Metacognitive interpersonal therapy

Sources. This guide draws on the pages listed under Sources: the publisher's description and table of contents for the 2015 MIT manual and for the 2020/21 experiential book; two open-access MIT papers that describe procedures (Salvatore, Ottavi, Popolo & Dimaggio 2019, on the therapeutic relationship; Centonze et al. 2021, on imagery and body work); the MIT-G study abstracts; and the ISRCTN record of the adolescent group adaptation (MIT-GA). The manuals' full text is not drawn on, so the step-by-step detail inside each procedure is partly reconstructed. That detail is tagged [uncertain] or [synthesis]. A claim tagged [source: …] is supported by the named page.

1. Introduction

Version followed: individual MIT as structured in the treatment manual by Dimaggio, Montano, Popolo & Salvatore (2015; chapter structure checked on the publisher's page). The guide adds the experiential techniques and the updated decision-making procedure from Dimaggio, Ottavi, Popolo & Salvatore, Metacognitive Interpersonal Therapy: Body, Imagery and Change (Routledge, released 2020, copyright 2021), and the group format MIT-G (Popolo et al. 2018–2019; Inchausti et al. 2020). A related but distinct protocol from the Terzo Centro di Psicoterapia Cognitiva in Rome (Carcione, Semerari and colleagues; Carcione et al. 2019) orders treatment by metacognitive function. It is noted in section 2 but not followed. [source: Routledge and Taylor & Francis book pages; Carcione et al. 2019]

Metacognitive Interpersonal Therapy (MIT) is an integrative cognitive therapy for personality disorders. It was developed in Rome by Giancarlo Dimaggio, Raffaele Popolo, Giampaolo Salvatore, Antonella Montano, Paolo Ottavi and colleagues. It grew out of the Italian cognitivist work on metacognition, states of mind and interpersonal cycles (Semerari, Carcione, Dimaggio, Nicolò, Procacci; Dimaggio et al. 2007), which drew on Liotti's attachment-informed cognitivism. [uncertain: the Liotti lineage and the institutional history were not checked] The 2020 book and the MIT-G papers aim it especially at clients with strong emotional inhibition and over-regulation: socially inhibited, over-controlled personality disorders other than borderline. Published cases and the Spanish group pilot also include narcissistic, paranoid, dependent and mixed presentations. [source: Routledge page for Dimaggio et al. 2020; Popolo et al. 2018 abstracts; Inchausti et al. 2020 abstract; Centonze et al. 2021]

The model of change has three linked targets:

• Poor metacognition. The client struggles to recognise and name their own emotions and thoughts, to link events to thoughts and emotions, to tell their mental states apart from reality, to integrate conflicting self-images and to understand others' minds. They also lack mastery: using this knowledge to regulate distress and solve social problems. [source: Salvatore et al. 2019]
• Maladaptive interpersonal schemas. A schema is a procedural representation of how one's wishes will be met in relationships. It has three parts. First, a wish from a basic motive: social rank and appreciation, attachment, caregiving, autonomy and exploration, or group inclusion. Second, an expected response of the other. Third, a response of the self: a dominant pathogenic self-image (for example, unworthy) and a healthier self-image kept in the shadow (for example, deserving appreciation). Example: wish to be appreciated, expect criticism, feel unworthy and withdraw. In personality disorders, schemas fire easily and give rigid readings. They drive intense or over-regulated emotion and push behaviour that draws the feared response, which confirms the schema. [source: Salvatore et al. 2019; Centonze et al. 2021]
• Dysfunctional coping and regulation. Examples are avoidance, rumination, worry, perfectionism and emotional over-control. [source: chapter titles of Dimaggio et al. 2020, 'Metacognition, Coping and Emotional Regulation' and 'Techniques for Restructuring Attention and Treating Cognitive Coping Strategies'] [examples partly synthesis]

Change comes in three steps:

• The client first builds a shared formulation with the therapist, grounded in specific autobiographical episodes: what happened, what they felt, thought and wanted, and how the schema shaped it.
• The client then takes critical distance from the schema, recognising it as a subjective representation rather than reality.
• Finally the client contacts and strengthens healthier self-aspects, feels them in body, emotion and thought, and tests them in daily behaviour. MIT does not try to argue the client out of their beliefs. Instead it evokes experiences of a capable self, using imagery, two-chair work, role-play, body work and the therapeutic relationship.

[source: Salvatore et al. 2019; Centonze et al. 2021; publisher pages for Dimaggio et al. 2015 and 2020]

2. Framework

The 2015 manual sets out MIT as step-by-step formalised procedures in two broad stages: shared formulation of functioning, then promoting change. These rest on assessment and case formulation, building the alliance and drafting a therapy contract. The 2020 book presents an updated decision-making procedure that tells the therapist which experiential technique to use and when. Metacognitive work is sequenced. Self-reflectivity comes first: awareness of emotions, links between events, thoughts and emotions, and what symptoms mean for the self-image. Understanding others' minds follows once self-reflectivity is good. Mastery is built in parallel, starting with regulating bodily arousal. The stages overlap. [source: Routledge page for Dimaggio et al. 2015 (chapter titles); Routledge page for Dimaggio et al. 2020; Salvatore et al. 2019] The labels A0–B4 below are this guide's, not the manual's. [synthesis]

Phase A — Shared formulation of functioning

• A0. Assessment, alliance and therapy contract. Assess personality functioning and metacognition, build the alliance, and agree a therapy contract around goals the client cares about. [source: manual description and chapter 4 title]
• A1. Enrich autobiographical memory. Collect specific episodes (where, when, who, why) and help the client re-experience them rather than summarise. [source: manual chapter 7 title; Centonze et al. 2021]
• A2. Improve access to inner states. Help the client identify the emotions, thoughts, wishes and bodily reactions within each episode. When narratives are poor, work from the relationship and from the body in the room. When states cannot be reached, use mindfulness, guided imagery or body techniques. [source: Centonze et al. 2021; Salvatore et al. 2019]
• A3. Reconstruct cause-and-effect links and schemas. Link triggers, thoughts and emotions. Compare episodes to find the recurring wish, response of the other and response of the self, plus the coping that follows. [source: Centonze et al. 2021; Salvatore et al. 2019]
• A4. Agree the formulation. Change promotion begins once the client recognises and agrees with the formulation. [source: Centonze et al. 2021] Writing the formulation down together is this guide's suggestion. [synthesis]

