Interpersonal psychoanalysis and detailed inquiry

1. Introduction

Version followed: Harry Stack Sullivan's interview structure and technique in The Psychiatric Interview (1954, lectures edited posthumously by Perry and Gawel), with his theory of anxiety and the self-system from The Interpersonal Theory of Psychiatry (1953). Clara Thompson's interpersonal view of transference and countertransference is added. Edgar Levenson's The Fallacy of Understanding (1972) and The Ambiguity of Change (1983) supply the later William Alanson White Institute (WAWI) model: frame, then detailed inquiry, then the same pattern enacted in the room. This is not Klerman and Weissman's Interpersonal Psychotherapy (IPT), which is a separate, manualised treatment.

What it is. Interpersonal psychoanalysis is a clinical tradition and method, not a manualised therapy. No session-by-session manual exists. What the founders describe is (a) a structured way of interviewing (Sullivan), (b) a stance, participant observation, and (c) a style of inquiry that asks for concrete detail about interactions, in the patient's life and between patient and therapist. Applications below that go beyond what the authors wrote are marked [synthesis]. [source: Sullivan 1954, Ch. 1–3; Levenson 1989; WAWI syllabus 2019]

How change happens. For Sullivan, anxiety in relationships drives "security operations". These are manoeuvres that protect self-esteem but blur what the person notices about their own dealings with others. Selective inattention is the main one. Change comes when a careful, anxiety-regulated inquiry with an expert makes those patterns of living clear. The aim is the patient's "success in living" [source: Sullivan 1954, Ch. 2]. For Levenson, understanding on its own is not what cures. Change comes when the therapist gets caught up in the patient's pattern as it repeats in the room, notices this, and the two of them work their way out of it together [source: Levenson 1972/1983, via Greenberg 1987 and Ingram 1985 review records and the WAWI syllabus; uncertain]. As Levenson put it, "To talk with someone is to behave with him" [source: Levenson 1989].

2. Framework

There are two frameworks. One is nested inside the other.

A. Sullivan's four stages of the interview. These apply to a single interview, to an assessment over several sessions, and, in rough form, to a whole course of treatment. Sullivan calls them hypothetical stages that help the interviewer use the time. [source: Sullivan 1954, Ch. 2]

Session count and format. Sullivan gives no fixed number. One interview may be a single consultation or the start of intensive psychotherapy [source: Sullivan 1954, Ch. 2]. In WAWI-tradition interpersonal psychoanalysis, treatment is open-ended, often at 1–4 sessions a week, and may be face-to-face or on the couch [uncertain: practice convention].

B. Levenson's process within treatment (WAWI model). [source: Levenson 1983, The Ambiguity of Change; structure reported from secondary literature; uncertain]

1. Frame. Set the arrangements: time, fee, frequency, how you will work. Then notice how the patient relates to these arrangements, and how you do. Experiences around the frame are themselves data.
2. Detailed inquiry. Find out what happened: the particulars of events, who said and did what, in what order, with what feeling. Do not jump to what it means.
3. Enactment / working through in the transference–countertransference. Notice how the pattern the patient has described is happening between the two of you right now ("what's going on around here?"). Use that to work free of it.

These are not strictly one after another. Phases 2 and 3 repeat in cycles. Move from 2 to 3 when the reported pattern starts to show up in your own participation, for example when you feel pulled, bored, irritated, rescuing, or interrogating [synthesis].

3. Guiding principles

[source: Sullivan 1954, Ch. 1–3 unless tagged otherwise]

