How to Know When Something Is Over

The difference between a bad patch and an ending.

“Endings are not always bad. Most times they are just beginnings in disguise.”

— Attributed to Anonymous

Almost nobody knows how to tell the difference between a relationship going through something difficult and a relationship that has ended, and the uncertainty itself does a great deal of damage. People stay for years in something that finished long ago, and they leave things that would have recovered given attention, and in both cases the mistake is the same: a decision made on the strength of how a particular fortnight felt rather than on any examination of the actual pattern. Uncertainty itself does damage independent of whatever the answer turns out to be.

What makes this genuinely hard is that both situations produce similar symptoms in the short term. A couple in a difficult year and a couple at the end of something both argue more, talk less, feel distant and wonder privately whether this is it. Feelings are a poor instrument here, because they are heavily influenced by whatever happened this week, and the question being asked is about a decade. A bad Sunday and a good Thursday produce opposite answers about the same decade.

This report is about the distinction. It covers the signals that reliably indicate something recoverable and the ones that indicate something finished, why the amount of time already invested is the least useful input available, what to do about the very common situation where the two people disagree about the answer, and how to reach a conclusion you can live with. It is deliberately not an argument in either direction. Nothing here argues for either outcome, which is unusual for writing on this subject.

Tip 1: Stop Asking How You Feel This Week

The most common approach is to consult your own feelings and treat the answer as data, and it is close to useless. Feelings about a long relationship swing enormously with sleep, work pressure, health, whatever was said on Tuesday and how the last conversation ended. Somebody asking themselves the question on a bad Sunday will get a different answer from the same person on a good Thursday, and neither answer is about the relationship. Health, money worries and a single bad night all contaminate the reading.

What is worth consulting instead is the pattern across a longer period. How has the last year been compared with the year before. Is the direction improving, stable or declining. Has anything actually changed, or has the same conversation been recurring since 2021. These questions have answers that do not move with the weather, and they are the only ones that address the timescale the decision actually concerns. Direction rather than position is the thing to establish.

Tip 2: Look for Contempt Specifically

Among everything that has been studied about couples, contempt is the single most reliable predictor of a relationship ending. Not anger, which is loud and passes, and not conflict, which happy couples have in similar quantities to unhappy ones. Contempt is the eye roll, the mocking imitation, the sarcasm, the sense that the other person is fundamentally beneath consideration. Happy couples argue at similar rates, which removes conflict frequency as a signal.

It matters because it is qualitatively different from other difficulties. Anger says that something is wrong. Contempt says that the person is wrong, and it is very difficult to come back from a position where one person has stopped respecting the other. Its presence does not mean something is over and it is the signal worth taking most seriously, and its absence is genuinely encouraging even where everything else looks poor. Respect once lost is very difficult to reassemble from anything.

Tip 3: Check Whether Repair Still Works

The more useful question than whether a couple argues is whether they can find their way back afterwards. In a functioning relationship, one person makes some small approach after a difficulty, a joke or a cup of tea or a remark about something else, and the other accepts it. That exchange happens constantly and almost nobody notices it, and it is the mechanism by which a relationship absorbs conflict. Almost nobody notices these exchanges happening, which is part of why they get lost.

When it stops working, the picture changes fundamentally. Approaches are made and rejected, or they stop being made at all, and each disagreement then has to be resolved by argument alone or left unresolved permanently. A couple who argue frequently and repair reliably is in considerably better condition than a couple who rarely argue and cannot reconnect afterwards, which is the opposite of how it looks from outside. Frequent arguing with reliable repair beats rare arguing with none, which looks wrong from outside.

Tip 4: Distinguish Absence From Presence

Some relationships end through something being present that should not be: contempt, cruelty, betrayal, control. Others end through something being absent that ought to be there: interest, warmth, any desire to know what the other person thinks. The second is considerably harder to identify, because nothing is happening and nothing can be pointed at, and people frequently stay for years in something that is merely empty. Emptiness is harder to name than cruelty and lasts considerably longer.

The absence version is worth testing before concluding anything, because it responds to attention far more often than people expect. A couple who have stopped being curious about each other are frequently not finished but under maintained, and three months of deliberate effort produces a clear answer. If genuine attention produces nothing at all, that is real information. If it produces something, the diagnosis was wrong. Three months of deliberate attention produces a clear answer either way.

