Surprisingly Blind
You might expect us to understand our romantic couple breakups very well. After all, breakups are usually chosen by one of the couple using a limited number of well known considerations, and our romantic relations are some of our most important relations, wherein we spend months and years focused on a single person, learning all about them and how they feel about us. How could we not know why we break up?
But it seems we don’t know remotely as much as this analysis suggests. I just asked six polls asking which of eight reasons “is more often why A breaks up with B in the romantic breakups you’ve experienced”. The priorities (relative to 100 max) from 640 responses:
Aside from #8, which few think plausible, the relative priorities of other 7 options seem remarkably compressed, compared to the many other polls I’ve done to generate relative priorities. And these results seem to express the biases we should expect given that most of my respondents are men, and that women initiate most breakups, namely blaming A more and B less. In fact, it isn’t clear to me that we see much more than these biases at work in these results.
I found one data study that prefers #4 as an explanation:
The top-rated reason participants selected for leaving the relationship in Studies 1 and 2 was something missing from the relationship (i.e., lacking satisfaction or need fulfillment), in Study 3, the only significant positive predictor of breakup was involvement with alternative dating partners. (more)
In our book The Elephant in the Brain, we argue that the job of our conscious minds is less to know why we really do things, and more to plausibly explain our actions in ways that protect us from accusations of norm violations. As a breakup offers many opportunities for norm violations, that theory predicts that we less know why we really break up, and more believe in and argue for stories that make us less at fault. That seems confirmed here.
Added 4May: This is ChatGPT4.5’s ranking of these as actual reasons: 4,7,6,1,3,2,5,8, and as the reasons people will say: 6,7,3,5,2,1,4,8. Aside from agreeing that no one finds #8 plausible, neither predicts my poll results well.
I don't reject the idea that there's some "elephant in the brain" stuff going on here, but it seems more likely simply that most break-ups are complex and involve multiple elements from your list. It's rarely clear which factor was dominant.
Take a friend’s break-up (told to me by A, so some bias there): A & B seemed happy at first. They had a child; B was hands-on early on but withdrew after a couple of years. A took on more parenting and worked more. B started partying, spent less time at home, and began overspending. Their personality clashes became more salient - A saw B as irresponsible, B saw A as less fun than before, and more controlling. A asked for serious changes; B agreed but felt constrained. Tension escalated until B slept with someone else quite openly. Then A ended it.
Several of your items apply:
1 (B novelty-seeking), 3 (B not improving), 5 (growing separation), 6 (hidden failing—B infidelity/A controlling personality), and 7 (B worsening—financial strain). Possibly 2 (A learning their own limits) and 4 (better options - B staying with the new partner).
Your theory is that we don't know why we break up, and choose to believe in and argue for stories that make us less at fault.
My theory is that break-ups typically contain at least 2 of your options, with both parties having different reasons, which explains the lack of coherence in responses equally well.
Why not just prompt an LLM? Hard to believe the training data is not full of romantic rants, betrayals, etc. and I’d take passionate posts in the moment vs. a questionnaire