Stonewalling is what happens when someone in a conversation stops responding — no reply, no eye contact, no visible reaction — while still physically present. It is one of the four conflict patterns the psychologist John Gottman named the “Four Horsemen,” and it is most often a shutdown under overwhelm rather than a deliberate punishment. But not always. The difference matters enormously, and most of what you will read online collapses it.
That collapse is the reason this article exists. Search for stonewalling and you will find it confidently described as a narcissist’s weapon, a silent-treatment tactic, a form of emotional abuse. Search a little further and you will find it described as an involuntary nervous-system response nobody chooses. Both descriptions are attached to real phenomena. They are not the same phenomenon, and telling them apart is the single most useful thing you can do if this is happening in your relationship.
Here is what the research actually supports, where it runs out, and what to do with the part that is left.
What stonewalling actually looks like
Stonewalling is withdrawal from an interaction while the interaction is still going on. In practice it looks like:
- One-word answers, or none at all
- Looking away, looking at a phone, looking through the other person
- A face that has gone completely still — no nodding, no frowning, no sign anything is landing
- Busying yourself with something else mid-conversation
- Leaving the room without saying you are leaving
The defining feature is not silence. It is unresponsiveness — the removal of every signal that would normally tell the other person they are being heard. You can stonewall while saying “fine.” You can stonewall while nodding, if the nodding has gone mechanical.
This is a real, measured behaviour, not a folk category. In Gottman and Levenson’s 1992 study of 73 married couples followed from 1983 to 1987, couples classified as “nonregulated” showed, among other things, “more stubbornness and withdrawal from interaction” and “greater defensiveness” than regulated couples, alongside lower marital satisfaction and a higher rate of separation. That study is observational — it watched couples and followed them — so it tells you withdrawal travels with poor outcomes. It does not, on its own, tell you withdrawal causes them.
What is happening in the body when someone stonewalls
The most common explanation you will read is “emotional flooding”: the person is so physiologically overwhelmed that their system shuts the conversation down to protect itself. That explanation is probably broadly right, and it is worth being precise about what supports it — because “flooding” is Gottman’s clinical term, not a validated measurement instrument. A search of PubMed for Gottman’s work on flooding returns no indexed papers using that construct.
What is measured is the arousal underneath it, and it is striking. In a 1983 study of 30 married couples, Levenson and Gottman tracked heart rate, skin conductance, pulse transmission time and somatic activity during conversations and found that “60% of the variance in marital satisfaction was accounted for using measures of physiological linkage alone” — that is, by how tightly the two partners’ bodies tracked each other, with no reference to what they said. The effect was strongest when couples discussed a real problem rather than the events of the day.
Read that carefully, because it is easy to overstate. It does not mean physiology causes 60% of marital happiness. It means that in this sample of 30 couples, a purely physiological measure statistically accounted for most of the variation in how satisfied they said they were. It is a small study from four decades ago. It is also the clearest available evidence that what happens in a difficult conversation is happening in two bodies, not just two minds.
So the mechanism is plausible and partly evidenced: arousal climbs, information stops going in, and the system does the one thing that reliably lowers arousal — it stops taking input. From the inside this feels like going blank, or like a wall coming down that you did not build. From the outside it looks like contempt.
Is stonewalling abuse?
This is the question people actually type, and it deserves a straight answer: sometimes, and the criteria are knowable.
Stonewalling as Gottman described it — a flooded partner going offline mid-argument — is not abuse. It is a bad conflict habit with real costs, most often performed by whoever is most overwhelmed. Deliberate, patterned silence used to punish, destabilise or control someone is a different behaviour that happens to look identical from across the room. That second thing can absolutely be abusive, and the research community has moved toward saying so explicitly.
A 2026 review in Current Opinion in Psychology, co-authored by Kipling Williams — the psychologist whose research programme defined the study of ostracism — argues that “behaviors like silent treatment, stonewalling, and emotional withdrawal are documented in intimate partner violence research” but have not been explicitly recognised as ostracism, and that these withdrawal-based behaviours “constitute a significant form of partner maltreatment.” The review notes that victims who experienced both physical abuse and ostracism rate the ostracism as more damaging, especially from a romantic partner, and that when withdrawal co-occurs with coercive control that isolates someone from outside support, it “blocks the recovery pathway.” This is a review article arguing a case, not a new trial — but it is a serious case, from the person best placed to make it.
