You Have Been Right for Thirty Years. That Is Why You Believe It.

You sent it at 4:10. Four lines, and one of them was “Happy to talk this through if useful.”

It is 6:58 now and you have opened the Sent folder for the fourth time to read it again. Not for a typo. For a tone. Something in that last line that might read as clipped to a person who was already having a bad day. The message has not changed. You have.

The short answer: thoughts are not facts, because a thought is something your mind produces rather than a report on the world. Under load your mind produces more of them, and more negative ones, the way a tired eye produces after-images. The trouble is never that the thought turned up. The trouble is that you read it as information. Research on cognitive defusion, the practice of holding a thought as a thought, found it lowered how believable a negative thought felt without changing how often it arrived.¹

Key takeaways

• A thought arrives in your own voice, in the second person, with no source line attached. That is why it reads as a report rather than as something produced.
• Bad information is processed more thoroughly than good, so the critical thought gets more of your attention per word than the reassuring one.²
• Cognitive defusion aims at the believability of a thought rather than its frequency. In one small study, briefly repeating a negative self-referential word reduced both its believability and the discomfort it caused, more than distraction or thought control did.³
• Arguing with a thought keeps it at the centre of the room. Labelling it as a thought moves it to the side.
• The whole practice is one sentence, said twice, twenty seconds.

In this article: Why you believe your own thoughts · Why knowing this changes nothing · Why it goes to the worst · Why it is loudest in a quiet house · Why it sounds like judgment · Why you go back over finished things · Why a thought decides what you do · The six-word prefix · FAQ

Why do I believe my negative thoughts?

Because a thought arrives without a source line. Everything else in your day comes labelled: an email has a sender, a figure in a deck has a footnote, a comment in a meeting has a face attached. A thought shows up in your own voice, already finished, with nothing telling you how much to trust it.

That missing label does more work than it looks like it should. When a colleague says the middle section ran long, you weigh it against what you know about the colleague. When your own head says the same thing at 6:58, there is nobody to weigh.

So the default is to trust it entirely. Not out of poor judgment. Out of habit, and out of a working life that rewarded exactly this. You have been right for thirty years about things other people missed. That is precisely why you believe yourself now.

There is a second reason, and it is measurable. A review drawing together research across everyday events, close relationships, social networks and learning found that bad information is processed more thoroughly than good, and that bad impressions form faster and resist correction harder.² So the thought that says the message read as clipped does not merely arrive. It gets read properly. The thought that says the message was fine gets skimmed, if it turns up at all. Nine people say it went well and you spend the evening with the tenth, which is the asymmetry that makes criticism stick and praise slide off, and it is not a flaw in how you are built.

Put those two together. A statement that arrives unlabelled, in your own voice, and gets the deepest read of anything in your day. Of course you believe it. It would be strange if you did not.

Here is what makes it worth examining anyway. The mind produces thoughts continuously, and the vast majority are not about anything. What flavour of crisps. Whether the bins went out. Whether that woman on the train is cold. You do not treat any of those as reports on reality, and you do not check them. You notice them and they go. The ones you check are the ones about you, about whether you are doing this well, about whether the last four years were a mistake. And the checking is what marks them out. You have quietly assigned a whole category of sentences the status of evidence, and only that category.

If thoughts are not facts, why does knowing that change nothing?

Because the sentence “thoughts are not facts” is itself a thought, arriving in the same voice as everything else and carrying the same lack of authority. You can agree with it in the abstract and still spend Thursday evening certain the message read as clipped. Knowing a thing in general changes very little at 6:58.

The usual advice at this point is to challenge the thought. Look for the evidence. Ask what you would say to a friend. Rate how likely it really is. That approach is well established and it genuinely helps some people, some of the time. It also has a cost that nobody mentions, which is that it keeps the thought at the centre of the room. You are now in a debate with it, on its terms, about whether it is true. And the sort of person who is good at their job is often very good at building the case for the prosecution.

There is a different move. Instead of arguing about whether the sentence is true, you change what kind of object it is.

Definition: cognitive defusion is the practice of relating to a thought as something your mind produced rather than as a report on the world, so that the sentence is still there but stops giving orders.

