Deputy Keeper
Anthropologists who have recently visited the Museum of Mankind (the
Ethnography Department of the British Museum) will be aware that records
of its collections are now stored in a computer. As these collections
are a major anthropological resource, BICA readers may be interested
in a summary of how and why this has been done, and in plans for the
future development of what is the largest computer project of its
kind yet undertaken in Britain.
The Museum's fast growing collections total some 300,000 specimens. They range in character from samples of raw materials used in technological processes to masterpieces of traditional art; in age, from items already in Sir Hans Sloane's collection, from which the British Museum was founded in 1753, to things collected yesterday; in origin, from archaeological cultures of Africa, Oceania and the Americas, to recent and extant tribal and peasant cultures of most parts of the world, together with some products of ³high civilisations² and of societies undergoing rapid change; and in type of acquisition, from isolated donations for which little or no background information exists, to comprehensive, well-documented field collections commissioned from anthropologists.
The variety of its collections is unarguably one of the great assets of the Museum, but their anthropological value depends largely on the quality and extent of accompanying documentation. Two qualifications are worth stressing here. First, even with a minimum of information, such as provenance in time and place, an artefact may provide important evidence for the interpretation of wider aspects of cultural activity, and further details enhancing its scientific interest may eventually accrue from comparison with collections elsewhere. Therefore whilst we now emphasize the importance of collecting a wide range of information along with the objects themselves, it by no means follows that collections with poor contextual data are unimportant. This leads to the second point. Any object can be more easily studied and compared with others if its documentation is ordered systematically and made accessible. The simple morphological description of an object records information about it even if all other contextual details are lacking; but that information is more likely to be enriched by comparison with other material if the record complies with convention and is available to researchers. To the same end, the security of the collections has to be carefully controlled; this is much easier with a systematic approach to record-keeping and a flexible method of retrieving information.
In 1979 an internal feasibility study recognised computerisation to be the only practical way of achieving these objectives in the Ethnography Department. The programme to carry this out was planned from the start as only the first step in the eventual computerisation of the collection records of the entire British Museum. If techniques were developed which could cope with such vastly miscellaneous and disorderly documentation as that of Ethnography, then it was felt that they could probably handle anything which the other antiquities departments might offer. Indeed, work is now under way in several of these. The decision of the Trustees of the British Museum to begin computerisation reflects an important aspect of broader context of the Museum's work, has already influenced the character of that work within the Ethnography Department, and will continue to do so in the future.
As a national institution, the British Museum receives an annual grant
from Parliament. Whilst the Trustees decide how this shall be allocated
among its different sectors and departments, the Museum remains answerable
to Parliament for the way it spends its money. Improving the Museum's
accountability means, in broad terms, being able to demonstrate its
effectiveness and efficiency in achieving the objectives specified
in the 1963 British Museum Act. These include the aim of making objects
available for inspection by members of the public. But although the
Act thoughtfully leaves it to the Trustees to judge how far this might
be practicable, the provision of objects in this way is seen as means
to the end of advancing learning, along the other ways of encouraging
the study of those academic disciplines to which the Museum's collections
are related. Such activities include carrying out fieldwork, building
the collections themselves and improving their storage conditions;
responding to enquiries from the public, including specialists; organising
exhibitions and educational programmes; producing scholarly works,
catalogues based on the collections and a wide range of materials
forpublic consumption; and liaising with colleagues. All of these
tasks have been or will eventually be affected by the computerisation
of records. Even before the first records were transferred onto the
computer a comprehensive inventory was compiled of the different forms
of object registration which have been in use at different times during
the past two centuries. This alone helped the search for information
about material in the collections by encouraging a more systematic
approach, and by recording for wider use, and enlarging the knowledge
of senior curators which had developed through personal but somewhat
haphazard experience. In addition to traditional documentary sources
which list collections in order by which they were acquired, computer
printouts now give us, for the first time, classified indexes to object
types, geographical and tribal origins; as well as alphabetical lists
of donors, vendors and materials. Special category lists such as
Indonesian weapons, tribal art bought at auction, 19th century American
masks, or everything acquired by Captain Cook on his third voyage,
can be produced on request, and assists enormously work on acquisition
programmes, exhibitions, publishing, and enquiries from the public.
