Pure Symmetric Outer Automorphisms

• Pure symmetric outer automorphisms are defined as outer automorphisms that preserve specified generators, free factors, or diagram data via conjugacy rather than arbitrary permutation.
• They appear in diverse contexts—including semisimple algebraic groups, right-angled Artin groups, and free products—facilitating rigorous constraints on group symmetry and unique cases like the non-inner automorphism of S₆.
• Their study employs geometric, cohomological, and combinatorial methods to elucidate diagram symmetries, triality in type D groups, and explicit generation theorems in automorphism groups.

Pure symmetric outer automorphisms are outer automorphisms subject to a purity constraint: they preserve specified generators, free factors, or diagram data up to conjugacy rather than permuting that data freely. The phrase is therefore context-dependent rather than universal. In the literature on semisimple algebraic groups it refers to $k$-defined diagram automorphisms acting nontrivially on the Dynkin diagram; in the literature on right-angled Artin groups it refers to basis-conjugating outer automorphisms $P\Sigma\mathrm{Out}(A_\Gamma)$; in the literature on free products it refers to outer automorphisms relative to a fixed splitting $G=G_1*\cdots *G_n$ that carry each factor $G_i$ to a conjugate of itself; and, as a classical benchmark for exceptional outer symmetry, the symmetric groups contribute the unique non-inner automorphism class of $S_6$ (Garibaldi, 2010, Corrigan, 7 Mar 2025, Iveson, 22 Jul 2025, McCammond, 2014).

1. Basic definitions and terminological scope

For any group $G$, the automorphism group $\mathrm{Aut}(G)$ contains the normal subgroup of inner automorphisms $\mathrm{Inn}(G)$, and the outer automorphism group is the quotient

$\mathrm{Out}(G)=\mathrm{Aut}(G)/\mathrm{Inn}(G).$

This quotient isolates symmetries not induced by conjugation. In the finite-group setting this is the standard notion used to state the exceptional theorem for $S_6$ (McCammond, 2014).

In the algebraic-group setting, one starts from the exact sequence of group schemes

$P\Sigma\mathrm{Out}(A_\Gamma)$0

and for a pinned semisimple group over $P\Sigma\mathrm{Out}(A_\Gamma)$1 one has $P\Sigma\mathrm{Out}(A_\Gamma)$2, where $P\Sigma\mathrm{Out}(A_\Gamma)$3 is the Dynkin diagram. In that language, “pure symmetric outer automorphisms” are precisely the diagram automorphisms represented by $P\Sigma\mathrm{Out}(A_\Gamma)$4-points of $P\Sigma\mathrm{Out}(A_\Gamma)$5 that act nontrivially on $P\Sigma\mathrm{Out}(A_\Gamma)$6 (Garibaldi, 2010).

For a right-angled Artin group

$P\Sigma\mathrm{Out}(A_\Gamma)$7

an automorphism is pure symmetric if each standard generator $P\Sigma\mathrm{Out}(A_\Gamma)$8 is sent to a conjugate of itself. The resulting groups are

$P\Sigma\mathrm{Out}(A_\Gamma)$9

and the notation $G=G_1*\cdots *G_n$0, $G=G_1*\cdots *G_n$1 is used in the same sense in recent cohomological work (Ardaiz-Gale, 15 Apr 2026).

For a fixed free product splitting $G=G_1*\cdots *G_n$2, a pure symmetric automorphism relative to $G=G_1*\cdots *G_n$3 is an automorphism $G=G_1*\cdots *G_n$4 such that $G=G_1*\cdots *G_n$5 for each $G=G_1*\cdots *G_n$6. Its outer class lies in $G=G_1*\cdots *G_n$7. Here “pure” means that the factors are not permuted, while “symmetric” is in the McCullough–Miller sense (Iveson, 22 Jul 2025).

