What is actually wrong with using Etymology to infer a word's meaning?

I mean other than semantics( or more subtle meaning, nuance) of what other use could studying etymology be.

I cannot see the similarities of Etymological Fallacy and Non Sequitur.

Theory serves the purpose of explaining and predicting phenomena.

Etymology could serve the purpose of explaining and predicting the meaning of a word.

It can also solve the Demarcation Problem. It could be the authority to decide what is right or wrong in order to avoid ambiguity. Like Oxford English Dictionary does for the modern English language. Only more general( every language has its words which have their etymology) and interlingual-intertemporal. Languages exchange words and Etymology is recorded through history.

Languages do Evolve but there can be no sudden or drastic Semantic changes.

I do not think I am a language purist just because I want to avoid ambiguity.

  • 3
    The etymological fallacy is fallacious because the speakers of a language don't usually know the histories of the expressions they use, which is what etymology tells you about.
    – Greg Lee
    Dec 27, 2019 at 14:36
  • 2
    If we were all using English words according to their reconstructed meaning as far back in time as we can (so, likely Proto-Indo-European), then the vast majority of common everyday words would acquire a completely different meaning than the one they actually have today.
    – LjL
    Dec 27, 2019 at 20:01
  • 2
    Also, if you find "no other use" in studying etymology than determining true meanings, then I you must have dealt very little with linguistics: etymology is used extensively to relate languages to one another, eventually building "family trees" of them, and to reconstruct languages or individual words that we have no direct attestation of. It's used a lot "backwards", if you like. I find that assertion, and others you made in this question, unwarranted and presumptuous (you don't get to decide what semantic changes are too sudden, and neither does the OED), and this is why I'm voting it down.
    – LjL
    Dec 27, 2019 at 20:09
  • @LjL Downvotes are purposed to show a question is badly written or doesn't fit the site, I would argue that the former is true, but I am upvoting the question to balance it out.
    – user22430
    Apr 6, 2020 at 0:21
  • @William if a question doesn't fit the site, I flag it as offtopic.
    – LjL
    Apr 7, 2020 at 1:58

3 Answers 3


Very ironically, understanding etymology today to be "true word meaning" is commiting to the etymological fallacy.

First, etymology is often uncertain and speculative, because speech in general. We can't be anymore certain about a words ancient usage than the scribe we learned it from. E.g.: The origin of Slav from slave is highly contested. Ignoring doubt in face of uncertainty would be fallaciou

Second, even if a word is well attested, at some ancient point in time, the way is often not clear that it took to arrive in English. E.g.: I have looked into "plague", found it refered to Latin plaga "wound", but no mention of where and when. That's not enough to work with. Accepting it as the last word would be fallacious.

Third, if a words development is well attested, we can be sure that the meaning has changed, or even come to be an auto-oxymoron. E.g.:

  • a) dysphemisms in recent times, like retarded, which was once a medical term, but now has been denigrated to a slur;

  • b) many Latin loans, e.g. camp, camping which didn't exist in that sense in classical Latin, but rather meant "open field, battle field, encampment", where part of the etymology transparently rests on the morphology, and part on semantic widening;

  • c) semantic narrowing is a thing as well, e.g. in car, cart, carry, as it's obvious that the definition for car today involves a motor; in that sense, motor, mecanismus etc has become polysemic.

  • d) figurative usage: a computer was a person only a century ago; The word for the computing machine only retains the animate aspect of the morpheme -er.

In these cases, it would be fallacious to stick to one ancient interpretation.

Fourth, in case of doubt it is often reasonable to assume that the core meaning of a word can still be inferred from usage.

  • e) even if reinterpretation happens, so that the morphology of a word changes, or if the original etymon has been forgotten, there's often one context in which the original sense is retained, which appears paradox after the change. I'll owe you an apt example. Trivially, the more complicated roots of Latin terms are commonly not well understood, e.g. the difference between re- "back, again, against", res- "thing, matter", but this is mostly inconsequential, because ...

... It would be fallacious to infer diachronic etymology on the basis of corrupted morphemes and folk etymology. But it would be likewise fallacious to infer synchronic etymology from diachronic etymology.

If Karl means "king" in Polish, Schärler "fighter" in German, and Karl, Charlotte personal names, and churl "knave" in English, then that's just that

I mean other than semantics( or more subtle meaning, nuance) of what other use could studying etymology be.

Historical linguistics is important for history, in various ways.The reconstruction of Proto-Indo-European religion is just one amazing example. But! The etymological fallacy means we can't rely purely on linguistic evidence and have to work in tandem with the fields of History, Archaeology and even Geology.

Languages do Evolve but there can be no sudden or drastic Semantic changes.

See a) above. A similar timespan of ten to hundred years looks pretty much like an instant through the lense of history. A single utterance is basically an instance. An utterance may have lasting effects if it reaches many listeners.

I do not think I am a language purist just because I want to avoid ambiguity.

Mere passive desire doesn't make you an -istor of any kind, but being proscriptive, activelly denying certain meaning and usage does. Indeed, proscriptivity is often based on authority, but that's a matter of Socio-Linguistics. Etymology only tries to chronicle the development. Etymology can inform authoritative decisions, but that's often not the basis, see 1. above. Vice Versa, Socio-Linguistics can try to describe the politics of etymology, whereby the basic assumption is scientific rigor and good intentions.

