6.1.1 Tensor Order Classification Scope

Tensor Order Classification Scope is the delimitation of the subject matter that belongs to the classification of tensors by their order, covering the definition of tensor order as the number of indices required to specify a component, the naming conventions attached to each order (scalar, vector, matrix, and higher-order tensor), the criteria used to distinguish one order from another, and the boundary separating order-based classification from type-based classification (covariant versus contravariant) and from rank-based numerical classification in the sense of matrix rank. This scope statement exists to fix what is, and what is not, discussed under the heading of tensor order classification, so that adjacent topics such as tensor type, tensor rank as a numerical invariant, and tensor transformation rules are treated elsewhere and only cross-referenced here.

What Falls Inside This Scope

The Definition of Order as an Index Count

Tensor order classification is concerned exclusively with the count of indices, k, needed to write a component of a tensor as T with subscripts and superscripts, for example T_{i₁...i_p}^{j₁...j_q} where k = p + q. Anything that follows from this count — naming, hierarchy, and boundary cases — is inside scope; anything that follows from the values the indices range over, or from how components transform, is outside scope.

Naming Conventions by Order

is classified as a scalar, a tensor with no indices at all.

is classified as a vector or covector, a tensor with exactly one index.

is classified as a matrix-like tensor, representable as a two-dimensional array once a basis is fixed.

is classified generically as a higher-order tensor, with no further special names attached beyond stating the value of k directly.

The Order Hierarchy Diagram

Criteria for Assigning an Order

Order classification within this scope depends only on the total number of indices, not on the dimension of the underlying vector space, not on the values of the components, and not on whether the indices are covariant or contravariant. A 3 × 3 matrix and a 100 × 100 matrix are both order-2 tensors under this classification, and a mixed tensor with one upper and one lower index is order 2 regardless of how the two indices transform.

What Falls Outside This Scope

Type Classification (Covariant vs. Contravariant)

The split of indices into p upper (contravariant) and q lower (covariant) positions, and the transformation behavior attached to each kind of index, belongs to tensor type classification, a distinct classification axis that is only referenced here insofar as k = p + q supplies the order.

Rank as a Numerical Invariant of a Matrix

The word "rank" is deliberately excluded from this scope when it refers to the classical linear-algebra invariant, the dimension of the column space of a matrix, since that notion applies only to order-2 tensors and measures something orthogonal to order: a 4 × 4 matrix of rank 1 and a 4 × 4 matrix of rank 4 have the same order.

Transformation Rules Under Change of Basis

How components of a tensor change when the basis of the underlying vector space changes is governed by the multilinear transformation law, which is a separate topic from the mere count of indices and is treated under tensor transformation, not under order classification.

Symmetry and Antisymmetry Properties

Whether a tensor is symmetric, antisymmetric, or has no particular symmetry under index permutation is an additional structural property layered on top of a tensor of a given order, and is classified separately from order itself.

Boundary Cases Addressed Within This Scope

Zero Tensors and the Order Convention

A zero tensor of any order retains the order assigned to the space it lives in; the value of the components (all zero) never overrides the order determined by the index count, so order classification remains well-defined even in degenerate cases.

Tensors Built from Products

is within scope as a stated consequence of the index-counting definition, since the tensor product of an order-a tensor and an order-b tensor concatenates their index lists into an order-(a+b) tensor.

Contraction and Order Reduction

is within scope as the counterpart fact for contraction, which removes one upper and one lower index and so lowers the order by exactly two; the mechanics of how contraction is performed belong to a separate operational topic and are only cross-referenced here for the effect on order.

Purpose of Fixing This Scope

Preventing Overlap With Adjacent Classification Schemes

Because tensors admit several independent classification axes — order, type, rank, symmetry — fixing the scope of order classification to the single criterion of index count prevents this topic from silently absorbing material that belongs under type or rank classification, keeping each classification scheme internally consistent and non-redundant.

Providing a Stable Reference Point for Higher-Order Discussions

Every later discussion of a specific tensor order (order 0 through order n) inherits its baseline vocabulary and boundary conditions from this scope, so that terms like "higher-order tensor" or "order-2 tensor" carry a fixed, unambiguous meaning throughout the surrounding material on tensor algebras.