True interoperability for digital scholarly editions

Abstract

If the digital scholarly edition (DSE) is ever to replace the print scholarly edition it must be made truly interoperable so it can be easily secured, moved, published, aggregated, distributed and sold. Current DSEs are customised for particular projects, and must be maintained by their creators. Their contents are also not easily reusable by others. However, digital editions can be made truly interoperable by writing them directly for the Web rather than first in XML, then converting them to HTML. For this to work, some changes to the organisation of the software and data of a DSE are needed. Instead of dividing the software into two parts that run on the server and the client, all software can be moved to the client. In this way, a DSE can become portable, durable and directly usable in any web-browser. Instead of XML, the textual data can use a simplified form of HTML consisting of only two elements: <P> and <SPAN>, controlled and customised by a standard CSS stylesheet. The current practice of encoding alternatives such as variants can be replaced by versions and layers: versions are complete texts written by the author and layers are notional transcripts of local changes ordered chronologically. In this way textual data can express most of the information formerly specified by the complex TEI-XML Guidelines, and the rest via other technologies, and reorganise it in a form that allows easy comparison, editing, searching and textual analysis using standard software tools.