Murray's filing system (OUP Museum)
Enter Keywords:
Thursday, 31 January 2013
Home arrow OED editions, updates and revisions arrow Review of OED3
Review of OED3
(Adapted excerpt from Brewer 2004, reproduced with permission.) Note that this article reviews an early stage of OED3, now replaced with the re-launch of the OED Online website in December 2010. For information about this latest stage of the OED, see our pages on OED Online and OED3 and  Re-launched OED Online.

The Oxford English Dictionary, 3rd edition, edited by John Simpson; published online at www.oed.com

This new edition of the Oxford English Dictionary (OED3) is a major leap forward in lexicographical practice and representation, unparalleled in the OED's history to date. The website offers a substantial range of information and resources, from extensive archival material on the history of the OED to sophisticated search options. The main elements of the printed dictionary's page and entry lay-out (designed by J. A. H. Murray, editor of OED1) are by and large preserved, in traditional lexicographical form, but users can choose to turn these features on or off – pronunciation, spellings, etymology and quotations – so as to clear the screen if desired. An additional optional feature, very useful, is a datechart which represents the distribution of quotations in graphic form, so that large gaps (or not) in documentation spring immediately to the eye.

The website fully exploits the new possibilities of viewing, accessing and cross-referencing information that the electronic medium provides, and there are many imaginative and helpful options such as the ability to view contiguous lemmas listed by date or by entry. The help menu includes a tour (free to non-subscribers) explaining quickly and accessibly how to use its main features. The site is constantly updated and supplemented, so that even a frequent user can be surprised by new features and new material appearing on-screen.

A particularly valuable feature of this edition and its ongoing revisions is its historical transparency: changes to the site are recorded as they made (at http://dictionary.oed.com/help/updates/ [accessed 8 August 2005]). This allows one to see at a glance how the revision is progressing through the alphabet, and to browse both revised and entirely new entries. If one takes the view, for example, that the OED is not an impartial record of language and its users, but instead a selection governed by all sorts of individual factors relating to the lexicographers concerned and the conditions under which they worked – including their own views on and assumptions about language and what their job of dictionary-compiling entailed – then one would want as much information as possible about who did what to the Dictionary and when. This is something OED Online implicitly recognizes by documenting its own actions in this way, and also by clarifying the bewilderingly various stages through which OED has passed since the completion of R. W. Burchfield's twentieth-century Supplement in 1986 – six different forms altogether, representing the progression from the printed 1989 second edition to the online medium, via various stages of addition and revision (see http://www.oed.com/public/guide/citing.htm [accessed 8 August 2005]). Helpful as this information is, however, the differences between these stages (up to OED3, that is, which completely recasts all entries) are nothing like as substantial as those between the first edition and Burchfield's four-volume Supplement, the undifferentiated merging of which took place in OED2. This means that anyone interested in the first edition's treatment of vocabulary, as compared with Burchfield's, continues to need the printed versions both of OED1 (including the 1933 Supplement) and of Burchfield's Supplement to hand, so as to be able to check whether apparently first-edition material, whether in the definitions or quotations, is truly that, or was added by Burchfield – and also so as to be able to distinguish between Burchfield's additions and those of the OED2 compilers.

So how do the new entries compare with the old? Examining the various different versions side by side – or rather, successively, since simultaneous viewing of the various chronological states of the Dictionary is not practicable on a standard-sized screen – indicates that the third edition of OED is delivering the root and branch reworking of the first edition (1884-1928) which has been so long required. A wholesale revision has taken place. The semantic structure of each entry has been reconsidered and in many cases recast, so that identification of the various senses of a word may be partially or completely different. In all necessary cases, which in practice means almost all cases, surviving definitions have been rewritten in contemporary English, replacing the late nineteenth- or early twentieth-century locutions that now look quaint, out-dated, and/or, for one reason or another, unsuitable.

The recasting of entries seems often to have been driven by the accumulation of substantial numbers of additional quotations. Every student of pre-contemporary texts will have come across examples of words which ante-date OED's quotation evidence, and many users have provided the lexicographers with supplementary material and corrections since publication of the dictionary began in the last decades of the nineteenth century. For the first time, this evidence has been drawn on and incorporated into the new edition, and also, evidently, backed up with independent research.[1] As a result of this process, the body of quotations has been reconfigured, with some alterations (mostly additions, and much less frequently subtractions), so as to illustrate more tellingly the words and senses exemplified, and also to even out the chronological spread of attestation. Existing quotations have been in many cases re-dated. In general, there has been a major drive to tidy up and standardise bibliographical referencing – no insignificant task.[2] Other categories of treatment – for example, on spelling forms, etymology, phonology, word labelling, and the occasional explanatory headnote – have all been overhauled and in many cases expanded, changed, or altogether rewritten. The result is that this edition of the OED has left clear water behind it. Readers wanting to get a representative idea of the scope and detail of the sweeping changes that have been introduced may wish to examine old and new versions of, for example, magic, martian, martyr – or indeed make, possibly the most substantially rewritten and reworked item to date (although the changes made to this verb are not susceptible to casual analysis, since the entry is too massive and too detailed to be apprehended on screen).[3]

Perhaps most pressingly of all, OED3 is engaged in bringing the record up to date. The huge number of entries for words still current, but for which the latest quoted evidence was still a hundred years or more out of date, has now – over the revised range that is – virtually disappeared. This anomaly had been a regrettable feature of OED2, belying its claims to be 'authoritative', 'up to date', and 'comprehensive', since for its twentieth-century component it had relied almost exclusively on Burchfield's evidence as printed in the four Supplement volumes of 1972 to 1986.

