresing, with no increase in color-naming errors. The Ss whose keypressing data appear in the upper part of Table 3 reduced their rate of color-naming with conscious key-pressing to 70 %I of their rate for colornaming alone; with subconscious key-pressing the rate was 77% of the base rate. The Merences are not significant.

Downloaded by [Memorial University of Newfoundland] at 02:29 13 November 2013

TASK INTERFERENCE AND DISSOCIATION

315

The hope of finding a few Ss whose behavior fits the classical interpretation of noninterference when a task is subconscious lies in this subfraction of the sample, where mean interferences are far below those reported for the whole sample. Noninterference among conditions appears to be the rule, rather than improvement owing to one of the task being subconscious. The possibility of improvement is limited because these Ss made few errors in the simultaneous tasks when both tasks were conscious, so that there waa little room for improvement in the subconscious condition. Four SS(Ss 9, 10, 11, and 15) reduced errors when key-pressing was subconscious, without decreasing their rates of performance beyond what they were in the conscious simultaneous tasks. The decreases in percent-errors were, however, only from 5 to 2 , 6 to 4, 4 to 0, and 4 to 0 for these Ss. While not impressive, at least those who made no errors did as well as they could have done. Also,Ss 6 and 7 both increased their rates of key-pressing simultaneously with color-naming when key-pressing was subconsciow, without increasing errors; this is what the classical interpretation would predict. For both, the rate of keypressing in this normally most complex condition returned to about the level found With conscious key-pressing alone. These results leave somewhat open the question about the circumstances in which subconscious performances may be noninterfering, or even advantageous in reducing interference. It is evident from the few errors made by these Ss in the simultaneous performance of both conscious tasks that this was easier for them to do than for Ss whose data are presented in the lower part of Table 3. Perhaps when tasks are easy enough the extra effort of maintaining dissociation will not cancel such benefits as it may bring. Because we did not alter objective difficulty in this investigation, the matter of dficulty has to be inferred from the daerential performances of Ss. The findings that led to the significant overall results are clearly attributable to the eight Ss of the high-error group, whose data are also presented in Table 3. Their mean results confirm in all respects those that have previously been mentioned as the general findings,except that the interferences are exaggerated because the noninterference Ss have been eliminated. Examination of the color-naming data showed the rate of color-naming to be only marginally different from that for the lowerror group, so that the differences between the high-error SS in keypressing and the low-error Ss does not rest upon a difference between them in the division of attention between colors and keys. This group reduced the rate of color-naming with conscious key-pressing to 77% of the original rate and with subconscious key-pressing to 84% of the original rate, not significantly different for the two conditions. The main burden of the interference waa carried by the key-pressing task.

Downloaded by [Memorial University of Newfoundland] at 02:29 13 November 2013

316

KNOX, CRUTCHFIELD, AND HILGARD

The data have been explored in other ways, Using regression analysis, but in view of the large variabilities most of the analyses have turned out uncertain in their implications and w i l l not be presented. It may be that the very fact of individual differences in the face of competing tasks was reapomible for the large variability under codictual conditions, despite the consistency of performance on the individual tasks. One way of gaining further information from the data is to classify Ss according to the number of conditions in which they reacted in accordance with the mean findings. For example, only 1 S (S 8 ) Showed clear deterioration of both k e y - p r b g and color-naming with each increase in complexity of conditions, that is from conscious key-pressing to subconscious key-pressing, from single color-naming to color-naming combined with conscious key-pressing, and additionally when subconscious k e y - p d g accompanied color-naming. As already noted, 7 Ss showed no appreciable changes in errors in key-pressing under the several conditions; of these, however, 5 reduced their rate of color-naming when conscious key-pressing and color-naming were undertaken simultaneously. Four of these 5 showed no further reduction in the speed of colornaming when key-pressing wm subconscious. For these Ss, the strategies used in simultaneously performing the two conscious tasks had more d e c t on the color-naming than the subconscious component did. In this connection, it is of interest that of the 11 Ss who reduced their colornaming rate when it waa combined with conscious key-pressing, 10 did not reduce their color-naming any further when key-pressing became subconscious. Thh confirms the previous statement that the main effects of the interferences were showri through errors in key-pressing rather than in a disruption of color-naming. This completes the presentation of the major quantitative data, but some further analyses are required to gain more understanding of what Ss were doing.

