Keeping attraction alive when it is not about looks
Long-term attraction tracks novelty and discovery more than appearance, and there is one unfamiliar thing worth trying this week.
- autopilot
- hermes
- attraction
- novelty
- long-term-desire
- shared-experiences
What you get from this
- Long-term attraction tracks how much you are still finding in each other, not how either of you looks.
- A small experiment with couples found novel, slightly nerve-racking shared activities moved relationship quality. Small sample, mixed replications.
- This week, book one thing you have never done together, chosen for the small nervousness rather than the romance.
- The first attempt will probably feel staged. That passes.

When did you last look at your partner and notice something you had not noticed before?
Early on the question is easy. A few years in it can take a while. The pause gets read as attraction fading, and fading gets blamed on time, on weight, on the ordinary grind of shared logistics. The first part is real. The second part is mostly wrong.
The insight
Long-term attraction is not a verdict on how either of you looks. It tracks how much each of you is still finding in the other.
That is an uncomfortable sentence, because appearance is the most visible variable in a relationship and the easiest one to blame. It is also the wrong lever. A face you have known for a decade will not surprise you again, however well it is kept. A person can, and only if that person is still showing you something you did not already have filed.
What the research actually shows
The clearest evidence here is about novelty rather than appearance. In a small experimental study published in the Journal of Personality and Social Psychology, the psychologist Arthur Aron and his colleagues gave about fifty couples a shared activity to do for ten weeks. One group was assigned something novel and mildly arousing, another something pleasant but familiar, and a third nothing in particular. The couples in the novel, arousing group reported the largest increase in relationship quality across the ten weeks.
Read that cautiously. The sample was small, and later attempts to repeat the result have been mixed. Two things survive the caution. The first is direction: the variable that moved was what the couple did together, not what either of them looked like. The second is the explanation underneath it. Aron and his co-authors built the work on what they call the self-expansion model, the proposal that a relationship stays compelling partly by continuing to widen what each person feels they are. Novelty does that work. Familiarity on its own does not.
The couples therapist Esther Perel makes a related argument from clinical observation rather than from research, and it is worth labelling it as exactly that. In her book Mating in Captivity she argues that desire needs some separateness to survive, that wanting someone depends on their not being entirely known or entirely available. Her phrase for it is that fire needs air. Treat it as the considered view of an experienced clinician, not as a finding from a study.
A second line of work looks at how partners see each other rather than how they look. In a longitudinal study of couples, the psychologists Sandra Murray and John Holmes found that people who rated their partners more favourably than the partners rated themselves reported higher satisfaction, and the pattern held over time. This is a correlation. It does not show that flattering thoughts cause a better relationship, only that the two travel together, and it should be read that way.
The experiment
One change this week, and it is not a gym.
Put one thing in the calendar that the two of you have never done together, and choose it for the small nervousness rather than for the romance. A climbing wall. A dance class neither of you can manage. A walk somewhere you have never been after dark. The test is not whether it sounds like a date. The test is whether you are both a little unsure you can pull it off.
Do it on the day you choose. Afterwards, notice two things, and consider saying them out loud. First, whether your partner seemed more like someone you were meeting than someone you were managing. Second, whether the ordinary evening that followed felt a little less ordinary than the ones before it.
The first attempt will probably feel staged, and that passes. What you are testing is not whether one evening goes well. You are testing whether doing something unfamiliar together changes how the two of you look at each other in the days after it.
Why this is a lever and not a trick
Nothing here asks you to pretend, and nothing here asks you to perform attraction you do not feel. The exercise is a test of one claim: that what drifts in a long relationship is not beauty but discovery. One unfamiliar afternoon is not proof of anything. It is a way to start finding out whether the claim holds for you, and which of the two you had actually been missing.
If you want to see where the two of you stand on this and a few other habits, the free Relationship Checkup takes about five minutes and stores nothing.
This is relationship education, not therapy or medical advice. If any part of this describes what is happening in your relationship, a qualified professional or a local support service is the right next step, and not an article.
One question a week.
Each Sunday: one question to ask your partner, the research behind it, and one thing to try. The free Relationship Checkup is open to everyone, any time.