The R∞  and S∞ properties for linear algebraic groups

Alexander Fel’shtyn; Timur Nasybullov

doi:10.1515/jgth-2016-0004

Article Publicly Available

The R_∞ and S_∞ properties for linear algebraic groups

Alexander Fel’shtyn and Timur Nasybullov

Published/Copyright: February 5, 2016

Published by

Become an author with De Gruyter Brill

Submit Manuscript Author Information Explore this Subject

From the journal Journal of Group Theory Volume 19 Issue 5

Abstract

In this paper we study twisted conjugacy classes and isogredience classes for automorphisms of reductive linear algebraic groups. We show that reductive linear algebraic groups over some fields of zero characteristic possess the R∞ and S∞ properties.

1 Introduction

Let φ:G→G be an endomorphism of a group G. Then two elements x,y of G are said to be twisted φ-conjugate if there exists a third element z∈G such that x=z⁢y⁢φ⁢(z)-1. The equivalence classes of this relation are called the twisted conjugacy classes or the Reidemeister classes of φ. The Reidemeister number of φ denoted by R⁢(φ), is the number of twisted conjugacy classes of φ. This number is either a positive integer or ∞ and we do not distinguish different infinite cardinal numbers. An infinite group G possesses the R∞-property if for every automorphism φ of G the Reidemeister number of φ is infinite.

The interest in twisted conjugacy relations has its origins, in particular, in Nielsen–Reidemeister fixed point theory (see, e.g. [7, 34]), in Arthur–Selberg theory (see, e.g. [1, 50]), in algebraic geometry (see, e.g. [28]), in Galois cohomology [49] and in the theory of linear algebraic groups (see, e.g. [52]). In representation theory twisted conjugacy probably occurs first in Gantmacher’s paper [20] (see, e.g. [51, 45]).

The problem of determining which classes of discrete infinite groups have the R∞-property is an area of active research initiated by Fel’shtyn and Hill in 1994 (see [13]). Later, it was shown by various authors that the following groups have the R∞-property: non-elementary Gromov hyperbolic groups (see [8, 38]); relatively hyperbolic groups (see [9]); Baumslag–Solitar groups BS⁢(m,n) except for BS⁢(1,1) (see [10]), generalized Baumslag–Solitar groups, that is, finitely generated groups which act on a tree with all edge and vertex stabilizers infinite cyclic (see [37]); the solvable generalization Γ of BS⁢(1,n) given by the short exact sequence 1→ℤ⁢[1n]→Γ→ℤk→1 (see [53]); a wide class of saturated weakly branch groups (including the Grigorchuk group (see [27]) and the Gupta–Sidki group (see [29])) (see [15]), Thompson’s groups F (see [2]) and T (see[3, 22]); the generalized Thompson groups Fn, 0 and their finite direct products (see [21]); Houghton’s groups (see [23, 35]); symplectic groups Sp⁢(2⁢n,ℤ), the mapping class groups ModS of a compact oriented surface S with genus g and p boundary components, 3⁢g+p-4>0, and the full braid groups Bn⁢(S) on n>3 strands of a compact surface S in the cases where S is either the compact disk D, or the sphere S2 (see [11]); some classes of Artin groups of infinite type (see [36]); extensions of SL⁢(n,ℤ), PSL⁢(n,ℤ), GL⁢(n,ℤ), PGL⁢(n,ℤ), Sp⁢(2⁢n,ℤ), PSp⁢(2⁢n,ℤ), for n>2, by a countable abelian group, and normal subgroups of SL⁢(n,ℤ), n>2, not contained in the center (see [40]); GL⁢(n,K), SL⁢(n,K), Sp⁢(2⁢n,K) and On⁢(K) if n>2 and K is an infinite integral domain which has zero characteristic and for which Aut⁢(K) is periodic (see [42, 44]); Chevalley groups over fields of zero characteristic which have finite transcendence degree over ℚ (see [43]); irreducible lattices in a connected semisimple Lie group G with finite center and real rank at least 2 (see [41]); non-amenable, finitely generated residually finite groups (see [17]) (this class gives a lot of new examples of groups with the R∞-property); some metabelian groups of the form ℚn⋊ℤ and ℤ⁢[1p]n⋊ℤ (see [12]); lamplighter groups ℤn≀ℤ if and only if 2|n or 3|n (see [25]); free nilpotent groups Nr⁢c of rank r=2 and class c≥9 (see [26]), Nr⁢c of rank r=2 or r=3 and class c≥4⁢r, or rank r≥4 and class c≥2⁢r, any group N2⁢c for c≥4, every free solvable group S2⁢t of rank 2 and class t≥2, any free solvable group Sr⁢t of rank r≥2 and class t big enough (see [47]); some crystallographic groups (see [6, 39]). Recently, in [5] it was proven that Nr⁢c, r>1 possesses the R∞-property if and only if c≥2⁢r.

Let Ψ be an element of Out⁢(G)=Aut⁢(G)/Inn⁢(G) (we consider Ψ as a collection of ordinary automorphisms a∈Aut⁢(G)). We say that two automorphisms a,b∈Ψ are similar (or isogredient) if b=φh⁢a⁢φh-1 for some h∈G, where φh⁢(g)=h⁢g⁢h-1 is an inner automorphism induced by the element h (see [38]). Let 𝔖⁢(Ψ) be the set of isogredience classes of automorphisms representing Ψ. Denote by S⁢(Ψ) the cardinality of the set 𝔖⁢(Ψ). The group G is said to possess the S∞-property if for every Ψ the set 𝔖⁢(Ψ) is infinite, i.e. S⁢(Ψ)=∞ (see [18]).

In this paper we study the R∞ and S∞ properties for linear algebraic groups. The first results in this direction were obtained for some classes of Chevalley groups by Nasybullov in [43].

In Section 3 we extend the previous result from [43] and prove the following.

Theorem 3.2

Let G be a Chevalley group of type Φ over a field F of zero characteristic. If the transcendence degree of F over Q is finite, then G possesses the R∞-property.

The following main theorem is proved in Section 4.

Theorem 4.1

Let F be an algebraically closed field of zero characteristic such that the transcendence degree of F over Q is finite. If a reductive linear algebraic group G over the field F has a nontrivial quotient group G/R⁢(G), where R⁢(G) is the radical of G, then G possesses the R∞-property.

These theorems cannot be generalized to groups over a field of non-zero characteristic, by the following theorem of Steinberg [52, Theorem 10.1].

Theorem

Let G be a connected linear algebraic group and let φ be an endomorphism of G onto G. If φ has a finite number of fixed points, then

G={x⁢φ⁢(x-1):x∈G}.

Indeed, any semisimple linear algebraic group over an algebraically closed field of positive characteristic possesses an automorphism φ with finitely many fixed points (the Frobenius morphism, see [48, Section 3.2]), therefore, this group coincides with the set {x⁢φ⁢(x-1):x∈G}=[e]φ, hence R⁢(φ)=1 and such a group cannot possess the R∞-property.

