MRI brief therapy

1. Introduction

Version followed: the Brief Therapy Center (BTC) model of the Mental Research Institute (MRI), Palo Alto, as set out in Fisch, Weakland & Segal (1982), The Tactics of Change: Doing Therapy Briefly, and operationalised in Rohrbaugh & Shoham's (2001) eight-step summary. The founding paper (Weakland, Fisch, Watzlawick & Bodin, 1974, Family Process) and the theory book Change (Watzlawick, Weakland & Fisch, 1974) are not drawn on. Later variants not followed here: Fisch & Schlanger (1999) on "intimidating" cases, Rohrbaugh and Shoham's family-consultation adaptations for health behaviours, and Nardone's Arezzo Brief Strategic Therapy. [source: Rohrbaugh & Shoham 2001; Springer encyclopedia entry; Fisch et al. 1982 foreword via Russian edition preview; Wiley table of contents]

Richard Fisch, John Weakland and Paul Watzlawick founded the BTC at MRI in 1966 to study rapid problem resolution. The model is also called the Palo Alto model, brief problem-focused therapy or brief strategic therapy. Its roots are Bateson's communication project (1952–62) and Milton Erickson's work; Tactics is dedicated to Erickson. Arthur Bodin co-wrote the 1974 paper and Lynn Segal co-wrote Tactics. [source: Springer entry; Fisch et al. 1982 preview]

The core idea of change is that a problem persists because of what people currently do to solve it. A difficulty is met with an attempted solution that keeps or worsens it, which prompts "more of the same" solution, and so on in a self-reinforcing loop. The origin and duration of the problem do not matter. Interrupting the loop, usually by getting the most concerned person to do "less of the same" or the reverse, can resolve the complaint without insight or any search for causes. The model is non-normative: if nobody complains, there is no problem. Treatment lasts at most ten sessions and aims for the smallest change that resolves the complaint. The authors frame therapy as a craft and compare strategy to chess. [source: Rohrbaugh & Shoham 2001; Carratalà et al. 2016; Fisch et al. 1982 foreword via preview]

It is used with individuals, couples and families, often seeing only one motivated member. It is listed under family therapy schools because its theory is interactional, not because whole families must attend. [source: Rohrbaugh & Shoham 2001]

2. Framework

Format. BTC cases were unselected and capped at 10 sessions. Sessions ran weekly, and one review describes them as one-hour sessions. Over 24 years the mean was about six sessions. A trial of a five-session cap gave worse results than the ten-session cap and was dropped. The team met weekly: one member worked as primary therapist while colleagues consulted from behind a one-way mirror. Couple partners are often seen separately. The client is told about the session limit at the start, and unused sessions can be "put on hold" (Wienands, observing Fisch in 2000). The time limit is meant to signal that change is expected. [source: Rohrbaugh & Shoham 2001; Crippledini review; Wienands 2002]

The phases follow the chapter order of Tactics (chs 2–8) and Rohrbaugh & Shoham's steps:

Phase	Goals	Typical timing	Move on when
0. Setting the stage (Tactics ch. 3)	Handle the first phone contact strategically: who calls, who is most concerned (the "customer"), who to invite, explaining the format. Keep the therapist's room to manoeuvre from the start.	First phone call, before session 1	You have an appointment with the person or people most concerned about the complaint.
1. Initial interview (Tactics chs 4–5; steps a–c, f)	Define the complaint in behavioural terms, set a minimum goal, map the attempted solutions, and learn the client's position (values, language, theory of the problem).	Usually session 1, sometimes into session 2	You can state who does what, to whom, and how it is a problem; what has been tried; a concrete small sign of improvement; and what the client values. [synthesis: the criteria are inferred from the steps]
2. Case planning (Tactics ch. 6; steps d–e)	Done by the therapist and team between sessions, not with the client. Formulate the loop, define the strategic objective (usually a 180° reversal of the main attempted solution), choose the situations to target, frame it in the client's position, and list the "mine field" positions to avoid.	Between sessions 1 and 2, then reviewed each week	A strategic objective has been written down and a frame chosen that this client can accept.
3. Intervention (Tactics ch. 7; step g)	Interdict the attempted solution through suggestions, tasks and reframes, using general interventions (go slow, dangers of improvement, U-turn) as needed. Review each task's result and adjust.	Sessions 2 to about 6	Some change on the minimum goal, or a task that clearly works. If not, re-examine the customer, position and objective.
4. Consolidation and termination (Tactics ch. 8; step h)	Nurture early change with restraint, reframe gains in the client's terms, predict or discuss relapse, and end without celebration. Leave unused "sessions in the bank." If new problems come up, suggest a break first.	The last 1–2 sessions, often spaced out [uncertain: spacing not confirmed]	The complaint is resolved to the client's satisfaction. The criterion is the complaint, not deeper change.
5. Follow-up	A team member other than the therapist phones at about 3 and 12 months. They ask about the presenting complaint, any new problems and any help sought elsewhere, and rate the case as success, significant improvement or failure.	3 and 12 months	(Research and service evaluation)

