Previous - Next
A simple computer 'network' can be created locally by linking together computers or computer terminals within a single institution. This is already common in archives and requires no special comment here.
A further stage is the setting up of direct links with a limited number of other users, for example in branch repositories, government departments or research centres.
The 8 offices of Australian Archives, for example, together with the Australian War Memorial, are linked. So also are the Presidential Libraries in the United States.
The PRIAM 3 application at Fontainebleau, France (see p.9 ) is linked to other offices of the Archives Nationales and to archivists based in government ministries. The Archives Nationales and the Direction des Archives are also linked by means of the LYNX telecommunications network to some sixty local archives.
In Switzerland, the first steps have been taken towards keying in data for the national archives at its point of origin in government ministries, a facility which, although under consideration, is not yet available in the United States or the United Kingdom.
The National Archives of Canada draws some of its computer power from a government administration computer, and the Archives Nationales du Quebec is linked (although at present only for off-line editing) to its provincial offices (MacDermaid 1990).
One of the databases of the Italian national archives is linked to the Italian Supreme Court.
Beyond this are still wider applications extending to networks with national and even international coverage. These are already well established in fields other than archives, but on the whole archives have been slow to take the initiative in creating networks. Generally speaking it is the universities and learned institutions acting on behalf of their researchers that have been keenest to obtain wide access to up-to-date information in their special fields, and have led the way in sponsoring computer networks for this purpose. Direct sponsorship of networks by national archives is still a rarity. Networking in any case may not be the only, or the best, means of pooling data for every country.
National networks come in a number of different configurations. It is unnecessary here to describe more than two types by way of example:
- where a single 'host' computer accepts data files from a number of participating institutions, the files being so stored that they may be searched either separately, institution by institution, or cumulatively, by certain fields or prescribed terms. (See below, RLIN).
- where there is no central database but the computers of each participant are linked by telecommunications in such a way that the network may be used to 'call up' the contents of any one of their databases, which may then be searched individually according to the rules appertaining to that database. (see below, JANET).
In response to the questionnaire, very few national archives reported direct involvement in such networking.
Exceptions include the US National Archives which is now participating in RLIN, the Canadian National Archives which contributes data on picture conservation to the International Conservation Information Network, and the national archives of Hungary one of whose databases is made publicly available through the national public education information network. In the United Kingdom both the Public Record Office and the National Register of Archives are in the early stages of exploring possible participation in networks, and experiments are beginning in the Soviet Union.
But it would be profoundly unsafe to base any appraisal of networking and the present extent of its coverage on the experiences of the national archives alone, when for reasons indicated above it is the academic institutions which have often taken the lead.
In the early phase of computerisation in the USA there was a yearning in some quarters for the creation of a centralised national information system, perhaps based on one computer network which would contain (at least in summary form) information about the holdings of every repository throughout the country. The 'great database in the sky' was soon seen to be a chimera, given the enormous quantities of data that would be involved, the great variety of institutions to be covered, each with its own particular needs from any such system, and the complexity of some of the issues to be addressed. What has emerged in its place is a more federalist approach in which any national information system will be a partnership of several, perhaps many, sub-systems. There is an interesting parallel here with early aspirations to centralised description standards contained within one code of practice, which are being diluted in favour of a federalist approach which recognises, for example, the varying specific requirements for the description of different archival media, and the local needs of local repositories, while still seeking to reduce or contain diversity of practice and keep the separate parts of the system linked.
National networks have both advocates and detractors. The issues can only be resolved nationally by detailed discussion between the major archival institutions and the professional associations of archivists, and perhaps with representatives of user organisations In part the outcome must hinge on the successful identification of the users and their requirements (see p.28). In part it will depend on the archivist's vision of 'information' and of current or future opportunities for its dissemination and use. Overwhelmingly, however, it will continue to be governed by practicalities such as the availability of financial and manpower resources.
Among the advantages canvassed for national computer networks in archives are the following:
Wider dissemination of information about the archival holdings of individual repositories, as a supplement to or substitute for traditional means of dissemination through publications and finding aids. This may be combined with the ability to output information in new ways such as subject guides.
