You are here: Home » About the OED » Archived documents » Second Edition (1989) » History of the OED » The New Oxford English Dictionary project
Search the site | Contact us
 
Preface to the Second Edition (1989)

The history of the Oxford English Dictionary (continued)

[First] [Previous] part 1 of 3 [Next] [Last]

The New Oxford English Dictionary project

Beginnings

Early in 1982, when the editing of the third and fourth volumes of A Supplement to the Oxford English Dictionary had reached an advanced stage, the Oxford University Press set itself to consider the future of the OED. The two questions of whether any further amplification or revision of the Dictionary should be undertaken, and whether an attempt should be made to combine the main OED with the four volumes of the Supplement, were recognized to be complementary. Publishing a supplement to the Supplement, or adding further material to it, had to be rejected as unsatisfactory expedients. The OED and Supplement should be combined before any further development was possible.

How should this amalgamation be performed? The two portions of the Dictionary had been typeset in hot metal. A new edition, whatever it might entail, would have to be typeset by computer. Conversion of the text into electronic form could be carried out either before or after the amalgamation. The option of creating copy for typesetting from the existing printed texts by means of cutting and pasting, or the marking-up of insertions and deletions, was dismissed. The technology of textual processing by computer was now at a stage of development that made it a highly appropriate tool for the task that OUP contemplated. The OED in machine-readable form, structured for use in a database management system, would be relatively easy to revise and, moreover, would be susceptible of a number of other applications, especially as a publicly available computer database. Indeed, the conversion of the Dictionary into electronic form, for just such a purpose, had already been suggested by parties both inside and outside the Press. It was therefore decided that the data conversion should be the first step taken, not only to lead into and facilitate the amalgamation, and subsequent editing, of the two parts of the Dictionary, but also to open up the possibility of its future development in electronic form. A preliminary study, carried out by the Oxford English Dictionaries Department in March 1982 under the supervision of Dr R. W. Burchfield, concluded that both the conversion of the texts by manual keyboarding and their integration by experienced editors were feasible;[9] the report also listed the main aspects of the OED which were ripe for revision and correction. The Senior Officers of the Press determined at once to pursue the idea. Since the OED could be regarded as a kind of national monument, it was felt to be quite proper to solicit assistance, whether financial or technical, from Government departments, research institutions, or industrial companies. Accordingly, Mr Richard Charkin, the then Head of Reference Publishing, initiated a large number of approaches in various quarters, and in the meantime assembled the elements of an appeal brochure. By the end of the year the first outlines had emerged of a project that would involve computerizing and merging the two parts of the Dictionary, revising and updating the merged text, and publishing it in both a new printed and an electronic form.

The identification of partners

In March 1983 a small team was set up within the Press to begin the planning of the project. Its first task was to write the appeal booklet. It was decided that this should combine two aims. The first part, a clear explanation of the background and purpose of the project, was intended both for general information and more specifically to arouse the interest of any institutions or individuals who might wish to enter into some kind of partnership in the project. The second part, giving detailed technical specifications, was to be used by firms wishing to tender for the work of computerizing and merging the text.

By June, the brochure, entitled A Future for the OED, was complete. Copies were sent to computer companies, data conversion firms, on-line database proprietors, universities, libraries, and the British Government. A deadline of 1 August was set, by which time four firms had submitted tenders. The project team, evaluating these, quickly found that, while each tender had its own particular strengths, none furnished the Press with grounds for confidence that one tenderer, alone, could carry out the entire task to the required standard.

The initial idea had been that the chosen supplier would convert the text into electronic form, merge the OED and Supplement, and supply the resulting computer database to OUP; then, using the text editing system newly installed at OUP, lexicographical staff would revise and correct the Dictionary interactively and pass it on for composition and filmsetting. It now became evident that to carry out integration, to create a fully searchable database system, and to revise the bulk of the text in a single step would be impracticable, and it would be far too long before any new edition of the Dictionary could be published. A revised approach was needed. The project had to be broken down into smaller components; a number of different project partners were required, each responsible for what it could do best; and OUP should act as overall manager of the whole process, co-ordinating the separate components centrally. By the end of 1983, partnerships had been established with three contrasting institutions. These were as yet on an informal basis, but during the succeeding months of joint exploration they rapidly crystallized into formal agreements.