The clinical work behind it comes from Steven Hayes and colleagues, whose one-line version of the aim is to see thoughts as what they are, rather than as what they say they are.⁴ In two small randomised trials for depression, this approach produced improvement equivalent to or better than a full course of cognitive therapy, and it got there by a different route: it reduced how believable the depressing thoughts were, more quickly than the traditional methods did, without reducing how often they occurred.¹

That last clause is the part to sit with. The thoughts kept coming. They just stopped being taken as evidence.

Two rooms in this house make the distinction unusually easy to see. The first is the 11 p.m. loop, where the same three seconds get re-run without ever producing a decision. There is a clean test for whether that thinking is going anywhere or simply circling, and it is worth learning: the difference between rumination and problem-solving is whether the thought moves toward something you would actually do.⁵

The second is the thought that begins “I should have caught that.” It feels like a fact because you now know how the thing turned out, and once you know the ending, the warning signs look obvious in a way they were not at the time. Psychologists call it hindsight bias, and a meta-analysis of 95 studies found the effect reliable enough that you can more or less assume it is operating whenever you review your own decisions after the fact.⁶ So the “I should have caught that” thought is the clearest case of all: a sentence that feels like recalled evidence and is actually a construction built after the event.

Why does my brain always go to the worst?

Because a mind under load does not generate neutral predictions. It generates the expensive ones, and it generates them first, because for most of human history the cost of over-preparing for a bad outcome was an afternoon and the cost of under-preparing was everything.

The prediction feels different from a memory, which is part of why it gets waved through. A memory is about something that happened and can in principle be checked. A prediction is about something that has not happened, so there is nothing to check it against, and the mind fills that gap with vividness instead of evidence. You do not merely think you might struggle at the next level. You can see the meeting where it becomes obvious to everyone. The detail feels like knowledge. It is production.

This is at its loudest around anything with a step up in it. A promotion, a new remit, a bigger room. The thought that says you are not built for the next level presents itself as a preview, and previews are footage of something that already exists. There is no footage. There is a sentence, produced by a mind that has been running hot for a while, about a version of you that has not been asked to do the job yet.

It also runs on a timetable. Sunday, roughly 4:40 in the afternoon, the light going, nothing wrong. What arrives then is not a memory of last week and not news about next week. It is anticipation, and it turns up reliably enough that you can almost set a clock by it. The Sunday dread that arrives on schedule is the same mechanism with a standing appointment.

Worth naming the honest limit here. Sometimes the worst-case thought is correct. Sometimes the restructure really is coming. Holding a thought as a thought settles nothing about whether it is accurate, and nobody here is asking you to be relaxed about a real risk. The ask is smaller. Notice when your mind has produced a possibility and handed it to you dressed as a finding.

Why is the thought loudest when the house is quiet?

Because volume is relative, and at 3 a.m. there is nothing to be relative to. The same sentence at 3 in the afternoon gets buried within seconds by a phone call, a doorway, a reminder firing. Nothing about the thought grew larger overnight. Everything around it got smaller.

That is worth sitting with, because it means the night version is not a more honest version. It is the same sentence with the competition removed. At 3 in the morning it has the stage, the lighting and an audience of one, and it performs accordingly.

The clock does its own damage. 3:14 is not a neutral number at 3:14. It is proof that you have four hours left, then three, then two, and the arithmetic becomes a second loop running alongside the first. The meeting that reconvenes at 3 a.m. is rarely new material. It is old material with no competition.

The daytime version of this is more interesting, and easier to miss, because it does not look like thinking at all. It looks like a cursor. The document is open, the deadline is real, and you have read the same paragraph three times without any of it going in. That is usually not a concentration problem in the ordinary sense. There is a sentence running underneath the work, quietly, using capacity, and what you experience is the shortfall rather than the sentence. The cursor blinking on a half-written email is the tax the loop charges during office hours.

If the voice sounds exactly like your own judgement, the Quiet Audit is ten minutes on where it learned to speak in your accent. Start here.

Why does it sound like my own judgment rather than a thought?

