3
$\begingroup$

Is there a generalization of Borel-Weil-Bott for partial flag varieties, i.e. homogeneous spaces of the form $G/P$ with $P$ parabolic and $G$ semisimple? If so, I would like a reference.

$\endgroup$
7
  • 7
    $\begingroup$ Doesn't this follow from the usual Borel-Weil-Bott by noting that the (derived) pushforward from $G/B$ to $G/P$ of a line bundle pulled back from $G/P$ is again that line bundle? $\endgroup$
    – Will Sawin
    Commented Nov 9, 2020 at 21:54
  • $\begingroup$ @WillSawin How does that help? $\endgroup$ Commented Nov 9, 2020 at 22:01
  • 2
    $\begingroup$ The cohomology of any equivariant line bundle on $G/P$ is equal to the cohomology of its pullback to $G/B$, which can be calculated using the usual Borel-Weil-Bott, by the Leray spectral sequence. $\endgroup$
    – Will Sawin
    Commented Nov 9, 2020 at 22:10
  • 1
    $\begingroup$ Section 2.3 of arxiv.org/abs/1203.2575 and references therein should be useful. $\endgroup$
    – Jef
    Commented Nov 9, 2020 at 22:48
  • 3
    $\begingroup$ @AviSteiner Because the fibers are connected (and have no coherent cohomology, if you're interested in the derived pushforward). This is a local question, so we can work locally on $G/P$, and in particular assume the line bundle is trivial - it's a question about $f_* \mathcal O_{G/B}$. We know this is trivial for a proper separable map with reduced connected fibers. $\endgroup$
    – Will Sawin
    Commented Nov 10, 2020 at 12:14

1 Answer 1

1
$\begingroup$

There are many places that give a complete answer to your question: One is a paper in the Annals of Math written by Kostant around the middle fifties, other more geometrical is due to Griffits-Schmid published in Acta Mathematica in the late sixties. best regards

$\endgroup$

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .