3
$\begingroup$

Let $K$ be the set of open-closed subsets of $\mathbb{Z}_p$. Let $M$ be the set of functions from $K$ to $C_p$ that are additive under disjoint unions. Then $M$ can be regarded as an elementary abelian pro-$p$ group: multiplication is pointwise in $C_p$, and a base of open subgroups is given by $\{ U_n \}$, where $U_n$ consists of those functions which map cosets of $p^n \mathbb{Z}_p$ to $0$. Moreover, $\mathbb{Z}_p$ has a translation action on $K$, and hence a continuous action on $M$ in which each of the $U_n$ is normalised. We use this action to construct a group $G = M \rtimes \mathbb{Z}_p$. This is realisable as an inverse limit of the groups $C_p \wr C_{p^n}$ (regular wreath product); in particular, it is a $2$-generator pro-$p$ group.

Now for the question: Is the group $G$ above isomorphic to a group arising from some standard construction, and if so does it have a nice name? I would be surprised if nobody has used this group before as an example of something or other. Also, there seems to be quite a general construction behind this.

Edit: the group described is a subgroup of the pro-$p$ group $C_p \wr_K \mathbb{Z}_p$. The unusual part is the extra condition that the functions should be additive (motivation: I wanted $G$ to have relatively few normal subgroups). Are there alternative conditions one could impose that would result in an isomorphic group? In particular, I'm relying on the fact that $C_p$ is abelian, which is bad for generalisations.

$\endgroup$
  • $\begingroup$ Is $C_p$ the cyclic group of order $p$? $\endgroup$ – Victor Protsak Jul 27 '10 at 18:05
  • $\begingroup$ Yes. The same construction would also work with any (profinite) abelian group in place of $C_p$. $\endgroup$ – Colin Reid Jul 29 '10 at 8:52
1
$\begingroup$

I would call this group the pro-$p$ completion of $C_p \wr \mathbb{Z}$. Alternatively, it is the group given by the pro-$p$ presentation <$ a, b| a^p, [a, a^b]>$, along similiar lines.

It looks a lot like $C_p\wr \mathbb{Z}_p$, except we have the product of as many copies of $C_p$ as open subgroups of $\mathbb{Z}_p$ rather than elements. This does seem a natural, interesting, general construction, and unless I am mistaken is a generalistion of the pro-$p$ wreath product. (This particular example is nice especially by virtue of having a balanced pro-$p$ presentation).

$\endgroup$
  • $\begingroup$ Yes, that sounds about right. It certainly contains a dense copy of $C_p \wr \mathbb{Z}$ (restricted wreath product), but I'm not sure if it contains a copy of the unrestricted wreath product. $\endgroup$ – Colin Reid Jul 29 '10 at 9:09
  • $\begingroup$ This pro-$p$-presentation $G=\langle a,b\mid a^p,[a,a^b]\rangle$ clearly does not work. Indeed, first consider the subgroup of index $p$ kernel of the map to $C_p$ mapping $(a,b)\mapsto (1,0)$, and write $c=a^p$, $b_i=t^ibt^{-i}$. It has the presentation $\langle c,b_1,\dots, b_p\mid b_i^p,[b_i,b_{i+1}],[b_p,tb_1t^{-1}]\rangle$. The quotient by the generators $b_2,\dots,b_p$ yields the presentation $\langle c,b_1\mid b_1^p\rangle$ of the free pro-$p$ product of $C_p$ and $\mathbf{Z}_p$. So $G$ is certainly not metabelian. $\endgroup$ – YCor Oct 22 '17 at 17:30
0
$\begingroup$

It appears to be a $p$-adic analog of the lamplighter group.

$\endgroup$
  • 1
    $\begingroup$ Indeed, but I suspect there are several claimants to the name of 'lamplighter group' in the world of pro-$2$ groups. $\endgroup$ – Colin Reid Jul 29 '10 at 9:27

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.