A locally convex topological vector space with the property that continuous linear maps from it to arbitrary Banach spaces factor in a way analogous to trace-class (nuclear) operators: bounded sets map to relatively compact sets under appropriately approximating finite-rank factorizations, yielding strong compactness and tensor-product regularity.

Principle

Nuclearity encodes a summability condition on the seminorm-dominated approximations of the identity so that operator ideals analogous to trace-class operators govern mappings from the space; equivalently certain canonical maps between completed tensor products coincide.

Demonstration

The Schwartz space S(R^n) of rapidly decreasing smooth functions is nuclear: continuous linear maps from S(R^n) to Banach spaces can be approximated by finite-rank maps with rapidly decaying singular values, and many kernel theorems (Schwartz kernel theorem) rely on nuclearity.

Misapplication

Assuming nuclearity for every Fréchet space or conflating nuclear maps (specific operators) with the global property of a space without checking the required summability/approximation conditions.

Consequence

Excellent mapping and duality behavior: tensor product topologies (projective and injective) agree in many cases, kernels represent continuous bilinear forms, distribution spaces admit kernel theorems, and compactness properties strengthen functional-analytic manipulations.

Reversal

Non-nuclear locally convex spaces—typical infinite-dimensional Banach spaces or many L^p spaces for 1≤p<∞—do not have the summability that forces approximability by trace-class-type operators and therefore lack many compactness and kernel-representation features.

Boundary

Nuclearity is a topological condition dependent on the chosen locally convex topology; it excludes most non-nuclear Banach spaces and must be verified by constructing suitable approximating sequences or checking tensor-product identifications.

Semantic Tension

The adjective 'nuclear' can refer either to nuclear operators (maps with summable singular-value representations) or to nuclear spaces (global topological property); confusing the two blurs local operator behavior with structural qualities of the domain.

Synthesis

A nuclear space is a locally convex space whose topology enforces summable finite-rank approximations of maps out of it, producing compactness-like behavior and regular tensor-product identities that underpin many kernel and duality theorems.