A logical fallacy is an argument that looks stronger than it is: the reasons offered do not carry the conclusion, but the gap is easy to miss. This list of logical fallacies gives the fifteen you will meet most in essays, debates and seminars, each defined from a named source, with an example of our own and a way to answer it.
EdCitation publishes this guide. We read every definition on 27 September 2026 in the Stanford Encyclopedia of Philosophy (SEP) entry by Hansen (2024), the Internet Encyclopedia of Philosophy (IEP) entry by Dowden (n.d.), the Writing Center at the University of North Carolina at Chapel Hill (n.d.) and Excelsior Online Writing Lab (n.d.-a). Where they disagree, we say so and say which we follow.
Four of these fallacies lean on sources, and the answer to them is to check the source. EdCitation's free Verify references confirms that a cited work exists, matches its record and has not been retracted. It writes no part of your essay or speech: the argument, and every answer to a fallacy, is yours.
What is a logical fallacy?
A logical fallacy is a defect in reasoning that makes an argument seem better supported than it is. Hansen (2024) takes as a working definition "an argument that seems to be better than it really is", and notes that the exact definition of each fallacy is often contested. Dowden (n.d.) calls a fallacy a kind of error in reasoning and lists 231 names. The North Carolina handout treats arguments as stronger or weaker on a scale, and aims to help you look critically at your own reasoning rather than to teach labels.
We follow that last approach. Naming a fallacy wins nothing on its own; Dowden (n.d.) puts the burden of proof on the person who alleges one. What answers a fallacy is showing the missing step.
Formal and informal fallacies
A formal fallacy is wrong in its structure; an informal fallacy is wrong in its content. Hansen (2024) describes formal fallacies as instances of recognisably invalid forms, such as denying the antecedent, and says informal fallacies are better analysed without symbolic logic. An example of ours: "If it rains, the match is off. It did not rain. So the match is on." The match could be off for another reason, whatever the weather.
The names differ too. Dowden (n.d.) treats "logical fallacy" as another name for a formal fallacy, while Excelsior's writing lab uses it for errors in reasoning of every kind. Everyday usage follows Excelsior, and so does this guide: all fifteen below are informal.
List of logical fallacies: 15 common ones, with examples
Each example is ours, and invented to show the fault. The source column names where we read the definition.
| Fallacy | What it is | Example (ours) | How to answer | Source |
|---|---|---|---|---|
| Ad hominem | Attacking the person instead of the argument; tu quoque ("you too") is the version that points at the speaker's own conduct | "Ignore the student rep's case for longer library hours: they failed a module last year." | Grant or set aside the point about the person, then return to the reasons for longer hours | SEP, IEP |
| Straw man | Distorting a view to make it easier to knock down | "You want a vegetarian option every day? So you want to ban meat." | Restate what was actually said and ask the other side to confirm it | SEP, Excelsior |
| Red herring | A side issue that pulls attention from the question | "Is the attendance rule fair? Lecturers work very long hours, you know." | Name the question again and ask how the point bears on it | IEP, UNC |
| Slippery slope | Claiming one step will lead, link by link, to disaster | "Allow one late essay and soon nobody will meet a deadline." | Ask for evidence for each link in the chain, not just the first | IEP, SEP |
| Appeal to authority | Treating someone's word as proof when they are not an expert on this, or are misquoted | "A famous footballer says this supplement boosts memory, so it will raise my marks." | Ask whether the source is expert in this field and says what is claimed; check it with Verify references | IEP, SEP |
| Bandwagon (appeal to the people) | Treating popularity as proof | "Everyone in my year uses this revision app, so it must work." | Ask what the evidence is that it works, not how many use it | IEP, UNC |
| False dilemma | Offering two options when there are more | "Either we cancel the trip or we put students at risk." | Name a third option | IEP, UNC |
| Hasty generalisation | A conclusion about a whole group from too few cases | "Two members of the debating society were rude to me, so the society is unfriendly." | Ask how many cases were seen and narrow the claim to fit them | IEP, UNC |
| Cherry-picking (suppressed evidence) | Leaving out the evidence that counts against you | An essay cites the one survey that found homework useless and omits the others it read | Ask what the rest of the evidence says; search it with Find sources | IEP |
| Post hoc | Assuming that because B followed A, A caused B | "I wore new trainers and won the race, so the trainers made me faster." | Ask for the mechanism, and whether anything else changed | IEP, UNC, SEP |
| Correlation as cause (cum hoc) | Assuming that two things found together are cause and effect | "Students at the front get higher marks, so move to the front to raise yours." | Ask whether a third factor, such as motivation, explains both | IEP, SEP |
| Begging the question | A premise that already assumes the conclusion | "The school rules are fair because they were written to be fair." | Ask for a reason that does not restate the claim | SEP, UNC |
| Appeal to ignorance | Treating a lack of proof as proof of the opposite | "No one has shown the new timetable harms learning, so it does not." | Ask whether anyone has looked, and how thoroughly | IEP, UNC |
| Appeal to emotion | Offering a feeling in place of evidence | "Think how let down our teachers will feel if this motion fails." | Acknowledge the feeling, then ask for the reason | IEP, SEP |
| Equivocation | One word used in two senses within an argument | "The result was significant, so it matters to everyone." | Point to the two meanings (here statistical and everyday) and hold one fixed | SEP, UNC |
Which fallacies are not always fallacies?
