Sleight Of Mouth

Language, often unknown to us, creates our mental models, our reality, and defines our meaning. But the model is not the terrain, and by engaging with the language we can change mental models. Sleight of mouth is a series of techniques for changing our models, reality and meanings, and thus our experiences. Our mental models are a frame, and we need help going from a ‘problem frame’ to an ‘outcome frame’, a ‘failure frame’ to a ‘feedback frame’, and an ‘impossible frame’ to an ‘as if’ frame. To experience the power of language, find anything that is invisible to you, and name it. Now that you’ve given it a name, see how you perceive it, and become aware of it.

These are my notes on Robert Dilts’ Sleight of Mouth, with Michael Hall and Bobby Bodenhamer’s Mind-Lines folded in. Dilts reverse-engineered fourteen moves by modeling how people like Lincoln, Gandhi, and Socrates argued their way through other people’s beliefs — and, more usefully, how you argue with your own. Hall and Bodenhamer took the same moves, renamed them mind-lines (because “sleight of mouth” sounds like a con), and sorted them into seven directions you can send a mind. The post follows the argument in order: words build the experience, a limiting belief is the target (an equation linking an external behavior to an internal state), the fourteen patterns are the tools, the seven directions are the map of where each tool sends your attention, and the payoff is running several of them against one belief at once.

• Interactive version
  • Run it on your own belief with an LLM
• Words build the experience
  • As small as ‘and’, ‘but’ or ‘even though’
  • What the patterns train
• The target: a limiting belief
  • Thought viruses
• The fourteen patterns
• Seven directions: where each pattern sends your attention
  • Deframing: how do you know?
  • Counter reframing: turn it on itself
  • Post-framing: play it forward
  • Outframing: wrap a bigger idea around it
• Frames: same facts, different picture
  • Outcome frame, not problem frame
  • Change frame size
  • Context reframing
  • Content reframing
  • Punctuation: where the period goes
  • Model of the world: take the second position
• Intention and redefining: reopening a closed conversation
  • Redefining, one word at a time
  • Turning critics into advisors
• Chunking: change the size of the thought
  • Chunking down
  • Chunking up
  • Chunking laterally: analogy
  • Finding isomorphisms
• Values and criteria: the deepest lever
  • Values and motivation
  • Criteria and judgment
• Levels of experience: where to aim
• Putting it together: one belief, ten patterns
• Resources

Interactive version

Pick a stuck belief from the dropdown; it shows as an equation, external behavior → link → internal state. Then try the patterns on it: the part a pattern pries at lights up (the external behavior, the link, the internal state, or the belief as a whole), the reframed sentence appears right underneath, and each pattern carries a one-line hint of what the move does. “About this pattern” explains the one in play, “By section” groups the patterns by the section of this post that teaches them, and “By direction” groups them by the seven directions a pattern can send your attention: down, inside, back on itself, back in time, forward, up, or sideways. (This replaces Dilts’ original 1987 diagram, which packs the same fourteen moves around a single cancer example.)

Run it on your own belief with an LLM

No explorer has your belief in it, so take the patterns to your own LLM. Copy this prompt into ChatGPT, Claude or whatever you use, put your belief where it says, and it runs all fourteen and shows which part of the belief each one pries at.

Help me loosen a belief that's holding me back, using Robert Dilts' fourteen Sleight of Mouth patterns.

My belief: [WRITE YOUR BELIEF HERE]

First, split my belief into external behavior → link → internal state (what happened, the "so" or "means" that joins them, and the meaning or feeling I attached), and show the split.

Then run each pattern below on it. For each one, give the pattern name, the part of the belief it targets (the external behavior, the link, the internal state, or the whole belief), and one reframe in my voice, one or two sentences, honest to the facts I gave. No pep talk.

