Empire of Words: The Reign of the ‘OED’ 
by John Willinsky.
Princeton, 258 pp., £19.95, November 1994, 0 691 03719 1
Show More
Show More

In these columns six years ago, among a chorus of praise for the new, revised Oxford English Dictionary, OED2, Charlotte Brewer entered a dissenting opinion (3 August 1989):

The riches stored in the two previous dictionaries are now much more accessible, and reviewers have already reported how delightful it is to browse through one of the 20 handsome volumes of OED2, and ponder, marvel or cavil at individual words and definitions among the extraordinary wealth of examples recorded on its pages. But what can such sample browsing tell us about the project as a whole, which purports to provide a systematic, reliable and comprehensive history of the English vocabulary? To make an assessment of the dictionary in its entirety, we need to judge its editorial premises and methodology. This boils down to three separate factors, crucial to lexicographical enterprise: the nature of the sources consulted to supply evidence on word usage, the thoroughness and accuracy with which these sources were read, and the use made by the lexicographers of the evidence which the source study provided. Unfortunately, it is surprisingly difficult for the average OED2 user to form an accurate opinion of any of these three factors.

According to Brewer, ‘only one study of OED1 makes any attempt to subject it to thorough-going methodological examination.’ (OED2 is a fusion of the 1933 OED1 and the post-1957 Supplements.) The study to which Brewer refers is a ‘little-known but revelatory book by Jürgen Schäfer on OED documentation’. Brewer endorsed Schäfer’s caveat that ‘the increasing discrepancy between the methods used at that time’ – i.e. when OED was compiled – ‘and those used now for evaluation calls for a detailed analysis of the nature and reliability of the OED documentation itself.’

Schäfer’s call, and Brewer’s echoing of it, are amply answered in Empire of Words. No longer will scholars use OED in the same way that Scrabble players use the Shorter Oxford and Webster, as a Bible on matters linguistic. Nor, one imagines, could a post-Willinsky scholar do what Raymond Williams did in the Fifties and base a major intellectual enterprise on the confident assumption that OED represents ‘a systematic, reliable and comprehensive history of the English vocabulary’ – raw lexical material. In the Foreword of Keywords Williams writes:

One day in the basement of the Public Library at Seaford, where we had gone to live, I looked up ‘culture’, almost casually, in one of the 13 volumes of what we now usually call the OED: the Oxford New English Dictionary on Historical Principles. It was like a shock of recognition. The changes of sense I had been trying to understand had begun in English, it seemed, in the early 19th century. The connections I had sensed with ‘class’ and ‘art’, with ‘industry’ and ‘democracy’, took on, in the language, not only an intellectual but a historical shape. I see those changes today [1976] in much more complex ways. ‘Culture’ itself has now a different though related history. But this was the moment at which an inquiry which had begun into trying to understand several urgent contemporary problems – problems quite literally of understanding my immediate world – achieved a particular shape in trying to understand a tradition. This was the work which, completed in 1956, became my book Culture and Society.

Willinsky’s title indicates his parti pris – the OED is as imperially tainted a production as the Raj. He reads the dictionary for its reigning ideology, picturing the 13 volumes as a great Victorian flagship, carrying with it, ‘as flagships do, the full weight of a nation consumed with the struggles of democracy at home, imperialism abroad, and a culture seeking to mediate the eclipsing of religion by the gospel of science’. The imperial foundations of the OED, as Willinsky uncovers them, are part racist (Max Müller’s pro-Aryan doctrines are discovered lurking beneath the floorboards), part nationalist (hence the pre-eminence, as cited sources, of Shakespeare, laureate of England, and Scott, laureate of the Union), and chauvinistically masculine. Jesperson’s luckless observation that ‘the English language seems positively and expressly masculine, it is the language of a grown-up man and has very little childish or feminine about it’ is hauled out for the flaying that such unreconstructed opinions can expect in the Nineties. Willinsky detects in the OED a consistent political agenda. All those 1,827,306 citations were mustered to establish ‘that in English literature there is an unbroken succession of authors from the reign of Alfred to that of Victoria, and that the language which we speak now is absolutely one, in its essence, with the language that was spoken in the days when the English first invaded the island and defeated and overwhelmed its British inhabitants.’

