Nonlinear System

A nonlinear system is a dynamical system whose equations of motion contain terms that are not proportional to the state variables. The superposition principle fails: the sum of two solutions is not itself a solution. This apparently minor mathematical difference separates the predictable world of linear systems from the rich, surprising world of chaos, bifurcations, and emergence.

Linear systems can be decomposed: solve each part independently, then add the results. Nonlinear systems resist decomposition. A small change in one variable can propagate through feedback loops to produce large, qualitative changes in the whole. This is why weather defies long-term forecasting, why financial markets crash unexpectedly, and why neural activity can shift abruptly from sleep to seizure.

The study of nonlinear systems is not merely the study of difficult equations. It is the study of systems that cannot be understood by understanding their parts. The whole is not just greater than the sum of its parts. It is different in kind — and that difference is what we call emergence.

Nonlinearity and Thermodynamics

Nonlinearity is not merely a mathematical complication. It is the signature of systems operating far from thermodynamic equilibrium. In equilibrium, systems are linear: small perturbations decay exponentially, and the response is proportional to the disturbance. Far from equilibrium, the entropy gradient itself becomes a nonlinear drive. Bénard convection — the spontaneous formation of hexagonal flow cells in a heated fluid layer — is not an exception to thermodynamic rules but their direct expression. The temperature gradient across the layer is a thermodynamic force; the convection pattern is a nonlinear response that emerges when that force exceeds a critical threshold.

The Second Law guarantees that nonlinearity is the default condition of systems that self-organise. Any system that exports entropy to maintain its internal structure — a living cell, a hurricane, a market — must operate far from equilibrium, and any system far from equilibrium is governed by nonlinear dynamics. Linearity is the special case: the approximation valid only near equilibrium, where Taylor expansion justifies treating response as proportional to perturbation. The real world is nonlinear everywhere except in the carefully engineered regimes we construct to make it behave.

Nonlinearity and Information

Nonlinear systems are information processors. A chaotic system amplifies small differences in initial conditions into exponentially diverging trajectories — this is not merely 'sensitive dependence' but a form of computation in which the system's state encodes information about its past. The logistic map, the simplest nonlinear dynamical system, demonstrates period-doubling bifurcations that encode binary information in the system's orbital structure.

The information-theoretic perspective reveals why nonlinearity is computationally expensive. Nonlinear systems explore their phase space in ways that cannot be compressed; their trajectories are incompressible strings in the sense of algorithmic information theory. This means that predicting the long-term behaviour of a nonlinear system requires storing and processing information at a rate that grows with the prediction horizon. In a thermodynamic universe where information erasure has an energy cost, this places fundamental limits on how far into the future any embedded intelligence can compute its own trajectory.

The deepest implication is structural: nonlinearity is what happens when a system is large enough and coupled enough that its parts cannot be treated independently. The brain is nonlinear because neurons are coupled by synapses that change strength with use. The climate is nonlinear because the atmosphere, oceans, ice sheets, and biosphere exchange energy and matter across multiple timescales. The economy is nonlinear because agents learn, adapt, and change the rules of interaction. In each case, nonlinearity is not a property of the equations we write but a property of the organisation that makes the system persist.