1. Intention (targets the whole belief): What is the positive purpose of this belief?
2. Redefining (targets the external behavior): What's another way to say this, with slightly different words?
3. Consequence (targets the whole belief): Where does holding this belief lead?
4. Chunk down (targets the external behavior): Get more specific. Which part, when, how much?
5. Chunk up (targets the link): What bigger class does this belong to?
6. Analogy (targets the link): What's a parallel situation that reads differently?
7. Change frame size (targets the external behavior): Zoom the time or the number of people in or out. Does it still mean the same?
8. Better outcome (targets the internal state): What outcome would actually be worth aiming at?
9. Model of the world (targets the link): Who would see the same facts and draw a different conclusion?
10. Reality strategy (targets the external behavior): How do you know this is true? What would you have to perceive?
11. Counter example (targets the external behavior): Where does the rule not hold?
12. Hierarchy of criteria (targets the internal state): What matters more than what this belief is measuring?
13. Apply to self (targets the whole belief): Turn the belief's own judgment back on the belief.
14. Meta frame (targets the whole belief): What belief about this belief changes how it looks?

Finally, ask me which reframe loosened the belief the most, and why.

Words build the experience

The words we reach for don’t just describe an experience, they assemble it. Swap the word and you change what the experience is — before you’ve changed a single fact.

As small as ‘and’, ‘but’ or ‘even though’

The connective between two facts decides which one wins. ‘But’ stresses what comes after it; if the first part is the one you want to land, use ‘even though’.

• A and B: It’s sunny today AND raining tomorrow
  • Equal weight A and B
• A but B: It’s sunny today BUT raining tomorrow
  • B is emphasized.
• A even though B: It’s sunny today even though it’s raining tomorrow.
  • A is emphasized.

What the patterns train

If words build the experience, then working the words deliberately is a skill. The techniques train us to:

• ‘re-punctuate’ and ‘rechunk’ our perceptions
• Identify and appreciate different perspectives and alternative models of the world
• Discover the internal process by which we assess ‘reality’, and through which we form and update our beliefs
• Explore how we build the mental maps by which we form expectations, determine cause, and give meaning to our experiences.
• Recognize the influence of our internal states on our beliefs and attitudes
• Pace the natural process of belief change
• Understand the impact of language and beliefs on different levels of our experience
• Become more aware of potential verbal ‘thought viruses’ and unspoken assumptions and presuppositions

The target: a limiting belief

At the core is a limiting belief, and it almost always has the same shape: a subject, a judgment, and a because/means/causes link to a consequence.

Hall and Bodenhamer write that shape as an equation. On the left is an external behavior: something that happened, which you could catch on video. On the right is an internal state: the feeling, judgment or significance you attached to it. Nothing means anything by itself; the equation is where meaning gets made, and the link comes in two flavors:

• Cause-effect: “When you raise your voice, you make me anxious.”
• Complex equivalence: “When you raise your voice, it means you don’t respect me.”

That link is where the patterns get their grip. Each one loosens the external behavior (is it really what you think it is?), the link (does it really lead there?), or the internal state (is that even the verdict that matters?). The explorer shows every belief in this shape, with the same colors, and aims each pattern at one of the three parts.

Two things to do before reaching for any pattern. First, get the whole equation: people usually say only half of it (“I’m so stressed”), so ask for the missing side — “what does that mean to you?” for the internal state, “how do you know?” for the external behavior. Second, notice that a belief is a thought you have said “yes” to. The “yes” sits one level above the thought, which is why a belief feels like a fact rather than an opinion, and reframing starts the moment any other frame goes around it: a “maybe,” a “since when?”, or a story.

Thought viruses

Dilts calls a limiting belief that’s sealed against evidence a thought virus — it ships with its own defenses (“well, you’d say that”), it spreads, and it runs in the background unexamined. The patterns are the antibodies. And the trick runs both ways: you can deliberately build a positive thought virus — an empowering belief wired with the same self-reinforcing structure (“every rep is a deposit”) so that it defends and propagates itself in place of the limiting one.

