Thank them now

The 20-minute version

20 min read · the full book is 55 min

You get home at half past six. In the car you were thinking about the evening, mostly about what still needs doing. You open the front door, walk into the kitchen, and see the pan on the hob, the one from breakfast with the porridge dried onto it. Your partner has been home for an hour.

"The pan's still there."

Or you say nothing and pick it up with a sound that says it for you. Your partner says, "Yes, I was going to do it." A pause. Then, to nobody, "You're welcome for the table." You look at the table. It's been cleared, wiped and laid for two. The post has been opened and sorted. The school bag is by the door with tomorrow's kit in it. You hadn't seen any of that. You'd seen the pan.

It doesn't only happen at home. At work you read the quarterly report your colleague spent three days on, and your first sentence is that the total on page four doesn't match page nine. You're right. The other forty pages are good, and you haven't mentioned them, because there was nothing to mention. Your daughter shows you her maths homework, twenty questions, and your eye goes to the two with a cross. You point at them. She puts the book away.

None of these are outbursts. You're not shouting and you're not unkind. You'd say you're paying attention, and you are. It's just that the attention only ever produces a remark in one direction.

My argument goes like this. The habit makes sense from the inside, because the faults are real, you're usually right about them, and someone with your standards is useful to have around. But it's a trick of attention and not a report on the world. What's fine sends no signal and what's wrong does, so a mind that goes where the signals are ends up with a list of faults and calls it the truth of the place. The people around you hear that list and nothing else, and draw their own conclusions about what you think of them. The practical answer is short. When you notice something's fine, say so, to the person, at the time, and specifically. When you notice something's wrong, decide whether it needs saying before you say it.

I'm not asking you to stop noticing faults, lower your standards, think positively, or keep a notebook of things you're grateful for. I'm asking you to notice that you've been keeping only half the record, and to say the other half out loud.

Two cautions. If the running list of faults is mostly about yourself, about what you did wrong, that's a related habit with its own book, You wouldn't say it to a friend. This one is about the scan turned outwards. And if the trouble is less that you miss what's fine and more that nothing has felt fine for a couple of weeks, if you've lost pleasure in things you used to enjoy and feel low most of the day, treat that as a change in your health and see a GP. Low mood flattens everything, and no amount of looking for the fine will find it while the mood is doing that.

One thing to try first. Pick a day, tomorrow if you can, and count. Every time you remark on something that's wrong, out loud or by a look, add one to the first column. Every time you remark on something that's fine, add one to the second. Don't change anything. Most people are surprised by the first column and can't remember filling in the second.

The half that rings

The scan earns its keep. That's why you still have it.

The wrong total would have gone to the board. The two crosses are two things your daughter doesn't understand yet. The pan will be harder to clean tomorrow. The scan doesn't invent problems. It finds them early, which is worth something, and it's what standards feel like from the inside. If you stopped noticing, you'd worry you'd stopped caring.

All of that is true enough. The trouble is the record the scan keeps.

You'd say you're just observant. Timothy Wilson and Daniel Gilbert, two psychologists who study how feelings fade, proposed a model of what happens to our response to things over time. We attend to what matters to us and what we can't yet explain. Once something is explained and understood, we attend to it less and feel less about it. Nothing about it has changed. It's been filed.

Apply that to the kitchen. The cleared table is explained. It's clear most evenings and there's nothing about it to account for, so it was filed long ago, and filed things don't get looked at. The pan is unexplained. It shouldn't be there. So your attention goes to the pan before you've decided anything, and faults are almost never filed, because a fault is by definition the thing that isn't as it should be.

A smoke alarm is silent while the house is fine and loud when it isn't, and nobody takes it for a report on the house. It's a report on smoke.

So you're observing half the room, the half that's ringing. The table, the post and the school bag are there too, doing nothing to attract you, and you'd have to turn and look at them on purpose.

Then there's weight. "I only mention what matters." But the belief that a fault outweighs the fine is a fact about how a fault feels to you. Paul Rozin and Edward Royzman, in a 2001 review, described negativity bias, the general tendency in people and other animals to give more weight to bad things than good things of the same size. Bad things are stronger, they loom faster, they get more thought, and when good and bad are mixed together, the result comes out more negative than adding them up would predict. The same year Roy Baumeister and three colleagues published a long review called "Bad is stronger than good", and across a wide range of everyday life they found almost no exceptions. Both are reviews rather than single experiments, and a later critic argued the picture is less uniform than they suggest, but the broad pattern is well established.

So the evening is unbalanced twice. By count, three fines to one fault, and the three weren't counted because they made no sound. And by weight, because the one that was counted felt heavier than it is. You walk in and the evening feels like a pan. Someone whose weights were level would see a cleared table with a pan in it.

