Other dictionaries have always been a hugely important influence on and source for the OED. After an Introduction, this page discusses Keeping up with contemporary dictionaries before outlining the disadvantages of relying on lexicographical evidence of past usage. An account of Critical discussion of OED’s use of dictionaries follows, with a final section on Major dictionaries and wordbooks used as sources by OED.
See also our section on Writers and dictionaries.
The Proposal for a Publication of a New English Dictionary (1859) – the dictionary which turned into the OED – had its immediate origin in the study by R. C. Trench on the deficiencies of existing dictionaries. From the start, therefore, the process of constructing the new dictionary by examining entries in other dictionaries was built into the OED’s methodology from the start.1
In other words, what would elsewhere be labelled plagiarism was a routine and essential element in making the OED. This was and continues to be true of all dictionary-making. From Cawdrey’s dictionary of 1604 onwards, monolingual English dictionaries had copied all kinds of content from each other, from bilingual dictionaries, and from other types of word-lists, whether those prepared for educational purposes like Edmund Coote’s English Schoole-maister, 1596, or Mulcaster’s Elementarie, 1582, or as glossarial aids to reading historical texts, as found for example in Thomas Speght’s edition of Chaucer (1598) or E. K.’s glosses to Spenser’s Shepheardes Calendar (1579), as well as much later works.2
In a lecture to the public in 1900, round about the time that his own dictionary had reached the letter J, James Murray, OED’s chief editor (1879-1915), described the result of this process as a ‘lexicographic cairn’, built stone by stone over many centuries.
(T)he English Dictionary, like the English Constitution, is the creation of no one man, and of no one age; it is a growth that has slowly developed itself adown the ages.Murray 1900: 6-7
The OED was the summit of this cairn. It ‘superadds to all the features that have been successively evolved by the long chain of workers’ (p. 47), registering ‘all omitted words and senses’, supplying ‘all the historical information in which [preceding] works were lacking’, and, ‘above all, giv[ing] quotations illustrating the first and last appearance, and every notable point in the life-history of every word’ (p. 46).
One of the most valuable of those features was the word lists themselves (see Initial practice). Naturally however the OED lexicographers were interested in all the other evidence provided by earlier editions too: definitions, illustrative quotations (if any), along with information on etymology, spelling, register and usage.
While preparing a new entry an OED editor began ‘by taking down the appropriate bundle of “slips” from its shelf’ to scrutinize the all the existing information that readers had gathered on the word in question. During this process he would be
continually turning to existing dictionaries – Dr. Johnson, the various editions of Webster, and the most recent supplements – gladly availing himself of any help or hint they offer in the wording of a definition, or in the record of new senses….Much of the toil of sifting and collecting fresh material consists in the examination of the Old English and Middle English dictionaries, the glossaries to early texts, and the concordances to the Bible, Shakespeare, and other poets.Onions 1928a: 15
Murray himself worked at a desk in his Scriptorium which ‘held the large Dictionaries of Johnson, Littre ́, Webster, and the Century open for reference’ (Murray 1977: 298). Webster’s ‘unabridged’ dictionary of 1864 would also have been close at hand. Throughout the 44 years it took to publish the first edition of OED in its entirety, OUP tried hard to keep the OED editors to within a ‘scale’ of comparison with this dictionary: the entries under no individual letter were supposed to exceed the size of the equivalent letter in Webster by more than five and a half times. Needless to say, this limit was breached time after time, with lexicographically positive results, but the comparison with Webster was a recurrent bone of contention between lexicographers and publishers (Brewer 2007b: 25-26).