The fourteen patterns

Run the belief through these questions (page 318). The sections after this list unpack them by family: frames, intention and redefining, chunking, and values.

1. Intention - What is the positive purpose of this belief?
2. Redefining - What is another way to say this that’s more positive?
3. Consequence - What is a positive effect of this belief?
4. Chunk down - Apply the belief to a more specific situation.
5. Chunk up - Apply the belief to a more abstract, general situation.
6. Analogy (chunk laterally) - Give a related metaphor that’s more positive.
7. Change frame size - Widen or narrow the time frame or the number of people involved, until the implication changes.
8. Better outcome - What would be a better outcome?
9. Model of the world - What different model of the world would you need for this to be positive?
10. Reality strategy - How would you need to perceive the world for this belief to be true?
11. Counter example - Find an exception to the rule.
12. Hierarchy of criteria - What criterion is more important than the one you are considering?
13. Apply to self - Apply the belief’s own judgment to the belief itself. “It’s too expensive” - how expensive is it to keep holding this belief?
14. Meta frame - What other belief about this belief could change or enrich the perception of it?

Try them on a belief in the interactive version above.

Seven directions: where each pattern sends your attention

Fourteen patterns is a lot to hold. What Hall and Bodenhamer add in Mind-Lines is the map: every move sends attention in one of seven directions relative to the equation. They count twenty lines, because several of theirs are finer cuts of one Dilts pattern (reframing the external behavior and reframing the internal state are both redefining; first outcome and outcome of the outcome are both consequence). The simplest version of the map has only two levels: the first two directions work inside the equation and change what the external behavior or the internal state is; the other five leave the content alone and wrap a frame around it (time, a value, an identity, a story), and the frame changes what the content means.

Chunking, intention, frames, values and analogy each have a section below. The four moves that don’t are here, run against one belief from the explorer: “I skipped the gym all month, so I’m just not a fitness person.”

Deframing: how do you know?

Chunking down asks which part. Its partner, reality strategy, asks how the belief gets built, step by step: “How do I know I’m not a fitness person? Do I see the empty calendar first, hear a voice, or feel the sinking and then find the word?” The evidence usually turns out to be one vivid memory or a feeling, and a belief standing on one memory has a short foundation. Hall and Bodenhamer moved deframing to the front of their second edition because it worked best in their workshops: a vague belief often falls apart as soon as someone asks “specifically?”

Counter reframing: turn it on itself

A belief is a rule, and rules can be tested. Counter example finds the case where the rule fails: “I walked to work every day last spring and loved it. Was that someone else?” One real exception shrinks “always” into “sometimes.” Apply to self turns the belief’s own standard on the belief: “If one month is enough to define me, then one good month redefines me just as fast — the belief can’t have it both ways.” A belief that fails its own test loses its authority. Hall and Bodenhamer also point the judgment at the listener (“so when you miss a week, you’re not a fitness person either?”), the same move with the speaker watching someone else wear it.

Post-framing: play it forward

Intention is the pre-frame: go back to the good purpose the behavior was serving. Its mirror is consequence: set aside whether the belief is true and look at what believing it does. “If I’m ‘not a fitness person,’ tomorrow’s workout becomes a test I expect to fail” — and a few failed tests later, the label is the reason to quit the habit I’d built. Limiting beliefs tend to cause the outcome they predict. Eternity framing is change frame size stretched all the way out: at 80, will I remember this month, or decades of mostly showing up?

Outframing: wrap a bigger idea around it

Outframing is the biggest family — model of the world, hierarchy of criteria, chunk up and change frame size all live here and have their own sections. The one that doesn’t is the meta frame: a belief about the belief. Hall and Bodenhamer’s lines 16 through 19 are all flavors of it. Is it a rule? (“Do I have to hit every session to count — and what happens if I don’t?”) Is it an identity? (“I’m a person who skipped a month, not a non-fitness person.”) Is it an instance of something I already know about? (“This is all-or-nothing scoring, and a streak counter isn’t a character test.”) Does it serve me? (“Does this belief get me to the gym more often, or just make rest days miserable?”) The belief may not change at all. Its standing does: once it’s “a thought I have on bad days,” it stops being the weather.

