The Milan Approach To Family Therapy

Working With Systems Instead of People

The Milan Approach To Family Therapy didn't become what it is by accident. It started in the early 1970s at the Milan Associated Service, run by Luigi Boscolo, Gianfranco Cecchin, Selvini Palazzoli, and Giuliana Prata. They were clinicians working in outpatient settings with families that other people had already given up on or weren't making progress with. The approach they developed was a direct reaction against the more directive, insight-oriented models that were common at the time. Instead of trying to make family members understand their dynamics, they focused on how those dynamics actually function in real time. The core technique is circular questioning. It sounds more complicated than it is. The therapist asks each family member about relationships between other family members rather than asking them directly about their own feelings or opinions. You might ask a teenager, "What does your mother think your father does when you're grounded?" This isn't a trick question designed to catch someone in a lie. It's designed to reveal the family's internal map of who is aligned with whom, who holds power, and where the patterns repeat.

Understanding The Milan Approach To Family Therapy

There are four principles that tie everything together. Neutrality is probably the most important one and the most misunderstood. Being neutral doesn't mean being passive or indifferent. It means the therapist consciously avoids taking sides or assigning blame. When a parent says the child is the problem, the Milan-trained therapist doesn't push back with "Actually, the parenting is the issue." They stay neutral enough that the family has to work out the contradiction themselves. This creates a specific kind of tension in the room that most inexperienced therapists find very uncomfortable. They want to fix it by being corrective. That instinct is usually wrong. The second principle is hypothesis. Before and during sessions, the therapist builds a working theory about what maintains the presenting problem. This isn't guesswork in the loose sense. It's a structured guess based on information gathered from referrals, intake questions, and early interaction patterns. The hypothesis is always provisional. You update it after every session based on what actually happened. If your hypothesis doesn't survive three sessions of observation, you discard it and start building a new one. This happens more often than you'd think. Rituals are the third principle. A ritual is a specific behavioral instruction given to the family between sessions. It's not a homework assignment in the traditional sense. It's a deliberate disruption of the family's usual pattern. For example, if a family's conflict follows a predictable sequence where the mother escalates and the father withdraws, the therapist might instruct the father to remain present and engaged for a set period when the escalation begins. The ritual is designed to be slightly absurd or uncomfortable enough that it forces the family system to respond differently.

The fourth principle is positive connotation. This means reframing problematic behavior as having a positive intention or function within the family system. If a child is acting out, the therapist might reflect that the acting out serves to keep the parents united in their concern for the child. This isn't about being fake or manipulative. It's about redirecting the family's interpretation of the behavior away from individual pathology and toward systemic function. The child isn't broken. The family structure has organized itself around the child's symptoms in a way that serves an unstated purpose. The team structure is another defining feature. Sessions are typically conducted with a behind-the-mirror setup where one therapist works with the family while the rest of the team observes and discusses. After the session, the team meets privately to develop their next intervention based on what they observed. The family then returns for a follow-up session where the lead therapist communicates the team's perspective. This structure has practical implications. It requires coordination and a shared theoretical language among team members. It also means the family is receiving a consolidated view rather than individual therapists pulling in different directions.

What It Actually Feels Like In Practice

I worked with a family where the presenting problem was a 14-year-old boy who refused to attend school. The parents were in the middle of a bitter divorce and both were using the school refusal as evidence against the other. The mother claimed the father's permissiveness was enabling the behavior. The father claimed the mother's anxiety was paralyzing the boy. Standard assessment would have led to individual therapy for the boy or co-parenting counseling for the adults. Instead of addressing any of that directly, the team used circular questioning to map the alliance structure. We asked the boy what his mother thought his father wanted him to do about school. We asked the father what the boy's grandmother believed was causing the refusal. We asked the mother what she thought the boy believed about her role in the situation. The pattern that emerged was clear. The boy's school refusal was maintaining the divorce conflict at a manageable intensity. If the boy got better, the parents would have to deal with their own relationship directly. The symptom was holding something else apart. The ritual we prescribed was simple but specific. Both parents were instructed to agree publicly on one small school-related decision per week and present it as a unified position to the boy. They were told to do this regardless of whether they actually agreed privately. The boy was instructed to respond to any such unified position with a single sentence: "I'll think about it." Nothing more. No argument, no negotiation, no escalation. This disrupted the triangulation pattern without anyone having to admit what was happening.

The edge case I ran into was when one parent refused to participate in the ritual. The father simply said he wouldn't present a unified position because there was nothing to unify. The mother was willing to go along with it. This is a common complication that most training materials don't address directly. The workaround I used was to shift the ritual target. Instead of requiring both parents, I instructed the mother to make one school-related decision per week and communicate it to the boy independently. The father was informed of this change but not required to participate. The key insight is that the ritual doesn't need to be symmetric to disrupt the pattern. A single parent acting consistently differently than the family expects is enough to create systemic perturbation. The boy still had to respond with "I'll think about it," which removed his ability to escalate either parent against the other. Another counter-intuitive thing about this approach is that neutrality can actually slow down initial progress. Families come in wanting the therapist to validate their version of events. When you don't provide that validation, they often feel misunderstood or even antagonized. I've had parents leave early sessions saying the therapist "doesn't get it." This is normal and usually temporary. The resistance to neutrality is itself data about the family's relational patterns. People who are used to having their grievances validated will treat neutrality as a personal rejection. Working through that reaction is often where the actual therapeutic movement begins. The approach also has significant limitations that aren't always discussed. It doesn't work well with families where there is active domestic violence or substance abuse being used as a control mechanism. The systemic framing can inadvertently minimize the experience of the abused or addicted person by suggesting their behavior serves a function in the family system. In those cases, individual intervention and safety planning take priority. The Milan model assumes a relatively stable family structure where patterns are maintained by interaction rather than by fear or addiction. When that assumption doesn't hold, the approach can do more harm than good.