Several of the best-known fallacies name a pattern that is sometimes perfectly sound. Hansen (2024) reviews scholars who doubt that textbook fallacies are common in real argument at all. Judge the particular argument, not its shape.
Appeal to authority
Most appeals to authority are not fallacious, Dowden (n.d.) says; much of what we know properly comes from experts. The IEP counts an appeal as a fallacy when the person is no authority on this subject, cannot be trusted, is misquoted, or when authorities disagree. Hansen (2024) gives a similar account, and notes that Locke's name for it, ad verecundiam, is better translated as "appeal to modesty". We follow the IEP, because it names misquotation, which is the version a reference check can catch. Citing an expert is how academic writing works; the fault is citing one who does not say what you claim, or does not exist.
Slippery slope
The slippery slope is where the definitions differ most. Excelsior Online Writing Lab (n.d.-b) makes it a fallacy when the chain of consequences is taken as certain. Hansen (2024) places the weakness in the causal links themselves. Dowden (n.d.) says the same form appears in good and bad arguments, and the verdict depends on how likely each step is: three steps each 80% likely leave only about a 50% chance of reaching the end. The North Carolina handout gives a chain it does not count as fallacious, from failing a required course to not graduating. We follow the IEP: ask for the probability of each link.
Ad hominem, post hoc and the rest
An attack on character is relevant when the argument is about character. The IEP's case is an argument for someone's fitness to lead, and North Carolina makes the same exception. Hansen (2024) reports that recent scholarship finds the modern ad hominem arguments sometimes fair and sometimes fallacious, and that the argument Locke first called ad hominem, holding someone to their own principles, is no fallacy at all. On cause, Dowden (n.d.) notes that correlations are often good evidence of causation; the fallacy is the hasty leap. Hansen (2024) also calls the appeal to force one of the most controversial entries on the list.
How do you answer a fallacy without committing one?
Answer the argument, not the label: saying what is missing moves a discussion on, where shouting "straw man" starts a quarrel.
- Restate the claim fairly. Excelsior Online Writing Lab (n.d.-c) advises clarifying questions and restating the terms of the argument when facing a straw man, and the habit helps with every fallacy.
- Find the missing step. Which link does not hold: the expert, the sample, the cause, the second option?
- Say it in plain words. "Two cases do not tell us about the whole society" persuades more than "hasty generalisation".
- Ask for the evidence that would close the gap. If it exists, the argument improves, and so does the discussion.
- Check any source the argument stands on. Paste the reference into Verify references, or look the work up in Find sources.
- Concede what is right. An argument with a fallacy in it can still have a true conclusion; the fallacy only shows that these reasons do not prove it.
Answering a whole opposing case is a larger task; the argumentative essay guide covers counterargument and rebuttal. In your own writing, the cure for most fallacies is a smaller claim, as hedging and cautious language shows.
How does checking the source answer appeal to authority, cherry-picking and false cause?
Four of the fifteen (appeal to authority, cherry-picking, post hoc and correlation as cause) rest on evidence, and each is answered by going to the evidence itself.
A fake or misquoted authority
An appeal to authority is only as good as its source. If the work does not exist, says something else, or has been retracted, the authority is gone. The first check is whether the reference is real and correct: Verify references looks each entry up in the publisher's record, and our guide on how to check a reference is real walks through the rest. The second is whether the work says what is claimed, and only reading it can settle that; citing a source you did not read explains why.
Cherry-picked evidence
Cherry-picking is answered by the evidence that was left out. Search the claim itself, then its opposite, and see what the rest of the literature says. EdCitation's free Find sources searches about 300 million published works by topic or by the claim a sentence needs to support, with filters for the kind of source and the years. A search returns works, not verdicts: which way each study points is for you to read.
