Is Hypnosis Real? Yes

The short answer

Hypnosis is real. It is a measurable state of narrowed, absorbed attention with reduced critical filtering, produced reliably under laboratory conditions for well over a century and visible in brain imaging as changed activity in the anterior cingulate cortex and altered coupling between the brain's executive control network and its default mode network. What is not real is the folklore attached to it — mind control, sleep, obedience against your will, amnesia on command. The state exists. The superpowers do not.

Everything the skeptic gets right

I start here on purpose. If I don't concede this part first, nothing after it will be believed, and rightly so.

Nobody is being controlled. There is no transfer of will. Hypnosis runs on the cooperation of the person in the chair, which is why it collapses the moment you push against someone. A suggestion that violates what a person actually wants gets declined — sometimes politely, sometimes by the person simply opening their eyes and asking what time it is.

Stage hypnosis is a casting process before it is a performance. The stage hypnotist does not walk on and overpower forty people. He runs quick group suggestibility tests, quietly discards everyone who doesn't respond, and keeps the handful who are both highly responsive and visibly delighted to be up there. By the time the show starts, the selection has already done most of the work. What you're watching is real trance in genuinely willing, genuinely unusual people — plus a lot of showmanship.

Nothing happens because the hypnotist has power. Irving Kirsch spent a career making this uncomfortable point: much of what hypnosis produces can be traced to response expectancy. What you believe will happen to you shapes what happens to you. That sounds like a debunking, and skeptics quote it as one. I'd call it an accurate description of the mechanism. Expectation is not the flaw in the machine. Expectation is the machine.

"Under" does not mean unconscious. Most people come out of their first session slightly disappointed, saying they heard everything and could have stood up at any point. Correct on both counts. That's what the state is. If you were expecting a blackout, you'll conclude nothing happened — and you'll be wrong for entirely understandable reasons.

The man who gave it its name was the first debunker

In 1841 a Manchester surgeon named James Braid went to a demonstration of animal magnetism intending to catch the performer cheating. He didn't catch him. What he concluded instead was that there was no magnetic fluid, no invisible force streaming from the operator's hands, and nothing whatsoever happening between the two people that wasn't happening inside the nervous system of the subject. Sustained fixation of attention did it. Within two years he had published Neurypnology and left us the word we still use.

He also came to regret that word. "Hypnosis" points at sleep, and sleep is exactly what it isn't, so Braid later proposed calling it monoideism — the condition of being occupied by a single idea. That name never caught on, which is a shame, because it would have prevented about a hundred and eighty years of misunderstanding. The founder of the field was its first skeptic, and his debunking is still substantially correct.

What a hypnosis brain scan actually shows

Imaging work from David Spiegel's group at Stanford, along with a broader literature built on the same approach, keeps recovering a similar picture when highly hypnotisable people are scanned in hypnosis and compared with themselves out of it. Activity shifts in the anterior cingulate cortex, the region that flags conflict and says hold on, check this. The coupling changes between the executive control network and the default mode network — roughly, between the part that acts and the part that narrates you to yourself. The insula, which handles how strongly body signals get weighted, joins in, which is presumably why suggested changes in sensation land as hard as they do.

Now the part that rarely makes the headlines. None of this amounts to a hypnosis detector. There is no dedicated hypnosis centre in the brain, the differences are statistical and appear at group level, and they show up most clearly in the responsive minority rather than in everybody. You cannot hand a scan to a technician and have them tell you whether that person was in trance. What the imaging establishes is narrower and still decisive: something reproducible is going on that is not politeness and not acting. If subjects were merely playing along, the pattern would not keep reappearing across labs.

"Does hypnosis actually work" is two questions in one coat

The first question is whether the state exists. That one is settled. The second is whether the state does anything useful, and there the answer has to be specific, because "hypnosis works" as a blanket claim is precisely the sort of thing that keeps the skeptics in business.

