A
growth dichotomy
for group
algebras
of free abelian
by
infinite
cyclic
groups
Christopher Briggs
Embry‐Riddle
Aeronautical
University
email:briggscl@erau.edu
Abstract
We
study growth
in
algebras, especially
uniform
exponential growth, extending
historical results
in thetopic
of
growth
in groups. Wegive
acondition under
which the
groupalgebra
ofafreeabelian
by
infinite
cyclic
grouphas uniform
exponential growth.
1
Introduction
Givena
finitely
generated
group G andafinitegenerating
setS=\{g_{1}^{\pm 1}, . . . , g_{T}^{\pm 1}\}
ofG,for eachelement g\in G, write
g=g_{i_{1}}^{n}\ldots g_{i_{k}}^{n_{k}}1
.Define thelength
ofgwithrespecttoStobethe minimalnonnegative
integer
k for which suchanexpression
ofgispossible
(the identity
is consideredtoUeanempty word,hence
oflength
0).
Thegrowthfunction
$\gamma$_{G,S}(n)
:\mathbb{Z}_{\geq 0}\rightarrow \mathbb{N}
ofG withrespecttoS is the number of elements of G oflength
atmostnwithrespecttoS.Wewrite$\gamma$_{G,S}(n)=$\gamma$_{S}(n)
ifG is understood. The
growth being exponential
orboundedby
apolynomial
isindependent
of thechosen
generating
set, sowespeak
of groups ofexponential
orpolynomial growth
respectively.
\mathrm{A}group which is neither is saidtohave intermediate
growth.
Growth in groupswasintroduced
independently
Uy
Schwarz in 1955 and Milnor in 1968[8,
9].
In 1968 Wolfproved
thatavirtualy nilpotent
group haspolynomial
growth,
andavirtually
polycyclic
group which is notvirtualy nilpotent
hasexponential growth
[11].
In 1981 Gromovproved
thatafinitely generated
group ofpolynomial
growth
isvirtually nilpotent [6].
In 1968 Milnor asked whether there exists groups of intermediate
growth.
Thequestion
wasanswered
affirmatively by
Grigorchuk
in 1983 [5]. In 1981 Gromov defineda group Gtohaveuniform exponential
growth
if\displaystyle \inf_{S}(\lim_{n}$\gamma$_{S}(n)^{1/n})
> 1. Inthe sameyear Gromov asked: if Ghas
exponential growth,
mustit have uniformexponential
growth [7]?
Thequestion
wasanswerednegatively by
Wilson 2004[10].
Inthemeantime,thequestion
wasansweredpositively
for severalclasses of groups. In2002
Alperin
proved
thatavirtually
polycyclic
group has eitherpolynomial
oruniform
exponential growth
[1].
In2005Eskin,
Moses, and Ohproved
alineargroupover afieldofcharacteristiczerohas
polynomial
oruniformexponential growth
[4].
In 2008Breuillardand Gelander
proved
thatalinear group of any fieId haspolynomial
oruniformexponential
growth
[2].
数理解析研究所講究録
2
Growth
in
algebras
Write
A=F[S]
todenote that A isgenerated (as
analgebra)
by
thesetS overthefield F. Thegrowth
function ofA withrespectto S is$\gamma$_{A,S}(n)
=\displaystyle \dim_{F}(\sum_{i=0}^{n}FS^{i})
. Growthtypeisindepen‐
dent of the
generating
set, soaswithgroups,wedefinepolynomial, exponential,
and intermediategrowth
foralgebras.
Also as with groups, acommutativealgebra
haspolynomial growth,
andthe free
algebra
on atleasttwoletters hasexponential growth.
We sayanalgebra
A hasuniform
exponential growth
if\displaystyle \inf_{S}$\gamma$_{A,S}(n)^{\frac{1}{n}}>1.
Thereare
examples
ofalgebras
of uniformexponential growth:
Golod‐Shafarevichalgebras,
groupalgebras
of Golod‐Shafarevich groups, and anyalgebra graded
by
\mathbb{N} withexponential growth
haveuniforn
exponential
growth
[3].
