Suggestion

Abstract

Suggestion, which MeSH classifies under hypnosis, is the process by which an idea communicated to a person produces a corresponding perception, belief, or action that the person does not experience as the product of their own deliberate reasoning. It operates across clinical, forensic, and everyday settings, and its effects can be indexed with standardized scales of suggestibility. Contemporary accounts frame suggestion less as a special altered state than as the operation of expectancy, imagination, and social demand on ordinary cognition. This article defines suggestion and its subtypes, surveys how interrogative and imaginative suggestibility are measured, contrasts the leading theoretical accounts, examines the suggestibility of memory, reviews the emerging cognitive neuroscience, and considers the methodological controversy that now surrounds the field.

Keywords: suggestion, suggestibility, response expectancy, misinformation effect, demand characteristics

Suggestion is one of psychology's oldest experimental problems and one of its most awkward. A suggested response — a raised arm that feels weightless, a symptom that lifts after an inert pill, a memory of an event that never happened — is produced by communication, yet it is experienced as involuntary, as something that happens to the person rather than something they do. Clark Hull's Hypnosis and Suggestibility (Hull, 1933) established that this phenomenon could be measured under laboratory control, and every subsequent account has had to explain the same puzzle: how an idea becomes an experience without passing through the sense of agency that normally accompanies deliberate thought. The modern answer, developed most fully in Irving Kirsch's response-expectancy theory (Kirsch, 1985), locates suggestion in the ordinary machinery of expectation rather than in a distinct hypnotic faculty.

What Suggestion Is

A suggestion is a communication — a word, a question, an implication, a gesture — that produces a matching response the recipient does not attribute to their own volition. The defining feature is the classic suggestion effect: the experience of the response as happening by itself. When a hypnotist suggests that an outstretched arm is growing heavy and the arm sinks, the person typically reports that it fell on its own, not that they lowered it. This felt involuntariness distinguishes suggestion from ordinary compliance, where a person knowingly does what is asked.

Suggestibility is the corresponding trait: the disposition to respond to suggestions, measured as a stable individual difference. It is not a single quantity. A person highly responsive to hypnotic suggestions of perceptual change need not be the person most easily misled by a leading question in an interview, because the two tap different processes. Suggestion also does not require hypnosis. Placebo responses, the persuasive force of advertising, the contagious spread of symptoms in mass psychogenic illness, and the reshaping of memory by post-event information are all suggestion effects that occur in fully ordinary waking states (Kirsch, 1985).

Types of Suggestion

In the MeSH classification, suggestion is filed beneath the broader descriptor hypnosis and carries one narrower descriptor of its own. MeSH is an indexing vocabulary — a scheme for cataloguing the literature — rather than a theory of how the mind is organized, so the placement below records how papers are indexed, not a claim that suggestion is a subtype of the hypnotic state. The subtype below is also not exhaustive of the ways suggestion is classified in research: psychologists routinely separate imaginative from interrogative suggestibility, and direct from indirect suggestion, cross-cutting the single MeSH child.

Measuring Suggestibility

Because suggestion is defined by response, it is measured by giving standardized suggestions and scoring how many take effect. Two measurement traditions grew up largely apart, and their near-independence is itself a substantive finding.

The first tradition scores imaginative suggestibility: a series of test suggestions — an arm rising, a hand locking shut, a fly buzzing that is not there, an inability to smell ammonia — is delivered and the number the person passes is counted. Scales in this family were long treated as measures of hypnotizability, on the assumption that responding to such suggestions required a hypnotic state. Wayne Braffman and Irving Kirsch broke that assumption apart empirically. Testing suggestibility with and without a hypnotic induction, they found that the induction added only a modest increment: most of a person's hypnotic responsiveness was already present as nonhypnotic imaginative suggestibility, and the two correlated highly (Braffman & Kirsch, 1999). Hypnotizability, on this evidence, is largely imaginative suggestibility with a small hypnotic bonus, not a separate capacity.