Throughout the paper, by an automorphism of a linear algebraic group G we mean an automorphism of the abstract group underlying G (without any extra conditions).

If T1,T2,… are algebraically independent over ℚ, the fields ℚ¯, ℚ⁢(T1,…,Tk)¯(k≥1) are algebraically closed fields of zero characteristic with the finite transcendence degree over ℚ. Then reductive linear algebraic groups over these fields possess the R∞-property.

In Section 5 we prove that an infinite reductive linear algebraic group G over a field F of zero characteristic and finite transcendence degree over ℚ which possesses an automorphism φ with finite Reidemeister number is a torus.

In Section 6 we prove the following theorem.

Theorem 6.3

2 Preliminaries

In this section we recall some preliminary statements which are used in the paper. A lot of the results we use are thoroughly presented in [43], this can be used as background material.

The symbols In and On×m indicate the identity n×n matrix and the n×m matrix with zero entries, respectively. If A is an n×n matrix and B is an m×m matrix, then the symbol A⊕B denotes the direct sum of the matrices A and B, i.e. the block-diagonal (m+n)×(m+n) matrix

([c|c]AOn×mOm×nB).

It is obvious that for a pair of n×n matrices A1,A2 and for a pair of m×m matrices B1,B2 we have

(A1⊕B1)⁢(A2⊕B2)=A1⁢A2⊕B1⁢B2 and (A1⊕B1)-1=A1-1⊕B1-1.

The symbols G×H and G∘H denote the direct product and the central product of the groups G and H, respectively.

If g is an element of a group G, then φg denotes the inner automorphism induced by the element g. The following lemma can be found in [16, Corollary 2.5].

Lemma 2.1

Let φ be an automorphism of a group G and let φg be an inner automorphism of the group G. Then R⁢(φ⁢φg)=R⁢(φ).

The next lemma is proved in [40, Lemma 2.1].

Lemma 2.2

Let

1→N→G→A→1

be an exact sequence of groups. Suppose that N is a characteristic subgroup of G and that A possesses the R∞-property. Then G also possesses the R∞-property.

Here we prove a similar result for the S∞-property.

Lemma 2.3

Let

1→N→G→A→1

be an exact sequence of groups. Suppose that N is a characteristic subgroup of G and that A possesses the S∞-property. Then G also possesses the S∞-property.

Proof.

Let φ be an automorphism of the group G. Since N is the characteristic subgroup of G, it follows that φ induces the automorphism φ¯ of the group A. Since the group A has the S∞-property, there exists an infinite set of elements g¯1,g¯2,… of the group A such that φg¯i⁢φ¯ and φg¯j⁢φ¯ are not isogredient for i≠j.

Suppose that S⁢(φ⁢Inn⁢(G))<∞. Then there exists a pair of isogredient automorphisms in the set φg1⁢φ,φg2⁢φ,…. Suppose that φgi⁢φ and φgj⁢φ are isogredient for i≠j. Then for some element h∈G we have

φgi⁢φ=φh⁢φgj⁢φ⁢φh-1.

From this equality we have the following equality in the group Aut⁢(A)

φg¯i⁢φ¯=φh¯⁢φg¯j⁢φ¯⁢φh¯-1,

this contradicts the choice of the elements g¯1,g¯2,…. ∎

Let ν be a map from the set of rational numbers ℚ to the set 2π of all subsets of the set of prime numbers π, which acts on the irreducible fraction x=a/b by the rule

ν⁢(x)={all prime divisors of⁢a}∪{all prime divisors of⁢b}.

The proof of the following lemma is presented in [43, Lemma 5].

Lemma 2.4

Let F be a field of zero characteristic and let x1,…,xk be elements of F which are algebraically independent over the field Q. Let xk+1 be an element of F such that the elements x1,…,xk+1 are algebraically dependent over Q. Let δ be an automorphism of the field F which acts on these elements by the rule

δ:xi↦t0⁢ti⁢xi,i=1,…,k+1,

where t0,…,tk+1∈Q and t1,…,tk+1 are not equal to 1. If ν⁢(ti)∩ν⁢(tj)=∅ for i≠j, then xk+1=0.

Using this lemma we prove the following auxiliary statement.

Lemma 2.5

Let F be a field of zero characteristic such that the transcendence degree of F over Q is finite. If the automorphism δ of the field F acts on the elements z1, z2,… of the field F by the rule

δ:zi↦α⁢ai⁢zi,

where α∈F, 1≠ai∈Q⊆F and ν⁢(ai)∩ν⁢(aj)=∅ for i≠j, then there are only a finite number of non-zero elements in the set z1,z2,….

Proof.

If all the elements z1,z2,… are equal to zero, there is nothing to prove. Hence we can assume that there exists a non-zero element in the set z1,z2,…. Without loss of generality we can assume that z1≠0. (Otherwise we can reorder the elements z1,z2,… such that the first element is not equal to zero. If the statement holds for the reordered set, then it holds for the original set z1,z2,…). Denote by yi=zi⁢z1-1. Then the automorphism δ acts on the element yi by the rule

δ⁢(yi)=δ⁢(zi⁢z1-1)=δ⁢(zi)⁢δ⁢(z1-1)

=α⁢ai⁢zi⁢α-1⁢a1-1⁢z1-1

=ai⁢a1-1⁢zi⁢z1-1=ai⁢a1-1⁢yi.

Since the transcendence degree of F over ℚ is finite, there exists a maximal subset of algebraically independent over ℚ elements in the set y2,y3,…, i.e. there exists a finite set yi1,yi2,…,yik of elements algebraically independent over ℚ such that the set yi1,yi2,…,yik,yj is algebraically dependent over ℚ for every j.

Without loss of generality we can assume that the set y2,…,yk is a maximal subset of elements algebraically independent over ℚ in the set y1,y2,….

If n>k is a positive integer, then the elements y2,…,yk,yn∈F satisfy the conditions of Lemma 2.4. Thus yn=0 for all n>k and since yn=zn⁢z1-1, we have zn=0 for all n>k and the only non-zero elements are z1,z2,…,zk. ∎

Let us recall some facts about Chevalley groups. We use definitions and notations from [4].

Let Φ be an indecomposable root system of rank l with a subsystem of simple roots Δ, |Δ|=l. The elementary Chevalley group Φ⁢(F) of type Φ over a field F is a subgroup in the automorphism group of a simple Lie algebra ℒ of type Φ, which is generated by the elementary root elements xα⁢(t), α∈Φ, t∈F. The dimension of the Lie algebra ℒ is equal to |Φ|+|Δ| and therefore the group Φ⁢(F) can be considered as a subgroup in the group of all (|Φ|+|Δ|)×(|Φ|+|Δ|) invertible matrices.

In the elementary Chevalley group, we consider the following important elements:

nα⁢(t)=xα⁢(t)⁢x-α⁢(-t-1)⁢xα⁢(t),hα⁢(t)=nα⁢(t)⁢nα⁢(-1),t∈F*,α∈Φ.

