Hypnosis

Abstract

Hypnosis, which MeSH classifies under psychotherapy, is a state of focused attention and heightened response to suggestion, entered after an induction, in which suggested changes in perception, memory, movement, and volition can be experienced as involuntary. This article sets out the modern definition adopted by the American Psychological Association, hypnotic suggestibility as a stable individual trait, and the two traditions that divide the field: the neodissociation account, on which hypnosis involves a genuine division of consciousness, and the sociocognitive account, on which hypnotic responding reflects ordinary expectancy, motivation, and imaginative absorption. It reviews the evidence that suggestion can reach normally automatic processes, the case for hypnotic analgesia, and the debates that remain open. Three demonstrations model the distribution of suggestibility, the reduction of Stroop interference under suggestion, and the effect size of analgesia.

Keywords: hypnosis, suggestion, hypnotic susceptibility, dissociation, analgesia

Hypnosis is a procedure, and a resulting state, in which one person responds to suggestions given by another for changes in subjective experience and behaviour that are experienced as happening by themselves rather than being deliberately produced. The revised definition adopted by the American Psychological Association's Division of Psychological Hypnosis describes it as a state of consciousness involving focused attention and reduced peripheral awareness characterized by an enhanced capacity for response to suggestion (Elkins et al., 2015). It is not sleep, and the older etymology that ties it to sleep is misleading; a hypnotized person is awake, attentive, and able to refuse a suggestion. What distinguishes the state is the quality of the response to suggestion, which acquires a sense of effortlessness and involuntariness, so that a suggested arm levitation or a suggested analgesia is felt as something that occurs rather than something the person does. The capacity to respond in this way varies widely between people and is one of the most stable individual differences in psychology (Piccione et al., 1989).

What Hypnosis Is

A hypnotic session has two parts: an induction, which orients the person toward the procedure and typically invites relaxation, focused attention, and a narrowing of awareness onto the operator's voice; and one or more suggestions, which propose specific changes in experience or behaviour. The induction is not the active ingredient it was once thought to be. Its removal changes hypnotic responding only modestly, and what predicts response is far more the person's suggestibility than the particular ritual used to induce the state (Kirsch, 1985). The suggestions themselves range from motor items, such as an arm growing heavy and lowering, through challenge items, in which the person is told they cannot perform an action such as bending an arm, to the cognitive-perceptual items that are the most demanding: a suggested hallucination, a suggested amnesia, or a suggested analgesia. The defining phenomenological mark across all of them is the classic suggestion effect, the experience that the response is occurring involuntarily.

Hypnosis is best understood not as a single thing a person is either in or out of but as the joint product of a procedure and a trait. The procedure sets the occasion; the trait, hypnotic suggestibility, determines how much of the suggested experience the person actually has. Because that trait is continuous and stable, most contemporary accounts treat hypnotic phenomena as lying on a dimension of responsiveness rather than as a discrete altered state that is switched on (Kihlstrom, 1985). This is why the study of hypnosis is inseparable from the measurement of the individual differences in consciousness and control that it reveals.

Types of Hypnosis

The Medical Subject Headings (MeSH) thesaurus, which indexes the biomedical literature, files hypnosis under psychotherapy and, in a parallel branch, under the mind-body therapies, and it lists two narrower descriptors beneath it (Table 1). These are the categories under which the literature is indexed, not a theory of how hypnotic phenomena are organized; the two subtypes are neither exhaustive of hypnotic technique nor mutually exclusive, since a suggestion is the elementary unit from which any hypnotic procedure, including a self-directed one, is built. MeSH is an indexing classification rather than a taxonomy of the mind, so this list marks how research is catalogued, not a claim about carving hypnosis at its joints.