Because it uses your voice, your vocabulary, and thirty years of accurate observations as collateral. The inner critic does not sound like an intruder. It sounds like the clear-eyed part of you, the part that does not flatter and has been right before, which is why the sentence arrives already credentialed.

When it says “you always do this,” it is drawing on real authority. You have in fact been the person in the room who saw the problem first, repeatedly, for decades. The credentials are genuine.

What gets done with those credentials is the sleight of hand. Real judgment is about a thing: this paragraph is weak, this number is wrong, this hire will not work. The critic starts there and then quietly promotes the claim. The paragraph is weak becomes you are sloppy. The number is wrong becomes you are slipping. Somewhere between the observation and the toothbrush, a comment about a middle section became a verdict on whether you should be doing this job at all, which is a considerably larger claim than the evidence was making.

Kristin Neff’s research on self-compassion is useful here, and not for the reason people expect. The finding is not that being kind to yourself feels nicer. It is that harsh self-criticism does not produce the accuracy or the motivation it promises.⁷ The critic sells itself as your quality control. It performs poorly in that role. Self-criticism is a rehearsed habit rather than a standard, and the two-minute colleague test in that piece is the fastest way to see the difference.

The comparison version works the same way and is even harder to spot, because it presents as data. She seems to have it together. He never seems rattled. But you are comparing the hour of somebody else that you can see against the whole of yourself that you cannot stop seeing, and that arithmetic has never once come out in your favour. Measuring yourself against a colleague is a thought wearing a spreadsheet.

Why do I keep going back over things I have already finished?

Because the mind guards whatever it has filed as unresolved, and a thought that says check it is hard to tell apart from a thought that says something is wrong. Going back over it feels like diligence. Decades of research on rumination found that it does not improve the outcome.⁵

That finding is worth reading twice, because everything about the experience argues against it. It is structured like work. It uses the same concentration work uses. And it measurably reduces your ability to solve the thing you are turning over. The replay resembles thinking without doing what thinking does.

The conversational version is the one most women in this position know best. The exchange ended at 4:10 and by 6:58 you have reconstructed it three times, each reconstruction slightly worse than the last, because memory is not a recording and every pass edits toward the version you fear. Replaying a work conversation after work is the daylight half of the 3 a.m. habit, and it has a four-minute interruption that works.

Checking is the physical form of the same thing, and it has a sting in it. You go back to confirm the door is locked, the email went, the number was right. It feels like it should settle the matter. It does the opposite: repeated checking reduces your confidence in the memory rather than strengthening it, because each new look overwrites the last and the recollection gets blurrier with every pass. The third time you opened the Sent folder is why the fourth time was never going to be the one that settled it.

Which is the actual answer to the 6:58 on the evening this article opened with. You were not gathering information. You were performing a reassurance ritual that removes reassurance at a steady rate.

Why does a thought get to decide what I do next?

Because a thought held as a fact is an instruction, and instructions get followed. This is where the idea earns its keep, because it stops being about how you feel in the evening and starts costing you hours, documents, decisions and the temper you had at 6:40.

Take the proposal that has not been opened. The inbox is spotless, the expenses are filed, three people got useful replies, and the document that matters is still closed. That is not a time management problem, because you plainly had the time and spent it on smaller things. What happened is that a thought about the proposal arrived, something in the register of “this needs to be good and you are not sharp enough today,” and the discomfort of that sentence got solved by doing something else. Procrastinating on the task that matters is a mood being managed, and the ten ugly minutes in that piece are how you get around it.

Then the cereal aisle. Four minutes, two boxes, a decision that costs nothing either way, from a woman who signed off six-figure calls before lunch. The thought there is usually about the deciding rather than the cereal: some version of “why can I not do this, what is wrong with me today.” That second sentence is doing more damage than the first, and it is what turns forty seconds into four minutes. Standing in the aisle unable to choose has a mechanism, and it is not willpower running out.

Rest is the same shape with higher stakes. You sit down, and before the sitting can start there is a small hearing about whether you have earned it. The list is done, and still the thought arrives to ask what makes you think you are entitled, and ten minutes go on the hearing rather than the rest. The guilt that shows up when you stop is a thought demanding a justification that no one else in the house has been asked for.

