Guilt by association

The bike rack

At a neighborhood association meeting, a member proposes installing a bike rack outside the community center. Another replies: “The people pushing bike projects are the same crowd who tried to close Main Street to cars last year. Let’s not go down that road.”

Whether a bike rack is a good idea depends on things like how many people cycle there and whether it blocks the entrance. The objection answers none of that. It links the proposal to an unpopular group and lets the dislike do the arguing, even though the rack and the street closure are separate proposals that can be judged separately.

“You know who else campaigned against smoking?”

A town council debates banning smoking near a children’s playground. A resident stands up: “The Nazis ran some of the first big anti-smoking campaigns. Is that really the company we want to keep?”

This is the form the philosopher Leo Strauss named the reductio ad Hitlerum. The history is real: the historian Robert Proctor describes Nazi Germany as having the world’s strongest antismoking movement in the 1930s and early 1940s. But the question before the council is whether smoke near a playground harms children, and that turns on evidence about smoke, not on who else once opposed it. As Strauss put it, “A view is not refuted by the fact that it happens to have been shared by Hitler.” Any policy has been supported, at some time, by someone terrible.

The conference program

A reviewer recommends rejecting a paper on songbird migration, noting that it was first presented at a conference where another session featured a researcher widely regarded as a crank.

This one sounds like professional caution, which is what makes it persuasive. But appearing on the same program links the paper to the crank only by a schedule. The paper’s data, methods and conclusions are in front of the reviewer and can be judged directly. Where a paper was presented can be a reason to read it carefully; it isn’t a finding about the paper.

• Guilt by a person’s associates: discrediting someone because of the company they keep. Bradley Dowden’s Internet Encyclopedia of Philosophy entry defines the fallacy in these terms, as the version of ad hominem “in which a person is said to be guilty of error because of the group he or she associates with.”
• Guilt by a shared view: discrediting an idea because a disreputable person or group also held it. The reductio ad Hitlerum is the best-known form. “Hitler loved dogs too.”
• Honor by association: the positive direction. “The founders of three successful companies swear by this morning routine, so it must work.” The link to admired people doesn’t show the routine had anything to do with their success.

Where the name “reductio ad Hitlerum” comes from. Strauss coined it in Natural Right and History (1953). He was arguing that the sociologist Max Weber’s view of values led to nihilism, and noted that following Weber’s thought to its end would reach “a point beyond which the scene is darkened by the shadow of Hitler.” He then warned against “the fallacy that in the last decades has frequently been used as a substitute for the reductio ad absurdum: the reductio ad Hitlerum.” The name plays on the reductio ad absurdum, which refutes a view by showing that it leads to something absurd; the Hitler version swaps that demonstration for a mere association.

Its place in the family. Dowden classifies guilt by association as a version of ad hominem, and notes that it is also called “circumstantial” ad hominem. When the target is an idea rather than a person, as in Strauss’s version, the move judges a claim by who else held it, which is the flaw shared by the whole Genetic fallacy family. The philosopher John McMurtry describes how the “smear” works: “by means of an alleged relationship between the position it seeks to discredit and an already discredited position.” He distinguishes it from a related diversion he calls the argumentum ad adversarium, which changes the subject to a common enemy without claiming any link between that enemy and the position under attack. It also differs from poisoning the well, which discredits a person in advance by any means; guilt by association discredits through a connection, whenever it’s raised.

A connection is evidence when it tells you something about the specific point at issue.

• The group is defined by the thing in question. Someone who joined an organization whose stated purpose is to promote a view probably holds that view. Membership is evidence of what the group exists to do, not of everything its other members have done.
• The association raises the likelihood of something specific, and it’s used as a reason to check. A payment sent to an account known to receive stolen money deserves a closer look. The link sets how hard to look; the looking settles it.
• The question is about the association itself. Who funds an organization, or whom an official appointed, is answered by the connections, because the connections are the subject.
• Choosing an associate reflects on judgment. Someone who repeatedly hires people with records of fraud may fairly be questioned about their hiring judgment. That still isn’t evidence that any particular hire was dishonest.

The test: does the connection give evidence about this particular claim or person on this particular point, or does it only pass along a bad feeling?

The flagged payment

A small business’s bank holds an outgoing payment for review. The receiving account has, in the past month, received money from four accounts already confirmed as stolen. The bank asks the business owner to confirm she knows the recipient before releasing the funds.

The bank is judging a transaction by what it’s connected to, which is the shape of guilt by association. But the connection bears directly on the specific risk (this account is probably collecting stolen money), and the bank treats it as a reason to check rather than a verdict: the payment goes through once the owner confirms it. That’s the reason to check condition.

The society’s speaker

A park committee is choosing a speaker for a talk on native plants. One candidate is president of a gardening society whose stated mission is to promote exotic ornamental species. A member says: “We should expect him to argue against replacing ornamentals with native plants. That’s what his society exists to do.”

This predicts a person’s view from the group he belongs to. It’s reasonable here because the group is defined by that view, and he chose to lead it. The member isn’t saying the speaker is wrong, only predicting which side he’ll take. That’s the group defined by the thing in question condition.

Feelings about a person or group spread easily to whatever is near them. The same spillover that lets one good trait color judgments of a person’s other traits, the Halo effect, lets a disliked group make everything near it seem suspect. Once a proposal “sounds like something those people would say,” it feels answered.

It also borrows the force of a real argument. A reductio ad absurdum shows that a view leads somewhere unacceptable, and a comparison to a notorious figure feels like the same demonstration. Strauss’s point was that it isn’t: it names the destination without showing the route. McMurtry adds that smear tactics work best when they invoke an adversary the audience already opposes, because then the audience’s existing hostility does the work. The more universally condemned the associate, the less anyone checks whether the link is real.

• Say the hidden premise out loud. “So anything that group supported is wrong?” usually shows how much the argument depends on it.
• Ask what the connection actually is. Shared membership, a shared view on one question, a single meeting, or a coincidence of timing are very different links.
• Ask whether it bears on this point. If the worry is real (a policy might lead somewhere dangerous), ask how it would get there. That turns an association into a claim that can be examined.
• Watch for the positive version too. A link to admired people is no more evidence than a link to disliked ones.