3.19 Tensor Covector Interpretation

Tensor Covector Interpretation is the collection of equivalent geometric and analytic viewpoints through which a covector, an element of the dual space V* of a vector space V, can be understood beyond its bare definition as a linear functional. These interpretations, spanning the algebraic, geometric, and analytic, give covectors concrete meaning as measuring devices, as families of parallel hyperplanes, as gradients of functions, and as the natural objects that transform under pullback along linear and smooth maps.

The Algebraic Interpretation

Covectors as Linear Functionals

At the most basic level, a covector alpha in V* is a linear map from V to the base field F, satisfying additivity and homogeneity.

This algebraic reading treats a covector as a machine that consumes vectors and produces scalars, and it is the interpretation from which all others derive, since geometry and analysis are recovered by studying the fibers and level sets of this map.

Covectors as Row Vectors

In a fixed basis, once vectors are represented as column vectors of components, a covector is represented as a row vector of components, and the pairing alpha(v) becomes ordinary matrix multiplication of the row by the column. This representation clarifies why covectors and vectors, although abstractly dual, behave differently under coordinate change: a change-of-basis matrix acting on column vectors induces the inverse-transpose action on row vectors, which is precisely the contragredient transformation rule of dual bases.

The Geometric Interpretation

Covectors as Families of Hyperplanes

A nonzero covector alpha determines a foliation of V into parallel affine hyperplanes, namely the level sets where alpha takes a constant value.

The kernel of alpha is the hyperplane H(0), a codimension-one subspace of V, and the remaining level sets are translates of this kernel. The spacing between consecutive integer-valued level sets, together with the orientation of increasing c, encodes the covector completely: doubling alpha halves the spacing between its level hyperplanes, since the same range of values is now traversed over a shorter distance in V.

Covectors as Gradients

When V is the tangent space of a manifold at a point and f is a smooth real-valued function, the differential df at that point is a covector whose geometric meaning is the direction of steepest increase of f, encoded dually rather than as an arrow. The level hyperplanes of df are precisely the tangent spaces to the level sets of f itself, so the hyperplane interpretation and the gradient interpretation coincide: the differential of a function is the covector whose kernel is tangent to the level surface of the function through that point.

The Analytic Interpretation

Covectors as Measuring Instruments

A useful physical intuition treats a covector as a measuring instrument or a stack of surfaces through which a vector, viewed as an arrow, passes; the value alpha(v) counts how many of these surfaces the arrow pierces. This intuition, common in differential geometry and physics, makes explicit the complementary roles of vectors and covectors: vectors are objects with magnitude and direction that move through space, while covectors are objects that measure such motion by counting crossings, and the pairing bracket formalizes this counting as a scalar.

Covectors as Rates of Change

Along a smooth curve gamma(t) in a manifold with velocity vector gamma'(t), applying a covector alpha to the velocity gives the instantaneous rate of change that alpha would assign to the curve's motion.

when alpha is a constant covector on a vector space, showing that covectors are naturally suited to differentiate linear quantities along paths, a fact that underlies their identification with differentials in calculus.

Interpretation via Pullback Behavior

Covectors as the Objects that Pull Back

A defining structural fact about covectors, distinguishing them from vectors, is how they transform under a map. If f is a linear map from V to W, or more generally a smooth map between manifolds, covectors on the target pull back to covectors on the source through composition.

Vectors, by contrast, are pushed forward, not pulled back, in the natural direction of the map f. This contravariant behavior is the deepest and most operationally significant interpretation of a covector: it is the type of object for which composition with a map naturally flows backward, against the direction of the map itself, in exact analogy with how a measuring instrument on the target space can be composed with a map to yield a measuring instrument on the source space, without needing the map to be invertible.

Interpretation as Members of the Cotangent Bundle

On a smooth manifold, collecting the covector interpretation at every point produces the cotangent bundle, whose sections are differential one-forms. Under this interpretation, a covector at a point is not merely an abstract dual vector but a locally defined measuring device compatible with the smooth structure, and pullback of one-forms along smooth maps is the direct globalization of the linear pullback of covectors described above, applied pointwise via the differential of the map at each point.