Devoting over a decade to investigating the impact of colonialism on Peru's Indigenous people during the 16th century, Parker VanValkenburgh, an associate professor of anthropology at Brown University, spearheaded intensive archaeological projects studying the aftermath of Spanish conquest. Focused on post-colonization changes, VanValkenburgh sought a broader perspective to contextualize his findings. Collaborating with Steven Wernke from Vanderbilt University, they conceived the Geospatial Platform for Andean Culture, History and Archaeology (GeoPACHA), an open-source tool enabling researchers to map archaeological sites using high-resolution satellite imagery.
VanValkenburgh emphasized archaeologists' proficiency at dissecting the minutiae but acknowledged a lack of systematic data for comprehensive insights into the larger historical picture. GeoPACHA addressed this gap, facilitating collaboration among researchers globally. Over a year, a diverse team utilized GeoPACHA to collect data on ancient sites, roads, livestock corrals, and agricultural terraces. Their collaborative effort, detailed in six papers published in Antiquity, unveils fresh insights into how Spanish colonization and forced resettlement transformed population density, landscapes, and even local climates in the Andean region.
Beyond deciphering 16th-century shifts, the GeoPACHA project exemplifies the potency of collaborative research, showcasing how archaeologists studying niche topics can collectively illuminate global historical changes in climate and culture. VanValkenburgh sees this approach as essential for addressing broader issues such as social inequality, colonial legacies, and societal responses to climate change. Archaeology, with its unique temporal depth, emerges as a crucial discipline offering insights into humanity's historical responses to challenges.
Mapping history
In 2018, GeoPACHA emerged from the collaboration between VanValkenburgh and Wernke, uniting a group of researchers dedicated to addressing specific archaeological inquiries through large-scale satellite mapping. One team focused on mapping hilltop fortifications to explore social conflict in the Andes from 1100 to 1400, while another aimed to pinpoint sites linked to lomas—lush, fog-covered oases amidst an arid landscape—to discern settlement patterns among Indigenous communities.
Leveraging high-resolution satellite imagery within GeoPACHA, researchers, armed with specialized knowledge, meticulously identified and tagged landmarks like hill forts and sheep corrals.
Describing the process as “brute-force” surveying, VanValkenburgh explained, “You take people who are trained to recognize archaeological sites, and you have them just click whenever they recognize something.”
Following the mapping phase, extensive analysis ensued, outlined in the recent Antiquity publications. Bethany Whitlock, a recent Brown Ph.D. graduate in anthropology, led a study examining road and sheep corral locations in Peru's central highlands. Whitlock's findings revealed a concentration of herding infrastructure around significant colonial and present-day settlements, signaling the enduring impact of colonialism on pastoralism and settlement in the Andes.
Another study, led by Giancarlo Marcone, a professor at Peru's Universidad de Ingeniería y Tecnología, delved into the analysis of lomas on Peru's central coast. Marcone highlighted the scarcity of documented activity in these dense vegetation pockets due to their remote, inaccessible locations. With GeoPACHA, Marcone emphasized the newfound ability to study how cultural and climatic changes have shaped lomas.
These research endeavors, as Marcone emphasized, are imperative for comprehending the long-term history of this segment of Andean South America. GeoPACHA's capacity to unveil hidden patterns and transform inaccessible landscapes into tangible data marks a pivotal step in advancing archaeological understanding and addressing urgent inquiries about the region's historical transformations.
GeoPACHA's next phase
VanValkenburgh, Wernke, and their colleagues are not concluding their archaeological mapping endeavors. Teaming up with Vanderbilt Assistant Professor of Computer Science Yuankai Huo, they envision expanding the project's horizons, incorporating the capabilities of artificial intelligence (AI). Through deep learning methods, the GeoPACHA team plans to conduct a comprehensive survey covering nearly the entire central Andean region.
In the upcoming phase, additional scholars from Peru, Chile, Bolivia, and Ecuador, countries integral to the former Inca Empire, will join forces. These experts will leverage AI-generated data, complementing it with their specialized knowledge to enrich the dataset.
“We're taking those data that were painstakingly curated during the first phase of the project, and we're using them to train deep learning models to identify sites en masse,” VanValkenburgh explained. “Using AI, we'll be able to cover an area that's 10 to 20 times larger than our initial research areas. The questions we can address with such a large data pool are pretty exciting.”
Anticipating years of analyses and fresh archaeological revelations, VanValkenburgh envisions the potential to locate and date every ancient terrace constructed in the Andes. This extensive dataset could unlock insights into how factors like climate change, population density, and imperial expansion influenced Indigenous settlement and agricultural decisions.
Reflecting on the transformative nature of large-scale digital projects like GeoPACHA, VanValkenburgh challenged the conventional narrative of archaeology centered on individual discoveries. He emphasized the essence of archaeology as a discipline that unveils relationships—how people interacted, the similarities and differences between societies, and the intricate connections between the past and the present. Moving beyond individual sites and fostering collaborative relationships across institutions and borders, GeoPACHA exemplifies the broader impact achievable through collective efforts in the field of archaeology.
Source: Brown University