Immunity to Change

1. Introduction

Version followed: Kegan & Lahey (2009), Immunity to Change (Harvard Business Press), Part Three "Over to You" (chapters 8 to 10: the four-column map and the Opening Moves / Middle Game / End Game process), read through the HarvardX course excerpt; plus Minds at Work's current description of the ITC Map. The authors' 2001 HBR article ("The Real Reason People Won't Change") is drawn on only as an abstract, so its question sequence is not used here [uncertain].

Immunity to Change (ITC) is a coaching and adult-development framework, not a psychotherapy. Robert Kegan and Lisa Laskow Lahey (Harvard Graduate School of Education) developed it; their firm Minds at Work trains and certifies ITC coaches. It is used one-to-one, in pairs and with leadership teams, mainly in organisations and education. [source: Kegan & Lahey 2009; Minds at Work, ITC page and Online Coach Development page]

How change happens, as the authors describe it. People repeatedly fail at goals they sincerely hold. The reason is not weak will. Their undermining behaviours serve a hidden, competing commitment, so they drive with "one foot on the gas" and one on the brake. That commitment is held in place by big assumptions: beliefs taken as plain truth. The authors call this self-protective system an anxiety-management system. Change becomes possible when the person moves the big assumption from something they are "subject to" (looking through it) to something that is "object" (looking at it). They then test it with small, safe experiments and revise it in light of the data. The method rests on Kegan's constructive-developmental theory of adult meaning-making. [source: Kegan & Lahey 2009 ch. 8–10; Minds at Work ITC page]

Use by a psychotherapist. The authors publish ITC for coaching and development, not for treating mental disorders. Using the map inside psychotherapy (for example as a formulation or behavioural-experiment tool) is a derived application [synthesis]. Section 5 sets out the coaching/therapy boundary as ICF, EMCC and AC publish it, and the referral signs.

2. Framework

Overall shape [source: Kegan & Lahey 2009 ch. 10, HarvardX excerpt]

Duration and dose: the authors plan for a few months, not a single session: about 30 minutes a week for about 12 weeks, the period over which most people notice significant change.
Format: alone, with a partner (ideally someone also working on their own immunity), or with a trained ITC coach who guides the whole process. Others who know the goal (colleagues, family) give cues, feedback and accountability.
Programme formats (Minds at Work): the Change Course is a six-week personal experience for anyone working on a goal. Foundations to Application trains coaches and facilitators to guide others through a full map and test design, one-to-one or in groups. Coach Certification is a selective 12-month programme for experienced coaches: each works through the full process as participant and as coach, with a certified mentor and a cohort. ITC with Teams also exists; no procedural detail was found [uncertain]. [source: Minds at Work Online Coach Development page; ITC page]

Phase 0: Preconditions (the three ingredients) [source: Kegan & Lahey 2009 ch. 8]

Goal: check the person has a "gut" source of motivation, will engage thinking and feeling together, and will act ("hand").
Moving on: the goal feels necessary, not just sensible.

Phase 1: Building the immunity map (diagnosis, the "X-ray") [source: Kegan & Lahey 2009 ch. 9; Minds at Work ITC page]

Column 1, improvement goal → Column 2, doing / not doing instead → Column 3, worry box then hidden competing commitments → Column 4, big assumptions.
Typical format: one long session or a workshop, often with prework gathering feedback on column 1. The authors report that a one-off group session with no prework had no lasting effect. Exact session length is not given in the sources [uncertain].
Criteria to move on: column 1 meets the goal criteria. Column 2 lists concrete behaviours. Column 3 makes every column 2 behaviour look sensible, and the person feels pulled in two directions. Column 4 holds assumptions each of which makes a commitment inevitable, which together sustain the whole system, and which point to a larger world the person has been avoiding. The map should show why simply stopping the column 2 behaviours will not work (the "adaptive formulation").