Tip 5: Ignore How Long It Has Been

Time invested is the most commonly cited reason for continuing and the least valid one. Twenty years already spent is not an argument for spending the next twenty, any more than a poor investment becomes a good one because you have held it a long time. The years are gone regardless of what happens next, and they do not become more or less worthwhile depending on the decision. A poor investment does not improve because it has been held for a long time.

This is genuinely difficult to internalise, because it feels like dismissing something that mattered, and it is not. A relationship that was good for fifteen years and has now ended was good for fifteen years, and that is not retrospectively cancelled. The question is only what the next decade should contain, and the previous ones cannot answer it however much weight they seem to carry. Gone regardless, and unaffected by whatever happens next.

Tip 6: Separate the Fear From the Assessment

A great many people who have concluded that something is over remain in it for reasons that are about fear rather than about the relationship. Fear of being alone, of the financial consequences, of the effect on children, of what people will say, of the practical enormity of unpicking a shared life. Every one of those is legitimate and none of them is an answer to the question of whether the relationship is finished. Financial obstacles in particular can be genuinely decisive rather than merely frightening.

Conflating the two produces years of paralysis, because the person keeps trying to resolve a question they have already answered. It is considerably more productive to separate them: decide first whether it is over, and then, separately, work out what to do about the practical obstacles. Some of those obstacles are genuinely decisive, particularly financial ones, and they should be faced as constraints rather than allowed to masquerade as uncertainty. Answered privately and then re asked forever is the usual pattern.

Tip 7: Do Not Decide Alone in Your Head

This is the single most common error and the most consequential. One person spends two years privately concluding that the relationship is over, never says anything, and then announces a decision the other person experiences as arriving from nowhere. From inside it feels like a long careful process. From outside it is an ambush, and it removes any possibility that the other person could have participated. Two years of private conclusion removes any chance of participation.

If something is wrong enough to be considering the end, that is a thing to say out loud, in those terms, before any conclusion has been reached. Some relationships change substantially when one person finally states plainly how serious it has become. Others do not, and at least the ending is then something both people were involved in rather than something delivered to one of them. Stating it plainly occasionally changes everything, which is worth knowing.

💡 Key Insight

Feelings about a long relationship swing with the week and cannot answer a question about a decade, so consult the pattern instead. Contempt is the strongest single predictor of an ending, and whether repair attempts still work matters more than how often you argue. Time already invested is the least valid reason to continue. Separate the assessment from the fear, and never reach a conclusion privately over two years.

Fast Action Steps

Three things to do before concluding anything in either direction.

Action 1: Write the Three Year Comparison

On paper, describe what the relationship was like three years ago and what it is like now, in specifics rather than in feelings. How you spent evenings, what you talked about, how disagreements went, how much you looked forward to seeing each other. Then note the direction of travel. This removes the week from the assessment, which is the single biggest source of error, and it frequently produces a surprise in either direction. Somebody convinced things are terrible discovers a slow improvement, or somebody assuming things are fine sees a decade of decline they had not registered. Direction of travel is what the exercise is actually measuring.

Is the direction across three years improving, flat or declining, and would your partner say the same?

Action 2: Run a Contempt and Repair Audit

For two weeks, note two specific things without changing anything. First, any instance of contempt in either direction: an eye roll, a mocking tone, sarcasm, dismissiveness. Second, every small repair attempt after a difficulty, and whether it was accepted. These two measures carry more information than any amount of reflection, because contempt is the strongest predictor of an ending and functioning repair is the strongest sign that a relationship can absorb difficulty. Two weeks of honest observation will tell you considerably more than two years of wondering. Two weeks of observation beats two years of wondering.

When somebody made a small approach after your last argument, what happened to it?

Action 3: Say the Serious Thing Out Loud

If you have been privately considering whether this is over, say so, in those words, before deciding anything. Not as an announcement or a threat, but as a statement of where you actually are: I have been wondering whether we are going to make it, and I did not want to be thinking that on my own. This is frightening and it is the difference between an ending your partner participated in and one that was delivered to them. It also occasionally changes everything, because some people have no idea how serious it has become. Frightening, and the difference between participation and delivery.

How long have you been thinking about this without your partner having any idea?

Relationship Love Secrets

How to Know When Something Is Over