Oxford University Press's
Academic Insights for the Thinking World

Monthly Gleanings: April 2009

By Anatoly Liberman

Heavy petting. The beauty of a blog is that one can always correct one’s mistakes. In my post on the etymology of pet, I suggested (on the basis of my reading) that, contrary to what is said in dictionaries, pet had come to English from Celtic, rather than the other way around. Unfortunately, I forgot to consult the OED online (the so-called 3rd edition of the OED), and a correspondent pointed out that there I would have found the same solution, which is true. But my oversight, embarrassing as it may be, will allow me to touch on an important aspect of etymological lexicography.

Among other things, I said that nowadays two streams—of books and articles on the origin of words and of English dictionaries (to the extent that they deal with etymology)—seldom meet and explained why this happens: the huge amount of work lexicographers face increases manifold because of the obscurity of countless publications in numerous languages. I became aware of a persuasive etymology of pet because I have a reasonably complete database, and two articles on this word (by O’Rahilly and Vendryes) turned up in it. At Oxford they undoubtedly have their own bibliographical resources; this case their information probably came from Vendryes’s etymological dictionary of Old Irish. The volume with the letter p was published in 1983. In the entry petta, Vendryes referred to O’Rahilly’s but not to his own 1927 article, which contained the same reconstruction he gave in the dictionary 56 years later. The etymologists at OED, quite naturally, consulted the dictionary, read O’Rahilly’s article and, like me, agreed with him. I am glad that our opinions coincided. But they went further than I did and tentatively accepted Vendryes’s distant etymology of the Irish word, and this is a point I would like to discuss.

English absorbed more foreign words than perhaps any other European language. However, an English etymologist is at best a specialist in Indo-European linguistics, and more often only in Germanic. Let us have no illusions about it. The field is vast. Even the history of one language (in our case, English) is so full of esoteric details that some people call themselves experts in Old English, others in Middle English, and still others in early Modern English, to say nothing of the latest period. It is unrealistic to expect that this person, who (for example) has spent years mastering every sound change and every semantic twist of Middle English will feel even half as comfortable in the history of Frisian, Dutch, German, Danish, and so forth. But English etymologists also deal with French, Latin, Greek, and Celtic. Borrowings from Finnish and Russian (as long as we stay in Europe), Native American, African, and Asian languages also occupy them. Obviously, they cannot have an informed opinion about such a mass of disparate material and depend on the works of specialists, who argue over the origin of “their” words, just as English etymologists argue over the derivation of English vocabulary. Copying their verdicts is fraught with danger. I can add that Skeat and James Murray boldly attacked Latin words in their dictionaries and often did not see eye to eye. Their suggestions are invariably clever, for their command of two classical languages left nothing to be desired (Skeat’s education was in Latin and Greek, and Murray seems to have mastered every old and new language), but I still think that they should have left Latin to people like Alois Walde. It is enough to compare the subsequent editions of his dictionary to see what a minefield Latin etymology is.

In 1927 Vendryes offered what he thought to be a good etymology of Old Irish petta, but the entry in his dictionary is short and rather dogmatic. O’Rahilly’s article but not his conjecture is mentioned. Historians of English are usually not in a position to agree or disagree with a distinguished Celtologist on the origin of an Old Irish word. I could afford the luxury of contradicting Vendryes because I was the author of an evanescent post and realized that, if I was wrong, the fault would be only mine and do no one any harm. The idea of tracing Old Irish petta to an ancient Indo-European root seemed unfounded to me, for petta looks like an example of early European slang (pet ~ put “small”), while Latin suesco “to get used” (the cognate suggested by Vendryes; sue– in it is from the historical perspective a prefix) and other related forms have little to do with pampered people and home animals.

By contrast, a semi-anonymous dictionary does not enjoy the freedom of an individual contributor. The OED is especially vulnerable, because most of its users take its pronouncements for absolute truth. The same holds for Webster and some other “thick” dictionaries. My advice to the editors is disappointing. I believe that bold hypotheses should be either discussed in them with the attention they deserve or avoided altogether. Research into the history of pet has reached a stage at which an expert in the area of English can say “from Scottish Gaelic” without fear. The origin of the Gaelic word is somebody else’s business. Etymological dictionaries of French cite the Latin etymons but never discuss them, and I think they are right. It is enough to read the old reviews of Kluge-Mitzka (a German etymological dictionary) with titles like “Slavic Words in KM” and “French Words in KM,” to see how hard it was for Mitzka to deal with the material he did not know in detail. Even discussion of the Frisian element in the latest etymological dictionary of Dutch fills one with dismay, though Frisian is much closer to Dutch than English is to Gaelic.