Phase B — Change promotion

• B1. Promoting differentiation. Help the client see the schema as a representation, not reality, and take critical distance from it. [source: manual chapter 8; Salvatore et al. 2019]
• B2. Constructing new self-aspects. Access healthy parts, explore, increase agency and overcome avoidances. Healthy parts can be noted from the start, but this is the stage where they are built up. [source: manual chapter 9]
• B3. Understanding the other's mind, and integration. Develop a more nuanced view of others' minds, and integrate self-states into a coherent narrative. [source: manual chapter 10; Salvatore et al. 2019]
• B4. Treating symptoms and mastering relational problems. Address symptoms and cognitive coping such as worry and rumination with attention-restructuring techniques. Help the client use their new understanding to solve relational problems. Behavioural exploration and activation supports this throughout. [source: manual chapter 11; Dimaggio et al. 2020, chapters 8–9]
• Ending. Consolidation and relapse planning are this guide's addition; the sources do not describe an ending phase. [synthesis]
• Terzo Centro variant (not followed). The Carcione et al. (2019) protocol runs five phases, one per metacognitive function: monitoring, integration, differentiation, decentration (interpersonal cycles), and self-agency, meaning an autobiographical continuity. [source: Carcione et al. 2019]

Format and dose

• Individual MIT: weekly sessions. Duration varies. The Terzo Centro sample attended weekly for one year (Carcione et al. 2019), and a published MIT case ran 50 sessions over two years (Centonze et al. 2021). The sources do not state session length; 45–50 minutes is the usual assumption. [uncertain: session length]
• MIT-G (group): a short-term, manualised group that combines psychoeducation with experiential work, chiefly role-play of significant episodes. It was first developed for young adults with over-regulated personality disorders other than borderline, and runs for 16 sessions. [source: Popolo et al. 2018 abstracts; Inchausti et al. 2020] The adolescent adaptation MIT-GA (Logroño, Spain) gives the fullest public structure. It has at least two individual pre-group sessions (alliance, motivation, shared formulation of schemas), two psychoeducational group sessions, 16 weekly group sessions of about 120 minutes, individual midpoint and final sessions, and five parallel sessions for parents. [source: ISRCTN64626533] Whether adult MIT-G uses the same session length and pre-group sessions was not confirmed. [uncertain]

Criteria for moving on

• A1 to A2: the client can produce at least some specific, time-limited episodes rather than generalities.
• A2 to A3: in most episodes the client can name at least one emotion and its trigger.
• A3/A4 to B: the client recognises the schema in new episodes, ideally while they are happening or soon after, and agrees with the formulation.
• B1 to B2/B3: the client can say something like "that was my schema" before or during an episode, and can recall at least one healthy-self episode. Work on others' minds is pushed only once self-reflectivity is good.
• When to step back: move back a step if narratives become generic again, if metacognition collapses under arousal, or if the alliance ruptures.

At all times, risk and alliance ruptures take priority over the planned step.

[source: Centonze et al. 2021 (agreed formulation starts change promotion); Salvatore et al. 2019 (self-reflectivity before others' minds; repair before technique)] The other operational thresholds are this guide's. [synthesis]

3. Guiding principles

[source: Salvatore et al. 2019; Centonze et al. 2021; Routledge page for Dimaggio et al. 2020, as tagged per item] The one-line practical meanings are this guide's wording. [synthesis]

• Work from specific episodes, not generalities. If the client says "people always reject me", ask for one concrete time and place. [source: Centonze et al. 2021; Salvatore et al. 2019]
• Sequence the metacognitive work. Build self-reflectivity first. Ask about others' minds only once self-reflectivity is good. Build mastery in parallel, starting with bodily arousal. [source: Salvatore et al. 2019]
• Pitch interventions at the client's current level. Don't ask for decentration or alternative views while the client cannot yet name what they feel. [synthesis from the sequencing above]
• Formulate before changing. Start change promotion only once the client agrees the formulation. [source: Centonze et al. 2021]
• Make the formulation shared and in the client's language. Co-write it, check it, and revise it when new episodes contradict it. [source for "shared": manual chapter 7 title] [written format: synthesis]
• Validate the underlying wish. Wishes for appreciation, attachment, autonomy, care or inclusion are healthy. The problem lies in the expected response, the self-image and the coping. [source: Salvatore et al. 2019 (motives)] [emphasis: synthesis]
• Don't argue with beliefs; build alternative experience. Evoke a strong, capable self-image, and let the client feel it in body, emotion and thought. Then help them keep it in memory and test it in behaviour. [source: Salvatore et al. 2019]
• Promote agency. Show the client that they can influence their own inner states and act on their wishes, especially if they are inhibited and over-regulated. [source: Centonze et al. 2021; Routledge page for Dimaggio et al. 2020]
• Look for healthy parts from the start. The healthier self-image is present but in the shadow. Note episodes of competence, pleasure and closeness from session one. [source: Salvatore et al. 2019; manual chapter 9]
• Keep the client's goals central while pushing gently. Nudge toward change, but around what the client wants. [source: Routledge page for Dimaggio et al. 2020]
• Use the therapeutic relationship as data and as a tool. Expect the schema to be enacted with you. Keep asking yourself which "Other" you have become for the client. [source: Salvatore et al. 2019]
• Monitor your own states. Your emotions and action tendencies show which interpersonal cycle is active. Use them to step out of it. [source: Salvatore et al. 2019]
• Repair ruptures first. If a schema becomes active in the relationship, set the technique aside and repair the relationship through metacommunication. [source: Salvatore et al. 2019]
• Use self-disclosure as a technique. Share your own thoughts and feelings when this builds a shared climate and helps the client's own experience emerge. [source: Salvatore et al. 2019]
• Combine experiential and narrative work. Imagery, two-chair work, role-play and body work bring states into the room that talk alone does not reach. Follow every exercise with reflection. [source: Routledge page for Dimaggio et al. 2020; Centonze et al. 2021]
• Experiential work needs agreement. Explain the rationale and get consent. Before you go on, watch how the client reacts to closing their eyes or to feeling pressured. [source: Salvatore et al. 2019]
• Regulate before exploring. If arousal is too high or too low for reflection, bring it back into a workable range first, with mindfulness, breathing, grounding or, if needed, distraction. [source: Centonze et al. 2021]

4. Intervention inventory

The procedures below follow the order of the manual's stages. Technique names follow the chapter titles of the 2015 manual and the 2020 book, which were checked on the publisher's pages. The main components of the experiential techniques were checked in Salvatore et al. (2019) and Centonze et al. (2021). The numbered steps are this guide's reconstruction and paraphrase, not manual text; where a step order is not in those sources, treat it as [synthesis]. All example language is illustrative. A source line under each heading says what was checked.

