Links, Letters and Lunaticks: a case study exploring the potential of Digital Historical Social Network Analysis

Abstract

The aim of this research was to explore the potential of Digital Historical Social Network Analysis (DHSNA) as a historical method. Five conditions were established on which DHSNA should be judged in order to determine if it qualifies as a promising method: added value, reliability, practicability, scope and prerequisites. In order to investigate these categories and to map the difficulties and the opportunities of applying social network analysis to bodies of historical data, a case study has been executed. This case study centered around the Lunar Society, an eighteenth-century informal scientific society in England’s Midlands. The question the case study focused on was to what extent DHSNA would be able to teach us more about the social and the organizational structure of the Lunar Society.

In order to investigate the question four types of network visualizations have been constructed based on the meta-data of letters exchanged between Lunar Society members. A number of exchanged letters is not one-to-one related to the strength of a social relation. Therefore the networks mapping the collected meta-data needed to be translated to observations about the social and organizational structure. The networks mapping the collected meta-data were converted to correspondence activity networks by two techniques of reconstruction, social and organizational structure was derived from the correspondence activity networks by interpretation. Different qualitative and quantitative methods - including the calculation of metrics - have been applied in order to analyze the network visualizations.

The network visualizations and their associated metrics did bring some added value to the Lunar Society's historiography, for instance by shedding light on historical discussions by unraveling them and by uncovering underlying principles. However, many shortcomings were encountered as well, it was for instance not possible to derive from the visualizations some basic social situations described in the historical literature. The shortcomings were mainly due to the fact that the reliability of the networks fell short. Because the aim was to investigate a (relatively new) method, the discussions of the research design and the evaluations of the research process are understood as part of research results, and they served as input to judge DHSNA on practicability. Combining the observations about added value, reliability and practicability, it was concluded that DHSNA is not a suitable historical method to study tight historical groups (scope), that is, not until more advanced methods have been developed, the historical research infrastructure has been improved, and more easily accessible tools for historians have been created (prerequisites). At this moment, the insights the network analyses could bring are not significant enough to outweigh the amount of effort the DHSNA research process asks for.