- Email: [email protected]
Left hemisphere superiority for event-related potential effects of hypnotic obstruction
Pergamon
0028 3932(95100149 2
Neuropsychologia,Vol. 34, No. 7, pp. 661-668, 1996 Copyright ,~ 1996 Elsevier Science Ltd. All rights reserved Printed in Great Britain 0028- 3932/96 $15.00+0.00
Left hemisphere superiority for event-related potential effects of hypnotic obstruction PAUL JASIUKAITIS,* BITA N O U R I A N I and DAVID SPIEGEL Department of Psychiatry and Behavioral Sciences, Stanford University Medical Center, Stanford, CA 94305, U.S.A.
(Received 13 February 1995; accepted 25 July 1995)
Abstract--Twenty-two highly hypnotizable subjects were run in a visual target detection task which compared hypnotic obstruction of the left and right visual fields over separate blocks. The visual event-related potentials (ERPs) to non-target stimuli revealed that hypnotic obstruction reduced the P200 component to stimuli in the right hemifield, but did not affect P200 for stimulation in the left hemifield. The earlier P100 and N100 were also reduced to hypnotic obstruction but not as preferentially for either hemifield, while the P300 was not significantly changed. Right visual field/left hemisphere P200 reduction predicted suppression of behavioral response (button press) to hypnotically obstructed targets in both hemifields. The results are discussed in terms of Farah's model of a left hemisphere mechanism for image generation, and how highly hypnotizable subjects might use this mechanism to comply successfully with the suggestion of a hallucinated visually opaque barrier. Copyright ,~ 1996 Elsevier Science Ltd. Key Words: visual event-related potentials; hypnotic obstruction; imagery; left hemisphere.
Introduction
gathering over the past decade that image generation engages left hemispheric functioning [20]. In a review of neurological cases, Farah [5] found 12 patients with deficits in image generation but with perception and recognition grossly intact. In these 12 cases the predominant site of brain damage was the posterior left hemisphere. Among neurologically intact volunteers, Farah [6] found that visual image and stimulus match in the right visual field speeded reaction time more than it did in the left visual field, suggesting more efficient image production in the left hemisphere. Relevant to ERP research is the study by Farah et al. [7]. They found that foveally presented word stimuli elicited greater positivity in the visual ERP, from about 600 to 900 msec post-stimulus, when subjects were required to form a concrete mental image of the word's referent as opposed to detecting misspellings. The scalp maximum of this enhanced positivity was over the left occipital and posterior temporal areas, suggesting an important role for the left visual cortex in the generation of images from word stimuli. It has been suggested that hypnosis is a "right brain" phenomenon [14]. However, this view is not firmly established, being based on EEG spectral and lateral eye movement data which are still open to various interpretations [4, 15]. It could be that there is no inherent laterality in hypnosis, but that whatever cognitive faculties {left or right hemisphere) are needed for performance of a given hypnotic instruction will be the ones employed. If so, it might be
It has been reported that hypnotic suggestion can reduce the amplitude of individual peaks in the visual eventrelated potential (ERP) of highly hypnotizable subjects [3, 18]. These two studies utilized hypnotic obstruction, i.e. the suggestion of an imagined opaque object between the subject and the visual stimulus, instead of hypnotic diminution or the suggestion that the evoking stimulus was less intense or not there. The fact that hypnotic diminution has often been used in attempts to demonstrate the effects of hypnosis upon visual ERPs may account for some of the negative results [1, 2, 21]. The suggestion of diminution would remain unconvincing to the subject as long as the stimulus continues to be perceived. The hypnotic production of a competing visual image does not focus attention upon the external percept to be eliminated, but instead focuses it on the internally generated image. In this manner hypnotic obstruction may provide more consistent results than hypnotic diminution [17]. Unlike the suggestion of stimulus diminution, hypnotic obstruction involves image generation. Evidence has been * Address correspondence and requests for reprints to: Paul Jasiukaitis, Room C231, MC 5544, Department of Psychiatry and Behavioral Sciences, Stanford University Medical Center, Stanford, CA 94305, U.S.A. 661
662
P. Jasiukaitis et al./Left hemisphere superiority for hypnotic obstruction
expected t h a t the left hemisphere i m a g e r y m e c h a n i s m described by F a r a h [6] would be used in the p e r f o r m a n c e o f h y p n o t i c visual o b s t r u c t i o n . The following visual h y p n o t i c o b s t r u c t i o n e x p e r i m e n t used visual h a l f field s t i m u l a t i o n as did F a r a h ' s [6] imagery a n d reaction time task. Subjects were instructed to i m a g i n e a solid b a r r i e r covering the visual hemifield o n the left o r right side o f a central fixation point. Visual stimuli occurred r a n d o m l y to the left a n d right o f fixation a n d subjects were required to press to a designated target that w o u l d r a n d o m l y occur a m o n g the m o r e frequent non-targets. The subjects h a d to r e s p o n d to the design a t e d target when it a p p e a r e d on either side, even t h o u g h they were to visualize an o p a q u e b a r r i e r o b s c u r i n g one half o f their visual field. The instruction to r e s p o n d to b o t h sides ensured t h a t subjects did n o t selectively a t t e n d to only the u n o b s t r u c t e d side. C o m p a r i n g s c a l p - r e c o r d e d visual E R P s from the same hemifield when it was o b s t r u c t e d a n d when it was n o t o b s t r u c t e d d u r i n g hypnosis p r o v i d e d a p s y c h o p h y s i o l o g i c a l metric o f the c a p a bilities o f each h e m i s p h e r e in p r o d u c i n g the h y p n o t i c h a l l u c i n a t i o n as evidenced b y E R P reduction. O n l y the n o n - t a r g e t E R P s were analyzed, eliminating a n y confusion between observed E R P changes a n d presence or absence o f m o t o r - r e l a t e d potentials. N o n - t a r g e t E R P changes also reflect " s t i m u l u s set" as o p p o s e d to " r e s p o n s e set" selection processes [9]. As p r o p o s e d by F a r a h [6], i m a g e r y enhances or r e t a r d s stimulus e n c o d i n g which is selection at the level o f stimulus set.
Methods
Subjects Subjects were recruited from the Stanford University campus and medical center by advertisement. All had normal vision with or without corrective lenses and reported themselves as right-handed without any left-handed first degree relatives. At a preliminary screening session they were administered the hypnotic induction profile [19]. If a subject scored greater than 7.5 on the 10 point scale of this test they were classified as "highly hypnotizable". The subjects used in this study were all highly hypnotizable, comprising 10 males (19-26 years of age, median 20 years) and 12 females (20-27 years of age, median 22 years). Informed consent was obtained after the nature and possible consequences of the study had been fully explained in accordance with the Stanford Human Subjects Committee guidelines. All subjects were reimbursed for participation in the ERP experiment at a rate of $10.00 per hour.
