Natural Agents, Complex Systems and Uncertain Environments: Results from Simulations on Heuristic Organizations

The primary objective of this work is to propose a naturalistic perspective on human cognition and agency as a foundation for the investigation and representation of microeconomic phenomena arising from the interactive behavior of organizations and of agents within organizations. This dissertation examines the evolution of bounded rationality in its ecological interpretation and explores both theoretically and practically the possibility of empirically grounding the dynamics of economics as an evolving complex system, starting from a realistic and descriptive attempt to approach the representation and modeling of economic agents. The thesis traces the development of Herbert Simon's concept of bounded rationality from its origins, focusing on the Adaptive Behavior and Cognition legacy that recognized the fundamental relationship between cognition and the agent's environment and interpreted rationality from an adaptive perspective. From the idealized homo economicus, highlighting the relevance of uncertainty, this theoretical journey traces the shift to a more realistic anthropology of homo heuristicus, by identifying the heuristics adopted in decision-making processes – encompassed within an embodied and enactive dimension of cognition – as building blocks for a mind-society microfoundation trajectory. Subsequently, complex social systems are presented as the fundamental units of economic reality, alongside a perspective from which to pursue microfoundation attempts starting from the representation of real-world agents. The epistemological and methodological coordinates oriented towards emergence and interactions, preparatory to the endeavor to merge Simon's analysis on cognition and that on organizations as hierarchical near-decomposable complex systems, are then outlined. A suitable paradigm that indicates which phenomena need to be explained and microfounded is then identified: a theoretical framework that combines evolutionary and complexity economics and conceptualizes the economy as an evolving complex system, denying the concept of equilibrium and focusing on dynamics, heterogeneity, and epistemological emergence. Within this paradigm, homo heuristicus finds its habitat in organizations, determining their resulting problem-solving behavior in uncertain environments. The second part of the thesis attempts to practically follow the traced theoretical route by incorporating the anthropology of homo heuristicus into Agent-based simulations. These are programmed on analytically tractable models that aim to provide a basis – consisting in the logic adopted and the results obtained – for implementing more sophisticated models such as Agent-Based or NK models. In this phase, populations of performing heuristic agents (organizations) in a turbulent environment are simulated, starting from the most fundamental heuristic: satisficing. Relative and absolute aspiration levels are set and turbulence is generated, and various results concerning observable properties and dynamics at the aggregate level are observed, analyzed, and interpreted. The behaviors of agents are progressively sophisticated and enriched by incorporating local search behavior, exploitation-driven organizational learning, exploration and exploitation strategies, social imitation heuristics and other secondary, more realistic decision-making and problem-solving traits, thus providing theoretical contributions and identifying practical orientations.