Somebody says something kind about you and your first move is to correct them. Not out loud necessarily, but the sentence forms anyway: they are being generous, they did not see the part that went wrong, anyone would have managed it. If that is familiar, the thing to know is that it is not a manners problem and not a confidence you can talk yourself into. It is a preference for information that matches what you already believe.
Psychologists call it self-verification, and it explains both why compliments feel bad and why almost all the standard advice about them does not work.
It is not modesty, it is consistency
The finding is counterintuitive enough that it took a run of studies to establish. People do not simply want to be thought well of. They want to be seen accurately, by their own definition of accurate, and when those two things conflict, a surprising number choose accuracy.
Swann and colleagues put people in a position to choose who they would interact with and tracked what they said as they decided. Participants with negative self-views gravitated towards partners who shared those negative views, and the reasoning they gave out loud was about predictability and control: a person who sees you the way you see yourself is someone you know how to be around. [swann-1992-self-verify]
That is worth sitting with, because it reframes the discomfort entirely. When a compliment lands badly, you are not failing to accept a gift. You are being handed a description of yourself that does not match your working model, and the mind treats a mismatch as something to resolve rather than something to enjoy. Resolving it means finding the error, and since the model feels solid, the error must be in the compliment.
The everyday version is the speed of it. The correction arrives before you have decided anything, which is the tell that this is not a social calculation.
Why “just accept the compliment” fails
The standard advice is to say thank you and stop there, and as a piece of etiquette it is fine. As a fix it does nothing, because it addresses the sentence and not what happens after it.
The stronger version of the same advice, telling yourself the good things directly, is worse than nothing for exactly the people it is aimed at. In a well-known experiment, participants asked to repeat “I am a lovable person” were measured afterwards on mood and self-regard. People who already had high self-esteem got a small lift. People with low self-esteem ended up feeling worse than a control group who had not done the exercise at all. [wood-2009-affirmations]
The mechanism is the same one. A statement far enough from your working self-view does not replace the view, it draws attention to the distance. You say the words, some part of you produces the counter-evidence, and the net effect is a review of the counter-evidence.
Which version of this is yours?
Tick what actually happens, not what you would like to happen.
0 of 6 ticked
What you are describing runs on after the interaction ends, which puts it outside the range of manners. It is also the pattern that responds best to work on self-worth rather than on confidence or social skills, because the belief being protected is the target.
Some of this is a learned reflex you could drop with practice, and some of it is genuine disagreement with the compliment. Worth separating them: the reflex responds to noticing, the disagreement responds to the reframe further down this page.
Finding praise a bit uncomfortable is close to universal and not a problem to solve. If it is not following you around afterwards, it is probably just the standard British reflex.
The reframe that got past it
Here is the useful part, and it comes from a study designed around exactly this obstacle.
Marigold, Holmes and Ross ran three studies on people in relationships, comparing how those with low and high self-esteem responded to a compliment from their partner. The manipulation was small: rather than recalling the compliment or repeating it, some participants were asked to describe what it meant and why it mattered that their partner had said it.
That single change closed the gap. People with low self-esteem who described the meaning and significance of the compliment felt as positively about it, about themselves and about their relationship as people with high self-esteem did. The effect was still detectable two weeks later. [marigold-2007-reframing-compliments]
Why it works is consistent with everything above. “You did that really well” is a specific claim about a specific event, and a specific claim invites specific counter-evidence, which you have plenty of. “My partner noticed and thought it was worth saying” is a claim about them and about the relationship, and it is much harder to argue with, because it is simply true that they said it.
Three minutes on one compliment
Pick something kind that somebody said to you recently, ideally in the last week. Do not write down the words. Write down what it means that they said it, and what it says about how they see you.
2:00
If you find yourself listing reasons they were wrong, that is the pattern doing its job. Note it and go back to the question of why they said it.
Done. Do this with the next one too, while it is fresh.
Two things about doing this properly. Write it rather than think it, because thinking allows the correction to run in parallel and writing does not. And do it soon after, while you can still remember the specifics, since the pattern erodes praise from memory considerably faster than criticism.
What it costs, and who pays
The framing so far has been about how this feels from inside, which understates it. The other person is in the interaction too.
From the outside, a deflected compliment is a rejected one. You told somebody their judgement was wrong about something they were trying to give you, and however gently that is delivered, it registers. Do it consistently and the compliments stop, not because the regard has gone but because offering it keeps being unrewarding.
In a relationship this becomes a closed loop, and it is the loop the Marigold work was built to interrupt: doubt your value to your partner, resist the reassurance, receive less reassurance, take that as confirmation. Nothing in the loop requires anyone to behave badly, which is why it can run for years.
The way out is not persuading yourself you are worth it. It is changing what you do with the words in the thirty seconds after they arrive, which is a much smaller ask and, unlike the belief, is under your direct control. Our guide to what self-worth actually is covers why the belief itself is the wrong lever, and what tends to move instead.
When to seek help
Most people can work on this alone, and the reframe above is a reasonable place to start. It is worth taking to a therapist when the pattern is costing you something concrete: a relationship where your partner has said they feel shut out, opportunities declined because you were certain you would be found out, or a level of self-criticism that has stopped being background noise.
Speak to a GP sooner if the picture underneath this is low mood rather than low self-worth specifically, meaning things you used to enjoy have stopped landing, sleep has changed, or you have had thoughts of not wanting to be here. Those are treatable and they are a different problem from finding praise awkward. The Rosenberg self-esteem scale is a reasonable way to see where you sit before that conversation, and the self-assessment hub has the rest of what we offer.
How MyFreud can help
The difficulty with this pattern is that it is invisible from inside a single day. You cannot see how much praise you deflected this month, and you certainly cannot compare it to how much criticism you kept. Writing down one compliment a week with what it meant, rather than what was said, gives you the record the research says is the active ingredient, and it gives you something to read back on the days when the working model feels like settled fact.
Download MyFreud and start today: App Store or Google Play.