Measuring Hypnotic Suggestibility

The trait that governs hypnotic response is measured much as any other ability is, by a standardized test administered under fixed conditions. The Stanford Hypnotic Susceptibility Scales, devised by André Muller Weitzenhoffer and Ernest Hilgard, present a standard induction followed by a graded series of suggestions, and score the person on how many produce the criterion response; the widely used Form C runs from zero to twelve. The distribution of scores in the general population is broad and roughly bell-shaped: most people are moderately responsive, a minority at each extreme are refractory or exceptionally responsive, and the small group scoring at the top, sometimes called high hypnotizables or hypnotic virtuosos, are the population in whom the striking phenomena are reliably obtained. The first demonstration presents a characteristic distribution of scores and lets the reader move the cutoff that defines high suggestibility, showing how the proportion classified as highly responsive depends on where that line is drawn.

What makes suggestibility a genuine trait rather than a transient mood is its stability. In a striking longitudinal study, participants retested on the Stanford scale after an interval of twenty-five years showed a test-retest correlation of about 0.71, a stability coefficient rivalling that of measured intelligence over comparable spans (Piccione et al., 1989). Suggestibility is not readily trained: attempts to raise it produce only modest and often temporary gains, and the ceiling a person can reach is largely fixed. This stability is a central fact any theory of hypnosis must accommodate, and it is one reason the trait is treated as a window onto durable differences in the organization of attention, imagination, and cognitive control.

Two Theories: Dissociation and Social Cognition

The enduring theoretical divide in hypnosis research is between accounts that treat it as a special state involving an alteration of consciousness and accounts that explain it through ordinary psychological processes. The state tradition's most developed form is Ernest Hilgard's neodissociation theory. Observing that a hypnotized person given analgesia would report little or no pain while a concurrent, separately elicited channel, the hidden observer, reported the pain at close to its normal level, Hilgard proposed that hypnosis divides the stream of control into parallel cognitive subsystems separated by an amnesic barrier, so that pain is registered by one system while being kept from the reporting consciousness of another (Hilgard, 1973). On this view the involuntariness of hypnotic response is real: executive control is genuinely dissociated from the subsystem executing the suggestion.

The sociocognitive tradition rejects the special-state premise. Nicholas Spanos and others argued that hypnotic behaviour is goal-directed action that the person interprets as involuntary, shaped by expectancies, attitudes, motivations, and the situational demand to enact the role of a hypnotized subject; even the hidden observer, on this reading, is a product of the instructions that elicit it rather than evidence of a pre-existing division (Spanos, 1986). Irving Kirsch's response-expectancy theory supplied the mechanism: responses to suggestion are mediated by the person's expectancy that they will occur, a self-confirming anticipation that generates the experience much as a placebo generates its effect, without any altered state being required (Kirsch, 1985). A third contemporary account, the cold control theory of Zoltan Dienes and Josef Perner, locates the phenomenon in metacognition rather than in a divided consciousness or in role enactment: the hypnotized person forms and executes the intention to act on a suggestion but holds an inaccurate higher-order thought about that intention, so a genuinely produced action is experienced as unintended. The account makes testable predictions about prefrontal control, and disrupting the left dorsolateral prefrontal cortex with low-frequency stimulation has been found to increase hypnotic suggestibility, as the theory anticipates (Dienes & Hutton, 2013). The traditions are not fully reconciled, but the modern position increasingly treats them as describing different levels of a single phenomenon, with top-down attention and expectancy shaping experiences that are, in the most responsive people, genuinely altered (Terhune et al., 2017).

Suggestion Reaching Automatic Processes

The sharpest evidence that hypnotic suggestion does something cognitively real comes from its effect on processes that are normally automatic and involuntary. The Stroop task is the standard case: reading is so overlearned that the word meaning intrudes on naming the ink colour, producing the robust interference that defines the effect, and this intrusion is taken to be beyond voluntary control. Amir Raz and colleagues gave highly suggestible participants a posthypnotic suggestion that the words would appear as meaningless symbols in an unfamiliar script, then measured Stroop performance. In the highly suggestible group the suggestion sharply reduced, and in some participants abolished, the Stroop interference effect, an outcome accompanied by reduced activity in the visual word-form area and the anterior cingulate (Raz et al., 2005). Interference did not fall in participants low in suggestibility given the same suggestion, tying the effect to the trait rather than to instruction alone. The second demonstration models this result, letting the reader set a suggestibility level and see how the suggested word-blindness scales down the interference.

