Policy and practice

Summary information on the treatment of literary sources in OED’s successive editions can be found on this page under the headings OED1, Supplements, OED2 and OED3 respectively. 


One of the principal founders of the project for the revolutionary new dictionary that turned into the OED, R. C. Trench, had been clear in his lectures on ‘On some Deficiencies in our English Dictionaries’ that ‘the ideal dictionary should be an ‘inventory’ of the language…It is no task of the maker of it to select the good words of a language. If he fancies that it is so, and begins to pick and choose, to leave this and to take that, he will at once go astray….He is an historian of [the language], not a critic’ (Trench 1857: 4-5).

Nevertheless, the notion that the OED was, or should be, a guardian of the nation’s language and a treasure-house of its great writers was early established and proved extremely difficult to resist. The 1859 Proposal which shortly followed Trench’s lectures made the new Dictionary’s expected reliance on literary sources clear, and although the main editor of OED1, James Murray, subsequently took a broader view of the lexicon when he became editor in 1879, for example in his enthusiasm for newspapers, he had to struggle hard against repeated pressure from the Oxford University Press Delegates to confine his sources to ‘great writers’.1

As can be seen by glancing through any page of the Dictionary literary quotations play a substantial role in furnishing the evidence for a word’s progress from ‘cradle to grave’, thus enabling ‘every word…to tell its own story’, as described by the first editor Herbert Coleridge, following Passow; see 19th-century historical lexicography). The reading lists and appeals sent out by the editors at various stages in the compilation of the first edition are dominated by literary works, and no doubt this reflected the reading preferences of the volunteers on whose labour the OED1 relied. See further Literature and the nation, OED1 compilation, and Initial practice, also The canon, Top sources, and pages in Period coverage.

The dominance of literary sources in the OED was additionally acknowledged by the lexicographers themselves when they explained that ‘all the great English writers of all ages’ were the first port of call for the OED’s quotations (Murray 1928, vol X, p. v), and referred to the 1933 Bibliography of quotation sources as a ‘guide to English literature’ (Murray 1933, Preface). And the intimate relationship between great writers and language more generally was emphasized in Oxford University Press’s publicity drive for the completed work in 1928. This claimed that the OED was ‘a Dictionary not of our English, but of all English: the English of Chaucer, of the Bible, and of Shakespeare is unfolded in it with the same wealth of illustration as is devoted to the most modern authors’ (quoted from the opening statement of OUP’s journal The Periodical, 15 February 1928).


No serious concern for literary sources is evidenced in the 1933 Supplement. But the editor of the 20th-century Supplement, R. W. Burchfield, admired the strong literary bent of the parent dictionary and considered it part of his job to quote and define the language of the great poets and prose writers of his day. In his view the OED was a ‘literary instrument’ (see Brewer 2007b: 184ff). Thus he

  • defied Oxford University Press’s internal objections to his inclusion of eccentric poetic vocabulary and usage in his 1962 sample material for the Supplement (e.g. T. S. Eliot’s ‘loam-feet’, from ‘East Coker’ in Four Quartets) (Burchfield 1989: 11-13)
  • believed that ‘the failure of descriptive linguists ‘to quote from the language of even our greatest modern writers, leave[s] one looking at a language with one’s eyes partly blindfolded’ (Preface to OEDS4: x)
  • several times named poets and writers whose ‘special vocabulary’ he felt deserved inclusion, writing in the same Preface that ‘OEDS [i.e. his Supplement], like its parent work, has been hospitable, almost from the beginning, to the special vocabulary, including the once-off uses, of writers like T. S. Eliot, Virginia Woolf, D. H. Lawrence, and others’

See further Literature and the lexicon, The canon, Auden in OED Supplement.


Although OED2 contained almost no new material – the edition simply merged the unrevised OED1 with Burchfield’s four-volume Supplement, adding a further 5,000 entries for new words – it was energetically marketed by its publishers Oxford University Press as a work of literary significance. A publicity brochure reproduced a series of small portraits of twelve quotation sources, calling them ‘”Contributors” to the OED’: Chaucer and Shakespeare, John Locke and John Wesley; Johnson and Noah Webster, Austen; Charles Darwin, Mark Twain, Thomas Hardy, Virginia Woolf, James Joyce.

OED2 publicity brochure, 1989 (author’s own copy)

Evidently, none of these individuals had in reality contributed to the OED; their works had simply been quoted in its entries – in varying quantities, ranging from c 33,300 times for Shakespeare to 227 for Woolf, the only female author mentioned in the entire brochure.2 See further our pages under Top sources, also Writers and dictionaries.