Further Analyses The purpose of the further analyses is to determine, if possible, the nature of the task interference which was involved under the various conditions, the strategies that were used to reduce the interference, and the manner in which hypnosis affected the results. Sources of Inhference. There were two main components to the task interferences: holding key-pressing subconscious and performing two tasks simultaneously. These appeared to have almost equal effects, as judged by the incresses in errors of key-pressing. When the two effects were combined (subconscious key-pressing with conscious color-naming) the mean effect on errors seemed to be additive, the increase in errors

Downloaded by [Memorial University of Newfoundland] at 02:29 13 November 2013

TASK INTERFERENCE AND DISSOCIATION

317

being the sum of the errom found in making key-pressing subconscious and the interference produced when both tasks were conscious. If one of these effects is independent of hypnosis, and the other dependent upon hypnosis, we might suppose them to be uncorrelated; to determine whether or not this the cam, we return to the original data by mbject, aa presented in Table 3. The errors made in key-pressing alone in the subconscious condition can be correlated with the approximately equal number of errors made in conscious key-pressing when done simultaneously with color-naming. The first of these error rates represents interference owing to the component of holding the task mbconscious, the second error rate is due to the second component, normal task interference, which has nothing particularly to do with hypnosis. These two errors scorea correlate significantly, r = .69 (p = .Ol). Hence, the hypothesis that the two components might be distinctive and uncorrelated is not supported; the simplest interpretation is that some Ss are more vulnerable to interference than others as a situation becomes more complex, and the source of the complexity, so far as represented by the conditions of this investigation, is immaterial. Therefore, when the task is made even more complex, by adding the two sources of interference, the vulnerable Ss are again more affected. In predicting the errors to be made in the most complex task (color-naming with simultaneous subconscious keypressing) two correlations are pertinent. The errors made in subconscious key-pressing alone, which estimate the contribution of the subconscious component, condate .71 (p = .OOl) with the errors made when subconscious k e y - p r d g is combined with conscious color-naming; the errors in conscious key-pressing simultaneous with conscious colornaming, representing the other normal task-interference component, correlate .50 (p = .06) with the errors made when the key-pressing is subconscious in the simultaneous tasks. This correlation does not reach higher significance because of one discordant S, S 14, who made 0 errors when both tasks were conscious and 85 % errors when simultaneous keyp r e g was subconscious. Omitting this S, the two correlations both become .69 indicating that without this one S the two components are equally predictive. Errors in the two components (subconscious component and t.askinterference component) correlate as high with each other as each does with the complex task criterion. Hence it must be concluded that S’s interference-susceptibility is responsible for the errom rather than anything specific to either kind of interference. There are still two components, however, and their effects are additive. Strategies of P&ming Two Tasks Simultanecrusly. The wide indi-

