Keynes understood money, and the US sociologist Talcott Parsons was much influenced by him. Money is a social construct, an institution. Not the result of rational individualistic bargains….
oh, and on ergodicity, Khintchine conjectured that any sufficiently complicated dynamical system would exhibit, for some observables, a kind of weak approximate ergodicity. I think mathematically it would be interesting to go in this direction and quantify the departures from ergodicity and when they matter based on whether the degrees of freedom of the system are structurally «clumped» together in a way that effectively reduces the complexity of the system in certain aspects so that Khintchine’s purporses, there are not yet enough degrees of freedom. See my paper in a recent copy of Revista Investigación Operacional on this conjecture of Khintchine.