TruthLens tools
Eight reasoning errors that show up constantly in political arguments, product claims and forwarded posts. Read the definitions first, then take the quiz — every answer explains itself.
Copy the post, comment, or message that made you pause. You can then match it against the eight fallacies below — this tool teaches you to recognise them, it does not analyse your text for you.
Nothing you type leaves this browser tab. 2000 character limit.
Rejecting a claim by attacking the person making it rather than the claim itself.
Why it fails: The person's character has no bearing on whether the evidence supports the statement.
Example: “You cannot trust that figure because the analyst who produced it once got a prediction wrong.”
What to say instead: Ask what the data shows, separately from who produced it.
Replacing someone's position with a weaker, exaggerated version, then refuting that.
Why it fails: Defeating a version nobody held proves nothing about the actual position.
Example: “They want safer streets, so they want to abolish public transport.”
What to say instead: Restate the original position accurately before responding to it.
Presenting two options as the only possibilities when more exist.
Why it fails: Real decisions usually have more than two branches, so the framing hides options.
Example: “Either we cut the budget or the service disappears.”
What to say instead: Ask what other approaches have been considered.
Drawing a broad conclusion from too little evidence.
Why it fails: A handful of cases cannot establish a pattern about a large group.
Example: “My two neighbours both lost money to that scheme, so it targets everyone.”
What to say instead: Ask how many cases would be needed to support the claim.
Treating someone's status as proof, especially outside their field of expertise.
Why it fails: A credential supports claims only within the area it covers.
Example: “A famous athlete said this supplement works, so it does.”
What to say instead: Check whether the authority is expert on this specific question.
Claiming one step will inevitably lead to an extreme outcome.
Why it fails: Each step requires a separate decision, so the chain is not automatic.
Example: “If we allow late opening once, the building will be open all night within a year.”
What to say instead: Examine what actually happens at each step rather than the far end.
Using the claim as its own evidence.
Why it fails: The conclusion never gets independent support; it only repeats the premise.
Example: “This policy is effective because it is the policy that works.”
What to say instead: Look for evidence that could have come from outside the argument.
Assuming that because one thing followed another, the first caused it.
Why it fails: Sequence alone is not causation; other factors may explain the change.
Example: “Traffic rose after the new sign went up, so the sign caused the rise.”
What to say instead: Check whether the effect also occurred where the cause was absent.
This does not judge whether a claim is true. A fallacy describes the shape of an argument, not whether its conclusion happens to be correct. Well-reasoned arguments can be wrong, and a fallacy can occasionally arrive at a true statement. It also does not analyse your text automatically.
A fallacy is not a trick or a lie. It is an argument that does not hold together, used by people who are sincere and by people who are not. Recognising the shape lets you ask a better question than “is this true?” — you can ask “does this reasoning work at all?”
The items are deliberately obvious once you have read the definitions. That is the point: the eight patterns are far easier to recognise in an exam question than in a real comment thread, where the argument arrives wrapped in other claims and good-faith disagreement. Read the examples first.
No. Keyword matching produces confident nonsense — an argument can contain any word and still be valid, or be a fallacy without any giveaway phrasing. This teaches you to see the structure yourself.
No, and treating it that way is its own failure. It identifies why the reasoning does not hold; it does not establish what is actually true. You still have to check the facts.
Yes. Bad reasoning can land on a true conclusion, and good reasoning can land on a false one. The quality of the argument and the truth of the claim are separate questions.
Because quoting invented statements and attributing them to real people would misrepresent them. Every example here is written for this tool.
Yes. The definitions, the why-it-fails line, and the worked examples are written to be read aloud or printed, and the quiz works as a self-check.
Plenty do. This is a starting set rather than a complete taxonomy — equivocation, false causes, black-and-white thinking and many others are covered in the linked references.