Double adjoints of submodules and densely defined operators #
The double adjoint of a closed submodule is the original submodule. As a consequence, the adjoint of a closed densely defined partial linear map has dense domain.
Main results #
Submodule.le_adjoint_adjoint: double adjoints contain the original submodule.Submodule.adjoint_adjoint_le: reverse inclusion for closed submodules.Submodule.adjoint_adjoint: equality for closed submodules.LinearPMap.dense_adjoint_domain: a closed densely defined operator has a densely defined adjoint.
Provenance #
Adapted from the Kitware DKPS formalization, originally authored in ForTauCeti
(copyright (c) 2026 Kitware, Inc.; authors Jon Crall and Claude Opus 5; Apache-2.0).
Every submodule is contained in its double adjoint, without closedness or completeness assumptions.
The double adjoint of a closed submodule is contained in the submodule when the ambient inner product spaces are complete.
Involutivity of the submodule adjoint on closed submodules.
This is the graph-level statement that makes the unbounded-operator adjoint
theory close on itself: with LinearPMap.adjoint_graph_eq_graph_adjoint it says
Γ(T††) = Γ(T) for closed densely defined T.
von Neumann's theorem. The adjoint of a closed densely defined partial linear map between complete inner product spaces has dense domain.