Abstract
This thesis delves into the intricate dynamics of digital writing processes to explore how eleven Dutch and Flemish fiction writers create their texts. Renée van Marissing, Jens Meijen, Arnoud Rigter, Ellen Van Pelt, Niels ’t Hooft, Jente Posthuma, David Troch, Roos van Rijswijk, Vincent Merckx, Aafke Romeijn, and Dirk Speelman wrote short stories, and Gie Bogaert made the composition of his novel Roosevelt available for research. Using a theoretical foundation which stems from genetic criticism and writing studies, the following core questions were addressed: How do cognitive and literary theories inform textual changes? How do writers structure their activities over time? What are the shared practices and unique methods of different writers?
After an introduction of keystroke logging as a tool for recording digital writing, four archetypical writing strategies were used to characterise these writers' approaches. The traditional division into phases (pre-compositional, compositional, and editorial) is still used by many writers. Additionally, revision is crucial during the composition process. Planning and revision are not extremes of the same spectrum but two dimensions by which processes can be characterised. Secondly, a newly developed measure of non-linearity was applied to the keystroke data. Several non-linearity profiles were found, from a more traditional phased approach to a variation on a smaller scale. The stages of working on a first draft (drafting) and editing that first draft (post-draft) were clearly identifiable through the non-linearity measure, providing a new link between textual development and process registration dynamics. This datadriven approach was followed by an in-depth look at the role that revision plays in writing. Revisions were categorised using a narratological taxonomy. The writers prioritised revision right from the start of their work processes. Although a shift was observed towards more stylistic revisions over time, all types of revisions still occured in the post-draft phase. Secondly, correlations were found between writing process dynamics and the semantic categorisation of the revisions. The great attention to style that many of the writers showed in their revisions is tied to the special role of language, and more elusively ‘voice’ or ‘tone’ in fiction compared to non-narrative non-fiction. This dissertation has established new links between the fine-grained information from keystroke logging and the cognitive and narratological aspects of writing strategies in experienced fiction writers who worked on their stories over multiple writing sessions.
After an introduction of keystroke logging as a tool for recording digital writing, four archetypical writing strategies were used to characterise these writers' approaches. The traditional division into phases (pre-compositional, compositional, and editorial) is still used by many writers. Additionally, revision is crucial during the composition process. Planning and revision are not extremes of the same spectrum but two dimensions by which processes can be characterised. Secondly, a newly developed measure of non-linearity was applied to the keystroke data. Several non-linearity profiles were found, from a more traditional phased approach to a variation on a smaller scale. The stages of working on a first draft (drafting) and editing that first draft (post-draft) were clearly identifiable through the non-linearity measure, providing a new link between textual development and process registration dynamics. This datadriven approach was followed by an in-depth look at the role that revision plays in writing. Revisions were categorised using a narratological taxonomy. The writers prioritised revision right from the start of their work processes. Although a shift was observed towards more stylistic revisions over time, all types of revisions still occured in the post-draft phase. Secondly, correlations were found between writing process dynamics and the semantic categorisation of the revisions. The great attention to style that many of the writers showed in their revisions is tied to the special role of language, and more elusively ‘voice’ or ‘tone’ in fiction compared to non-narrative non-fiction. This dissertation has established new links between the fine-grained information from keystroke logging and the cognitive and narratological aspects of writing strategies in experienced fiction writers who worked on their stories over multiple writing sessions.
| Original language | English |
|---|---|
| Place of Publication | Amsterdam |
| Publisher | University of Amsterdam |
| Number of pages | 349 |
| Publication status | Published - 17 Jan 2025 |
Keywords
- creative writing
- keystroke logging
- textual scholarship
- dutch literature