Recording The EEG was recorded from l0 to 20 system sites Fz, Cz, Pz, F4, C4, P4, 02, F3, C3, P3, O1, F8, T6, F7 and T5 referenced to linked earlobes. Ground electrode was placed on the forehead. Resistance at all sites was brought down to less than 5 kf~ by mild abrasion. The EEG channels were amplified with a bandpass of 0.1-100 Hz. In addition, the combined vertical and horizontal electro-oculograph (EOG) was monitored from sites 1 inch above the left canthus and 1 inch below the right
canthus, referenced to each other, and amplified with the same parameters as the EEG. The subject was asked to make repeated eye movements away from and back to a central fixation point in the left, right, up and down directions. All such movements produced EOG channel amplitudes in excess of _+35 #V. All 16 channels of amplified data were continuously digitized at 256 Hz and stored on optical disk along with an additional channel of stimulus event markers.
Stimuli The subjects sat approximately 14 inches away from a 11 x 8 inch PC monitor. They were instructed to fixate on a small cross in the center of the screen. The stimuli were the uppercase letters A, E, F, H, 1, L, T, Y and Z which appeared for 100 msec on either the left or right side of the fixation cross. The letters were white on the black background of the monitor and 1 inch high. The center of each letter was displaced 2.75 inches to the side of and 1.5 inches above the fixation cross, subtending a visual angle of approximately 13'" from the center of each letter stimulus to the fixation point. The IS! was 950 msec with a jitter of _+50 msec. Within each block, nine of the 10 letters were presented in random order, 13 times each, on both left and right sides of fixation. A tenth letter, designated as the target stimulus for the particular block was randomly presented 26 times on both left and right sides. There were thus a total of 286 stimuli per block with an overall probability for target occurrence of 18%. The target letter stimuli differed randomly over blocks.
Procedure Subjects were given a series of eight blocks all comprised of a basic target detection paradigm in which they were required to respond to a designated target letter, as quickly and accurately as possible, with a right forefinger button press. It was emphasized to the subjects that during stimulus presentation they were to maintain visual fixation upon the central cross. The subjects were given a preliminary practice demonstration with stimuli presented on both sides of the fixation cross in order to familiarize them with the task. After 20 trials or so, all subjects reported that they could identify the lateralized letter stimuli in their peripheral vision. For all subjects, the hypnotic obstruction right visual field and hypnotic obstruction left visual field blocks were randomized over block positions 3-8. The other blocks comprised selective attention, hypnotic stimulus enhancement and passive and active attentional blocks which are not included in the present analysis. For the two hypnotic obstruction blocks, subjects were given a short self-hypnotic induction and instructed to imagine, with their eyes closed, a solid wall being built up and covering one half of the monitor. They were then told to open their eyes and maintain the image of the wall as best as possible. They were to respond to a designated target letter when it appeared on either side of the monitor and were told that " . . . even though the (left/right) side of the computer monitor will be blocked out by the wall, some letters might come through" but not to worry if they did not see any. The hypnotist operator sat directly behind and slightly above the subject during testing. They were able to see the subject's eyes in a oneway mirror behind the monitor and thus monitor the subject's compliance with the central fixation instruction.
Event-related potentials The digitized EEG was segmented into 1 sec epochs starting 100 msec prior to stimulus onset and continuing for 900 msec
P. Jasiukaitis et al./Lefl hemisphere superiority for hypnotic obstruction after. Trials on which EOG channel amplitude exceeded _+35 #V, were excluded from the averaging process. Difference waveforms were computed for non-target stimuli by subtracting hypnotically obstructed from unobstructed visual hemifield ERPs. The unobstructed ERPs were obtained from the other hypnotic obstruction block when the same hemifield was not obstructed during hypnosis, i.e. left unobstructed would be obtained from the hypnotically obstructed right block. Mean voltages were calculated in these difference waves over time periods corresponding to the P 100 (70-120 msec), N 100 (120 190 msec), P200 (190-300 msec) and P3 (300~410 msec).
663
hemifields. For the right visual field, hypnotic obstruction reduces positivity in the visual E R P from about 70 to 400 msec post-stimulus. This reduction appears greatest over the left posterior scalp. In the left visual field hypnotic obstruction reduces the PI00 peak at electrode T6, but seems to have little effect on other components or at other electrodes.
Non-target stimulus difference wavel'orms Results
The mean difference voltages were submitted to four separate repeated measures multivariate analysis of variance ( M A N O V A ) omnibus tests for the P100, N100, P200 and P300 components. Multivariate analysis was used in order to control for type 1 error due to nonorthogonality of the electrode factor. Neighboring electrodes, being highly correlated with each other, would inflate significance if included as independent variables. In M A N O V A they are treated as multiple dependent variables. After Bonferroni correction (L = 4) the nora-
Event-related potentials, non-target stimuli Figure 1 superimposes the grand mean (N = 22) ERPs for right visual field stimuli when the right hemifield was unobstructed over those ERPs when it was hypnotically obstructed. Figure 2 makes the same comparison for left visual stimuli. There appears to be a marked difference in the effect of hypnotic obstruction between the two
RIGHT VISUAL FIELD NOT OBSTRUCTED RIGHT VISUAL FIELD HYPNOTICALLY OBSTRUCTED
.........................................
EO~
F4
V'~
'~
ry
P3
\
"'4.
PZ ,.,
W',.,Y
P4 ,.
W k"F
/ ,//
Fig. 1. Grand mean (N = 22) visual ERPs to stimuli in the right hemifield. Heavy lines indicate ERPs when the right hemifield was unobstructed during the left hypnotic obstruction block, light lines indicate ERPs when the right hemifield was obstructed during the right hypnotic obstruction block. Positive voltage is plotted downwards. The area between the heavy and light lines constitutes the grand mean difference voltages displayed in Tables 1--3 and in Fig. 3. The P100 and P200 peaks are indicated at electrodes T5 and P3.
664
P. Jasiukaitis et al./Left hemisphere superiority for hypnotic obstruction
LEFT VISUAL FIELD NOT OBSTRUCTED LEFT VISUAL FIELD HYPNOTICALLY OBSTRUCTED E06
, F3
FZ
C3
CZ
P3~t
PZ
/
\
)
C4
500 re=tee
P4 ~ c -
VV
//
'~'~
'vg
P oo
Fig. 2. Grand mean (N = 22) visual ERPs to stimuli in the left hemifield. Heavy lines indicate ERPs when the left hemifield was unobstructed during the right hypnotic obstruction block, light lines indicate ERPs when the left hemifield was obstructed during the left hypnotic obstruction block. Positive voltage is plotted downwards. The area between the heavy and light lines constitutes the grand mean difference voltages displayed in Tables 1-3 and in Fig. 3. The P100 peak is indicated at electrode T6.
inal 0.05 level of significance for each MANOVA was set to 0.0125. The independent variables within each MANOVA were two levels of visual hemifield (left vs right) and cerebral hemisphere (left vs right). Multivariate observations comprised central (C3 or C4), parietal (P3 or P4), posterior temporal (T5 or T6) and occipital (O1 or 02) recording sites within each hemisphere. Since both independent factors (hemifield and hemisphere) had only two levels, the F statistic and associated P value for all multivariate tests were exact. For the P 100 component, the unobstructed-obstructed residuals were significantly different from zero [F(4, 18) = 9.33; P < 0.001] indicating that hypnotic obstruction did reduce PI00 overall. The hemifield by hemisphere interaction was significant [F(4, 18) = 4.66; P = 0.009]. Inspection of Table 1 and Fig. 3 suggests that the source of this interaction is that the P100 residual tended to be larger at electrodes over the hemisphere contralateral to the hemifield of stimulation. This is not surprising given that visual hemifields project to contralateral hemispheres and stimulus selection processes might appear maximal there.
