I CONSULT THE OXFORD ENGLISH DICTIONARY almost every day. The binding on my first edition, the last installment of which was published in 1928, is disintegrating. Shreds of vellum flutter onto desk or carpet every time I open one of the 12 massive volumes, which can weigh as much as 15 pounds. Because I'm researching the history of the OED, I need to compare the first edition with the second. But truth be told, I also have a sentimental attachment to these cream-colored pages, stained by age and use, with the complex yet clear patterning of each element in an entry (headword, pronunciation, etymology, definition, quotations), which James Murray, the first chief editor of the OED, designed to be "eloquent to the eye."
Bibliophilic considerations aside, however, the OED Online is my dictionary of choice. This remarkable resource displays both the second edition of the OED, published with great fanfare in 20 volumes in 1989, and the gradually accumulating third edition, begun in 2000 and due to be completed some decades hence. The great value of the OED's third edition is that it is the first revision ever undertaken of this vast dictionary. The 1989 edition merely spliced the first edition with supplements produced during the previous two decades, but it did not venture to revisit the outdated Victorian and Edwardian scholarship of its elderly parent. That makes OED3, as aficionados call it, the hottest English-language lexicographical product around.
But could the online edition spell the end of the OED as we know it? Earlier this year the OED's U.S. editor told The New York Times Magazine, "We have about 20 years' more work to do revising and adding entries. Who knows what will happen with technology in 20 years? We certainly don't." Bibliophiles and technophobes greeted this remark with intense anxiety, speculating that the OED's publisher, Oxford University Press, would never issue a printed edition of the OED again.
At first, I wondered whether the sackcloth and ashes were warranted. True, books do furnish a room, and the 20 volumes of the second edition of the OED fulfill tiffs purpose admirably. (The photographic reductions of the OED with which many dictionary lovers are familiar--two volumes for the first edition and three for the second, accompanied by a magnifying glass--aren't on the same scale, but still look quite handsome on the shelf.) The fact is, however, that the OED Online is the last word in space saving and portability, as well as lexicography. And it is now so much easier to look up words. Instead of determining which of the 20 volumes you need, pulling the heavy tome off the shelf, finding an uncluttered and sturdy surface on which to lay it, and fumbling through the pages for the right entry, you can tap the keyboard and skip blithely from one end of the alphabet to the other in the blink of an eye, finding 10 words in the time it used to take to track down one.
So why all the hand-wringing about the loss of this unwieldy behemoth? Is it the sheer physical substance of this great work, the size and heft of it, that makes the prospect of its disappearance into the ether a cause for alarm? That's part of the answer, not least because the OED's history is one of agonizingly slow emergence into physical form. Reviewing the second edition in 1989, the novelist William Golding cast his mind back to its heroic and idealistic origins: "In the high days of Queen Victoria a dictionary was conceived, not to say dared, which matched her iron bridges, her vast ships and engines." A characteristically Victorian project, the OED set out to encompass the entirety of the English language, recording within its pages every single word. Of course, the editors had to relinquish this ideal fairly quickly. Such a thing was as impossible then as it would be now, even with all the electronic aids we have to hand. The vocabulary of English, as of all languages, appears to be infinitely variable.
To many of those who contributed to the first edition of the OED, from around 1860 onward, their task must initially have seemed endless. The editors hoped to read as much as they could of everything that had ever been printed in English, relying on armies of volunteers to scour libraries and private collections and to write down, on slips of paper measuring four by six inches, the words they thought were worthy of inclusion, along with the authors and rifles and dates of the works in which they had found them, and the sentences in which those words appeared. These quotation slips, thousands of them, were posted back to the editors and stored haphazardly in sacks and boxes. Only when James Murray took over the editorship in 1879 were they thoroughly sorted and filed in pigeonholes. Not one page of the dictionary had yet been published, and it took another five years under the methodical and painstaking Murray before the first installment appeared, in 1884, covering the entries A-ant. The dictionary was finally completed 44 years later, by which time much of the earlier part of the edition was out of date and the editors were already compiling a supplement (published in one volume in 1933).
The array of massive volumes on the shelf is literally a monument to this protracted gestation period and to the vast quantity of material from which the OED was assembled. It also testifies to the scope of the English language. Right from the dictionary's first appearance, readers and reviewers loved it for its awesome compendiousness, and its slowly increasing mass gave an appropriate impression of its scholarly substance. The novelist Arnold Bennett reported in 1928 that he had "been buying it in parts for nearly 40 years," and judged it "the longest sensational serial ever written!"
