Let $H$ a subgroup of $S_n$ such that for all $a,b\in \{1,2,...,n\}$ there exist $\tau\in H$ such that $\tau(a)=b$. Show that if $H$ contains a transposition and a $n-1$ cycle then $H=S_n$.

  • $\begingroup$ If the first fact is true then I see no need for the second part. Since every element in $S_n$ can be rewritten as the product of transpositions and the fact that all transpositions in $S_n$ are contained in $H$($\exists\tau\in H$ such that $\tau(a)=b$). Would be enough to prove the statement. (Okay ignore this I see that the way you used $\tau$ does not imply that it is a transposition ) $\endgroup$ – user160110 Oct 25 '15 at 21:39
  • $\begingroup$ @user160110 : The fact that there exist an element of $\tau \in H$ such that $\tau(a) = b$ does not imply that $\tau$ is a transposition. For example, the subgroup of permutations of the form $\tau_a : k \mapsto k + a \mod n$ (where ${1,\ldots, n}$ is identified with $\mathbb{Z}/n\mathbb{Z}$) has that property (take $\tau_{b-a}$) but does not contain any transpositions (except if $n = 2$ of course). $\endgroup$ – Joel Cohen Oct 25 '15 at 21:49
  • $\begingroup$ @JoelCohen I know this, but the symbol $\tau$ is often used for transpostions where as symbols like $\alpha, \beta,$ and $\phi$ might be used for general permutaions. I just misread the second part. $\endgroup$ – user160110 Oct 25 '15 at 21:52
  • $\begingroup$ @user160110 $\tau$ not is necesarilly a transposition $\endgroup$ – Roiner Segura Cubero Oct 26 '15 at 0:38

Okay since you have a transposition (lets call it $\tau=(i,j)$ ) in H and for any $a,b\in\{1,...,n\}$ their exist a permutation in $\beta_{a,b}\in H$ such that $\beta_{a,b}(a)=b$, it follows that for any $a\in \{1,...,n\}$ we have that $\beta_{i,a}\tau \beta_{i,a}^{-1}=(a,j)$. In this way we generate all transpositions and thus all permutations.

  • $\begingroup$ We generates all transpositions? $\endgroup$ – Roiner Segura Cubero Oct 26 '15 at 1:18
  • $\begingroup$ @RoinerSeguraCubero Let $a,b\in \{1,...,n\}$ then $\beta_{j,b}[\beta_{i,a}(i,j)\beta_{i,b}^{-1}]\beta_{j,a}^{-1}=(a,b)$ $\endgroup$ – user160110 Oct 26 '15 at 1:20
  • 1
    $\begingroup$ $\beta$ might move $j$ too, so you only get that $(a,k(a))\in H$ for all $a$. $\endgroup$ – user138530 Oct 26 '15 at 2:49
  • 1
    $\begingroup$ But you can finish up as follows: do what you did for the $a$ not in the $n-1$ cycle, and then move $k(a)$ around at will by conjugating with the $n-1$ cycle. This shows that $(a,b)\in H$ for some fixed $a$ and all $b$, and these generate $S_n$. $\endgroup$ – user138530 Oct 26 '15 at 2:53

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.