3.5.1 Tensor Basis Covector Index Role

Tensor Basis Covector Index Role is the specific function performed by the single upper index attached to a basis covector e^i, namely serving simultaneously as a label distinguishing e^i from the other members of the dual basis and as the position that, when matched against a lower index of the same letter elsewhere in an expression, triggers summation under the Einstein convention. Understanding this dual function, labeling and summation-triggering at once, is essential to reading index notation correctly whenever a basis covector appears.

The Upper Index as a Labeling Device

Distinguishing Members of the Dual Basis

The index i on e^i ranges over 1, ..., n, and each distinct value of i picks out a different, specific member of the dual basis {e^1, ..., e^n}; in this role, the index functions exactly as a subscript or superscript would in any enumerated list, with no summation implied when i is fixed and appears only once in an expression, as in the single statement e^3(v) = v^3.

Upper Placement Signals Covariant Membership

The choice to place this labeling index in the upper position, rather than the lower position used for basis vectors e_i, is not arbitrary: it signals that e^i belongs to V* and transforms covariantly under a change of basis, in contrast with the lower-index basis vectors e_i, which belong to V and transform contravariantly relative to vector components. The upper-index placement on a basis covector is therefore already carrying variance information, in addition to serving as a mere label.

The Same Index as a Summation Trigger

Free Versus Summed Occurrences

When the index i on e^i appears exactly once in an expression, it is free, and the expression denotes one specific covector or one specific value, depending on context. When the same letter i appears a second time, as a lower index elsewhere in the same term, such as in v^i e_i or in the coordinate expansion ω = ω_i e^i, the repeated upper-lower pair triggers implicit summation under the Einstein convention, and the index is no longer labeling a single member of the dual basis but ranging over all of them within the sum.

Determining the Role From Context Alone

Because the notation for a free index and a summed index looks identical at the level of a single symbol, i, the distinction is made entirely by counting how many times the letter occurs, in which positions, within the surrounding term; a reader of tensor index notation must track occurrences across an entire expression, not just examine e^i in isolation, to know whether its index is currently playing the labeling role or the summation-triggering role.

Index Role in Coordinate Expansions

Basis Covectors as the Summed Factor

In the expansion ω = Σ_i ω_i e^i, the basis covectors e^i are the objects being summed, with the index i running over the full range 1 to n, while the coefficients ω_i supply the scalar weight for each term; the index role of e^i here is to mark which basis covector, among all n of them, a given term of the sum corresponds to, with the actual identity of e^i as a specific functional becoming relevant only once the sum is expanded.

Contrast With a Single Fixed Basis Covector

This summed role is distinct from an expression like e^2(v), in which the index 2 is a definite numeral rather than a free letter available for matching, and no summation is possible or intended; e^2 denotes one particular, fixed member of the dual basis, and the expression evaluates to a single number, the second coordinate of v, with no ambiguity about index role at all.

Consistency With General Tensor Index Conventions

The Basis Covector as the Building Block of Upper-Index Notation

Every upper index appearing in the coordinate expression of a general tensor traces back, ultimately, to a basis covector e^{i_k} occupying that position in the tensor's expansion as a sum of simple terms; the labeling-versus-summation duality described here for a single e^i is therefore the atomic instance of the same duality that governs every upper index appearing anywhere in tensor index notation, including the multi-index expressions used for coordinate transfer and contraction of general (p, q)-tensors.