Deepening is what you do after the induction: someone is already in trance, and you help them go further into it. In practice a deepener is a suggestion. You tie the idea of “going deeper” to something the person is already doing or noticing, and their response carries them there.
That raises two questions. How much depth does the work in front of you actually need? And how do you know how deep the person already is? Those turn out to matter more than knowing more scripts. So this is a piece about reading depth and matching it to the goal, with the techniques sitting inside that, where they belong.
What “depth” actually is
Depth isn’t a single dial you turn up. It’s a bundle of separate responses: relaxation, focus, amnesia, anesthesia, the ability to hallucinate. They don’t all move together. One person can hallucinate vividly while barely relaxed. Another can be deeply relaxed and only dimly aware of the room, yet still can’t let go of the numbers when you suggest amnesia. So “how deep is she?” is really shorthand for “how much of what is available right now?” There aren’t tidy floors you pass through one at a time, and deepening isn’t a one-time act that locks a level in. It’s ongoing.
The classic labels (hypnoidal, light, medium, somnambulism) are useful shorthand for roughly how much is on offer, but don’t take them as rungs on a ladder everyone climbs the same way. And you can’t read any of it directly. There’s no gauge. Anesthesiologists have tried the reverse experiment, wiring hypnotized people to the depth-of-anesthesia monitors used in surgery, and found no correlation between the monitor values and the person’s rated hypnotic depth. There’s no established instrument that reliably reads trance depth off the body. What you have instead is behavior, and your read of it.
So everything below about “how deep” is inference. Skilled inference, the kind you get better at, but inference. Keep that in your back pocket. It changes how hard you lean on any single sign.
Somnambulism, and where the idea came from
The deep working state most people mean has a name: somnambulism. It’s the responsive, sleep-like-yet-lucid state where the interesting phenomena become available: amnesia, anesthesia, positive and negative hallucination, and vivid regression work. When people say they need a client “deep,” somnambulism is usually what they mean.
The concept is older than the word “hypnosis,” and it arrived by accident. In 1784 the Marquis de Puységur, a student of Franz Mesmer, was working on a young man named Victor Race on his estate at Buzancy. Mesmer’s method aimed at a dramatic convulsive “crisis.” Victor did the opposite. He fell into a calm sleep-like state in which he could still talk and respond, and seemed, if anything, sharper than in his ordinary waking life. Puységur, struck by how much it resembled natural sleepwalking, called it artificial somnambulism.
Two details from that first case are worth carrying forward. Puységur was so overrun with people wanting the treatment that he magnetized a large elm tree in the village, ran cords from it, and had people tie themselves on to receive the “fluid” in groups. It’s a good reminder of how much of the early theory was scaffolding that turned out not to be load-bearing. The second detail is the one that matters for practice: when Victor came out of it, he remembered nothing. Amnesia on waking was noticed as a hallmark of the deep state from the very first description of it. (If you want the careful scholarly version of this history rather than my thumbnail, Henri Ellenberger’s The Discovery of the Unconscious is the standard source.)
Somnambulism isn’t the floor, at least not in the traditions that think in terms of depth. Some name states beyond it. The best known is the Esdaile state, or “coma state,” after James Esdaile, a 19th-century surgeon who operated on patients in deep mesmeric trance in 1840s India, before chemical anesthesia existed. Dave Elman revived the idea and described a state past somnambulism: profound lethargy, an anesthesia that appears without being specifically suggested, waxy catalepsy, and a comfort so complete the person often doesn’t want to leave. A few methods name deeper levels still, mostly proprietary systems like Sichort’s “Ultra-Depth” work, with claims attached that outrun anything I’d state as fact. Two honesty notes. “Coma” here is a legacy label, not a medical coma: the person is aware, comfortable, and can emerge whenever they choose. And the whole “deeper than somnambulism” ladder is a craft-and-clinic idea. The validated research scales stop at somnambulism, or drop the depth ladder entirely for measured responsiveness. So take these as states the depth traditions describe, not as settled categories.
Hold on to that link between somnambulism and amnesia. It runs straight through to a test still taught today, and to a useful argument about how much any of these markers really tell you.
How much depth does the goal actually need?
Deeper is not always better. Deeper is deeper. The right question is what the work in front of you requires.
As a rough guide, direct suggestion, motivation work, habit change, and a fair amount of comfort and pain-softening work tend to land without deep trance. The phenomena that genuinely seem to need the deep end are the dramatic ones: near-total amnesia, surgical-grade anesthesia, and positive hallucination (perceiving something that isn’t there).
Treat that mapping as a guide, not a law, because the phenomena don’t line up cleanly against any single scale. What one person can do in an apparently light trance, another can’t manage while looking far more deeply absorbed. The variation tracks the individual more than the moment.
The measurement research shows hypnotic responsiveness behaves like a stable personal trait. The Elkins Hypnotizability Scale correlates about .91 with the older Stanford scale and is highly stable on retest. That’s a different thing from the shifting in-session depth you’re trying to read. And it’s different again from helping a client reach their full potential of hypnotic responsiveness. Some of your clients arrive able to do things others may never manage, at any apparent depth. Plan around the person in front of you, not around a level.
How to tell someone is going deeper
You read depth the way you read whether two people at the next table are in love, or whether someone is enjoying their meal. Usually you can tell. Not with certainty, and not from one signal, but from a cluster of them, read against what that particular person looked like a few minutes ago. It’s pattern recognition on a baseline, and it’s a skill that sharpens with reps.
The signs worth watching, as a subject settles deeper, tend to include:
- The face goes slack. Muscle tone drops, the jaw loosens, features flatten out.
- Breathing slows and moves lower, toward the belly.
- Swallowing, blinking, and small spontaneous movements slow down or stop.
- Responses come on a delay. Ask for a finger to lift and it takes a beat, sometimes with a slightly “literal” quality to how instructions get taken.
- Catalepsy: a limb stays where you place it, heavy and comfortable.
- Color changes, watering eyes, a general look of having “gone somewhere.”
Read those together, and against the person’s own starting point, not against some textbook ideal. A naturally still person deepening looks different from a fidgety one.
Here’s why you hold all of this loosely. I was once running arm levitation with a client, suggesting his arm would float up, on and on for what felt like several minutes. Nothing. No movement I could see, so I gave up on it and moved to something else. After he emerged, he asked me why I’d kept at the levitation so long, because his arm had been floating up high the whole time. To him it had lifted. To me it hadn’t moved an inch.
That’s very likely a deep trance. His internal experience had come loose from what his body was doing, and I misread it as no response, because the sign I was watching for didn’t show up. The marker can be absent while the depth is right there. And the reverse happens too: a sign you’re counting on can show up in someone who isn’t especially deep. Signs inform your guess. They don’t settle it.
The amnesia test, and what it’s really worth
The most famous depth marker is the one from Victor Race: amnesia. Dave Elman built a working test around it, and it’s still widely taught. The subject counts backward from a hundred, with the suggestion that each number relaxes them twice as much and that the numbers will fade and be gone. Most people stop somewhere in the high nineties. Then you check, carefully, without saying the word “numbers” (naming them can hand them back), something like “and are they gone now?” A yes, in Elman’s framing, says the person has reached somnambulism. You can see how this fits into the Elman induction as its final deepening step.
As a practical signpost it’s reasonable, as long as you’re clear on what it tests. The number test asks for a small, specific amnesia: just the numbers. In my experience most people can do that. That’s worth sitting with, because it means a “yes” doesn’t single out a rare, deep state; mostly it tells you the person will take a suggestion, which is genuinely useful. And if a client can’t let the numbers go, that isn’t a dead end either: it says they haven’t taken that one suggestion, not that they’re beyond help, and plenty of good work happens without anyone reaching that mark.
It helps to separate two things that both wear the name amnesia. One is suggested amnesia: you asked for it, whether that’s the numbers or the whole session. The other is spontaneous amnesia: the person emerges and finds part of the trance simply isn’t there, though no one suggested it would go.
Suggested amnesia is, as you’d expect, a product of the suggestion. The research on it is fairly settled: it’s produced by the suggestion rather than welling up on its own, the fuller version of it tracks hypnotizability so it shows up mainly in highly responsive people and only when it’s actually suggested, and it’s reversible, the memories returning on a cue without any need to re-induce hypnosis. As a facet of the hypnotizability trait, it tells you more about the person than about how deep a given trance went.
Spontaneous amnesia is the more interesting case, and harder to pin down. It may have less to do with depth than with a change of context. Think of how an acquaintance’s name comes easily at the party where you met them, then won’t surface when you run into them at the grocery store. Or how a lecture you followed fine is mostly gone by the time you’re home. A deep trance is a very different context from ordinary waking life, and sometimes the memories just don’t cross the boundary. Most hypnotherapists read that as a good sign rather than a problem. The aim was never for the client to remember to exercise. It was for them to find themselves exercising without thinking about it.
So the number test measures “is this person responding to my suggestions right now” rather than “how deep is the trance.” Amnesia is Elman’s proof of responsiveness, not a validated measurement of depth. And whether hypnosis is even a distinct state with identifiable markers, and so whether a single “depth” exists to measure at all, is itself disputed in the research. None of that means the test is useless. It means you should read a yes as “good, they’re responding,” not as a depth gauge reading you can bank on.
