1800-1929 in OED3

This page discusses the changes now being made to the OED1/2 quotation documentation of the years 1800-1929, so far as is visible using the search tools available on today’s OED Online.

Outline changes to the Dictionary’s quotation totals per decade 1800-1929 are considered first, followed by a look at individual totals per decade for OED3’s ‘new’ quotations and at what that might mean for an assessment of Lexical productivity. Next, some more General observations include commentary on the changing record for Top individually authored sources in the OED and the continued dominance of canonical male literary writers in today’s revised version of OED Online. It appears that the OED3 revision has attempted to offset the predominance of these culturally distinctive single-authored sources by adding enormous numbers of new quotations from multi-authored sources. This is discussed in the page’s final sections on Newspapers, journals, and periodicals and Multi-authored encyclopaedias/dictionaries.

Outline changes to OED1/2 record

Chart 19 below displays the total number of quotations over 1800-1929 recorded in OED Online (searched December 2018) in comparison with the equivalent totals for OED2. As always, users of OED Online must remember that this version of the Dictionary combines revised entries (i.e. the embryo OED3) with unrevised entries (i.e. from OED2), and that it is impossible to distinguish between these two categories electronically.

Chart 19: Total quotations in OED1/2 & OED Online (Dec 2018) 1800-1929

The obvious comment here is that OED3’s new collection of quotations for the revision seems to be following the original OED1 line with remarkable fidelity – including the post-1899 decline (discussed in the next section).

This pattern was early established: compare evidence recorded on our Archived site for comment on revision for this period up to 2007: OED3: 1500-1899; also Period coverage: 19th century (scroll down pages for relevant charts).

Also notable is the high number of quotations that OED3 has added to OED2’s record. OED2’s total number of quotations for 1800-1899 came to 747,119, larger than for any preceding century (as can be seen in Chart 7b). The OED3 revision, by December 2018, had increased that total to 982,847 in OED Online, an addition of 235,728 quotations. For the entire period examined on this page, 1800-1929, OED2’s total was 872,922, which in turn had been expanded to 1,254,777 quotations in OED Online by December 2018, an addition of 381,855 quotations in all. As already mentioned, many of these new quotations have been derived from recently available historical databases for Newspapers, journals, and periodicals.

Quotations added by OED3

We can get a more nuanced view of the changes the OED3 revision is making to the OED’s chronological representation of 1800-1929 by hiving off the numbers of new quotations that today’s lexicographers have added to each successive decade, then comparing these new totals with each other separately from the OED2 ones. (Unfortunately, it is impossible to identify OED3’s additional quotations individually, so we can’t tell which words or alphabet ranges they represent).

Looking just at the OED3 increases for this period, as displayed in Chart 29 below, allows us to see that the lexicographers are certainly making efforts to remedy OED2’s post-1889 drop. Each post-1889 decade has had more quotations added to it over the course of the revision (as recorded in December 2018, i.e. coming up to 50% of entries) than any of the pre-1889 decades. Nevertheless, the new totals are some way from being sufficient to remedy the existing deficit and the OED3 revision will have to work hard to correct this in the coming years. It is implausible that the vocabulary as a whole, or its usage, was shrinking over this period, as indicated both by Chart 19 above and Chart 30 below. On the contrary, over the fin de siècle and first decades of the 20th century there was a general consciousness of the growth of new words and senses to meet rapidly changing social, cultural, and technological changes – as can be seen in the reception of the 1933 Supplement.

Chart 29:  Total ADDITIONAL quotations in OED3 (Dec 2018) 1800-1929 

Lexical productivity 1800-1929

We can dig deeper into the limited evidence afforded by the current OED Online website in another way too, by considering just the first quotations recorded between 1800 and 1929. These tell us how many new words or senses have been captured for this period, both in OED2 and in OED Online as of December 2018 (the latter constituting the usual undifferentiated mixture of OED2 quotations with those so far added by OED3 in the revision process).

Chart 30: First quotations in OED2 compared with OED Online (Dec 2018) 1800-1929

Once again, the OED Online totals track the OED2 ones – not surprisingly, since in general, the number of new words found in any period by OED is proportional to the number of quotations recorded overall for that period. The more quotations amassed, the more likely it is that a new word will be first detected there.

