0
$\begingroup$

Let $Y$ be a reduced noetherian $1$-dimensional scheme such that the normalization morphism $f:X \longrightarrow Y$ is finite. Let $g:Y\longrightarrow Z$ be a finite flat morphism, where $Z$ is a connected (1-dimensional) Dedekind scheme.

Suppose that the morphism $g\circ f$ from $X$ to $Z$ is etale.

Question. Is the morphism $g:Y\longrightarrow Z$ etale?

Remark. The hypothesis on the dimension is not necessary probably.

Remark. One may assume $Y$ to be integral.

Remark. To assure that $f$ is finite one may suppose that $Y$ is excellent.

Interesting cases one may consider are $Z\subset \mathrm{Spec} \mathbf{Z}$ or $Z\subset \mathbf{P}^1_k$ non-empty open.

$\endgroup$
1
$\begingroup$

If $g$ is étale, then $Y$ is regular because $Z$ is regular. Thus $X=Y$. So the anwser to your question is yes if and only if $X=Y$.

$\endgroup$
  • $\begingroup$ If $Y$ is not normal then can the hypothesis of the question ever be true? i.e. does there ever exist $Z$ and some map $Y\to Z$ as in the question such that the induced map $X\to Z$ is etale with $X$ the normalization? $\endgroup$ – Kevin Buzzard Jul 27 '11 at 17:43
  • 2
    $\begingroup$ @Kevin: Yes, start with an étale double cover $X\to Z$ with a fibre consisting of two points. Construct $Y$ as the nodal "curve" with these two points identified. $\endgroup$ – Torsten Ekedahl Jul 27 '11 at 19:00
  • 1
    $\begingroup$ @Kevin: Yes: $Y=$ union of the two diagonals in the plane, $Z=$ the $x$-axis, map= the $x$-projection. $\endgroup$ – Laurent Moret-Bailly Jul 27 '11 at 19:03
  • 1
    $\begingroup$ @Torsten: You beat me by 3 minutes on this one. $\endgroup$ – Laurent Moret-Bailly Jul 27 '11 at 19:05
  • $\begingroup$ Very nice -- thanks to both of you. $\endgroup$ – Kevin Buzzard Jul 28 '11 at 20:43

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.