But Burchfield's brief had been only to identify and record new words (or new senses of existing words) omitted by the first edition of OED, or emerging only after its publication, and not to update the quotation record of words for which there was at least nineteenth-century evidence – for as he himself pointed out, 'for the earlier letters of the alphabet such a policy would have entailed the addition of late-nineteenth-century or of twentieth-century examples for virtually every word and sense listed in the Dictionary'. Consequently, OED2 's attestation and treatment of pre-existing words – i.e. the bulk of the English lexicon – during the course of the twentieth century was extraordinarily thin, and its reprinting of out-of-date definitions sometimes disconcerting or even absurd. (An egregious example is the definition of the Conservative Party as 'one of the two great English political parties' – the other being, evidently, the Liberal Party. First published in 1891, this statement has been faithfully reproduced in every edition of the OED since, up to and including OED2).[4] This major defect in the OED has now begun to be remedied. Entry after entry in the revised alphabet range displays new quotations from twentieth- and indeed twenty-first-century sources, repairing the gaps left in documentation for the last 100-odd years. Taking the range m-monnisher as a sample, and using OED Online 's search mechanism, it is possible to calculate that, whereas Burchfield supplied 19,563 quotations for this section of the alphabet, the new revisers have supplied 37,639. In other words, for this portion of the alphabet they have nearly doubled the number of quotations from recent sources.[5] It is difficult to overstate the value of this material (or the extent to which it was overdue).

Burchfield went on to say, 'Our policy depends upon the realization by users of the Dictionary that any word or sense not marked "obs." or "arch." is still part of the current language'. This advice, however, was unsatisfactory. One has only to turn over a few pages of OED2 to find example after example of words quite unfamiliar to a mid- to late-twentieth-century user, not marked "obs." or "arch." but quite clearly not 'part of the current language' of the time. Most of these (so it appears) are now being caught by the third revision. The lexicographers have presumably looked for and failed to find subsequent quotations for numerous such words documented in OED1 with nineteenth-century quotations but left untouched by Burchfield in the 1970s and 1980s, and consequently also by OED2 in 1989 (e.g. magiric, magirist, magism, magnase, magnetiferous, maidenism, manificative, and countless others), and at long last applied the label 'obsolete' (in some instances, especially where Murray or his co-editors marked the word 'rare', it may well have been obsolete for a hundred years or more). The same has been done for those words (far fewer in number) documented in OED1 with only pre-nineteenth-century quotations (sometimes only one), not then identified as obsolete, and also left unmarked by Burchfield (e.g. miskenning a, last quotation in OED1 (& 2 &3) 1608, melonist, 2 quotations in OED1 (& 2 & 3) dated 1629 and 1727).[6]

Conversely, some words labelled archaic in the first edition, and again left untouched by Burchfield, have had the label removed – thus misdoubt (the noun, not the verb), is supplied with three twentieth-century quotations, and misenter now has its single OED1 quotation, dated 1675, sandwiched between one of 1598, and one of 1999 (leaving a very odd gap?)[7] And some words which, one might have thought, should have been labelled obsolete or rare in the second edition have now been shown to have had a new lease of life. Thus magnanerie, a silk-worm house, had two quotations (1887 and 1885) in OED 1, was passed over without comment by Burchfield, and now in the third edition is demonstrated as having been both earlier and later used, with additional quotations from 1835, 1966 and 1969.[8]

The range and variety of differences and revisions between OED3 and its predecessors make it difficult to form a clear idea of the character of the new Dictionary as it unfolds before us. One way of grappling with this problem is to examine a particular area of revision, however limited, to see what hints it can give us about the Dictionary as a whole. The varying policies on labelling in the OED prove a promising place to start: the most significant change that OED3 revisers have made in this respect may be the avoidance of the terms 'erroneous' and 'catachrestic', which Murray and Burchfield had freely employed to stigmatise objectionable usages, and the corresponding decision not to use the paragraph mark (¶) as an indicator of incorrect usage. This change – along with the introduction of far more consistent and rigorous notes on contested or offensive usage (e.g. s.v. masterful, media, nigger) – indicates that for the first time, OED is succeeding in the often repeated lexicographical aim, originally articulated by Johnson, 'not [to] form, but register the language...not [to] teach men how they should think, but relate how they have hitherto expressed their thoughts'.

For more on OED and labels, see Brewer 2004, 2005, and Mugglestone 2000b.

