Seed Grant Summer 2025 Report

Gillian MacDonald and Morgan Fox

Background

The launch of the beta version of Networking the Letters of Revolution project on Github pages has so far been a success. The project itself, still nascent, is building upon the idea that communication and relationships during conflict are incredibly important in terms of political capital during chaos. The project’s success is reflected in the fact that the authors were invited to unveil at the Omohundro Institute’s Digital Project Coffee Hour in April 2025. Using one main corpus of letters, Leven and Melville Papers, the project visualized this world’s letter communications, places of importance, and most connected people. We were able to get all the data from the 599 letters contained in the digitized copy of the Leven and Melville papers into a usable dataset. This approach allows us to focus our attention on the interactions and relationships. The letters, memorandums, warrants, and petitions exchanged in this period are more than just networks of exchange. In the late seventeenth century, people were bound together through community, print, and dialogue. From the success of this first endeavor we were inspired to expand the project to include at least one more corpus of letters and any extant manuscripts from the Leven and Melville Papers that existed archivally.

Project Description

While the first phase of the project has been a success, we sought to expand and improve the project in a number of ways. Not only would this include adding more material but also building out the accessibility of the website and creating some materials for future users. For summer 2025 we aimed to commit another 650 letters from the manuscripts of the Duke of Hamilton (another important seventeenth century statesman) into the project as well as expand our analysis, explore more tools (primarily graphcommons), migrate the website to its own domain, and build an API.

Our goals for the project remain consistent:

• Visualize the world of communication from a corpora of letters.
• Understand communication patterns during conflict and a regime change.
• Parse group dynamics and information distribution from a chaotic period of transition.
• Democratize the pursuit of knowledge and expand the potential for research in this area of research.

This project brings together British transnational history and computational data science. The visualizations continue to show a story of connectivity over time because all those involved in this conflict have manuscript and print sources that do a very good job at capturing the difficulties in the reconstruction and administration of Scottish governance during such a chaotic period.

We followed a similar process in compiling manuscript data into csv data for use in the tool. This time our data followed a different format to accommodate the new tool: graphcommons. The data was collected as follows: nodes (ID, name, description) and edges (from type, from name, edge type, to type, to name, weight). After creating a master spreadsheet information file, we then set about creating different sheets for different visualizations including people, places, keywords, nodes, and then edges. We also converted the Leven and Melville CSV data into this new format as well for importation into the tool. After we imported the data into graphcommons, the result was a very complicated web of connections.

Figure 1. Combined Melville and Hamilton Graph

Outcomes

Given our previous experience with networking from the previous iteration of the project, we elected to do some calculations using python and others in graphcommons. These tools were the foundation of building our API and integrating the Leven-Melville and Hamilton correspondence datasets. The groundwork established through Python libraries such as Networkx and Matplotlib, along with prior documentation and workflow design, provided a smooth transition into expanding this project.

API

Figure 2. API Letter Network Viewer

The API was developed to provide access to two collections of letters, the Hamilton correspondence and the Leven-Melville correspondence. Each dataset was converted into a JSON file with consistent fields for sender, recipient, date, and other relevant metadata. In addition, the API includes filtering and download features, allowing users to retrieve subsets of the data as needed.

Hamilton Letters

Figure 3. Exploratory Hamilton Correspondence Heatmap

An exploratory analysis of the Hamilton dataset consisted of using python tools and libraries such as Pandas, Matplotlib, and Networkx. This included examining patterns, identifying the most frequent correspondents, and visualizing the structure of Hamilton’s network. These analyses provided insight into the dynamics of his correspondence and served as a model for examining additional datasets as well as a comparison to the Melville network.

Combining Hamilton and Leven Melville Data

Figure 4. Combined Networking Spreadsheet

In order to analyze both networks, the next step was to combine the Hamilton and Leven and Melville letters into a combined dataset. This required some data processing and filtering of names and formats so that the two sets could be analyzed together. Once integrated, the data allowed cross-network comparisons and offered a broader view of correspondence across different figures and contexts.

Leaflet Map

Figure 5. Combined Leaflet Maps

Developing a Leaflet map of the combined networks provided a more visual and interpretable view of the correspondence. The map incorporates both the Hamilton and Melville networks, with each node representing a correspondent. Users can also toggle between datasets to view them individually. This visualization reveals geographic and relational patterns that are not immediately apparent in the raw CSV data.

Future Directions

Networking Letters of the Revolution has had the opportunity to expand with the addition of the Hamilton correspondence. With the initial network consisting of approximately 600 pages of material from the Leven-Melville papers; it has now grown considerably with about 650 pages of new content. This expansion and the integration of the two corpora have enabled new analyses of late seventeenth century correspondence dynamics. Looking ahead, future directions in this project include the continuation of expanding the network with the addition of the privy council and committee for plantations. This would allow a more comprehensive view of the data as well as expanding the scope of the network. We also plan to pursue additional funding to support labor costs, enhance digital infrastructure and tools, and cover domain expenses.