Invited Symposium: Neural Bases of Hypnosis
According to the synergistic approach to hypnosis (Laurence, 1991), individual differences in cognitive processing predispose a subject to the experience of hypnotic phenomena. These attributes interact with social factors such as subject's attitudes, belief and expectations towards hypnosis (Nadon, Laurence & Perry, 1991) to shape the degree to which an individual will respond to an hypnotic context. When a subject undergoes an hypnotic induction, these differences in cognitive processing are present "a priori" and, as such, should be measured both in and out of the hypnotic context. This way, it becomes easier to disentangle the respective role of social demands and cognitive factors. Absorption and imagery, for instance, have been found to correlate more highly with hypnotizability when measured within the hypnotic context then when removed from it. This variability (which is predictable) is a good example of what we mean by the synergy of hypnosis. When it comes to neuro-physiological data, however, the evidence supporting the relation between individual difference variables and hypnosis has been difficult to interpret mainly because the data has been gathered either just before hypnosis or under conditions that were clearly related to hypnosis testing.
A series of recent studies conducted in our laboratory have investigated individual differences and neurocognitive factors at play in hypnotizability and have emphasised the role of automaticity in various perceptual and memory tasks as a major component of the hypnotic response. Direct evidence from a neurophysiological study investigating event related potentials (ERPs) as well as the results of other studies investigating cognitive measures obtained completely divorced from the hypnotic context, are reviewed and discussed in terms of their ability to predict hypnotizability.
Hypnotizability and Automaticity
Automaticity is defined as a way of processing information effortlessly, rapidly and involuntarily (Shiffrin & Schneider, 1977). The automaticity hypothesis predicts that subjects' capacity to automatize perceptual tasks is directly related to their capacity to experience hypnosis. As such it is seen as a state of preparedness to the elicitation of hypnotic phenomena and other anomalous experiences as well (age regression to prior lives, extraterrestrial kidnappings, false memories of all sorts).
A first line of evidence in support of the automaticity hypothesis came from Dixon, Brunet & Laurence's (1990) study. Automatic and strategic effects on the Stroop task were assessed on subjects previously classified as high, moderate or low on hypnotizability. The subjects were kept unaware about the link between the Stroop task and hypnosis. Automatic effects were measured by establishing individual thresholds for word perception using a staircase method. Colour words were presented followed by a colour patch to be named. The colour word preceding the patch was presented at subthreshold and at suprathreshold time exposure to all subjects. Strategic effects were assessed by manipulating and informing the subjects regarding the probability of congruent trials (e.g. red word and red colour patch) and incongruent trials (e.g. red word and blue colour patch). During a first testing session, the word would predict the colour patch on 25 % of the trials. During a second testing session, the word would predict the colour patch on 75% of the trials. The results indicated that compared to low and moderately hypnotizable subjects, highly hypnotizables (HH) displayed shorter mean reaction times on congruent trials in all conditions. When differences in mean reaction time between congruent and incongruent trials were examined, HH showed a significantly larger discrepancy than that of the moderate and low hypnotizable subjects. The size of the discrepancy was observed to vary according to expected probability and suprathreshold word perception. Thus, HH exhibited the largest discrepancy between their mean reaction time when they expected the word and colour patch to match on 75% of the trials, and when the word preceding the presentation of the colour patch was consciously perceived.
This larger discrepancy was explained as greater facilitation and interference effects experienced by HH subjects. These effects were interpreted at the time as resulting from stronger connection strengths along verbal pathways. HH subjects were found to process information with greater automaticity and to be more likely to learn and implement a strategy to improve their performance. Thus, the size of the HH subjects' discrepancy resulted from a mixture of automatic and strategic effects.
A second study used a paradigm to separate automatic and strategic effects (Dixon & Laurence, 1992). Only two colour words and patches were used (red and blue). All subjects were told that if the word was blue than 3 out of 4 times, the colour patch would be red, or if the word was red, than 3 out of 4 times the colour would be blue (75% incongruent probability). Time delays between the presentation of the colour word and the colour patch were manipulated to control for strategic effect. At very short delays, subjects had no time to implement a strategy. The results indicated that HH responded faster to congruent trials. As inter stimulus intervals increased, HH exhibited a reversal of the standard Stroop pattern. Their mean reaction time became shorter for incongruent trials. Although Low Hypnotizable (LH) subjects also showed the reversal effect, HH were able to implement the strategy more rapidly and the reversal of the Stroop effect was observed at shorter presentation delays than it was for LH subjects.