You can get a sense of the OED's intellectual capaciousness just by turning the pages in any of the enormous volumes and casting your eye down the extraordinarily detailed entries. What made the dictionary revolutionary when it appeared, and makes it revolutionary still, is that every definition it contains is based on a study of the empirical data: those masses of original quotation slips that recorded a word's use in real historical sources from 1150 to the present day. From these scraps of paper (eventually numbering more than five million for the first edition) the lexicographers constructed their picture of the history of a word's usage from the beginning of its life to its end--from cradle to grave, as they themselves said.
A dictionary, wrote the French man of letters Anatole France, is "the universe in alphabetical order." Perhaps above all others, the OED encourages the idea that it contains everything that has ever been thought or said by anybody speaking English, and is hence a record of the language's culture and history. As one reviewer wrote of the OED in 1899, "Everything is to be found here, but one feels that human faculties are inadequate to penetrate the details of so vast a collection." Virtually every entry of the OED munificently displays quotations of real historical usage, often derived from the works of the great writers in the language--Chaucer, Shakespeare, Milton, Pope. The quotations are the reason the OED is so very long, and they are certainly the reason it is the greatest dictionary ever written (and such a fabulously good read).
Understandably, lovers of the OED find it alarming that this record of human labor, which stretches over so many years and records such a vital aspect of our culture, might sublimate into a form without physical dimension. Still more alarming is the notion that the latest, and best, edition of the great work exists only on the Internet. Even OED Online enthusiasts concede that many things immediately evident in the printed book are obscured or not apparent on the screen: the length of an entry, which may stretch over several columns and pages; the relationship between one entry and its neighbors; the variation in page count among the letters--in the first edition, C has a massive single volume all to itself, the same size as the one devoted to all of V, W, X, Y, and Z.
It's hard to avoid the conviction that such an accumulation of knowledge and erudition should have a physical dimension: The dictionary's sheer size is important to our understanding of its value. As his poetry attests, W. H. Auden was a lifelong lover of the OED, and visitors to his flat in New York City often remarked on the battered copy of the dictionary that took up so much space on his shelves. After he moved to Austria in 1972 he kept it in his otherwise sparsely furnished workroom, except for the volume he used to sit on at table, as if (so one guest reported) he were a child too small for the nursery furniture. Schoolchildren and college students who encounter the great work electronically have no way to fathom its physical might and bulk, and its weighty difference from all those other online dictionaries-whether today's Merriam-Webster or its forebear, Noah Webster's 1828 American Dictionary of the English Language, which in its online versions is sadly stripped of all quotations save those from the Bible.
The fate of Webster's dictionary points up another problem with Internet reference works. You can never be sure of what you are getting. Material can be silently subtracted and added. The editors add a new tranche of words and entries to OED3 every quarter. These take two forms: first, slow and steady revision of each old entry (so that they might avoid cut- l ting their teeth on the vagaries of the first edition's treatment of A--on which the 19th-century lexicographers cut their teeth--today's revisers began in the middle of the alphabet, at the letter M); and second, new words and corrections from across the alphabet. No need for subscribers to squeeze a new volume of the revised work onto their shelves next to existing volumes. Instead, each fresh batch of cutting-edge scholarship miraculously materializes on everyone's screen. In June subprime made its way into the OED for the first time, with the current meaning attested from 1993. So did cookie cutter and wantaway, a British word usually used to describe a professional soccer player who wants to transfer to a different dub. Anyone familiar with the old way of doing things, when dictionary revisions took decades to appear or came out piecemeal in printed supplements, will agree that the swiftness, convenience, and neatness of Internet production is simply wonderful.
The new OED may be a revision-minded editor's dream, but it's trickier for readers seeking a truly definitive definition. The entry you consult in January may be different by March. The editors first revised the entry for make, one of the most complex verbs in the language, in 2000. On several occasions since, they have made changes--I can't give you chapter and verse, because the first version, and all the subsequent ones through June, have been expunged from the record. (I printed out the entry on two occasions, around 2002. The first time it came to about 98 pages, the second to about 102. But I mislaid these piles of paper, and now they are lost to me forever.) Scholars find this evanescence upsetting and infuriating; even the casual reader may find it disconcerting. By contrast, the printed book is (more or less) permanent and unchanging. If a new edition supersedes the old, the old does not disappear.
