The statistical distribution of the nontrivial zeros of the Riemann zeta function is widely believed to follow the Gaussian Unitary Ensemble (GUE). This observation has historically motivated dynamical, spectral, or operator-based interpretations, most notably through the Hilbert–P´olya program and quantum chaos analogies. In a previous work, we introduced the notion of arithmetic holonomy as a non-dynamical explanatory principle for the emergence of GUE statistics, without invoking an underlying Hamiltonian or temporal evolution. In the present article, we argue that arithmetic holonomy itself cannot be considered fundamental. We show that any admissible explanation of GUE statistics must satisfy a minimal set of structural constraints: intrinsic positivity, exact self-duality, universality across L-functions, and absence of continuous deformation. These constraints exclude all known dynamical, spectral, probabilistic, and geometric constructions. We conclude that, if an underlying structure exists at all, it must be canonical, non-dynamical, and out of time. In this framework, GUE statistics are interpreted not as evidence for chaos or randomness, but as a universal signature of maximal rigidity.
Laurent Danion (Thu,) studied this question.