Methods
What exactly is this project?
Created as a part of the program in digital humanities at Northeastern University, this digital archive traces the diagnosis of shell-shock across popular culture.
A digital collection of primary sources (including twenty-eight textual and ephemeral artifacts) aids in providing a digital exhibit on war poetry, to show how this genre influenced depictions of shell-shocked people after the First World War.
This digital archive is hosted on Omeka, on a domain purchased through Reclaim Hosting, and utilizes collection tools, Dublin Core metadata, a tagging system, and showcases open-access resources for text analysis, such as the Project ArcLight Timeline Search, HathiTrust's Bookworm timeline search, and HathiTrust's Data Analytics center.
What does this archive do?
This digital archive traces how shellshock was represented through medical, literary, and popular culture.
As the public became aware of shell-shock through avenues such as newspapers, newsreels, and literature, their perceptions of shell-shocked and disabled men also changed. This archive is intended as a means of visualizing that paradigm shift in representations across poetry (in the work of war poets), across popular culture (in forms like comics and ragtime scores), and through medical rhetoric.
What are the central questions of this project?
My central questions are:
- How do documents of the First World War reflect shellshock as a disability?
- How are gendered messages, especially messages around femininity and masculinity, included in these documents?
- How do these documents cross over into popular culture?
- How can I offer methods to model texts which are freely available, do not have a huge learning curve for mastery, and create a dialogue with the texts in my Omeka collection to show how latent textual elements reflect messages around disability, gender, and war?
What is the scope of this project?
I limit my scope from 1915, when Charles Myers coined the diagnosis in the pages of The Lancet, and the end date for my analysis is September 1, 1939, when the Second World War officially began. For the scope of this project, the term “shell-shock” refers to a form of war trauma, first named during the first world war, which became a term of contention for the emerging field of psychology.
This project builds upon medical studies of masculinity and shell-shock in the aftermath of World War One to draw together representations of shell-shock. Through shell-shock as a focal point, this project shows how this medical diagnosis of the First World War altered questions around masculinity and gender, disability and ability, and how literature and popular culture communicated these messages.
What are my primary sources?
My primary sources consist of the archival scans of The Hydra, the war hospital magazine of Craiglockhart Hospital, where Wilfred Owen served as an editor. During the First World War, war hospitals produced their own magazines, highlighting work from soldiers who were recovering from their wounds. Alongside these archival scans, I draw on critical editions of “Dead Man’s Dump” and “Break of Day in The Trenches” by Isaac Rosenberg, “Strange Meeting” by Wilfred Owen, “The Rear Guard” by Siegfried Sassoon, and “Song of the Mud” by Mary Borden, which my digital exhibit on shell-shock in popular culture references.
I also draw upon the comic Keep Smiling!: More News By Liarless for German Homes. I also include 15 books for children to show how they demonstrate messages around gender and war. Together, these sources enrich and add further archival context for my exhibit on war poetry and shell-shock.
What are the tools I used?
Drawing on publicly available tools including the Media History Digital Library’s ArcLight search, HathiTrust’s Bookworm tool, and HathiTrust’s Digital Analytics Center, I offer three case studies to show how digital humanities methods can change the frameworks through which we view texts, including distance reading across a keyword timeline in the Media History Digital Library, a keyword timeline of eighteen million volumes from HathiTrust, and topic modeling and sentiment analysis of historical newspapers on HathiTrust, which I place alongside a smaller corpus of poetry, comics, and music.
Through these case studies, I create a dialogue with the texts in this Omeka collection to show how latent textual elements display messages around disability, gender, and war.
What is a "collection"?
For my four collections, "Shell-shock," "Popular Culture,” “Children’s Books and the First World War” and "War Poetry," each item description has been fleshed out to include relevant historical context for all entries, effectively creating short-form curation statements, so that if a user only clicks on the collection, they can still learn about the items included. Think of the collection descriptions as museum labels, offering more context for a reader to discover.
Within Omeka, collections exist to organize items by a chosen imposed theme or pattern. You can go to my Data Dictionary for more detail, or check out the collections themselves. For the purpose of this project, my collections exist to showcase the items themselves.
For the most productive view of this archive, a user should go to each collection and navigate through the items, to explore what can be found.
What is an "exhibit"?
If we think of the item descriptions in the collection area as museum labels, these exhibits are closer to historical tours. Essentially, these exhibits are my way of taking you on a tour through the archive.
Within Omeka, exhibits exist as a way to incorporate items from collections into longer-form written content.
For this project, "exhibits" are curated, long-form mini-essays which provide more context for the historical context of how these items tell the story of shell-shock and its medical status through poetry and popular culture, both during and after the First World War.
That's great! Why do you have case studies within your exhibits? And what are they intended to show?
Because Omeka features longer-form writing within the "exhibit" section of the website, I found these case studies were best represented as exhibits throughout.
Across three case studies, I explored two varying textual searches, which index collections through differing rates of access, created datasets within a research capsule, removed stopwords, and validated my worksets. My first case study performs textual analysis of shellshock in popular culture via Project Arclight, which searches the nearly 2 million page collection of the Media History Digital Library and graphs the results, a case study using HathiTrust’s Bookworm tool to search shellshock across its collections, and a case study tracing how I applied HathiTrust’s Analytics tools to a corpus of war poetry, novels by Mary Borden, medical journals, and periodicals, to show greater trends across all four corpora.
What's the value of these case studies for an archival project like this one?
These case studies exist to model tools which fulfill my criteria: to 1) offer methods to model texts which are freely available 2) to showcase tools which do not have a huge learning curve for mastery, and 3) create a dialogue with the texts in my Omeka collection to show how latent textual elements reflect messages around disability, gender, and war.
- https://search.projectarclight.org
In my first case study, I tested the Arclight search available through the Media History Digital Library through a variety of terms, including "war" and "Hollywood" alongside "shellshock." I chose to focus on the Media History Digital Library, a “non-profit initiative dedicated to digitizing collections of classic media periodicals that belong in the public domain for full public access." - https://bookworm.htrc.illinois.edu/develop/
In my second case study, I drew on the HathiTrust Bookworm tool as an entry point to deepen my engagement with textual analysis through a contiguous system, so that I was querying within the same collection where I was planning to do further textual analysis. - https://analytics.hathitrust.org/statisticalalgorithms
For my third case study, I used HathiTrust's search engine to build collections, which served as my corpora-building tool. Through HathiTrust’s analytics research center, I downloaded every public collection as a workset through the HathiTrust Analytics research center website, validated each workset, and created topic models through their InPho Topic Explorer, which trains multiple Latent Dirichlet Allocations, or LDAs, to extract topics which occur across a given corpus.
Taken together, these projects represent a number of applications across methods, applying close-reading alongside distance reading, in order to show a range of different ways that literary scholars can address textual trends throughout historical texts.
What are some limitations of the project?
There are a number of potential limitations in relying so completely on HathiTrust for a digital humanities project. One limitation lies in the reliability of servers (there is a limit to storing your files). At one point, while I was still learning how to navigate the research center, the server went down for maintenance, leaving me unable to continue my research for the day. I am hoping to figure out a more reliable form of exporting the .json files for interactive visualizations. Another limitation lies in HathiTrust’s search system and its tendency to include duplications of texts. To avoid duplications, I manually sorted out doubles of texts in my worksets.
For my collections, I was also limited to English language sources from the United States and Britain, as materials in French and German are still being translated and thus the translations are not in the public domain. In Further Sources, visitors to this site can find readings from Senegalese writers Bakary Diallo and Lamin Senghor, whose work was only just translated from French and made available for public access in 2021.