Given: $|G|=n$, $H$ is a subgroup of $G$ and $|G/H|=k$, where $n$ does not divide k!.

WTP: The left action map on $G/H$ has a nontrivial kernel.

I have not put the entire problem I am trying to work on here, just the part I am stuck on. I think this should be true, but I am having trouble showing nontriviality. I know if we take an element of G/H then the kernel of the map are all elements a such that $agH=gH$, but I cannot prove that there are any $a$'s that will do this other than the identity. Can anyone help, or tell me if this is not true?

  • 1
    $\begingroup$ Work this out with $S_3$ and a subgroup of order $2$. $\endgroup$ – John Brevik Apr 20 '15 at 3:17
  • $\begingroup$ The kernel are all elements $a$ such that $agH=gH$ for every $g$, i.e., the intersection of the stabilizers. $\endgroup$ – Dustan Levenstein Apr 20 '15 at 3:18
  • $\begingroup$ I'm sorry I mistyped, n should not divide k!, obviously it does not divide k,k-1,k-2,...,1. I have modified the question. $\endgroup$ – Math Student Apr 20 '15 at 3:22

Hint: The left action map of $G$ on $G/H$ induces a homomorphism $\phi: G \to S(G/H) \simeq S_k$, given by $a \mapsto \{gH \mapsto agH\}$.

The kernel of the action is the kernel of this homomorphism. If the kernel were trivial, then $\phi$ would be injective, giving an embedding of $G$ as a subgroup of $S_k$. Why can't this happen?

  • $\begingroup$ I think the homomorphism I am looking at is G/H on phi(G/H), would that be any different? Also, I have figured out what the kernel would be, but I am having difficulty proving the kernel is necessarily nontrivial, which I would assume follows from the second given, that n does not divide k!. Do you know how to prove nontriviality? Or maybe this is an obvious fact? As stated above, the overall question is actually a lot more than what I wrote, this is just one step of it that I am having trouble with. $\endgroup$ – Math Student Apr 20 '15 at 3:38
  • $\begingroup$ No idea what you mean by "I think the homomorphism I am looking at is..."; remember that $G/H$ is not necessarily a group. The induced homomorphism I speak of is given by $a \mapsto \{gH \mapsto agH\}$. What can be said about this homomorphism if the kernel is trivial? $\endgroup$ – Dustan Levenstein Apr 20 '15 at 3:44
  • $\begingroup$ It means it is injective, is that what you are going for? $\endgroup$ – Math Student Apr 20 '15 at 3:49
  • $\begingroup$ If the kernel is trivial, then the map is injective. I assume that's what you meant to say, and yes. $\endgroup$ – Dustan Levenstein Apr 20 '15 at 3:50
  • $\begingroup$ But I want the kernel to be nontrivial, so you are saying that what I said above is not true? $\endgroup$ – Math Student Apr 20 '15 at 3:50

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.