You Could Look It Up: The Reference Shelf From Ancient Babylon to Wikipedia

by Jack Lynch

  AREA: 21,000 ft2 (1,963 m2); 22,600 ft2 (2,113 m2) including supplements

  WEIGHT: 110 lb. (50.2 kg); 119 lb. (54 kg) including supplements

  As a young man, de Vries was fascinated by all things Dutch, especially the literature of the Middle Ages. He had studied classics at Leiden University, earning a Ph.D. in 1843, and became a private tutor in classical studies at Leiden. In 1846 he took another position, as a teacher at the city’s grammar school, but he was determined to bring the study of Dutch to the level of scientific inquiry that had been achieved in the classical languages. In 1849 he published De Nederlandsche taalkunde, beschouwd in hare vroegere geschiedenis, tegenwoordigen toestand en eischen voor de toekomst (Dutch philology: past, present, and future), and was appointed professor of Dutch language and literature at Groningen University, but a few years later he returned to Leiden, where he remained for nearly forty years, until his retirement in 1891.

  One of de Vries’s contemporaries wrote that “Language is the soul of the nation, it is the very nation itself,” and de Vries was so impressed with that formulation that he made it the motto for the entire dictionary.2 He valued the historical study of a living language, for in a living language he could discern the true spirit of a nation. But there was something special about Dutch. De Vries believed that the Romance languages were now “benumbed” in the people’s minds, but that the Germanic languages were still vigorous, still vital. De Vries’s most important predecessor as both lexicographer and Leiden-based philologist, Matthijs Siegenbeek, believed that the point of studying the Dutch language was to achieve eloquence. Not so de Vries, who thought it was to achieve a scientific understanding of the language, including the old, the hackneyed, the awkward, “just as to the botanist the most insignificant weed is as important as the most splendid flower.”3 By paying attention to the language in all its registers, he hoped to extract the laws that guided its development—and that could be achieved only through a solid grounding in linguistics. In this de Vries was channeling the spirit of August Schleicher, a German linguist who had made similar calls for a scientific approach to language.

  The practical work on the WNT began in 1852, with the publication of de Vries’s Ontwerp van een Nederlandsch woordenboek (Proposal for a Dutch dictionary). De Vries was originally planning a dictionary only of contemporary—which is to say nineteenth-century—Dutch, with some background going back to the early seventeenth century. He was eventually persuaded to cover the whole of the modern form of the language, and to that end he built up a database that eventually reached 1.7 million quotations.

  In 1863, de Vries teamed up with Lamert Allard te Winkel, another philologist eleven years his senior. Te Winkel lived only another five years, so he had a short tenure with the WNT, but the work he did was fundamental. He developed a new systematic approach to spelling. Most of the European languages had achieved a high degree of standardization in their orthography by the mid-nineteenth century, but Dutch was still chaotic. The system worked out, still known as the de Vries–te Winkel spelling, became official—mandated by law—in both countries.

  The fruits of their efforts appeared in 1864, when the first part, covering A–Aanhaling, came out. But it was not an entire volume, just a fascicle. Another eighteen years passed before the first full volume appeared, A–Ajuin, in 1882. Already the character of the dictionary was visible. Each entry provided information on the part of speech and, for nouns, the gender and the plural forms; historical information, with minimal speculation on etymological questions; definitions, given in Dutch; and quotations drawn from Dutch literature. Unlike the Grimms, with whom he traded ideas, de Vries decided to forgo Latin definitions in favor of the vernacular. There was information on pronunciation only in difficult or ambiguous cases, but no dates of quotations appeared in the text proper, though a long bibliography provided them.4 De Vries also included usage notes, and he was not shy about offering value judgments: from time to time he declared his own opinion was better than widespread usage, even the usage of the best authors. Nonstandard dialect was systematically excluded. He laid out policies, not always strictly followed, for including or excluding foreign borrowings; borrowings from French that he considered insufficiently Dutch often did not make the cut.

