Milan Systemic Family Therapy

Running a Milan Team Session

The Milan approach to family therapy doesn't come naturally to most clinicians trained in direct individual work. It requires a different posture entirely. You sit behind a one-way mirror with a team, observing the family while they talk in the room. Your job is to detect the patterns, not to fix them in the moment. The model relies on circular questioning, neutrality, and later, a set of curated messages delivered back to the family. That structure sounds clean on paper. It gets messy fast when a family starts fighting in front of your entire team. Milan Systemic Family Therapy originated in the 1970s at the Milan Associated Center, founded by Selvini Palazzoli, Boscolo, Prato, and Cyulatti. The core idea is that family symptoms are maintained by relational patterns, not by individual pathology. The therapist's role is to map those patterns and introduce new information that disrupts the homeostasis holding the symptom in place. Unlike structural or strategic models, the Milan team doesn't try to reorganize the family hierarchy directly. Instead, they use indirect interventions. The key technique is circular questioning. You ask each member about their relationships with everyone else, which forces the family to see the system from angles they usually avoid. For example, you might ask the mother how she thinks her husband feels about the daughter's anorexia. The mother then has to articulate the father's position, which often reveals alliances and triangles the family pretends don't exist. The questions aren't designed to gather information. They're designed to introduce doubt into the family's fixed narrative.

After several sessions, the team formulates a positive connotation for the symptom. This means reframing the problematic behavior as serving a protective function for the family system. The starving daughter isn't rebellious. She's keeping her parents' marriage intact by occupying all their attention. That reframing isn't comfort. It's a deliberate strategic move to reduce blame and open space for change. Families often resist this because it removes their moral certainty about who is at fault.

Setting Up the Team Format

You need at least two people to run this properly. One therapist sits in the session room with the family. The rest form the observing team behind a mirror or via video feed. The in-room therapist takes notes but rarely speaks except to pose circular questions. After the session, the team confers privately for twenty to forty minutes. They debate interpretations, challenge each other's assumptions, and converge on a set of messages to deliver. The in-room therapist returns and presents these without taking a definitive position. The neutrality requirement is brutal. You cannot side with anyone. You cannot validate one member's story over another's. When a father tells you his wife is the problem, you don't agree or disagree. You ask him how his wife would describe the situation. That redirect is uncomfortable for therapists who came up learning to build rapport through alignment. But alignment is exactly what maintains the symptom. The family has already found an alliance structure. You need to be the unpredictable element. Session frequency matters. The original protocol used weekly meetings for roughly twelve sessions. Extensions happen when the family's resistance pattern is deeply entrenched. I've seen cases stretch to thirty sessions before the positive connotation landed. More than that and you're just running the same loop. At that point, either the intervention worked or the model doesn't fit the case.

The Circular Questioning Sequence

There are three main types of circular questions. Relational questions probe the relationship between two people through a third person's eyes. Difference questions highlight changes over time. Hierarchical questions address power and authority within the system. You move through these in a loose sequence rather than a rigid script. Start with behavioral mapping. What exactly does the symptom look like day to day? How does each family member respond? This grounds the questioning in observable behavior before you pivot to meanings and alliances. Then move into relational circles. Ask each person how they think the others feel about the problem. This usually exposes mismatched perceptions that fuel the conflict cycle. The critical pivot comes when you introduce hypothetical questions. What would happen if the symptom suddenly improved? Who would be most worried? This reveals the hidden payoffs and often surfaces the family's unspoken fear that removing the symptom would remove something else important, like parental attention or structural stability.

I ran into a specific edge case early in my training that still sticks with me. We were working with a family where the teenage son had been diagnosed with oppositional defiant disorder. The parents were locked in a blame spiral that had been playing out for four years. Every circular question we asked just bounced off. The mother would answer, the father would contradict, and the son would go silent. The team kept trying to find the triangle but couldn't locate one. The system wasn't triangular. It was binary and rigid. The parents had formed a united front of disagreement, each playing an extreme position so the son didn't have to choose sides. Our questions were feeding the very pattern we were trying to disrupt. The workaround was to shift from circular questioning to U-shaped questioning. Instead of asking about relationships between members, we asked each person directly about their own position in relation to the other's position. How do you feel about how your wife feels about your son's behavior. How do you feel about how your son feels about your wife's complaints. This forced self-reflection rather than reactive blaming. The son finally spoke about feeling trapped between two opposing realities. That single statement changed everything. It took six sessions to get there. We had wasted about four sessions asking the wrong shape of question.