The core deepeners, and when each earns its place
Every deepener runs on the same engine: you tie “deeper” to a cue the person is already giving you, and let their own response do the work. That’s the principle behind the Vogt deepener and behind most of the rest. Here are the ones worth having ready, and what each is actually good for.
Fractionation. Bring the person up part way, or all the way out, then back into trance, repeatedly. Each return tends to go faster and further, and you’re also building an association: my voice, that chair, means slipping deeper easily. It’s the workhorse, and it’s especially handy with someone who deepens slowly, because the second and third pass usually beat the first.
Countdown and staircase. Count down and attach going deeper to each number, or use imagery of a staircase, escalator, or lift descending. These are the same tool in different clothes: a steady, predictable cue with “deeper” married to every step. Clean and reliable for most people, and easy to combine with anything else. I’ve written up a few simple deepeners of this kind separately if you want scripts.
Compounding. Stack suggestions so each one builds on the last. Every breath takes you deeper, and each word I say takes you deeper still. You’re layering cues so the responses accumulate. Good glue between other techniques.
Imagery. A beach, a staircase, a quiet familiar place. The value isn’t the scenery, it’s the absorption. The more fully someone is inside an imagined scene, the less attention is left over for monitoring the room. Worth tailoring to the person, since a client who hates the beach won’t sink into one.
Direct suggestion. Sometimes you simply suggest it. “Deeper now, further with every breath.” With a responsive subject who’s already moving, plain direct suggestion is often all the deepening you need. Don’t over-engineer what’s already working.
Phenomena as deepeners. An arm that levitates, an eye-catalepsy that holds: these can double as deepeners. The person experiences their own body responding to suggestion, that experience is convincing, and conviction feeds responsiveness. More on that in a moment, because it’s really about convincing, and it comes with a catch I learned the hard way.
Silence. An underrated one. Once someone is settling deeper, quiet lets them keep going on their own. You don’t have to fill every second with patter. Suggest deepening, then get out of the way and let them do it.
Convincers: verify and deepen in one move
A convincer is anything that shows the subject, unmistakably, that something real is happening: an arm too heavy to lift, hands that won’t come apart, a number that’s genuinely gone. Its first job is to convince them, and that’s the point: once a person is persuaded they’re responding, they respond more. Doubt is the thing that keeps people shallow, and a good convincer dissolves it.
That’s why the same moves show up on both lists. A convincer verifies responsiveness and deepens it in a single step. A suggestibility test run before the induction does a related job at the other end, telling you what you’re working with before you start. Just remember which way the information runs. A convincer that “fails” isn’t proof the person is shallow. It might mean you reached for the wrong one, or reached too early, or that their response looked nothing like what you were watching for, as my levitating client would happily tell you.
Sometimes it’s framing, not depth
A lot of what gets filed under “deep trance phenomena” has less to do with how deep the person went and more to do with how you framed the thing you asked for.
When we do timeline work in NLP, we’ll often just ask the person to imagine their life stretched out on the floor, from birth over here to the present over there, and to step onto it. No formal induction. With decent rapport, most people simply do it — they walk their timeline, make changes, and get real results. What made that possible wasn’t ten minutes of deepening. It was the frame. The request was concrete and permission-giving, rapport carried it, and the person’s own critical factor, the part that would ordinarily stop and say “this is silly,” stepped aside.
That’s a critical factor bypass, and it’s closer to how a child plays than to anything mysterious. A kid told the couch is a boat doesn’t need convincing or deepening. They’re already absorbed; they just go. A cooperative adult, framed well and in rapport, does the same thing. The absorption you’re after is often available for the asking, if you ask in a way that gives permission and doesn’t invite the analytical part to dig in.
So before you reach for more depth to produce some phenomenon, ask whether depth is even the lever. Often the faster route is a better frame, more rapport, and getting out of your subject’s way.
When someone won’t deepen
Sometimes the person seems stuck shallow. Before you pile on more deepeners, run through a short list, because the fix is usually diagnostic, not more force.
Check rapport first. Thin rapport is the most common brake. People don’t hand over responsiveness to someone they haven’t settled with. If depth isn’t coming, the problem is often upstream of technique.
Ask whether they’re testing instead of deepening. Analytical clients, the “am I doing this right?” monitors, stay busy checking. A convincer can help here by giving the analytical part something undeniable to chew on, and fractionation helps by not leaving time to analyze. Sometimes simply telling them there’s nothing to get right, and that thinking about it is fine and they can drift anyway, takes the pressure off. An Ericksonian “that’s right,” offered warmly whatever they happen to be doing, can do a surprising amount here: it tells them they’re succeeding without handing the analytical part anything to argue with.
Reconsider what you’re seeing. Remember there’s no gauge to read. “Won’t deepen” sometimes means “is deep in a way I’m not reading.” Test for a response (a small catalepsy, an ideomotor signal) rather than trusting your read of their face.
