Theorem 2.33.2. (Isometric tensor embeddings). For any normed -spaces , the projective seminorm on is a norm and satisfies . If and are isometric linear embeddings, then are isometric embeddings; the image of the second map is closed.
Proof
Proof. Algebraic injectivity follows by extending vector-space bases. We first compare seminorms. Fix and finite-dimensional subspaces , containing the factors of one expression for . An arbitrary competing expression Let be spanned by and the , and let be spanned by and the . The equality holds already in , by algebraic injectivity. For a fixed , Lemma 2.33.1 gives projections , of norm at most . Applying yields Taking the infimum over competing expressions and then letting gives . The reverse inequality follows from the inclusion of the sets of expressions. Only projections between the displayed finite-dimensional spaces have been used.
On a finite-dimensional tensor , choose bases and their coordinate functionals. These functionals are bounded by Lemma 1.12.1. If the -coordinate of is nonzero, their tensor gives Thus the finite-dimensional projective seminorm is definite. Every nonzero tensor lies in such a subspace, whose inclusion is isometric by the preceding argument; this proves definiteness in general. For one-dimensional factors the norm is the product norm, and the same isometric inclusion proves the pure-tensor formula. An isometry extends to an isometry of metric completions. Its image is complete, hence closed. ◻