Diagram Automorphisms in Representation Theory

• Diagram automorphisms are permutations of Dynkin diagram nodes that preserve edge structure and underpin the folding from simply-laced to non-simply-laced types.
• They extend to Lie and Kac–Moody algebras, inducing invariant subalgebras and twisted modules that connect combinatorial and geometric representation theory.
• Their influence spans quantum groups, quiver varieties, and spherical buildings, providing canonical bases, dualities, and symmetry in diverse algebraic contexts.

A diagram automorphism is a permutation of the nodes of a Dynkin diagram that preserves the edge structure, thereby inducing an automorphism of the associated Cartan matrix. This action extends to automorphisms of Lie algebras, Kac–Moody algebras, related quantum groups, representation categories, quiver varieties, and combinatorial and geometric structures such as generalized cluster complexes and spherical buildings. Diagram automorphisms underlie the folding of simply-laced root data to forms of non-simply-laced type, interlinking representation theory across types, and producing symmetries in a broad swath of algebraic, geometric, and physical contexts.

1. Definition and Basic Properties

Let $\Gamma$ denote a Dynkin diagram with node set $I$ and an integrally-valued Cartan matrix $A=(a_{ij})_{i,j\in I}$. A diagram automorphism is a permutation $\sigma:I\to I$ such that $a_{ij} = a_{\sigma(i),\sigma(j)}$ for all $i,j$. This guarantees that connectivity and labeling of the diagram are preserved. For a Lie algebra $\mathfrak{g}$ or Kac–Moody algebra $\mathfrak{g}(A)$ with Chevalley generators $e_i,f_i,H_i$, these automorphisms lift to Lie algebra automorphisms via

$$\sigma(e_i) = e_{\sigma(i)},\quad \sigma(f_i)=f_{\sigma(i)},\quad \sigma(H_i) = H_{\sigma(i)}$$

preserving the simple root system and the associated Serre relations (Mukhopadhyay, 2013). The extension commutes with the involutive bar symmetry, and in the affine case, can be consistently extended to the loop variables, central, and grading elements (Liu et al., 2018).

Diagram automorphisms permute roots, weights, and modules, yielding twisted module structures and fixed-point subalgebras $\mathfrak{g}^{\sigma}$.

2. Classification and Foldings

Nontrivial diagram automorphisms are classified by the symmetries of the underlying Dynkin diagrams. For simple finite types:

Type	Nontrivial Automorphisms	Folded Type
$A_n$ ($n\geq1$)	Order-2 involution (reversal about center)	$A_{2n-1}\to C_n$, $A_{2n}\to D_{n+1}$
$D_n$ ($n\geq4$)	Triality $S_3$ for $D_4$, inv. for $n>4$	$D_{n+1} \to B_n$, $D_4\to G_2$
$E_6$	Unique order-2 involution	$E_6\to F_4$

In affine types, diagram automorphisms $\bar{\sigma}$ correspond to permutations of the extended index set $I=\{0,1,\ldots,\ell\}$ such that $a_{\bar{\sigma}(i),\bar{\sigma}(j)}=a_{ij}$. Foldings by $\sigma$—by averaging over orbits and forming new Cartan matrices—yield new Dynkin diagrams of affine type when the so-called "linking numbers" $b_i\leq 2$ for all orbits (Liu et al., 2018). For example:

• $A_{2n-1}^{(1)} \to C_n^{(1)}$ via order-2 involution,
• $E_6^{(1)} \to F_4^{(1)}$.

Folded algebras inherit the affine property if all node orbits have size at most 2, leading to only specified affine types possessing such automorphisms (Liu et al., 2018).

3. Representation Theory and Rank-Level Dualities

Diagram automorphisms act functorially on representation categories. They generate isomorphisms between highest weight representations, define twisted modules $V(\lambda)^\sigma$, and induce symmetries on spaces of conformal blocks: $$U_\sigma:V(\lambda) \longrightarrow V(\sigma(\lambda)),\quad U_\sigma(Xv) = \sigma(X)U_\sigma v$$ A significant application is in the context of rank-level duality: if $\phi:g_1\oplus g_2\to g$ is an embedding, then the associated block isomorphisms commute with diagram automorphisms of $g_i$ and $g$. This functoriality underpins the parabolic strange duality and new symplectic rank–level dualities, as well as invariance of Verlinde numbers under diagram automorphism actions (Mukhopadhyay, 2013, Hong, 2016).