Downloaded by [Memorial University of Newfoundland] at 02:29 13 November 2013

,318

KNOX, CRUTCHFIELD, AND HILGARD

vidual differences reported in Table 3 may depend upon aspects of skill having little to do with hypnosis, such as the problem solving ability that results in organizing the two tasks 80 as to make their simultaneous performance easier. If the total task is easier, there should be leas problem in carrying it out with one of the components out of awarenw. Because the voice records and key-pressing were recorded on a common polygraph record, some inferences regarding strategies can be made. Six strategies can be ideneed, all involving an integration between the key-pressing and the color-naming. One strategy was used most commonly and rather mcceasfully by 8 of the 15 Ss. Because some decision-time is involved in naming colors, emh of these 8 SEapparently selected the color name, and then held it in readiness to announce simultaneously with key-pressing, for the record showed that he preased a key once, then a ~ o u n c e dthe color and pressed the remaining two of the triplet, for example, 3L. Apparently deciding on the next color while permitting the relatively automatic act to continue, he turned to the first press of the next triplet, 3R, pressing once, announcing the color name, and then continuing the two additional presses. If this integration was performed consistently, the result was that two colors would be named for each full set of six key-presses, a ratio that was often found. The next most often used strategy, used by 3 of the 15 subjects, alternated color-naming and key-pressing, so that for one unit of six keypresses there would be six colors named. The consequence wm a slower rate of key-pressing when tasks were simultaneous than when keypressing was done alone. The other strategies were used by only 1S each: naming the color between triplets, rather than simultaneously with pressing, naming the color after having pressed the keys twice, and naming a color a t each key-press rather than between them. When both tasks were comcious, Ss tended to use the preferred strategy that each had chosen a mean of 85% of the time. When the key-pressing was subconscious, the same strategy was used a mean of 59 % of the time, owing to some break in the integration between the tasks. Four Ss maintained a common strategy consistently throughout all of the tasks. The three 5 s who used the strategy of one color name for every three key-presses were the most successful in reducing errors throughout (Ss 4,6, and 7). The one S who invariablyused thestrategy of attempting t o name a color between each single key-press greatly reduced her rate of key-pressing in order to do this and made numerous errors (S 2). It is very possible that had she used another strategy her results might have been different. However, the most successful S (5 10) also used this strategy most of the time, without making errors. It does not follow, however, that the adopted strategy accounts for all of the individual

Downloaded by [Memorial University of Newfoundland] at 02:29 13 November 2013

TASK INTERFERENCE AND DISSOCIATION

319

d8erences; the successful strategy of SS4,6, and 7,m referred to above, was also used by S 9 but not invariably, though he made few errors, and by Ss 5,8,12, and 13, all of whom made many errom. The effects of various gtrategies on gimultaneous task performance would be an intsresting study in itself, but it does not appear to help very much in the interpretation of these data. Hypnotic Dissocidion from S’s StandpOint. All of the analyses which have been presented thus far have been made on the assumption that the instructions were appropriately carried out and that a subconscious condition was indeed one of which S was totally unaware. Although those Ss who could not come up to expectation during the practice .session were excluded, it does not follow that all the requirements of the experiment were fulfilled, so that it is important to examine the results of the inquiries which were conducted within the experimental day. At the end of each of the three sets of trials on the experimental day (single tasks,simultaneous tasks, single tasks again), S was asked about what he had been doing, then he was released from amnesia and questioned further. Before going ahead he was quickly rehypnotized, the posthypnotic suggestions again implanted, and then the experiment proceeded. There was no evidence that these interim questions affected the subsequent performances. All questioning by E, and the replies by S, were tape recorded. The main question to be answered for the single tasks is what happened when S was told to rest but actually was given the signal for subconscious key-pressing. Of the 15 Ss, 11 indicated absolutely no awareness that they were pressing the keys, although 3 of the 11 occasionally heard clicking sounds that they did not know how to account for. Of the 4 Ss who were somewhat aware of key pressing, 2 knew dimly that they were doing something With their right hand, but they were unclear about it and not paying attention. Two others showed poor dissociation, 1 S feeling somewhat altered from waking, but aware of presaing the keys; the other inferred that she was pressing the keys by interpreting the sounds that she heard. The results, while imperfect, were reasonably satisfactory in showing that most Ss were carrying out the assignment without awareness. The 4 ‘aware’ Ss can be identified in Table 3 as Ss 1, 8, 14, and 15. Of these, 3 were the highest errormakers in simultaneous key-pressing and only 1 (5’ 15) was in the lowerror group. The reaction to naming the colors while simultaneously key-pressing subconsciously showed no consistent pattern. Only two Ss noticed that. it was harder to name the colors alone (that is, while pressing keys sub-