[1] Lists and accounts of antedatings have been published in many places, notably the periodical Notes & Queries. In 2002 I conducted a sample check of N & Q articles to find almost all the relevant antedatings appropriately dealt with in OED3. I passed on a small list of omitted items to the OED editors, who immediately incorporated them in the Online edition.
[2] As every user will know, the first (and hence second) edition of the OED, appearing as it did over many years, and under often difficult conditions and successive hands, could not maintain consistent bibliographical standards. Thus many works were allotted different dates on the different occasions on which they were cited, and authors and works were often referred to by different titles and/or abbreviations. Burchfield was more consistent than his predecessors, but by no means perfectly so (see e.g. Brewer 1993: 329 n.5). Simpson discusses bibliographical regularization of OED citations at http://www.oed.com/public/guide/preface_6.htm#bib [accessed 8 August 2005].
[3] The OED3 version of martian is evidently superior to that of its predecessors, but should the revisers have recorded its use as a poetic term, current in the wake of the publication of Craig Raine's A Martian Sends a Postcard Home (1979)? Good quality print examples of its usage can be found in a range of standard works on poetic diction and technique (e.g. Blake Morrison and Andrew Motion, eds, The Penguin Book of Contemporary British Poetry (Harmondsworth: Penguin, 1982)), and the OED has in previous versions made a point of recording literary critical terminology, especially in relation to the writer with whom a term is associated. See Burchfield 1989: 70 (on T. S. Eliot) and his Supplement's treatment (reproduced in OED2) of negative capability (s.v. negative 8c), ambiguity (s.v. 3b), practical criticism (s.v. practical 6), and the like.
[4] Burchfield's treatment of political parties (and consequently that of OED2) is not straightforward. There is no entry for 'Conservative Government' or Conservative Party', for example, although the former term occurs 12 times, and the latter 61 times, in the second OED (in definitional text s.v. conservative (n.) 2a but elsewhere usually in quotations), and the party is treated appropriately by the first OED s.v. conservative (adj.) 2a. Labour Party was omitted from OED1 (the 'l-leisurely' fascicle appeared in 1902), but was treated by Craigie and Onions in the 1933 Supplement, who placed the term and its definition in the ragbag category of attributive uses of labour and provided 6 quotations dated between 1886 and 1922. Burchfield reproduced their definition, added four more quotations, and kept the term in the same minor position (sandwiched between labour-pains and labour-relations).
[5] The 'M' fascicles for OED1 were published in 1906-1907, and Burchfield's Supplement volume covering words beginning with 'M' was published in 1976, so any OED2 quotations for words over the range M-monnisher between 1907 and 1976 must have been inserted by Burchfield (barring a few possibly added by the OED2 compilers; I have disregarded these as numerically insignificant). I searched for the date range 1908-1976 in 'quotation date' on OED Online, and counted the number of quotations for the range M-monnisher, to get the number of quotations inserted by Burchfield. I then clicked on the button giving the corresponding quotations for the new edition, and again counted up the number occurring within that alphabet range.
[6] This is how the matter is treated in the Preface to the Third Edition, written by the editor John Simpson and posted in March 2000 – so his figures presumably apply only to the range of revised material then published, i.e. M-mahurat: 'One effect of the extension to the available documentation is to show that many words formerly regarded by the Dictionary as still current are now apparently obsolete (e.g. machopolyp, macrophagocyte, and magnetoscope), and that many which were formerly unlabelled can now be reliably labelled as being rare (e.g. machinization, magnolious, and Mahdism). In all, the revised sample shows 52% more words and meanings marked 'obsolete' and 242% more marked 'rare' than was the case in the equivalent range of the Second Edition of the Dictionary. It is just as important to monitor how and when terms fade from the language as it is to record their arrival' (http://www.oed.com/public/guide/preface_3.htm#documentation [accessed 8 August 2005]).
[7] Surprising in view of the recent rebirth of the verb enter (not yet treated by the revisers) in relation to electronic data – a sense recognized in the rewriting of the definition for misenter to read 'To enter erroneously, esp. in a book, register, database [my italics], or other record'. And couldn't misentering, vbl noun, which OED3 also antedates with a sixteenth-century quotation but for which it retains the OED1 label 'obs.', have been documented with contemporary evidence? The latest quotation is still that of OED1, dated 1607.
[8] The word is unlabelled, though that the 1966 quotation is from a volume of poetry by Kenneth White, and the 1969 one from Nabokov's Ada , may give one pause for thought: are these isolated examples of literary resuscitation, a phenomenon occasionally identified by Burchfield (cf. his remarks on Auden's revival of baltering, or Joyce's of peccaminous and cessile, reproduced in OED2)? It would be helpful to have some editorial comment on this possibility, as with many other unlabelled words whose use is sparsely illustrated, from unusual or eccentric sources, over the last 100 years.
Last Updated ( Wednesday, 28 December 2011 )
< Previous

Built with Mambo. Any comments or feedback are welcome.
All responsibility for views and data published on this site is that of the author, Charlotte Brewer.
Copyright © 2005-13 Charlotte Brewer. All rights reserved.