Medieval SNA

Social Network Analysis Researchers of the Middle Ages (SNARMA)

Esther Lewis

My PhD focuses on devotional attitudes and practice in the late medieval port town of Bristol, England. This study has considered the relationships between testators and ecclesiastical institutions in the town, and the relationships between ‘heretics’ and the rest of the town. A large part of the source base for this research has been last wills and testaments. There are 393 of these that survive for fifteenth-century Bristol. I would argue that these are rich documents in terms of what they can tell us about testators’ social networks. Although they do not show us the full network, they give an indication of who the testator’s family were, who were the people they most trusted (to be their executors, for example) and which ecclesiastical institutions they felt they needed to, or wanted to, patronise at their deaths. Throughout my doctoral research, I have used Network Analysis as an analytical tool to give a quantitative overview of the data I have used and to identify and visualise trends in the data set.

Initially, I looked at relationships between testators and parish churches in the town. I found that some testators were leaving legacies to more than one parish church within the town of Bristol. These legacies were things like bequests for unpaid tithes or donations to the upkeep of the church’s fabric: things that one might expect only a parishioner of the church to give.

By inputting the relationships between testators and multiple parish churches, I produced this network graph. This graph was created on Gephi. It has 150 nodes and 259 edges. Edges are directed from the testator to the institution.

Testators giving to multiple parish churches in Bristol, 1400-1500 (Figure 3 in thesis).

This analysis showed a pattern in testamentary giving: that testators appeared to give to an additional parish church on the basis of neighbourhood. This is most apparent in the southern suburb of the town that was made up of three parishes. The cluster on the left of the graph shows a relationships between the neighbouring parishes of St Thomas the Martyr and St Mary Redcliffe, and the other parish church in the suburb, Holy Cross of Temple, can be seen on the edge of the cluster below the main cluster of Redcliffe and St Thomas. Throughout the graph, neighbouring parishes are close to each other. The graph is not intended to replicate geography, but the association between neighbouring parishes suggests that topography and neighbourhood had an impact on testators’ relationships with additional parish churches in the town.

In-degree centralities of giving to additional parish churches in Bristol, 1400-1500 (Figure 8 in thesis).

As the edges were directed, it is possible to measure the relative popularity of the parishes. In the graph above, in-degree centrality has been measured and is represented by the size of the nodes in the graph. What is apparent is that St Mary Redcliffe is by far the most popular church in the town. This may be the case because it was regarded as a merchant church and many of the testators under study here were merchants.

However, parish churches were not the only type of ecclesiastical institution that testators gave to in their wills. Below is a graph that shows the full network of testamentary giving across the fifteenth century. As well as parish churches, it includes friaries, hospitals, fraternities and chapels. It has 609 nodes and 1525 edges, and the edges are directed from the testator to the institution.

Full network of testamentary giving to religious institutions in Bristol, 1400-1500. (Figure 10 in thesis).

One clear issue with this network graph is that its complexity renders it difficult to read compared to the previous graphs. Therefore, I have used the metrics produced by Gephi to look at the in-degree centralities of various pious institutions. What is clear from the graph is that the friaries (represented by the four green circles in the centre of the graph) were popular with testators. This was probably because they were involved with funeral services and other commemorative acts for the testators soul, so testators left generic bequests to each of the four orders of friars in the town to secure their prayers. However, there are some cases where testators appear to have had a more meaningful with the friaries.

Throughout the full network graph, parish churches continue to be important and the graph continues to sort them into neighbourhoods. This shows that neighbourhoods were important for the organisation of lay piety in the late medieval town of Bristol.

Fraternities appear to have taken testators out of their parish for their devotional activities in some cases. In cases where a testator left a legacy to a fraternity, we can assume that they were a member of that fraternity. There are different types of fraternity/guild in late medieval Bristol: there were craft guilds with a vocational focus, such as St Katherine’s and St John the Baptist’s, and parish guilds, such as the Fraternity of St Mary of the Bellhouse or the Fraternity in the Crypt.

Relationship between testators and fraternities in Bristol, 1400-1500. (Figure 11 in thesis).

The above graph shows testators connections to their home parish churches and their connections to fraternities in the town. It shows that the craft guilds drew their membership from all over the town, but it also shows that some parish guilds drew their membership from outside of their parishes. This is especially the case for the Fraternity in the Crypt of the parish church of St Nicholas. While many of its members appear to have come from the parish, evident in its closeness to St Nicholas’s church in the graph, some came from other parishes. However, some parish fraternities, such as St Mary of the Bellhouse in the parish of St Peter, only had patronisation from parishioners in the testamentary records.


The above is a snippet of the work I have been doing with SNA in my thesis. Largely, the methodology has provided a strong basis for my research, a quantitative overview and a basis for comparing different types of networks.

Of course, SNA does not tell me why there was a trend in individual testamentary giving. This emphasises the importance of continued qualitative analysis to go alongside the SNA. This analysis has formed the basis of qualitative enquiry and prompted me to search for the answered to explain the trends seen in the network graphs.

One of the problems that I have come across is that network graphs are difficult for the reader to see when printed out on an A4 page in a PhD thesis. To try and tackle that, the figures in my PhD thesis are linked by URLs to this blog post so that the reader can click through from the PDF version of the thesis to this blog post.

Esther Lewis

March 2020

%d bloggers like this: