Circular questions: seeing a relationship pattern differently

START HERE

The essential idea

Learn to ask and follow relational questions through a worked family-coordination example.

BUILD THE REASONING · 01

Describe the sequence before explaining it

A circular question explores how people notice, interpret, and respond to one another. It is useful when a story has become a fixed accusation: one person is “controlling,” another is “unmotivated,” and everyone already knows who should change. The question should produce information that could revise the counselor’s understanding. It is not a clever way to persuade the family to accept the counselor’s preferred explanation. Tomm’s interviewing framework makes the interviewer’s assumptions and intentions part of the clinical work, rather than treating wording as a neutral tool.

Begin with a specific recent occasion. Suppose Mira asks her brother Leo whether he has arranged transport for their father’s appointment. Leo gives a short answer and leaves the room. Mira sends several messages. Leo stops replying. “Mira nags because Leo is irresponsible” compresses the whole episode into two personality labels. A useful sequence keeps the actions separate: a transport question, a short response, leaving, repeated messages, and no reply. Ask what happened immediately before and after each step. The order matters because a general account of what people “always” do can conceal occasions when the pattern changes.

Circularity does not require an endless search for who started it. The counselor can begin with any observable point and follow the next responses. Equally, abandoning a single starting cause does not mean abandoning facts. If a transport booking was missed, that matters. If one sibling has been carrying more unpaid care work, that matters too. The purpose is to understand the process well enough to choose a useful next step, while keeping practical burdens and individual responsibilities visible.

BUILD THE REASONING · 02

Match the question to the information you need

A sequence question asks about linked actions: “After Leo leaves, what do you do next, and what does he seem to notice?” It is appropriate when the order is unclear. A difference question compares occasions: “What is different on days when one message is enough?” It is appropriate when an exception might reveal conditions that reduce the difficulty. A perspective question asks one person how another may understand an event: “Leo, what do you imagine Mira concludes when you leave?” It is useful when the meanings assigned to the same behavior differ. These are practical question designs, not rigid categories that determine the answer in advance.

A perspective answer is a report about the speaker’s understanding; it is not evidence that the other person actually thinks that way. Return the interpretation to the person described: “Mira, does that fit your experience, or is something missing?” If Mira says she assumes Leo is overwhelmed rather than uncaring, the original hypothesis changes. Without that check, the counselor could accidentally turn one family member’s mind-reading into the official story.

A hypothetical question can open a possible response: “If the booking were visible to both of you before dinner, who would notice the change first?” This is different from collecting a past sequence. Make the shift explicit so a prediction is not recorded as an observed fact. A useful hypothesis remains small and revisable: repeated checking may communicate urgency to Mira while signaling mistrust to Leo. That account still needs to be tested against their words, observed responses, and occasions that do not fit.

BUILD THE REASONING · 03

Follow the answer instead of protecting the question

Good systemic inquiry is responsive. Imagine Leo says, “I leave because that is when my work call starts.” The counselor should investigate scheduling before interpreting withdrawal as resistance. Ask whether the same pattern occurs when Leo is not working. Imagine Mira adds that Leo has never been given access to the transport booking app. A shared access problem is now part of the formulation. Repeating a communication exercise while ignoring the practical barrier would make the intervention look sophisticated while leaving the mechanism unchanged.

Use an exception to refine the mechanism, not to disprove someone’s distress. On a week when the family agreed the booking before Leo’s shift, Mira sent one reminder and Leo confirmed it. The difference may be timing and a visible agreement. It does not establish that all tension is solved or that Mira’s worry was unnecessary. A small experiment could set one named person to make the booking, one confirmation time, and a backup if no confirmation arrives. At the next meeting, ask what actually happened and how each person understood it.

A clear note separates observation, report, and interpretation. “Mira reported sending four messages after Leo left; Leo linked leaving to a work call; both agreed to test an earlier confirmation time” describes the information and plan. “Family displays dysfunctional pursuit and withdrawal” is too broad to show what the counselor learned or how the next step follows. Theoretical language can be useful in formulation, but it should not replace an account that another clinician can examine.

BUILD THE REASONING · 04

Keep power, consent, and safety in the formulation

Questions about reciprocal influence do not establish equal responsibility. A person’s response to intimidation is not equivalent to the intimidation itself. If a client describes threats, fear of consequences for speaking, surveillance, or coercive control, ordinary conjoint pattern work may be inappropriate. Follow the setting’s safety procedures, assess separately when clinically and legally appropriate, and seek qualified supervision or specialist consultation. Do not use a circular question to invite someone to explain how they “caused” another person’s violence.

Consent also concerns the format of the conversation. Clarify who is participating, who the client is, and how information will be handled before inviting sensitive comparisons. A young person may not experience an apparently open family question as optional. An interpreter, disability accommodation, or separate conversation may change what participation requires. Ask who has not yet had a reasonable opportunity to speak, while avoiding an assumption that equal speaking time automatically means equal power.

The decision at the end of an interview is whether the new information supports the proposed next action. A practical coordination problem may call for a concrete agreement. Unclear meanings may call for further checking. A safety concern changes the assessment and care plan. Circular inquiry earns its place by helping the counselor distinguish these possibilities, not by making every problem into the same relational cycle.

FROM CONCEPT TO DECISION

One missed ride, three different interpretations

1. 1

   Establish the recent sequence

   Ask each sibling for the last actual occasion. They agree that Mira asked during Leo’s shift change, Leo left to answer a work call, and four messages followed. The sequence is shared; their interpretations differ.

   “Let’s slow down the last occasion. What happened just before you left, and what happened after the first message?”

2. 2

   Check a perspective without declaring it true

   Leo predicts that Mira sees leaving as indifference. Mira instead describes uncertainty because the previous booking was missed. Her concern contains a concrete reliability problem, not only a negative attitude toward Leo.

   “Mira, when Leo says you may read that as not caring, what fits and what would you change?”

3. 3

   Use a difference to choose an experiment

   Both recall that the previous week worked when the booking was confirmed before the shift. Agreeing a confirmation time and backup is more closely tied to the available evidence than assigning a general listening exercise.

   “What was different on the day one message was enough? Which part could you realistically repeat this week?”

4. 4

   Define what would change the hypothesis

   If conflict continues even when bookings are visible and confirmed, coordination alone is insufficient. If either sibling cannot disagree safely, reconsider the conjoint format. Review the experiment without treating a poor result as proof that someone refused to improve.

MAKE THE REASONING YOURS

Work through a new situation

READ WITH A PURPOSE

Best references

These readings deepen the specific models, evidence, and decisions in this lesson.

1. 01
   Tomm (1988): Interventive interviewing, Part III

   Primary conceptual paper on how interviewers’ assumptions and intentions shape lineal, circular, strategic, and reflexive questions. The PubMed record provides the abstract and publication details.

2. 02
   SAMHSA TIP 39: Family counseling approaches

   An extended clinical teaching chapter comparing family approaches and showing how relational sequences enter substance-use treatment; use it to distinguish approaches rather than treating all family work as one technique.

3. 03
   AAMFT Code of Ethics, effective January 2026

   Professional standards for informed consent, confidentiality, and responsibilities when working with multiple people. Apply the laws and professional code governing your own role and setting.