Why couples fight about the same thing

In an argument the truth is lost. The one who ends the argument is the smarter one.

— L. N. Tolstoy

Seen from a distance, conflict looks like a sign of malfunction. A good couple should understand each other, a bad one should fight; good love smooths differences out, bad love keeps running into them. On that logic every serious argument looks like a crack, and a recurring fight looks like proof that these people got the wrong address. But where two people spend years sharing space, money, time, bodies, relatives, illnesses, children and their own bad days, a complete absence of friction is no less suspicious than constant screaming. If the fire alarm never goes off, there are two possible reasons: there are no fires, or the alarm is disconnected.

Conflict on its own diagnoses nothing. It is how a system reports a mismatch of interests, expectations, information or workload. In business, partners argue about strategy, investment, dividends and acceptable risk, and nobody treats the fact of an argument as proof of mutual hatred. At home the subjects look less dignified: who gets up for the child, whose job matters more when you move, how much to spend on a holiday, whether you can help your parents out of the shared budget, why one person calls spontaneity being alive and the other calls it badly organised logistics. Behind the domestic scenes there are usually different models of fairness.

An incomplete contract for an ordinary Tuesday

Most recurring fights grow in the gaps between agreements, not in their complete absence. A couple may have decided that housework is split evenly, but never decided who notices that something needs doing, who reminds, who checks and who picks up the failure. They may have agreed to support both careers, but set no rule for a simultaneous deadline. The contract exists at the level of principle and stays empty at the moment when a specific loss has to be assigned.

The gap is normally filled by the status quo. The work is done by whoever noticed first, tolerates mess worse, calms the child faster or fears being late more. In the short run this is efficient: the task is done. In the long run one person’s reliability becomes a free reserve for the other. The next conflict is no longer about a cup: the system has once again picked the same payer.

A promise to “fight less” rarely changes the circuit. What is needed are new decision rules: a rota, a minimum standard, an area of responsibility, a budget for outside help, a review date. This sounds prosaic, because conflict often is a bad protocol for a recurring load rather than an absence of love.

What exactly repeats

The commonest mistake is to take the content of a fight for its mechanism. A couple can argue about an unwashed cup for ten years, although the cup became an indicator of invisible labour long ago. You can argue about messages with an ex when the real subject is the loss of exclusive status; about money when the subject is safety, power or a fear of dependence; about sex when sex is already carrying exhaustion, resentment, self-esteem and the question of whether I am wanted at all. Daily life supplies convenient pretexts, but the mechanism lives deeper.

A recurring argument is either technical or constitutional. The technical kind comes back until duties, budget, schedule or help have been assigned. The constitutional kind is about the right to set the rules: who decides, whose discomfort counts for more, whether refusal is allowed and what counts as a sufficient contribution. A technical conflict is cured by rebuilding the task; a constitutional one reproduces itself even after a particular episode has been settled well.

The second type is a stable difference between people. One needs more closeness, the other more autonomy; one calms down by talking, the other by pausing; one grew up in a family where love was expressed by intervening, the other where love was proved by not touching anything unless necessary. Such differences do not disappear after one correct conversation. They can be acknowledged, fenced in with rules, compensated for, and then you can stop demanding every month that your partner become a different species of animal.

Couples often get the genre of the problem wrong. They try to solve a stable difference as a technical problem — “why can’t you just be more attentive?” — or declare a technical problem to be a feature of character: “that’s the sort of person I am”, when what is actually happening is a habit of offloading work. A stable difference calls for acceptance and management, a technical problem for a change of procedure; with the wrong diagnosis the treatment is useless.

Demand and withdrawal

The demand–withdraw pattern was long described as an almost natural pairing of female insistence and male silence. Research by Neil Jacobson, Andrew Christensen and colleagues did find an average sex difference, but it also showed the role of the topic and of who wanted the change. Whoever wants to change the current order pushes for the conversation more often; whoever finds the status quo more profitable or safer withdraws more often. When the problem under discussion was one the husband raised, the direction could reverse.

That is an important shift from character to position. Demandingness may be not a stable property of a person but the strategy of the party who needs a revision. Withdrawal is not simply an inability to talk but a way of protecting the existing distribution or reducing overload. Both strategies can be understandable and together produce a bad equilibrium: pressure makes retreat more profitable, retreat makes pressure stronger.

Breaking the cycle means changing the payoff of the moves. A pause must have a time to return, otherwise it becomes a veto by silence. A demand must contain a solvable request, otherwise the conversation turns into an endless trial of a personality. The parties are not obliged to agree on the substance straight away; they have to make taking part in the procedure more profitable than flight or escalation.

Repetition tells you that this is not a random failure but a feedback loop. One makes a move, the other answers predictably, the first gets his fears confirmed, the second hers, and the system returns to its starting point with a new layer of resentment. The person who fears distance demands conversation ever more sharply and gets still more distance. The person who fears pressure retreats into silence and gets still more pressure. The demand–withdraw pattern has been studied considerably better than most family labels: a meta-analysis of 74 studies and more than 14 thousand participants found a stable link between the pattern and worse individual, communicative and relational outcomes. Each sees in the other’s reaction the proof of being right, although both are producing exactly the result they fear.