• Participant observation. "The data of psychiatry arise only in participant observation" (Sullivan 1954). → You cannot observe from outside the relationship. Include your own behaviour in what you observe.
• The self as instrument. The interviewer's "principal instrument of observation is his self" (Sullivan 1954). → Track your own reactions during the session and treat them as data, not as noise.
• Expert role. The therapist is an expert in interpersonal relations. They use that knowledge for the patient and do not seek personal satisfaction or prestige from the patient. → Stay task-focused. Do not chat for your own enjoyment or seek admiration.
• Expertise is shown in the here and now. A referral gets the patient into the room. It does not make you an expert to them. → Earn credibility in the session through accurate, respectful, relevant questions.
• The patient must gain something. If the patient senses no benefit, the interview fails. → In every session, give something back: a clarification or a summary.
• Respect for security. Respect for the patient's worth and sensitivity to their security are core to expertise. → Before a sensitive question, give a small warning. Help the patient out when they lack information.
• Anxiety is necessary but must be dosed. Some anxiety is needed for an interview to become "psychiatric". Acute anxiety makes the interview go vague. → Approach anxious areas, then move on to reassuring material. Back off when answers lose detail.
• The interview is mainly vocal communication. → Listen to tone, rate, and hesitation as much as to content.
• Mixed motives are expected. Patients come with mixed motives. Reluctance is data, not a fault. → Inquire into reluctance. Do not argue with it.
• Entitlement to relevant data. The expert is "entitled to" relevant information and discourages irrelevance. → Ask plainly. Cut off trivia and repetition once you have what you need.
• No psychiatric banalities. Avoid clichés and jargon (Sullivan gives "mother-fixation" as an example). → Use plain words that are specific to this person.
• No tacit agreement with error. Do not let a delusion or a serious misstatement pass through silence. → Gently note where you disagree or are unsure.
• Parataxic distortion. The patient may be relating to an imagined person rather than to you. → Ask yourself who you are being treated as.
• One-genus postulate. "We are all much more simply human than otherwise" [source: Sullivan 1953; widely cited; uncertain wording]. → Approach every difficulty as a human pattern you could share, not as something alien.
• Particulars over meaning (Levenson). Pursue "what happened" before "what it means". Concrete inquiry draws transference and countertransference into view [source: Levenson 1989; WAWI syllabus lists Levenson 1988, "The pursuit of the particular"]. → Ask for the specific scene before you interpret.
• The patient's perceptions are plausible (Levenson). The patient is taken to distort experience in trying to grasp reality, not out of fantasy [source: WAWI syllabus 2019 paraphrasing Levenson]. Accepting the patient's view is a matter of conviction, not of nurturing them [source: Levenson 1989]. → Check what is accurate in the patient's reading of you or of others before you call it distortion.
• The analyst is a real person (Levenson; Thompson). "We have a real person in the room with us" [source: Levenson 1989]. Thompson treated transference as reactions to the analyst's real attributes as well as to past figures [source: Thompson 1945, 1950; uncertain]. → Expect the patient to notice real things about you. Acknowledge them where it helps.
• Understanding is not the cure (Levenson). Insight alone is a "fallacy". What matters is the experienced change in the relationship [source: Levenson 1972 title and thesis; uncertain]. → Do not stop at explaining. Look for the pattern repeating between you and work free of it.
• Influence provokes refusal. Any effort to influence provokes resistance. A patient given working space largely cures herself [source: Levenson & Slomowitz 2018, publisher description]. → Avoid pushing advice. Make room for the patient.

4. Intervention inventory

This section lists every technique Sullivan, Thompson and Levenson describe in the sources consulted. Sullivan's techniques are mostly interviewing operations rather than named exercises. Interpersonal psychoanalysis uses no worksheets, homework protocols or structured exercises. All example language is illustrative unless it is a quoted fragment with a citation. [source: Sullivan 1954, Ch. 1–3; Ch. 4–9 known only from chapter headings; Levenson 1989]

4.1 Formal inception

• Purpose / when: Opening of the first contact, and of any consultation. It establishes the expert–client relationship and a legitimate reason for meeting. [source: Sullivan 1954, Ch. 2–3]
• Steps: (1) Receive the person formally: greet them, introduce yourself, seat them. (2) Ask about, or refer to, the circumstances of their coming. (3) State briefly what you already know, such as referral information, so they can correct it. (4) Establish a reason for the meeting that both of you accept. (5) From the first moment, notice the impression you are making and the patient's effect on you.
• Example language (illustrative): "Dr. Ahmed told me you've been having trouble at work since the reorganisation, and that sleep has been bad. That's all I know. Is that roughly right, or would you put it differently?"
• Watch for: What the patient corrects, and how. That is early data on their security operations. Note also any expectation of a "miracle", which Sullivan warns makes the work harder. It is working if the patient adds to or corrects your account and agrees on a reason to continue.