The second tradition, developed by Gisli Gudjonsson for forensic use, scores interrogative suggestibility: the tendency to accept misleading information embedded in questions and to change answers under interpersonal pressure. The Gudjonsson Suggestibility Scale reads a witness a story, then asks a set of questions of which several are subtly leading. It yields two scores. Yield counts how many leading questions the person accepts on first asking. The examiner then delivers negative feedback — a firm instruction that errors were made and the questions will be repeated — and re-asks everything. Shift counts how many answers change under that pressure. Yield and Shift sum to a Total Suggestibility score, and the two components are partly separable: Yield tracks the uncritical acceptance of a false premise, while Shift tracks the effect of social pressure (Gudjonsson, 1984). The scale is used to assess whether a defendant is unusually likely to have produced an unreliable statement under interrogation.

Theories of Suggestion

Two broad theoretical families dominate. The first is Kirsch's response-expectancy account. A response expectancy is an anticipation of one's own automatic reaction — expecting to feel relaxed, to feel pain lift, to be unable to bend an arm. Kirsch's claim is that such expectancies are directly self-confirming: expecting a nonvolitional response tends to produce it, through the same processes that generate placebo effects (Kirsch, 1985). On this view a suggestion works precisely when it installs a confident expectation, and the felt involuntariness is genuine, because the expectancy engages automatic rather than deliberate systems.

The second family is the sociocognitive account associated with Nicholas Spanos and Theodore Barber. It treats suggested behavior — including hypnotic amnesia, analgesia, and the odd literalism of trance logic — as strategic, goal-directed action shaped by the participant's interpretation of the situation, their attitudes and expectations, and their reading of what the experimenter wants, rather than as the output of an altered state (Spanos, 1986). The two families are not wholly opposed: both reject a special trance as the explanation, and both give expectancy and interpretation the central role. They differ over how much of the response is automatic self-confirmation and how much is enacted performance. Neither treats suggestibility as a mysterious weakness of will; both root it in normal cognitive and social processes.

The Suggestibility of Memory

Suggestion's most consequential effect is on memory. In Elizabeth Loftus and John Palmer's foundational demonstration, participants watched a film of a car accident and were then asked how fast the cars were going when they contacted, hit, bumped, collided, or smashed into each other. The single changed verb shifted the average speed estimate systematically upward, and a week later participants who had heard smashed were more than twice as likely to report — falsely — having seen broken glass in the film (Loftus & Palmer, 1974). The leading question did not merely bias a guess; it altered the reported memory itself. This misinformation effect is now one of the most replicated findings in cognitive psychology and the empirical foundation for reforms to eyewitness interviewing.

Interrogative and mnemonic suggestibility meet here. A leading question can install false information (a Yield-type effect), and repeated or pressured questioning can push a witness to revise an accurate memory (a Shift-type effect). The practical lesson drawn across forensic psychology is that the form of a question is part of the evidence it produces, and that a confident, detailed memory report is not, by itself, proof that the remembered event occurred as described.

The Cognitive Neuroscience of Suggestion

Neuroimaging has made suggestion a tool as well as an object of study. Because a verbal suggestion can reliably produce or remove a specific experience — color where there is none, pain where there is a noxious stimulus, blindness with open eyes — suggestion offers a controlled way to manipulate conscious contents and study their neural correlates. David Oakley and Peter Halligan argued that suggested states can serve as instruments for cognitive neuroscience, creating clinically realistic analogues of conditions such as functional paralysis or hallucination under experimental control (Oakley & Halligan, 2013). The recurring finding is that highly suggestible individuals, responding to suggestion, show brain-activity patterns resembling those of the corresponding genuine percept, and altered coupling between prefrontal control regions and the networks that generate the experience.

