0
$\begingroup$

Let $X$ be a Brauer–Severi variety over $k$.

I understood that the automorphism scheme $\mathrm{Aut}_{X/k}=G$ as a group scheme acting on $X$ via $$(m,pr_2):G\times X \rightarrow X\times X$$ (the multiplication morphism), and that this morphism is surjective.

Questions:

What about the action of $\mathrm{Aut}_{X/k}(k)$ on $X$?

Is there for example a transitive action of $\mathrm{Aut}_{X/k}(k)$ on the closed points of $X$? If not, what kind of assumption is needed to get this?

$\endgroup$
7
  • $\begingroup$ It is clear that there cannot be a transitive action on the closed points in general: the isomorphism class of the residue field, in particular also degree, of a closed point must be preserved. $\endgroup$
    – naf
    Nov 19, 2012 at 8:38
  • $\begingroup$ I see, but supposing this is the case what can we say? $\endgroup$
    – Peter
    Nov 19, 2012 at 8:56
  • $\begingroup$ So suppose there are closed points x and y such that the residue fields are isomorphis as k-algebras does there exsist an element of Aut_{X/k}(k) mapping x to y? $\endgroup$
    – Peter
    Nov 19, 2012 at 9:13
  • $\begingroup$ @Peter: Have you tried looking at what happens for the simplest non-trivial Brauer-Severi varieties, namely conics which do not contain a rational point? $\endgroup$ Nov 19, 2012 at 9:29
  • $\begingroup$ Or you could even consider the action on $\mathbb{P}^1$, say over a finite field, and then counting will show that the action is not transitive. $\endgroup$
    – naf
    Nov 19, 2012 at 14:33

0

Your Answer

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

Browse other questions tagged or ask your own question.