7.18.5 Tensor Component Type Dependence

Tensor Component Type Dependence is the fact that although the Tensor Component Total Entry Calculation depends only on the rank of a tensor and the dimension of the underlying space, the actual number of independent components a tensor possesses, and the way its components are organized, depend additionally on how many of its indices are contravariant, how many are covariant, and how those indices are grouped for the purpose of applying symmetry or antisymmetry relationships.

Why Type Enters the Enumeration

Variance Type Does Not Affect the Raw Count

The Tensor Component Total Entry Calculation, expressed as n raised to the power r, treats every index identically regardless of whether it is contravariant or covariant, since each index independently contributes the same Tensor Component Index Range no matter its type. A rank-two tensor with two contravariant indices, a rank-two tensor with two covariant indices, and a rank-two tensor with one of each type all share the identical raw entry count of n squared.

Variance Type Restricts Which Indices Can Be Paired for Symmetry

Although the raw count is unaffected by variance type, the Tensor Component Symmetry Pattern that a tensor may exhibit is directly constrained by it, since a Tensor Component Symmetric Index Pair or a Tensor Component Antisymmetric Index Pair can only be formed between two indices of matching variance type. A mixed tensor with one contravariant and one covariant index cannot exhibit either symmetric or antisymmetric behavior between those two indices, because exchanging indices of different variance type is not a meaningful operation within the tensor's transformation law.

Illustration

Each of the three tensors depicted shares the same rank and the same raw entry count, yet only the first two, whose indices match in variance type, can be tested for a symmetric or antisymmetric relationship between those two indices.

Consequences for Enumeration

Independent Component Counts Diverge by Type

Because symmetry and antisymmetry reductions apply only between indices of matching variance type, two tensors of identical rank and dimension can have different counts of independent components depending solely on how their indices are distributed between contravariant and covariant type. A fully contravariant rank-two tensor known to be symmetric has n times the quantity n plus one, divided by two, independent components, while a mixed rank-two tensor with one contravariant and one covariant index generally has the full n squared independent components, since no symmetry reduction applies to a pair of mismatched type.

Grouping of Indices Matters for Higher Rank Tensors

For tensors of rank three or higher, the way indices are grouped by type determines which pairs are even eligible to be tested for symmetric or antisymmetric behavior. A rank-four tensor with two contravariant and two covariant indices may exhibit symmetry within its contravariant pair, within its covariant pair, or in both simultaneously, but never across a contravariant and a covariant index taken together.

Persistence of Type Across Coordinate Change

The distinction between contravariant and covariant indices, and therefore the eligibility of any two indices to form a symmetric or antisymmetric pair, is itself unaffected by a change of coordinates, since the transformation law assigns a fixed variance type to each index independent of the coordinate system chosen. A pairing of indices that is ineligible for symmetry testing due to mismatched type in one coordinate system remains ineligible in every other coordinate system as well.

Relationship to Other Tensor Concepts

Tensor Component Type Dependence refines the Tensor Component Total Entry Calculation by identifying which reductions from the Tensor Component Symmetric Reduction or the Tensor Component Antisymmetric Reduction are actually applicable to a given tensor, based on the variance type of its indices. It is a necessary consideration within the broader process of Tensor Component Enumeration whenever a tensor mixes contravariant and covariant indices rather than consisting entirely of one type or the other.