Visualizing Social Networks: Palladio and the Encyclopédistes, Pt II

This post is a part of a series entitled “(In)forming Revolution: Information Networks in the Age of Revolutions.”

By Melanie Conroy

In my first post on Palladio, we explored points-based and point-to-point based mapping. In this post, we will focus on how we can use Palladio to visualize networks. A network can link people, places, books, or any other entities that are connected to one another. People or groups are commonly nodes in graphs of social networks. The Age of Revolutions provides many examples of networks that can be studied from a theoretical or a technical perspective: courts, Masonic temples, literary circles, Jacobin clubs, militias, scholarly and correspondence networks—to name a few. Many of these networks have been studied as networks insofar as social historians have studied their memberships prosopographically, or demographically; other scholars have taken a profound interest in the quality and the character of relations within these groups.[1]

Network Graphs

Network graphs are useful for seeing connections between people or groups. They are also helpful for understanding how groups are structured. A network graph is a set of points (called nodes), connected by links (called edges).[2] There are other software packages—notably Gephi, a network analysis computer program designed for the quantitative study of networks. Palladio is particularly good for humanities research but lacks the ability to do complex mathematical analysis.

For the sake of this demo, I recommend opening the JSON document—that is, the preformatted version—in Palladio by following these steps:

1) Download the JSON document: https://drive.google.com/open?id=0B4hYJNXGw_WbeU5XQ3BhRW1Mamc

2) Open Palladio in your browser: http://hdlab.stanford.edu/palladio/.

3) Click “Start” and choose “Load an existing project” and then choose file “Encyclopedistes-Kafker-Conroy-2017.palladio.1.2.9.json”

4) Select the tab that you would like to see. For this demo, please choose “Graph.”

The first network graph shows how the major social networks of the French Enlightenment are linked to knowledge networks (Fig. 7). The nodes in this graph represent groups that are sized according to the number of encyclopédistes in that group (not the total number of members of the group). The largest node is, thus, “Encyclopédistes,” which represents all 146 contributors. This node is connected to all of the other nodes since all of the other networks contain encyclopédistes. The next largest network is the “Letters_Literary” network which contains 56 encyclopédistes. Other large knowledge networks include “Letters” for people whose publications are unknown or difficult to classify; “Letters_Antiquarian” for people whose publications address ancient or learned subjects like Classics or numismatics; and “Letters_Philosophical” for authors whose works, including articles in the Encyclopédie, are philosophical. Smaller categories include “Letters_Religious” for the authors of religious, pious, or theological writings; “Sciences” for the authors of general or hard-to-classify scientific works; as well as “Sciences_Natural,” “Sciences_Mathematical,” and “Sciences_Medicine.”

These knowledge networks appear as nodes sized according to the number of members of those networks who were also encyclopédistes. The other networks in this diagram are social networks of the French Enlightenment represented among the encyclopédistes. Some of these networks are correspondence networks: the correspondents of Voltaire, D’Alembert, and Rousseau are relatively large and appear close to the encyclopédistes node because they have many other connections (high degree). Unsurprisingly, the correspondents of Adam Smith, Hume, and Bentham (bottom right corner) are far less connected to the encyclopédistes and the other networks of the French Enlightenment. Salons and social circles like the circle of Holbach and madame Geoffrin’s salon are more linked to the Encyclopédie than the salons of madame de Lambert, madame de Deffand, or madame Graffigny.

Screen Shot 2017-09-20 at 1.06.28 PM.pngFigure 1: Network Graph of Enlightenment and Knowledge Networks of the Encyclopédistes

In this diagram, we see that the most central groups to the encyclopédistes network are the correspondents of D’Alembert, Voltaire, and Rousseau. Groups like the encyclopédistes in the “Sciences_Medicine,” “Sciences_Natural” and “Letters_Literary” are more connected to other groups than the “Letters_Antiquarian” and “Letters_Religious” networks. Despite the large number of people in the “Sciences_Natural” and “Letters_Literary” networks, they have fewer connections, or lower degree. This quick comparison between correspondence networks, social networks, and knowledge networks suggests that the “celebrities” of the French Enlightenment were more fundamental to connecting people together than were knowledge networks.

The second network graph shows the same Enlightenment networks (correspondence networks and salon networks) among the 146 encyclopédistes. Again, the size of the circles represents the number of encyclopédistes who were members of each group. This time, I have also included the major social networks of the French Enlightenment (Government, Aristocracy, Elite, Military, Court). Unsurprisingly, there are more encyclopédistes who were members of these larger networks than of the much smaller individual salons. Once again, the correspondents of the major figures of the French Enlightenment occupy the center of the graph because they have high degree (both are connected to 12 other groups). The larger social networks, “Military” and “Aristocracy,” are connected to fewer groups (5 and 7 other groups, respectively); salons and foreign correspondence networks are still more eccentric.

Screen Shot 2017-09-20 at 1.06.43 PM.pngFigure 2: Network Graph of Enlightenment and Social Networks

These network diagrams show some expected results—for example, the association of Deffand’s salon with courtly high society; the association of Bentham and Adam Smith’s correspondents with government; Geoffrin’s greater association with the elite than the titled nobility. Yet there are surprises: the association of Deffand’s salon with government; the association of Voltaire’s network and the military; Rousseau’s correspondence connections with government officials. Network diagrams allow us to see such connections quickly and to investigate peculiarities.

Conclusions

Mapping smaller social networks, such as the network of the encyclopédistes, is one way to uncover unexpected or anomalous connections. In the case of the encyclopédistes, these preliminary results show that correspondence networks around celebrities like Voltaire, Rousseau, and d’Alembert were more significant to the network than were knowledge networks such as the “Letters_Literary” network or the “Letters_Philosophical” network. The toolkit of network analysis provides humanists with one more valuable means of uncovering unexpected new avenues of research, especially in larger datasets where connections are difficult to track.

Melanie Conroy is assistant professor of French at the University of Memphis. She received her doctorate from Stanford University in 2012. Her research explores the intersection of literature, visual studies, and social networks in modern French culture. She is the co-director of the Salons Project, a part of Mapping the Republic of Letters.

Title image: Un dîner de philosophes painted by Jean Huber. Denis Diderot is the second from the right (seated). (Palladio Network Graph Overlay)

Further reading

Maria Teodora Comsa, Melanie Conroy, Chloe Edmondson, Dan Edelstein, and Claude Willan, “The French Enlightenment Network,” The Journal of Modern History, 88:3 (2016), 495-534.

Miriam Posner, “Getting started with Palladio.”

Christina Prell, Social Network Analysis: History, Theory, and Methodology, Sage Publications.

Endnotes:

[1] Gephi is open-source visualization software for networks, maintained by a team of developers including Mathieu Bastian and Sébastien Heymann. For more information, see https://gephi.org/about/.

[2] For an introduction to social network analysis, see Christina Prell, Social Network Analysis: History, Theory, and Methodology, Sage Publications, 2011.

 

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s