I’ve always been a bit regretful for not training myself to use Geographical Information Systems (GIS) in my research earlier in my career, when I had more time to master the technology. GIS is not a method that lends itself to half-assed approaches; it doesn’t add much to a study unless it is carefully applied, and requires considerable care in the framing of research questions and in the selection of data to be inputted. It can pay off magnificently, as in Colin Gordon’s study of the economic hollowing out of Saint Louis, and historians like Amy Hillier (with the Mapping DuBois project at Penn) and John Logan (with the Urban Transition Project at Brown) offer great examples of the deep and collaborative effort required to render historical urban maps into usable GIS shapes and to assemble broad and systematic data sets that can be mapped. These efforts hold tremendous promise for considering the spatial dimensions of urbanization processes (which, in my definition, includes both traditional centripetal urbanism and the related pattern of suburban sprawl) and the relationship of urbanization processes to racial, gender, and socioeconomic phenomena.
As David Bodenhamer argues, GIS, despite its usefulness for historians concerned with the spatiality of the past, have had a hard time fully integrating the method into their work. The reasons for this are complex, and, Bodenhamer argues, entail technological, data access, and conceptual barriers to integrating GIS within the disciplinary practices of historians (see Amy Hillier in the Journal of Planning History for more on this.
Aside from the somewhat daunting technological learning curve, users face the problem of access to software. The most common professional GIS platforms are proprietary, and limited by and large to Windows operating systems, a problem for the considerable numbers of academic researchers on Mac OS. There is, however, an open-source and free GIS client that runs on Mac OS (as well as Windows and Linux) called Quantum GIS (QGIS).
One of the benefits of being on fellowship is the opportunity to “steal” a bit of my own time for professional development, and playing with QGIS has been, if not yet professionally rewarding, quite in interesting use of an hour here and there. I’ve begun working through a few online tutorials for QGIS that walk the user through some basic tasks–selecting and modifying a shapefile (the data that creates a visual map) and “joining” tabular data (e.g. demographic data or an inventory of toxic releases) to the shapefile so that the map presents a visual argument about some aspect of social reality. The tutorial I’ve been using (from Len de Groot at the University of California-Berkeley’s Knight School of Journalism) includes a data set for download; the National Historical GIS project at the University of Minnesota provides a free platform for accessing GIS shapefiles and census data tables that include data fields allowing the tabular data to be joined to the shapefile.
Another tutorial series from Baruch College also looks promising, and since QGIS is open-source, using it allows one to connect to the communitarian trend in software use and to a vast array of resources for self-training at little or no cost.
For further reading on GIS in historical research, check out: