When the Wanting Goes Quiet in a Marriage That Works

Two people do everything right. They build a life that is predictable, transparent and safe. They know each other’s moods from the sound of a key in a door. Nothing is hidden and very little is surprising, which was the entire objective and took fifteen years of patient work. And somewhere in there, without either of them deciding it, the wanting quietened.

They will almost certainly conclude that something is wrong with them, or with each other, or with the marriage. The less comfortable possibility is that they performed the task correctly and the task had a cost nobody mentioned.

Esther Perel put the tension into general circulation: security and eroticism draw on conditions that are partly opposed. Security is built from predictability, familiarity, transparency and proximity — you know where the other person is, what they will do, and what they are thinking. Desire, in most accounts of it, wants some of the opposite: distance, a degree of unfamiliarity, the capacity to be surprised by someone, a person who is still partly unknown.

It is worth flagging that her account is influential rather than settled — it is widely argued with, and nothing here depends on it being the final word. What it does well is explain a pattern that people otherwise experience as an unexplained personal failure, and an explanation that makes a common experience ordinary is worth having even while the theory behind it is contested.

The practical consequence is what matters. If the two run on partly opposed conditions, then a couple can be excellent at one and lose ground on the other without any failure of love, effort or attraction — and the more thoroughly they succeed at merging, the more reliably it happens.

The account above is the best-known one and it is not the only one, and a page that gave it a clear run without saying so would be selling a thesis rather than examining one. Perel’s security-versus-eroticism tension comes out of clinical practice: it is an influential reading of what therapists see, argued about at length, and not an experimental result.

The experimental literature that exists points, in places, the other way. Work by Gurit Birnbaum, Harry Reis and colleagues on **perceived partner responsiveness** — the sense that the other person understands what you care about, values it, and acts on it — finds responsiveness associated with HIGHER sexual desire, and the association is reported as stronger in women. The mediators they identify are worth having by name: feeling special, and rating the responsive partner as more desirable generally. On the face of it that is an intimacy variable increasing desire, which is the opposite of what a simple distance thesis predicts.

Both cannot be straightforwardly true, so the useful move is to ask what each is actually measuring. And the reconciliation, once looked at, is not a fudge: they are measuring different axes that the word "closeness" runs together. Predictability of BEHAVIOUR — knowing what will be said, what will happen, how the evening goes — is what a long partnership accumulates and what the first account says costs something. Separateness of PERSONHOOD — the other being somebody with their own standing, their own week, their own life that is not an extension of yours — is a different variable entirely, and it is what responsiveness actually requires in order to mean anything.

Read that way both findings say the same thing from opposite ends. Being valued by somebody whose valuation carries weight requires that they be somebody — a separate person, with their own judgement and their own alternatives, whose attention was given rather than owed. What kills desire is therefore not intimacy and not familiarity. It is the other person ceasing to be a person and becoming a fixture: reliably present, entirely known, and therefore no longer choosing anything. Which is also why the separateness in the next section works, and it is not distance that makes it work. It is that a person who has been somewhere you were not comes back as somebody again.

One piece of arithmetic belongs here because its absence is what turns an ordinary trajectory into a private emergency. Frequency in most long partnerships declines, the decline is steepest early rather than late, and this is one of the most consistently reported patterns in the whole literature on couples. A pair at year twelve comparing themselves to year one are comparing themselves to a period that does not recur for anybody.

And the part that does the work: frequency and satisfaction are only loosely coupled. Couples at the same frequency report wildly different levels of contentment, and couples whose frequency fell report both relief and grief and sometimes neither. So a number is a poor instrument for the question it is usually being used to answer, which is not "how often" but "is this all right with both of us".

Which gives the one substitution this section is for. The question that produces an answer is not how often do we, which invites a comparison to a population nobody has met. It is: is either of us going without something we want, and has that been said out loud. Two people can be at any frequency at all and have nothing wrong; two people can be at a frequency that would look unremarkable on any chart and have one of them quietly doing without for years.

The standard advice is to add: a weekend away, lingerie, a new technique, a date night in the calendar. Some of that is pleasant and none of it is wrong, but it is all working the accelerator, and what changed in a long marriage was usually not the accelerator.

Two things did change, and both are structural. The brakes came down — the shared logistics, the unresolved conversation, the child awake down the hall, the visible undone tasks, all of which the chamber’s essay on responsive desire handles in full. And the distance closed, which is the specific loss this page is about. A weekend away sometimes works, and when it does it is usually not the novelty doing it: it is a room in which neither person is a household manager and neither can see anything that needs doing.

If the distance closed, the intervention is distance, and it is worth saying plainly because it sounds like a symptom rather than a treatment. Separateness. Time apart that is not a rupture. Interests that the other person is not part of and does not fully understand. Friendships that are yours. A day where nobody knows exactly where you are.

The mechanism is not mysterious. You cannot be surprised by someone you are currently merged with, and a person who has been somewhere you were not comes back slightly unfamiliar. The couples who report this working describe the same small thing: the moment of seeing the other person across a room, before being recognised, as a stranger might.

