Given the interplay of the texts and the computational tools for their study, this section reports on the results of experimenting with new methods of analysis and visualization. Some of these experiments are in progress and attached to existing, upcoming, or planned publications.
Some are so medium-specific that they resist publication. Galileo’s Virtual Library is installed on one computer in Bowdoin’s Virtual Reality Lab. GaLiLeO: Galileo’s Library and Letters Online was demonstrated at Harvard to a group of specialists in 2018, but awaited not just the establishment of the Journal of Digital History (2020/2021) that publishes articles in Jupyter notebook form, but also the ability to publish notebooks in the coding language R (~2023) in order to find an appropriate publication venue.
Regardless of the relationship between traditional publication forms in the humanities like articles or books, making is a form of argumentation in digital humanities. By doing, we are both making choices about data, design, and processes and also drawing conclusions about the digital technology and our object of study. Each experiment described in these pages makes an argument about the existing digital tools and arguments about the texts in Galileo’s library.
- Network metrics tend to express nodes (i.e. a book, a person, a letter) as a unique item. What happens if we treat them as molecules (amalgams of possible books with several potential publishers or places of publication)?
- How do the features of specific books in Galileo’s library overlap or depart from the books for which lack specific information but can gather the amalgamated data about possible editions?
GaLiLeO: Galileo’s Library and Letters Online
- Computational text analysis methods tend to separate documents by form or genre (letters vs. long-form books, poetry vs. prose), yet literary scholars and historians work across these boundaries fluidly. Can a computational method put these texts into conversation despite the low-quality plain text available for primary sources authored by Galileo and his contemporaries?
- What can we learn about the insider-language of Galileo’s circles when the letters are contextualized against a much larger backdrop?
- Are there strengths for using LLMs to study pre-modern, non-Anglophone historical and literary texts despite the limitations of their training data?
- What patterns in Galileo’s texts could be identified (and also worth exploring)?
- What happens when we prioritize the features provided in primary source material that describe Galileo’s books instead of modern library database metadata?
- Why do certain books resist this kind of datafication?
- What happens when uncertainty becomes the design feature for a virtual space that by default requires precision in order to visualize its components and make them interactive?
- Why do certain books resist this kind of datafication?
- How can computational text analysis be adapted for more literary analyses that are not just quantitative?
- What did the sections of Galileo’s text sound like when compared to the entire text or the broader cultural context in which he was writing?