Personal construct therapy and fixed-role therapy

1. Introduction

Version followed: George Kelly's original personal construct psychotherapy and fixed-role therapy (Kelly, 1955, The Psychology of Personal Constructs, vols 1–2; 1991 Routledge reprint). Later British personal construct psychology (PCP) counselling (Fransella & Dalton, 2000; Winter, 1992) supplements it where Kelly is silent, and each such addition is named.

Sources and status: this guide draws on the pages listed under Sources: the Internet Encyclopaedia of Personal Construct Psychology entries on fixed-role therapy, self-characterisation, group therapy and loosening; Routledge's contents listing for Kelly's vol. 2; the Wikipedia overview; and secondary summaries of Kelly's clinical procedures (Bay Path open textbook, Cote 1995 on ERIC, a UOW repository abstract, a practitioner summary). Kelly's 1955 text itself is not drawn on. Claims tagged [source: …] are supported by the pages listed under Sources. Claims tagged [uncertain] include chiefly Kelly's exact lists of loosening and tightening techniques, his group-therapy stage names, the professional-construct definitions, session frequency during fixed-role enactment, the self-characterization analysis headings, and the later methods of Hinkle, Landfield and Tschudi (4.5–4.8). Check those against Kelly vol. 2 and Fransella & Dalton (2000).

Personal construct therapy rests on Kelly's view that each person is like a scientist. People make sense of events by building bipolar personal constructs (e.g. 'reliable vs lets people down') and use them to anticipate what will happen. Distress follows when a construct system keeps failing to anticipate events, yet the person can't revise it, often because revising it would threaten core identity. Change happens when the client tests new or reorganised constructions through behavioural experiments and revises them on the evidence. In Kelly's words, therapy is about opening up "the possibility of continual change in the client's construct system" (vol. 2, p. 582). The therapist acts as a co-experimenter who understands the client's construing from the inside and offers ways to loosen, tighten, test or replace it. Fixed-role therapy is Kelly's best-known procedure. The client spends two weeks enacting a written character sketch of a different person, to learn in practice that the self can be construed in other ways. [source: Wikipedia, Personal construct theory (person-as-scientist, bipolar constructs, experimental cycle); Cote 1995 quoting Kelly 1955 vol. 2 p. 582; PCP Encyclopaedia, Fixed role therapy (two-week enactment)] [uncertain: 'threat to core identity' as the usual reason a system isn't revised is from recall of Kelly's professional constructs]

2. Framework

Kelly set no fixed phase sequence for personal construct therapy as a whole. Therapy is open-ended and steered by a transitive diagnosis: a working formulation of where the client's construing could move next, rather than a label. Fixed-role therapy is the one tightly sequenced procedure. Kelly's vol. 2 (Routledge edition) runs from the psychotherapist's role and approach, through appraisal and diagnosis ('Disorders of Construction', 'Disorder of Transition'), to 'Elaborating the Complaint', 'Elaborating the Personal System', 'Loosening and Tightening', 'Producing Psychotherapeutic Movement' and 'Special Techniques in Psychotherapy'. The phases below follow that order. Kelly lists the therapist's tasks in sequence: first accept the client's construct system as it stands, then help the client select new elements, speed up the tempo of experience, design and carry out experiments, and validate the client's experiments. The client is ultimately to become their own therapist. [source: Routledge contents listing for vol. 2; Bay Path textbook ch. 14, summarising Kelly 1955 vol. 2] [uncertain: the names of phases A–D are a working arrangement, not Kelly's headings]

Phase A: Elaborating the complaint and the construct system (often sessions 1–3 or more). Goal: understand the problem as the client construes it, how it began, what it means to them, and what it lets them anticipate. Find out early what the client expects therapy to do; Kelly held that the complaint itself shows this. Tools: credulous listening, Kelly's complaint-elaboration questions, the self-characterization, the repertory grid, and (from later PCP) laddering and pyramiding. Move on when the therapist can state the main constructs the problem sits on, the client's core role constructs, and which pole the client sees as 'me'. The therapist should also be able to say whether the system is too loose, too tight, too constricted or too preemptive. [source: Kelly 1955 vol. 2 chapter titles via Routledge listing; Bay Path textbook on client expectations; Fransella & Dalton 2000] [uncertain: the move-on criteria are from recall]

Phase B: Transitive diagnosis and choosing a strategy. Goal: decide which kind of movement is needed. Options include trying the contrast pole of an existing construct (Kelly's 'slot movement'), adding or choosing a different construct, making preverbal constructs verbal, loosening a tight system, tightening a loose one, or broadening a constricted field. Kelly's professional diagnostic constructs (anxiety, threat, guilt, hostility, aggressiveness, the CPC cycle, the creativity cycle, dilation/constriction, permeability) guide the choice. Move on when the therapist and client have agreed a direction and a first experiment. [source: Kelly 1955 vol. 2; Winter 1992]

Phase C: Reconstruction through experiment. Goal: the client tries new constructions in and out of the session and judges them against outcomes. This phase includes loosening and tightening work, enactment (role-play), homework experiments, and fixed-role therapy when a broad, rapid shift in self-construing is wanted. Move on when new constructions predict events usefully and the client can produce further alternatives without the therapist. [source: Kelly 1955; Epting 1984]

Phase D: Consolidation and ending. Goal: tighten and validate the new constructions, review what was learned about one's capacity to reconstrue, and plan further experiments. In PCP, ending is itself an experiment the client can revise. [source: Fransella & Dalton 2000] [synthesis: specific ending tasks]

Fixed-role therapy sequence (Kelly 1955, vol. 1, ch. 8):

The client writes a self-characterization (see 4.3). [source: PCP Encyclopaedia, Fixed role therapy]
The therapist writes a fixed-role sketch of a new character with a new name, drawing on the self-characterization, interview material and, where available, the Rep Test. [source: Bay Path textbook] Kelly's use of a panel of clinicians to draft it is [uncertain: recalled].
The sketch is presented, and the client is asked to enact it only if they find the character plausible. [source: PCP Encyclopaedia]
The client enacts the role for two weeks. [source: PCP Encyclopaedia] Secondary accounts vary ('a period of weeks', Bay Path; 'perhaps 3 weeks', hypnomanchester summary); this guide follows the two weeks the Encyclopaedia attributes to Kelly.
During this period several sessions are held in which the client rehearses the role, concentrating on situations they are likely to meet outside therapy. [source: PCP Encyclopaedia; 'frequent' sessions, hypnomanchester summary] Ordering the rehearsals from less to more personal (work or study, peers, intimate relationships, family, life outlook) is [uncertain: recalled].
The role is withdrawn and the experience reviewed. Kelly did not intend the client to become the character permanently. [source: PCP Encyclopaedia]