2. The exceptional outer symmetry of $G=G_1*\cdots *G_n$8

Among symmetric groups, the outer automorphism group is trivial except in degree $G=G_1*\cdots *G_n$9: $G_i$0 Equivalently,

$G_i$1

The elementary proof proceeds by tracking the conjugacy class $G_i$2 of transpositions among the involution classes $G_i$3 of cycle type $G_i$4. If an automorphism sends transpositions to transpositions, it is inner; if it sends $G_i$5 to some $G_i$6 with $G_i$7, then the product-order argument forces $G_i$8 and $G_i$9. Thus only $S_6$0 admits a non-inner automorphism, and it must interchange transpositions with triple disjoint transpositions (McCammond, 2014).

The rigidity of the class of transpositions is encoded in the algebraically characterized subsets

$S_6$1

Each $S_6$2 is a maximal subset of $S_6$3 with the property that any two distinct elements in it do not commute and $S_6$4 for any noncommuting $S_6$5. An automorphism preserving $S_6$6 must permute the family $S_6$7; after conjugation by the corresponding permutation, it fixes every transposition and hence is inner, because transpositions generate $S_6$8 (McCammond, 2014).

The exceptional case is visible numerically. In $S_6$9, the class of transpositions and the class of triple disjoint transpositions both have centralizer size $G$0 and class size $G$1: $G$2

$G$3

The icosahedron model gives an explicit automorphism realizing the swap $G$4: label the $G$5 antipodal vertex pairs of a regular icosahedron by $G$6, identify the resulting $G$7 labelings modulo rigid motions as $G$8 dual pairs $G$9, and let permutations of $\mathrm{Aut}(G)$0 act on these dual pairs. After identifying $\mathrm{Aut}(G)$1 with $\mathrm{Aut}(G)$2, one obtains an automorphism of $\mathrm{Aut}(G)$3 that sends transpositions to products of three disjoint transpositions. Graph-theoretically, it exchanges the $\mathrm{Aut}(G)$4 edges of $\mathrm{Aut}(G)$5 with the $\mathrm{Aut}(G)$6 perfect matchings and swaps the $\mathrm{Aut}(G)$7 stars with the $\mathrm{Aut}(G)$8 factorizations of the edge set into perfect matchings (McCammond, 2014).

A later construction replaces the icosahedron by a complex Hadamard matrix of order $\mathrm{Aut}(G)$9 with third roots of unity and the algebra of split quaternions. In that construction, a subgroup $\mathrm{Inn}(G)$0 acts through two inequivalent $\mathrm{Inn}(G)$1-point permutation representations obtained as permutation projections of conjugate monomial actions, and the resulting automorphism $\mathrm{Inn}(G)$2 sends $\mathrm{Inn}(G)$3 to $\mathrm{Inn}(G)$4. It also swaps the conjugacy classes $\mathrm{Inn}(G)$5, $\mathrm{Inn}(G)$6, and $\mathrm{Inn}(G)$7, while fixing the identity and the $\mathrm{Inn}(G)$8-cycle class, thereby exhibiting again the unique nontrivial element of $\mathrm{Inn}(G)$9 (Gillespie et al., 2018).

3. Diagram symmetries and triality in algebraic groups

For a connected semisimple algebraic group $\mathrm{Out}(G)=\mathrm{Aut}(G)/\mathrm{Inn}(G).$0 over a field $\mathrm{Out}(G)=\mathrm{Aut}(G)/\mathrm{Inn}(G).$1, the outer automorphism problem is controlled by the Dynkin diagram. The map

$\mathrm{Out}(G)=\mathrm{Aut}(G)/\mathrm{Inn}(G).$2

records the induced action on the based root datum, and the main cohomological criterion states that, for semisimple simply connected $\mathrm{Out}(G)=\mathrm{Aut}(G)/\mathrm{Inn}(G).$3,

$\mathrm{Out}(G)=\mathrm{Aut}(G)/\mathrm{Inn}(G).$4

where $\mathrm{Out}(G)=\mathrm{Aut}(G)/\mathrm{Inn}(G).$5 is the Tits class. Under the exactness conditions formulated in the paper, a diagram automorphism $\mathrm{Out}(G)=\mathrm{Aut}(G)/\mathrm{Inn}(G).$6 is realized by a $\mathrm{Out}(G)=\mathrm{Aut}(G)/\mathrm{Inn}(G).$7-automorphism of $\mathrm{Out}(G)=\mathrm{Aut}(G)/\mathrm{Inn}(G).$8 if and only if $\mathrm{Out}(G)=\mathrm{Aut}(G)/\mathrm{Inn}(G).$9. In this setting, the pure symmetric outer automorphisms are exactly the nontrivial diagram automorphisms defined over $S_6$0 (Garibaldi, 2010).