More general observations

The discussions above are restricted by the difficulties of wrestling with the current OED Online website. It is more rewarding to look at how the record is changing on individual sources. Here we can compare OED Online’s treatment of top 19th-century quotation sources with that of OED1/2 (as in Chart 17 on the previous page). Chart 31 below displays the results. But we can do more than that, too, since so many 19th and early 20th-century texts are among the most quoted sources in OED Online. That means they are usefully listed for us in the table OED Online now provides of the Dictionary’s 1,000 top sources, updated every quarter. We therefore have much more information to work on, as discussed further down the page at OED Online top sources 1800-1929 – individual authors

Chart 31: Major 19th-century authors in OED2 compared with the same authors in OED Online (June 2019)

Each of these changes looks individually interesting. It is frustrating not to be able to identify the new quotations OED3 appears to have added to the Dictionary, notably in the case of Scott and Dickens. Scott’s record in OED1 was disproportionately prominent, so one would have expected quotations from this writer to have declined not increased over the course of revision – and one would also have expected the promised new attention to 18th-century sources (to remedy OED1’s comparative neglect of this century) to have provided multiple antedatings for words initially recorded as first used by Scott, as has been the case with Shakespeare.¹

Similarly, it would be fascinating to discover whether the additional quotations for Dickens and Carlyle are for distinctive uses often typical of these authors (whether eccentric, or colloquial), or for more standard vocabulary. 

These questions are currently impossible to answer, given the absence of appropriate searching tools on OED Online. Nor can we know whether the variations in treatment between authors in the process of today’s revision (the increase in quotations from George Eliot, the decline in those from Macaulay) represent decisions taken by the revisers as to which sources to read, or alternatively tell us something genuinely distinctive about the linguistic qualities of the vocabulary in question.

Fortunately, other matters come to light on examining OED Online’s list of top sources, as described in the next section.

OED Online top sources 1800-1929 – individual authors

Using the information in OED Online’s own list of the 1,000 most quoted sources in the OED (in its June 2019 updating), we can see that alongside the major authors over 1800-1929 identified above there sit many others from the same cultural demographic. Chart 32 below lists the top individual sources of this period in the current version of OED Online, with a cut-off of 3,000 quotations chosen for ease of display.

Chart 32: Top individually-authored sources in OED Online (June 2019) 1800-1929

(The number in brackets after each author’s name indicates his respective ranking in the June 2019 OED Online ‘Top sources’ list).

The chart shows clearly that where individual sources are concerned, a Victorian canon of ‘great writers’ – literary, cultural and historical authors – continue to dominate today’s OED quotations. Of the 18 individuals included in the chart (i.e. with over 3,000 quotations dated 1800-1929), 15 fall into this category, of whom only one (George Eliot) is a woman. Six are principally poets (Tennyson, Southey, Coleridge, Byron, Shelley, Robert Browning), six are principally novelists (Scott, Dickens, Thackeray, George Eliot, Bulwer-Lytton, Kipling), two are principally historians (Carlyle and Macaulay), while one (Ruskin) might be described as a cultural commentator and essayist. The remaining three non-literary individuals turn out to be authors of reference works containing systematic lists of specialist terms: Edward Knight (1824–1883), quoted from his various dictionaries of mechanical terms, John Lindley (1799–1865), similarly quoted from dictionaries and other works on botany, and the distinguished physician Thomas Clifford Allbutt (1836–1925), almost all of whose 3,300-odd quotations are from his multi-volume System of Medicine, published over 1896–1911.

All 18 sources were also quoted in OED1 – as one can see from random searches of their citations in OED Online, which turn up regularly in unrevised entries. Some have been additionally cited in OED3 – again, as one can see from random OED Online searches revealing newly added examples of their quotations in revised entries. The search tools do not however permit any systematic comparison of quotations between revised and unrevised entries, so we can’t know whether the record of these authors is changing in any significant way as the OED is revised. It is probable, nevertheless, that the vast majority of quotations from these individual authors – a culturally skewed group – have simply been inherited from OED1.²