Assessment aid. Teams in the MIT tradition assess metacognition with clinician-rated tools. The Terzo Centro study used the Metacognition Assessment Interview (MAI), which rates monitoring, integration, differentiation and decentration. [source: Carcione et al. 2019] The Metacognition Assessment Scale and its revision (MAS-R; Semerari, Carcione and colleagues) rate session transcripts for understanding one's own mind, understanding others' minds, decentration and mastery. [uncertain] The 2015 manual has an appendix of the diagnostic instruments usually used in MIT. [source: Routledge page] No items are reproduced here. Use these tools informally to judge which step the client can manage.

4.0 Building the alliance and drafting a therapy contract

[source: manual description ("forming an alliance, drafting a therapy contract"); MIT-GA pre-group sessions (ISRCTN64626533)] [steps: synthesis]

• Purpose: To engage clients who are often hard to engage, and to agree goals that matter to them, before and alongside formulation work.
• Steps:
  1. Ask what the client wants from therapy, in their own words, and what has stopped them getting it.
  2. Link the request to one or two concrete life goals, such as a relationship, work, or a pleasure the client wants back.
  3. Explain MIT simply: together you will look at specific episodes to understand what the client wants, what they expect from others and how they react.
  4. Agree the practicalities: frequency, expected length, how to raise disagreement with you, and what happens after missed sessions.
  5. Agree that experiential exercises will always be proposed and explained first, and done only with the client's consent.
  6. Revisit the contract when goals change or the alliance is strained.
• Example language (illustrative): "What would you like to be different a year from now? ... Let's start from that, and look at some recent moments where it went wrong."
• Watch for: Compliant agreement to goals the client thinks you want. Vague goals. It's working when the client states a goal of their own and brings episodes related to it.

4.1 Eliciting narrative episodes

[source: manual chapter 7, "Enriching Autobiographical Memory"; Centonze et al. 2021 (where, when, who, why; re-experience rather than summarise)] [step detail: synthesis]

• Purpose: To get specific, time-bounded memories rich enough to reflect on. This is the entry point for every later step. Use it from session one, and whenever talk turns abstract.
• Steps:
  1. When the client makes a generalisation, ask for one recent occasion.
  2. Fix the time and place.
  3. Ask who was present, what each person said and did, in order.
  4. Slow down at the peak moment: "what happened inside you then?"
  5. Ask how it ended and what the client did afterwards (coping).
  6. If the client cannot find an episode, ask about the last 24–48 hours, or use the client's reaction to the therapist in the room.
• Example language (illustrative): "You said your colleagues always ignore you. Can we take one time this week? Where were you, and who was there? What did he say, word for word as best you remember?"
• Watch for:
  • Intellectualising.
  • Summary narratives ("it's always like that").
  • Long scene-setting with no inner states.
  • It's working when the client volunteers episodes unprompted, with dialogue and feelings.

4.1a Working from the relationship and the body when narratives are poor

[source: Salvatore et al. 2019] [step order: synthesis]

• Purpose: To reach mental states and episodes with clients whose narratives are thin, when direct requests for episodes keep failing.
• Steps:
  1. Stop pressing for episodes after repeated attempts fail.
  2. Move to a relaxed, genuinely interested conversation about something the client cares about, such as their work, studies or an interest.
  3. Track bodily signals in the room: posture, breathing, facial expression.
  4. Name a signal tentatively.
  5. Ask what emotion and thought go with it.
  6. Ask whether it brings a specific moment to mind, and move into that episode (4.1).
• Example language (illustrative): "As you talk about your thesis, your face looks sad for a moment. What's that? ... Does it bring a particular moment to mind?"
• Watch for: The client feeling scrutinised, so keep observations tentative and warm. It's working when a concrete scene emerges from a state that showed in the session. In one published example, a sad expression during a chat about the client's field of study led to a memory of her grandfather, and then to a perfectionistic schema. [source: Salvatore et al. 2019]

4.2 Promoting monitoring (naming inner states)

[source: manual chapter 7, "Improving Access to Inner States"; Centonze et al. 2021 (emotions, thoughts, wishes, bodily reactions)] [step detail: synthesis]

• Purpose: To help clients, especially inhibited or alexithymic ones, identify the emotions, thoughts, bodily sensations and wishes inside an episode.
• Steps:
  1. Return to the peak moment of an episode.
  2. Ask about the body first if the client can't name feelings: where they felt it, and what kind of sensation.
  3. Offer tentative emotion labels as hypotheses, and let the client accept, reject or modify them.
  4. Separate the emotion from the thought and from the action tendency.
  5. Ask what the client wanted in that moment (the wish).
  6. Summarise in the client's words.
• Example language (illustrative): "When she said that, what did you notice in your body? Tight in the chest... If that tightness had a name, would it be closer to anger, fear, or shame? Or something else?"
• Watch for:
  • Labels the client accepts only to please you.
  • Retreat into "I don't know".
  • It's working when the client starts naming states spontaneously and with nuance, for example telling irritation apart from humiliation.