For an arbitrary Chevalley group G of type Φ over a field F we have the following short exact sequence of groups:

1→Z⁢(G)→G→Φ⁢(F)→1,

where Z⁢(G) is the center of the group G, and by Lemma 2.2 we are mostly interested in the study of the R∞-property for elementary Chevalley groups.

Detailed information about automorphisms of Chevalley groups can be found in [43, 30]. Every Chevalley group has the following automorphisms:

Inner automorphism φg, induced by an element g∈G,
φg:x↦g⁢x⁢g-1.
Diagonal automorphism φh,
φh:x↦h⁢x⁢h-1,
where the element h can be presented as a diagonal (|Φ|+|Δ|)×(|Φ|+|Δ|) matrix. If F is an algebraically closed field, then any diagonal automorphism is inner [43, Lemma 4].
Field automorphism δ¯,
δ¯:x=(xi⁢j)↦(δ⁢(xi⁢j)),
where δ is an automorphism of the field F.
Graph automorphism ρ¯, which acts on the generators of the group G by the rule
ρ¯:xα⁢(t)↦xρ⁢(α)⁢(t),
where ρ is a symmetry of a Dynkin diagram. The order of the graph automorphism is equal to 2 or to 3.

Any field automorphism commutes with any graph automorphism. All diagonal automorphisms form a normal subgroup in the group which is generated by diagonal, graph and field automorphisms.

A theorem of Steinberg says that for every automorphism φ of the elementary Chevalley group G=Φ⁢(F) there exists an inner automorphism φg, a diagonal automorphism φh, a graph automorphism ρ¯ and a field automorphism δ¯ such that φ=ρ¯⁢δ¯⁢φh⁢φg (see [30]).

3 Chevalley groups

In this section we extend the following result from [43, Theorem 1].

Theorem 3.1

Let G be a Chevalley group of type Φ over a field F of zero characteristic and the transcendence degree of F over Q is finite. Then:

If Φ has one of the types Al(l≥7), Bl(l≥4), E8, F4, G2, then G possesses the R∞-property.
If the equation Tk=a can be solved in the field F for every element a, then G possesses the R∞-property also for the root systems Al(l=2,3,4,5,6), Bl(l=2,3), Cl(l≥3), Dl(l≥4), E6, E7, where k is a positive integer from the following table.

In particular, this theorem says that if F is an algebraically closed field of zero characteristic such that the transcendence degree of F over ℚ is finite, then a Chevalley group of any normal type over the field F possesses the R∞-property.

Here we exclude the condition of solvability of equations from the second item of Theorem 3.1. We prove the following result.

Theorem 3.2

Let G be a Chevalley group of type Φ over a field F of zero characteristic. If the transcendence degree of F over Q is finite, then G possesses the R∞-property.

Proof.

Since G/Z⁢(G)≅Φ⁢(F), by Lemma 2.2 it is sufficient to prove that the elementary Chevalley group Φ⁢(F) possesses the R∞-property. It suffices to assume that G=Φ⁢(F).

Let us consider an arbitrary automorphism φ of the group G and prove that the number of φ-conjugacy classes is infinite. By the theorem of Steinberg there exists an inner automorphism φg, a diagonal automorphism φh, a graph automorphism ρ¯ and a field automorphism δ¯ such that

φ=ρ¯⁢δ¯⁢φh⁢φg.

By Lemma 2.1 the Reidemeister number R⁢(φ) is infinite if and only if the Reidemeister number R⁢(φ⁢φg-1) is infinite, and we can consider that φ=ρ¯⁢δ¯⁢φh.

Suppose that R⁢(φ)<∞ and consider the following elements of the group G:

gi=hα1⁢(pi⁢1)⁢hα2⁢(pi⁢2)⁢…⁢hαl⁢(pi⁢l),i=1,2,…,

where p11<p12<…<p1⁢l<p21<p22<⋯ are prime numbers. In matrix representation the element gi has diagonal form

gi=diag⁡(ai⁢1,ai⁢2,…,ai⁢|Φ|,1,…,1⏟|Φ|),

for certain rational numbers ai⁢j such that ν⁢(ai⁢j)≠∅ and ν⁢(ai⁢j)∩ν⁢(ar⁢s)=∅ for i≠r since ν⁢(ai⁢j)⊆{pi⁢1,…,pi⁢l} (see [43]).

Since R⁢(φ)<∞, there exists an infinite subset of φ-conjugated elements in the set g1,g2,…. Without loss of generality we can assume that all the elements g1,g2,… belong to the φ-conjugacy class [g1]φ of the element g1. Then there exists an infinite set of matrices Z2,Z3,… from G such that

g1=Zi⁢gi⁢φ⁢(Zi-1),i=2,3,…⁢.

Acting on these equalities by iterations of the automorphism φ we have

g1=Zi⁢gi⁢φ⁢(Zi-1),

φ⁢(g1)=φ⁢(Zi)⁢φ⁢(gi)⁢φ2⁢(Zi-1),

φ2⁢(g1)=φ2⁢(Zi)⁢φ2⁢(gi)⁢φ3⁢(Zi-1),

⁢⋮

φ5(g1)=φ5(Zi)φ5(gi)φ6(Zi-1). i=2,3,….

If we multiply all of these equalities, we conclude that

(3.1)g1⁢φ⁢(g1)⁢…⁢φ5⁢(g1)=Zi⁢gi⁢φ⁢(gi)⁢…⁢φ5⁢(gi)⁢φ6⁢(Zi-1).

Since the matrix gi has a diagonal form and the automorphism φh acts as conjugation by the diagonal matrix, we have φh⁢(gi)=gi. Since the matrix gi has rational entries, it follows that δ¯⁢(gi)=gi and therefore φ⁢(gi)=ρ¯⁢(gi). If we denote g~i=gi⁢φ⁢(gi)⁢…⁢φ5⁢(gi)=gi⁢ρ¯⁢(gi)⁢…⁢ρ¯5⁢(gi), then

g~i=diag⁡(bi⁢1,bi⁢2,…,bi⁢|Φ|,1,…,1⏟l),i=1,2,…,

since ρ¯ permutes diagonal elements of the matrix gi. Moreover, ν⁢(bi⁢j)≠∅ and ν⁢(bi⁢j)∩ν⁢(br⁢s)=∅ for i≠r, since ν⁢(bi⁢j)⊆ν⁢(ai⁢1)∪…∪ν⁢(ai⁢|Φ|).

Since graph and field automorphisms commute and diagonal automorphisms form a normal subgroup in the group, which is generated by graph, field and diagonal automorphisms, for a certain diagonal automorphism φh~ we have

φ6=(ρ¯⁢δ¯⁢φh)6=φh~⁢δ¯6⁢ρ¯6.