This is likely to be true of the quotations from those immediately beneath these 18 on the OED Online’s ‘Top 1,000 sources’ list too. Extending the cut-off point for Chart 32 to 2,500 quotations would add three more English novelists (Charles Kingsley at 115 on the ‘1,000’ list, R. L. Stevenson at 122, and the popular novelist of sea-faring fiction Frederick Marryat at 140), two American men of letters (R. W. Emerson, placed 113th, and Mark Twain, at 137), the Scottish poet Robert Burns (116), and the distinguished populariser of science Thomas Huxley (133). One might guess that the vocabulary used by these writers is beginning to become more representative of general usage of English of the period (however defined), but it is nevertheless remarkable, given the hugely increased prominence of female novelists and writers at the time, that male-authored as well as literary sources continue to be so disproportionately favoured in today’s OED. This is the case for early 20c authors too, not just 19th century ones. James Joyce, George Bernard Shaw, and D. H. Lawrence immediately follow Marryat on the top 1,000 list. The first 20c female author on OED Online’s ‘Top 1,000’ list, Virginia Woolf, occupies the 736th place. Reliance on such a skewed choice of quotations does not seem appropriate for a dictionary describing itself as ‘the definitive record of the English language’. See further our pages on Fe/male sources.

The OED lexicographers have yet to produce their own account of the decisions that have gone into selecting quotation sources for the OED3 revision. But one can guess that they must have been concerned to correct the distorting effects on the linguistic record of this mass of evidence from (mostly) canonical male writers, over the 19th century as over others. One way of dealing with these vast blocks of OED1 quotations from individual writers would have been to jettison some or most of them. That might well have seemed unattractively wasteful. An alternative solution might have been to quote in equally high numbers from other individual sources, representing more diverse social and educational backgrounds. That might have been rejected as unrealistically laborious and time-consuming. Instead, the OED3 revisers seem to have decided to focus their quotation-searching in the newly available databases of multi-authored newspapers, journals and periodicals. The quotations from ‘great male writers’ remain, but – as Chart 33 in the next section shows – these now sit side by side huge numbers of citations from anonymous sources, almost certainly also male, in most cases, but from less identifiable (or obvious) cultural backgrounds, albeit all written in a formal style. These newly dominant sources in the OED are discussed in the next section.

Increase in multi-authored sources: periodicals, journals, newspapers

As just described, in tackling revision of the 19th-century portion of OED the revisers would have been well aware of the existence of large numbers of quotations from individual ‘great writers’. Walter Scott (over 15,000 quotations), Dickens (over 8,000) and Tennyson (c 6,700) were all among the top ten most quoted individual authors in OED in its entirety, with Carlyle (c 6,240) and Macaulay (just under 5,500), both among the top twenty most quoted authors, relatively closely at their heels.³

A considerable proportion of these quotations were likely to be quite unrepresentative of 19th-century current vocabulary. Scott’s novels and poetry contain many historical, archaic, and dialectal words, Tennyson’s works are often characterised by poetic diction probably not typically found in contemporary non-poetic language, and even Macaulay, a prose writer, prided himself on his frequent consultation of Johnson’s dictionary (1755), published over 100 years before the first fascicle of OED (1844).⁴

The issue facing the lexicographers, therefore, given their decision to retain the original OED1 quotations rather than start from scratch, was how to meet the deficiencies or distortions of these banks of literary quotations in producing a representative account of the state of vocabulary in English over the this period. Turning to newspapers, periodicals and journals as quotation sources looks in many ways a sensible recourse, not least because of their increasing availability from the early 2000s onwards in searchable databases. Charts 33 and 34 below show that the OED3 revisers have done this on a significant scale.

Both charts also include the totals of quotations in OED Online dated 1800-1929 from three other multi-authored publications now also available in electronic form: the Encylopaedia Britannica (1768-), the Century Dictionary (1889-1991), and the Penny Cyclopaedia (1833-1858), for comment on which see Multi-authored encyclopedias/dictionaries below.

The 19th century saw an enormous expansion in newspaper and periodical publication. Then as now, newspapers covered politics, regional and national stories, legal cases, reviews and articles on cultural and sporting events, opinion pieces, reader’s letters, reports on specialised areas such as shipping news, business, the social calendar, etc. As can be seen from the list of titles in Chart 33 below, periodicals and journals were similarly wide-ranging, with some focussing on additionally specific disciplines such as medicine (e.g. The Lancet) or academic research (e.g. The American Naturalist, which as its website says covers research in ecology, evolutionary biology, population, and integrative biology). 

