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

let $M_n(c)$ denote the n times n matrices over the complex number field. $N$ be a subspace of


1 If there is no unitary lies in $N$, what is the maximum of the dimension of $N$ can be?

It's easy to see that it is not less than n(n-1), I guess it's also tight, but I don't know if I am correct.

2 If all the rank of $M$ lies in $N$ are greater than a fixed integer $k$, what is the maximum of the dimension of $N$ can be?

share|cite|improve this question
Very closely related question: – Scott Morrison Mar 30 '10 at 15:43
MathOverflow has very good $\text{\LaTeX}$ support. So there's really no excuse to have bad formatting. This looks cut-and-paste from something, what with the random line break in the first line. – Theo Johnson-Freyd Mar 30 '10 at 21:59
As I think I've said on a previous comment, I personally have no problem for leaving some LaTeX unformatted. I think that we should beware giving an impression that the visual presentation of a question is what's important here. – Yemon Choi Mar 30 '10 at 22:30
Sorry, this is my first problem... I think these questions are interesting, and guess they might be considered before, but I don't find... – gondolf Mar 31 '10 at 6:51

In $M_n({\mathbb R})$, I studied Question 2, with $k=2$ (subspaces whose matrices are not of rank $1$) in: Formes quadratiques et calcul des variations. J. Math. Pures Appl. (9) 62 (1983), no. 2, 177--196. See a short version in: Condition de Legendre-Hadamard; espaces de matrices de rang $\not=1$. C. R. Acad. Sci. Paris Sér. I Math. 293 (1981), no. 1, 23--26. I don't remember whether my study was specific to the scalar field $\mathbb R$.

share|cite|improve this answer
It's probably easier if the field is $\mathbb{C}$. For example take $k=n$ so the question is about subspaces such that any nonzero matrix is invertible. You see that any $2$-diml space of matrices contains noninvertible matrices (because if $A,B$ are two invertible matrices, then $\det(A+X \cdot B)$ has roots in $\mathbb{C}$). However, the corresponding problem over $\mathbb{R}$ is much harder, and is related to the maximal number of linearly independent vector fields on a sphere, which involves $K$-theory. See for details. – Laurent Berger Sep 21 '10 at 11:44

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.