The Greek philosopher and polymath Theophrastus (ca. 371 - 287 B.C.E.) was a prolific scholar whose works had considerable influence in his own time and the centuries succeeding. A student of Plato and associate of Aristotle, Theophrastus became the second head of the Peripatetic School in Athens, and a central figure in Hellenistic intellectual culture: his authority spanned logic and metaphysics, physical and natural science, ethics, psychology, politics and rhetoric. With the exception, however, of a series of psychological profiles known as the Characters, two treatises on botany, and a number of smaller scientific writings, Theophrastus's work is attested only in fragmentary form, in quotes, paraphrases and references by a wide range of other authors.
This makes the study of Theophrastus's work and influence a challenging but instructive exercise: his works, words and ideas cannot simply be edited and presented, but have to be reconstructed by the careful relating of various texts, from different sources, with one another. As a result, a study of Theophrastus inevitably entails a broad study of ancient texts and intertextuality.
Since the nineteenth century a number of print editions have appeared in which the editor has sought to present a single area of Theophrastus's work; but largely because of the breadth of the material, there has been no attempt to present Theophrastus's fragments comprehensively until Project Theophrastus was conceived at a conference on the School of Aristotle held at Rutgers University in 1979.
Its ambitious aim was "collecting, editing, translating and writing commentaries on all the ancient and medieval sources which quote, report or otherwise refer to Theophrastus by name" (I, p. 3). Twelve years later the two-volume work Theophrastus of Eresus: Sources for his Life, Writings, Thought and Influence was published.
A strong example of print scholarship at its most complex, the edition of Sources brings together in two volumes fragments of text from all available sources which cite Theophrastus. By compiling the chosen texts in one location and providing a normative reference scheme (each fragment gets its own number), the print edition facilitates the work of relating them to one another.
As with similar scholarly print sources, this reference structure serves to frame and organize the supporting scholarly material. Translations are supplied on pages facing the original texts. Two forms of apparatus, showing textual variants and pointing the reader to parallel texts, appear as footnotes to the fragments; and of course commentaries in additional chapters or volumes can refer to the fragments by number. In addition to the topical arrangement of the texts, supplied indexes and cross-referencing mechanisms work to chart the complex body of represented work.
The technology of print has been highly developed over centuries to provide many of the qualities scholars demand from texts, notably portability, stability and longevity. Furthermore, by collecting materials together and providing them with a transparent, documented arrangement (through tables of contents, indexes and other structures), a book provides access to materials; but this access comes at the price of foreclosing other possibilities of arrangement or use.
In stabilizing a particular arrangement, establishing and finalizing it, print has the effect of masking from students and non-specialists the considerable labor, expertise and judgment that has gone into collecting and assessing source materials: it tends to create an impression of certainty and finality even where scholars' judgments are tentative and provisional.
Likewise, there are limits introduced by the once-for-all design of the printed page, which can become saturated by references, and so less useful to a wide range of audiences with different needs. Such issues of knowledge, representation and utility are fundamental to scholarly work, and editions of fragments will always be experiments in scholarship because the nature of the material is so resistant to presentation: no original source is extant, and these are "bodies" of work which have no intrinsic form other than what editors design for them. This makes an electronic version of Project Theophrastus an especially interesting prospect, as these problems will have to be addressed within the very different conditions of the electronic medium.
Given the formidable scholarship already provided by the scholars and editors of Project Theophrastus and "encoded" in the Fortenbaugh edition of the fragments, an electronic edition does not have to start from scratch. But it is not quite a straightforward conversion either, inasmuch as the electronic form offers a different context and new possibilities for reading and manipulating texts.
The first phase of an electronic Theophrastus is, in fact, to perform a straightforward conversion. It involves taking the text of the print edition and providing it with TEI-compliant encoding of the various complex structures in it, including textual and critical apparatus, facing translations and tables of cross-references.
Provided with such encoding, the fragment of Theophrastus can be delivered electronically with hypertext links to connect the various elements of text, translation, apparatus and commentary.
There are several objectives to this phase of work. First, we are developing and implementing a consistent tag structure for the fragments, conformant with the TEI and supporting a range of functionalities in searching and display.
Secondly, Phase 1 will result in a stable SGML rendition of the text, in which not only the texts of the fragments themselves (the information) but also the considerable knowledge about these texts (the "meta-information") provided by its editors are available.
Although the TEI encoding of these structures is fairly straightforward (the text is complex but its complexities can be described very well by TEI tags), this phase does provide significant related challenges, primarily in handling character sets. The source texts of the fragments are in Greek, Latin and Arabic, and all these texts require stable representations in SGML -- ideally independent of platform or application, and themselves capable of conversion as further standards emerge.
A further objective of Phase 1, therefore, is the creation of a generic filtering program which we can use to perform conversions between various non-standard character set mappings presently used on different platforms, and which will aid us in creating a stable reference text.
The second phase of the Electronic Theophrastus will begin as we are concluding with Phase 1. This involves shifting our attention from the encoding of the text itself, to developing a means for its delivery and use. For this purpose we are reviewing the possibility of using Electronic Book Technologies' DynaText browser, an industrial strength SGML browser designed to handle "libraries" of electronic texts.
In configuring DynaText to work with an SGML Theophrastus, we will not only be developing the use of the TEI structures in practice; we will also be testing the limits of the "book" model of access to information.
One reason why it is critically important to provide high-quality and stable electronic texts is that such work also underlies more experimental approaches to representing and handling information: an authoritative SGML version of a text can be put to any number of uses in different applications.
In further phases of the project, an electronic Theophrastus will be a suitable candidate for a range of experimental approaches: for example, supported by SGML encoding and a consistent reference scheme, it will be possible to "remove the binding from the book," allowing students and scholars to track their work, document newly-perceived relations between texts and produce comments of their own, experiment with other possible arrangements for the fragments, or integrate other on-line resources such as lexicons or the Thesaurus Linguae Graecae. Technologies to come will support SGML textbases with advanced analytical functions and new kinds of representation and access.