Visual Analytics and the Roman Empire

For the past four centuries, scholars have analysed the Peutinger Chart by talking and writing about it. This late antique chart of the Roman Empire and the Orient has been described at great length and with great erudition.

But can that ever be adequate when the subject of analysis is an information visualization? Are we not likelier to discern its method of composition, its purpose and perhaps even its late antique origins if we analyse it visually? That is why I am developing some new visual analytical tools, which you can see in my two new regional charts (click on each small version to go to my website):

These focus on the Italia and Africa sections of the Tabula Peutingeriana. The first of the two was introduced in a previous post, and the latter got its first public exposure in a draft article at Academia.edu.

These graphic tools provide you with useful new ways to look at the Tabula and associated texts. You can summon up or dismiss a range of views by tapping or clicking on the radio buttons provided with the two graphics.

The view offered at first sight is of the Tabula converted to a subway-style diagram. This involves straightening out the horizontal lines and reducing to them to 1/5 or 1/20 in overall length so that everything can be seen at a glance. This transformation is entirely consistent with what the Tabula is: a chart that selects parallel, east-west lines for especial prominence and manipulates their length. I have also categorized the connections into long-distance and local ones and differentiated them by color.

My conversion of the Tabula to a spider or spider-web diagram omits the distances and contains only the place names. That is not to say the route-section lengths are irrelevant. Recently I have been reading the work of Emil Schweder, who proposed that the chart is not a diagram of routes, but a diagram of distances between places. His revisionism has a surprisingly modern feel, but even so, the graphic substrate of the network is its nodes and links, and that is what we must study first of all.

The second view compares these two spider diagrams to the outlines of the landmasses they represent. These will assure the user that my reformulations can be accommodated with modern geographical knowledge and that the Tabula itself possesses a certain integrity, only departing from a scaled representation of these landmasses in a rule-bound, not a chaotic way.

The third view is a visualization of the Ravenna Cosmography in terms of the Tabula. I have already introduced this textual work, which is in effect a second recension of the Tabula. There is a link on my SVG files to the Archive.org image of the Pinder & Parthey edition of the Cosmography. The greatest problem for the scholar comparing these two recensions is that there has been no published, complete tabulation of them side by side. I am told there is plan to create such a resource, but I think a visual comparison is an even more urgent desideratum. Here it is.

The Antonine Itinerary overlay offered for Africa is a similar comparison, though less productive. There is, I think, no longer any general doubt that the Itinerary (ItAnt) and the Tabula have quite different origins. The comparison shows you quickly why: The ItAnt not only peregrinates around the African provinces in a fashion that suggests its compilers wanted to visit a great many places on the way, not reach an ultimate destination by the most direct means. It actually omits the main highway from Carthage to Setif which passes through the wastelands of the arid high plateau (marked blue on the plot). There could not be a clearer indication that long-distance routes were of little interest to the ItAnt author, whereas long-journey itineraries were a resource that the Tabula author exploited wherever possible.

The fourth view probes for north-south alignments with two hypotheses in mind. The first is my own that some kind of graphic mechanism for easy reproduction is built into the Tabula. It would be surprising if the designer had paid no mind to methods that would minimize the risk of copying errors. In the case of the Great Stemma, a late antique chart with  many similarities to the Tabula, the mechanism is a grid of 10 by 70 spaces into which the individual entries or roundels had to be inserted. The second hypothesis I wanted to explore was one by Kurt Guckelsberger which proposes the forerunner of the Tabula was a high, narrow chart with the Orient at top and Atlantic below.

To compose this fourth view, I combined two sets of numbers from my database. The vertical positions (in the y axis) conform with the spider diagram (view one). But the horizontal positions (in the x axis) are determined by the distribution of the place-names in the 12/13th-century manuscript in Vienna. These numbers are taken from my exact facsimile, the Tabula Peutingeriana Digital Plot. This data-mix, as well as rotating the names by 90 degrees, allows you to focus on the scribally transmitted left-right positions, not the swerves or the writing, and judge if there is any kind of regularity between the rows

So far I can only say the fourth view neither confirms nor refutes the two hypotheses I mention. Keep looking and exploring and you may discover something I have missed.

No comments :

Post a Comment