Method

This paper is part of a larger, ongoing, geospatial study of the Russian Empire which I have dubbed "the Imperiia Project." This project is an historical GIS (geographic information system) that seeks to compile, analyze, and visualize the infrastructure of mobility in the Russian Empire.

There are many challenges and frustrations inherent in the application of GIS methods to historical subjects, but one of the goals of this project is to demonstrate the payoffs of working on a scale and in a manner somewhat unfamiliar to historians - particularly to historians like me, who came to Russian history through the poems of Pushkin, the stories of Turgenev, and the novels of Tolstoy.

Developing an historical GIS of any scale requires a multi-phase approach. First, the data must be compiled and arranged in geodatabases (think of them as spreadsheets organized according to spatial units). In some cases this process is relatively straightforward, as a great deal of the information generated by imperial ministries was indeed organized by administrative and other spatial units. The statistical data shaping this paper, for example, describes regions, provinces, rivers, and wharves.

The trick then is to "join" the historical data to vector data (points, lines, and polygons that describe locations on the surface of the Earth). This requires knowing where the places mentioned in your sources 1) are located, and/or 2) where they were when the primary sources were generated. This stage can get complicated astonishingly quickly. Why? To take but one example: provincial boundaries were subject to (in some cases repeated) revision and to make matters worse, most available GIS boundary data describes the contemporary political world. GoogleMaps will happily show you where Moscow is located, but it will not - and can not - show you the boundaries of Ufa province in the 1870s. For my purposes here I have generated maps using polygons derived from the Geograficheskij atlas produced by the General Staff between 1821 and 1827. I have adapted the shapefile to reflect major revisions in the southeast, but otherwise left the boundaries alone. (Producing a comprehensive geodatabase of boundary revisions in on the 5-year plan.)

The next stages - symbolization, layering, and spatial analysis - are deeply intertwined. To make a map is to make an argument and thus developing the visual language of a map (or a series of maps) is as involved a process as writing an academic paper. I am not (to my increasing chagrin) a trained cartographer. Nor am I a trained geographer. It is inescapable then that the maps I have shared here are problematic in any number of ways. However, I have found the experience of conducting even a rudimentary kind of spatial analysis of large amounts of quantitative data - quantitative data that otherwise I would have been at a loss to engage - highly productive. Spatial analysis excels at identifying patterns, idiosyncrasies, and correlations. (And it can do much more than that too.) It is an excellent tool, therefore, for testing both new hypotheses and long-entrenched arguments about place and space. This paper then is an attempt to reconsider, recontextualize, and possibly reconceptualize the structures and meanings of economic exchange.