Collecting the evidence

One of the guiding principles of historical lexicography is that entries should be based on the ‘real’ facts of the language, drawn principally from examples illustrating contemporary usage over the centuries. By analysing this information, the dictionary’s editors are able to form a picture of the development and gradual shift in meaning of each word.

The OED has a number of different ways of collecting information of this sort:

  1. running directed and voluntary reading programmes
  2. analysing the content of published historical dictionaries, especially ones written to describe a variety of English
  3. searching through large databases of historical and modern text
  4. asking for and responding to contributions from the public

Nowadays most of the data that arrives at the OED‘s offices has been captured in electronic form. But handwritten letters still arrive from time to time, and their contents are stored in the dictionary’s quotation files.

Different reading programmes for different types of text

The OED has always used the results of targeted reading programmes as a major source of evidence. Reading programmes are particularly useful for isolating new usages, and for collecting evidence which is not readily available on online databases.

For these reading programmes, readers scrutinize texts such as novels, poetry, magazines, newspapers, scientific journals, television scripts, etc., mainly from the 19th and 20th centuries. They look for examples of new words, but also for good, quotable examples of older words. Each word they find is then keyed, with a line or two of surrounding context, into the OED‘s reference database, which is used by all Oxford dictionaries in their selection of new entries.

Recent texts ‘read’ include: Art Quarterly, Ultra Fit, New Yorker, New York Times Magazine, Vanity Fair, and many others … David Starkey’s Henry: virtuous prince (2008), Simon Armitage’s Gig: the life and times of a rock-star fantasist (2008), Una McGovern’s Lost Crafts: rediscovering traditional skills (2008), Ian Thomson’s Dead Yard: a story of modern Jamaica (2009), Jane Struthers’ Red Sky at Night (2009) – etc., etc.

The readers of the HRP look mainly at texts not available to the editors of the original OED, such as diaries and collections of letters. They write out slips by hand in the traditional way, or key directly on to laptops, often in the library where the text is held.

Examples of texts recently read (with examples) include:

Examples of articles read by the SRP recently include:

Historical dictionaries: standing on the shoulders of giants

Many historical dictionaries have been produced (and others are still being prepared) which are based on the principles of the OED. Typically, they provide a thorough coverage of a particular variety of English, whether based on a chronological period of the language (such as Middle English), a region in which English is spoken (such as New Zealand), a specific subject area (e.g. baseball), or some other aspect of the language.

Material from these dictionaries is an invaluable resource for OED editors, who review the contents of these remarkable volumes before embarking on their own work. The dictionaries listed here form just a small subset of all those that have been produced, and are mentioned for illustrative purposes only.

The OED has provided much of the raw materials for many of these dictionaries in the past, and is now able to make use of their findings to enhance its own entries.

Databases: mining new caverns of English

One of the most remarkable changes in the world of historical lexicography since the 1980s and 1990s has been the proliferation of large-scale, searchable databases of historical text. It is now possible to search much of the raw data of lexicography from one’s pc or laptop (whether one is an OED editor or a general researcher).

It is invidious, as usual, to select particular databases, but some of the oldest and most valuable databases for lexical research include (many accessible on a subscription-only basis):

This change has had far-reaching effects for historical lexicography, dramatically improving the quality of documentary evidence available to editors. As a result, it is possible to have more confidence in the reliability of lexical documentation than it was in the past.

Largely as a result of the availability of this data, OED editors have moved from being able to antedate roughly 30% of the words and senses of the dictionary in 2000 to (now) providing earlier attestations for around 60% of all words and senses in the dictionary.

Contributions: words from the Democratic Republic of English Worldwide

Beginning in 1859 the OED has appealed to the public for information about words, and since then the general public (as well as scholars and researchers) have been generous in their contributions to the dictionary.

Recent additions to the dictionary from public contributions:

atom bomb: back from 1945 to 1920 (referring to H. G. Wells)

meet-and-greet (as a noun): back from 1981 to 1960

Molly dancer: back from 1903 to 1881

pork-barrelling: back from 1941 to 1921, and then further back to 1910

Each week more material for the dictionary arrives by letter at the OED‘s offices, or electronically through the OED Online‘s submission form. This data is filtered into the dictionary electronic files or physical quotation cabinets. Every piece of information stored is available to the editors when they come to revise their entries or to add new words to the dictionary. We don’t take kindly to the suggestion that evidence could be thrown away!

The next stage in the editorial process is the sorting of the material that has been collected.

Sorting of quotations

Back to top