Hypnosis

In brief

Hypnosis is a procedure, not a substance and not sleep. One person (the hypnotist) offers suggestions, and another (a willing subject) responds with changes in perception, memory, mood, or the sense of control over their own actions. In responsive subjects these effects are genuine and measurable: a suggested numbness that lets a person tolerate a hand in ice water far longer than they otherwise could, an arm that feels like it is rising by itself, a suggested inability to smell ammonia. What research has repeatedly failed to establish is that any of this requires a special trance state. The central scientific question is not whether hypnotic responses happen (they do) but what produces them, and the honest answer, after more than a century of study, is that the field is still divided.

The full treatment

The problem it answers

People differ enormously in how they respond to suggestion, and a subset respond dramatically: told their arm is paralyzed, they cannot lift it; told a color is absent, they report not seeing it. Hypnosis is the label for the procedure that reliably elicits and studies these responses. It answers a practical question, how to harness suggestion for clinical benefit, and a theoretical one, what such responses reveal about the ordinary architecture of attention, volition, and belief.

How it works

A standard session begins with an induction, a set of instructions to relax and focus attention, though research shows the relaxation is largely stage dressing: alert, even exercising, "inductions" work about as well. What follows are specific suggestions, and the subject's responsiveness is the whole phenomenon. That responsiveness is a stable trait, measured since the 1960s by the Stanford Hypnotic Susceptibility Scales (Andre Weitzenhoffer and Ernest Hilgard), which score how many standardized suggestions a person passes. Scores are roughly normally distributed and impressively stable over time, with a minority of "highs" who show the vivid effects and a comparable group of "lows" who show almost none. The trait, not the ritual, predicts the response.

What it claims

The strong claim, associated with "state" theorists, is that hypnosis induces an altered state of consciousness in which suggestions bypass normal critical judgment and act more directly on experience. The weaker, near-universal claim is descriptive: suggestions produce real, sometimes profound alterations in perception and the experienced sense of agency, and the responses are not mere lying or conscious playacting. The disputed territory is everything between those two.

The key demonstration

Ernest Hilgard's "hidden observer" studies are the signature experiment for the state view. A highly hypnotizable subject given a suggestion for pain analgesia, then asked to submerge a hand in circulating ice water (the cold pressor task), reports little or no pain. But Hilgard found that if he instructed a "hidden part" of the person to report separately, often by automatic writing or a key press, that channel reported pain at close to normal levels while the person's spoken report stayed calm. Hilgard read this as evidence that consciousness had literally divided: an amnesic barrier separated the part registering the pain from the part reporting to the experimenter. This became the empirical anchor of his neodissociation theory.

Related distinctions

Hypnosis is not sleep (EEG shows waking patterns), not unconsciousness, and not a state in which people can be made to act against deep values. It should be kept distinct from suggestibility in general (hypnosis is one route to it), from the placebo response (overlapping but not identical), and above all from the popular fantasy of the hypnotist's control. It is also distinct from "recovered memory" work, where suggestion under hypnosis is now understood as a way to manufacture confident false memories rather than retrieve buried true ones.

Lineage

The direct ancestor is Franz Anton Mesmer (1734 to 1815), whose "animal magnetism" a 1784 French royal commission, which included Benjamin Franklin and Antoine Lavoisier, tested and demolished, concluding the effects came from imagination, not any magnetic fluid. That verdict, that the power lay in the subject's expectation, foreshadows the whole modern debate. The word hypnosis comes from the Scottish surgeon James Braid (1795 to 1860), who in the 1840s recast the phenomenon as a neurological and psychological one rather than an occult force. In late nineteenth-century France it split the leading neurologists: Jean-Martin Charcot at the Salpetriere treated hypnosis as a sign of pathology tied to hysteria, while Hippolyte Bernheim in Nancy argued, correctly as it turned out, that it was ordinary suggestibility present in normal people. Sigmund Freud used hypnosis early, then abandoned it for free association, a turn that pushed it to the margins of respectable science for decades until the experimental revival led by Hilgard, Orne, Barber, and Spanos from the 1950s onward.

