MathOverflow is a question and answer site for professional mathematicians. Join them; it only takes a minute:

Sign up
Here's how it works:
  1. Anybody can ask a question
  2. Anybody can answer
  3. The best answers are voted up and rise to the top

Let $R$ be a ring of characteristic $p$. Let $G$ be the kernel of the natural map $\pi_1^{et} (\operatorname{Spec} R[[x]] [1/x]) \to \pi_1^{et}( \operatorname{Spec} R)$. $G$ has a natural map to $\prod_{l\neq p} \mathbb Z_l(1)$, coming from the etale coverings adjoining the $n$th roots of $x$ for $n$ prime to $p$. Is the kernel of this natural map always a pro-$p$ group?

For $R$ a field, this is standard. But I'm not sure when $R$ is not a field.

share|cite|improve this question
If $R$ is a henselian regular local ring (in particular, noetherian) then this follows from Abyhankar's Lemma. The next case to wonder about is a normal noetherian domain (maybe henselian), since then "normalization" makes sense and $x$-adic separatedness and completeness is inherited by finitely generated $R[[x]]$-modules (as $R$ is noetherian). Perhaps try to understand that before the case of more general $R$. – user30379 Apr 4 '13 at 4:25
Can you elaborate on the Abyhankar's Lemma argument a bit? As far as I know it's just about local rings of dimension one. I did not mean to imply, in asking the general question, that I was not interested in special cases. I've been thinking mostly about the normal case. For the application I'm considering, regular and Noetherian are fine assumptions, but local is problematic. – Will Sawin Apr 4 '13 at 5:24
Whoops, the argument I had in mind only shows (via Abyhankar's Lemma) that the kernel has no nontrivial prime-to-$p$ quotients, which is of course much weaker than being pro-$p$, so I have nothing useful to say about your question. More specifically, assuming regularity and excellence but not "henselian local", the maximal prime-to-$p$ quotient of the kernel is the quotient $\prod_{\ell} {\mathbf{Z}}_{\ell}(1)$ that you know. Please talk to experts near you in Princeton to learn about Abhykanar's Lemma beyond the 1-dimensional case (as higher dimensions is where the real strength of it lies). – user30379 Apr 5 '13 at 4:34
Or read SGA 1 XIII section 5 to learn more about Abhyankar's lemma, then talk to experts. – user31960 Apr 5 '13 at 13:50
up vote 0 down vote accepted

The answer is no.

Take $R=\mathbb F_2[y]$ and consider the cover defined by the equation $t^3+yt+x=0$. By computing the discriminant, this cover is etale. The inertia group of this cover is a subgroup of $S_3$. By reducing mod $(y)$, it contains a $3$-cycle. By reducing mod $y-1+x^2$, the equation factors into $(t^2-xt+1)(t+x)$, so the inertia group contains a $2$-cycle. Thus the inertia group is full $S_3$.

So $S_3$ is a quotient of $G$. But clearly if $G$ has the property I describe, every finite quotient of it must have a normal $2$-Sylow subgroup, which $S_3$ does not.

share|cite|improve this answer

Your Answer


By posting your answer, you agree to the privacy policy and terms of service.

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