4.9.1 Tensor Scalar Valued Output

Tensor Scalar Valued Output is the single number, an element of the base field $F$, that results from evaluating a scalar-valued multilinear map on one specific, fully filled tuple of arguments. It is the terminal, irreducible result of a tensor evaluation in the scalar case: once produced, it carries no further internal slot or component structure, and it stands as the concrete numeric fact that the entire multilinear apparatus, domain, slots, and linearity, was built to compute.

Formal Definition

The Output as a Single Field Element

For a scalar-valued multilinear map $T : V_1 \times \cdots \times V_k \to F$ and a specific tuple $(v_1, \ldots, v_k)$, the scalar valued output is

Unlike a vector-valued or algebra-valued output, $c$ has no further internal structure to expand: it is not itself a tuple, an array, or an operator, but a bare element of the field, subject only to the ordinary arithmetic of $F$ (addition, multiplication, and, when nonzero, inversion).

Well-Definedness Given Slotwise Linearity

Because $T$ satisfies slotwise linearity in every argument, the scalar output is completely determined once the tuple $(v_1, \ldots, v_k)$ is specified: there is no ambiguity or additional choice involved in producing $c$, in contrast to, say, choosing a basis expansion, which is a separate act performed only if a component representation is desired.

Properties of the Scalar Output

Basis-Independence of the Value Itself

While the components of a tensor depend on a chosen basis, the scalar output produced by evaluating the full multilinear map on a specific, basis-independent tuple of vectors and covectors does not depend on any basis choice: the same geometric or algebraic input tuple produces the same numeric output regardless of which coordinate system is used to compute it, provided the computation correctly accounts for how components and coordinates transform together.

Additivity and Homogeneity of the Output as a Function of Any One Argument

Because $T$ is multilinear, the scalar output, viewed as a function of any single argument with the rest fixed, itself satisfies additivity and homogeneity: doubling one input vector doubles the resulting scalar output, and replacing one input by a sum of two vectors splits the output into the sum of two separately computed scalar outputs, in accordance with the slotwise linearity property underlying the map.

Degree-k Homogeneity Under Joint Scaling

Scaling every argument in the tuple simultaneously by a constant $c_0$ scales the scalar output by $c_0^{k}$, where $k$ is the arity of the map; this degree-$k$ behavior is a direct, characteristic signature of the scalar output arising from a genuinely multilinear, rather than linear, evaluation.

Role of the Scalar Output in Applications

Aggregation Into Component Arrays

Systematically evaluating $T$ on every combination of basis vectors, one per slot, produces a finite collection of scalar outputs that together constitute the tensor's full component array; each individual scalar output in this collection is one entry of that array, and the array as a whole is nothing more than an organized listing of scalar outputs across all basis combinations.

Termination Point of Contraction Chains

When a higher-type tensor is repeatedly contracted until every slot is saturated, the final result of the entire chain of contractions is itself a single scalar valued output, obtained without any further intermediate tensor remaining; this is the sense in which the scalar valued output represents a fully reduced, maximally simplified tensorial computation.

Direct Physical or Geometric Readout

In applications, the scalar valued output is frequently the immediately usable end result of a calculation: the numeric length produced by the metric tensor, the numeric work produced by a force-displacement pairing, or the numeric probability amplitude produced by a suitably normalized multilinear pairing, each obtained as the bare scalar output of the corresponding tensor evaluation.

Summary of Key Points

• A scalar valued output is the single field element produced by fully evaluating a scalar-valued multilinear map on one specific tuple of arguments.
• It carries no further internal structure, unlike vector-valued or algebra-valued outputs, and is subject only to ordinary field arithmetic.
• The value is basis-independent when the input tuple itself is basis-independent, even though the tensor's components used to compute it may depend on a chosen basis.
• Scaling all arguments simultaneously by a constant scales the scalar output by that constant raised to the arity, reflecting the map's degree-$k$ homogeneity.
• Systematic evaluation over basis tuples produces the tensor's component array, and full contraction of any tensor terminates in a single scalar valued output.