  De Vries retired in 1891 at age seventy, having reached only the word gitzwart ‘pitch-black’. The project was taken over by the Instituut voor Nederlandse Lexicologie, under the aegis of the Dutch Language Union, and over the next hundred years a series of editors worked on it. One of the most important was Jacoba van Lessen, who joined the editorial staff in 1929 and was promoted to chief editor in 1946, a position she held until her death in 1951, making her one of the very few women at the helm of a major reference project.

  The Woordenboek has a prominent place among the other great historical dictionaries of that era, the Deutsches Wörterbuch, the Oxford English Dictionary, and Émile Littré’s Dictionnaire de la langue française; of the three, it is the largest, but also the least consistent in carrying out its policies.5 Over the course of the dictionary, for instance, the period it covered changed several times: the starting date for quotations was originally 1637 (the publication of the Statenbijbel, the most important Bible translation), then it got pushed back to 1580, then even further to 1500; the ending date was always the present, but the present kept extending, and when it reached 1971, it was suddenly pushed back to 1921. The nature of the source material also changed over time. De Vries imagined a dictionary of literary usage, but the dictionary was eventually opened to nonliterary sources: legal writing, diaries, letters, and so on. Even foreign and dialect words that are widely used in Dutch-speaking countries made their way in. In the 1940s and ’50s a supplement appeared, bringing the early part of the dictionary into line with the new editorial policies.

  At long last, 134 years after the first volume appeared and 146 years after the dictionary was proposed, the team presented the fortieth volume, Zuid–Zythum, to Albert II of Belgium and Queen Beatrix of the Netherlands on June 16, 1998. Three supplementary volumes, bringing the early parts of the dictionary up to date through 1976 and bringing the page count up to 49,255, appeared in 2001. It was the work of five generations.

  The story of the Oxford English Dictionary begins in 1857. Johnson’s was still the reigning English dictionary, and it had appeared in dozens of editions. Webster was also taken seriously, especially in the United States, and was in some ways superior to Johnson—his definitions are often more precise, and of course he was able to include words from the seventy-five years that came between Johnson’s Dictionary and his own. Another major English dictionary had appeared in 1835–37, written by Charles Richardson, a protégé of the eccentric philologist John Horne Tooke. Richardson adopted Horne Tooke’s strange notion that every word has one and only one meaning, which can be determined by looking at the etymology. He therefore wrote a dictionary without definitions. With the space he saved, he expanded the etymologies, the subject closest to his heart. That most of those etymologies were pure bunk did little for the quality of the dictionary.

  By 1857, Johnson’s Dictionary was too old, Webster’s too weak in his etymologies, and Richardson’s too eccentric to serve the growing English-speaking world. Worst of all, none of them was designed on a scientific plan. The Germans had shown what was possible when scientific method was brought to bear on a language. So in Richard Chenevix Trench’s lecture that year before the Philological Society, later published as On Some Deficiencies in Our English Dictionaries, he demanded a proper historical dictionary of English, from the earliest days to the present. At its heart should be quotations from English literature. Johnson, of course, had included quotations, and Webster followed, but neither of them was concerned with finding the first and (in the case of obsolete words) last occurrences, or of showing how senses developed over time. Remembering Passow’s edict that “Every word should be made to tell its own story,” Trench argued that “the study of language is … the most p
otent means of all for planting in us the true past of our country; and of this it is proposed in great part to deprive us by those who would make our Dictionaries the representations merely of what the language now is, and not also of what it has been.”6

  The Philological Society released a Proposal for the Publication of a New English Dictionary in 1859, and in it the debt to Passow was explicit. Passow’s mantra about words telling their stories was again invoked and expanded upon:

  In the treatment of individual words the historical principle will be uniformly adopted;—that is to say, we shall endeavour to show more clearly and fully than has hitherto been done, or even attempted, the development of the sense or various senses of each word from its etymology and from each other, so as to bring into clear light the common thread which unites all together.7

  Macmillan was to be the publisher, but they had little idea of what they were signing on to. They imagined a moderate-sized dictionary, perhaps four quarto volumes, and were prepared to devote ten years to the project. Little did they know the length would be exceeded by a factor of four and the time by a factor of seven. The lexicographers needed a quarter century simply to get to the letter A. The snail’s pace eventually led Macmillan to back out and to hand the project over to Oxford University Press, which had more experience dealing with overextended scholars.

