2

I'm having some trouble understanding this answer to If a maximal subgroup is normal, it has prime index.

Notation: We denote the normal subgroup by $N$ instead.

By the Correspondence Theorem, there exists a bijection from the set of all subgroups $H$ such that $N\subseteq H\subseteq G$ onto the set of all subgroups of $G/N$. Since the only such subgroups are $H=N$ and $H=G$, $G/N$ has only two subgroups, namely $N/N$ and $G/N$.

Let $xN$ be a nontrivial element in $G/N$. $\langle xN\rangle$ is a nontrivial subgroup of $G/N$, thus $\langle xN\rangle=G/N$. This means $G/N$ is cyclic. If $|G/N|$ is infinite, then $G/N\cong\mathbb{Z}$ which is a contradiction as $\mathbb{Z}$ has infinite subgroups of the form $n\mathbb{Z}$. Therefore $[G:N]=|G/N|$ is finite.

Question:

I understand that $xN$ is a non-trivial element in $G/N$. However, what proof do we have that $(xN)^2 = x^2N$, $(xN)^3 = x^3N$ and other higher order terms exist in the group $G/N$? And without proving that first, I don't think it makes sense to say that $\langle xN\rangle$ is a subgroup of $G$.

  • 2
    I'm not sure I get your problem. If $N$ is normal then $G/N$ is equipped with the quotient group structure given by $gN\cdot hN:=ghN$. And so $(xN)^m=x^mN$ by the definition (plus some trivial induction). – freakish Apr 14 '20 at 07:26

2 Answers2

2

For any element $g\in G$, $gN$ is some coset of $G$, and thereby is some element of $G/N$, which is by definition the group of all cosets of the normal subgroup $N$ in $G$.

Parcly Taxel
  • 105,904
1

For any $x\in G$, we have $xN\in G/N$. But once you have your hands on an element of a group, you can always consider the cyclic subgroup it generates.