Requirements and tools for a modern genetic edition of Goethe's Faust

The Goethe–Schiller archive in Weimar, the Goethe-Faust edition and the problems of finding and adapting
tools and making them interoperable; as such it offers
an interesting parable on the beauty of theory and the
pitfalls of reality.
References
TEI Special Interest Group on Manuscripts. http://
www.tei-c.org/wiki/index.php/SIG:MSS (accessed 14
November 2008).
TextGrid. http://www.textgrid.de/ (accessed 14
November 2008).
Vanhoutte, Edward (2006). Prose Fiction and Modern
Manuscripts: Limitations and Possibilities of Text
Encoding for Electronic Editions. In Burnard, Lou,
O’Keefe, Katherine O’Brien, and Unsworth, John (eds),
Electronic Textual Editing. New York: Modern Language
Association, pp. 161–180.
Van Hulle, Dirk (2006). Authorial Translation: Samuel
Beckett’s Stirrings Still / Soubresauts. In Burnard, Lou,
O’Keefe, Katherine O’Brien, and Unsworth, John (eds),
Electronic Textual Editing. New York: Modern Language
Association, pp. 150–160.
Haus in Frankfurt and my group in Darmstadt are
preparing a digital critical edition of Goethe’s drama
Faust; work started at the beginning of 2009. Astonishingly
enough, there is no modern critical edition. Thus
scholars still have to use the edition created 100 years
ago as part of the Weimarer Ausgabe of Goethe’s work,
which makes an understanding of the existing material
and the genetic processes rather difficult. Goethe worked
for fifty years on the drama, and about 1000 manuscript
pages exist, documenting this long genesis; most of these
pages are part of Faust II, which Goethe finished a short
time before his death.
The new edition of Faust will allow access to all manuscripts
as facsimile and as diplomatic transcription. In
this kind of edition the facsimile and transcription should
be presented side by side; moreover, in contrast to the
manner of a number of such editions currently on the
web, the transcription should be closely linked to the
facsimile, first by preserving many aspects of the look
of the facsimile and by being linked bidirectionally to
it on a micro level. It should also be possible to view
the facsimile alone and render transcriptions of verses
as mouse-over activated tooltip; or, vice versa, to view
solely the transcription and additionally view parts of the
facsimile.
A detailed genetic analysis of each document and of the
text will allow viewing the genetic process of a document
and of structural units like verse or scene. The
archival (or document-centric) view of the material is
thus complemented by a conceptual view of the text as a
drama and its units.
These basic requirements determine the architecture
of the markup, which will be mostly layered stand-off
markup to avoid problems with overlapping hierarchies
and to allow easier processing. (On the problems with
the markup of genetic editions, see: Vanhoutte, 2006;
Van Hulle, 2006.) To name some of these layers:
• Linking facsimile and text
This can be done using the facsimile markup in the
TEI P5 guidelines.
• Diplomatic transcription
Features of the manuscripts that must be recorded
include the direction of the text, shift of hand, textual
alterations, gaps, non-textual marks on the page
etc. Some of these features are covered by the TEI
guidelines; some of them, which describe imagelike
qualities of a manuscript and of the writing,
have to be expressed by other means such as SVG.
• Genetic markup
This is an interpretation of the document which
is described by the facsimile and the transcription
under a genetic perspective. At the moment no
markup schema for this type of edition exists, but
there is one under development by a working group
which is part of the TEI Special Interest Group on
Manuscripts. The genetic markup will be split into
different levels: a functional analysis of textual
alterations, sequences and time lines, grouping
changes—like revisions on a page or changes of
conceptual units of the text, a coordinate system
which links together chunks of text which belong
together, a way to add the interpretation which is
the basis for a markup decision, and uncertainty.
One of the problems for an edition like this is the lack of
editors and other tools which are easy for scholars to use
and allow them to concentrate on the editorial work; and
which are robust and integrate easily with other tools.
The project TextGrid, which has reached early beta status
after three years of development, has as its aim the
creation of such an environment for text-centered studies.
If its second work phase is funded, it will offer some
basic tools, but all tools have to be adapted for the Faust
edition to allow editors to enter the different layers of
markup, and new tools will have to be integrated into
the workflow. The special requirements of each digital
edition seem to make it impossible (or at least unaffordable)
to offer a complete solution which allows their creation,
change and enhancement, and publication out of
the box. Therefore some time of an editorial project has
to be spent on setting up a workflow and overcoming the
problems of data conversion between different tools or
merging the outcome of different tools. As the ease of
use of tools has a direct impact on the speed with which
a digital edition is created, a solid knowledge of existing
tools and of the difficulties to adapt them is crucial for
a new project, especially since up to now there have not
been enough reviews of specialized tools in the digital
humanities. (And there is almost no culture of publication
for the concepts and algorithms of specialized tools
for the digital humanities, either.)
The talk will concentrate on the requirements for the