Let:
$$\kappa(G,H):=\cap_{\phi \in {\rm Hom}(G,H)} \ker\phi$$ where ${\rm Hom}(G,H)$ is the set of all homomorphisms from $G$ to $H$.
I am trying to prove that if $G$ is a finitely generated group, and $H$ a finite group, then $[G:\kappa(G,H)]<\infty$.
Since I had no idea on how to start proving this generally, I tried proving a simplified version of this, where both $G$ and $H$ are abelian. In this case, I can use the Fundamental Theorem of Finitely Generated Abelian Groups and assume that $$G = \mathbb{Z}/(g_1)\oplus\mathbb{Z}/(g_2)\oplus\dots\mathbb{Z}/(g_k)\oplus\mathbb{Z}^r$$
$$H = \mathbb{Z}/(h_1)\oplus\mathbb{Z}/(h_2)\oplus\dots\oplus\mathbb{Z}/(h_m)$$
I then observed that for all homomorphisms $\phi:G\to H$ we have that $\ker\phi$ has the form: $$N_1 \oplus N_2 \oplus\dots\oplus N_k \oplus m_1\mathbb{Z}\oplus\dots\oplus m_r\mathbb{Z}$$ for some finite normal subgroups $N_i \leq \mathbb{Z}/(g_i)$ and normal subgroups $m_1\mathbb{Z}\oplus\dots\oplus m_r\mathbb{Z} \leq \mathbb{Z}^r$ (we can see this by defining each possible homomorphism using the standard generators (e.g., $(0,\dots,0,1,0\dots0)$). Which means that when we take the intersection of all these kernels, we still get something of that form (e.g., instead of $m_1\mathbb{Z}$ we will get the ${\rm lcm}$ of all $m_1$'s taken across all $\phi$'s), so that $[G:\kappa(G,H)]<\infty$).
I am, however, none the wiser about how to generalize this to the non-abelian case.
I do have at my disposal two potentially related facts (although I do not see how they might be helpful... so I may be completely off here):
- For any two groups $G, H$, we have that $H$ is isomorphic to a quotient of $G$ iff $H$ is isomorphic to a quotient of $G/\kappa(G,H)$.
- For any groups $G, G', H$ we have: $\kappa(G\times G', H) = \kappa(G,H)\times\kappa(G', H)$
Any advice?
(Thanks to user1729 and Lee Mosher)
It is sufficient to prove that $[G:\ker\phi]<\infty$ for all homomorphisms $\phi: G \to H$, since:
Indeed, let $\phi: G \to H$ be a homomorphism. Then by the first isomorphism theorem, we have that $[G:\ker\phi] = |{\rm Im}\, g(\phi)|<|H|<\infty$.
(as a side note - one way in which I could have realized how to generalize the abelian case was to think of which part of the observation:
was essential and generalizable, and that is that the intersection retains the essential property of being of finite index [and stripping the non-essential parts, e.g. of being in $\mathbb{Z}$]).