Abstract

Advances in observational techniques now allow us to apply statistical methods to study a range of many-body biological phenomena previously considered only, on the one hand, in narrative or descriptive form or, on the other, in terms of toy models that could not handle the large-N and the non-mean-field system features. I here discuss some of the theoretical aspects of biological systems, show the powers and limitations of physical models for 'life-like' systems such as the spin-glass, and show new results for how these concerns come together in a study of unprecedentedly detailed time-series data from a primate society.