A correlation sold as a cause
A correlation is answered by the study's own method. The North Carolina handout puts it simply: correlation is not the same thing as causation. Before accepting a causal claim, read how the study was designed: were groups assigned at random, or observed after they had chosen? Were other factors measured? The methods section usually settles it.
What happened when we ran Verify references on a list with mistakes in it?
We pasted three references into Verify references on 27 September 2026. The works are real. The first entry was correct; the mistakes in the other two are ours, made on purpose: we moved one paper's year from 2012 to 2014, and changed a word in another's title. The correct entry came back verified; the other two came back doubtful, which the page shows as "check this". None was reported not found and none was retracted.
| What we pasted | What came back, word for word |
|---|---|
| Tversky and Kahneman, Science, 1974, with its DOI (correct) | Verified: The DOI resolves to this record and the title matches. |
| Messerli, The New England Journal of Medicine, given as 2014, no DOI (our wrong year) | Doubtful: A record with this title exists, but it was published in 2012, not 2014. |
| Ioannidis, PLoS Medicine, 2005, title given as "…are wrong" (our wrong title) | Doubtful: The DOI is right, but its record's title is "Why Most Published Research Findings Are False". Copy the title from the record. |
Both of our mistakes were caught, each with the correction beside it, and the correct entry passed. The third result also listed a correction notice from 2022 on that paper's record; a correction is not a retraction, but it is worth reading before you rely on a paper.
How EdCitation helps you argue from sources
For an argument that leans on sources, EdCitation is the best check we know of, because it looks every reference up in the publisher's record and never writes one; no chatbot can do that, and a chatbot asked for support can produce a reference that does not exist, as our guide to why AI tools invent references shows. In the run above it caught a wrong year and a wrong title, and named the right ones.
- Verify references, free with no account: paste a reference list or upload a paper, and each entry comes back verified, doubtful or not found, with retracted papers flagged. A reference it could not check is never shown as not found.
- Cite a source, free: builds a reference from a DOI, title, PubMed ID, ISBN or web address in six styles, and screens the source for retraction.
- Find sources, free: the evidence on both sides of a claim.
- References from a file, part of Pro at $8 a month: every reference in an uploaded paper checked, and every in-text citation matched to the list (pricing).
EdCitation never writes any part of an essay or a speech, and we are glad of it: an argument you built yourself is one you can defend when someone spots a weak link.
Quick questions
What are the most common logical fallacies?
Ad hominem, straw man, red herring, slippery slope, appeal to authority, bandwagon and false dilemma lead most lists; the table above gives all fifteen, with an example of each.
What is the difference between a straw man and an ad hominem?
A straw man distorts the other person's argument and attacks the distorted version; an ad hominem attacks the person instead of the argument. Both avoid the real reasons.
Is appeal to authority always a fallacy?
No. The Internet Encyclopedia of Philosophy says most appeals to authority are not fallacious; one becomes a fallacy when the person is not expert in the subject, is misquoted, cannot be trusted, or when experts disagree.
Is a slippery slope argument always a fallacy?
No. A chain of consequences can be sound if each step is likely. It is a fallacy when the likelihood of reaching the bad end is exaggerated.
Can EdCitation check that an authority I cite is real?
Yes. EdCitation's free Verify references checks that each cited work exists, that the entry matches the publisher's record, and whether it has been retracted. Whether the work says what you claim still needs you to read it.
References
- Dowden, B. (n.d.). Fallacies. Internet Encyclopedia of Philosophy. https://iep.utm.edu/fallacy/
- Excelsior Online Writing Lab. (n.d.-a). Logical fallacies. https://owl.excelsior.edu/argument-and-critical-thinking/logical-fallacies/
- Excelsior Online Writing Lab. (n.d.-b). Slippery slope fallacy. https://owl.excelsior.edu/argument-and-critical-thinking/logical-fallacies/logical-fallacies-slippery-slope/
- Excelsior Online Writing Lab. (n.d.-c). Straw man fallacy. https://owl.excelsior.edu/argument-and-critical-thinking/logical-fallacies/logical-fallacies-straw-man/
- Hansen, H. (2024). Fallacies. In E. N. Zalta & U. Nodelman (Eds.), The Stanford Encyclopedia of Philosophy (Fall 2024 ed.). https://plato.stanford.edu/archives/fall2024/entries/fallacies/
- The Writing Center, University of North Carolina at Chapel Hill. (n.d.). Fallacies. https://writingcenter.unc.edu/tips-and-tools/fallacies/