Mathematics Stack Exchange is a question and answer site for people studying math at any level and professionals in related fields. 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

I would like to know the proof/explanation for the following three properties of the representation of $SO(6)$,

  • What is the importance of symmetric traceless tensors of arbitrary rank w.r.t $SO(6)$ representations?

  • Why can the $6$ dimensional vector representation of $SO(6)$ be thought of as the rank-$2$ anti-symmetric representation of $SU(4)$?

  • Why can the tensor product two of $6$ dimensional vector representations of $SO(6)$ be thought of as a sum of a symmetric traceless, anti-symmetric and a one-dimensional representation?

I would also like to know if there are some general properties of $SO(n)$ representations from which the above follow. Then may be someone can may be kindly also write down the explanation for $SO(n)$ in general!

share|cite|improve this question
I don't understand how the first statement constitutes a property. Can you be more precise? – Qiaochu Yuan Aug 12 '12 at 18:04
I'm confused on your third question-- doesn't the vector representation usual refer to the defining rep of $SO(6)$ on $\mathbb R^6$? But this is irreducible. – Eric O. Korman Aug 13 '12 at 1:06
@Qiachu Yuan I am not sure how to make it precise. From certain literature I get the feeling that symmetric traceless tensors of arbitrary rank form a (irreducible?) representation of $SO(6)$. It would be great if you can elaborate on that point. Are these somehow the "fundamental representations"? If so then how? – user6818 Aug 13 '12 at 22:57
@Eric I think I had a typo in the question. I tried editing it. Does it now make sense? – user6818 Aug 13 '12 at 22:58
up vote 5 down vote accepted

I'm not sure how to answer question 1-- it is a little vague. The second question is unique to $SO(6)$ and is answered by understanding how $SU(4)$ is the double cover of $SO(6)$. This double cover is constructed as follows: first notice that $\Lambda^2 \mathbb C^4$ is six dimensional and has a Hermitian inner product induced by the one on $\mathbb C^4$. This can be defined by saying if $\{e_1,\ldots,e_4\}$ is an orthonormal basis for $\mathbb C^4$ then $\{e_i \wedge e_j\}$ is an orthonormal basis for $\Lambda^2 \mathbb C^4$ or, more invariantly, $$ \langle v_1 \wedge v_2, w_1 \wedge w_2 \rangle = \det\langle v_i, w_j\rangle $$ where $\langle v, w\rangle$ is the Hermitian inner product on $\mathbb C^4$. Then it is easy to see that the action of $SU(4)$ on $\Lambda^2 \mathbb C^4$ preserves this inner product, giving a homomorphism $SU(4) \to U(\Lambda^2 \mathbb C^4) = U(6)$. But there is also a symmetric inner product on $\Lambda^2 \mathbb C^4$ $$ \Lambda^2 \mathbb C^4 \otimes \Lambda^2 \mathbb C^4 \to \Lambda^4 \mathbb C^4 \simeq \mathbb C $$ which is also preserved by the action of $SU(4)$ since elements in $SU(4)$ have determinant 1. Thus the image of the above homomorphism lies in $U(6) \cap SO(6, \mathbb C) = SO(6)$. One checks that the kernel of this map is $\{\pm 1\}$ and therefore, by dimensionality reasons, must be a surjection. Thus this is the double cover $SU(4) \to SO(6)$ but now it is transparent that the $\Lambda^2 \mathbb C^4$ rep of $SU(4)$ corresponds to the vector representation of $SO(6)$.

Your third question holds for all of $SO(n)$ (indeed for any self-dual representation). Since there is an invariant symmetric inner product on $V = \mathbb C^n$ preserved by $SO(n)$, it follows that the fundamental representation is self-dual. Thus $V\otimes V \simeq V^* \otimes V \simeq End(V)$. Now for any rep $V$ of any group, $V \otimes V$ always reduces into a direct sum of symmetric plus antisymmetric (this is immediate from the definition of the tensor product representation). Similarly, for any rep $V$ of any group, $End(V)$ has a one-dimensional subrepresentation spanned by the identity transformation. Now under the correspondence of $End(V) \simeq V\otimes V$, the symmetric tensors will contain the identity endomorphism but a complementary subspace is given by traceless endomorphisms.

share|cite|improve this answer
Thanks for the reply. Can you kindly make it explicit as to what is the $SU(4)$ action on $\wedge ^2 \mathbb{C}^4$ that you have in mind? Also given the 6 dimensional real representation of $SO(6)$ being spanned by say $\{ \phi^a \}_{a=1}^6$ how does one write down the basis of antisymmetric rank 2 tensors which span the same when thought of as a representation of $SU(4)$? Also if you could explain as to how the symmetric traceless tensors become irreducible representations of the $SO$ groups...and how/if they are all that can be gotten inside tensor powers. – user6818 Aug 24 '12 at 15:28
@user6818: $A \in SU(4)$ acts on $v\wedge w \in \Lambda^2 \mathbb C^4$ by $(Av) \wedge (Aw)$ where $Av$ is the normal action of $SU(4)$ on $\mathbb C^4$. For the second question I suggest you work out what's happening with a basis of $\Lambda^2 \C^4$ to make a specific identification of $\mathbb R^6$. For the last question, I think it may take a little work to see its irreducible (unless you compute the highest weight and compare with the Weyl dimension formula): any symmetric matrix is diagonalizable via a matrix in $SO(6)$ so given two diagonal tracelss matrices try to find a matrix... – Eric O. Korman Aug 25 '12 at 2:02 $SO(6)$ conjugating one into the other. – Eric O. Korman Aug 25 '12 at 2:03
Dear Eric, A small comment/query on your nice answer: is it standard to write $S^2 V$ for $V\otimes V$? I would rather have thought that $S^2 V$ stood for $Sym^2 V$ (although the intended meaning of your usage was clear in context). Regards, – Matt E Oct 21 '12 at 20:08
@MattE : No, that was a mistake-- thanks for pointing it out. I also wrote "symmetric" somewhere where I should have said "tensor." Not sure why I kept conflating the two. I have edited it accordingly. – Eric O. Korman Oct 21 '12 at 21:40

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.