# Aspherical homotopy orbit space of configurations on the 2-sphere

The group SO(3) acts naturally on $S^2$ and thus on $Conf(S^2, q)$, the configuration space of $q$ distinct points on the 2-sphere, via the diagonal action on $S^2 \times...\times S^2$. This is a question about $Conf(S^2, q)_{hSO(3)} = ESO(3) \times _{SO(3)} Conf(S^2, q)$, the $SO(3)$-equivariant homotopy type of this space.

Theorem: $Conf(S^2, q)_{hSO(3)}$ is aspherical for $q\geq 3$.

Proof (sketch): I will just write $C_q$ for $(Conf(S^2, q))_{hSO(3)}$. Projection to the first three coordinates $p:C_q \to C_3$ is a fibration (an equivariant version of the Fadell-Neuwirth fibration) with fiber $Conf(S^2-$ {3 points}, $q-3)$. But $C_3 \simeq *$, so the inclusion of the fiber $Conf(S^2-$ {3 points},$q-3)$ into $C_q$ is a homotopy equivalence. The fiber is a $K(\pi, 1)$ by an easy induction argument (using Fadell-Neuwirth again). QED.

Question: So $Conf(S^2, q)_{hSO(3)}$ is a $K(\pi, 1)$. What's $\pi?$

-
Are you distinguishing the "SO(3) fundamental group" from the usual fundamental group? If so, would the former be defined as the fundamental group of the Borel construction? – David Roberts Nov 19 '10 at 4:41
Right, the former; so I'm looking at $ESO(3) \times _{SO(S)} Conf(S^2, q)$. – Romeo Nov 19 '10 at 4:44
Errr, $...\times _SO(3)...$, of course. – Romeo Nov 19 '10 at 4:44
ugh, cant edit: $ESO(3) \times _{SO(3)} Conf(S^2, q)$. – Romeo Nov 19 '10 at 4:46
We need to project down to the 1st 3 coordinates, so q$\geq3$. – Romeo Nov 19 '10 at 16:53

$\pi$ is an iterated extension of free groups by free groups again using Fadell-Neuwirth. This is also called the pure braid group of the sphere minus three points on q-3 braids.
$\pi_1(Conf(S^2, q)) is the "q-strand pure braids on$S^2$", but I'mquotienting by the SO(3) action here. – Romeo Nov 19 '10 at 4:49 In your question I just noticed the symmetric group action is also present. Fundamental group of which space you are looking for...? Sorry I do not understand your notations. – Roushon Nov 19 '10 at 5:19 I added the clarification from D Robert's comment to the main post. – Romeo Nov 19 '10 at 5:40 I think Roushon's answer is correct (if you take ordered configurations). Proof: The fibre bundle$Conf^{q}(S^2) \to Conf^3 (S^2)$is$SO(3)$-equivariant and the fibre is$Conf^{q-3} (C \setminus \{0,1\})$. So there is a fibre sequence$Conf^{q-3} (C \setminus \{0,1\}) \to Conf^q (S^2)_{hSO(3)} \to Conf^3(S^2)_{hSO(3)}$. The base is contractible (essentially because Möbius transformations on$S^2=CP^1$act simply transitively on$Conf^3 (S^2)$. – Johannes Ebert Nov 19 '10 at 16:41 Right, this is the argument in the post. So we are looking at$\pi_1$of configurations of$q-3\$ points in the thrice-punctured sphere. – Romeo Nov 19 '10 at 17:16