As the brochure makes clear, Johnson and Webster were included in the portraits for the quotations from their own dictionaries subsequently incorporated into the OED. Literary writers were the backbone of Johnson’s Dictionary: his seven most quoted sources include Shakespeare, Dryden, Milton, Addison and Pope – those seven between them furnishing nearly half the quotations in the work. Johnson’s major influence on the OED is treated in our Johnson pages under Historical background; see also Major sources in 1700-1799 in OED1 (Period coverage). The influence of successive editions of Webster and the Webster imprint has been less clearly literary; for example, Webster’s Third International Dictionary played a crucial role in opening up Burchfield’s eyes to the importance of covering colloquial vocabulary and slang – i.e., for the most part, non-literary vocabulary – in the 20th-century version of OED. This is a major shift in OED policy which continues vigorously today.3


OED3’s editorial policy on literary sources was indicated in the Preface to the Third Edition published online in 2000, written by the then chief editor John Simpson. Observing that earlier editions of OED had been criticized for their ‘apparent reliance on literary texts’, Simpson stated that ‘closer examination of earlier editions shows that this view has been overstated, though it is not entirely without foundation.’ He went on to explain that the text so far revised was already making use of ‘many non-literary texts which were not available to the original Victorian readers and their immediate successors, particularly social documents such as wills, inventories, account books, diaries, journals, and letters’. These and similar sources were allowing the editors ‘to provide a fuller picture of the vocabulary of (especially) the Early Modern period’, while ‘further reading of similar sources [would] doubtless result in additional significant discoveries’.4

Notably, however, OED3’s description of its new Reading Programme refers to continued excerpting of literary as well as non-literary sources:

In addition to the ‘traditional’ canon of literary works, today’s Reading Programme covers women’s writing and non-literary texts which have been published in recent times

‘The Reading Programme’, OED Online at https://public.oed.com/history/reading-programme/ [accessed July 2019]

and the Third Edition has certainly continued to add literary quotations to the Dictionary, including from male canonical authors already generously represented. For example, by June 2019, OED Online had added over 1,000 additional quotations from S. T. Coleridge, Daniel Defoe, and Charles Dickens, along with over 800 more from Samuel Richardson and just over 540 more from Dryden – all of them poets or novelists already liberally quoted in OED1, with totals ranging from c 9,350 (Dryden) to c 2,800 (Richardson). These writers are all in the top 100 of the June 2019 OED Online list of 1,000 most quoted authors and sources. The only female author to figure in the top 100 is also literary: George Eliot, to whose OED1 tally an additional 380-odd quotations has been notched up.5

Other literary sources have also greatly expanded. James Joyce’s stock of quotations has leapt from 1,749 in OED2 to 2,472 in OED (as of June 2019); Jane Austen’s from 1,044 to 1,733.

First 18 items on the list of ‘Top 1000 sources in the OED’ after the June 2019 update. Source: screenshot from OED Online 5 August 2019

To be sure, the OED3 revisers would appear to have added far more quotations in total from non-literary sources than from literary ones, most obviously (according to their top-1,000 list, from which the first 18 items are shown in the image above) from (multi-authored) newspapers and periodicals. So far, however – after nearly 20 years of revision – this has not reduced the prominence of creative writers in their 1,000 most quoted authors and works list. For more details, see our pages on Top sources.

The problem is the same as that identified in the EOED Leverhulme study of 2009 (as discussed in the second half of our page on Men and women compared). OED3 is a revision of OED1 and has not re-made the Dictionary from scratch. In consequence, it has inherited the huge quantities of literary quotations gathered for the first edition. It cannot dislodge this literary bias without either discarding many of these quotations – an option regarded as unthinkable, it would appear – or rebalancing the Dictionary by collecting even greater numbers of non-literary quotations from single-authored sources.

Last updated on


  1. See K. M. E. Murray 1977: 221ff.; a page on quotations from Newspapers in the OED is currently in preparation.
  2. Hardy certainly used and read the OED (see Taylor 1993), and possibly James Joyce too (Deane 1998); Woolf’s active use of OED, if it occurred, has yet to be demonstrated; she refers only to the Concise Oxford Dictionary. See Fowler 2002, and for Burchfield’s Supplement quotations from Woolf, Brewer 2009. On Webster see Brewer 2008.
  3. On Webster’s Third and its influence on Burchfield, see Brewer 2007b: 167-8; on the importance of American slang for the first Supplement, Brewer 2007b: 66-7 et seq; on the role of Webster’s ‘Unabridged’ Dictionary of 1964 on the ‘scale’ of OED1 entries, Brewer 2007b: 25-6, 32, 33, 34.
  4. Quoted from John Simpson, section on Documentation in Preface to the Third Edition, 2000 [accessed July 2019].
  5. OED1/2 data is taken from EOED’s pre-2010 searches of the OED2 database available on the OED website up to 2010, and must therefore be regarded as approximate; see Sources of OED data. The specific figures are as follows: Coleridge: 3,577 quotations in OED1/2, 4,731 in OED Online as of June 2019 (OEDO); Defoe: 4,288 in OED1/2, 5324 in OEDO; Dickens: 8,200 in OED1/2, 9,229 in OEDO; Richardson: 3,606 in OED3; Dryden: 8,805 in OED1/2; 9,347 in OEDO; George Eliot: 3,490 in OED1/2, 3493 in OEDO.