Since the order of the automorphism ρ¯ is equal to 2 or to 3, it follows that ρ¯6=id and φ6=φh~⁢δ¯6. Then equality (3.1) can be rewritten

g~1=Zi⁢g~i⁢φ6⁢(Zi-1)=Zi⁢g~i⁢φh~⁢δ¯6⁢(Zi-1)=Zi⁢g~i⁢h~⁢δ¯6⁢(Zi-1)⁢h~-1,i=2,3,…⁢.

If we multiply this equality by the element h~ on the right and denote g^i=g~i⁢h~, then we have

(3.2)g^1=Zi⁢g^i⁢δ¯6⁢(Zi-1),i=2,3,…⁢.

From this equality we have

(3.3)δ¯6⁢(Zi)=g^1-1⁢Zi⁢g^i,i=2,3,…⁢.

If we denote

h~=diag⁡(c1,c2,…,c|Φ|,1,…,1⏟l),

then

g^i=g~i⁢h~=diag⁡(bi⁢1⁢c1,bi⁢2⁢c2,…,bi⁢|Φ|⁢c|Φ|,1,…,1⏟l),i=2,3,…⁢.

Let

Zi=([c|c]QiRiSiTi),

where Qi=(qi,m⁢n) is a |Φ|×|Φ| matrix, Ri=(ri,m⁢n) is a |Φ|×|Δ| matrix, Si=(si,m⁢n) is a |Δ|×|Φ| matrix, Ti=(ti,m⁢n) is a |Δ|×|Δ| matrix. Then by equality (3.3) for all m=1,…,|Φ|, n=1,…,|Φ| we have

δ6⁢(qi,m⁢n)=(b1⁢m⁢cm)-1⁢bi⁢n⁢cn⁢qi,m⁢n=dm⁢n⁢bi⁢n⁢qi,m⁢n,i=2,3,…,

where dm⁢n=(b1⁢m⁢cm)-1⁢cn. Since ν⁢(bi⁢n)≠∅ and ν⁢(bi⁢n)∩ν⁢(bj⁢n)=∅ for all i,j with i≠j, we can apply Lemma 2.5 to the set q2,m⁢n,q3,m⁢n,…. Therefore by Lemma 2.5 there exists a positive integer Nm⁢n such that qi,m⁢n=0 for every i>Nm⁢n.

If we denote by N the value

N=maxn,m=1,…,|Φ|⁡Nm⁢n,

then for every i>n we have Qi=O|Φ|×|Φ|. Using the same arguments for the matrices S2,S3,…, we conclude that for sufficiently large indexes i all the matrices Si are matrices with zero entries only, and therefore the matrix Zi has the form

Zi=(O|Φ|×|Φ|RiO|Δ|×|Φ|Ti).

The determinant of this matrix is equal to zero, therefore Zi cannot belong to G. This contradiction proves the theorem.∎

4 Linear algebraic groups

If G is a linear algebraic group over an algebraically closed field, then it has a unique maximal solvable normal subgroup R⁢(G), called the radical of G. A connected linear algebraic group G is called reductive if its radical is a torus, or, equivalently, if it can be decomposed G=G′⁢T′ with G′ a semisimple group and T′ a central torus [52, Section 6.5].

The quotient group G/R⁢(G) has a trivial radical, i.e. is a semisimple group [31, Section 19.5].

Theorem 4.1

Proof.

For the group G we have the following short exact sequence of groups:

1→R⁢(G)→G→G/R⁢(G)→1.

Since G is reductive, the radical R⁢(G) is a central torus and therefore is a characteristic subgroup of G. Hence by Lemma 2.2 it is sufficient to prove that the semisimple group G/R⁢(G) possesses the R∞-property and we can assume that G is a semisimple linear algebraic group. Every semisimple linear algebraic group is a product, with some amalgamation of (finite) centers, of its simple subgroups H1,H2,…,Hk (see [31, Section 14.2])

G=H1∘…∘Hk.

Every simple linear algebraic group Hi is a Chevalley group of (normal) type Φi over the field F. Factoring the group G by its center we have the following short exact sequence of groups:

1→Z⁢(H1∘…∘Hk)→H1∘…∘Hk→Φ1⁢(F)×…×Φk⁢(F)→1,

where Φi⁢(F) is the elementary Chevalley group of type Φi over the field F. Hence, by Lemma 2.2 we can assume that G=Φ1⁢(F)×⋯×Φk⁢(F) and prove that this group possesses the R∞-property. Permute the groups Φ1⁢(F),…,Φk⁢(F) so that all groups with the same root system form blocks

G=Φ1⁢(F)×⋯×Φ1⁢(F)⏟k1×Φ2⁢(F)×⋯×Φ2⁢(F)⏟k2×⋯×Φr⁢(F)×⋯×Φr⁢(F)⏟kr,

where k1+k2+⋯+kr=k. Denote by

Gi=Φi⁢(F)×⋯×Φi⁢(F)⏟ki.

Every group Gi is a characteristic subgroup of G=G1×⋯×Gr. Therefore by Lemma 2.2 it is sufficient to prove that some group Gi possesses the R∞-property. Thus we can consider that G=Φ⁢(F)×…×Φ⁢(F)=Φ⁢(F)k.

Every element g∈G=Φ⁢(F)k can be presented as a direct sum of k matrices g1,…,gk of the size (|Φ|+|Δ|)×(|Φ|+|Δ|) each of which belongs to Φ⁢(F).

The automorphism group of G has the form

(4.1)Aut⁢(G)=(Aut⁢(Φ⁢(F)))k⋋Sk,

where Sk is the full permutation group on k symbols.

To prove that the group G=Φ⁢(F)k possesses the R∞-property, consider an arbitrary automorphism φ of the group G and we will prove that R⁢(φ)=∞. By equality (4.1) the automorphism φ can be written in the form

φ=(φ1,…,φk,σ),

where φ1,…,φk∈Aut⁢(Φ⁢(F)), σ∈Sk, and φ acts on the group G by the rule

(4.2)φ:x1⊕x2⊕…⊕xk↦φ1σ⁢(x1σ)⊕φ2σ⁢(x2σ)⊕…⊕φkσ⁢(xkσ),

where iσ denotes the image of i by the permutation σ.

Every automorphism φi∈Aut⁢(Φ⁢(F)) can be presented as a product of an inner automorphism φgi, a diagonal automorphism φhi, a graph automorphism ρ¯i and a field automorphism δ¯i. Since F is an algebraically closed field, every diagonal automorphism φhi is inner [43, Lemma 4], hence for every i we can assume that φi=φxi⁢ρ¯i⁢δ¯i. Then the automorphism φ can be presented as a product of two automorphisms

φ=(φx1σ,φx2σ,…,φxkσ,id)⁢(ρ¯1⁢δ¯1,ρ¯2⁢δ¯2,…,ρ¯k⁢δ¯k,σ),

where (φx1σ,φx2σ,…,φxkσ,id) is an inner automorphism. By Lemma 2.1 we can consider that φi=ρ¯i⁢δ¯i and

φ=(ρ¯1⁢δ¯1,ρ¯2⁢δ¯2,…,ρ¯k⁢δ¯k,σ).