"I'm being realistic." This is the belief that feels most like a virtue, so look at what a realistic report of the kitchen contains. Table cleared and laid. Post sorted. Bag ready. Pan on the hob. Four items, one of them a fault. The scan's report is one item long, three-quarters of the evening left out. Someone who only sees faults is exactly as unrealistic as someone who only sees what's fine. The difference is that one of them feels rigorous and the other feels naive, and neither feeling is evidence. Your tally from tomorrow will be a count of what rang. The day itself held several hundred things that went as they should. Meals happened, people turned up, the car started.

Now the people around you, who don't live in your record. They live in what you say.

You may hold a rule you've never put into words, that silence means approval. If you didn't mention it, it was fine, and surely everyone knows that. But test it against what crosses the room. Your partner receives one remark, about the pan. There's no second message saying the table was noticed. You may feel that message, but a felt message doesn't travel. The forty good pages produced silence, and silence has no content. People fill it with whatever they're already inclined to believe. Over months most settle on one of two readings. Either you didn't see what they did, or you saw it and didn't think it worth a word. Neither is what you meant. Both are reasonable conclusions from the only evidence you've given them.

From where they stand, you're the sum of your remarks. "You're welcome for the table" was that reading, spoken aloud. It was a report on what it's like to clear a table for someone whose only remark was the pan, and it surprised you, because in your version of the evening the table had been seen.

What crosses the room

They don't know. That's the finding.

You'd say they know you appreciate them, that after years together it would be embarrassing to spell it out. Amit Kumar and Nicholas Epley ran three experiments on exactly this. People wrote a letter of thanks to someone who'd done something for them and predicted how the recipient would feel. Then the letters were sent, and the recipients said how they actually felt. The writers were wrong in a consistent direction. They underestimated how surprised the recipients would be by what they were being thanked for, and underestimated how good they'd feel. They overestimated how awkward it would be. And the people who expected it to be awkward and fall flat were the ones least willing to do it.

The surprise is the part to hold onto. The recipients hadn't known that what they'd done had registered. That's "they know" tested and failed. The study used letters, and taking it into a kitchen is my extension. I think it holds, because the mistake is about how the other person feels, whatever the medium.

"Praise makes people slack." The worry is that the standard is held up by the commentary, so if you start saying what's fine, next quarter's report will have thirty-eight good pages. But the standard lives in the fault list, and the fault list stays. The wrong total still gets raised. What changes is that you also tell your colleague which forty pages are right, which is more information about the standard. Kumar and Epley followed up with experiments on small kindnesses between strangers. Givers again underestimated how good recipients would feel, and in the last experiment, recipients were more generous to a third person afterwards than givers expected. It's a lab result about strangers and doesn't show that thanking a colleague makes them work harder. My own reading is simply that people repeat what's noticed. If you only notice faults, your commentary can only teach which faults to avoid. Thanks can teach what to keep doing.

"It would sound fake."

You imagine saying "thanks for laying the table" and it sounds like something from a training course. That feeling is the forecast Kumar and Epley measured, the expected awkwardness, and it mostly stayed on the writer's side. What sounds fake is vagueness. "You're wonderful" is a verdict with nothing under it. "Thanks for sorting the post, I'd been dreading that pile" is about a pile that was there and now isn't, and you can only say it if you saw it. Lisa Williams and Monica Bartlett had people receive a note from a stranger, with or without a line of thanks for a favour they'd done. Those who were thanked were more likely to give their contact details to keep in touch, because the thanks made the writer seem warmer. It's one small study, so hold it lightly. It fits, though. Thanks reads as warmth to the person hearing it.

"Thanks is for big things." You'd thank someone for a kidney, or for a week off to help you move. A laid table is Tuesday.

That's exactly the problem. Kennon Sheldon and Sonja Lyubomirsky followed four hundred and eighty-one students for three months after a good change in their lives. The happiness from it faded in the usual two ways. The good feelings declined, and expectations rose until the gain became the new floor. What slowed the fading was continued appreciation of the change, and variety in how it was experienced. Those were students and one change, and my extension is this. Everything fine in your life was once a change. The first time your partner laid the table unasked, you noticed. Then it happened again and again and became the floor, and now only the pan shows above it. What nobody does on their own is thank people for Tuesday, and Tuesday is where they live.

The honest reader objects here. Some of it isn't fine. The total is wrong. You won't pretend otherwise.

You don't have to. You don't thank the pan. You keep two lists instead of one. The fault list stays a fault list, and the second sits alongside it. "Thanks for doing the table" when you walk in and "the pan's still there" ten minutes later are both true, and together they're a fair report of the evening. The second list makes the first more usable, too. A fault from someone who also mentions what's fine is heard as information. The same fault from someone who only mentions faults is heard as the weather, and people stop listening to weather.