The clinical evidence is not evenly spread. It is densest and most consistent in pain — from experimental pain in the lab through to procedural and chronic pain — and in irritable bowel syndrome, where gut-directed hypnotherapy has been studied seriously enough to enter treatment guidelines in some countries. That is a description of where the research record is strongest. It is not a recommendation, not a promise, and no substitute for a doctor. Everywhere else the picture is patchier, and anyone who tells you otherwise is selling something.

There's a second kind of evidence that I find quietly more persuasive than any outcome study. Peter Halligan and David Oakley have used hypnotic suggestion as a laboratory instrument — producing states like paralysis or altered perception in healthy volunteers so they can be studied under controlled conditions. You would not build a research method on a phenomenon that consisted of people being agreeable. The suggestion has to reliably do something, in people who aren't trying to help you get a result, or the method is worthless.

Not everyone gets there, and that is not a character flaw

Ernest Hilgard's work at Stanford gave the field its standardised susceptibility scales, and the result has held up for decades: hypnotisability behaves like a trait. It is fairly stable across years, it is roughly normally distributed, and it sorts people into a small group who respond dramatically, a small group who barely respond at all, and a large middle who respond moderately with a decent guide.

This single fact explains most of the argument about whether hypnosis is real or fake. The person who went under like a stone and the person who sat there thinking about their shopping are both reporting honestly. They are describing different experiences of the same procedure, and each assumes the other is lying. If you want a rough sense of where you fall before you form an opinion, take a suggestibility test — it will tell you more in ten minutes than a week of reading.

Hilgard is also responsible for the strangest observation in the field, the so-called hidden observer: subjects given a suggestion for analgesia who, when asked in a particular way, could report a part of them that had registered the pain all along. Interpretations of that finding are still contested, and I mention it precisely because it is contested. Anyone who presents the science of hypnosis as tidy hasn't read much of it.

The cheapest experiment you can run tonight

In 1833 Michel-Eugène Chevreul, a chemist with no patience for the occult, took apart the divining pendulum. Hold a small weight on a thread, think steadily about it swinging left to right, and it swings left to right. Chevreul demonstrated that the movement came from the holder's own imperceptible muscle activity: brace the hand properly, or arrange things so the person can't see the weight, and the effect dies.

Both halves of my answer live inside that one demonstration. Nothing supernatural is happening — and something is happening that the person genuinely is not doing on purpose. That gap between not magic and not voluntary is where this entire craft operates. It has a name, the ideomotor effect, and if you want to watch it work on your own hands, the pendulum is the classic way in.

If you actually want to test it

Read the mechanism before you judge the phenomenon: how hypnosis works covers the four ordinary ingredients that build the state, and what trance actually is covers what it feels like from inside, which is far less dramatic than anyone expects. Then look at a real induction and notice how mundane the components are. If you've ever lost forty minutes to a loop on a dancefloor — I've written about hypnotic techno for exactly this reason — you have already been in the state you're doubting. You just weren't calling it that.

Skepticism, done properly, is not a conclusion. It is a method, and this happens to be one of the few subjects where you own the laboratory. What two consenting adults can build on the state once they stop arguing about whether it exists is what my books are about.

— Mira

FAQ

Is hypnosis real or fake? Real, with a large fake layer on top. The state is reproducible in laboratories and shows consistent changes in brain activity and network connectivity. Stage mind control, forced obedience and commanded amnesia are the fake layer, and dropping them costs the real thing nothing.

Does hypnosis actually work? For producing the state, reliably — in most people to some degree, in a minority very strongly. For clinical outcomes, the evidence is strongest in pain and in irritable bowel syndrome and thinner elsewhere. Anyone claiming it works for everything is describing marketing, not research.

Can a brain scan prove someone is hypnotised? Not for an individual. Imaging shows reproducible group-level differences in regions like the anterior cingulate cortex and in the coupling between executive and default mode networks, which is strong evidence that the state is genuine. There is still no scan that certifies one particular person is in trance.

Why do some people say it did nothing for them? Because hypnotisability is a stable trait, spread out roughly like height. A minority barely respond, and there is nothing wrong with them. Expectation, trust and a patient guide move most people further along the scale than they assume, but not everyone lands in the same place.

All blog posts · Suggestibility test · Books