  The work—titled simply A New English Dictionary—was plagued by a series of problems in the early days. The first editor, Herbert Coleridge, was the grandson of the poet and a literary boy wonder. Unfortunately, he was never given the chance to show what he was capable of, because he died almost as soon as he took the reins, having just turned thirty years old. The next in line was Frederick Furnivall, a distinguished lawyer and the secretary of the Philological Society. While he loved establishing clubs and societies of every description, he could not bring himself to do any real lexicographical work, and what little he did in those early days was sloppy.

  The next editor, though, was a winner: a forty-two-year-old Scot named James Murray. (Later he would pick up a knighthood and a pair of middle initials, becoming Sir James A. H. Murray, but the initials do not seem to stand for anything.) While still a young child, Murray taught himself to read several languages, and he kept up the passion the rest of his life. As he boasted, “I at one time or another could read in a sort of way 25 or more languages.”8 Murray and his team decided to produce a historical dictionary, covering the English language from the year 1150, traditionally recognized as a boundary between Old and Middle English, to the present. Words that had died out during the Old English period would be excluded, but anything that survived past 1150 would go in. None of the other historical dictionaries presumed to cover so many centuries.9 The Grimms dealt with a canon of about three centuries, and de Vries was initially interested in the Dutch language over just two centuries, though the scope was widened to four. But the OED was dealing with a continuous literary tradition more than a thousand years long.

  The dictionary was prepared according to a much-expanded version of Johnson’s method. Everything had to be based in actual written sources, so the first step was to read widely in English. Words would be included and defined based solely on how they had been used in writing by others; the compilers would resist the urge to single out “good” or “bad” words, and on principle they refused to make up or tinker with quotations. All of those things were characteristic of Johnson’s Dictionary, but the OED editors were determined to be as close to comprehensive as possible. And this involved reading exceedingly widely—a reasonable approximation of everything in the English language.

  TITLE: A New English Dictionary on Historical Principles; Founded Mainly on the Materials Collected by the Philological Society

  COMPILER: Sir James A. H. Murray (1837–1915)

  ORGANIZATION: Alphabetical, a to zyxt

  PUBLISHED: Oxford: Oxford University Press, 1 Feb. 1884–19 April 1928; supplement, 1933

  VOLUMES: 12 + 1 vol. supplement (1933); 4 further supplements (1972–86)

  PAGES: 15,487; 21,217 including supplements

  ENTRIES: 252,200, defining 414,800 word forms

  TOTAL WORDS: 38 million; 52 million including supplements

  SIZE: 13″ × 10¼″ (33 × 26 cm)

  AREA: 14,200 ft2 (1,328 m2); 19,500 ft2 (1,820 m2) including supplements

  WEIGHT: 145 lb. (66 kg); 198 lb. (90 kg) including supplements

  PRICE: £52 10s. to £57 15s., depending on binding

  LATEST EDITION: 2nd ed., 1989, in 20 vols. and 21,730 pages, running to 59 million words; 3rd ed., in progress, online

  No single human being could hope to read all this, of course, so Murray organized a small team of professional lexicographers, who in turn energized a vast team of volunteer readers, and they were able to do an impressive approximation of the task. The readers were encouraged to read everything they could get their hands on, with occasional specific requests and advice coming from the editors; it would hardly make sense, after all, to have everyone reading Hamlet. As they read, they were told, they should keep an eye out for interesting words, and not merely obscure ones; the dictionary needed instances of the usage of ordinary words such as tree and walk and little. Whenever the readers saw something worthy of notice, they were told to write the word in the upper left corner of a slip of paper, to transcribe the whole quotation below it, to provide a precise citation, and to send the slip to Murray’s office in Oxford. This office, a prefab building, was fondly known as the Scriptorium, an echo of and tribute to the monastic scriptoria that Cassiodorus had developed nearly a millennium and a half before.

