MathOverflow is a question and answer site for professional mathematicians. Join them; it only takes a minute:

Sign up
Here's how it works:
  1. Anybody can ask a question
  2. Anybody can answer
  3. The best answers are voted up and rise to the top

We say $L< (V\oplus V^{*})\bigotimes \mathbb{C}$ is isotropic when $< X,Y>=0$ for all $X,Y\in L$

Why $O(4n,\mathbb{C})$ (orthogonal group) acts transitively on the space of maximal isotropics of $V\bigotimes \mathbb{C}$ ?

(here $V$ is a vector space of finite dimention $2n$)

share|cite|improve this question
You need to explain what you mean by $V$. Is it $V = \mathbb{R}^{4n}$? – Robert Bryant Dec 6 '12 at 13:17
I'm afraid that just saying that $V$ is a vector space of finite dimension doesn't give us enough information. You have to tell us how you think $\mathrm{O}(4n,\mathbb{C})$ is acting on $V\otimes\mathbb{C}$. Obviously, this transitivity won't hold for any finite dimensional $V$ and any action of $\mathrm{O}(4n,\mathbb{C})$ on $V\otimes\mathbb{C}$. What are you not telling us? – Robert Bryant Dec 6 '12 at 13:48
I revised my question – baba ab dad Dec 6 '12 at 15:14
Would your question be more simply phrased as "How can one prove that $\mathrm{O}(4n,\mathbb{C})$ acts transitively on the maximal isotropic subspaces of $\mathbb{C}^{4n}$?" What I'm getting at is whether it's important for you that your complex vector space be written as $V\otimes\mathbb{C}$ for some (I presume) real vector space $V$ of dimension $2n$ over $\mathbb{R}$. If the 'real structure' is not important to you, then the above version of your question would be adequate. (By the way, in this case, you'd have transitivity for all $n$, so I'm not sure why you are assuming $4n$.) – Robert Bryant Dec 6 '12 at 16:33
In fact maximal isotropic $L$ corresponds to a generalized complex structure on $V$ and so the dimension of $V$ have to be even, – baba ab dad Dec 7 '12 at 11:14
up vote 2 down vote accepted

I think a more general statement can be proved along the following line:

Let $V= \mathbb R^n$. Then on $W:= V\oplus V^*$ the symmetric bilinear form $((v,v^*)|(w,w^*)) = \langle w^*,v\rangle + \langle v^*, w\rangle$ has signature $(n,n)$. Now we are quite similar to a symplectic vector space. Given isotropic $L$, choose a basis $w_1,\dots w_n$ of $L$. Then $(w_i|w_j)=0$ for all $i,j$. We can extend this to a basis $w_1,\dots,w_n, w^1,\dots,w^n$ of $W$ such that $(w_i,w_j)=0$, $(w^i,w^j)=0$ for all $i,j$ and $(w_i,w^j)=\delta_i^j$. Then $w^1,\dots, w^n$ spans a complementary isotropic subspace. The group $O(n,n,\mathbb R)$ acts transitively on the set of all such bases. Thus it acts transitively on the set of pairs of complementary isotropic subspaces. Thus also transitively on the set of isotropic subspaces. Now $O(2n,\mathbb C)$ is the complexification of $O(n,n,\mathbb R)$.

share|cite|improve this answer
And this idea can be extended to give a proof of Witt's Theorem, that in a non-degenerate "formed space" (quadratic form, hermitian form, symplectic, etc.) an isometry from one subspace to another extends to an isometry of the whole space. – paul garrett Dec 6 '12 at 17:41
Dear Paul, if possible for you please more explain your second comment – baba ab dad Dec 8 '12 at 1:42

Your Answer


By posting your answer, you agree to the privacy policy and terms of service.

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