Under that is a quieter worry. You don't want to become one of those people who thank the waiter three times and tell the team they're amazing. What you distrust in them is that everything is amazing, so amazing means nothing. When the person known for finding the wrong total says "the north region section is really clear," the colleague knows it was read the way you read everything, looking for faults, and this is what you found instead. Your reputation is what makes your thanks believable.

What about a gratitude journal? I won't recommend one. The research behind it is real but smaller than its reputation. A 2016 meta-analysis by Don Davis and colleagues found gratitude exercises beat doing nothing by a modest amount, beat a filler activity by less, and did no better than a comparison designed to help. And a notebook is private. Nothing in it crosses the room. Nathaniel Lambert and colleagues had people spend three weeks either expressing gratitude to a friend, thinking grateful thoughts about them, or doing other things, and the ones who said it out loud ended up seeing the friendship as stronger than every other group, including the ones who only thought it. That finding is about the person doing the saying, and it was the saying that did it. The NHS puts "saying thank you to someone for something they have done for you" under giving to others, which is the right category. If the notebook helps you see the fine, keep it, and then say what's in it to the person it's about.

Many readers learned the scan from a parent. A good school report got "what happened in geography?" You turned out capable, so the method seems to have worked. What you can do is remember how it felt to do eighteen things right and be asked about two, and to start showing your work braced, or not at all. The scan was learned because it was the only kind of attention anyone demonstrated, and the second look was never modelled. That's an omission, and omissions can be filled at any age. You can be where the second kind starts.

And there's the feeling that you're the only one holding it all together. That feeling is made by the scan. Your own effort is visible to you in full, because you were there for every pan you washed. Other people's effort shows up only when it fails. Over a year the ledger shows your work whole and everyone else's only where it went wrong, and of course it looks as if you're alone. Meanwhile the people around you feel nothing they do is seen. Both sides are reading the same ledger. One of you kept it and the other one is in it.

Try a fairer count for a week, if you're willing. Each evening, list what other people in your house or team did that went as it should. Only theirs, and only the fine. You'll have to go looking, because none of it rang. Most people find a longer list than they expected, and feel less alone at the end of the week with nothing in the household changed.

Left alone, the scan has a slow cost. People stop producing the fine for someone who never sees it. The colleague who used to send drafts early now sends them at the deadline, because early only ever bought an earlier list of faults. Your daughter shows her homework to the other parent. Your partner still does the table, without mentioning it, until one day they don't, and you notice that, of course.

Look past the pan

Pick the moments when the scan runs hardest. For most people that's walking into a room, opening a piece of someone else's work, and the first minute of seeing someone. At those moments, before you say anything, ask one question.

What's fine here?

Nothing good or wonderful is required. Just what's as it should be. In the kitchen, the table, the post, the bag. You'll have to look past the pan to find them, and looking past the pan is the whole skill. The fine will never ring the way the pan does, so it has to be looked for on purpose.

Don't try to see all of it all the time. You'll fail and stop. Choose three moments, the same three every day for a fortnight. Walking in from work, sitting down to a meal, opening the first piece of someone's work. At each one, find one thing. Three repetitions a day is how a look gets built until it runs on its own.

Expect it to feel like nothing. The pan comes with a small charge, a flicker, and that's how you know it's there. The table comes with none, so seeing it will feel flat, as if you've noticed something that doesn't matter. That's the filing at work. The charge measures what's unexplained, and it has never measured what matters.

Some rooms really are a mess and some work really is poor. Ask anyway. The answer might be that the person turned up, or tried, or got the first half right. If there's truly nothing, say nothing about the fine. Don't invent it, because the second list has to be as honest as the first, or neither will be believed.

Once you've seen it, say it. That's the only part that reaches anyone.

  • To the person, looking at them, not to the room or a third party later.
  • Now, while the thing is still in view. "Thanks for laying the table," while the table is laid.
  • Specific. "Thanks for sorting the post." "The section on the north region is really clear." "Eighteen out of twenty, and you got the fractions ones." Specific thanks shows them the evidence. "You're great" asks them to take your word for it.
  • Short, one sentence or two. The awkwardness you're forecasting comes mostly from imagining a speech, so don't give one.
  • No "but". "Thanks for the table, but the pan's still there" is a fault with a run-up, and it teaches people to brace when they hear thanks.
  • No "finally", no "about time", no "see, you can do it". If you can't say it without the sting, wait until you can.
  • No comparison. "That's more than your brother ever does" turns thanks into a ranking.
  • Out loud. If you can't manage that yet, a text is a start. "Just saw the post is sorted, thanks." It's a poor second to a face, but it crosses the room.

