Magazine "History" · Reading time approx. 9 min

Around 1970, four Milanese psychoanalysts left the couch for the family table. Their work behind the one-way mirror gave rise to a form of questioning that still shapes every systemic intake session today.

Milan, Late 1960s: A Couch Is No Longer Enough

Mara Selvini Palazzoli (1916–1999) was a psychoanalyst in Milan who had specialized in a condition classic individual therapy regularly failed to treat: anorexia. For years she sat across from her patients, interpreting, analyzing — and watched many of them, no sooner discharged, return to the same families and fall ill again. In 1967, together with seven colleagues, she founded the "Centro per lo Studio della Famiglia e delle Tecniche di Gruppo" in Milan, still psychoanalytic in orientation at first, but with a gaze that widened from the individual patient to the system around her. A trip to Philadelphia, where Selvini Palazzoli watched family therapists such as Salvador Minuchin, Ivan Boszormenyi-Nagy, and James Framo at work, confirmed her direction: in the United States, what she was only beginning to conceive of in Italy was already being practiced — treating not just the individual, but the family as a whole.

In 1971 she took a further step. She left her own institute and, together with three physicians likewise trained in psychoanalysis — Luigi Boscolo, Gianfranco Cecchin, and Giuliana Prata, all three psychiatrists — founded a new Centro per lo Studio della Famiglia. This four-person team became what the field would soon call the "Milan School."

What united the four was a shared disappointment: psychoanalysis had given them tools to understand individuals, but none to understand why entire families kept their sick members sick with such persistence. They found their answers not in Vienna but in California — in Gregory Bateson's cybernetic thinking about communication, feedback, and contradictory messages, as it was being developed at the Mental Research Institute in Palo Alto. Translated into the family therapy room, this became, within a few years, an approach in its own right.

The Team behind the Mirror

How the Milan group worked was unusual for the early 1970s and is still recognizable today as the basic template in many systemic practices. Two therapists sat in the room with the family; two more observed from behind a one-way mirror. Before the session, the whole team agreed on an initial hypothesis: what might be holding this family together, what is keeping the symptom alive? During the session, the conversation could be interrupted — the observers would send in notes or briefly call a colleague out of the room. At the end, the entire team withdrew to confer while the family waited alone, so they could jointly formulate an intervention, which was then communicated to the family to close the session.

And then, for a long time, nothing happened. The Milan group left unusually long gaps between sessions; people spoke of a "long brief therapy." The idea behind it: a family that is just beginning to change needs time to carry out that change itself, without the next session interrupting it again. For this team, therapy was not a weekly ritual but a rare, carefully prepared intervention.

Hypothesis, Circularity, Neutrality: Three Rules That Changed Everything

In 1980, Selvini Palazzoli, Boscolo, Cecchin, and Prata published the paper "Hypothesizing — Circularity — Neutrality: Three Guidelines for the Conductor of the Session" in the journal Family Process — to this day one of the most cited texts in the history of systemic therapy. Five years earlier, in 1975, "Paradosso e controparadosso" had already appeared in Italian, followed by a German edition in 1977 titled "Paradoxon und Gegenparadoxon"; in it, the team had documented its work to date with families in "schizophrenic transaction." The 1980 paper distilled from this three working principles — concise enough to remember, and profound enough to carry an entire school of therapy.

Hypothesizing means the team never begins a session without a working assumption about what function the symptom might serve within the relational system — well aware that this hypothesis will be disproven or revised during the session. The hypothesis is not a verdict but a tool that makes it possible to start asking questions at all. Anyone entering a conversation without an assumption has nothing against which to test the family's answers.

Circularity is at the heart of what made the Milan School famous: the conviction that behavior in families cannot be explained linearly — not "A causes B" — but as a cycle of mutual reactions, in which every member responds to every other and in turn triggers further responses. This basic premise changed the way questions are asked: instead of asking who is to blame, one asks how one behavior relates to another — and who sees it differently from whom.

Neutrality, as the Milan group understood it, was not indifference but a deliberate stance: the therapist forms no lasting alliance with anyone during a session, passes judgment on no one, and does not position themselves as a judge within the family system. This stance was meant to enable the family to observe itself, rather than have the clinicians take that observation away from them through their own partisanship.

