2.3 Tensor Underlying Vector Space Structure

Tensor Underlying Vector Space Structure is the complete package of interrelated components, the vector space V, its field of scalars F, its dimension n, a chosen basis and the corresponding dual basis on V*, considered together as the single structural foundation that every tensor in a given construction is built from and expressed relative to. Rather than treating the vector space, the field, and a basis as separate, loosely related prerequisites, this structure emphasizes how they function jointly, as one coordinated package whose parts constrain and determine one another.

The Components of the Structure

The Vector Space and Its Field Together

V and F are not independent choices made in isolation, since V is by definition a vector space over F; specifying V without specifying F leaves the structure incomplete, and the same set of vectors can in principle support scalar multiplication by more than one field, producing genuinely different vector space structures on what appears to be the same underlying set.

Dimension as a Consequence, Not an Independent Input

The dimension n is not chosen independently of V and F, but is instead a derived property, the size of any basis of V, so that once V and F are fixed, n follows automatically rather than being supplied as a separate, freely chosen input.

Basis and Dual Basis as a Linked Pair

A chosen basis of V and the corresponding dual basis of V* are not two independent choices either; fixing a basis of V automatically determines the dual basis through the defining pairing condition, so the two must always be introduced, and changed, together.

Why the Components Must Be Treated as a Single Package

Changing One Component Affects the Others

Because the components of this structure are linked, changing the field F while holding the underlying set fixed generally changes the very notion of dimension and linear independence available on V, and choosing a new basis automatically forces a corresponding change in the dual basis, so no single component can be varied in complete isolation from the rest.

Why This Matters for Every Tensor Built From the Structure

Every tensor of every type (p, q) in the construction is expressed relative to this entire package at once, its component count depends on n, its arithmetic depends on F, and its specific numerical components depend on the linked basis and dual basis, so a tensor's full description implicitly carries a dependence on all four components simultaneously.

Fixing the Structure Once for an Entire Construction

One Structure Shared Across All Tensor Types

A single choice of this underlying structure, V, F, a basis, and the induced dual basis, is fixed once and then shared by every tensor type built from it, vectors, covectors, and every mixed type in between, rather than being re-chosen independently for each type as it is introduced.

Consistency Requirements When Combining Tensors

Because tensors intended to be combined, through contraction or linear combination, must be built from the same underlying structure, confirming that two tensors share the identical V, F, and basis convention is a prerequisite check before any operation between them is attempted, extending the scalar compatibility requirement to the full structural package rather than the field alone.

Distinguishing the Structure From Any Particular Representation

The Structure Versus a Specific Set of Components

The underlying structure itself is what makes component representations possible at all, but the structure is not the same as any one specific component array; the same structure supports infinitely many different tensors, each with its own components relative to the fixed basis the structure provides.

The Structure Versus a Change of Basis

A change of basis alters only the basis and dual-basis part of the structure, leaving V, F, and n unchanged, which is precisely why the transformation law connects two different component representations of the same tensor rather than producing an entirely new, unrelated tensor space.

Why Naming This Structure Explicitly Is Useful

A Single Reference Point for Later Foundations

Later material on fluency, problem-solving, and error patterns routinely presupposes that this entire structure, not merely the bare set V, has already been fixed, so naming it explicitly as a single package provides a compact reference point rather than requiring each of its four linked components to be restated individually every time.

Preventing Partial or Inconsistent Assumptions

Treating the structure as one coordinated package, rather than four separate and only loosely related choices, guards against inconsistent assumptions, such as silently switching the field partway through a derivation while assuming the dimension and basis remain unaffected, when in fact the entire structure must be reconsidered together whenever any one of its components changes.