Red herring

A true point about something else

Dana: The new project-tracking tool is slower than the old one. Most pages take several seconds to load.

Morgan: The company that made the old tool was bought by a big conglomerate last year, and their customer support has gone downhill since.

Dana’s claim is about speed. Morgan’s point may be true, and it might even matter when deciding which tool to keep, but it says nothing about whether the new tool is slower. Offered as a reply, it changes the question from “is it slower?” to “is the old vendor any good?”, and Dana’s claim is left standing, unanswered.

A tangent that never comes back

A parent asks the youth soccer coach why their daughter has played only about five minutes in each of the last three games. The coach says: “You know, when I started coaching twenty years ago, we didn’t even have proper fields. The kids played on gravel. These kids have no idea how lucky they are, and the volunteers who run this league put in hundreds of unpaid hours every season.”

Everything the coach says may be true, and it’s hard to argue with. It is also about the league, so it feels like a reply. But the question was about one player’s minutes, and the answer never returns to it. Douglas Walton’s term for this kind of case is diversionary irrelevance: the reply heads toward a more stirring issue and never arrives at any conclusion about the original one.

Answering a question about method with a point about importance

Reviewer: The study had only 12 participants, so the result may not hold in a larger group.

Author: This condition affects millions of people every year. Research on it is badly needed.

This one sounds like a rebuttal because it defends the study. But the reviewer’s point was about whether 12 participants can support the result, and how important the condition is doesn’t change how many people were studied. The importance is a reason to do the research well, not evidence that this study did.

• Diversion to a stirring issue: steering toward something vivid, emotional or scandalous that the audience will want to talk about instead. Walton calls this diversionary irrelevance.
• Pointing at a different wrong (“but what about...?”): answering a criticism by raising some other problem. When the other problem is the critic’s own behavior, it becomes the tu quoque form of ad hominem.
• Rambling off the point: wandering through several loosely related topics so that no single diversion can be named, and the question is never answered. Walton calls this pettifogging.

Is it the same as “missing the point”? Not quite, and textbooks disagree about how to file them. The older Latin name ignoratio elenchi (roughly, “ignorance of refutation”), also called irrelevant conclusion or missing the point, covers arguments that set out to prove one conclusion and end up supporting a different one. Some textbooks treat the red herring as a kind of ignoratio elenchi; Patrick Hurley’s A Concise Introduction to Logic treats missing the point and the red herring as two separate fallacies. The Stanford Encyclopedia of Philosophy describes ignoratio elenchi as misdirection in argument rather than weak inference, and counts the straw man as one variation of it.

Walton (2004) argues they are distinct. Missing the point arrives at a definite conclusion, just the wrong one. A red herring leads away from the question without arriving anywhere in particular, and what marks it is a diversion, not a wrong conclusion. He also notes that the line is hard to draw in practice: several textbook examples of the red herring can also be read as missing the point, and the deciding question is whether there is evidence of a deliberate diversion. That’s why this site doesn’t list those names as other names for the red herring.

Changing the topic isn’t a fallacy in itself. It’s a red herring only when the new topic is used as if it answered the original one.

• The new point undermines something the claim relies on. A reply that sounds off-topic can go straight to a premise the claim quietly depends on. Then it isn’t a change of subject at all.
• The first point is granted and weighed against others. In a decision with several factors, “yes, that’s true, but this other thing matters more” answers the decision, not the claim, and says so openly.
• The change of subject is announced. “Let’s come back to that; first I want to raise something else” doesn’t pretend to be an answer.
• A digression that doesn’t claim to settle anything. Conversation wanders, and a tangent offered as a tangent misleads no one.

The test: if the new point were true, would it make the original claim any more or less likely to be true, or does it only make the claim feel answered? Walton’s suggested question works in the moment: “Where are you going with this?”

A detour that goes straight to a premise

Sam: Let’s go to the lake on Saturday. It’s only an hour away.

Riley: The bridge on Route 9 is closed for repairs until October.

Riley seems to be talking about a bridge instead of the lake. But Sam’s argument rests on the trip taking an hour, and if the usual route crosses that bridge, the trip will take longer. Riley’s point bears directly on a premise Sam gave. This is the undermines something the claim relies on condition.

Granting the point and weighing it

Dana: The new project-tracking tool is slower than the old one. Most pages take several seconds to load.

Morgan: You’re right, it’s slower. But it’s the tool our biggest client uses, and copying everything between two systems costs us more time than the slow pages do. I think we should keep it anyway.

Morgan brings in a different consideration, just as in the first example, but doesn’t use it to dodge Dana’s claim. The claim is granted, and the new point is offered for a different question, whether to keep the tool, which both factors really do bear on. This is the granted and weighed condition.

A reply on a nearby topic feels responsive. Walton distinguishes topical relevance (being about the same subject) from probative relevance (actually making a claim more or less likely to be true). A red herring usually has the first and lacks the second, and in a quick exchange the two are easy to mistake for each other.

It also takes attention to keep track of a question while something more engaging is being discussed. A vivid story, a grievance or a point everyone agrees with invites a response on its own terms, and once the conversation has followed it, going back to the original question can feel like the awkward move.

• Restate the original question and ask how the new point bears on it. Sometimes it does, and the answer will show how.
• Grant the new point if it’s true, and offer to discuss it separately. That keeps the original question from being lost without dismissing the other person’s concern.
• In yourself, notice when you answer a question with something that’s true but easier to talk about, and ask whether it actually changes the answer.