Type $S_6$1 is the main example. For the simply connected $S_6$2 with $S_6$3 even, the center is

$S_6$4

and the Tits class is

$S_6$5

The unique nontrivial diagram symmetry of $S_6$6 swaps the two spinor nodes and therefore the two $S_6$7-factors. It is realized over $S_6$8 if and only if

$S_6$9

For quasi-split groups, $P\Sigma\mathrm{Out}(A_\Gamma)$00, so every diagram automorphism exists over $P\Sigma\mathrm{Out}(A_\Gamma)$01. For non-archimedean local fields, $P\Sigma\mathrm{Out}(A_\Gamma)$02 for semisimple simply connected $P\Sigma\mathrm{Out}(A_\Gamma)$03, and the same Tits-class criterion becomes exact (Garibaldi, 2010).

Type $P\Sigma\mathrm{Out}(A_\Gamma)$04 is exceptional because $P\Sigma\mathrm{Out}(A_\Gamma)$05, the triality group. This is where symmetric composition algebras enter. For an eight-dimensional composition algebra $P\Sigma\mathrm{Out}(A_\Gamma)$06 with norm $P\Sigma\mathrm{Out}(A_\Gamma)$07, one associates two strongly outer automorphisms $P\Sigma\mathrm{Out}(A_\Gamma)$08 of $P\Sigma\mathrm{Out}(A_\Gamma)$09. If $P\Sigma\mathrm{Out}(A_\Gamma)$10 is symmetric, then $P\Sigma\mathrm{Out}(A_\Gamma)$11 has order $P\Sigma\mathrm{Out}(A_\Gamma)$12, $P\Sigma\mathrm{Out}(A_\Gamma)$13, and $P\Sigma\mathrm{Out}(A_\Gamma)$14. More generally, if $P\Sigma\mathrm{Out}(A_\Gamma)$15 is an isotope of a symmetric composition algebra, then

$P\Sigma\mathrm{Out}(A_\Gamma)$16

The paper formulates this as an equivalence of categories

$P\Sigma\mathrm{Out}(A_\Gamma)$17

where $P\Sigma\mathrm{Out}(A_\Gamma)$18 is the $P\Sigma\mathrm{Out}(A_\Gamma)$19-action groupoid of trialitarian pairs. In this sense, the symmetric case gives the pure order-$P\Sigma\mathrm{Out}(A_\Gamma)$20 triality automorphisms, while arbitrary composition algebras correspond to trialitarian pairs modulo weakly inner conjugacy (Alsaody, 2015).

4. Pure symmetric outer automorphisms of right-angled Artin groups

For a finite simplicial graph $P\Sigma\mathrm{Out}(A_\Gamma)$21, the right-angled Artin group $P\Sigma\mathrm{Out}(A_\Gamma)$22 has one generator for each vertex and one commutation relation for each edge. A pure symmetric automorphism is one that sends each standard generator $P\Sigma\mathrm{Out}(A_\Gamma)$23 to a conjugate of itself. The group $P\Sigma\mathrm{Out}(A_\Gamma)$24 is generated by partial conjugations, and its outer quotient is

$P\Sigma\mathrm{Out}(A_\Gamma)$25

When $P\Sigma\mathrm{Out}(A_\Gamma)$26, there is a short exact sequence

$P\Sigma\mathrm{Out}(A_\Gamma)$27

In the notation of recent work, $P\Sigma\mathrm{Out}(A_\Gamma)$28 and $P\Sigma\mathrm{Out}(A_\Gamma)$29 are the same groups denoted $P\Sigma\mathrm{Out}(A_\Gamma)$30 and $P\Sigma\mathrm{Out}(A_\Gamma)$31 (Ardaiz-Gale, 15 Apr 2026).

