Recasting the Past: Digital Histories
For the last three years the RHS has held a seminar in memory of Gerald Aylmer, the purpose of which is an exchange of ideas between historians and archivists. Digitisation was to be the subject for 2004. The AHC proposing a conference on the same subject, the two bodies came together to present this joint event. The format of the conference was that of four sessions, each with two panels, except for the last. I attended Panel A in each instance.
Welcome
The proceedings were opened on behalf of the AHC by Dr Ian Anderson, convenor of the AHC-UK Conference, and lecturer at the Humanities Advanced Technology and Information Institute (HATII), University of Glasgow. Ian pointed out that digitisation presents considerable possibilities for the presentation of history, and transcends all traditional boundaries; but he also asked what implications this has for the interpreter of history. Moreover he voiced three concerns commonly expressed:
- how to select and then prioritise material to be digitised
- which techniques to adopt, including the possibility of losing or inserting data
- access considerations and sustainability
Dr Christopher Kitching, Chairman of the Research Policy Committee, then welcomed participants on behalf of the RHS, to what promised to be an interesting and thought-provoking day.
Session 1:
Chaired by Christopher Kitching
200 Years of Migration History on One Web Site :
Elizabeth Lovell, TNA, Project Manager for the Moving Here Project
This session commenced with a very informative presentation about the Moving Here Project [1], hosted on TNA's Web site. Moving Here is an exhibition of digital images with supporting narrative on the subject of migration to the United Kingdom. It provides availability and interpretation, and is particularly aimed at a traditionally under-represented group among those who access TNA and its record holdings. Some 30 archives, museums and libraries contributed to Moving Here, and there was also a large contribution of stories from community groups. It has also generated country-wide exhibitions, and has, thus, provided outreach far beyond the traditional activities of an archive. The impact has been enormous, and feedback good. Looking to the future, sustaining the site, and medium and long-term strategies were big issues, and financial support would be vital here.
The National Archives, the State Papers, and the World Wide Web
Helen Good, University of Hull
Helen Good is a PhD student digitising the State Papers Domestic of the reign of Elizabeth I (series SP12) which are held in TNA. Other Elizabethan State Papers, not subject to this exercise, are held in the British Library and at Hatfield House. Helen's intention was to move presentation of the records from reliance on the original calendars to the presentation of digital images of the originals. Her thought-provoking talk encompassed details of the project, the archive user's view of this procedure, and problems encountered. She felt that the calendars would never be adequate for the academic user, even where they were scholarly productions, as users did not want their material pre-digested. She did not even feel, for example, that the calendars relating to the SP 12 records were particularly adequate. She reported she was taking digital images of all the records, transcribing them and putting them up, under non-commercial licence from TNA, on the Hull University Web site [2] [3]. While she accepted that TNA was a large and complex organisation with many considerations that needed to be prioritised against each other, Helen said she would like to see TNA undertaking this type of activity, thus giving due consideration to academic users, as well as to the more numerically significant other users.
Funding, Availability and Access
Aidan Lawes, TNA, Academic Publications Manager
Aidan Lawes began his informative and measured presentation by describing the academic publications remit and programme within TNA. The purpose was to produce resource-based packages for the academic audience. Projects within this required external funding and were co-published either with an academic or commercial partner. Internal publications within TNA covered the more popular market. There was a fine tradition of academic publication within TNA, and one question now being asked was whether there still existed a market for the traditional paper publication, especially as these were expensive and slow to produce. Clearly much thought was now being given to electronic publications as well as digitised images of documents, but one of the big questions here was sustainability - a calendar would still be going strong after 200 years, but would a CD? Furthermore, not all academics felt that digitised images addressed all their requirements. Moreover these images were not cheap to produce. There was even the feeling that some academics bore hostility towards electronic sources. Charging mechanisms were a factor here. Another big issue for TNA was how to prioritise among the products of 1000 years of record keeping. Advisory Panels and an Editorial Board now helped to inform and advise in this area, and Aidan welcomed any suggestions which might be fed back to these bodies. Most partnerships in TNA were with medievalists, who seemed to be more 'joined up' than others, and, again TNA would welcome overtures from other areas.
