When I was growing up, I read Paul de Kruif’s book Microbe Hunters so many times that I still remember some pages by heart. Two chapters in that book are devoted to Pasteur. The second is called “Pasteur and the Mad Dog.” A book about great word hunters would similarly enthrall the young and the old. Think of the chapters: “Jacob Grimm and an Enchanted Castle of Roots” (Jacob and Wilhelm Grimm did not only collect folk tales: they, especially Jacob, were the founders of comparative philology), “Wedgwood beyond Porcelain” (etymologist Hensleigh Wedgwood was related to the porcelain makers but had nothing to do with cups and plates), “Walter Skeat at Home and on the Skating Rink,” a spoof on the analogy of Jean-Jacque Brousson’s memoirs Anatole France en pantoufles (the author of our best etymological dictionary happened to be an excellent skater, and outside the university, folks at Cambridge knew him mainly in that capacity), “Frank Takes His Chance” (about Frank Chance, one of the most sagacious English etymologists of the second half of the 19th century), “James A.H. Murray: The Man Who Was Monday-to-Monday”—what a field for a fertile mind, what a joy for an inquisitive reader! In any book on word hunters, some space should clearly be allotted to the Swiss linguist Wilhelm Oehl. Thus, “Wilhelm Oehl and the Butterfly.”
Butterfly is a puzzling word. Why butter? The few best-known hypotheses have been recycled in countless popular articles and books. Perhaps butterfly, originally the name of the yellow species, such as the cabbage butterfly or the brimstone butterfly, was extended to all lepidopterous insects. Variations on this theory are that the first butterflies one sees in spring are butter-creamy or that butterflies come in “butter season,” from March to November. Or vats containing butter have special attraction for butterflies. Or their excrement resembles butter. Or the popular superstition has it that witches turn into butterflies and steal milk and butter (hence the name). Or butterfly is a “corruption” of flutter-by. In the past, some of those who had the courage to translate Dutch boterschijte (the main support for the excrement etymology) into English bemoaned the degradation of the butterfly: from Psyche (in Greek, psyche is both “soul” and its incarnation “butterfly”) to a butter-shitter. Except for the flutter-by joke, none of the conjectures cited above is fanciful. The butterfly is at the center of numerous superstitions the world over, and in some parts of Germany it is called “milk thief.” Schmetterling, the Standard German word for “butterfly,” sounds so much like the Slavic word for “(sour) cream” (consider the family name of the Czech composer Smetana, the author of the opera The Bartered Bride) that it is usually classified with borrowed words, and we get one more tie between butterflies and dairy products. Russian babochka “butterfly” (nearly the same form elsewhere in Slavic) seems to be a diminutive of baba “(old) woman,” a doublet of babushka “grandmother”—a fact that seems to strengthen the alleged connection between witches and butterflies. (In Russian, both nouns are stressed on the first syllable.) Romance scholars also have trouble explaining the origin of their word for “butterfly,” such as French papillon (Latin papilio), Italian farfalla, and Spanish mariposa. And this is where Oehl comes in, his net cast widely.
A learned linguist, Oehl studied what he called elementary, that is, primitive, word formation. He was inspired by the gifted but controversial scholar Alfredo Trombetti, whom he called a genius. Today we know that languages form families and that every family goes back to a proto-, or mother language. For example, English is a member of the Indo-European family. It remains unclear who the Indo-Europeans were, where they came from, and how they managed to spread from Ceylon to Norway, but the fact that such a family exists is certain. Other families are smaller, and their early history is sometimes more transparent, but regardless of size, comparison within families follows the same principles: related forms (cognates) are recognized because their meanings match, while their sounds coincide or differ according to certain rules. For example, if a word of Latin begins with d, its English cognate should begin within t, as happens in duo ~ two, and so on. For this reason, Engl. dome cannot be a cognate of Latin domus, and indeed, dome is a loanword from Romance; real cognates of domus are tame (“domesticate,” so to speak) and timber (the material of which houses are made). One can see that real cognates tend to differ in their sound shape; perfect look-alikes are seldom related. Yet words occur that, although they sound amazingly similar, violate sound correspondences, and cannot be looked upon as borrowed (the languages in question have never been in contact), must be in some way connected, and Trombetti considered a broader framework than a set of mutually isolated families. Oehl examined hundreds of such words and ascribed their similarity to people’s identical verbal reaction to certain movements. For instance, when speakers, wherever they live, catch a ball, they cry out gop, kap, hop, or something like it. Such words may be short-lived, but after they disappear, they are coined again. Creating them is like following an instinct. This is Oehl’s “primitive word formation.”
The Old English for butterfly was buttor-fleoge, for all intents and purposes the same word as today. Since in English –fly is added freely to the names of various winged creatures (dragonfly is one example of many), only buttor– “butter” has to be explained. While investigating the names of the butterfly from all continents, Oehl noticed that it tends to be a symbolic representation of the butterfly’s opening and closing its wings. The “ideal” names are simple reduplication: popo, pepe, pupu, dudu, buom-buom, kupu-kupu, firi-firi, and so forth. Some reduplicated words have additional syllables: bebele, palpalut, liblikas, kukupo, yakukek, odefufu, and the like. We immediately recollect Russian babochka (-ochka is a suffix), Latin papilio, and Italian farfalla. Papilio is not an exact counterpart of bebele (the vowels in it differ from syllable to syllable), and in farfalla -r and -l alternate, but those look like insignificant complications. The main difficulty with Oehl’s lists is that classic types never exhaust his material, and “mixed types” are pressed into service. Peme and lablok are slightly less persuasive than pepe and liblikas, while pete (disyllabic), mbudi, futie, and naphitka do not even make one think of reduplication. However, Oehl was convinced that butterfly does not owe its origin to butter. He had shown that once words like babochka arise, they tend to be reshaped or reinterpreted under the influence of folk etymology: babochka, naturally, aligned itself with baba, milk-stealing witches appeared on the scene, and the word’s “primitive” origin was forgotten. According to Oehl, butter- in butterfly should be understood as an alteration of the “mixed” pete ~ bete type. The cause of the alteration might be the belief that butterflies steal butter or the color of their excrement, or whatever. The implication is that at one time the insect was called approximately buto or boto. Such hypotheses cannot be proved, but Oehl’s material makes his reconstruction not entirely improbable. He also listed a type based on the combination m + t, widespread in Slavic, and believed that German Schmetterling is a borrowing of some such word. Initial sch-, he suggested, was added to the met-word later, to connect it with smetana. Schmetterling retained the status of a loanword from Slavic, but its history emerged in a new light.
It will cause no surprise that Romance etymologists accepted Oehl’s conclusions, for farfalla and papilio are such obvious cases of reduplication, while those who attempt to trace the origin of Schmetterling and butterfly ignore them. Oehl was the creator of a most impressive panorama, but, not unlike an oil painting, its effects are lost if the onlooker stands too close to the canvas. No one will bet on his derivation of butterfly, but, who knows! it may be right, for, when all is said and done, we should admit that butterflies have nothing to do with butter. To an etymologist, beautiful, fluttering butterflies are sometimes more dangerous than mad dogs: they don’t bite but are harder to catch.
Anatoly Liberman is the author of Word Origins…And How We Know Them. His column on word origins, The Oxford Etymologist, appears here each Wednesday. Send your etymology question to firstname.lastname@example.org; he’ll do his best to avoid responding with “origin unknown.”