Phase 2: Opening Moves (setting the stage) [source: Kegan & Lahey 2009 ch. 10]

Hone the map until it feels powerful and holds at least one testable big assumption.
Run an initial survey of others for a baseline on the column 1 goal.
Moving on: there is a testable assumption written in "if… then…" form, and baseline ratings exist.

Phase 3: Middle Game (digging in) [source: Kegan & Lahey 2009 ch. 10]

Continuum of progress → self-observation of the big assumption → biography of the big assumption → a series of tests that run against the assumption, each larger than the last.
Moving on: test results have been explicitly linked back to the assumption, and the assumption has been revised or its limits mapped. The excerpt breaks off during test design, so the authors' rule for how many tests to run before consolidating was not read [uncertain].

Phase 4: End Game (consolidating) [source: Kegan & Lahey 2009 ch. 10]

Follow-up survey of the same people; compare it with the self-assessment. Identify hooks and releases. Plan future progress, and re-engage the process for goals not yet met.

3. Guiding principles

[source: Kegan & Lahey 2009 ch. 8–10 unless marked; Minds at Work ITC page]

Immunity, not weakness. The undermining behaviours are doing a protective job. Practically: never treat column 2 as a fault list to eliminate; be curious about what each behaviour protects.
Adaptive, not technical. Trying harder to stop the column 2 behaviours is a technical fix and fails, because the behaviours serve column 3. Practically: resist action plans until column 4 is visible.
Gut. Motivation must be visceral: distress at the cost of the status quo, hope and self-efficacy, or a felt gap between what one wants and what one does. Practically: if the goal is "should", not "must", rework it before mapping.
Head and heart together. The immunity manages anxiety, so purely intellectual mapping does not change it. Practically: in column 3, stay with the feeling until a real fear is named.
Hand. Mindsets change through acting against the assumption and reading the data. Insight alone is "mere reflection" and action alone is a technical fix. Practically: every session after mapping ends with an observation or test task.
Subject to object. The aim is to see the assumption rather than see through it. Practically: write assumptions down, in the person's words, as hypotheses.
Information, not improvement. A test exists to gather data about the assumption, not to perform better. Practically: judge a test by what it taught, not by whether it went well.
Safe and modest tests. Start small, choose situations where a poor outcome would not be costly, and break catastrophic assumptions into smaller links. Practically: never set a test whose worst case is serious harm.
The goal is about the self. Column 1 must be important to the person and to people around them, centre on the person's own change rather than other people's behaviour, and be stated affirmatively. Practically: redirect "I want my team to…" to "I want to become…".
Concreteness. Column 2 lists observable behaviours, not states of mind. Practically: ask "what would I see you do?".
Social context. Outside input on the goal, and surveys before and after, keep the work honest. Practically: build feedback gathering into the plan from session 1.
Hidden commitments are reasonable. The authors hold that failing to carry out visible commitments often reflects success at invisible ones (paraphrase of the quote on the Minds at Work ITC page). Practically: treat column 3 with respect, not as sabotage.
Development takes time. Plan for months of weekly practice, alone, with a partner or with a coach. Practically: contract for about 12 weeks rather than one session.
Work within competence and contract clearly (coaching bodies, not the ITC authors). Practically: explain the nature of the work, its confidentiality limits and when you would refer, before you begin. [source: ICF Code of Ethics 2025, standards 1.1, 2.3, 5.1; Global Code of Ethics, Section 4]

4. Intervention inventory

The items below are the exercises named in Kegan & Lahey (2009) chapters 8 to 10 and on the Minds at Work ITC page, in the order the authors give. Steps are paraphrased. All example language is the guide's own (illustrative). The authors' worked examples are summarised briefly: Fred (listening), Sue (saying no), Claus (over-preparing), Peter's leadership team. [source: Kegan & Lahey 2009, HarvardX excerpt]