And the evening. You snapped at the person who asked what was for dinner, and it was not about dinner, and you knew that while it was happening. Holding a thought as a fact all day is work, and the capacity for doing that work is finite. By 6:40 there is none left, which is why the family gets the version of you the office never sees. The irritability that only ever lands at home is the bill for a day of taking every sentence seriously.

Fifteen rooms, one mechanism. That is the whole argument of this article. The 3 a.m. replay, the inner critic, the cereal aisle, the unopened proposal, the tenth piece of feedback: every one is a thought arriving and being read as a report on reality instead of as an event in a mind.

How do you stop believing every thought you think?

Not by getting better at arguing with it. By changing what kind of object the sentence is, which takes about twenty seconds. You say the thought as it arrived, then say it again with six words in front of it, and notice what changes.

The six-word prefix

The next time a sentence lands and you feel the pull to check something, do this before you check.

1. Catch the sentence in its actual words. Not the summary. The real thing, with its real wording: that message read as clipped.
2. Say it once, out loud if you are alone, exactly as it arrived.
3. Say it again with six words in front of it: I’m having the thought that that message read as clipped.
4. Notice what changed. Not whether it is true. Whether it is still giving orders.

That prefix is a standard defusion technique, one of a long list assembled by Hayes and colleagues for exactly this purpose.⁴ It works because of what it does to the grammar. “That message read as clipped” is a claim about the world, and claims about the world require investigation. “I’m having the thought that the message read as clipped” is a description of something happening in your head, and a description of a mental event does not need to be settled before bed.

The nearest thing to hard evidence for a single defusion move on its own comes from a small study that tested an even more basic technique: repeating a one-word version of a negative self-referential thought, quickly, out loud, for thirty seconds. Compared against distraction and against active thought control, the repetition reduced both the discomfort and the believability of the thought for every participant.³ It was eight people. It is a preliminary finding rather than a settled one, and the authors say so plainly. It is offered here as a small piece of support for a very old idea, not as proof of anything.

Three honest notes, because this is where these articles usually oversell.

It will feel ridiculous. Adding six words to a sentence to change your relationship with it sounds like something that should not work, and the first two or three times you will be aware of yourself doing it. That fades. The clumsiness is not evidence against the technique.

It does not make the thought go away, and it is not supposed to. The research is specifically about believability rather than frequency.¹ The sentence may well turn up again tomorrow. The change is what happens in the ninety seconds after it does.

It is the wrong tool for information. If the thought is “the filing deadline is Friday,” that is a fact and it needs a calendar, not a prefix. This is for the sentences that hand down verdicts about you: the ones that say what you are, what other people concluded, how this will go. Those are the ones dressed as reports.

Dr Julie Smith, a clinical psychologist who has spent years explaining this to people who are not in therapy, puts the whole thing in two lines worth keeping. Thoughts are not facts. They are a mix of opinions, judgments, stories, memories, theories, interpretations and predictions about the future.⁸ The word doing the work there is mix. Some of what your mind produces is accurate. Some of it is a story you formed in 1998. It arrives in one undifferentiated stream, in one voice, with nothing marking which is which, and you are not going to sort it by thinking harder.

Frequently asked questions

What does “thoughts are not facts” actually mean?
It means a thought is an event in your mind rather than a report on the world. Your mind produces sentences the way it produces images and memories, continuously, without checking them first. The sentence may turn out to be accurate or not. Either way, its arrival is not evidence.

Is “thoughts are not facts” just a slogan?
As a phrase on its own, more or less, and that is fair to say out loud. The phrase does nothing. What does something is the specific move underneath it, where you restate the sentence as a thought you are having and notice whether it still commands anything. The research is on the move, not the motto.

What if the thought is actually true?
Then it stays true after you have labelled it, and nothing has been lost. This is not a technique for deciding thoughts are false. It is for separating the question of whether a sentence is accurate from the reflex to obey it immediately, which are two different questions that usually arrive fused together.

