10
$\begingroup$

There are many non-homeomorphic 3-manifolds with isomorphic fundamental groups, for example the lens spaces $L(p,q_1)$ and $L(p,q_2)$ with $q_1 \ne \pm q_2^{\pm 1}\mod p$. Also, Seifert fibered spaces over the disk with two exceptional fibers can have isomorphic fundamental groups. A third type of example can be constructed by using the connect sum of knot of a chiral knot $K$, $S^3-(K\#K)$ and $S^3-(K\#-K)$ (think of the square and granny knots). Even more examples can be constructed by connect summing the examples in this paragraph (in the sense of 3-manifolds as opposed to knots).

However, there are also non-orientable and orientable manifolds with isomorphic fundamental groups, which motivates my question:

Is there a list of all non-homeomorphic pairs of 3-manifolds (orientable and non-orientable) with isomorphic fundamental groups?

Given, the likely complexities of reducible manifolds and graph manifolds, it would be fine if it was only a list of geometric 3-manifolds.

$\endgroup$
3
  • 3
    $\begingroup$ Aschenbrenner, Friedl and I discuss this question at the beginning of our survey '3-manifold groups'. In the closed case, non-homeomorphic pairs only arise as lens spaces or connect sums, but the case with boundary is more complicated. $\endgroup$
    – HJRW
    Apr 29, 2015 at 11:05
  • $\begingroup$ Thanks Henry. Isn't assumed throughout your paper that you guys discuss only oriented 3-manifolds? I was really hoping to find a reference that deals with the non-orientable case as well. I fully understand why one would like to avoid the pathologies of non-orientable manifolds especially when considering virtual properties. However, I wondering if anyone has a nice approach to 3-manifolds with the same fundamental group regardless of orientablity. $\endgroup$ Apr 29, 2015 at 12:07
  • $\begingroup$ Sorry, Neil, I missed your emphasis on the non-orientable case. Yes, we don't discuss that. Indeed, I think it's not even clear that geometrization has been worked our properly in the non-orientable case, so you may want to assume that. $\endgroup$
    – HJRW
    Apr 29, 2015 at 12:12

1 Answer 1

11
$\begingroup$

(Edit: Everything what follows is about closed and orientable $3$-manifolds.)

Non-spherical geometric $3$-manifolds are determined by their fundamental group. This is proved in Peter Scott's paper "There are no fake Seifert fiber spaces with infinite $\pi_1$" http://www.jstor.org/stable/2006970 (He does the case of Seifert fibrations. For hyperbolic $3$-manifolds the uniqueness of course follows from Mostow rigidity.)

For irreducible manifolds with nontrivial JSJ-decomposition one can apply Waldhausen's rigidity theorem (because these manifolds are necessarily Haken) to conclude that the are also determined by their fundamental groups.

For reducible $3$-manifolds things are a bit more complicated because one can choose orientations of the prime components separately. Section 2 of the survey paper of Aschenbrenner-Friedl-Wilton http://www.uni-regensburg.de/Fakultaeten/nat_Fak_I/friedl/papers/3-manifold-groups-final-version-031115 explains why there are only three ways to get 3-manifolds with isomorphic fundamental groups.These three ways are

  • the examples of lens spaces mentioned in the question.

  • manifolds of the form $M\sharp N$ and $M\sharp \overline{N}$, where $N$ and $\overline{N}$ are the same manifold with opposite orientations

  • examples constructed from the above by taking connected sums.

$\endgroup$
2
  • $\begingroup$ I think you might want to be a little careful citing Scott. He also requires that the manifolds are closed, right? There are examples of non-homomorphic Seifert fibered spaces over the disk with two exceptional fibers that are non-homeomorphic. D^2(3,1)(3,1) and D^2(3,1)(3,2) form a pair of these type of examples. One can see they are not homeomorphic by performing surgery along the peripheral curve that bounds an annulus, which results in L(3,1)#L(3,1) and L(3,1)#L(3,2). $\endgroup$ Apr 29, 2015 at 12:03
  • $\begingroup$ Yes, I was talking about closed manifolds. In section 2.2 of Aschenbrenner-Friedl-Wilton you find a discussion of $3$-manifolds with incompressible boundary. $\endgroup$
    – ThiKu
    Apr 29, 2015 at 12:05

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.