[source: Rohrbaugh & Shoham 2001; Wiley table of contents; Wienands 2002; Carratalà et al. 2016]

3. Guiding principles

Non-normative stance. No complaint means no problem. Do not treat problems you infer but the client does not raise. [source: Rohrbaugh & Shoham 2001]
The problem is the complaint. Work only on what the client brings, and stop when that improves. [source: Wienands 2002]
Problems are maintained in the present. Spend session time on current sequences, not history or causes. [source: Rohrbaugh & Shoham 2001; Carratalà et al. 2016]
The attempted solution is the target. Ask what everyone has tried, and look for the "more of the same" pattern. [source: Rohrbaugh & Shoham 2001]
Less of the same, ideally a 180° reversal. Plan interventions that move people in the opposite direction from their main solution. [source: Rohrbaugh & Shoham 2001]
Interdict rather than forbid. Give a task that cannot be done alongside the old solution, rather than just asking the client to stop. [source: Rohrbaugh & Shoham 2001]
Minimal change. Aim for the smallest concrete change that shows improvement, because change comes step by step. That a small change spreads further is a commonly stated MRI premise. [source: Rohrbaugh & Shoham 2001; Wienands 2002; uncertain: the spreading premise comes from the 1974 paper]
Work with the customer. See the person most bothered by the complaint, even if someone else's behaviour is at issue. Reluctant family members need not attend. [source: Rohrbaugh & Shoham 2001; Wienands 2002]
Therapist maneuverability. Keep your options open, avoid early commitments, and keep control of the direction of treatment. Fisch et al. say the therapist must "maintain his own options while limiting those of the patients." [source: Fisch et al. 1982, quoted in Rohrbaugh & Shoham 2001]
Patient position. Speak the client's language: accept their view of self, problem and others, and extend it to support the task. Wienands quotes Watzlawick: "You have to talk the language of your client." [source: Rohrbaugh & Shoham 2001; Wienands 2002]
Non-confrontational, non-expert stance. Avoid arguing or pushing. Take an equal or one-down position, and do not become the referee between partners. [source: Carratalà et al. 2016; Rohrbaugh & Shoham 2001; Crippledini review]
Restraint. Counsel going slowly, which eases the urgency that drives more-of-the-same. [source: Rohrbaugh & Shoham 2001; Wienands 2002]
Don't step on the mine field. Check that you are not doing the same thing as the client's failed solution, for example pushing a client who resists being pushed. [source: Rohrbaugh & Shoham 2001]
Specificity. Ask for behavioural detail every time; a general description does not show you the loop. [source: Rohrbaugh & Shoham 2001; Crippledini review]
Theory is a tool. Keep the theory simple and close to practice, and do not copy an intervention without understanding why it worked. [source: Fisch et al. 1982 ch. 1 via preview]
Ending is set by the client's complaint. End quietly once the complaint has improved, and keep unused sessions available. [source: Rohrbaugh & Shoham 2001; Wienands 2002]
Transparency and consent. The original model often did not explain its rationale to clients, and critics have called this manipulative. Today, get general consent to an "experimental tasks" approach, and never say anything false about risk, diagnosis or facts. [synthesis; source for the criticism: Rohrbaugh & Shoham 2001; Crippledini review]

Weakland's three rules, as Wienands gives them, are: if it isn't broken, don't fix it; if it doesn't work, do something different; if it works, do more of it. [uncertain: these maxims are usually credited to de Shazer and Berg's solution-focused therapy; Wienands credits Weakland]

4. Intervention inventory

The sources give the eight-step procedure and name the general interventions; chapter 7 of Tactics itself is not drawn on. Steps for items that are only named in the sources are marked. All example wording is illustrative unless it is quoted and cited.

