Data Model
Data Model
Model of Connections Between Characters
To the right is the model of connections we designed to record and analyse how and when the characters we capture in our sources relate to each other, as described by each text’s author. At the start of the project, we sampled a variety of our sources to establish a robust model of possible connections occurring within our texts and germane to the historical context.
Our model has 21 types of relationship, which allows us to analyse our texts in sufficient depth without excessive diffusion. The different categories mean that we can create networks of all connections or focus on specific relationship types. We investigate our connection data employing qualitative and quantitative methods, including digital network analysis using a software code developed by our team. For some results of our quantitative analysis see the Graphs & Charts page.
We started with 9 categories some of which were subsequently sub-divided leading us to 21. Our 9 main categories are: Family, Friendship, Religious, Patronage, Politics, Hostility, Transmission of Information, Post Mortem, and Supernatural.
The full 21 categories are found in the table to the right.
Example of Connections between Characters
Below is one of our spreadsheets in which we connect the characters that appear in each of our texts based on the data model above. This spreadsheet shows the connections as described in the Life of Radegund by Venantius Fortunatus (d. c. 609).
We list characters as they appear in the course of the narrative in Column A. We treat every chapter as a sub-unit of the story so that we can trace narrative progression and weight characters' centrality to the story. To do so, we record all characters in each chapter (Column B), which means some characters (e.g. Radegund) appear multiple times in Column A. In Column C we record a character's gender: M (male), F (female), V (varied = groups with, or potentially with, mixed genders), and X (unspecified - e.g. angels and spirits). Columns from D onwards show our different types of relationships based on the Data Model explained above: not all 21 categories are included in this spreadsheet as not all relationship categories were represented in this text, e.g. there were no examples of religious hostility. The connections of each character in Column A can be traced by reading horizontally along the rows of the spreadsheet.
The graph visualises all characters and connections in the spreadsheet and was generated by our algorithm. The graph is ordered by degree centrality. Female characters are magenta nodes and male nodes are turquoise: there are no varied groups or characters of unspecified gender in this text. Radegund is the largest node as she has the most connections in the text and therefore the highest degree centrality. The graph does not distinguish between the different types of connections in the spreadsheet. For more on this see here.
Venantius' Life of Radegund is the only source we examined that has a majority of female nodes: 23 women (51%) to 22 men (49%). The most important connections of women in this text are within the monastic community, including with servants. There are fewer connections of kinship and marriage for women than in many of our other texts.
Venantius Fortunatus, Radegund sharing file.xlsxVenantius Fortunatus, Life of Radegund, spreadsheet