Move from step 3 to 4 only when the client finds the character plausible [source: PCP Encyclopaedia]. Also asking whether it is someone they'd like to know follows Kelly as recalled [uncertain: wording of Kelly's acceptance check]. Move to withdrawal at the agreed end date, not on 'success'. [synthesis] [uncertain: exact session frequency during enactment, which Kelly may have given as several a week]

Format and dose. Kelly treated interview planning (frequency, length, tempo and termination) as decisions the therapist plans and stays in control of, adjusted when the client's outside circumstances require it. [source: Bay Path textbook, summarising Kelly 1955 vol. 2] Individual weekly sessions are typical outside fixed-role therapy, and length is open-ended [uncertain]. Kelly also described group psychotherapy (see 4.20). Fixed-role therapy has been adapted for groups (Epting & Nazario 1987; Beail & Parker 1991) and for a couple (Kremsdorf 1985) (see 4.13a). [source: PCP Encyclopaedia, Fixed role therapy] PCP has been run as brief counselling and as long-term therapy, and no standard session count exists. [uncertain]

3. Guiding principles

[source for this section: Kelly 1955 vols 1–2; Kelly 1964 'The language of hypothesis'; Fransella & Dalton 2000; Epting 1984] Supported by sources: constructive alternativism, person-as-scientist and bipolarity (Wikipedia, Personal construct theory); the credulous approach and Kelly's first principle, rendered as "if you do not know what is wrong with someone, ask them" (PCP Encyclopaedia, Self-characterisation, citing Kelly 1955/1991 p. 323, vol. 1 p. 241); the fundamental postulate (Kelly 1955 vol. 1 p. 46, via Cote 1995); 'not the opposite' and make-believe as a route to experimenting (PCP Encyclopaedia, Fixed role therapy); never confining a client to the stereotype of their culture (Kelly 1955 vol. 2 p. 833, via Bay Path). [uncertain: the wording of the other principles is paraphrase]

Constructive alternativism. Any event can be construed in other ways. In practice, never treat the client's current account, or your own, as the only possible reading. Your job is to help generate alternatives.
Person-as-scientist. The client already forms hypotheses and tests them. Frame change as experiments with predictions and evidence, not as advice or insight alone.
Fundamental postulate (anticipation). Behaviour is organised around anticipating events. Ask what a symptom lets the client predict or avoid having to predict.
Credulous approach. Take the client's statements as true as the client's construction, even when they're factually doubtful. Listen for what the account means to them before evaluating it.
'Ask the client'. Kelly's well-known first principle: if you don't know what is wrong, ask the person, who may tell you. Start from the client's own formulation. [paraphrase]
Bipolarity (dichotomy corollary). A construct gets its meaning from its contrast. Always ask 'as opposed to what?' A client may resist a change because, for them, the contrast pole implies something unacceptable.
Choice corollary (elaborative choice). People choose the pole that seems to offer more chance to extend or define their system. A symptom may be the most meaningful choice available. Look for what would be lost by giving it up.
Sociality / role. Relating means construing the other's construing. The therapist plays a role in this sense, by subsuming the client's outlook within professional constructs rather than imposing their own.
Transitive diagnosis. Diagnose pathways of movement, not fixed categories. Each formulation should say what could change next and how.
Professional diagnostic constructs. Use Kelly's terms to read the client's process: anxiety, threat, fear, guilt, hostility, aggressiveness, loosening/tightening, dilation/constriction, permeability, preverbal construing, and the CPC and creativity cycles. They tell you whether to slow down, loosen, tighten or broaden.
Pacing by threat. Changes in core role constructs are threatening. Start with peripheral, low-stakes experiments, and approach core constructs only with support.
Invitational mood. Kelly (1964) proposed speaking in the language of hypothesis ('suppose…', 'what if…'). Phrase interpretations and tasks as invitations to try something, not as statements of fact.
Therapist as co-experimenter. Kelly compared the therapist to a research supervisor. Help design experiments, protect the client from overly costly ones, and treat outcomes as data, not as pass or fail.
Reflexivity. The theory applies to the therapist too. Watch your own preemptive or hostile construing of the client, for example insisting they are 'resistant'.
Technical eclecticism within a coherent theory. Kelly allowed any technique the theory can account for. Choose techniques for the movement you want (loosening, tightening, testing), not for their brand.
Make-believe as protection. In fixed-role therapy and enactment, the 'as if' frame lets the client try new behaviour without committing core identity. Keep the frame explicit.
Not the opposite, not the ideal. A fixed-role sketch should not be the mirror image or an 'improved' version of the client. It should offer a different main theme that opens new questions.

4. Intervention inventory

All example language is illustrative unless quoted. Kelly's own interventions come first (4.1–4.4, 4.9–4.20). Later PCP methods are labelled with their originators. [source: Kelly 1955 unless stated] Supported by sources: the fixed-role procedure (4.10–4.13a), self-characterization (4.3), the Rep Test's role titles and triad method (4.4), the definition of loosening and the creativity cycle (4.15), the existence of Kelly's staged group method (4.20) and the rotating-dyad format of the interpersonal transaction group (4.21). [uncertain: the rest of this section, including Hinkle's, Landfield's and Tschudi's methods (4.5–4.8) and Kelly's technique lists in 4.15–4.17 and 4.19, is not verified]

4.1 Credulous elaboration of the complaint

Purpose / when: First sessions, and whenever a new problem appears. Sets out the problem as the client construes it and shows where change is possible. [source: Kelly 1955 vol. 2]
Steps: (1) Listen credulously and accept the account as the client's reality. (2) Work through Kelly's areas of inquiry: what the client wants help with; when the problem was first noticed; the conditions under which it appeared; what has been tried; what has changed with time or treatment; when it is most and least noticeable. (3) Note the client's own constructs and words, and ask for the contrast pole of key terms. (4) Ask what the problem lets them predict, and what life would be like without it. (5) Summarise in the client's terms and check. [uncertain: exact list of Kelly's complaint questions]
Example language (illustrative): "When you say you're 'weak', what would the opposite be for you?" / "When is this least noticeable?" / "If it went away tomorrow, what would you have to deal with that you don't now?"
Watch for: Problems framed in time ('since I left home') suggest a construct system overwhelmed by change, which points to anxiety. Unusual contrast poles ('calm vs. a doormat') explain resistance to apparently desirable change. It's working if the client gives richer, more differentiated accounts and starts naming alternatives.