For the N 100 component there was a significant hemifield by hemisphere interaction [F(4, 1 8 ) = 4.46; P = 0.011]. This can be accounted for in the same manner as the hemifield by hemisphere interaction for P100. Table 2 indicates a tendency for N100 residuals to be more positive over both hemispheres at occipital and parietal sites for right hemifield obstruction. This may be reflected in the main effect of hemifield which approached significance [F(4, 18) = 3.90; P = 0.019]. The MANOVA on the P200 unobstructed-obstructed residuals found a significant main effect of hemifield [F(4, 18) = 6.53; P = 0.002] along with an interaction of hemifield and hemisphere [F(4, 18) = 5.37; P = 0.005]. Table 3 and Fig. 3 reveal that the P200 residuals, like those for P100 and N100, were larger at electrode sites contralateral to the hemifield of stimulation. However, by this time post-stimulus the right hemifield unobstructedobstructed residuals are overall more positive than those for the left hemifield. This is the source of the significant hemifield main effect. For the P300 residuals there were no significant main effects or interactions. This may indicate that the overall
665
P. Jasiukaitis et al./Left hemisphere superiority for hypnotic obstruction Table 1. Grand mean (N = 22) difference voltages, in #V, between unobstructed and obstructed ERPs for the P100 component area (70-120 msec). Standard deviations in this and following tables are in parentheses Left visual field Left hemisphere Occipital Temporal Parietal Central
1.6
~uv
Right hemisphere
0.24 (1.27) -0.21 (1.28) -0.15 (1.44) -0.032(1.36)
0.78 (1.79) 0.58 ( 1 . 2 6 ) 0.39(1.75) 0.22(1.67)
PIO0
'
Right visual field Left hemisphere
Right hemisphere
1.07 (1.50) 0.82(1.13) 1.04(1.50) 0.48(1.31)
0.71 (1.42) 0.14(1.03) 0.53(1.44) 0.09(1.30)
1.e
1.2,
1.2,
0.8
0.8'
0.4
P200
0.4
..e
®D.O
RIGHT
HEMIFIELD
LEFT
"""
0.0
HEMIFIELD G ........ E)
-0.4
I
I
01
02
-o.4
"
1.6
1.2,
1.2
D.8'
0.8
D,4,
0.4
1,'6
I
I
01
02
~V .-O 0.0' -D.4
0.0"
I
-o.4
"1
P3
P4
I
I
P3
P4
Fig. 3. Grand mean (N = 22) difference voltages for the P100 (70-120 msec) and P200 (190-300 msec) peaks plotted separately by hemifield of stimulation over left and right hemispheres for the occipital and parietal electrodes. Voltages were calculated by subtracting the obstructed ERP from the non-obstructed ERP within hemifield.
Table 2. Grand mean (N = 22) difference voltages, in pV, between unobstructed and obstructed ERPs for the N100 component area (120-190 msec) Left visual field Left hemisphere Occipital Temporal Parietal Central
-0.34(1.62) -0.46(1.31) -0.60 (1.48) - 0.29 (1.46)
Right hemisphere 0.07(1.70) 0.11 (1.30) -0.02 (1.49) 0.01 (1.19)
Right visual field Left hemisphere 0.90(1.79) 0.70(1.38) 0.92 (1.73) 0.45 (1.63)
Right hemisphere 0.60(1.55) -0.08(1.03) 0.51 (1.51) 0.19 (1.87)
666
P. Jasiukaitis et al./Left hemisphere superiority for hypnotic obstruction Table 3. Grand mean (N = 22) difference voltages, in /~V, between unobstructed and obstructed ERPs for the P200 component area (190-300 msec) Left visual field
Occipital Temporal Parietal Central
Right visual field
Left hemisphere
Right hemisphere
Left hemisphere
Right hemisphere
-0.03(1.41) -0.09(1.11) - 0.05 (1.44) 0.13 (1.51)
0.41 (1.44) 0.39(1.22) 0.34 (1.15) 0.26 (1.62)
1.49(1.22) 1.29(1.43) 1.64 (1.16) 1.05 (1.06)
0.96(1.10) 0.41 (1.00) 1.00 (1.00) 0.51 (0.93)
reduced positivity in the non-target ERP due to hypnotic obstruction does not extend significantly beyond 300 msec post-stimulus.
Behavioral data
The mean percentage missed (no button press) targets for the right and left visual fields when they were hypnotically obstructed were 55.6 and 57.4% respectively. When not obstructed they were 8.2 and 13.4% for the right and left hemifields. This is reflected in a highly significant effect of obstruction [F(1,21) = 53.95; P < 0.001]. However, the difference between percentage missed in the right and left hemifields when obstructed was not significant I F ( l , 2 1 ) = 0.39; P = 0.54, which stands in contrast to the N100 and P200 non-target ERP effects. To see how the non-target ERP effects might relate to behavior, percentage missed in each hemifield was linearly regressed with the right hemifield unobstructedobstructed P200 residual at electrodes O1, P3 and T5 as the dependent variables. The electrodes were chosen since they seemed to show the maximal hypnotic obstructive effect (Fig. 1). For the percentage missed targets in the right hemifield the results were: at O1, r = 0.63 and P = 0.002; at P3, r = 0.54 and P = 0.01; and at T5, r = 0.48 and P = 0.02. For percentage missed targets in the left hemifield: at O1, r = 0.50 and P = 0.02; at P3, r = 0.47 and P = 0.03; and at T5, r = 0.46 and P = 0.03. Thus it appears that while overt behavior during hypnotic obstruction did not show a hemifield effect by itself, those subjects who did show a right hemifield ERP (P200) effect were more likely to show response suppression to target stimuli in either hemifield when that hemifield was hypnotically obstructed.