Change the channel. If imagery isn’t landing, try a physical deepener; if counting feels mechanical to them, try fractionation or straight suggestion. People vary in what absorbs them. Match the tool to the person.
Myths about depth worth dropping
“Deeper is always better.” Covered above, but it’s the big one. Match depth to the goal. Most work doesn’t need the deep end, so treating more depth as the goal aims you at the wrong target.
“If they’re aware, they’re not really hypnotized.” Plenty of deeply responsive people are fully aware and remember everything. Awareness isn’t the enemy of trance, and a client narrating their experience to you can still be doing remarkable things. Somnambulism can come with amnesia; it doesn’t require it.
“There’s a way to measure how deep they are.” There’s a self-rating the person can give you, and there are signs you learn to read, but there’s no objective instrument I know of. The anesthesia monitors don’t track it, and even the classic amnesia marker turns out to measure responsiveness more than depth. Your calibrated read of behavior is the best tool available. Trust it, and hold it loosely.
Frequently asked questions
How deep does hypnosis need to be to work?
For most change work (direct suggestion, motivation, habit change, comfort work), light to medium is plenty. The deep end matters mainly for dramatic phenomena like near-total amnesia, strong anesthesia, positive hallucination, and some regression. Match the depth to the job rather than treating more depth as the goal.
What are the signs someone is deeply hypnotized?
A slack face, slower and lower breathing, delayed responses, reduced swallowing and blinking, catalepsy, and a general “gone somewhere” look, read as a cluster and against that person’s own baseline. No single sign is decisive, and a subject can be deeply absorbed while showing few outward signs.
Can you make someone go too deep?
In ordinary practice, no lasting harm comes from depth itself; people can come up on their own and will if left alone. The practical issues are more mundane: someone drifting past your instructions, or so relaxed they doze rather than respond. Keep contact, keep testing for response, and you stay in a working range.
Is the number test a reliable way to know someone’s in somnambulism?
It’s a reasonable working signpost. A “yes, the numbers are gone” tells you the person is responding to suggestion. But it’s best read as proof of responsiveness rather than a precise depth reading. Suggested amnesia is produced by the suggestion and shows up mainly in already-responsive people, so treat a positive test as good news about responsiveness, not a guaranteed depth gauge.
Wrapping up
Deepening is a suggestion tied to a cue, and the techniques are variations on that one idea. The craft that separates a good operator isn’t knowing more deepeners. It’s reading the person well enough to know how deep they already are, honestly enough to admit you’re inferring, and clearly enough to know whether the goal in front of you needs depth at all, or just a better frame and more rapport.
If you want to build out the induction side of this, our induction training is where deepening naturally lives, and there’s a free “Secret Success with Hypnotic Inductions” guide on the homepage worth grabbing. And for taking your own trance further, there’s a piece on deeper self-hypnosis too.
Enjoy, Keith.
Sources
- Zech, N., Seemann, M. & Hansen, E. (2023). Hypnotic depth and its relation to depth-of-anesthesia monitors. PMC10798158. (No correlation found between anesthesia-monitor values and rated hypnotic depth.)
- Elkins, G., Fisher, W. & Johnson, A. (2012). The Elkins Hypnotizability Scale. PMC3373787. (Hypnotic responsiveness as a stable, measurable trait; r ? .91 with the Stanford scale.)
- Kihlstrom, J. F. Posthypnotic Amnesia. International Society of Hypnosis Encyclopedia. ishhypnosis.org. (Suggested amnesia is a product of suggestion, correlates with hypnotizability only when suggested, and is reversible without re-inducing hypnosis.)
- Facco, E. et al. (2021). Posthypnotic amnesia in hypnotizability assessment. PMC7850173. (Amnesia modeled as a facet of the hypnotizability trait; note on older depth scales.)
- Matthews, W. J., Conti, J. & Starr, L. (1999). Ericksonian hypnosis: a review of the empirical literature. Sleep and Hypnosis. (Little support for hypnosis as a state with identifiable markers.)
- “Mesmerism.” Encyclopedia.com / Gale. encyclopedia.com. (Puységur, Victor Race, the magnetized tree, and amnesia on waking.)
- Ellenberger, H. F. (1970). The Discovery of the Unconscious. Basic Books. archive.org. (Standard scholarly history of Puységur and artificial somnambulism.)
- Elman, D. (1964). Hypnotherapy. Westwood Publishing. (Primary source for the amnesia/number-test criterion for somnambulism, and for the modern “Esdaile / coma state” concept.)
- Esdaile, J. (1846). Mesmerism in India, and its Practical Application in Surgery and Medicine. (Primary historical account of surgery performed under mesmeric trance; origin of the “Esdaile state.”)
