The community of the realm in Scotland, 1249–1424
History, law and charters in a recreated kingdom

CITE THIS PAGE: Matthew Hammond, 'Robert I's charter witnesses', The community of the realm in Scotland, 1249 -1424: History, law and charters in a recreated kingdom, 21 Nov. 2019, https://cotr.ac.uk/social-network-analysis-political-communities-and-social-networks/networks-robert-i-king-scots-1306-29/robert-is-charter-witnesses/

The People of Medieval Scotland database contains document pages for 542 documents of Robert I, king of Scots (1306-29). Of these, 228 are charters with witness lists. We can analyse these 228 witness lists to see who the most prominent witnesses of Robert's charters were, who witnessed alongside whom, and to calculate who were the most central actors within the context of the royal court and household as expressed in the act of charter witnessing, according to social network analysis calculations.

I. Robert's network of charter witnesses

As the following graph showing the number of royal documents of King Robert I by year of his reign illustrates, the witnessed documents used in the social network analysis (orange line) are a subset of the overall document production (blue line), but they mirror the shifting nature of document production/ survival that we see year-on-year. Very few were produced in the early years when the fledgling king was struggling to gain a military foothold, but the numbers grow significantly in regnal year 9 with the political stability created by the victory at the Battle of Bannockburn (June 1314).

The network comprises 99 individual witnesses to Robert's 228 surviving charter texts with witness clauses. We can view this network as a Gephi sociogram:

This sociogram shows the status of each of the witnesses, the majority of whom were laymen (70%). The colour code breaks these down into specific categories of laymen and clerics.

A zoom view:

There is an interactive version of this sociogram here. Hovering over each person highlights their own connections.

The following table lists the 99 witnesses of Robert I's charters, in order of degree centrality. Degree is the measure of how many other actors to which each witness is directly connected. In the context of a network of charter witnesses, this means how many of the 99 known witnesses each actor witnessed alongside; we may think of them as 'co-witnesses'. Robert Keith, the king's marischal, witnessed alongside 85 out of 99 people. In that context, he had 'contact' with a remarkable (nearly) 85% of witnesses. This centrality suggests a hitherto unrecognised significance within Robert's household.

II. Raising the threshold

A. Method One: Co-witnessing instances

Raising the co-witnessing threshold to more than 20 co-witnessing instances. (Netdraw sociogram)

more than 40 co-witnessing instances...

and more than 75 co-witnessing instances. The number of times a pair witnessed alongside each other is labelled on the edge.

B. Method Two: Degree Centrality

We can raise the threshold for inclusion in the network by limiting to actors with a degree of 20 or above, i.e., those who have witnessed with 20 or more others in the network.

The network with actors of 30 or more degrees...

Explore the interactive visualization of Robert's witness network, filtered to degree of 30, click here.

And the network constrained to those of only 50 or higher degree.

A. Eigenvector centrality

Eigenvector centrality is a measurement which adds a layer of nuance to degree, in terms of thinking of power elites. Eigenvector takes into account the degree centrality of those actors who are immediately adjacent to ego. This allows us to see not simply how many people with whom each individual co-witnessed, but also whether those connections were themselves central players or more peripheral figures within that network. Eigenvector centrality as calculated in Gephi sets the person with the highest degree, in this case Robert Keith, at a value of one (or we can think of this in terms of percentage, with Robert at 100%), and reckons the value of all the other actors in relation to him. The eigenvector values of the most central figures in Robert's reign do not differ too greatly from the picture given by degree centrality. The fact the top four actors, all laymen, have eigenvector values over 90% is reflective of the importance of the close-knit group of Keith, Randolph, Hay, and Douglas around King Robert.

B. Betweenness centrality

Betweenness centrality is a measurement intended to reflect influence in a network. The number is calculated by adding the shortest paths between two nodes in the network that pass through ego. In theory, if individuals have to go through ego to get to other nodes, this places ego in a position of power, although the importance of this in terms of witness networks is less obvious than other kinds of networks. The betweenness network of Robert I's charter witnesses reveals the same familiar top actors we have already encountered. One small difference is that the importance of Bernard, abbot of Arbroath, the king's chancellor, increases at the expense of James Douglas, lord of Douglas: more individuals must pass through Abbot Bernard than James Douglas, despite James co-witnessing alongside more individuals (70 compared to 66).

C. Weighted degree

Weighted degree is calculated by adding up all the individual links between nodes, as opposed to the number of contacts. In that sense, it is more concerned with edges (the lines between nodes) than the nodes themselves. It is calculated by adding up all the instances of co-witnessing, so the number of times one has co-witnessed matters more than just how many people one has appeared alongside. In this list, Abbot Bernard has jumped to a strong lead position, with 914 instances as opposed to 824 for James Douglas and only 770 for the otherwise most central person,. Robert Keith. This means that Abbot Bernard, and to a lesser extent James, do not witness alongside as many other people as Robert Keith or Gilbert Hay, but they witness more often with that smaller group. Bernard's position as Robert I's chancellor explains why he appears so many times alongside a smaller group of co-witnesses.