Willinsky’s thrust at the OED is three-pronged. He first recalls the origins of the dictionary as a coalition between the utopian Philological Society, based at the University of London, the amazing sticktoitiveness of James Murray, editor and mastermind of the project from 1879 to his death in 1915, and the (initially) nervous patron, OUP, who inherited the project 22 year after its launch. Willinsky tells much the same story as that given by Elisabeth Murray (Caught in the Web of Words, 1977) and Peter Sutcliffe (in his 1978 ‘informal history’ of OUP), from their biographical and house-historical perspectives. Willinsky’s Murray, however, is less the innocent gnome in his funny hat in the scriptorium at the bottom of the garden in the Banbury Road, and more the ideologue, sustained as much by nationalistic zeal as benevolent lexicographical mania.

The predispositions of Murray, his Philological Society adjutants, and his shifting corps of a thousand or so postally-networked volunteer helpers are anatomised by Willinsky for the false consciousness in which they undertook their great task, and their collective (if well-meaning) betrayal of the project’s initial aim, to cite ‘every word occurring in the literature of the language’. The ‘literature of the language’, that is, not ‘the language’ or, as promised in the dictionary’s title, ‘English’. It is a significant difference. Willinsky demonstrates the chilling effect of restricting citation to printed literary sources, and predominantly the high-literary sources, of which Shakespeare is the loftiest and most-cited. The OED, judged by its incidence of citation, is less a comprehensive inventory of English as it has been written down over the past seven centuries, than the ‘Victorian coronation of Shakespeare as the King of Poets’. Willinsky devotes two chapters to the OED as ‘Shakespeare’s Dictionary’. Elsewhere he notes the exclusion of the vernacular, the dialects of business, trade, music-hall, barrack-room, bar-room and those professional discourses not represented in the old universities. The OED has generally disregarded the lower classes and their rich argots, and relegated women and minorities as providers of source material. (Women, especially genteel spinsters, have, however, always been invaluable as gatherers of citations; the most diligent of 20th-century ‘carders’ was Marghanita Laski, who hoovered up as many as 200,000 words for the Supplements, many of them from the detective fiction which she reviewed in bulk.) Willinsky records the thinness of 18th-century citation, the skittishness about ‘Americanisms’, and systematic omissions – such as the clerks of the 15th-century Court of Chancery and Romantic writers, ‘with the debatable exception of Scott’. These omissions can be explained in terms of the untimely deaths of key contributors, oversights in farming out chores and specifying source texts, and Victorian literary taste. But the main point is that the OED is built on a tendentiously skewed sample.

Willinsky’s second prong is a close scrutiny of representative citations – those with symptomatic evidence on such things as racism. His longest and most subtly illuminating analysis is of the subterranean currents swirling under the various definitions of ‘Jew/jew’, as verb, noun and adjective. The third prong is fieldwork at present-day Oxford, to examine the work in progress which will eventually form into the largely computerised OED3. In his visits to 37 St Giles and his evidently courteous exchanges with recent custodians of the project (Robert Burchfield, John Simpson, Edmund Weiner) Willinsky detects a quaint mixture of ‘afternoon tea and high-speed computer searches’. His conclusion is friendly, but a little condescending: ‘All told, the OED’s literary, prosaic and omitted citations authorise a definition of the English language that was part of a nation’s hegemonic brief in the last century, both abroad and at home. It is currently finding a rich afterlife in a slowly decolonising world.’

The implication in Willinsky’s verdict is that future mutations of OED will be much less self-mystified than their deluded predecessors. The problem is, however, that works of such magnitude are only undertaken by those deluded to the point of fanaticism (like the Mormons and their gigantic genealogical database in Salt Lake City). The cool, post-colonial mood of today’s Oxford enables a sceptical scholar to assess the achievement of Oxford’s dictionary-makers with refreshing clarity. But will it generate energies which are equal to those that were needed for the completion of OED1? Like the soldier, the lexicographer must be prepared to give his life for the great cause, without thought of reward. Who would make that great sacrifice, if they understood that they were merely tools of an imperial ideology which later generations would at best sneer at, and at worst hold up as criminal? Would amateurs with the professional skills of James Murray, Frederick Furnivall, Richard Chenevix Trench – if they exist today – be foolish enough in this post-Thatcherite world to give away the fruits of their labour? Why slave to make future generations of OUP publishers rich?

