As I’ve settled more into the idea of an online format for my project, I’ve been thinking about the specifics of how readers would interact with it. I have a personal website, so I could easily add another page for the project. I have some experience with HTML and CSS for basic web design, but would need to learn Javascript and either jQuery or another framework like Svelte, Nuxt, or Vue in order to enable the interactive and dynamic elements. 

Another option would be to use a pre-built tool for creating interactive narratives, like Twine (which can also be customized with CSS and JavaScript as desired). Their reference guide seems clear and readable, but I have some concerns about potential limitations. In particular, it is designed for branching narratives, but I’ve been considering other dynamic and interactive possibilities for my project, such as methods of layering text, allowing reader revisions (or reader-prompted revisions through embedded AI or algorithmic tools), and/or representing successive changes to the text (the idea of a digital palimpsest). I also found this list from the British Library of other similar tools, which could be helpful to explore.

Delving more into the computational elements, given that I’m still considering different options for what interactive mechanics I’d like to offer for readers and whether/how I’d like to use computational techniques in my own writing process, I haven’t yet been able to do much research into specific tools or techniques. In general, however, the possibilities would fall into a few different categories:

  • Creative and generative processes while writing the text
    • ChatGPT, Google Bard, and similar tools are the first thing that comes to mind. They are easily accessible and usable, but I’m still unsure whether they align with my thematic interests within this project (namely, considering computational processes and how they can contribute new dynamics to writing and reading, rather than simply echoing and reflecting human processes).
    • Other less “smart” tools could be potentially more interesting from the perspective of introducing elements that I as a human alone couldn’t create. In Text Remix, I’ve been enjoying some of my experiments with randomized poetry, particularly those where I’ve been able to begin with a chosen corpus (e.g., a certain poem, novel, etc.) and assemble a poem from that — so that the particular combinations of words/phrases are often unexpected, but the overall theme and focus can be predetermined. 
    • Combining the two approaches, I’ve also enjoyed some of the poetry that melds randomization with interaction, so some of the back-and-forth between ChatGPT and me as the human writer could be used as the basis for new combinations.
  • Audience interaction with the text
    1. As with the generative processes, this could go the direction of AI or of rules-based processes. The pros and cons would be similar, but integrating either of these within an accessible UI would be more of an undertaking. According to some preliminary searching, I would need to learn some server-side programming and encryption to embed ChatGPT on a webpage. I could also use PyScript to embed Python programs and interactive elements. This would allow me to use and build upon any of my work from Text Remix. 
  • Dynamically updating content within the text
    1. In addition to the same possibilities as for audience interaction, I could also introduce an analytical layer that audiences could explore, perhaps simulating analysis conducted by the scholar character. This might take the form of evolving layers of different approaches to computational text analysis, changing depending on societal and scholarly priorities and concerns. Some possibilities here include stylometry (study of style, often used to study authorship and/or genre), principal component analysis (used to deduce a relatively small number of features that capture most information about a text), network studies (used to analyze the structure of a text and connections within it), and topic modeling (used to identify groups of similar words within a text). I want to consider more deeply which of these would fit technically and thematically before digging deeper into the specifics of how to implement them. 

In terms of disciplinary connections, in addition to drawing on my creative writing background, I think it will be useful to do some reading in the digital humanities, interactive fiction (including hypertext fiction), perhaps game design, digital poetry, and other forms of electronic writing.