0
$\begingroup$

Here's the question: is every finite group a subgroup of a symmetric group of sufficiently large order? More specifically, if a group $G$ has order $n$, then is it true that $G \le S_{n}$?

For instance, both groups of order $4$ can be found in $S_4$. $\{(1), (1 2 3 4), (1 3)(2 4), (1 4 3 2)\} \cong C_4$ and $\{(1), (1 2), (3 4), (1 2)(3 4)\} \cong \mathbb{Z}_2 \times \mathbb{Z}_2$. Also, in general, $C_n, A_n, D_{2n}, S_n \le S_n$.

After a short investigation, I was unable to find a subgroup in $S_4$ that was congruent to $Q$, the unit quaternions, but I would be quite surprised if $Q \not\le S_8$.


Okay, the answer is "yes", and this is known as Cayley's Theorem. However, the link points to a proof-less page and the proofs given on Wikipedia aren't very clear, so I would really, really appreciate a clear, intuitive, and conceptual proof of Cayley's Theorem.

$\endgroup$
6
  • $\begingroup$ What is the action of each element of $Q$ on the elements of $Q$. I.e., how does "multiplication on the left by the element $x$" permute the elements of $Q$ as you iterate $x$ through $Q$? This should suggest to you that the multiplication table of a group can be embedded in permutations of the elements of the group... $\endgroup$ – Eric Towers Jul 8 '14 at 1:36
  • 2
    $\begingroup$ Cayley's Theorem. $\endgroup$ – David Jul 8 '14 at 1:37
  • $\begingroup$ @David: Cayley's Theorem is exactly what I was suspecting. $\endgroup$ – El'endia Starman Jul 8 '14 at 1:41
  • $\begingroup$ See math.stackexchange.com/a/854561/589. $\endgroup$ – lhf Jul 8 '14 at 2:03
  • $\begingroup$ Please explain why the proofs at Wikipedia do not work for you. $\endgroup$ – lhf Jul 8 '14 at 2:36
1
$\begingroup$

The intuation is that $$gG=G$$ for any $g\in G$ which means that every $g$ of permutes the elements of $G$. Surprisingly the corresponding permutations elements constitute a subgroup of $S_G$ which is isomorphic to $G$.


If $G$ is simple, $G$ can be embeded into $A_n$ where $n$ is the smallest index of nontrivial subgroup of $G$.

$\endgroup$
1
  • 1
    $\begingroup$ Ah! That's just the sort of intuition I was looking for! Much appreciated! $\endgroup$ – El'endia Starman Jul 8 '14 at 2:38
2
$\begingroup$

In any group $G$, any element $g \in G$ defines a function from $G$ to itself denoted $L_g : G \to G$, called "left multiplication": $$L_g(h) = gh $$ The group axioms show that this function is injective $$L_g(h) = L_g(h') \iff gh = gh' \iff h=h' $$ and that it is onto $$L_g(g^{-1} h) = g g^{-1} h = h $$ Therefore the formula $g \mapsto L_g$ defines a function $G \mapsto Sym(G)$, where $Sym(G)$ denotes the "symmetric group" of the set $G$, namely the group of self-bijections of $G$. The group axioms also show that the function $g \mapsto L_g$ is a homomorphism $$L_{g_1} \circ L_{g_2} (h) = g_1(g_2(h)) = (g_1 g_2)(h) = L_{g_1 g_2}(h) $$ and that it is injective $$L_g(h)=h \iff gh=h \iff g=Id $$ Therefore, $G$ embeds in $Sym(G)$ which is isomorphic to $S_{|G|}$; that isomorphism is obtained by enumerating the elements of $G$.

$\endgroup$
1
$\begingroup$

Yup. $G$ is always a subgroup of $S_{|G|}$.

On a related note, the question of the smallest $n$ such that $G\leq S_n$ is, in general, not known.

$\endgroup$

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.