Borrowing from other dictionaries had many positive aspects, not least in enabling the lexicographical cairn, in Murray’s words quoted above, to achieve considerable heights. However, not all editors were content to have their work pillaged (as they saw it) by others. Noah Webster, whose major dictionary of 1828 had borrowed extensively from Johnson’s Dictionary, engaged in bitter copyright battles to protect his own work, a struggle continued by his successors in the so-called Dictionary Wars. Both Murray himself and his publishers OUP were extremely disturbed by the Century Dictionary (1889–1891), the multi-volume work edited by W. D. Whitney which initially looked to be a serious competitor to the OED enterprise and whose scholarship, Murray feared, might benefit from OED’s preceding labours. In the event, OED’s progress was so slow as to rule out out this possibility altogether.3
Keeping up with contemporary dictionaries
Keeping up with contemporary dictionaries was not such a pressing issue for OED1, as this revolutionary endeavour was so clearly conceived as a work of historical lexicography with no effective rival. However, as just mentioned above, the relatively swift appearance of the multi-volume Century Dictionary in the US was initially a worry to the Press, especially since it was feared that subscribers would be lured away with serious economic consequences for OUP.
Over the twentieth century and since, contemporary dictionaries have influenced OED much more directly. OUP itself began the trend of smaller desk dictionaries of present-day English with its publication of the Concise Oxford Dictionary in 1911, a phenomenally successful publication which encouraged further investment in this new market: the Pocket and Little Oxford Dictionaries followed (1924 and 1930 respectively; see Brewer 2007b: 77-9; Brewer 2015c). Other dictionary houses followed suit. Over the years that Burchfield’s Supplement was being prepared and produced (1957-1986), a number of important contemporary dictionaries appeared whose lexicographical practices challenged OED’s own, notably Webster’s Webster’s Third New International Dictionary in 1961 (Gove 1961). Marketed as ‘the greatest vocabulary explosion in history’, this work treated the vocabulary of colloquial and slang English far more generously than had any general English dictionary in the past. As Burchfield himself described, ‘the sheer quantity of words included in [Webster’s Third] made it apparent at once that I had seriously underestimated the task of collecting modern English vocabulary wherever it occurred. The whole editorial process had to be delayed—in the event by several years – until my editorial assistants and outside readers had assembled evidence on this majestic scale’ (Burchfield 1984: 117).
More than any preceding editions and versions of the OED, today’s OED Online presents itself as the last word on contemporary as well as historical language in English, for example by rushing treatments of new Covid-19 vocabulary to the screen within months of the onset of the virus in the UK and elsewhere.4 Consequently the editors keep a close watch on the policies and practices of other contemporary dictionaries. Occasionally it appears that OED’s decisions on when to include new vocabulary, or new senses of existing words, are influenced by decisions taking by other publishing houses.
A case in point is the inclusion of the non-binary use of the pronouns they, them, and their. In July 2020 this use was still unrecorded in most online (and print) general dictionaries other than the forward looking Dictionary.com. Merriam-Webster had been monitoring the pronouns for some time, as described in their ‘Words We’re Watching‘ blog, and in September 2019 decided to go ahead and include the new sense in their online dictionary. The event was widely reported along with Merriam-Webster’s subsequent choice of they as ‘word of the year’.5 By 3 March 2020, the new sense had also been added to the OED3 entries for they, them, and their, with supporting quotations stretching back to 2009 Twitter evidence in all three cases.6
Disadvantages of dictionaries as evidence of usage
There were (and are) lexicographical disadvantages as well as advantages in dictionaries’ use of each other’s evidence. Many dictionaries list and define words without instancing specific examples of their use, i.e. quotations from verifiable textual sources. This is of course contrary to one of the founding principles of the OED, that the definitions of words should be derived from examples of their use in ‘real’ contexts: i.e. not meta-textual ones.
This screenshot from part of OED3’s entry for allowable, updated from OED1 in September 2012, reproduces typical entries from two heavily cited seventeenth-century dictionaries, Huloet’s 1552 Abecedarium, an English-Latin dictionary, and Baret’s 1574 Alvearie, a dictionary of English, Latin and French with occasional illustrations from Greek (alveary is derived from the Latin word for bee-hive). As can be seen, both dictionaries offer no more than glosses which simply state what a word means rather than providing credible evidence for its current usage and examples of how it is used in context.
Where possible, therefore, the OED uses contextual rather than dictionary quotations to substantiate a word’s use. It is not clear why both these two quotations were carried over from the original OED1 entry given the 1561 quotation from a source of ‘real’ usage, and the fact that all three quotations are so close in date.