Frames: same facts, different picture

A frame is how we experience the world. Think about picture frames: the frame bounds what you perceive, and thus how you experience and react to reality. Most of the patterns are ways of moving the frame.

Outcome frame, not problem frame

Not “I have a problem,” but “I want this outcome.” The outcome frame is more powerful and action oriented.

• Outcome frame

  • What do I want?
  • How do I get it?
  • What resources are available?

• Problem frame:

  • What is wrong?
  • Why is that a problem?
  • What caused it?
  • Whose fault is it?

The same move works on the other two frames from the top of the post: a failure becomes feedback (“what did this teach me?”), and an impossible becomes an ‘as if’ (“if it were possible, what would I do first?”).

Change frame size

Zoom the lens in or out until the problem changes size. Will this matter in a week? A year? On my deathbed? Zooming out shrinks most crises into footnotes. Zooming in does the opposite good work: “I’ll never get in shape” is too big to fight, but “can I do one push-up right now?” is a frame small enough to actually win.

Context reframing

Every behavior is useful somewhere — find the context where it’s a gift. The friend who’s exhaustingly anal-retentive is exactly who you want packing the moving truck or reading the contract. The question stops being “how do I get rid of this trait” and becomes “where is this trait an asset?”

Content reframing

Keep the facts, change what they mean. Instead of judging the behavior, look for the positive intention underneath it — the intention is almost always good, it’s the expression that’s clumsy. A kid who won’t stop arguing isn’t “defiant,” they’re practicing how to hold a position. My back pain isn’t “my body failing me,” it’s my body sending a signal I finally learned to read.

Punctuation: where the period goes

Experience arrives as a run-on stream. We punctuate it — deciding where one event ends, where the next begins, and which clause caused which. Framing is the punctuation of objective reality into our subjective reality. Move the punctuation and the meaning moves with it, even though not a single fact changed.

• The classic demo flips meaning on comma placement alone: “A woman without her man is nothing” versus “A woman: without her, man is nothing.” Same words, repunctuated into opposites.
• Meaning isn’t in the event, it’s in the frame the event lands in. A castle under siege, its people starving, catapults its last food over the walls in defiance. The besieging army — already demoralized, low on supplies — reads the flying food as proof the castle can outlast them, and goes home. The people inside meant “defiance,” the army outside read “they can outlast us.” The meaning lived in the punctuation, not in the food.
• Applied to a belief: “I bombed the talk. I’m not a speaker.” becomes “I bombed the talk, and I now know the three slides that lost the room.” Same facts — the period just moved from a verdict to a comma in front of feedback.

Model of the world: take the second position

Seek first to understand. Stand in each of three positions and the same event turns into three different events:

• First position — my frame. How it looks from inside my own map.
• Second position — their frame. Step into the other person’s shoes and argue their case as if it were mine.
• Third position — the observer. Watch the two of us from the balcony, like a documentary of the conversation.

Most stuck arguments are two people jammed in first position, each certain the other is just wrong. Genuinely visiting second position — not performing it, actually going there — dissolves more conflict than any clever comeback.

Intention and redefining: reopening a closed conversation

Two patterns keep coming up, because they’re the ones that turn a wall back into a conversation:

• Intention: direct attention to the person’s intention — the desired outcome underneath the complaint.
• Redefining: swap in words that are similar but slightly different, usually more positive.

Together they drive the three big shifts: problem → outcome, failure → feedback, critic → advisor. Convert the problem into the intention and you’re back in the outcome frame. Watch a good salesperson do it:

• Customer: I’m afraid that’s too expensive.
• Salesperson (thinking): need to understand the intention, take a few stabs at it.
• Salesperson: Do you think the item is overpriced, or are you worried you can’t afford it?