4.1 Readiness check: gut, head and heart, hand

Purpose / when: before mapping. Confirms the person has a felt need for change and will both reflect and act. [source: ch. 8]
Steps: (1) Ask what the status quo is costing the person or people they care about. (2) Ask what gives them hope that they could change. (3) Name the gap between what they want and what they do. (4) Explain that the work needs both honest feeling and small real-world actions over about 12 weeks. (5) Agree a support arrangement: partner, coach or trusted others. [source: ch. 8, ch. 10]
Example language (illustrative): "What does it cost you, and people you care about, if this stays exactly as it is for another year?"
Watch for: goals chosen because someone else says so, which tends to produce a flat, intellectual map. It is working when the person shows some urgency or emotion about the goal. A person with no energy for any goal, or with marked hopelessness, may need clinical assessment rather than ITC (see 5.8) [synthesis].

4.2 Column 1: the improvement goal (with outside input)

Purpose / when: first column; sets the one goal the whole map serves. In Peter's team each executive chose a single goal that excited them, added value and needed personal growth rather than a learnable technical skill (without requiring a "personality transplant"). [source: ch. 9]
Steps: (1) Draft a goal for the person's own change. (2) Check it against the four criteria: important to the person, with urgency; important to people around them; about changing themselves, not others; stated affirmatively, as what they want to become rather than stop. (3) Get outside input before finalising: ask people at work and at home whether the goal would matter to them and whether they see a better one. In Peter's team this was one-to-one feedback from the boss, peers and a direct report, followed by colleagues sharing and challenging each other's goals and checking with family. (4) Revise. First drafts often fail the criteria. [source: ch. 9]
Example language (illustrative): "Say it as something you want to get better at, rather than something you want to stop. What would the people who work closest with you say it should be?"
Watch for: goals about other people ("get my team to…"); purely technical goals ("learn the software"); negative phrasing. The authors' Fred example (to be a better listener, more present and patient) shows the target form. It is working when the person says "yes, that's the one" and others confirm it matters.

4.3 Column 2: the fearless inventory (doing / not doing instead)

Purpose / when: second column; lists what the person actually does, or fails to do, that works against the goal. [source: ch. 9]
Steps: (1) List concrete, observable behaviours and omissions. (2) Exclude states of mind ("I get impatient"); translate them into behaviour ("I interrupt"). (3) Be thorough and honest; aim for an exhaustive list. (4) Do not explain, justify or plan fixes yet. (5) List only undermining behaviours, not things already done well. (6) Ask trusted colleagues for items, or use feedback already gathered. [source: ch. 9]
Example language (illustrative): "If I were a camera in the room, what would I record you doing that pulls against this goal? Don't fix it yet; just list it."
Watch for: solutions creeping in ("I should…"); self-attack. The coach keeps the tone descriptive. It is working when the list is long, specific and recognisable to others.

4.4 Column 3, step 1: the worry box

Purpose / when: after column 2; surfaces the fear that the column 2 behaviours manage. [source: ch. 9]
Steps: (1) Take each column 2 behaviour and imagine doing the opposite. (2) Ask what the most uncomfortable, worrisome or frightening feeling would be. (3) Treat surface feelings such as boredom or impatience as possible "covers" and keep going until a deeper fear appears (the authors give humiliation, helplessness and making a big mistake as examples). (4) Write the fears in the worry box. [source: ch. 9]
Example language (illustrative): "Picture yourself doing the exact opposite, right now, in that meeting. What's the feeling you'd least want to have? And under that?"
Watch for: this is the most emotionally charged step. Mild discomfort is expected and useful. Panic, dissociation, intrusive trauma memories or self-harm thoughts mean pausing the exercise (see 5.3 and 5.8) [synthesis]. It is working when the person names a fear with visible recognition.

