 |
 |
|
|
||
| Subscriber home page |
| |
|
||
Selecting the images; Organization and staffSelecting the imagesIn recommending the inclusion of images, Matthew diverged from the DNB in practice but not in spirit. In the 1880s it would have been difficult to include reproductions of likenesses on any large scale, he felt, 'both for technical printing reasons and because portrait research was at a very elementary stage'. None the less, the DNB had given 'as much visual information as it could about subjects, short of actually printing an image', through quoting descriptions of their physical appearance as a standard item and by providing much information about the character and location of their portraits. So the new dictionary's policy would be 'a natural extension of the first DNB's approach to the definition of a subject's individuality'. (32)In 1995 Peter Funnell of the National Portrait Gallery was appointed consultant editor for likenesses. Responsible to the Dictionary's editor for policy and practice in selecting 10,000 images, he and two academic picture researchers based at the Gallery (listed here) undertook a project that ran in parallel with the editorial work in Oxford. 'We have the opportunity of creating a national iconography', he wrote, 'and it is an opportunity which should be embraced with great excitement.' (33) The editor and research director identified the articles which would (if feasible) be illustrated, so that the selection of illustrations would reflect the Dictionary's editorial aims, not the availability of images. There were two particular requirements: first, that the Oxford DNB should include images of the more familiar figures in the national past for whom readers would expect illustrations; and second, that new emphases within the Dictionary should be reinforced by highlighting some of the newly included—for example, people from the regions of the British Isles and British territories overseas, people in business, and women. Each article chosen for illustration was to receive only one image (see Table 3 below). In selecting likenesses, the first objective was to ensure authenticity by establishing that the portrait did indeed represent the person intended. In the case of painted portraits, the aim was to locate the prime version—the original portrait done from the life—falling back on a contemporary version if documented, and only in exceptional cases settling for a later copy. Where there were several types of representation to choose from (paintings, photographs, prints, and sculpture, for example), selection reflected several considerations. Where possible, likenesses were chosen from the time of life when the persons concerned were most active in the role that had justified their inclusion in the Dictionary; named artists were usually preferred to the unknown; the work of the best artists was chosen above the inferior; and in some cases the rare image was preferred to the familiar. Caricatures were avoided. The likenesses reproduced in the Dictionary were to be 'visually rich as well as informative', (34) so selections were not restricted to head-and-shoulder or half-length studies. Full-length portraits were chosen where these seemed more revealing. Costume, background, accessories, and period stylization were seen, along with likeness, as part of what 'constitutes the portrait as a representation of a given individual'. (35) Consequently the reproductions included in the Dictionary show in almost all cases the complete original composition. Lists of the articles selected by the editor and research director in Oxford for illustration were sent to Funnell's team. Suitable images were then sought, and a final recommendation of a single likeness was made. The editor accepted the team's recommendations with very few exceptions. Continuous monitoring ensured that the balance of likenesses reflected the Dictionary's new emphases for inclusion of subjects. Contributors provided information about likenesses, and this was sent regularly to the picture researchers; the information proved most fruitful where authors had access to private information about images that was not available in the usual scholarly resources. Research and selection of images began in 1996 and was completed on schedule in March 2002. Half the images came from the National Portrait Gallery's own collections, or from collections administered by them, and the rest from many other sources—ranging from major galleries and museums to much smaller collections, specialist societies, and private individuals. Separate research was undertaken to select images for subjects active before 1500, and this secured different types of iconographic material, including coins, effigies, stained-glass windows, and illustrated manuscripts. Clearing rights and obtaining original photographic materials from which to reproduce images from non-Gallery sources constituted a major task, which was undertaken as a sub-project from September 2000 by the Gallery's picture library team. The later stages were completed by the Dictionary's administrative staff, who also managed the design, scanning, and final approval of the 10,057 images published, finishing in March 2003. In all, some 1500 rights-holders gave permission for likenesses to be reproduced. The Dictionary is deeply indebted to all rights-holders for their co-operation and generosity. Research by the National Portrait Gallery's team illuminated the Gallery's own collection, as well as the complexity of the relationship between historically noteworthy activity, celebrity, and portraiture. Organization and staffThe challenge confronting Matthew was not only scholarly but organizational. Like the DNB, the Oxford DNB was planned as an integrated writing and publishing venture. Newcomers to the project were routinely reminded that the central task was still, in the new venture, the same as it had been in the old: to persuade the right people to write the selected articles in the desired manner. Personal relationships remained of central importance within an inevitably complex administrative structure. But whereas late-Victorian intellectual life flourished within a relatively tight London-based world of gentlemanly leisure, co-ordinated through the London clubs and informal associations, the new dictionary operated within a much wider, international world of professional scholarship conducted mainly through universities. Co-ordination on a larger scale was now needed if a dictionary nearly twice the size of George Smith's DNB was to be produced in little more than half the time by fifteen times as many contributors distributed worldwide.The publishing process also posed fundamental questions for OUP. Encouraged by the Oxford English Dictionary's success when it was published on CD-ROM in 1992, plans for the Oxford DNB from 1993 envisaged publication in both printed and electronic forms. Yet in a period of rapid technological change, nobody could predict exactly how the Dictionary would best be published twelve years hence; nor could OUP wait until the entire text had been written before preparing it for publication. The strategy decided upon in 1992–4 was at that time ambitious: to prepare the new dictionary's material so that it could be used to produce any future form of publication, and to ensure that after 2004 the contents could be continuously corrected and expanded. No