2.6 Tensor Basis Vector Structure

Tensor Basis Vector Structure is the overall architecture by which a single basis of a vector space, together with its dual, generates a coherent family of bases across every tensor space built over that vector space, unifying the vector set, the dual covector set, the spanning and independence properties, and the coordinate assignment into one integrated framework. It is the organizing structure that shows why choosing one basis of the underlying vector space is enough to fix explicit, computable coordinates for tensors of every type simultaneously.

The Generating Data

A Single Choice Fixes Everything

Let $$V$$ be a vector space of dimension $$n$$ over a field $$F$$. The entire tensor basis vector structure originates from a single choice: a basis

of $$V$$. This single choice determines the dual basis $$e^1,\cdots ,e^n$$ of $$V^{*}$$ uniquely, through the duality relation $$e^i\left(e_j\right)={\delta }_j^i$$, so that no further independent choice is required to proceed.

The Two Elementary Ingredient Sets

The full structure is generated from exactly two elementary sets: the basis vector set, supplying contravariant building blocks, and the dual basis covector set, supplying covariant building blocks. Every basis tensor at every type is assembled from these two sets and no others.

Assembling the Basis at a Fixed Type

Formation Rule

For each type $$\left(r,s\right)$$, the tensor basis vector structure produces the induced basis of $$T_s^r\left(V\right)$$ by taking every tensor product of $$r$$ elements from the basis vector set and $$s$$ elements from the dual basis covector set, in every possible combination:

Structure Across Types

Because this construction is uniform in $$r$$ and $$s$$, the same pair of ingredient sets generates a compatible family of bases across the entire graded collection of tensor spaces

rather than requiring a separate, unrelated basis to be constructed independently for each type.

The Two Structural Guarantees

Spanning

At every type $$\left(r,s\right)$$, the assembled set spans $$T_s^r\left(V\right)$$, since multilinearity allows any tensor to be reconstructed from its values on inputs drawn from the basis vector set and dual basis covector set.

Independence

At every type, the assembled set is also linearly independent, since the duality relation permits any single coefficient in a vanishing combination to be isolated by evaluation, forcing it to vanish. Spanning and independence together confirm that the tensor basis vector structure produces a genuine basis at each type, of dimension $$n^{r+s}$$.

Indexing and Bookkeeping

Multi-Index Labeling

Each basis tensor in the structure is labeled by a multi-index, one contravariant slot per upper index and one covariant slot per lower index, and this labeling is what allows the components of an arbitrary tensor, once expanded in the induced basis, to be referenced individually and manipulated using the Einstein summation convention.

Upper and Lower Placement as Structural Information

The placement of an index as upper or lower is not cosmetic within this structure; it records, directly and unambiguously, whether the corresponding factor was drawn from the basis vector set or the dual basis covector set, and this same placement governs how that factor transforms under a change of the underlying basis.

Coordinate Assignment as the Structure's Output

From Structure to Computation

The tensor basis vector structure is what makes the coordinate assignment map, sending each tensor to its array of components, well-defined and computable: without a confirmed basis at the relevant type, there would be no coordinate system relative to which components could be recorded in the first place.

One Structure, Many Coordinate Systems

Because the structure depends on the initial choice of basis of $$V$$, a different choice produces a different, but equally valid, instance of the same structural pattern, related to the first by the transition matrix and its inverse acting index by index.

Structural Role Within Tensor Algebra

Foundation for Higher Operations

The tensor basis vector structure is the foundation on which explicit computation of tensor products, contractions, and symmetrization operations rests, since each of these operations, when expressed in coordinates, is defined by its effect on the labeled basis tensors of the structure and extended linearly to all tensors.

Basis-Independence of the Underlying Algebra

While the specific basis tensors making up the structure depend on the initial choice of basis, the algebraic relationships among tensor spaces of different types, closure under addition and scalar action, the grading by type, and the dimension formula, are properties of the tensor algebra itself, and the tensor basis vector structure serves only as the coordinate scaffolding through which these properties are verified and computed, not as a source of new structure beyond what the abstract tensor spaces already possess.