How and why Hypnosis Reveals the Nature of Consciousness
Hypnosis is the only state that has been systematically induced in hundreds of thousands of people, in controlled laboratory settings, for more than two centuries, in which the ordinary structure of consciousness is reliably and dramatically rearranged by nothing more than words. No substance is ingested, no electrode implanted, no brain region removed; only a human voice (or, in self-hypnosis, oneโs own inner voice) speaks, and within minutes the stream of experience can be narrowed, widened, split, inverted, or apparently abolished altogether. If we want to know what consciousness is made of, how plastic it is, and how tightly or loosely it is tied to the rest of the mind, hypnosis is the most powerful natural experiment we possess.
The Phenomenology of the Hypnotic Trance
When a highly hypnotisable person (a โhighโ on the Stanford or Harvard scales, roughly the top 10โ15 % of the population) enters deep trance, the first thing that usually changes is the sense of agency. A suggested arm levitation feels as though the arm is rising โby itself,โ yet the movement is perfectly normal on EMG and fMRI; the motor cortex, parietal cortex, and cerebellum are active exactly as in voluntary movement. The difference is that the forward model of intention (the predictive signal that normally accompanies willed action) is attenuated or absent. The action is executed, but the feeling โI am the authorโ is subtracted. Conversely, in suggested paralysis the motor cortex is inhibited downstream by prefrontal and anterior cingulate mechanisms, yet the subject often reports that the arm โwants to move but cannot,โ as if an external force is holding it.
Time distorts. Ten real minutes can feel like two or ninety. In deep somnambulism, subjects may later insist that โno time passed at all.โ The default-mode network (DMN) shows reduced activity and altered connectivity with the salience network, exactly the pattern seen in meditators and under psilocybin, suggesting that the ordinary narrative self, which is heavily DMN-dependent, is partially suspended.
Sensory experience becomes extraordinarily malleable. Positive hallucinations (โYou now see a pink elephant in the cornerโ) recruit the same occipital, temporal, or parietal areas as real perception; negative hallucinations (โYou no longer see the chair in front of youโ) produce top-down inhibition of those same areas even when the eyes are open and fixating the object. Pete Lush and Zoltan Dienes (2020โ2024) have shown that the neural signature of suggested blindness is indistinguishable from the signature of a real blind person looking at the same stimulus: V1 is suppressed by prefrontal feedback before the signal even reaches awareness.
Pain can be abolished or created on command. In the cold-pressor test, highly hypnotisable subjects who receive the suggestion โyour hand has no feeling, it belongs to someone elseโ show no pain-related activation in anterior cingulate, insula, or somatosensory cortex, and they report zero pain despite clear autonomic arousal (heart rate, galvanic skin response). Conversely, suggested pain in the absence of noxious stimulation activates exactly the same โpain matrix.โ Hypnotic analgesia is now used routinely in surgery (burn debridement, dental work, childbirth) when pharmacological anaesthesia is contraindicated.
Memory becomes optional. Post-hypnotic amnesia can erase entire trance episodes; the suggestion โwhen you open your eyes you will remember nothing that happenedโ produces a reversible functional ablation of episodic recall that is deeper than any pharmacological amnesia. Yet the memories are not destroyed; a pre-agreed cancellation cue restores them instantly and completely. This demonstrates that the contents of consciousness during trance are laid down in long-term memory normally; what is altered is the higher-order metacognitive tagging that normally renders them accessible to narrative self.
Identity itself can be rewritten. In deep somnambulism, age regression can convince a 45-year-old that she is literally six years old again; handwriting, vocabulary, and even IQ revert to childhood levels, and later memory tests show source amnesia for adult knowledge. Suggested past-life regression (whether one believes the content or not) produces vivid, detailed phenomenological states that subjects experience as real memory. Multiple personalities can be installed and removed at will; the famous case of โSybilโ is now widely regarded as iatrogenic, created by an overzealous therapist using hypnotic techniques.
The Architecture of Consciousness Revealed by Suggestion
From these phenomena a remarkably clear picture of the modular, hierarchical, and heavily top-down nature of ordinary consciousness emerges:
a) The โselfโ is not a primitive; it is a constructed model Hypnosis shows that the feeling of being a single, continuous, volitional agent is a high-level inference generated by prefrontalโparietalโcingulate networks. Turn down the gain on those networks (as hypnotic suggestions do) and the sense of self fragments or vanishes without loss of perception, memory, or intelligence. This is precisely what Grazianoโs โattention schemaโ theory and Metzingerโs โself-modelโ theory predicted: the self is a useful fiction maintained for the sake of social prediction and motor control.
b) Phenomenal experience is under direct prefrontal control The fact that a spoken sentence can abolish pain, create colour, erase objects, or implant false memories proves that the final contents of consciousness are not determined bottom-up by sensory transduction but top-down by expectation and suggestion. Predictive-coding models (Friston, Hohwy, Clark) receive their strongest confirmation here: perception is the brainโs best hypothesis, and in hypnosis the hypothesis is being edited in real time by linguistic input.
