Converting a textbook to hypertext

Reviewer: Richard Furuta

Debating the similarities and dissimilarities of documents printed on paper and documents represented electronically in hypertext form is fascinating for philosophers, authors, and software designers, to name but three interested groups. A particularly complex topic is whether and how a document prepared for one medium can be transformed into a document represented in another medium, particularly when the transformation occurs with no or little manual intervention. Earlier, Rada reported on the transformation of a large document from a research hypertext system into printed book form [1]. In this paper, Rada semi-automatically transforms the book's markup back into hypertext form, with the four targets being three generally available systems (Emacs-Info, Guide, and Hyperties) and one sophisticated research system (Bellcore's SuperBook). The paper focuses on the mechanisms by which the transformation is effected. Although the relative success of the transformation effort is not evaluated, Rada has made the resulting hypertexts available free of charge for evaluation by the reader. The paper, while giving welcome detail of a case study in document transformation, does not help much with the larger question: whether document transformation is a generally useful technique or whether its successes are restricted to a small, limited, stylized set of documents. That the document being transformed was itself originally automatically converted from hypertext form suggests that, stylistically, its concepts may already be chunked in a fashion appropriate for hypertext and that its markup may be more regularly structured than one might expect from a manually-created source. The broader question will remain a topic of strong debate in the hypertext community.