TEI AI4 P9 Final Report of the TEI's Working Group on Historical Studies (AI WG4) 5 September 1991 The report of the working group on historical studies (AI WG4) is available in print in Daniel Greenstein ed., <> (St Katharinen, 1991), and in machine-readable form from Dr Manfred Thaller. As explained in the introduction to the volume, (Daniel Greenstein, v-x1), the objectives set out for the working group in AI4 P1 (9 Oct 1990) and the aims of the working groups generally as described in EDW 14 (19 Sept 90) required agreement amongst historians represented in the working group about the list of documents, sources and text types which were of particular interest to historians, and about the analytical and interpretive procedures and types of encodable information required and used by computer-literate historians in their research. Thus, before the aims of the TEI could be met, it was essential to make some headway in constructing some general conceptual model of historian's data modelling and processing requirements. On behalf of the working group, therefore, I would like to thank the steering committee of the TEI for providing a concrete focus for similar investigations which have been ongoing without much success since the early to mid 1980s. In presenting our report to the TEI it is perhaps best to follow concretely the objectives set out for the working group on historical studies as they were set out in AI4 P1 and to show how and where these objectives were met. I. Formulate a list of document, source, or text types of particular interest to historians. This was done at the meeting of the group which took place at Southampton University and for which minutes have already been furnished to the TEI. There, members of the working group drew up a short-list of documents whose structure and contents were distinctive because they presented sufficiently different problems to computer-aided historical research. That list of sources is set out in the minutes of the Southampton meeting and in the volume <> on p.104. Many of the sources appearing in the list received more extensive treatment by members of the working group and this is presented in part II.A. of the volume, pp.111-94. II. Formulate an account of important analytical or interpretive procedures, and the types of (encodable) information they require, which are of particular importance to the historical disciplines. Treatment of this subject is available in the volume in two places. In part I some considerable effort is made by Thaller and Greenstein to give such an account of analytical and interpretive procedures and types of encodable information which is likely to be encountered across a wide range of distinctive documents which are of particular to historians. We recommend in particular Dr Thaller's, "A Draft Proposal for a Standard for the Coding of Machine Readable Sources", 19-64 and Dr Greenstein's "Encoding Standards for Computer-Aided Historical Research: The Problems Reassessed", 93-110. The former article is aimed at the development very specific processing requirements which may or may not be compatible with the TEI's aims to develop the appropriate SGML tags and DTDs. None the less, it is the best account to date of the structure and processing requirements associated with particularly problematic elements found in most historical texts (e.g. names, currency measures, place names, etc.). Although the processing recommendations may or may not be of any use, the conceptual work which underlies them needs to be taken into account by the steering committee. The latter article constructs a conceptual model of historian's data modelling and processing requirements into which the elements considered by Dr Thaller are integrated. In part II.A. of the volume, further consideration is given to the specific types of encodable information which is found in the documents considered there. III. Make a preliminary evaluation of TEI P1 as to the facilities it provides for handling these text types, procedures, and ancillary information. Most of the chapters in section II.A. make some general comments about the strengths and weaknesses of the proposals set out in TEI P1 with respect to the document types being considered. In section II.B., Dr Greenstein's "Conceptual Models and Model Solutions: A Summary Report of the TEI's Working Group on Historical Studies", 195-204, summarizes the recommendations which appear in section II.A. and which emerged in the meeting of the working group at Southampton. IV. Formulation of a list of work groups needed to create document type descriptions, tags, and documentation for the encoding of the text types not now adequately handled by TEI P1, or tags and documents for analysis and interpretive information of a particular interest to historians. For each working group, specify a list of concrete objectives and requirements, and suggest deadlines for the work. There are two directions in which future work might proceed the first of which is most likely to meet the TEI's production schedule and the second of which is likely to take place anyway as an activity of the international Association for History and Computing and from which the TEI will undoubtedly benefit. Following the first direction, experts in SGML who are closely associated