"Can I just check this with you?"

The email from chapter 1 didn't go straight to the client. You wrote it, polished it, and then forwarded it to a colleague with "does this sound OK?" She wrote back, "Looks fine to me." You felt your shoulders drop, and you sent it. The next day there was another email, and you forwarded that one too.

There are versions of this at home. You're choosing between two coats and you send photos to your sister. You've more or less decided about a course, and you put the question to five friends and go with the majority. Your partner asks where you'd like to eat and you say, "I don't mind, you choose."

It all looks like plain good sense. Two heads are better than one, other people often know more, and checking costs nothing.

A good deal of that is true. A second pair of eyes catches things. In plenty of jobs, review is a required step, and rightly. Asking people who know more than you is one of the best habits a person can have. None of what follows is an argument against it.

The distinction I'd draw is between asking for information and asking for reassurance. When you ask for information, there's something you don't know and the other person might. Their answer could change what you do. Once you have it, you don't need to ask again.

When you ask for reassurance, you already have your answer. The email is fine and you know it's fine. What you want is for the uneasy feeling to be settled, and only one reply will do that. If she said "hmm, I'm not sure about the second line", you wouldn't be grateful for the information. You'd be thrown. And when she says "looks fine", the comfort lasts until the next email.

There's a simple test you can use before you ask. What do you think the answer is? If you know what they'll say, and you'd send the thing as it stands if nobody were around to ask, then it's reassurance you're after.

The reason it doesn't work is the one from chapter 2. The relief is quick, so the habit gets stronger. And the credit goes to the check. The email was all right "because she looked it over". You never find out that it would have been all right anyway. Your own judgement, which was the thing in doubt, never gets tested alone, so it never gains any standing with you.

Clinicians have described this most carefully in people with obsessive and anxious problems, where seeking reassurance is a recognised part of what keeps the problem going. In one study, people with these problems reported a brief drop in anxiety after being reassured, followed later by the return of both the discomfort and the urge to ask again. People without such problems didn't report that return. The researchers are careful to say that this rests on what patients remember and hasn't been shown by experiment. They also warn against simply refusing reassurance to someone in distress. Ordinary self-doubt is a much milder thing, and carrying the lesson across is my step. But the shape looks the same to me. Reassurance seems to wear off quickly, and the more often you ask, the more you need to.

The costs fall in a few places. Your judgement stays untested, as I've said. Your colleagues form a view. Someone whose work always arrives with "can you just check?" comes to be seen as someone whose work needs checking, and that affects what they're trusted with. There's also a small, steady tax on the people you ask, who are usually too kind to mention it.

The cost at home is quieter and I think larger. Each "you choose" is tiny. Over years, a good part of your life ends up chosen by other people. Preferences that are never voiced get fainter, and after a while "I don't mind" is no longer a politeness. You don't know what you'd like any more.

You may be thinking, "But she does catch things sometimes." Then keep the part that works, and make it a proper review. Ask a specific question. "Is the second paragraph clear about the price?" will get you a useful answer. "Does this sound OK?" will get you "looks fine". It also helps to decide in advance, by the type of work and not by the feeling, what gets a second look. Anything going to the board or anything with a contract attached might qualify. A routine email to a client you've dealt with for two years doesn't.

Here's an experiment for the next fortnight. Each time you're about to ask someone to check something, first write down what you expect them to say. Then ask, if you still want to, and compare. If your prediction matches their answer nine times out of ten, you've learned something solid, which is that your own judgement was already telling you what they would. After that, start sending the low-stakes things unchecked, and see what happens to them.

For decisions, the first step is smaller still. When you're asked what you'd like, give a preference before you hand the choice over, even a slight one. "I'd rather the Italian, but I'm easy." You can still be easy. The point is to keep in practice at knowing what you think.