Still, even those who caress their dog-eared OEDs must acknowledge that the dictionary is flawed. For instance, why did it take coo/de cutter so long to reach its pages? The new June entry records the word's first use in 1864, and as early as 1922 it was being used in a derived sense--"Characterized by homogeneity or lack of originality; conformist, unimaginative, generic"--indicating that the initial meaning was firmly established in the language. (The OED's evidence is a quotation from the Chicago Sunday Tribune: "There are always 'cookie cutter' tendencies among us. One of these this year is the caracul trimmed coat which every other woman in New York wears.") But the OED lexicographers passed the word by: Inevitably, in building so vast an edifice of scholarship, they have sometimes missed the occasional brick As one of them wrote in 1951, when he was trying to persuade the publishers to take on the expense of revising the OED afresh, this greatest of dictionaries, despite its public reputation for unimpeachable authority, has "hosts of wrong definitions, wrong datings, and wrong crossreferences."
So the electronic OED enables comparatively easy correction of past errors on the one hand, and swift addition of new words and usages on the other. But it also does something just as important, undreamt of by the OED's first makers. However much we may lament the loss of the material book in all its comfortable solidity, those thousands of pages of dense print were largely impregnable to any kind of systematic analysis. In the wake of digitization we can, for the first time, bring to light, and utilize, the rich linguistic and literary treasures previously scattered piecemeal among individual entries. In one respect, alphabetically organized dictionaries, or encyclopedias, are arranged arbitrarily: All that data--the quotations themselves, and information about etymology, pronunciation, definitions, spelling forms, and so on--is ordered not according to sense or date or provenance but by the letter with which the headword begins. Now we can run successions of searches and see all the words first recorded in the language in 1599 or 1776 or 1968, or all the quotations from Emily Dickinson, or all the hapax legomena (one-off coinages) that the second edition of the OED quoted from James Joyce's Ulysses. (The total for the last of these is 54, down to 44 in OED3 because the lexicographers have recently found fresh examples of these words, some from earlier sources and some from later--meaning that they aren't hapax legomena any more.)
What's more, we can begin to assess the nature of the primary information from which the OED was constructed--its quotations--and the inevitable biases of selection and interpretation that went into its making. We might guess that both the Victorian and Edwardian lexicographers favored a particular literary canon from which to draw their quotations. This, as it turns out, was the case: The most quoted individual sources in the OED's first edition were Shakespeare, Chaucer, Milton, Walter Scott, the Bible, and a medieval historical poem called Cursor Mundi. So were these works the giants that constructed the English language? Or were they the ones the lexicographers most delighted in quoting from?
Digitization of the OED has been an extraordinary gift, enabling us to better understand both the strengths and the weaknesses of the dictionary, as well as to look up lots of words quickly. This is due to the search tools Oxford University Press has provided, and to the enormous expense it lavished, in the 1980s, on transforming the physical object into electronic form in the first place. (All the different elements in each entry were electronically tagged, so that they could be subsequently retrieved according to different taxonomic criteria. Keying in all the information, and checking it, took well over 200 person-years.)
Does a more comprehensive understanding of the OED compensate for the loss of those handsome volumes? Will it really matter if the OED is never printed again? Given the OED's likely length (or size) when complete, would any of us be able to afford it if it were? On balance, I remain convinced that the advantages of digitization dwarf the disadvantages, for scholars, and even more so for the thousands of people who now have electronic access to this dictionary--many more than could ever have been envisaged for the printed form--whether at home or at an academic institution. (In the United Kingdom, you can access the OED free of charge at your local library; in the United States, an individual subscription costs $295 a year.) The OED's transformation is one more example of the democratization of knowledge in the digital age.
To say this is not to dismiss the attachment to books as mere sentiment. We are now, involuntarily and unceasingly, it often seems, assailed by a superabundance of electronic information, which can confuse and repel as much as it enlightens us. By contrast, when we pick up a book, we are making a deliberate choice that is limited to the contents between the cover, and we can see, feel, and smell what we are getting. As a material object, a book bears its own physical history of use (whether our own or other people's), without which, arguably, we cannot fully comprehend its social and cultural significance. I recognize that electronic resources can never replicate the range and character of experiences that accompany the consultation of a printed volume, and I can see the argument that the switch to digital resources is dehumanizing. Nevertheless, I think the trade is worthwhile, if not for all books then certainly for the OED. Anyone who knows and loves this work, one of the greatest of human endeavors, must agree that the more fully and intensely we can engage with its contents, the better. Indeed, that is the best way to repay the successive editors and contributors for the years of devotion they have poured into it. Whether or not the OED is printed again, the computer tells us more about this extraordinary intellectual achievement than those heavy volumes will ever do. So let us embrace digitization, not deplore it.
CHARLOTTE BREWER is the author of Treasure-House of the Language: The Living OED (2007). Her website Examining the OED, at http://oed.hertford.ox.ac.uk/main, publishes recent research and historical material on the OED. She is a fellow of Hertford College, Oxford, and professor of English language and literature at Oxford University.