Abstract

The MLCD Project seeks to illuminate issues arising in the development of markup languages for complex documents. The examples are from literary texts or philosophical manuscripts, but the problems addressed are generally applicable: any document type may become complex when examined in enough detail. One of the most persistent problems involved in applying SGML and XML to complex materials is overlap. When the material to be represented exhibits several distinct conceptual hierarchies, or frequent occurrences of non-hierarchical structures which overlap each other, the usual rules for applying SGML and XML seem not to apply. The mechanisms proposed for handling overlap in SGML and XML (concurrent markup, milestones, fragmentation of elements and reconstitution of virtual elements from the fragments, and standoff markup) are ingenious and can be effective, but they do not always exhibit the simplicity and clarity which make SGML and XML attractive in simpler cases. The MLCD project has accordingly been experimenting with markup and data structures in which elements may overlap each other.