PROVIDENCE, R.I. [Brown University] — The American Council of Learned Societies has awarded Brown University Assistant Professor of Anthropology Parker VanValkenburgh and colleagues a $150,000 grant to develop a digital platform for archaeological survey in the Andes.
With Steven Wernke, associate professor of anthropology at Vanderbilt University, VanValkenburgh is developing the Geospatial Platform for Andean Culture, History and Archaeology (GeoPACHA). GeoPACHA will connect satellite imagery from a variety of sources, as well as photos from historic aerial surveys, to build a detailed inventory of archaeological remains in the Andes — including many that have not been discovered yet.
The platform will enable a broader understanding of Andean culture over a large area, VanValkenburgh said.
“The project began with a particular problem in mind,” he said. “Namely, to understand how the Spanish viceroyalty’s forced resettlement of at least 1.4 million indigenous people in the 1570s transformed the social and political landscapes of Peru.”
Many researchers have conducted studies on individual resettlement sites, called reducciones, or small regions including several reducciones, VanValkenburgh said, but there is little sense of the overall resettlement program’s collective impact.
The tool will provide a broad regional view of the impact on the landscape of the mass involuntary migration and might be useful for mapping other kinds of sites, including elements of Inka imperial infrastructure.
“We set about trying to put as many reducciones on a map as we could, based in part on scattered textual descriptions of their locations in three extant census documents,” VanValkenburgh said.
But because many of these towns were never recorded in their own time, he and Wernke needed to build a tool for discovering them. That became GeoPACHA.
Jeremy Mumford, an assistant professor of history at Brown, helped to develop a closely related project as well as the design for the prototype for GeoPACHA.
The platform will incorporate relatively recent satellite imagery and archived historical aerial photographs, including high-resolution images taken by the Peruvian Air Force between the 1940s and 1980s. Those images, VanValkenburgh said, helped found the field of “landscape archaeology” as archaeologists started systematically recording the sizes, locations and surface characteristics of sites in order to chart population history, as well as political and environmental change.
Once the satellite imagery is collected, the next step in the project is a crowdsourcing effort to locate thousands of archaeological sites over nearly 150,000 square kilometers. Using a browser-based tool, researchers including students will directly participate in a “virtual archaeological survey.”
VanValkenburgh said that his team is utilizing this crowdsourcing effort carefully.
“We’ve put together a team of regional editors who have deep knowledge of the archaeology of different regions of Peru and Bolivia, who have conducted on-the-ground surveys in those regions and who will be assembling their own teams of students and collaborators,” he said. “Wernke and I are both regional editors as well. This distinguishes our effort from strictly crowdsourced site identification initiatives. We’re creating highly curated content that's trying to address a series of very specific research questions.”
For the third stage of the project, the team will use these manually identified datasets as training data for a machine-learning algorithm that will be able to identify sites automatically over even larger areas.
The project points to a form of archaeology that bears little relation to popular images of researchers conducting regional settlement surveys — a technique that involves walking systematically spaced areas to map sites, artifacts and features that appears on the ground. That technique is still an indispensable part of the archaeological toolkit, VanValkenburgh said, but like excavation, it is incredibly time-intensive and even the largest surveys cover limited areas.
Instead, GeoPACHA will allow researchers to map massive phenomena and expand the scale of regional datasets, pushing beyond valley-sized silos to consider continental scale patterns.
“When the general public and students think of the type of work archaeologists do, they typically think of intensive research conducted at single sites — excavations that focus on uncovering the layered histories of settlements,” VanValkenburgh said. “But there’s an entire other branch of the field, landscape archaeology, that focuses on understanding the material dimensions of human life beyond the settlement. GeoPACHA will create datasets that enable us to map these massive phenomena and put variations at the level of sites and regions in context.”