Networks of Hugh le Shipward

It can be challenging to view Medieval people as complex individuals. Or rather, one can imagine them having inner lives, whether illustrating manuscripts as scribes or making a tool as a blacksmith. But thinking of them as occupying complex positions within their communities is more difficult. Oftentimes, the general understanding of a Medieval society is based on the Feudal system, which necessarily reduces individuals to their roles in society. In doing so, it also presents a problem for conceptualizing the intricacies of individuals’ roles in Medieval England. This research project, though, could be helpful in uncovering the realities of Medieval life.

Our project begins with a dataset: a group of documents collected in or concerning Herefordshire around the turn of the 14th century. The bulk of this dataset, and the entirety of the records used in this particular investigation, is comprised of the Hereford Cathedral Muniments (HCM). These muniments could document any type of transaction or event, from a land grant to a memorandum to an ordinance. Simply perusing the HCM documents thus reveals the everyday transactions and interactions between members of Hereford society, and sheds light on the complexities of Medieval lives beyond basic places in the social hierarchy. But we can still go beyond the dataset itself. Specifically, this project uses network analysis to visualize the connections between people, and calculate the degree of their connections. To demonstrate the usefulness of network analysis, as well as the HCM documents in general, this post focuses on one middle-class man in
14th century Herefordshire: Hugh le Shipward.


In many ways, Hugh leaves us with more questions than answers, but we might first start by determining what exactly we do know about him. Hugh was an artisan, though it is impossible to determine exactly what kind of artisan he was. His family name, Shipward, indicates that he may come from a family of shipbuilders (as in they were the wards of ships). However, whether or not he carried on that legacy is unclear from our documents. Hugh is often listed as a witness in the HCM documents in which he appears, usually land releases or grants. Interestingly, he is occasionally placed last in the list of witnesses, a position that may be reserved for the scribe of the document. So, it is possible that he may have served as scribe on occasion, though even that is not absolutely certain. We do know that Hugh lived in Breynton, along with his father, John, and his sister, Isabella (HCM 11 refers to Hugh as the son of John le Shipward, and HCM 755 refers to land associated with Isabella, daughter of John le Shipward). Finally, we know that Hugh held land in Breynton. This last piece of information is certain because the mention of his land shows up several times in the HCM. For example, HCM 11: “Release, in consideration of two marks, of two acres of arable land in the field of Chirchebreynton’ between the land of Hugh the son of John le Sypward and the highway and extending as far as the hedge of the said Hugh and containing eight selions” (other mentions of Hugh’s land are in HCM 19, 9, and 23).

So, we do not have much information available about the actual life of Hugh le Shipward. But we do know that Hugh was a fairly well-connected person. He appears in 22 documents in the Hereford Cathedral Muniments, always as a witness and thrice in the position of a potential scribe. His connectedness makes sense given his two dominant roles as an artisan and a landowner. The former would likely make him a known figure in the community, for whatever his craft happened to be. It is unclear whether or not there was major guild activity in his area, but if there was, the connections he made there would have brought him even closer to the center. The documents that record him are also exclusively related to the transfer of land, either through a grant or a release. That he often appears in such documents, and on several occasions
did so with regards to his own land’s barriers, highlights his role as a landowner, and the
centrality of that role to his importance in the community.

Network Analysis

But perhaps it’s better to show his connections rather than discuss them in abstract terms. This is where network analysis has been beneficial. Looking at Hugh as an isolated individual, it’s hard to say whether or not he’s really as well-connected as we might like to think he is based on documentary research alone. A network allows us to visualize the connections between things based on real data, which helps to eliminate biases we might accumulate while reading through the archives. In these networks specifically, a dot represents a person, or sometimes a document, while a line between two dots indicates that two people are identified in the same document. Figure 1 (below) shows a visualization of Hugh’s ego network, or a network specific to himself and the people mentioned in the same documents as him (all network visualizations generated by Gephi, an open source network visualization software).

