Appeal to authority

The physicist and the mushrooms

On a hike, a friend picks a wild mushroom. “It’s safe. My neighbor is a physics professor, and he told me this kind is edible.”

The neighbor may be a real expert, but in physics. Nothing in that training bears on identifying mushrooms, so his opinion here is worth about as much as anyone else’s. This is the textbook case: an authority on one subject quoted on another. Celebrity endorsements work the same way.

One expert from a divided field

A beekeeping club is deciding whether to wrap its hives for winter. A member who has read up on it reports that the extension bulletins she found give opposite advice. Then she adds: “But Dr. Alvarez at the state extension office says to wrap them, so that’s the expert answer.”

This one looks respectable. Dr. Alvarez is real, named and working in the right field. But the member has just said the experts disagree, so quoting one of them doesn’t settle anything. When authorities are divided, picking one and calling that “the expert answer” hides the disagreement instead of resolving it. The club still needs to weigh the reasons each side gives.

“Are you questioning my expertise?”

A home inspector says a damp patch in a basement is harmless condensation. The homeowner asks why the patch grows after every rainstorm. The inspector replies: “I’ve been doing this for twenty-five years. Are you questioning my expertise?”

The inspector is qualified and speaking within his field, which is what makes this case subtle. The problem is what the appeal is used for. The homeowner asked a fair question about the evidence, and the credentials are offered in place of an answer. An expert’s opinion is a strong reason to believe something, but it doesn’t make the question about the evidence go away. Douglas Walton and Marcin Koszowy describe this as a distinct form of the fallacy: pressing an expert opinion so forcefully that further questioning is blocked.

Douglas Walton’s standard analysis treats the legitimate version as plausible reasoning: a reason to accept a claim for now, which has to be given up if certain questions can’t be answered satisfactorily.

The legitimate form has a premise the fallacy leaves out (the claim belongs to the expert’s field), and a weaker conclusion: plausible, not proven. Walton pairs it with six critical questions:

1. Expertise: how credible is E as an expert?
2. Field: is E an expert in the field that A is in?
3. Opinion: what did E actually say that implies A?
4. Trustworthiness: is E personally reliable as a source (for example, free of bias)?
5. Consistency: is A consistent with what other experts say?
6. Backup evidence: is E’s opinion based on evidence?

If a question can’t be answered satisfactorily, the argument loses its force.

• Irrelevant authority: an expert or celebrity quoted outside their field. “This famous chef says the vitamin supplement boosts memory.”
• Unnamed authority: “experts agree”, “scientists say”, “studies show”, with no way to find out who, or what exactly they said. “Botanists agree you should water succulents with ice cubes.”
• Misquoted or invented authority: a real expert’s words stretched, taken out of context, or made up. This fails Walton’s opinion question.
• One side of a divided field: citing one authority as if their view were the consensus.
• Authority as the last word: using credentials to shut down questions instead of answering them.

The Latin name argumentum ad verecundiam goes back to John Locke’s Essay Concerning Human Understanding (1690), and verecundia means modesty or shame. Hans Hansen’s encyclopedia entry notes that Locke’s language suggests he meant deference to people’s social standing, not their expertise: accepting a claim because contradicting an eminent person would seem immodest. The modern name has shifted toward expertise, but the pressure Locke described survives in the “are you questioning my expertise?” move.

The mirror image of ad hominem. An appeal to authority and an ad hominem are the same kind of argument pointed in opposite directions. Argumentation theorists call arguments from a person’s character or credibility ethotic arguments (from the Greek ethos, character). In the account by Walton, Reed and Macagno, summarized by Katarzyna Budzynska and colleagues, the positive form reasons from someone’s good character or credibility to accepting what they say (arguments from expert opinion and witness testimony are examples), and the negative form reasons from bad character to rejecting it (ad hominem). Walton models both as raising or lowering a speaker’s credibility. P. T. Mackenzie traced the fallacious versions to the same impulse: people want to share the beliefs of those they admire and distance themselves from the beliefs of those they dislike.

Its wider family. Both judge a claim by where it came from, which is what the genetic fallacy names on its broader definitions, the ones that count judging a source favorably as well as unfavorably. An appeal to authority is the favorable direction, an ad hominem the unfavorable one. The appeal to popularity is a close relative, where the source is the crowd instead of an expert; in snob appeal it’s a small, admired group.

Deferring to experts is how non-specialists reasonably form beliefs about almost everything they can’t check. An appeal to authority is good reasoning when:

• The source really is an expert, with training or a track record in the subject.
• The claim is in their field, not an adjacent or unrelated one.
• You know who the source is and what they actually said, so it can be checked.
• There’s no strong reason to doubt their honesty, such as a stake in the answer.
• Other experts in the field broadly agree, or at least the disagreement is acknowledged.
• Their view rests on evidence, even if you can’t evaluate it yourself.
• It’s held as a strong presumption, not a final verdict: open to revision if better evidence or a better-qualified disagreement turns up.

The test: would this person’s say-so be good evidence on this exact question, and is it being offered as evidence, or as a reason to stop asking?

The engineer and the deck

A licensed structural engineer inspects a backyard deck. She shows the owner two joists where a screwdriver sinks into soft wood and says the deck shouldn’t hold a crowd. The owner moves the party inside without checking the other joists himself.

The owner accepts a conclusion he can’t evaluate, purely on an expert’s word. But she is qualified in exactly this question, examined the deck directly, and pointed to evidence. Nothing suggests she has a stake in the answer. Deferring to her is the reasonable thing to do. This meets the expert, in their field, with evidence conditions above.

The distance to the Sun

Asked how she knows the Sun is about 93 million miles away, a student says: “That’s what astronomers have measured. I couldn’t do the measurements myself.”

The student is relying entirely on authority, and that’s the right call. The authority isn’t one person but a field, its members agree, and their figure rests on measurements that others in the field can repeat. The student can’t check them, and doesn’t need to in order to hold the belief reasonably. This is the other experts broadly agree condition above.

Trusting experts usually works. Nobody has time to verify medicine, engineering and history for themselves, so deferring to people who know more is a rational habit. The fallacy is that habit applied without checking whether its conditions hold.

The signs of authority are easy to see; the fit between the expertise and the question isn’t. A title, a white coat, confidence or fame is visible at a glance, while knowing whether a physicist is qualified to judge mushrooms takes a moment’s thought. Competence in one area also tends to spill over into judgments of competence in others, the pattern described by the Halo effect.

There’s a social pull too. Questioning an expert, especially in person, can feel presumptuous. That is close to what Locke originally described: accepting a claim out of modesty rather than because of the reasons for it.

Walton’s critical questions work well as a checklist:

• Who, exactly? A named source can be checked; “experts say” can’t.
• Expert in what? Is the claim in their field?
• What did they actually say? Find the original wording if you can.
• Do others in the field agree? One expert against a consensus is weak support.
• What’s it based on? Ask for the reasons the expert would give. A good expert can usually explain them.
• Do they have a stake in the answer? If so, look for independent confirmation, without assuming they’re wrong (that would be the mirror-image mistake).