MathOverflow is a question and answer site for professional mathematicians. It's 100% free, no registration required.

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

First I want to review some concept from quadratic form. Let $V$ be quadratic space over finite field $F$ and $char(F)\neq 2$ with quadratic form $q$. For exmaple

$q:V\rightarrow V$ and $|F|=q$ and $\dim V=n$.

We may assume the vectors of $V$ are represented as column vectors $v=(a_1,b_1,…)\in V$, and that $q$ takes one of the three forms




or if we want to have matrix representation we have




Finally for subspace $W$ we have $W^{\perp}=\{v\in V\mid b(v,w)=0 ,\forall w \in W \}$ where $b(x,y)=(\frac{1}{2})[q(x+y)-q(x)-q(y)]$.

The problem I am interested in :

Compute the number of $k$-dimensional subspaces $W$ such that $W\subseteq W^{\perp}$; this subspace is called totally isotropic subspace (T,I). For $k=1$, it is prominent theorem. That comes back to find solution of these equations



3.$q(v)={x_1}^2-{y_1}^2+…+{x_k}^2-{y_k}^2-{x_{k+1}}^2=0$ So the number of 1-dimensional (T,I) is




Now what can I do for $k\geq 2$.

share|cite|improve this question
evidently from page79 classic group and geometric algebra by Larry Grove,… – Will Jagy Jul 19 '12 at 5:21
anyway, copied from… – Will Jagy Jul 19 '12 at 5:27
up vote 0 down vote accepted

Let $q$ denote the order of the finite field, which assume has odd characteristic. If $Q$ is a non-degenerate quadratic form in $n$ variables, for $n$ odd (this is your case 3), we have that the number of totally isotropic subspaces of dimension $k$ (for $k \leq \frac{n-1}{2}$, otherwise the number is $0$) is $$ \frac{\prod_{i=0}^{k-1} (q^{2n-2i-1 } -1)}{\prod_{i=1}^{k} (q^i -1)}.$$

If the number of variables, $n$, is even then the answer depends on the signature of the quadratic form. Let $\epsilon = \{+1,-1\}$ denote this signature. The number of totally isotropic subspaces of dimension $k$ is then

$$ \frac{(q^{n-k} - \epsilon q^{n/2-k} + \epsilon q^{n/2} -1) \prod_{i=1}^{k-1} (q^{n-2i} -1 ) }{\prod_{i=1}^{k} (q^i -1) }.$$

Again, we assume that $k\leq n/2-1$ if $\epsilon =-1$ and $k \leq n/2$ if $\epsilon = +1$ (otherwise the number is $0$). Note that $\epsilon=+1$ corresponds to your case $1$ and $\epsilon=-1$ corresponds to your case $2$.

A short proof of these identities can be found in:

V. Pless, The number of isotropic subspaces in a finite geometry. Atti Accad. Naz. Lincei Rend. CI. Sci. Fic. Mat. Natur. (8) 39 1965 418-421.

although I believe these results were known previously.

share|cite|improve this answer

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.