Ad hominem, tu quoque, and the genetic fallacy all do the same basic thing: they go after the source of an argument instead of the argument itself. That shared move is why people mix them up. They are cousins, not synonyms, and the differences are sharp once you see them.
The distinction is worth getting right. Cite the wrong one in a thread and you hand the other side a clean rebuttal: "that wasn't an attack on me, that was a question about my consistency." A precise label sticks. A sloppy one bounces. This guide gives you a one-line distinction, a side-by-side table, a worked example showing all three from the same opening sentence, and a decision tree you can run in your head.
The one-line distinction
Each fallacy targets a different part of the source.
| Fallacy | What it attacks |
|---|---|
| Ad hominem | The arguer's character, qualifications, or person |
| Tu quoque | The arguer's own conduct ("you do this too") |
| Genetic fallacy | The origin of the claim (institution, study, funding) |
Read the table as a narrowing of focus. Ad hominem is about who you are. Tu quoque is about what you have done. The genetic fallacy is about where the claim came from, which need not be a person at all.
Same scenario, three different fallacies
Start with one argument and hit it three ways. The opening line stays fixed:
"We should increase funding for public libraries because they reduce illiteracy in low-income neighborhoods."
Ad hominem response:
"You only think that because you're a former librarian. Anyone who actually understands budgeting wouldn't agree."
This attacks the speaker's background and competence. Whether the speaker once worked in a library has no bearing on whether libraries reduce illiteracy. The illiteracy data stands or falls on its own.
Tu quoque response:
"You voted against the parks bill last year. Suddenly you care about public funding?"
This points at the speaker's past behavior and implies hypocrisy. Even if the speaker is wildly inconsistent, that does nothing to the library claim. A hypocrite can still be right about libraries.
Genetic fallacy response:
"That study came out of a library industry trade group. Of course it says libraries reduce illiteracy."
This rejects the claim because of where it originated. The funding source is a fair reason to check the study's methods, but it is not a reason to dismiss the finding outright. Biased sources sometimes produce accurate results.
Three responses, one untouched argument. None of them engaged with whether the funding would actually reduce illiteracy.
Why the distinctions matter
The labels are not interchangeable, and treating them as such weakens your citation.
If you call something ad hominem when it was tu quoque, the other side has a fair counter. They can say, correctly, that they never attacked your character, they asked about your consistency. Now you are defending a bad label instead of pressing a good point. The conversation drifts to the meta-argument and your original objection gets lost.
There is a second reason precision matters. Each of these moves has a legitimate twin, and the line between fallacy and fair play sits in a different place for each one:
- A consistency challenge is not tu quoque. Asking someone to apply their own stated standard evenly is fair. It only becomes a fallacy when "you do it too" is offered as proof the claim is false.
- A credibility challenge is not ad hominem. When a person's honesty or expertise is directly the issue, like a witness on the stand or an expert testifying, their reliability is on the table. It becomes ad hominem when the attack replaces the argument rather than addressing a relevant qualification.
- A source check is not the genetic fallacy. Noting that a study was funded by an interested party is good practice. It tips into fallacy only when origin alone is treated as a refutation.
Name the move precisely and you keep the burden where it belongs. Name it loosely and you invite a tangent.
Quick diagnostic
Run the response through four questions, in order. Stop at the first one that fits.
- Did the response say something about the person making the argument, or about where the claim came from? If no, you are not looking at any of these three. Check whataboutism or red herring instead.
- Did it point at the speaker's own past behavior or apparent hypocrisy? If yes, it is tu quoque.
- Did it point at who made or funded the claim, the institution or study rather than the person? If yes, it is the genetic fallacy.
- Otherwise, an attack on the person's character or competence is ad hominem.
The order matters. Tu quoque and the genetic fallacy are specific kinds of source attack, so check for them first. Ad hominem is the catch-all for personal attacks that are not about past conduct.
When each is not a fallacy
Every entry on this site carries a "Counter-example (not a fallacy)" section because the line is where the real work happens. Pulled together, the three lines look like this:
- Ad hominem is fair when the person's reliability is the actual subject. "He has lied under oath three times" is a relevant objection to a witness, not an attack that dodges the argument.
- Tu quoque is fair as a consistency check. "You demanded sources from me, so where are yours?" presses a standard evenly. It only fails when used to prove the original claim wrong.
- The genetic fallacy is fair as grounds for scrutiny. A tobacco-funded study on smoking deserves a harder look. The funding is a reason to verify, not a reason to dismiss unread.
In each case, the difference is the same: are you using the source to evaluate the claim, or to avoid evaluating it? The first is reasoning. The second is the fallacy.
Cite the one that fits
Name the move precisely or do not name it. A sloppy citation makes you look less reliable than no citation at all, and it gives the other side a free rebuttal. When the response is about who someone is, reach for fallacy.is/ad-hominem. When it is about what they have done, fallacy.is/tu-quoque. When it is about where the claim came from, fallacy.is/genetic-fallacy. For more on citing without turning a thread into a war, see using fallacy.is in a discussion. The right label does the arguing for you.