Anexample
ofanalgebra
of nonuniformexponential
growth
isthe group
algebra
of Wilsons groupof nonuniformexponential growth
over afieldof characteristic0.
We are
particularly
interested in thegrowth
of groupalgebras,
as FG hasexponential,
poly‐
nomial,
orintermediategrowth
if andonly
ifG doesrespectively.
IfFG has uniformexponential
growth,
sodoes G. Itisunclear, and it isamotivating question
of the authorsresearch,
whethertheconverseistrue.
3
Main result
The main resultconcerns acondition under whichthe group
algebra
ofafreeabelianby
infinitecyclic
group has uniformexponential growth.
The desired end result is an extensionto groupalgebras
ofAlperins
result thatapolycyclic
group has eitherpolynomial
oruniformexponential
growth.
The free abelianby
infinitecyclic
groupisabuilding
block of the poycyclic
group, soitmakessense to
begin Uy
attempting
toestablish thatagroupalgebra
ofafree abelianby
infinitecyclic
group is ofpolynomial
oruniformexponential
growth.
Henceforth,
let$\Gamma$=G\rangle\triangleleft_{ $\sigma$}\mathbb{Z}
beafree abelianby
infinitecyclic
group. LetF beafield. Sincethe action of $\sigma$on GcanUedescribed
by
amatrix,we canrefertotheeigenvalues
of $\sigma$.By
aresultofAlperin [1],
combined with the fact that F $\Gamma$ haspolynomial growth
if andonly
if $\Gamma$does,
wehavethe resultTheorem 1
Ifall eigenvalues of
$\sigma$havenorm1,then thegroupalgebra
F $\Gamma$ haspolynomial
growth.
The
following simple
result is often useful:Lemma 2
Uniform exponential growth
lifts
from
homomorphic
images
for
groups andalgebras
[31
\cdotThe main result is:
Theorem 3
If
$\sigma$hasarealeigenvalue
$\lambda$ with| $\lambda$|>1
,then F $\Gamma$ hasuniform exponential growth.
Theyetunsolvedcasetoestablish the
growth dichotomy (polynomial
oruniformexponential)
for free abelian
by
infinitecyclic
groupsis thecasewhen alleigenvalues
arecomplex
butnotalleigenvalues
havenorm1.References
[1]
R.Alperin,
Uniform exponential growth of polycyclic
groups, Geom. Dedicata 92(2002)
105−113[2]
E. Breuillard andT. Gelander,Uniform
independence for
lineargroups, Invent.Math. 173(2008)
no.2,225−263[3]
C.Briggs, Examples of Uniform Exponential
Growth inAtgebras,
J.Algebra
Appl.,
DOI:http:
//\mathrm{d}\mathrm{x}.doi.org
/10.1142/\mathrm{S}0219498817502413[4]
A. Eskin, S Mozes and H. Oh, Onuniform exponential growth for
lineargroups, Invent.Math. ló0
(2005)
no. 1,1−30[5]
R.I.Grigorchuk,
On Milnorsproblem of
groupgrowth,
Soviet. Math. Dokl. 28(1983
\rangle23−26[6]
M. Gromov,Groups
of polynomial growth
andexpanding
maps, Inst. HautesÉtudes
Sci.Publ. Math. 53
(1981)
53−73[7] M.
Gromov,
Structuresmetriques
pourlesvarietesriemanniennes,CEDIC,
Paris,1981,
MR85e:5305l
[8]
J.Milnor,
Growthinfinitely generated
solvablegroups, J. Diff. Geom. 2(1968)
447−449[9]
A. S.Schwarz,
A volume invariantofcoverings,
Dokl. Ak. Nauk USSR 105(1955)
32−34[10]
J.Wilson,
Onexponential growth
anduniformly exponential growth
for
groups, Proc. Amer.Math. Soc.60
(1976)
22−24[11]
J.Wolf,
Growthoffinitely generated
solvablegroupsandcurvatureofRiemannian manifolds,
J. Diff. Geom. 2