The value of such sources as witnesses to language in current use was recognized as early as 1855 by the lexicographer Hyde Clarke, whose dictionary published in the same year, A New and Comprehensive Dictionary of the English Language, pointed out the advantages of this material as opposed to the more traditional sources favoured by dictionaries of the period, including the future OED:

A great merit, and thereby a deficiency in our standard dictionaries, so far as their general use by the population is concerned, is that they are founded upon literary considerations, under which a printed authority is required for a word….it has seemed to me that the English people require something more than a dictionary of book words.

Clarke 1855: iv.

Clarke continued,

It is the growth of the newspaper press which has given importance to the English oral language, the influence of which cannot long be neglected. While the lexicographer is hesitating, weighing, suspending, harshly rejecting, or tardily admitting, a language is being worked out, which will react again on our literature.

Clarke 1855: iv.

By ‘literature’, Clarke must mean writing of high cultural value published in books as opposed to ephemeral publications such as newspapers and journals. Hyde Clarke went on to assert that ‘The Times ought to be as eligible an authority as some book long since defunct, and known only by its epitaph or the title on its coffin-plate’ (iv-v). Our page on Newspapers, due to be published in 2020, will discuss this subject further, meanwhile see Brewer 2007b: 117-18. 

The two charts below clearly illustrate OED Online’s additional, and notably heavy, dependence on periodical and newspaper sources over 1800-1929. Figures in brackets after the name of source indicate that source’s ranking in OED Online’s list of 1,000 top sources (October 2019). As ever, readers should be aware that OED Online merges revised with unrevised entries, so the totals given here represent the combination of unrevised OED2 entries with revised OED3 ones at a stage when OED3’s revision of OED2 entries stood at just over 50%.⁵

Chart 33 lists the top 12 such sources – bar one, The Times – by citation over the period 1800-1929 – searched in OED Online in October 2019. The most remarkable feature both of this chart and of Chart 34 below should be the enormous number of increased citations from The Times. Unfortunately, however, we have had to exclude this newspaper from both charts. OED Online’s search tools do not permit one to distinguish between this publication and numerous others with similar titles – Musical Times, New York Times, etc. Manual elimination of the false hits is not a practicable undertaking given the number involved and the difficulty of manipulating the website’s search results. Since OED Online’s total number of quotations from The Times in October 2019 came to 43,185, considerably higher than for any other source in the Dictionary at any stage in its life, it is almost certainly the top quoted source for 1800-1929 in OED Online too. See the screenshot of the June 2019 update on our page on Top sources in OED3.  

Please note that an earlier version of Chart 33 on this page did include The Times, reporting the figure of 28,753 quotations for 1800-1929. We have since (June 2020) spotted the online search malfunction, applying to searches made in the Advanced search box in the form Quotations/Quotation title (= “The Times”)/Date of quotation (= “1800-1929”)/All quotations.  

Similar problems arose with the periodical Science, which we have also had to exclude.⁶

Chart 33: Top periodicals, journals and newspapers in OED Online (October 2019) 1800-1929 (The Times omitted)

Chart 34: Top periodicals, journals and newspapers in OED Online (October 2019) 1800-1929 compared with OED1/2 equivalents over all dates where known (The Times omitted)

Note on OED1/2 totals in Chart 34. It is not now possible to search OED2 data electronically; see Loss of OED2. Figures for the totals in chart 34 are taken from Willinsky 1994, Tables 7.1 and 7.2, pp. 213-14, where available, with corresponding totals added from Table 10.2 (‘Top Twenty Periodicals for 5,000 NEWS Items Added to OED2), p. 219, where also available. The newspaper Daily News (published from 1846 onwards) was the second most cited periodical in OED1, according to Willinsky. Total quotations from this publication in OED1 were 8,832, compared with 9,993 in OED Online searched October 2019. The difference between these two figures, 1,161, suggests this publication has not figured as significantly in newly added OED3 quotations from newspapers and periodicals over 1800-1929 as have other such titles.

For more on Phil Trans (Philosophical Transactions of the Royal Society) see 1700-1799 in OED3.

