Enough Isn't a Number You Can Borrow

She read the number in an article at midnight, alone with her phone, and before she could stop it, it landed like a grade posted somewhere she couldn't see the sheet. Once a week. Twice. Whatever the number was, it wasn't the number that unsettled her — it was the immediate, involuntary act of comparison, the fast subtraction of what she and her partner actually did from what the number said they were supposed to do.

That grading reflex is the actual subject of this essay, not the number itself. Somewhere along the way, a count of how often two people have sex got quietly promoted from a fact about their week to a verdict on their relationship — proof, or its opposite, that something between them is working. The promotion happened so gradually that most couples never noticed it happen, and almost nobody asked whether a count was built to carry that weight in the first place.

It was not. What follows is not a case for having more sex, or less, or a specific amount. It is a case for taking the number back off the witness stand — for separating the plain fact of how often something happens from the verdict it has been quietly deputized to deliver, and putting something sturdier in the verdict's place.

Ask a couple why the number matters to them and the honest answer is rarely about the number itself. It's about what the number is standing in for — attraction, commitment, fairness, whether the early feeling is still there, whether one partner has quietly stopped wanting the other. The count becomes a single, checkable figure that promises to answer five much harder questions at once, which is exactly why it is so tempting and so unreliable: a number asked to answer five questions rarely answers any of them correctly.

This is a case of what statisticians would call construct validity failure and what everyone else has simply lived through: a number can be accurately recorded and still measure the wrong thing. Frequency records occurrence. It does not record why an occasion happened, what either person felt during it, whether it was wanted equally by both, or what either partner's life looked like that week. Two couples can report an identical number and be living in entirely different relationships.

None of this means frequency is meaningless — a number that has fallen sharply and stayed there is real information, and this essay is not arguing otherwise. It means the number cannot do, by itself, the interpretive work it is routinely handed. It needs a verdict standing next to it before it can mean anything, and the verdict has to come from somewhere other than the number.

The single most useful piece of evidence available on this question comes from psychologists Amy Muise, Ulrich Schimmack and Emily Impett, who examined survey data from more than thirty thousand people across three separate studies to test a simple question: does more sex reliably mean more well-being. Their answer, published in 2016, was more precise than either side of the usual argument expects. The relationship between sexual frequency and well-being was curvilinear, not straight-line — well-being rose steadily with frequency up to roughly once a week, and then flattened. People having sex several times a week were not measurably happier, on average, than people having it about once.

The mechanism the researchers identified runs through relationship satisfaction rather than sex itself: frequency predicted well-being largely because it predicted feeling satisfied in the relationship, and that satisfaction is what did the actual work. Past the roughly-weekly mark, additional frequency stopped adding satisfaction, which is a specific, falsifiable finding rather than an impression, and it is the finding that most directly undercuts the belief that a higher number is always the healthier answer.

The honest caveats matter here. This is population-level survey data, not a controlled experiment on any particular couple, and an average plateau says nothing about where any two specific people's own comfortable range sits — some couples are well above it and thriving, some well below it and thriving, because the study measured a population curve, not a prescription. What it reliably establishes is narrower and still useful: past a fairly modest point, chasing a bigger number stops being the lever anyone thinks it is.

The impulse to settle this with an external, official figure is not new, and looking at what previous attempts actually produced is instructive. Historically, several legal and religious traditions tried to write a version of this question into law — not as an ongoing rate, but as a single binary threshold: did consummation happen at all. In Catholic canon law, a marriage that was validly entered into but never consummated is termed ratum sed non consummatum, and under the 1983 Code of Canon Law such a marriage can be dissolved by papal dispensation in a way an otherwise valid, consummated marriage cannot — the presence or absence of that one occasion carried real legal weight.

For centuries, the mechanism for establishing that absence was startling by any modern standard: documented cases between the fifteenth and eighteenth centuries describe a formal "trial by congress," in which a panel of clergy and physicians was convened to witness an attempt at consummation directly, in order to certify whether it had occurred. The practice was eventually abandoned as the indignity it plainly was, but the impulse behind it — settle the private question with an external, official verdict — did not disappear with it.

What every version of this history has in common is instructive by its limits: none of it ever attempted to specify an ongoing rate. The law could certify that something had happened once, or had never happened at all, because a binary is verifiable by outsiders in a way a private rhythm between two people is not. The number this essay is actually concerned with — how often is enough, on an ongoing basis, for these two specific people — has never once been successfully legislated, borrowed from a population average, or certified by a panel, because it is not the kind of fact that lives outside the relationship producing it.

Absent an official figure, couples import an unofficial one, and the sources are worth naming plainly because none of them describe the two people doing the comparing. A population average collapses everyone from newlyweds to grandparents, from long-distance couples to people who share a house with three children, into a single mean that describes nobody's actual week. A friend's disclosed number is filtered through the specific kind of exaggeration or minimisation that particular friendship rewards. A quiz or an app supplies a number engineered to be shareable, not accurate. A past self or a past relationship compares a current, real partnership against a memory that has had years to edit itself into something uncomplicated.

Every one of these comparison classes shares the same defect: it locates the answer somewhere other than in the two people it is supposed to be about. A couple who imports a number this way is not measuring their own relationship. They are measuring the distance between their relationship and a stranger's average, which is a real distance but not a meaningful one, because nothing about a population mean is entitled to set the terms for any specific pair inside it.

