Thirty people at the talk. Twenty-nine engaged, good questions at the end, someone stayed behind to thank you. One person looked at their phone for most of it.
Guess which one you’re still thinking about on the train home.
What it is
Burns describes it in one line: “You pick out a single negative detail and dwell on it so that the entire experience seems negative” [1].
The two halves both matter. Pick out — the detail is real, and you selected it. Dwell — the selection isn’t a glance, it’s where your attention stays. And the result is the strange part: not a negative detail sitting inside a decent experience, but a whole experience that has taken on the color of one moment inside it.
This is why filtering is the hardest of the ten patterns to catch. With most of the others there’s something in the thought you can argue with. Catastrophizing makes a prediction you can test. Overgeneralization uses the word never, which one counterexample breaks. Filtering makes no false claim at all. The person really was on their phone. Everything you’re thinking is true. It’s just that it’s four percent of what happened, occupying a hundred percent of the account.
What it sounds like
- “The presentation went terribly” — about a talk that got good questions.
- “The whole trip was a write-off” — because of one bad afternoon.
- “They hated it” — based on one lukewarm line in a page of positive feedback.
- “I can’t stop thinking about what I said at dinner” — after four hours of a dinner that went well.
The tell isn’t a word, which is what makes it slippery. It’s a ratio. A verdict about the whole thing that rests on a very small part of the thing. When you catch yourself summarizing an event, ask what fraction of the actual time the summary is describing.
The second tell is the detail that won’t leave. If one moment is still being replayed hours later while the rest of the day has gone blurry, that moment has been selected, and the selection is doing the work.
Why it’s convincing
Because it’s accurate. Every individual thing you’re thinking about survives fact-checking, so the ordinary defense — is this actually true? — comes back clean and the filter stays in place. The distortion isn’t in the content. It’s in the editing.
It also has help from how attention works. Negative details are stickier: they’re more surprising, they feel more informative, and they’re the ones the mind flags as worth learning from. Beck’s account of automatic thoughts describes this machinery running below the level of deliberate thought — you don’t decide to dwell on the phone person, you simply find that you are [2].
And it compounds with mood. Feeling bad makes the negative detail easier to reach, and reaching for it makes you feel worse, which makes it easier to reach again.
How to catch it
Ask what percentage. Of the ninety minutes, how many were the bad bit? Two? The ratio question works where the truth question fails, because it doesn’t dispute the detail — it re-sizes it.
Write down the whole thing, not the summary. List what actually happened, in order, including the boring parts. Filtering depends on the summary; it does not survive a full inventory. This is the single reason writing beats thinking here — you cannot really list a day in your head, and the list is the whole intervention.
Name what you left out. Not to cheer yourself up. Just to put it back in the record. The good questions were as real as the phone.
Notice which detail got picked. Over time this becomes the interesting part. Most people’s filters are consistent — always the one critical comment, always the awkward silence, always the thing that might have looked unprofessional. The filter has a shape, and the shape tells you something.
What to do once you’ve caught it
The fairer version isn’t a happier verdict, it’s a complete one. “One person was disengaged and the rest went well” — both halves, in one sentence. You’re not arguing the detail away; you’re restoring its proportions.
Then keep the entry, because this pattern is nearly impossible to see in a single instance and obvious across ten. When the same kind of detail is the one you flagged every time, you’re no longer looking at a series of bad events. You’re looking at a filter — and that’s a much more useful thing to know.
The other nine patterns are in the guide to cognitive distortions. It pairs most often with overgeneralization — one detail becomes the whole event, and then the event becomes a rule.
How aMind helps with mental filtering
Filtering is the one pattern that survives the usual question, because everything you are thinking is true. It does not survive a full inventory — and an inventory is not something you can really do in your head, which is why this pattern needs writing more than any of the others.
aMind’s evidence step is that inventory. It asks for both sides: what supports the thought, and what argues against it. Filling in the second column is the whole intervention, because the good questions and the thank-you afterwards were as real as the one person on their phone, and they only make it into the account if something asks for them. The rest of the flow keeps the proportions honest — you name what you feel and how strongly, tap the pattern, and write a version that includes both halves. And because entries stay in History, the shape of your particular filter becomes visible over time: if the detail you flagged was the same kind of detail ten times running, you are not looking at ten bad events.
