Self-organization
Self-organization is the process by which global order and structure emerge from local interactions among components of a system — without any central planner, external template, or explicit program specifying the outcome. The organization is not imposed from outside; it is a property of the dynamics.
Self-organization occurs in systems as different as Bénard cells (convective rolls in heated fluids), termite mound architecture, slime mold aggregation, neural synchrony, market price formation, and language change. What these cases share is the same structural feature: local interactions governed by simple rules, repeated at scale, produce macroscopic patterns that are not derivable from the rules alone without running the dynamics. The pattern is real, but it is nowhere in the rules.
The conceptual stakes are higher than they first appear. Self-organization is the mechanism by which complex adaptive systems produce emergent structure — and therefore the mechanism by which explanation at the systems level is irreducible to explanation at the component level. A gas's temperature is a statistical property of molecular motion: you can derive it from the components. A termite mound's architecture is a self-organized outcome: you cannot derive the architecture from the rules governing individual termites without simulating the population. This is not a failure of reduction in principle; it is a failure of tractability that has the same practical effect.
The physics of self-organization draws on dissipative structures (Ilya Prigogine's Nobel Prize-winning work on systems far from thermodynamic equilibrium), symmetry breaking (the selection of one structural outcome from among equally probable alternatives), and bifurcation theory (the qualitative change in a system's behavior when a control parameter crosses a threshold). In all cases, the ordering is purchased at the cost of entropy export: self-organizing systems maintain low entropy locally by dissipating it into their environment.
The skeptic's challenge: self-organization is sometimes invoked to explain away rather than explain. When a theorist says 'the market self-organizes,' they may mean something precise (local price signals coordinate decentralized decisions into an equilibrium) or they may mean something vacuous (nobody is in charge). The distinction matters, because the precise claim is falsifiable and the vacuous claim is not. Any time 'self-organization' appears as an explanation without a specified mechanism for how local rules produce the observed global pattern, it is not an explanation. It is a label for ignorance that sounds like a discovery.