Fallacy Fallacy

In plain terms

Someone argues badly for something. You catch the flaw. Then you treat the underlying claim as refuted.

That last step is the fallacy fallacy, also called argument from fallacy. A bad argument for a claim is a fact about the argument. It is not a fact about the world. The claim goes back to unproven, not to false.

This site has an obvious interest in being honest about that. Naming a fallacy proves less than most people assume: it removes one reason to believe something, and better reasons may be sitting right behind it. The social version of the same warning is in Why Calling Out a Fallacy Usually Backfires. This entry is the logical version.

Why it's fallacious

An argument is a route from premises to a conclusion. Showing that the route is broken shows you have not arrived. It says nothing about whether the destination is there. Other routes may reach it.

The mistake is treating "this doesn't follow from these premises" as if it were "this is false." Only one of those is a conclusion you have earned. Weak arguments are extremely common for true claims, because most people reach for the first reason that comes to mind rather than the best one available.

The correct update after spotting a fallacy is a drop in support, not a flip in sign. If the failed argument was the only support on offer, your confidence should fall a long way. It should fall toward uncertainty, which is where appeal to ignorance waits for anyone who mistakes "unproven" for "disproven."

Canonical example

"Sitting at a desk all day is bad for your back. My uncle did it for thirty years and now he can barely stand up straight." "One relative isn't evidence. That's anecdotal. So no, sitting is fine."

The first line is weak support: a single case, no comparison group, thirty years of other possible causes. The second speaker is right about that, and then keeps going one step too far.

Prolonged sitting and back pain have been studied properly, and the literature is not settled by the uncle in either direction. The anecdote was never load-bearing. Removing it leaves the question open, and the person who removed it has not answered it.

Counter-example (not a fallacy)

"You've given me three reasons and all three have problems. I'm not saying you're wrong. I'm saying you haven't shown you're right, so I'm where I started."

Rejecting an argument is not the same as asserting the opposite conclusion. Someone can take apart every case put in front of them and still hold no position at all, which is a defensible place to stand.

There is a second legitimate case. When a claim rests on one argument and that argument fails, confidence in the claim should genuinely fall. If the only reason for a revenue forecast turns out to be a spreadsheet error, doubting the forecast is reasonable. You have not proved the forecast wrong. You have removed the thing that made it credible, and credibility was doing real work.

The line: did the flaw send the claim back to unproven, or did you promote it to disproven?

How to fix it

If you've been linked here, someone thinks you won a point about form and then claimed the whole argument. The repair is one sentence long: state what you actually established. "That reasoning doesn't hold up" is defensible and often useful. "Therefore the opposite is true" needs its own evidence, and you haven't offered any. If you believe the claim really is false, argue for that directly, because it is a separate job with its own burden.

If you're on the receiving end, there is a decent chance you are the one who argued badly, and both things can be true at once. The productive move is to offer a better reason rather than to fight about the label. "Fair, that was a weak example, here is the study I should have led with" ends the argument about arguing and returns everyone to the actual question.