The Milton Model: Language Patterns From Milton Erickson, With Examples
The Milton Model is a set of language patterns, drawn from the work of the psychiatrist Milton H. Erickson, that use deliberately general wording so the listener fills in the meaning from their own experience. Where most instructions tell people what to think, Milton Model language leaves a space and lets them find their own content to put in it.
Richard Bandler and John Grinder analysed Erickson’s language in Patterns of the Hypnotic Techniques of Milton H. Erickson, M.D. (Volume I, 1975; Volume II, 1977, with Judith DeLozier), and the patterns were organised into what NLP calls the Milton Model from that work. They now sit alongside the Meta Model at the centre of NLP’s language patterns work. This note covers where the model came from, every core pattern with an original example, how the patterns combine, where they are useful in ordinary work, and the ethical and evidential limits you should keep in view.
What is the Milton Model?
Milton H. Erickson (1901–1980) was an American psychiatrist known for an unusually indirect, conversational style of clinical hypnosis. He rarely gave blunt commands. He told stories, asked odd questions, and spoke in a way that seemed vague but was carefully constructed. Bandler and Grinder analysed his language in Patterns of the Hypnotic Techniques of Milton H. Erickson, M.D. (Volume I, 1975; Volume II, 1977, with Judith DeLozier), and the patterns were organised into what NLP calls the Milton Model from that work.
The key observation was that Erickson’s language was the inverse of the Meta Model. The Meta Model hears deletion, distortion and generalisation and asks questions to recover the detail. Erickson used the same deletions, distortions and generalisations on purpose. A sentence such as “You can learn something important here” names no specific thing, so each listener supplies the most relevant thing for them. That is what is meant by artfully vague language: vague enough to fit everyone, structured enough to point somewhere.
Milton Model language patterns with examples
The first group mirrors the Meta Model categories. The second group comes from Erickson’s particular way of pacing, leading and embedding suggestions. Every example sentence here is original and written for a teaching or coaching context rather than formal hypnosis.
| Pattern | What it does | Example sentence |
|---|---|---|
| Mind read | Claims to know the listener’s thoughts, in a way that is likely to be true | “You may be wondering how much of this you already know.” |
| Lost performative | States a judgement without saying whose it is | “It’s good to take your time with something new.” |
| Cause–effect | Links something happening now to the response you want | “As you read this line, you can begin to notice what already makes sense.” |
| Complex equivalence | Treats one thing as meaning another | “Being curious about this means you’re already learning it.” |
| Presupposition | Assumes the point and asks about something else | “Which part of this will you find most useful first?” |
| Universal quantifier | Generalises to everyone or always | “Everyone has known more than they could put into words.” |
| Modal operators | Uses possibility or necessity words to soften or frame | “You can let your attention settle wherever it needs to.” |
| Nominalisation | Turns a process into an abstract noun the listener defines | “You have a lot of understanding and new learnings available.” |
| Unspecified verb | Describes action without saying how | “You can consider this in your own way.” |
| Tag question | Adds a short question that invites agreement | “You’ve noticed that already, haven’t you?” |
| Lack of referential index | Leaves the subject unnamed so the listener applies it | “People can learn things quickly when they stop forcing it.” |
| Comparative deletion | Compares without saying with what | “And it’s easier now.” |
| Pattern | What it does | Example sentence |
|---|---|---|
| Pacing current experience | Describes things the listener can verify, building agreement | “You’re reading these words, sitting where you are, aware of the sounds around you…” |
| Double bind | Offers a choice in which both options lead the same way | “Would you rather start with the hardest item on the list or the easiest?” |
| Conversational postulate | A yes/no question that invites action rather than an answer | “Can you picture how the finished project will look?” |
| Extended quote | Nests a message inside layers of reported speech | “My first manager used to quote her mentor, who told her, ‘The answer arrives once you stop chasing it.’” |
| Selectional restriction violation | Gives a quality to something that cannot have it | “Let your notes tell you what matters.” |
| Phonological ambiguity | Words that sound alike but differ in meaning | “You can hear / here, now, what’s right for you / write for you.” |
| Syntactic ambiguity | A grammatical structure that can be read two ways | “Coaching managers can be demanding.” |
| Scope ambiguity | Unclear how far a word applies | “The experienced coaches and trainees will recognise this.” |
| Punctuation ambiguity | Runs two sentences together through a shared word | “You can look at the screen your thoughts from distraction.” |
| Utilisation | Uses whatever happens in the moment as part of the message | (A phone rings.) “And that sound can remind you how easily attention comes back.” |
| Embedded command | Places an instruction inside a longer sentence | “I wonder whether you’ll take a slower breath before you answer.” |
| Analogue marking | Marks the embedded command with a change of tone, pause or gesture | Saying “take a slower breath” slightly lower and slower than the rest of the sentence |
Recognising these patterns at speed is part of Master Practitioner-level work. The NLP Master Practitioner knowledge quiz tests it, with an explanation for each answer.
