A Halt for Hearsake: Towards a Digital Genetic Edition of Finnegans Wake II.2§6.

This poster provides a short overview of the challenges of building a digital genetic edition of James Joyce’s
Work–in–Progress
. In doing so, it uses the textual genetic analysis of
Finnegans Wake
Chapter II.2 Section 6 (FWII.2§6) to demonstrate how the writer’s composition habits impact editorial decisions in data modelling. This process includes determining how exactly the text grew, what constitutes building blocks of the genetic development, and establishing an optimal level of granularity for draft representation and textual collation.

This research addresses one of the central dissonances in digital editing between generalisation and customisation. The Text Encoding Initiative, for example, as encoding guidelines for digital editors, presents a certain way of dealing with text modelling, but it is by no means all–inclusive. The impetus of the present project is to promote a double movement toward not only process–over–product editions, but also to apply the same approach to the editorial process itself. This will contribute to the idea of a robust digital ecology, where the end–user has more control over the text, its constituent elements and analysis, engendering further research and recycling of the material.

Operating from this standpoint, the poster demonstrates a test–case analysis of FWII.2§6. The inceptive assumption is that the architecture of a digital genetic edition should reflect the writer’s compositional style and the text’s genesis. This requires an inductive analysis of the genetic material of FWII.2§6, which determines prerequisites for its effective digital editorial model.

One of the key prerequisites is the level of granularity of the given data. Considering that the nature of the data is not only textual, i.e. the edition aims to represent not only the text of
Work–in–Progress
but also the manuscripts, their physical attributes and sequentiality, the established smallest unit must accommodate multiple functions from representation to collation. In particular, Joyce’s Buffalo–notebooks material and its “transmigration” from sources to the final text, which contributes to the greater part of his revisions, adds additional complexity to this task.

Another challenge is the “ghost drafts” (textually restorable but physically absent manuscripts) and how to account for their semi–existence in a database where their physical attributes are as important as the text they carry. Some of these attributes allow to reconstruct a better picture of the various aspects of Joyce’s compositional process, such as the time, agent, or even the purpose of textual changes. Notably, it is possible to employ Joyce’s multicoloured crayon deletions as a guide for distinguishing draft levels: how many times a manuscript underwent changes or so–called revision campaigns. These and many other aspects of the rich data set presented by the manuscript sources need to be ascertained before the data model of FWII.2§6 is fully functional.