4.1 Setting the stage: the first telephone contact

Purpose / when: before session 1, to start treatment with the right person, with the format explained and without commitments that limit the therapist. [source: Fisch et al. 1982 foreword (ch. 3 synopsis) via preview; Rohrbaugh & Shoham 2001 note that the BTC used control tactics from the first phone contact]
Steps: (1) Find out who is calling and who is most bothered. (2) Get a one-line statement of the complaint, without interpreting it. (3) Invite the person or people most concerned. Do not insist that a reluctant identified patient comes. (4) Explain the format: up to 10 sessions, possibly fewer, and an observing team if you use one. (5) Do not promise outcomes or agree to the caller's terms on who must attend or what you will do. Say you will decide on the basis of the first meeting. [synthesis for steps 1, 2 and 5; uncertain: the detail of ch. 3]
Example language (illustrative): "It may help most if I start with whoever is most concerned about this, which sounds like you. We can see later whether anyone else should come."
Watch for: a caller who sets conditions ("you must see my son and fix him"). Agreeing gives away maneuverability. Instead, offer to meet the concerned caller first. [synthesis]

4.2 Explaining the brief format

Purpose / when: at the start of session 1, to set the expectation that change will be quick and to keep a reserve of sessions. [source: Wienands 2002; Rohrbaugh & Shoham 2001]
Steps: (1) Say that there are at most ten sessions and there may be fewer. (2) Say that if things improve early, the remaining sessions can be kept for later. (3) If you use a team, mention it and get consent to observation or recording. (4) Move straight to "What is the problem?" [source: Wienands 2002 (Fisch's practice); synthesis for step 3]
Example language (illustrative): "We work in a limited way, up to ten meetings. Often it takes fewer, and any you don't use stay available."
Watch for: relief, or a sharper focus, in response to the limit. If a client protests that this is too little, use go-slow framing (4.15), not reassurance. [synthesis]

4.3 Identifying the customer and deciding whom to see

Purpose / when: from the first contact onward, and again whenever treatment stalls.
Steps: (1) Ask who sees the behaviour as a problem and who is most distressed or most motivated. (2) Treat that person as the customer, even if the problem behaviour belongs to someone else (a parent about a child, a spouse about drinking, relatives about a patient who refuses treatment). (3) Do not require the identified patient or a reluctant spouse to attend. (4) Consult others who are also concerned, such as teachers, grandparents or probation officers. (5) In couple work, consider separate sessions, so that you can side with each partner's view and avoid becoming the referee. [source: Rohrbaugh & Shoham 2001; Wienands 2002; Crippledini review describes the therapist commiserating with each partner seen separately]
Example language (illustrative): "Since you're the one most worried about this, let's start with what you can do, whether or not he ever comes in."
Watch for: a "window shopper" who is there at someone else's insistence and has no complaint of their own. Weakland treated the parents of a young man diagnosed with schizophrenia, who refused all services. There were two sessions, the son never attended, and at follow-up the parents rated the problem as resolved. [source: Carratalà et al. 2016; the window-shopper term is from Fisch et al. 1982 and is not confirmed, so it is uncertain]

4.4 Defining the complaint behaviourally

Purpose / when: session 1. Without a concrete complaint you cannot map the loop or measure change.
Steps: (1) Ask what the problem is. (2) Ask for a recent example and get the sequence: who said or did what, then what happened next. (3) Ask where, when, how often and how long. (4) Use the video question. (5) Ask who else sees this as a problem and why. (6) Ask "why now?" (what led them to seek help at this point). (7) Ask what distinguishes the problem from normal behaviour, which turns a self-diagnosis into an observable pattern. Tactics frames the target as who is doing what, to whom, and how that makes it a problem. [source: Rohrbaugh & Shoham 2001; Wienands 2002 (Fisch asking what makes a "compulsive" look different from a normal one); Crippledini review paraphrasing Tactics]
Example language: "If we had a video of this, what would I see?" (Rohrbaugh & Shoham 2001). Illustrative: "Take me through last Tuesday's argument from the first thing said."
Watch for: abstractions ("we don't communicate", "he's depressed") or diagnostic labels. Keep asking for specifics, politely and without accepting the label as a description. [source: Rohrbaugh & Shoham 2001]