In conformal field theory, the trace of a diagram automorphism on spaces of conformal blocks is computed by a twisted Verlinde formula, involving the maximal torus and Weyl group of the folded algebra (Hong, 2016): $$\mathrm{Tr}(\sigma|V_\ell(\mathfrak{g};C,\vec{p},\vec{\lambda})) = |T_{\sigma,\ell}|^{g-1} \sum_{[t]\in T_{\sigma,\ell}^{\mathrm{reg}}/W^\sigma}\mathrm{Tr}(t|W_{\vec{\lambda}})A_{\mathfrak{g}^\sigma}(t)^{g-1}$$

4. Canonical Bases and Quantum Groups

Diagram automorphisms admit a profound influence on the theory of quantum groups and their canonical (global crystal) bases. For $U_q^-(\mathfrak{g})$, $\sigma$ acts as an algebra automorphism and permutes Lusztig's canonical basis $B$. The $\sigma$-fixed part $B^\sigma$ is canonically in bijection with the canonical basis of $U_q^-(\mathfrak{g}^\sigma)$, the negative part for the folded algebra (Shoji et al., 2018, Ma et al., 2022): $B^\sigma \longleftrightarrow \underline{B}$ This bijection extends to signed canonical bases and holds in finite, affine, admissible, and non-admissible (with noted exceptions) type (Ma et al., 2022). PBW-type bases adapted to $\sigma$ or modified elements are essential to derive and transport this bijection in detail.

In the affine setting, the diagram automorphism identifies the canonical bases of $U_q^-(\mathfrak{g})$ and the folded quantum algebra via the structure of invariant PBW-bases; the method extends to affine types using Beck–Nakajima constructions (Shoji et al., 2018).

5. Geometry: Quiver Varieties and Hyperplane Arrangements

Diagram automorphisms act on quivers and their representation varieties, and their fixed-point loci can be described as unions of Nakajima quiver varieties for split-quotient quivers (Henderson et al., 2013). For admissible automorphisms $a$ of a quiver $Q$, the fixed-point variety $M(v,w)^a$ decomposes as a union of quiver varieties for a folded quiver $Q^a$ defined in terms of orbits and eigenvalues. In the special case of the type $A$ involution, this process recovers varieties of type $D$ from those of type $A$.

Diagram automorphisms further induce identifications between quiver varieties and Slodowy slices for orthogonal or symplectic Lie algebras in small self-dual cases. Such correspondences generalize the McKay–Slodowy correspondence in this geometric context (Henderson et al., 2013).

Cyclotomic discriminantal arrangements arise when folding the usual hyperplane arrangements by cyclic diagram automorphisms. The associated flag and Orlik–Solomon complexes correspond via explicit isomorphisms to cohomology and chain complexes for subalgebras fixed by the automorphism, with applications to cyclotomic Gaudin models and the construction of Bethe vectors (Varchenko et al., 2016).

6. Combinatorics and Automorphism Groups

The automorphism group of generalized cluster complexes, as defined by Fomin–Reading, is generated by diagram automorphisms (arising from the underlying Coxeter/Dynkin diagram) together with a canonical dihedral symmetry (rotation and reflections). The overall automorphism group decomposes as a direct or semidirect product: $$\mathrm{Aut}~I(m) \cong \mathrm{Dih} \times \left(\mathrm{Diag}/\langle C\rangle\right)$$ with explicit order formulas dependent on the Coxeter number and symmetry group order (Josuat-Vergès, 2024). Diagram automorphisms may commute or intertwine with the dihedral generator depending on preservation or swapping of the bipartite structure of the diagram.

In the theory of AR–quivers and Coxeter combinatorics, diagram automorphisms lead to twisted Coxeter elements, twisted and folded cluster points, and AR–quivers that encode the combinatorics of representations in folded types (e.g., $A_{2n+1} \to B_{n+1}$), governing denominator formulas for $R$-matrices and Dorey's rule in the context of quantum affine algebras (Oh et al., 2016).

7. Spherical Buildings and Opposition Diagrams

In incidence geometry, diagram automorphisms correspond to automorphisms of the underlying spherical building, with detailed combinatorial encoding via opposition diagrams. For exceptional types ($E_6$, $F_4$, $G_2$), opposition diagrams classify automorphism types by which simplex-types are sent to their opposites, using Dynkin diagrams with marked nodes orbits. Domestic automorphisms (those not sending chambers to opposites) are rigidly classified by these diagrams, with connections to long root elations and unipotent or toral elements (Parkinson et al., 2020).

These opposition diagrams offer an exact, group-theoretically and geometrically canonical language to describe the action and classification of automorphisms in the geometric representation theory of algebraic groups of exceptional type.

---

In summary, diagram automorphisms unify algebraic, geometric, combinatorial, and physical phenomena by folding simply-laced structures to their non-simply-laced counterparts, controlling fixed point loci, invariant subalgebra structures, canonical bases, dualities in representation theory, and combinatorial symmetries across a wide array of mathematical disciplines. Their role is essential in the analysis and explicit construction of integrable hierarchies, quantum groups, moduli of sheaves, cluster algebras, quiver varieties, spherical buildings, and related structures (Liu et al., 2018, Mukhopadhyay, 2013, Josuat-Vergès, 2024, Shoji et al., 2018, Ma et al., 2022, Hong, 2016, Varchenko et al., 2016, Henderson et al., 2013, Oh et al., 2016, Parkinson et al., 2020).