Discussion The right hemifield main effect of hypnotic obstruction on the P200 is in agreement with Farah et al.'s [7] results. They found increased positivity in the ERP to word stimuli when visualization of the word's referent was required. Although they did not use lateralized stimulus presen-
tation, they found a posterior left hemisphere maximum for the enhanced positivity and attributed this to a specialized left hemisphere mechanism for the production of visual imagery. Visual imagery in the form of an imagined obstructive barrier was used in the present experiment. Unlike Farah et al.'s [7] imagery, it was constant and ongoing, i.e. not time-locked to stimulus presentation. Farah et al. [7] found increases rather than decreased positivity because engagement of the imagegenerating mechanism in their experiment was timelocked and related to their visual stimulus presentation. In the present study left hemisphere image generation would have subtracted positivity from the ERP to right hemifield visual stimuli because left hemisphere capacities would be divided between maintaining the image and processing the right hemifield visual stimulus. The maintenance of the obstructing image would thus have competed with processing of visual stimuli, in effect acting like a secondary task in a dual task paradigm (detecting designated targets while imagining a barrier). Such a reduction was not found for left visual field stimulation, as right hemisphere capacity could be more devoted to visual stimulus processing since the image generating mechanism is not resident in the hemisphere. Farah et al.'s [7] increased positivity took place much later, around 600 msec, than the decreased positivity observed maximally around 200 msec in the present experiment. This could reflect the additional time required to process word stimuli lexically and then formulate the referent image. In the present study, the image of the barrier was already existent prior to stimulus presentation and thus could exert its dampening effect much earlier in the stimulus evoked potential. Alternative explanations of the obtained P200 reduction might invoke attentional mechanisms rather than imagery. However, from what we know about the laterality of visual attention (for review, see Jutai [10]), hemifield superiority for attentional effects should appear in the left, rather than right, visual field. Harter et al. [8] found that when subjects selectively attended to one visual field, occipitally recorded N270 was larger over the right hemisphere regardless of the location of the flashed stimulus. Harter et al. [8] suggested that the right hemisphere may be specialized for selective spatial attention, if the suggestion for hypnotic obstruction used in the present
P. Jasiukaitis et al./Left hemisphere superiority for hypnotic obstruction study is construed to be a type of inattention to one hemifield, then a left visual field superiority for ERP effects should emerge if the right hemisphere model for visuo-locational attention is to hold. Exactly the opposite was found in the present study. This would suggest that the effects of hypnotic obstruction on the P200 component are not due to selective attention to location. The second important finding in this study was that the magnitude of the right hemifield ERP effect was positively correlated with behavioral compliance (button press omission) to the suggestion of an obstructive hallucination in either hemifield. This indicates a positive relationship between the left hemisphere imagery generating mechanism and overt hypnotic behavior. Previous attempts to find correlation between imagery and hypnotizability have not been successful [12, 16]. However these studies have all used self-report of imagery vividness as their imagery measure. The present work measured the effect of imagery upon the ERP. Our subjects were not instructed to manipulate their ERP amplitudes directly. The right hemifield ERP reduction thus constitutes a performance effect of imagery. Self-report, subject as it is to attribute and demand characteristics, may be inherently more inaccurate than performance measures. The lateralization of a hypnotic effect to the left hemisphere would seem to go against the trend to see hypnosis as a right hemisphere function. However, as mentioned in the introduction, viewing hypnosis as a function of one or the other hemisphere may be an oversimplification. Highly hypnotizable subjects may mobilize faculties of either hemisphere in order to meet the demands of the hypnotic task. Which faculties are mobilized will depend on what the hypnotic task is. In the case of positive visual hallucination, imagery would be very useful. But there may be some face validity in regarding hypnosis as a left hemisphere function. Highly focused attention, simultaneous with a relative independence of behavior from context, has long been considered a fundamental aspect of hypnotic behavior [19]. Such narrow attentional focus would seem to be a function of the left hemisphere's detailed analytical and sequential processing capacities. This was suggested in a theoretical paper by Kinsbourne [1 l] in which he writes: "A narrow range of awareness suffices for specific acts of discrimination or execution (left hemispheric); automalized, the act calls for little more . . . . Goldberg and Costa [13] advocated the view that the right hemisphere is specialized for novelty detection. If so, its functions, more than those of the left hemisphere, would appear to invoke awareness. If at any time a hemisphere works like an automaton, it is the left."
Acknowledgements- -This work was supported by funding from the John D. and Catherine T. MacArthur Foundation. The authors wish to thank the members of the MacArthur Mind-
667
Body Network for their discussion and criticism which played an important role in the development of this work.
References
1..~ndreassi, J. L., Balinsky, B., Gallichio, J. A., DeSimone, J. L. and Mellers, B. W. Hypnotic suggestion of stimulus change and visual cortical evoked potential. Percept. Motor Skills 42, 371 378, 1976. 2. Beck, E. C., Dustman, R. E. and Beier, E. G. Hypnotic suggestions and visually evoked potentials. Electroenceph. Clin. Neurophysiol. 20, 397400, 1966. 3. DePascalis, V. Event-related potentials during hypnotic hallucination, hTt. J. Clin. Exp. Hypnosis 42, 39 55, 1994. 4. Ehrlichman, H. and Weinberger, A. Lateral eye movements and hemispheric asymmetry: A critical review. Psvchol. Bull. 85, 1080--1101, 1978. 5. Farah, M. J. The neurological basis of mental imagery: A componential analysis. Cognition 18, 245-272, 1984. 6. Farah, M. J. The laterality of mental image generation: A test with normal subjects. Neuropo'chologia 24, 541-551, 1986. 7. Farah, M. J., Weisberg, L. L., Monheit, M. and Peronnet, F. Brain activity underlying imagery: Event-related potentials during mental image generation. J. Cognit. Neurosci. 1,302 316, 1990. 8. Harter, M. R., Aine, C. and Schroeder, C. Hemispheric differences in the neural processing of stimulus location and type: Effects of selective attention on visual evoked potentials. Neurop,~vchologia 20, 421 ~438, 1982. 9. Hillyard, S. A., Hink, R. F., Schwent, V. L. and Picton, T. W. Electrical signs of selective attention in the human brain. Science 182, 177 180, 1973. 10. Jutai, J. W. Cerebral asymmetry and the psychophysiology of attention. Int. J. Psychophysiol. 1,219 225, 1984. I I. Kinsbourne, M. Integrated field theory of consciousness. In Consciousness & Contemporao' Science, A. J. Marcel and E. Bisiach (Editors), pp. 239-256. Clarendon Press, Oxford, 1992. 12. Glisky, M. L., Tataryn, D. J. and Kihlstrom, J. F. Hypnotizability and mental imagery, bTt. J. Clin. Exp, Hypnosis 43, 34 54, 1994. 13. Goldberg, E. and Costa, L. D. Hemispheric difference in the acquisition and use of descriptive systems. Bra& Lang. 14, 144 173, 1981. 14. Morgan, A. H., MacDonald, H. and Hilgard, E. R. EEG alpha: Lateral asymmetry related to task and hypnotizability. Psvchophysiology 11,275-282, 1974. 15. Perlini, A. H. and Spanos, N. P. EEG alpha methodologies and hypnotizability: A critical review. P,D,chophysiology 28, 511 530, 1991. 16. Spanos, N. P. Imagery, hypnosis and hypnotizability. In Mental Imagery, R. G. Kunzendorf (Editor), pp. 79 88. Plenum Press, New York, 1991, 17. Spiegel, D. and Barabasz, A. F. Effects of hypnotic instructions on P300 event-related potential amplitudes: Research and clinical implications. Am. J. Clin, Hypnosis 31, 11 17, 1988.
668
P. Jasiukaitis et al./Left hemisphere superiority for hypnotic obstruction
18. Spiegel, D., Cutcomb, S., Ren, C. and Pribram, K. Hypnotic hallucination alters evoked potentials. J. Abnormal Psychol. 94, 249-255, 1985. 19. Spiegel, H. and Spiegel, D. Trance and Treatment: Clinical Uses of Hypnosis. Basic Books, New York, 1978.
20. Tippett, L. J. The generation of visual images: A review of neuropsychological research and theory. Psychol. Bull. 112, 415-432, 1992. 21. Zakrzewski, K. and Szelenberger, W. Visual evoked potentials in hypnosis: A longitudinal approach. Int. J. Clin. Exp. Hypnosis 29, 77-86, 1981.
