Practice

Quick Reflex

A word flashes up. Name the pattern before you think about it.

Detect

Full client sentences. Select every pattern you can hear.

Decode

Write what the sentence assumes to be true, then compare with the model answer.

Belief Decoder

Awareness → Cause & Effect → Complex Equivalence → Label. Sort it, then challenge it.

Challenge Lab

Identify the pattern, then formulate the question you would actually ask.

Pattern Hunt

Find every pattern in a natural statement, then choose which one to challenge first.

Create Lab

Build purposeful language for a desired internal representation.

Client autonomy first: offer, never impose.

Sentence Gym

Same outcome, different presupposition — transform and strip sentences.

AI Client

Practise with a simulated client across eight themes.

Guided Simulation — AI connection can be enabled later.

Study the Trainer

Watch a dialogue, spot what was used or challenged, then reveal the analysis.

Voice Practice

Speak or write for 60 seconds using at least four different presuppositions.

Uses your browser's own speech recognition where available, with a text fallback.

What Did I Presuppose?

Write a sentence you might say to a client, tag your own patterns, then check the cues.

Quick Reflex

A trigger word or phrase appears. Name the Meta Model violation instantly.

Detect

Real client sentences. Select every Meta Model pattern present.

Presupposition or Mind Read?

Two distortions that are easy to confuse. Decide which one is doing the work.

Decode

Say what has been distorted, generalised or deleted — and what information is gone.

Deep Structure Lab

Surface structure → undelete, undistort, ungeneralise → deeper, specific structure.

Deterministic layers: each item is the next layer down the same conversation.

Challenge Lab

Write the Meta Model question you would actually ask, then compare.

Pattern Hunt

Find every pattern in a natural statement, then choose which one to challenge first.

Coach / Client Practice

The trainer's pairs exercise: vague client statements, your Meta Model questions.

The class loop: client sentence → identify patterns → choose the primary pattern from your prediction → challenge → client answer → follow-up → deeper structure. Self-assessed; model challenges and follow-ups are shown after you answer.

Challenge Chain

Every client answer gets a follow-up question. Three levels, one thread.

AI Client

Practise a full exchange against a scripted client. AI can be connected later.

Voice Practice

A vague client line appears. Say your Meta Model question out loud, fast.

Speak aloud, or type if you prefer. No recording leaves this device.

Study the Trainer

Watch a questioning sequence, name what was used, then reveal the analysis.

Prediction First

Name what you intend to recover, then write the question that recovers it.

Advanced. The prediction comes before the question — a question without a target is a guess.

Quick Reflex / Name It

A cue appears. Name the Milton pattern it belongs to.

Study the Trainer

Trainer demonstrations with the pattern stack annotated afterwards.

Find It

A practitioner sentence. Select every Milton pattern inside it.

Pattern Stack X-Ray

The trainer's full paragraph, clause by clause. Label each move.

Clauses run in the order the trainer delivers them, so the whole stack builds as you go.

Pace vs Mind Read

Verifiable experience, or a claim about a private state?

Build It

One target pattern. Write one clean, natural example of it.

Blend It

Two patterns, one sentence that still sounds like a human being.

Conversational Postulate Lab

Turn a direct command into a respectful indirect question.

If you would not accept “no”, the question is not honest. Say it directly instead.

Extended Quote Builder

Track who is speaking at each layer, then build your own nest.

Ambiguity Prism

Phonological, syntactic, scope, punctuation — then both meanings.

Double Bind Ethics

Name the structure, then judge whether it preserves real choice.

Two separate judgements: pattern accuracy first, communication quality second. A correct pattern can still be the wrong thing to say.

Utilize the Unexpected

Something happens mid-session. Use it instead of resisting it.

Scored on three things: did you acknowledge what actually happened, did your next line follow coherently, and did the client keep their autonomy?

AI Client

A benign coaching conversation. Choose your Milton language turn by turn.

Guided Simulation — deterministic scripts. AI connection can be enabled later.

Voice / Delivery Practice

Deliver the language aloud with pace, pause and emphasis cues.

Delivery cues are there to practise against. Nothing is recorded, and the app does not analyse your tone.

Daily Five

Five patterns: recall the definition, recognise it, construct it, integrate it.