A partial conjugation is determined by a vertex $P\Sigma\mathrm{Out}(A_\Gamma)$32 and a union, or in the standard generating set a connected component, of $P\Sigma\mathrm{Out}(A_\Gamma)$33. If $P\Sigma\mathrm{Out}(A_\Gamma)$34 is such a union, then

$P\Sigma\mathrm{Out}(A_\Gamma)$35

Laurence proved that partial conjugations generate $P\Sigma\mathrm{Out}(A_\Gamma)$36. Koban–Piggott’s presentation organizes the relations in terms of shared, dominant, and subordinate components for nonadjacent pairs $P\Sigma\mathrm{Out}(A_\Gamma)$37: commuting relations occur when the supports are suitably separated or subordinate, and the genuinely nontrivial relations are the SIL-type relations attached to separating intersections of links (Ardaiz-Gale, 15 Apr 2026).

The larger symmetric outer automorphism group $P\Sigma\mathrm{Out}(A_\Gamma)$38 allows graph symmetries and inversions. It is generated by graph symmetries, inversions $P\Sigma\mathrm{Out}(A_\Gamma)$39, and $P\Sigma\mathrm{Out}(A_\Gamma)$40-Whitehead automorphisms $P\Sigma\mathrm{Out}(A_\Gamma)$41 defined by symmetric $P\Sigma\mathrm{Out}(A_\Gamma)$42-partitions. The pure symmetric outer group $P\Sigma\mathrm{Out}(A_\Gamma)$43 is a finite-index subgroup of $P\Sigma\mathrm{Out}(A_\Gamma)$44, and in the language of relative outer automorphism groups one has

$P\Sigma\mathrm{Out}(A_\Gamma)$45

Consequently,

$P\Sigma\mathrm{Out}(A_\Gamma)$46

This places pure symmetric outer automorphisms simultaneously in the RAAG-automorphism and relative-outer-automorphism frameworks (Corrigan, 7 Mar 2025).

5. BNS invariants, cohomology, geometry, and graded Lie theory for RAAGs

The structure of $P\Sigma\mathrm{Out}(A_\Gamma)$47 is highly sensitive to SIL combinatorics. Day–Wade define, for each vertex $P\Sigma\mathrm{Out}(A_\Gamma)$48, a support graph $P\Sigma\mathrm{Out}(A_\Gamma)$49 whose vertices are the connected components of $P\Sigma\mathrm{Out}(A_\Gamma)$50, and whose edges record dominating–shared pairs across SILs. They prove that $P\Sigma\mathrm{Out}(A_\Gamma)$51 is a RAAG if and only if every support graph $P\Sigma\mathrm{Out}(A_\Gamma)$52 is a forest; if some $P\Sigma\mathrm{Out}(A_\Gamma)$53 contains a loop, then the first homology of the BNS-derived subspace arrangement is nontrivial, $P\Sigma\mathrm{Out}(A_\Gamma)$54, and $P\Sigma\mathrm{Out}(A_\Gamma)$55 is not a RAAG. For the automorphism group rather than the outer quotient, Koban–Piggott show that $P\Sigma\mathrm{Out}(A_\Gamma)$56 is a RAAG if and only if $P\Sigma\mathrm{Out}(A_\Gamma)$57 has no separating intersection of links (Day et al., 2015, Koban et al., 2013).