The theoretical weight of the finding is that a suggestion appears to modulate an early, obligatory stage of processing rather than merely biasing a late decision. It shows that the top-down control marshalled under hypnosis can reach into the machinery of selective attention and alter what is usually automatic, which is why hypnosis has become a tool for probing the limits of cognitive control and the boundary between automatic and controlled processing (Oakley & Halligan, 2013). It also connects hypnotic responding to the wider study of metacognition and volition, since what the suggestion changes is not only performance but the person's sense of authorship over their own response.

The Neural Signature of Hypnosis

If hypnosis is a distinctive state, it should leave a signature in brain activity, and imaging has begun to describe one, though with less consistency than early enthusiasm implied. A systematic meta-analytic review by Landry and colleagues found that hypnosis and hypnotic suggestion are associated with changes in a network including the prefrontal cortex, the anterior cingulate cortex, and the insula, regions supporting executive control, conflict monitoring, and the sense of agency, together with altered connectivity between the executive-control and default-mode networks (Landry et al., 2017). Figure 1 sketches this cortical network and the functions each node contributes.

The most reliable functional finding is the one from the Stroop work: when highly suggestible people are given a suggestion that reduces cognitive conflict, activity in the anterior cingulate, the conflict-monitoring hub, falls in step with the behavioural reduction (Raz et al., 2005). Oakley and Halligan reframed this pattern by treating hypnotic suggestion as a controlled way of producing and studying alterations in experience and volition, so that suggested paralysis, suggested blindness, and suggested pain become experimental models for the neural basis of belief, agency, and subjective experience (Oakley & Halligan, 2013). Whether there is a single neural signature of the hypnotic state itself, as opposed to signatures of particular suggestions, remains unsettled, and the heterogeneity of the imaging literature is a live methodological concern (Landry et al., 2017).

Clinical Hypnosis and Analgesia

The best-supported clinical use of hypnosis is in the control of pain, where suggestion can be turned directly on the reward and aversion circuitry that registers noxious input. A meta-analysis by Montgomery and colleagues of hypnotic analgesia across controlled studies found a mean effect size of about 0.67, meaning the average person given hypnotic analgesia obtained more relief than roughly three-quarters of untreated controls, with the largest benefits in the most suggestible participants (Montgomery et al., 2000). Later syntheses extended the finding to clinical pain: Jensen and Patterson reviewed the evidence for hypnosis in chronic pain and concluded that it produces reliable reductions in pain intensity that can outlast the sessions themselves (Jensen & Patterson, 2014). A large meta-analysis of eighty-five controlled experimental trials confirmed a moderate-to-large analgesic effect and, again, its dependence on hypnotic suggestibility (Thompson et al., 2019). The third demonstration makes the effect-size language concrete, translating a Cohen's d into the percentage of treated people who do better than the average control.

Beyond pain, the clinical evidence is more mixed. Hypnosis has respectable support as an adjunct in irritable bowel syndrome and in procedural distress, and as an add-on that improves the outcomes of cognitive-behavioural treatment, but claims for its power to recover buried memories are not merely unsupported but actively harmful, since hypnotic procedures inflate confidence in recollections without improving their accuracy and can implant false memories. The clinical and forensic literatures accordingly distinguish sharply between hypnosis as a legitimate tool for symptom control, where its evidence base for analgesia is strong, and hypnosis as a supposed route to accurate memory retrieval, where its use is contraindicated.

Worked Example

The claim that hypnotic analgesia has an effect size of about 0.67 is easy to misread as a small number, so it is worth converting into a more intuitive quantity, exactly as the third demonstration does. Cohen's d expresses the difference between two group means in units of their pooled standard deviation. Suppose a trial of experimental pain reports a mean pain reduction of 4.7 points on a scale under hypnosis and 2.0 points under a no-treatment control, with a pooled standard deviation of 4.0 points. Then d is the difference of the means, 4.7 minus 2.0, which is 2.7, divided by the pooled standard deviation of 4.0, giving 0.675, or about 0.67, matching the meta-analytic mean.

