© Copyright 2025 MarriageFamilyTherapist.org
Assess, pace, and repair male shutdown without pressure, and spot when silence signals risk
In many couples sessions, one partner carries the conversation while the other offers two words: "I'm fine." Many clinicians read that silence as resistance or disinterest, but this is one of the most costly couples therapy misconceptions: the behavior more often functions as a learned safety behavior. That reading gap is where early-career LMFTs lose traction with the quiet partner.
This matters because the standard move, pressing harder for words, usually deepens the shutdown instead of opening it. The more useful clinical question is not why a partner won't talk, but what taught him that talking wasn't safe.
A recent Yahoo Creators interview with LMFT Scott Mallon models that reframe well and serves as a starting case study here. It is one clinician's perspective, not peer-reviewed evidence, so treat it as a teaching example to pair with primary research.
Why does my partner shut down in couples therapy when he talks fine at home? Usually because the therapy room changes the stakes. The same person who chats easily over dinner can go silent when a stranger is watching and the topic turns to blame or feelings.
Several pressures stack up at once:
In many cases, shutting down is a learned protective move, not a refusal to participate.
He may also be quietly scoring the room. If the therapist nods more at his partner's account, he may conclude the therapist has taken a side. He may also doubt that what he says will stay among the three of them, or wonder whether it will resurface as ammunition at home, concerns that make therapist boundaries in couples counseling feel especially important. Until those questions are settled, saying less feels like the sensible option.
Home offers things the session does not. There is no audience, the rhythms are familiar, and he can leave the room, put on the TV, or change the subject when a conversation gets too hot. A therapy hour removes those exits, so shutdown may be the only one left.
Notice the contrast. A partner who is relaxed and talkative about work, the kids, or logistics, then freezes when the conversation moves to emotion or fault, is showing a topic-specific pattern. That differs from someone who is quiet across every subject, which calls for a different line of assessment.
If you are the partner sitting beside someone who has gone silent, hold onto this: shutting down in session is often about the setting and the stakes, not a lack of love. Treating it as a verdict on the relationship tends to raise the pressure and deepen the quiet, which can erode human connection in therapy.
In a recent Yahoo Creators interview, LMFT Scott Mallon recasts the common complaint "Why won't he open up?" as "What taught him that he shouldn't open up?" For trainees, that shift is the skill worth noticing. A complaint about a partner's behavior becomes a developmental question about where the behavior came from.
Mallon points to childhood messages such as "be tough," "walk it off," and "suck it up," which accumulate over time. Many boys, he says, become fixers instead of emotional communicators. He adds that most men are privately carrying worries about marriage, kids, money, career, aging parents, and health. Silence in session often sits on top of a heavy load, not an empty one.
Psychologist Ronald F. Levant proposed normative male alexithymia in 1992, based on clinical observation rather than prevalence data.1 The idea is that masculine socialization can leave men with mild to moderate difficulty identifying and describing feelings, especially hurt, fear, sadness, and loneliness, because those states conflict with traditional male expectations. It differs from general alexithymia, and it is not a psychiatric diagnosis.
Measurement came later. The Normative Male Alexithymia Scale (2006) has 20 items rated on a 7-point scale and targets difficulties tied to male-role socialization. The Toronto Alexithymia Scale-20 measures general alexithymia instead.1
A 2025 study in Sex Roles (Mancini and colleagues) linked masculine ideals, including emotional restraint, self-reliance, and putting work first, with alexithymia, poorer emotion regulation, and anxiety and depression symptoms.2 A systematic review in the American Journal of Men's Health found that boys hear messages like "boys don't cry" from an early age, and that these messages create barriers to emotional openness even in serious distress.1
Both findings describe group-level patterns. They do not mean every man is shut down, and they cannot explain any one client. Evidence on the construct is also mixed, and the full study details here should be verified against the original papers.
The interview itself is a single clinician's account, not peer-reviewed evidence. Programs should pair it with the primary literature above.
Treat shutdown as a learned strategy that once protected him. Going quiet, solving the problem, or saying "I'm fine" may have kept him safe from mockery or dismissal. Mallon notes that some men did open up before and were dismissed, mocked, or had their honesty used against them later.
That changes the work. You are not fixing a deficit or pushing someone to perform openness. You are building enough safety in session that an old protective habit becomes one option among several, instead of the only one.
Instead of asking "Why won't he open up?" ask "What taught him that he shouldn't open up?" Silence is often learned safety behavior, not a deficit.
