4
$\begingroup$

Let $V$ be a vector space of dimension $>n$, and define the subset $$ K:=\{ ([\omega],v)\mid v\wedge\omega=0 \}\subset\mathbb{P}(\Lambda^nV)\times V\, . $$ Denote also by $\pi:K\longrightarrow \mathbb{P}(\Lambda^nV)$ the restriction of the canonical projection onto the first factor of $\mathbb{P}(\Lambda^nV)\times V$. Then, $\pi$ is a linear bundle over $\mathbb{P}(\Lambda^nV)$ with variable fibre dimension.

QUESTION 1: is it true that $$\{[\omega]\in \mathbb{P}(\Lambda^nV)\mid \dim\pi^{-1}([\omega])=n\}\quad\quad\quad (^*)$$ is precisely the Plucker image of the Grassmanniann $\mathrm{G}(n,V)$?

In the affirmative case, I'd like to know whether such a "bundle" $\pi$ ever appeared, perhaps in other guises, and ask the next

QUESTION 2: what if I replace $n$ with an arbitrary $i\neq n$ in the above $(^*)$? What do I obtain?

I'm expecting some subset of $\mathbb{P}(\Lambda^nV)$ intrinsically related to $\mathrm{G}(n,V)$, like its tangent/secant variety.

(I'm aware of another MO question, concerning the Plucker embedding and the tautological bundle, and another one, concerning its ring of functions, though neither can help here.)

Improve this question
$\endgroup$
2
  • 2
    $\begingroup$ The answer to your first question is yes (you can see this e.g. by taking taking a basis $v_i$ of $V$ with $v_1,v_2,\cdots v_n$ a basis of $\pi^{-1}([\omega])$ and writing $\omega$ in terms of the induced basis of $\wedge^nV.$) $\endgroup$
    – dhy
    Jul 30, 2015 at 6:04
  • $\begingroup$ Ok, let me try. If $[\omega]\in \mathrm{G}(n,V)$, then $\omega=v_1\wedge\cdots\wedge v_n$, and $\pi^{-1}([\omega])=\langle v_1,\ldots,v_n\rangle$, which is the easy implication. Conversely, knowing that $\dim \pi^{-1}([\omega])=n$, you choose a basis of $V$ such that $\{v_1,\ldots,v_n\}$ generates $\pi^{-1}([\omega])$: now you suggest to write down $\omega$ as a linear combination of $v_1\wedge\cdots\wedge v_n$ and other terms, and show that the coefficients in front of the latter must necessarily be zero. Yes, it should work, thanks! Do you have any clue about question #2? $\endgroup$
    – Giovanni Moreno
    Jul 30, 2015 at 17:13

0

Reset to default

Your Answer

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