Session 2:
Chaired by Dr Sarah Richardson, University of Warwick
Digital London: Creating a Searchable Web of Interlinked Sources on Eighteenth Century London
Professor Robert Shoemaker and Jamie Maclaughlin, University of Sheffield
This presentation provided a fascinating introduction to the proceedings of the Old Bailey online [4], and the very valuable linkages of data employed. This project was now almost complete. In addition to the proceedings both local authority and charity records have been used to provide additional related information. Links have been made from the trial proceedings to these associated records, thus enhancing the basic Web product. Robert Shoemaker highlighted the problem of dealing with name spellings, and how these were standardised, but linked to variant spellings. Place names have been mapped, as partial and exact matches, using Geographic Information System (GIS) maps. This meant that all instances of a much used name, such as 'Stable Yard', could be brought up. It was also possible to bring up all the trials relating to a certain place. Linkages could also be made via occupation. The emphasis has been on providing the maximum amount of information, with the minimum of assumptions and allowing for full interpretation and use of judgement on the part of the user or researcher. Care has been taken to preserve the original archival structure of the documents and to avoid taking them out of context.
Digitisation Selection Strategies - Preserving and Uncovering Hidden Histories
Jacqueline Spence, University of Wales, Aberystwyth, Archives Administration and Records Management School
Jacqueline Spence's paper looked generally at some of the issues surrounding both decisions to digitise, and born-digital records. She highlighted the fact that keeping all data in the 21st Century is a real issue. The challenges were to use technology productively, expand human understanding and widen our scope. At the same time we had to protect history and prevent censorship, as well as creating a cultural memory. However, there was also the opportunity to bring disparate material together, facilitate comparative research and allow for deeper analysis more easily than ever before. This was aided by the fact that many projects are international. Expectations might well be considerable, and had to be matched. Again the issue of sustainability was raised. Finance to set up projects was not hard to come by, but sustaining them might be. The main issue with born-digital records was access not preservation; they were a new form of primary source for historians to deal with, and could be manipulated in new ways. They were often disconnected and unfocused, and their ownership was not always clear, and would provide challenges for individuals and small organisations. The catchword for their preservation ought to be devolution, not de-selection. Technical service centres to help with migration and access would be welcomed. She concluded that history was now much more immediate, and this gave the user a new role; more social responsibility was required than previously.
Preserving the Historical Record
Dr Matthew Woollard, Head of Arts and Humanities Data Service (AHDS), History, University of Essex
Matthew Woollard provided a very useful and informative description of the remit and activities of the history strand of the AHDS [5], which is one of five strands, and is funded by the Arts and Humanities Research Board (AHRB). The AHDS provided the balance between preservation functionality and the historical functionality. Mostly AHRB-funded projects were preserved here among the 40 years' worth of data collection in more than 600 collections. The aim is to preserve the bits, the information content, and the experience (layout, display etc), with the remit of preserving the content and enough explanatory material to enable it to be used. The AHDS strategy is to keep the original version of material presented, create a preservation version, and, from this, to create dissemination versions. Part of all this is the history of computing. Material is also acquired which is then re-formatted to provide data to the academic historian. An example of this is the database of the Prerogative Court of Canterbury wills, digitised by TNA. Matthew also spoke about the themes represented. Most material was statistical; currently, political, social and economic material was variously in decline.
Session 3:
Chaired by Prof Ludmilla Jordanova, University of Cambridge
Building an Ontology for History: Lessons Learned and A 21st Century History Portal: The Semantic Web and Contextualising European History
Richard Deswarte, University of East Anglia, and Dr Jan Oosthoek, University of Newcastle-upon-Tyne
These were two very thought-provoking papers inviting participants to consider the complexities of digital searching and hyper-linking. Jan Oosthoek summarised the problem as one of how you make searching sophisticated yet helpful, and how you know whether you have all sources. The European Union-funded Visual Contextualisation of Digital Content Web site (VICODI) [6] addressed this issue. This was a visual contextualisation portal, providing contextualisation for history and the ability for the users to build their own queries using a historical ontology. Contextualisation considered how terms or documents related to each other. For storage, retrieval, presentation and navigation, the system used a methodology called LATCH, standing for Location, Alphabetical Order, Time, Category and Hierarchy. It weighed the relevance of references and scored them. (For example, in a search for 'Newton', 'Newton', '1660-1680' and 'gravity' would have high ratings, while 'apple' would have a low rating). Afterthoughts on this project were that it was over-complicated and made some false assumptions about history, ontologies and thesauri. There should have been input from historians, and different components should have been separately developed.
