Complex system
A complex system is a system composed of many interacting components whose collective behavior cannot be predicted from the properties of the individual parts in isolation. The term does not mean 'complicated' — a Boeing 747 is complicated but not complex, in the technical sense, because its behavior is largely predictable from the designed interactions of its subsystems. Complexity arises when local interactions generate global patterns that feed back to alter the local interactions themselves, producing dynamics that are non-linear, path-dependent, and historically contingent.
The study of complex systems spans physics, biology, ecology, economics, and computer science, but its conceptual unity lies in a shared set of organizational principles: emergence, feedback, self-organization, and adaptation. These principles do not reduce to a single formalism. They constitute what the physicist Philip W. Anderson called 'more is different' — the claim that quantitative increases in scale or connectivity produce qualitative changes in behavior that require new concepts, new mathematics, and new modes of explanation.
Key Properties
Non-linearity is the defining mathematical feature of complex systems. Small changes in inputs can produce disproportionately large changes in outputs; conversely, large changes may produce no effect at all if they fall within a stable basin of attraction. This sensitivity makes prediction difficult and control dangerous. A policy designed to stabilize a market may, through feedback effects, amplify the very volatility it was intended to dampen.
Emergence names the phenomenon in which system-level properties arise from interactions among components and are not possessed by any component individually. The stability of a food web, the coherence of a flock of birds, the crash of a financial market — these are emergent properties. The debate over whether emergence is merely a failure of computational tractability or a genuine ontological feature remains unresolved, and the answer likely depends on the specific system and the level of description. See Community Ecology for a case study in ecological emergence, and Emergence for the broader philosophical debate.
Self-organization is the capacity of a system to produce and maintain its own structure without external direction. Dissipative structures — such as convection cells, autocatalytic chemical networks, and living cells — extract energy from their environment and use it to maintain ordered states that would spontaneously decay in equilibrium conditions. Self-organization is not teleological; it does not require a blueprint or a goal. It requires only energy flux, local interaction rules, and a mechanism of positive and negative feedback.
Complex Systems and Other Disciplines
Complex systems theory has been applied to social systems, where the diffusion of beliefs and practices is modeled as an epidemiological process; to algorithmic game theory, where the interaction of simple strategic agents produces market dynamics that no individual agent intends; and to Memetics, where cultural evolution is understood as a complex adaptive system in which ideas compete, mutate, and co-evolve with the cognitive architectures that host them.
The connection to analogical reasoning is less obvious but equally deep. Complex systems are understood not by direct analysis but by mapping them onto other systems that share their relational architecture. The Ising model in physics, the Lotka-Volterra equations in ecology, and the SIR model in epidemiology are all formally analogous — they describe different domains with the same mathematical skeleton. To study complex systems is to practice analogy at the highest level of abstraction.
The most seductive error in complex systems thinking is the belief that because a system is complex, it is therefore mysterious. It is not. Complexity is a property of formal systems, not a negation of formalism. The equations that describe a complex system may be intractable, but they are not magical. To invoke complexity as an explanation is to mistake a description for a cause. The challenge of complex systems science is not to celebrate mystery but to build better mathematics — and better analogies — for systems that current tools cannot yet capture.