Logical Fallacies Cheat Sheet
16 Common Ones, and When They Are Not Fallacies

A logical fallacy is a flaw in the reasoning behind a claim. It is not the same as the claim being false. That distinction is the most useful thing on this page, so it comes first.

A bad argument can still land on a true conclusion. If someone argues badly that the Earth goes round the Sun, the Earth still goes round the Sun. Spotting a fallacy tells you the argument does not support its conclusion. It tells you nothing about whether the conclusion is true.

Each entry below has a plain definition, an invented everyday example, and — where the sources support it — when the same move is not a fallacy at all. The examples are made up and do not describe anyone real.

Where the list comes from

The study of fallacies is old. The Stanford Encyclopedia of Philosophy puts it simply: “The history of the study of fallacies begins with Aristotle’s work, On Sophistical Refutations.” That short work describes thirteen ways an argument can look like a refutation without being one, split into those that depend on language and those that do not.

At least four entries in this cheat sheet have ancestors on Aristotle’s list. In W. A. Pickard-Cambridge’s English translation they are “ambiguity” (equivocation), “assuming the original conclusion” (begging the question), “stating as cause what is not the cause” (false cause), and “the making of more than one question into one” (the loaded question).

Attacks on the person

1. Ad hominem

What it is. In the Internet Encyclopedia of Philosophy’s definition: “Your reasoning contains this fallacy if you make an irrelevant attack on the person arguing and suggest that this attack undermines the argument itself.”

Example. “Why should we listen to his plan for the library opening hours? He can’t even keep his own desk tidy.”

When it isn’t. The encyclopedia draws a useful line: a person’s characteristics may affect their credibility as an authority without affecting the quality of their argument. Our reading: facts about a person matter when you are relying on their word, not when you are judging their reasoning.

2. Tu quoque (“you too”)

Tu quoque is Latin for “you too”. The Internet Encyclopedia of Philosophy classes it as “a kind of Ad Hominem Circumstantial Fallacy”: dismissing an argument because the person making it does not follow it themselves.

Example. “You say I should stop checking my phone at dinner, but you check yours at breakfast.”

When it isn’t. Hypocrisy is a fair reason to look harder at what someone says. The encyclopedia puts the limit precisely: discovering that a speaker is a hypocrite is a reason to be suspicious, “but it is not a sufficient reason to discount it.”

A related modern word is whataboutism. Merriam-Webster dates whataboutism to 1978. The Irish Times traces the idea to a letter it published on 30 January 1974 describing “the Whatabouts”, and the word Whataboutery to its columnist John Healy three days later.

3. Genetic fallacy

What it is. Judging a claim by where it came from when that origin is beside the point. The encyclopedia’s wording: it “attempts to discredit or support a claim or an argument because of its origin (genesis) when such an appeal to origins is irrelevant.”

Example. “That recipe tip came from a cartoon, so it can’t work.”

When it isn’t. The definition contains its own exception — “when such an appeal to origins is irrelevant.” Our reading: a source’s track record is often relevant, so a claim that comes only from an account known for invented stories deserves more checking than one from a careful reference work.

Distorting the question

4. Straw man

What it is. To “attribute an easily refuted position to your opponent, one that the opponent would not endorse, and then proceed to attack the easily refuted position.”

Example. “She wants the office to recycle more? So she thinks we should sort every staple into its own bin.”

Origin. The image is old. The Stanford Encyclopedia notes that the idea may go back to Isaac Watts’s Logick (1724), which describes disputers setting up weak versions of opponents’ views and defeating them. Merriam-Webster dates the phrase straw man in this sense to 1840, so the term itself is not Watts’s coinage.

5. Red herring

What it is. “A digression that leads the reasoner off the track of considering only relevant information.”

Example. Asked why the group project is late, someone replies with a long account of how hard the software update was on everyone’s laptops.

Origin — DISPUTED. Merriam-Webster traces the phrase to the practice of drawing a red herring across a trail to confuse hunting dogs. The etymologist Michael Quinion, drawing on the Oxford English Dictionary, credited the figurative use to a story William Cobbett published on 14 February 1807. But the Wordorigins.org entry on the phrase cites figurative uses from the 1680s onwards, and concludes that “both the figurative use and the idea that it is grounded in distracted dogs predates” Cobbett. The same entry says the real hunting practice was dragging a red herring along a trail “to train or exercise hounds and horses”, and that stories of using one to distract hounds seem to have been invented. We lean towards the Wordorigins account: it cites dated uses older than Cobbett’s, and both it and Quinion reject the idea that throwing hounds off the scent was a real practice.