4.2 Reconnaissance

• Purpose / when: After the inception. Get a rough social and personal history: who this person is and how they came to be where they are. This is not a full history. Sullivan says skill at this stage often decides how easy the detailed inquiry will be. [source: Sullivan 1954, Ch. 2; Ch. 4 not drawn on]
• Steps: (1) Ask ordinary identifying questions: age, place in the family, siblings, parents' occupations, schooling, work, marriage or partnership, children. (2) Keep the questions natural and conversational, not a form. (3) Keep it brief. Note areas to return to rather than chasing them now. (4) Form tentative hypotheses about the patient's main patterns of living. (5) At the end, Sullivan is reported to offer a summary statement for the patient to correct [uncertain: secondary report only; Ch. 4 not drawn on].
• Example language (illustrative): "Before we go into the problem itself, I'd like a quick sketch of your life: who was in the family, where you grew up, what you've done since. Just the outline for now."
• Watch for: Gaps, blurred periods, and abrupt changes of tone. These are leads for the detailed inquiry. It is working if a sketch emerges and you can state two or three hypotheses.

4.3 Detailed inquiry (Sullivan)

• Purpose / when: The main body of the work. It tests and refines your hypotheses by asking for the particulars of the patient's interpersonal experience, past and present. [source: Sullivan 1954, Ch. 2; WAWI syllabus 2019 (Ch. 5 assigned); Kanter 2013. Ch. 5 itself not drawn on]
• Steps: (1) Pick an area linked to a hypothesis or to the presenting difficulty. (2) Ask for a specific episode, not a generalisation. (3) Go through it in sequence: who was there, what was said and done, what happened next, what the patient felt and did at each point. (4) Note where the account goes vague, jumps ahead, or turns to abstraction. These are likely points of anxiety and security operations (see 4.8). (5) Compare episodes to see the recurring pattern. (6) Revise the hypothesis aloud with the patient (see 4.7).
• Example language (illustrative): "Take me through that evening. You came home, and then?" … "What exactly did she say?" … "And what did you say back?" … "You've skipped from the argument to the next morning. What happened in between?"
• Watch for: Kanter's contrast between asking "What happened?" and asking "How does it feel?" [source: Kanter 2013]. Ask for the facts first, and feeling will come with them. Watch also for the inquiry turning into a one-sided interrogation, which Sullivan warns against. It is working if accounts become more specific, patterns become visible to the patient, and the patient starts adding detail without being asked.

4.4 Developmental survey ("A Suggested Outline for Obtaining Data")

• Purpose / when: A checklist of developmental and life areas to cover during the detailed inquiry, so that important areas are not selectively missed. [source: Sullivan 1954, Ch. 7 title only; content uncertain]
• Steps: Sullivan's developmental epochs [source: Sullivan 1953] are infancy, childhood, the juvenile era, preadolescence, early adolescence and late adolescence. Move through these and through adult domains, asking for concrete interpersonal events in each. The areas commonly covered in this tradition include early care and training, school and peers, the preadolescent close friend ("chum"), the onset of lust and early sexual and dating experience, work history, intimate partnerships, use of alcohol and drugs, sleep and dreams, and attitudes to one's body and health [uncertain: this list is my reconstruction, not checked against Ch. 7; do not treat it as Sullivan's exact outline].
• Example language (illustrative): "Around 9 to 12, was there one particular friend you were close to?" … "When did sexual interest first become something you had to deal with with other people?"
• Watch for: Areas the patient dismisses quickly ("nothing much to say"). Sullivan's model predicts that anxious areas get skimmed. It is working if each epoch yields at least one concrete scene.

4.5 Transitions: smooth, accented, abrupt

• Purpose / when: Every change of topic. Each type controls anxiety and keeps one topic from contaminating the next. Interviewers should notice when they change subject. [source: Sullivan 1954, Ch. 3]
• Steps:
  • Smooth: link the old topic to the new with a bridge that is true on its face, so the patient can follow the path. Sullivan's example: "Well, now, that brings up the topic of so-and-so. Eh?"
  • Accented: signal clearly that a change is coming, drop the previous topic emphatically, and pause. This keeps the earlier discussion from "poisoning" the new data.
  • Abrupt: raise a relevant topic, without warning, at a socially awkward moment. It can be used either to provoke anxiety or to avoid it. Sullivan advises using it sparingly.
• Example language (illustrative): Smooth: "You mentioned your brother took over the business. That brings me to your own work history." Accented: "All right. Let's leave that completely for now. [pause] I want to ask about something different: your health."
• Watch for: Whether the patient can "at least dimly" follow where you are going. If they can't, they will feel interrogated. With abrupt transitions, watch for a spike in anxiety and vagueness. It is working if responses stay specific across topic changes.