The possibility of merging or comparing information from several, or many, different archives. Apart from the obvious benefits to searchers, this may bring practical opportunities for archivists to compare and perhaps improve records management practices and standards of archival description.
The further possibility of merging or comparing archival information, irrespective of medium, with bibliographical information, and combining this facility with selective subject indexing to increase the points of access for the benefit of researchers as well as archivists.
Some potential problems, mainly associated with the first type of networking, need to be investigated more closely in relation to particular requirements and particular networks. They include:
Cost. The need to become a subscriber to the network, which may involve not only direct costs of membership and of data capture but also the need to conform to the network's requirements as to data standards. For small archives in particular, the cost of introducing new standards might be prohibitive in relation to any likely benefits from use of the system. And there is little incentive to contribute data, with a charge for input, if the contributor will not otherwise use the network.
Possible problems of over-centralisation, leaving participants at the mercy of the host organisation and dependent upon its continued commitment. There may be inconvenient periods when the entire system is shut down for maintenance, or 'crashes'.
Lack of 'policing'. Where the host organisation is not in a position to control or standardise the quality of the input data, or to verify or maintain it, there may be unevenness of approach among participants and possibly confusion over fundamentals such as the desired levels of description or depth of detail to be given.
Possible loss of control over the data and its onward use by other users of the network.
The research libraries information network (RLIN)
[This section is substantially based on information kindly supplied by Anne Van Camp, archivist of the Hoover Institution, including issue 18 of the Research Libraries Group Newsletter (winter 1989), an RLG press release of 16 March 1990, 'Vatican archives records to be accessible through RLIN', and a number of information leaflets on RLG and RLIN. Further information is obtainable from the RLIN Information Center, The Research Libraries Group Inc. 1200 Villa Street, Mountain view, CA 94041-1100 USA. For additional information and assistance the author is indebted to Nancy Sahli, David Sutton and Lisa Weber. (See also Reed 1985; Wallach 1990)].
The exchange of bibliographical information among national and academic libraries had already become commonplace well before any similar process was considered in respect of archives and manuscripts (Weber 1988).
In North America a number of national and international networks had been established with this primary purpose, including the Online Computer Library Center (OCLC), the Research Libraries Information Network (RLIN), the University of Toronto Library Automation System (UTLAS) and the Washington Library Network (WLN). The advantages of exchanging cataloguing data, to achieve standardisation, and reduce toil where the ground had been ploughed already, needed no further justification.
In 1983 Yale, Cornell and Stanford universities, members of the non-profit-making Research Libraries Group (RLG) that hosted RLIN, obtained a grant from the US Office of Education to enable them to undertake a pilot project entering data in the newly devised AMC format into RLIN. The network has never looked back. Many other libraries and archives have followed suit, and although other networks, especially OCLC, have taken the same path, at the time of writing RLIN, based at Palo Alto, California (Stanford University), has established a commanding lead in this field with about 250,000 records, generally but not exclusively at collection or series level, for the holdings of 90 institutions.
Established initially to serve the bibliographical needs of members of RLG, covering many of the larger academic and special libraries, RLIN through its adoption of AMC has enabled archivists not only to record their holdings for wider benefit but also, if they choose, to manage them using the archive and record control segment of AMC. The Hoover Institution, for example, has now recorded all its series by this means and uses the ARC segment to manage its holdings. It is also able to produce some 180 up-to-date subject guides on demand from information in the database. Other users too have commented on the redundancy of, for example, manuscript accession registers and registers of depositors in the wake of automation.
AMC has enabled RLIN to take on board data not only from the members of RLG, but, for example, from two successive projects on selected state records, sponsored by the National Historical Publications and Records Commission (the Seven States Project completed in 1988 and, more recently, the Government Records Project). RLIN also provides on-line access to the name and subject-heading authority files of the Library of Congress, whose use is encouraged throughout the network. In 1988 the National Union Catalog of Manuscript Collections began to supply current data for inclusion in RLIN.
In the same year the US National Archives became a member of RLG. Its published evaluation of a pilot project using RLIN to test the Archives and Records Control segment of MARC: AMC (Holmes 1986) remarked on the value of RLIN as a potential means for the National Archives to exchange information with other federal institutions, other archive repositories and national bibliographical databases, and for achieving wide general circulation at less cost than by traditional methods for information from the National Archives. Although, owing to the complexity of its own management requirements and its huge volume of archives, the National Archives felt it impractical to use the ARC facility, the state archive repositories and other participants have claimed substantial benefits including the ability to compare and standardise records management and archival description practices.