c) There is no hard boundary between imagination and perception Highly hypnotisable individuals do not merely โimagine vividlyโ; they literally perceive the suggested reality. fMRI shows that suggested visual hallucinations activate V1 and higher visual areas to the same degree as real vision; the only difference is the source of the top-down signal (prefrontal cortex instead of thalamus). This collapses the ordinary distinction between imagination and perception and strongly favours representationalist theories of qualia: what makes red โfeel like redโ is the activation of the same neural representation, regardless of distal cause.
d) Conscious will is a post-hoc interpretation Wegnerโs work on the illusion of conscious will (2002) finds its most dramatic confirmation in hypnosis. Automatic writing, ideomotor responses, and post-hypnotic acts are executed flawlessly, yet the subject experiences them as happening โby themselves.โ The readiness potential appears before the subject is aware of any intention, exactly as in Libetโs classic experiments, but in hypnosis the feeling of authorship can be attached, detached, or reversed at will.
e) There is a deep continuity between hypnosis, meditation, and psychedelics All three states involve reduced DMN activity, disrupted self-model, increased prefrontalโsensory coupling, and heightened suggestibility. The difference is only in the direction of control: meditation trains endogenous top-down regulation, hypnosis uses exogenous (hetero- or auto-) suggestion, psychedelics temporarily dismantle the regulatory hierarchy altogether. Hypnosis is therefore the only one of the three that is fully reversible, on command, within seconds.
Theoretical Implications
- Illusionism receives powerful support. If a few sentences can make pain vanish, objects disappear, or an arm feel alien, then the ordinary conviction that qualia are intrinsic, ineffable, and privately certain begins to look like a user-interface trick. The brain is running a controlled hallucination at all times (Anil Seth); hypnosis simply gives the experimenter the password to edit the hallucination directly.
- The hard problem shrinks dramatically when you realise that the very same neurons that fire during real pain also fire during imagined pain, and that hypnotic suggestion can toggle the phenomenal experience on and off while leaving the neural firing pattern largely unchanged (or vice versa), the question โwhy does this neural process feel like anything?โ starts to sound misdirected. The feeling is the process, seen from the inside under a particular attentional framing. Hypnosis is the only state in which we can watch the framing being installed and removed in real time.
- Global Workspace and Higher-Order theories both survive, but with modifications, Dehaeneโs GNW is confirmed: suggestions must be broadcast globally to take effect; localised suggestions (e.g., to a single finger) fail. But higher-order monitoring is radically altered: subjects in deep trance often report no inner voice, no metacognition, yet they follow complex suggestions perfectly. This suggests that HOT theories must allow for unconscious higher-order representations, or else abandon the claim that all consciousness requires meta-awareness.
- Integrated Information Theory struggles IIT predicts that deep hypnosis should have lower ฮฆ than waking consciousness because effective connectivity is reduced (fewer modules are talking to one another). Yet highly hypnotisable subjects in deep trance report richer, more vivid phenomenology than normal waking (synesthesia, hypermnesia, profound emotional release). This is the opposite of what IIT expects and remains one of the theoryโs most serious empirical challenges.
- Non-dual traditions are vindicated in a surprising way. Advaita Vedฤnta and Kashmir Shaivism insist that ordinary waking consciousness is already a trance, a self-induced hypnosis maintained by habitual language and belief. The guruโs or textโs function is to issue a counter-suggestion (โyou are not the body-mindโ) that cancels the original trance. Modern hypnosis demonstrates that such counter-suggestions can indeed produce temporary states indistinguishable from the classical descriptions of samฤdhi or sahaja-avasthฤ: no self, no time, no inside/outside boundary, yet perception and cognition remain perfectly functional.
Hypnosis as the Royal Road to the Nature of Consciousness
Hypnosis is the only altered state that is (a) fully reversible on command, (b) inducible in the laboratory with no drugs or equipment, (c) measurable at every level from single-neuron to whole-brain to verbal report, and (d) capable of producing the entire range of phenomenological distortions that philosophers have ever claimed to be metaphysically impossible (zombie-like action without agency, perception without a perceiver, pain without suffering, identity without continuity).
It therefore functions as a kind of philosophical solvent. Any theory of consciousness that cannot explain how a few well-chosen words can abolish pain, create colour, erase memory, fracture identity, or induce the conviction that one has always been a cat must be incomplete. And any theory that treats ordinary waking consciousness as the immutable ground of reality has already failed the hypnotic test.
In the end, hypnosis does not solve the hard problem; it relocates it. The mystery is no longer โHow can neurons give rise to experience?โ but rather โHow is the brain able to maintain the illusion of a stable, unitary, privately certain self in the face of evidence that the entire edifice can be dismantled and rebuilt by suggestion in minutes?โ The answer that emerges from two centuries of hypnotic research is both sobering and liberating: consciousness is not a deep, hidden property added to the world by mysterious extra ingredients; it is a controlled, high-speed, self-updating story that the brain tells itself about what it is doing right now; and stories, as every hypnotist knows, can be rewritten.
Tanmoy Bhattacharyya
29th November 2025
