ArcUser Online

Reconstructing Aztec Political Geographies
Continued...

The output table from the name-matching algorithm was joined to the NGA place-name layer, creating a table that contained data keys for both the modern place-name layer and historical location records. With these links established, attribute information derived from historic research could be rendered. Depending on the type of analysis, historical records that could not be located were dropped from the join.

Once the historic attribute information was joined to the modern place-name point layer, attribute renderings were created based on specific research needs. The ArcGIS Maplex extension was used for developing automated label schemes that intelligently placed ancient and modern place-name labels for each location. Custom label expressions of contrasting styles for modern and historical names were derived from a single data table. The following code is an example of a label expression that displays the modern place name in Arial, and the ancient name is shown in parentheses in a smaller, italicized font.

Map-based annotation layers were used extensively for fine tuning of label placement after the initial label layout was generated. Given the large number of annotations, the ability to make manual graphic adjustments saved many hours of map production time.

An important feature of Aztec political geographies is how places interacted through political, social, and economic systems. Conceptual representations of place relationships were based on temporal period and visualized using line-direction symbols. Interpretation of the historical record ranged from clear indications of place interactions confirmed by explicit references to place-names to generalized, inferred, or uncertain references concerning place interactions. A directional arrow showed the interaction between Aztec capitals and subjects. The certainty or uncertainty of an interaction was recorded as an attribute as well as the name of the TO and FROM points of the interaction.

To effectively show these interactions, an automated spider diagram creation procedure was developed using ArcObjects. This procedure helped draw lines of interaction by querying capital/subject relationships stored in the historical database. This saved time and provided an automated, exploratory method for finding sites when visual inspection was no longer effective such as when multiple locations with the same name were located near each other. Approximately 25 percent of the time, computer-generated connections were wrong and subsequently dropped from the historic database, but sometimes key links were found.

Future Research

As the analysis moves further back in time, the historical record no longer provides evidence and archaeological investigation will be required. Regional-level place interaction analysis can determine which sites should receive a higher priority when developing research agendas for archaeological field verification of locations.

Improved modern place-name database development, another area for future research, could include a systematic comparison of NGA place-name points with annotation labels shown on 1:50,000-scale quadrangle maps from the Instituto Nacional de Estad�stica Geograf�a e Inform�tica (INEGI). In less populated places, place-names don't exist in the NGA database but are shown on the INEGI maps. Locating a broader range of modern place-names makes more effective historic reconstructions possible. Finding a Mexican basemap series that predates the 1988 INEGI series would aid historic name identification by providing additional information on variations in place-name spelling.

Summary

GIS can aid interdisciplinary research in history, archaeology, and geography. The use of GIS and relational database management technology and datasets helped manage information and increased understanding of where ancient Aztec places were located and how they interacted with each other over time. Through improved data structuring, visualization, and analysis, GIS has also helped manage issues of uncertainty that exist in the historical record. For more information, contact

Brian Tomaszewski
Department of Geography
Pennsylvania State University
302 Walker Building
University Park, Pennsylvania 16802
E-mail: bmt139@psu.edu

References

Berdan, Frances, and Patricia Rieff Anawalt, Editors. 1997. The Essential Codex Mendoza. Los Angeles: University of California Press.

National Geospatial-Intelligence Agency: GEOnet Names Server (GNS), earth-info.nga.mil/gns/html.

Acknowledgments

Portions of this research were funded by master's thesis research awards from the Association of American Geographers (AAG) Cartography Specialty Group and the Terry G. Jordan-Bychkov Student Research Award from the AAG Historical Geography Specialty Group.

The author thanks Dr. Michael E. Smith of Arizona State University (formerly of State University of New York, Albany) for providing the idea for the project, many of the critical datasets used in the analysis, and general direction for the project. The author would also like to thank Dr. Alan M. MacEachren and Dr. Deryck W. Holdsworth of Pennsylvania State University, Department of Geography, for their assistance.

About the Author

Brian Tomaszewski, a Ph.D. candidate in the Department of Geography at Pennsylvania State University, conducted the research and software development discussed in this article. He has a master's degree in geography from the University at Buffalo, New York, and his research interests include GIScience, visualization, historical GIS, and spatial data modeling and implementation.