So both things are true, and the honest version of this article gives you criteria rather than a verdict. English and Welsh law offers a useful frame here, because the offence of controlling or coercive behaviour under section 76 of the Serious Crime Act 2015 turns on exactly the right questions: is the behaviour repeated or continuous; does it have a serious effect — meaning fear of violence on at least two occasions, or serious alarm or distress with a substantial adverse effect on day-to-day activities; and does the person know or ought to know it will have that effect.
| Ask | Overwhelm shutdown | Coercive silence |
|---|---|---|
| When does it start? | Mid-conflict, once things heat up | Often before conflict, or after a boundary you set |
| How long? | Minutes to hours; comes back on its own | Days; ends when you concede or apologise |
| Is it targeted? | Happens with most difficult conversations | Deployed selectively, for particular topics |
| What happens when you name it? | Embarrassment, explanation, some effort to change | Denial, or the silence extends as a consequence |
| Does it stand alone? | Yes — one bad habit in a mixed relationship | No — sits alongside monitoring, isolation, finances, threats |
The last row does most of the work. Isolated shutdown is a communication problem. Silence that is one instrument in a wider pattern of control is not a communication problem, and no amount of better conflict technique will fix it. If the right-hand column is describing your relationship, the useful next step is a domestic abuse service, not a communication skill: in the UK, the National Domestic Abuse Helpline is 0808 2000 247 (24 hours); in the US, the National Domestic Violence Hotline is 1-800-799-7233.
Stonewalling, the silent treatment, and ghosting
People ask how these differ, and the differences are real.
| Present? | Usually | |
|---|---|---|
| Stonewalling | Yes — in the room, not responding | Reactive shutdown during conflict |
| Silent treatment | Yes — withholding contact while available | A deliberate withdrawal of connection, often as leverage |
| Ghosting | No — disappears entirely | Ending a relationship by exit rather than by conversation |
They shade into each other, and one can become another: a shutdown that lasts three days has stopped being a shutdown. But the distinction to hold on to is intent and duration. Stonewalling is something that happens to a conversation. The silent treatment is something done to a person.
Either way, being on the receiving end is not a small thing. Williams’s 2007 review of the ostracism literature found that being ignored or excluded “signals a threat for which reflexive detection in the form of pain and distress is adaptive for survival”, and that chronic exposure “appears to deplete coping resources, resulting in depression and helplessness.” If you feel disproportionately wrecked by someone going quiet, that reaction is doing exactly what it evolved to do.
Why do men stonewall? The honest answer
You will see a specific claim repeated constantly: that 85% of stonewallers are men. That figure comes from Gottman’s popular writing, and it does not trace to a peer-reviewed finding we could verify. We are not going to repeat a number we cannot source.
What is in the literature is more interesting anyway. In a 1994 study of 151 couples in long-term marriages, Levenson, Carstensen and Gottman found that “husbands reported feeling more negative the more they were physiologically aroused; for wives, affect and arousal were not correlated.” In other words, for the men in that sample, physical arousal and feeling bad moved together; for the women, they did not. If your body’s alarm and your sense of “this is going badly” are welded together, shutting down the conversation is a much more attractive move.
But the pattern is not a male one. Stonewalling usually appears inside the broader demand/withdraw pattern — one partner pressing for engagement, the other retreating — and that pattern has been meta-analysed properly. Schrodt, Witt and Shimkowski’s 2014 review of 74 studies (N = 14,255) found a moderate association between demand/withdraw and poorer individual, relational and communicative outcomes overall (r = .360), and — this is the part that matters here — near-identical effects whichever way round the roles fell: r = .380 for wife-demands/husband-withdraws, and r = .392 for husband-demands/wife-withdraws. The effect was larger in distressed and clinical samples (r = .413) than in non-distressed ones (r = .345). These are correlations, not causal estimates.
The takeaway: whoever is withdrawing, the pattern costs about the same. It is a role in a dynamic, not a gender trait.
What often decides who takes which role is what each person’s alarm system does with an unresolved conversation. Someone whose attachment system runs hot will find an unfinished argument close to unbearable and will pursue; someone whose response to the same threat is to go quiet will retreat. Neither of them is choosing the role on purpose, which is why the pattern is so hard to argue your way out of from inside it.
Does stonewalling always wreck a relationship?
No — and the evidence against the doom framing comes from Gottman’s own primary research.
In a 1993 study following 73 couples over four years, Gottman identified five types of couple and found that three of them were stable: “validators, volatiles, and avoiders”. Avoiders — couples who characteristically do not engage conflict head-on — were one of three viable adaptations, not a failure mode. The two unstable types were the hostile and the hostile/detached. Detachment paired with hostility predicted trouble. Avoidance on its own did not.