One of the most common training errors in couples therapy communication is treating every silent partner as a deliberate stonewaller, when the behavior often reflects an overwhelmed nervous system rather than a strategic choice.
Stonewalling, a term popularized through John Gottman's observational research, describes a pattern of withdrawing from interaction: turning away, going unresponsive, withholding engagement. It sits alongside criticism, contempt, and defensiveness in the Gottman Four Horsemen model.1 Emotional shutdown, by contrast, is frequently involuntary. It can accompany what Gottman's lab called diffuse physiological arousal, or flooding: intense autonomic activation that narrows attention and makes listening, perspective-taking, and problem-solving genuinely difficult.2
In plain terms, a flooded partner is not refusing to engage so much as temporarily unable to. Clinical materials often cite a heart rate at or above 100 bpm as a marker of flooding, but treat this as a rough rule of thumb, not a diagnostic cutoff.2 It shifts with fitness, medication, baseline heart rate, and context.
You will encounter the widely repeated figure that roughly 85% of stonewallers are men. Honor its limits. The number is tied to Gottman's observational research program, including work often traced to Gottman and Levenson's 1985 study, with the laboratory tradition beginning in the 1970s using videotaped conflict synchronized with physiological measures.2 At least one review could not trace the 85% to a verifiable peer-reviewed finding, and its replication is not established.3 Reported this way, it is a conditional proportion within an observed sample, not a population prevalence, and it does not tell you whether it counts individuals, episodes, or coded segments. Present it as a report of Gottman's research tradition, with caveats attached.
Never assume the same behavior carries the same cause across couples. Intent and physiology are different things. A useful in-session read: does the partner look flooded (rapid or pressured speech, flushed skin, a glazed or faraway expression) or deliberately disengaged (calm, cold, pointedly turning away)? The first suggests overload that calls for pacing and a break; the second warrants closer attention to what the withdrawal communicates, which later sections address.
Treat each row below as a working hypothesis to test across sessions, not a diagnosis, because the same quiet partner can look identical across very different drivers. Rely on observable behavior, history, and differential assessment rather than polyvagal labels: a 2026 critique argues that major premises of Polyvagal Theory lack support, and the brainstem pathway often cited for "freeze" has not been shown to mediate emotional freezing or dissociation in humans. When depression, trauma, or neurodivergence is suspected, refer for individual assessment before building the couples treatment plan around it.
| Pattern | Observable signs in session | Likely driver | Suggested first move |
|---|---|---|---|
| Flooding-based shutdown | Goes quiet mid-conflict and struggles to think or respond. Stillness or flatness may coexist with jaw tension, shallow rapid breathing, restless hands, or effortful control. | Acute conflict-related emotional overload and high physiological arousal, not necessarily deliberate disengagement. | Slow the pace and reduce escalation before any problem-solving. Name the pattern and offer a calming break of at least 20 minutes with a set time to return. |
| Habitual stonewalling | Minimal answers, reduced eye contact, changing the subject, or physically leaving, often recurring whenever the other partner pursues. | Repeated flooding plus no alternative way to stay engaged. A learned demand-withdraw pattern, not necessarily intentional cruelty. | Work both sides of the cycle: build regulation and a return path for the withdrawing partner while helping the pursuing partner raise concerns in less triggering ways. |
| Dissociation or trauma response | May appear absent, lose track of the conversation, or respond after long delays. No single behavior confirms dissociation. | Possible trauma-related response. Do not assume a specific vagal mechanism explains it. | Assess clinically through history and direct, gentle questions rather than inferring it from a polyvagal state label. Refer for individual trauma assessment. |
| Depression | Flatness and low energy that persist across topics, slowed speech, hopeless statements, and withdrawal reported outside of conflict as well. | A mood problem affecting energy and engagement broadly, not a response tied to the couple's conflict. | Screen for depressive symptoms and safety, lower session demands, and refer for individual evaluation. |
| Neurodivergent processing | Long pauses before answering, more fluent in writing than speech, shutting down during rapid or noisy exchanges, reduced eye contact as a baseline. | Slower verbal processing or sensory overload rather than avoidance of the partner. | Ask how each partner communicates best, allow processing time, use written or structured prompts, and reduce sensory load. Refer if not previously assessed. |
| Deliberate punitive silence | Silence used selectively after the partner displeases them, days-long silent treatment, contempt cues, or a partner who reports fear. | Control or punishment rather than overwhelm. | Pause conjoint problem-solving and screen each partner individually for coercion and safety before continuing couples work. |
Scott Mallon, a licensed marriage and family therapist quoted in an October 2026 Yahoo Creators interview, named five separate meanings that can hide behind the two words "I'm fine." That single observation reframes assessment: the quiet partner is rarely withholding nothing. He is usually withholding something he has no practiced language for.