Using induction on r, we will prove that

(4.3)φr:g1⊕…⊕gk↦ψ1⁢(x1σr)⊕…⊕ψk⁢(xkσr),

where ψi=φiσ⁢φiσ2⁢…⁢φiσr.

The basis of induction (r=1) is obvious (equality (4.2)). If we suppose that equality (4.3) holds for some r, then

φr+1⁢(g1⊕…⊕gk)=φ⁢(φr⁢(g1⊕…⊕gk))

=φ⁢(ψ1⁢(x1σr)⊕…⊕ψk⁢(xkσr))

=φ1σ⁢ψ1σ⁢(x1σr+1)⊕…⊕φkσ⁢ψkσ⁢(xkσr+1).

Noting that φiσ⁢ψiσ=φiσ⁢φiσ2⁢…⁢φiσr+1 we obtain equality (4.3).

Consider the set of elements g1,g2,… of the group Φ⁢(F) from Theorem 3.2

gi=hα1⁢(pi⁢1)⁢hα2⁢(pi⁢2)⁢…⁢hαl⁢(pi⁢l),i=1,2,…,

where p11<p12<…<p1⁢l<p21<p22<⋯ are prime integers. These elements are presented by diagonal matrices

gi=diag⁡(ai⁢1,ai⁢2,…,ai⁢|Φ|,1,…,1⏟l),i=1,2,…,

where ai⁢j are rational numbers such that ν⁢(ai⁢j)≠∅ and ν⁢(ai⁢j)∩ν⁢(ar⁢s)=∅ for i≠r.

As already shown in Theorem 3.2, for every automorphism φj=ρ¯j⁢δ¯j we have φj⁢(gi)=ρ¯j⁢(gi).

Let us consider the set of elements g~1,g~2,… of the group G=Φ⁢(F)k, where g~i=gi⊕…⊕gi. Then by the arguments above

φ⁢(g~i)=ρ¯1σ⁢(gi)⊕…⊕ρ¯kσ⁢(gi).

Suppose that R⁢(φ)<∞. Then there is an infinite subset of φ-conjugated elements in the set g~1,g~2,…. Without loss of generality we can consider that all matrices g~1,g~2,… belong to the φ-conjugacy class [g~1]φ of the element g~1. Then for certain matrices Z2,Z3,… we have

g~1=Zi⁢g~i⁢φ⁢(Zi-1),i=2,3,…⁢.

Denote by s the order of the permutation σ and act on this equality by iterates of the automorphism φ

g~1=Zi⁢g~i⁢φ⁢(Zi-1),

φ⁢(g~1)=φ⁢(Zi)⁢φ⁢(g~i)⁢φ2⁢(Zi-1),

⋮

φ6⁢s-2⁢(g~1)=φ6⁢s-2⁢(Zi)⁢φ6⁢s-2⁢(g~i)⁢φ6⁢s-1⁢(Zi-1),

φ6⁢s-1⁢(g~1)=φ6⁢s-1⁢(Zi)⁢φ6⁢s-1⁢(g~i)⁢φ6⁢s⁢(Zi-1).

If we multiply all of these equalities, we obtain the following equality:

(4.4)g~1⁢φ⁢(g~1)⁢φ2⁢(g~1)⁢…⁢φ6⁢s-1⁢(g~1)=Zi⁢g~i⁢φ⁢(g~i)⁢φ2⁢(g~i)⁢…⁢φ6⁢s-1⁢(g~i)⁢φ6⁢s⁢(Zi-1).

The element g~i⁢φ⁢(g~i)⁢φ2⁢(g~i)⁢…⁢φ6⁢s-1⁢(g~i) can be rewritten in details as

g~i⁢φ⁢(g~i)⁢φ2⁢(g~i)⁢…⁢φ6⁢s-1⁢(g~i)=(gi⊕…⊕gi)⁢(ρ¯1σ⁢(gi)⊕…⊕ρ¯kσ⁢(gi))

…⁢(ρ¯1σ6⁢s-1⁢(gi)⊕…⊕ρ¯kσ6⁢s-1⁢(gi))

=gi⁢ρ¯1σ⁢(gi)⁢…⁢ρ¯1σ6⁢s-1⁢(gi)⊕…⊕gi⁢ρ¯kσ⁢(gi)

…⁢ρ¯kσ6⁢s-1⁢(gi)

=g^i⁢1⊕…⊕g^i⁢k,

where g^i⁢j=gi⁢ρ¯jσ⁢(gi)⁢…⁢ρ¯jσ6⁢s-1⁢(gi).

Since every graph automorphism ρ¯j permutes elements on the diagonal of the matrix gi, we have for every j=1,…,k and i=1,2,… that

(4.5)g^i⁢j=diag⁡(bi⁢j⁢1,bi⁢j⁢2,…,bi⁢j⁢|Φ|,1,…,1⏟l),

where ν⁢(bi⁢j⁢r)≠∅ and ν⁢(bi⁢j⁢r)∩ν⁢(bu⁢v⁢w)=∅ for all i,u with i≠u since ν⁢(bi⁢j⁢r) is a subset of {pi⁢1,…,pi⁢l}.

From (4.3) we have φs=(ψ1,ψ2,…,ψk,id), where ψi=φiσ⁢φiσ2⁢…⁢φiσr. Since all automorphisms φ1,φ2,…,φk are products of graph and field automorphisms (φi=ρ¯i⁢δ¯i) and graph and field automorphisms commute, it follows that every automorphism ψi is a product of graph and field automorphisms ψi=ξ¯i⁢θ¯i for certain ξ¯i, θ¯i. Therefore

φ6⁢s=(φs)6=(ξ¯1⁢θ¯1,…⁢ξ¯k⁢θ¯k,id)6=(ξ¯16⁢θ¯16,…⁢ξ¯k6⁢θ¯k6,id)=(θ¯16,…⁢θ¯k6,id).

Using this fact, denoting by Zi=Zi⁢1⊕…⊕Zi⁢k projecting equality (4.4) to the first group Φ⁢(F) we obtain the equality

g^11=Zi⁢1⁢g^i⁢1⁢θ¯16⁢(Zi⁢1),i=2,3,…⁢.

This equality is the same as equality (3.2) from Theorem 3.2. Using the same arguments as in Theorem 3.2, we conclude that for sufficiently large number N the matrix ZN is degenerate but this contradicts the fact that ZN belongs to G.∎

We use the fact that the group G is a reductive linear algebraic group in order to say that the radical R⁢(G) is a characteristic subgroup of G. Even Theorem 4.1 holds for every connected linear algebraic group such that the radical R⁢(G) is a characteristic subgroup. For example, if every automorphism of the group G is a morphism of the group G (as in the case of an affine manifold), then the radical R⁢(G) is characteristic [52, Theorem 7.1 (c)] and Theorem 4.1 holds for such groups.