0028 3932(95100149 2
Neuropsychologia,Vol. 34, No. 7, pp. 661-668, 1996 Copyright ,~ 1996 Elsevier Science Ltd. All rights reserved Printed in Great Britain 0028- 3932/96 $15.00+0.00
Left hemisphere superiority for event-related potential effects of hypnotic obstruction PAUL JASIUKAITIS,* BITA N O U R I A N I and DAVID SPIEGEL Department of Psychiatry and Behavioral Sciences, Stanford University Medical Center, Stanford, CA 94305, U.S.A.
(Received 13 February 1995; accepted 25 July 1995)
Abstract--Twenty-two highly hypnotizable subjects were run in a visual target detection task which compared hypnotic obstruction of the left and right visual fields over separate blocks. The visual event-related potentials (ERPs) to non-target stimuli revealed that hypnotic obstruction reduced the P200 component to stimuli in the right hemifield, but did not affect P200 for stimulation in the left hemifield. The earlier P100 and N100 were also reduced to hypnotic obstruction but not as preferentially for either hemifield, while the P300 was not significantly changed. Right visual field/left hemisphere P200 reduction predicted suppression of behavioral response (button press) to hypnotically obstructed targets in both hemifields. The results are discussed in terms of Farah's model of a left hemisphere mechanism for image generation, and how highly hypnotizable subjects might use this mechanism to comply successfully with the suggestion of a hallucinated visually opaque barrier. Copyright ,~ 1996 Elsevier Science Ltd. Key Words: visual event-related potentials; hypnotic obstruction; imagery; left hemisphere.
Introduction
gathering over the past decade that image generation engages left hemispheric functioning [20]. In a review of neurological cases, Farah [5] found 12 patients with deficits in image generation but with perception and recognition grossly intact. In these 12 cases the predominant site of brain damage was the posterior left hemisphere. Among neurologically intact volunteers, Farah [6] found that visual image and stimulus match in the right visual field speeded reaction time more than it did in the left visual field, suggesting more efficient image production in the left hemisphere. Relevant to ERP research is the study by Farah et al. [7]. They found that foveally presented word stimuli elicited greater positivity in the visual ERP, from about 600 to 900 msec post-stimulus, when subjects were required to form a concrete mental image of the word's referent as opposed to detecting misspellings. The scalp maximum of this enhanced positivity was over the left occipital and posterior temporal areas, suggesting an important role for the left visual cortex in the generation of images from word stimuli. It has been suggested that hypnosis is a "right brain" phenomenon [14]. However, this view is not firmly established, being based on EEG spectral and lateral eye movement data which are still open to various interpretations [4, 15]. It could be that there is no inherent laterality in hypnosis, but that whatever cognitive faculties {left or right hemisphere) are needed for performance of a given hypnotic instruction will be the ones employed. If so, it might be
It has been reported that hypnotic suggestion can reduce the amplitude of individual peaks in the visual eventrelated potential (ERP) of highly hypnotizable subjects [3, 18]. These two studies utilized hypnotic obstruction, i.e. the suggestion of an imagined opaque object between the subject and the visual stimulus, instead of hypnotic diminution or the suggestion that the evoking stimulus was less intense or not there. The fact that hypnotic diminution has often been used in attempts to demonstrate the effects of hypnosis upon visual ERPs may account for some of the negative results [1, 2, 21]. The suggestion of diminution would remain unconvincing to the subject as long as the stimulus continues to be perceived. The hypnotic production of a competing visual image does not focus attention upon the external percept to be eliminated, but instead focuses it on the internally generated image. In this manner hypnotic obstruction may provide more consistent results than hypnotic diminution [17]. Unlike the suggestion of stimulus diminution, hypnotic obstruction involves image generation. Evidence has been * Address correspondence and requests for reprints to: Paul Jasiukaitis, Room C231, MC 5544, Department of Psychiatry and Behavioral Sciences, Stanford University Medical Center, Stanford, CA 94305, U.S.A. 661
662
P. Jasiukaitis et al./Left hemisphere superiority for hypnotic obstruction
expected t h a t the left hemisphere i m a g e r y m e c h a n i s m described by F a r a h [6] would be used in the p e r f o r m a n c e o f h y p n o t i c visual o b s t r u c t i o n . The following visual h y p n o t i c o b s t r u c t i o n e x p e r i m e n t used visual h a l f field s t i m u l a t i o n as did F a r a h ' s [6] imagery a n d reaction time task. Subjects were instructed to i m a g i n e a solid b a r r i e r covering the visual hemifield o n the left o r right side o f a central fixation point. Visual stimuli occurred r a n d o m l y to the left a n d right o f fixation a n d subjects were required to press to a designated target that w o u l d r a n d o m l y occur a m o n g the m o r e frequent non-targets. The subjects h a d to r e s p o n d to the design a t e d target when it a p p e a r e d on either side, even t h o u g h they were to visualize an o p a q u e b a r r i e r o b s c u r i n g one half o f their visual field. The instruction to r e s p o n d to b o t h sides ensured t h a t subjects did n o t selectively a t t e n d to only the u n o b s t r u c t e d side. C o m p a r i n g s c a l p - r e c o r d e d visual E R P s from the same hemifield when it was o b s t r u c t e d a n d when it was n o t o b s t r u c t e d d u r i n g hypnosis p r o v i d e d a p s y c h o p h y s i o l o g i c a l metric o f the c a p a bilities o f each h e m i s p h e r e in p r o d u c i n g the h y p n o t i c h a l l u c i n a t i o n as evidenced b y E R P reduction. O n l y the n o n - t a r g e t E R P s were analyzed, eliminating a n y confusion between observed E R P changes a n d presence or absence o f m o t o r - r e l a t e d potentials. N o n - t a r g e t E R P changes also reflect " s t i m u l u s set" as o p p o s e d to " r e s p o n s e set" selection processes [9]. As p r o p o s e d by F a r a h [6], i m a g e r y enhances or r e t a r d s stimulus e n c o d i n g which is selection at the level o f stimulus set.
Methods
Subjects Subjects were recruited from the Stanford University campus and medical center by advertisement. All had normal vision with or without corrective lenses and reported themselves as right-handed without any left-handed first degree relatives. At a preliminary screening session they were administered the hypnotic induction profile [19]. If a subject scored greater than 7.5 on the 10 point scale of this test they were classified as "highly hypnotizable". The subjects used in this study were all highly hypnotizable, comprising 10 males (19-26 years of age, median 20 years) and 12 females (20-27 years of age, median 22 years). Informed consent was obtained after the nature and possible consequences of the study had been fully explained in accordance with the Stanford Human Subjects Committee guidelines. All subjects were reimbursed for participation in the ERP experiment at a rate of $10.00 per hour.
