6.18 Tensor Scalar Zero Order Classification

Tensor Scalar Zero Order Classification is the categorization of a scalar as a tensor of type (0, 0), meaning it carries no contravariant indices and no covariant indices at all, making it the unique tensor of total order zero and the simplest possible member of the entire hierarchy of tensors built on a vector space. This classification establishes the base case from which every higher-order tensor is ultimately built through tensor products, and it fixes the essential property that a genuine scalar must remain completely unchanged under any change of basis, in contrast to a mere real number extracted arbitrarily from a coordinate-dependent calculation.

Defining the Zero Order Case

No Indices at All

A type (0,0) tensor T has no free indices whatsoever; it is written simply as a single number, without superscripts or subscripts, since p = 0 and q = 0 in the general notation T^{i_1 ... i_p}_{j_1 ... j_q}. Coordinate-free, a scalar is regarded as an element of the trivial one-dimensional tensor product consisting of zero factors of V and zero factors of V*, which is simply the field of scalars itself, typically the real numbers.

Only One Component

Since the component count formula n^{p+q} gives n^0 = 1 regardless of the dimension n of the underlying vector space, a type (0,0) tensor has exactly one component in every basis, and that single component is the scalar's value.

The Transformation Law for Scalars

Invariance Under Every Change of Basis

Applying the general transformation pattern for type (p, q) tensors to the case p = 0, q = 0 leaves no factors of A or B to apply at all, so the transformation law degenerates to:

A scalar's single component is therefore identical in every basis, which is precisely the defining property that separates a true scalar from a quantity that merely happens to be a single number in one particular coordinate system, such as a single component of a vector, which does change under a change of basis and is not itself a scalar.

Why This Matters for Verification

This invariance provides a simple test: to check whether a computed number is a genuine scalar invariant of some tensor construction, one recomputes the same construction in a different basis and confirms the result is unchanged; if the value differs, the quantity was not actually a type (0,0) tensor but rather a basis-dependent component masquerading as one.

How Scalars Arise from Higher-Order Tensors

Full Contraction as the Source of Scalars

The most common way a scalar arises within tensor algebra is by fully contracting a type (p, p) tensor, pairing each of its p upper indices with one of its p lower indices and summing over all of them, leaving zero free indices and hence a type (0,0) result. The trace of a type (1,1) operator, T^i_i, is the simplest nontrivial instance of this process.

Evaluation of Multilinear Forms

A type (0,q) multilinear form evaluated on a full complement of q vectors, or a type (p,0) multilinear form on V* evaluated on p covectors, likewise produces a type (0,0) scalar, since every index of the tensor is consumed by contraction against the supplied vectors or covectors, leaving none free.

Diagram of the Zero Order Position in the Hierarchy

Distinguishing Scalars from Trivial Cases in the Type Hierarchy

Scalars Versus the Zero Tensor of Higher Type

A type (0,0) scalar should not be confused with the additive identity element within a higher tensor space, such as the zero vector in V or the zero operator in V ⊗ V*; the scalar zero is the unique element of the zero order classification itself, while the zero vector and zero operator are elements of entirely different, higher-order tensor spaces that merely happen to also be called "zero."

Scalars as the Ground Ring of the Tensor Algebra

Within the full tensor algebra built from a vector space, the type (0,0) scalars form the ground ring over which every other tensor space is defined as a module, meaning scalars are used to scale tensors of every other type through ordinary scalar multiplication, and this scaling operation is what gives each space of type (p, q) tensors its structure as a vector space in the first place.