A breakup is not a failure: when ending a relationship is the healthiest choice

“I failed.” That is the sentence I hear most often in the therapy room when someone says out loud for the first time that they are thinking about ending their relationship. Not “it isn’t working any more”, not “we’ve grown apart” — but I failed. As though it were an exam they had flunked, and as though the only mark that counted were the number of years the relationship lasted.

The sentence has its own logic. We grew up with a story in which the happy ending is always the same picture: two people who stay together. Everything else, in that story, reads as failure — someone who didn’t try hard enough, couldn’t talk things through, gave up too soon. And so people often stay not because they are happy in the relationship, but because they are afraid of what leaving would mean about them.

In this piece I am not going to argue for separation, and I am not going to argue for staying. What I will do is describe as precisely as I can the part that is genuinely hard: the signs that ending is the healthiest choice; the trap of the years already invested; what the research actually shows about “staying for the children”; how such a decision is made at all, if not in the heat of the moment — and why it hurts even when it is right.

Why we experience the end of a relationship as a personal defeat

Separation is one of the few losses in which a person grieves and feels ashamed at the same time. When someone close to us dies, nobody asks what we did wrong. With a divorce that question hangs in the air, often even when nobody says it out loud.

There is a simple reason for this: we experience a relationship as something within our control. If it fell apart, someone must be to blame, and if you don’t want to blame the other person, you are what’s left. The shame that follows isn’t shame about something you did — it is shame about what you supposedly now are: a person who couldn’t manage. I’ve written about this difference between guilt for an act and shame about your own worth in the piece on when it feels as if something is wrong with us.

Then there is everyone else. Relatives who “always knew it would end this way”. Acquaintances who awkwardly step back. Questions that aren’t really questions: “Did you two really try everything?” Social pressure around divorce has eased over the decades, but it hasn’t disappeared — it has only become more polite.

This is where a different measure is worth setting. A relationship is not an endurance race. Its job is not to last, but to be a space in which both people can live. A relationship that lasts thirty years and in which one person systematically shrinks is no more successful than one that lasts seven, ends decently and leaves both people whole. The measure is not length; the measure is what happened to the people inside it.

Signs that seriously suggest it is time to end things

There is no list that will decide for you. There are, however, patterns that relationship research and clinical practice consistently link with a relationship that can no longer be repaired from within:

• Violence and control. Physical, sexual or systematic psychological violence — humiliation, threats, control over money, movement, contacts. Here there is nothing to weigh up and there is no couples therapy: the first thing is safety, not the relationship.
• Repeated broken agreements. Not a one-off mistake, but a pattern: promises that recur and collapse every time — untreated addiction, repeated affairs, hidden debts. The key question is not whether they are sorry, but whether anything concrete has changed in their behaviour.
• Contempt. Not arguments — contempt. More on this in the next section.
• Growth that has stalled. When the relationship requires that one of the two never becomes what they could be: doesn’t study, doesn’t move forward, doesn’t see their people, doesn’t develop, because it would upset the balance. I write about when growth pulls partners apart and when that is still workable in the piece on when a partner grows in another direction.
• One person carries the relationship. Not for a few months, but for years: one person suggests conversations, reads books, books a therapy appointment, repairs things; the other doesn’t even reach the door. One-sided effort can hold a relationship up for a while, but it cannot change it.
• The other person’s worth has permanently gone out. When your partner’s good qualities are no longer visible even in calm moments — not in anger, but in indifference.

And one measure people often overlook because it is too simple: what you are like in this relationship. Not what he is like. If you have become a person you don’t like — suspicious, cynical, sharper with the children, distant from your friends — that is information about the relationship, not only about you.

Contempt: the most reliable sign that a relationship is in serious danger

John Gottman and his colleagues spent decades recording couples’ conversations and then following up on who stayed together. Out of that research came the famous “four horsemen” — criticism, defensiveness, stonewalling and contempt — which I describe in more detail in the piece on the four horsemen of relationships. For a decision about ending things, it is above all the fourth that matters.

Contempt is not anger. Anger means you still care and you still want something. Contempt means you have placed the other person beneath you: eye-rolling, a mocking tone, insults dressed up as jokes, correcting them in front of others. In Gottman’s research it was precisely contempt that was the strongest single predictor of a relationship breaking down — and, less well known, also a predictor of more frequent physical illness in the partner it was aimed at.

Contempt is not a final verdict. With couples who come early enough and who both recognise it as a problem, it can be worked with. But it does mean that “more talking” alone will not be enough, and it means time is short: where contempt has become the main language of a relationship, respect rarely returns on its own. If you recognise your own relationship here, it is worth also reading about when couples therapy makes sense — and when it is for weighing things up rather than for saving them.

The trap of invested years: why “we’ve been together so long” is not a reason

“I can’t leave. Fifteen years are in there.”

This is probably the most common argument for staying — and one of the worst. Economics knows it as the sunk cost effect. Arkes and Blumer described it in 1985 in a classic series of experiments: people persist with decisions they have already invested heavily in, even when it is clear the investment will not pay off. A ticket already bought makes us go to a show we don’t want to see. Money already spent convinces us to carry on with a failing project.

Relationships are no exception. Researchers have demonstrated the sunk cost effect in partnership decisions too: the more time, money and effort people had put into a relationship, the longer they were willing to stay in it — regardless of how satisfied they were.