Recording The EEG was recorded from l0 to 20 system sites Fz, Cz, Pz, F4, C4, P4, 02, F3, C3, P3, O1, F8, T6, F7 and T5 referenced to linked earlobes. Ground electrode was placed on the forehead. Resistance at all sites was brought down to less than 5 kf~ by mild abrasion. The EEG channels were amplified with a bandpass of 0.1-100 Hz. In addition, the combined vertical and horizontal electro-oculograph (EOG) was monitored from sites 1 inch above the left canthus and 1 inch below the right
canthus, referenced to each other, and amplified with the same parameters as the EEG. The subject was asked to make repeated eye movements away from and back to a central fixation point in the left, right, up and down directions. All such movements produced EOG channel amplitudes in excess of _+35 #V. All 16 channels of amplified data were continuously digitized at 256 Hz and stored on optical disk along with an additional channel of stimulus event markers.
Stimuli The subjects sat approximately 14 inches away from a 11 x 8 inch PC monitor. They were instructed to fixate on a small cross in the center of the screen. The stimuli were the uppercase letters A, E, F, H, 1, L, T, Y and Z which appeared for 100 msec on either the left or right side of the fixation cross. The letters were white on the black background of the monitor and 1 inch high. The center of each letter was displaced 2.75 inches to the side of and 1.5 inches above the fixation cross, subtending a visual angle of approximately 13'" from the center of each letter stimulus to the fixation point. The IS! was 950 msec with a jitter of _+50 msec. Within each block, nine of the 10 letters were presented in random order, 13 times each, on both left and right sides of fixation. A tenth letter, designated as the target stimulus for the particular block was randomly presented 26 times on both left and right sides. There were thus a total of 286 stimuli per block with an overall probability for target occurrence of 18%. The target letter stimuli differed randomly over blocks.
Procedure Subjects were given a series of eight blocks all comprised of a basic target detection paradigm in which they were required to respond to a designated target letter, as quickly and accurately as possible, with a right forefinger button press. It was emphasized to the subjects that during stimulus presentation they were to maintain visual fixation upon the central cross. The subjects were given a preliminary practice demonstration with stimuli presented on both sides of the fixation cross in order to familiarize them with the task. After 20 trials or so, all subjects reported that they could identify the lateralized letter stimuli in their peripheral vision. For all subjects, the hypnotic obstruction right visual field and hypnotic obstruction left visual field blocks were randomized over block positions 3-8. The other blocks comprised selective attention, hypnotic stimulus enhancement and passive and active attentional blocks which are not included in the present analysis. For the two hypnotic obstruction blocks, subjects were given a short self-hypnotic induction and instructed to imagine, with their eyes closed, a solid wall being built up and covering one half of the monitor. They were then told to open their eyes and maintain the image of the wall as best as possible. They were to respond to a designated target letter when it appeared on either side of the monitor and were told that " . . . even though the (left/right) side of the computer monitor will be blocked out by the wall, some letters might come through" but not to worry if they did not see any. The hypnotist operator sat directly behind and slightly above the subject during testing. They were able to see the subject's eyes in a oneway mirror behind the monitor and thus monitor the subject's compliance with the central fixation instruction.
Event-related potentials The digitized EEG was segmented into 1 sec epochs starting 100 msec prior to stimulus onset and continuing for 900 msec
P. Jasiukaitis et al./Lefl hemisphere superiority for hypnotic obstruction after. Trials on which EOG channel amplitude exceeded _+35 #V, were excluded from the averaging process. Difference waveforms were computed for non-target stimuli by subtracting hypnotically obstructed from unobstructed visual hemifield ERPs. The unobstructed ERPs were obtained from the other hypnotic obstruction block when the same hemifield was not obstructed during hypnosis, i.e. left unobstructed would be obtained from the hypnotically obstructed right block. Mean voltages were calculated in these difference waves over time periods corresponding to the P 100 (70-120 msec), N 100 (120 190 msec), P200 (190-300 msec) and P3 (300~410 msec).
663
hemifields. For the right visual field, hypnotic obstruction reduces positivity in the visual E R P from about 70 to 400 msec post-stimulus. This reduction appears greatest over the left posterior scalp. In the left visual field hypnotic obstruction reduces the PI00 peak at electrode T6, but seems to have little effect on other components or at other electrodes.
Non-target stimulus difference wavel'orms Results
The mean difference voltages were submitted to four separate repeated measures multivariate analysis of variance ( M A N O V A ) omnibus tests for the P100, N100, P200 and P300 components. Multivariate analysis was used in order to control for type 1 error due to nonorthogonality of the electrode factor. Neighboring electrodes, being highly correlated with each other, would inflate significance if included as independent variables. In M A N O V A they are treated as multiple dependent variables. After Bonferroni correction (L = 4) the nora-
Event-related potentials, non-target stimuli Figure 1 superimposes the grand mean (N = 22) ERPs for right visual field stimuli when the right hemifield was unobstructed over those ERPs when it was hypnotically obstructed. Figure 2 makes the same comparison for left visual stimuli. There appears to be a marked difference in the effect of hypnotic obstruction between the two
RIGHT VISUAL FIELD NOT OBSTRUCTED RIGHT VISUAL FIELD HYPNOTICALLY OBSTRUCTED
.........................................
EO~
F4
V'~
'~
ry
P3
\
"'4.
PZ ,.,
W',.,Y
P4 ,.
W k"F
/ ,//
Fig. 1. Grand mean (N = 22) visual ERPs to stimuli in the right hemifield. Heavy lines indicate ERPs when the right hemifield was unobstructed during the left hypnotic obstruction block, light lines indicate ERPs when the right hemifield was obstructed during the right hypnotic obstruction block. Positive voltage is plotted downwards. The area between the heavy and light lines constitutes the grand mean difference voltages displayed in Tables 1--3 and in Fig. 3. The P100 and P200 peaks are indicated at electrodes T5 and P3.
664
P. Jasiukaitis et al./Left hemisphere superiority for hypnotic obstruction
LEFT VISUAL FIELD NOT OBSTRUCTED LEFT VISUAL FIELD HYPNOTICALLY OBSTRUCTED E06
, F3
FZ
C3
CZ
P3~t
PZ
/
\
)
C4
500 re=tee
P4 ~ c -
VV
//
'~'~
'vg
P oo
Fig. 2. Grand mean (N = 22) visual ERPs to stimuli in the left hemifield. Heavy lines indicate ERPs when the left hemifield was unobstructed during the right hypnotic obstruction block, light lines indicate ERPs when the left hemifield was obstructed during the left hypnotic obstruction block. Positive voltage is plotted downwards. The area between the heavy and light lines constitutes the grand mean difference voltages displayed in Tables 1-3 and in Fig. 3. The P100 peak is indicated at electrode T6.
inal 0.05 level of significance for each MANOVA was set to 0.0125. The independent variables within each MANOVA were two levels of visual hemifield (left vs right) and cerebral hemisphere (left vs right). Multivariate observations comprised central (C3 or C4), parietal (P3 or P4), posterior temporal (T5 or T6) and occipital (O1 or 02) recording sites within each hemisphere. Since both independent factors (hemifield and hemisphere) had only two levels, the F statistic and associated P value for all multivariate tests were exact. For the P 100 component, the unobstructed-obstructed residuals were significantly different from zero [F(4, 18) = 9.33; P < 0.001] indicating that hypnotic obstruction did reduce PI00 overall. The hemifield by hemisphere interaction was significant [F(4, 18) = 4.66; P = 0.009]. Inspection of Table 1 and Fig. 3 suggests that the source of this interaction is that the P100 residual tended to be larger at electrodes over the hemisphere contralateral to the hemifield of stimulation. This is not surprising given that visual hemifields project to contralateral hemispheres and stimulus selection processes might appear maximal there.