The project was initially divided into two stages. The first, which was completed between 1980-3, involved the recruitment of full-time temporary staff who would have the task, in consultation with a computer specialist based in the Research Laboratory and senior curators in the Department, of preparing and transferring all collection records onto the computer. Only in the second stage would the computerised records be checked back to the collections themselves and amended accordingly, and location codes and other details added. The reason for this approach was two-fold. First, it was recognised that in order to retain the interest and necessary involvement of curators, the project would have to quickly produce useful results. This it has done by generating indexes and other tools, such as a specialised thesaurus, deriving from the work of creating consistent terminologies used to describe objects, materials, tribes and geographical entities. By using the indexes, curators discover further information which is fed back to the computer, or errors, which are corrected, and where needed, additions are made to the original registers or slips, which remain the primary sources of documentary evidence. The second reason for dividing the project into stages was the equally practical one of cost. Progress in the first stage could therefore be assessed before extra funds were committed to the second. In the event the first stage was judged sufficiently successful for the project to continue, and the second year of Stage Two has now begun.
It is now increasingly obvious that the Museum's existing minicomputer, a Hewlett Packard 21 MX E-Series machine, will need to be replaced by a mainframe computer in order to cope with the vast increase of information and complexity of activities which the further development of the full British Museum programme entails. For the two stages of the project in the Ethnography Department increased capability has been achieved using two Hewlett Packard 7925 disc drives, various kinds of magnetic and paper tape equipment, printers and terminals, VDU's and other hardware. To input collection records at the point where the computer team was based, a terminal was installed in the Department's store at Shoreditch, and linked by the GPO after a frustrating six-month delay, to the computer in the Research Laboratory in the main British Museum.
The software used is a slightly modified version of the GOS package,
developed by the Museum Documentation Advisory Unit, which has supplied
it to several other museums and galleries. This was judged not only
to be the best available package for the Ethnography Department but
also to be the most likely to be used by other British national museums.
A clinching factor was the cost. It is estimated that the British
Museum would have had to spend about £30,000 to create its own
software; happily, GOS comes free with the corporate membership subscription
by which the MDA is supported by the Museum and other institutions.
It will be many years, if ever, before optimum accuracy and completeness is achieved in the Department's collection records. The task of correcting and refining these is a routine curatorial responsibility. But the enormous flexibility of computerised methods makes this a much easier and more immediately rewarding activity than it has ever been in the past. The GOS package facilitates the addition of numerous supplementary fields to those which have already been established. In the main the techniques used will be simply cross-reference individual records to related information elsewhere, such as bibliographical sources, photographic data, and examples of comparable items in other collections or museums, without duplicating all the details. Records of objects' conservation histories are already kept by the Museum's Conservation Department, and it is a simple matter to insert cross-references to these. Details can also be easily recorded about loan, photographic, exhibition and publishing histories.
Even more interesting possibilities are emerging in the area of visual data. At the time of writing, the Department is contemplating the use of video equipment to record new acquisitions quickly, conveniently and securely. For initial registration, this has distinct advantages over the current system of photographing new material. Neither photography, nor the long-established, traditional and undoubtedly educational method of drawing specimens one by one in a bound register, appears able to keep pace with the rate of acquisitions, and the time is approaching when a verbal description of an object's morphology will be made redundant by new forms of recording, sorting and retrieving visual images. Such developments as laser discs, holograms, and automatic analysis and comparison of visual forms raise quite new prospects for many aspects of the Museum's work and for the use which members of the public, including anthropologists and other scholars, may make of its resources.
Eventually the application of computer technology in the Museum is likely to extend beyond even such efficient 'housekeeping' functions as the co-ordination of collection data on a worldwide scale. Recording and organising factual information always involves choice and interpretation; but the more this systematising process develops, the more practicable it will become to use computer techniques to formulate and investigate anthropological problems in relation to artefact-based collections. It would probably be better for such work to begin prematurely than to be deferred, since it will be more by inspiration, trial and error than by simply waiting for more systematic data to be amassed, that interesting new ideas are likely to emerge. This is the area in which the best long-term prospects for computer-aided collaboration between the Museum and academic anthropologists can be anticipated.*
*Further technical details about hardware, software, vocabulary, choice
of fields and use of keywords are available upon request and a fuller
account of the project so far is currently in preparation. Contact
the author at: Museum of Mankind, 6, Burlington Gdns., London W1X
2EX; tel (01-) 437 2224.