Further evidence for individual differences in automatic processing came from a neurophysiological investigation of evoked potentials during a Stroop-like task. Subjects who had been previously classified as HH or LH (one year earlier in a different laboratory) were presented with a Stroop task while their Event Related Potential (ERPs) were recorded at central, parietal and frontal sites ( Baribeau, Le Beau, Roth,& Laurence, 1994). The Stroop test consisted of four tasks; manually recognising 1)a colour word written in black (word recognition, e.g. RED), 2) the colour of a series of XXXs (neutral, e.g. XXX), 3) the colour when the colour word and the colour stimulus were the same (congruent), and 4) the colour when the word and the colour stimulus differed (incongruent). Contrary to the two earlier studies that used randomized block presentations, trials were grouped by type (e.g. all congruent trials together, etc.) Although HH were found to display shorter mean reaction times on most tasks, the expected larger discrepancy between mean reaction time on congruent and incongruent trials observed in Dixon et al.(1991) and Dixon & Laurence(1992) disappeared. Because of the homogeneity of the block presentations, it was hypothesised that HH subjects had the time to develop strategies to improve their performance on incongruent trials as had been the case in Dixon and Laurence(1992).
The main finding of the neurophysiological study, however, involved HH displaying significantly shorter latencies of a pre-300 negativity at the frontal site for the word (mean for HH: 240ms, for LH: 291ms) and neutral conditions (mean for HH: 250ms, for LH: 291ms) and nearly significant also in the congruent condition. These results provided further support for the automaticity hypothesis (this negativity preceded the P300, thus indexing non-conscious processing, as had been shown in the previous studies). It suggested also that the greater automaticity exhibited by HH subjects was not exclusively verbal but rather perceptual in nature.
A recent correlational study investigating individual differences on several cognitive tasks yielded results in agreement with the automaticity hypothesis. The Target Detection test (D2) derived from Brichenkamp (1966) was administered in combination with the Stroop test and some explicit and implicit memory tasks to a group of subjects under the pretence of a cognition/perception experiment). After completing the cognitive tasks, subjects were asked to return a week later to be assessed for hypnotizability. Hypnosis was not part of the picture prior to the end of the cognitive assessment. The cognitive variables were found to account for 32% of the variance of subjects' hypnotizability scores over and above the contribution of imagery and absorption (Slako, Lepage, & Laurence, 1996). The D(2) test is a neuropsychological measure used to assess frontal lobe damage in the clinical population. In this task, subjects are given a strategy to recognise a complex target and asked to identify and cross out this target among distracters. In addition to this task, subjects were presented with a list of words to memorise and were later tested on an implicit (stem completion) task and an explicit (recognition) one. Indexes of intentional versus automatic memory processes were established using Jacoby's Process Dissociation paradigm (Jacoby, 1991). For a first series of stems, subjects were asked to use words from the studied list when ever possible. For a second series of stems, subjects were told not to use the words from the studied list as much as possible. Following the stem completion tasks, subjects were asked to identify all the words from the studied list among an equal amount of distracter words (recognition task).
Hypnotizability was found to be predicted in part by accuracy on the D2 test. The more hypnotizable subjects were, the less omission errors they made on this task. More hypnotizable subjects were also found to have a better index of intentional recall, that is, they were more effective at controlling the use of words from the studied list to complete stems. Finally, more hypnotizable subjects were observed to be more likely to confuse words from the studied list with distracter words after completing the stems. In summary, hypnotizable subjects may have been more effective at recognising and crossing out targets on a perceptual task, however, the activation of words self-generated to complete stems may have interfered with the activation of words from the studied list. The results suggest that HH are more alert if the perceptual task is simple and if a strategy is applicable. However, when greater familiarity is involved, their perceptual automaticity may make them more vulnerable to confusion.
The investigation of automaticy in memory research has been based on measurement of implicit (automatic) and explicit (conscious) recall. Implicit and explicit forms of recall have been viewed as distinct and separate memory processes (Graf, Mandler & Haden, 1982; Graf & Schacter, 1985; Jacoby, 1991). Traditionally, implicit memory measures have relied on stem and fragment completion tasks while explicit memory measures have relied on recognition and cued recall tasks. The major distinction between explicit and implicit memory processes involves the subject's awareness of the learning episode. The results of our studies designed to investigate subjects' degree of awareness and its relation to hypnotizability have consistently pointed to a view of implicit and explicit memory as end points of a continuum of memory. Awareness was found to be more likely an individual threshold reached through increasing implicit activation ( Slako, Lepage, & Laurence, 1996; Laurence and Tremblay,1996).