Newer work says something similar with more precision. In two studies (162 and 151 couples), Sasaki and Overall found that a partner’s withdrawal predicted lower perceived responsiveness and lower relationship satisfaction specifically when the withdrawing partner’s other half was behaving in a hostile, critical way and being pressed to change — the effect did not show up in every conflict context. Withdrawal is not universally corrosive. It is corrosive when it reads as I am not going to respond to you.
Which points at the actual problem. The damage is rarely the pause. The damage is the message the pause sends — that the person in front of you does not merit a reaction — and that message can be corrected without either of you learning to stay in a conversation you cannot physically stay in.
What to do if you are the one shutting down
This is the half almost nobody writes, which is odd, because it is the half that can actually change.
Name it before you go, not after. The single highest-value sentence in this whole topic is some version of: “I want to keep talking about this and I have hit my limit. I need twenty minutes. I will come back at nine.” That converts unresponsiveness into a stated intention. It is not a magic phrase; it is simply the difference between a wall and a door.
Come back when you said you would. The promise is what makes the break tolerable for your partner. A break you do not return from is the silent treatment, whatever you meant by it.
Actually lower the arousal during the break. Not by rehearsing the argument — that keeps it high. Walk, breathe out slowly for longer than you breathe in, do something with your hands. Our guide to de-escalating a heated conversation goes into the mechanics.
Look at what the shutdown is protecting. For a lot of people this is old. If conflict at home once reliably went somewhere frightening, going blank was a good adaptation that has outlived its usefulness. That is worth understanding in its own right — and it is often the same root system behind the relationship patterns that keep repeating.
What to do if you are on the receiving end
Stop pursuing during the shutdown. This is the hardest instruction in the article and the most reliable. Once someone is flooded, more words do not get in; they raise the arousal that caused the shutdown. You are not conceding by stopping. You are ending a loop.
Raise it later, as a pattern, not as an incident. “You did it again last night” invites a defence of last night. “When conversations end with you going quiet, I end up feeling like I don’t matter, and I want to find a version of this that works for both of us” invites a conversation about the pattern.
Ask for the return time, not for the conversation. Negotiating when you come back to it is far easier than negotiating whether someone can stay in a conversation their body has already left.
Watch what happens next. A partner who is flooding will often be relieved that you have named it. A partner using silence as leverage will treat your naming it as an escalation. That response is diagnostic, and it is worth paying attention to.
If the person going quiet is also the person you would normally process this with, you are in a genuinely lonely position. Thinking it through out loud — with a friend, a therapist, or an AI coaching and therapy tool like aidx.ai — is a reasonable way to work out what you are actually dealing with before you decide what to do about it. What none of that can do is be couples therapy: there is no evidence base for AI-mediated work with both partners in the room, and we are not going to pretend otherwise.
What the evidence says about getting help
Two honest notes, because this is where the internet overclaims hardest.
First, the famous claim that Gottman’s research can predict divorce with over 90% accuracy should not be relied on. Those figures came from statistical models fitted to samples already collected, and when Heyman and Smith Slep tested such models properly, they found that “accuracy and predictive value drops precipitously during crossvalidation”, concluding that results without crossvalidation “should be interpreted with extreme caution, no matter how impressive the initial results appear to be.” The observational findings above stand. The prediction headline does not.
Second, withdrawal is changeable, and the best evidence for that does not come from the Gottman method. In a randomised trial of 134 distressed couples assigned to traditional or integrative behavioural couple therapy, observed negativity and withdrawal decreased from the end of therapy through a two-year follow-up, with integrative behavioural couple therapy showing better maintenance of gains — though the same authors found only limited links between communication and outcomes five years out.
A later analysis of that trial contains the most useful practical finding in this whole literature: relationship outcomes were “more consistently linked with constructive communication than with destructive communication.” Adding the good tracked outcomes better than removing the bad. If you take one thing from this article beyond the abuse distinction, take that: the goal is not to become a couple that never shuts down. It is to become a couple with enough working conversation that a shutdown is an event rather than the weather.
Structured couples education can help too, with a caveat about what kind of evidence supports it. A 2024 study of the 12-hour Gottman Seven Principles course, with 490 participants and 242 controls, found it “improves couple relationships and is equally effective whether delivered in person or online” — and, interestingly, that it did not matter whether a clinically trained therapist delivered it, “suggesting that the material itself is sufficient.” That study used propensity-score matching rather than randomisation, so it controls for measured confounders only. Note also that this is the psychoeducational course. Gottman Method Couples Therapy itself has no randomised controlled trial in the PubMed-indexed literature, which is not proof none exists — PubMed does not index the entire family-therapy field — but it does mean the phrase “evidence-based” is doing more work than the evidence supports.
