On an Infrastructure for Historical Spatial Analysis

Peter K. Bol, October 2012

References to a "spatial" turn—a linking of geography and history, temporal change and spatial variation—have been increasing across the historical disciplines. A decade ago the geographer Anne Knowles connected a spatial turn in history to historical geographic information systems (GIS), and she has since edited two volumes illustrating the application of GIS to historical subjects. Since then, a number of major historical studies using geospatial analysis have appeared and there is even more interest on the part of the geographers in historical topics. For example, the Association of American Geographers had sixty sessions on "Space-Time Integration in Geography and GIScience" at its 2011 meeting. Geographers are clearly very interested in looking at the past. Why should historians be interested in looking at space through the geographer's lens?

A tool such as GIS helps the historian see what was not apparent before. It enables the use of large quantities of data, the visualization and measurement of the spatial patterns in that data, the tracking of change over time, and the correlation of those patterns with information from different domains. Learning GIS does not involve much more than learning a piece of software and reading a textbook (see the sidebar). The real obstacles to making geospatial analysis useful for historians are infrastructural. In what follows I note some of the major cyberinfrastructural challenges and some early attempts to overcome them.

An Adequate World-historical Gazetteer

GIS is about geographic space, but in the historical record "place" matters more than "space." A person is of a place, a religious site is located at a place, tax is reported by place, a postal station is a place in itself. Places are nodes in networks (for premodern times it is easier to find the nodes than the routes between them and reliable sources for administrative boundaries before 1800 are few). A historical gazetteer, at its simplest a listing of place names with their locations in space and time, is the bridge between history's places and spatial analysis.

Figure 1. Getting Around Ancient Rome. Two routes from Roma to Constantinopolis--shortest and fastest--as provided by ORBIS: The Stanford Geospatial Network Model of the Roman World. Users may query ORBIS for a route between two ancient cities, selecting from a range of conveyances, times of year, and costs of travel. The resulting route shows time spent on roads, rivers, coastal seas, and open seas, and estimates cost of travel depending on whether the traveler is using a donkey, wagon, or carriage.

Naming serves as the interface between humans and the physical world. Naming allows humans to position themselves relative to their geography and make it intelligible, to organize knowledge of it, and preserve the memory of their acts within it. Naming—of a mountain or a river, town or a building—creates an intelligible layer between the geophysical world and human culture; the name makes it a place. If we can capture names in written/drawn sources and locate them in space, then we can locate historical data with spatial attributes—such as tax records, population, religious activities, battles—in space.

Naming pertains to all aspects of human life and, like everything in human culture, names are not stable. A name links the historical and the geographic but its validity is temporally bounded. The historical record is, ultimately, finite, so in principle we can imagine collecting all names in one giant historical gazetteer that tells us when a name is valid, what system of naming it belongs in, and in most cases where it is. The gazetteer is fundamental to the geographic ordering of our human past and to making it accessible. In practice, however, a gazetteer should be able to accommodate place names (even imaginary and fabulous landscapes) that have sources but lack spatial locations.

The major online world gazetteers are invaluable but share a common flaw: they ignore time. This is true of the Geographic Names Information System; the National Geospacial Intelligence Agency's GEOnet Names Server, and GeoNames, the largest non-governmental gazetteer (see sidebar). These systems provide between two million and ten million place names, but, incredibly, do not track name changes over time.

The U.S. federally mandated gazetteers will not include change over time unless that is part of their legal charge. This has consequences. The lack of a record about when a name is changed or a jurisdictional line redrawn eventually will result in the loss of knowledge about when the attributes of that place name (population, area, etc.) are valid. In the past, territorial administrations managed their records through texts, archiving past records and thus creating a written trail. Now, many information management systems simply overwrite earlier data, sacrificing the enormous benefits of a longitudinal record to clerical efficiency. Thus a first-order infrastructural need in linking history and geography, time and space, is a world historical gazetteer. While funding for such a project remains elusive, there are a number of examples that could serve as this project's foundations (see sidebar).

Computationally mining existing databases, dated digitized texts, and digitized georeferenced and dated maps could populate a world-historical gazetteer. Beginning in the 1790s, mathematically accurate cartography has mapped the globe. These maps provide information on routes, boundaries, physical features, and locations that texts cannot provide. For a limited historical period—but one which saw global growth at a pace unparalleled in human history—geo-referenced maps allow us to link place names, locations, and time and thus provide a foundation for georeferencing place names that appear in earlier texts. Manual data extraction will always be limited to specific projects; a systematic approach requires the extension of optical character recognition technology to maps. Until such a project is funded, the best hope may lie with projects now beginning for crowd sourcing the extraction of data from scanned maps.

Access to Georeferenced Maps and Spatial Data

Short of an historical gazetteer, georeferenced maps are the most important source for historical information about space and place. Georeferencing is the process of linking a piece of data to a space. There are several illustrations of projects that do this well. The David Rumsey Map Collection is a leader in the georeferencing of historical maps, with 29,000 scanned maps available online of which some 22,000 have rough georeferencing. A federated system for registering and sharing historical map scans now exists through Old Maps Online, a UK-based project. In effect it creates a union catalog of georeferenced historical maps, including those of the Rumsey collection, the New York Public Library, and soon the Harvard University Library and the Boston Public Library. The user interface allows for searching on a map of the world, and for narrowing the search by time and space.

For spatial datasets, OpenGeoportal.org, led by Tufts, Harvard, and MIT with many partner institutions, is now developing a portal which has the potential to provide a single entry for searching and previewing collections of data.