“Just stop doing that” rarely works: it is addressed to one participant while the problem lives between two. Responsibility does not disappear, and some actions are unacceptable regardless of the dynamic, but for repairable conflicts it is more useful to work out which of your own actions raises the odds of the reaction you then hate. This is unpleasant, because it removes the luxury of being only the injured observer.

Repair or erosion

There are repairing conflicts and eroding ones. After a repairing one there is more clarity: people may argue harshly, get tired and take offence, but they come back to the task — what happened, which rule needs changing, who misread what. An eroding conflict does not solve the problem, it wears away respect. It produces no new agreement, but it leaves an extra layer of contempt, fear or the wish to show less of yourself. Research on joint coping with stress supports precisely this couple-level unit of analysis: a meta-analysis links a couple’s ability to notice a load, respond to it and act together with higher relationship satisfaction. These data do not allow you to rank repair and fight frequency universally. They show something narrower: a count of conflicts is not enough; you have to watch whether the couple produces new coordination after a load.

A bad conflict quietly changes the goal: instead of solving the task, the parties try to confirm that they are right. The partner turns into an opponent and then into a threat. Out come old mistakes, family histories, diagnoses and the words “always” and “never”, which rarely describe statistics and almost always add petrol. A win in an argument after which your partner is colder, further away and less willing to cooperate is a formal victory and an expensive defeat.

A relationship is a repeated game. In a one-off game it pays to squeeze out the maximum now; in a repeated one today’s victory can destroy tomorrow’s cooperation. If a person always gets their way through pressure, silence, the threat of leaving, financial leverage or sarcasm, they may win individual episodes. Meanwhile the partner learns that the system is unsafe and starts to resist, to hide, or to leave from the inside; victories that cost too much destroy the field of cooperation itself.

Silence does not always mean improvement. Sometimes an argument preserves at least a bad channel of communication, and the moment when one person stops arguing means not agreement but the loss of any expectation of being heard. A house goes quiet after the residents have emotionally evacuated, not after the fire has been put out. Conversely, not every irritation deserves negotiation: people get tired, get hungry, misread faces and sometimes want a fight simply because the nervous system needs an outlet. Noise has to be told apart from signal.

When talking does not solve the conflict

Talking is advertised as the universal solvent for conflict: find the non-violent phrasing, name the feelings, hear the needs, and the contradiction starts to dissolve. That belief did not come from nowhere. But research on couples also produces a more awkward result: the same behaviour can have different effects depending on how serious the problem is. Gentle smoothing helps with a small and temporary conflict, whereas with a serious repairable problem direct dissatisfaction and pressure to change sometimes predict better outcomes, because they stop the system from keeping a convenient defect. Some recurring fights really are fed by missing or distorted information. One takes silence for respect of the other’s space, the other takes it for punishment; one waits for help that seems obvious, the other takes the absence of a request for the absence of a task; one hides a financial problem so as not to worry a partner and thereby produces exactly the distrust he wanted to protect the union from. Here talking changes the map: the participants finally see what was happening in the system and get material for a new procedure.

But communication fixes a shortage of information and is under no obligation to remove a conflict of interest. Two people can understand each other flawlessly and still want incompatible things: one children, the other a childless life; one to move, the other to stay near a sick parent; one monogamy, the other some other form of union; one to put the coming years into a career, the other into a family project. Good words make positions clearer, but they do not create half a child, two permanent countries of residence, or a form of exclusivity that both exists and does not. Sometimes the best conversation does not save the agreement but ends a convenient ambiguity and shows that the argument was not the result of poor speech technique but a real choice between two futures.

Before the next discussion it is worth telling three tasks apart. In the first the parties lack facts, or they read the same signal differently; here talking really is the main instrument. In the second the goals coincide but the procedure produces a bad result; then you need a schedule, a budget, a rule about pauses, a division of labour or some other way to coordinate. In the third the parties understand each other well enough, but the gain and the loss are distributed incompatibly; talking can help agree on compensation, an order of priority, a boundary or an exit, but the conflict itself will not disappear because of an empathetic phrasing. To demand in that situation that people “just learn to communicate” is to treat a bargaining problem as a defect of vocabulary.

After a discussion of a recurring problem, two things should change: the emotional temperature and the procedure. If people have said everything they said before but have not changed the schedule, the rule, the boundary, the way of warning about overload or the permitted length of a pause, the system will almost certainly reproduce the same conflict. A good reconciliation without a change of procedure is sometimes necessary, but it only postpones the next run of the same family play.

Main sources

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.

Falconier, M. K., Jackson, J. B., Hilpert, P., & Bodenmann, G. (2015). Dyadic coping and relationship satisfaction: A meta-analysis. Clinical Psychology Review, 42, 28–46.

McNulty, J. K., & Russell, V. M. (2010). When “negative” behaviors are positive: A contextual analysis of the long-term effects of problem-solving behaviors on changes in relationship satisfaction. Journal of Personality and Social Psychology, 98(4), 587–604.

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; Christensen, A., & Shenk, J. L. (1991). Communication, conflict, and psychological distance in nondistressed, clinic, and divorcing couples. Journal of Consulting and Clinical Psychology, 59(3), 458–463.