How friendships end

Interview surveys on the end of friendships distinguish two forms: rupture, following an identified event, and fading, where contacts space out until they cease without any decision being taken.

The second form is by far the most frequent in the accounts collected. It is also the least studied, precisely because it produces no event to date. Respondents often place the end long after the fact, when they notice they have not seen each other for a long time.

In this, friendship differs from other ties. Unlike marriage or an employment contract, it rests on no institution marking its beginning or end, and no social rite accompanies its dissolution, as sociologists of friendship have long pointed out.

Longitudinal work on network turnover shows that leaving a relationship most often follows a change of context rather than a disagreement. The surveys of Claire Bidart, who followed young adults over several years, thus describe ties that come undone with moves, studies and the start of couple relationships.

That observation explains why respondents rarely attribute a cause to the end of a friendship: there is in fact no relational cause to find.

When both members of a faded relationship are interviewed separately, their accounts differ sharply, including on whether the relationship is over. One may regard it as closed while the other considers it merely dormant.

That asymmetry echoes the result on non-reciprocal nominations: a tie may exist in one person's declaration and not in the other's, without either knowing it.

Where a break is explicit, the motives most often reported in surveys are betrayal of a confidence, a serious disagreement, and the sense of a lasting imbalance in exchanges. That last motive points to an expectation of reciprocity that friends rarely voice but that surveys regularly find.

Those motives are collected after the fact, which exposes them to reconstruction: a coherent account is produced to explain a situation, which is not the same as a cause. A disagreement cited as the motive may have been no more than the occasion for a distance already in place.

The persistence of ties on platforms maintains visibility without contact, an unprecedented configuration: a faded relationship stays displayed, and resumption remains technically possible at any time. In the past, fading also meant losing news of the other person; it can now coexist with detailed knowledge of their life.

Available work on this point is recent and inconclusive. It mainly describes new practices of graduated distancing, without measuring whether network turnover is thereby altered.