15.13 Tensor Symmetric Rank Structure

Tensor Symmetric Rank Structure is the overall framework describing how a totally symmetric tensor of rank n can be decomposed into a sum of pure power terms, and how the smallest number of such terms needed for a given tensor defines an invariant, the symmetric rank, that measures the tensor's structural complexity independent of any chosen basis. This structure draws together the notion of a pure power term as the atomic building block, the minimality condition that turns a mere term count into a well-defined invariant, and the decomposition count as the concrete numerical value under study, presenting them as three interlocking facets of a single coherent theory.

The symmetric rank structure occupies a distinct place within the broader study of symmetric tensors: where the independent component structure and the basis structure address how a symmetric tensor is represented and stored using a fixed, chosen basis, the rank structure instead asks a basis-independent question, namely how efficiently the tensor can be expressed using the most economical possible building blocks, pure powers of arbitrary vectors rather than combinations restricted to a fixed set of basis directions.

The Building Blocks and the Decomposition Problem

Pure Power Terms as Atoms

Every decomposition considered in this structure is built from pure power terms, scalar multiples of a single vector's n-th symmetric power, as detailed in the discussion of the pure power term; these terms are the smallest possible nonzero pieces from which a general symmetric tensor can be assembled, corresponding under the polynomial analogy to perfect n-th powers of individual linear forms.

The Decomposition Problem Itself

Given a symmetric tensor T, the central problem addressed by this structure is finding the smallest collection of pure power terms whose sum equals T exactly; this problem is nontrivial precisely because a generic symmetric tensor's associated polynomial does not factor into a small number of perfect powers, forcing a genuine sum of several distinct terms rather than a single one.

Minimality as the Organizing Principle

Distinguishing Rank From Mere Decomposability

As established through the minimality condition, simply exhibiting some decomposition of T into pure power terms says little on its own, since padding or splitting terms can always increase the count arbitrarily; it is the requirement of minimality, taking the smallest achievable count over every possible decomposition, that elevates the term count into the well-defined invariant called the symmetric rank.

Rank as a Basis-Independent Measure of Complexity

Because the vectors appearing in a decomposition are drawn from the entire vector space rather than restricted to a fixed basis, the resulting rank invariant does not depend on any particular choice of coordinates, distinguishing it sharply from component-counting invariants tied to a specific basis, such as the dimension of the independent component selection.

Concrete Values and Bounds

The Decomposition Count as a Computable Quantity

The actual numerical value of the minimal term count, examined as the decomposition count, is bounded above by the dimension of the graded piece Sym^n(V) and bounded below by matrix-flattening arguments, with the generic value typically approximated by comparing the dimension of the symmetric tensor space to the dimension spanned by pure power terms, though exact values require case-by-case analysis for rank three and higher.

Extremes of the Rank Spectrum

At one extreme, a tensor expressible as a single pure power term achieves the minimum possible symmetric rank of one; at the other extreme, tensors requiring a decomposition count approaching the full dimension of Sym^n(V) represent the most structurally complex symmetric tensors of their given rank and dimension, with most tensors falling at or near the generic value between these two extremes.

Relation to the Broader Symmetric Tensor Framework

Complementary to the Basis and Component Structures

While the independent symmetric component structure and the symmetric basis structure describe how to represent a symmetric tensor completely and unambiguously using a fixed reference frame, the symmetric rank structure instead describes how to represent the same tensor as economically as possible without regard to any fixed frame, offering a genuinely different, complexity-oriented lens on the same underlying objects.

A More Difficult and Less Fully Resolved Theory

Unlike the dimension and basis questions resolved completely by the count relation and the explicit basis construction, the symmetric rank structure remains, for rank three and above, an area where exact values and general determination procedures are considerably harder to obtain, marking it as the point in the theory of symmetric tensors where the transition from fully computable, matrix-like behavior to genuinely higher-complexity, less completely understood behavior becomes most apparent.