4.5 Setting a minimum goal

Purpose / when: session 1, after the complaint has been defined. It gives a benchmark for change and conveys that change is expected.
Steps: (1) Ask what would be the first small sign that things are improving, in terms of the complaint itself. (2) Push for presence of behaviour, not absence ("doing more of X" rather than "no longer Y"). (3) Keep the goal small and visible. (4) Write it down. This is also the termination criterion. [source: Rohrbaugh & Shoham 2001; Wienands 2002]
Example language: "What will you… be doing differently…?" (fragment quoted in Rohrbaugh & Shoham 2001). Illustrative: "What's the smallest thing that would tell you this is starting to get better?"
Watch for: utopian goals, such as never feeling anxious again. Scale the goal down to something they could notice within a week or two. [synthesis]

4.6 Inquiring into attempted solutions

Purpose / when: session 1, which is the core assessment. Find the more-of-the-same that keeps the problem going.
Steps: (1) Ask what the client has done to try to solve or control the problem. (2) Ask what others (partner, parents, doctors) have done. (3) Get specifics of what was said and done, and how the other person responded. (4) Look for a common theme across different-looking attempts; they often share one direction, such as reassuring, pushing or avoiding. (5) Ask about solutions that worked and were dropped, which can point to what to restore. (6) Treat deliberate inaction or avoidance as a solution too. (7) Do not point out the loop to the client or blame them at this stage. [source: Rohrbaugh & Shoham 2001; Wienands 2002]
Example language (illustrative): "What have you tried so far to get him to cut down? What do you say, and how do you say it?" … "Has anything ever helped, even for a while?"
Watch for: the case in Rohrbaugh & Shoham. A wife had dropped a firm stance on her husband's drinking because she thought it was controlling, yet questioning showed the firm stance had worked. It was relabelled as caring and she went back to it. [source: Rohrbaugh & Shoham 2001]

4.7 Assessing patient position

Purpose / when: throughout session 1 and after, so that any task is framed in terms the client will accept.
Steps: (1) Listen for how the client sees themselves and wants to be seen, for example flexible or principled, strong, protective, responsible. (2) Note what they value. (3) Note their theory of the problem, for example inner turmoil versus wilful disobedience, and why they think others behave as they do. (4) Note whether they want direction or prefer to find their own way. (5) Note their exact words and phrases so you can reuse them. Do not challenge the position; plan how to extend it. [source: Rohrbaugh & Shoham 2001; Wienands 2002]
Example language (illustrative): "You've clearly put a lot of care into protecting her. That matters to you."
Watch for: strong positions, which are the best levers. A client who insists the therapist doesn't understand may accept a task framed as something that only they can do. [synthesis]

4.8 Case planning: loop, strategic objective and mine field

Purpose / when: between sessions, with the team if possible. This is internal work and is not shared with the client.
Steps: (1) Write the loop: complaint, then attempted solution, then more of the complaint, then more of the solution. (2) Name the common direction of the solutions. (3) State the strategic objective, which is the 180° opposite, for example from persuading to an apologetic one-down stance, from avoiding to bringing it on, or from encouraging to accepting gloom. (4) Pick the recurring situation in which to apply it, such as bedtime power struggles. (5) Choose a frame from the client's position (4.10). (6) Write the mine field: what you must not do because it repeats the loop, for example pushing abstinence on a client who flees pressure, or warmly befriending a child in a way that undermines the parent. [source: Rohrbaugh & Shoham 2001]
Example language: none; this is internal.
Watch for: a strategic objective that is just "stop doing X." Turn it into an incompatible positive action. [source: Rohrbaugh & Shoham 2001]

4.9 Maneuverability tactics

Purpose / when: every session, to keep control of treatment without confronting the client.
Steps (as named in the sources): (1) Avoid firm positions and early commitments. (2) Take a one-down, non-expert stance. (3) See parties separately when that keeps options open. (4) Defer decisions ("let me think about that before next time"). (5) Use tentative, qualified phrasing that can be revised later. [source: Rohrbaugh & Shoham 2001; Crippledini review; uncertain for 4 and 5: these are likely contents of Tactics ch. 2]
Example language (illustrative): "I'm not sure yet. It may be too early to say what would help, so let me hold off on a suggestion."
Watch for: being drawn into promises, diagnoses or alliances ("tell her she's wrong"). Step back to curiosity or a one-down stance. [synthesis]