Embedded commands and analogue marking
Embedded commands are the most quoted part of the Milton Model and the most overclaimed. The structure is straightforward: an instruction phrased as a direct command (“notice what’s changed”) placed inside a softer frame (“I don’t know when you’ll notice what’s changed”). Analogue marking — a slight shift in voice, a pause, a glance — sets the command apart from the rest of the sentence.
“You might find yourself getting curious about the second half of the report.”
“There’s no need to decide right now, of course.”
“Some people feel more settled once they’ve written the first paragraph.”
What these patterns do reliably is change the tone of a request from demand to invitation. Claims that embedded commands bypass conscious judgement or compel behaviour have not been demonstrated in controlled research, and you should be wary of anyone who sells them that way.
How the patterns work together: pace and lead
Individual patterns do little on their own. Erickson’s skill was in sequencing them. The basic structure is pace and lead: first describe what is already true for the listener (pacing), then link it to what you would like to become true (leading). Several verifiable statements build agreement; a cause–effect link carries that agreement towards something new.
“You’ve found a seat, you’ve got your coffee, and some of you are still thinking about what you left on your desk. (pacing)
And as you settle in, you may already be wondering which part of today will be most useful to you. (cause–effect, mind read, presupposition)
Because everyone here has handled a difficult conversation at some point, haven’t they? (universal quantifier, tag question)
So you can begin to notice what you already do well. (modal operator, embedded command)”
Nothing in that opening is false or hidden. It simply meets people where they are before asking them to move. Pacing only works if the descriptions are accurate, which is why careful observation — see sensory acuity — comes before any of the language.
Everyday uses of Milton Model language
- Presenting. Open with pacing statements the room can verify, and use presuppositions (“As you apply this next week…”) to move attention from whether to how.
- Teaching. Nominalisations and unspecified verbs (“you’ll find your own understanding of this”) let learners at different levels each take something relevant.
- Coaching. Double binds and conversational postulates offer direction without instruction: “Would you rather explore what’s working or what’s in the way first?” Pair this with the precise questions of the Meta Model once the person is talking.
- Calming a tense conversation. Pacing what the other person is feeling (“You’ve been waiting a long time, and you’re frustrated, and you want this sorted today”) before leading towards a next step lowers heat far more than “Please calm down”.
Ethics and limits: influence or manipulation?
Any language that shapes attention is a form of influence. The line between influence and manipulation is not in the pattern but in the purpose and the consent. A useful test: would you be comfortable if the listener knew exactly what you were doing and why? A presenter using pacing to help a room settle passes that test. A salesperson stacking double binds to rush a vulnerable buyer does not.
- Get consent for anything hypnotic. Formal trance work, especially around health, phobia, anxiety or trauma, belongs with a qualified clinician. Milton Model language is not a substitute for treatment.
- Keep the listener’s outcome central. Erickson’s indirectness served his patients’ goals. Borrowing the technique without that orientation misses the point.
- Be honest about the evidence. Erickson’s clinical influence is well documented, but NLP’s formalisation of his language has limited and contested scientific support. The patterns are a useful description of persuasive, permissive speech, not a proven mechanism for controlling behaviour.
If you want to practise these patterns with feedback in a coaching or leadership setting, see work with me.
Free instruments for this note
NLP Master Practitioner Knowledge Quiz
For those beyond the basics: belief change, time orientation, values hierarchies, modelling, strategies and integration patterns.
Open itQuestions people ask
What is the Milton Model in NLP?
The Milton Model is a set of language patterns drawn from the psychiatrist Milton H. Erickson. Richard Bandler and John Grinder analysed his language in Patterns of the Hypnotic Techniques of Milton H. Erickson, M.D. (Volume I, 1975; Volume II, 1977, with Judith DeLozier). It uses deliberately general, “artfully vague” wording — mind reads, presuppositions, nominalisations, embedded commands — so that listeners fill in meaning from their own experience rather than being told what to think.
What is the difference between the Milton Model and the Meta Model?
They are inverses. The Meta Model hears where language has deleted, distorted or generalised experience and asks precise questions to recover the detail. The Milton Model uses those same patterns deliberately to keep language open. Practitioners typically use the Meta Model to understand a problem and the Milton Model to help someone imagine a different response.
What are embedded commands?
An embedded command is an instruction placed inside a longer, softer sentence, such as “You might find yourself relaxing as we go.” It is often marked by a slight change in voice or a pause, called analogue marking. In practice it makes a request feel like an invitation. Claims that it bypasses conscious choice are not supported by controlled research.
Is the Milton Model manipulative?
It can be used manipulatively, as can any persuasive language. The patterns themselves are common in ordinary speech. What matters is purpose and consent: whether you are serving the listener’s outcome and would be comfortable with them knowing exactly what you were doing. Hypnotic or therapeutic use belongs with qualified, consenting practice.
Does the Milton Model actually work?
Milton Erickson’s influence on clinical hypnosis and brief therapy is well established. The NLP formalisation of his language has little controlled research behind it, and NLP’s evidence base overall is limited and contested. The patterns are best treated as an accurate description of permissive, persuasive speech that you can test in your own presenting, teaching and coaching.
Related
One useful observation,
no motivational sludge.
One useful observation about human behaviour, language, or decision-making — sent when there’s something worth saying. Unsubscribe any time.