The friendships that do not survive matrescence

You used to speak every week, and now it is twice a year, and nothing happened in between.

That is the shape of it. Not a falling out, not a betrayal either of you could name and be angry about. A gradual lengthening of the gaps until the gaps were the relationship, and then a Christmas message, warm and slightly formal, from a woman who used to know what you were thinking before you said it.

I have three of those. I did not see any of them happening.

There was no falling out, which is the part that disorders you

The absence of an event is what makes this impossible to hold. Grief needs a cause to be legible. Something has to have happened, or there is nothing for anyone to be sorry about, including you.

So you do the only thing available to you and file it under your own failure of effort. You did not reply to the message in March. You canceled twice in a row that summer. You let it go, and therefore it is yours.

I have written before about grief that has no social standing, the kind where the loss is real and the permission to mourn it never arrives. This is a quieter version of the same problem. Nobody asks how you are about a friendship that merely got further apart.

My own framing here may be wrong, and I would rather say so than build on it. A 2022 study of call records covering 644,170 mobile subscribers and 178,592 friendship links looked at whether friendships tend to tail off gradually or stop abruptly, and found no significant difference between the two patterns (Roy, Bhattacharya, Dunbar and Kaski, 2022). It measured call frequency rather than felt closeness, in one country. But it means the quiet fade I keep describing is not demonstrably the usual way a friendship ends. It is the way mine ended.

What actually ends is the shared present tense

A friendship runs on a common now. Not shared history, which is what people assume. What you are each turning over this week, what you can both afford to be careless about, which of you is currently the one with the problem.

Matrescence rewrote mine without asking. The things I could be careless about had gone. The thing I was turning over was so large and so unsayable that everything else sounded like small talk to me, which is an unbearable way to sit across a table from someone you love.

Diaries are the cover story. Everyone agrees it is diaries, because diaries are a blameless explanation and both of you can use it. What is true is that you no longer arrive at a conversation from the same place.

There is arithmetic underneath this, and I find it steadying. In a 2015 study of 540 people, the average innermost support network ran from 6.2 people for single respondents, to 5.4 for those with a partner, to around 4.2 for those with a partner and a child (Burton-Chellew and Dunbar, 2015). That first gap of 0.8 is a net figure, and the churn underneath it is larger: the partnered respondents named about 1.6 fewer of the people the single ones named, against about 0.8 that only they did. The ones who also had a child sat more than a person lower again.

The study was not longitudinal and the authors say so plainly: it did not record changes to anyone's network, it compared different people at one moment. So the honest sentence is that people with a partner and a child name about two fewer, never that you will lose two. Even in that weaker form it did something for me. The number is not a measure of how hard I tried.

Only one of you moved

She has not changed. That is the part I could not get anyone to understand, including myself.

She did nothing wrong. She kept being the person you liked, at the pace you liked her at, with the same subjects and the same appetite for an evening decided at six o'clock. She is waiting, kindly and without resentment, for you to come back.

You are not coming back, because there is no back. This is not a scheduling problem with a scheduling solution. It is the same rearrangement I keep circling in the whole Losing Yourself chapter. Postpartum identity loss is nearly always described as something happening inside a woman, privately. This is the one place where another person can watch it happen.

Here the evidence complicates me, which I would rather report than quietly leave out. A Swiss study that followed the same households across twelve waves between 1999 and 2010 found that becoming a parent decreases contact with friends and increases contact with neighbors, and that the earlier you have children the greater the loss (Rözer, Poortman and Mollenhorst, 2017). The decline was larger for men than for women. Women regained contact with friends as the child got older. Men did not.

I had assumed this was something that happens to mothers in particular. On the best longitudinal evidence I could find, it is not, or not only. I do not have a tidy account of what that means.

You are also the one doing the leaving

I let some of them go, and I want that on the record before this becomes a thing that happened to me.

There were calls I could have returned and did not. There was a woman I had known for eleven years who wanted to talk about a promotion, at length, in the way we always had, and I found the subject unbearable and I never told her so. I got slower at replying until the problem solved itself.

Some of that was capacity. Some of it was not. Some of it was that being around the people who knew the earlier version of me made her absence specific, and it was easier to sit with people who had never met her.

That is not a flattering sentence and I have not found a way to make it one.

The ones that survived had nothing in common

I looked for the pattern for about two years. There is not one, or I never found it.

It was not the mothers. Two of the friendships I lost were with women who had babies the same year I did, and we were no use to each other at all. We had the identical problem and nothing left over to hold anyone else's version of it.

It was not the oldest ones. It was not the ones who tried hardest either, and that is the one that still bothers me, because it means effort was not the variable, and effort is the only lever anybody has.

Personal observation rather than anything established: the ones that lasted were with people who could tolerate me being uninteresting for a long time without requiring an explanation. That is not a quality you can screen for in advance, and it is not a quality I would have valued at twenty-five.

Where this stops being matrescence

A few particular friendships thinning is not the same thing as having stopped answering everybody, and that difference matters more than anything else on this page.

If the withdrawal is general, if you are not replying to your sister or your oldest friend or anyone at all, if the thought of being seen by someone who loves you produces dread rather than tiredness, that is a different question and an essay cannot settle it. It belongs with a doctor, a midwife, a child health nurse or a psychologist.

I am not a clinician. I wrote more carefully about where the ordinary version ends and something treatable begins, and the position in that piece holds here. I cannot tell you which side of the line you are on, and neither can a website.

None of that is a checklist to grade yourself against, and please do not use it as one. It is the distinction I use on myself, and it is the only one I am qualified to draw.

What I can say is that selective is different from total. Losing three friendships while others hold is a shape. Losing all of them at once is not a shape, and it is worth saying out loud to one person qualified to hear it.

What I did not do, and would not tell you to do

I never had the closure conversation. I have thought about it for years and I still think it would have been for me rather than for her.

Consider what it asks. You call a woman who has done nothing wrong, tell her something has ended that she may not have registered as ending, and hand her the work of managing your feeling about it. She then has to either reassure you, which costs her, or agree with you, which costs you both. It resolves your discomfort by transferring it.

What replaced those friendships is a separate question, and I wrote about what actually helped afterwards rather than answering it here.

The one thing I still do is answer. Not quickly, and not with anything shaped, but I answer. Twice a year, warmly, with no apology for the gap and no attempt to account for the eleven months in between. A friendship conducted at twice a year is not the thing it was. It is also not nothing, and I stopped grading it against the version it used to be.

What I still do not know

Whether any of it is recoverable.

I cannot tell whether these friendships are over or dormant, and I notice that the word dormant is doing comforting work I have not earned. I do not know whether this is matrescence at all, or simply what your thirties do to everybody, with a baby standing in as the explanation for something that was coming anyway.

I also do not know whether I would want them back on the old terms. The woman those friendships were built around is not available. Some of what I miss may be her rather than them.

The books piece I wrote named this as one of the holes in the literature: the quiet attrition, the friendship that ends without a rupture, one of the most common losses of this period and one of the least documented. This is my attempt at it. An attempt by one woman with notes, not a finding.

What to carry

It was not a failure of effort. That is the whole of what I have, and it took me four years to believe it about myself.

A friendship runs on a shared present tense, and yours was rewritten without your consent. You can be attentive, generous and entirely blameless, and still watch a weekly conversation become a twice-yearly one, because attention was never the mechanism.

So you are allowed to miss her without concluding anything about your character. You are also allowed to have been the one who stopped replying. Both are usually true at the same time, and neither of them is the story of a woman who let people down.