Devin Terhune and colleagues synthesized this work into a framework in which suggestion operates through top-down regulation: a suggestion recruits high-level expectation and attention to modulate lower-level perceptual and motor processing, so that the suggested content is generated and yet not tagged as self-produced (Terhune et al., 2017). This account connects suggestion to mainstream theories of predictive processing and cognitive control, and treats high suggestibility as an unusual efficiency in letting expectations reshape experience.

Clinical Uses

Suggestion is the active ingredient in several therapies and a component of many others. Suggestive procedures are used for pain control, for habit change, and as an adjunct to psychotherapy. A meta-analysis by Kirsch and colleagues pooled studies that added hypnotic suggestion to cognitive-behavioral treatment and found that the augmented treatment outperformed the same treatment without it, with the average client receiving hypnosis improving more than about 70% of comparable clients who did not — a moderate but reliable benefit, most consistent for outcomes such as weight loss (Kirsch et al., 1995). The placebo effect is the same mechanism operating without deception's opposite: an inert treatment relieves symptoms to the degree that it generates a confident expectancy of relief, exactly as response-expectancy theory predicts (Kirsch, 1985).

Worked Example

Consider a witness assessed on an interrogative-suggestibility scale modeled on the Gudjonsson procedure. The examiner reads a short narrative, waits, and then asks 15 questions, of which several contain false premises. On the first pass the witness accepts 6 of the leading premises: the Yield 1 score is 6.

The examiner then states firmly that the witness has made errors and that all the questions will be asked again, and re-administers them. This time the witness changes 5 of their earlier answers under the pressure: the Shift score is 5.

Total Suggestibility is the sum of the two components:

Yield 1 + Shift = 6 + 5 = 11.

Interpretation depends on the reference norms, but the two components carry different meaning. A high Yield with a low Shift describes a person who uncritically absorbs false premises but holds their ground when challenged; a low Yield with a high Shift describes a person who initially resists but capitulates under interpersonal pressure. Two witnesses can reach the same total by opposite routes, which is why the components are scored and interpreted separately rather than collapsed into one number.

Discussion

The trajectory of suggestion research is a steady demystification. What began as the study of a special hypnotic power has become the study of how expectancy, imagination, attention, and social context shape ordinary cognition. The convergence is striking: response-expectancy and sociocognitive theories, developed as rivals, now agree that no altered state is needed and that suggestion draws on processes present in everyone. The measurement work reinforces this, showing that hypnotic responding is mostly nonhypnotic imaginative suggestibility (Braffman & Kirsch, 1999), and the neuroscience situates suggestion within general models of top-down control (Terhune et al., 2017).

Yet the same feature that makes suggestion powerful makes it methodologically treacherous. A suggested response is certified only by the participant's report of an inner change, and reports are sensitive to what participants believe is expected of them. Distinguishing a genuinely altered experience from a compliant or reinterpreted report has been the field's central difficulty since its beginning, and it is unresolved. The suggestibility of memory sharpens the stakes: because leading questions can manufacture confident false memories (Loftus & Palmer, 1974), the same mechanisms studied in the laboratory operate in interrogation rooms and clinics, where a suggested memory can have irreversible consequences.

Current Directions

The liveliest current debate concerns whether some celebrated suggestion effects are, in part, artifacts of the very demand the field has always worried about. Peter Lush, Zoltan Dienes, and colleagues introduced the construct of phenomenological control — the trait capacity to generate experiences to meet the demands of a situation — and showed that it is essentially the same disposition as imaginative suggestibility. Crucially, individual differences in phenomenological control predict the strength of effects widely treated as automatic and involuntary, including the rubber-hand illusion and mirror-touch responses (Lush et al., 2020). If a participant's susceptibility to a bodily illusion tracks their suggestibility, then part of what such paradigms measure may be suggestion rather than the low-level multisensory process the paradigm was designed to isolate.