Downloaded by [Memorial University of Newfoundland] at 02:29 13 November 2013

320

KNOX, CRUTCHFIELD, AND HILGARD

consciously) than while consciously tapping the keys, while four thought it was easier when they were naming the colors alone. These comments were unrelated to the quantitative findings, however, since three of the four who found it easier made considerably more errors. The rest of SS reported the color-naming as generally easy or generally hard, without discriminating between conditions. None of Ss had difficulty in recalling the instructions under which they had operated, once amneaia was removed, although several of them spontaneously remarked how weird it seemed that they were able to do what WM suggested. Those who were less succea~fuloften made very illuminating remarks about what it meant to try to do what was expected, and how d B c u l t it was.

DI~CUSSION The quantitative results showed clearly that tasks of some measure of complexity can be carried out simultaneously, even though one of them is out of awareness. However, such dissociation from awareness does not mean that the tasks do not interfere. We found, in fact, two main sources of interference: first, the interference produced when a task practiced in the conscious condition had to be carried out subconsciously, and second, the interference produced when both tasks were carried out consciously but simultaneously. Under the conditions of the experiment, these two sources had approximately the same mean effects. When one of the simultaneous tasks was kept out of awareness, the two sources of interference operated at once and summed to produce the maximum interference. The inference may be drawn that each of the added requirements causes more cognitive effort, whether the effort is that of holding a task out of awareness, or attempting to do two things a t once. If cognitive capacity is Limited, it may be expected that the greater the demands made upon S the more the required performances will deteriorate. Interference between dissociated tasks contradicts a form of dissociation theory which would posit an absolute. separation of the tasks, but it is coherent with the neodissociation theory that we favor. This theory accepts the fact that the tasks may be dissociated from a conscious aspect-ne of them not being available to conscioUsnes?-while they may still interfere at some other level. Such a conception is comparable, for example, to the psychoanalytic theory of repression as an active process, which may indeed keep some memories out of awareness, but not without demonstrable cognitive ef€ort.Such effort may be revealed, for example, in the rituals of compulsive behavior that can be interpreted as maintaining repression. This does not mean that the mech-

Downloaded by [Memorial University of Newfoundland] at 02:29 13 November 2013

TASK INTERFERENCE AND DISSOCIATION

321

anisms of dissociation are always those attributed to repression, but the interaction Of conscious and nonconscious material evident in both dissociation and repression. The possibility exists that these generalizations do not hold for all hypnotizable Ss. Within the sample studied there were those for whom t,he more complex tasks did not produce interference, but there was only marginal indication that any Ss were helped in performing simultaneous tasks by holding one of them out of awareness. The fact that even making the single key-pressing task subconscious tended to interfere with its accuracy is coherent with the position that hypnosis is an active process, requiring cognitive effort on the part of the hypnotized S. We have inferred this to be the case on the basis of other observations, such 88 the aroused heart rates and blood pressure increases in anticipation of a difficulthypnotic task, such m maintaining analgesia to circulating ice water (Hilgard, Macdonald, Marshall, & Morgan, 1974). Some Ss appear to have the ability to carry out simultaneous tasks very well outside of hypnosis, and there is evidence that this ability played a part in producing the variability found in our results. The source of this ability is not known to us, but the ability may have been acquired, a8 through practice in playing a musical instrument. We did not Show hypnosis to generate this ability, although use could be made of it within hypnosis. When the hypnotic S is required to hold out of awareness a task which for him would ordinarily interfere with a simultaneous task, the interference may be magnified by the effort required to maintain the dissociation.

REFERENCES CAE&W.A. An experimental investigation of the dissociation hypothesis, utilizing a post-hypnotic technique. Unpublished master’e thesis, Univer. of Oregon

1942. ~ L Q A E D ,E.