A data conversion firm of great experience and capacity, International Computaprint Corporation (ICC), a subsidiary of Reed International situated in Fort Washington, Pennsylvania, was selected to carry out the conversion of the two texts into electronic form. IBM United Kingdom Ltd. undertook to supply computer hardware and software, and to second a group of experts as the nucleus of a team of system designers: their task would be to build a computer system that would facilitate the integration of the two electronic texts into one. Early in 1984 it was confirmed that this assistance would take the form of a donation under the auspices of IBM's Academic Programme. The University of Waterloo in Ontario, Canada, had expressed great interest in the research aspects of the project as early as 1982. They undertook the task of designing a database system suitable for the dissemination of the OED in electronic form and for the phase of updating and editing which would follow integration. The geographical distribution of this partnership was felt by OUP to symbolize quite suitably the international significance of the OED.

Recognition of the national importance of the project followed soon afterwards. A copy of the brochure had been favourably received by the British Government, and as a result a subvention towards the cost of the lexicographical research was announced by the Department of Trade and Industry in May 1984.

There was one other development of great importance towards the end of 1983. Fifty years before, the remainder of the team responsible for the OED, having completed the original Supplement, dispersed, and there followed an interval of a whole generation during which no original historical lexicography was carried on at OUP. Because of this, the new Supplement had to be started virtually from scratch, and needed many years to make up the lost ground. It was imperative to avoid the repetition of such a hiatus at the completion of the Supplement. Accordingly, a small editorial group who had been engaged in the drafting of entries in Volume IV was set to continue the work of compiling new entries, starting again from A, but also taking in new words and senses anywhere in the alphabet for which entries were clearly needed. This new series of entries was called by the convenient acronym 'NEWS', standing for the 'New English Word Series'. It immediately became a valuable source of information for the other Oxford Dictionaries. Although a complete updating of the Dictionary was now postponed to a second phase of the project, it was decided during 1984 that approximately 5,000 items from this series should be included in the new edition of the OED in order to compensate for the temporal gap between the earlier and later volumes of the Supplement. The selection and editing of these articles was set in motion, and their text converted to machine-readable form at the end of 1986. This subject is treated more fully in the foregoing Introduction.

Planning

In January 1984 a department was established within OUP to manage the project. It was now possible to establish the objectives of the project, as follows: the conversion of the Oxford English Dictionary and Supplement to machine-readable form, ensuring that all information contained in the one form was carried over into the other; the integration of the two texts into one; the addition of articles on a selection of new words and senses; and the publication of an integrated version of the Dictionary within an acceptable time. These objectives constituted the first phase of the New Oxford English Dictionary Project. The revision, updating, and enhancement of the Dictionary (of which more is said below) would be the business of a subsequent phase.

Detailed planning was essential to the attainment of these objectives. An overall plan (known as the 'Plan of Record') was drawn up that identified all the major activities within the project, their interrelationships, the time each would take, and the resources of staff, equipment, and finance each required. These were: conversion of the data (or 'data capture'), initial proof-reading, computer development, automatic processing of the machine-readable text, editing of entries on the screen, composition of galley proofs, final proof-reading, and final page composition. For each of these a detailed plan was made. In July 1985, when the outline design of the computer system was complete, it became possible to estimate the times required by the process of building and using that system; these times were added into the plan, and a firm Plan of Record was established. Thereafter, the target dates for the completion of each main activity were fixed.

Proprietary software designed for project planning and spreadsheet operations was of vital help in developing and monitoring each of the interlocking detailed plans which made up the overall Plan of Record. This computer assistance immediately revealed the effects on that plan of changing any value (number of staff, amount of time, or cost). Hence it was possible to project time and cost quite accurately and to monitor progress against these projections. The use of such technically sophisticated methods, more redolent of engineering than lexicography, and unprecedented in the history of the Oxford Dictionaries, was necessitated by the scope and scale of the project. The latter may be roughly illustrated by some figures for the resources used in each main activity. Data capture, the keying of about 350,000,000 characters over 18 months, took 120 person-years; computer development took 14 person-years; automatic processing of the text took 10 months; interactive integration took 7 person-years; the two rounds of proof-reading, undertaken by over 50 people, each took 60 person-years; and final composition of the integrated text involved the setting of approximately 20,000,000 characters per week. The adoption of rigorous planning and adherence to strict monitoring of progress contributed significantly to the work's completion in full accordance with the schedule and expenditure forecast which had been established four years previously.

9. The feasibility of using an optical scanner to convert the text of the Dictionary into machine-readable form was also investigated by OUP at this point, as also by others later. It was generally agreed that the complexity of the structure and the irregularity of the type would require an excessively large amount of editorial intervention in the scanning process; and it was not clear how an adequate framework of structural mark-up could be introduced into the text alongside this method of data conversion. [return]

[First] [Previous] part 1 of 3 [Next] [Last]