Bringing the Dunkirk urban intelligence network to life

The digital humanities landscape has changed substantially in the past year. Google Earth, the most user-friendly and widely available GIS-style platform, will cease operation at the end of 2015. Some of its services will be rolled into Google Maps. But the future for Google Earth’s technical drawing and polygon functions is unknown. Support for Gephi, the most widely used network analysis software in the humanities, has unofficially ceased; no new updates have been available since January 2013. Gephi does not operate well with Apple’s latest OSX El Capitain update, and no word has yet been forthcoming as to when such operability issues will be addressed. In the interim humanities and social science experts are increasingly shifting to Cytoscape. Originally designed with National Institute of General Medical Sciences funding to map genome, molecular, and biological pathways, Cytoscape now enjoys growing popularity in the digital humanities community. 

In a 2012 article, for instance, Sandra van Ginhoven recounted using Cytoscape to recover, visualize, and analyze the world of famed Flemish artist, furniture maker, and art dealer Guilliam Forchondt (1608-1678). Forchondt’s reach and legacy is difficult to calculate using traditional textual methodology. An average artist, Forchondt’s true talents lay in networking and marketing. His pieces – usually made by talented apprentices – found their way to Spain and New Spain. Competing artists filed litigation against him in civil court, contending that his works should not have the reach that they commanded, and that he was “unduly interfering in the art market.”[1] In Cytoscape Ginhoven modeled three distinct spreadsheets: people-network; documents; and transactions. Each spreadsheet included such characteristics as “seller, business partner, employee, factor, sales agent, and buyer.”[2] As part of her visualizations, Ginhoven overlaid networks onto topographical base layers, providing a true spatial framework to her analysis. 

With this in mind, Designing Empire is currently constructing its first network visualization: the Dunkirk Urban Intelligence Network, 1713-1731. In the 1680s Sebastièn le Prestre, Marquis de Vauban, France’s pioneering urban engineer, reimagined the sleepy coastal community of Dunkerque (Dunkirk) as an important fortress city, capable of supporting a sustainable local economy, defending itself from naval and land attack, and threatening English and Dutch shipping in the English Channel. The English, in particular, were threatened by Dunkirk, so much so that in the 1713 Treaty of Utrecht they demanded its deconstruction.[3] Within months newly-enthroned King George I commissioned John Armstrong, an upstart surveyor officer in the British Army, to oversee Dunkirk’s demolition and collect any information on the city’s fortifications, organization, or infrastructure.

This first spreadsheet (figure 1) incorporates three groups: individual-network, city, and person. Within the individual-network database, columns have been allotted for sender, recipient, location from, location to, date sent, date received, and notes. Figure 2 documents cities with geographical coordinates according to WGS84 codification. Figure 3 identifies each individual involved in some capacity in the network. This means that they could be a sender, recipient, or someone referenced or discussed in the correspondence or maps.

Figure 1: The Dunkirk database.

Figure 2: Locations (WGS84 coordinate system). The second iteration of this database (currently under development) will require greater geographical specificity. E.g., “Kensington” and “Windsor” instead of simply “London.” This will permit accurate distinction between individuals and between formal and informal networks.

Figure 3: Individuals (sender, recipient, mentioned in correspondence). The second database iteration will need greater profession specificity.

When building this database, a first round of data cleaning must occur. Writing a date in the eighteenth century for instance proved to be an extraordinarily personal choice, appearing in varying formats: month, day, year; day, month, year; year, month, day; year, day, month; and sometimes simply season, year. Titles and postnominals require systematization. Comte and Count translated as the same title, albeit in French and English. To further complicate matters, Dunkirk correspondents, too, struggled to gradually switch from the Julian (“Old Style” or “O.S.”) to Gregorian (“New Style” or “N.S.”) calendars. 

Historians soon learn that not all that is present is easily visible. This initial database hides two issues. First, both formal and informal networks were active. The formal, or official network was readily apparent: letters moved two and from Armstrong to British and Dutch plenipotentiaries at Versailles, before continuing on the Secretary of State’s quarters in London. As needed, correspondence skipped Paris and was sent directly to London. However, other letters were not sent to Paris or London. Instead Armstrong sent them directly to Charles Townshend, 3rd Viscount Townshend, who was at the King’s Court in Windsor. Townshend, a progressive with a keen interest in colonial urbanization schemes and engineering reforms, had the personal ear of the King. Second, this database is currently limited by focusing only on those individuals who corresponded with each other by letter. It does not highlight the equally important communications between persons who were married, part of a team, or within walking distance (and who would therefore not write letters to one another). For instance, few if any letters existed between Armstrong and his deputies, Thomas Lascelles and Jacob Ackworth. These connections will have to be manually added.

[1] Bert de Munck, “Construction and Reproduction: The Training and Skills of Antwerp Cabinetmakers in the Sixteenth and Seventeenth Centuries,” in Bert de Munck, Steven L. Kaplan, and Hugo Soly, eds., Learning on the Shop Floor: Historical Perspectives on Apprenticeship (New York: Berghahn Books, 2007), 100.
[2] Victoria Szabo, “Transforming Art History Research with Database Analytics: Visualizing Art Markets,” Art Documentation: Journal of the Art Libraries Society of North America 31.2 (Fall, 2012): 158—75, at 161—63. Also see Sandra van Ginhoven, “The Role of the Antwerp Painter-Dealer Guillam Forchondt in the Large-Scale Distribution of New Imagery in Europe and the Americas during the Seventeenth Century” (Duke University: Ph.D. diss., 2015).
[3] See Richard Steele, The importance of Dunkirk consider’d: in defence of the Guardian of August the 7th (London: A. Baldwin, 1713); John R. Moore, “Defoe, Steele, and the Demolition of Dunkirk,” Huntington Library Quarterly 13.3 (May, 1950): 279—302.

Subscribe: RSS | ATOM