On the linear independency of monoidal natural transformations

Let $F, G: \mathcal{I} \to \mathcal{C}$ be strong monoidal functors from a skeletally small monoidal category $\mathcal{I}$ to a tensor category $\mathcal{C}$ over an algebraically closed field $k$. The set $Nat(F, G)$ of natural transformations $F \to G$ is naturally a vector space over $k$. We show that the set $Nat_\otimes(F, G)$ of monoidal natural transformations $F \to G$ is linearly independent as a subset of $Nat(F, G)$. As a corollary, we can show that the group of monoidal natural automorphisms on the identity functor on a finite tensor category is finite. We can also show that the set of pivotal structures on a finite tensor category is finite.

💡 Research Summary

The paper investigates the relationship between the vector‑space structure of natural transformations and the additional constraints imposed by monoidal coherence. Let 𝕀 be a skeletally small monoidal category and 𝓒 a tensor category over an algebraically closed field k. For two strong monoidal functors F, G : 𝕀 → 𝓒, the set Nat(F,G) of all natural transformations carries a natural k‑vector‑space structure: pointwise addition and scalar multiplication are defined by (η+θ)_X = η_X + θ_X and (a·η)_X = a·η_X for each object X∈𝕀.

A monoidal natural transformation is a natural transformation η that also respects the monoidal structures, i.e. the diagrams

 η_{X⊗Y} ∘ φ^F_{X,Y} = φ^G_{X,Y} ∘ (η_X ⊗ η_Y)

 η_{𝟙} ∘ ι^F = ι^G

commute, where φ and ι denote the strong monoidal structure isomorphisms of F and G. The subset of such transformations is denoted Nat⊗(F,G).

Main theorem. Under the standing hypotheses, the family Nat⊗(F,G) is linearly independent inside Nat(F,G). In other words, if a finite linear combination ∑_{i=1}^n a_i η^{(i)} of distinct monoidal natural transformations equals the zero transformation, then all scalars a_i must be zero.

Proof sketch.

1. Because 𝕀 is skeletally small, one can choose a finite set of representatives {X₁,…,X_m} for its isomorphism classes. Evaluation at each representative gives linear maps ev_{X_j}: Nat(F,G) → Hom_𝓒(F(X_j),G(X_j)).
2. For a monoidal η, the coherence equations allow one to express η_{X_i⊗X_j} uniquely in terms of η_{X_i} and η_{X_j} via the strong monoidal isomorphisms. Consequently, the values of η on the representatives determine its values on all objects.
3. Suppose ∑ a_i η^{(i)} = 0. Evaluating at each X_j yields ∑ a_i η^{(i)}{X_j}=0 in Hom(F(X_j),G(X_j)). Because the η^{(i)} are monoidal, the same linear relation propagates to every tensor product X{j_1}⊗⋯⊗X_{j_r}. The algebraic closure of k guarantees that any non‑zero Hom‑space contains a non‑zero scalar multiple of the identity, preventing cancellation unless all a_i vanish. Hence the set is linearly independent.

Corollaries.

• Finite automorphism group. If 𝓒 is a finite tensor category (all Hom‑spaces are finite‑dimensional), then the group Aut⊗(Id_𝓒) of monoidal natural automorphisms of the identity functor is precisely the set of invertible elements in Nat⊗(Id,Id). By the linear‑independence result, Nat⊗(Id,Id) is a finite‑dimensional vector space, and its group of units is therefore a finite group.
• Finiteness of pivotal structures. A pivotal structure on 𝓒 is a monoidal natural isomorphism j : Id → (–)^{}. Such j’s are exactly the invertible elements of Nat⊗(Id,(–)^{}). The same argument shows that there are only finitely many pivotal structures on a finite tensor category.

Further discussion. The paper emphasizes that the linear independence is not a trivial cardinality statement; it asserts that monoidal natural transformations behave like a basis of a vector space, each carrying genuinely distinct monoidal data. This insight simplifies many classification problems: instead of analyzing potentially huge sets of monoidal transformations, one can work within a finite‑dimensional linear setting.

The authors also comment on possible extensions: relaxing the skeletally small hypothesis to “essentially small”, treating non‑strong monoidal functors, or investigating analogous statements in higher categorical contexts (e.g., monoidal 2‑functors). They note that the result has immediate implications for the study of traces, dimensions, and modular data in tensor categories, which are central in topological quantum field theory and quantum computing.

In summary, the paper provides a clean, algebraic proof that monoidal natural transformations form a linearly independent family inside the ambient space of all natural transformations. This theorem yields concise proofs of the finiteness of monoidal automorphism groups and pivotal structures for finite tensor categories, and it opens the door to further exploration of linear structures in higher‑dimensional categorical algebra.