Appeal to consequences

The pier can’t be cracked

An inspector reports hairline cracks in two support piles under the town pier. A festival organizer responds: “That can’t be right. If the pier has to close, the summer festival is ruined, and half the shops on the waterfront depend on that weekend.”

Everything the organizer says about the festival may be true, and it’s a good reason to hope the report is wrong. But whether the piles are cracked depends on the piles, and the only way to dispute the report is to look at them again. A costly consequence is a reason to check carefully, not evidence of anything about the concrete.

It would save so much money

Priya: The new supplier’s parts are going to pass the stress tests. If they do, we save forty thousand dollars a year, and we can finally hire the extra technician.

This is the hopeful direction, sometimes called wishful thinking. The savings and the new hire are real benefits, and they explain why Priya wants the parts to pass. They say nothing about how the parts will do under load. If anything, a large reward for one answer is a reason to be more careful in testing, because it’s the answer everyone is hoping to see.

“We shouldn’t accept that”

A local historian finds letters showing that the famous rescue by the town’s lighthouse keeper in 1893 involved one boat, not the five in the legend. At a meeting, a council member says: “We shouldn’t accept this. The rescue is on the town seal, the school is named after him, and the tourist trade is built on that story.”

This one sounds like a practical point, and part of it is. The consequences are real, and they bear on practical questions: whether to change the seal, how to tell the story at the museum, when to announce the finding. But “we shouldn’t accept this” slides between two meanings. If it means “we shouldn’t rename the school yet”, the consequences are relevant. If it means “we shouldn’t believe the letters”, they aren’t, because the tourist trade can’t change what happened in 1893. The philosopher Charles Kielkopf pointed to this ambiguity: a “should” conclusion can be read as a claim about what is the case or as a call to act, and consequences are relevant only to the second.

The right-hand column is the form Douglas Walton calls argument from negative consequences. It concludes that something should be done, and it is often reasonable. Walton’s example: “You ought not to take this drug, because it is known to have the following side effects, one of which in particular is very dangerous for you.” The left-hand column has the same premises but a conclusion about what is true, and there the premises don’t support it.

A third shape looks like the fallacy but isn’t: if P were true, we would observe O; we don’t observe O; so P is probably false. Here O is a prediction, not a cost, and the argument is an instance of Modus tollens. The question is never whether O would be unwelcome, only whether it would be observed.

• Negative: rejecting a claim because its truth would be bad. “The soil test can’t be contaminated; we’d have to move the whole garden.”
• Positive (wishful thinking): accepting a claim because its truth would be good. The Internet Encyclopedia of Philosophy lists wishful thinking as its own fallacy: suggesting a claim is true or false merely because one strongly hopes it is.
• Consequences of believing it: “if people believed that, they’d lose hope, so it can’t be true.” Believing something can have real effects, but those effects are still not evidence about the claim.

Argumentation theorists in the pragma-dialectical tradition describe the fallacy as defending a factual claim with the wrong kind of argument: undesirable consequences are offered in support of a descriptive claim, confusing facts with values (as summarized by Francisca Snoeck Henkemans). The moralistic fallacy, which concludes that something isn’t so because it ought not to be, makes a similar move from values to facts.

Several neighboring errors are built on consequences. A Slippery slope argues from a chain of bad consequences. An appeal to force makes the bad consequence one the speaker threatens to cause; Walton analyzes threat arguments as having the underlying structure of an argument from negative consequences. Fear-based arguments can combine this with an appeal to emotion.

Consequences are relevant whenever the question is what to do, or when they tell you something about whether the claim is true.

• The conclusion is a decision. Whether to close the pier, take the drug or rename the school depends on what would happen if you did. This is the ordinary argument from consequences, and Walton notes that it is extremely common and often quite reasonable.
• The claim is itself about consequences or value. “This policy would hurt small shops” is answered by looking at its consequences.
• The consequence is a testable prediction. If a claim implies something you’d be able to observe, not observing it counts against the claim, whether or not the prediction is pleasant.
• The consequence follows logically and is absurd. Henrike Jansen distinguishes this kind of refutation, reductio ad absurdum, which shows that a claim logically implies something unacceptable, from an argument from consequences, which appeals to outcomes that a claim or proposal would cause. The two are often confused, but the reductio is a legitimate way to refute a claim.
• You’re deciding how to act while the facts are uncertain. “Treat the ladder as broken until someone checks it” uses consequences to choose an action. It doesn’t claim to know whether the ladder is broken.

The test: would the good or bad outcome make the claim more or less likely to be true, or only more or less welcome?

Closing the pier anyway

The town engineer reads the same inspection report. “We don’t yet know whether those cracks go deep. If they don’t and we close the pier, we lose a festival. If they do and we stay open, the pier could fail with a crowd on it. We’re closing it until the divers finish.”

The engineer reasons from consequences, just as the organizer did, but reaches a decision, not a verdict about the concrete. The report’s findings are accepted as uncertain, and the consequences are used to choose what to do under that uncertainty. This is the conclusion is a decision and acting while the facts are uncertain conditions.

The newsletter nobody reads

Sam: Honestly, nobody reads the staff newsletter.

Dana: If nobody read it, nobody would have noticed last week’s misprinted phone number. Six people emailed about it within an hour.

Dana argues from what would follow if Sam’s claim were true, which can look like an appeal to consequences. But the consequence isn’t a reason to want the claim false; it’s a prediction that can be checked, and the check came out against the claim. This is the testable prediction condition, in the form of Modus tollens.

Most everyday reasoning is about what to do, and for decisions, consequences are exactly what matters. The habit carries over easily to questions of fact, especially when the two are tangled together, as they are when a fact would force an unwelcome decision. Words like “accept” and “should” make the slide easier, since they can mean either “believe” or “act on”, as in Kielkopf’s analysis.

Desire also shapes belief. The Stanford Encyclopedia of Philosophy discusses work connecting fallacies with motivated reasoning, in which what people want to be true affects how they weigh the evidence. An unwelcome claim gets scrutinized hard and a welcome one waved through, so the conclusion that happens to be convenient can feel like the one the evidence supports.

• Ask which question is on the table. Are we deciding whether it’s true, or what to do about it? Answer them separately, in that order.
• Grant the consequences. Saying “yes, closing the pier would be a disaster for the festival” often makes it easier to look at the evidence on its own.
• Treat high stakes as a reason to check, not to decide. When the truth would be costly, that’s a reason for a second inspection or an independent test, and for planning, not for concluding the claim is false.
• Watch the hopeful direction in yourself. When you notice you badly want a result, look hardest at the evidence for it.