Let $V$ be a vector space of dimension $n$. Let $S^k V$ be a representation of $GL(n)$. I would like to know if there exists some characterization of finite dimensional $GL(n)$ modules $V_1,V_2$ such that exists $GL(n)$ mapping $T:S^k V \rightarrow V_1\otimes V_2$ such that for some $x$ matrix $T(x)$ is invertible .

Added: Just to clarify: The mapping $T$ itself may not be invertible I am asking that $T(x)$ will be invertible in $V_1\otimes V_2$. For example the mapping $S^2 V \rightarrow V\otimes V$ is not invariable, but $T(x)$ is invartable for most $x$'s.

I am interested for a base field $\mathbb{C}$. And of course this may happen only in case when $S^kV$ is a sub-representation of $V_1 \otimes V_2$. In fact I do not know for which irreps $V_1, V_2$ representation $V_1 \otimes V_2$ have $S^kV$ as a component. I will be happy if you can give me a reference for this question.

  • $\begingroup$ If you work over a field of characteristic $0$, then the category of polynomial representations is semisimple (I assume that you are talking about polynomial representations), and there exists a non-zero $T$ if and only if $V_1\otimes V_2$ has a simple summand isomorphic to a simple summand of $S^kV$. Now such a $T$ will never take an invertible value unless $S^kV$ is isomorphic to $V_1\otimes V_2$, so really, you are asking whether $S^kV$ can be isomorphic to a tensor product of two representations. $\endgroup$ Jan 11, 2013 at 6:31
  • $\begingroup$ Dear Prasad, Thanks for your comment, but I am not asking for $T$, but for $T(x)$ to be invertible in $V_1\otimes V_2$. $\endgroup$ Jan 11, 2013 at 9:11
  • 2
    $\begingroup$ @Klim: what does it mean for an element of $V_1\otimes V_2$ to be invertible? Are $V_1$ and $V_2$ of the same dimension? What do you mean by invariable in the example $S^2V\to V\otimes V$? $\endgroup$ Jan 11, 2013 at 9:32
  • $\begingroup$ @Dragos "invariable" should be a misprint for invertible. $\endgroup$ Jan 11, 2013 at 10:59

1 Answer 1


Regarding the question of pairs, here is the answer (hope, I'm not mistaken).

Let me first change the notation a little bit. Suppose, we are looking for irreps $V_1=\Sigma^{\lambda}V$ and $V_2=\Sigma^{\mu}V$ where $\lambda = (\lambda_1,\ldots,\lambda_n)$ and $\mu = (\mu_1,\ldots,\mu_n)$ resp., such that there exists an embedding $S^kV\to V_1^*\otimes V_2$.

It's immediate that such an embedding exists iff $V_2=\Sigma^{\mu}V$ is an irreducible factor in $V_1\otimes S^kV=\Sigma^{\lambda}V\otimes S^kV$. The Littlewood-Richardson rule gives us the answer. Firstly, $k=|\mu|-|\lambda|$, where $|\lambda| = \sum_i\lambda_i$. Secondly, one has the following inequalities:

  • $\mu_1\geq \lambda_1$,
  • $\lambda_1\geq \mu_2 \geq \lambda_2$,
  • $\ldots$
  • $\lambda_{n-1}\geq \mu_n \geq \lambda_n$.

Finally, if such an embedding exists, it's unique by the very same L-R rule.

UPD: Forgot to mention that the main question still seems to be very hard as even checking that $\dim V_1=\dim V_2$ is a huge problem.

UPD 2: Let me answer your main question (despite UPD). Let's forget about representation theory and do some linear algebra. What we have is a surjective map $V_1\otimes U\to V_2$, where $\dim V_1 = \dim V_2$. Let the corresponding bilinear map be $h:V_1\times U\to V_2$. We want to find such $u\in U$ that $h_u = h(-, u)$ is invertible, equivalently, of maximal rank.

Well, this seems to be quite obvious: let $r$ be the maximal rank of $h_u$ among all $u\in U$. Suppose $r<\dim V_2$. Let $u_0\in U$ be such that $\mathrm{rk}\ h_{u_0}=r$. Take any $v\in V_2\setminus\mathrm{Im}\ h_{u_0}$. Then there exists such $u_1\in U$ that $v\in \mathrm{Im}\ h_{u_1}$. Now it's easy to see that the rank of a general linear combination $\alpha h_{u_0} + \beta h_{u_1} = h_{\alpha u_0+\beta u_1}$ will be greater than $r$.

  • $\begingroup$ Thanks for answer. Why do you say that $k=|\mu|-|\lambda|$? For example take $k$ even then definitely $S^k$ is subreps of $S^{k/2}\otimes S^{k/2}$ $\endgroup$ Jan 11, 2013 at 13:20
  • $\begingroup$ @Klim Due to dualization, in my notation $V_1=S^{k/2}V^*$. In particular, $\lambda = (0,\ldots,0,-k/2)$. $\endgroup$ Jan 11, 2013 at 13:34

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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