The reason this advice is rarely given is that it is easy to hear as a threat, and in a relationship already strained it can be received as the opening move of a withdrawal. Which is a real risk and worth naming rather than skating past: separateness restores something only when it is arranged together and said out loud. Taken unilaterally and unexplained, it is not distance. It is the beginning of leaving, and the difference between the two is entirely in whether it was agreed.

One thing about this problem is specific to duration, and it is usually missed because it looks like the problem rather than a separate one sitting on top of it. In a short relationship a decline is an event. In a long one there is a record, and after enough of them neither person is reacting to tonight any more.

What happens is that both sides start managing a ledger. The one who initiates less often does so because each attempt now carries the weight of the last several, and a refusal costs more than it did when there was no history to add it to. The one who declines begins to feel every ordinary evening as a test she is failing, and a hand on a shoulder stops being a hand on a shoulder and becomes an opening move requiring an answer. At that point affection itself gets withdrawn, not because it is gone but because it has become load-bearing.

This is worth separating out because it has its own remedy and the remedy is not about desire at all. What has to be restored first is touch that is guaranteed to lead nowhere — said in words, in advance, and meant. Not as a technique for eventually getting somewhere, which both people will detect immediately and which converts the guarantee into a longer opening move. As the actual arrangement. The loop this breaks is mechanical: the ledger runs on ambiguity, and an unambiguous evening has nothing to enter into it.

And it is worth saying who this most often costs, because the ledger is not symmetrical in practice. The person who declines more usually ends up carrying the label of the one with the problem — regardless of why she is declining, and frequently without anybody ever having asked. Exhaustion, pain, resentment about something entirely unrelated, a body that changed and was never discussed: each produces the same observable behaviour, and the label absorbs all of them into a single fact about her. This library has separate essays on most of those causes. The point here is narrower: a ledger records outcomes and cannot record reasons, which is why the conversation below cannot be about the ledger.

Everything above is useless without a conversation, and the conversation is the part that does not happen, for a reason worth naming rather than treating as cowardice. The subject has no neutral opening. Every available first sentence sounds like either a complaint or a request, both of which put the other person into a defensive position before anything has been established, and both people know this, which is why nine years pass.

Three things make it possible, and none of them is courage. First, it is not held in a bedroom, not at night, and never within a day of a refusal — because in any of those settings it is a move in the ledger rather than a conversation about it. A walk, a car, a restaurant: somewhere neither person has to look at the other continuously and nothing is going to be proposed at the end.

Second, it is not about frequency, which is the trap. A negotiation about a number produces a number that neither person wanted and that becomes its own obligation. The subject is what each of you is actually short of, and the answers are frequently not the same kind of thing at all — one person short of being wanted, the other short of being left alone, a third thing short of an hour in which nobody needs anything. Those have different solutions and none of them is a number.

Third, and this is the one that actually unlocks it: it opens with the speaker’s own half rather than with an observation about the other person. Not "you never" and not "we have not", both of which are findings about somebody who is now going to dispute them. Something with no verdict in it at all — I have been missing this, and I have not known how to say so. There is nothing in that sentence to defend against, which is the entire mechanism, and it is also the only version that is reliably true.

Japanese has a word English lacks, and the absence in English is not neutral. 卒婚 was coined by 杉山由美子 in 2004: a marriage that continues while two people increasingly live their own lives, by mutual agreement. Graduation from marriage — not from the other person.

It should be held apart from 家庭内別居, which is the older and much bleaker term for two people living separately inside one household. The two get collapsed constantly and they are opposites in the respect that matters: 家庭内別居 is what happens when nothing was agreed and both parties stopped negotiating; 卒婚 is an arrangement, discussed, with terms. One is a failure state. The other is a design.

Having the word available does real work. In English, a couple who wanted this arrangement would have to describe it from scratch, against a vocabulary in which every available term — separated, estranged, drifting apart — is a failure word. In Japanese it can be named in two syllables and then discussed. Which is a small illustration of a large thing: what a language has a positive word for is what its speakers can propose without sounding like they are giving up.

None of which is a recommendation. 卒婚 suits some people and would be a disaster for others, and this page has no view on which. The point is narrower and holds either way: an arrangement you can name is an arrangement you can negotiate, and most couples who end up in 家庭内別居 did not choose it. They arrived there because the only alternatives they had words for were staying exactly as they were, or leaving.

Moonlight offers no marital repair, no counselling and no diagnosis of a relationship, and would refer anyone looking for those rather than improvise them. Where a couple needs a therapist, saying so is part of this work.

What exists is narrower and is the thing this page has been describing: hours in which neither person is a household manager, nothing visible needs doing, the logistics are somebody else’s, and the ending was settled in advance so nobody is watching for it. For a couple, the Experience for two is arranged that way on purpose — and separately, and equally legitimately, a woman may arrange an evening that is hers alone and not about the marriage at all. Both are ordinary. Neither requires anything to be wrong.

The useful thing to take is the smallest one. If the wanting went quiet in a marriage that is otherwise working, the first hypothesis should not be that somebody stopped loving somebody. It should be that two people did the merging thoroughly, as instructed, and nobody told them that the second thing needed a little room left over. That is a repairable situation, and it is a considerably kinder account of fifteen years of effort than the one most couples reach for at two in the morning.