# Abstract Nonsense

## Review of Group Theory: Homomorphisms

Point of post: In this post we review basic facts about homomorphisms. Simple stuff like the image of a subgroup is a subgroup, the inverse image of a subgroup is a subgroup, etc. etc.

Motivation

Just as linear transformations are the structure preserving maps for vector spaces homomorphisms are the structure preserving maps for groups. They, in essence, preserve most the qualities that interest us in the theory of groups. And, if the homomorphism happens to be bijective then we get a group isomorphisms (not to be confused with a linear isomorphism) which preserve ‘all’ the qualities that interest us. So, let us begin

Homomorphisms

Let $\left(G,\star\right)$ and $\left(G',\ast\right)$ be groups. Then a map $\theta:G\to G'$ is a group homomorphism if

$\theta\left(g_1\star g_2\right)=\theta\left(g_1\right)\ast\theta\left(g_2\right)$

for all $g_1,g_2\in G$. We denote the set of all homomorphisms $G\to G'$ by $\text{Hom}\left(G,G'\right)$. We prove the immediate theorems

Theorem: Let $G$,$G'$, and $G''$ be groups and $\theta:G\to G'$ and $\phi:G'\to G''$ be homomorphisms then:

\displaystyle \begin{aligned}&\mathbf{(1)}\quad \theta(e_G)=e_{G'}\\ &\mathbf{(2)}\quad \theta\left(g^z\right)=\theta(g)^z,\text{ }z\in\mathbb{Z}\\ &\mathbf{(3)}\quad H\leqslant G\implies \theta\left(H\right)\leqslant G'\\ &\mathbf{(4)}\quad H\leqslant G'\implies \theta^{-1}\left(H\right)\leqslant G\\ &\mathbf{(5)}\quad\textit{If }\theta\textit{ is bijective, then }\theta^{-1}\textit{ is a homomorphism}\\ &\mathbf{(6)}\quad \textit{If }G=\left\langle S\right\rangle\textit{ and }\theta,\phi:G\to G'\textit{ are homomorphisms such that }\theta_{\mid S}=\phi_{\mid S}\textit{ then }\theta=\phi\\ &\mathbf{(7)}\quad\textit{If }G\textit{ is abelian and }\theta\textit{ surjective, then }G'\textit{ is abelian}\\ &\mathbf{(8)}\quad\textit{If }G\textit{ is cyclic, then }\theta(G)\textit{ is cyclic}\\ &\mathbf{(9)}\quad\phi\circ\theta\textit{ is a homomorphism}\end{aligned}

Proof:

$\mathbf{(1)}$: This is immediately clear since $\theta(e_G)=\theta(e_Ge_G)=\theta(e_G)\theta(e_G)$ and so upon cancellation $\theta(e_G)=e_{G'}$.

$\mathbf{(2)}$: It’s clear by induction that $\theta\left(g^n\right)=\theta(g)^n$ for all $n\in\mathbb{N}$. Note though that by $\mathbf{(1)}$ we have that

$\theta(g)\theta\left(g^{-1}\right)=\theta\left(gg^{-1}\right)=\theta\left(e_G\right)=e_{G'}$

and thus $\theta\left(g^{-1}\right)=\theta(g)^{-1}$. Thus, it follows that

$\theta\left(g^{-n}\right)=\theta\left(\left(g^{-1}\right)^n\right)=\theta\left(g^{-1}\right)^n=\left(\theta(g)^{-1}\right)^n=\theta(g)^{-n}$

for all $n\in\mathbb{N}$ from where the problem easily follows.

$\mathbf{(3)}$: Let $\theta(g),\theta(h)\in \theta\left(H\right)$ then

$\theta(g)\theta(h)^{-1}=\theta(gh^{-1})\in\theta(G)$

Thus, noting that $\theta(e_G)=\theta(e_{G'})\in \theta(H)$ we may conclude from a previous theorem that $\theta(H)\leqslant G'$.

$\mathbf{(4)}$: Suppose that $g,h\in\theta^{-1}\left(H\right)$, then

$\theta\left(gh^{-1}\right)=\theta(g)\theta(h)^{-1}\in H$

so that $gh^{-1}\in\theta^{-1}\left(H\right)$. Noticing then that $e_G\in\theta^{-1}(\{e_{G'}\})\subseteq\theta^{-1}\left(H\right)$ allows us to conclude.

