Why Smart People Fall for These
Fallacies survive because they are structurally close to valid reasoning. Each one is a near-miss of a pattern that works, which is why they feel right in the moment.
Ten of them, in two groups: the formal ones, where the structure itself is broken, and the informal ones, where the structure is fine and something else has gone wrong.
The Formal Fallacies
These are errors in the shape of the argument. They can be identified without knowing anything about the subject matter — which is why they appear on tests.
1. Affirming the consequent
If it rained, the ground is wet. The ground is wet. Therefore it rained.
The conditional runs one way only. A wet ground is consistent with rain and with a burst pipe, a sprinkler, or a spilled bucket.
The valid version affirms the antecedent: it rained, therefore the ground is wet. One direction works, the other does not, and the gap between them is a single word.
2. Denying the antecedent
If it rained, the ground is wet. It did not rain. Therefore the ground is not wet.
The mirror error. Rain guarantees wetness; its absence guarantees nothing, because other causes remain available.
The valid version denies the consequent: the ground is not wet, therefore it did not rain. This one is watertight, and it is the form scientific falsification runs on.
3. Confusing necessary and sufficient
You need a degree for this role. She has a degree. Therefore she qualifies.
A necessary condition is one you cannot do without. A sufficient condition is one that settles the matter. Treating the first as the second is the most common conditional error in professional life.
The degree is a hurdle, not a guarantee — and the two get collapsed constantly in hiring, admissions, and compliance.
4. Undistributed middle
All accountants are numerate. She is numerate. Therefore she is an accountant.
The shared term — numerate — covers a far larger group than accountants. Sharing a property with a category does not place you inside it.
This is affirming the consequent wearing categorical clothing, and syllogism items exploit it heavily.
The Informal Fallacies
Here the structure is sound and the content has been substituted, misrepresented, or smuggled in.
5. Ad hominem
Attacking the person rather than the claim. You would say that, you work for them.
The subtlety worth knowing: an interest is legitimate grounds for scrutinising evidence, and never grounds for concluding the claim is false. "Check this carefully" is reasonable; "therefore wrong" is the fallacy.
6. Straw man
Refuting a weaker version of the position than the one actually held, then declaring the argument won.
It usually appears in the paraphrase — "so what you're saying is…" followed by something nobody said. The test is whether the person holding the view would accept the restatement.
7. False dilemma
Presenting two options as exhaustive when others exist. Either we cut the budget or we go under.
Genuine dilemmas exist, so this one requires judgement rather than pattern-matching. The question is whether the missing options were omitted because they are unavailable or because they are inconvenient.
8. Circular reasoning
Assuming the conclusion among the premises. The policy is fair because it treats everyone equitably.
The argument is technically valid — the conclusion does follow — and completely uninformative, because you have to accept the conclusion to accept the premise. It is the clearest illustration that validity is not enough.
9. Appeal to authority
Treating a claim as established because someone eminent asserted it.
Expert testimony is legitimate evidence when the expert is speaking within their field and the field has consensus. It becomes a fallacy when the eminence is borrowed from an unrelated domain, or when the consensus does not exist.
10. Correlation and causation
Departments using the new tool have higher output. Therefore the tool raises output.
Three alternatives are always available: reverse causation, a common cause driving both, or coincidence in a small sample. High-performing departments may simply adopt new tools sooner.
This is the most consequential fallacy in commercial life, because it is the one that gets acted on with a budget attached.
Which Is Most Common?
In everyday argument, correlation-as-causation and the straw man. In formal testing, affirming the consequent by a wide margin.
The conditional fallacies dominate test performance because they are invisible from the inside. Nothing about affirming the consequent feels like an error — the reasoning runs smoothly and produces a confident answer.
Can Spotting Them Be Taught?
The formal ones, yes and quickly. They are a short list of named patterns, and naming a pattern is most of the work of recognising it.
The informal ones are harder, because identifying a straw man requires knowing what the original position was, and identifying a false dilemma requires knowing what the third option is. That is domain knowledge, not logical training.
The transfer problem
Wason's selection task established that reasoning performance depends heavily on framing — near-universal failure on an abstract conditional, easy success on the same structure dressed as a familiar social rule.
The practical implication is that recognising affirming the consequent in a textbook does not guarantee recognising it in a meeting, where it arrives in unfamiliar clothing and nobody has labelled it.
The Habit Worth Building
One question does most of the work: what exactly is being claimed, and what would have to be true for it to follow?
Applied to a conditional, it catches both formal fallacies. Applied to a causal claim, it surfaces the alternative explanations. Applied to a paraphrase, it exposes the straw man.
A final caution worth stating plainly: an argument containing a fallacy is a bad argument, not a false conclusion. Dismissing a claim because someone defended it badly is itself an error, and it is the one people make immediately after learning this list.