"I'm learning from it"

This is the most respectable reason for a replay, so I'll start with what's right about it.

Thinking about how things might have gone differently is one of the more useful things a mind does. Psychologists call these thoughts counterfactuals, and you'd know them as "if only". Kai Epstude and Neal Roese have reviewed a large body of experiments on them. Their conclusion is that such thoughts exist to improve future behaviour. "If only I'd left ten minutes earlier" turns into "next time I'll leave ten minutes earlier". The thought about the past produces an intention, and the intention changes what you do. On their account that's the useful route. The thought does its work by handing you something specific to do next time.

So ask what the photocopier has to teach. I'd put it like this. When someone has been bereaved, say you're sorry and then stop talking. Don't reach for a silver lining.

That's one sentence, and I'd guess you had it within a day of the event, certainly by the second or third replay. It's a good lesson and you've probably followed it ever since. The question for this chapter is what replays four to four hundred were for. They can't have been for learning that sentence, since you already had it. If there were a second lesson in the scene, eleven years is long enough to have found it.

There's a reason the later replays don't yield anything, and it's to do with the kind of thinking they involve. Edward Watkins and Michelle Moulds ran a study with forty depressed patients and forty people who'd never been depressed. They tested everyone's ability to solve everyday social problems. Then they had people dwell on themselves in one of two ways. One way was abstract, turning over why things are as they are and what it all means. The other was concrete, attending to the particulars of how things happen, step by step. Among the depressed patients, the concrete kind of thinking improved their problem solving compared with the abstract kind. Both groups had thought about the same subject. They differed only in the way they'd thought about it.

Now listen to what a replay sounds like once it's past the opening scene. "Why did I say that? What is wrong with me? Why do I always do this? What must she have thought of me?" Those are abstract questions about causes and meanings and your character, and they don't have answers you can act on. The useful version would ask different things. What happened, in order? What did I do? What would I do in the same spot now? That version is over quickly, because the questions have answers. Replays run long because nothing in them can be answered.

In a related set of studies, Lyubomirsky and Nolen-Hoeksema had people in a low mood either dwell on themselves or turn their attention elsewhere for a few minutes. Then everyone was given interpersonal problems to solve. The ones who'd been dwelling produced less effective solutions. The ones who'd been distracted did as well as people who weren't low at all. A spell of dwelling made people worse at the very thing the dwelling claims to be for.

You might object that some lessons take time. "I didn't understand what went wrong in my marriage until years afterwards." I believe you. But look at how that understanding arrived. My guess is that it came from something new. It may have been a conversation, a book, a second relationship that showed you the pattern, or just being older. It didn't come from the four-hundredth pass through unchanged material. New understanding needs new information, and a replay has none. It can only show you what's already in it.

So there's a simple check you can use whenever a replay starts. Ask what this one has told you that the last one didn't. If the answer is nothing, then it isn't study, whatever it feels like. You've been learning from this mistake for eleven years, and the lesson took a day.