Last week, my post was devoted to the origin of race. Here I think we are in a better situation than with pet. Contini’s reconstruction is not an “opinion,” not an attempt to connect a Romance word with a hypothetical root; it is based on evidence. Von Wartburg, the great compiler of a multivolume etymological dictionary of French, challenged Contini’s colleagues to boost the derivation of race from Old French haras “stud” (he preferred to trace race to ratio), and this is exactly what they did. In the entry race, the online edition of the OED is cautious: it lists two other conjectures and mentions Contini’s idea (though the reference is to dictionaries). Here I think we may be less circumspect. Just as we now know that Engl. pet is a borrowing from Celtic, we know that race did not come to French from Italian, but since French etymology is not what the OED (or any English dictionary) is expected to elucidate, had the editors said “from French” and stopped, they would have been justified. If some users of the OED are curious about the history of razza ~ race in the Romance languages, they are welcome to find out. Even the great OED is not a clearing house of world etymology. No dictionary is.

Small fry. 1) Generic they. In my old polemical notes on the pronouns they and their being used with reference to a single person, I should have made it clear that this usage (the price we pay for purging English of gender-specific words when gender-neutral ones are expected), apart from being inelegant and often silly, leads to people’s stultification, for the simplest sentences begin to verge on self-parody. Here are two such sentences from a local (University of Minnesota) student newspaper. “As someone who has been pro-life all their life, I believe life begins at the point of conception.” The letter writer did not dare to use my, with I being the antecedent; their seemed to be safer. “Can someone in Dateopia tell me [the correspondence was addressed to Dr. Date, a collective expert on “relationships,” loss of virginity, and unrequited love] why it is that we have so many brooding male writers out there? … I am struck by the feeling that the writer has spent a lot of time thinking about their situation in an academic sense….” A male surely has their problems. Don’t all of us?

2) Ghetto. Those interested in the origin of the word ghetto should pay special attention to the comment by Anonymous (I happen to know the identity of the commentator but see no reason to make it public.) He is a proponent of the borghetto etymology defended by such outstanding experts as Alfredo Prati and Leo Spitzer. Unfortunately, etymologies cannot be proved (even disproving a palpably wrong etymology is sometimes hard.) I thought that the meaning of the clipped form ghetto was too general for naming the newly established Jewish quarter in Venice, but “street” (gata) is not much better, except that Jewish quarters were called “street” all over Eastern Europe, whereas ghetto did not mean “street.” My extremely modest goal was to enliven the discussion by offering what I though was a novel idea. I will take refutation in stride, even gratefully. As for why I cited the Gothic and the Swedish form rather than those from Danish (gade) and Norwegian (gate), the question has been answered by our correspondent Paul Peterson: gade and gate show traces of later sound changes; gata is more archaic. 3) Latin pax ~ English fair, Engl. war ~ French guerre. I mentioned these cognates in my post on “Peace and War.” Our correspondent wonders how –ax (in pax) and –air (in fair), as well as war and guerre, are connected. Pax is pak-s; its root is pak-. The early history of fair becomes clear from its Old Icelandic cognate fag-r (-r is part of an ancient suffix). Latin p corresponds to Germanic f, and k corresponds to h, but there is a rule by which voiceless spirants may, depending on the ancient place of stress, become voiced; hence g in fagr. In certain positions, Old Engl. g was pronounced like y in Modern Engl. yes, and this is how fair acquired its pronunciation. Romance speaking people borrowed a word meaning “strife” with the Germanic root wer-, but, since the sound w was alien to them, they replaced it with gv (spelled gu). In Old Northern French (the dialect of William the Conqueror), the pronunciation was evidently werre (without g-), so that the French noun returned to English in almost the same shape in which it at one time had existed. Yet war is not a direct continuation of the Germanic form.