6. Loaded question

What it is. “Asking a question in a way that unfairly presumes the answer.”

Example. “Have you stopped leaving the kitchen in a mess?” — either answer admits the mess.

When it isn’t. A question that assumes something both sides already accept is just a question. The problem is presuming something controversial.

7. False dilemma

What it is. A reasoner “who unfairly presents too few choices and then implies that a choice must be made among this short menu of choices.”

Example. “Either we repaint the whole house this weekend or we accept that it will look shabby forever.”

When it isn’t. The key word in the definition is unfairly. Some choices really do have only two options — a light switch is on or off. Our reading: it is a fallacy only when real alternatives are being hidden.

Bad reasoning about evidence

8. Appeal to authority

What it is. Treating a claim as true because of who says it.

When it isn’t — which is usually. This entry needs the biggest caveat of all. The Internet Encyclopedia of Philosophy is blunt: “Most reasoning of this kind is not fallacious, and much of our knowledge properly comes from listening to authorities.” It becomes a fallacy when the “authority” is speaking outside their field, is not trustworthy, is misquoted, or when genuine experts disagree.

Example of the fallacy. “A famous footballer says this phone camera takes the sharpest photos, so it must.”

9. Appeal to popularity (bandwagon)

What it is. Suggesting “too strongly” that a claim is correct simply because many people believe it.

Example. “Everyone in the neighbourhood waters their lawn at noon, so that must be the right time.”

When it isn’t. The encyclopedia notes that what most people believe is, for that reason, somewhat likely to be true, all things considered. Popularity is weak evidence, not no evidence. The fallacy is treating it as proof.

10. Post hoc ergo propter hoc

The Latin means “after this, therefore because of this”: B followed A, so A caused B.

Example. “I wore my blue socks and the bus came on time. Blue socks it is.”

When it isn’t. The encyclopedia is careful here: “Correlations are often good evidence of causal connection, so the fallacy occurs only when the leap to the causal conclusion is done” too hastily. Something happening first is a reason to investigate, not a conclusion.

11. Hasty generalisation

What it is. In the encyclopedia’s words, “the key error is to overestimate the strength of an argument that is based on too small a sample for the implied confidence level or error margin.”

Example. “I’ve been to two cafés in this town and both were slow. The whole town is slow.”

When it isn’t. Generalising is how we learn anything. Our reading: it is fine when your confidence matches the size and fairness of the sample — “both cafés I tried were slow” is honest; “the whole town” is not.

12. Burden of proof

What it is. “Committing the error of trying to get someone else to prove you are wrong, when it is your responsibility to prove you are correct.”

Example. “There’s a hidden room behind the school gym. Prove there isn’t.”

When it isn’t. It is reasonable to set aside, at least provisionally, an unlikely claim that comes with no adequate evidence. Our reading: declining to believe it is not the same as claiming to have disproved it.

13. Slippery slope

What it is. In the Stanford Encyclopedia’s description, arguing that from a starting point “one can by a series of incremental inferences arrive at an undesirable conclusion”, so the first step must be refused.

Example. “If we let the children stay up ten minutes later tonight, next they’ll want to stay up all night, and then they’ll never go to school again.”

When it isn’t. Not every chain of consequences is a fallacy. The Internet Encyclopedia puts the test clearly: “Arguments of this form may or may not be fallacious depending on the probabilities involved in each step.” A chain of steps that are each very likely can be a sound warning. A chain of shaky steps multiplies the doubt at every link.

Tricks with words

14. Begging the question (circular reasoning)

What it is. “A form of circular reasoning in which a conclusion is derived from premises that presuppose the conclusion.” Its Latin name is petitio principii.

Example. “This is the most reliable guidebook, because it says so in the introduction, and a reliable guidebook wouldn’t lie.”

When it isn’t. In everyday English, “that begs the question” is also used to mean “that raises the question”, and Merriam-Webster lists both senses. That everyday use is not a fallacy — it is just a different meaning of the phrase.

15. Equivocation

What it is. “The illegitimate switching of the meaning of a term that occurs twice during the reasoning.”

Example. “A feather is light. What is light cannot be dark. So a feather cannot be dark.”