4.5 Column 3, step 2: hidden competing commitments

Purpose / when: turns each fear into the commitment the behaviour serves. [source: ch. 9; Minds at Work ITC page]
Steps: (1) Rewrite each fear as a commitment to prevent it, in the form "I am also committed to not …" (Fred: not looking stupid, not being humiliated). (2) Check the authors' criteria: each commitment ties clearly to a fear and to self-protection; each makes the column 2 behaviours look sensible; it becomes clear why simply stopping them won't work; and the person feels a foot on the gas and a foot on the brake. [source: ch. 9]
Example language (illustrative): "So alongside wanting to listen better, you are also committed to never being caught not knowing. Does that make sense of everything in column 2?"
Watch for: commitments phrased as virtues ("committed to being efficient") that do not carry the fear. Push them back to self-protection. It is working when the person sees column 2 as smart rather than stupid.

4.6 Column 4: big assumptions

Purpose / when: the beliefs that make column 3 commitments necessary. The authors say this is where change becomes possible. [source: ch. 9; Minds at Work ITC page]
Steps: (1) Brainstorm beliefs held as unquestioned truth that make each commitment feel necessary, e.g. Fred's: if he made a big mistake he could not recover from it. (2) Write them as "I assume that if… then…" (the Harvard Extension summary also uses "if ___, then ___"). (3) Check the criteria: some feel true, some false, some uncertain; each makes at least one commitment inevitable; together they keep the whole system in place; they point to a bigger world the person has been avoiding. [source: ch. 9–10; Harvard Extension blog]
Example language (illustrative): "What would you have to believe for that commitment to make total sense? Finish this: 'I assume that if I…, then…'"
Watch for: assumptions that are really facts (real danger, e.g. an abusive partner or a hostile employer); see 5.7. It is working when the person says something like "I've never said that out loud, but I live as if it's true."

4.7 Reading the whole map (the adaptive formulation)

Purpose / when: at the end of mapping; checks that the map is "powerful". [source: ch. 9]
Steps: (1) Read the map left to right as one sentence-chain: goal, behaviours, fears and commitments, assumptions. (2) Confirm it shows why a technical fix (eliminating column 2) would fail. (3) Ask whether it feels powerful. If not, revisit column 1 (importance) or column 3 (depth). [source: ch. 9–10]
Example language (illustrative): "Reading across: you want X, you do Y, because you're also protecting yourself from Z, given the assumption W. Does that feel like you?"
Watch for: intellectual agreement without emotional recognition, which usually means column 3 is too shallow. It is working when the person sees the "immune system" as one coherent, protective system.

4.8 Honing the map (Opening Move)

Purpose / when: first step after the map is built; makes it sharper and testable. [source: ch. 10]
Steps: (1) Revisit each column against its criteria. (2) Make sure at least one big assumption is specific enough to test, in "if… then…" form. (3) Revise wording until the person feels the map "has them". [source: ch. 10]
Example language (illustrative): "Which of these assumptions, if it turned out to be wrong, would loosen everything else?"
Watch for: vague assumptions ("people won't like me"). Narrow them to who, when and what exactly.

4.9 Initial survey of others (baseline)

Purpose / when: Opening Move; gets outside ratings of where the person stands on the column 1 goal before change work begins. [source: ch. 10]
Steps: (1) Choose a handful of people who see the behaviour (colleagues, family). (2) Ask them for input on the column 1 goal and how the person currently does on it. (3) Record the baseline and keep it for the follow-up survey (4.15). The excerpt does not give a survey form or rating format [uncertain]; a simple rating plus one example comment is a reasonable format [synthesis].
Example language (illustrative, to a colleague): "I'm working on becoming a better listener. How would you rate me on that now, and can you give me one recent example?"
Watch for: the person choosing only friendly raters, or catastrophising about asking. Asking can itself be an early, mild test.

