Let $g \geq 2$. Let $S = \langle a_1,b_2,...,a_g,b_g | [a_1,b_1] \cdots [a_g,b_g] \rangle$ be the fundamental group of a genus $g$ surface and let $F_g$ be a free group with $g$ generators. Given two surjections $f_1,f_2 : S \to F_g$ is there a way to determine if there are automophisms $\phi: S \to S$ and $\psi: F_g \to F_g$ so that $f_1 = \phi \circ f_2 \circ \psi$?

  • $\begingroup$ You can rephrase your question in terms of equations over free groups: each such surjection corresponds to a solution to the equation $[x_1, y_1]\cdots[x_g, y_g]=1$ in the free group $F_g$, and you want to know if two such solutions are "equivalent". There is a well-developed theory of equations over free groups, using gadgets called "Makanin-Razborov" diagrams. These describe all solutions using automorphisms of "limit groups", and your group is a limit group (although may not be a limit group for the equations you want...). $\endgroup$
    – user1729
    Mar 21, 2018 at 10:26
  • $\begingroup$ For a reference, see: Sela, Zlil (2001), "Diophantine geometry over groups. I. Makanin-Razborov diagrams", Publications Mathématiques de l'IHÉS, 93 (1): 31–105, doi:10.1007/s10240-001-8188-y, MR 1863735 Also, this theory comes with a health warning attached - it is all very complicated and slightly controversial. But also extremely important! $\endgroup$
    – user1729
    Mar 21, 2018 at 10:27

1 Answer 1


Reduction of problem, but not an answer:

Suppose $N=\ker f_1 = \ker f_2$. Then $f_1$ and $f_2$ both induce isomorphisms, $f_1' : S/N\to F_g$ and $f_2' : S/N\to F_g$. Then $\phi =f_1'f_2^{\prime -1}$ is an automorphism of $F_g$, and $f_1=\phi f_2$. Thus we only need to find an automorphism $\psi$ of $S$ taking $\ker f_1$ to $\ker f_2$, since then we'll have $\ker f_2\psi = \ker f_1$. Thus the desired result is true if and only if the automorphism group of $S$ acts transitively on normal subgroups $N$ of $S$ with $S/N \cong F_g$.

I'm not sure how to prove that the automorphism group either does or does not act transitively on such subgroups though.

  • $\begingroup$ I think that the desired result is not whether the automorphism group acts transitively on such subgroups. Instead it is whether there is a way (a method, or an algorithm) which, given two such subgroups, determines whether there exists an automorphism taking one of those two subgroups to the other. $\endgroup$
    – Lee Mosher
    Mar 23, 2018 at 16:24

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.