A computer programme for processing Greek and Latin epigraphical texts.

The world of epigraphy has a long scientific tradition, which goes back to the 19th century. This was precisely the time when the Berlin Academy launched the huge enterprise of the Corpus Inscriptionum Latinarum and Corpus Inscriptionum Graecarum, (which later became Inscriptiones Graecae). The corpora gathered, or were supposed to do so, all the documents of a specific region. This brilliant project has never been abandoned and continues to be actively pursued, but a variety of circumstances, primarily political ones, have prevented it from fully reaching its goal. Over time, many other corpus volumes and also some very serviceable Repertoria-style collections of inscriptions have been added, which help immensely in the daily work of scholarship, and offer convenient access to material otherwise scattered in an infinity of books, periodicals, and other publications. Publication in the form of a book, which until recently has been the only method available, suffers, however, from its necessarily static character. New inscriptions are constantly being discovered but updating is impossible, except in the form of a new edition - a possibility excluded in the majority of cases by the difficulty and cost of the enterprise.

In the computer age, new forms of publication are becoming available which make it conceivable to reach the original goal of providing up-to-date collections of documents from all the different parts of the ancient world. The PETRAE programme was launched by the Centre Pierre Paris, now the Ausonius Institute, of the Université Michel de Montaigne-Bordeaux III, to help towards this aim.

The project takes the form of a database programme, written in the 4th Dimension software environment for Macintosh. This programme is supplied at no charge, provided that the recipient, individual researcher or institution, agrees to contribute for a specific region to the above mentioned database by making available the file when it is completed. The PETRAE team will then do its best to publish the information in an electronic format - the Internet seemingly being the most appropriate form to achieve this goal. Meanwhile, users will always have full liberty to produce books, articles, and other publications at their own expense and responsibility, using the automatic edition routines included in the PETRAE programme, making sure that in doing so copyrights will not conflict with the electronic Internet database publication eventually envisaged. A commitment to regular up-dating of the database would also be highly appreciated.

The PETRAE programme allows a full treatment of all the aspects of an epigraphic document. Every form is identified by its author, with a record of date of entry and, if carried out, of revision of the document. The basic data concern the stone and text. The place of origin, material, decoration, state of preservation and dimensions, are the main components of the description of the stone (or other object). Photos or other graphic documents (at this stage and in the present state of the database, in PICT format) can be stored and accessed at will. The text is then described by defining its origin (which may be different from the place where the stone was set up, viz. a decree of Histiaea in Delos), letter shapes and dimensions, and principal contents (fig. 1). The text itself is accompanied by an apparatus criticus, translation and commentary on its main aspects. A full bibliography can also be added. No field needs to be filled in if the information cannot be obtained (for instance if a stone is lost) or is irrelevant. The general philosophy is to use a key world field, then if necessary to provide a fuller (but concise) comment in a supplementary field. Key words are in French, but texts and comments may be entered in any modern language using a Roman character font.

The text of the inscription, Greek or Latin, must be stored using the fonts Hellas V4 and Latin V4 (provided with the database). The Greek font has the special advantage of supplying all the characters currently used in epigraphy, numeral signs included.

Fig. 1: PETRAE viewed in a Netscape Navigator browser window. This example shows how a fully lemmatised text of an inscription can be generated.

Texts of inscriptions can then be semi-automatically processed to produce indexes: a first step (fully automatic) generates a concordance of all the words in the inscriptions, sorted by document, line number and position in the line (as in fig. 2); a second step (semi-automatic) offers the opportunity to define the dictionary entry of the different words, and also to specify, in accordance with a definition list, the historical definition of all the words beginning with a capital letter; a third step (optional) offers the possibility of supplementary historical encoding (identity of persons mentioned, rank or function, such as magistrates, kings or Roman emperors, etc.). All the information can be corrected, sorted and accessed by specific menus. The information in the different files can be made the object of separate or combined searches.

Fig. 2: PETRAE viewed in a Netscape Navigator browser window. This example illustrates PETRAE's ability to generate a Greek index

All the information in the database - stone, text, indexes, and bibliographical concordance - can be fully automatically edited in two formats: either in RTF format for word processor (an option available to any user), or in HTML format for the Internet (an option restricted to the administrators of the database, who will take charge of the diffusion of the information). The RTF edition programme offers a high quality output format which can generate a document ready to print. In HTML format, text, indexes and bibliographical concordance are linked by hypertext references, so that PETRAE HTML documents effectively become fully dynamic corpora, a fact which in itself constitutes a dramatically new philosophy in epigraphic corpora. The Greek texts may be edited in either Hellas V4 or Kadmos fomat, the Latin texts in LatinV4 or LatinÉpigraphique. Other font translation programmes are under consideration. Also under development is an IBM-PC version of the system for Internet, which we hope to have available in 1998. Fully automatic multilingual editions are also envisaged.

The PETRAE team, which has already produced a series of books using this method, now expects to install its first HTML files for Internet access in summer 1997. Files from different parts of Gaul (Aquitania), Northern Spain, Sardinia, Tunisia, Greece, Cyprus and Turkey will be made available as soon as they are ready. We hope that this example will be followed by new proposals of collaboration and that in the long run the landscape will be ever more densely populated.

Professor Alain Bresson
Université Michel de Montaigne-Bordeaux III
F-33405 Talence Cedex

Return to Table of Contents