4.10 Continuum of progress

Purpose / when: Middle Game; makes the goal concrete by describing what full success would look like. [source: ch. 10]
Steps: (1) Describe, in observable terms, what fully achieving the goal would look like. (2) The exercise name suggests a scale from first steps to full success. Intermediate anchor points are not detailed in the excerpt [uncertain]; defining a "good first step" and "significant progress" point is a useful addition [synthesis].
Example language (illustrative): "Twelve weeks from now, if this had really worked, what would people see you doing differently on a Tuesday afternoon?"
Watch for: perfectionistic endpoints that recreate the immunity (an over-preparer defining success as "never unprepared").

4.11 Self-observation of the big assumption

Purpose / when: Middle Game, before active testing; the person watches the assumption at work in daily life. [source: ch. 10]
Steps: (1) For a week or more, notice when the big assumption is activated: situations, thoughts, feelings, what they then did. (2) Also notice moments when it proved inaccurate or did not hold. (3) Make no attempt yet to change the behaviour. (4) Review the notes with the coach or partner. [source: ch. 10] A simple log (situation, what the assumption said, what I did, what happened) is a practical format [synthesis].
Example language (illustrative): "This week you're a naturalist. Your job is to catch the assumption in the act, not to beat it."
Watch for: turning observation into self-criticism. It is working when the person can predict the situations that trigger the assumption.

4.12 Biography of the big assumption

Purpose / when: Middle Game; loosens the assumption's status as plain truth by locating it in personal history. [source: ch. 10]
Steps: For each big assumption ask: (1) When did it begin, or when do I first remember living by it? (2) What is its history; where has it been reinforced? (3) Does it still hold in my current life and roles? [source: ch. 10]
Example language (illustrative): "When did this rule first make sense? What has changed since then?"
Watch for: this step can pull a person into painful early memories (family conflict, bullying, abuse). In coaching, keep the focus on how the assumption functions now. If traumatic material emerges, contain it and consider referral (5.3, 5.8) [synthesis]. Inside therapy, a clinician may go further within their own model, but ITC itself does not supply trauma-processing procedures [synthesis].

4.13 Designing a test of the big assumption

Purpose / when: Middle Game; plans a deliberate action that runs against the assumption, to collect data. [source: ch. 10]
Steps: (1) Choose a powerful assumption that is testable, not catastrophic. If its feared consequence is extreme (e.g. being fired), trace back to an earlier, smaller link in the chain and test that. (2) Write it as "I assume that if… then…". (3) Choose a specific behaviour that runs against it, including how to act, not just what to do: a test of spontaneity cannot be over-prepared, as with Claus. (4) Choose a safe situation where a poor outcome would not be costly. (5) Decide in advance what data (others' reactions and one's own feelings) would count for or against the assumption. If nothing could count against it, redesign. (6) Fix when and where. [source: ch. 10]
Example (authors' case, summarised): Sue tested the assumption that saying no would make others see her as cold and uncaring by declining specific requests in chosen settings and tracking her feelings and others' responses. [source: ch. 10]
Example language (illustrative): "What will you do, with whom, and by when? What would you have to see or hear to doubt the assumption a little?"
Watch for: the authors warn against an "event-focused" stance, where doing the action counts as success. The aim is information, not improvement. Tests must fit the person; a test that is too big triggers the immunity and gets avoided.

4.14 Running and interpreting the test

Purpose / when: after each test; converts experience into a revised assumption. [source: ch. 10]
Steps: (1) Run the test as designed. (2) Record what happened: others' responses and one's own feelings. (3) Link the results back to the assumption: what do they say about how certain, or how general, it is? (4) Design the next, larger test. The authors describe a series that starts small and grows in scope. [source: ch. 10] The excerpt breaks off before the authors' full interpretation guidance, so rules for a test that seems to confirm the assumption were not read [uncertain]. A sensible approach is to ask what conditions the assumption is true under, rather than whether it is true [synthesis].
Example language (illustrative): "What happened, and what does that tell you about the assumption? Is it always true, sometimes, or only with certain people?"
Watch for: discounting disconfirming data ("they were just being polite"). Ask what evidence would convince them, and build that into the next test. It is working when the assumption gets qualified ("sometimes, with some people") rather than held as absolute.