For the N 100 component there was a significant hemifield by hemisphere interaction [F(4, 1 8 ) = 4.46; P = 0.011]. This can be accounted for in the same manner as the hemifield by hemisphere interaction for P100. Table 2 indicates a tendency for N100 residuals to be more positive over both hemispheres at occipital and parietal sites for right hemifield obstruction. This may be reflected in the main effect of hemifield which approached significance [F(4, 18) = 3.90; P = 0.019]. The MANOVA on the P200 unobstructed-obstructed residuals found a significant main effect of hemifield [F(4, 18) = 6.53; P = 0.002] along with an interaction of hemifield and hemisphere [F(4, 18) = 5.37; P = 0.005]. Table 3 and Fig. 3 reveal that the P200 residuals, like those for P100 and N100, were larger at electrode sites contralateral to the hemifield of stimulation. However, by this time post-stimulus the right hemifield unobstructedobstructed residuals are overall more positive than those for the left hemifield. This is the source of the significant hemifield main effect. For the P300 residuals there were no significant main effects or interactions. This may indicate that the overall
665
P. Jasiukaitis et al./Left hemisphere superiority for hypnotic obstruction Table 1. Grand mean (N = 22) difference voltages, in #V, between unobstructed and obstructed ERPs for the P100 component area (70-120 msec). Standard deviations in this and following tables are in parentheses Left visual field Left hemisphere Occipital Temporal Parietal Central
1.6
~uv
Right hemisphere
0.24 (1.27) -0.21 (1.28) -0.15 (1.44) -0.032(1.36)
0.78 (1.79) 0.58 ( 1 . 2 6 ) 0.39(1.75) 0.22(1.67)
PIO0
'
Right visual field Left hemisphere
Right hemisphere
1.07 (1.50) 0.82(1.13) 1.04(1.50) 0.48(1.31)
0.71 (1.42) 0.14(1.03) 0.53(1.44) 0.09(1.30)
1.e
1.2,
1.2,
0.8
0.8'
0.4
P200
0.4
..e
®D.O
RIGHT
HEMIFIELD
LEFT
"""
0.0
HEMIFIELD G ........ E)
-0.4
I
I
01
02
-o.4
"
1.6
1.2,
1.2
D.8'
0.8
D,4,
0.4
1,'6
I
I
01
02
~V .-O 0.0' -D.4
0.0"
I
-o.4
"1
P3
P4
I
I
P3
P4
Fig. 3. Grand mean (N = 22) difference voltages for the P100 (70-120 msec) and P200 (190-300 msec) peaks plotted separately by hemifield of stimulation over left and right hemispheres for the occipital and parietal electrodes. Voltages were calculated by subtracting the obstructed ERP from the non-obstructed ERP within hemifield.
Table 2. Grand mean (N = 22) difference voltages, in pV, between unobstructed and obstructed ERPs for the N100 component area (120-190 msec) Left visual field Left hemisphere Occipital Temporal Parietal Central
-0.34(1.62) -0.46(1.31) -0.60 (1.48) - 0.29 (1.46)
Right hemisphere 0.07(1.70) 0.11 (1.30) -0.02 (1.49) 0.01 (1.19)
Right visual field Left hemisphere 0.90(1.79) 0.70(1.38) 0.92 (1.73) 0.45 (1.63)
Right hemisphere 0.60(1.55) -0.08(1.03) 0.51 (1.51) 0.19 (1.87)
666
P. Jasiukaitis et al./Left hemisphere superiority for hypnotic obstruction Table 3. Grand mean (N = 22) difference voltages, in /~V, between unobstructed and obstructed ERPs for the P200 component area (190-300 msec) Left visual field
Occipital Temporal Parietal Central
Right visual field
Left hemisphere
Right hemisphere
Left hemisphere
Right hemisphere
-0.03(1.41) -0.09(1.11) - 0.05 (1.44) 0.13 (1.51)
0.41 (1.44) 0.39(1.22) 0.34 (1.15) 0.26 (1.62)
1.49(1.22) 1.29(1.43) 1.64 (1.16) 1.05 (1.06)
0.96(1.10) 0.41 (1.00) 1.00 (1.00) 0.51 (0.93)
reduced positivity in the non-target ERP due to hypnotic obstruction does not extend significantly beyond 300 msec post-stimulus.
Behavioral data
The mean percentage missed (no button press) targets for the right and left visual fields when they were hypnotically obstructed were 55.6 and 57.4% respectively. When not obstructed they were 8.2 and 13.4% for the right and left hemifields. This is reflected in a highly significant effect of obstruction [F(1,21) = 53.95; P < 0.001]. However, the difference between percentage missed in the right and left hemifields when obstructed was not significant I F ( l , 2 1 ) = 0.39; P = 0.54, which stands in contrast to the N100 and P200 non-target ERP effects. To see how the non-target ERP effects might relate to behavior, percentage missed in each hemifield was linearly regressed with the right hemifield unobstructedobstructed P200 residual at electrodes O1, P3 and T5 as the dependent variables. The electrodes were chosen since they seemed to show the maximal hypnotic obstructive effect (Fig. 1). For the percentage missed targets in the right hemifield the results were: at O1, r = 0.63 and P = 0.002; at P3, r = 0.54 and P = 0.01; and at T5, r = 0.48 and P = 0.02. For percentage missed targets in the left hemifield: at O1, r = 0.50 and P = 0.02; at P3, r = 0.47 and P = 0.03; and at T5, r = 0.46 and P = 0.03. Thus it appears that while overt behavior during hypnotic obstruction did not show a hemifield effect by itself, those subjects who did show a right hemifield ERP (P200) effect were more likely to show response suppression to target stimuli in either hemifield when that hemifield was hypnotically obstructed.
