Internet Archaeol 15. Diana Murray. Philosophy of the computerisation strategy

2. Philosophy of the computerisation strategy

In Scotland, the decision was taken early on to ensure that the indexes and catalogues to all the material held in the NMRS, however diverse, was integrated so that users could cross-search or easily navigate between different elements of the data. This formed the philosophy behind the computerisation of the record when it began in the 1990s and all new initiatives and projects have conformed since that date.

What seems obvious now was a difficult concept to promote in the early years of computerisation when databases were first used to replace card indexes and paper catalogues. There was strong pressure to find different solutions to each part of the record that would allow the manual systems to be replicated using computers, and there were some that believed that no computer system could ever cope with the manual methodologies that were in operation. The differences between archaeological and architectural recording methods were of particular concern at the time. Without having real technical solutions, it took a great leap of faith to visualise an integrated system that one day would have links to other records, and incorporate not just the information on the card indexes but also the maps and images. What seemed impossibly out of reach only ten years ago is now almost taken for granted, and recording practices have changed as the opportunities and benefits of technological innovation have become apparent.

In Scotland, it was possible to integrate from the start and this was the data that determined the structure of the first computerised record in NMRS. In England and Wales, separate databases were set up for different sectors of the records, but today, the advances in technology are allowing these records to become integrated also.

The first data structure set the scene for the future, incorporating important principles of site location, classification, authority for the record, informative text, catalogue details relating to archive. It was more than a basic index but less than a fully digitised record. It incorporated the principle of text descriptions, firstly derived from the OS cards but subsequently derived from RCAHMS field visits, and now includes text describing items in the collection from aerial photographs to architectural drawings.

The computerisation of the archaeological records into database format opened up retrieval possibilities that were formerly impossible. For the first time data could be analysed in an efficient manner. The impact on research has most significantly been to enable users to create a corpus of information prior to more in-depth research. An early benefit of this was the revision of the Roman Britain Map in the early 1990s. Former revisions had involved going through every card in the index and creating a separate card index of Roman material that could then be used to revise the map. Following computerisation, this information could be retrieved rapidly and time could be spent analysing the results rather than sifting out the data. The most recent revision (published 2001) used GIS technology. Another example, using GIS technology, was the creation of a corpus of archaeological sites that fell in wetland areas, a task completed in a few hours that would previously have taken weeks. The NMRs encourage feedback from detailed research of this kind to ensure that the record is updated.

The computerisation of the catalogue to the architectural collections started in 1996 and was fully integrated into the database structure. It was decided to undertake a major upgrade to the data content as data entry progressed so that the work is still only 50% complete.

One of the consequences of the initial computerisation was that the manual indexes and catalogues form only part of a much more complex collection of items that operate together to make up the complete documentary record. Significantly, the annotated maps play a key role in information retrieval and indexing as well as in spatial illustration. GIS was the obvious solution to the integration of spatial data and research and implementation began in 1992.

At the same time, solutions were being sought for the complex issues of recording landscapes in the field. The limits of recording technology had been reached, while pressure from the archaeological community to recognise and record landscapes of field systems and settlements was increasing. Technological solutions of EDM and later GPS surveys have come to the rescue, not only allowing the desired recording to take place but themselves opening possibilities not previously conceived.

The use of GIS allows such survey material to be loaded directly into the NMRS where it no longer requires cataloguing through the constraints of the database and cataloguing methodologies, but becomes the record itself, able to be linked through the technology to other elements. Desk-based studies are also in place to develop an Historic Landuse Assessment for the whole of Scotland and similar historic land characterisation programmes are beginning in England and Wales.

It is remarkable that, while the landscape of Britain is now being recorded in more detail and more effectively, heritage management both at national and local level is still mainly centred on unitary monument protection.

GIS also provides the solution to the integration of vertical air photography collections. By digitising the flight plots and displaying them as a GIS layer, it is possible to use them interactively with all other parts of the computerised record.

The GIS frees the data from the tyranny of the database. The NMR records that lend themselves initially to computerisation are the indexes and catalogues, and the database constrains the data into fields in order to make retrieval efficient. This is not suitable for all data, especially complex landscape studies. While text searching is possible, it is inefficient and random unless the text is formulaic. As spatial information is at the heart of all the NMRs it is axiomatic that a well-designed GIS should also be at the heart of the computerised data, as indeed it always was in the manual systems. GIS is now fully operational with the NMR database though not yet web-enabled to work with CANMORE — a task scheduled for the current year. It is being actively used for all kinds of analyses, most recently to plot the distribution of the work of architects represented in the collections that are part of the Scottish Architects' Papers Preservation Project (SAPPP). This is possible because the cataloguing system set up for the SAPPP project conforms to the integrated model and included site locations for every item catalogued.

So exciting are the possibilities that, initially, the downside was not appreciated. As computerisation was developed, more users wanted data to be analysed, and no visitor could access the information held in the computer without the intervention of a member of staff to operate the database on their behalf. The two consequences of this were that visitors either used out-of-date card indexes and catalogues for their research or staff were increasingly tied up with enquiries.

This alone provided the impetus to develop a self-help system for the public, initially in-house and subsequently online — CANMORE.