Figure 2. Mapping Civil Service Degree Holders in China. A map of China produced on ArcGIS, showing the distribution of civil service degree holders. The data is from the China Biographical Database, the physiographic macroregions are by G. William Skinner published by the China Historical GIS, and the underlying digital elevation model is from the China Historical GIS.

A concomitant to a federated map/spatial dataset catalog is a system for archiving and searching historical datasets that could be joined to spatial objects in a GIS. The World-Historical Dataverse of the Center for Historical Information and Analysis at the University of Pittsburg is creating a system for the collection, preservation and analysis of world historical data.

In contrast, there are valuable projects that allow interactive data query visualization online, but are not necessarily geared to making their datasets available for user download. Examples include two Stanford projects: Mapping the Republic of Letters (a visualization of early modern letter exchanges) and ORBIS, which tracks the time and cost of transportation and communication in the Roman Empire. Harvard's Digital Atlas of Roman and Medieval Civilization is another example.

Flexible and user-friendly, the Social Explorer platform provides free and subscription editions for the visualization of spatialized data (such as historical US census data). MapStory, also aimed at the individual researcher, allows users to upload their own GIS data with time attributes, display them dynamically on an online map, and permits others to download the data.

Online Platforms for Accessing and Sharing Spatial Data

The final piece of cyberinfrastructure needed for a wider use of GIS is an online platform for sharing spatialized data (both as online visualizations and downloadable data files). The goal is to make the maps and data layers a researcher creates accessible to others, so that others can take advantage of the accumulation of spatialized data in their own work.

In this area, there has been significant progress. ESRI's ArcGIS Online and ArcGIS Explorer Online allow users to create, store and share maps and datasets, manage geospatial content, and control access to volunteered content. Geocommons, developed by the GeoIQ Companyis similar. In collaboration with the open-source web-mapping developer community, the Center for Geographic Analysis at Harvard is developing the WorldMap platform, allowing users to explore, visualize, edit and publish geographically referenced information. WorldMap has an expanding list of functionalities it wishes to add, but it already allows researchers to upload large datasets and combine them with those shared by others, create and edit maps and link map features to rich media content, grant edit permission to small or large groups, export data to standard formats, georeference paper maps scans online, and share data with just a few collaborators or with the entire world. The promise of WorldMap is that it is cumulative; nearly five thousand registered users are already uploading and sharing their data and maps, and 180,000 others are viewing their work. Working with MapStory, it will soon add animations, allowing vivid displays of change over time.

Historical georeferenced data is part of the world of big data that the digital environment has made part of our lives. As the cyberinfrastructural impediments give way, GIS technologory, federated geodata systems, and online mapping are likely to become part of the toolbox for the next generation of historians.

Peter Bol is the director of the Harvard Center for Geographic Analysis and the Charles H. Carswell Professor of East Asian Languages and Civilizations. He works on China's intellectual history and historical GIS.

Geospatial Information Analysis: Some Sources and Tools

The "Spatial Turn"

Baker, Alan R. H. 2003. Geography and History: Bridging the Divide, Cambridge Studies in Historical Geography 36. Cambridge, UK; New York: Cambridge University Press.

Bol, Peter K. 2007. "Creating a GIS for the History of China" in Placing History: How Maps, Spatial Data, and GIS Are Changing Historical Scholarship, edited by A. K. Knowles and A. Hillier. Redlands, CA: ESRI Press.

Dear, Michael, Jim Ketchum, Sarah Luria, and Doug Richardson, eds. 2011. Geohumanities: Art, History, Text at the Edge of Place. London and New York: Routledge.

Gregory, Ian, and Paul S. Ell. 2007. Historical GIS: Technologies, Methodologies and Scholarship, Cambridge Studies in Historical Geography 39. New York: Cambridge University Press.

Guldi, Jo. The Spatial Turn in History. Institute for Enabling Geospatial Scholarship at the Scholars' Lab at the University of Virginia Library, 1 April 2011 [cited January 16, 2012).

Knowles, Anne Kelly, ed. 2002. Past Time, Past Place: GIS for History. Redlands, Calif.: ESRI Press.

Knowles, Anne Kelly ed. 2000. Special Issue: Historical GIS: The Spatial Turn in Social Science History: Social Science History 24.3.

Knowles, Anne Kelly, and Amy Hillier. 2008. Placing History : How Maps, Spatial Data, and GIS Are Changing Historical Scholarship. Redlands, Calif.: ESRI Press.

Warf, Barney, and Santa Arias. 2009. The Spatial Turn: Interdisciplinary Perspectives, Routledge Studies in Human Geography. London; New York: Routledge.

Some Major Historical Studies Using Geospatial Analysis

The Valley of the Shadow: Two Communities in the American Civil War.

Gordon, Colin. 2008. Mapping Decline: St. Louis and the Fate of the American City, Politics and Culture in Modern America. Philadelphia: University of Pennsylvania Press.

White, Richard. 2011. Railroaded: The Transcontinentals and the Making of Modern America. New York: Norton.

GIS Software and Online Platforms

QuantumGIS

Geocommons

WorldMap platform

The Social Explorer

MapStory

ArcGIS

Online World Gazetteers and GIS with Historical Attributes

American Association of Geographers' Historical GIS Clearinghouse

Geographic Names Information System

GeoNames

Great Britain Historical GIS

National Historical GIS [for the US]

China Historical GIS

German Historical GIS