Discussion The right hemifield main effect of hypnotic obstruction on the P200 is in agreement with Farah et al.'s [7] results. They found increased positivity in the ERP to word stimuli when visualization of the word's referent was required. Although they did not use lateralized stimulus presen-
tation, they found a posterior left hemisphere maximum for the enhanced positivity and attributed this to a specialized left hemisphere mechanism for the production of visual imagery. Visual imagery in the form of an imagined obstructive barrier was used in the present experiment. Unlike Farah et al.'s [7] imagery, it was constant and ongoing, i.e. not time-locked to stimulus presentation. Farah et al. [7] found increases rather than decreased positivity because engagement of the imagegenerating mechanism in their experiment was timelocked and related to their visual stimulus presentation. In the present study left hemisphere image generation would have subtracted positivity from the ERP to right hemifield visual stimuli because left hemisphere capacities would be divided between maintaining the image and processing the right hemifield visual stimulus. The maintenance of the obstructing image would thus have competed with processing of visual stimuli, in effect acting like a secondary task in a dual task paradigm (detecting designated targets while imagining a barrier). Such a reduction was not found for left visual field stimulation, as right hemisphere capacity could be more devoted to visual stimulus processing since the image generating mechanism is not resident in the hemisphere. Farah et al.'s [7] increased positivity took place much later, around 600 msec, than the decreased positivity observed maximally around 200 msec in the present experiment. This could reflect the additional time required to process word stimuli lexically and then formulate the referent image. In the present study, the image of the barrier was already existent prior to stimulus presentation and thus could exert its dampening effect much earlier in the stimulus evoked potential. Alternative explanations of the obtained P200 reduction might invoke attentional mechanisms rather than imagery. However, from what we know about the laterality of visual attention (for review, see Jutai [10]), hemifield superiority for attentional effects should appear in the left, rather than right, visual field. Harter et al. [8] found that when subjects selectively attended to one visual field, occipitally recorded N270 was larger over the right hemisphere regardless of the location of the flashed stimulus. Harter et al. [8] suggested that the right hemisphere may be specialized for selective spatial attention, if the suggestion for hypnotic obstruction used in the present
P. Jasiukaitis et al./Left hemisphere superiority for hypnotic obstruction study is construed to be a type of inattention to one hemifield, then a left visual field superiority for ERP effects should emerge if the right hemisphere model for visuo-locational attention is to hold. Exactly the opposite was found in the present study. This would suggest that the effects of hypnotic obstruction on the P200 component are not due to selective attention to location. The second important finding in this study was that the magnitude of the right hemifield ERP effect was positively correlated with behavioral compliance (button press omission) to the suggestion of an obstructive hallucination in either hemifield. This indicates a positive relationship between the left hemisphere imagery generating mechanism and overt hypnotic behavior. Previous attempts to find correlation between imagery and hypnotizability have not been successful [12, 16]. However these studies have all used self-report of imagery vividness as their imagery measure. The present work measured the effect of imagery upon the ERP. Our subjects were not instructed to manipulate their ERP amplitudes directly. The right hemifield ERP reduction thus constitutes a performance effect of imagery. Self-report, subject as it is to attribute and demand characteristics, may be inherently more inaccurate than performance measures. The lateralization of a hypnotic effect to the left hemisphere would seem to go against the trend to see hypnosis as a right hemisphere function. However, as mentioned in the introduction, viewing hypnosis as a function of one or the other hemisphere may be an oversimplification. Highly hypnotizable subjects may mobilize faculties of either hemisphere in order to meet the demands of the hypnotic task. Which faculties are mobilized will depend on what the hypnotic task is. In the case of positive visual hallucination, imagery would be very useful. But there may be some face validity in regarding hypnosis as a left hemisphere function. Highly focused attention, simultaneous with a relative independence of behavior from context, has long been considered a fundamental aspect of hypnotic behavior [19]. Such narrow attentional focus would seem to be a function of the left hemisphere's detailed analytical and sequential processing capacities. This was suggested in a theoretical paper by Kinsbourne [1 l] in which he writes: "A narrow range of awareness suffices for specific acts of discrimination or execution (left hemispheric); automalized, the act calls for little more . . . . Goldberg and Costa [13] advocated the view that the right hemisphere is specialized for novelty detection. If so, its functions, more than those of the left hemisphere, would appear to invoke awareness. If at any time a hemisphere works like an automaton, it is the left."
Acknowledgements- -This work was supported by funding from the John D. and Catherine T. MacArthur Foundation. The authors wish to thank the members of the MacArthur Mind-
667
Body Network for their discussion and criticism which played an important role in the development of this work.
References
1..~ndreassi, J. L., Balinsky, B., Gallichio, J. A., DeSimone, J. L. and Mellers, B. W. Hypnotic suggestion of stimulus change and visual cortical evoked potential. Percept. Motor Skills 42, 371 378, 1976. 2. Beck, E. C., Dustman, R. E. and Beier, E. G. Hypnotic suggestions and visually evoked potentials. Electroenceph. Clin. Neurophysiol. 20, 397400, 1966. 3. DePascalis, V. Event-related potentials during hypnotic hallucination, hTt. J. Clin. Exp. Hypnosis 42, 39 55, 1994. 4. Ehrlichman, H. and Weinberger, A. Lateral eye movements and hemispheric asymmetry: A critical review. Psvchol. Bull. 85, 1080--1101, 1978. 5. Farah, M. J. The neurological basis of mental imagery: A componential analysis. Cognition 18, 245-272, 1984. 6. Farah, M. J. The laterality of mental image generation: A test with normal subjects. Neuropo'chologia 24, 541-551, 1986. 7. Farah, M. J., Weisberg, L. L., Monheit, M. and Peronnet, F. Brain activity underlying imagery: Event-related potentials during mental image generation. J. Cognit. Neurosci. 1,302 316, 1990. 8. Harter, M. R., Aine, C. and Schroeder, C. Hemispheric differences in the neural processing of stimulus location and type: Effects of selective attention on visual evoked potentials. Neurop,~vchologia 20, 421 ~438, 1982. 9. Hillyard, S. A., Hink, R. F., Schwent, V. L. and Picton, T. W. Electrical signs of selective attention in the human brain. Science 182, 177 180, 1973. 10. Jutai, J. W. Cerebral asymmetry and the psychophysiology of attention. Int. J. Psychophysiol. 1,219 225, 1984. I I. Kinsbourne, M. Integrated field theory of consciousness. In Consciousness & Contemporao' Science, A. J. Marcel and E. Bisiach (Editors), pp. 239-256. Clarendon Press, Oxford, 1992. 12. Glisky, M. L., Tataryn, D. J. and Kihlstrom, J. F. Hypnotizability and mental imagery, bTt. J. Clin. Exp, Hypnosis 43, 34 54, 1994. 13. Goldberg, E. and Costa, L. D. Hemispheric difference in the acquisition and use of descriptive systems. Bra& Lang. 14, 144 173, 1981. 14. Morgan, A. H., MacDonald, H. and Hilgard, E. R. EEG alpha: Lateral asymmetry related to task and hypnotizability. Psvchophysiology 11,275-282, 1974. 15. Perlini, A. H. and Spanos, N. P. EEG alpha methodologies and hypnotizability: A critical review. P,D,chophysiology 28, 511 530, 1991. 16. Spanos, N. P. Imagery, hypnosis and hypnotizability. In Mental Imagery, R. G. Kunzendorf (Editor), pp. 79 88. Plenum Press, New York, 1991, 17. Spiegel, D. and Barabasz, A. F. Effects of hypnotic instructions on P300 event-related potential amplitudes: Research and clinical implications. Am. J. Clin, Hypnosis 31, 11 17, 1988.
668
P. Jasiukaitis et al./Left hemisphere superiority for hypnotic obstruction
18. Spiegel, D., Cutcomb, S., Ren, C. and Pribram, K. Hypnotic hallucination alters evoked potentials. J. Abnormal Psychol. 94, 249-255, 1985. 19. Spiegel, H. and Spiegel, D. Trance and Treatment: Clinical Uses of Hypnosis. Basic Books, New York, 1978.
20. Tippett, L. J. The generation of visual images: A review of neuropsychological research and theory. Psychol. Bull. 112, 415-432, 1992. 21. Zakrzewski, K. and Szelenberger, W. Visual evoked potentials in hypnosis: A longitudinal approach. Int. J. Clin. Exp. Hypnosis 29, 77-86, 1981.