Formulating the Positive Connotation

The positive connotation is where most trainees stumble. It's not praise for the symptom. It's a systemic hypothesis presented as an idea rather than a truth. You frame the behavior as a solution the family accidentally created, not a problem they need to eliminate. The language needs to be careful. You're offering a perspective, not diagnosing a motive. A common error is making the connotation too sympathetic. If you sound like you're comforting the family, you've abandoned neutrality. The positive connotation should feel slightly unsettling because it challenges the family's preferred story about who the problem belongs to. The anorexic daughter protects the parents' marriage. That's not kind. It's analytical. It also absolves everyone of direct blame, which can frustrate families who want someone held responsible. The final message component is usually a ritual or an assignment. This isn't a behavioral contract. It's a directive that restructures interactions in a way the family can't easily predict or resist. The original Palazzoli protocols sometimes used elaborate rituals like having family members swap roles or perform specific sequences of interaction. Modern applications tend toward simpler interventions. A family with a school-phobic child might be asked to leave the child at home for one weekend while the parents go out together. The ritual communicates the hypothesis through action rather than words.

Limitations and When to Step Away

Milan Systemic Family Therapy fails in several predictable scenarios. Acute crisis intervention is not this model's strength. If a family is dealing with active suicidality, domestic violence, or substance withdrawal, you need a more structured, directive approach. The Milan model assumes a certain level of family stability and capacity for reflection that simply doesn't exist in crisis. The model also struggles with families who lack the verbal capacity to engage with circular questioning. Severe cognitive impairment, acute psychosis, or limited literacy can make the questioning sequence ineffective. I worked with one family where the father had borderline intellectual functioning. Every attempt at a difference question collapsed because he couldn't hold a hypothetical in mind. We switched to a more concrete structural approach with clear boundaries and rules. The family improved faster under that model than they ever did under Milan. Another hard limitation is cultural mismatch. The Milan model emerged from a specific European context with particular assumptions about family structure and individual autonomy. Families from collectivist backgrounds or non-nuclear arrangements sometimes find the indirect questioning style confusing or evasive. The indirectness reads as dishonesty rather than clinical technique. In those cases, a more direct model like Bowenian or brief solution-focused therapy tends to land better.

The model also requires institutional support. Running a proper Milan team demands a consultation room, a one-way mirror or reliable video system, and at least two trained clinicians per case. Many community mental health centers can't sustain that infrastructure. You end up adapting the model into a single-therapist version, which loses the team's corrective function. The observing team isn't decorative. They prevent the in-room therapist from collapsing into Alliance with any single family member.

Practical Workflow for a First Session

Begin by mapping the presenting problem with all members present. Don't let one person monopolize the narrative. Circle back to each person and ask them to describe the same event from their perspective. Note the discrepancies. Those discrepancies are your entry points for circular questioning. Build your questions around the relational map. Identify the triangles, coalitions, and exclusions. Ask questions that reveal who is aligned with whom and what happens when that alignment shifts. Watch for emotional reactions more than verbal content. The family will tell you what they think you want to hear. Their affect will show you what they're actually experiencing. After the session, the team should debate for at least twenty minutes before settling on a formulation. Disagreement within the team is productive. It prevents premature closure. If everyone agrees too quickly, someone probably missed something. The positive connotation should address the symptom's function while leaving room for alternative interpretations. Avoid language that sounds like moral judgment or therapeutic certainty.

The return to the family requires the in-room therapist to present the team's hypotheses as tentative ideas. Use phrasing like our team wondered whether or we're considering the possibility that. Neutrality lives in the hedging. The family should feel invited to accept, reject, or modify the formulation rather than persuaded to adopt it.

Reading the Room During Intervention

The moment of truth comes when the family receives the positive connotation. Watch for the specific reactions that indicate the intervention landed. Silence followed by a shift in posture usually means the family is processing something that contradicts their automatic responses. Defensiveness followed by a gradual softening often indicates the connotation touched a real underlying tension. Flat affect or immediate agreement can signal resistance disguised as compliance. The family is absorbing the message without integrating it. Between-session contact is rare in pure Milan work but sometimes necessary. A brief phone call can prevent a crisis without derailing the therapeutic process. Keep it short and avoid giving advice. If you find yourself calling families regularly, the team format has broken down and you've slipped into a more traditional individual-oriented stance. The model works best when applied consistently across multiple sessions. Switching mid-stream to a different framework confuses the family and weakens the intervention. If the Milan approach isn't producing movement after six to eight sessions, step back and reassess whether the case fits the model or whether you need a different theoretical orientation entirely.