When Communication Breaks Down: The Patterns That Keep Couples Stuck
Most couples in distress aren't failing to communicate — they're communicating in patterns that consistently make things worse. Understanding those patterns is the first step to changing them.
By Dr. Julie Rashkis, Psy.D. · Licensed Psychologist · Menopause Society Certified Practitioner · therapyformidlife.com
When couples describe their communication problems, they often use phrases like 'we just can't talk anymore' or 'every conversation turns into a fight.' But in my clinical experience, the more accurate description is usually this: they can talk — they just keep having the same conversation, in the same way, with the same outcome. The content changes. The pattern doesn't.
This distinction matters enormously, because it shifts the question from 'what are we fighting about?' to 'what are we doing when we fight?' The research on couples communication has been unambiguous for decades: it is not the presence of conflict that predicts relationship outcomes, but the specific patterns couples use to navigate it. Some patterns are corrosive. Others are reparative. And the difference between them is learnable.
This article draws on two of the most robustly replicated bodies of research in relationship science — Gottman's observational studies of couples in conflict, and the demand-withdraw literature — to explain what the most damaging patterns look like, why they are so hard to interrupt, and what the evidence says about changing them.
The Four Horsemen: What Gottman's Research Actually Found
Dr. John Gottman spent over four decades observing couples in his University of Washington research lab — a converted apartment where couples came to have ordinary conversations and disagreements while being filmed and monitored for physiological arousal. From this work emerged one of the most widely cited findings in relationship science: by observing just 15 minutes of a couple's interaction, Gottman could predict with approximately 94% accuracy whether they would divorce — not based on the topics they argued about, but on the presence of four specific communication behaviors.
He named them the Four Horsemen:
Horseman: Criticism
What it looks like: Attacking a partner's character rather than their behavior. Uses 'you always' and 'you never.' Implies something is fundamentally wrong with the person.
The antidote: Gentle startup — 'I feel...' statements that address specific behaviors without character attack.
Horseman: Contempt
What it looks like: Communicating from a position of moral superiority. Eye-rolling, sarcasm, mockery, name-calling. The single strongest predictor of divorce in Gottman's research.
The antidote: Building a sustained culture of appreciation and respect — noticing and expressing what's going right.
Horseman: Defensiveness
What it looks like: Responding to perceived attack with counter-attack or victimhood. Deflects responsibility. Escalates rather than resolves.
The antidote: Taking responsibility, even for a small part. 'You're right that I...' is a repair, not a capitulation.
Horseman: Stonewalling
What it looks like: Withdrawing from the interaction — shutting down, going silent, leaving. Often a response to physiological flooding, not indifference.
The antidote: Structured self-soothing: a mutually agreed 20-minute break with commitment to return to the conversation.
It is worth pausing on contempt specifically. Gottman describes it as the single most corrosive pattern — 'sulphuric acid for a relationship,' in his words. Partners exposed to contempt over time not only report lower relationship satisfaction; research has found they also show measurable immune suppression, suggesting that the physiological costs of a contemptuous relationship extend beyond emotional distress.
The Four Horsemen rarely arrive independently. More commonly, they cascade: criticism triggers defensiveness, which escalates to contempt, which eventually produces stonewalling. Over time this cascade becomes the default mode of interaction, and the couple enters what Gottman calls 'negative sentiment override' — a state in which even neutral behaviors from a partner are interpreted as hostile. In this state, even a partner's repair attempts fail, because they are not trusted.
"By observing 15 minutes of a couple's conflict, Gottman could predict with approximately 94% accuracy whether they would divorce — not based on what they argued about, but on how they argued." — Gottman Institute
The Physiology of Getting Stuck: Flooding and the Nervous System
One of Gottman's most important contributions was demonstrating that communication breakdown is not just psychological — it is physiological. When conflict activates the stress response, heart rate climbs, cortisol spikes, and the body enters a state of diffuse physiological arousal that Gottman called 'flooding.' At heart rates above approximately 100 beats per minute, the prefrontal cortex — the seat of reasoning, empathy,
and perspective-taking — begins to go offline. Productive conversation becomes neurologically impossible, not because someone is being stubborn or indifferent, but because their nervous system has shifted into a mode that prioritizes self-protection over connection.
Research on heterosexual couples has consistently found that approximately 85% of stonewallers are men — not because men care less, but because men's cardiovascular systems tend to activate more quickly and take longer to return to baseline during conflict. A man who stonewalls mid-argument is often not choosing to shut his partner out; he is overwhelmed in a way that makes continued engagement genuinely impossible in that moment.
Understanding the physiology of flooding has practical implications. It means that taking a break during a heated conflict is not avoidance — it is a biological necessity. It means that the timing of difficult conversations matters. And it means that for couples who have been in high-conflict or contemptuous patterns for years, the nervous system dysregulation itself has to be addressed, not just the communication skills.
The Demand-Withdraw Pattern: The Most Researched Cycle in Couples Conflict
Alongside Gottman's work, the demand-withdraw pattern is one of the most extensively documented communication cycles in the relationship science literature. A meta-analysis of 74 studies involving over 14,000 participants found an effect size of r = .36 for demand-withdraw predicting negative relationship outcomes overall, with the strongest associations for relationship satisfaction (r = .42).
The pattern is straightforward to describe and surprisingly hard to interrupt: one partner pursues — raising concerns, seeking discussion, expressing dissatisfaction — while the other withdraws — deflecting, going quiet, leaving the room. The pursuer experiences the withdrawal as abandonment or dismissal. The withdrawer experiences the pursuit as attack or pressure. Both responses intensify the other, creating a self-reinforcing loop.
Research by Christensen and colleagues established that in heterosexual couples, the woman-demand/man-withdraw pattern is more common than the reverse — likely a product of both attachment style differences and the socialization of emotional expression in men. However, longitudinal studies of middle-aged couples found that the pattern operates with equal frequency in both directions, and that its association with marital dissatisfaction holds regardless of which partner is in which role.
Crucially, the demand-withdraw pattern is associated not just with relationship dissatisfaction but with individual health outcomes: higher rates of depression, elevated cortisol and norepinephrine levels in the demanding partner, and increased physiological stress responses in the withdrawing partner. The pattern is corrosive for both people.
"A meta-analysis of 74 studies found that the demand-withdraw pattern predicted relationship outcomes with an effect size of r = .42 — one of the strongest associations in couples research."
Why These Patterns Are So Hard to Stop
Understanding a negative pattern and changing it are not the same thing, and it is worth being honest about why. Several factors make these cycles remarkably persistent.
First, the patterns are typically well-established by the time couples recognize them as problematic. Research has found that couples wait an average of six years after problems develop before seeking help. By that point, the negative interaction cycles have often become the default mode of relating — automatic rather than deliberate.
Second, both the demanding and withdrawing behaviors feel, from the inside, like reasonable responses to what the other person is doing. The partner who pursues is not trying to attack — they are trying to connect, to resolve, to be heard. The partner who withdraws is not trying to abandon — they are trying to regulate, to avoid escalation, to protect the relationship from further damage. Both people believe they are responding to the other's behavior. Neither recognizes themselves as originating the cycle.
Third, midlife stress amplifies these patterns. Research on aging and couples communication found that avoidance behaviors tend to increase over time as couples move from middle to later adulthood — a developmental shift that may reflect an attempt to reduce conflict in service of relationship preservation, but which, when combined with unresolved issues, can entrench the distance and isolation cascade described in the previous article.
Finally, for couples in which contempt has become established, the research is unambiguous: the pattern is unlikely to resolve without external support. Negative sentiment override is particularly resistant to self-correction, because the very goodwill required to attempt repair has been eroded by the contempt itself.