Those aren’t the same worry — overpriced → unable to afford → expensive for the value are three different beliefs with three different answers. Same move as “how much pain are you in?” versus “how much discomfort do you feel?”: the word sets the size of the thing.

Redefining, one word at a time

Take a statement and slide it one notch more positive, then one notch more negative. Same underlying thing, three different verdicts — which proves the verdict was never in the thing, it was in the word you picked.

The same act, cooling down or heating up:

• It doesn’t piss me off → I’m OK with it → I’m accepting it
• I made a mistake → you twisted my words → he’s a damn liar

The same trait, spun up and spun down (the middle word is the neutral one):

• Responsible → Stable → Rigid
• Stable → Comfortable → Boring
• Fiscally responsible → Frugal → Stingy
• Friendly → Nice → Naive
• Big imagination → Telling fairy tales → Lying

Whoever gets to pick the word gets to pick the meaning. Notice when someone hands you theirs, and notice when you can hand back a better one.

Turning critics into advisors

There are three roles in any plan: dreamers, realists, and critics. Critics usually assert a generalized judgment — “this won’t work.” That puts the listener in a polarized position, where the only answers are yes or no. But notice the critic’s intention is good: it’s to evaluate, and to help the dreamer and the realist.

The fix is to restate the criticism as a positive statement of its positive intention. Criticism usually comes in the negative form — “avoiding stress” instead of “feeling relaxed and comfortable.” Switch from what you don’t want to what you do want. If I don’t want X, what is it that I do want? Say that instead:

• It is… → I want to make sure…
• Too expensive → I want it to be affordable
• Fear of failure → Desire to succeed
• Unrealistic → Concrete and achievable
• Too much effort → Easy and achievable

Then turn it into a “how” question: “it’s too expensive” → “how will you make sure it’s affordable?” As a recipe, ask three questions, especially when you’re the critic:

1. What is your criticism or objection?
2. What do you want to achieve via your criticism (the thing that’s desired, not the thing that’s bad)?
3. What “how” question can be asked about how it will be achieved?

Notice how questions shift the critic from a “spoiler” into an advisor. Pro tip: before going into what needs to be achieved, call out the things that are good.

Chunking: change the size of the thought

Chunking changes the specificity of a thought:

• Down → more specific
• Up → more abstract
• Lateral → analogy

Chunking down

Get specific until the generalization can’t survive the detail. A big vague judgment only holds up from a distance.

• “I’m ugly.” Ugly which part — nostril, arm, legs? Ugly to whom — my mom, myself, a stranger? The word can’t survive the questions.
• “I’m learning disabled.” Disabled at which step — taking information in, storing it, or reproducing it? Now it’s a pipeline with a fixable stage, not an identity.
• “I’m a failure.” Failed at goal-setting, planning, or execution? Name the stage and it becomes a skill to drill, not a verdict to carry.

Chunking up

By going up, we see whether it’s really as bad, and whether there’s something better to chunk back down to. “I failed this project” chunks up to “I’m learning to run projects” — a class in which one failure is a single data point, not the verdict. Then chunk back down to the one skill worth fixing next.

Chunking laterally: analogy

By finding an analogy, we understand the current statement through the new one. “My code keeps breaking, so I’m a bad engineer” → “A surgeon’s first incisions bled too; the reps are how the hand steadies.” The analogy smuggles in a reframe the direct statement fights off.

Finding isomorphisms

An isomorphism is the deep version of an analogy: a problem with the same shape as yours in a completely different domain — and the borrowed domain often carries a solution your own is too close to see. A game for finding one:

• Three people: A, B, and C.
• A describes a problem they want help with.
• B and C hunt for its critical elements — the underlying structure, stripped of A’s specifics.
• B and C build a metaphor with that same structure from another world — fantasy, sports, war, folk tales, Star Trek — and hand it back to A.

