Going abroad. FASTI and the “Only Connect”-workshop on historical databases for prosopography, Oxford, June 28 2001.

The unfortunate historian forced to work with advanced computing techniques for the managing and analysis of his data cannot get enough feedback. Therefore, the FASTI Scientific Network Group gathered an international team of researchers in University History at the Vrije Universiteit in Amsterdam on March 30-31, 2001. The goal of this workshop was to provide an answer to the question whether it is possible and desirable to create more uniform database structures for university prosopography. The advantages of such an endeavour, as the participants agreed, are twofold: first, using a common core structure for the storage of their data on academic populations, historians can save a considerable amount of time for actual historical research, instead of reinventing the wheel again and again when creating their own, project related, database structures. Secondly, it might facilitate, up to a certain degree, data exchange between collections, henceforth being structured according to common principles. Several presentations on existing databases showed that there were, indeed, common features in the more Model-Oriented Database Structures at hand, and that these elements might be useful when talking about a more uniform way of working with biographical data in university history. A steering committee was commissioned with the task to elaborate the meeting’s conclusions. For more details on this fruitful meeting, I refer to the report by Ad Tervoort elsewhere on the FASTI Web page.

An opportunity to present these results to historians active in other research fields, and to enrich the work in progress in university history with their experiences, was offered by the fourth edition of the “Only connect. Resourcing Sources” colloquium in Oxford, organised by Dr. Katharine Keats Rohan (Unit Prosopography, Linacre College). As the program announced, the meeting focussed upon the issues raised in electronically-assisted prosopographical analysis in general. Indeed a wide range of historians were gathered in the Centre for Humanities Computing, where electronic tools for late antique, medieval and early modern topics were presented. As some of the presentations were rather methodological and had nothing to do with history and computing as such (Altay Coksun and Ruth Paley), they will not be included in this report here.

A very decent introduction to the day’s topic was supplied by Hamish James of the History Data Service of Essex, on the designing of historical databases. Discussing familiar problems confronted by historians when starting to build their database, he stressed - as did the IT-specialists in Amsterdam - that technical decisions are often the least important. The main focus of the creation of a new database should always be its structure. But the importance attached by James to the structure of the source is significantly different from the FASTI point of view, that a prosopographical database aims rather at the creation of a meta-source than at the electronic transcription of the sources themselves. This difference is accentuated in James’ three layer model (sources, standardisation, interpretation), where he states that an accurate digital representation of the source should be the primary aim of every database. This concern for detail and context is, however, not at all irreconcilable with a Model Oriented Database Structure. All participants to the FASTI workshop had indeed agreed that direct links to source material are highly desirable. The difference in accent might be, partly, explained by the fact that a lot of sources used by university historians are quite repetitive, thus orientating the researchers in this domain rather to the standardisation and interpretation layers.

The next lectures brought the audience back to the hard practicalities of daily life in history and computing. The work of Paolo Renzi (Siena) will be familiar to the participants to the workshop in Amsterdam. He presented a vast, project-oriented, electronic card-box on Siena and Perugia professors 1250-1500, all fields on the one flat file per person being the reflection of a predefined set of questions. Michael Jeffreys (Oxford and London) and Dion Smythe (Belfast) respectively explained the materials of, and demonstrated, the Prosopography of the Byzantine Empire, a digital version of the books with the same title that have been printed for years. Their presentations showed that digitalisation of standardised biographical articles offered new possibilities, next to the advantage of more sophisticated indexes. By introducing ‘factoids’ (any phenomenon, action, … accompanied by at least one reference, in contrast with a ‘fact’ which is the product of the historian’s construction), categorised under several headings such as ‘locations’, ‘activities’, ‘ethnicity’, ‘education’, etc the Prosopography of the Byzantine Empire provides the scholar with quite advanced retrieval strategies, as a demonstration on the possible links between the participants to the Council of Nicea showed. Nevertheless, because of the centralisation of its production and its project-oriented character, the PBE cannot, in my opinion, serve as an example for a Repertorium Academicum that aims simultaneously at providing university historians with a core database structure for prosopographical research.

A very different approach was presented by Bente Opheim (Bergen). She tags her data on networks in medieval Scandinavia as they appear in literary sources in XML-codes (Extended Markup Language), thus preparing them directly for a publication in a web environment and swinging around the first step of storing them in a traditional database and ditto software program. As all data were derived from one source, reference facilities were not provided. Although this problem does not seem insurmountable, one might wonder whether this very flexible system is fit for less advanced users of computing techniques, and, therefore, for the public aimed at by FASTI.

Last but not least, Dr. Katharine Keats-Rohan (Oxford) demonstrated a version of the famous COEL database, initially set up for a prosopography of medieval land owners in England; later on, its structure has been adapted for the scrutiny of the obituaries of Mont Saint Michel for prosopographical purposes. Based on the principle that data should always be linked to their sources and that the interpretative role of the historian cannot be allowed to corrupt the data themselves, COEL pays a lot of attention to the edition of the original sources, reflecting, again, the (not inappropriate) preoccupation of historians, when working with less repetitive sources, with an exact representation of the context from which biographical data are derived.

The confrontation of these very different approaches on data and the way they should be managed with the ambitions of FASTI proved to be very interesting. In the course of demonstrating the Lovanienses Database - one of the MODs presented in Amsterdam – the present meeting consistantly switched back to the Amsterdam gathering. That way the possibility of the integration, up to a certain level, between individual needs and common features when working with electronic databases in a specific research field, was highlighted. Nevertheless, the audience did not seem prepared to accept the idea, that a more Model Oriented approach on data management might work in another branch of the historical métier. This is, in my opinion, partly due to the fact, that not in every room of Clio’s spacious house the same needs are necessarily felt. As a matter of fact, the continuous and time-consuming “reinvention of the wheel” did not seem to bother most of the participants of the workshop in Oxford at all. And data-exchange, an important issue among university historians, is, apparently, not one of the deepest concerns of all prosopographists.

Bruno Boute

Research Assistent

Fund for Scientific Research

University of Leuven