4.3 Linking states to triggers

[source: Centonze et al. 2021 (reconstructing cause-and-effect links); Salvatore et al. 2019 (event, thought and emotion links)] [step detail: synthesis]

• Purpose: To build understanding of what produced the state. Use it once some states can be named.
• Steps:
  1. Ask what happened just before the emotion rose.
  2. Look for the precise interpersonal trigger: a tone, a silence, a look.
  3. Ask what the client thought the other person meant.
  4. Write the trigger, the meaning and the emotion as a short chain.
• Example language (illustrative): "So the shame came the moment he paused before answering. What did you read into that pause?"
• Watch for:
  • The client attributing states to no cause ("it just comes").
  • Over-attributing everything to one person.
  • It's working when the client can tell you the trigger without being asked.

4.4 Recognising recurring patterns (schema reconstruction)

[source: manual chapter 7, "Reconstructing Schemas"; Salvatore et al. 2019 (wish, response of the other, response of the self)] [step detail: synthesis]

• Purpose: To identify the maladaptive interpersonal schema: wish, then expected response of the other, then response of the self, then coping. Use it after several monitored episodes.
• Steps:
  1. Put two or three monitored episodes side by side.
  2. Ask what they have in common.
  3. Identify the shared wish.
  4. Identify the response the client expected or perceived from the other.
  5. Identify the client's emotional and behavioural reaction.
  6. Note the coping that follows (for example withdrawing, over-working, ruminating).
  7. Explore, without forcing it, where the pattern may have been learned. Keep the focus on the present.
• Example language (illustrative): "In all three stories you wanted your effort to be recognised. Each time you expected to be judged inadequate, felt ashamed, and then worked late alone. Does that sound like you?"
• Watch for:
  • Premature or therapist-imposed formulations.
  • The client agreeing compliantly.
  • It's working when the client spots the pattern in a fresh episode.

4.5 Shared written formulation

[source: manual chapter 7, "Shared Formulation of Functioning"; Centonze et al. 2021] [written map and copy for the client: synthesis]

• Purpose: To make the schema explicit and portable, and to agree on goals.
• Steps:
  1. Draft the formulation together on paper or a whiteboard, using the client's phrases.
  2. Include the wish, the core self-images (negative and healthy), the expected other, the self-response and the coping.
  3. Add how this might show up between client and therapist.
  4. Derive goals from the formulation, for example acting on the wish despite fear.
  5. Give the client a copy, and revisit it periodically.
• Example language (illustrative): "Let's write this down in your words, so we both have the same map. Where would you put what you want most here?"
• Watch for:
  • Formulations that blame the client or others.
  • It's working when the client uses the map's terms between sessions.

4.6 Identifying and enriching healthy self-parts

[source: manual chapter 9, "Access to Healthy Parts"; Salvatore et al. 2019 (healthier self-image in the shadow; experiential contact with self-aspects free from pathology)] [step detail: synthesis]

• Purpose: To find self-states in which the client felt competent, connected or free. These are resources for change. Do this throughout, and in particular before differentiation.
• Steps:
  1. Ask for episodes in which the client did something they wanted, enjoyed something, or felt at ease with someone.
  2. Elicit them in the same detail as problem episodes, including bodily sensations.
  3. Name the self-image active then.
  4. Link it to the wish.
  5. Note it in the formulation.
• Example language (illustrative): "Tell me about a moment, however small, when you did something just because you liked it. What was that like in your body?"
• Watch for:
  • The client disqualifying positive episodes ("that doesn't count").
  • It's working when the client can call up healthy-self memories when needed.

4.7 Therapist self-monitoring of interpersonal cycles

[source: Salvatore et al. 2019 (the therapist keeps asking which "Other" they have become and monitors their own reactions)] [step detail: synthesis]

• Purpose: To detect when the therapist is being pulled into the client's schema, for example becoming critical, bored, protective or deferential.
• Steps:
  1. After or during the session, notice your emotion and your action tendency.
  2. Ask which schema role this matches.
  3. Choose to step out of it rather than act it out.
  4. Consider whether naming it would help the client.
  5. Take it to supervision.
• Example language (internal, illustrative): "I'm getting impatient and want to push them. That's the critical other they expect."
• Watch for: Repeated stuck sessions with the same feeling in you. It's working when you can pull out of the cycle and the session regains movement.

4.8 Therapist self-disclosure

[source: Salvatore et al. 2019 (technically oriented disclosure of the therapist's thoughts and feelings); Centonze et al. 2021 (case use)] [step detail: synthesis]

• Purpose: To validate, to normalise, or to correct the client's expected response of the other, through a brief and relevant disclosure.
• Steps:
  1. Choose disclosures that are short and in the client's interest.
  2. Usually disclose your here-and-now reaction ("I didn't feel bored, I felt moved").
  3. Occasionally disclose a similar life experience.
  4. Return the focus to the client at once.
  5. Check how it landed.
• Example language (illustrative): "You expected me to think you were weak for saying that. What I actually felt was respect."
• Watch for: Disclosure that serves the therapist, or that burdens the client. It's working when the client tests the expectation against what you actually said.

4.9 Metacommunication and rupture repair

[source: Salvatore et al. 2019 (metacommunication; withdrawal and confrontation ruptures after Safran & Muran); Centonze et al. 2021 (repair before experiential work)]

• Purpose: To repair alliance ruptures and to use them as live examples of the schema. MIT draws on Safran & Muran's model, which distinguishes withdrawal ruptures from confrontation ruptures.
• Steps:
  1. Notice the markers: the client goes vague or compliant (withdrawal), or criticises you (confrontation).
  2. Stop the planned work.
  3. Name what you noticed, without blame.
  4. Acknowledge your own contribution.
  5. Invite the client to explore what happened inside them.
  6. Link it to the formulation only if the client is ready.
  7. Agree on how to continue.
• Example language (illustrative): "I noticed you went quiet after my suggestion. I may have pushed too fast. What happened for you just then?"
• Watch for: Defensive explaining on your part. It's working when the client can express disagreement directly, and the session continues.