Richard Deswarte described the development of the ontology, defined by him here as a formal knowledge structure. The challenges were the quantity of historical sources, how to populate the ontology, and carrying out the quality control. The successes of the project were the opportunities it provided for a wide grasp of the complexities of history, the use of 2000 historical documents, and good quality control. The problems were that it was very labour-intensive and expensive, the concepts were very complex, there were limited property relationships and extensibility was an issue.
The Role of Hypermedia History
Dr Isobel Falconer, Open University, and Dr Ian Anderson, University of Glasgow
This was a very interesting presentation of a concept which was, I suspect, novel to many of the audience. Hypermedia represents a conflation of hypertext and multimedia. It offers an alternative to traditional linear narrative, using all the potential of Web-enabled developments, and is, above all, a teaching tool. Isobel Falconer commenced with a concept of 'good history', which was that which included no more than was needed to explain past events. Whether the historian was regarded as an interpreter or enabler affected the criteria by which we evaluated the understanding of the audience. It also had implications for the ways in which historians would communicate. Different modes of understanding were best suited to different sorts of media, of which hypermedia was clearly just one. By their choice of the medium and architecture of their communication, historians would be seeking to control their audience. Use of hypermedia may seem to be donating control to the audience in a large number of areas, but these were effective only if the historian had built in hidden controls, including an appropriate architecture and adherence to protocols and standards. For communicating a large number of interacting causes, or for historians wishing to act as enablers, the multiple links of hypermedia might be appropriate. Hypermedia were good at linking multiple signs, but currently were weak at showing the nature of the links. The cost of designing the architecture meant that hypermedia history was very expensive compared with narrative history. Realising the potential in this area was still in its infancy.
Session 4:
Chaired by Dr Ian Anderson, University of Glasgow
The Effectiveness of Various Retrieval Strategies for Digital Archives in Ireland
Dr Andrew White, Archivist, Public Record Office of Northern Ireland
This paper described several digitisation exercises, and some interesting conclusions were drawn from these. The records were digitised to preserve them, to widen access, and to improve research efficiency through sophisticated search strategies. There was a perceived democratisation of archives through bringing together geographically diffuse archives. This saved users physical travel, but it also generated a demand for original records. It was also perceived that the historian was able to add multimedia features without compromising academic value. The input of historians was very important, as they determined the research needs of users, developed a controlled vocabulary, quality assured the products, and highlighted areas which could be enhanced with annotations. Retrieval was by bibliographic elements, by special features, and by use of natural language searching incorporating a controlled vocabulary. Fuzzy searching was also incorporated. These projects were deemed a success as the use of OCR (Optical Character Recognition) and re-keying guaranteed a high level of accuracy, a wide interest was generated, and audiences had been broadened.
Re-exploring Alnwick: How Digitisation of Data Can Give a New Insight on Methodology
Dr Elwin Koster, Groningen University, Department of Humanities Computing and Department of Art History
This last paper was a very interesting insight into something totally different from previous presentations. The urban geographer MTG Conzen carried out two surveys in Alnwick in 1953 and 1964. Elwin Koster digitised these two surveys using a GIS, and then mapped these against each other for comparison. He then carried out a survey of his own in 2003-2004 for further comparison. In particular, he highlighted difficulties with descriptions particularly, such as trying to match fast food outlets or charity shops against the 1953 and 1964 retail use categories.
Summing up
There were, fortunately, opportunities for comments and questions at the end of each session. The main issues raised in the summing up were that funding is a crucial issue, particularly for sustainability; in particular, what would happen if funding for digitisation were to dry up? Preservation is more important than digitisation for its own sake. It would be useful if figures of influence were to be drawn into the debate and proceedings in the area of digitisation of historical records. The choice of content, and prioritisation of content are major issues. Historians need to take the lead here or commercial considerations will dictate decisions. In particular, there is a need for input from contemporary historians.
Conclusion
This was an interesting and wide-ranging conference, and set an agenda for future discussion and consideration. Not only was the subject matter extremely varied, but the areas covered, from the factual and practical to the theoretical and technical. I'm sure that there was something here for every participant, and I only wish it had been possible to attend all the sessions!
References
- Moving Here Web site http://www.movinghere.org.uk/
- Transcriptions Web site http://www.hull.ac.uk/sp12/
- Images Web site http://www.hull.ac.uk/sp12/images
- Proceedings of the Old Bailey Web site http://www.oldbaileyonline.org/
- AHDS Web site http://www.ahds.ac.uk/history/
- VICODI Web site http://www.vicodi.org/