5 Finite Reidemeister number in linear groups

Following [47], we define the Reidemeister spectrum ofG as

Spec⁡(G)={R⁢(φ):φ∈Aut⁢(G)}.

In particular, G possesses the R∞-property if and only if Spec⁡(G)={∞}.

It is easy to see that Spec⁡(ℤ)={2}∪{∞}, and, for n≥2, the spectrum of ℤn is full, i.e. Spec⁡(ℤn)=ℕ∪{∞}. For free nilpotent groups we have the following:

Spec⁡(N22)=2⁢ℕ∪{∞} (N22 is the discrete Heisenberg group) ([32, 14, 47]),
Spec⁡(N23)={2⁢k2:k∈ℕ}∪{∞} ([47]),
Spec⁡(N32)={2⁢k-1:k∈ℕ}∪{4⁢k:k∈ℕ}∪{∞} ([47]).

Recently, in [5, 47] it was proven that the group Nr⁢c(r>1) admits an automorphism with finite Reidemeister number if and only if c<2⁢r.

In [24], examples of polycyclic non-virtually nilpotent groups which admit automorphisms with finite Reidemeister numbers have been described. In these examples G is the semidirect product of ℤ2 and ℤ by the Anosov automorphism defined by the matrix (2111). The group G is solvable and of exponential growth. The automorphism φ with finite Reidemeister number is defined by (01-10) on ℤ2 and as -id on ℤ.

Metabelian (therefore, solvable) finitely generated, non-polycyclic groups have quite interesting Reidemeister spectrum [12]: for example, if the homomorphism θ:ℤ→Aut⁢(ℤ⁢[1p]2) is such that θ⁢(1)=(r00s), then we have the following cases:

If r=s=±1, then
Spec⁡(ℤ⁢[1/p]2⋊θℤ)={2⁢n:n∈ℕ,(n,p)=1}∪{∞},
where (n,p) denotes the greatest common divisor of n and p.
If r=-s=±1, then
Spec⁡(ℤ⁢[1/p]2⋊θℤ)={2⁢pl⁢(pk±1),4⁢pl:l,k>0}∪{∞}.
If r⁢s=1 and |r|≠1, then
Spec⁡(ℤ⁢[1/p]2⋊θℤ)={2⁢(pl±1),4:l>0}∪{∞}.
If either r or s does not equal to ±1, and r⁢s≠1, then
Spec⁡(ℤ⁢[1/p]2⋊θℤ)={∞}.

In [33] Jabara proved that if a residually finite group G admits an automorphism of prime order p with finite Reidemeister number, then G is a virtually nilpotent group of class bounded by a function of p.

On the other hand, we have described in the Introduction (Section 1) a lot of classes of non-virtually-solvable, finitely generated, residually finite groups which have the R∞-property. Taking these ideas together was a motivation for the following conjecture.

Conjecture 5.1

Conjecture 5.1 ([18, Conjecture R])

Every residually finite, finitely generated group either possesses the R∞-property or is a virtually solvable group.

Here we study this question for infinite linear groups.

Proposition 5.2

Let G be a reductive linear algebraic group over a field F of zero characteristic and finite transcendence degree over Q. If G possesses an automorphism φ with finite Reidemeister number, then G is a torus.

Proof.

Since G possesses an automorphism φ with finite Reidemeister number, by Theorem 4.1, the quotient group G/R⁢(G) is trivial, therefore G=R⁢(G) and hence G is a central torus (therefore, is solvable).∎

6 Groups with property S∞

Let Ψ∈Out⁢(G)=Aut⁢(G)/Inn⁢(G), and let 𝔖⁢(Ψ) be the set of isogredience classes of Ψ. Then 𝔖⁢(Id) can be identified with the set of conjugacy classes of G/Z⁢(G) (see [18]).

The definition of similarity (isogredience) from Section 1 goes back to Jacob Nielsen. He observed (see [34]) that conjugate lifting of homeomorphisms of surfaces have similar dynamical properties. This led Nielsen to the definition of the isogredience of liftings in this case. Later Reidemeister and Wecken succeeded in generalizing the theory to continuous maps of compact polyhedra (see [34]).

The set of isogredience classes of automorphisms representing a given outer automorphism and the notion of index Ind⁡(Ψ) defined via the set of isogredience classes are strongly related to important structural properties of Ψ (see [19]).

One of the main results of [38] is that for any non-elementary hyperbolic group and any Ψ the set 𝔖⁢(Ψ) is infinite, i.e. S⁢(Ψ)=∞. Thus, this result says: any non-elementary hyperbolic group possesses the S∞-property. On the other hand, finitely generated abelian groups are evidently non-S∞-groups.

Two representatives of Ψ have forms φs⁢a, φq⁢a for some s,q∈G and fixed a∈Ψ. They are isogredient if and only if

φq⁢a=φg⁢φs⁢a⁢φg-1=φg⁢φs⁢φa⁢(g-1)⁢a,

φq=φg⁢s⁢a⁢(g-1),q=g⁢s⁢a⁢(g-1)⁢c,c∈Z⁢(G)

(see [38, p. 512]). So, the following statement is proved.

Lemma 6.1

Lemma 6.1 ([18, Lemma 3.3])

Let φ∈Ψ be an automorphism of the group G and let φ¯ be an automorphism of the group G/Z⁢(G) which is induced by φ. Then the number S⁢(Ψ) is equal to the number of φ¯-conjugacy classes in the group G/Z⁢(G).

Since Z⁢(G) is a characteristic subgroup, we obtain the following statement.

Theorem 6.2

Theorem 6.2 ([18, Theorem 3.4])

Let G be a group with finite center. Then G possesses the R∞-property if and only if G possesses the S∞-property.

A more advanced example of a non-S∞-group is Osin’s group [46]. This is a non-residually finite exponential growth group with two conjugacy classes. Since it is simple, it cannot possess the S∞-property (see [18]).

Theorem 6.3

Proof.

Since R⁢(G) is a characteristic subgroup of G, by Lemma 2.3 it is sufficient to prove the theorem for semisimple group G/R⁢(G). The result follows immediately from Theorem 4.1 and Theorem 6.2 and from the fact that a semisimple linear algebraic group has finite center. ∎

Proposition 6.4

Let G be a reductive linear algebraic group over a field F of zero characteristic and finite transcendence degree over Q. If G possesses an outer automorphism Ψ with finite number S⁢(Ψ), then G is a torus.

Proof.

Since G possesses an outer automorphism Ψ with finite number S⁢(Ψ), it follows from Theorem 6.3 that it has a trivial quotient G/R⁢(G), thus G=R⁢(G) and is a central torus. ∎

The conjecture of Fel’shtyn and Troitsky from Section 5 can be rewritten in terms of the S∞-property by the following way.

Conjecture 6.5

Conjecture 6.5 ([18, Conjecture S])

Every residually finite, finitely generated group either possesses the S∞-property or is a virtually solvable group.