4.2 Controlled elaboration of the construct system

Purpose / when: To make the client's system clearer and more internally consistent so that inconsistencies become visible and can be tested. Use it when the client's account is vague or contradictory. [source: Kelly 1955 vol. 2]
Steps: (1) Pick a construct the client uses. (2) Ask them to elaborate it: examples, who fits each pole, what follows from each pole, and where the construct does not apply (its range). (3) Ask how it relates to other constructs they use. (4) Reflect back implications without contesting them, and let the client notice conflicts. (5) Agree small checks on whether the construct predicts well.
Example language (illustrative): "Who do you know who is 'genuinely confident'? What do they do that you'd notice?" / "Does 'selfish' apply at work as well as at home?"
Watch for: Too fast or too wide elaboration can raise anxiety. It's working if the client starts to qualify, sub-divide or question the construct on their own.

4.3 Self-characterization

Purpose / when: To see how the client structures their world, sees themselves within it and deals with it, from their own vantage point. It can be done before therapy starts or at any later point, in session or as homework. It is the first step of fixed-role therapy and can be repeated to review change. Kelly presented it as an application of the credulous approach and of his first principle (ask the client). [source: Crittenden & Ashkar 2012 (UOW abstract); PCP Encyclopaedia, Self-characterisation; Kelly 1955 vol. 1 ch. 7]
Steps: (1) Ask the client to write a character sketch of themselves in the third person, under their own name, as if they were the principal character in a play, written as a friend who knows them intimately and sympathetically might write it. [source: Bay Path textbook; hypnomanchester summary] (2) Keep instructions minimal so the client chooses the content. (3) Read it credulously. The interest is in how the client sees themselves and their relationships, not in whether the account is true, and it is not scored by counting positive and negative statements. [source: PCP Encyclopaedia, Self-characterisation] Kelly suggested attending to the opening and closing statements, the sequence and transitions between topics, how it is organised, repeated terms, the contexts or areas of life mentioned, recurring themes, and the dimensions (contrasts) implied. [uncertain: exact list of Kelly's analysis headings; the Encyclopaedia refers readers to Fransella, Bell & Bannister for them] (4) Draft hypotheses about the client's core constructs and what the client is struggling to anticipate. (5) Optionally discuss it with the client.
Example language (illustrative): "I'd like you to write about yourself as if a close, understanding friend were describing you, starting with your name: 'Anna is…'. There's no right length or content."
Watch for: Writing in the third person makes the task less threatening. Note what is missing, such as whole areas of life. A second sketch later in therapy can show change. Variants include writing it as one might be in five years, or as a significant other would describe the client. [uncertain: variants recalled from Fransella & Dalton 2000]

4.4 Role Construct Repertory Test and repertory grid

Purpose / when: To elicit the client's constructs about people (elements) and map how they relate. Use it for assessment, for formulation, and to measure change across therapy. [source: Bay Path textbook; Wikipedia, Personal construct theory; Kelly 1955 vol. 1 chs 5–6; Fransella, Bell & Bannister 2004]
Steps: (1) Choose elements. Kelly's Role Title List has 25 role titles, such as mother, father, siblings, spouse or partner, a liked and a disliked teacher or employer, a person who seems to dislike the client, and the most successful, intelligent or interesting person the client knows. The client names a real person for each. [source: Bay Path textbook] Later practice often uses fewer elements and adds 'self', 'self as I would like to be' and 'self in the future'. [uncertain] (2) Present elements in threes (triads) and ask how two are alike in some important way and different from the third. The shared quality is the emergent pole. [source: Bay Path; Wikipedia] (3) Ask what the opposite is, which gives the contrast (Kelly's 'implicit') pole. [source: Bay Path] (4) Repeat with new triads until no new constructs emerge. [source: hypnomanchester summary] (5) In the grid form, constructs are rows and elements columns, and the client places every element on every construct (ticks, ratings or rankings). [source: Wikipedia; hypnomanchester summary] (6) Analyse by inspection, for example counting cells where two constructs match to see how closely they are linked, or with software: which constructs cluster, where 'self' and 'ideal self' sit, and which constructs are lopsided or unused. [source: hypnomanchester summary for matching counts; the rest uncertain] (7) Share the findings as hypotheses. Kelly's formal analysis looked at the number and range of constructs; a skilled examiner can also judge which constructs are functionally equivalent. [source: Bay Path] Kelly described several forms (full-context, minimum-context, sequential, self-identification, personal-role). [uncertain: form names]
Example language (illustrative): "Think about your sister, your old manager, and yourself. In what important way are two of them alike and different from the third?" / "And what would be the opposite of 'pushy' for you?"
Watch for: Clients sometimes give superficial constructs ('both tall'). Ask for ones about how people behave or think. A very tight grid (one dominant component) suggests a rigid system, and a very loose one suggests loose construing. A large self–ideal distance links to low self-esteem. It's working if the grid gives you and the client usable hypotheses and later shows movement. [uncertain: interpretive rules from recall]

4.5 Laddering (Hinkle, 1965)

Purpose / when: To move from a specific construct to the superordinate, core values it implies. This shows why a change would be threatening. [source: Hinkle 1965; Fransella & Dalton 2000]
Steps: (1) Take an elicited construct. (2) Ask which pole the client prefers. (3) Ask why that matters, what the advantage of being that is, and get the contrast ('as opposed to?'). (4) Repeat with the new construct until the answers become very general or self-evident, such as core values. (5) Write the ladder out. (6) Use it to see what change would imply for the client's identity.
Example language (illustrative): "Why would you rather be 'organised' than 'chaotic'?" → "What does being 'in control' give you, compared with 'at others' mercy'?"
Watch for: It can bring up core issues fast. Check for distress and stop or slow down. Answers that loop back mean you have reached a core construct. It's working when the client recognises the ladder as important ('that's what it's really about').

