Examining the OED - OED3 1500-1599

Thursday, 31 January 2013

Home

Initial results

OED3

OED3 1500-1599

This page reports

some recent discussion of sixteenth-century lexis
the early analysis (pre-2002), by OED3's Principal Etymologist Philip Durkin, of changes in OED's documentation of the sixteenth century made by the revision of the OED currently underway (OED3) to that of the second edition of 1989. (It is important to remember that OED2's documentation of the sixteenth century replicates that of OED1, completed in 1928, since OED2 made virtually no changes to pre-nineteenth-century material in OED1. See further OED editions.)

For OED1's treatment of this period, and work up to 1996 or so, see our page on Early Modern English under Period coverage of OED1.

Examples of the dangers inherent in relying on OED documentation for judgements on the development of the English language were pointed out most strikingly in Schäfer 1980 (and cf. Schäfer 1989). But they continue to be made, not least because in the absence of an Early Modern English Dictionary there is no other comparable collection of lexical information for this period, whether satisfactory or not (see e.g. Bauer 2006). In a masterly survey of 1999, 'Early Modern English Lexis and Semantics', Terttu Nevalainen discussed dictionary evidence for this period, with reference to Schäfer's research, to conclude that for 'the Early Modern English period as a whole, the imbalance in primary sources cannot be ignored when assessing lexical growth on the basis of the dictionary [i.e. the OED]' (Nevalainen 1999: 336-9).

In her latest book on Early Modern English, Nevalainen (2006a: 46) again refers to OED's 'uneven' coverage, and cites a number of studies, some recent, analysing Early Modern English lexis on the basis of OED evidence. Nevalainen herself makes extensive use of corpus evidence to analyse sixteenth-century language. Her university, Helsinki, has been the leader over many decades in developing historical language corpuses to create a sound basis for the understanding of language growth and change, though so far the flood of studies based on corpuses has largely concentrated on morphology and syntax rather than lexis.[1]

One of the reasons for this is that such corpuses contain many more examples of morphological and syntactical forms than they do of any individual words: to date they are all far smaller than the OED. The great advantage of corpuses is that they are carefully chosen according to stated principles to reflect particular categories of texts produced over the relevant period. Conclusions based on them are transparently grounded in evidence that can be fully described. The OED - a vast treasure-house of linguistic data - is based on evidence that varied according to what was available to the lexicographers and what they and their readers chose to record (see e.g. our pages on Outline material and elsewhere). Despite its tempting super-abundance it is dangerous to draw large-scale (or indeed small-scale) conclusions from this evidence, which may be partial or eccentric in ways difficult to second-guess. (For some comparisons between corpuses and the OED see Hoffmann 2004 and cf. Brewer 2006.)

The OED3 lexicographers likewise show themselves fully aware of the problem of depending too much on OED for information on the lexical development of the language. In 2002 Philip Durkin published an analysis of a small sample of the revised alphabet range in OED3 (M-mamzer) 'to illustrate how revision work on all areas of the text...is transforming the [OED's] record of the vocabulary of English' (Durkin 2002: 66). Noting (p. 67) that non-literary texts had been especially fruitful sources for revising OED1's record - not surprising given the early lexicographers' special attention, for a variety of reasons, to literary sources (see our page on Literature and the nation under OED1 intellectual climate) - he began and ended his article by referring to OED's role in charting the development of the English language, and the reliance by scholars on the OED for this purpose:

Attempts to characterize the development of the vocabulary of English in various historical periods have, understandably, often taken as their basis the documentation provided by the Oxford English Dictionary. (p. 65)

A major aim of OED3 is to make the dictionary's methodology more transparent at all levels...it is to be hoped that this, together with the revised data, will provide a powerful tool for future studies of the development of English lexis. (p. 76)

His conclusions contain two warnings:

'the overall rate of change [sc. between OED2 and OED3, as the latter revises the former] is sufficient to demonstrate that considerable caution should be exercised when using OED2 dates for sixteenth-century items for statistical purposes', since 'approximately a third of OED2 words and senses are being antedated during the course of work on OED3'
'any dictionary dates should be treated with a certain amount of caution' (p. 70; cf. pp. 75-6).

Durkin's lists of date-changes for words treated in OED2 and OED3 make fascinating reading but present a complex picture. While many words identified in OED2 as first used in the sixteenth century are being antedated to the fifteenth century, at the same time many words identified in OED2 as first used in the seventeenth century (or later) are being shifted back to the sixteenth century. It is too early to draw firm conclusions from this evidence: we shall have to see how the pattern of documentation develops as the third edition progresses. (See previous page for analysis of OED3's quotation distribution in 1500-1599 between September 2003 and December 2005 - i.e. subsequent to Durkin's article.)

Durkin also lists (pp. 67-8) the sources which the OED3 revisers have found most fruitful for antedating OED2's records of Early Modern English vocabulary:

quotations printed elsewhere in OED
OED's own file collections: i.e. the antedatings submitted by numerous members of the public over decades and also printed in periodicals like Notes & Queries or in scholarly books/articles, together with those yielded by OED's own reading programmes, especially non-literary sources such as wills, inventories, etc.
historical dictionaries, especially MED and DOST
(identified as particularly significant) Bailey's EMED materials (for more on which go here)
electronic corpora, 'especially for this period Chadwyck-Healey's Literature Online'.