14.23 Tensor Map Product Boundary

Tensor Map Product Boundary is the delineation of where the theory of the tensor product of maps, as developed for linear maps between finite-dimensional vector spaces, stops applying in its ordinary form, encompassing both the degenerate special cases that sit at the edge of the construction's own definition and the adjacent theories that extend or replace it once its standing hypotheses no longer hold.

Two Different Kinds of Boundary

Internal Boundary: Degenerate Cases Within the Theory

One kind of boundary arises entirely within the finite-dimensional linear theory itself, at parameter values or inputs that are technically permitted but at which the general formulas take on special, sometimes trivial, forms; these are not failures of the theory but limiting cases that must be checked separately whenever a general argument is applied to them.

External Boundary: Where the Hypotheses of the Theory Fail

The other kind of boundary marks the edge past which the standing hypotheses of the theory, linearity of the maps, finiteness of the dimensions, or the vector space structure itself, no longer hold, requiring passage into an adjacent theory such as module theory, infinite-dimensional or topological tensor products, or category theory, each of which retains some but not all of the finite-dimensional apparatus, as detailed in the tensor map product theory relation boundary.

Degenerate Cases at the Internal Boundary

Zero Maps and Zero Ranks

If either $f$ or $g$ is the zero map, the rank formula $\text{rank}(f\otimes g)=\text{rank}\left(f\right)\cdot \text{rank}\left(g\right)$ forces $f\otimes g$ to be the zero map as well, a degenerate instance of the general multiplicativity of rank that nonetheless requires no separate proof, since zero multiplied by any rank is zero.

Zero-Dimensional Factor Spaces

If $V$ is the zero vector space, then $V\otimes W$ is the zero vector space regardless of $W$, and the only linear map out of the zero space is the zero map, so $f\otimes g$ is forced to be the unique map on a zero-dimensional domain; the basis input elements described elsewhere in the theory form an empty set in this case, and formulas expressed as sums over such elements reduce to empty sums equal to zero, consistent with, but requiring separate acknowledgment within, the general theory.

One-Dimensional Factor Spaces

If $V$ and $W$ are both one-dimensional, $f$ and $g$ are each given by a single scalar, and $f\otimes g$ reduces to multiplication by the product of those two scalars on the one-dimensional space $V\otimes W$; the general Kronecker product machinery still applies formally in this case but produces a single-entry, one-by-one matrix rather than genuinely exhibiting any block structure.

Boundary at the Level of Map Type

Linearity Is a Hard Requirement

The entire construction of the tensor product of maps presupposes that $f$ and $g$ are linear; a nonlinear map between vector spaces has no tensor product defined by this theory at all, since the universal property, the basis formula, and every transformation law rely on linearity at every step. Constructions extending tensor-like ideas to nonlinear maps exist in other areas of mathematics but are unrelated extensions rather than boundary cases of this theory.

Semilinear and Other Weakenings

Maps that are only additive, or that are conjugate-linear rather than linear, sit just outside the ordinary boundary of the theory; a separate, parallel construction is needed to combine such maps via a tensor-product-like operation, and the formulas of the ordinary linear theory do not apply to them without modification.

Boundary at the Level of Field

Behavior Over Non-Algebraically-Closed or Finite Fields

The theory of the tensor product of maps as developed here places no restriction on the base field beyond it being a field, but certain auxiliary facts occasionally invoked alongside it, such as the existence of eigenvalues for every operator, depend on the field being algebraically closed; over fields lacking this property, the eigenvalue-based consequences of operator matrix transformation, such as the product-of-eigenvalues description of the spectrum of $T\otimes S$, may fail to apply directly even though the tensor product construction itself remains perfectly well defined.

Summary of the Boundary's Role

A Map of the Theory's Own Limits

Understanding these internal and external boundaries is what allows a general statement proved for the tensor product of maps to be applied with confidence: internal boundary cases confirm that degenerate inputs do not silently break the general formulas, while external boundaries mark precisely where a different, related theory must be consulted instead of extrapolating the finite-dimensional linear results without justification.