Landscape-based Null Models for Archaeological Inference

Abstract

How do we, as humans and as scientists, learn about the world around us? In this dissertation, I explore how models--epistemological tools that connect theory and reality--not only structure scientific inquiry (including the social sciences), but also reflect how humans experience and understand the world. Using this insight enables anthropologists and other social scientists to build more ontologically powerful understandings of human behavior. Here, I focus on how humans experience physical and social landscapes--the environments in which they live and with which they interact. The dissertation consists of three studies, each of which build on the previous by adding to the complexity of modeled landscapes. The first concerns static landscapes--those that are unchanging over the temporal timescales relevant to human experience. I develop a topographically-derived index of defensibility and use it to infer defensive behavior among prehistoric populations in the Northwest Coast of North America. The second paper introduces dynamic landscapes--those that change at scales experienced by humans, but whose changes are primarily driven by external forces. An example relevant to agrarian societies is climate change. I develop a new method for reconstructing past climate landscapes and explore the potential impacts of those changes on Ancestral Pueblo maize farmers in the southwestern United States over the past two millennia. Finally, the third paper grapples with complex landscapes--dynamic landscapes in which human behaviors play important and recursive causal roles. I highlight the coevolution of locally-adapted maize varieties and human selection and cultivation strategies as an example of these types of landscapes, and develop frameworks for modeling maize paleoproductivity that can better honor the realities of Pueblo agricultural strategies.