"Better safe than sorry"

This is the defence that feels most like adult sense. Even if the recheck mostly finds nothing, the consequences of the one time it would have found something are so bad that the habit is worth it. Ten seconds against a house fire. Who wouldn't take that deal?

Rachman, who spent a career on this, set out how the deal gets sized. Compulsive checking, he proposed, happens when a person feels a special responsibility for preventing harm, mostly to others, and isn't sure the threat has been dealt with. How much they check is set by three multipliers. How responsible they feel. How likely the harm seems. How serious it would be. Raise any of the three and the checking rises with it. That was written about clinical checking, but the multipliers are ordinary, and I think you'll recognise them in your own mornings. The straighteners aren't just a fire risk. They're a fire that would be your fault, that feels quite possible on the days you're tired, and that would take the house.

Look at what the recheck does to each multiplier. It does nothing to the probability. The door was locked before you pushed it. Pushing it didn't lock it more. It does nothing to the seriousness. A fire is as bad after the check as before. What it acts on, briefly, is the feeling of responsibility, which is why it gives a moment's relief. And then, on Rachman's account, something strange happens. He describes a self-perpetuating loop in which checking produces paradoxical increases in the sense of responsibility and in the perceived likelihood of harm, along with the reduced memory confidence the last chapters described. The more you check, the more it feels like yours to prevent and the more possible the harm seems. The check doesn't shrink the multipliers. It grows them.

Van den Hout and Kindt found a hint of this too. Across five experiments with their stove animation, repeated checking not only undermined trust in memory. There were indications that it also reduced the sense of having acted responsibly. The person who's checked five times feels less like someone who's dealt with it than the person who checked once. That's backwards, and it's what the experiments show.

So "better safe than sorry" is right, and it's an argument for the once. Safe is the door locked and checked properly at the moment of locking it. Sorry is what the rechecks make you feel, because each one leaves you with a fainter picture and a larger sense that the house depends on you. The safe act was done at twenty past eight. Everything after it was the opposite of safety, dressed as care.