9.23.1 Tensor Basis Definition Boundary

The Tensor Basis Definition Boundary outlines the limits and structure of basis elements in tensor algebra, essential for understanding tensor spaces and their operations.

Tensor Basis Definition Boundary is the precise line separating what qualifies as a valid basis for tensor algebra from collections of vectors that fail to meet the requirements, marking exactly which properties a set of vectors must satisfy before tensor components can be defined relative to it.

The Two Defining Requirements

Linear Independence as a Boundary Condition

A set of vectors ({e_1, \dots, e_n}) qualifies as a basis only if no nontrivial linear combination of them equals the zero vector.

c^{i} e_{i} = 0 \Rightarrow c^{1} = c^{2} = \dots = c^{n} = 0

A set that fails this condition sits outside the definition boundary: some vector in the set can be written as a combination of the others, so the set carries redundant information and does not admit a well-defined, unique component expansion for every vector in the space.

Spanning as a Boundary Condition

The second requirement is that every vector in the space be expressible as some linear combination of the candidate set; a set that spans only a proper subspace lies outside the boundary of a valid basis for the full space, since some vectors would have no expansion at all.

V = span {e_{1}, \dots, e_{n}}

What Falls Outside the Boundary

Too Few Vectors

A set with fewer than (n) vectors, where (n) is the dimension of the space, cannot span the entire space regardless of how the vectors are chosen, and therefore always falls outside the definition boundary of a basis for that space.

Too Many Vectors

A set with more than (n) vectors in an (n)-dimensional space is automatically linearly dependent, since no more than (n) linearly independent vectors can exist in an (n)-dimensional space; such an over-complete set also falls outside the strict boundary of a basis, though it may still be useful as a frame in specialized contexts that explicitly relax the independence requirement.

Exactly n Independent, Spanning Vectors

Only a set of exactly (n) vectors that is simultaneously linearly independent and spanning sits inside the boundary; remarkably, for a finite-dimensional space, satisfying either one of the two conditions with exactly (n) vectors automatically guarantees the other, which is why checking just one condition suffices in practice once the count of vectors is confirmed to equal (n).

Uniqueness of Expansion at the Boundary

Why the Boundary Guarantees Unique Components

Precisely at the boundary defined by independence and spanning, every vector has one and only one expansion in terms of the basis, which is what allows components (v^i) to be defined unambiguously as functions of the vector and the chosen basis.

v = v^{i} e_{i}

If the set were dependent, at least one vector would have infinitely many valid expansions, making the notion of "the" components of (v) ill-defined; if the set failed to span, some vectors would have no expansion at all.

Visual Illustration

Boundary Relative to the Ambient Space

Basis of a Subspace Versus Basis of the Full Space

A set that is independent and spans a subspace correctly qualifies as a basis of that subspace, but sits outside the definition boundary for a basis of any larger ambient space containing it; the definition boundary is always relative to a specifically named vector space, and confusing the two is a frequent source of error when subspaces are involved.

Why the Boundary Must Be Stated Precisely

Fixing the exact boundary of what counts as a tensor basis is what guarantees that every subsequent construction, including dual bases, component expansions, and basis change formulas, rests on solid ground: any set of vectors admitted as a basis must guarantee unique, well-defined coordinates for every vector in the space, and this guarantee holds only within the precise boundary set by independence together with spanning.