4.6 Regulating anxiety: warning, normalising, rescuing, moving to reassuring material

• Purpose / when: Before and during anxiety-laden topics. The aim is to keep enough anxiety for the interview to be "psychiatric", but not so much that it goes vague. [source: Sullivan 1954, Ch. 2–3]
• Steps: (1) Warn: before a sensitive question, give "a little warning to brace himself". (2) Normalise: if the patient is anxious about something ordinary, say so. Sullivan's own version is that he has rarely heard anyone speak honestly without mentioning it. (3) Rescue: supply information the patient lacks, rather than leaving them stranded. (4) Move through: after upsetting material, lead on to definitely reassuring material, so that the anxiety is lived through rather than abandoned midway.
• Example language (illustrative): "I'm going to ask about something many people find awkward, the sexual side of the marriage." … "Almost everyone who talks honestly about their parents mentions some resentment. It would be odd if you didn't."
• Watch for: Signs of rising anxiety: answers become vague, the subject shifts, voice and rate change, or the patient suddenly agrees too readily. It is working if the patient re-engages with detail after the warning or the normalising remark.

4.7 Checking communication (consensual validation in the session)

• Purpose / when: Throughout, and especially after a long account or when you form an inference. It prevents one-sided interrogation, keeps the patient aware that they are gaining something, and tests whether your understanding is shared. [source: Sullivan 1954, Ch. 2–3; the term "consensual validation" is from Sullivan 1953]
• Steps: (1) Restate in plain words what you think you heard. (2) Ask directly whether you have it right. (3) Accept corrections and fold them in. (4) Do not let a serious misstatement or a delusion pass through silence. State your own view tactfully.
• Example language (illustrative): "Let me see if I have this straight: each time your manager gives feedback, you go quiet in the meeting and then rewrite everything that night. Is that right?"
• Watch for: Automatic agreement. That may itself be a security operation (compliance), so ask for an example that would contradict you. It is working if the patient corrects or refines you and adds new detail.

4.8 Tracking security operations and selective inattention

• Purpose / when: Continuously. Security operations are manoeuvres that reduce anxiety and protect self-esteem. Selective inattention is the main one: not noticing what would make you anxious. Both obscure communication and keep people from learning from experience. The detailed inquiry aims at exactly these blind spots. [source: Sullivan 1953; Sullivan 1954 introduction and Ch. 5 setting; Kanter 2013; Wikipedia, Interpersonal psychoanalysis]
• Steps: (1) Mark the moment the account changes: it goes vague, the patient changes topic, gets technical, jokes, deflects with flattery, or offers a stock explanation. (2) Note what was about to be said or noticed. (3) Go back to that point gently and concretely. (4) Ask about what the patient did not notice or ask about in the scene, such as other people's reactions or their own part. (5) Respect the pacing in 4.6: if anxiety climbs sharply, warn and approach in steps.
• Example language (illustrative): "Just as you got to what your father said, we ended up talking about the traffic. Can we go back?" … "What did your wife's face look like when you said that?" … "Did you ever wonder why he stopped calling?"
• Watch for: Your own selective inattention as well. Selective inattention can be seen in the countertransference, for example when you find yourself losing interest at the same point every time [source: Wikipedia, Interpersonal psychoanalysis]. It is working if the patient begins to notice the gaps themselves ("I just did it again, didn't I?").

4.9 Controlling relevance: insisting on data, discouraging trivia

• Purpose / when: When the account drifts, repeats, or becomes padded. The expert is entitled to the relevant data and discourages irrelevance. Patients appreciate it when the interviewer cuts off detail he no longer needs. [source: Sullivan 1954, Ch. 2–3]
• Steps: (1) Treat your entitlement to the information as a given rather than something to negotiate. (2) Once you have the point, end the account politely. (3) Redirect with a transition (4.5). (4) Stop your own curiosity once the clinical question is answered.
• Example language (illustrative): "I've got the picture of the office layout, thanks. What I need now is what happened between you and her."
• Watch for: Whether the trivia is itself a security operation, for example a flood of detail around an avoided point. If so, combine this with 4.8. It is working if sessions stay on interpersonally important material.