R. A neodissociation interpretation of pain reduction in hypnosis. Psychol. Rev., 1973, 80, 396-411.(a) BLOW, E. R. Dissociation revkited. In M. Henle, J. Jaynes, & J. Sullivan (Fds.), Historical conceptions of psychology. New York: Springer, 1973. Pp. 205-219. (b) HILQARD, E.R. Toward a neo-dissociation theory: Multiple cognitive controls in human functioning. Perspect. Biol. Med., 1974,17,301-316. HILOARD,E.R.,MACDONAID, H., MAREHALL, G., & MORQAN, A. H. Anticipation of pain and of pain control under hypnosis: Heart rate and blood pressure responses in t h e cold pressor test. J . abnorn. Psychol., 1974, S 8 , 56-568 KNOX,V. J., MORGAN, A. H., & HILGARD,E. R. Pain and suffering in ischemia: The paradox of hypnotically suggested anesthesia aa contradicted by reports from the “hidden observer.” Arch. gen. Psychiat., 1974,S O , 840-847. MEEEERSCHMIDT, R . A quantitative investigation of the alleged independent

Downloaded by [Memorial University of Newfoundland] at 02:29 13 November 2013

322

KNOX, CRUTCHFIELD, AND HILGARD

operation of conscious and subconscious processes. J. abnornz. S O C . Psychol., 1927,$9, 32.5-340. MITCHELL, M.B. Retroactive inhibition and hypnosis. J. gen. Psychol., 1932, 7 , 343-359. SEARS,R. R. Functional abnormalities of memory with special reference to amnesia. Psychol. Bull., 1936,SS, 229-274. SHOR,R. E., & ORNE,E. C. Harvard Group Scale of Hypnotic Susceptibility, Form A . Palo Alto, Calif. : Consulting Psychologists PM, 1962. STEVENSON, J. H. The effect of hypnotic and posthypnotic dissociation on the performance of interfering tasks. (Doctoral h e r t a t i o n , Stanford Univer.) Ann Arbor, Mich.: Univer. Microfilms, 1973. No.73-4601. WEITZENHOFFER, A. M., & HILQ~RD, E.R. Stanford Hypnotic Susceptibility Scale, Form C . Palo Alto, Calif.: Consulting Psychologists Prees, 1962. WHITE, R. W., & SHEVACH, B. J. Hypnosis and the concept of dissociation. J . A b n m . 8oc. Psychol., 1942,56,309-328.

Die Natur der Aufgabeinterferenz bei hypnotischer Dissoziation: Eine Untersuchung des hypnotischen Verhaltens

V. Jane Knox, Lila Crutchfield und Ernest R. Hilgard Abstrakt: Die V p w . waren 15 h i h s t hypnotisierbare Studenten, die in der Lage waren, eine manuelle Aufgabe (das Herunterdriicken von Tasten im Rahmen eines einfachen Musters) als Folge einea posthypnotischen Signals, dessen Befehl ausserhalb des Bewusstseins (“Unterbewusatsein”) lag, durchzufaren. Manchmal wurde die “unterbewusste” Aufgabe allein ausgefiihrt, wiihrend die V p . scheinbar ruhte, und manchmal zu gleicher Zeit mit dem Nennen von Farben. dies h e r “bewusat”. Kontrollbedingungen schlossen “bewusstes” Tastendriicken entweder allein oder mit zugleichem Farbennennen ein. Die meisten m e k t e bestanden in Irrtiimern des Tastendriickens jenaeits der in der allein “bewussten” Bedingung: wenn “bewusstes” Driicken zu gleicher &it mit Farbennennen versucht wurde, ebenso. wenn “unterbewusstes” Tastendriicken allein durchgefiihrt wurde und in griisserem Masze, wenn es “unterbewusst” mit Farbennennen ausgefiihrt wurde. Urn eine Dissoziation aufrecht zu erhalten. wird scheinbar eine kognitive Anstrengung beniitigt. die daher zur Aufgabeinterferenz beitrfgt. Individuelle Unterschiede konnten in Abweichungen von durchschnittlichen Befunden beobachtet werden, doch fand sich kein Beweis dafiir, dass die Ausfiihrungen von irgendeiner der Vpn. substantiell durch hypnotische Dissoziation verbessert wurdea.