Really, if S⁢(φ⁢Inn⁢(G))<∞ for some automorphism φ∈Aut⁢(G), then by Lemma 6.1 we have R⁢(φ¯)<∞, where φ¯ is an automorphism of the group G/Z⁢(G) induced by φ. Since G is a residually finite finitely generated group, G/Z⁢(G) is also finitely generated and residually finite and by Conjecture 5.1 is a virtually solvable group.

This means that there exists a solvable subgroup H¯≤G/Z⁢(G) of finite index. Let n be a derived length of H¯, i.e. H¯=(n)1. Let H be the full preimage of H¯ under the canonical homomorphism G→G/Z⁢(G). Then H(n)≤Z⁢(G) and H(n+1)=1, therefore H is a solvable group. Since

G/H≃(G/Z⁢(G))/(H/Z⁢(G))=(G/Z⁢(G))/H¯,

the index of H in G is equal to the index of H¯ in G/Z⁢(G), i.e. is finite, therefore G is a virtually solvable group.

In all, we have proven that [18, Conjecture S] can be formulated without the restriction that the group under consideration has finite center.

Communicated by Evgenii I. Khukhro

Funding source: Russian Science Foundation

Award Identifier / Grant number: 14-21-00065

Funding statement: The second author is supported by Russian Science Foundation (project 14-21-00065).

Acknowledgements

The authors are grateful to Andrzej Da̧browski, Evgenij Troitsky and Evgeny Vdovin for the numerous important discussions on linear algebraic groups. The first author would like to thank the Max Planck Institute for Mathematics (Bonn) for its kind support and hospitality while a part of this work was completed.

References

[1] Arthur J. and Clozel L., Simple Algebras, Base Change, and the Advanced Theory of the Trace Formula, Princeton University Press, Princeton, 1989. 10.1515/9781400882403Search in Google Scholar

[2] Bleak C., Fel’shtyn A. and Gonçalves D., Twisted conjugacy classes in R. Thompson’s group F, Pacific J. Math. 238 (2008), no. 1, 1–6. 10.2140/pjm.2008.238.1Search in Google Scholar

[3] Burillo J., Matucci F. and Ventura E., The conjugacy problem in extensions of Thompson’s group F, preprint 2013, http://arxiv.org/abs/1307.6750. 10.1007/s11856-016-1403-9Search in Google Scholar

[4] Carter R., Simple Groups of Lie Type, Wiley, London, 1989. Search in Google Scholar

[5] Dekimpe K. and Gonçalves D., The R∞ property for free groups, free nilpotent groups and free solvable groups, Bull. Lond. Math. Soc. 46 (2014), no. 4, 737–746. 10.1112/blms/bdu029Search in Google Scholar

[6] Dekimpe K. and Penninckx P., The finiteness of the Reidemeister number of morphisms between almost-crystallographic groups, J. Fixed Point Theory Appl. 9 (2011), no. 2, 257–283. 10.1007/s11784-011-0043-2Search in Google Scholar

[7] Fel’shtyn A., Dynamical zeta functions, Nielsen theory and Reidemeister torsion, Mem. Amer. Math. Soc. 147 (2000), no. 699. 10.4064/-49-1-77-116Search in Google Scholar

[8] Fel’shtyn A., The Reidemeister number of any automorphism of a Gromov hyperbolic group is infinite, Zap. Nauchn. Sem. S.-Peterburg. Otdel. Mat. Inst. Steklov. (POMI) 279 (2001), no. 6, 229–240. 10.1023/B:JOTH.0000008749.42806.e3Search in Google Scholar

[9] Fel’shtyn A., New directions in Nielsen–Reidemeister theory, Topology Appl. 157 (2010), no. 10–11, 1724–1735. 10.1016/j.topol.2010.02.018Search in Google Scholar

[10] Fel’shtyn A. and Gonçalves D., The Reidemeister number of any automorphism of a Baumslag–Solitar group is infinite, Geometry and Dynamics of Groups and Spaces, Progr. Math. 265, Birkhäuser, Basel (2008), 399–414. 10.1007/978-3-7643-8608-5_9Search in Google Scholar

[11] Fel’shtyn A. and Gonçalves D., Twisted conjugacy classes in symplectic groups, mapping class groups and braid groups, Geom. Dedicata 146 (2010), 211–223. 10.1007/s10711-009-9434-6Search in Google Scholar

[12] Fel’shtyn A. and Gonçalves D., Reidemeister spectrum for metabelian groups of the form Qn⋊ℤ and ℤ⁢[1/p]n⋊ℤ, p prime, Internat. J. Algebra Comput. 21 (2011), no. 3, 505–520. 10.1142/S0218196711006297Search in Google Scholar

[13] Fel’shtyn A. and Hill R., The Reidemeister zeta function with applications to Nielsen theory and a connection with Reidemeister torsion, K-Theory 8 (1994), no. 4, 367–393. 10.1007/BF00961408Search in Google Scholar

[14] Fel’shtyn A., Indukaev F. and Troitsky E., Twisted Burnside theorem for two-step torsion-free nilpotent groups, C*-Algebras and Elliptic Theory II, Trends Math., Birkhäuser, Basel (2008), 87–101. 10.1007/978-3-7643-8604-7_4Search in Google Scholar

[15] Fel’shtyn A., Leonov Y. and Troitsky E., Twisted conjugacy classes in saturated weakly branch groups, Geom. Dedicata 134 (2008), 61–73. 10.1007/s10711-008-9245-1Search in Google Scholar

[16] Fel’shtyn A. and Troitsky E., Twisted Burnside–Frobenius theory for discrete groups, J. Reine Angew. Math. 614 (2007), 193–210. 10.1515/CRELLE.2007.097Search in Google Scholar

[17] Fel’shtyn A. and Troitsky E., Twisted conjugacy classes in residually finite groups, preprint 2012, http://arxiv.org/abs/1204.3175. Search in Google Scholar

[18] Fel’shtyn A. and Troitsky E., Aspects of the property R∞, J. Group Theory 18 (2015), no. 6, 1021–1034. 10.1515/jgth-2015-0022Search in Google Scholar

[19] Gaboriau D., Jaeger A., Levitt G. and Lustig M., An index for counting fixed points of automorphisms of free groups, Duke Math. J. 93 (1998), no. 3, 425–452. 10.1215/S0012-7094-98-09314-0Search in Google Scholar

[20] Gantmacher F., Canonical representation of automorphisms of a complex semi-simple Lie group, Rec. Math. (Moscow) 47 (1939), no. 5, 101–146. Search in Google Scholar

[21] Gonçalves D. and Hristova-Kochloukova D., Sigma theory and twisted conjugacy classes, Pacific J. Math. 247 (2010), no. 2, 335–352. 10.2140/pjm.2010.247.335Search in Google Scholar

[22] Gonçalves D. and Sankaran P., Twisted conjugacy in Richard Thompson’s group T, preprint 2013, http://arxiv.org/abs/1309.2875. Search in Google Scholar

