8
$\begingroup$

I have two questions:

Question 1. Suppose that $K$ is a knot in $S^3$. Let $\Sigma(K)$ be the double branched cover of $S^3$ branched along $K$. If $\Sigma(K)=\#_{i=1}^n\Sigma(2,3,5)$, then $K=\#_{i=1}^nT_{3,5}$?

Question 2. There is a conjecture of Ozsvath-Szabo that the only $L$-spaces which are integral homology 3-spheres are connected sums of Poincare homology 3-spheres. If this conjecture and Question 1 is true, then can we (re)prove that Khovanov homology is an unknot-detector using Ozsavth-Szabo's spectral sequence?

$\endgroup$
7
$\begingroup$

As Ian Agol points out the Orbifold Theorem will answer question 1 affirmatively. Although the original result is due to Thurston, the common references in the literature are:

Boileau, Michel, Sylvain Maillot, and Joan Porti. Three-dimensional orbifolds and their geometric structures. Vol. 15. Paris: Société mathématique de France, 2003.

and

Cooper, Daryl, Craig David Hodgson, and Steve Kerckhoff. Three-dimensional orbifolds and cone-manifolds. Vol. 5. Mathematical Society of Japan, 2000.

However, both of these results would still require one to understand the quotients of $\Sigma(2,3,5)$ by its order 2 symmetries. Dunbar provides such a classification, which serves a translation of Threlfall and Seifert's work, see:

Dunbar, William D. "Geometric orbifolds." Revista Matematica de la Universidad Complutense de Madrid 1.1 (1988): 67-100. (available here)

and

Threlfall, William, and Herbert Seifert. "Topologische Untersuchung der Diskontinuitätsbereiche endlicher Bewegungsgruppen des dreidimensionalen sphärischen Raumes I and II." Mathematische Annalen 104 and 107 (1931 and 1932): 1-70 and 543-586.

More specifically Dunbar gives a proof that if a knot $L$ has the property that $\Sigma(L)=\Sigma(2,3,5)$ then $L=T_{3,5}.$

| cite | improve this answer | |
$\endgroup$

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.