“Digital Proxies” for Validating Models of Past Socio-ecological Systems in the Mediterranean Landscape Dynamics Project

Abstract

All representations of the human past are models, whether they are in the form of narratives, equations, or computer algorithms. While we can never know the “true” past, archaeologists seek to create more reliable and useful models of the dynamics of ancient lives and societies. One of the most widely accepted ways to scientifically establish a model’s validity is to compare its results or predictions against observable, empirical, data. However, most archaeological models deal with richly dynamic, living human behavior, social relationships, and interactions with the environment, while the archaeological and associated paleoecological records are composed of fragmentary, altered, static, and discarded material culture; sediments and soils; and dead plant and animal remains. This apparent incommensurability between archaeological models and the empirical data needed to validate them has long created significant challenges for establishing the credibility of archaeological explanation. In spite of ongoing advances in data collection and analysis methods, we cannot change the nature of the archaeological record. But we can modify models to generate outputs more directly comparable with this record. The Mediterranean Landscape Dynamics Project (MedLanD) has created a sophisticated computational laboratory to simulate long-term dynamics of agropastoral land use and landscape evolution. To better evaluate these simulation models, we also have developed a validation instrument that creates a “digital proxy” record based on model results. The digital proxy is analogous to extracting a digital core at specified points in the gridded, digital landscape. It simulates the accumulation over time of a proxy-like record for modeled human land use, vegetation, landscape fire, and surface processes. Digital proxy “cores” can be compared directly with empirical samples taken from analogous points in real-world landscapes, improving our ability to validate complex models. We present an overview of our digital proxy modeling method and a test case of comparing digital and empirical data from locales in Mediterranean Spain.