4.24.5 Tensor Algebraic Structure Boundary

The Tensor Algebraic Structure Boundary marks where tensor algebra transitions into broader mathematical structures, defining operational limits within formal systems.

Tensor Algebraic Structure Boundary is the delineation of what algebraic structure the collection of multilinear maps of a fixed arity and fixed domain and codomain naturally carries, fixing that this collection is always a vector space under pointwise operations while marking the point beyond which no further structure — such as a product between two arbitrary multilinear maps of the same type — is automatically present without additional data.

The Structure That Lies Inside the Boundary

Pointwise Vector Space Structure

For fixed vector spaces V_1, ..., V_k and a fixed codomain, the set of all multilinear maps

T : V_{1} \times \dots \times V_{k} \to F

is closed under pointwise addition,

(T_{1} + T_{2}) (v_{1}, \dots, v_{k}) = T_{1} (v_{1}, \dots, v_{k}) + T_{2} (v_{1}, \dots, v_{k})

and under scalar multiplication defined the same way, and this closure is what places the set of multilinear maps inside the boundary as a genuine vector space, isomorphic to the tensor product V_1* ⊗ ⋯ ⊗ V_k* under the tensor product factorization boundary once the spaces are finite-dimensional.

Grading by Arity and Type

Multilinear maps of different arity, or of different pattern of primal versus dual factors, belong to distinct vector spaces under this boundary; there is no addition defined between a bilinear map and a trilinear map, since pointwise addition requires the two maps to accept the same number and type of arguments. The boundary therefore organizes multilinear maps into a family of separate vector spaces indexed by arity and type, rather than a single undifferentiated collection.

What Lies Outside the Boundary

No Automatic Product Between Two Multilinear Maps

Given two multilinear maps T_1 and T_2 of the same type, there is no canonically defined product T_1 · T_2 that is itself again a map of that same type; multiplying pointwise values, where meaningful, generally produces a map of higher arity or a map that fails one of the linearity conditions, and so falls outside the vector space in which T_1 and T_2 live. Any product structure — such as the tensor product of the maps themselves, producing a map of combined arity, or a contraction producing a map of reduced arity — is additional data introduced deliberately, not a consequence of the vector space structure already present.

No Ring or Algebra Structure Without Further Construction

Because no multiplication internal to a fixed type is available, the set of multilinear maps of a given type does not form a ring or an algebra under this boundary alone; it remains only a vector space. Promoting it to an algebra requires either restricting to endomorphism-valued maps, where composition supplies a product, or passing to the full tensor algebra, where the tensor product of maps of varying arity supplies a graded multiplication — both of which are separate constructions lying outside this boundary.

Role Within the Tensor Framework

Foundation for the Tensor Algebra's Vector Space Layer

This boundary establishes the vector space layer that every graded piece of the tensor algebra rests on: each graded component of the tensor algebra is, by this boundary, a vector space of multilinear maps or tensors of fixed type, and it is only the subsequent introduction of the tensor product between components of different type that upgrades the whole graded collection into an algebra.