4.10 Listening to the vocal channel

• Purpose / when: All the time. Sullivan calls the interview primarily vocal communication: tone, rate, and hesitation carry meaning. [source: Sullivan 1954, Ch. 1]
• Steps: (1) Note changes in pace, pitch, volume, hesitations, and laughter that don't fit. (2) Link them to the topic of the moment. (3) Either make a note and come back later, or comment on it gently.
• Example language (illustrative): "You sped up just then, talking about the move."
• Watch for: Using vocal cues as accusations. Offer them as observations. It is working if the patient connects the change to a feeling or to something left unsaid.

4.11 Participant observation: using your own reactions

• Purpose / when: All the time, and especially when you notice a strong or unusual reaction in yourself. The interviewer's self is the main instrument of observation. You are part of what you observe. [source: Sullivan 1954, Ch. 1; Levenson 1989; Thompson 1952, "Countertransference" (uncertain)]
• Steps: (1) Register your reaction: bored, pulled to rescue, irritated, interrogating, charmed. (2) Ask what in the interaction just now called it out. (3) Ask whether it resembles what the patient describes others doing to them, or doing with them. (4) Use it to guide the inquiry (4.3, 4.14) rather than acting on it. (5) Check your blind spots against your own patterns, which Sullivan calls the interviewer's "technical gaps".
• Example language: Usually not said aloud. When it is shared, see 4.14 (illustrative): "I notice I've been asking more and more questions and you've been answering less and less. I wonder what's happening between us."
• Watch for: Mistaking your own patterns for the patient's. Supervision or personal analysis is the standard safeguard [synthesis]. It is working if your reactions lead you to concrete hypotheses that the patient can confirm.

4.12 Identifying parataxic distortion

• Purpose / when: When the patient responds to you, or to others, as if to a different, imagined person. This is the interpersonal counterpart of transference. [source: Sullivan 1954, Ch. 1; Sullivan 1953; Wikipedia, Harry Stack Sullivan]
• Steps: (1) Note a response that doesn't fit what you actually said or did. (2) Ask in detail what the patient perceived. (3) Work out together what in your behaviour was accurately noticed (see 4.15), and what was added. (4) Ask whom it is reminiscent of, through concrete episodes rather than interpretation.
• Example language (illustrative): "When I asked about your sister you went very still, almost as if you expected me to be critical. What did you hear in my question?"
• Watch for: Labelling a perception as "distortion" when it is accurate. Levenson's correction applies (4.15). It is working if the patient can tell apart what you did from what they expected.

4.13 Levenson's detailed inquiry: pursuing the particular

• Purpose / when: Throughout treatment, whenever the patient gives a summary, an abstraction, or an interpretation in place of an event. Specific inquiry focuses the work and brings transference and countertransference into view. [source: Levenson 1989; Levenson 1988, "The pursuit of the particular" (listed in the WAWI syllabus)]
• Steps: (1) Notice gaps in the narrative, including things that "don't add up" or that the story drops. Levenson's example question is "What happened to the cat?" (2) Ask about the real circumstances: did things actually happen as described, what did the other person actually do? Levenson asks of a remote father: "Did they talk?" (3) Prefer questions about events to interpretations about inner meaning. (4) Let the specifics, not theory, decide what you conclude.
• Example language (illustrative): "You said your mother 'was always disappointed'. Give me a moment when you saw that. What did she do?" … "You mentioned you had a dog, and then it vanished from the story. What happened to it?"
• Watch for: Two opposite failures. One is detail used defensively by the patient. The other is inquiry used by the therapist to keep a safe distance [synthesis]. It is working if new material appears that surprises both of you.

4.14 "What's going on around here?": working with the enactment in the room

• Purpose / when: When the pattern described in the patient's life starts to happen between patient and therapist. Levenson's shorthand points attention at the actual events of the patient's life and of the session, in contrast to "what does it mean?" [source: Levenson 1989; Kanter 2013 uses the phrase for building social perception; Levenson 1972/1983 on transformation and enactment (uncertain)]
• Steps: (1) Hold side by side the story being told and what is happening between you as it is told. Levenson treats them as transforms of the same pattern [uncertain: from secondary knowledge of Levenson 1972]. (2) Identify your own part in the repetition, using 4.11. (3) Describe the interaction concretely and without blame, including your contribution. (4) Inquire into it together using the method in 4.3: what happened, in what order, what each of you felt. (5) Do something different and examine it together. Levenson's idea of "working free" of the pattern [paraphrase; uncertain].
• Example language (illustrative): "You've been telling me how your partner keeps 'fixing' things for you, and I notice I've spent the last ten minutes giving you advice you didn't ask for. I think we've just done it here. What was that like from your side?"
• Watch for: Using this as a disguised accusation ("you're doing it to me"). Own your part first. It is working if the patient engages with the here-and-now account and the pattern loosens in later sessions.