Nor, with the economic short-termism that they must nowadays work to, can one expect modern academic publishers to take as benignly long a view of the project as OUP did of OED. Not that they exactly knew at the beginning what they were getting into. Initially, when the Press inherited it from the Philological Society, the dictionary was conceived as a ten-year haul. But ‘as late as 1896, after the publication of the letter “D”, the balance sheet was still in the red, with £50,000 invested, compared to £15,000 recovered from sales.’ By this date it was ‘the largest single engine of research working anywhere in the world’. By the time it was finished, in 1928, it was probably the longest-running such engine. Some of the OUP-OED editorial staff had been employed full-time on the project for more than forty years, and one compositor had worked on setting its type for almost sixty.

Serial projects which drag on for decades are notoriously prone to peter out, denying their originators the profits which come with completion. OUP took almost sixty years not to finish the Oxford History of English Literature. The Delegates have since drawn a line under the project, and embarked on OHEL2, under the brisk management of Jonathan Bate. Arden 2 similarly never reached the end of the Shakespeare corpus in the half-century during which its volumes were published. Arden 3, one may be sure, will be pushed along at a much faster clip, driven by the economic imperative for prompt return on Routledge’s investment.

OUP’s patience with the OED must have been handsomely rewarded – in the very long run. The 125 ‘fascicles’ stanched some of the losses over the decades. But with the completed work for sale as an A-Z set in 1933, substantial profits could at last be reaped. The 13-volume 1961 republication of OED advertised itself as merely a reprint. The two-volume 1971 ‘Compact’ edition (with its all-too-necessary magnifying glass) was nothing more than a miniaturised reprint of the OED-1933 stereotype plates, but claimed a new copyright. The ‘Compact’ was hugely distributed, being used as loss-leader bait for book clubs. There were at least five printings in the UK and 24 in America. The nuclear OED copyright was renewed again in 1989, with the seamless integration of the Supplements, and wholesale changes to the phonetic notation. In OED2 OUP has a property which should earn it royalties until 2055, under new copyright legislation.

All this must be very satisfactory to the accountants at Walton Street. But the investment costs involved in assembling OED3 are staggering to contemplate. And with the accelerating pace of technological advance, it is unlikely that OUP will again have seventy years with little more than repackaging costs to factor in. OED3 will need to adapt to an Information Technology and Information Systems environment that is changing faster in a decade than book-based systems did in half a millennium. Put another way, if by time machine you brought Caxton forward to 1933 he could – given a few years and an army of medieval guildsmen – reproduce OED1 using his 15th-century apparatus. However much time and however many hands Caxton had, he could not reproduce OED2 in its CD-ROM form. And not even the young computer visionaries in Silicon Valley can plausibly guess what forms information systems will take twenty years from now. New scanners which allow vast inputs of material, automatic parsing and convergence with the English usage corpora – like the ongoing British National Corpus and the International Corpus of English – are on the horizon. With the quantity of source material growing exponentially, the architecture of OED3, whose nucleus is still substantially based on a library of canonical literary texts, will need to change accordingly. But what will ‘accordingly’ mean? Investment needs to be huge, and it is hair-raisingly risky.

These considerations highlight the peculiar role that OUP now occupies as curator of many of Britain’s more significant cultural monuments – of which the OED and DNB are probably the most significant. The Press originated neither. They took over the lexical dictionary from the Philological Society and the biographical dictionary from the commercial publisher George Smith. With these two projects OUP inherited a curatorial responsibility which has, over the years, become increasingly burdensome. Unlike the National Heritage Commission, OUP has the task not just of conserving important national properties – that is to say, keeping these reference works in print at affordable (but not unremunerative) cost – the Press also has the responsibility of modernising them. It’s as if it were compelled to rebuild Blenheim Castle every few decades, so to speak, using the latest construction materials and techniques.