4.6 Pyramiding (Landfield, 1971)

Purpose / when: The reverse of laddering: to move from an abstract construct down to concrete, observable, subordinate constructs. Use it to plan experiments or when the client speaks only in abstractions. [source: Landfield 1971; Fransella & Dalton 2000] [uncertain]
Steps: (1) Take a construct pole. (2) Ask what kind of person that is, or how you'd know someone was like that. (3) For each answer ask the same question, branching down. (4) Stop when you reach behaviours that can be observed or tried.
Example language (illustrative): "What does a 'warm' person actually do, the kind of thing you'd see?"
Watch for: It gives material for homework experiments and for fixed-role sketches. It's working if concrete behaviours emerge that the client could try.

4.7 Implications grid and resistance-to-change grid (Hinkle, 1965)

Purpose / when: To show how constructs imply one another, and which constructs the client would least want to change, so you can tell what is core and threatening. [source: Hinkle 1965; Fransella, Bell & Bannister 2004] [uncertain: procedural details]
Steps (implications grid): (1) Use laddered constructs. (2) For each pair, ask whether a change from the preferred to the non-preferred pole on construct A would imply a change on construct B. (3) Count the implications per construct; heavily implicated constructs are more superordinate.
Steps (resistance-to-change): (1) Pair constructs. (2) Ask which of the two the client would rather change to the non-preferred pole, if forced to change one. (3) Tally; constructs rarely given up are core.
Example language (illustrative): "If you had to move from 'reliable' to 'unreliable' or from 'clever' to 'dull', which would you rather change?"
Watch for: It's time-consuming; use it selectively. It helps when a client keeps 'not changing' a symptom whose loss would cost a highly implicated identity.

4.8 ABC model (Tschudi, 1977)

Purpose / when: For 'stuck' problems where the client wants change but doesn't change. It shows the hidden advantages of the problem pole and the disadvantages of the desired pole. [source: Tschudi 1977; Fransella & Dalton 2000] [uncertain]
Steps: (1) A: name the problem as a construct, a1 (present, unwanted) vs a2 (desired). (2) B: ask for the disadvantages of a1 and the advantages of a2. These usually come easily. (3) C: ask for the advantages of a1 and the disadvantages of a2. These are often surprising. (4) Discuss how C keeps a1 in place. (5) Design experiments to get the C1 benefits by other means, or to test whether the C2 fears come true.
Example language (illustrative): "What might be the downside of being 'assertive', for you specifically?" / "Is there anything 'staying quiet' protects or gets you?"
Watch for: Present it as normal logic ('it makes sense'), not as secret gain, so it doesn't feel like an accusation. It's working when the client sees the problem as a reasonable choice and starts generating alternatives.

4.9 Transitive diagnosis using professional constructs

Purpose / when: Ongoing formulation that chooses the next move. [source: Kelly 1955 vol. 2; Winter 1992]
Steps: Map the client's state onto Kelly's constructs and choose a strategy for each. Anxiety (events outside the system's range) calls for helping build or extend constructs to cover them. Threat (imminent comprehensive change in core structure) calls for slowing down and working peripherally. Guilt (being dislodged from one's core role) calls for exploring the core role and alternative roles. Hostility (forcing evidence to fit a failed prediction) calls for looking at what invalidation would cost, not confronting. Constriction calls for gradual dilation. Loose construing calls for tightening, and tight construing for loosening. Being stuck in the circumspection phase of the CPC cycle calls for help with preemption and choice; impulsive cycling calls for help with circumspection.
Example language (illustrative): Mostly internal to the therapist. To the client: "It sounds as if part of you is holding on to the old view because letting it go would mean rethinking a lot. Shall we go slowly?"
Watch for: Use these terms as working hypotheses about process, not as labels.

4.10 Fixed-role therapy, step 1: writing the fixed-role sketch

Purpose / when: To give the client a fully drawn alternative self to try out. Kelly's aim was experimentation and "one good, rousing, construct-shaking experience" (Kelly 1955, p. 412), not a permanent change of character. It suits clients whose self-construing is stuck, who can engage in make-believe, and for whom a short intensive experiment is practical. Kelly saw it as one technique among many, not the whole of PCP. [source: PCP Encyclopaedia, Fixed role therapy; Kelly 1955 vol. 1 ch. 8; Epting 1984; Fransella & Dalton 2000]
Steps: (1) Study the self-characterization, interview material and any Rep Test data. [source: Bay Path textbook] (2) Choose a major theme for the new character that sets new processes in motion rather than fixing a trait. Kelly's sketches often centred on curiosity about how other people see things, which uses the sociality corollary. [uncertain: recalled] (3) The character is not the client's complete opposite. It is likely to introduce a new construct orthogonal to the client's main construct dimensions ("at 90 degrees" to the self-portrait). [source: PCP Encyclopaedia; hypnomanchester summary] Kelly also reported that a sharply different role was easier for clients to play than a minor adjustment of their usual self. [source: Bay Path textbook] Reconcile the two: write a clearly different character, but different along a new dimension, not the mirror image on the client's own dimension. A pure contrast-pole character invites 'slot rattling'. [synthesis] (4) Don't write an ideal or 'healthy' person, but a plausible one with an inner life. (5) Include how the character thinks and feels, not only what they do. Cover areas the client will meet: work or study, friends, intimate relationships, family, and outlook on life. (6) Build in ways the character gets feedback (validation) from others, such as asking questions and noticing reactions. (7) Give the character a new name. In Kelly's illustrative case, 'Ronald Barrett' was given the role of 'Kenneth Norton'. [uncertain: case names recalled from Kelly 1955] (8) Kelly used a panel of clinicians to draft and agree the sketch [uncertain]. A solo therapist drafts it and adjusts it until the client is comfortable with it. [source: hypnomanchester summary]
Example language (illustrative): Sketch extract in your own words: "Sam is someone people find easy to talk to, mostly because Sam is quietly curious about how they see things, and keeps finding out how differently people view the same event."
Watch for: Sketches that read as a reproach ('unlike now, Sam is confident') are a mistake. Too much detail leaves no room for the client to invent. A sketch built on the contrast pole of the client's core construct may be rejected or produce guilt. [synthesis]