4.15 Treating the patient's perceptions of the therapist as plausible

• Purpose / when: When the patient comments on you ("you seemed bored", "you're judging me"), or reads other people in a way you are inclined to call distorted. The interpersonal view holds that the patient distorts experience while trying to grasp reality. Taking their view seriously is a matter of conviction, not nurturance. [source: Levenson 1989; WAWI syllabus 2019; Thompson 1945/1950 on the analyst's real attributes (uncertain)]
• Steps: (1) Ask what the patient noticed: tone, words, timing. (2) Check honestly whether it was there. (3) Acknowledge what was accurate, briefly, without a confession. (4) Then ask what they made of it and what they expected next.
• Example language (illustrative): "You may be right that I sounded impatient just now. What did you hear, exactly?" … "Given that, what did you think I'd say next?"
• Watch for: Defensive denial (yours), or over-disclosure that shifts the focus onto you. It is working if the patient can test their perceptions against yours without either of you having to be wrong.

4.16 Self-revelation and self-disclosure (later interpersonal practice)

• Purpose / when: Selectively, when telling the patient about your experience of the interaction helps the inquiry in 4.14. [source: Levenson & Slomowitz 2018, chapter "Aspects of self-revelation and self-disclosure" (title only); Ehrenberg 1982/1992 on engagement (listed in syllabi). Steps below: synthesis]
• Steps [synthesis]: (1) Disclose your reaction within the session, not your personal life. (2) Tie it to a specific moment. (3) Offer it as data to examine together, not as a verdict. (4) Return straight away to inquiry: "What do you make of that?"
• Example language (illustrative): "When you described the accident so flatly, I felt a jolt of alarm that you didn't seem to feel. I'm curious about that gap."
• Watch for: The patient caretaking you, or the session becoming about you. It is working if the disclosure opens new material about the patient's experience.

4.17 Dreams: taking the manifest content seriously

• Purpose / when: When the patient brings a dream. Levenson reads the manifest content literally and in relation to the real relationship. In his example, the patient is amazed that a straw man bleeds, which he reads as the patient discovering that the analyst is real and can be hurt. [source: Levenson 1989]
• Steps: (1) Get the dream in detail, as you would an event (4.3). (2) Before reaching for symbolism, ask what in current relationships, including this one, the dream's actual scenes correspond to. (3) Ask about the real situations that the dream figures resemble.
• Example language (illustrative): "In the dream I'm behind glass and can't hear you. Has anything in here felt like that lately?"
• Watch for: Losing the manifest content by jumping to "deeper" meaning too soon. It is working if the dream links to a concrete interaction you can then explore.

4.18 Prescription for the interval (interruption)

• Purpose / when: End of a session when more sessions will follow. It consolidates the gains and carries the inquiry on between sessions. [source: Sullivan 1954, Ch. 2]
• Steps: (1) Pick a specific point from the session that is still open. (2) Give the patient something to consider or try to recall, for example the details of a particular period or relationship. (3) Keep it small and concrete. (4) Ask about it at the next session.
• Example language (illustrative): "Between now and Thursday, see what you can remember about the year you moved schools: who your friends were and how it ended with them."
• Watch for: The prescription turning into homework the patient feels judged on. Treat whatever happens, including forgetting, as data [synthesis]. It is working if the patient returns with recalled material or new observations.

4.19 Termination of a consultation: final statement, prescription of action, final assessment, formal leave-taking

• Purpose / when: End of a consultation, or of a treatment, when no further meeting is expected. The purpose is to consolidate "what has been achieved" as a durable benefit. [source: Sullivan 1954, Ch. 2; Ch. 9 headings only; detailed steps partly uncertain]
• Steps: (1) Final statement: summarise what you have learned about the patient's difficulties in living, in plain language, without jargon or "psychiatric banalities". (2) Prescription of action: say what the patient might do next, such as a referral, treatment, or a specific change to try, with its likely consequences. (3) Final assessment: invite the patient's reaction and correct misunderstandings [uncertain: content of this sub-step inferred from the heading]. (4) Formal leave-taking: close clearly and respectfully, so the patient leaves feeling more capable even if the findings are unwelcome.
• Example language (illustrative): "Here's what I've understood. The trouble isn't the job itself. It's that you go silent whenever someone senior disagrees with you, and then the resentment comes out sideways at home. I'd recommend weekly therapy focused on exactly that. What fits and what doesn't?"
• Watch for: Leaving without a summary, which wastes the consolidation, or overloading the patient. It is working if the patient can restate the gist and the next step in their own words.