MIT-specific detail: when you share your own reaction during repair, first name what in the exchange set it off, and only then the feeling. Don't voice irritation or anger unless you first link it to what set it off, and never frame it as blame. [source: Salvatore et al. 2019]

4.10 Differentiation (critical distance from the schema)

[source: manual chapter 8, "Promoting Differentiation"; Salvatore et al. 2019 (critical distance; MIT does not try to refute beliefs)] [step detail: synthesis]

• Purpose: To help the client see the schema-driven expectation as one subjective reading, not reality. Use it once the shared formulation is agreed.
• Steps:
  1. In a fresh episode, find the moment the schema fired.
  2. Invite the client to label it as the schema ("the old map").
  3. Ask how the scene looked from inside the schema, and how it looks now from a calmer or healthier state. Prefer readings the client generates.
  4. Don't dispute the belief head-on. If the schema keeps its grip, move to experiential work that lets the client feel an alternative self-image (4.11–4.15), then test it in behaviour (4.18).
  5. Practise noticing the schema in real time between sessions.
• Example language (illustrative): "When he didn't reply, the map said 'he despises me'. Looking from the part of you that felt confident last week, what else could the silence mean?"
• Watch for: Intellectual agreement without any felt change. Moving to this step too early. The client feeling that their experience is being denied. It's working when the client says, during an episode, something like "that's my schema" and reacts less.

4.11 Guided imagery of an episode (re-experiencing for monitoring)

• Purpose: To bring an episode alive in the room, so that clients who report it flatly can reach their states. MIT also uses imagery to rehearse future situations and to stage memories or fantasies. [source: Salvatore et al. 2019; Centonze et al. 2021; Dimaggio et al. 2020, chapter 5 title]
• Steps:
  1. Explain the rationale, get consent and agree a stop signal. [synthesis]
  2. Open with a brief mindfulness or breathing exercise to settle attention. [source: Centonze et al. 2021]
  3. Invite the client to close their eyes, or to lower their gaze if closing them feels unsafe. Notice how they react to this before going on. [source: Salvatore et al. 2019]
  4. Have them relive the scene in detail, in the present tense: place, people, sensory details, what is said.
  5. Slow down at the key moment, and ask about body, emotion, thought and wish.
  6. If the client becomes dysregulated, use distraction or grounding until regulation returns. [source: Centonze et al. 2021]
  7. Come back to the room through grounding: feet on the floor, eyes open, describe the room.
  8. Reflect together on what the client noticed and on new links to the formulation. [source: Centonze et al. 2021 (reflection after each exercise)]
• Example language (illustrative): "Go back to that office. Tell it as if it's happening now. What do you see? He's speaking... what do you feel in your chest as he says it?"
• Watch for: Dissociation, flooding or numbness: stop and ground the client. Going along with the exercise out of shame or fear of disappointing you. If a schema activates in the relationship, set the technique aside and repair (4.9). [source: Salvatore et al. 2019] It's working when the client reaches emotions they couldn't reach in plain narration.

4.12 Imagery rescripting

• Purpose: To change the emotional meaning of a schema-laden scene by activating a healthy self-part. The aim is to change the scene's meaning, not the past. The client relives a scene and then changes its course. MIT uses it for childhood memories where the schema was learned and for recent episodes. Use it once the formulation is agreed. [source: Centonze et al. 2021; Salvatore et al. 2019; Dimaggio et al. 2020, chapter 5 title]
• Safety and consent: Assess trauma history and dissociation first. Avoid starting with the most severe trauma. Agree a stop signal and leave time for grounding. For clients with severe trauma, consider trauma-focused protocols or specialist referral. [synthesis]
• Steps:
  1. Choose an episode central to the schema.
  2. Have the client re-enter it in imagery, as in 4.11, until the core feeling is active.
  3. Ask what was needed in that moment.
  4. Bring in a change. Options: the client's healthy or adult self enters the scene, the therapist supports the client's healthy part, or the client says and does what they wanted. [source for a healthy part voiced with the therapist's support: Salvatore et al. 2019] [the full menu: synthesis]
  5. Use the body: have the client adopt a different posture and tone of voice for the new response, for example upright and grounded. [source: Centonze et al. 2021]
  6. Check the felt shift in body and emotion, and let the client stay with it.
  7. Reflect afterwards, and link the new experience to differentiation and to behavioural exploration. [source: Centonze et al. 2021 (reflection phase)]
• Example language (illustrative): "What did that younger you need right then? ... Can the you of today step in? Stand the way she would stand. What do you want to say?"
• Watch for: Rescripts that stay intellectual, or that turn punitive or violent beyond the client's values. Overwhelm. It's working when distress drops, the client feels relief or strength, and the new self-image carries into later sessions. In one published MIT case, rescripting a childhood memory of being lost was followed by less social anxiety and procrastination. [source: Centonze et al. 2021]

4.13 Chair work

• Purpose: To externalise the dialogue between self-states, or between the self and a significant other. Examples are the inner critic against the vulnerable self, or the client speaking to the person who triggers the schema. [source: Dimaggio et al. 2020, chapter 6 "Drama Techniques: Two Chairs, Role-Play and Enactment", and the publisher's description] [step detail: synthesis from general two-chair practice; how closely it matches the book is uncertain]
• Steps:
  1. Set up two chairs.
  2. Have the client voice one part from one chair, for example the critic saying "you're pathetic".
  3. Switch chairs, and ask the attacked part how it feels.
  4. Let it respond and express its need.
  5. Bring in the healthy self, if needed, to set limits on the critic.
  6. Process the experience afterwards.
• Example language (illustrative): "Sit here and be the voice that tells you you're not good enough. Speak to him directly. ... Now come back to this chair. Hearing that, what do you feel?"
• Watch for:
  • Embarrassment. Normalise it, or start with lower-intensity versions such as imagining the other person in the chair.
  • A critic who escalates unchecked.
  • It's working when the vulnerable part gains a voice and the critic softens or is limited.

4.14 Role play and enactment (including drama-informed exercises)

[source: Dimaggio et al. 2020, chapter 6; Salvatore et al. 2019 (role-play and dramaturgical techniques: the client enacts scenes with the therapist); ISRCTN64626533 (role-play of personal episodes; problem-solving role-play for metacognitive mastery)] [step detail: synthesis]

• Purpose: To rehearse new interpersonal behaviour that serves the wish, and to experience the schema enacted live. Used in individual work and in MIT-G.
• Steps:
  1. Pick a target situation from the formulation.
  2. Act it out first as it usually goes.
  3. Swap roles so that the client plays the other person.
  4. Discuss how the other person might really be feeling (decentration).
  5. Rehearse a new response that acts on the wish.
  6. Process the bodily and emotional reactions.
  7. Plan a real-life attempt.
• Example language (illustrative): "Let's try it. I'll be your boss, and you ask for the day off the way you'd really like to."
• Watch for: Performance anxiety or perfectionism. Keep the stakes low. It's working when the client spontaneously varies the response and reports less fear.

