2
$\begingroup$

Suppose I have a morphism of Noetherian schemes over a field $k$ (if one needs this then assume $k$ is algebraically closed) $f:C'\rightarrow S$ which is finite with geometrically connected and reduced fibers. Is $f$ an isomorphism?

$\endgroup$
2
  • $\begingroup$ Note that this is true if $S$ is normal, by Zariski's main theorem. $\endgroup$ Nov 26, 2012 at 8:07
  • $\begingroup$ Also note that if you do not assume that $S$ is surjective, then any closed immersion gives a counterexample (Jason Starr's answer is a closed immersion but you can see that he assumed that you wanted the morphism to be surjective). $\endgroup$ Nov 26, 2012 at 8:10

3 Answers 3

9
$\begingroup$

No: $\text{Spec} k[\epsilon]/\langle \epsilon \rangle$ mapping to $\text{Spec} k[\epsilon]/\langle \epsilon^2 \rangle$.

$\endgroup$
3
$\begingroup$

Though any closed embedding will give a counter example, it is true if you ask whether it is an isomorphism to the (scheme theoretic) image. In other words, the map is an isomorphism from $C'\to f(C')$ where $f(C')$ is thought of as the scheme theoretic image (which makes sense since the map is assumed to be finite), with $k$ algebraically closed.

$\endgroup$
2
$\begingroup$

Or, perhaps more to the point, the map from the normalization to a cusp on a curve. Note that such a map is an isomorphism on points without being an isomorphism of schemes.

$\endgroup$
2
  • 2
    $\begingroup$ Fiber above the cusp is not reduced. $\endgroup$
    – R.P.
    Nov 26, 2012 at 9:39
  • $\begingroup$ Yes. You are right. I was thinking only about the isomorphism on points, not the fiber! $\endgroup$ Nov 28, 2012 at 17:24

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

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