PS: Your post is missing a direct question. It appears like a request for comments. So I picked out the last statement and interpreted it as a question.

PPS: I think what I really wanted to say about the etymological fallacy is the problem of two competing perspectives. For one, the native matrix language naturally directs the etymological discourse and language learning. The logically chronological development of language is secondary, at odds with the immediate top-down approach by which we learn about it. The basic fallacy lies with confusion of these two perspectives, interrelated as they are.

  • This is essentially the fundamental problem of linguistics: How do we learn language, or what is language? related to the philosophical problem of consciousness, who am I? the epistemic logic problem of what is truth? and the ages old question: what's up?

The problem with using etymology to infer the meaning of a word is that language is always changing, and how a word is used now is not necessarily how it has been used historically.

Linguists do not choose the "correct" way to use a word, but rather record and explain language as it is used. If etymology can explain how a word's meanings boundaries make sense in the context of its etymology, then it is useful, but if the current use does not adhere to those historical boundaries, then etymology is not useful for explaining the current use.

The Etymological Fallacy is similar to a Non Sequitur, because older meanings (without the context of how they changed to modern meanings) are irrelevant to what we mean when we say words now. When a French speaker refers to a "pomme de terre" they are not referring to a ground apple, but rather a potato, and when an Italian speaker refers to a "pomodoro," that is likewise a tomato, and not an apple made out of gold. If you get asked for a "pomme de terre" and put an apple on the ground and pick it up, and hand it to the other person, you will get a funny look. Similarly, in English, "naughty" doesn't mean "has nothing (naught)," but shifted through the intermediate meaning "has no morals" to the current meaning. And so bringing back the original meaning, and calling a glass "naughty" for containing no water, you have not used language in a way that people understand. For an even quicker semantic shift, "The texts are her water gate" would have meant something different before 1974 (when "water gate" became a word that specifically referring to damning evidence in a scandal, and not a temporary dam) and now (when texts usually refers to text messages, and not paper documents).

Unfortunately for anyone who wants to avoid ambiguity, it seems to always keep arising, as language changes to fit new situations. There are better (more effective, less annoying, less likely to end up ignoring marginalized groups) ways to do this than trying to proscribe that language be used now as it was in the past.

  • 1
    Re ambiguity, there's a saying (I first saw it on LanguageLog) that ambiguity is a feature of language, not a bug Dec 27, 2019 at 23:26
  • Which alternative is there to avoid ambiguity? When the cost of solving ambiguity(transaction/ communication cost) is less than the cost of the ambiguity in itself( I would like to call it an "externality" or any other market failure would do) the parts themselves will solve the ambiguity. The problem is when it costs to much to solve the ambiguity. The parts will incur the costs of the ambiguity in itself rather than solving it. It is costly to be clear. One will always need an Authority as I see to optimize the utility. The problem is what should the authority be. The one to proscribe. Dec 27, 2019 at 23:52
  • Ambiguity fixes occur naturally! For example, words can drop meanings (the rise of mechanical computers got rid of the ability to use "computer" to refer to human computers. In the southern USA, the vowels in "pin" and "pen" merged, and so to deal with the ambiguity that it created, people in that dialect region say "ink-pen" and "stick-pin" Dec 28, 2019 at 22:51
  • And we said nothing different. People will only fix ambiguity when the cost of solving it is less that then cost of suffering with it. That is Coase Theorem applied in communication. Unless there is some authority to proscribe the use ambiguity will never be extinguished. Bona fide communicators very well understand that ambiguity harms them but fixing it would simply harm them more. Who/What can be the authority which will be faced with the least opposition? Jan 2, 2020 at 21:11
  • What I am saying is that people do these things without an authority. Ambiguity is not a "problem" that you need to mandate out of existence, but something that is a part an parcel of every language. It can even be seen as an efficiency, as it allows people to use more short words, and more minimally worded constructions. Jan 2, 2020 at 23:44

There are plenty of cases where etymology is irrelevant to understand the meaning of a word. An example is "porcelain", now a type of ceramics. Originally, the word "porcelana" was created to describe a kind of sea-shell, that looks like the vulva of a sow, hence the root porc- of "porcelain". This sea-shell is indeed whitish and looks like porcelain. It's of no help to know the origin of the word "porcelain" to understand what it now means in present times.

  • Etymologies like this pork vulva is why OP brings up "the authority to decide […] only more general"--because everyone with a shred of reason left has to recognize that as a folk etymology. For one, the etymon is merely supposed to be porcella f. "piglet". Second, I doubt even this, because -cella reminds me of En shell, PIE *(s)kel(H)* "to cut" (supposedly from a sense "to peel" > "peel", viz Ger v. schälen, n. f. Schale; but even this is doubtful, cf *ske-w- "to cover"; also cp skull ~ Ger Schädel; cp Ger Muschel "sea shell, mollusc", also Muschi "vulva", Mulle "vulva")
    – vectory
    Dec 27, 2019 at 14:38
  • 5
    @vectory The -c- in porcella is part of the root porc-, and has nothing to do with English "shell".
    – Draconis
    Dec 27, 2019 at 16:49
  • @Draconis if you say so.
    – vectory
    Dec 27, 2019 at 16:54

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge that you have read and understand our privacy policy and code of conduct.

Not the answer you're looking for? Browse other questions tagged or ask your own question.