This has reopened a sixty-year-old question about demand characteristics — the cues in a study that tell participants what response is expected of them, a concept Martin Orne introduced in 1962 to argue that an experiment is a social occasion in which participants actively try to discern and confirm the hypothesis (Orne, 1962). Olivier Corneille and Lush have argued that the field lacks an adequate conceptual framework for the subjective experiences that demand characteristics elicit, and that many studies do not distinguish a genuine effect from a suggested one because they never measure participants' expectations about what the study is for (Corneille & Lush, 2023). The constructive proposal is methodological: routinely measure expectancies and phenomenological control, use them as covariates, and design paradigms that can separate a suggested response from the process under study. The controversy is unsettled, and it places suggestion — long a niche topic — at the center of a broad reckoning with how subjective effects are measured across psychology.

Common Misconceptions

Only weak-minded or gullible people are suggestible.

Suggestibility is not a deficiency of intelligence or willpower. Imaginative suggestibility correlates with the capacity for absorbed, vivid imagination, not with low intelligence, and it is distributed across the population as a normal individual difference (Braffman & Kirsch, 1999). The trait reflects how readily expectation reshapes experience, which is a capacity, not a weakness.

A suggested response is faked or consciously chosen.

The felt involuntariness is real, not pretense. Response-expectancy theory holds that a confident expectation engages automatic processes and genuinely produces the response, which is why placebo analgesia relieves real pain rather than merely being reported (Kirsch, 1985). Even sociocognitive accounts, which stress goal-directed enactment, do not treat the response as deliberate faking (Spanos, 1986).

Suggestion requires hypnosis.

Most suggestion occurs in ordinary waking states. A hypnotic induction adds only a small increment to the suggestibility a person already shows without it (Braffman & Kirsch, 1999), and placebo, advertising, and eyewitness misinformation are all suggestion effects with no hypnosis involved.

Glossary

Autosuggestion.

Suggestion that a person directs at themselves, as in the silent formulae of autogenic training, rather than receiving it from another person.

Classic suggestion effect.

The experience of a suggested response as occurring involuntarily, by itself, rather than as a deliberately produced action.

Compliance.

Knowingly doing what is requested while privately unconvinced; distinguished from suggestion by the absence of any felt change in experience.

Demand characteristics.

Cues in a study that convey what is expected, which can lead participants to produce the anticipated response independent of the process under investigation.

Hypnosis.

A procedure in which suggestions are administered, often after an induction; its effects are now understood as largely continuous with nonhypnotic suggestibility.

Imaginative suggestibility.

Responsiveness to suggestions of perceptual, motor, or cognitive change, scored by the number of such suggestions passed; largely the same trait as hypnotizability.

Interrogative suggestibility.

The tendency to accept misleading information in questions and to alter answers under interpersonal pressure, measured by the Gudjonsson Suggestibility Scale.

Misinformation effect.

The distortion of a witness's memory by information encountered after the event, typically introduced through leading questions or post-event narratives.

Phenomenological control.

The trait capacity to generate experiences that fit the demands of a situation; empirically close to imaginative suggestibility.

Placebo effect.

Symptom relief produced by an inert treatment through the expectancy of relief it generates; a suggestion effect requiring no hypnosis.

Response expectancy.

An anticipation of one's own automatic reaction which tends to be self-confirming, proposed as the core mechanism of suggestion.

Shift.

On the Gudjonsson scale, the number of answers a person changes after receiving negative feedback; an index of susceptibility to interpersonal pressure.

Sociocognitive theory.

The view that suggested behavior is strategic, goal-directed action shaped by attitudes, expectations, and the participant's reading of the situation, rather than the product of a trance.

Suggestibility.

The stable individual-difference trait describing how readily a person responds to suggestion; not a single quantity but a family of partly independent capacities.

Suggestion.

A communication that produces a matching perception, belief, or action which the recipient does not experience as the product of their own deliberate reasoning.

Yield.

On the Gudjonsson scale, the number of leading questions a person accepts on first asking; an index of uncritical acceptance of a false premise.