To make that value meaningful, translate it into a superiority percentage: the proportion of the treated distribution that lies above the mean of the control distribution, which for normally distributed outcomes is the standard normal cumulative probability evaluated at d. Evaluating the standard normal at 0.67 gives about 0.749. So a person receiving hypnotic analgesia obtains, on average, more relief than roughly 75 percent of untreated controls. Put the other way, if a treated patient and a control patient were drawn at random and compared, the treated patient would report the greater relief about three times in four. The number 0.67, which looks modest as a raw d, is a clinically substantial effect once its overlap is unpacked, and it is why hypnosis is taken seriously as an analgesic technique even by researchers sceptical of the special-state theory of how it works.

Discussion

Hypnosis occupies an unusual position in psychology: a phenomenon whose reality is not in doubt, whose central measurement is exemplary, and whose explanation remains genuinely contested. The trait of suggestibility is one of the most reliably measured and most stable individual differences in the field, and the phenomena it predicts, from suggested analgesia to the modulation of Stroop interference, are robust and replicable (Piccione et al., 1989; Raz et al., 2005). What is unsettled is the mechanism. The neodissociation tradition reads the involuntariness of hypnotic response as evidence of a real division of control, while the sociocognitive tradition reads it as a compelling interpretation the person places on goal-directed action shaped by expectancy (Hilgard, 1973; Spanos, 1986; Kirsch, 1985).

The contemporary synthesis is less a resolution than a division of labour. Expectancy and social context clearly modulate hypnotic responding, as the sociocognitive tradition insists, yet the imaging evidence that suggestion reaches early, obligatory processing in the most suggestible people is hard to fold entirely into a role-enactment account (Terhune et al., 2017). The likeliest picture is that top-down attention and expectancy set up experiences that are, in high hypnotizables, genuinely altered at the level of perception and control, so that the two traditions describe complementary parts of one process. Two problems keep the field honest. The first is the heterogeneity of high hypnotizables, who are not a uniform group but appear to reach high scores by different cognitive routes, which blurs the search for a single neural signature (Terhune et al., 2017). The second is methodological: much of the older literature is underpowered and inconsistently controlled, and the imaging findings in particular are more variable than summary statements suggest (Landry et al., 2017). What is not in doubt is that a verbal suggestion, given to the right person, can change what is seen, felt, and controlled, and that this makes hypnosis a uniquely direct instrument for studying the top-down construction of experience.

Current Directions

The most active current programme treats hypnosis less as a curiosity to be explained than as an instrument for manipulating consciousness under experimental control. A consensus statement by Jensen and a large group of collaborators set out a research agenda framing hypnosis as top-down regulation, and urging that studies stratify participants by suggestibility, use adequately powered designs, and distinguish the neural correlates of the hypnotic state from those of specific suggestions (Jensen et al., 2017). A parallel theoretical move, developed by Terhune and colleagues, reconceives hypnotic phenomena as instances of the general capacity for top-down regulation of consciousness, placing hypnosis within mainstream cognitive neuroscience rather than treating it as a special case, and emphasizing that highly suggestible individuals are heterogeneous, reaching high suggestibility through distinct underlying mechanisms (Terhune et al., 2017). On the clinical side, the largest and most rigorous meta-analysis to date, covering eighty-five controlled experimental trials, has put the analgesic effect on a firmer quantitative footing and confirmed that its magnitude scales with suggestibility, sharpening the case for matching the technique to responsive patients (Thompson et al., 2019). The unifying thread is a shift from asking whether hypnosis is real to using it, as a controlled and reversible perturbation, to test theories of attention, agency, and the perception of pain.

Common Misconceptions

Hypnosis is a form of sleep or unconsciousness.