4.11 Fixed-role therapy, step 2: presenting the sketch and acceptance check

Purpose / when: To make sure the role is credible before enactment. The client is asked to enact the character only if they find it plausible. [source: PCP Encyclopaedia, Fixed role therapy; Kelly 1955 vol. 1 ch. 8]
Steps: (1) Give the client the sketch to read, often more than once. (2) Ask whether this sounds like a believable person, and whether it is someone they'd like to know. The question is about plausibility and liking, not 'would you like to be them'. [source for plausibility: PCP Encyclopaedia] [uncertain: the 'like to know' wording] (3) Revise any parts the client finds implausible or unappealing until they are comfortable with it. [source: hypnomanchester summary] (4) Explain the experiment. For two weeks the client will act as if they were this person in every situation possible. [source for two weeks: PCP Encyclopaedia] Kelly framed the client's usual self as being away on holiday, with the client thinking, eating and sleeping as the character. [uncertain: recalled] (5) Make clear it is temporary and that the old self can be taken back. [source: PCP Encyclopaedia, not a permanent change] (6) Agree the schedule of sessions and the end date.
Example language (illustrative): "Does Sam seem like a real person to you, someone you'd like to know?" / "For two weeks, let your usual self take a break. As far as you can, be Sam. Then we'll look at what happened."
Watch for: If the client wants to be Sam permanently, explain that the goal is to learn about change, not to swap identities. If they find Sam unlikeable, revise; don't persuade. Get consent for the time commitment and the social exposure involved. [synthesis]

4.12 Fixed-role therapy, step 3: enactment period and rehearsal sessions

Purpose / when: To support and shape the client's living of the role, and to process what they learn. [source: PCP Encyclopaedia, Fixed role therapy; Kelly 1955 vol. 1 ch. 8; Epting 1984]
Steps: (1) Hold several sessions during the enactment period, frequently enough to discuss experiences and how to play the role. [source: PCP Encyclopaedia; hypnomanchester summary] [uncertain: Kelly's exact frequency] (2) Each session, ask how the character got on and what others did in response. Use the character's name. (3) Rehearse the role in session, concentrating on situations the client is likely to face outside the therapy room. [source: PCP Encyclopaedia] The therapist plays the other person (a colleague, friend, partner, parent) and the client plays the character. Then reverse roles so the client experiences how the character comes across. [uncertain: role reversal within fixed-role rehearsal recalled] (4) Move through situations roughly in order of increasing personal intimacy: work, peers, romantic or intimate situations, family, then life direction and values. [uncertain: recalled order] (5) Coach small specific changes ('what would Sam ask here?'). (6) Ask the client to notice others' responses as evidence. (7) Expect awkwardness; treat it as information.
Example language (illustrative): "How did Sam handle the team meeting?" / "I'll be your manager. You be Sam… Now let's swap: I'll be Sam and you be your manager. What was that like from the other side?"
Watch for: It's working if the client reports that people reacted differently to them [source: hypnomanchester summary], notices they could think differently, or finds parts of Sam 'already in me'. Problems include the client dropping the role in high-stakes settings (agree which settings first), shame, and over-reliance on the script. Tell the client they can ask 'what would Sam do?' instead of following the text literally. [synthesis]

4.13 Fixed-role therapy, step 4: withdrawal of the role and review

Purpose / when: At the agreed end date. To end the role cleanly and consolidate the learning that the self can be reconstrued. Kelly did not intend the client to keep the new character; the point is to show the client they are not trapped in their own autobiography and habitual ways of construing. [source: PCP Encyclopaedia, Fixed role therapy; hypnomanchester summary; Kelly 1955 vol. 1 ch. 8; Fransella & Dalton 2000]
Steps: (1) Formally drop the role: the client's usual self 'returns'. (2) Review what was experienced, what others did, and what the client learned about themselves and others. (3) Ask which bits the client wants to keep, adapt or discard, in their own name. (4) Optionally ask for a new self-characterization or repeat the grid. (5) Plan further experiments or ongoing therapy. [uncertain]
Example language (illustrative): "Sam's two weeks are up. Welcome back. What did you find out, about others and about yourself?" / "Which of Sam's ways might you want to try as you?"
Watch for: The aim is not adoption of Sam but discovery that 'I can change how I construe and act'. A client who feels the result was 'fake' can be helped to see that every self is constructed and tested. That's a central PCP idea. [synthesis: wording of this response]

4.13a Fixed-role variants: groups and couples

Purpose / when: Kelly and later practitioners applied fixed-role therapy in groups (Epting & Nazario 1987; Beail & Parker 1991) and with a couple (Kremsdorf 1985). Use a group version when several clients can rehearse with one another; use the couple version when each partner's construing of the other keeps a problem going. [source: PCP Encyclopaedia, Fixed role therapy, for the existence of these variants] [uncertain: procedural details below]
Steps: [synthesis, adapting 4.10–4.13; not the authors' published procedure] (1) Each member (or partner) writes a self-characterization. (2) The therapist writes an individual sketch for each, following the rules in 4.10. (3) Check plausibility with each person privately. (4) During the enactment period, members rehearse their roles with one another in sessions, playing the people each will meet. With a couple, agree beforehand whether each partner knows the other's role. (5) Withdraw all roles together and review as a group or couple.
Example language (illustrative): "This week, each of you will be trying out a character. In here you can practise on each other: Lee, you play Jo's flatmate while Jo tries out 'Alex'."
Watch for: Members commenting on each other's 'real' selves rather than the roles; keep the make-believe frame explicit. In couples, a partner using the role as a weapon ('you're only acting'). [synthesis]

4.14 Enactment (brief role-play)

Purpose / when: In ordinary sessions, to let the client try constructions, see another person's perspective, or rehearse experiments. Kelly saw it as a tightening and testing technique and as a way to elaborate the role of another person. [source: Kelly 1955 vol. 2]
Steps: (1) Pick a specific situation. (2) Set it up briefly without long preparation. (3) Role-play for a short time; Kelly favoured short enactments. (4) Exchange parts so the client plays the other person and the therapist plays the client. (5) Step out and discuss what each person anticipated. (6) Optionally repeat with a new construction.
Example language (illustrative): "Let's try it now. I'll be your father. Say what you'd like to say… OK, swap: you be him."
Watch for: Keep it short so the client isn't overwhelmed. Avoid making the client play a deeply threatening role early. It's working if the client says something new about how others see things.