4.15 Embodied, body-oriented techniques

• Purpose: To use the body to reach inner states, to regulate arousal, and to activate and anchor alternative self-states. Body interventions have their own chapter in the 2020 book. [source: Dimaggio et al. 2020, chapter 7 title; Salvatore et al. 2019; Centonze et al. 2021]
• Steps:
  1. Track bodily cues in session (posture, breathing, facial expression) and name them tentatively. Then ask for the linked emotion and thought, and then for an episode. [source: Salvatore et al. 2019]
  2. To centre the client, use diaphragmatic or slow deep breathing. [source: Salvatore et al. 2019; Centonze et al. 2021]
  3. For grounding, invite an upright standing posture with the legs slightly apart and the knees bent. [source: Centonze et al. 2021]
  4. During imagery or rescripting, try out a different posture and tone of voice that fit the healthier self-state. Ask what changes in emotion and thought. [source: Centonze et al. 2021]
  5. Link the agentic bodily state to a real situation, and practise it as homework. [synthesis]
• Example language (illustrative): "Your breathing seems to have sped up. ... What happens if you stand, feet a little apart, knees soft, and breathe out slowly? Anything shift?"
• Watch for: Trauma-related body memories. The client feeling exposed or watched. Go ahead only with consent and at the client's pace. It's working when the client reports a felt shift and can recreate it.

4.16 Mindfulness and attention-regulation exercises

• Purpose: To regulate arousal, interrupt rumination and narrative flow, and help the client observe states without being fused with them. MIT uses an interpersonal ("relationally integrated") mindfulness taken from the sensorimotor approach: attention goes to bodily experience, in the presence of the therapist. Mindfulness also opens experiential exercises. [source: Salvatore et al. 2019; Centonze et al. 2021; Routledge page for Dimaggio et al. 2020] [uncertain: the exact practices and their length]
• Steps:
  1. Introduce a brief practice: attention to the breath, to posture, or to sensations happening now.
  2. If the client starts ruminating or telling a story, gently bring their attention back to what the body is doing now.
  3. Debrief together what was noticed.
  4. Link it to schema recognition ("there's the old map, as a thought"). [synthesis]
  5. Agree a short daily practice if useful. [synthesis]
• Example language (illustrative): "Let's pause the story for a moment. Notice your feet on the floor and your breath... What do you notice right now, here with me?"
• Watch for: Practice used as avoidance. Trauma clients who find a body focus distressing. Shame at being observed. It's working when the client spends less time ruminating and notices states sooner.

4.17 Regulating dysfunctional coping

[source: Dimaggio et al. 2020, chapter 9 "Techniques for Restructuring Attention and Treating Cognitive Coping Strategies"; the publisher's description (steering attention away from rumination); Salvatore et al. 2019 (attention-shifting to reduce symptoms)] [step detail: synthesis]

• Purpose: To reduce coping that maintains the schema, such as worry, rumination, perfectionism, avoidance, over-control and substance use.
• Steps:
  1. Identify each coping strategy in the formulation.
  2. Explore its short-term function and its long-term cost.
  3. Help the client decide to reduce it.
  4. Agree on a concrete alternative. Examples: postponing rumination to a set time, a "good enough" standard, or approaching instead of avoiding.
  5. Monitor it between sessions.
• Example language (illustrative): "Re-checking the report calms you for a moment. What does it cost you over a week?"
• Watch for: Covert substitution of one coping strategy for another. It's working when the time spent on coping falls and the wish becomes more reachable.

4.18 Behavioural exploration and activation (behavioural experiments and exposure)

[source: Dimaggio et al. 2020, chapter 8 "Behavioural Exploration and Activation"; manual chapter 9 (exploration, increase in agency, overcoming avoidances); Salvatore et al. 2019 (bodily or exposure exercises, which may activate schemas)] [prediction and belief-rating format: synthesis from general CBT practice]

• Purpose: To test the schema's expectation in real life and to act on wishes, such as asking for something, expressing an opinion, or pursuing an interest. Use it in change promotion, often following imagery or role play.
• Steps:
  1. Choose a wish-driven action.
  2. Write down the predicted response of the other, with a belief rating.
  3. Plan the action concretely, and anticipate safety behaviours to drop.
  4. Carry it out.
  5. Record what actually happened and how the client felt.
  6. Compare the result with the prediction.
  7. Update the formulation, and grade the next step.
• Safety and consent: Grade the difficulty. The client chooses the steps. Avoid experiments that are unsafe in the client's real relationships, for example with abusive partners. [synthesis]
• Example language (illustrative): "Your map predicts that if you say you disagree in the meeting, they'll look down on you. How sure, 0–100? Shall we test it with a small disagreement?"
• Watch for:
  • Discounting a good outcome ("they were just being polite").
  • Hidden safety behaviours.
  • It's working when the belief rating drops and the client initiates new experiments.

4.19 Promoting decentration (understanding others' minds)

[source: manual chapter 10, "Promoting the Understanding of the Other's Mind"; Salvatore et al. 2019 (only once self-reflectivity is good)] [step detail: synthesis]

• Purpose: To help the client understand others' minds as separate from their schema. Use it later, once self-reflection is stable.
• Steps:
  1. In an episode, ask what the other person might have been thinking or feeling, apart from the client.
  2. Generate several possibilities.
  3. Look for information about the other's own situation.
  4. Use role reversal (4.14) if needed.
• Example language (illustrative): "Setting aside what it meant to you, what might have been going on for her that day?"
• Watch for: Premature use, which clients can feel as invalidating. It's working when the client offers non-self-referential explanations.

