16.6 Tensor Sign Change under Swap Behavior
Tensor Sign Change under Swap Behavior describes how tensor signs flip when indices are swapped, a fundamental property in multilinear algebra and tensor symmetry.
Tensor Sign Change under Swap Behavior is the general phenomenon, observed across every alternating tensor regardless of rank, dimension, or geometric interpretation, whereby swapping any two of the tensor's arguments negates its output value — the single unifying behavior from which every other property of alternating tensors is derived.
The Behavior in Its Simplest Form
Statement for Two Arguments
For a tensor T exhibiting this behavior, swapping its first two arguments produces exactly the negative of the original value:
No scaling other than −1 is permitted; the behavior is not "the value changes" in some vague sense, but specifically that it is negated, exactly and without exception.
Extension to Any Pair of Slots
For tensors accepting more than two arguments, the same swap behavior applies independently to every pair of slots, holding all other arguments fixed:
Physical and Geometric Reading of the Behavior
Orientation Reversal
Geometrically, swap behavior models the reversal of orientation: if T(u, v) represents the signed area spanned by u then v, tracing the boundary counterclockwise, then T(v, u) represents the same shape traced clockwise, and the sign change captures exactly this reversal of traversal direction.
Physical Instance: Torque and Angular Quantities
Cross-product-derived quantities such as torque or angular momentum inherit swap behavior directly: reversing the order of the two vectors defining the moment arm and the force reverses the sign of the resulting torque, reflecting a real physical reversal of rotational sense rather than a mere bookkeeping convention.
Swap Behavior Compounding Under Multiple Applications
Two Swaps Cancel
Performing the same swap twice restores the original value, since the sign factor −1 squares to +1:
Chains of Swaps and Parity
When several swaps are chained together, the net sign is determined entirely by whether the total number of swaps is even or odd — an even chain returns the original sign, an odd chain returns the negated sign — regardless of which specific pairs were swapped along the way.
The Behavior's Boundary Case
Swapping Identical Arguments
If the two arguments being swapped happen to be equal, swap behavior demands T(v, v) = −T(v, v), forcing the tensor's value at that input to be exactly zero. This boundary case is not an exception to the swap behavior but a direct logical consequence of it, applied to a degenerate input.
No Swap Behavior Below Two Arguments
Tensors that accept fewer than two arguments — scalars and covectors — have no pair of slots to exchange, so the swap behavior is vacuously present rather than actively exhibited; there is simply nothing to test.
Swap Behavior Across Composed Operations
Determinant as an Instance
The determinant of a matrix, viewed as a function of its column vectors, exhibits swap behavior directly: exchanging any two columns negates the determinant, which is precisely why a matrix with two identical columns has determinant zero.
Wedge Product Inheritance
When two alternating tensors are combined via the wedge product, the resulting tensor inherits swap behavior automatically from its factors, since the wedge product is constructed specifically to preserve this sign-reversal property across the combined set of arguments.