4.15 Loosening techniques

Purpose / when: When construing is too tight, rigid or literal and the client can't generate new ideas. Kelly defined loosening as characteristic of constructs "that lead to varying predictions" (vol. 2, p. 330): an element can sit at one pole today and the other tomorrow while the construct keeps its identity. Loosening is the first half of the creativity cycle, which starts with loosened construing and returns to tight construing; the C-P-C cycle, by contrast, starts with propositional construing. [source: PCP Encyclopaedia, Loosening, citing Kelly 1955/1991 vol. 2 pp. 330, 351] Kelly's vol. 2 has a chapter on 'Loosening and Tightening'. [source: Routledge contents listing]
Steps: Kelly described four. (1) Relaxation: help the client be physically relaxed. (2) Chain association: invite the client to follow ideas freely without checking them. (3) Recounting dreams: ask for dreams and explore them without fixing meanings. Kelly regarded dreams as loosely construed material. (4) Uncritical acceptance by the therapist: receive whatever the client says without evaluation. Use loosening for part of a session, then tighten before the end. [uncertain]
Example language (illustrative): "Don't worry about whether it makes sense. Just say what comes next." / "Tell me the dream as it comes; we won't try to work it out yet."
Watch for: Over-loosening can produce confusion or anxiety. Some people are 'locked in' to loose construing and can't tighten enough to deal with their world, which Bannister linked to schizophrenic thought disorder; with such clients loosening is contraindicated or needs great care. [source: PCP Encyclopaedia, Loosening, citing Bannister 1962, 1963] Tighten before the client leaves. It's working if new, tentative constructions appear. [synthesis]

4.16 Tightening techniques

Purpose / when: When construing is too loose to produce testable predictions, or to firm up new ideas from loosening so they can be tested: the second half of the creativity cycle. [source: Kelly 1955 vol. 2; Winter 1992] [uncertain: exact list]
Steps: Use one or more. (1) Judging / superordinating: ask the client to evaluate or classify an idea. (2) Summarising: have the client summarise the session or an idea. (3) Explaining to the therapist: ask the client to explain an idea as if to someone who doesn't know it. (4) Historical explanation: ask how it came about. (5) Asking for validating evidence: what would show this to be true or false? (6) Word-binding: fix a construct to specific words; tie preverbal or vague feelings to a label. (7) Time-binding: tie a construction to the time it was formed ('that's how you saw it at nine'). (8) Enactment: act it out to make it concrete. (9) Concept formation / direct challenge sparingly.
Example language (illustrative): "Could you put that into one sentence I'd remember?" / "What would you expect to see this week if that were true?" / "That was how it looked to you when you were nine. Is it still the best way to see it?"
Watch for: Premature tightening in a client who needs loosening produces rigid restatement. It's working if the client can state a clear prediction and test it.

4.17 Producing alternative constructions (reconstruction moves)

Purpose / when: In Phase C, to help the client find another way of construing a problem. [source: Kelly 1955 vol. 2; Winter 1992] [uncertain: Kelly's exact taxonomy]
Steps: Choose the kind of movement. (a) Slot movement: trying the other pole of an existing construct. It is quick but often unstable and can reverse. (b) Choosing a different existing construct to apply to the situation. (c) Developing a new construct, often through experiments or through the therapist offering new elements and contrasts. (d) Making preverbal constructs verbal, giving words to feelings or reactions formed before language. (e) Changing permeability, so a construct applies to new events or ceases to. Then (f) frame the new construction as a hypothesis and (g) design a test.
Example language (illustrative): "You've been seeing your silence as 'cowardly vs brave'. Is there another way you could see it, say as 'listening vs speaking'?"
Watch for: If the client simply swaps poles ('I'll be aggressive now'), help them develop a new dimension rather than pursue the swap.

4.18 Experiments and homework

Purpose / when: Throughout. This is the core change mechanism: test constructions against events. [source: Kelly 1955; Epting 1984; Fransella & Dalton 2000]
Steps: (1) State the construction to test. (2) Make a specific prediction. (3) Choose a small, low-risk real-life test. (4) Agree how to record the result. (5) Review the outcome as data and revise the construction. (6) Escalate as confidence grows.
Example language (illustrative): "So your prediction is that if you ask Jo a question, Jo will look bored. Shall we test that this week and see?"
Watch for: Disconfirmation can be met with hostility (explaining it away). If so, explore what accepting the evidence would cost.

4.19 Reassurance and support (used selectively)

Purpose / when: Kelly described reassurance as a temporary device to keep a client's system working (reducing anxiety by making things look more consistent). He described support as broad acceptance that lets the client experiment safely. He warned that reassurance blocks change if overused. [source: Kelly 1955 vol. 2] [uncertain]
Steps: (1) Use reassurance briefly at points of acute anxiety or before a frightening experiment. (2) Withdraw it once the client is steadier. (3) Give support continuously by validating the client as an experimenter, not by validating particular constructions.
Example language (illustrative): "It makes sense that this feels unsettling. Rethinking things usually does. You don't have to decide anything today."
Watch for: A dependent client seeking constant reassurance. Shift toward support for experimenting.