4) Engl. pet and German Petze “female dog.” Those are most probably unrelated. Several similar-sounding words for “female dog” have been recorded in old and later European languages: Engl. bitch, Old Icelandic bikkja (from which, I think we have the verb bicker), as well as baka (in greybaka), and German Petze. Petze surfaced only in the 15th century. It is anybody’s guess how those words are related (if at all). One can go even further and cite Russian sobaka “dog” (stress on the second syllable), an irritatingly obscure word that has a history of its own. A similar common European litter accompanies us the moment we whistle to a tyke. Perhaps some of the names mentioned above imitate the bark. Neither sounds like bow-wow, barf-barf, or yap-yap, but there was an Old English word tife “bitch,” and one of the Russian words for “bow-wow” is tiaf-tiaf! Dogs, as I understand, were domesticated to guard herds, rather than as pets. 5) Caius of Cambridge. In discussing the origin of the word key, I mentioned John Caius, the founder of a college at Cambridge. The question was about the man and the pronunciation of his name (which is keez). All the eminent bearers of the names Kayes and Keys seem to have Latinized their name as Caius. The Oxford man was also Caius (keez). Googling for the biography of both produces several worthwhile results. As for the pronunciation, I consulted the book Laut und Leben. Englische Lautgeschichte der neueren Zeit by Wilhelm Horn and Martin Lehnert. Berlin: Deutscher Verlag der Wissenschaften, 1954, p. 288. (I am indebted to this book for many of my posts on spelling.)

6) Acronyms. I said in connection with the etymology of tip “gratuity” that all attempts to explain common words as acronyms are usually wrong. Our correspondent cited NABISCO and others. But those are not “common words”: they were coined as acronyms and never concealed their origin. Of such formations we have hundreds. 7) Shed light. There is little to say about the origin of this phrase. Light can be shed, thrown, or cast. According to the OED, the figurative use of shed light (that is, to illuminate some matter) does not antedate the middle of the 19th century. 8 ) I was pleased to get support of my views on Shakespeare. Somebody wrote his plays. This person was probably William Shakespeare.


Anatoly_libermanAnatoly 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.”

Recent Comments

  1. John Cowan

    When referring to an unknown or unspecified individual, or referring distributively (rather than collectively) to the unknown members of a class, they is a grammatical and appropriate pronoun. This is a rule of English “grammatical number”. It is no more inelegant, silly, or self-stultifying than the French rule which requires that if we speak of la personne qui a écrit sur le mur, any future referring pronouns must be elle, even if the wall in question is in a men’s room. And it has been the rule since at least the 19th century, as comments to your earlier posts have firmly established.

  2. B.G.Sanford

    Than you for your informative article. Being a romance writer, I deal in words so to speak. I suppose, before I leave this academic site, I should mention my new book, if you will permit me. “Beth:Love Along The Way…by B.G.Sanford is the story of one woman who overcomes all odds and two ugly divorces to find real Love…..Along The Way,” and just released by Eloquent Books. It’s an amazing and entertaining story and study in relationships, and if you’re so inclined, you can order it off the internet or have your favorite bookstore order it for you.
    Good reading friends,

  3. […] Monthly Gleanings: April 2009 : OUPblog […]

  4. J P Maher

    Anatoly Liberman and I are in agreement that the “foundry” etymology is junk or joke, in the vein of Jose Jimenez (born Szathmary): “I’m a Jungarian Hew”. Dictionaries outside Italy are not much help/ And in Italy the dictionaries have to be supplemented by gazetteers, Yahoo and Google maps, as well as figurative speech. GHETTO means, as also CHIASSO, noise, disorder, a slum (street, quarter), a whorehouse (literal & figurative), a racket (acoustic or criminal). See the Vocabolario degli Accademici della Crusca (Vocabulary of the Academics of the Crusca, i.e winnowing). Translation mine): “BORGHETTO: dim[inutive] of BORGO = Lat[in] VICULUS ‘street’.”

  5. J P Maher

    Isn’t Russian SOBAKA ‘dog’ from Iranian SPAKA ‘bitch’? Dogs get metaphoric names. In German lands NERO is a favorite. Spanish PERRO is from the name PYRRHUS, a ruthless tyrant of Hellenic Sicily (not him of the Victory). Polish SMOK ‘dragon’ appears in German and Yiddish as name of a dangerous watchdog, then a nasty man and in Yiddish and Yinglish by totum pro parte – SCHMOCK and SHMUCK. If tyrants and watch dogs, why not bears? PETZ ~ BETZ is the nickname of the bear, BAER…

Comments are closed.