8.16.5 Tensor Collision Resolution Rule

Tensor Collision Resolution Rule is the prescribed procedure for correcting any instance of an index collision in a tensor expression, consisting of identifying which occurrences of a letter constitute the excess or conflict, selecting a replacement letter that is absent from the entire relevant scope, and substituting that replacement into precisely the occurrences that must change while leaving every other index in the expression untouched. It is the operational counterpart to the various collision patterns: where those patterns describe what has gone wrong, the resolution rule specifies exactly what must be done to put it right.

The Procedure in Order

Step One: Isolate the Colliding Term

Because index collisions and the scopes governing dummy indices are both confined to individual terms, resolution begins by isolating the specific term in which the colliding letter appears, since corrections made elsewhere in a larger expression are irrelevant to a collision contained within one term. In a multi-term equation, only the offending term needs to be examined and modified.

Step Two: Classify Every Occurrence of the Colliding Letter

Within the isolated term, every occurrence of the colliding letter must be classified as either part of an intended free index or part of one of the dummy pairs it has become tangled with. This classification determines which occurrences are eligible for renaming — free index protection specifies that legitimately free occurrences must not be touched — and which occurrences constitute the dummy component of the collision that renaming is meant to fix.

Step Three: Select a Replacement Letter Free of Collisions

Once the dummy occurrences requiring renaming have been identified, a replacement letter is chosen according to the requirements of collision avoidance: it must not coincide with any free index in the term, nor with any other dummy index already present, nor with the letter being replaced. Any letter satisfying these conditions is an equally valid choice, since the specific replacement letter carries no meaning of its own.

Step Four: Substitute Consistently Within the Correct Scope

The chosen replacement letter is then substituted into every occurrence of the dummy pair being renamed — both the upper and the lower occurrence — while every occurrence outside that pair, including any legitimately free occurrence of the original letter, is left exactly as it was. This substitution must be applied in full, per scope preservation, rather than to only one of the two occurrences forming the dummy pair.

Worked Application

Resolving a Free–Dummy Collision

Given the colliding term $A^{i}B_{i}C_{i}$, the procedure first isolates the term (it is the entire expression here), classifies the three occurrences of $i$ — one as the intended free index attached to $C$, two as the dummy pair between $A$ and $B$ — selects an unused letter such as $k$, and substitutes it into the dummy pair alone, yielding the resolved expression $A^{k}B_{k}C_{i}$, in which $i$ remains free and untouched while $k$ now unambiguously marks the contracted pair.

Resolving a Dummy–Dummy Name Reuse

Given the colliding product $A^{i}B_{i}C^{i}D_{i}$, formed by combining two originally separate sums, the procedure classifies all four occurrences as belonging to two distinct, previously independent dummy pairs, selects a replacement letter such as $j$ for one of the two pairs, and substitutes it into that pair only, yielding $A^{i}B_{i}C^{j}D_{j}$, restoring two clearly separated summations.

Guarantees Provided by Following the Rule

The Resolved Expression Is Equivalent to the Intended One

When the resolution rule is followed correctly — respecting scope, avoiding new collisions, and protecting genuinely free indices — the resulting expression is guaranteed, by the same reasoning that underlies dummy index renaming equivalence, to represent exactly the value that was originally intended before the collision arose, since only the arbitrary label of a bound variable has been changed. The rule does not merely produce some well-formed expression; it produces the well-formed expression corresponding to the correct, disambiguated interpretation.

The Rule Never Introduces a New Collision

Because step three explicitly requires checking the replacement letter against every other index already present in the term, correctly applying the resolution rule cannot itself generate a fresh collision; the procedure is closed under its own conditions, meaning it can be applied repeatedly, if multiple independent collisions exist within a larger expression, without any risk of the fixes interfering with one another as long as each is checked against the current, evolving state of the expression.

When the Rule Cannot Be Applied Directly

Collisions Involving Only Free Indices

If a collision arises because two indices that were both meant to remain free have been given the same letter — rather than one free and one dummy occurrence — the resolution rule as stated for dummy indices does not directly apply, since free indices cannot be renamed locally within a single term without a corresponding change everywhere else that index appears in the governing equation; resolving this variety of collision instead requires re-examining the original derivation to determine what distinct letters the two indices should have carried from the outset.

Ambiguity That Predates the Notation

If the collision reflects a genuine error in the underlying mathematics — for instance, an intended tensor contraction that does not actually correspond to any valid pairing of upper and lower slots — no renaming operation can resolve it, since the resolution rule only repairs notational collisions between otherwise correct pairings; correcting a substantive mathematical error requires revisiting the derivation itself, not merely relabeling its indices.

Role Within the Index Collision Pattern

The collision resolution rule is the constructive procedure that closes the loop opened by the various collision patterns and the ambiguity they cause: having identified that a collision exists and characterized its type, the resolution rule supplies the concrete, checkable steps — isolate, classify, select, substitute — that convert an ambiguous or ill-formed expression back into one with a single, well-defined meaning, fully consistent with the value the notation was originally meant to represent.