Social Network Analysis

The Leverhulme Trust generously funded a three-year project (2013-2016) called ‘The Transformation of Gaelic Scotland in the Twelfth and Thirteenth Centuries’. The project team included Prof Dauvit Broun and Dr Matthew Hammond of the University of Glasgow School of Humanities and Mr John Bradley and Dr Cornell Jackson of Kings College London Department of Digital Humanities. This project provided for the mapping facilities now available on the website (thanks to Neil Jakeman) as well as exploratory research on using Social Network Analysis on the PoMS database.

Social Network Analysis is an interdisciplinary bundle of methods, techniques and concepts taken from matrix algebra, graph theory, and decades of sociological and anthropological research. The work conducted by this project marks the first time SNA has been applied to medieval society on such a large canvas. We hope that our findings will provide the disciplines of medieval history and digital humanities with new models, methods and ideas. In particular, our results offer insight into both exciting possibilities and perplexing challenges stemming from the application of SNA methods to a large-scale digital prosopographical database which has been carefully designed to represent the social context of documentary production.

This project is also unprecedented in the extent to which our workings are open and accessible to the public. First, the database from which the datasets were drawn is fully available on this website. Second, we have created a number of interactive SNA visualizations using the program Gephi, so that users can explore the some of the graphs (known as ‘sociograms’) in greater detail. Third, we are publishing most of our results in a nearly-500-page e-book, Social Network Analysis and the People of Medieval Scotland 1093-1286 (PoMS) Database, available for free on this website. This explains in detail which SNA methods were most fruitful, describes the datasets and techniques fully, with dozens of tables and illustrations giving the results. We hope this will give medieval historians and historical social network researchers both a lot to engage with. Finally, we have provided over 60 Excel spreadsheets, laying out the makeup of datasets and the  full results for centrality and other calculations. We hope that making our work so transparent and accessible will help encourage the greatest possible level of scholarly engagement.