4.20 Group personal construct psychotherapy (Kelly)

Purpose / when: For clients who benefit from many others to experiment with. In Kelly's original method the group passes through a series of stages, with a different therapeutic approach at each stage; the method is now rarely used, and later PCP group methods mostly use structured exercises (see 4.21). [source: PCP Encyclopaedia, Group therapy; Kelly 1955 vol. 2; Winter 1992]
Steps: Move the group through: (1) initiating mutual support; (2) initiating primary role relationships (members construe each other's construing); (3) initiating mutual primary enterprises (joint experiments within the group); (4) exploring personal problems; (5) exploring secondary roles (with people outside the group); (6) exploring secondary enterprises (experiments in the outside world). Use enactment, role exchange and between-session experiments. [uncertain: stage names and order; sources confirm only that there are stages]
Example language (illustrative): "Before we look at problems, let's each guess how someone else here sees today's topic, and then check with them."
Watch for: Pushing personal disclosure before mutual support exists. [synthesis]

4.21 Interpersonal transaction group (Landfield & Rivers, 1975)

Purpose / when: A structured group format developed within PCP, in which part of each session is run in rotating dyads, building sociality through repeated one-to-one exchanges. Neimeyer (1988) published clinical guidelines for running it; it has been used, for example, with depression and with survivors of abuse. [source: PCP Encyclopaedia, Group therapy (format, Landfield & Rivers 1975, Neimeyer 1988)] [uncertain: client groups]
Steps: (1) Members rate their mood or note a mood tag at the start. [uncertain] (2) Each member talks in turn with every other member in pairs, a few minutes per pair, on a theme the therapist provides; themes move from less to more personal over sessions. [source: PCP Encyclopaedia for rotating pairs and therapist-set theme] [uncertain: timing and theme progression] (3) Pairs listen to understand rather than to advise. [uncertain] (4) The dyadic exchanges are then discussed in a plenary phase. [source: PCP Encyclopaedia] (5) Closing mood check. [uncertain]
Example language (illustrative): "In this pair, take turns: one talks about a time you felt understood, the other just tries to understand. Then switch."
Watch for: Topics escalating in intimacy too quickly. [synthesis]

4.22 Serial validation for loose construing (Bannister et al., 1975)

Purpose / when: A research-derived procedure for clients whose construing is very loose (thought disorder). It aims to tighten construing by repeatedly validating specific predictions. It grew from Bannister's work on schizophrenic thought disorder, which he explained as the result of serial invalidation. [source: PCP Encyclopaedia, Loosening, citing Bannister 1962, 1963] [uncertain: the 1975 serial validation study itself (Bannister, Adams-Webber, Penn & Radley) is not drawn on; steps below are recalled]
Steps: (1) Use the grid to identify construct areas that remain relatively tight. (2) Elicit simple predictions within those areas. (3) Arrange for them to be validated, building gradually toward less tight areas. (4) Repeat the grid to monitor change.
Watch for: Use within a broader treatment plan; the evidence base is limited. [synthesis: positioning]

Not covered: PCP methods for children and families, including Ravenette's techniques and Procter's family approaches. These are separate developments and were not researched for this guide. [uncertain]

5. Handling problems

PCP reframes most 'problems' as understandable construing. Kelly and later authors explain resistance mainly through threat, guilt and hostility. [source: Kelly 1955 vol. 2; Fransella & Dalton 2000; Winter 1992] [uncertain: from recall] Specific responses are [synthesis] unless tagged.

'Resistance' / client doesn't change. Treat it as threat. The change would require rebuilding core constructs, or the alternative pole carries unacceptable implications. Response: use laddering, ABC or resistance-to-change grids to find what the change would imply. Work on peripheral constructs first, and offer a third alternative instead of the contrast pole. [source: Kelly 1955; Tschudi 1977; Hinkle 1965] Drop the word 'resistance' with the client: "It makes sense not to change something that holds a lot up." (illustrative)
Hostility (explaining away disconfirmations). Response: don't argue with the evidence. Make the hostility understandable ("If that were wrong, what else would have to be wrong?"). Lower the stakes by testing smaller predictions, and work with the guilt or threat underneath. [source: Kelly 1955] [synthesis: specifics]
Stuck / going round in circles. Check whether construing is too tight (the same arguments repeated) or too loose (vague, shifting). Loosen with 4.15 or tighten with 4.16 accordingly. Check whether the client is stuck in circumspection; if so, help with a time-limited choice and experiment (the CPC cycle). Consider a fixed-role experiment for a global shift. [source: Kelly 1955] [synthesis]
Overwhelm / acute anxiety. In Kelly's terms, events lie outside the client's construct system. Response: slow down, tighten (summarise, word-bind), offer brief reassurance and steady support, and narrow the focus to one area. Stop loosening work and set small, predictable experiments. Use standard risk assessment if distress includes self-harm or suicidal thinking. [source: Kelly 1955 on reassurance] [synthesis]
Guilt after change ('that's not me'). Change has dislodged the client from their core role. Response: explore the core role: what kind of person 'I' must be and who defined it. Use make-believe framing (fixed role, enactment) so experiments don't count as identity betrayal. Explore other roles that preserve key values. [source: Kelly 1955 definition of guilt] [synthesis: response]
Non-adherence to experiments or the fixed role. Treat it as information, not failure. Ask what the client anticipated. Check whether the role or experiment was too threatening, too intimate too early, or implausible. Revise the sketch or scale down. For fixed-role therapy, agree in advance which situations are in scope. [source: Kelly 1955 on revising sketches] [synthesis]
Client finds the fixed role 'fake' or dishonest. Restate the as-if frame: it's an experiment, not deceit, and all roles are constructed. Ask what Sam would say about feeling fake. If ethical discomfort persists (e.g. with a partner), let the client tell others they're trying something new. [synthesis]
Client wants to keep the fixed role or idealises it. Return to the purpose, which is learning to reconstrue. Withdraw the role on schedule and ask which parts they want to keep as their own. [source: Kelly 1955 on withdrawal] [synthesis: specifics]
Slot rattling / flip-flopping. The client jumps to the opposite pole, for example from doormat to aggressive, then back. Response: help build a new construct dimension (4.17) so change isn't limited to two poles. [source: Kelly 1955 on slot movement] [synthesis]
Dependency and transference. Kelly saw the client as trying out constructions of other people on the therapist. He distinguished transference that the therapist can use to help the client elaborate roles from preemptive, dependent transference that stalls change. Response: name the pattern hypothetically, use enactment and role exchange, and gradually widen the client's network of validating figures. [source: Kelly 1955 vol. 2] [uncertain: Kelly's exact terms; synthesis: response]
Ruptures. Read the rupture as the client invalidating your construction, or as you preempting theirs. Response: return to the credulous approach. Ask how the client saw what happened and own your part. Treat the repair as an experiment in a new kind of relationship. Leitner's experiential PCP stresses deep (ROLE) relating and managing therapeutic distance. [source: Leitner 1988 — uncertain] [synthesis]
Loose construing / thought disorder. Avoid loosening techniques. Prioritise tightening and serial validation within a wider psychiatric plan. [source: Bannister et al. 1975] [synthesis]
Clients who can't or won't engage with make-believe (e.g. very literal, or with a severe social anxiety response to performance). Use smaller enactments in session, pyramided concrete behaviours and graded experiments before or instead of a full fixed role. [synthesis]
Grid fatigue or intellectualising. Shorten grids and stop treating them as tests. Discuss the results conversationally. If the client talks only in abstractions, pyramid down to behaviour. [synthesis]
Cultural and linguistic fit. Elicit constructs in the client's own language and idioms. Don't impose role titles or contrasts from your culture. Adapt the fixed role's social situations to the client's context. [synthesis]
Therapist preemption. The therapist fixes on one view of the client, for example 'avoidant'. Response: apply reflexivity and constructive alternativism to yourself, generate at least one alternative formulation, and use supervision. [source: Kelly 1955 reflexivity] [synthesis]

