The thermal roots of correlation-based complexity

TL;DR

This paper explores how thermal physics and information theory, through Bayesian maxent and net surprisal, provide a framework for understanding the emergence and measurement of complexity and correlations in natural systems.

Contribution

It introduces a thermodynamic approach to analyzing correlations and complexity, emphasizing the role of reversible thermalization and multiscale strategies in natural processes.

Findings

Net surprisal as a free energy analog constrains mutual information.

Thermalization influences the evolution of complexity in natural systems.

Correlations can be viewed as delocalized physical structures across scales.

Abstract

Bayesian maxent lets one integrate thermal physics and information theory points of view in the quantitative study of complex systems. Since net surprisal (a free energy analog for measuring "departures from expected") allows one to place second law constraints on mutual information (a multi-moment measure of correlations), it makes a quantitative case for the role of reversible thermalization in the natural history of invention, and suggests multiscale strategies to monitor standing crop as well. It prompts one to track evolved complexity starting from live astrophysically-observed processes, rather than only from evidence of past events. Various gradients and boundaries that play a role in availability flow, ranging from the edge of a wave-packet to the boundary between idea-pools, allow one to frame wide-ranging correlations (including that between a phenomenon and its explanation) as delocalized {\em physical} structures.