Key Researchers

Gisli H. Gudjonsson (b. 1947). Forensic psychologist at King's College London who created the Gudjonsson Suggestibility Scale and established interrogative suggestibility as a measurable trait central to the assessment of unreliable confessions. ORCID: 0000-0002-8817-0797. Faculty · Wikipedia

Clark L. Hull (1884-1952). Yale experimentalist whose Hypnosis and Suggestibility (1933) brought suggestion under quantitative laboratory control and set the agenda for its measurement. Wikipedia · Wikidata

Irving Kirsch (b. 1943). Associate Director of the Program in Placebo Studies at Harvard Medical School and author of response-expectancy theory, which reframed suggestion, hypnosis, and placebo as expressions of self-confirming expectation. No public ORCID was found on an individual search. Scholar · Wikipedia

Elizabeth F. Loftus (b. 1944). Distinguished Professor at the University of California, Irvine, whose work on the misinformation effect demonstrated that leading questions can reshape eyewitness memory itself. ORCID: 0000-0002-2230-6110. Faculty · Wikipedia

Nicholas P. Spanos (1942-1994). Carleton University psychologist and leading proponent of the sociocognitive theory of hypnosis and suggestion, arguing that suggested behavior is strategic action rather than the product of a trance state. Wikipedia · Wikidata

Devin B. Terhune (contemporary). Cognitive neuroscientist at King's College London studying individual differences in suggestibility and the top-down regulation of consciousness that underlies suggested experience. ORCID: 0000-0002-6792-4975. Faculty · Wikipedia

Frequently Asked Questions

What is suggestion in psychology? Suggestion is the process by which a communicated idea produces a matching perception, belief, or action that the recipient experiences as involuntary rather than as the product of their own deliberate reasoning. It spans hypnotic responses, placebo effects, persuasion, and the reshaping of memory by leading questions.

What is the difference between suggestion and compliance? In compliance a person knowingly does what is asked while privately unconvinced; there is no change in their experience. In suggestion the response feels involuntary, as something that happens by itself. This felt involuntariness, the classic suggestion effect, is the defining marker that separates a suggested response from mere going-along.

Are imaginative and interrogative suggestibility the same thing? No. They are largely independent. Imaginative suggestibility is responsiveness to suggestions of perceptual and motor change and is closely tied to hypnotic responding. Interrogative suggestibility is the tendency to accept misleading questions and to change answers under pressure, measured for forensic purposes by the Gudjonsson Suggestibility Scale. A person high on one need not be high on the other.

Does suggestion require hypnosis? No. Most suggestion occurs in ordinary waking states. Braffman and Kirsch found that a hypnotic induction adds only a small increment to the suggestibility a person already shows without it, and placebo responses, advertising, and eyewitness misinformation are all suggestion effects with no hypnosis involved.

How is suggestibility measured? Imaginative suggestibility is scored by counting how many standardized test suggestions a person passes. Interrogative suggestibility is measured with the Gudjonsson Suggestibility Scale, which yields a Yield score (leading questions accepted on first asking) and a Shift score (answers changed after negative feedback), summed to a Total Suggestibility score.

Can suggestion create false memories? Yes. In Loftus and Palmer's classic study, changing a single verb in a question about a filmed accident raised speed estimates and later made witnesses more than twice as likely to report broken glass that was never there. This misinformation effect shows that leading questions can distort the reported memory itself, not merely bias a guess.

Is a suggested response faked? No. The felt involuntariness is genuine. Response-expectancy theory holds that a confident expectation engages automatic processes and actually produces the response, which is why placebo analgesia relieves real pain. Even sociocognitive theories, which emphasize goal-directed enactment, do not treat the response as deliberate pretense.

Why is suggestion controversial in current research? Because a suggested response is certified only by self-report, effects thought to be automatic may be partly suggested. Recent work shows that susceptibility to bodily illusions such as the rubber-hand illusion tracks a person's phenomenological control, a trait close to suggestibility, reopening a long-standing debate about demand characteristics and how subjective effects should be measured across psychology.

References