Courtship Is a Signalling Game, and the Signals Have Reasons

Most of what is written about courtship is either romance or advice, and both are useless for understanding it. Romance describes how it feels from inside. Advice tells you which buttons to press. Neither asks the interesting question, which is why the buttons exist at all.

Here is the question asked properly. Two people are each trying to make a decision with enormous consequences, a long partnership, possibly children, on the basis of very little information, from a counterparty who has every incentive to present well. That is a signalling problem under asymmetric information, and the behaviour that looks irrational or cruel from inside is, mostly, what you would design if you were solving that problem from outside. I want to walk through four pieces of it.

Why Sampling Several Candidates Is Rational

The complaint that a woman "tries out several men before choosing" is usually made as an accusation. It is a description of a correct search strategy.

Consider the problem she faces. The quality of a partner, reliability, kindness under stress, what he is like after the first six months, is not observable at the first meeting. It reveals itself slowly and only under varied conditions. Worse, the distribution is wide: the difference between a good partner and a poor one is enormous, and the cost of the poor one, historically, fell overwhelmingly on the woman. A pregnancy is not symmetric.

When the variance is high, observation is slow, and the cost of error is asymmetric, the optimal policy is to sample. You look at several candidates, you invest a little in each, you drop the ones that reveal a problem, and you commit only when the remaining evidence is strong. This is how anyone hires for an important job. The only reason it looks different in courtship is that the rejected candidates have feelings about it.

The sampling is not fickleness. It is the correct response to a decision that cannot be undone cheaply. A man who is offended by being one of several is offended by arithmetic.

Why the Cost of Leaving Has to Be Real

The second piece of the note is uglier as usually stated: create a feeling of loss in the other party if they leave. Put like that it is advice for a manipulator. Put analytically it is something else.

A partnership only works if both parties are actually committed, and commitment cannot be observed directly; anyone can say it. What can be observed is whether leaving would cost the other person something. If it would cost nothing, the commitment was never there, whatever was said. So each party is, throughout the courtship, trying to answer one question: would this person actually lose something if I walked away?

This is why the signals that matter are costly ones. Time spent that could have been spent elsewhere. Other options declined. Plans made that depend on the relationship continuing. Each of these is a stake placed on the table, and the stakes are what make the words credible. A person who has placed no stake and professes devotion is making a claim with nothing behind it, and the other party correctly discounts it.

So "making them feel they would lose something" is not a trick. It is the only honest way to demonstrate that you yourself have something to lose. The manipulative version fakes the stake. The honest version places it. From outside they look similar, which is why the manipulation works for a while; from inside, over time, they do not, which is why it stops.

Why Defection Must Be Terminal

The third point in the note is that a defection on the other side has to be understood as final: there is no coming back.

This is the iterated game, and it is the same logic that governs cooperation everywhere. A partnership is a long sequence of moments at which either party could defect, in small ways and large. What keeps defection rare is not goodwill but consequences. If defection is forgiven automatically, defection is free, and things that are free get consumed.

So the credible threat of exit is load-bearing. Not the threat made aloud, which is itself a kind of defection, but the understood fact that a real betrayal ends the game. A person who will forgive anything will, in the end, be given everything to forgive, not because the partner is wicked but because the incentive is there and incentives get taken. The relationships that last are the ones in which both sides know the floor is real, and therefore never go looking for it.

This is also why the credibility has to be established early and rarely tested. A threat that has to be repeated is a threat nobody believes.

Reading Emotional Statements in Context

The last point is the one most often gotten wrong, in both directions. Emotional statements, "I need space", "I do not know what I want", "you are too much", are treated either as literal reports to be acted on or as noise to be ignored. They are neither. They are signals, and signals mean what they mean in context.

A statement made in the middle of a fight is a move in the fight. A statement made calmly the following morning is a report. The same words carry different information depending on when they are said, to whom, and what came before, and a listener who decodes them without the context will get them systematically wrong. He will take a tactical retreat as a verdict and a verdict as a mood.

The analytic attitude here is not coldness. It is the recognition that a person under emotional pressure is not a reliable narrator of her own state, and neither is he, and that the reliable information is in the pattern rather than in any single sentence. What she does over a month tells you more than what she says in a minute.

What This Is Not

None of this is a manual. Understanding why a signal exists does not entitle you to forge it, and forgery is exactly what the "techniques" literature sells: how to look like you have a stake without placing one, how to appear to have options you do not have. That works about as well as counterfeit money works, which is to say until someone checks.

What the analysis buys you is the opposite of technique. It lets you stop being offended by behaviour that is merely correct, stop mistaking honesty for coldness and warmth for commitment, and place real stakes rather than fake ones. That is not a way to win the game. It is a way to understand which game you are in, and most of the misery in this area comes from people who do not.