with and ideally chosen by the co-editors of TEI P1 will develop SGML tags and outline DTDs based on the recommendations contained within this volume. Tags and DTDs once developed should be re-submitted to the members of the working group for evaluation and comment. In addition, contacts have been made with at least two computer-scientists interested in developing and testing the TEI's proposal against historical sources. Peter Flynn at the Computer Centre, University College, Cork, Ireland (e-mail=cbts8001@iruccvax.ucc.ie) represents a project which is already funded and about to get underway creating a machine-readable version of early and medieval Irish literature. Since the project hasn't any other analytical aims than to make a machine-readable corpus available to researchers of whatever complexion, the working group feels that it is ideally suited to test the TEI's proposals. Mr Giordano's plans are rather more fluid at the moment. He is eager to develop and apply the TEI's guidelines to machine-readable versions of historical documents and is currently seeking funding and the appropriate documents with which to carry out his research. The second direction recognizes that historians were on the whole so lacking in any conceptual underpinnings of their computer-aided historical research as to make the work presented in <> very preliminary indeed. Consequently, we would hope to convene several working groups to consider areas which are only touched upon very briefly in the volume but which can build upon the work already presented therein. Under this scheme, six working groups would be set up to consider each of the types of information identified by Thaller ("A Draft Proposal") and referred to by Greenstein as primitive datatypes ("Encoding Standards"). Thus, a working group would be set up to consider the modelling and processing problems presented by geographical place names, currency measures, dates, personal names, etc. Each of the working groups (there would be six) would be set up along lines similar to that of WG AI4. Every group should consist of approximately six people (including an organizer) with a medievalist, an early modernist and a modern historian represented amongst them. Each of the working groups should consider its work at two meetings - a preliminary and a working meeting. The working meetings would be divided into two sessions: one for modelling and one for processing problems respectively. At the working meeting, each member would pre- circulate other members with a paper on a topic agreed jointly by the working group at the preliminary meeting. At the working meeting each paper would be presented by the author in 5 or so minutes and discussion initiated by an agreed commentator. The papers so presented and discussed would then form the basis of publications such as the one presented here with the working group organizer acting as editor. Each publication would then be pre-circulated to AHC members participating in the annual international conferences and discussed at these conferences in public workshops. One additional working group should also be created to flesh out the general treatment given to document types (Section II.A) in the present volume. Whenever the publications of the seven working groups were available, their convenors would meet together and plan to return to the themes touched upon in <>, and to re-write that book based upon the more extensive specialist investigations of each of the working groups. At the recent conference of the AHC in Odense, Denmark, it was agreed that future work such as that outlined above was essential to the integrity of historical informatics as a sub-field of history and some considerable efforts were made to enlist individuals to host, organize and finance particular working groups. Over the next several months much more organizational work will be conducted and we hope to have the seven working groups established and some already in train by 30 September 1992. Clearly this second and more comprehensive scheme for continuing the work begun by WG AI4 will continue well after the TEI has produced its final proposals. None the less, it is hoped that convenors of the aforementioned working groups will continue to inform the TEI of its work wherever possible and likewise to be kept abreast of developments within the TEI. V. To solicit participation in WG AI4 by North Americans. Here the working group has failed to meet the objectives set out by the TEI in AI4 P1. It is hoped that responsibility for this failure might be shared jointly between the working group chairman Dr Greenstein who did not seek contacts within US historical societies with which he has very little contact, and with members of the steering committee who, according to Dr Greenstein's notes from the meeting of working group heads in Oxford, offered assistance in placing him in touch with relevant US historians. It was agreed this summer, however, that Mark Olsen at the University of Chicago or any other US historians known to the steering group might be approached to evaluate the minutes of working group's Southampton meeting and <> and to comment on these in light of TEI P1. Daniel Greenstein Convenor, TEI Working Group on Historical Studies (WG AI4) Glasgow University Glasgow