Ongoing change in the practice of economic development has led to an increased emphasis on projects designed to liberate people from the uncertainty and instability created by locally specific articulations of environment, economy and society. Critical to these libratory goals are longitudinal understandings of these articulations and how they shift over time in response to various economic and environmental changes. The growing attention given to the study of these local articulations at the sociospatial margins of globalization highlights issues in anthropology, geography and other disciplines surrounding the use and interpretation of data from such contexts. The study of such contexts has, to this point, been dominated by the gathering of data through ethnographic and archival means, sources of data that become difficult to interpret in light of the social contestations that often arise in contexts undergoing rapid and dramatic transformations of economy, environment and society. This dissertation addresses that interpretive difficulty through an approach that incorporates ethnographic, archaeological and archival/documentary data into a methodological triangulation that privileges no single source of data, but relies on a tension between all three to rigorously evaluate and interpret all sources. The application of this triangulation to the study of village growth and abandonment in coastal Ghana provides new opportunities for the examination of past social settings, and therefore enables the construction of provisional longitudinal understandings of these systems in the context of global capital and environment that can be used to shape future research directions and current development projects in a manner otherwise not possible.