5.14 Tensor Canonical Product Map

Tensor Canonical Product Map is the standard name for the map ⊗: V × W → V ⊗ W that sends a pair of vectors (v, w) to their simple tensor v ⊗ w, serving as the fixed, structure-preserving bridge between the raw pairing of two vector spaces and the algebraic object built to encode bilinear behavior. It is "canonical" because it arises directly from the construction of the tensor product itself, requiring no arbitrary choices such as a basis, and it is this map, rather than any auxiliary map, that carries the defining universal property.

Definition and Basic Properties

Formal Definition

For vector spaces V and W over a field F, the canonical product map is defined by:

where the right-hand side denotes the equivalence class of (v, w) in the quotient construction F(V × W) / R used to build V ⊗ W.

Bilinearity, Not Linearity

The canonical product map is bilinear rather than linear when V × W is regarded with its direct sum vector space structure. This means it satisfies additivity and scalar compatibility separately in each argument, but does not satisfy the single linearity condition ⊗(a·(v,w) + b·(v',w')) = a·⊗(v,w) + b·⊗(v',w') that would be required of a linear map on V ⊕ W.

Canonicity: Why No Choices Are Involved

Independence from a Choice of Basis

Unlike many maps used to compute with tensor products, such as those defined by fixing bases for V and W and mapping basis vectors to basis tensors, the canonical product map requires no such choice. It is defined purely from the pairing operation and the quotient construction, making it intrinsic to V and W themselves.

Naturality Under Linear Substitution

Given linear maps φ: V → V' and ψ: W → W', the canonical product maps for (V, W) and (V', W') satisfy the compatibility condition:

confirming that the map behaves consistently, or naturally, as the underlying vector spaces are transformed.

The Map's Role in the Universal Property

Serving as the Universal Bilinear Map

The canonical product map is precisely the universal bilinear map through which every other bilinear map on V × W factors uniquely. It is not merely one bilinear map among many; it is the specific, distinguished map around which the entire theory of the tensor product's universal property is built.

Uniqueness of the Canonical Map's Role

If a different bilinear map τ: V × W → T also satisfied the universal factorization property with the same uniqueness guarantee, then τ and ⊗ would necessarily correspond under a canonical isomorphism T ≅ V ⊗ W, showing that the canonical product map is unique up to this isomorphism, not in an absolute sense but as the representative of a uniquely determined universal role.

Structural Diagram of the Map

Image and Kernel-Like Behavior

The Image Is the Set of Simple Tensors

The image of the canonical product map is exactly the set of simple tensors in V ⊗ W. Since simple tensors span but generally do not exhaust the additive structure of the tensor product, the canonical map is typically not surjective as a map between vector spaces, even though its image generates the codomain.

No True Kernel, but Fibers over Simple Tensors

Because ⊗ is a map of sets rather than a linear map on V ⊕ W, it does not have a kernel in the usual linear-algebra sense; instead, different pairs (v, w) and (v', w') can map to the same simple tensor whenever they are related by the scalar compatibility identity, for instance (cv, w) and (v, cw) map to the same element for any nonzero scalar c.

Applications and Significance

Foundation for Defining Tensors Concretely

The canonical product map provides the concrete link between abstract elements of V ⊗ W and pairs of vectors from V and W, which is essential when tensors are used to represent physical or geometric quantities built from combinations of vectors, such as stress tensors or multilinear forms.

Building Block for Iterated Tensor Constructions

The same canonical map, applied repeatedly, underlies the construction of higher tensor powers V ⊗ V ⊗ ... ⊗ V and the full tensor algebra, where the canonical product map at each stage supplies the universal bilinear pairing needed to extend the construction by one more factor.