16.3.2 Tensor Alternating Argument Exchange Behavior

Tensor Alternating Argument Exchange Behavior describes how alternating tensors flip sign when arguments are swapped, a fundamental property in multilinear algebra.

Tensor Alternating Argument Exchange Behavior is the precise pattern by which the value of an alternating tensor changes when its input arguments are exchanged among themselves, describing not just that a sign change occurs but exactly how that sign change accumulates as multiple exchanges are performed in sequence or as arbitrary combinations.

Single Exchange Behavior

The Basic Swap Rule

Exchanging any two arguments of an alternating tensor, adjacent or not, negates its value:

T (\dots, v_{i}, \dots, v_{j}, \dots) = - T (\dots, v_{j}, \dots, v_{i}, \dots)

This holds regardless of how far apart i and j are among the argument slots; the behavior is not limited to neighboring positions.

Non-Adjacent Exchange Decomposed

An exchange of two non-adjacent arguments separated by d positions can be realized as an odd number, 2d − 1, of adjacent transpositions, and each contributes a factor of −1; the net result is still exactly one sign flip, confirming the single-exchange behavior is independent of positional distance.

Repeated Exchange Behavior

Double Exchange Returns the Original Value

Performing the same exchange twice in succession restores the original value, since two sign flips cancel:

T (\dots, v_{j}, \dots, v_{i}, \dots) \to T (\dots, v_{i}, \dots, v_{j}, \dots) = T (\dots, v_{i}, \dots, v_{j}, \dots)

Cyclic Exchange Behavior

Cycling k arguments through a k-cycle produces a sign equal to (−1)^{k−1}, since a k-cycle decomposes into k − 1 transpositions:

T (v_{2}, v_{3}, \dots, v_{k}, v_{1}) = (- 1)^{k - 1} T (v_{1}, v_{2}, \dots, v_{k})

so a 3-cycle leaves the value unchanged while a 4-cycle reverses it, a distinction that follows purely from parity rather than from the number of arguments moved.

Exchange Behavior Under Composite Rearrangements

General Permutation Response

For an arbitrary rearrangement σ applied simultaneously to all k arguments, the exchange behavior collapses to a single multiplicative factor:

T (v_{σ (1)}, \dots, v_{σ (k)}) = sgn (σ) T (v_{1}, \dots, v_{k})

This is the composite exchange behavior: no matter how complicated the rearrangement, the outcome depends only on whether σ is even or odd, never on any finer structural detail of the permutation.

Order-Independence of Composition

Because the sign function is a group homomorphism, composing two rearrangements and then exchanging is equivalent to exchanging under each rearrangement in sequence:

sgn (σ \circ τ) = sgn (σ) \cdot sgn (τ)

so the exchange behavior is consistent whether the rearrangement is applied all at once or built up from smaller steps applied one after another.

Degenerate Exchange Behavior

Exchange Involving a Repeated Argument

If two of the arguments being exchanged happen to be equal, the exchange behavior forces the tensor to equal its own negative, and hence to be zero:

T (\dots, v, \dots, v, \dots) = - T (\dots, v, \dots, v, \dots) \Rightarrow T = 0

No Exchange Available at Degree Below 2

For degree 0 and degree 1 alternating tensors, there are fewer than two argument slots, so no exchange operation exists to test; the exchange behavior is trivially satisfied by default rather than by an actual sign computation.

Diagram of Exchange Chains

Each successive exchange multiplies the running sign by −1, so tracking the parity of the number of exchanges applied is sufficient to predict the resulting sign at any point in the chain, without needing to recompute the tensor's value from scratch.