The SECOND EDITION of the OED, compiled by John Simpson and Edmund Weiner, was published in 20 volumes in 1989. This edition, though of great significance in the history of the OED since it marked the Dictionary’s transformation to digital form, has led to some considerable misunderstanding amongst both public and academic users.

OED2 (as advertised by Barnes&Noble)

This is because it was not a new edition in the commonly understood sense (i.e. a revised version of an earlier edition, in this case OED1).1 Instead, it was a merging together of OED1, almost completely unchanged, with the Second Supplement (1972-86), with the addition of a further 5,000 new words and senses (less than 1% of the original total; see Additions and changes below). Despite its relatively modern publication date, therefore, very nearly all the entries in OED2 reproduced entries already published, and most of them reproduced, without change, definitions and explanatory matter 60 years and more out of date.

OED2 nevertheless remains the only version of OED which is currently in print. It is found as the work of authoritative reference on the shelves of countless public and academic libraries throughout the English-speaking world. It is nothing of the sort, owing to its unscholarly mixture of old and new lexicography with no indication to the reader which is which. Online, it has been superseded by the version of the Dictionary available at www.oed.com, described at our page on OED Online, though this continues to present interpretative problems to users in its unsignalled mixture of new and unrevised content.

For a time (March 2000 to December 2010) an electronic version of OED2, with sophisticated search tools identical to those then available for OED3, was also available at www.oed.com. Regrettably, OUP withdrew this version of OED2 from its website, so it is now impossible to compare different versions of the Dictionary systematically, i.e. by using comparable search tools. Much of the data and analysis on Examining the OED derives from searches of OED2 (and comparisons with OED3) made before this version of OED2 was withdrawn.

Old wine in new bottles

To create OED2, OUP joined forces with the computer company IBM. The content of both OED1 and Burchfield’s Supplement was tagged and digitized, and the two separate works digitally merged into a seamless entity. This was an innovative project for its day and consumed the imaginations of both editors and lexicographers, as is clear from the many pages of introductory material devoted to describing the ins and outs of the technical issues involved as opposed to the actual content of the ‘new’ work.

The processes that created OED2 established the digital basis of all future forms of the Dictionary. But they also established a dangerous precedent: the mashing together of dictionary entries deriving from two wholly distinct editorial periods and processes – that of 1884-1928 with that of 1972-86 – without alerting the unsuspecting user.

This was an injustice to the integrity of scholarship and purpose that had characterized the OED up to that point. Its legacy persists in the additionally hybridized version of the Dictionary presented to us today by OED Online. This version exacerbates the problem further. Instead of show-casing the superb up-to-date lexicography of OED3 in a coherent and accessible form, OED Online mixes the revised portions of the Dictionary, entry by random entry, with the already mixed lexicography of OED2.

See further OED Online and OED3. Accounts of the laudatory public reception of OED2 can be found in Gilliver 2016 and (more critically) Brewer 2007b, who both discuss OED2 in terms of its initiating OUP’s ‘New OED project’, resulting in OED3. Only specialist OED users of the time were in a position to identify the presentational and lexicographical inconsistencies of this new version of out-dated lexicography: see Algeo 1990, Stanley 1990, Brewer 1993.

Another short but interesting account was recently published online by Richard Charkin, Head of Reference at OUP in 1980 and originator of the OED2 project. Charkin identifies the publisher OUP’s two-fold motivation in developing the so-called Second Edition of OED: firstly, to recoup the enormous losses sustained in supporting and publishing Burchfield’s Supplement, and secondly, to protect the commercial value of OUP’s flagship product, given that OED1 was then about to go out of copyright. While partly tongue-in-cheek, this is an interesting glimpse into the struggle between commercial and scholarly interests in creating OED that has characterised the Dictionary since OUP took over the project in 1878, with financial considerations often trumping lexicographical ones.

In common with the Introduction to OED2 described above, Charkin also, and presumably inadvertently, makes clear the extent to which the digital innovation initiated by OED2 took precedence over scholarly considerations of its lexicographical content, a power relationship that continues today in the makeup of OED Online. See Charkin 2018 (https://publishingperspectives.com/2018/09/richard-charkin-short-history-new-oxford-english-dictionary/ [accessed 7 May 2020]).

Additions and changes between OED1 and OED2

Since no electronic version of OED1 exists, it is impossible to identify all the changes between OED1 and OED2 – and the text of OED2 does not indicate whether an entry (or portion of an entry) derives from OED1, Burchfield’s Second Supplement, or is new to OED2 itself. We know for certain that OED2 contained all OED1 entries, all the new entries in Burchfield’s Supplement, and all Burchfield’s additions to existing OED1 entries. OED2’s Introduction also explained that it had added 5,000 new entries to this mix – i.e. new words appearing or noted since the relevant volume of Burchfield’s Supplement for which Dictionary entries had meanwhile been prepared. No list exists of these new entries, however.2 

In the absence of digital aids, one can only come across changes between OED1, Burchfield’s Supplement and OED2 serendipitously, turning the pages of the large printed volumes side by side. It is hard to identify clear and consistent principles behind the various alterations and interpolations made by OED2 which are detectable from such an exercise.

For example, occasionally it seems that anachronistic definitions relating to sex or gender were an issue for the OED2 editors. Thus OED2 rewrote OED1’s definition of jury to include women and removed OED1’s mention of ‘unnatural’ sexual relations and of ‘vice’ in definitions for Sapphism and tribade. Vice continued to be specified in the etymology section of the entry for Sapphism, however, and much other potentially offensive material in OED1 was left unchanged (e.g., references to ‘unnatural’ practices in definitions for homosexual relations between men).3

It’s possible to spot changes to editorial matter in OED1 entries resulting in similar inconsistencies and/or discrepancies. For example, in OED1, Murray insisted that words of classical origin beginning ‘ps-’ should retain the initial ‘p’ pronunciation. His strictures were reproduced in OED2 (sv. ps-), but not identified as deriving from 100 years earlier – while at the same time a new editorial note was added to state that in ‘current English’ the ‘p’ was silent. The two positions sit side by side in silent contradiction, while some OED2 entries for words beginning ‘ps-’ are marked as having a silent ‘p’, some a pronounced ‘p’ (Brewer 2007c; see Stanley 1990 and Brewer 1993 for further examples).

Last updated on


  1. ‘A particular version of a text maintained by regular revision’, as defined by Lexico.com [accessed 20 September 2019], the new platform for Oxford Dictionaries online. The OED Online entry of edition, n., which reproduces that of OED2, is unchanged since 1891 when it was first published in OED1, other than by the addition of a single quotation of 1943 deriving from Burchfield’s Second Supplement.
  2. In order to compile these 5,000 new entries, we are told, the editorial team supplemented traditional methods of gathering new material (e.g. the reading programme) with ‘large computer databases containing research abstracts, newspaper and periodical texts, and legal reports’. These were ‘invaluable in […] providing examples to complete the lexicographical record’ (OED2 vol 1, ‘Introduction’: xxii).
  3. For details see Brewer 2013a: 107-8.