The first week will feel strange, and the strangeness is yours. The other person may look surprised, which is the point, since they didn't know. If they ask what's brought this on, the honest answer works fine. "I noticed I only ever mention what's wrong, and I'm trying to mention the rest." Most people, told that, will help you.

The fault list stays. Now that it isn't carrying the whole conversation, you can look at it with a little more care, through three questions.

  • Does it need saying? The wrong total does, or it goes to the board. The pan needs washing, which is a different thing from needing saying. If a remark leads to a fix, it's information. If it only tells the other person you noticed, it's commentary.
  • To whom? Your daughter, shown the two wrong answers, will learn more than if she's told about them. The colleague, not the meeting.
  • When? Not in the same breath as a thanks, not as the first thing on walking in, not at bedtime. Faults land best when people aren't already braced.

For the pan-sized ones there are two honest responses. Fix it, and there's nothing to say. Or drop it, and there's nothing to say. Leaving it and mentioning it costs the most and does the least, and it's what the scan reaches for by default. You may have heard there's a right ratio of fines to faults. I won't give you one. The ratios that get quoted come from disputed research, and a target turns thanks into bookkeeping. Before something goes on the fault list at all, ask whether it's wrong or just not yours. The dishwasher stacked another way, a paragraph in a style you wouldn't use, a child's idea of tidy. A lot of commentary is preference read out as standards, and the people hearing it can tell even when the person saying it can't.

All of it fits in one sentence you can carry into a kitchen.

When I notice something's fine, I say so to the person, now, and specifically. When I notice something's wrong, I decide whether it needs saying before I say it.

Peter Gollwitzer and Paschal Sheeran reviewed ninety-four studies of plans in the form "when this happens, I'll do that," and people who plan that way follow through much more often than people who only intend to. So write three, and put them on the fridge.

  • When I walk in from work, before I say anything, I'll look for one thing that's fine and say it.
  • When I open someone's work, I'll find one thing that's good in it and say that first, and raise the faults separately.
  • When I'm about to point out a fault, I'll ask whether it needs saying, and if it's pan-sized I'll fix it or drop it.

The table was laid

A partner is where the commentary has usually run longest, so start there if you can. The first specific thanks may get a suspicious look, because someone who has lived under the scan for years hears a fault behind every opening. Keep it short and don't explain unless asked. If they say "what do you want?", the honest answer is "nothing, I just saw it." It may take a few weeks before thanks is heard as thanks.

With children, say the eighteen before the two, and specifically. "You got all the fractions ones" tells her what she can do. Then look at the two with her, as something to understand. Children whose fine gets seen keep showing their work, and that's the only way you'll get to help with the two.

At work, the first good thing goes first and separately, and the wrong total after. If you manage people, this is most of managing, and the scan is the commonest reason a capable manager is called hard to work for.

If someone in your life never thanks you, you can't make them, and this isn't a way to get thanks back. Say yours anyway, because the rule is about your record. If it matters, ask them, plainly, once.

If you're the one under someone else's scan, their remarks are a record of what rang, and the silence about your fine is an omission. It's no more pleasant for knowing that, but it's less personal. If the commentary has moved past faults into contempt, or you find yourself afraid of it, that's a different matter from anything here.

If the scan has recently got much worse, or you see the table week after week and feel nothing, that may be low mood, which makes the fine invisible in a way no method reaches. The NHS guide is that low mood or loss of pleasure for most of the day, for two weeks, is a reason to see a GP. Go.

And you'll slip. You'll walk in one evening and say "the pan's still there" before you've looked at anything. It's a scan that has run for decades doing what it does. Treat it like any small fault and fix it on the spot. "Sorry, that came out first. The table looks great." The rule is about the next room.

It's half past six on a Tuesday, some weeks on. You walk into the kitchen and the pan is on the hob with this morning's porridge on it. You see it first. You always will.

Then the question asks itself, as it does about half the time now. The table is laid. The post is in its pile. The bag is by the door. "Thanks for doing the table," you say, and your partner, at the sink, says "That's all right," with nothing else in it. There was surprise, the first few times. You wash the pan. It takes a minute and there's nothing to say about it, so you don't.

Later your daughter brings her homework, three wrong this time. You find the long division among the seventeen, which she couldn't do last month, and you say so. Then she shows you the three. One is a slip and two are the same thing not understood, and you sit down with that one thing. She doesn't put the book away.

On Thursday the report arrives two days early, which it hasn't for a while. You tell your colleague the north region section is the clearest thing you've read all quarter. An hour later, in a separate message, you mention the wrong total on page six. It's fixed in ten minutes, and the reply says "good catch", which it hasn't said for a year.

The scan is still running. The pan still rings and the table still doesn't. The record just has two columns now, and the people around you can hear both.

Tomorrow, when you walk into a room, look for one thing that's fine, and say it to the person who did it before you say anything else.