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

1) I was wondering if there exists a criterion (of (a) combinatorial or (b) geometric nature) for a sum of simple vectors $V\in\wedge^k(\mathbb R^n)$,

$V=e_{a_{11}}\wedge\cdots\wedge e_{a_{1k}} + \cdots + e_{a_{m1}}\wedge\cdots\wedge e_{a_{mk}}$, where $e_1,\cdots, e_n$ is a basis of $\mathbb R^n$

to be a simple vector: in other words, I want to know what should one verify in order to be sure that $V$ can be written as $V=v_1\wedge\cdots\wedge v_k$ for some vectors $v_i\in\mathbb R^n$.

2) What happens if we allow noninteger coefficients? In other words, how does the answer change if we consider question 1) for

$V=c_1e_{a_{11}}\wedge\cdots\wedge e_{a_{1k}} + \cdots + c_me_{a_{m1}}\wedge\cdots\wedge e_{a_{mk}}$, where now $c_i\in\mathbb R$.

3) any reference or hint about some (name of) branch of mathematics dealing with this is also welcome!

share|cite|improve this question
up vote 13 down vote accepted

You are looking for the Plucker relations, a collection of quadratic polynomials in the $c_m$'s which vanish if and only if $V$ can be written as $v_1 \wedge v_2 \wedge \cdots \wedge v_k$. You should be able to read about them in most books on algebraic geometry. For example, Griffiths-Harris or Miller-Sturmfels will definitely discuss this.

If you are comfortable with representation theory of $GL_n$, they can be described very briefly: Let $\omega_k$ be the weight of the highest weight vectors in $\bigwedge^k(\mathbb{R}^n)$. There is (up to scalar multiple) a unique $GL_n$ equivariant map from $\bigwedge^k(\mathbb{R}^n)^{\otimes 2}$ to the representation of $GL_n$ with highest weight $2 \omega_k$. Let the kernel of this map be $K$. The plucker relations say that $V \otimes V$ is in $K$.

It is not difficult to write them down explicitly, but there are a lot of signs to get right, so I'll just refer you to the sources cited above.

There is no significant simplification in assuming that the $c$'s are $0$ or $1$.

Since you tagged this combinatorics and hyperplane arrangements, you may want to know the following: If $V$ is simple, then the set of $k$-tuples $(a_{m1}, \ldots, a_{mk})$ for which $c_m$ is nonzero form the bases of a matroid. There is no useful converse to this statement. Essentially by definition, a matroid $M$ is realizable if and only if you can find vectors $v_1$, $v_2$, ..., $v_k$ such that the bases of $M$ correspond to the nonzero terms in $v_1 \wedge v_2 \wedge \cdots \wedge v_k$. But there is no good criterion for a matroid to be realizable.

share|cite|improve this answer
Thanks for the answer! Actually I found the question from the theory of calibrations, I don't know representation theory. Matroid theory is a little more familiar. Actually the straightforward conditions one would have to impose on the $v_i$ would reduce to some conditions on the subdeterminants (is it possible that these are connected to "Plucker's relations"?) of the matrix $(v_i^j)_{i=1}^k_{j=1}^n$ (where $v_i^j$ is the component of $v_i$ along $e_j$). Anyway, in terms of representation theory this looks more manageable, I will try to study that! – Mircea Apr 26 '10 at 13:38

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.