Closure Dynamics, Reopening, Hysteresis, and Fixed-Point Evolution

Advances CSM from a static graph to a time-indexed dynamical system, evolving through event-driven updates. Core claim: evidence accumulation can be monotonic, but the closure state usually is not — old evidence is never deleted, but BLOCKED/CLOSED/EXHAUSTED states can be downgraded to STALE/REOPENED. Defines closure events, closure scheduling, event commutation/non-commutation, closure hysteresis (the same final evidence arriving in a different order can produce path-dependent, different outcomes), reopening waves, debt discharge, frontier drift, local closure fixed points, relative equilibrium, closure attractors/cycles, metastability, and closure oscillation. Central non-collapse principle: $\mathfrak{C}_{t+1}=\mathfrak{C}_t \not\Rightarrow \mathfrak{C}_t=\Omega^{math}$ — reaching a fixed point under a fixed theorem library/grammar/representation/bridging policy is only a relative closure fixed point, and never means “the mathematical space is complete.”