longer could it be allowed to suffer the DNB's fate of cumulating decay. The text would first be edited, keyed, proofread, and stored in a permanent electronic form that could be edited within the project. Versions of it could later be produced for publication in whatever form required. In planning the project's administrative structure, a complete view of the publishing cycle had to be taken: from first considering a person for inclusion in the Dictionary to completing the individual article and from there to producing a complete set of articles for publication. Central to the work was the project's database. Conceived by Robert Faber, it was designed and built by specialist OUP computing teams, who from 1998 formed a dedicated technology group on the project staff. As with the Dictionary's administration in general, the database was designed to achieve three aims: to clarify the sequence of tasks and decisions from which the completed article emerged; to remove routine tasks from editors and specialist staff; and to make all the necessary information, including the entire Dictionary text, immediately available to everyone within the project. From 1992 database records were kept of every individual who had been included in the DNB and of every new candidate proposed for inclusion; the database helped staff to perform and monitor every subsequent stage of work. Each of the Dictionary's 50,000 articles could be seen as an individual project—to be commissioned, delivered, edited, approved, and prepared for publication at the same time as thousands of other articles were proceeding through every stage in their journey towards publication. The Dictionary's text was an integral part of the database. OUP had keyed the Concise DNB's articles in electronic form to create the consolidated print edition published in 1992, and when the new project began these texts were used to set up the initial database records for each person in the DNB. This made a brief digest of the DNB's articles readily accessible to staff while the subject matter of the new dictionary was being sorted into research areas. The 33 million words of the complete DNB (including the supplements) were rekeyed for the new project from November 1992, so that from late in 1993 this text too could be incorporated into the database. It enabled staff to read, group, and print out any DNB articles that were needed for editorial review or for sending to contributors; this facility greatly accelerated assessment and commissioning. It also enabled OUP to prepare an electronic version of the DNB on CD-ROM, from which valuable lessons were learned about how to present biographical material in that form. From 1997 Oxford DNB text was also entered into the database, forming the primary version of completed material from which printed and electronic editions could later be prepared. Multiple forms of publication were made possible by the design and implementation of electronic coding; unlike traditional typesetting methods, this procedure remained within editorial control throughout. The coding was used to identify not only typographical appearance, but particular aspects of content, and it enabled detailed searching of text both by editors within the project and eventually also by readers of electronic editions. Matthew had highlighted in his report of April 1993 the central importance of computer technology for completing the project. Immediate electronic access to both text and administrative information was essential to his strategy of research by subject area rather than by alphabet. Not only could previous texts be extracted from the alphabetical run: new texts could be entered, in any order, as and when articles were completed, and editors could survey the accumulating mosaic of the new dictionary when addressing its special editorial problems. Like the Dictionary itself, these problems were unusual in scale. First, given that the Oxford DNB covered the whole of British history, it had to accommodate all the requirements specific to detailed biographical research in every subject area and period. Second, inconsistencies of treatment between different fields of research, rendered more visible by modern electronic searching, frequently posed substantial academic questions. These could be solved only through close collaboration between production and research staff. Because the accumulated material remained manipulable in the database until shortly before publication, conventions could be revised at a late stage in the publication process, with access to much of the completed dictionary—an opportunity which few editors of previous large-scale works had enjoyed. Such a structure of sequential but interlocking concerns called for staff who were both expert in their respective fields and willing to collaborate on the broader aims of the work. The project's full-time staff (listed here) was divided formally between research staff, who were employees of the University of Oxford funded by OUP and the British Academy, and publishing staff employed by OUP. Yet both groups operated as a single project under the direction of the editor, who had oversight of both editorial and production concerns. In this he worked closely with the project director, whose role from start to finish was to manage the project's organization and resources, and to oversee its publication. At its greatest extent, in 2002, the project employed fifty-one people (twenty-nine University and twenty-two OUP staff) on a full-time basis, as well as up to a hundred freelance, part-time, and temporary staff (listed here). The research director was responsible for overall research policy and university staffing matters. In addition to the research editors, whose numbers rose to twenty-three at the height of the work in 2001–2, the research staff included a research co-ordinator and two bibliographic editors. The research co-ordinator organized the central provision of research materials to contributors and staff, and later the retrospective editing of reference lists, while the bibliographic editors focused on standardizing citations in the lists of sources. From 2001 one research editor assumed additional responsibility, as publication editor, for organizing the research side of the overall editing process in its later stages. The publishing staff, working under the project director, consisted of four groups overseen by the publishing administrator (responsible for all aspects of the Dictionary's administration), the chief copy-editor (responsible for editorial conventions and their implementation through copy-editing and proofreading), the data manager (responsible for the structure of the Dictionary's electronic data and its implementation through setting the text), and the technology manager (responsible for building and maintaining the project's information systems to specifications agreed with staff). The administrative and secretarial teams backed up both the research and production staff at every stage and provided contributors with an important point of contact. They organized the flow of work, tracked the progress of articles, and gathered illustrations; with each of the 50,000 articles passing through their hands at least ten times, they had to confront the demands involved in archiving all the project's materials in the 600 metres of filing that embody the Dictionary's history.
|