  Over the course of five decades the list of volunteers grew to more than a thousand people, and they submitted around 5 million slips. These, combined with the more focused reading project carried out by the in-house lexicographers, formed the raw material from which the dictionary was compiled. Once they had sorted the millions of quotation slips into pigeonholes, first by headword and then by year of publication, the lexicographers worked through the alphabet, pulling all the slips for a given word. By reading through the accumulated evidence they determined how many different senses a word had. Murray’s approach to defining owes much to Johnson and Webster, both of whom believed in long and precise definitions, with many meanings broken out into numbered senses. Definers can be “lumpers,” who try to cover the whole range of meaning in a small number of broad definitions, or “splitters,” who propose a separate definition for every slightly different meaning. Richardson was the ultimate lumper, insisting every word had just one meaning, which he believed would be obvious from context. Murray, on the other hand, was a card-carrying splitter, preferring a large number of distinct definitions.

  Murray believed that a proper historical dictionary should tell the story of every word, and every story should start at the beginning. He took seriously Trench’s recommendation: “the first authority for a word’s use in the language which occurs should be adduced; … the moment of its entrance into it, … the register of its birth, should be thus noted.”10 (Trench was thinking of both Passow’s principle and John Jamieson’s Etymological Dictionary of the Scottish Language, which in 1808 was the first to offer coverage of the birth of every word.) And so Murray resolved to track down the first written occurrence of every word in the English language, along with at least one citation for every sense for each century, and, in the case of obsolete words, the last known citation.

  How to recognize, in the Victorian era, the first occurrence of a word was a challenge. Trench was a keen reader, and he was among the first to state a principle that has guided lexicographers every since: “But if it be thus desirable to note in every case, so far as this is possible, the first appearance of a word, then all those tokens which will sometimes cleave to words for awhile, and indicate their recent birth, ought also to be diligently noted. None are more important in this aspect than what one may fitly call ‘marks of imperfect naturalization.’ ”11 In other words, readers
were asked to look for signs that a word was still novel to the audience. When someone read in one of Lord Chesterfield’s letters of 1768 that “I feel what the French call a general mal-aise, and what we call in Ireland an unwellness,” the phrase “what the French call” suggested that the English did not: it is the first known occurrence of malaise. Likewise, when a reader of W. Abney’s Treatise of Photography (1878) came across the sentence “The next lens … is what is known as a ‘wide angle’ doublet,” the telltale phrase “what is known as” suggested that the term was still unfamiliar to readers, and so it appeared as the first occurrence of wide angle.

  All these pieces came together to form the dictionary. For 252,000 entries, covering 414,800 words and word forms, Murray and his team provided the headword, along with all its attested spellings and when they were current; pronunciation; an etymology; a series of numbered definitions; and under each definition, quotations to illustrate the word in use. (Whereas the Grimms gave leisurely passages from the great German writers, Murray trimmed his quotations carefully, leaving enough to get the sense of a sentence but no more.)12 Altogether there are 1,861,200 quotations, covering 4,500 works by 2,700 authors.

  The information in each entry includes a pronunciation, then a list of accepted spellings used over the centuries:

  Forms: 3–4, 7 elemens (pl.), 4 ela-, elemente, 5 elymente, 6 elyment, elemente, 4– element.

  Space is at a premium, so single digits indicate centuries (“3–4,” for example, means “current in the thirteenth and fourteenth centuries”). For some words, the list of spellings can go on for many lines. Then comes a discussion of the word’s origin. Curt, even single-word etymologies are usually adequate, but when necessary, Murray provided long and detailed descriptions of a word’s derivation: