2.20.1 Tensor Linear Functional Context

Tensor Linear Functional Context is the specific treatment of an individual linear functional as the atomic covector object that a tensor's covariant slots accept as input, focusing on the internal structure, kernel geometry, and representation properties of a single map f : V → F, as distinct from the dual space V* taken as a whole. Where the dual vector space context establishes V* as a space and describes how it pairs with V, the linear functional context zooms in on one element of that space at a time, examining what a functional is, how it acts, what its kernel looks like, and how it is represented concretely once a basis is fixed.

The Linear Functional as a Map

Formal Definition

A linear functional on a vector space V over a field F is a function f : V → F satisfying additivity and homogeneity:

for all vectors u, v in V and all scalars α in F. A linear functional is thus a linear map whose codomain is one-dimensional, the field itself, which is what distinguishes it from a general linear transformation between two vector spaces of arbitrary dimension.

Functionals as Rank-One Covariant Objects

In the tensor context, an individual linear functional is treated as the elementary building block of covariant behavior: it is a type (0, 1) tensor. Every covariant index attached to a larger tensor can, when all other indices are held fixed, be regarded as producing a linear functional acting on the remaining free vector argument. This makes the linear functional the unit from which covariant tensor slots are conceptually assembled, one index at a time.

The Kernel of a Linear Functional

Kernel as a Hyperplane

The kernel of a nonzero linear functional f, the set of vectors v such that f(v) = 0, is always a hyperplane through the origin, a subspace of codimension exactly one in V. This is a direct consequence of the rank-nullity theorem applied to a map with one-dimensional image: if f is not the zero functional, its image is all of F, so the kernel has dimension dim(V) - 1.

Level Sets and Geometric Interpretation

The full family of level sets {v : f(v) = c} for varying scalars c forms a stack of parallel hyperplanes, all translates of the kernel. Two nonzero functionals f and g have the same kernel if and only if one is a nonzero scalar multiple of the other, so a functional's kernel determines the functional up to overall scale. This geometric picture is why covectors are often visualized as families of parallel planes rather than as arrows: applying a covector to a vector counts how many of these planes the vector crosses.

Representation of Functionals

Component Representation

Once a basis e_1, ..., e_n of V is fixed, every linear functional f is uniquely determined by the n scalars f_i = f(e_i), so that for an arbitrary vector v = v^i e_i, linearity gives

using the Einstein summation convention. This expresses the abstract action of a functional as a concrete contraction between the functional's components and the vector's components, the prototypical example of an index contraction in tensor algebra.

Row Vector Analogy in Finite Dimensions

In a finite-dimensional setting with a fixed basis, a linear functional behaves exactly like a row vector acting on a column vector by matrix multiplication: the tuple (f_1, ..., f_n) multiplies the tuple (v^1, ..., v^n)^T to produce the scalar f(v). This concrete picture underlies why lower-index quantities are so often drawn or coded as row arrays, in contrast to the column-array convention for upper-index vector components.

Basis Dependence of Components, Basis Independence of the Functional

The scalars f_i change when the basis changes, according to the contragredient transformation law inherited from the dual space, but the functional f itself, as an abstract map, does not change. The linear functional context is precisely the discipline of treating f as prior to any coordinate description while still being able to compute with its components whenever a basis is chosen.

Operations on Individual Functionals

Linear Combinations of Functionals

Because the set of linear functionals on V forms a vector space, V*, any two functionals f and g can be combined into a new functional αf + βg, defined pointwise by (αf + βg)(v) = αf(v) + βg(v). The kernel of a nontrivial linear combination is generally different from the kernels of f and g individually, except in special alignment cases.

Pullback of a Functional Under a Linear Map

Given a linear map T : W → V, a functional f on V induces a functional T*f on W by composition, (T*f)(w) = f(T(w)). This pullback operation, restricted to a single functional, is the elementary case of the more general pullback of covariant tensors along linear maps, and it shows why covariant objects transform "backward" relative to the direction of a map between spaces, while vectors transform "forward" by pushforward.

Annihilator of a Single Functional

The annihilator of a subset S of V is the set of functionals vanishing on all of S. For a single vector v, the annihilator {f ∈ V* : f(v) = 0} is itself a hyperplane in V*, exhibiting the dual role each individual functional and each individual vector play with respect to one another under the natural pairing.

Role Within Larger Tensor Constructions

Slot-Wise Evaluation

When a general (p, q) tensor is evaluated, fixing all arguments except one covariant slot produces, in that remaining slot, exactly an individual linear functional on V. This slot-wise reduction is what allows multilinear tensors to be analyzed one functional at a time, and it is the mechanism by which properties established for a single linear functional, such as its kernel structure and component representation, extend to the covariant behavior of tensors of any rank.

Distinguishing the Functional Context from the Full Dual Space

While the dual vector space context describes V* as a whole, together with its pairing and basis relationship to V, the linear functional context isolates the behavior of one element of V* in detail, including its kernel, its representation as a contraction, and its transformation under pullback, providing the concrete computational tools that the broader dual space context uses only abstractly.