There are projects in research, which are frequently discussed at conference sessions, and even more frequently outside of the sessions, as prototypes of a given kind of method. One of these has over the years, well, decades, undoubtedly been Michael Gervers' work on the computer supported analysis of medieval charters, leading to the DEEDS project at the university of Toronto. At the time of the writing of the book to be reviewed here, it contains 3353 dated Latin charters from twelfth- and thirteenth- century England. When projects are discussed that way, it is always valuable to present them to a forum of other specialists to evaluate their findings and compare them systematically to similar--or intentionally dissimilar-- approaches. A meeting like that has been held in Budapest in 1999; these are its acta.
In the case of the present volume, this strategy has worked admirably. This reviewer has to admit that most of his recent expertise is in the area of computer applications in the Humanities and his medievalist expertise is less than recent; so necessarily his view is somewhat biased. Nevertheless, or maybe just because of that, he is deeply impressed by this volume as a publication which is admirably fitted to introduce students to what research about medieval history is about, once it goes beyond the stage of enthusiastic story telling. Indeed, there may be few collections which would allow (admittedly advanced) students to find on a reasonably limited number of pages such a broad introduction into the diversity of methodological tools available for tackling clearly focused substantial problems like dating the undated. And, while the academic quality of the papers is unquestionable, the huge majority of them are written in such a clear style that even non-native English- speaking students did actually enjoy them, when asked to read them for seminars.
I started with this, because it is probably the most surprising when hearing about a volume on the chronological location of documents, not normally considered a very glamorous topic.
The collection starts with a description of methodology and major findings of the DEEDS project under the self- explanatory section heading Dating by Word-Pattern Matching. "Word patterns" in the sense used here, is a very good translation of formulae as the older hilfswissenschaftliche literature uses them; formulae or word patterns, however, which are not derived from the rules of a specific chancellery, but from the observed fads and fashions of the total corpus of available charters, that is, by a thorough analysis of the frequency of words and the patterns of their co-occurence in the whole corpus. In the section Michael Gervers himself gives a general overview of the methods used and the results achieved by DEEDS, while Rodolfo Fiallos explains some of the technical and statistical details.
The following section, entitled Dating by Formulae and Vocabulary provides a very obvious base of comparison for the first approach presented: precisely the methodological tradition, from which the generalization via database usage has been derived, and a possibility for comparison, which in itself makes the volume very worthwhile reading. Specifically, that comparison involves the contributions of Marjorie Chibnall, working on charters of smaller religious houses in Suffolk, Veronique Gazeau discussing Norman material and Nicholas Vincent, revisiting the charters of Henry II. Benoit-Michel Tock in this section deals himself with a project which uses another database-based approach, applied to Belgian material.
The Diktatvergleich, from which the basic concepts come, was originally not introduced for purposes of dating, but rather for the detection of forgeries; a context which is provided by the following section Identifying Forgeries. Georges Declercq experiments with harnessing the DEEDS method of comparison to that goal, applying it to a forged charter of an abbey in Ghent. Zsolt Hunyadi and Attila Zsoldos discuss forgeries of Hungarian documents, in the first case in the charters issued by a Hungarian convent of the Knights of St. John of Jerusalem, in the second studying the charters issued by Hungarian queens of the Arpad dynasty in the 11th-13th centuries.
Section 4 enters a new element by leaving methods based on the analysis of the vocabulary or the formulae, and discussing Dating by the Association of Names. Maria Hillebrandt retells the story of the analysis of the charters of Cluny, developed at Munster by the research group headed by Joachim Wollasch in the early eighties. Here the formulae are replaced by patterns of co- occurences of names in lists, e.g., lists of witnesses. It is a bit sad, that not very much more than the achievements of the eighties can be recapitulated here. So it is very satisfying, that Trevor Chalmers can continue with a similar approach for the analysis of such co-occurences of names, which seems to be rather promising as a possible extension of the model proposed by DEEDS. This approach claims to be more modest than the Munster one; in any case, it works today. A distinguishing mark of this contribution is an appendix with examples, which makes this approach particularly transparent.
The next section, Palaeography and Sigillography, is clearly described by its title as the domain of two classical non-textbased methodologies applied to the solution of dating problems. Laszlo Veszpremy shows the paleographical handling of dating problems based on Hungarian manuscripts from the 11th to 13th century, P.D.A Harvey provides a 3 page statement on the relationship between seals and dating.
A final section Comparative Methodologies leaves the realm of historical methodology in the usual sense. Andras Grynaeus gives a very well written introduction into dendrochronology which is very instructive for this link between document derived and "objective" timelines. The next contribution on the role of fossils and geological stratigraphy for the chronology of the period roughly 180 million years before the present extends the traditional definition of the Middle Ages somewhat and a 2.25 page contribution on the existence of a collection of medieval charters in the National Archives of Hungary--thou, who never hast edited a conference volume, throw the first stone.
Leaving the geologically biased one aside, all the contributions relate to the 10th-13th centuries, with a clear focus on the 12th and 13th.
One potential audience, in the hands of which this reviewer would like to see this volume very much, has already been mentioned: students, particularly in the early postgraduate stage. Again: well written, diverse, but with a clear focus on a well understood area of medieval studies methodology, where traditional and modern approaches blend most convincingly, this can be a model introduction into what scholarly work is about.
That didactic praise should not obscure the fact, that at the same time it is a very good overview over an area, which should be of interest to anybody who is interested in the systematic study of medieval charters as a category of sources worth studying beyond the merit of an individual document.
This reviewer is first and foremost a specialist of computer applications in the Humanities and historical studies in particular. The reason that this review does not say very much about computer techniques is very simple: they are mentioned only very lightly in the volume, basically as background for the substantial results; a reader who would want to know, how exactly to start her or his own DEEDS project technologically would be disappointed. One fact, however, should not be overlooked: the title of Michael Gervers' introductory paper starts: The Deeds Project and the Development of a Computerized Methodology.... The less-computer-oriented medievalist may not notice that this in some ways reads very unusually today. To use computer- based analytical tools, you need data and you have to have software for their analysis: you may even have to develop it yourself. Quite opposed to that theory we have been told in recent years with increasing insistency, that encoding documents in SGML (and more recently XML) was all a Humanities scholar would have to do, before waiting patiently that somebody would create the software needed without the intervention of the Humanities researcher her- or himself. Some projects have waited for that somebody to come along since the late eighties now, and still, he has not arrived yet. Michael Gervers and his team have.