4.20 Integration and mastery (consolidation, relapse prevention)

[source: manual chapters 10–11 (integration; mastery of relational problems)] [relapse planning and tapering: synthesis]

• Purpose: To build a coherent self-narrative across states, and to help the client become their own MIT therapist.
• Steps:
  1. Review the formulation and the changes.
  2. Have the client write their own account of the schema, the healthy self, and the strategies that work.
  3. Identify high-risk situations.
  4. Plan the steps to take when the schema fires (notice, name, ground, choose an action consistent with the wish).
  5. Taper sessions, and offer booster sessions.
• Example language (illustrative): "If this pattern shows up next year, what will you notice first, and what will you do?"
• Watch for: Fear of ending, which itself activates the schema, so address it explicitly. It's working when the client handles new episodes independently.

4.21 Group format (MIT-G) components

• Purpose: To deliver MIT as a short-term, manualised group that combines psychoeducation with experiential work, chiefly role-play of significant episodes, to improve metacognition and social problem-solving. It was developed for young adults with over-regulated, socially inhibited personality disorders, and has also been piloted across the range of personality disorders in Spanish public services. [source: Popolo et al. 2018 abstracts; Inchausti et al. 2020; Dimaggio et al. 2020, chapter 10 title]
• Structure, as published for the adolescent adaptation MIT-GA:
  1. At least two individual pre-group sessions: alliance, motivation, assessment, a draft shared formulation of the client's schemas, and affect-regulation strategies.
  2. Two psychoeducational sessions introducing the treatment principles.
  3. Sixteen weekly group sessions of about 120 minutes, in blocks of two to three sessions. Each block covers one motive: social rank and competition, group inclusion, attachment, caregiving, exploration, sexuality and cooperation.
  4. Within each block: plain-language psychoeducation on the motive; film or cartoon clips showing it; each member writes an autobiographical memory linked to it; role-play of one member's episode, followed by group reflection on mental states and on verbal and nonverbal cues. In the second session on each motive, a problem-solving role-play aimed at metacognitive mastery.
  5. An individual midpoint session (group issues, alliance ruptures) and a final session (review of strategies learned, with a reformulation letter).
  6. For adolescents, five parallel sessions for parents.
• [source: ISRCTN64626533] [uncertain: how far adult MIT-G follows the same order and timing]
• Delivery: The adult studies ran 16 sessions. The number of facilitators and the group size were not confirmed. [source: Inchausti et al. 2020; Popolo et al. 2018] [uncertain]
• Watch for: Members' schemas triggering each other, for example criticism or rejection within the group. Manage this through metacommunication at group level and in the midpoint individual session. [synthesis; midpoint session from ISRCTN64626533]

5. Handling problems

The MIT sources address some of these problems: thin narratives and poor monitoring, ruptures, the therapist being drawn into the cycle, and dysregulation or shame during experiential work. Those items carry their source. Everything else is this guide's own guidance. [synthesis]

• The client cannot produce episodes, or speaks only in generalities. [source: Salvatore et al. 2019; Centonze et al. 2021]
  • Lower the demand: ask about yesterday, or use what is happening in the room now.
  • Stop pushing for scenes. Shift to a relaxed conversation about something the client cares about, track bodily signals and name them, then link them to a concrete scene (4.1a).
  • When states still can't be reached, use mindfulness, guided imagery (4.11) or body techniques (4.15).
  • Accept fragments, and show interest in any specific detail. [synthesis]
• Very poor monitoring (alexithymic, over-regulated clients). [source: Salvatore et al. 2019 (self-reflectivity first; begin mastery with bodily arousal)]
  • Stay longer in Phase A.
  • Offer emotion labels as hypotheses.
  • Use bodily cues (4.15) and brief mindfulness (4.16).
  • Don't move to differentiation or to others' minds yet.
• Stuck: the formulation is agreed but nothing changes. [synthesis, consistent with MIT's preference for experiential activation over argument (Salvatore et al. 2019)]
  • Check whether agreement is only intellectual. If it is, move to experiential work (imagery, rescripting, two chairs, role-play).
  • Check whether healthy-self episodes have been found. If not, look for them.
  • Check whether coping such as rumination or avoidance is still running unaddressed.
  • Check whether therapy itself has become a safe avoidance, and agree a behavioural experiment.
• Resistance as the schema enacted. [source: Salvatore et al. 2019]
  • Treat compliance, withdrawal or arguing as the schema playing out with you, and ask which "Other" you have become.
  • Notice your own reaction first. In one published case the therapist noticed his irritation at the client's dramatic help-seeking, reframed it to himself as a test, validated the suffering and then elicited a scene.
  • Metacommunicate (4.9), name the cycle, and validate the wish underneath. For example, a wish for autonomy may lie behind refusing homework. [example: synthesis]
• Withdrawal ruptures (lateness, absences, less communication, over-ready agreement). [source: Salvatore et al. 2019]
  • Name the shift gently.
  • Own your part, and invite disagreement.
  • Slow the pace. [synthesis]
• Confrontation ruptures (open disagreement or criticism of you). [source: Salvatore et al. 2019]
  • Don't defend.
  • Acknowledge what is valid in the criticism.
  • Explore what the client felt and expected.
  • Link it to the formulation only after the alliance has settled. [synthesis]
• Overwhelm or dissociation during imagery or chair work. [source: Centonze et al. 2021 (distraction until regulation returns; mindfulness to regulate arousal); Salvatore et al. 2019 (check reactions to closed eyes and to pressure)] [the rest: synthesis]
  • Stop the exercise and have the client open their eyes. Ground them: name things they can see, feel their feet, slow their breath. Use distraction if needed until regulation returns.
  • Next time, lower the intensity: eyes open, the scene seen on a screen, the observer perspective, or a less charged episode.
  • Before such work, screen for dissociation and trauma, and agree stop signals.
• Non-adherence to homework or experiments. [synthesis]
  • Explore what not doing it means, often perfectionism or fear of being judged by you.
  • Shrink the task, and design the next one together.
  • Treat a non-attempt as a new episode to explore, not a failure.
• The therapist is drawn into the cycle (critical, rescuing, overprotective, bored). [source: Salvatore et al. 2019 (including a therapist who recognised his own overprotective pull during imagery)]
  • Self-monitor (4.7), take it to supervision, and consider disclosing your reaction transparently (4.8).
• The client feels invalidated by differentiation ("you're saying I'm imagining it"). [synthesis, consistent with MIT not refuting beliefs (Salvatore et al. 2019)]
  • Validate the real elements of the episode first.
  • Make clear that the aim is to look at several readings, not to deny the client's experience.
  • Return to monitoring, or to experiential work.
• Shame or compliance during experiential exercises. [source: Salvatore et al. 2019 (a rupture during imagery when the client's shame and compliance surfaced, repaired by metacommunication)]
  • Pause the technique and metacommunicate about what is happening between you.
  • Normalise the reaction, start small (eyes open, short scenes), and give the client control over the pace. [synthesis]
• Suicidality, self-harm or crisis. [synthesis]
  • Pause the MIT steps, and assess risk following current standard practice.
  • Build a safety plan, including lethal-means counselling on safe storage and removal.
  • Increase contact as needed, and coordinate with psychiatric care. Resume MIT work once the client is stable.
• Comorbid substance misuse or a severe symptom disorder. [synthesis; symptom work has its place in manual chapter 11]
  • Address symptoms in parallel.
  • Frame substance use as coping in the formulation.
  • Refer for specialised care when needed.
• Group-specific problems (MIT-G). [synthesis; midpoint individual session from ISRCTN64626533]
  • When members trigger each other's schemas, the facilitators metacommunicate at group level and model repair.
  • Use the individual midpoint session to address group issues and ruptures, and add individual sessions for members who become overwhelmed.
• Risk of premature ending. [synthesis]
  • Clients with avoidant or paranoid schemas may drop out when they feel exposed. Name this possibility early, and include it in the formulation as a predicted pattern.