It is not. A hypnotized person is awake and attentive, with electroencephalographic activity unlike that of sleep; the state is defined by focused attention and heightened response to suggestion, not by a loss of consciousness, and the sleep connotation is a historical artefact of the name coined from the Greek for sleep (Elkins et al., 2015).

A hypnotized person is under the operator's control and can be made to act against their will.

Hypnotic response depends on the person's own suggestibility and cooperation, not the operator's power; suggestions can be refused, and on the response-expectancy account the effect is generated by the person's own anticipation rather than imposed from outside (Kirsch, 1985).

Hypnosis can recover accurate buried memories.

It cannot. Hypnotic procedures raise a person's confidence in recovered recollections without raising their accuracy, and can implant false memories, which is why hypnotically refreshed testimony is treated with grave caution in forensic settings and why memory recovery is a discredited use of the technique (Kihlstrom, 1985).

Glossary

Absorption.

A trait disposition to become deeply engaged in imaginative and perceptual experience, modestly correlated with hypnotic suggestibility and often invoked to characterize the cognitive style of high hypnotizables.

Analgesia, hypnotic.

The reduction of experienced pain produced by hypnotic suggestion; the best-evidenced clinical effect of hypnosis, with a meta-analytic effect size around 0.67 that scales with suggestibility.

Autogenic training.

A self-directed relaxation method using repeated formulaic suggestions of warmth and heaviness to induce a calm autonomic state; a MeSH subtype of hypnosis used for stress-related complaints.

Challenge item.

A hypnotic suggestion that the person cannot perform an action, such as bending an arm held rigid; scored by whether the person fails the challenge, indexing intermediate suggestibility.

Cold control theory.

Dienes and Perner's higher-order-thought account on which a hypnotic response is a genuinely executed intention accompanied by an inaccurate higher-order thought about that intention, so the action is experienced as involuntary; it locates hypnosis in metacognition rather than a divided or role-enacted consciousness.

Hidden observer.

In Hilgard's neodissociation theory, a concurrently accessible cognitive channel that reports experiences, such as pain under hypnotic analgesia, kept from the person's reporting consciousness by an amnesic barrier.

Hypnosis.

A state of focused attention and reduced peripheral awareness, entered after an induction, characterized by an enhanced capacity to respond to suggestion with experiences felt as involuntary.

Hypnotic suggestibility.

The measurable, stable trait determining how strongly a person responds to hypnotic suggestion; assessed by standardized scales and roughly normally distributed in the population.

Induction.

The procedure that opens a hypnotic session, typically inviting relaxation and focused attention; its specific form contributes little to responding beyond orienting the person to the task.

Neodissociation theory.

Hilgard's account on which hypnosis divides the stream of cognitive control into parallel subsystems separated by an amnesic barrier, explaining the felt involuntariness of hypnotic response as a real dissociation.

Response expectancy.

Kirsch's construct: a person's anticipation of an automatic subjective response, which is self-confirming and generates the expected experience; the proposed mechanism of hypnotic and placebo effects.

Sociocognitive theory.

The tradition explaining hypnotic behaviour through ordinary processes, expectancy, motivation, attitudes, and imaginative involvement, as goal-directed action interpreted by the person as involuntary, without a special altered state.

Stanford Hypnotic Susceptibility Scales.

The standard instruments, devised by Weitzenhoffer and Hilgard, that administer a fixed induction and graded suggestions and score suggestibility, with Form C running from zero to twelve.

Suggestion effect.

The defining phenomenological mark of hypnotic response: the experience that a suggested action or perception occurs involuntarily, by itself, rather than being deliberately produced.

Suggestion.

The elementary process by which an accepted idea produces a corresponding change in experience or behaviour; the unit of hypnotic response, and a MeSH subtype of hypnosis.