16. No true Scotsman

What it is. Protecting a generalisation from a counterexample by redefining the group — “no true member would do that.” The Internet Encyclopedia describes it as a kind of ad hoc rescue of a generalisation.

Example. “No real gardener uses weedkiller.” — “My aunt has won prizes for her roses and uses it.” — “Then she isn’t a real gardener.”

Origin. The name comes from a story told by the philosopher Antony Flew. The Fallacy Files lists his discussions of it in several books, including An Introduction to Western Philosophy (1971).

When it isn’t. Our reading: sometimes a person is simply clarifying what they meant all along. The warning sign is a definition that changes only after a counterexample appears.

Two warnings before you use this list

1. The fallacy fallacy. As The Fallacy Files puts it, calling an argument fallacious “says nothing about the truth or falsity of the conclusion.” Winning the argument about the reasoning is not winning the argument about the facts.

2. Most fallacies sit next door to good reasoning. In a 2015 paper in the journal Argumentation, Maarten Boudry, Fabio Paglieri and Massimo Pigliucci argued that “almost every known type of fallacy is a close neighbor to sound inferences or acceptable moves in a debate.” That is why every entry above has a “when it isn’t” line. A list of names is a starting point for thinking, not a replacement for it.

Does learning about manipulation make you better at spotting it?

Disputed It depends on what you measure.

• Recognition in tests: yes, on the evidence so far. A 2022 study in Science Advances by Roozenbeek and colleagues found that short videos explaining manipulation techniques — including false dichotomies and ad hominem attacks — improved people’s ability to recognise them, across six randomised experiments and a field study run on YouTube. The authors also reported better discernment of trustworthy from untrustworthy content and better-quality sharing decisions.
• Behaviour in a realistic feed: not shown. A 2025 study in PNAS Nexus taught people about one technique, emotional manipulation, and measured engagement in a simulated social media feed. The lesson reduced engagement only when the feed held nothing but relevant synthetic posts; adding real posts, or posts using other techniques such as ad hominem attacks, “appeared to nullify the effect.” The authors note there is “substantial evidence that inoculation increases technique recognition when directly assessed” — the open question is whether it transfers.
• Telling reliable from unreliable news: DISPUTED, leaning positive. A 2023 reanalysis of five studies by Modirrousta-Galian and Higham found two inoculation games “did not improve discrimination” between true and fake news, but made people answer “false” more often to everything. A 2026 reanalysis by Simchon and colleagues, covering 33 experiments with 37,025 participants and using a different statistical method, found that “both gamified and video-based interventions consistently improve discrimination between reliable and unreliable news, without increasing response bias.”

Our reading: the larger 2026 reanalysis carries more weight than the five-study critique, because it covers far more experiments — so learning to name manipulation techniques probably does help, at least in study conditions. What has not been shown is that the benefit survives the pace and mix of a real social media feed.

Sources

• “Fallacies”, Stanford Encyclopedia of Philosophy (substantive revision 2024).
• Bradley Dowden, “Fallacies”, Internet Encyclopedia of Philosophy (entries for each fallacy named above).
• Aristotle, On Sophistical Refutations, chapters 4–5, trans. W. A. Pickard-Cambridge.
• Merriam-Webster, entries for “straw man”, “red herring”, “whataboutism”, “tu quoque” and “beg the question”.
• Michael Quinion, “Red herring”, World Wide Words.
• “In a Word… Whataboutism”, The Irish Times.
• The Fallacy Files, “No True Scotsman”.
• The Fallacy Files, “Fallacy fallacy”.
• Maarten Boudry, Fabio Paglieri and Massimo Pigliucci, Argumentation (2015).
• Roozenbeek and colleagues, “Psychological inoculation improves resilience against misinformation on social media”, Science Advances (2022).
• S. Y. N. Wang, S. C. Phillips, K. M. Carley, H. Lin and G. Pennycook, “Limited effectiveness of psychological inoculation against misinformation in a social media feed”, PNAS Nexus (2025).
• A. Modirrousta-Galian and P. A. Higham, “Gamified inoculation interventions do not improve discrimination between true and fake news”, Journal of Experimental Psychology: General (2023).
• Wordorigins.org, “Red Herring”.
• Simchon and colleagues, “A Signal Detection Theory Meta-Analysis of Psychological Inoculation Against Misinformation”, Current Opinion in Psychology 67 (2026).

Checked September 2026.

Related: How to check something online: the SIFT method · Overcoming survivorship bias · First principles thinking