"The biggest error in our view would be to see in models accurate representations of specific real-world social systems and reliable tools to predict their behaviour, except in the most constrained and bounded cases, and certainly not in the case of anything as large and complex as international or national politics.
"There are several reasons for this position, some of which pertain to complex
systems, others being of a more general nature. Sensitivity to initial conditions,
non-linearity and feedback preclude any reliable prediction of a given social
system beyond the shortest timescales, in the same way that the weather still
eludes trustworthy forecasts beyond a few weeks. The study of social systems is
further complicated by the reflexivity of actors capable of absorbing and adjusting
to the very knowledge produced about them. Furthermore, any model requires the
drawing of system boundaries that complexity has shown to be highly porous,
along with necessarily arbitrary choices in the scale of analysis. Finally, models of
real-world social systems frequently rely on implicit, unexamined or unacknowledged
epistemological and ontological assumptions."
Bousquet, Antoine and Curtis, Simon(2011) 'Beyond models and metaphors: complexity theory,
systems thinking and international relations', Cambridge Review of International Affairs, 24: 1, 43 — 62
But: surely sensitivity to initial conditions, non-linearity and feedback do not preclude any reliable prediction of any given system beyond the shortest timescales? Is that correct? (Cf. examples of physical systems that are linear & deterministic).
Is "the social" being understood as necessarily complex in the mathematical sense? If so, is it true?
Some systems are nonlinear and yet predictable (e.g. multistability).
Are there non-complex systems in the world? Or only in mathematics? Does sensitivity to initial conditions mean that an apparently non-complex system may "transform" at any moment into a complex one? (Cf. "boundaries" q.v.).
How does the idea that complex systems are inherently unpredictable relate to the idea that no predictable system is 100% predictable?
Why further complicated by the reflexivity of actors? Isn't this better understood as an aspect of the complexity of the social?