In Laurence and Tremblay's 1996 study, the subjects were presented with a list of words and asked to either rate the words for pleasantness (semantic encoding) or to count the number of "T junctions" (structural encoding). The subjects were unaware that they would be later tested for recall on the latter list. Priming was assessed by the number of stems completed using words from the initial list. Level of encoding was not found to affect priming, however, semantic encoding was found to be related to higher performance on a cued recall test. The subjects' level of awareness during and after the task was assessed using post experimental questionnaires. The results indicated that subjects who used more words from the list to complete stems were more likely to become aware of the link between the list of words and the stem completion task. Although priming was present in self-reportedly unaware subjects, awareness was observed to be an outcome of this higher implicit responding. When subjects were later assessed for hypnotizability, more hypnotizable subjects were found to have a larger direct priming performance. These results are consistent with the hypothesis of a higher degree of automatic processing in hypnotizable subjects. However, this priming effect had further measurable consequences: it led to an increased activation of previously learned material, awareness and thus a superior performance on a memory task involving conscious recall.
Another interesting dimension of automaticity in relation to hypnotizability involves the acquisition of automatization, a question that was raised by HH's responses to homogeneous blocks of trials during the ERP study. We investigated individual differences in acquisition of automatization between HH and LH subjects using a Stroop-like perceptual task (Laurence, Blatt, Maestri, & Khodaverdi-Khani, 1997). The perceptual task consisted of learning associations between four colours and four irregular shapes. Subjects learned the task over five sessions spread over 5 days, and received over 2,000 shape-naming practice trials. The shape stimuli were four low-association value random polygons selected from a set developed by Vanderglass and Garvin (1959). The shapes were called Red, Blue, Green and Yellow. Following an acquisition phase, subjects were asked to either name the actual colour of the shape or to name shapes by using their corresponding colour name. When subjects had to name the shape using their corresponding colour name, they exhibited greater facilitation (shorter mean reaction time) on congruent trials, and greater interference (longer mean reaction time) on incongruent trials regardless of their degree of hypnotizability. HH subjects, however, were observed to be significantly slower than LH subjects, at naming a shape or a colour when the two stimuli were incongruent. HH subjects were also found to display the automatic response sought (interference of shape on colour naming) faster than their LH conterparts. Again, these findings are consistent with the hypothesis that HH subjects are faster at learning a task and automatizing it.
Several implications with regards to individual differences in cognitive processing can be foreseen from these studies. Greater facilitation and interference effects on the Stroop tasks for HH subjects has been found quite consistently across studies conducted outside of the hypnotic context. When subjects are aware of the relevance of the hypnotic context, the size of the discrepancy between mean reaction time on congruent and incongruent trials in HH and LH subjects was found to vary. Under such conditions, LH subjects were found to be less motivated to perform on the Stroop and their mean reaction time increased across congruent, neutral and incongruent trials (Labelle, 1994). HH subjects, on the other hand, were not found to vary across contexts. These findings clearly indicate that the subjects' perception of their "good" or "bad" performance during hypnosis modulates their attitude and motivation to perform on the Stroop task (Labelle, 1994).
As other researchers have shown using different methodologies, the results of the studies reviewed here suggest that individual differences in cognitive processing speed at the level of the frontal lobes play a role in hypnotizability. Frontal activity may be involved in the mediation of attentional resources. Because of their ability to process information more rapidly and with less effort, HH subjects may be able to invest more attention in the application of strategies to improve their performance on demanding cognitive tasks. Performance on cognitive tasks associated with frontal lobe activity, attention, and memory performance, out of the hypnotic context, deserve further research effort. Current research findings indicate that such cognitive processing differences are likely to be new (maybe even better) predictors of the hypnotic experience than well-established cognitive correlates of hypnosis such as imagery and absorption. On a more general note, the capacity to process perceptual and verbal information more automatically as well as the ability to automatise attentional strategies may provide an important clue in the understanding of anomalous experiences reported by high hypnotizable subjects (Nadon, Laurence and Perry, 1987). Because both the automatic response and the automatisation of strategic responses happen in a non-aware, involuntary fashion, subjects may have a tendency to confuse the source of their experiences. As the reality monitoring literature has shown, the awareness of the cognitive processing involved in remembering is an important cue to differentiating externally from internally generated events. Perceptual automaticity could then explain in part why hypnotisable subjects are more prone to memory creation and confabulation in general. The hypnotic context, by depotentiating conscious awareness, may amplify the intensity of the subjective experience created by a greater perceptual automaticy.
| Discussion Board | Previous Page | Your Symposium |
|Laurence, J.R.; Slako, F.; Le Beau, M.; (1998). Automaticity, Hypnotizability and the Creation of Anomalous Experiences: Neuro-physiological Indicators. Presented at INABIS '98 - 5th Internet World Congress on Biomedical Sciences at McMaster University, Canada, Dec 7-16th. Invited Symposium. Available at URL http://www.mcmaster.ca/inabis98/woody/laurence0642/index.html|
|© 1998 Author(s) Hold Copyright|