Isn’t this the same as positive thinking?
No, and in a sense it is the opposite. Positive thinking swaps the content of a thought for nicer content. Defusion leaves the content alone and changes your relationship to it. Hayes and colleagues found improvement without the negative thoughts becoming less frequent.¹

Why does saying it out loud matter?
Because hearing a sentence as sound is what makes it audible as a construction rather than as truth. The word-repetition work rests on the same principle: say something enough times and the meaning thins out while the sound comes forward.³ You do not need thirty seconds of repetition for most thoughts. One spoken pass is often enough.

How long until this works?
The effect, when it comes, tends to arrive inside the same minute rather than after weeks of practice, which is unusual and part of why it is worth trying. What takes longer is remembering to do it at all. The prompt to attach it to is the moment you reach to check something you have already checked.

When is this worth taking to someone?
When the thoughts are constant rather than occasional, when they spread across every subject rather than clustering around work, or when they come with weeks of flat mood and sleep that will not hold. A technique is for a habit. That pattern is worth a conversation with your doctor rather than a search bar.

Read next

• The 3 a.m. Replay: Why Racing Thoughts Bring the Meeting Back at Night. The same mechanism with the competition removed.
• Self-Criticism Is a Habit, Not a Fact. When the sentence arrives in your own voice and sounds like standards.
• The Thought That Won’t Land: Rumination or Problem-Solving?. The test for whether the loop is going anywhere.
• Twenty-Five Years In, and You Still Think They Will Find Out. The same machinery, pointed at your competence instead of your day.
• The Three Good Days That Made You Uneasy. What the mind is doing when nothing is wrong and you brace anyway.
• You Got the Thing. You Felt Almost Nothing. Why a win stops registering, and the two routes that erode it.
• You Did Not Change. Your Life Got Narrower. When the thought is about who you are rather than what you did.

Two other things drive the same exhaustion from different directions. If what you are carrying is a job nobody can see rather than a thought you cannot drop, start with the work that never appears on any list. If the lost words and the fog arrived alongside everything else, what the brain scans actually found is the better place to begin.

If you want the next quiet step

A mind produces more of these when it is carrying more. That is the part no technique reaches, because the sentences at 6:58 are downstream of a week that had more in it than anybody could see. The Quiet Audit is a short, private set of questions that names what is actually on your ledger at the moment: the deadlines, the calendars that are not even yours, the things you agreed to in March. Nobody else sees your answers. Start here.

The message you sent at 4:10 was four lines long and one of them was kind. It has been the same four lines all evening. You can close the tab.

References
1. Luoma, J.B. & Hayes, S.C. (Scholar) Cognitive defusion. In O’Donohue, Fisher & Hayes (Eds.), Empirically Supported Techniques of Cognitive Behavior Therapy. Wiley. personal.kent.edu
2. Baumeister, R.F. (Wikipedia), Bratslavsky, E., Finkenauer, C. & Vohs, K.D. Bad is stronger than good. Review of General Psychology. 2001;5(4):323-370. assets.csom.umn.edu
3. Masuda, A., Hayes, S.C., Sackett, C.F. & Twohig, M.P. Cognitive defusion and self-relevant negative thoughts: examining the impact of a ninety year old technique. Behaviour Research and Therapy. 2004;42:477-485. contextualscience.org
4. Association for Contextual Behavioral Science, Cognitive Defusion (Deliteralization); clinical materials assembled by Gifford, Hayes & Strosahl. contextualscience.org
5. Nolen-Hoeksema, S. (1959–2013; in memoriam), Wisco, B.E. & Lyubomirsky, S. Rethinking rumination. Perspectives on Psychological Science. 2008;3(5):400-424.
6. Guilbault, R.L., Bryant, F.B., Brockway, J.H. & Posavac, E.J. A meta-analysis of research on hindsight bias. Basic and Applied Social Psychology. 2004;26(2-3):103-117.
7. Neff, K. (Wikipedia, Scholar) Self-Compassion: The Proven Power of Being Kind to Yourself; research summaries at self-compassion.org.
8. Smith, J. (LinkedIn, Linktree) Why Has Nobody Told Me This Before? Goodreads.

Written by Nora Whitfield for Blue Leaf Journal. Updated: August 12, 2026.