Not included as interventions

• Free association is not a core technique in Sullivan's method. Some secondary accounts say he used it in a limited way and examined the results through dialogue [uncertain: secondary web summary only, not verified in a primary source].
• Interpretation in the classical sense is played down in favour of inquiry. Sullivan and Levenson give no formal protocol for interpretation [source: Levenson 1989; Kanter 2013].

5. Handling problems

Items tagged [source] are from the founders' texts. Items tagged [synthesis] are my own guidance, built from the principles above.

• The interview goes vague (stuck). Sullivan notes that many interviews drift from informative to vague once the patient becomes acutely anxious [source: Sullivan 1954, Ch. 3]. Response: (1) Recognise this as anxiety, not lack of material. (2) Use an accented transition to a safer area, then come back with a warning (4.5, 4.6). (3) Ask for one specific episode rather than a general account (4.3).
• Stuck in a repeating story or impasse. Levenson treats standoffs, impasses and stalemates as part of the process [source: Levenson & Slomowitz 2018, chapter title only]. Response [synthesis]: stop pushing content. Ask "what's going on around here?" (4.14), check how you have been taking part in the stalemate, and describe it concretely to the patient.
• Resistance and reluctance. Patients come with mixed motives, and reluctance is data rather than a fault [source: Sullivan 1954, Ch. 1]. Any effort to influence provokes refusal [source: Levenson & Slomowitz 2018, publisher description]. Response: inquire into the reluctance itself ("what's the risk in talking about this?"). Do not argue, persuade, or give unrequested advice. Make room for the patient.
• Overwhelm or flooding. Response: follow Sullivan's anxiety handling [source: Sullivan 1954, Ch. 2–3]. Warn before you go deeper. Normalise. Supply missing information. Lead through the upsetting material to definitely reassuring material, and don't stop midway. [synthesis]: with trauma material, inquire into the sequence of events at the patient's pace. Stop asking for detail when the patient becomes dissociated or disorganised, and ground them in the present relationship before going on.
• The inquiry feels like an interrogation. Sullivan warns against one-sided interrogation and says the patient must sense some benefit [source: Sullivan 1954, Ch. 2–3]. Response: use smooth transitions so the patient can follow your path. Check your understanding aloud (4.7). Give something back each session: a summary, a link, or a clarification. [synthesis]: if the patient seems to be answering "correctly" to please you, name that and ask what they think you want.
• Non-adherence (interval prescription not followed; missed or late sessions). Response [synthesis]: do not treat it as a failure to comply. Inquire into it in detail (4.3): when they thought of it, what got in the way, what they expected you to say. Missed sessions and fee problems are frame events, and how they are handled is part of the data (Levenson's frame) [uncertain].
• Ruptures, and the therapist's mistakes. Levenson treats mistakes, errors and oversights as part of the work [source: Levenson & Slomowitz 2018, chapter title only]. Response [synthesis]: (1) Ask what the patient experienced, specifically (4.3). (2) Acknowledge what was accurate in their perception (4.15). (3) Examine whether the rupture repeats the patient's pattern with you as a participant (4.14). (4) Don't apologise in a way that closes off the inquiry.
• The patient expects a "miracle" or demands answers. Sullivan notes this makes the interviewer's job harder [source: Sullivan 1954, Ch. 1–2]. Response [synthesis]: at the inception, set out what the work actually involves: a joint, detailed look at how things go with people. Then show its value through useful summaries rather than promises.
• The patient wants interpretation or theory ("what does it mean?"). Response: Levenson moves the question from "what does it mean?" to "what's going on?" [source: Levenson 1989]. Explain that you want to understand what actually happens first (4.13). Avoid jargon and "psychiatric banalities" [source: Sullivan 1954].
• The therapist's own selective inattention. You will skip what makes you anxious. Response: watch where you lose interest, change the subject, or reassure too fast (4.11). Bring recurring patterns to supervision [synthesis]. Countertransference is a route to spotting both selective inattention and selective non-participation [source: Wikipedia, Interpersonal psychoanalysis].
• Getting caught in the enactment and not noticing. This is not a problem to avoid. In Levenson's model it is unavoidable and is how change happens [source: Levenson 1972/1983; uncertain]. Response [synthesis]: routinely compare the story the patient is telling with what you are doing at that moment. The signals include sudden advice-giving, coldness, rescuing, over-questioning, and boredom.
• Mistaking accurate perceptions for distortions. Response: see 4.15. Before you name a parataxic distortion, check what you actually did [source: Levenson 1989; Thompson (uncertain)].
• Delusions or serious factual errors. Do not tacitly accept them through silence [source: Sullivan 1954]. Response [synthesis]: register your differing view calmly and briefly, then inquire into how the patient came to the belief. With psychotic presentations, follow standard psychiatric assessment and consider shared care.
• Suicide risk or crisis. The sources consulted give no crisis protocol. Response [synthesis]: the detailed inquiry style fits well with structured risk assessment: ask concretely about intent, plans, means, and protective factors. Follow your service's standard suicide risk assessment and safety planning procedures, and leave the exploratory stance until the patient is safe.
• Confusion with IPT. Patients or referrers may expect Interpersonal Psychotherapy, a time-limited, manualised treatment. Response: explain the difference at the inception, and refer for IPT if a manualised, time-limited treatment is what is wanted [synthesis].

