Can hypnosis displace the threshold for visual consciousness?

Abstract

To test the specific effects of hypnosis on the attentional components of visual awareness, we developed a posthypnotic suggestion for peripheral visual inattention inspired on the "tunnel vision" symptom of the Balint Syndrome. We constructed a dual-target visibility and discrimination paradigm, in which single-digit numerical targets were placed both on the hypnotically affected peripheral space and on the remaining undisturbed central area. Results were 3-fold: (i) when compared to participants of Low hypnotic susceptibility (Lows), highly susceptible participants (Highs) presented decreased subjective visibility; (ii) Highs did not show dual-task interference from peripheral targets (an effect of unconscious processing) during hypnotic suggestion to not attend them, but Lows did; (iii) nevertheless, when asked to execute a discrimination task over these same targets, Highs performed with the same accuracy as Lows. These results suggest that the hypnotic manipulation of visuospatial attention did produce an experiential change in Highs, but not one that could be mapped onto interference at a single (conscious or unconscious) level of processing. Rather, we posit that Highs simultaneously displayed (i) a fluctuation in awareness of peripheral targets coherent with the suggestion and (ii) a control strategy that involved removing hypnotically unattended targets from the task set whenever task instructions would allow for it. In light of these findings, we argue that hypnosis cannot be used as a tool to restrict the processing of otherwise supraliminal stimulation to subliminal levels.

Keywords: cognitive control; hypnosis; hypnotic induction; posthypnotic suggestion; semantic priming; subjective visibility; task set; visuospatial attention.

Figure 1.

Graphical outline of the trial structure. All trials presented an identical structure, consisting of a Peripheral Target of variable duration (0 for a control baseline, i.e. no target, 16, 33, 67, and 84 ms) presented at either one of the four pointers set around the center of the screen, immediately followed by a Central Target of fixed duration (50 ms) displayed inside of the central ellipse. Only the task changed across blocks. Block order was balanced across participants.

Figure 2.

Visibility rating task. Subjective visibility (PAS scale) for all Peripheral Target durations and hypnotizabilities. Visibility rose globally with stimulus energy (P < 0.0001). Differences between hypnotizability groups increased with Peripheral target duration. SE bars calculated over grand mean minus participant mean, Morey-corrected.

Figure 3.

Visibility rating task. Subjective visibility (PAS scale) for all Peripheral Target durations, for highly susceptible participants with and without a hypnotic induction. Visibility was lower for those who went through the Induction process (P < 0.05) and increased with Peripheral Target duration (P < 0.0001). Visibility differences between Induction and No Induction groups increased significantly as a function of Peripheral Target Duration (P < 0.0001). SE bars calculated over grand mean minus participant mean, Morey-corrected.

Figure 4.

Visibility rating task. Subjective visibility (PAS scale) for all Peripheral Target durations, for highly susceptible participants who received a non-hypnotic suggestion and Low susceptibility participants who underwent the full hypnosis process. Visibility differences between both groups increased as a function of Peripheral Target duration (P < 0.0001). SE bars calculated over grand mean minus participant mean, Morey-corrected.

Figure 5.

Central discrimination task. Accuracy scores (percentage correct) for congruent and incongruent trials, across Low and High susceptibility participants. Accuracy differences related to Congruency were larger for the Low group (P < 0.05). Accuracy remained unaffected by Congruency for the High group (P > 0.08, BF = 40.4 in favor of the null model; model on the High population alone). SE bars calculated over grand mean minus participant mean, Morey-corrected.

Figure 6.

Peripheral discrimination task. Accuracy scores (percentage correct) for all collapsed Peripheral Target durations, across both groups of participants. Results showed a significant Interaction between Hypnosis × Congruency (P < 0.0001). SE bars calculated over grand mean minus participant mean, Morey-corrected.