[23] Gonçalves D. and Sankaran P., Sigma theory and twisted conjugacy-II: Houghton’s groups and pure symmetric automorphism groups, preprint 2014, http://arxiv.org/abs/1412.8048. 10.2140/pjm.2016.280.349Search in Google Scholar

[24] Gonçalves D. and Wong P., Twisted conjugacy classes in exponential growth groups, Bull. Lond. Math. Soc. 35 (2003), no. 2, 261–268. 10.1112/S0024609302001832Search in Google Scholar

[25] Gonçalves D. and Wong P., Twisted conjugacy classes in wreath products, Internat. J. Algebra Comput. 16 (2006), no. 5, 875–886. 10.1142/S0218196706003219Search in Google Scholar

[26] Gonçalves D. and Wong P., Twisted conjugacy classes in nilpotent groups, J. Reine Angew. Math. 633 (2009), 11–27. 10.1515/CRELLE.2009.058Search in Google Scholar

[27] Grigorchuk R., On Burnside’s problem on periodic groups, Funct. Anal. Appl. 14 (1980), 41–43. 10.1007/BF01078416Search in Google Scholar

[28] Grothendieck A., Formules de Nielsen–Wecken et de Lefschetz en géométrie algébrique, Séminaire de Géométrie Algébrique du Bois–Marie 1965–1966. SGA 5, Lecture Notes in Math. 569, Springer, Berlin (1977), 407–441. 10.1007/BFb0096811Search in Google Scholar

[29] Gupta N. and Sidki S., On the Burnside problem for periodic groups, Math. Z. 182 (1983), 385–388. 10.1007/BF01179757Search in Google Scholar

[30] Humphreys J., On the automorphisms of infinite Chevalley groups, Can. J. Math. 21 (1969), no. 4, 606–615. 10.4153/CJM-1969-099-7Search in Google Scholar

[31] Humphreys J., Linear Algebraic Groups, Springer, New York, 1975. 10.1007/978-1-4684-9443-3Search in Google Scholar

[32] Indukaev F., The twisted Burnside theory for the discrete Heisenberg group and for the wreath products of some groups (in Russian), Vestnik Moskov. Univ. Ser. I Mat. Mekh. 71 (2007), no. 6, 9–17; translation in Moscow Univ. Math. Bull.62 (2007), no. 6, 219–227. 10.3103/S0027132207060022Search in Google Scholar

[33] Jabara E., Automorphisms with finite Reidemeister number in residually finite groups, J. Algebra 320 (2008), 3671–3679. 10.1016/j.jalgebra.2008.09.001Search in Google Scholar

[34] Jiang B., Lectures on Nielsen Fixed Point Theory, Contemp. Math. 14, American Mathematical Society, Providence, 1983. 10.1090/conm/014Search in Google Scholar

[35] Jo J. H., Lee J. B. and Lee S. R., The R∞ property for Houghton’s groups, preprint 2014, http://arxiv.org/abs/1412.8767. Search in Google Scholar

[36] Juhász A., Twisted conjugacy in certain Artin groups, Ischia Group Theory 2010, World Scientific, Hackensack (2011), 175–195. 10.1142/9789814350051_0014Search in Google Scholar

[37] Levitt G., On the automorphism group of generalised Baumslag–Solitar groups, Geom. Topol. 11 (2007), 473–515. 10.2140/gt.2007.11.473Search in Google Scholar

[38] Levitt G. and Lustig M., Most automorphisms of a hyperbolic group have very simple dynamics, Ann. Sci. Éc. Norm. Supér. (4) 33 (2000), 507–517. 10.1016/S0012-9593(00)00120-8Search in Google Scholar

[39] Lutowski R. and Szczepański A., Holonomy groups of flat manifolds with R∞ property, Fund. Math. 223 (2013), no. 3, 195–205. 10.4064/fm223-3-1Search in Google Scholar

[40] Mubeena T. and Sankaran P., Twisted conjugacy classes in abelian extensions of certain linear groups, Canad. Math. Bull. 57 (2014), 132–140. 10.4153/CMB-2012-013-7Search in Google Scholar

[41] Mubeena T. and Sankaran P., Twisted conjugacy classes in lattices in semisimple Lie groups, Transform. Groups 19 (2014), no. 1, 159–169. 10.1007/s00031-014-9249-xSearch in Google Scholar

[42] Nasybullov T., Twisted conjugacy classes in general and special linear groups, Algebra Logic 51 (2012), no. 3, 220–231. 10.1007/s10469-012-9185-6Search in Google Scholar

[43] Nasybullov T., Twisted conjugacy classes in Chevalley groups, Algebra Logic 53 (2015), no. 6, 481–501. 10.1007/s10469-015-9310-4Search in Google Scholar

[44] Nasybullov T., The R∞-property for Chevalley groups of types Bl, Cl, Dl over integral domains, J. Algebra 446 (2016), 489–498. 10.1016/j.jalgebra.2015.09.030Search in Google Scholar

[45] Onishchik A. and Vinberg È., Lie Groups and Algebraic Groups, Springer Series in Sov. Math., Springer, Berlin, 1990. 10.1007/978-3-642-74334-4Search in Google Scholar

[46] Osin D., Small cancellations over relatively hyperbolic groups and embedding theorems, Ann. of Math. (2) 172 (2010), no. 1, 1–39. 10.4007/annals.2010.172.1Search in Google Scholar

[47] Roman’kov V., Twisted conjugacy classes in nilpotent groups, J. Pure Appl. Algebra 215 (2011), no. 4, 664–671. 10.1016/j.jpaa.2010.06.015Search in Google Scholar

[48] Seitz G., Topics in the theory of algebraic groups, Group Representation Theory (Lausanne 2005), Fund. Sci. Math., CRC Press, Boca Raton (2007), 355–404. Search in Google Scholar

[49] Serre J., Galois Cohomology, Springer Monogr. in Math., Springer, Berlin, 2002. Search in Google Scholar

[50] Shokranian S., The Selberg–Arthur Trace Formula, Lecture Notes in Math. 1503, Springer, Berlin, 1992. 10.1007/BFb0092305Search in Google Scholar

[51] Springer T., Twisted conjugacy in simply connected groups, Transform. Groups 11 (2006), no. 3, 539–545. 10.1007/s00031-005-1113-6Search in Google Scholar

[52] Steinberg R., Endomorphisms of linear algebraic groups, Mem. Amer. Math. Soc. 80 (1968). 10.1090/memo/0080Search in Google Scholar

[53] Taback J. and Wong P., Twisted conjugacy and quasi-isometry invariance for generalized solvable Baumslag–Solitar groups, J. Lond. Math. Soc. (2) 75 (2007), no. 3, 705–717. 10.1112/jlms/jdm024Search in Google Scholar

Received: 2015-6-17

Published Online: 2016-2-5

Published in Print: 2016-9-1

Articles in the same Issue

https://doi.org/10.1515/jgth-2016-0004