Sources of OED data

The data represented in our various charts and analyses is derived from electronic searching of OED Online (www.oed.com) in both its pre-2010 and its current form, as described below under the main headings OED1, OED2, OED3, OED Online, including a discussion of Variations in search results under the last heading. A note on Willinsky 1994 follows. For more information on the different stages and versions of the OED, please see Which edition contains what?

Given the vagaries of electronic searches of OED in all its forms, all quotation figures and other data should be regarded as approximate.

OED1

Our estimated quotation numbers for year-spans, authors, works etc recorded from OED1 are derived from searches made of the electronic version of OED2 available at www.oed.com from 2000 to 2010, made as described under OED2 below. In the late 1980s, both OED1 and Burchfield’s 1972-86 Supplement were digitized, then merged together to create OED2. This hybrid version of the OED was subsequently published in 20 printed volumes and also made available in electronically searchable form, both in successive CD Roms and (from 2000) online at www.oed.com. The pre-1928 component of OED2 is usually therefore a reliable guide to the content of OED1, which has never been released in a searchable format. It must be remembered, however, that Burchfield’s Supplement incorporated much of the 1933 Supplement (which contained pre-1928 material) and in addition added a now-unverifiable number of quotations from some 19th-century – and occasionally earlier – texts and authors. Care must therefore be exercised in interpreting search results.1 

OED2

Quotation figures for year-spans, authors, works etc come from the electronic version of OED2 available at www.oed.com up to 2010. Our methods of searching this resource were fully described at the time and can be read on archived pages from the previous site (also accessible via the link to Archived site at the bottom of each page of the new site).

OED3

It is not possible to search OED3 electronically. We therefore consult the individual OED3 entries accessible at OED Online (www.oed.com) – or, in attempting to draw inferences about OED3 on a larger scale (as illustrated in our charts), we compare OED2 data with equivalent searches of OED Online. Since the version of the OED at this resource is subject to constant change, we record the date of our online access. See further the explanations (on charts) at our Outline page under Period coverage and (more generally) on our pages at OED Online.

OED Online

When searching OED Online (www.oed.com), whether for year-spans, authors, works or anything else, we use the advanced search link on the front page of the website – see screenshot below – and then select search criteria as appropriate. It is sometimes necessary to eliminate false returned entries from the results of advanced searches. We do not use the Categories, Timelines or Sources portals also visible in the screenshot, unless specifically stated – for example, for Charts 9 and 15, which report data taken from OED Online’s list of top cited authors and works cited. This is because search results can vary according to which method is used: see Variations in search results below.

Screenshot of OED Online front page (top left portion) taken 21 August 2019. Source: OED Online
Variations in search results

When making OED Online searches we use the advanced search link on the front page of the OED Online website, as explained above, unless otherwise stated. It should be noted that search results can vary according to which resource is used, sometimes for reasons which are hard to identify.

For example, searching on 21 August 2019 for quotations from John Palsgrave’s Lesclarcissement de la langue francoyse (1530) via Advanced search/Quotations/Quotation title/Lesclarcissement, specifying 1530 as Quotation date, returned 5,190 results in 4,095 entries, as in the screenshot below:

Screenshot of result for Advanced search selecting Quotations option, with “Lesclarcissement” in Quotation title and “1530” in Date filter, 21 August 2019. Source: OED Online

Information on Palsgrave’s Lesclarcissement searched for on the same day via the Sources portal (top 1,000 authors and works cited in the OED) returned a different result: 5,443 quotations in total, not 5,190 as in the first search (screenshot below). The most obvious reason for this difference is that Lesclarcissement is cited not just from the 1530 edition but from later ones too.

Screenshot of result for John Palsgrave via Sources portal (top 1,000 authors and works cited in the OED), 21 August 2019. Source: OED Online

However, clicking on the underlined link in the screenshot above yielded a third result, 5,202 results in 4,105 entries (screenshot below). This indicates that the variation between search results is not due just to citations from different editions but to something else as well – but without combing in detail through the 5,000 items, it is difficult to identify what is causing the disparity.

Screenshot of result of clicking on link for “Lesclarcissement” in entry for John Palsgrave via Sources portal (top 1,000 authors and works cited in the OED), 21 August 2019. Source: OED Online

Where large searches such as this are concerned, yielding thousand of results, it is not practicable to work through them individually. We therefore quote the Advanced search result, making clear if a specific condition applies (a date of 1530, in the case of Lesclarcissement), and leave it at that. Where varying search results for authors and works quoted in smaller numbers are concerned, we work through the lists of individual entries to try to identify disparities and eliminate false results if possible. These are sometimes due to a confusion between different authors owing to shared names or initials, or other confusions of shared titles/dates/editions, etc.

Note on Willinsky 1994

Other data recorded from the different stages of the OED is available in the Appendixes to John Willinsky’s Empire of Words: The Reign of the OED, published by Princeton University Press in 1994. These supply invaluable guidance to a range of OED statistics but need to be used with caution. Willinsky had pre-1994 access to electronic versions of the OED and was able to make separate searches of OED1 and of Burchfield’s Supplement which ceased to be possible after the 1990s. In making our pre-2010 searches of OED2 we occasionally found it difficult to replicate his results, which may indicate flaws in the software or various searching mechanisms he was using – though some of Willinsky’s own reports of his findings are also inconsistent. For example, his Table 9.1 (Willinsky 1994: 215) states that the Supplement contained 1,251 quotations from Joyce’s Ulysses, while Table 9.2 (p. 216) gives the figure as 1,302. Pre-2010, we found 1,285 quotations from Ulysses in OED2 when searching the OED2 CD-ROM version 3.1, compared with 1,266 when searching the then-available OED2 database at www.oed.com

Last updated on 9 October 2019

Footnotes

  1. On Burchfield’s editorial policy in this respect see Brewer 2007a: 165 and Brewer 2015a: 751-4.