From Spins to Birds to Traders: Model Adaptation in the Study of Critical Behavior

Key data of the project

Description of the project

Is there a common ground between the behavior of a flock of starlings and the polarization of opinions within human communities? Can we represent traders in a financial market just like spins in a magnet? For years, scientists have conjectured that a certain type of collective behavior ubiquitous in nature adhere to common principles, which can be captured by the physics of criticality. Originally, this idea was largely speculative, but the last three decades have witnessed an increase in the number of concrete approaches to biological and socioeconomic modelling inspired by the physics of criticality. Although plenty of resources are currently spent with the hope that physics can help us prevent unwanted coordinated events, such as population collapse or political radicalization, the foundations for extrapolating the physics of criticality into other sciences remain unclear.

MACBeh’s ambition is to tackle this timely challenge by addressing three outstanding questions: (Q1) What is transferred from physics to other sciences in the case of criticality? (Q2) What explains the predictive and explanatory power of some models of criticality outside of physics? (Q3) How can we improve the practice of model building in the study of criticality? We will address these questions from an interdisciplinary perspective, by connecting relevant debates in philosophy of science with a systematic study of existing models of criticality in biology and socioeconomics. The underlying thesis is that models of criticality outside of physics are cases of “model adaptation,’’ a powerful new framework stating that successful models in the new domain are constructed based on universal properties and causal mechanisms borrowed from physics plus the incorporation of domain dependent properties, which have no analog in the original physical models. Our results will lead to a step change in our understanding of cross-disciplinary knowledge transfer for the case of critical behavior and beyond.