Sources

Sources:

• Carcione et al. 2019, Metacognition as a predictor of improvements in personality disorders, Frontiers in Psychology
• Inchausti et al. 2020, MIT in group for personality disorders: pilot study in a public mental health setting, J Contemp Psychother (abstract page)
• Centonze, Inchausti, MacBeth, Popolo & Dimaggio 2021, Changing embodied relational patterns in Metacognitive Interpersonal Therapy, Papeles del Psicólogo (PDF)
• Salvatore, Ottavi, Popolo & Dimaggio 2019, La relazione terapeutica nella terapia metacognitiva interpersonale, Quaderni di Psicoterapia Cognitiva (PDF)
• Popolo et al. 2018, Metacognitive interpersonal therapy in group: a feasibility study (DOAJ record)
• Popolo et al. 2018, MIT in groups for over-regulated personality disorders: a single case study (University of Edinburgh research portal)
• ISRCTN64626533, Metacognitive Interpersonal Group Therapy for Adolescents (MIT-GA) with personality disorders (trial record PDF)
• Routledge: Metacognitive Interpersonal Therapy for Personality Disorders: A treatment manual (description and contents)
• Taylor & Francis eBook page for the 2015 manual
• Routledge: Metacognitive Interpersonal Therapy: Body, Imagery and Change (description and contents)

Further reading (not drawn on): the Research in Psychotherapy article page and DOI landing page and the Europe PMC full text of PMC12420030.

Books and papers cited (books are not drawn on in full; their contents are as described on the pages above):

• Dimaggio, G., Semerari, A., Carcione, A., Nicolò, G., & Procacci, M. (2007). Psychotherapy of Personality Disorders: Metacognition, States of Mind and Interpersonal Cycles. Routledge. [uncertain]
• Dimaggio, G., Montano, A., Popolo, R., & Salvatore, G. (2015). Metacognitive Interpersonal Therapy for Personality Disorders: A Treatment Manual. Routledge. [description and contents checked]
• Dimaggio, G., Ottavi, P., Popolo, R., & Salvatore, G. (2020; copyright 2021). Metacognitive Interpersonal Therapy: Body, Imagery and Change. Routledge. [description and contents checked]
• Carcione, A., Riccardi, I., Bilotta, E., et al., & Semerari, A. (2019). Metacognition as a predictor of improvements in personality disorders. Frontiers in Psychology, 10, 170.
• Centonze, A., Inchausti, F., MacBeth, A., Popolo, R., & Dimaggio, G. (2021). Changing embodied relational patterns in Metacognitive Interpersonal Therapy. Papeles del Psicólogo, 42(1), 21–32. doi:10.23923/pap.psicol2020.2936
• Salvatore, G., Ottavi, P., Popolo, R., & Dimaggio, G. (2019). La relazione terapeutica nella terapia metacognitiva interpersonale. Quaderni di Psicoterapia Cognitiva, 45.
• Inchausti, F., Moreno-Campos, L., Prado-Abril, J., Sánchez-Reales, S., Fonseca-Pedrero, E., MacBeth, A., Popolo, R., & Dimaggio, G. (2020). Metacognitive Interpersonal Therapy in group for personality disorders: preliminary results from a pilot study in a public mental health setting. Journal of Contemporary Psychotherapy, 50(3), 197–203.
• Popolo, R., MacBeth, A., Brunello, S., et al., & Dimaggio, G. (2018). Metacognitive interpersonal therapy in group: a feasibility study. Research in Psychotherapy: Psychopathology, Process and Outcome, 21(3). doi:10.4081/ripppo.2018.338
• Popolo, R., MacBeth, A., Canfora, F., et al., & Dimaggio, G. (2018). Metacognitive interpersonal therapy in groups for over-regulated personality disorders: a single case study. Journal of Contemporary Psychotherapy. doi:10.1007/s10879-018-9401-3
• Popolo, R., MacBeth, A., et al. (2019; online 2018). MIT-G for young adults with personality disorders: pilot randomised controlled trial. Psychology and Psychotherapy: Theory, Research and Practice. doi:10.1111/papt.12182 [DOI from the Edinburgh portal; exact title uncertain]
• Semerari, A., Carcione, A., Dimaggio, G., et al. (2003). How to evaluate metacognitive functioning in psychotherapy? The Metacognition Assessment Scale and its applications. Clinical Psychology & Psychotherapy. [uncertain]
• Safran, J. D., & Muran, J. C. (2000). Negotiating the Therapeutic Alliance. Guilford. (The rupture model MIT draws on, as cited by Salvatore et al. 2019.)