Willem Borkgren is a Scholars Lab Graduate Research Assistant (GRA) in his second year of the School of Information’s MSIS program. During the 2023-2024 academic year, he undertook a Digital Scholarship Project as part of his GRA role. Using tools in the Scholars Lab, he researched a comparative literary analysis of Jean Rhys’s Wide Sargasso Sea and Charlotte Brontë’s Jane Eyre. The outcomes of this analysis are available in a digital exhibit “Intemperate & Unchaste.”
When I had the opportunity to pursue a Digital Scholarship Project in the Scholars Lab, I was eager to apply the new digital tools available to me on materials I love. Charlotte Brontë’s Jane Eyre has been the subject of literary discourse for centuries, canonized in controversy, with fair points made for its landmark importance in women’s authorship in English literature but also its casual colonial attitudes that reduce marginalized identities to symbols. Jean Rhys’s Wide Sargasso Sea is a 1960s retelling of the events of Eyre but from the perspective of the abused wife of Rochester (Antoinette in WSS, Bertha in JE). Rhys’s own feelings of liminality about her intersectional identity color her characters, and critics have argued she underwrites black characters in her story just as Brontë had.
My goal in this project was to apply modern quantitative literary analysis to these texts with the purpose of evaluating how these texts interact with themes of gender and empire, how they compare in their handling of these topics, and where a modern reader may find them lacking. From these outcomes, I hope to explore their canonicity in spite of their limitations as well as the progression of women’s authorship in English literature.
The overhead scanner in the ScanTech Studio.
I used the tools and resources of the Scholars Lab from start to finish on this project. I began by digitizing a copy of Sargasso Sea I brought from the PCL stacks to the Scholars Lab. Before this I was connected with Colleen Lyon, our Head of Scholarly Communications who also advises researchers on copyright. On her recommendation, I limited my digital copy of Rhys to 3 sections of the novel to remain well within fair-use. With that in mind, I utilized the Scan Tech Studio’s overhead scanner to digitize the selections. With the book imaged, I could send the files directly to Google Docs, where I used the built-in Optical Character Recognition (OCR) software to transform the images into text. In future projects, I would consider a more robust OCR tool as, despite the convenience of the scanner-to-Google-Drive workflow, the OCR’d pages required a good amount of manual cleaning. I downloaded the pages as plain text, then accessed a plain text copy of Jane Eyre from Project Gutenberg. I imported the texts to RStudio, where I cleaned them of stopwords using a custom list containing the Creole patois of Sargasso Sea as well as the Snowball list of standard stopwords. The text was then tokenized for analysis. The tidyverse modules and GGPlot2 were used heavily in my scripting for text mining, the former for organizing data and the latter for visualizing it. With my data visualized, I moved onto Scalar to create my digital exhibit.
Bar graph of words within 5 of ‘dress’ in WSS.
I conducted my analysis in R Studio, scripting to find skipgrams of thematically significant words and assess their sentiment. Ngrams are word phrases where one word is within ‘n’ of a target word, and skipgrams are Ngrams where words between the target word and its pair are skipped over rather than maintained. These are useful in literary analysis as they can indicate how a word or idea is connotated within a text that may not be immediately apparent with a traditional close-read. They also are useful in filtering for words whose meanings are changed by a negating modifier. I represented the outcomes of these skipgrams in word clouds and bar graphs, using size to indicate the frequency of words and thus the relative importance of their connotation. In this way, the juxtaposed contradictions of Eyre and the cynicism of Sargasso Sea can instantly be visually compared.
This word cloud shows ngrams of marriage words in both texts. Larger words are more frequent. Dark blue represents ‘husband’ & ‘groom’ skipgrams in JE, dark red for WSS. Likewise, light blue represents ‘wife’ & ‘bride’ in JE, which are pink for WSS.
From these outcomes and more, I conclude that Rhys is writing back to the assumptions made in Brontë’s work with a more nuanced and modern perspective, but is still limited by her position in time and relative privilege of the white Creole perspective in a Caribbean context. Elements like Antoinette’s status as the daughter of former slave owners and her wealth compared to Tia’s fetishized poverty do indicate some acknowledgment of this privilege.
Skipgrams of 5 for the words “Slave” and “Servant.” Note that ‘slave’ did not occur in the limited selection of JE used in this project.
I hope that this post has piqued your interest and that you will explore my digital exhibit on Scalar. If it has, I invite you to visit the Scholars Lab and see what research you can come up with. From digitization to scripting to publication, all the resources I used in the scope of my Digital Scholarship Project can be found there! If you are curious about any of these tools or spaces, schedule a consultation with the Scholars Lab staff and we’ll help you accomplish your research goals.
