8.8 Tensor Repeated Index Notation Pattern

Tensor Repeated Index Notation Pattern is the general typographic pattern, underlying both dummy index summation and its distinction from free indices, in which a given index letter appears more than once within a term of a tensor expression, and the specific way in which it repeats, in position, in variance, and in count, determines whether that pattern signals a legitimate summation, an error, or a special convention requiring additional care.

The Pattern in Its Standard Form

Two Occurrences, Opposite Variance

The canonical repeated index pattern consists of an index letter appearing exactly twice within a term, once as an upper index and once as a lower index, which is the pattern recognized by the Einstein summation convention as denoting an implicit sum.

This is the pattern most tensor notation is built around, and recognizing it reliably is the single most important skill in reading index expressions correctly.

Single Occurrence as the Complementary Pattern

The complementary pattern, a single occurrence of an index letter within a term, signals a free index rather than a repeated one, and the two patterns together, single occurrence and paired repetition, exhaust the standard cases that arise in ordinary tensor notation.

Non-Standard Repeated Patterns

Same-Variance Repetition

A pattern in which an index letter repeats twice but with the same variance, both upper or both lower, falls outside the standard summation convention and typically signals that a metric tensor is needed to properly pair the indices before summation can be justified, or that the expression contains an error.

Repetition Beyond Two Occurrences

A pattern in which an index letter appears three or more times within a single term is never covered by the standard convention and always indicates that the expression must be rewritten, typically by renaming one or more of the colliding occurrences, before it can be interpreted unambiguously.

Repeated Pattern Across Different Structural Contexts

Repetition Within a Single Symbol

A repeated index pattern can arise entirely within one tensor symbol carrying both an upper and a lower slot with the same name, producing a trace-type contraction rather than a contraction between two distinct factors.

Repetition Across Distinct Factors

More commonly, the repeated pattern spans two separate tensor factors multiplied together, with one occurrence belonging to each factor, which is the pattern underlying matrix multiplication, inner products, and general tensor contraction.

Recognizing the Pattern in Practice

Correctly parsing any tensor expression begins with scanning each term for repeated index letters and classifying the pattern found: only the two-occurrence, opposite-variance pattern licenses an implicit sum under the standard convention, while every other repetition pattern signals that the expression either requires an additional operation, such as raising or lowering an index with the metric, or requires outright correction through renaming before it can be considered well formed.