Use these sparingly, one or two per session, with permission to decline:
Mallon's list gives trainees a working differential rather than a script. "I'm fine" may mean he is overwhelmed and cannot explain it, that he does not want his partner to worry, that he expects judgment or disbelief, that he is exhausted and has no energy for the conversation, or that he genuinely does not know where to start. Offer one hypothesis at a time, tentatively, and let him correct you. Being corrected is progress; a partner who says "no, it's not that, it's more like..." has just produced emotional language.
Ask directly about prior cost. Mallon noted that some men have opened up before and been dismissed, mocked, or had their honesty raised against them later in an argument. A partner who learned that disclosure gets weaponized is not avoidant by temperament; he is running a reasonable risk calculation, and that calculation belongs in your case formulation.
Screen for trauma history, immigration and acculturation stress, military or first-responder culture, religious norms about male stoicism, family rules about who gets to show feeling, and domestic violence screening in family therapy when shutdown is paired with contempt or control. This part of MFT risk assessment family dynamics shapes expression far more than personality does. When answers stay flat or when you suspect disclosures he will not make in front of his partner, offer a brief individual check-in with a clear confidentiality policy stated to both partners in advance.
Pace all of it. Assessment that fires question after question becomes one more interrogation, and interrogation is exactly what taught him to go quiet.
What do you actually do in the room when one partner goes quiet, stares at the floor, and the other partner keeps pressing for an answer?
The first move is pacing, not persuasion. When you see the shut-down partner's eyes drop, arms cross, or speech shorten, name the shift gently: "I notice we've lost your voice. We don't need to solve this right now." Offer a short pause, a few slow breaths, or a grounding question about where the tension sits in the body. If the pursuing partner keeps escalating, interrupt kindly but firmly: "Let me pause us here. The temperature is too high for [partner] to track what you're saying, and I want him to actually hear you." Cutting the other partner's turn short when flooding shows is protective, not punitive.
Replace open-ended processing with short, observable turns. Have each partner speak for two or three sentences while the other paraphrases before responding. This reduces the demand to produce fluent emotion language on the spot. Then use Narrative Therapy techniques to externalize the pattern: name the pursue-withdraw cycle as the shared problem, not the person. Say something like, "When you raise a concern, the cycle pulls you toward more questions and him toward less response. We're working against the cycle, not each other."
Avoid the three moves that reinforce shutdown: do not rescue with a long therapist summary that lets the quiet partner off the hook, do not let the pursuer fill the silence, and do not shame the quiet partner for having few words. Let the silence sit briefly while tracking safety. Offer a lower-demand route such as writing one sentence, rating distress on a 0-10 scale, or nodding to a guess.
If silence persists despite pacing, pause conjoint work. A separate individual session can assess safety, depression, substance use, or coercion, and whether couples addiction therapy is needed, without an audience, and can rebuild alliance with a partner who feels outnumbered. For men who distrust "talk therapy," use concrete engagement: a written agenda, brief rating scales, psychoeducation about flooding and self-soothing, and behavioral tasks before emotional disclosure. Frame his protection as learned safety behavior, not refusal to care.
Emotionally focused couples therapy treats withdrawal as protection against threat, shame, or attachment injury; re-engagement moves from grief and hurt toward agency, not simply more talking.1 Older EFCT reviews reported mean effects around 1.28 and stable estimates near 1.31 with recovery rates around 70%; a recent meta-analysis of couple and family therapy for couples approaching divorce reported a fixed-effects effect size of 0.73.3 Those are overall model effects, not isolated re-engagement changes. Gottman's flooding and self-soothing protocols are clinically specified, but no Gottman-specific meta-analytic effect size emerged in the available evidence, so use them as structured tools rather than citing a precise outcome number. Bowen-informed coaching targets differentiation and reduced reactivity through process questions, the "I" position, and genogram work; systemic therapy broadly shows modest effects (overall g=0.30, family functioning 0.34),4 but Bowen-specific estimates are not established here.
Every clinical move described so far assumes the silence is defensive. The harder judgment call is distinguishing a partner who goes quiet because he is overwhelmed from one who goes quiet because silence works as a weapon. Get that distinction wrong and conjoint therapy stops being neutral: it becomes a stage where one partner discloses and the other retaliates later.
