Collaborative Research: Methodological Challenges and Interpretations in Network Analysis of Artifact Data

University Of Wisconsin-Milwaukee, Milwaukee WI

Investigators

Abstract

Across many fields, scientists increasingly conceive of diverse research problems as involving networks of some kind, and tackle these problems via methods for network analysis. Archaeology is no exception. One particularly fruitful application of the network perspective in archaeology has involved the construction of networks representing ties between inhabited sites in a region, with the presence or strength of a network link between two sites inferred from the observed similarity between the sites' artifact assemblages. For instance, research on the pre-Hispanic American Southwest has examined the similarity between distributions of ceramic wares collected at the sites, with micro- and macro-regional networks of sites derived from these ceramic similarities. Investigations of these networks, and their change over time, via network analytic methods provide important insights into social processes that are of intense interest in archaeology, such as migration, population and depopulation, and diffusion of ideology. However, this approach also carries with it grave risks of inappropriate application of network analytic tools. These tools were developed in other disciplines and may sometimes be ill-suited to the archaeological context, and the archaeological setting can pose particular challenges for interpretation of results from these methods. The current research project undertaken by PI John M. Roberts, Jr. (University of Wisconsin-Milwaukee) and co-PI Matthew A. Peeples (Arizona State University) will involve an interdisciplinary team working to develop in-depth assessments of the validity of network analytic tools in the archaeological context, the implication of methodological choices for resulting archaeological interpretations, and the investigation of new approaches to analyzing archaeological network data. This work will be carried out in collaboration with a graduate research assistant at UW-Milwaukee, for whom the project will provide in-depth training in network analysis and its application to substantive research questions. Additional colleagues from the University of Arizona will participate as consultants, and online tools will be developed to help spread the project's benefits to other researchers, graduate and undergraduate students, and the public. This research will address a number of specific technical issues that are at present inadequately understood in the archaeological literature, such as measurement of uncertainty in network measures due to sampling variability in collected artifact data, methods for creating graphs representing archaeological networks that distill complex information into accessible visual displays, and comparisons between results of network-based analyses with those of other analytic approaches to artifact similarity data. The project will also explore more directly archaeological questions, including the relationship between the spatial arrangement of sites and the similarity of their artifact assemblages, and how network change corresponds to other dimensions of regional chronology. Some data for the project are derived from the team's previous work on the American Southwest, but the project will also examine other network data from a wide variety of archaeological contexts, times, and places. In sum, the project will greatly improve understanding of the technical and substantive issues surrounding the current practice of archaeological network analysis, while also creating a roadmap for new approaches that will yield deeper archaeological insights. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.