Added in the verification pass:

Mismatched expectations of therapy. Kelly asked the therapist to stay alert to what the client expects from psychotherapy. The complaint itself shows what the client thinks therapy can do. Clients may construe the therapist as a parent, an absolver of guilt, a companion, a threat or a representative of reality, and may be "bitterly disappointed" when the therapist doesn't enact the expected role. Response: elicit expectations early, say which roles you can and can't play, and use the 'representative of reality' role to make sessions a safe place to experiment. [source: Bay Path textbook, citing Kelly 1955 vol. 2 pp. 575, 618] [synthesis: response]
Stereotyping the client by culture. Kelly held that a client should never be confined to the stereotype of their culture (vol. 2 p. 833). Response: use cultural knowledge as a source of hypotheses about shared construing (the commonality corollary), then check each one credulously with this client. [source: Bay Path textbook] [synthesis: response]

Sources

Sources.

Further reading (not drawn on): Encyclopaedia entries on tightening and the C-P-C cycle. All Kelly page numbers above are as cited by the sources listed.

Books and papers cited. Those marked [citation verified] appear in the reference list of a source page; the rest are not so marked.

Kelly, G. A. (1955). The Psychology of Personal Constructs (Vols 1–2). New York: Norton. Republished 1991, London: Routledge. [primary] [citation verified]
Kelly, G. A. (1964). The language of hypothesis: man's psychological instrument. Journal of Individual Psychology, 20, 137–152. Reprinted in Maher, B. (Ed.) (1969), Clinical Psychology and Personality: The Selected Papers of George Kelly. New York: Wiley. [primary] [uncertain: page numbers]
Hinkle, D. N. (1965). The change of personal constructs from the viewpoint of a theory of construct implications. Unpublished PhD thesis, Ohio State University.
Landfield, A. W. (1971). Personal Construct Systems in Psychotherapy. Chicago: Rand McNally.
Landfield, A. W., & Rivers, P. C. (1975). An introduction to interpersonal transaction and rotating dyads. Psychotherapy: Theory, Research and Practice, 12, 366–374. [citation verified]
Neimeyer, R. A. (1988). Clinical guidelines for conducting interpersonal transaction groups. International Journal of Personal Construct Psychology, 1, 181–190. [citation verified]
Bannister, D. (1962). The nature and measurement of schizophrenic thought disorder. Journal of Mental Science, 108, 825–842. [citation verified]
Bannister, D. (1963). The genesis of schizophrenic thought disorder: a serial invalidation hypothesis. British Journal of Psychiatry, 109, 680–686. [citation verified]
Bannister, D., Adams-Webber, J. R., Penn, W. I., & Radley, A. R. (1975). Reversing the process of thought disorder: a serial validation experiment. British Journal of Social and Clinical Psychology, 14, 169–180. [uncertain: citation details]
Tschudi, F. (1977). Loaded and honest questions. In D. Bannister (Ed.), New Perspectives in Personal Construct Theory. London: Academic Press.
Epting, F. R. (1984). Personal Construct Counseling and Psychotherapy. Chichester: Wiley.
Epting, F. R., & Nazario, A., Jr. (1987). Designing a fixed role therapy: issues, technique, and modifications. In R. A. Neimeyer & G. J. Neimeyer (Eds.), Personal Construct Psychotherapy Casebook. New York: Springer. [citation verified]
Kremsdorf, R. (1985). An extension of fixed-role therapy with a couple. In F. Epting & A. W. Landfield (Eds.), Anticipating Personal Construct Psychology. Lincoln: University of Nebraska Press. [citation verified]
Beail, N., & Parker, C. (1991). Group fixed role therapy: a clinical application. International Journal of Personal Construct Psychology, 4, 85–96. [citation verified]
Leitner, L. M. (1988). Terror, risk and reverence: experiential personal construct psychotherapy. International Journal of Personal Construct Psychology, 1, 251–261. [uncertain: citation details]
Winter, D. A. (1992). Personal Construct Psychology in Clinical Practice: Theory, Research and Applications. London: Routledge. [citation verified]
Fransella, F., & Dalton, P. (2000). Personal Construct Counselling in Action (2nd ed.). London: Sage.
Fransella, F., Bell, R., & Bannister, D. (2004). A Manual for Repertory Grid Technique (2nd ed.). Chichester: Wiley. [citation verified; the Encyclopaedia dates it 2003]
Crittenden, N., & Ashkar, C. (2012). The self-characterization technique: uses, analysis and elaboration. In P. Caputi, L. L. Viney, B. M. Walker & N. Crittenden (Eds.), Personal Construct Methodology (pp. 109–128). Chichester: Wiley-Blackwell. [citation verified]

← Previous · II-082Multimodal therapy Next · II-084 →Post-rationalist cognitive therapy