You feel guilty. So you must have done something.
Notice there was no step in the middle. You didn’t recall what you did, weigh it, and conclude you were at fault. The guilt showed up first, and the wrongdoing was inferred backwards from it — as though the feeling were a receipt for something.
What it is
Burns’ definition: “You assume that your negative emotions necessarily” reflect the way things really are, and “you take your emotions as evidence for the” truth [1]. His own example is exactly the one above: “I feel guilty. Therefore, I must” have done something bad.
He rates it highly among the ten, writing that “emotional reasoning plays a role in nearly all your depressions” [1] — his claim, about the pattern, not ours about you.
The reason it deserves that billing is that it runs underneath the other nine rather than beside them. Catastrophizing makes a prediction. Overgeneralization makes a rule. Emotional reasoning makes any of them feel confirmed, because the anxiety that follows a catastrophic thought gets read as proof the catastrophe is coming.
What it sounds like
- “I feel anxious, so this situation must be dangerous.”
- “I feel like a fraud, so I probably am one.”
- “I feel guilty, so I must have done something wrong.”
- “I don’t feel like doing it, so it must be too hard.”
- “It feels hopeless, so it is.”
The tell is a two-part sentence: “I feel X” followed by “so it must be Y.” Sometimes the second half is spoken. More often it’s silent, and you only see it by writing the first half down and asking what you concluded from it.
A useful version of the test: would this hold if someone else felt it? If your friend said “I feel like a fraud”, you would not accept that as evidence they were one. You’d ask what happened. The standard you apply to other people’s feelings is the right one; it’s just not the one you apply to your own.
Why it’s convincing
Because feelings genuinely are information — just not about the thing you’re pointing them at. Anxiety reliably tells you your alarm system fired. It does not tell you whether there was anything to fire at. Smoke detectors are useful and they go off when you make toast.
It’s also faster than thinking. A feeling arrives whole and instantly, while evidence has to be assembled, and the fast thing wins when you’re tired. Beck’s account of automatic thoughts describes precisely this — plausible, unbidden thoughts that get accepted without examination because they never present themselves as claims to be checked [2].
And it’s circular in a way that’s hard to see from inside. The thought produces the feeling, and the feeling is then read as proof of the thought. Nothing external ever enters the loop.
How to catch it
Split the sentence in two. Write the feeling on one line and the conclusion on another. “I feel anxious.” / “This is dangerous.” Two separate claims, only one of which you have evidence for. On a page the join is visible; in your head it isn’t there at all.
Name the feeling precisely, then stop. Not “bad” — dreading, ashamed, resentful, lonely. Naming it accurately does two things: it makes the feeling smaller, and it finishes the sentence, so there’s no trailing “…so it must be” left hanging.
Rate it. Low, medium or high. A feeling with a size is a feeling you can compare to yesterday, and comparison is the thing certainty can’t survive.
Then ask the evidence question separately. What actually happened? What would you tell a friend who described these facts, without knowing how you felt about them?
What to do once you’ve caught it
The fairer version keeps the feeling and drops the inference. “I feel anxious about this, and I don’t yet know whether there’s a real problem” is accurate, believable, and — importantly — it doesn’t ask you to stop feeling anxious. You’re not arguing with the emotion. You’re revoking its promotion.
How aMind helps with emotional reasoning
The structural fix for this pattern is to stop the feeling and the conclusion from arriving as one object, and that is how aMind’s reflection is laid out. Naming the emotion is a step of its own — you pick from a list of 26 and rate each one Low, Medium or High — and it sits apart from the step where you weigh the thought. The feeling gets fully recorded. It just doesn’t get to count as evidence, because the evidence question is asked on a different screen.
That separation is the whole intervention, and it’s very hard to perform on yourself unprompted. Then, on the paths where it applies, aMind asks you to re-evaluate the emotion at the end, after you’ve worked through the evidence and written a fairer version. Seeing an emotion you rated High come back down to Medium once the thought was examined is about the most direct demonstration there is that the feeling was tracking the thought, not the world.
The other nine patterns are in the guide to cognitive distortions. If this one is yours, naming a feeling precisely is the skill that does the most work on it.