The corrective is not to stop comparing — comparison is how humans orient themselves to almost everything — but to change what gets compared against what. The number worth measuring a couple's frequency against is not a borrowed one. It is the number the two actual people involved would each, independently and honestly, describe as comfortable, which is a figure that exists nowhere outside the relationship and has to be produced by it rather than looked up.

The practical version of this correction is a specific, two-step exercise, and the order of the steps matters more than it looks. Each partner first states their own comfortable range privately — a number or range they would be genuinely content with, arrived at without looking at the other person's face, without guessing what answer is wanted, and without the number they think a healthy couple is supposed to report. This step fails if done out loud together, because the moment one partner speaks first, the second partner is no longer reporting a private number; they are negotiating in real time, which is a different and premature exercise.

Only once both numbers exist independently do the two get compared, and what usually turns up is not a dramatic gap but a workable overlap that neither partner expected, because most of the anxiety leading up to the comparison was manufactured by not knowing rather than by any real incompatibility. Where a genuine gap does appear, it is at least now a specific, nameable gap between two honest figures, rather than a vague, unmeasured sense that something is wrong — and a specific gap is something two people can actually negotiate, where a vague feeling is not.

This is a negotiation, not an average imposed on both people. A number exactly between two ranges satisfies neither person if it sits outside what either one called comfortable; a workable target usually sits inside the overlap where the two ranges actually meet, and where no overlap exists at all, that is real, useful information about where the actual conversation needs to happen — not about frequency as a topic, but about whatever is producing the wider gap.

A second, complementary exercise addresses a separate and common problem: couples frequently argue about a number neither of them has actually counted. Memory of frequency is unreliable in a specific, predictable direction — it tracks emotional weather more than it tracks the calendar, so a rough patch gets remembered as sexless even when it wasn't, and a good stretch gets remembered as more frequent than the record would show.

The corrective is mechanical rather than emotional: for thirty days, note the date each time it happens, without commentary, without a target in mind, and without showing the tally to the other person until the month is over. The rule that makes this work is that the counting has to happen before any conversation about what the number should be, not during one — a count taken while trying to prove a point is not a count, it is evidence-gathering for an argument already underway.

What the exercise usually produces is a number closer to the middle than either partner's memory suggested, because memory rounds toward whichever story is currently being told about the relationship. A couple negotiating from an accurate thirty-day count is negotiating about something real. A couple negotiating from two competing memories is arguing about two different, half-invented relationships that happen to share an address.

This is not a case that more is always better, or that a couple below the roughly-weekly mark identified above is missing something they need to fix — the same research this essay leans on found well-being flattening past that point precisely because more frequency stops being the lever, in either direction. Nor is it license to present a partner with a study and an ultimatum; a citation is not a form of pressure this essay endorses, and a partner who hears "the research says we should" is being handed a chore with footnotes, not an invitation.

It is also not a substitute for the different, harder question of why a number dropped in the first place, where the drop traces to something other than two comfortable ranges failing to align — resentment, exhaustion, an unaddressed hurt, or a genuine and specific desire mismatch of the kind this Library has already examined in detail elsewhere, in the essay on why she has more veto conditions rather than a different kind of desire. That essay's mechanism — an unexplained no as complete information rather than a puzzle — is credited here rather than repeated, and this essay's exercises assume that mechanism is not the live issue; where it is, the inventory in that essay is the more useful starting point than anything offered here.

And it is not medical or relationship therapy. Where a gap persists after an honest negotiation from real numbers, where desire has dropped for reasons connected to health, medication, past harm, or a relationship strain deeper than a frequency mismatch, that belongs with somebody trained, and an essay about interpreting a count is not equipped to replace that conversation.

This house sells single, bounded hours, and it is worth stating plainly what that structure can and cannot say about the question this essay actually addresses.

A session records nothing about a couple's own frequency question, one way or the other — it is not a data point in that negotiation, favourable or unfavourable, because it describes a transaction between different people under different terms entirely, and treating it as evidence about what two partners owe each other would be a category error this house has no interest in encouraging.

What a session can occasionally supply is something narrower and more honest: an uncomplicated hour against which a person can notice her own actual baseline, away from anyone else's number, anyone else's need, or anyone else's disappointment — which is useful only insofar as it clarifies her own comfortable range for the exercise above, and not because it settles anything about the relationship the exercise is actually meant to serve.

This piece takes up a question this Library deliberately set aside until it could be given a dedicated evidence pass rather than a quick restatement, consolidating several strands on interpreting sexual frequency into one essay, with real citation discipline throughout.

The claim draws on Muise, Schimmack and Impett's (2016) three-study finding that sexual frequency predicts well-being only up to roughly once a week, after which further frequency stops adding to it; on the historical record of canon law's ratum sed non consummatum provision and the documented "trial by congress" practice, offered as color illustrating that external, official thresholds have a long history of certifying only whether something happened once, never an ongoing rate; and on this essay's own negotiated-comfort-range and thirty-day-count exercises as practical applications of that research rather than directly tested interventions.

Explicit limits carried throughout: this is not a case for more frequency as a default good, not license to weaponise research in a negotiation, and not a substitute for the different question of why a number dropped, which this Library has already addressed at length elsewhere — credited here, not restated. Where a persistent gap traces to health, past harm or deeper relational strain, this essay routes that to somebody trained rather than attempting to resolve it. The house section discloses precisely why a single session is not, and does not claim to be, evidence in a couple's own frequency negotiation.