8.23.5 Tensor Index Notation Error Boundary

Tensor Index Notation Error Boundary is the line separating a legitimate boundary case of index notation — one that is merely unusual and requires extra structure or care to handle correctly — from an outright notational error, in which an expression violates the basic syntactic rules of the notation itself and is therefore meaningless as written, regardless of any additional structure that might be supplied. It is the diagnostic distinction between "this expression needs something more before it is well-defined" and "this expression is simply malformed," and correctly locating which side of this line a given piece of notation falls on is what determines whether the fix is to extend the calculation or to rewrite the expression outright.

The Triple-Repetition Error

More Than Two Occurrences of a Single Letter

The Einstein summation convention is defined for an index letter occurring exactly twice, once upper and once lower; an expression in which the same letter appears three or more times within a single term, such as A^i B_i C_i, does not correspond to any standard interpretation under the convention and is a genuine notational error rather than a boundary case requiring extension. Unlike a same-variance repeated pair, which can be legitimately repaired by introducing the metric, a triple repetition has no standard corrective structure to supply — it must simply be rewritten, typically by introducing a second, distinct index letter for one of the three occurrences.

Distinguishing Triple Repetition From a Legitimate Chain of Contractions

A genuine chain of contractions, such as A^i_{\ j}B^j_{\ k}C^k, uses each letter exactly twice, even though three factors and three letters are involved; this is not an error, since every letter individually satisfies the standard repetition rule. The error boundary is crossed only when one single letter itself is reused a third time, not merely when several different contracted pairs appear within one expression.

The Free/Dummy Collision Error

Reusing a Free Index Letter as a Dummy Index

An expression in which a letter already serving as a free index on one side of an equation is separately reused as a dummy (summed) index elsewhere in the same expression creates a genuine ambiguity about whether that occurrence is meant to vary freely or to be summed away; unlike a legitimate same-variance repair or a chain of distinct contractions, this collision cannot be resolved by supplying additional structure and must instead be corrected purely by renaming, choosing an unused letter for the dummy index so the collision no longer occurs.

Why This Sits Firmly on the Error Side

Because dummy index letters are entirely arbitrary placeholders with no meaning beyond marking a summation, and because renaming them never changes the value of an expression, there is no legitimate mathematical content that a free/dummy collision could be preserving; it is purely a labeling accident, which is exactly why it counts as a notational error to be fixed by renaming rather than a boundary case reflecting some genuine subtlety in the underlying mathematics.

The Unbalanced Equation Error

Free Index Sets That Fail to Match

An indexed equation whose two sides have different free index sets — a stray free index on one side with no counterpart on the other, or matching letters with mismatched variance — is a notational error in the strict sense that the expression, as written, cannot be interpreted as a coherent claim about tensors at all; this differs from a boundary case because there is no additional structure (metric, connection, transition map) that could repair a genuine free-index mismatch, since the mismatch reflects a comparison between two objects of different tensorial type rather than a well-defined comparison lacking one extra ingredient.

Diagram Separating Error Cases From Boundary Cases

Why the Distinction Matters in Practice

Different Corrective Actions Follow From Each Diagnosis

Misdiagnosing a genuine notational error as a boundary case leads to a fruitless search for some additional mathematical structure to supply, when the actual fix is simply to rewrite the offending expression correctly; conversely, misdiagnosing a legitimate boundary case as an outright error can lead to abandoning or drastically rewriting an expression that was in fact salvageable once the appropriate extra structure — a metric, a connection, an atlas of charts — was introduced.

The Error Boundary as a First-Pass Filter

Because notational errors are, by construction, repairable by inspection alone (renaming a letter, splitting a triple repetition, adding a missing free index), checking an expression against the error boundary is properly done before any deeper investigation into whether a subtler boundary case is at play; only once an expression is confirmed free of outright notational errors does it make sense to ask whether it instead sits at one of the more substantive boundaries — of contraction, of definition, of transformation, or of component validity.