Figure 1: Hugh le Shipward’s ego network, Hugh circled in red.

Here, the yellow dots represent the different people in Hugh’s network (Hugh’s own dot is circled in red) while the blue lines represent a connection between people. This network is interesting because everyone in it is connected to Hugh, thus allowing us to visualize the full extent of his social connections as preserved in the HCM. Furthermore, it shows us which connections exist within Hugh’s network, which might reveal small communities within the greater Herefordshire one. For example, there seem to be many more lines in the middle, which tells us those names in the middle are more connected and might be involved in the same types of business, or have land nearby each other. Figure 2 (below) shows a zoomed in version of this central part of the ego network, again with Hugh circled in red.

Figure 2: Central area of Hugh le Shipward’s network, Hugh circled in red.

Still, though, we cannot tell how connected Hugh truly was based on this singular ego network. To figure that out, we can use network statistics, generated by Gephi. For the purposes of this project, I will examine three different statistics: weighted degree, betweenness, and eigenvector centrality. Unweighted degree measures the number of connections a node has. So, for Hugh’s ego network, he would have the most, since he is connected to everyone, while not everyone is connected to him. Weighted degree takes into account the weight of an edge, or connection. So, if someone is mentioned in four separate documents with Hugh, that connection would be counted four times instead of once. Betweenness measures how well a node connects other unrelated groups of nodes, and provides a decent measurement for how integral one person is to sustaining a community. And finally, eigenvector centrality measures influence and power in a network.

Table 1 below shows Hugh le Shipward’s statistics compared with the range of statistics present in the dataset. However, average values were not available for every statistic so in each case, it is important to remember that there are far more people with incredibly low (that is, near zero) statistics. Conversely, there are very few people with a statistic at the high end of the range. So, even though some of the ranges seem drastically large, the average is usually much closer to the bottom of the range than the top. Hugh’s weighted degree, for example, is 486. The highest in the dataset is 2450, but very few people have a weighted degree that high. The average weighted degree, in fact, is 56.1, making Hugh’s weighted degree, and therefore his connectivity to the rest of the community, pretty high for the general population of Herefordshire. Hugh’s betweenness is 0.0022, only a factor of 10 away from the highest scorers. That score indicates that, while he is not the key bridge between absolutely everyone in Herefordshire, it may be safe to bet that he is a key connector between people in his social group of middle class landowners, and possibly even a key connector between those people and different social groups. Hugh’s eigenvector centrality is 0.274. This is actually a pretty average score; Figure 3 shows the network in its entirety, and Figure 4 shows the entire network of people in our dataset with a score higher than 0.274. The two are vastly different, with very few people obtaining a score equal to or greater than Hugh’s. So, Hugh was likely both an important member of his social circle, and pretty central to the activities of Herefordshire as a whole.

Weighted DegreeBetweennessEigenvector Centrality
Lowest Value00.00.0
Highest Value24500.0131.0
Hugh le Shipward4860.00220.274
Table 1: Hugh le Shipward network statistics compared to range of the entire dataset of people
Figure 3: Herefordshire network, unedited
Figure 3: Herefordshire network, unedited

Hugh le Shipward and William le Corviser

Now we know that Hugh le Shipward was an incredibly well-connected and integral member of Herefordshire society. However, we may still wonder whether or not his presence was common or someone of his standing. For the sake of comparison, we might turn to another middle class artisan, William le Corviser. Also often called William Seym (as in a seam) (see HCM 398) or even simply William the Cobbler (HCM 19), William was an artisan of the shoemaking variety working around Herefordshire at the same time as Hugh le Shipward. Unfortunately, we know even less about William than Hugh. He has a son, William le Corviser II (HCM 218), and he may have also been a scribe: of the 13 documents he appears in, he is positioned in the place of a scribe for 5. There are no records relating to any land he may have owned, so even though he appears with Hugh in many land-related documents, there is no proof to confirm that he was a landowner as well. Most importantly for the sake of comparison, he appears in many of the same
documents as Hugh le Shipward.

