The Sandpile Model of Japanese Empire Dynamics

Abstract

Global empires take on the same pattern of life cycles (rises and falls). As a typical history of empires, Japan has an obvious life cycle pattern within 2246 years. Natural and human societies (empires) are complex systems, which is why we witness cross-system patterns. Individuals’ local (micro-level) interactions determine or shape a macro-level life cycle pattern. Individuals’ local (micro-level) interactions determine or shape a macro-level life cycle pattern. When a critical threshold is reached, any local changes can lead to phase transitions of systems. To investigate this pattern, internal (microscopic) mechanisms should be applied. Self-organized criticality and complexity theory are applied to reveal similar mechanisms. Based on related theories, we build our agent-based model. Deeming Japanese empires as the target function, we traverse parameters to obtain the highest matching degree (optimal solution). For 13 Japanese empires, the best and multiple simulations can perfectly match real empires in history. Besides, we infer counterfactual, for 11, 12, 14 & 15 empires, and achieve a perfect matching. So, our model’s validity, robustness, and generality can be confirmed. It implies that natural and human systems (Japanese Empires) share similar mechanisms.