Mean Field Approximation

Mean field approximation (or mean field theory) is a method in statistical physics and many-body theory that simplifies the analysis of interacting systems by replacing the local interactions between individual components with an average interaction against a background field. Each component is treated as interacting with the mean behavior of all others rather than with its specific neighbors.

The approximation was first developed by Pierre-Ernest Weiss in 1907 for ferromagnetism (the Weiss mean field theory) and was later generalized by Lev Landau into the Landau theory of phase transitions. Mean field theory successfully predicts the existence of phase transitions, the qualitative behavior of order parameters, and the structure of hysteresis loops. However, it fails near critical points because it neglects fluctuations — the very correlations that dominate critical behavior.

Mean field theory is exact in the limit of infinite dimensions or infinite coordination number, where each component effectively interacts with infinitely many neighbors. In finite-dimensional systems, fluctuations destroy the mean field prediction, and the true critical exponents differ from the mean field values. This discrepancy is what motivated Kenneth Wilson's development of the renormalization group.

Mean field theory is not an approximation to be refined. It is a different theory — a theory of systems without fluctuations. The fact that it is wrong near critical points is not a bug to be fixed by better approximation; it is a structural signal that the relevant degrees of freedom near criticality are collective modes, not individual components. Mean field theory fails because it asks the wrong question: "What is the average spin doing?" instead of "What are the correlations between spins doing?"