A second problem with citing dictionary evidence was that entries were often copied from one work to another without any evidence that the vocabulary recorded had ever genuinely formed part of the language. A special term, ghost word, was coined by Murray’s friend W. W. Skeat in 1886 to refer to vocabulary of this sort, i.e. ‘Words which have no real Existence’. Skeat urged that ‘We should jealously guard against all chances of giving any undeserved record of words which had never any real existence, being mere coinages due to the blunders of printers or scribes, or to the perfervid imaginations of ignorant or blundering editors’ (Transactions of the Philological Society (1885-7: 350-1)).
OED1 tried to avoid risks of this kind since it aimed always to print quotations as evidence of usage – indeed, quotations were, in Murray’s words, the essence of the work, forming the methodological basis of the new Dictionary. Nevertheless OED1 (and hence OED2) did include a number of headwords without evidence of usage outside dictionaries, generally in the earlier stages of the Dictionary when quotation evidence was still often deficient.
Examples of possible ‘ghost words’ still present in OED Online today include
- agroten (v), ‘to surfeit’, illustrated only by the reproduction of an illustrative example in an entry from the English-Latin dictionary Promptorium Parvulorum (1440); the entry was revised in 2012 and this remains the only example of use (as of September 2020)
- communy, v, ‘to communicate’, illustrated only by an entry from Palsgrave’s Lesclaircissement (1530); the entry was revised in 2009 and again this remains the only example of use (as of September 2020)
- craches, ‘pimpernel or chickweed’, this time derived from an unrevised entry originally published in 1893, which again relied on Palsgrave’s Lesclaircissement, which in turn had relied on Cotgrave’s French-English dictionary of 1611. Today’s OED entry marks the entry with a cross to indicate the word is obsolete and prints the same single illustrative text as in 1893, expanding the previously abbreviated reference to Palsgrave (but not to Cotgrave): ‘J. Palsgrave Lesclarcissement 210/1 Craches herbe, movron [Mouron, mourron = pimpernel, also chickweed (Cotgr.)].
This chain of dictionary evidence unsupported by ‘real’ usage can also be found in some of the entries OED1 took from Johnson’s Dictionary (see Major dictionaries below) – e.g. besputter, carrotiness and carcelage, all supported by listings in other dictionaries as well as Johnson’s.
Google searches suggest that all of the words listed above have only very rarely if ever appeared outside a dictionary: i.e. they had no real claim to be included in the OED in the first place.
A third problem with dictionary evidence for a word’s use is that successive editions of any individual word book often retain a word long after it has ceased to be in general use, without at the same time marking it as obsolete (or obsolescent). In this way obsolete words can be passed on from one dictionary to another over years or even centuries, with no indication that they are not still current in the language, a phenomenon one might call ‘entry inertia’. Somewhat surprisingly, entry inertia can be found in the OED itself, which in past and present forms has long listed words as current whose last recorded date of use – itself taken from a dictionary reference – was many years before.
This can be seen in today’s OED in the treatment of nealing, meaning ‘the process of tempering’ a substance. The word was last recorded in ordinary (or ‘contextual’) use in OED1 in 1799, but the actual last quotation was from Andrew Ure’s Dictionary of Arts. Manufactures, and Mines, published in 1839: ‘Annealing or Nealing, a process by which glass is rendered less frangible, and metals..are again rendered malleable’. Given that OED1’s entry for nealing was published in 1906, that suggests the word had already fallen out of use by the 1900s (and possibly by 1839, the date of Ure’s dictionary).
This OED1 entry, along with thousands of others, was carried over unchanged into OED2 in 1989 and in June 2003 revised for OED3. The two quotations already cited (the 1799 one and Ure’s 1839 entry) were retained as the penultimate and last quotation respectively, from which one can infer that the OED3 lexicographers searched for post-datings (i.e. subsequent evidence of use) but did not find any. In turn, that looks like pretty clear evidence that the word is now obsolete (a judgement supported by internet searches, as any reader can confirm). It is clear therefore that nealing has not been in use since 1799, and the word’s retention in Ure’s 1839 Dictionary was probably the result of inertia. As of September 2020, however, the OED3 entry labels nealing not ‘obsolete’ but rare’.
