Fading attraction

Why do the very traits that first attracted someone to a partner stop being noticed?

Someone falls for a partner's calm. Ten years later they can list what they love, and calm is not on the list — not because the partner became agitated, but because the calm has become invisible. The trait did not go anywhere. So what changed, if not the trait? And is anything actually lost, or merely re-catalogued?

Start with something more basic than romance. Put your hand in warm water and it feels warm; leave it there and within a minute it feels like nothing. The water has not cooled. What happened is that your nervous system stopped reporting the level and started reporting only departures from it. That is habituation, and it is close to a universal property of perception: constant stimuli stop producing responses. Ask why that would be built in, and the answer is almost forced. Information lives in change. A signal that never varies tells you nothing you do not already have, and continuing to spend attention on it is waste.

Now apply that to a person. The trait that attracted you was, at the time, news — unexpected, uncertain, evidence about a stranger. Every instance of it updated something. Ten years on, each instance updates nothing, because you already knew. So it stops recruiting attention, not because you value it less but because there is nothing left to learn from it — and the fade should therefore be sharpest for the traits that were most striking at the start, since those carried the most surprise.

There is a second thread, about arousal rather than attention. Early attraction is not a calm appraisal; it is agitated — preoccupation, intrusive thinking, physiological arousal, and, importantly, uncertainty about whether it is returned. Take the uncertainty away and a large part of the machinery has nothing to do. What made the trait thrilling was partly not knowing whether you would get it. Security is the goal of the whole enterprise, and it is also the thing that quiets the alarm.

Psychologists have a standard way of carving this up. Berscheid and Hatfield distinguished passionate love — intense, absorbing, arousal-laden, focused on union with the other — from companionate love, the affection and deep attachment felt for someone whose life is thoroughly entwined with your own. The typical trajectory has passionate love running high early and declining over months to years, while companionate attachment can grow across the same period. If that is right, the person who says "I don't notice his calm any more" may not have lost anything; they have moved to a mode where the calm is infrastructure rather than event.

There is a sharper version of the fade worth naming, because it is not habituation at all. Sometimes the trait keeps being noticed and changes sign: the spontaneity becomes unreliability, the steadiness becomes dullness. Diane Felmlee studied these "fatal attractions", where the quality that drew someone in is the one they cite on the way out — the same behaviour re-described once its costs have been lived with.

Now the honesty this topic needs, because decline is often described as a law and it is not. Acevedo and Aron's work on long-married people reporting continued intense romantic love found reward-related brain activity resembling that of new love, but without the anxious, obsessive component. So the finding is not "passion always fades" but something more interesting — the obsessive part reliably goes, the intense part need not. Two cautions: this evidence is largely self-report, and couples still together are a selected sample, so we are partly observing survivors.

And it responds to what people do. Aron's self-expansion model holds that relationships reward us partly by extending the self, and that this is what runs low once a shared life is fully known. Experiments in which couples do novel and arousing activities together, rather than pleasant familiar ones, report higher relationship quality afterwards — a modest laboratory result, not a guarantee, but real, and it points the right way: novelty is the input adaptation has been consuming.

The misconception to clear is that fading attraction is a verdict on the partner, or evidence that the feeling was shallow. Neither follows. Habituation is a property of the observer's perceptual and attentional system, and it operates on things we still value — you do not stop needing the calm, you stop registering it. The right test of whether something still matters is not whether you notice it but what happens when it is withdrawn, and by that test these traits usually turn out to be load-bearing.

Two refinements connect the pieces. First, adaptation acts on steady states, not on relationships as such, which is why anything variable — humour, conflict, a partner in an unfamiliar setting — resists it far better than a constant virtue does. Second, the decline of passionate love and the growth of attachment are separate curves, not one curve; treating a fall in the first as a fall in the whole is a measurement error that people make about their own lives all the time.

• Why misattributing arousal from an unrelated source can be read as attraction.
• Whether the obsessive component of early love is better understood as an attachment-formation process than as an emotion.

Every term the collection defines is gathered in the glossary.