There's also the question of cultural applicability. The circular questioning technique assumes a family structure where members are comfortable discussing relationships with each other. In families from cultures where direct discussion of interpersonal dynamics is considered inappropriate or shameful, the questions can feel invasive or disrespectful. I've seen therapists adapt by using more indirect forms of circular questioning, framing questions around what "people in our culture" might think rather than asking family members directly. This is a practical modification but it requires genuine cultural competence rather than a superficial checklist approach. The time commitment is another practical consideration. The behind-the-mirror team structure requires at least three to four clinicians available for the same time slot. This isn't feasible in most solo practice settings or small community clinics. Therapists working alone have to adapt the approach significantly, usually by doing more self-reflection between sessions rather than team consultation. This is doable but it changes the quality of the intervention. The team process catches blind spots that individual therapists miss. Working alone means you're more likely to stick with a hypothesis that should have been abandoned. If you're considering this approach, the most useful starting point is understanding that it's fundamentally about changing the conversation rather than changing the people. The Milan team was influenced by systems theory and cybernetics more than by psychoanalysis or humanistic psychology. Their references lean toward Bateson, von Foerster, and the Palo Alto group rather than Freud or Rogers. Reading the original work by Selvini Palazzoli et al. in "Paradox and Counterparadox" gives you the full picture, though it's dense. The later work by Cecchin ("Hypothesizing, Circularity, Neutrality") is more accessible and covers the operational details more thoroughly.

The basic session structure follows a recognizable pattern. The therapist opens with an update on the previous ritual and gathers information about what happened during the week. This is where circular questioning typically gets most of its material. The therapist then communicates the team's hypothesis and introduces any new rituals. The session ends with a clear summary so there's no ambiguity about what was agreed upon. Each session usually lasts between 60 and 90 minutes, with the team consultation happening immediately after while the observations are still fresh. The total number of sessions varies widely. Some families show meaningful change within eight to ten sessions. Others require longer engagement, particularly when the presenting problem has been maintained for many years and the family system is quite rigid. The biggest mistake I see beginners make is treating the four principles as a checklist rather than as an integrated framework. Neutrality without hypothesis is just politeness. Hypothesis without circular questioning is speculation. Circular questioning without rituals is just interesting conversation. Rituals without positive connotation can feel like behavioral conditioning. The principles reinforce each other. Removing one weakens the whole structure. The approach also tends to get diluted when therapists pick and choose techniques without understanding the underlying systems thinking. Using circular questions in an otherwise directive therapy doesn't make it Milan work. It makes it something else entirely, and usually something less effective than either approach on its own.

Practical Considerations For Getting Started

Training in the Milan Approach To Family Therapy typically involves supervised practice with real families, not just role plays or case studies. The skill of maintaining neutrality while building and testing hypotheses in real time is something you develop through doing it, not through reading about it. Observation of experienced practitioners is particularly valuable because much of the work happens in what the therapist doesn't say as much as in what they do say. The pauses, the refusals to engage with provocations, the deliberate redirection of questions — these are all technical choices that are hard to learn from texts alone. If you're working in a setting without access to a Milan-style team, you can still apply the core principles individually. The hypothesis-building and circular questioning work without behind-the-mirror consultation. Rituals can be prescribed and monitored individually. Positive connotation is a verbal technique that doesn't require a team. What you lose is the corrective function of the team process. You're more likely to miss your own blind spots and cling to hypotheses that aren't serving the family. Self-monitoring through supervision or peer consultation can partially compensate for this, but it's not the same as having multiple perspectives in real time. The approach has evolved since the 1970s. Later Milan practitioners moved away from some of the more controversial aspects of Selvini Palazzoli's work, particularly the use of convulsive rituals designed to produce dramatic shifts. The newer emphasis is on collaborative hypothesis-building with the family rather than interventions imposed from outside. This is a reasonable development. The original approach sometimes had a tendency to treat families as systems to be manipulated rather than people to be understood. The updated direction corrects for that without abandoning the systemic foundation.

The research base for the Milan Approach is mixed. Some studies show favorable outcomes for certain presenting problems, particularly eating disorders and adolescent behavioral issues. Other studies find no significant difference compared to other family therapy modalities. The methodological problems in much of the family therapy outcome research — small samples, inconsistent treatment fidelity, difficulty blinding participants — affect the Milan work the same way they affect everything else. What the research doesn't capture well is the clinical nuance: which families respond to which interventions, under what conditions, and why. That kind of knowledge comes from sustained practice rather than controlled trials. The main takeaway is that this is a rigorous approach that demands intellectual discipline and emotional restraint. It's not the most intuitive therapy model to learn because it requires resisting the natural tendency to take sides, offer advice, or resolve tension. The work is in the questioning, the observing, and the timing of interventions. Families tend to improve not because they suddenly understand themselves better but because the therapist's interventions create small disruptions that force the system to reorganize. The reorganization, not the insight, is what produces change. That's a distinction that matters when you're deciding whether this approach fits your practice or your clients.