Examples of all three problems identified above can be found by browsing OED editions in print and online form – unfortunately the OED Online search tools in their current form (September 2020) do not allow systematic ways of identifying such words (it is not possible to search for entries with only one quotation, for example, or by date of last quotation).
In contrast, some other ‘dictionary’ words in the OED turn out to have been used in non-meta-linguistic contexts too. See next section.
Critical discussion of OED’s use of dictionaries
The principal work here is Jürgen Schäfer’s Early Modern English Lexicography, posthumously published in two substantial volumes in 1987. While Starnes and Noyes’s seminal work on early monolingual dictionaries had demonstrated these works’ dependence not only on each other but also on earlier wordbooks and bilingual dictionaries, Schäfer showed that they had also drawn heavily on glossaries to literary and other works. So not all of the strange words that found their way into early monolingual dictionaries were learned freaks found only in meta-linguistic treatises of one form or another. Instead, some ostensibly ‘dictionary’ words had genuinely been used in literary and translated contexts: they were not limited to specialist word-books alone. Accordingly, it would be worthwhile for a future revision of OED to trace such contextual uses of words.
One striking example of the value of such research, taken from a post-Early modern dictionary entry that was copied into the OED without supporting quotations, is atroscopy. As shown in the screenshot above, OED1’s entry – borrowed without any other evidence from Johnson’s Dictionary of the English Language of 1755 – was carried over without alteration into OED2.
When the OED3 revisers tackled this word in March 2012, however, they found that the word had been used in a ‘real’ context over a hundred years before Johnson and had continued in occasional use ever since:
Such fruitful research must characterise many other OED3 entries. At the moment we can only come across them by happenstance, however, since there is no systematic way of comparing the new OED3 entries with their equivalents in OED2 and OED1.
Major dictionaries and wordbooks used as sources by OED
Two of the most important dictionaries influencing the OED were Samuel Johnson’s A Dictionary of the English Language (1755) and Charles Richardson’s New Dictionary of the English Language (1836-7). Our EOED sections on Johnson and Richardson respectively supply more information on these works.
Johnson’s is the only dictionary on which figures are available to quantify the influence concerned. Describing how Johnson’s work influenced not only the choice of headwords in OED1 but also its wording of definitions and inclusion of quotations, the former Director of OED, Penny Silva, reported in 2005 that 769 definitions had been carried over from Johnson’s work into the first edition of OED (attributed to ‘J’ or ‘Johnson’), along with a total of 2,976 quotations, several of them supplying the sole example of use not only in Johnson but also in the OED. Among the latter she instanced astroscopy, which as we have seen was later discovered by OED3 revisers to be much more widely used, along with the so far under-evidenced besputter, carrotiness and carcelage – these three also supported by other examples of dictionary use (Silva 2005: 234-7; see Disadvantages above).
Many other dictionaries have been extensively mined by OED but are not always acknowledged in its text, often because their content has been independently accessed and verified, and on occasion corrected or amplified. These works include the Dictionary of Old English and Middle English Dictionary as well as the Dictionary of the Older Scottish Tongue, Green’s Dictionary of Slang and many contemporary works (see Keeping up with contemporary dictionaries above).
Dictionaries and wordbooks continue to be heavily quoted in OED3, with many citations carried over from OED1. A selection of titles is listed below, in each case with their tally of quotations in OED Online as of September 2020. Copies can be consulted on Lexicons of Early Modern English (at https://leme.library.utoronto.ca), EEBO, ECCO, and the Internet Archive.
