Epigraphic Documents in TEI XML (EpiDoc)

Under the rubric of ‘EpiDoc’ one finds a growing community of scholars intent upon using the TEI for the digital encoding and interchange of epigraphic
documents; i.e., modern digital editions—both print-
derived and ‘born digital’—of inscribed and incised texts
emanating from the areas associated with the Roman
empire at any period before the medieval. This community’s
chief deliverable to date has been a set of guidelines
setting out agreed standards for the application of TEI P4 to such work, together with a customized DTD and various
software tools. The success of the effort is demonstrated, and has been advanced by, three flagship projects: one at King’s College London under the direction of Charlotte Roueché (Aphrodisias in Late Antiquity (2d. ed.) and The Inscriptions of Aphrodisias) another at Brown University under the direction of John Bodel (The U.S. Epigraphy
Project) and a third at Oxford University under the
direction of Alan Bowman, Charles Crowther, and John Pearce (The Vindolanda Tablets Online).
The EpiDoc community is poised to move from P4 to P5. At P4, our extensive and growing guidelines were developed under the auspices of the same customized DTD that we promulgated for epigraphic encoding, with various resulting infelicities and shortcomings. Our strategy for P5 is to separate the schemata for the EpiDoc
Guidelines from that for the encoding recommended by those guidelines. In moving the EpiDoc Guidelines to P5 first, we create a standard, but highly structured, TEI ODD that will, in turn, instantiate a P5 schema to govern EpiDoc-compliant encoding practice.
We shall begin with a brief overview of the EpiDoc
Guidelines, focusing on those features that are dictated either by the requirements of epigraphic markup itself or by the community’s open development practices, in which software tools and the EpiDoc Guidelines themselves are multi-authored by individuals whose TEI expertise
varies considerably one to another. Given this mode of collaboration, the flexible application of constraint possible with the RelaxNG schema technology has significant
advantages over the Backus-Naur content models of P4. We shall also discuss the elements we have added as
extensions or adaptations to the P5 tagset, such as tags for typographic forms and those supporting the integration
of linked typographic examples, encoding examples and regular expressions. These data ‘tuples’ form the basis for automated verification of EpiDoc-specific XSLT
stylesheets and custom text conversion tools, as well
as the internal consistency of the EpiDoc Guidelines themselves