RLIN has the facility not only to establish on-line connections with mainframe computers elsewhere, but to accept data on tape from mini- and microcomputers, provided that it conforms to the MARC formats. This is currently opening the way for the importation of data from a number of sources outside the USA.
Discussion is under way towards the inclusion of the existing database of the Location Register of English Literary Manuscripts [Reading University, UK], and developments concerning archives in the Far East are likely, owing to RLIN's facility to handle Chinese, Japanese and Korean characters.
One of the members of RLG, the Bentley Historical Library of the University of Michigan, obtained funding support from the National Endowment for the Humanities, the Getty Grant Program and the American Friends of the Vatican Library, to send two staff members to Rome in September 1989 to gather series-level information about the Vatican Archives and those of the Propaganda Fide, together with additional information about other archives relating to the Papacy and St Peter's. Using a microcomputer equipped with the MicroMARC: AMC software developed by the university, they have entered data from the Vatican's disparate existing finding aids, returning the disks to the university for eventual loading into RLIN. As well as providing a unified database approach to these internationally important archives the project, scheduled for completion in 1992, has demonstrated the great advantage of combining private with public funding and of seeking international cooperation among archivists. (RLG press release 1990).
Reviewing the first five years of MARC: AMC in RLIN, one of its key officers noted that 'It has increased interinstitutional access to a very broad spectrum of primary source material nationally. It has provided an invaluable forum for cooperative activities among archivists and special collections people. And it has proved conclusively that you can integrate information about archival records and manuscripts into a bibliographic database'.
RLIN is accessible to individuals as well as institutions upon payment of a subscription. It can be accessed via a modem link through the GTE/Telenet communications network for a charge.
The joint academic network (JANET)
An example of the second kind of network is the Joint Academic Network (JANET) serving mainly universities and polytechnics in the United Kingdom. Participants can make available any computerised, databases under their control for the scrutiny of subscribers to JANET and can impose their own conditions of access and charges.
The format in which data is presented is left entirely to each participating organisation It is not necessary to use specified hardware or software, and the only requirements for extracting data are a modem link, the network's telephone number and a computer terminal equipped to receive and interrogate the data. The last point has caused some initial difficulties to users with incompatible terminals but this problem is easily resolved.
At present, in comparison with a network like RLIN, there are few archival databases available on the system, which mainly concerns the sciences and social sciences. But the number is steadily rising. Among them is the Wellington papers project, in progress at the University of Southampton, where the database contains not just series or item level descriptions of the first Duke of Wellington's papers held by the University Library, but extensive transcripts, insofar as they have been edited. The system at Southampton uses STATUS software which allows complete free text interrogation if required.
International networks?
Archivists, unlike librarians, have not yet established any fully international networks, although a number of examples cited above (eg pp. 8,23) demonstrate that national boundaries need be no obstacle to collaboration and to the exchange of data, and that networks of primarily national scope such as RLIN may have the capacity to take in data from other countries, and accept foreign subscribers.
In its response to the questionnaire Switzerland, alone among all the respondents, looked to the future creation of a network to link national archives. As more and more archival data is automated and databases held by individual institutions become more comprehensive and more representative of their total holdings, it is difficult not to see the demand for international networking increasing, both among archivists and perhaps more particularly among scholars using the archives for their research.
Some of the more recent developments in computer technology, and the effects which they may soon have upon the storage and retrieval of archival information, have been reviewed in the professional literature (Dollar 1986, 1990; Bearman, Optical media). Others, such as the application of artificial intelligence in expert systems, mentioned by Switzerland and the USA, offer scope for further exploration in future.
Image-processing systems
Optical and video disk applications, already widely established in business, are gaining a strong (if still experimental) foothold, both as archival media in their own right and, linked to computer databases, as a means of retrieving information from archives. They are being used respectively to store images of original paper documents, electronic data converted to this medium from a number of different computer applications (UK), and sound recordings (Soviet Union).
