This paper describes the novel ways in which the Orlando Project, based at the Universities of Alberta and Guelph, is using SGML to create an integrated electronic history of British women's writing in English. Unlike most other SGML-based humanities computing projects which are tagging existing texts, we are researching and writing new material, including biographies, items of historical significance, and many kinds of literary and historical interpretation, all of which incorporates sophisticated SGML encoding for content as well as structure. We have created three DTDs, for biographies, for writing-related activities and publications, and for social, political and other events. A major factor influencing the design of the DTDs was the requirement to be able to merge and restructure the entire text base in many ways in order to retrieve and index it and to reflect multiple views and interpretations. In addition a stable and well-documented system for tagging was deemed essential for a team which involves almost twenty people, including eight graduate students, in two locations.