Multi-authored encyclopaedias/dictionaries

The three reference books in Charts 33 and 34 are an interestingly different category of evidence on the use of language at this period, albeit also from multi-authored sources. Only one of the three works concerned – the Encylopaedia Britannica (1768-), the Century Dictionary (1889-1991), and the Penny Cyclopaedia (1833-1858) – is a dictionary, but significant numbers of quotations from the two encyclopedias are taken from their definitions or technical descriptions of the terms they include, as can be seen by clicking randomly on search results for these sources to see the quotations concerned – for example, from Encyclopaedia Britannica, under aba n.2 (OEDO updated December 2011) , ‘Aba, a form of altazimuth instrument, invented by, and called after, Antoine d’Abbadie’ (1910), or from Penny Cyclopedia, under achord (OEDO updated December 2011), ‘The acrochord is covered with scales like all other serpents’ (1833). In general, all editions of OED have avoided such quotations since they are self-conscious statements on the meaning of a word or term – and hence less likely to be typical – as opposed to disinterested examples of its contextual use.⁷

On the other hand, both encyclopaedias also provide many examples of contextual use. Penny Cyclopaedia illustrates the noun abnormity (OEDO updated Sept 2009) thus: ‘Without those capricious abnormities which so frequently offend us even in the buildings of Palladio’ (1839), while Encyclopaedia Britannica provides the following under safe door (s.v. the noun safe, OEDO updated Sept 2011): ‘Where larger quantities of valuables had to be preserved,..a safe-door of larger dimensions would be made and attached to a masonry or brick room’ (1911).⁸ Without better search tools it is impossible to investigate OED3’s treatment further. The chief value of these three works for the revisers may be guessed to be the same as that of the newspapers, journals and periodicals: multi-authored non-literary sources for an extraordinarily wide range of vocabulary, covering all aspects of human activity, thought, endeavour etc.

Helpful information on all three works can be found in Wikipedia.

In an overwhelming number of cases the authors of all the intensively quoted multi-authored sources discussed in the last two sections will have been males using formal and highly formal language. The range of styles and registers of language will therefore be limited. How the revisers have set out to supplement OED2’s account of the vocabulary of slang and colloquial language from this period is not easy to ascertain from the OED Online website. Such language is more likely to be found in letters, diaries, and other informal documents closer to oral language. We can but hope that the editors of OED3 will at some stage give us a full account of their lexicographical principles and practice in this respect as in many others.

Last updated on 13 August 2020

1. For OED1’s neglect of the 18th century, see 1700-1799 in OED1/2; for Shakespeare antedatings see Brewer 2012 and Goodland 2018.
2. Willinsky’s list of ‘Top Twenty Books by Citation in OED1‘ tells us that Knight’s Pract. Dict. of Mechanics was quoted 3,813 times in OED1, a total which by June 2019 had increased to 4,037 in OED Online, while Allbutt’s System of Medicine was quoted 3,629 in OED1, a total by June 2019 since decreased in OED Online to 3,293. See Willinsky 1994: 213, Table 7.1.
3. See Top sources. Willinsky’s table of ‘Top Twenty Authors by Citation in OED1’ (i.e. not including sources such as newspapers and periodicals) lists these authors as 2^nd (Scott), 9^th (Dickens), 10^th (Tennyson), 12^th (Carlyle), and 17^th(Macaulay); Willinsky 1994: 211 (Table 6.1).
4. Derwent Coleridge describes how, visiting his ‘lamented friend Lord Macaulay’ a few months before his death, ‘I noticed a newly-bound copy of Johnson’s Dictionary lying upon the table. He told me that it was the fourth livery in which he had invested this trusty servant. And on my asking, with some surprise, in what service he had found so much employment for such a valet-de-libraire, he replied, to keep his diction up to the classical standard, and to prevent himself from slipping into spurious modernisms’. D. Coleridge 1860: 155.
5. See David-Antoine Williams’s charts at ‘More precisions on revisions’, https://thelifeofwords.uwaterloo.ca/precisions-on-revisions/ [accessed 26 April 2020].
6. We do not know how many quotations OED1 recorded from Science, first published 1883; given the date, probably not many. OED Online’s December 2019 total of quotations from Science since first publication was a striking 8,207 (as listed in the website’s ‘Top 1,000 sources’ and therefore – presumably – excluding other periodicals with Science in their title). The bulk of these are likely to be post 1929.
7. See discussion of quotations from Bailey’s dictionaries at 1700-1799 in OED1. An EOED page on OED’s use of Dictionaries is in preparation.
8. All OEDO entries cited in this and the previous paragraph were accessed 22 June 2020.