Natural Transformations

Morphisms between functors and the naturality square

Natural Transformations

If functors are maps between categories, natural transformations are maps between functors. Saunders Mac Lane, co-inventor of category theory, said: "I didn't invent categories to study functors; I invented them to study natural transformations."

Naturality Square

A natural transformation α : F ⇒ G assigns to each object A a morphism α_A : F(A) → G(A) such that the naturality square commutes. Watch both paths arrive at the same place!

F(f)G(f)α_Aα_BF(A)F(B)G(A)G(B)
Path 1: F(f) then α_BPath 2: α_A then G(f)

Key insight: The naturality condition says: α_B ∘ F(f) = G(f) ∘ α_A for every morphism f. Both paths through the square give the same result — that's what makes the transformation "natural" (not dependent on arbitrary choices).

Determinant as Natural Transformation

The determinant det : GL_n(k) → k* is a natural transformation! The naturality condition says det(AB) = det(A) · det(B). Verify it with concrete matrices below.

Path 1: Multiply first, then take det
A
[ 2   1 ]
[ 0   3 ]
×
B
[ 1   0 ]
[ 1   1 ]
=
A·B
[ 3   1 ]
[ 3   3 ]
det(A·B)
6
Path 2: Take dets first, then multiply
det(A)
6
×
det(B)
1
=
6
det(A·B) = 6 = 6 = det(A)·det(B) — Naturality holds!

Key insight: The determinant is "natural" because it doesn't depend on a choice of basis — it works uniformly across all dimensions. This is a concrete example of the naturality square commuting: the two paths always give the same answer.

Functor Category [C, D]

Functors themselves form a category! In the functor category [C, D], objects are functors F, G, H : C → D and morphisms are natural transformations between them.

αββ∘αFA↦X, B↦YGA↦Y, B↦ZHA↦Z, B↦Z
α: α_A : F(A)→G(A) = X→Y, α_B : F(B)→G(B) = Y→Z

Key insight: Natural transformations compose component-wise: (β∘α)_A = β_A ∘ α_A. This gives us a whole new category whose "objects" are functors! This is one of the most powerful ideas in category theory.

Gallery of Natural Transformations

Natural transformations appear throughout mathematics. Here are some famous examples that show why "naturality" is such a powerful concept.

Double Dual

IdV ↦ V**

The canonical embedding of a vector space into its double dual. Natural because it doesn't require choosing a basis.

Components: η_V : V → V** sends v to the evaluation functional "evaluate at v"

Key insight: The common thread: all these transformations work "uniformly" — they don't depend on arbitrary choices. Eilenberg and Mac Lane invented category theory precisely to make the word "natural" mathematically precise!

Key Takeaways

  • Naturality Square — α_B ∘ F(f) = G(f) ∘ α_A must hold for every morphism f.
  • Determinant Example — det(AB) = det(A)·det(B) is a naturality condition.
  • Functor Categories — Functors and natural transformations form a new category [C, D].
  • Famous Examples — Double dual, Hurewicz, unit/counit — all are natural.