What the Research Says About Interrupting These Patterns
The evidence base for changing negative communication patterns is genuinely encouraging. The key findings include:
Gentle startups change the trajectory of conversations measurably. Gottman's research found that the way a conflict conversation begins predicts its outcome with high accuracy. Conversations that start with 'I felt hurt when...' land differently than those that begin with 'You always...' — and that difference compounds over time.
Repair attempts work when there is enough positive sentiment to trust them. Small moments of humor, connection, or acknowledgment during or after conflict — what Gottman calls repair attempts — are among the strongest predictors of couple resilience. But they only work when the relationship's emotional bank account has enough positive balance to make the repair credible.
Self-soothing during flooding is both necessary and effective. Taking structured breaks during heated conflict — a minimum of 20 minutes of genuine physiological calming, not mental rehearsal of arguments — allows the nervous system to return to a state where productive conversation is possible.
Attending to emotional experience, not just behavioral patterns, accelerates change. Research comparing behavioral and emotion-focused approaches suggests that targeting the emotional experience driving the demand-withdraw cycle — the fear beneath the pursuit, the shame beneath the withdrawal — produces more durable change than skill-building alone.
For couples who recognize these patterns in their own relationship, this last point has a practical implication: the goal is not just to argue more cleanly, but to understand what the argument is really about. Beneath most communication breakdowns is an unmet need, an attachment fear, or an injury that hasn't been acknowledged. When couples begin to hear each other at that level, the patterns that once seemed immovable often begin to shift.
About the Author
Dr. Julie Rashkis is a licensed psychologist and Menopause Society Certified Practitioner with over 20 years of clinical experience. She is the founder of Therapy for Midlife, a virtual practice offering individual and couples therapy for adults navigating the psychological, relational, and hormonal complexities of midlife. She is licensed in California and Wisconsin and sees clients across all PSYPACT-participating states.
www.therapyformidlife.com · Book a free consultation
References
1. Gottman, J. M. (1994). What Predicts Divorce? The Relationship Between Marital Processes and Marital Outcomes. Erlbaum.
2. Gottman, J. M., Coan, J., Carrere, S., & Swanson, C. (1998). Predicting marital happiness and stability from newlywed interactions. Journal of Marriage and the Family, 60(1), 5–22.
3. Gottman, J. M., & Levenson, R. W. (2000). The timing of divorce: Predicting when a couple will divorce over a 14-year period. Journal of Marriage and the Family, 62(3), 737–745.
4. Levenson, R. W., & Gottman, J. M. (1983). Marital interaction: Physiological linkage and affective exchange. Journal of Personality and Social Psychology, 45(3), 587–597.
5. Christensen, A., & Heavey, C. L. (1990). Gender and social structure in the demand/withdraw pattern of marital conflict. Journal of Personality and Social Psychology, 59(1), 73–81.
6. Schrodt, P., Witt, P. L., & Shimkowski, J. R. (2014). A meta-analytical review of the demand/withdraw pattern of interaction and its associations with individual, relational, and communicative outcomes. Communication Monographs, 81(1), 28–58. 7. Holley, S. R., Haase, C. M., & Levenson, R. W. (2013). Age-related changes in demand–withdraw communication behaviors. Journal of Marriage and Family, 75(4), 822–836. PMC3728718.
8. Eldridge, K. A., & Christensen, A. (2002). Demand-withdraw communication during couple conflict. In P. Noller & J. A. Feeney (Eds.), Understanding Marriage. Cambridge University Press.
9. Kiecolt-Glaser, J. K., Malarkey, W. B., Chee, M., Newton, T., Cacioppo, J. T., Mao, H. Y., & Glaser, R. (1996). Negative behavior during marital conflict is associated with immunological down-regulation. Psychosomatic Medicine, 55(5), 395–409. 10. Baucom, B. R., McFarland, P. T., & Christensen, A. (2010). Gender, topic, and time in observed demand-withdraw interaction in cross- and same-sex couples. Journal of Family Psychology, 24(3), 233–242.
