2.1.1 Tensor Algebra Vector Space Requirement

Tensor Algebra Vector Space Requirement is the specific set of conditions a candidate space must satisfy before it can serve as the underlying V in a tensor construction, finite dimensionality, a well-defined field of scalars, closure under linear combination, and the existence of a well-behaved dual space, distinguishing spaces that are suitable as a foundation for tensor algebra from those that are not, regardless of how superficially similar they may appear.

Finite Dimensionality

Why an Infinite Basis Breaks the Standard Construction

The standard tensor construction relies on describing a tensor's components as a finite array indexed by a fixed number of basis vectors, n values for a single index, n squared for two, and so on; if V has infinite dimension, no such finite array exists, and the entire apparatus of finite index ranges and finite summation no longer applies without substantial modification.

Finite Dimensionality and the Identification of the Double Dual

Finite dimensionality is also what guarantees the natural identification of V with its double dual (V*)*, an identification that the standard treatment of vectors and covectors as interchangeable arguments of multilinear maps depends on; this identification can fail or require extra structure in the infinite-dimensional setting.

A Well-Defined Field of Scalars

Scalars Must Support Ordinary Arithmetic

The field F underlying V must support addition, multiplication, and division by nonzero elements, since tensor operations, contraction, forming linear combinations, computing determinants used in the transformation law, all require these operations to be available and to behave predictably.

Why a Mere Ring of Scalars Is Insufficient

If the scalars come only from a ring rather than a field, missing guaranteed division, key steps such as normalizing a vector or inverting a change-of-basis matrix, both essential to the transformation law, may fail to be well defined, which is why tensor algebra as standardly developed restricts its scalars to a field.

Closure Under Linear Combination

The Basic Vector Space Axioms

V must satisfy the ordinary vector space axioms over F, closure under addition and scalar multiplication, existence of a zero vector, existence of additive inverses, associativity and distributivity, since these are what allow a general element of V to always be expressed as a linear combination of basis vectors.

Why This Guarantees Well-Defined Components

Closure under linear combination is what guarantees that once a basis is fixed, every vector in V has a unique, well-defined set of components relative to that basis, the starting point from which the components of every higher-type tensor are ultimately assembled.

Existence of a Well-Behaved Dual Space

The Dual Space Must Be Equally Well-Structured

Because covectors and higher mixed-type tensors are built using V*, the space of linear functionals on V, the requirement extends to V* itself being a finite-dimensional vector space over the same field F, with a dual basis that pairs correctly with any chosen basis of V.

Why This Follows Automatically in Finite Dimensions

For finite-dimensional V over a field, this requirement is automatically satisfied, the dual space always exists, always has the same dimension as V, and always admits a dual basis, which is part of why finite dimensionality is listed as the primary requirement: once it holds, the dual space requirement follows without any additional assumption.

Consequences When a Requirement Fails

Infinite-Dimensional Spaces Require Modified Machinery

When V is infinite-dimensional, as in many function spaces, tensor-like constructions still exist but require additional structure, topological completeness, careful treatment of convergence, that goes beyond the purely algebraic, finite-index machinery standard tensor algebra relies on.

Non-Field Scalar Structures Require Modified Machinery

When the scalars come only from a ring, as in modules over a ring rather than vector spaces over a field, an analogous but distinct theory, that of tensor products of modules, is needed, and results that rely on invertibility, such as certain normalization or basis-change arguments, may no longer hold without extra hypotheses on the ring.

Why These Requirements Matter for the Rest of Tensor Algebra

Every Later Construction Presupposes These Conditions Hold

Rank, type, the transformation law, contraction, and the identification of tensors with multilinear maps are all developed under the standing assumption that V is a finite-dimensional vector space over a field with a well-behaved dual, so verifying these requirements at the outset, rather than assuming them implicitly, prevents later constructions from being applied outside the setting in which they are actually valid.

A Checklist for Applying Tensor Algebra to a New Setting

When tensor algebra is applied to an unfamiliar space, a space of matrices, a space of polynomials up to a given degree, a tangent space at a point on a manifold, checking these requirements explicitly, finite dimension, field of scalars, vector space closure, dual space existence, confirms that the space is a legitimate candidate for the standard construction before any tensor-specific reasoning is applied to it.