The cohomology of the pure symmetric outer group is computed combinatorially from the $P\Sigma\mathrm{Out}(A_\Gamma)$58-Whitehead poset $P\Sigma\mathrm{Out}(A_\Gamma)$59. If $P\Sigma\mathrm{Out}(A_\Gamma)$60 denotes the number of rank-$P\Sigma\mathrm{Out}(A_\Gamma)$61 essential vertex types, then

$P\Sigma\mathrm{Out}(A_\Gamma)$62

and $P\Sigma\mathrm{Out}(A_\Gamma)$63 is generated in degree $P\Sigma\mathrm{Out}(A_\Gamma)$64 by classes $P\Sigma\mathrm{Out}(A_\Gamma)$65 dual to canonical outer automorphisms. If $P\Sigma\mathrm{Out}(A_\Gamma)$66 is the number of $P\Sigma\mathrm{Out}(A_\Gamma)$67-cliques in $P\Sigma\mathrm{Out}(A_\Gamma)$68, then under $P\Sigma\mathrm{Out}(A_\Gamma)$69,

$P\Sigma\mathrm{Out}(A_\Gamma)$70

again with ring generation in degree $P\Sigma\mathrm{Out}(A_\Gamma)$71. The paper also formulates the Generalized Brownstein–Lee Conjecture and proves it in dimension $P\Sigma\mathrm{Out}(A_\Gamma)$72 (Ardaiz-Gale, 15 Apr 2026).

The corresponding outer space is now available. There exists a contractible cube complex $P\Sigma\mathrm{Out}(A_\Gamma)$73, the symmetric spine, equipped with a proper and cocompact action of $P\Sigma\mathrm{Out}(A_\Gamma)$74. More generally, for any finite set $P\Sigma\mathrm{Out}(A_\Gamma)$75 of conjugacy classes, the subcomplex $P\Sigma\mathrm{Out}(A_\Gamma)$76 is contractible, and the subgroup $P\Sigma\mathrm{Out}(A_\Gamma)$77 that permutes $P\Sigma\mathrm{Out}(A_\Gamma)$78 is of type $P\Sigma\mathrm{Out}(A_\Gamma)$79. The virtual cohomological dimension satisfies

$P\Sigma\mathrm{Out}(A_\Gamma)$80

where $P\Sigma\mathrm{Out}(A_\Gamma)$81 is the set of principal vertices and $P\Sigma\mathrm{Out}(A_\Gamma)$82 is the maximal size of a compatible family of symmetric $P\Sigma\mathrm{Out}(A_\Gamma)$83-partitions based at vertices of $P\Sigma\mathrm{Out}(A_\Gamma)$84. Since $P\Sigma\mathrm{Out}(A_\Gamma)$85 has finite index, the same virtual cohomological dimension holds for the pure symmetric outer group (Corrigan, 7 Mar 2025).

Representation-theoretically, $P\Sigma\mathrm{Out}(A_\Gamma)$86 admits a homomorphism

$P\Sigma\mathrm{Out}(A_\Gamma)$87

where $P\Sigma\mathrm{Out}(A_\Gamma)$88 is a RAAG and the image is surjective onto each factor. This yields CAT(0) actions in which every partial conjugation has infinite-order image. For connected $P\Sigma\mathrm{Out}(A_\Gamma)$89, there is also a RAAG $P\Sigma\mathrm{Out}(A_\Gamma)$90 with

$P\Sigma\mathrm{Out}(A_\Gamma)$91

and $P\Sigma\mathrm{Out}(A_\Gamma)$92 is free abelian. If $P\Sigma\mathrm{Out}(A_\Gamma)$93 has no SIL, then $P\Sigma\mathrm{Out}(A_\Gamma)$94 is itself a RAAG and $P\Sigma\mathrm{Out}(A_\Gamma)$95 is abelian (Aramayona et al., 2017).