Such renovation raises sharp strategic and financial issues. How radical should the Press be? With the Oxford Shakespeare, they attempted a ground-up, total remodelling job. Stanley Wells and Gary Taylor were recruited into the Press for a substantial portion of their scholarly careers. The expenditure must have been huge. The results, judging by reviews of the edition, may not have been entirely satisfactory. The Press seems to have had better luck with a smaller-scale renovation, that of Paul Harvey’s Oxford Companion to English Literature (1932). Margaret Drabble’s 1985 fifth edition retained a substantial nucleus of the original text (all the many Scott entries, for instance), representing, as well as one can estimate, about 60 per cent of the whole – revised but pretty much in its original form. On this Harveyan base, a contemporary superstructure was erected. The mixture proved amazingly successful, propelling OCEL to mass-market sales, well beyond the traditional reference library market. So, too, with the new DNB, currently being undertaken under the general editorship of Colin Matthew. According to promotional material there will be a general tidying up and an integrated supplement of new contributions, amounting to about 40 per cent of the work. These will spread the biographical coverage of women (principally), business-people, and those who previously fell outside the DNB’s definitions of what constituted national significance. DNB2 will appear in traditional book and contemporary electronic forms. The judicious mixture of formal conservatism and novelty is symmetrical with the 60/40 nature of the contents. From the requests broadcast through the columns of learned journals, it looks as if this policy of 60/40 supplementation, integration, and adaptation (but not surrender) to new electronic forms is the way in which OED3 is moving forward.

Another interesting question is how these huge endeavours, and others like them, will be paid for in future and how gifted scholars can be induced to devote themselves to such work, even at the faster tempo in which they are completing. The current regime of four-year research assessment has made academics chary of involving themselves in projects which do not produce immediate and tangible pay-offs. There is little career advantage in being even a key member of a team with, at the end of it all (the end being as much as a decade in the future), group billing alongside a hundred others and a few hundred pounds for many more hundreds of hours of work.

Send Letters To:

The Editor
London Review of Books,
28 Little Russell Street
London, WC1A 2HN


Please include name, address, and a telephone number.


Vol. 17 No. 15 · 3 August 1995

In his review of John Willinsky’s Empire of Words: The Reign of the ‘OED’ (LRB, 8 June), John Sutherland asserts that no ‘post-Willinsky scholar’ could ‘do what Raymond Williams did in the Fifties and base a major intellectual enterprise on the confident assumption that the OED represents a “systematic, reliable and comprehensive history of the English vocabulary" – raw lexical material’. Sutherland describes Williams as possessing a ‘confident assumption’ concerning the OED. But Williams was at least as explicit as Willinsky evidently is in describing the shortcomings of the OED. He writes that he could summarise its major shortcomings ‘in three ways’.

I have been very aware of the period in which the Dictionary was made: in effect from the 1880s to the Twenties … this has two disadvantages: that in some important words the evidence for developed 20th-century usage is not really available; and that in a number of cases, especially in certain sensitive social and political terms, the presuppositions of orthodox opinion in that period either show through or are not very far beneath the surface … The air of massive impersonality which the Oxford Dictionary communicates is not so impersonal, so purely scholarly, or so free of active social and political values as might be supposed from its occasional use. Indeed, to work closely in it is at times to get a fascinating insight into what can be called the ideology of its editors … Secondly, for all its deep interest in meanings, the Dictionary is primarily philological and etymological; one of the effects of this is that it is much better on range and variation than on connection and interaction … Thirdly, in certain areas I have been reminded very sharply of the change of perspective which has recently occurred in studies of language: for obvious reasons … the written language used to be taken as the real source of authority, with the spoken language as in effect derived from it; whereas now it is much more clearly realised that the real situation is usually the other way round.

In other words, as Willinsky notes twenty years later, the OED excludes ‘the vernacular, the dialects of business’, and so on. It is, notes Williams, a narrowly-focused, ideologically driven work, and its value to scholars, for these and other reasons, is limited. ‘Few inquiries into particular words end with the great Dictionary’s account, but even fewer could start with any confidence if it were not there.’

