The name of this fallacy comes from the sport of fox hunting. According to one story3, dragging a dried, smoked herring, which is red in color, across the trail of the fox would throw the hounds off the scent4. Thus, in general, a "red herring" is anything that can be used to distract attention5. In the context of argumentation, a red herring is something which distracts the audience from the issue in question. This frequently occurs during debates when there is an at least implicit topic, yet it can be easy to lose track of it. By extension, it applies to any argument in which the premisses are logically irrelevant to the conclusion.
This fallacy is one of Aristotle's thirteen fallacies identified in his pioneering work On Sophistical Refutations, which dealt with fallacious refutations in debate. It is often known by the Latin name "ignoratio elenchi", which is a translation of Aristotle's Greek phrase for "ignorance of refutation". The ignorance involved is either ignorance of the conclusion to be refutedeven deliberately ignoring itor ignorance of what constitutes a refutation, so that the attempt misses the mark. As with all of Aristotle's original fallacies, its application has widened to include all arguments, not just refutations or those occurring in the context of a debate.
Red Herring is the most general fallacy of irrelevance. Any argument in which the premisses are logically unrelated to the conclusion commits this fallacy. A set of premisses is logically irrelevant to a conclusion if their truth does not make it more likely that the conclusion is true.
- Because it is the most general fallacy of irrelevance, most fallacious arguments will be identified as some more specific type of irrelevancy. This is why I provide no example of this fallacy: for an example, see any of its subfallacies, above.
- Most other logicians seem to use Red Herring, under whatever name they call it―see the Aliases, above―as a catch-all category for any irrelevant argument that doesn't fit into one of the other types of fallacy of irrelevance6. In contrast, I consider all fallacies of irrelevance as types of Red Herring, though I would usually classify individual arguments as some subfallacy thereof. In practice, these two approaches have similar results, but mine does not treat Red Herring as a sort of bin for examples of logical irrelevance that don't fit elsewhere.
- Logical relevance is a vague and ambiguous notion. It is ambiguous in that deductive and inductive reasoning seem to involve distinct types of relevance: Deductive relevance is semantic in nature, whereas inductive relevance is causal. Both types are vague in that there are no precise, language-independent definitions of semantic or causal relevance.
Another ambiguity of the term "relevance" is that logical relevance can be confused with psychological relevance. The fact that two ideas are logically related may be one reason why one makes you think of the other, but there are other reasons, and the stream of consciousness often includes associations between ideas that are not at all logically related. Moreover, not all logical relations are obvious, so that a logical relationship may not cause a subjective feeling of relatedness. This is why proofs are sometimes surprising: a logical proof shows that the conclusion is logically related to the premisses, even though this fact may not have been apparent. When logical relations are immediately obvious, proofs are usually considered unnecessary.
- Many other types of fallacy involve irrelevance: for instance, in fallacies of ambiguity, the premisses are logically irrelevant to the conclusion, but this fact is disguised by ambiguous language. However, Aristotle classifies "ignorance of refutation" as a language-independent fallacy, though he does say: "One might, with some violence, bring this fallacy into the group of fallacies dependent on language as well."7 However, this would make Red Herring so wide that just about every fallacywith the exception of Begging the Questionwould be a subfallacy of it. This is too wide to be useful, so I will follow Aristotle in restricting it to non-linguistic fallacies, excluding those where the irrelevance is disguised by ambiguity or vagueness.
- All of these aliases, except for "ignoratio elenchi" but including "red herring", come from: S. Morris Engel, With Good Reason: An Introduction to Informal Fallacies (6th edition, 2000), p. 190.
- Translation: "Ignorance of refutation", Latin. This is a translation of Aristotle's classical Greek phrase; see: On Sophistical Refutations.
- William & Mary Morris, Dictionary of Word and Phrase Origins (1962).
- Engel tells a different story―see Note 1, above:
Red herring may seem a puzzling name. It derives from the fact that prison escapees have been known to smear themselves with a herring (which turns brown or red when it spoils) in order to throw dogs off their track.
Whatever the truth about the origin of the name, a "red herring" in the metaphorical sense is a distraction or diversion.
- Readers of mystery stories will be familiar with the term: in that context, it refers to a clue designed to mislead the reader.
- See, in particular, Engel's treatment, pp. 190-3―see Note 1.
- Section 1, Part 5.