Critical Realism, Local Knowledge, and Autonomous Knowledge Systems in Economics: An Agent-based Modeling Investigation

This paper aims to adopt a relatively formalized approach to explore the potential advantages of autonomous knowledge systems within the field of economics compared to alternative epistemological frameworks. Drawing upon theoretical perspectives from Critical Realism, Social Positioning Theory, and Local Knowledge Paradigms, we construct a multi-layered behavioral logic embedded in an Agent-based Model. Employing this computational model, we systematically analyze the relationship between autonomous knowledge systems and other epistemological approaches—including universalist knowledge systems—in terms of their respective capacities for economic explanation and prediction under varying social ontological assumptions. The study finds that when economic systems manifest any of the following conditions: (1) macro-structural influences on agent behavior, (2) socially embedded relational configurations, or (3) complex networked interactions among individual actors, autonomous knowledge systems consistently exhibit superior explanatory and predictive power. These findings provide theoretically generalizable support for the systematic construction of autonomous knowledge frameworks in economic research.