Can You Trust Your (Etymological) Dictionary?

By Anatoly Liberman

When the title of a scholarly article contains a general question (that is, a question beginning with a verb), the author’s answer is almost always “no” “Will people ever Speak One Language?” (No, they won’t). “Can Pidgins Tell Us Anything about the Beginning of Human Speech?” (No, they cannot). But don’t expect sensations from this post. Can you trust the etymologies provided by your dictionary? Yes, you can, though not unconditionally.

Most of the information we find in dictionaries is the result of consensus. What is the meaning of the noun moon? “A planet, a satellite of the earth.” And what is to moon? “To show signs of infatuation.” Anything else? “To spend one’s time in idle reverie.” Is that all? Not quite: also “to expose one’s bare buttocks.” How do we know those meanings? From experience: English speakers have agreed to assign them to that object and those actions. Is there a family name Moon? Sure. One can find it in any directory. The pronunciation of the word moon is another observable fact. But the origin of moon has to be recovered, and a plebiscite is not part of the procedure. In similar fashion, it matters not at all how many people believe that the moon is made of green cheese. Even if everybody is positive on this point, the conclusion will be wrong.

Language reconstruction is never a hundred percent secure. In a typical etymological thriller, the main witnesses have been dead for a long time or refuse to speak. Historical linguists are doomed to play the game of probabilities and are often satisfied with the least improbable rather than the most probable solution. The authors of etymologies are disadvantaged detectives, not seers. The public stays away from the lexicographical kitchen and has blind faith in dictionaries. This is an excellent thing, for despite the fashionable opinion that all knowledge is relative and that there is no reality (everything is allegedly a construct and depends on the point of view of the observer), we pine for the absolute. Rational human beings, when in doubt about a word, do not indulge in mooning (reverie); they look it up in a dictionary. Oh, yes, of course: a modern dictionary should be descriptive, not prescriptive. You, like, say irregardless, and this is your right, but good dictionaries (even though they appreciate your feelings, feel your pain, and are full of sympathy-empathy) should, like, gently advise you against such usage, and, as a rule, they do. However, when it comes to etymology, the best lexicographer can only say what is supposed to be right or express an informed opinion, and it is instructive to observe the change of these opinions.

Those who read this blog with some regularity must have noticed how often my discussion resolves itself into listing conjectures and trying to say something beyond “origin unknown.” Explanatory English dictionaries never, and specialized etymological dictionaries almost never, present a full picture of the debate surrounding word origins. This is due to the limitation of space, the editors’ natural wish to avoid technicalities that will scare the uninitiated (for etymology is as technical as chemistry, but dictionaries are published to be sold), the absence of a database that comprises everything researchers have written about the origin of English words, and the fear of beginning a comprehensive dictionary and never finishing it. It is irritating that we often have conflicting reports even on the origin of words created in recent memory; consider the history of Jeep, glitch, and most slang. Old words present graver problems. The list at my disposal is disconcertingly long, but two examples of what may be called etymological games will probably suffice.

Boy. In Old English, the proper name Boia was recorded, but its connection with boy remains a matter of debate. Boy surfaced in English texts only in the 13th century and at that time it meant “menial servant.” The sense “male child” emerged (or developed) later. Since boy goes back to Middle English, it might be a borrowing from French. However, despite the lack of clarity, most etymologists believed (note: believed) that the English word had a Germanic origin. The sole dissenting voice (we are in the year 1900 with it) made no impression on dictionary makers. But in 1940 a distinguished scholar, unaware of his predecessor’s work, again suggested that boy continues a French etymon. I deliberately skip all the forms and names, because it is only lexicographical practice that interests me here. The 1940 publication was hailed like a great discovery, and some of our most authoritative dictionaries changed tack. The Concise Oxford Dictionary is one of the best products of 20th-century lexicography. It authors thought that boy was Germanic, but the dictionary has been revised and updated many times. The 5th edition states that the origin of boy, “subject of involved conjectures,” is still undiscovered. The next two editions cite a French source. In the meantime, strong objections were raised to the 1940 article. As a result, the 8th and the 9th editions reproduced the statement of the 6th and the 7th with a question mark. The 10th edition says “origin unknown.” Sic transit… (Sorry for the unbearable cliché.)

Girl. Here is another Middle English word, and its origin has also been the “subject of involved conjectures.” This should not come as a surprise. Words for “child, boy, girl” often trace to metaphors and metonymies that are hard to trace. Especially common are the equations “child” = “twig, branch, offshoot; stump, piece of wood.” In Greek, Latin, Germanic, Romance, Celtic, and Slavic, etymologists face similar problems when it comes to the designation of children. Attempts to discover the etymology of girl and boy have in many respects been similar. As regards girl, only one thing is incontestable: -l is a diminutive suffix, so that girl is a little gir, whatever gir means. The root gir- is not isolated in Germanic, and it competes with gor- and gur-, both being the basis of the names of young creatures. Girl appears to have been borrowed from Low (that is, northern) German. But as early as 1855, it was suggested that girl continues Old English girla “dress,” a word that surfaced in several forms. Such a metonymy would not be unusual, for words for “girl” and “woman” often derives from the names of clothes (compare he runs after every skirt). The girla/girl etymology had little currency, though it was not forgotten. It received a second lease on life in a 1967 article by a leading American scholar (who—a familiar story—did not realize that he had rediscovered an old but usable wheel). The second edition of The Random House Dictionary of the English Language (let it be noted, a splendid dictionary) incorporated it, and The Barnhart Dictionary of English Etymology treated it with respect. By contrast, the many modern offshoots of the OED hedge a little but prefer the traditional view (they say “probably related to the Low German form”). I think, if you are interested in my opinion, that boy is Germanic and girl was a borrowing from German.

Two examples, as I have said, will be enough. The situation is always the same. Dictionaries reproduce a certain view that is supposed to be safe. Then some iconoclast offers a different hypothesis. Some editors ignore it, while others jump on the bandwagon. Etymology is like medicine in that its prescriptions (recommendations) reflect not the truth but the state of the art. Should you trust your doctor? Indeed you should. And the same is true of your etymologist. May a clinic and a dictionary live long and be available to all, even though neither guarantees survival.

Anatoly Liberman is the author of Word Origins…And How We Know Them as well as An Analytic Dictionary of English Etymology: An Introduction. His column on word origins, The Oxford Etymologist, appears here, each Wednesday. Send your etymology question to [email protected]; he’ll do his best to avoid responding with “origin unknown.”