4.10 Reframing: accepting and extending the client's view

Purpose / when: to make a 180° task acceptable, and to relabel problem-keeping or problem-solving behaviour in the client's own values.
Steps: (1) State the client's view back without disputing it. (2) Add one extension that opens a new action, for example: "he's uncommunicative" becomes "his defensiveness may show how vulnerable he feels." (3) Link the new action to what the client values, for example firmness as caring and protective. (4) Use the client's words. [source: Rohrbaugh & Shoham 2001; Carratalà et al. 2016]
Example language (illustrative): "You're right that he shuts down. I wonder if the harder you press, the more exposed he feels. Someone that defensive may be more fragile than he looks."
Watch for: the client spontaneously using the new frame, or acting differently in the target situation. If they argue with it, you have stepped outside their position, so return to it. [synthesis]

4.11 Interdicting the attempted solution: specific interventions by type

The main intervention gives a task or suggestion that cannot be done alongside the old solution. [source: Rohrbaugh & Shoham 2001] Rohrbaugh & Shoham name three solution types from Tactics; the first two below come from Tactics ch. 6 and are not confirmed. [uncertain]

a. Forcing something that can only happen spontaneously (sleep, erection, arousal, memory, spontaneity, "not thinking about it"). Reverse: stop the deliberate effort or prescribe the opposite, for example staying awake in bed, or scheduled times to entertain intrusive thoughts on purpose. [source: Rohrbaugh & Shoham 2001 for the examples]
b. Mastering a feared event by postponing or avoiding it (anxiety, performance fears). Reverse: approach the feared situation with permission to fail, or even to be anxious. Rohrbaugh & Shoham report that socially anxious men told to "be anxious" with a positive rationale improved more. Safety (exposure-type task): screen first for medical conditions, panic with cardiac concerns, trauma, dissociation and substance use. Explain the approach and get consent. Agree an exit. Start with situations the client chooses, and never assign approach to genuinely dangerous situations. [source: Rohrbaugh & Shoham 2001; synthesis for the safety steps]
c. Reaching accord through opposition (arguing, persuading or lecturing to win agreement). Reverse: an apologetic, one-down stance, or acknowledging that the other person may be right. [source: Rohrbaugh & Shoham 2001]
d. Gaining compliance through voluntarism (wanting the other person to do it because they want to, so the request is never made directly). Reverse: make a clear, direct request, even an arbitrary one, without needing the other to want it. [source: Rohrbaugh & Shoham 2001 name the type; uncertain: the reverse described is from my recollection of Tactics]
e. Confirming an accuser's suspicions by defending oneself (accusation, then denial, then more suspicion). Reverse: the accused stops defending, for example by agreeing or not engaging, or both partners are helped to break the accuse-defend sequence. Tactics calls this "jamming." [source: Rohrbaugh & Shoham 2001 name the type; uncertain: the term and procedure are not verified]
f. Interpersonal variants: pushing for someone else's change becomes declaring helplessness, accepting, or simply observing. Cheering up a depressed partner becomes less encouragement. Exhorting or overprotecting a stroke patient becomes less direct support, as in the Tactics case study of the stroke victim's family. [source: Rohrbaugh & Shoham 2001; Wiley table of contents]
Steps (all types): (1) Name the target situation. (2) Give the new action as a positive task, not a ban. (3) Frame it in the client's position (4.10). (4) Rehearse briefly what they will do and say. (5) Ask them to notice the effects and report back. [source: Rohrbaugh & Shoham 2001; synthesis for step 4]
Example language (illustrative, type c): "This week, when the subject comes up, try something unusual for a strong person like you. Say, 'You may be right. I hadn't thought of it that way,' and leave it there."
Watch for: the other person changing in response, the complaint easing, or the client reporting that it felt "odd but different." A partial try is still useful information. [synthesis]

4.12 Observation task (diagnostic intervention)

Purpose / when: when a pattern is hard to interrupt directly, or the client isn't ready to change. It gives a wedge in the usual solution, because watching takes the place of acting.
Steps: (1) Present the client's previous behaviour as understandable. (2) Frame observation as an active, important task, not as doing nothing. (3) Ask them to watch what happens in the target area without reminding, discussing or intervening. (4) Ask them to keep brief notes. (5) Review the notes next session. [source: Rohrbaugh & Shoham 2001]
Example language: "Let's see what he does this week on his own…" and "…at least then we'll know what we're up against" (both quoted in Rohrbaugh & Shoham 2001). Illustrative: "Your reminders made sense, but we need to know what he does without them. Can you watch and jot it down?"
Watch for: the other person changing once the pressure lifts. That often happens, and it shows the old solution was keeping the problem going. [source: Rohrbaugh & Shoham 2001]