At the level of descending central series, both $P\Sigma\mathrm{Out}(A_\Gamma)$96 and $P\Sigma\mathrm{Out}(A_\Gamma)$97 are $P\Sigma\mathrm{Out}(A_\Gamma)$98-formal and have explicit quadratic presentations for their graded Lie algebras. For $P\Sigma\mathrm{Out}(A_\Gamma)$99, one obtains the graded Lie algebra of $G=G_1*\cdots *G_n$00 by adding the linear relations $G=G_1*\cdots *G_n$01 over the components $G=G_1*\cdots *G_n$02 of $G=G_1*\cdots *G_n$03. A combinatorial condition

$G=G_1*\cdots *G_n$04

lying in four distinct connected components of $G=G_1*\cdots *G_n$05 governs Koszulness: if $G=G_1*\cdots *G_n$06 holds, then $G=G_1*\cdots *G_n$07 and $G=G_1*\cdots *G_n$08 are iterated extensions of RAAGs, hence poly-RAAG and poly-free, and the associated graded Lie algebras are Koszul. By contrast, for $G=G_1*\cdots *G_n$09, $G=G_1*\cdots *G_n$10 and $G=G_1*\cdots *G_n$11 are not poly–finitely generated free (Martínez-Pérez et al., 14 Oct 2025).

6. Free products and relative Whitehead generation

For a fixed splitting $G=G_1*\cdots *G_n$12 with $G=G_1*\cdots *G_n$13 and all $G=G_1*\cdots *G_n$14, the pure symmetric outer automorphism group $G=G_1*\cdots *G_n$15 consists of those outer automorphisms that have a representative $G=G_1*\cdots *G_n$16 with $G=G_1*\cdots *G_n$17 for every $G=G_1*\cdots *G_n$18. Two natural families of generators appear. A factor automorphism relative to $G=G_1*\cdots *G_n$19 restricts to an automorphism of each factor $G=G_1*\cdots *G_n$20, so factor outer automorphisms form a subgroup isomorphic to $G=G_1*\cdots *G_n$21. A Whitehead automorphism relative to $G=G_1*\cdots *G_n$22 is determined by an operating factor $G=G_1*\cdots *G_n$23, an element $G=G_1*\cdots *G_n$24, and a subset $G=G_1*\cdots *G_n$25; it conjugates each factor in $G=G_1*\cdots *G_n$26 by $G=G_1*\cdots *G_n$27 and fixes the others. The operating factor is well-defined at the outer level (Iveson, 22 Jul 2025).

The main theorem states that every pure symmetric outer automorphism of the splitting can be written as a product of factor outer automorphisms relative to $G=G_1*\cdots *G_n$28 and Whitehead outer automorphisms relative to $G=G_1*\cdots *G_n$29. The proof is geometric. One constructs an Outer Space $G=G_1*\cdots *G_n$30 from $G=G_1*\cdots *G_n$31-labelled $G=G_1*\cdots *G_n$32-trees and then a “nice” $G=G_1*\cdots *G_n$33-dimensional subcomplex $G=G_1*\cdots *G_n$34 spanned by $G=G_1*\cdots *G_n$35-graph classes and $G=G_1*\cdots *G_n$36-graph classes. The star joining the basepoint $G=G_1*\cdots *G_n$37 to the vertices $G=G_1*\cdots *G_n$38 is a strict fundamental domain for the action of $G=G_1*\cdots *G_n$39 on $G=G_1*\cdots *G_n$40. A volume-decreasing argument in the Bass–Serre tree produces an $G=G_1*\cdots *G_n$41–$G=G_1*\cdots *G_n$42–$G=G_1*\cdots *G_n$43 path from any $G=G_1*\cdots *G_n$44-vertex to $G=G_1*\cdots *G_n$45, proving that $G=G_1*\cdots *G_n$46 is path-connected and enabling an inductive factorization of any element of $G=G_1*\cdots *G_n$47 into the stated generators (Iveson, 22 Jul 2025).

Under the additional hypothesis that the factors are non-trivial, not infinite cyclic, freely indecomposable, and pairwise non-isomorphic, the splitting is Grushko and every automorphism of $G=G_1*\cdots *G_n$48 is pure symmetric. In that case,

$G=G_1*\cdots *G_n$49

so the same theorem becomes a generation theorem for the full outer automorphism group. This relative free-product picture is the direct analogue of Whitehead-type generation in $G=G_1*\cdots *G_n$50, with factor automorphisms replacing basis permutations and relative Whitehead automorphisms replacing basis-conjugating moves (Iveson, 22 Jul 2025).