Method in Focus: How a Circular Question Asks Differently

The difference between linear and circular questioning can be shown in a single, fictional scene.

A family comes in because a teenager has barely been attending school for months. A linear question would go straight for the symptom: "Why aren't you going to school anymore?" Such a question looks for the cause within the person being asked and usually ends in explanations, justifications, or silence.

A circular question, by contrast, starts with the relationships in the room. For instance: "If your son went back to school — who in the family would be happiest about it? And who would barely notice?" Or, addressed to the mother: "What do you think your husband makes of the way you're handling his school refusal?" Or to the younger brother: "What's different at home on the days your brother doesn't go to school?"

None of these questions demand an explanation from the teenager. Instead, they prompt the family to reflect on itself as a system: who reacts to whom and how, what coalitions exist, what would shift if the symptom disappeared. The answers often reveal that staying away from school serves functions within the family system that no one had previously named — bringing the father home more often, say, or uniting the parents in a shared concern that would otherwise have divided them. It was information of exactly this kind, not the answer to "why," that helped the Milan team test and refine their hypotheses.

Split and Further Development

Around 1980 — the same year the landmark paper appeared — the four-person team split up. Mara Selvini Palazzoli and Giuliana Prata took a path described as more structural-diagnostic, extending their work beyond the individual family to larger systems such as schools and clinics; in 1982, Selvini Palazzoli founded a new Centro with further colleagues, including her son Matteo Selvini. Luigi Boscolo and Gianfranco Cecchin struck out in a constructivist-dialogical direction and founded the Centro Milanese di Terapia della Famiglia in 1980, from which they built training programs aimed primarily at professionals in healthcare and social work.

On this path, Cecchin took one further step beyond the shared legacy in 1987. In his paper "Hypothesizing, Circularity, and Neutrality Revisited: An Invitation to Curiosity," again in Family Process, he declared neutrality an ideal that could never be fully realized in an actual session. In its place, he put curiosity: a stance that forces therapists to keep admitting new hypotheses rather than settling for an explanation once found — his warning being that anyone who stops staying curious turns from a therapist into the family's overseer.

At nearly the same time, the Canadian psychiatrist Karl Tomm took up the Milan questioning technique and expanded it systematically. In a three-part series of papers titled "Interventive Interviewing," published in Family Process between 1987 and 1988, Tomm distinguished four types of questions — lineal, circular, strategic, and reflexive — according to whether they were more geared toward gathering information or deliberately prompting change. What began as the Milan question thus became a differentiated set of tools that still shapes training curricula today.

In this way, a Milan team that had worked together for only about a decade gave rise to an entire field of successor schools: the structural-diagnostic line around Selvini Palazzoli and Prata, the constructivist-dialogical line around Boscolo and Cecchin, and Tomm's Canadian refinement of question typology. What united these lines, despite all their differences, was the conviction with which it had all begun: that the right question can achieve more than the right explanation.

What Remains in Every Intake Session Today

What began in 1971 in a Milan consulting room can be found today in almost every intake session conducted with the aim of understanding a family as a whole. The question of who would be happiest about a change. The question of what someone else might think about a particular situation. The question of what would shift if the problem suddenly disappeared. All of these are direct descendants of the circular question as it was developed in Milan.

For genogram work, too, the legacy of the Milan School is no coincidence but a foundation. A genogram initially shows only facts: births, marriages, separations, deaths, generations layered on top of one another. Only the circular question turns it into a conversation about relationships. If a drawing shows the same distance between fathers and eldest sons across three generations, that observation can be translated directly into a question: "If your father were sitting next to you and saw this pattern — what do you think he would say about it?" Or: "Who in your family broke this pattern most successfully, and what did that person do differently?" The structure provides the occasion; the circular question opens access to what it means for those involved. And the hypothesis — that tool the Milan group placed at the start of every session — has found a visual counterpart in the genogram: a constellation of symbols and lines is itself already a working hypothesis, one that is free to change again with the next question, the next conversation. That was exactly the idea with which a handful of Milanese psychoanalysts left the couch more than fifty years ago.

Sources

Note: The original Italian title "Paradosso e controparadosso" (Feltrinelli, Milan), the 1975 publication date, and the 1977 German first edition are corroborated convergently across several independent library and antiquarian records.