Etymological fallacy

Decimate

A news report says a late frost decimated the county’s peach crop, killing about 80% of the blossoms. A reader writes in: “Decimate means to destroy one in ten. It comes from the Latin for ten. If 80% was lost, the crop wasn’t decimated, and the paper should use words correctly.”

The Latin history is real: Crystal gives “destroy one tenth of” as the older sense, from Latin decem, “ten”. But, as Crystal notes, “decimate” is nowadays widely used to mean “destroy a lot of”, and that’s what readers understand by it. He points out the irony of insisting on the old sense: it virtually bans the word from everyday use, since it’s hard to imagine contexts where anyone needs to destroy exactly one tenth of something. This is the clear-cut case, and the textbook one; Crystal and the corpus linguist Michael Stubbs both use it.

Which “nice”?

“Calling someone ‘nice’ is really a backhanded compliment. The word comes from a Latin word for ignorant.”

Crystal traces “nice” through its history: it meant something like “fastidious” in Shakespeare’s day, “silly” in Old French, and it goes back to Latin nescius, “ignorant”. If the older meaning is the real one, which older meaning? Crystal asks whether we must conclude that we don’t know what “nice” means, because its original use in Indo-European is obscure or lost. The example shows the fallacy’s edge: it has no principled place to stop, and the meaning it picks is usually just the oldest one the speaker happens to know.

“Awful” praise

After a friend describes a hotel stay as “awful”, someone replies: “Actually, awful means full of awe. So technically you’re saying it was impressive.”

The linguist Laurie Bauer notes that “awful” used to mean “full of awe, inspiring awe”, just as “terrific” used to mean “causing terror”. But the friend was speaking modern English to a modern listener, and nobody in the conversation took “awful” as praise. Applying the old meaning to what someone said today gets the speaker’s meaning wrong. Compare the look-alike below, where the same old meaning is exactly right.

• Insisting on the oldest sense: “decimate” must mean one in ten.
• Reading meaning off a word’s root: assuming that a word means whatever its root meant. The biblical scholar James Barr, in The Semantics of Biblical Language (1961), named a related error “the root fallacy”: giving excessive weight to the meaning of a word’s root, which he said in many cases “comes to much the same thing as ‘etymologizing’”, giving too much weight to a word’s origin. His subject was the study of ancient Hebrew and Greek, but the point about language is general.
• Correcting others’ usage: the everyday form, in which the older sense is used to call a common modern sense an error.

Names. The term comes from linguistics rather than logic textbooks. John Lyons’s Semantics (1977) indexes “etymological fallacy” and states the underlying principle: “The etymology of a lexeme is, in principle, synchronically irrelevant”, that is, irrelevant to what the word means at any given point in time. Stubbs describes it as a “faulty line of argument”, precisely because words change in meaning over time.

The fallacy is close to the Genetic fallacy, which judges a claim by where it came from; this one judges a meaning by where it came from. It also resembles an Appeal to tradition: the older use is treated as the proper one because it came first. Neither classification is standard in the sources, so this page lists them as related rather than as parents.

• Reading an old text. What a word meant when it was written is what the writer meant. Reading “awful” in an eighteenth-century letter as “awe-inspiring” may be exactly right.
• Explaining how a meaning developed. Etymology is good at showing how a word got its current sense. Stubbs notes that the etymological fallacy disposes only of some poor reasons for studying etymology, leaving the good ones intact.
• A deliberately coined term. When a technical term was defined at its coinage and the field still uses that definition, going back to the original definition is checking current usage.
• Usage really is divided, and clarity is the point. When a word is used in two ways and some readers object to one of them, history can explain the split, and choosing the less contested word for the sake of being understood is sensible. It becomes the fallacy when the history is treated as proof that one current usage is wrong.

The test: is the older meaning being used to explain the word, or to overrule how people use it today?

The letter from 1770

A historian reads a 1770 letter in which a traveler calls a mountain storm “awful” and writes: “He isn’t complaining here. In the eighteenth century ‘awful’ could mean awe-inspiring, and the rest of the letter is full of wonder at the scenery.”

This uses an older meaning of “awful” to overrule the modern one, just like the hotel example. The difference is that the text is old. The question is what the traveler meant, and the right meaning is the one current when he wrote. The historian also checks the reading against the rest of the letter rather than relying on the dictionary alone.

How “decimate” got its meaning

A teacher explains: “’Decimate’ comes from the Latin for ten, because it first meant destroying one in ten. Over time it came to mean destroying a large part of something, which is how people use it now.”

The teacher brings up the same etymology as the letter-writer to the newspaper. But the etymology explains how the meaning developed, and it’s used to describe the modern meaning, not to reject it. Nothing in the explanation says the current use is wrong.

Many people have an intuition that words have a fixed, correct meaning, and that change is a kind of decay. The oldest meaning seems like the purest one, and knowing it feels like having inside knowledge. A classical origin adds prestige: Latin and Greek roots can sound like the word’s true identity, with modern usage a corruption of it.

Etymology is also a genuinely interesting and legitimate subject, and the facts it turns up are usually true. The mistake is a step beyond those facts: from “this word used to mean X” to “this word really means X”. Stubbs’s point is that it’s precisely because words change meaning over time that this step fails.

• Separate history from current meaning. “That’s what it meant then. What do people mean by it now?”
• Ask which origin. Most words have several older meanings. Why stop at this one?
• Check a current dictionary, which records how a word is used today, often alongside its history.
• Allow that a usage point can still be fair. A word used in a way many readers will misunderstand may be worth avoiding for clarity. That’s an argument about being understood, not about the word’s “real” meaning.