Attacking the person

Attacking competence instead of the work

Sam: The new bridge design exceeds its load rating. Here are my calculations.

Riley: You failed your engineering exams twice. Why would anyone listen to you?

Riley hasn’t said anything about the calculations. If they’re correct, they’re correct no matter who did them; if they’re wrong, the way to show it is to find the mistake. Sam’s exam history might be a reason to double-check, but it isn’t a reason to throw the calculations out unread.

Pointing at motives

A researcher employed by a drug company publishes the full trial data, methods and statistical analysis for a new medication. A critic responds: “She works for the manufacturer. Obviously the drug doesn’t work.”

The data and methods are published and can be checked by anyone. The researcher’s employer is a good reason to check them carefully, but it isn’t evidence that the drug doesn’t work. The critic has skipped the checking and gone straight to the conclusion. This is the circumstantial form, and it’s the one that sounds most reasonable.

“You do it too”

A parent tells their teenager that smoking causes lung disease. The teenager replies: “You smoked for twenty years, so you can’t tell me what to do.”

The parent’s history doesn’t change the medical evidence about smoking. It shows only that the parent didn’t follow their own advice. This is tu quoque (“you too”): hypocrisy is treated as if it refuted the claim.

• Abusive: attacking character or intelligence directly. “You’re an idiot, so you’re wrong.”
• Circumstantial: claiming the person only believes it because of their interests or situation.
• Tu quoque: rejecting a claim because the person making it doesn’t live by it.
• Poisoning the well: discrediting someone in advance, so that anything they say afterward is dismissed before it’s heard.

Douglas Walton’s standard treatment also separates out accusations of bias as their own subtype.

Who someone is matters whenever you’re relying on them rather than on their reasons.

• Testimony. If you can’t check a claim yourself and would believe it because a person says so (a witness, an expert, a reviewer), then their honesty, competence and incentives are directly relevant.
• Conflicts of interest in unverifiable claims. When you can’t inspect the evidence, a source’s stake in the answer is a fair reason to lower your confidence and look for independent confirmation.
• When the question is about the person. If you’re deciding whether to trust someone with a job, a loan or public office, their character is the subject, not a distraction from it.
• Pointing out inconsistency. Showing that someone’s argument conflicts with other things they’ve said or committed to is fair. (This was the older, original meaning of argumentum ad hominem.) It becomes a fallacy only if you conclude the claim is false, rather than that their position is inconsistent.
• An insult alongside a real rebuttal. Rudeness isn’t a fallacy. It’s only an ad hominem when the attack does the work of the argument.

The test: does the point about the person bear on the reasons you’re being asked to accept, or replace them?

Challenging a witness

In a trial, the only evidence that the defendant was at the scene is one eyewitness. The defense shows the witness has two convictions for perjury and was paid by the other side. “You shouldn’t convict on this testimony.”

This attacks the person, but the argument is the person: the claim is believed only because the witness says so. With no independent evidence to evaluate, the witness’s reliability is exactly the right question. This is the testimony condition above.

An insult with the argument attached

“Your budget is wrong again. You counted the rent twice, once in row 4 and again in row 12.”

The first sentence is a jab, but the argument is in the second. Delete the jab and the rebuttal is unchanged: the rent is double-counted. That’s rude, not fallacious.

Evaluating an argument takes effort. Evaluating a person is quick and feels relevant, because we constantly and reasonably judge information by its source. That makes it easy to apply the habit where it doesn’t belong: to arguments that stand or fall on their own. Once someone seems untrustworthy, their arguments tend to seem weaker too, the same spillover described by the Halo effect.