The strongest case for it

Two things are firmly established, and they are not small. First, hypnotic responding is a real, measurable, stable individual difference, not compliance or faking. Martin Orne's real-simulator design is the decisive tool: subjects who cannot be hypnotized are instructed to fake it convincingly, and blind experimenters cannot tell them apart on gross behavior, yet on fine measures the genuine highs differ. Orne's classic finding was "trance logic," a tolerance for contradiction (seeing a hallucinated person and the real person occupying the same chair) that simulators, trying to be convincing, do not produce. That dissociation between the truly responsive and the merely compliant is hard to explain if nothing distinctive is happening.

Second, hypnosis works clinically for some conditions, and the evidence is not anecdotal. It produces reliable, replicated reductions in acute and chronic pain in laboratory and clinical settings; meta-analyses find medium-to-large analgesic effects, larger in high-suggestibility patients. It has controlled support for irritable bowel syndrome (Peter Whorwell's gut-directed hypnotherapy) and as an adjunct in surgery and childbirth. Neuroimaging by David Spiegel, Pierre Rainville, and others shows that hypnotic suggestions to change pain unpleasantness produce corresponding changes in the anterior cingulate cortex, evidence that subjects are altering perception, not merely their reports.

The strongest case against it

The case against the special-state view is that everything hypnosis does can be produced without any trance, and the "state" adds no explanatory power. This is the sociocognitive (or nonstate) tradition of Theodore X. Barber, Nicholas Spanos (1942 to 1994), and Irving Kirsch. Their evidence is formidable. Barber showed that motivated, non-hypnotized subjects given the same suggestions and encouragement respond about as strongly as hypnotized ones, so the induction contributes little. Spanos reinterpreted the hidden observer as an artifact of the experimenter's instructions: tell subjects the hidden part will feel less pain and it does, tell them it feels more and it feels more. The "divided consciousness" tracks the demand, not any real split, which is a textbook problem of demand characteristics. Spanos recast hypnotic behavior as strategic, goal-directed role enactment by people who have absorbed cultural expectations about how a hypnotized person acts, not passive automatism.

Kirsch pressed the point further with response expectancy theory: hypnotic responses, like placebo responses, are produced by the subject's expectation that they will occur, which then becomes self-fulfilling. On this account hypnosis is a nondeceptive placebo, powerful precisely because the person believes. The apparent involuntariness ("my arm rose by itself") is a genuine experience but a misattribution, not evidence of an altered state. Critics also note that the state camp has never produced a physiological signature that reliably marks "being hypnotized" independent of the suggestions given, which is what a genuine distinct state ought to have. A review by Devin Terhune and colleagues concluded the neuroimaging literature shows changes tied to specific suggestions and to trait suggestibility, but no consistent marker of a unitary hypnotic state.

Where it stands now

The war between state and nonstate camps has cooled into an uneasy integration. Almost everyone now accepts the sociocognitive point that expectation, motivation, and interpretation are central, and that no single physiological trance state has been demonstrated. Almost everyone also accepts that trait hypnotic suggestibility is real, stable, and increasingly linked to individual differences in brain connectivity and executive control, which is closer to a dispositional than a purely social account. Contemporary researchers such as Amir Raz and Terhune tend to treat hypnosis less as a mystery to be explained and more as a tool: a reliable way to manipulate perception, attention, and the sense of agency in the laboratory, useful for studying volition and belief regardless of which theory of the underlying mechanism prevails. Clinically, the American Psychological Association recognizes hypnosis as a legitimate adjunctive technique, strongest for pain, with the firm caveat, following the false-memory research of Elizabeth Loftus and others, that it must never be used to "recover" memories, because it reliably inflates confidence in memories that are false.

Test yourself

Picture yourself as the vivid stage-hypnosis subject clucking like a chicken, and ask which explanation you find more natural: that your critical mind was switched off, or that, wanting to be a good subject and half-believing it would work, you let yourself go and then experienced the result as something happening to you. Notice that the second story, the one that keeps you in charge the whole time, is the one most of the evidence points to, and that it is somehow the harder one to believe.