The reframe rides in on the metaphor. “Your co-founder standoff is two gunslingers who both know whoever draws first loses” lands somewhere the direct advice can’t reach.

Values and criteria: the deepest lever

Underneath every belief is a value it’s trying to serve and a set of criteria for whether it’s being served. This is the deepest place to reframe: change what someone measures by, and the whole surface argument reorganizes itself.

Values and motivation

Values are the reasons you get out of bed — the filter you see reality through. Someone who values fun automatically asks “how much fun is in this?” of everything they meet. Goals are values wearing work clothes: if my goal is to build a product, I probably value making new things. Find the value under a goal and you’ve found the lever that actually moves the person.

Criteria and judgment

• Value — what we desire and want (health, freedom, respect).
• Criteria — the standard and evidence we use to decide whether we’re getting it.

The leverage is that criteria are negotiable. Two people can share the value “success” and measure it by completely different evidence — one counts titles, the other counts free afternoons. Argue the criteria, not the value, and you can move someone without ever attacking what they care about. The hierarchy of criteria pattern is this move: find a criterion that matters more than the one the belief is defending.

Levels of experience: where to aim

Our experiences filter through these layers (Dilts’ Neuro-Logical Levels, borrowed from Gregory Bateson):

• Environment
• Behavior
• Capability
• Beliefs and Values
• Identity

We assign blame across the levels inversely: for ourselves we attribute fault to the environment, but for others we attribute it to their identity. This is the fundamental attribution error.

The levels also tell you where to aim a reframe. Answering an identity attack (“I’m a failure”) at the level of identity just trades verdicts; answer it a level or two down — at behavior and capability (“here’s the specific skill, here’s the next rep”) — and the frame actually moves.

Putting it together: one belief, ten patterns

No single pattern is the magic bullet — the power is running one belief through several at once, from different angles, until the frame loosens and stops feeling like reality. Take one of my own limiting beliefs, straight from the table above: “My back has hurt forever, so I’ll never get better.” Here it is hit with ten of the fourteen:

• Intention: the belief is trying to protect me from getting my hopes up and being let down again.
• Redefining: “hurt forever” is really “hurt through a long stretch I didn’t yet know how to train through.”
• Consequence: believing it guarantees I skip the rehab that’s the one thing that could change it.
• Chunk down: which movement, on which day, at what load? “Forever” dissolves into specific reps.
• Chunk up: bodies are systems that adapt to load — I’m inside a whole class of problems that respond to the right stimulus.
• Counter example: there was a month last year it didn’t hurt. If it’s “forever,” where did that month go?
• Analogy: a rusty hinge isn’t a broken door, it’s an un-oiled one.
• Apply to self: how much has holding “I’ll never get better” cost me — has that belief ever bought me a single good rep?
• Hierarchy of criteria: being strong enough to play with my kids matters more than being right about the prognosis.
• Change frame size: will this specific flare-up matter across a year of consistent training?

Fire enough of these and the belief stops feeling like the truth and starts feeling like one sentence I happened to write — which is the thesis of the whole book: the words built the experience, so different words can rebuild it. (My actual back saga is in physical pain.)

Four habits from Hall and Bodenhamer that make the stack work: get the whole equation before you reframe anything; start by going down, since a vague belief often dissolves at “specifically?”; watch the other person’s face, because the shift shows up there before it shows up in words; and keep the outcome kind — pried open against someone’s interest, these moves are manipulation, and offered as a more useful frame, they’re a gift. The easiest place to practice is on your own inner critic, which speaks in tidy equations.

Resources

Robert Dilts’ Sleight of Mouth is the source of the fourteen patterns. L. Michael Hall and Bobby G. Bodenhamer’s Mind-Lines: Lines for Changing Minds (second edition) is the source of the equation and the seven directions; later editions grow the list to 26 lines, the same engine with finer cuts. For the internal-negotiation cousin of these moves, see get to yes with yourself and the voices in my head.