The Lokalbericht project (2013-2016) 1 aims to deliver a scholarly digital edition of Swiss author Hermann Burger’s first novel Lokalbericht. Accompanied by a reading edition in print (fall 2016), the digital edition will include all extant precursory stages of the novel as well as Burger’s final revision of the text, written between 1970 and 1972 (Wieland et al., 2012). The project relies on document oriented encoding, as supported by the TEI guidelines since P5 v. 2.0 (cf. Burnard et al., 2012), in order to aptly capture the materiality and the genetic dimension of the work.
Whereas the recommendations of other textgenetic edition projects as well as the tradition of genetic editions in print were very helpful for the encoding of local, microgenetic phenomenons (i.e. intradocumental) and their presentation to the reader, much less information was available on how to document interdocumental textgenetic relations in a structured manner. In order to better understand the challenges and needs of what we dubbed macrogenetic editing, a workshop was convened in Bern in 2014. 2 As Hans Walter Gabler concluded, ‘a digital genetic edition will only tap its full potential by capturing and visualizing “macrogenetic” aspects’ (Gabler, 2014).
Incited by the outcomes of the workshop we investigated ways that offer immediate insight into Burger’s way of writing – the novel was crafted as a mosaic consisting of numerous tesserae that were worked over repeatedly. This writing technique results in an f avant texte consisting of many segmented and related texts, of which at times more than a dozen copies are extant. Taking a bird’s eye view of the corpus, the reader should get a clear understanding of the writing process and be able to follow the development of specific segments or chapters and their integration into the final typescript, before swooping down to the text level to discover the textgenetic minutiae.
A Visual Approach to Macrogenetic Relationships
In printed and early digital genetic editions, macrogenetic relations are often visualized as a stemma. Such stemmata leave much to be desired – they often abstract on a high level (e.g. witnesses or versions of a text) as opposed to smaller textual entities such as paragraphs or sentences, require considerable intellectual and manual effort, and remain comparatively static, perhaps linking to representations of related documents, but lacking in precision. Yet their genealogical tree structure serves still as a powerful basis to represent editorial changes.
Taking the underlying idea to represent temporal development on the vertical axis and introducing finer levels of granularity in a stemma tree on one hand threatens to lead to over-complexity and problems of positioning. Resorting to force directed plotting as it is often used to display and investigate graph data on the other hand does not satisfy the sequentiality inherent in written text. For the manageable amount of typescript pages and drafts of the novel at hand, we found that a prearranged layout that entails a temporal dimension (y-axis) and material evidence of the texts (x-axis) works best (cf. figure 1). 3 Genetic relations between the typescript pages, the smallest units of display, 4 are highlighted on mouse actions. Interacting with this visualization, the user of the digital edition can trace the development of the early and final drafts of the novel.
Figure 1: Display of two genetically related sheets: genetic relations (r.) and navigational component (top l.)
Accessing Data through Visual Navigation
Good textgenetic visualizations should not be mere by-products of digital editions, but closely tie in with the structure of the digital edition, i.e. serve as a subsidiary navigation to its contents. By way of two panels that yield lists of related typescript pages the user can access detail views in the main interface of the edition (single or synoptical views) directly from the visualization. Vice versa, the visualization initially highlights the current selection of the main interface in order to provide specific views of the dossier génétique.
While the analytical approach based on genetic paths defined by the editors (on sheet level) should provide the most valuable knowledge about the production of the novel, it would be desirable to complement the visualization by an explorative mode based on relations encoded on more granular levels. 5 A visualization of this kind would be less useful to pursue particular authorial changes, but it would help to identify patterns that deserve closer attention and facilitate the isolation and display of the least interpretation-ridden relations and consequently better suit a reader-centered concept of a digital edition.
The wish for visualizations that abstract from the detailed genesis of a text to a more approachable overview is shared by a number of genetic edition projects. 6 After all, imparting his or her knowledge and insight to the readership lies at the core of a scholarly editor’s role. We hope to contribute to the debate by sharing our approach to this problem at DH 2016 and finally by launching the digital Lokalbericht edition in October 2016.
Cf. the preliminary project website at http://lokalbericht.unibe.ch.
More complex genetic corpora and different choices of granularity would however require more sophisticated zooming and panning mechanisms or ways of three-dimensional stacking and it remains to be seen whether the chosen approach would serve them well.
The development of an explorative visualization of the entire Lokalbericht corpus as initially planned will presumably only be implemented for selected texts due to limited resources.
Examples are the digital historical critical edition of Arthur Schnitzler’s works (1905–1931), the digital edition of Goethe's Faust, or the Samuel Beckett Digital Manuscript Project.