Sources

Sources

• Sullivan H.S., The Psychiatric Interview (PDF; text accessible through Ch. 3 plus table of contents)
• The psychiatric interview, Internet Archive record (borrow-only; metadata)
• Levenson E., "Whatever Happened to the Cat? Interpersonal Perspectives on the Self" (WAWI archive PDF)
• WAWI course syllabus 2019, Langan & Bohm (Sullivan detailed inquiry session)
• IPSSNY syllabus, Interpersonal to Relational
• Kanter J., Helping, Healing and Interpreting: Sullivan, the Interpersonal School and Clinical Social Work (Academia)
• Interpersonal Psychoanalysis and the Enigma of Consciousness (Routledge page)
• PEP-Web: Greenberg, review of Levenson, The Ambiguity of Change (record only)
• PEP-Web: Ingram, review of The Ambiguity of Change (record only)
• PEP-Web: Foehl, Follow the Fox: Levenson's Pursuit of Psychoanalytic Process (record only)
• Interpersonal psychoanalysis (Wikipedia)
• Harry Stack Sullivan (Wikipedia)
• Clara Thompson (Wikipedia)

Further reading (not drawn on unless listed above)

• Sullivan, H. S. (1940). Conceptions of Modern Psychiatry. New York: Norton.
• Sullivan, H. S. (1953). The Interpersonal Theory of Psychiatry. New York: Norton.
• Sullivan, H. S. (1954). The Psychiatric Interview (eds. H. S. Perry & M. L. Gawel). New York: Norton.
• Thompson, C. (1945). Transference as a therapeutic instrument. Psychiatry, 8.
• Thompson, C. (1950). Psychoanalysis: Evolution and Development. New York: Hermitage House.
• Thompson, C. (1952). Countertransference. Samiksa, 6 [uncertain: journal not verified].
• Green, M. R. (Ed.) (1964). Interpersonal Psychoanalysis: The Selected Papers of Clara M. Thompson. New York: Basic Books.
• Levenson, E. A. (1972). The Fallacy of Understanding. New York: Basic Books.
• Levenson, E. A. (1983). The Ambiguity of Change. New York: Basic Books.
• Levenson, E. A. (1988). The pursuit of the particular: On the psychoanalytic inquiry. Contemporary Psychoanalysis, 24, 1–16.
• Levenson, E. A. (1989). Whatever happened to the cat? Interpersonal perspectives on the self. Contemporary Psychoanalysis, 25, 537–553.
• Levenson, E. A. (1992). Harry Stack Sullivan: From interpersonal psychiatry to interpersonal psychoanalysis. Contemporary Psychoanalysis, 28.
• Levenson, E. A., & Slomowitz, A. (2018). Interpersonal Psychoanalysis and the Enigma of Consciousness. Routledge.
• Ehrenberg, D. B. (1992). The Intimate Edge. New York: Norton.
• Kanter, J. (2013). Helping, healing and interpreting. Journal of Social Work Practice, 27(3), 273–287.