- Promptorium parvulorum (‘storehouse for children’) (1440). First Latin-English dictionary. Extensively quoted in OED1, presumably from the Camden Society edition by A. Way (1843-65), which was based on the Harley 221 MS. OED Online currently (September 2020) indicates that the work is quoted 5,526 times in all in OED, which now cites the Harley MS itself as source text. Strangely the work is not listed in the OED Online list of top 1,000 most quoted sources.7
- John Palsgrave, Lesclarcissement de la langue francoyse (1530). French-English dictionary. Sept 2020: 5,173 results in 4,076 entries. See discussion of OED’s use of this work at 1500-1699 in OED1/2 and compare its OED3 ranking (June 2019) in Chart 9
- Richard Huloet, Abecedarium Anglico Latinum (1552). Sept 2020: 1,870 results in 1,686 entries
- John Florio, Worlde of Wordes (1598). English-Italian dictionary. Sept 2020: 1,818 results in 1,704 entries
- Robert Cawdrey, A Table Alphabeticall (1604, 1609, 1616). The first monolingual English dictionary. Sept 2020: 354 results in 353 entries
- Randle Cotgrave, A dictionarie of the French and English tongues (1611, 1632). Sept 2020: 5,816 results in 5,091 entries
- Henry Cockeram, The English dictionarie; or, An interpreter of hard English words (1623, 1626, 1632). Sept 2020: 1,668 results in 1,615 entries
- Thomas Blount, Glossographia (editions from 1656-1674). Sept 2020: 3,516 results in 3,403 entries
- Nathan Bailey, The universal etymological English dictionary (1727) and other dictionaries. Sept 2020: 4,573 results in 4,329 entries
- The Century Dictionary, 1889–1891. Sept 2020: 4,926 results in 3,992 entries
Numerous other reference works have been and continue to be quoted in OED, presumably because they are such an easily accessible source of evidence notwithstanding the limitations on their value discussed in Disadvantages above. Most notable among these is the Encyclopedia Britannica, whose huge number of quotations – coming up to 15,000 in 2020 – put it In the top five of the OED’s most cited sources. Other encyclopedic works are also high on this list, not just the two other works detailed below but also publications such as Andrew Ure’s A dictionary of arts, manufactures, and mines (various editions from 1810; Sept 2020: 1502 results in 1177 entries), Edward Knight’s dictionaries of mechanics (1873-1891; Sept 2020: 4,201 results in 2738 entries) and Thomas C. Allbut’s System of Medicine (1896, 1899; Sept 2020: 3,298 results in 2726 entries).
- Encyclopaedia Britannica (successive editions from 1768 onwards). Sept 2020: 14,718 results in 10,933 entries
- Ephraim Chambers, Cyclopædia; or, an universal dictionary of arts and sciences (1728). Sept 2020: 3,703 results in 3,083 entries, though this includes later editions
- The Penny Cyclopaedia (1833–1858). Sept 2020: 4,052 results in 3,410 entries
Last updated on 7 September 2020
- Trench 1857, available on Google Books [accessed 2 September 2020].
- See further Starnes and Noyes 1991, first published 1946, to which Schäfer 1989 provides invaluable additional and contextualising material, Green 1996, Osselton 2009.
- For Webster and the Dictionary Wars, see Micklethwait 2000 and Martin 2019; on OED’s rivalry with the Century, see Bailey 1996, Bailey 2000a: 2017-20 and Gilliver 2016: 2012-17 and elsewhere (as listed in Gilliver’s index).
- See https://public.oed.com/blog/circuit-breakers-ppes-veronica-buckets-world-englishes-covid-19/, https://public.oed.com/blog/using-corpora-to-track-the-language-of-covid-19-update-2/, https://public.oed.com/blog/july-2020-update-scientific-terminology-of-covid-19/ [all accessed 3 September 2020].
- See https://www.merriam-webster.com/words-at-play/woty2019-top-looked-up-words-they; https://time.com/5679132/merriam-webster-gender-neutral-dictionary/ [accessed 3 September 2020].
- All three OED entries are dated September 2013, though in each case the latest quotation for the sense under discussion is 2019.
- Advanced search 2 Sept 2020 in ‘Quotations’ for ‘Promptorium Parvulorum’ in ‘Quotation Title’ yields ‘5,526 results in 4,716 entries’ .