Wayne Templeton
Surrey, British Columbia

Vol. 17 No. 16 · 24 August 1995

John Sutherland’s review of Professor Willinsky’s Empire of Words; The Reign of the OED (LRB, 8 June) is perceptive in its discussion of the financing and management of long-term scholarly projects. But the reader may be left with the view that Willinsky ‘s own conclusions about the Oxford English Dictionary as a cultural icon should be accepted unchallenged. The problem with being a national institution is that people respond not to the institution itself, but to the idea of institutional authority, which is a very different thing. The OED is an excellent case in point. It is a book which records our language: the meanings of words, their history, their derivation. But meaning is an imprecise concept: dictionaries attempt an approximation of meaning according to the evidence available to the lexicographers and their ability to decipher it. The history of any word is also, in an absolute sense, insecure: the evidence simply is not available to document any word comprehensively. Derivation is an inexact science: for any root word there comes a point in its history at which we have to say that we have no idea why or from whence it evolved as it did.

But the eager reader has been led to expect certainties, in religion, in science, in medicine, and in almost any sphere of life. Although it is a commonplace that succeeding generations overturn the certainties of their predecessors, still the myth of certainty remains. The need to invest authority can be seen everywhere, as can the opposing need in others to topple the established authorities, either gratuitously or to replace them with a new temporary orthodoxy. It would be a disaster if this were to endanger work which is valuable. Criticism achieves a certain authority by the very fact of being in print, but may itself be as Hawed as the work it attempts to describe. The documentation of the English language in the OED is valuable: there is no other work which attempts the same tremendous goal, or which gets so close to achieving it. Authority is relative: errors should be corrected, imbalances eradicated, better information incorporated. But the sporadic undermining of a valuable record is dangerous, when the purposes of scholarship are better served by collaborative assistance. If a definition could be improved, then the scholar should submit an improvement; if new documentation is discovered, then this should be presented; if a point of policy is questioned, a better alternative should be proposed. We are fortunate that the Oxford University Press has shouldered the burden of revising the OED: in many ways the success of the enterprise depends on the contributions of those who are prepared to regard authority as relative.

John Simpson
Chief Editor, OED

Vol. 17 No. 18 · 21 September 1995

John Simpson complains that the attempt ‘to topple the established authorities’ can be disastrous if it endangers valuable work and that ‘the sporadic undermining of a valuable record is dangerous’ (Letters, 24 August). But the attempt to prop up a national institution such as the OED also has dangers, since the defence is itself open to obvious attack. In my own experience, the OED is far from satisfactory, and the new edition is more unsatisfactory than ever. Several scholars have publicly pointed out errors and omissions, and I have found all sorts of serious mistakes in my own areas of interest – mainly social and political thought. Moreover these are not just incidental but essential to the way the work has been done. For example, technical terms in anarchist and pacifist discourse have been traced back not to the specialist publications where they first appeared but to general periodicals where they appeared some time later. One search I made – into the term ‘libertarian’ – had the embarrassing conclusion of finding an article by myself in the Listener of 1969 cited, although it merely followed established usage.

Nicolas Walter
London N1

Vol. 17 No. 17 · 7 September 1995

Professor Sutherland, in his review of John Willinsky’s Empire of Words: The Reign of the ‘OED’ (LRB, 8 June), informs us that ‘no longer will scholars use OED in the same way that Scrabble players use the Shorter Oxford and Webster, as a Bible on matters linguistic.’ I do not know about Webster and the US, and I would not quibble about the Shorter Oxford as a Bible on matters linguistic but, according to the information inside the cover of my copy of the 1993 edition of Chambers Dictionary, it is the official reference dictionary for the UK National Scrabble Championship. Chambers is, therefore, the Bible for Scrabble players.

L.B. McKenzie
West Pennant Hills, New South Wales

send letters to

The Editor
London Review of Books
28 Little Russell Street
London, WC1A 2HN


Please include name, address and a telephone number

Read anywhere with the London Review of Books app, available now from the App Store for Apple devices, Google Play for Android devices and Amazon for your Kindle Fire.

Sign up to our newsletter

For highlights from the latest issue, our archive and the blog, as well as news, events and exclusive promotions.

Newsletter Preferences