It is tempting to see the life of an extended society as a species of chaos - normally it is restricted chaos in that the observed actions constitute only a small fraction of the possibilities.  The analogy becomes especially persuasive when one recalls that ancient commonplace that no action, however small, but has consequences that spread through the community : 'for want of a nail the shoe was lost, for want of a shoe...'.⁽³⁴⁾  The implication is that one must not accept the theories of social science to be predictive except for short times ahead, or for certain processes that can develop only slowly, like the fundamental principles of a political party.  It may be, for instance, that grand overall economic theories can be devised, but that the limited economic theories that governments long for will never have predictive validity for far enough ahead to be any use.  Be that as it may, it is desirable that those social scientists who seem to strive to make their discipline conform as closely as possible to the perceived ideal of physics, should recognize that they may be imitating physical procedures at the very point where they are least reliable.  And for their part physicists should not allow themselves to be gratified overmuch when philosophers of science select physics as the typical science.  The phenomenon of chaos is a salutory reminder of the frailty of human endeavour, and it may be that the recognition of the limitations of mathematical prediction will prove the most typically scientific aspect of physics.

-- Alfred Brian Pippard, Response and stability: an introduction to the physical theory (p. 128)