9.23.3 Tensor Component Theory Boundary

Tensor Component Theory Boundary is the delimiting condition marking where the theory of tensor components, built on fixed-rank arrays satisfying strict transformation laws, ceases to apply and gives way to related but distinct structures such as tensor densities, non-tensorial arrays, and objects requiring additional or weaker transformation behavior.

What Counts as a Genuine Tensor Component

The Transformation Law as the Defining Test

An array of numbers indexed by a basis qualifies as the components of a genuine tensor only if it transforms under a change of basis exactly according to the standard tensor transformation law, with one factor of the change-of-basis matrix or its inverse for every index.

Any array failing to satisfy exactly this rule, with exactly this combination of factors, falls outside the boundary of what tensor component theory covers, no matter how similar it may look notationally.

Fixed Rank as a Boundary Condition

Tensor component theory applies only to arrays with a definite, fixed number of contravariant and covariant indices; objects whose effective rank changes from basis to basis, or that mix tensorial and non-tensorial behavior across their indices, sit outside the boundary of standard tensor component theory.

Tensor Densities as a Boundary Case

The Extra Jacobian Factor

A tensor density transforms like an ordinary tensor except for an extra power of the Jacobian determinant of the change-of-basis matrix, placing it just outside the strict boundary of tensor component theory while remaining closely related to it.

The exponent (w), called the weight, is zero exactly for ordinary tensors; any nonzero weight places the object across the boundary into the separate but related theory of tensor densities, which requires its own transformation bookkeeping.

Symbols That Look Like Tensors But Are Not

The Kronecker Delta as a Boundary Exception

The Kronecker delta (\delta^i_j) happens to have numerically identical components in every basis, which might suggest it lies outside ordinary tensor behavior, yet it is in fact a genuine mixed tensor because its invariance is a consequence of, not an exception to, the standard transformation law; this makes it a useful test case for the boundary, since verifying it satisfies the transformation rule despite unchanging components confirms the rule rather than contradicting it.

The Levi-Civita Symbol as a True Boundary Case

The Levi-Civita symbol, by contrast, does not transform as a tensor under general changes of basis; it picks up an extra factor of the determinant sign or magnitude, placing it outside the tensor component boundary and correctly classifying it instead as a tensor density or pseudotensor, depending on convention.

Coordinate Patch Limits on Component Theory

Where Components Cease to Be Defined

Tensor component theory presupposes a valid, non-degenerate basis at the point in question; at coordinate singularities where the basis vectors fail to be linearly independent, the theory of tensor components simply does not apply, and statements about components there require switching to a regular coordinate system covering that region.

Visual Illustration

Why This Boundary Must Be Drawn Precisely

Marking the exact boundary of tensor component theory prevents the accidental misapplication of tensor identities, contraction rules, or invariance arguments to arrays that merely resemble tensors in index notation but do not actually satisfy the transformation law. Objects on the far side of the boundary, such as tensor densities or connection coefficients, still play essential roles in the broader subject, but they require their own separate transformation rules precisely because they fall outside this boundary.