Five patterns per round, ending with one integrated sentence. Everything stays on this device.

Memory Lock

Light revision of the manual sequence, once the concepts already make sense.

Order is a revision aid, not a prerequisite. Learn the patterns first, the numbers later.

Chunking Elevator

Move up, across and down the manual hierarchy and name what becomes visible at each level.

Pair this with the interactive elevator on the Hierarchy overview page.

Direction Decision Lab

Read a real statement, choose the direction the objective needs, and say what you are going for.

Category or Part?

Tag each downward relationship as 'is a kind of' or 'is a part of' — and spot the genuinely ambiguous ones.

Structure Match

Four parallel worlds, one original relationship pattern. Pick the world whose structure actually matches.

Three-Directions Drill

One concept, three moves: broader category, same-level sibling, and a specific type or part with the relation tagged.

Scored on relational direction, not exact wording. Several answers are usually defensible.

Same-Level Siblings

Find the shared class, then step to a different branch at the same level.

Metaphor Skeleton

Map Original Structure → Parallel Structure element by element, before writing a word of story.

Complete the Metaphor

You get the setup and the obstacle. Supply the shift and resolution that stay true to the structural pattern.

Build From a Client Statement

A benign client line. Chunk up to the structure, move laterally to a world, and write the short metaphor.

The full guided eight-screen Metaphor Builder lives on the Hierarchy overview page.

Intuition Bridge

Two or three different situations. Chunk up to the shared relationship — then you MUST come back down and apply it.

An abstraction that never returns to the situation does not count.

Nit-Picking Recovery

Spot the spiral of increasing detail plus mismatch, then make the upward recovery move.

Agreement Search

Two positions. Chunk each up separately until a shared nominalization appears — then get it confirmed.

Rapid Lateral

Chunk up, then generate three parallel examples against the clock.

AI Client — Chunking

A deterministic practice client. Choose your direction turn by turn and compare with the suggested move.

Scripted simulation on this device — no external or paid AI.

Overwhelm Reducer

Take a goal nobody could act on and chunk it down until somebody could start today.

Repair a Bad Metaphor

Remove the preaching, restore the isomorphism, and give the listener their choice back.

Metaphor X-Ray

Read a metaphor and name the original relationships underneath — not the surface objects.

Study the Trainer

Annotated demonstrations: each chunking move named, with the reason it was chosen.

Content or Context?

Eighteen client statements. Content/Meaning, Context, or not enough information yet — with the structural evidence explained.

What Changed?

A statement and a response. Did the meaning move, the context move — or did nothing move at all?

Frame Shift

One behaviour, several situations. Predict how usefulness and the likely response change.

Study the Trainer

Trainer demonstration transcripts, with the analysis held back until you commit.

Meaning Multiplier

Fixed event. Generate three or more plausible meanings, then rank them: strongest fit, plausible, weak, unsupported.

Plausibility is the target, not positivity.

Reframe Generator

One statement: produce three context reframes and three content/meaning reframes where logically possible, then rank the fit.

Reframe Builder

Said → fact → added meaning → route → three possibilities → best fit → how you will test it.

The step-by-step builder also runs on the Reframing overview page.

Bad Reframe Detector

Positive thinking, argument, mind reading, too early, irrelevant cleverness, invalidation — or a genuinely good reframe.

A REFRAME MUST FIT THE CLIENT — NOT THE COACH'S EGO.

Meta Model → Reframe Bridge

Four steps: recover the event, test the equivalence, recover alternatives, and only then offer a frame.

Uses the Meta Model module for the precision questions rather than repeating them.

Intention ≠ Behaviour Lab

Split behaviour from possible intention, list several, and design the question that would actually find out.

Never assert an intention as fact. Exploring intention does not excuse harmful behaviour.

Reframe Repair

The client rejects your reframe. Choose the move that keeps the work going.

Rapid Reframe

Statement → classify → one plausible question or reframe → self-check. Speed drill, deterministic model answers.

Live Guided Simulation

Benign coaching scenarios: hear the statement, choose context / meaning / explore first, respond, then continue or repair.

Scripted on this device — no external or paid AI.

Memory Lock

Same behaviour — change context. Same event — change possible meaning. Separate intention from behaviour.