Key Researchers

Ernest R. Hilgard (1904-2001). Professor of Psychology at Stanford University; with André Muller Weitzenhoffer he devised the Stanford Hypnotic Susceptibility Scales, and he formulated neodissociation theory and the hidden observer to explain hypnotic analgesia as a division of consciousness. Wikipedia - Wikidata

John F. Kihlstrom. Professor Emeritus of Psychology at the University of California, Berkeley; he brought hypnosis into the study of the cognitive unconscious, integrating posthypnotic amnesia and implicit memory into mainstream cognitive psychology and providing an influential critical synthesis of the field. Faculty Page - Google Scholar - Wikipedia - ORCID

Irving Kirsch. Associate Director of the Program in Placebo Studies at Harvard Medical School and emeritus professor at Connecticut and Plymouth; his response-expectancy theory recast hypnosis as a nondeceptive placebo, mediated by anticipation rather than an altered state. Google Scholar - Wikipedia - Wikidata

Steven Jay Lynn (1946-2024). Distinguished Professor of Psychology at Binghamton University; a leading proponent of the sociocognitive perspective, he documented the malleability of hypnotic responding, the risks of hypnotically recovered memory, and the myths that surround the field. ORCID

Amir Raz. Professor and Chair in Clinical Neuroscience at Chapman University; using a Stroop paradigm under posthypnotic suggestion, he showed that suggestion can reduce or abolish the normally automatic Stroop interference effect in highly suggestible people. Google Scholar - Research Group

David Spiegel. Willson Professor of Psychiatry and Behavioral Sciences at Stanford University School of Medicine; he co-authored the revised APA definition of hypnosis and led clinical and neuroimaging research on hypnotic analgesia and the functional signatures of the hypnotic state. Google Scholar - Wikipedia - Wikidata

Devin B. Terhune. Reader in Psychology at King's College London; he advanced the measurement and cognitive neuroscience of hypnotic suggestibility, showing that highly suggestible people are heterogeneous, and co-authored the reframing of hypnosis as top-down regulation of consciousness. Faculty Page - Google Scholar - Wikipedia - ORCID

Frequently Asked Questions

What is hypnosis? Hypnosis is a state of focused attention and reduced peripheral awareness, entered after an induction procedure, characterized by an enhanced capacity to respond to suggestion with experiences felt as involuntary (Elkins et al., 2015).

Can everyone be hypnotized? No. Hypnotic response depends on suggestibility, a stable trait that varies widely; most people are moderately responsive, a minority are highly responsive, and a minority respond little, so the striking phenomena are reliably obtained only in high hypnotizables (Piccione et al., 1989).

Is hypnosis a kind of sleep? No. A hypnotized person is awake and attentive, with brain activity unlike that of sleep; the sleep connotation is an artefact of the term's Greek root, and the defining feature is heightened response to suggestion, not loss of consciousness (Kihlstrom, 1985).

Can a hypnotist make someone act against their will? No. Responding depends on the person's own cooperation and expectancy rather than the operator's power, and suggestions can be refused; the response-expectancy account holds that the effect is generated by the person's own anticipation (Kirsch, 1985).

Does hypnosis actually reduce pain? Yes. Meta-analyses of controlled trials find that hypnotic analgesia produces moderate-to-large reductions in experimental and clinical pain, with a mean effect size around 0.67 that is largest in the most suggestible participants (Montgomery et al., 2000).

Can hypnosis recover lost memories? No. Hypnotic procedures increase confidence in recovered memories without increasing their accuracy and can implant false memories, which is why hypnotically refreshed testimony is treated with great caution and memory recovery is a discredited use (Kihlstrom, 1985).

Is hypnosis a special altered state or just social role-play? The field is divided: state and neodissociation accounts posit a real division of consciousness, while sociocognitive accounts explain responding through expectancy and motivation, and the modern view treats these as complementary levels of one phenomenon (Terhune et al., 2017).

Can suggestion change automatic processes like reading? In highly suggestible people it can: a posthypnotic suggestion that words are meaningless symbols reduces or abolishes the normally automatic Stroop interference effect, with a matching drop in anterior cingulate activity (Raz et al., 2005).

References