In a special venture to commemorate the quincentenary of Columbus's first voyage to the Americas, the Spanish ministry of culture with sponsorship from IBM (Spain) and the Fundacion Raṃn Areces has embarked on a project to scan, and capture on optical disk, initially some 9 million pages of text and all the maps of the Archive of the Indies' Seville, at an estimated cost of one billion pesetas, and requiring a storage capacity of around 5 gigabytes (Spain 1989).
The system will in addition absorb and assimilate by means of a relational database of 600 megabyte capacity, all the existing finding aids to the Archives, in printed, typescript or manuscript form. Access points will include provenance, repository call number, and document description, controlled by means of a specially constructed subject thesaurus based on terms and concepts intrinsic to the material.
Some logistical problems seem likely to arise in handling the many optical disks which will be generated by the project, but its results will be eagerly awaited by the archival world. The objectives of the project are to facilitate the wider dissemination elsewhere in Spain of information about the archives by providing copies of the disks or print-outs of documents as required, and to reduce wear and tear on the originals since they will normally be consulted from the optical disks at high resolution screens.
Although operations on this scale cannot readily be contemplated by most archives, a number of other archives, both national and local, reported using or being about to use similar technology.
The National Archives of Canada's ArchiVISTA system holds on optical disk the archives of the Canadian Centre for Caricature, some 20,000 cartoons whose scanning as a pilot project took 7 months. In digital form, however, these already occupy 30 optical disks. For practical viewing, video disks will be used for economy of storage. As in Spain, the data content as well as the images are processed by the same integrated system. Optical disks are also now being used in Canada for the storage of data from magnetic tapes.
In the USA an experimental optical disk system (ODISS) at the National Archives has been successfully tested on the military service records of Tennessee units that had served in the Confederate army in the Civil War (Dollar 1990). The equipment is capable of scanning not only the written or printed word but also images on microfilm, microfiche or aperture card.
The City Archives of Utrecht, Netherlands, has developed with WANG The Netherlands BV an archive information system ARIS for use on microcomputers, initially for the construction of a database of notarial records, linked to a Wang Integrated Image System containing the images of all items described in the database.
Similar developments or plans are reported in a number of other countries including Norway, Singapore and the United Kingdom.
Image processing systems offer a number of advantages including ease of distribution, the option to enhance the image of a faded or damaged original, the preservation of that original from wear and tear, the ready supply of copies and fast access to the data.
But apart from considerations of the initial capital cost, which are common to all such equipment, considerable difficulties of a practical as well as a technological nature may arise. A disk may be consulted by only one user at a time. One manufacturer's equipment is still normally incompatible with another's and today's 'reading' equipment may be replaced by a new series of machine in a few years, with the risk that disks may have to be copied or transcribed if they are to remain readable (Hedstrom 1988). Prototype erasable optical disks (in contrast to the conventional WORMS) have been announced by several manufacturers only to be withdrawn. For the moment this remains a field into which archivists should venture with some circumspection.
Multi-media systems
As the technology advances, a few national institutions are at the forefront of innovations, seeking new means of collating and presenting information for a public which, it seems, is increasingly image-dependent. This may result in the further weakening of some of the conceptual barriers hinted at elsewhere in this report, especially those between library and archival holdings on the one hand and between different archival media on the other.
At the Library of Congress (USA) the first steps have been taken in a multi-media presentation, entitled American Memory, on which a six-year pilot project was scheduled to begin in 1990 after trials in 1989. This will link LOC holdings of manuscripts, sound recordings, books, motion-pictures and graphic and photographic materials thematically in series. A master catalogue on CD-ROM linked to a microcomputer will be the initial means of access to associated compact optical disks for digitally stored information and video disks for analogue signals.
CD-ROM
Little other mention was made of CD-ROM applications by respondents to the questionnaire, but some see this medium as an alternative to hard copy publications of texts and finding aids including databases, and some as an alternative to networking.
The Genealogical Society of Utah, for example, now offers its International Genealogical Index and Ancestral File on CD-ROM. It identifies as some of the advantages for its own constituents' needs the ability to keep family research independent of a complex computer system and of networks, to avoid widespread inconvenience in the event of computer failure on-line, and to protect system access and important data.