Why Continuum Dynamics Are Not Semantically Closed

Continuum physics represents states as real- or complex-valued fields and dynamics as operators on infinite-dimensional function spaces. Under an ontic interpretation, however, fundamental evolution must be semantically total: it must take every admissible state specification to a successor state specification in which all admitted magnitudes remain denoting. We make this requirement explicit using standard admissible representations, in which denotation is characterized by bounded finite-precision input dependence on state descriptions. We show that standard continuum dynamics can violate this semantic closure requirement even when the underlying PDE is linear, well posed, and uniquely solvable. The obstruction is forced extensional totalization: a regional readout can require joint control over countably many individually well-behaved contributions in a way that destroys bounded input dependence and hence denotation. Because locality commits a field ontology to treating spatial restrictions on compact regions as single physical magnitudes, this failure mode cannot be avoided by appealing to pointwise readouts or sequential access. Reanalyzed in these terms, the Pour-El and Richards Theorem 2 construction for the three-dimensional wave equation provides a concrete witness: admissible finite-energy initial data evolve to states whose regional field restrictions do not admit admissible names in the required function-space topology. The obstruction is semantic rather than epistemic: it does not concern what can be computed or observed, but the failure of continuum evolution to preserve denotation under realist locality commitments. The conclusion is eliminative: unrestricted continuum field ontology cannot serve as the basis of a semantically closed fundamental dynamics.