"If I stop, I'll do it again"

This belief is the quietest of the five and, I think, the one with the firmest grip on people. The replays feel like a guard on duty. As long as the photocopier still stings, you'll never again say something crass to a grieving person. If it stopped stinging, you might get careless. So you keep it sore deliberately, or at least you don't dare let it heal.

Here are two pieces of evidence from your own life.

First, you haven't done it again. In eleven years you've spoken to other bereaved people, and I'd bet you've been gentle and brief and said you were sorry. What was steering you in those conversations? It wasn't a replay. You weren't running the corridor scene in your head while you talked to them. It was the one-sentence lesson from chapter 3, which has long since become how you behave. The replays happen at washing lines and in the shower, when no bereaved person is anywhere near. They're not present at the moments when you'd need a guard.

Second, think of the mistakes you made once, learned from, and never replay. Years ago you backed into a gatepost, and now you check the mirror. You once sent an email in anger and had to apologise, and now you wait till morning. Those lessons hold perfectly well with no replays to maintain them. You don't have to keep flinching about the gatepost in order to look in the mirror. The guard theory says lessons only hold if they're kept painful, and your own record shows plenty of lessons holding with no pain at all.

A mistake is kept from recurring by two things, and both are dull. One is a lesson stated clearly enough to act on. The other is a plan for the situation where it applies. "When I'm about to speak to someone who's grieving, then I'll say I'm sorry and ask how they're doing, and nothing else." You don't remember to buy milk by chanting "milk" for a week. You write it on the list, and then you're free to forget it until you're in the shop.

There's research suggesting that the writing down is what lets the mind stand down. E. J. Masicampo and Roy Baumeister ran a series of studies on unfinished goals. They gave people a goal and left it unfulfilled. Those people then had intrusive thoughts about it during an unrelated reading task, and did worse on other work, because part of their mind was still on the unfinished thing. That much was already known. What Masicampo and Baumeister added was the next step. When people were allowed to make a specific plan for the unfinished goal, saying when, where and how they'd deal with it, the intrusions stopped. They hadn't completed the goal. They had only planned it, and that was enough for the mind to stop raising the matter.

Their studies were about tasks, not about old mistakes, so applying them here is my interpretation. But it fits the features from chapter 1. Replays arrive uninvited, in idle moments, and they never conclude. That's how an open item behaves. Some part of your mind has the photocopier filed under unfinished business and brings it up whenever there's spare attention, in the hope that this time it'll get settled. The replay never settles it, because a replay contains no verdict and no plan. So it stays open and comes round again.

If that's right, then the way to reduce replays isn't to fight them. You give the open item what it's been waiting for. That means a clear account of what happened, a lesson you can state, a plan for next time, and a decision about whether anything remains to be done. Chapter 12 sets out how to do it. The guard can stand down because there'll be something better in his place.

You may worry that you'll become careless. A person with a written lesson and a plan is more careful than a person with a recurring flinch, because the lesson and plan are there when they're needed. The flinch only ever turned up at the washing line.