Common questions
What does stonewalling do to a person? Being ignored threatens belonging, self-esteem, a sense of control and a sense of meaningful existence, and chronic exposure depletes coping resources — Williams’s review associates it with depression and helplessness. Feeling badly affected by it is not oversensitivity.
Is stonewalling toxic behaviour? Unmanaged, it is a genuinely destructive habit: it reliably communicates unresponsiveness, which is what the research links to declining satisfaction. Whether it is toxic in the sense people usually mean — deliberate harm — depends entirely on the criteria in the table above.
Is stonewalling a trauma response? It can be. Shutting down under threat is a well-recognised protective pattern, and someone whose early experience of conflict was frightening may have learned it young. That explains the behaviour; it does not excuse its effects, and it is not a diagnosis you can make from a checklist.
How long does stonewalling last? A flooding-driven shutdown typically resolves within minutes to a couple of hours once arousal falls. Silence that runs for days is doing something else.
What are the long-term effects of stonewalling? In the couple, an accumulating sense of unresponsiveness that tracks with falling relationship satisfaction. In the person on the receiving end, the ostracism literature points to depleted coping resources and, with chronic exposure, depression and helplessness.
How do I deal with a partner who stonewalls? Stop pursuing in the moment; raise the pattern later when you are both calm; agree a return time rather than arguing about whether the conversation can continue; and pay attention to how they respond when you name it.
Can stonewalling be fixed? Withdrawal decreased measurably in a randomised couples-therapy trial and stayed lower two years later, so yes, it is a changeable behaviour — provided it is a communication problem rather than one arm of a control pattern.
The line worth holding
Stonewalling is not a personality type and it is not automatically abuse. Most of the time it is a body that has run out of room, doing the only thing it knows how to do — which is a fixable problem, and a much less frightening one than the internet suggests. Sometimes it is silence used as an instrument, which is a different problem entirely and does not respond to better communication skills.
You are allowed to work out which one you are in. The criteria above are the honest version of how.
Last reviewed: August 2026
This article is general information about relationship communication, not professional advice, diagnosis or treatment. If conflict in your relationship involves fear, control, or your safety, contact a domestic abuse service — the National Domestic Abuse Helpline (UK) on 0808 2000 247, or the National Domestic Violence Hotline (US) on 1-800-799-7233. If you are in immediate danger, call your local emergency number.
References
- Gottman JM, Levenson RW. Marital processes predictive of later dissolution: behavior, physiology, and health. J Pers Soc Psychol. 1992;63(2):221-33.
- Levenson RW, Gottman JM. Marital interaction: physiological linkage and affective exchange. J Pers Soc Psychol. 1983;45(3):587-97.
- Levenson RW, Carstensen LL, Gottman JM. The influence of age and gender on affect, physiology, and their interrelations: a study of long-term marriages. J Pers Soc Psychol. 1994;67(1):56-68.
- Gottman JM. The roles of conflict engagement, escalation, and avoidance in marital interaction: a longitudinal view of five types of couples. J Consult Clin Psychol. 1993;61(1):6-15.
- Heyman RE, Smith Slep AM. The hazards of predicting divorce without crossvalidation. J Marriage Fam. 2001;63(2):473-479.
- Schrodt P, Witt PL, Shimkowski JR. A meta-analytical review of the demand/withdraw pattern of interaction and its associations with individual, relational, and communicative outcomes. Communication Monographs. 2014;81(1):28-58.
- Dvir M, Beşikci E, Williams KD. The quiet cruelty: ostracism as intimate partner violence. Curr Opin Psychol. 2026;68:102229.
- Williams KD. Ostracism. Annu Rev Psychol. 2007;58:425-52.
- Sasaki E, Overall NC. Partners’ withdrawal when actors behave destructively: implications for perceptions of partners’ responsiveness and relationship satisfaction. Pers Soc Psychol Bull. 2021;47(2):307-323.
- Baucom KJW, Sevier M, Eldridge KA, Doss BD, Christensen A. Observed communication in couples two years after integrative and traditional behavioral couple therapy. J Consult Clin Psychol. 2011;79(5):565-76.
- Baucom KJW, Baucom BR, Christensen A. Changes in dyadic communication during and after integrative and traditional behavioral couple therapy. Behav Res Ther. 2015;65:18-28.
- Zahl-Olsen R, Thuen F, Bertelsen TB. The effectiveness of the in-person and online Gottman Seven Principles Couple Enhancement Program: a propensity score matching design. J Marital Fam Ther. 2024;50(4):882-898.
- Serious Crime Act 2015, section 76: controlling or coercive behaviour in an intimate or family relationship. legislation.gov.uk.