4.15 Follow-up survey

Purpose / when: End Game; measures change from others' point of view. [source: ch. 10]
Steps: (1) Go back to the same people as the initial survey. (2) Ask for their current view on the goal. (3) Compare with the baseline and with the person's own self-assessment, and discuss any gaps. [source: ch. 10]
Example language (illustrative): "Where do others see more change than you do, and where less?"
Watch for: large gaps between self and others' ratings; explore them, don't argue them.

4.16 Hooks and releases

Purpose / when: End Game; relapse prevention. [source: ch. 10]
Steps: (1) Identify hooks: situations, people or states that pull the old assumption back into force. (2) Identify releases: what helps the person notice and recover when they slip. (3) Write both down as a short personal plan. [source: ch. 10]
Example language (illustrative): "When are you most likely to slide back? And when you do, what gets you out fastest?"
Watch for: expecting zero slips. The authors frame slippage as expected and recoverable.

4.17 Future progress and re-engaging the process

Purpose / when: at close; continues the work on goals not yet met. [source: ch. 10]
Steps: (1) Review which parts of the goal are achieved. (2) Re-run the process (new map or refined assumptions) for what remains, alone, with a partner or with a coach. [source: ch. 10]
Example language (illustrative): "What's the next assumption in the chain that you'd want to test?"

4.18 Group and team applications

Purpose / when: leadership teams, school leaders, colleague groups. [source: ch. 9; Minds at Work programmes page; Hayward review]
Steps (as described): (1) Prework: each member gathers one-to-one feedback on a single improvement goal. (2) Members share their goals and challenge each other, and consult coaches and family. (3) Each builds their own map, privately or with affected others present; a review by Hayward notes columns 2 and 3 can be more powerful when colleagues or family take part. (4) Pairs or a coach support the testing phase. [source: ch. 9; Hayward 2009 review] A collective map for a team's shared goal exists in the authors' work ("ITC with Teams"; a "Team Tip" in the book according to a 2009 review), but its procedure was not read [uncertain].
Watch for: psychological safety. Disclosing column 3 fears in front of a boss can be risky, so let members choose what they share [synthesis].

5. Handling problems

The ITC sources say little about problem handling beyond the map and test criteria. Most of what follows is marked [synthesis], built on the authors' principles.

5.1 Stuck at mapping: flat or intellectual map

Signs: column 3 lists virtues; the person agrees with the map but feels nothing.
Response: go back to the worry box and the opposite-behaviour image. Ask for the feeling underneath the first one, and treat boredom or impatience as covers [source: ch. 9]. Check column 1 for gut-level importance [source: ch. 8]. If the goal was imposed by an employer, rework it until the person owns it [synthesis].

5.2 "Resistance" and non-adherence (tests not run, observations not done)

Frame: treat skipped tasks as the immunity doing its job, and as data, not defiance [synthesis, consistent with the authors' frame of the immunity as self-protection, ch. 8–9].
Response: (1) Ask what the person was protecting by not doing it; this often reveals a further commitment or assumption [synthesis]. (2) Shrink the test: a smaller link in the chain, a safer setting [source: ch. 10]. (3) Check the test is about information, not performance [source: ch. 10]. (4) Reconfirm the 30-minutes-a-week structure and the partner or coach support [source: ch. 10].

5.3 Overwhelm during the worry box or biography work

Signs: rising panic, tearfulness that escalates rather than settles, shutting down, spacing out, intrusive images of past events [synthesis].
Response [synthesis]: stop the exploration. Slow down and orient to the present: feet on the floor, name what is in the room, slow breathing. Name what happened without probing it. Move the session to a contained task, such as re-reading column 1, and agree a gentler next step. Afterwards, ask whether this material belongs in coaching at all. If the person is a coaching client, offer referral (5.8). If you are their therapist, decide whether to switch to your clinical model and say so explicitly (5.9).

5.4 Disconfirming data discounted

Response: before the next test, agree what result would count. The authors require the data to be specified in advance [source: ch. 10]. Reframe the question from "is it true?" to "when and with whom is it true?" [synthesis].

5.5 Test "confirms" the assumption

Response [uncertain: the authors' guidance on this was not in the excerpt; this is synthesis]: (1) Check the test was well designed: was the behaviour carried out as planned, or done anxiously and over-prepared? (2) Check whether the setting was representative. (3) Treat a real confirmation as useful: the assumption is accurate in that context. Map its limits and choose a different context or a different assumption.

5.6 Ruptures with the coach, partner or team

Response [synthesis]: name the strain directly. Check whether the coach has become the "other person" in the client's assumption (for example, the coach as judge). Invite the client to test the assumption with the coach in the room. Adjust pace. In teams, never require disclosure of column 3 or 4 content.

5.7 Assumptions that are realistic, or tests that could cause harm (ITC-specific)

Problem: some "big assumptions" describe real danger, such as "if I say no, my partner will become violent" or "if I report this, I will be retaliated against". Discrimination, harassment and insecure employment can make a feared outcome realistic [synthesis].
Response: the authors require safe tests in situations where a poor outcome would not be costly, and catastrophic assumptions to be broken into smaller links [source: ch. 10]. Go further: never design a test that exposes the person to abuse, violence, serious financial or legal harm, or loss of essential care. Switch to safety planning and appropriate referral (domestic-abuse services, legal or HR advice) [synthesis].

5.8 The coaching–psychotherapy boundary and referral signs

What the coaching bodies publish

ICF: the ICF defines coaching as partnering with clients in a thought-provoking and creative process. Its Code of Ethics (Board approved October 2024, revised May 2026) does not define therapy. It does require coaches to: explain the nature of coaching and co-create an agreement before starting (1.1); agree in advance on confidentiality exceptions, including imminent or likely risk of danger to self or others and legal requirement (2.3); disclose when acting in another role such as therapist (3.7); stay alert to shifts in the value received and explore a different coach, professional or resource (4.3); and work within their competence (5.1). [source: ICF Code of Ethics 2025]
ICF referral guidance (Hullinger & DiGirolamo, 2018) states that psychotherapy is outside the coaching scope of work and treats referral as an ethical responsibility. [source: ICF client-referral overview]
ICF case study on Standard 3.7: a coach who is also a former therapist names the role switch, explains the difference between coaching and therapy, asks permission, refers the client, then returns to the coaching agenda. [source: ICF coaching-vs-therapy case study]
EMCC / AC Global Code of Ethics (V4, updated June 2021 / December 2025): coaching "is not counselling, psychotherapy or medical intervention". Members work within their professional competence and refer where appropriate to a more experienced or suitably qualified practitioner. They encourage ending the work when the client would be better served by a different form of professional help. They state where confidentiality cannot be kept (illegal activity, danger to self or others, legal requirement). They discuss ethical dilemmas in supervision. [source: Global Code of Ethics PDF] The PDF did not list its signatories. The Association for Coaching's co-signing is shown only by a search-result headline from Coaching at Work [uncertain].
ITC authors: none of the Kegan & Lahey or Minds at Work pages contains an explicit statement on the therapy boundary or on referral [uncertain]. Their programmes are presented as coaching and personal development, and certification requires prior coaching experience. [source: Minds at Work programmes]

Concrete signs a coachee needs referral to clinical care

The ICF referral guidance gives the following. [source: ICF client-referral overview]

The issue is outside the coach's competence; it interferes with daily functioning; it blocks progress in coaching; or it is psychological and involves deep-seated emotions.
Marked mood changes (irritability, anger, anxiety, sadness).
Declining performance at work or school.
Withdrawal from relationships and activities.
Changes in weight or appearance, including neglected hygiene.
Sleep disturbance.
Expressions of hopelessness or suicidal thoughts.
Presentations that commonly need referral: anxiety, depression, eating disorders, posttraumatic stress, substance misuse, suicidal ideation and thought disorders.
Consider duration, frequency and intensity, and consult a supervisor or mental health professional when unsure.

ITC-specific signs to add [synthesis]:

The worry box repeatedly produces panic, dissociation or flashbacks.
The biography of the assumption leads into abuse or trauma history.
Big assumptions are global and self-condemning ("I am worthless") rather than situational.
The goal is itself a symptom-driven target, for example weight or eating control in someone showing disordered eating.
Tests are avoided because of intense fear, beyond ordinary reluctance.
Alcohol or drug use is used to manage the anxiety the immunity protects against.
Unusual beliefs or perceptions suggest a thought disorder.

How to refer (ICF guidance): be direct and calm; express concern and care; point to specific behaviours; listen closely; encourage help-seeking; normalise therapy; make the referral. In an emergency with imminent risk to self or others, contact local emergency services and stay with the person until help arrives; use crisis lines for suicidal thoughts. [source: ICF client-referral overview] Follow current standard suicide risk-assessment practice; any lethal-means discussion is limited to safe storage and removal [synthesis].

5.9 For psychotherapists using ITC inside therapy

[synthesis] Name the frame. Say whether you are using ITC as a coaching exercise or as part of treatment, mirroring the ICF role-disclosure principle (3.7).
[synthesis] Screen first. Use ITC once acute symptoms, risk and active trauma symptoms are stabilised. The map's fears and assumptions can feed your existing formulation, and big-assumption tests resemble behavioural experiments, but ITC adds no clinical safety procedures of its own.
[synthesis] Treat the biography of the assumption as the client's own account of how a belief developed, not as historical fact to be uncovered.
[synthesis] Do not use ITC as a stand-alone treatment for a diagnosed disorder. It is published as a development method, not a treatment.

Sources

Sources

Immunity to Change, Part Three, chapters 8–10 (HarvardX GSE1x course excerpt, PDF): the primary procedural source. The excerpt ends partway through chapter 10.
Minds at Work: Immunity to Change
Minds at Work: Our Thinking
Minds at Work: Online Coach Development / programmes
Harvard Extension School: The surprising reason we don't keep our resolutions (secondary)
Morris, R. (2009). Review of Immunity to Change, The Systems Thinker 20(9) (secondary)
Hayward, P. Review of Immunity to Change (Swinburne Research Bank) (secondary)
Kegan & Lahey (2001). The Real Reason People Won't Change, HBR: only the introduction is drawn on; not used for procedure.
ICF Code of Ethics (Board approved Oct 2024, revised May 2026)
ICF: When and how to refer a client to therapy (research overview)
ICF ethics case study: coaching vs therapy boundaries (Standard 3.7)
Global Code of Ethics for coaches, mentors and supervisors, V4 (EMCC France copy, Dec 2025)

Further reading (not drawn on) [uncertain]: the Minds at Work FAQ and Change Course pages; Coaching at Work news item on EMCC and AC launching a single Global Code of Ethics (headline only).

Books and papers cited

Kegan, R., & Lahey, L. L. (2009). Immunity to Change: How to Overcome It and Unlock the Potential in Yourself and Your Organization. Harvard Business Press.
Kegan, R., & Lahey, L. L. (2001). The real reason people won't change. Harvard Business Review, 79(10).
Hullinger, A. M., & DiGirolamo, J. A. (2018). Referring a Client to Therapy: A Set of Guidelines. International Coaching Federation.
Helsing, D., Howell, A., Kegan, R., & Lahey, L. (2008). Putting the "development" in professional development: Understanding and overturning educational leaders' immunities to change. Harvard Educational Review, 78(3), 437–465.

← Previous · VI-149Appreciative inquiry and appreciative coaching Next · VI-151 →Integral and systemic coaching