A Capacity Functional Framework for the Structural Classification of Constraint Satisfaction Systems

This paper introduces a unifying framework for the structural classification of constraint satisfaction systems using the capacity functional \Phi(A) = \log |\Omega_A|, where \Omega_A denotes the set of configurations satisfying a constraint subset A. The curvature properties of \Phi over the lattice of constraint subsets provide a structural trichotomy: Class I (Additive / Rank-Based Systems): Systems whose capacity functional is submodular and corresponds to a matroid rank structure. In these systems, global constraint counting combined with local sparsity conditions is sufficient for saturation. Class II (Multiplicative / Entropic Systems): Systems governed by exponential pruning of solution space. Modular functionals correspond to independent constraints, while supermodular functionals capture correlated regimes. A general theorem shows that supermodularity implies the first-moment (modular) threshold is an upper bound on the true saturation threshold. Class III (Hybrid Systems): Systems combining additive and multiplicative channels through a product-lattice structure. An interaction functional formalizes coupling between rank and entropic mechanisms, and a conjecture is proposed regarding optimal mixed saturation strategies. The framework does not introduce new universal constants or complexity separations. Rather, it provides a functional and curvature-based language that unifies matroid theory, entropy methods, and probabilistic threshold analysis under a common structural lens.