La nature de I’interfCrence avec la tiche dans la dissociation hypnotique: une recherche sur le comportement hypnotique

V. Jane Knox. Lila Crutchfield, e t Ernest R. Hilgard RksumC: Quinze Btudiants de niveau colldgial participent ri l’exp6rience. Tous sont hautement hypnotisables et capables d’accomplir, de fagon

Downloaded by [Memorial University of Newfoundland] at 02:29 13 November 2013

TASK INTERFERENCE AND DISSOCIATION

323

“subconsciente” une tiche manuelle (appuyer sur un levier selon un s c h h e simple) conskutivement P un signal posthypnotique. La tiche “subconsciente” est parfois exbcut6e seule, alors que le S est prtsumbment nu repos, et parfois en m&metemps que la nomination de couleurs, toujours “consciente”. Lee conditions de contrile incluent la mise en action du levier de faGon “consciente”, thche exkcutte t a n t i t seule, tan& en m i k e temps que la nomination de wuleurs. La plupart des effets se traduisent par des erreurs dans l’action du levier, erreurs qui d6paesent. dans lee conditions suivantes. celles de la condition “consciente” seule: quand la mise en action du levier e t la nomination de couleurs sont effectu6es s i m u l t a n b e n t ; quand la mise en action “subconsciente” du levier eet e d c u t t k seule; et & un d e g 6 plus marqut, quand la mise en action du levier e a t exbcutte de fagon “subconsciente” en m&me temps que la nomination de couleurs. Le maintien de la dissociation requiert apparemment un effort cognitif, ajoutant ainsi & l’interf6rence avec la tiche. L’btude des dS6rences individuelles par rapport aux rtsultats moyens ne fournit aucune preuve P I’effet que le rendement de quelque S ait t t 6 ambliorb de faqon substantielle par la dissociation hypnotique.

La naturaleza de la interferencia en la tarea en la disociaci6n hipn6tica : una investigacidn del comportamiento hipnbtico

V. Jane Knox. Lila CrutchGeld, y Ernest R. Hilgard Resumen: Los Ss son 15 estudiantes de “college” altamente hipnotizables, capacee de realizar una tarea manual (apretar un dispositivo s e d n una pauta sencilla) tras una seiial posthipnbtica. sin que dicha tarea (“sub. consciente”) sea ejecutada conscientemente. A veces, la tarea “subconsciente” es realizada aisladamente, mientras el S est&probablemente descansando; otras veces. va acompaaada de la denominaci6n. siempre consciente, de colores. Las situaciones de control incluyen la presi6n ‘I consciente” de un dispositivo, aisladamente y acompafiada de denominaci6n de colores. La mayor parte de 10s efectos se registran en 10s errores en la preai6n del dispositivo, a1 margen de 10s observados en la situaci6n “Consciente” considerada aisladamente: es decir. cuando la presi6n “consciente” se intentaba a1 mismo tiempo que la denominaci6n de colores; anatlogos resultados se observan en cuanto a la presi6n “subconsciente” del dispositivo. Mantener la disociaci6n. exige aparentemente un esfuerzo cognitivo. que viene asf a sumarse a la interferencia en la tares. Se encuentran diferencias individuales relativas a1 comienzo de la tar-, respecto de la media de 10s resultados, sin evidencia alguna de mejora sustancial, por disociaci6n h i p d t i c a , en 10s rendimientos de 10s

Ss

.

The nature of task interference in hypnotic dissociation: an investigation of hypnotic behavior.

This article was downloaded by: [Memorial University of Newfoundland] On: 13 November 2013, At: 02:29 Publisher: Routledge Informa Ltd Registered in E...
1MB Sizes 0 Downloads 0 Views