The logical trap is simple. Those fifteen years are already spent. Nothing you decide today will give them back or erase them. The only question your decision actually affects is what the next fifteen will look like. If you turn the question around like this — “would I choose this relationship, exactly as it is today, all over again, if nothing tied me to it?” — you usually get an answer rather faster than you would like.

Which of course doesn’t mean the years are worthless. It only means they are not evidence about the future. A shared history is a reason for respect and for serious reflection — it is not a reason for a sentence of continuation.

“We’re staying for the children” — what we actually know

This is the most common and the most honourable reason for staying. It deserves a precise answer, not a reassuring one.

Research on children after divorce is extensive and relatively consistent: divorce in itself represents a risk for a child, but most children adjust within a few years and show no serious long-term difficulties. What matters most, though, is what emerged from longitudinal studies that followed families before the divorce. Paul Amato and Alan Booth and their colleagues found that the effect of divorce on children is not the same in every family: in marriages with high, open and enduring conflict, children generally do better after a divorce than if the parents stay together, while in low-conflict marriages — the kind where everything looked fine to the child — divorce affects the child more.

From this follows an uncomfortable but useful conclusion: the question is not “divorce or not”, but “what is the child living through right now”. If a child is growing up amid arguments, contempt, silence or a cold war, a shared household is not giving them protection — it is giving them a daily model of what they will imagine love to be. If, on the other hand, the conflict is quiet and the child doesn’t see it, the separation will be a major shock and they will need more support.

The second key finding: what matters most for a child’s adjustment is how much conflict remains after the separation and whether they keep a good relationship with both parents. Which means divorce is not an answer in itself — the answer is how it is carried out. I write about this in more detail in the pieces children and divorce and co-parenting after separation.

And one more thing I say often: children miss almost nothing. A child growing up in a house where their parents neither touch nor look at each other knows. They just have to put the explanation together themselves — and as a rule they put it together so that somewhere in the middle of it, they are to blame.

How such a decision is actually made

The worst time to decide about ending things is in the middle of an argument. The second worst is the third day of an illness, after a sleepless night, after one nasty message. Decisions are not made in the heat of the moment — in the heat of the moment we only do things that will be hard to repair tomorrow.

What actually works for people is slower and more boring:

• Separate exhaustion from the decision. Many people who arrive with the question “should I go” are in fact worn out — by work, small children, a parent’s illness. A burnt-out person cannot assess a relationship, because they cannot assess anything. The first step is often sleep and relief from load, not a decision (I write about the signs of exhaustion in the piece on burnout).
• Test the decision in three different moods. What still holds when you are calm, when you are angry and when you are sad is probably the truth; what appears in only one of them is an emotion.
• Ask a concrete question about the future. Not “do I love him”, but: “If in five years everything is exactly as it is today — no worse, no better — can I do that?”
• Check whether you really did try. Not in the sense of “I made an effort”, but concretely: did the two of you talk about it openly? Was outside help ever on the table? Did your partner even know how close to the edge you were?
• Don’t decide alone in your head. A decision that goes on for months purely inside your own thoughts turns into rumination — more on that in the piece on decision paralysis.

For couples where one wants to leave and the other wants to save the relationship, there is a particular form of work developed by Bill Doherty called discernment counseling. It is not relationship therapy and it is not preparation for divorce; it is a limited set of sessions whose only aim is clarity: whether the couple moves into serious work on the relationship, into separation, or stays as they are for now. Surprisingly often it turns out that the greatest relief is simply having both possibilities said out loud in the same room for the first time.

Why it hurts even when the decision is right

People often expect that the right decision will bring peace. For a few days it sometimes does — and then comes a wave, the same as with any loss, and the person begins to doubt: “If it was right, why does it hurt this much?”

Because a decision and grief do not cancel each other out. You decided to end a particular relationship, not to end attachment, not to end the memories, not to end the plans the two of you had. All of that now dies at once, however well founded the decision was. Those who did the leaving especially often have no space for that grief — everyone expects them to be pleased, since it was what they wanted. I write about this particular form of grieving in the piece divorce as grieving.

So the pain is not evidence that you were wrong. It is the price paid for every large change — and the only way to avoid it is to change nothing.

Separation is not a defeat — only the manner of it can be

If there is anything in a separation that deserves the word defeat, then, it is not the fact that the relationship ended. The defeat is if it ends in such a way that two people who once loved each other spend the rest of their lives communicating through lawyers and children. The defeat is if they learn nothing from it and repeat the same pattern with the next person. The defeat is if someone walks away from it convinced that they are not worth anything.

But there is also an ending we can call decent: where each person admits their own share, where the children don’t become a battlefield, where enough respect remains that years later the two can meet at a school leavers’ ball and not look away. Such an ending is not a given — it is work. For many people that work is easier when they are not doing it alone: support through separation is meant for exactly this, whether you come while you are still deciding, during the separation itself, or afterwards, once everything is formally settled but inside it still isn’t. If you want to know how such sessions go, there is more in the piece on therapy through separation; and if it feels as though the right moment for a conversation is already here, you can write to me.

A relationship that has ended does not erase the years you spent in it. It is a chapter that has closed — and chapters do close in a life. The question is not whether you managed to stay. The question is what kind of person you are when you leave.