$\mathbf{(5)}$: This is similar to the proof that set-theoretic inverses of linear transformations are linear transformations. Namely, let $g',h'\in G'$. Then, since $\theta$ is surjective there exists $g,h\in G$ such that $\theta(g)=g'$ and $\theta(h)=h'$. Then,

$\theta(\theta^{-1}(g')\theta^{-1})=\theta(gh)=\theta(g)\theta(h)=g'h'$

and thus

$\theta^{-1}(g')\theta^{-1}(h')=\theta^{-1}(g'h')$

from where the conclusion follows.

$\mathbf{(6)}$: This follows since if $g\in G$ is arbitrary there exists $s_1,\cdots,s_n\in S$ such that

$\displaystyle s_1^{\epsilon_1}\cdots s_n^{\epsilon_n}=g$

where $\epsilon_n\in\{-1,1\}$. Then,

\displaystyle \begin{aligned}\theta(g) &= \theta\left(s_1^{\epsilon_1}\cdots s_n^{\epsilon_n}\right)\\ &= \theta(s_1)^{\epsilon_1}\cdots\theta(s_n)^{\epsilon_n}\\ &= \phi(s_1)^{\epsilon_1}\cdots\phi(s_n)^{\epsilon_n}\\ &= \phi\left(s_1^{\epsilon_1}\cdots s_n^{\epsilon_1}\right)\\ &= \phi(g)\end{aligned}

from where the conclusion follows.

$\mathbf{(7)}$:  Let $g',h'\in G'$ be arbitrary. Since $\theta$ is surjective there exists $g,h\in G$ such that $\theta(g)=g'$ and $\theta(h)=h'$. So,

$\displaystyle g'h'=\theta(g)\theta(h)=\theta(gh)=\theta(hg)=\theta(h)\theta(g)=h'g'$

since $g',h'$ were arbitrary the conclusion follows.

$\mathbf{(8)}$: We merely note that

\displaystyle \begin{aligned}\theta\left(\left\langle g\right\rangle\right) &= \left\{\theta\left(g^z\right):z\in\mathbb{Z}\right\}\\ &= \left\{\theta(g)^z:z\in\mathbb{Z}\right\}\\ &= \left\langle \theta(g)\right\rangle\end{aligned}

$\mathbf{(9)}$: This follows from noticing that for any $g,h\in G$ we have that

$\left(\phi\circ\theta\right)(gh)=\phi(\theta(gh))=\phi(\theta(g)\theta(h))=\phi(\theta(g))\phi(\theta(h))$

$\blacksquare$

References:

1.  Lang, Serge. Undergraduate Algebra. 3rd. ed. Springer, 2010. Print.

2. Dummit, David Steven., and Richard M. Foote. Abstract Algebra. Hoboken, NJ: Wiley, 2004. Print

December 28, 2010 -

1. […] Point of post: This is a continuation of this post. […]

Pingback by Review of Group Theory: Homomorphisms (Pt. II) « Abstract Nonsense | December 28, 2010 | Reply

2. […] Let be a finite group and be a finite-dimensional complex vector space. Then a representation of on is a homomorphism […]

Pingback by Representation Theory: Definitions and Basics « Abstract Nonsense | January 18, 2011 | Reply

3. […] Let and be any two groups. Then a homomorphism has the property that is abelian if and only if can be factored through by , in other words if […]

Pingback by Review of Group Theory: The Commutator Subgroup and the Abelianization of a Group « Abstract Nonsense | February 27, 2011 | Reply

4. […] morphisms to preserve the ring structure. In other words, we’d like it to be additive (be a group homomorphism for the group structure) and multiplicative (be a semigroup homomorphism for the semigroup […]

Pingback by Ring Homomorphisms (Pt. I) « Abstract Nonsense | June 18, 2011 | Reply

5. […] Let . We know from basic group theory that  and so it suffices to show that has the left absorption property. To see this let and . […]

Pingback by Ideals and Homomorphisms « Abstract Nonsense | June 24, 2011 | Reply

6. […] the endomorphism ring of an abelian group would be to note that for a given abelian group and endomorphisms  there are two ‘natural’ operations to perform on them, namely you can add them (in […]

Pingback by Endomorphism Ring of an Abelian Group « Abstract Nonsense | July 10, 2011 | Reply

7. […] for every and . Since is, in particular, a group homomorphism between it enjoys all the pursuant benefits. We define, as per usual, the kernel of a -morphism to be group-theoretic kernel of (i.e. the […]

Pingback by Module Homomorphisms « Abstract Nonsense | October 28, 2011 | Reply