4
$\begingroup$

Let $\bar{X}$ be a compact Riemann surface of genus $g>0$. Let $X$ be $\bar{X}$ minus a finite set of points $\{a_1,\ldots,a_n\}$ ($n\geq 1$). Let $X^{(r)}$ be the configuration space of $r$ ordered distinct points on $X$: $$X^{(r)}=X^r-\bigcup_{i\neq j}\{(x_1,\ldots,x_r): x_i=x_j\}$$

(1) Is there an explicit description of the Lie algebra associated to the lower central series of the fundamental group of $X^{(r)}$ (=pure braid group of $r$ strands on $X$) in terms of generators and relations?

(2) Are there cases (i.e. specific curves) for which an explicit description of $H^1_{dR}(X^{(r)})$ is known?

Any relevant references will be appreciated. For (1), a paper of Bezrukavnikov addresses $X=\bar{X}$ case. A paper of Nakamura, Takao and Ueno studies the non-compact situation but works with the weight central series (as opposed to the lower central series). In (2), by an explicit description I mean an explicit basis of $H^1_{dR}(X^{(r)})$.

$\endgroup$

1 Answer 1

5
$\begingroup$

I can at least answer your second question. I'll be a bit brief, but let me know if you need more details. Let $M$ be an oriented manifold and $M^{(r)}$ the configuration space as in your question. Then it follows from the Totaro spectral sequence, i.e. the Leray spectral sequence for $M^{(r)} \to M^r$, that the map $M^{(r)} \to M^r$ induces an isomorphism on $H^1(-,\mathbf Z)$ whenever $\dim(M) >2$ or when $M$ is a surface of positive genus with $n\geq 0$ punctures.

To see this use the exact sequence of low degree terms: $$ 0 \to H^1(M^r,\mathbf Z) \to H^1(M^{(r)},\mathbf Z) \to E_2^{0,1} \to E_2^{2,0}$$ If $\dim(M)>2$ then $E_2^{0,1}=0$. When $\dim(M)=2$ we have $E_2^{0,1} = \mathbf Z^{n(n-1)/2}$ -- a copy of $\mathbf Z$ for each small diagonal -- which injects into $E_2^{2,0} = H^2(M^r,\mathbf Z)$ because each diagonal maps to a different Künneth component of the form $H^1(M,\mathbf Z) \otimes H^1(M,\mathbf Z)$ (the class of the diagonal in $H^1(M,\mathbf Z) \otimes H^1(M,\mathbf Z)$ is nonzero precisely when $g>0$). The result follows.

$\endgroup$
1
  • $\begingroup$ Thanks. I read parts of Totaro's paper after this answer and found your explanation clear and helpful. $\endgroup$
    – P.E.
    Mar 7, 2017 at 18:29

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.