4.13 Symptom prescription

Purpose / when: for self-referential problems that get worse with effort, such as insomnia, anxiety, intrusive thoughts and sexual performance. Also useful for reactant clients.
Steps: (1) Check that the attempted solution really is a struggle against the symptom. (2) Give a positive rationale that fits the client's position, for example "to learn exactly how it works" or "to get better control". (3) Prescribe deliberately producing or allowing the symptom at a set time or in a set situation, for example trying to stay awake, or a 15-minute worry slot. (4) Ask for observation and a report. Rohrbaugh & Shoham stress that the point is to interrupt the loop, and that non-paradoxical routes to that are equally valid. [source: Rohrbaugh & Shoham 2001]
Example language (illustrative): "Tonight, rather than trying to sleep, I'd like you to lie there and try to stay awake as long as you comfortably can, and notice what your mind does."
Watch for: relief, or the symptom fading when the effort stops. Contraindications: never prescribe self-harm, suicidal ideation acted on, violence, substance use, purging or restriction, risky driving, or any dangerous symptom. [synthesis]

4.14 Behavioural tasks and homework review

Purpose / when: at the end of each intervention session. In Weakland's recorded case, a task was given at the end of each session. [source: Carratalà et al. 2016]
Steps: (1) Give one task, tied to the strategic objective. (2) Specify when, where and how. (3) Ask for a report. (4) Next session, start with what happened before anything else. (5) If it worked, do more of it. If it was partly done, ask what got in the way and keep the direction. If it backfired, revise the frame or the target. [source: Carratalà et al. 2016; Wienands 2002 (Fisch's case: the client told his wife about each incident instead of hiding it, and they agreed how often to discuss it); synthesis for step 5]
Example language (illustrative): "Before we go further, how did the experiment go?"
Watch for: non-compliance. Never push, which is the mine field; see section 5. [source: Rohrbaugh & Shoham 2001]

4.15 Go slow (restraint)

Purpose / when: (a) to prepare for change, (b) to accept reluctance, and (c) to consolidate early change. It eases urgency and increases cooperation. [source: Rohrbaugh & Shoham 2001; Wienands 2002]
Steps: (1) When the client pushes for quick results, or reports early gains, advise caution. (2) Give a plausible reason in the client's frame, for example that change needs time to settle, or that rushing has not helped so far. (3) Suggest a small step, or no change yet. [source: Rohrbaugh & Shoham 2001 name the uses; synthesis for the steps]
Example language (illustrative): "I'm glad, and also a bit cautious. Things that change this fast can be unsettling, so let's not hurry the next step."
Watch for: the client arguing for faster progress, which shows ownership. A client who relaxes and then improves anyway shows the urgency was part of the loop. [synthesis]

4.16 Dangers of improvement

Purpose / when: with clients who are reluctant or ambivalent about change. In the case opening Tactics, the therapist tells a depressed client who keeps entering short, unsatisfying relationships that improving too quickly could be harmful, before she knows what she needs. [source: Rohrbaugh & Shoham 2001; Fisch et al. 1982 ch. 1 via preview]
Steps: (1) Ask the client to consider what problems improvement might bring, for themselves or others. (2) Explore these seriously. (3) Let any wish to change come from the client. [synthesis; uncertain: the procedural detail from Tactics ch. 7]
Example language (illustrative): "Before we try to fix this, I'd like to understand what might get harder for you if it went away."
Watch for: the client arguing for change, which often follows. Do not use this with clients at risk, for whom framing improvement as dangerous could feed hopelessness. [synthesis]

4.17 Prescribing a delay or making the problem worse

Purpose / when: when a client rushes into the pattern that keeps the problem going. In the Tactics opening case, the therapist advised the woman to look less attractive if she went out, so she would not rush into another relationship. Two weeks later she reported feeling better. [source: Fisch et al. 1982 ch. 1 via preview]
Steps: (1) Identify what the client does to make things worse, which is usually the attempted solution itself. (2) Either prescribe a deliberate brake on it, or ask the client what they would do to make the problem worse, then note that this is what they are already doing. [uncertain: the second variant, "how to make the problem worse", is listed in Tactics ch. 7 by my recollection]
Example language (illustrative): "If you wanted to guarantee another week like the last one, what would you do?"
Watch for: recognition, or a smile. Do not prescribe worsening of anything risky. [synthesis]

4.18 The U-turn

Purpose / when: when a direction is failing, or you have stepped on the mine field and need to reverse your own advice without losing face or credibility. [source: Rohrbaugh & Shoham 2001 (named only); Crippledini review]
Steps: (1) Openly take back your earlier direction. (2) Credit the change to new understanding or to a consultant, team member or "expert." (3) Give the new direction. The reviewer of Tactics describes reporting a point an expert raised that you don't fully understand, which can put the client in a position superior to the therapist. [source: Crippledini review; synthesis for step 1]
Example language (illustrative): "I think I steered you wrong last time. My colleague pointed out something I'd missed, and I'd like to change course."
Watch for: relief or engagement. If clients feel tricked, be honest that you changed your mind. [synthesis]

4.19 Using the team and consultant

Purpose / when: throughout, for planning, observation and maneuverability, if you have a team.
Steps: (1) The primary therapist interviews while colleagues watch behind a one-way mirror or on video. (2) The team helps formulate the loop, objective, position and mine field. (3) The therapist can use a consultant's view to frame or reverse a suggestion (see 4.18). (4) A non-treating team member does the follow-up. [source: Rohrbaugh & Shoham 2001; Crippledini review]
Solo adaptation: take a break before the end of the session to plan the task. Use peer consultation between sessions, and attribute U-turns to it. [synthesis]
Watch for: consent to observation and recording. [synthesis]

4.20 Predicting or prescribing relapse

Purpose / when: with clients who are overly optimistic or anxious after early change, to stop a slip from turning into a return to the old solution. [source: Rohrbaugh & Shoham 2001]
Steps: (1) Predict that a setback is likely and is part of change. (2) Discuss how the problem could come back, which usually means a return to the old solution. (3) Optionally ask the client to allow or produce a small slip on purpose. [source: Rohrbaugh & Shoham 2001; uncertain for step 3: named but not detailed]
Example language (illustrative): "Things may well slide back a bit. If they do, what would you be tempted to do again?"
Watch for: the client calmly handling a slip without going back to the old solution. [synthesis]

4.21 Nurturing early change

Purpose / when: whenever the client reports improvement, however small.
Steps: (1) Respond with restraint, not celebration. (2) Ask exactly what was done differently. (3) Reframe the change in the client's values, for example "that was the protective thing to do." (4) Build on it: if it works, do more. (5) Go slow (4.15) and discuss relapse (4.20). [source: Rohrbaugh & Shoham 2001; Wienands 2002]
Example language (illustrative): "Hm. That's interesting. What exactly did you do differently on Thursday?"
Watch for: the client giving credit to luck. Gently tie the change to what they did, without pushing. [synthesis]

4.22 Termination

Purpose / when: once the minimum goal is reached and the client is satisfied with the change in the complaint.
Steps: (1) Note the change soberly. (2) Ask the client's view on stopping. (3) Offer to keep unused "sessions in the bank." (4) If the client raises new problems, suggest a break to adjust to the change, then reassess. (5) Leave the door open to call. [source: Rohrbaugh & Shoham 2001; Wienands 2002 (Fisch ending after four sessions)]
Example language (illustrative): "It seems things have moved enough that we could stop here. You still have sessions if you need them."
Watch for: anxiety about stopping. Offer the banked sessions, and don't press the client to stay. [source: Rohrbaugh & Shoham 2001]

4.23 Follow-up

Purpose / when: at about 3 and 12 months, for evaluation, and it also offers some contact.
Steps: a non-treating colleague phones and asks: (1) whether the original complaint has changed, (2) whether new problems have appeared, and (3) whether help was sought elsewhere. The case is rated as success, significant improvement or failure. [source: Rohrbaugh & Shoham 2001; Carratalà et al. 2016 report an MRI outcome scale rating]
Watch for: clients describing the follow-up call as itself helpful. Keep it neutral. [synthesis]

4.24 Reported but cautionary: contingent ordeal contract

The review of Tactics reports a smoker's contract: he was allowed to smoke, but any puff obliged him to smoke a whole pack in the middle of the night. [source: Crippledini review] This sits closer to Erickson and Haley's ordeal work than to standard interdiction. Do not use health-damaging ordeals, and choose benign, consensual alternatives instead. [synthesis]

5. Handling problems

The sources deal directly with reluctance, the mine field and consolidation. Most of the other responses below are my own synthesis from the model's logic, marked [synthesis].