Figure 5 below shows William le Corviser’s ego network with le Corviser circled in light blue. From his own ego network, we can tell that William was also fairly well-connected, and in similar geometries to Hugh. That is, there also exists a more concentrated group of connections in the middle of the network, with less well-connected people skirting the edges of the network, with William tying them together. It could be that the well-connected group at the center of both ego networks is the same. Figure 6 below shows the intersection of their networks, with Hugh circled in red and William in blue. While the intersection between their networks obviously eliminates the bulk of each individual ego network, it also confirms what we already
hypothesized: that there is a common group of connections between them. Indeed, many of the names that came up in the centers of each separate network show up in the intersecting one, including John de Seuenak, John de la Berne, and John de Kemeseye. These overlaps tell us that the two did operate in the same main social circles, and held similar social standings.

Figure 5: William le Corviser’s ego network, William circled in light blue

Figure 6: Intersection of William and Hugh’s ego networks, Hugh circled in red, William circled
in light blue

In addition to comparing network visualizations, we may also compare network statistics between the two (see Table 2 below).

Weighted DegreeBetweennessEigenvector Centrality
William le Corviser3700.00100.279
Hugh le Shipward4860.00220.274
Table 2: Le Corviser and Le Shipward statistics.

Based on these statistics, the two artisans seem to have a similar number of connections. That makes sense, given the shared extent of their documentary appearances. The two men’s betweenness scores are very close, at least to the same degree. So, they probably both acted as bridges between social groups to similar extents. Looking back at their ego networks, that makes sense—both people had fairly centralized networks, but always maintained a large network of more loosely connected people as well. Their eigenvector centralities are also astoundingly close, almost to the thousandth of a decimal. And if we recall the above evaluation of Hugh le Shipward’s eigenvector centrality (Figures 3 and 4), we might note that Hugh’s score indicated an integral nature to Herefordshire society. Due to the closeness of William’s score, we can
assume the same extent of influence and status for him.

Thus it seems that for someone of this middle class status, specifically one involved in land transactions, it is normal to become an integral part of a shire community. It makes sense, too; Hugh le Shipward literally formed a bridge between people by occupying land between a highway and a hedge, near that of one John Thwyti and the dean of Hereford. His very presence on the land gave him the unique opportunity to engage in the dynamic processes of land transfer, connecting people who might otherwise not end up in each other’s spaces and networks, thus creating and connecting microcosmic communities within Herefordshire.

Limitations: What we may never know

While this process of network analysis can shed light on the intricacies of Medieval communities, it is important to note that we may miss important contexts and histories. That is to say, our dataset is limited. The HCM documents we included only cover a specific period of time and scope of documentary practices in the entire history of Herefordshire, which means that certain historical networks, which may be useful in creating full context, might be inaccessible. For example, while I spoke on the Shipwards of Breynton, there were also Shipwards in the nearby Castle Goodrich at the same time. Castle Goodrich is across Herefordshire from Breynton, and is accessible from Breynton through the River Wye. The documents we have access to do not make a connection between the Shipwards, but it is of course possible that they were ancestrally related. Because both Breynton and Castle Goodrich are on the Wye, an earlier migration would have been entirely possible. Furthermore, the family in power at Castle Goodrich around Hugh’s era at the beginning of the 14th century was the Talbots. Two of its major members, Gilbert Talbot and Richard Talbot, were associated with Credenhill, another village near Breynton. When the Talbots took over Castle Goodrich near around the 1300s, then, it is entirely possible that some of the Shipward clan went with them, while some stayed in Breynton. This is, granted, speculative, but still remains demonstrative of the histories we could easily be missing. Network analysis is an incredible tool to further our understanding of historical people and societies, but only with the caveat that we always interrogate our datasets to identify what may still be hidden.