Involuntary shutdown usually carries distress with it. The partner looks physiologically activated, reports feeling overwhelmed, and wants to come back to the conversation once he can. Punitive silence looks different. It is deployed, not suffered. It tends to travel with other controlling behaviors: financial abuse such as restricting money or contact, monitoring the other partner's phone or movements, threats, and a pattern of the other partner changing her behavior to avoid a reaction. Coercive control can be present with no recent physical assault at all, and on its own it contraindicates conjoint work.
Screen each partner in a separate, private interview rather than only with the couple in the room. Disclosure in front of a controlling partner can raise retaliation risk and is often suppressed by fear. Explain confidentiality and its limits to each person individually. Use both written and verbal assessment, and ask behaviorally specific questions rather than "is there any violence?" Ask directly about fear: fear of serious injury or death, whether disagreement feels safe, whether she expects payback for what is said here. Assess pattern, severity, timing, and escalation, plus injuries, strangulation, threats to kill, weapons, stalking, protective-order violations, violence involving children, and uncontrolled substance use.
Note that the AAMFT Code of Ethics does not spell out an IPV-specific conjoint screening protocol or a single contraindication list. It establishes duties: safety, informed consent, competence, avoiding harm, appropriate referral. The contraindication criteria above come from the broader clinical literature, not from one codified standard.
If safety is in doubt, do not run a confrontational disclosure session. Conjoint work is potentially appropriate only when both partners want the violence to end, are committed to ending it, and do not fear the conjoint conversation itself. Do not use couples therapy as the tool to stop ongoing violence. If violence occurs during treatment, stop conjoint sessions and assess each partner individually, with attention to therapist safety. Throughout, watch for retaliation, surveillance, pressure to recant, compelled attendance, and intimidation. Refer to specialized IPV services and complete safety planning first.
For readers in this situation: silence by itself is not abuse. Silence paired with fear, intimidation, or control is a reason to seek help outside the couple.
Pushing for an answer in the moment and agreeing on a pause in advance are two very different ways to handle silence, and only the second tends to keep a partner talking. Homework works best when it builds that agreement before the next shutdown arrives. Three short assignments cover most couples.
The talking partner's job is to lower the pressure, not raise the volume. Offering a pause and a return time works better than demanding words. For example: "I can see this is a lot. Let's take 20 minutes and come back." The return time matters. Without it, a break can feel like abandonment to the pursuing partner, and the pursuit starts again.
The quieter partner then keeps the other half of the deal: coming back when the time is up, even if all they can say is "I'm still sorting it out."
Good repair is brief and does not reopen the argument. Coach couples to skip who said what and focus on the moment of disconnection. Useful openers include:
A repair attempt does not require a resolution. Its job is to show both partners that the relationship survived the silence.
Tell couples to keep a simple running list of what worked and what didn't: which signal they used, whether the return time held, and how the check-in felt. Bring it to the next session. A failed time-out is useful clinical data, not a failed assignment, and it shows you where to adjust pacing or structure.
"I'm fine" can mean "I'm overwhelmed and don't know how to explain it," "I don't think you'll understand and will judge me," or "I don't even know where to start."
Couples therapy training increasingly treats male emotional shutdown as a learned safety response rather than a character flaw, and the best programs now teach it that way. Knowing where that training happens helps you judge programs before you apply.
Three places do most of the work. Couples therapy courses teach pursue-withdraw dynamics and in-session pacing. Assessment courses teach you to tell shutdown apart from depression, dissociation, or control. Gender and socialization coursework supplies the developmental lens, including Ronald Levant's concept of normative male alexithymia. Practicum is where the learning becomes real, because a silent partner across from you is nothing like a case vignette.
Bring recordings of the moments when a partner went quiet. Then ask your supervisor one specific question: what did I do in the 30 seconds before the silence? Often the answer is a rapid-fire question, a reflection that sounded like criticism, or a pause you filled too quickly. Those patterns are hard to see in the moment and easy to see on playback. Supervisors can also help you notice when your own discomfort with quiet is driving the session.
A clinician interview like the one discussed earlier is a good teaching prompt. Scott Mallon's reframe from "Why won't he open up?" to "What taught him not to?" is useful precisely because it is memorable. But it is one therapist's account, not peer-reviewed evidence. Read it alongside:
If you are weighing how to become a couples therapist, start with accreditation. Programs accredited by COAMFTE (the Commission on Accreditation for Marriage and Family Therapy Education) have to meet defined standards for clinical training, which matters when you want supervised couples hours. Then compare graduate programs on practicum structure, supervision access, and how directly the curriculum covers gender, multicultural counseling competencies, and relational violence.