Reluctance to change. Do not argue for change. Use go slow (4.15) and dangers of improvement (4.16), work from patient position, and offer an observation task (4.12) instead of a change task. [source: Rohrbaugh & Shoham 2001]
Task not done. Pushing harder is the mine field. Take a one-down stance and take the blame: the task may have been badly chosen or explained, or the timing wrong. Ask what happened in neutral detail. Then either reframe the same direction in terms that fit the client's position better, make it smaller, or make a U-turn (4.18). Repeated non-completion often means you are not talking to the customer (see next item). [synthesis; mine-field logic from Rohrbaugh & Shoham 2001]
No customer: client sent by someone else, or a mandated client. Find out who is actually complaining (a parent, school, court or partner) and work with them, or consult them. With the attending client, ask what they would like to get out of the meetings, for example getting the referrer off their back, and treat that as the complaint. [source: Rohrbaugh & Shoham 2001 on consulting probation officers and others; synthesis for the rest]
Stuck: no change by sessions 4–5. Re-check four things: (1) Is the complaint concrete? (2) Did you find the main attempted solution, or a side one? (3) Is the strategic objective a true reversal, or more of the same in a new form, for example a new way to persuade? (4) Is the frame outside the client's position? Use team or peer review. Consider using the remaining sessions with someone else in the system. [synthesis]
Therapist caught in the loop. Examples are reassuring an anxious client who is reassured by everyone, encouraging a depressed client, pressing abstinence, or befriending the child against the parent. Stop, name the mine field to yourself, and do the reverse, for example express doubt or restraint. [source: Rohrbaugh & Shoham 2001]
Couples: the therapist becomes the referee. See partners separately, accept each one's view and give each an individual task. Do not share one partner's task with the other unless that is planned. [source: Rohrbaugh & Shoham 2001; Crippledini review] Be wary of creating secrets that could later rupture trust. Agree a rule on confidentiality at the start. [synthesis]
Rupture: the client feels manipulated or disbelieved. This is the main criticism of the model. Do not defend the intervention, since defending is itself the accuser-defender pattern. Take a one-down stance, acknowledge the client's experience and say plainly why you suggested it. Never misstate facts, risks or your own views about safety. Get advance consent that you will sometimes suggest experiments that may sound odd. [synthesis; criticism from Rohrbaugh & Shoham 2001 and the Crippledini review]
Overwhelm, or improvement too fast. Go slow, make tasks smaller, predict a slip, and space out sessions. [source: Rohrbaugh & Shoham 2001; synthesis for the spacing]
Client asks for explanations or insight, or wants to explore the past. Accept this as part of their position. Offer short interactional explanations that support the task, for example "we need to see how it works now." Do not spend sessions on history. If the client strongly prefers depth work, refer on rather than fight. [synthesis]
New problems after success. Suggest a break to adjust, then reassess. Do not roll straight into a new contract. [source: Rohrbaugh & Shoham 2001]
Risk: suicidality, self-harm, violence, child protection, severe substance withdrawal, psychosis with danger. The model's minimal-change, complaint-only stance does not override your duty of care. Carry out a standard risk assessment, safety planning and lethal-means counselling (secure storage or removal) where relevant. Make mandated reports, and coordinate medical care. Never use paradox, restraint, dangers of improvement or symptom prescription on risk behaviours. In intimate partner violence, see partners separately, prioritise safety, and do not frame abuse as a mutual loop. [synthesis]
Severe or chronic conditions (psychosis, eating disorders, chronic illness). The model can still be used with concerned relatives, as in the Weakland case of parents of a young man with a schizophrenia diagnosis and the stroke-family case study. Keep the identified patient's medical and psychiatric care in place, and target relatives' attempted solutions, such as overprotection or exhortation. [source: Carratalà et al. 2016; Rohrbaugh & Shoham 2001; synthesis for the cautions]
Session limit reached without change. Stop as agreed, review what was learned, and refer or offer a new contract after a break. An open-ended extension weakens the expectancy effect. [synthesis]
Cultural fit. Patient position is the model's built-in cultural adaptation. Learn the client's values and idioms, and frame tasks in them. Check whether one-down or indirect framing could seem disrespectful or evasive in the client's culture. [synthesis]

Sources

Sources: