Monodromy Representations of Mixed Braid Groups
Chitrabhanu Chaudhuri, Saswati Mukherjee · 2026
Read Monodromy Representations of Mixed Braid Groups by Chitrabhanu Chaudhuri, Saswati Mukherjee (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2604-00146.txt.
Introduction
Chaudhuri and Mukherjee give unitary representations of mixed braid groups on cohomology groups of Abelian branched covers over the complex projective line. Complex reflections generate the image of each representation. The authors also link the construction to the multivariate Burau representation, giving a full account of this monodromy picture. The paper works out the maps in full detail.
Catalog record
| CREATOR | Chitrabhanu Chaudhuri, Saswati Mukherjee |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 72324 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2604.00146
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Monodromy Representations of Mixed Braid Groups Chitrabhanu Chaudhuri, Saswati Mukherjee Abstract . We explicitly describe unitary representations of mixed braid groups on the cohomology of Abelian branched covers of ℂ ℙ 1 \mathbb{CP}^{1} . We show that the image of the representation is generated by complex reflections and relate it to the multivariate Burau representation. 1. Introduction The action of the braid group on the cohomology of covering spaces of the punctured plane is a classical subject. Note that a tuple p=(p_{1},\ldots,p_{n}) in \mathbb{C}^{n} of n distinct points determines a compact Riemann surface X_{p} with affine part \{(x,y)\in\mathbb{C}^{2}\mid y^{d}=(x-p_{1})(x-p_{2})\cdots(x-p_{n})\}. The map X_{p}\to\mathbb{CP}^{1} extending (x,y)\mapsto x is a Galois (branched) covering with cyclic Galois group \mathbb{Z}/d\mathbb{Z} . Now if we vary p over the unordered configuration space \conf_{n}(\mathbb{C})=\frac{\{(z_{1},\ldots,z_{n})\in\mathbb{C}^{n}\mid z_{i}\neq z_{j}\}}{S_{n}} we have a (smooth) fiber bundle \pi:\mathfrak{X}\to\conf_{n}(\mathbb{C}) with fibers \pi^{-1}(p)=X_{p} . Let \mathbb{C}_{\mathfrak{X}} be the (locally) constant sheaf on \mathfrak{X} . Taking the first higher direct image sheaf: \mathcal{V}=R^{1}\pi_{*}\mathbb{C}_{\mathfrak{X}} we obtain a local system \mathcal{V} over \conf_{n}(\mathbb{C}) with fibers \mathcal{V}_{p}=H^{1}(X_{p},\mathbb{C}). Consequently we have a monodromy representation of the Braid group \mathcal{B}_{n}=\pi_{1}(\conf_{n}(\mathbb{C}))\to\mathrm{GL}(H^{1}(X_{p},\mathbb{C})). In [ 14 ] , McMullen gives a complete account of this representation, and the geometric structures on moduli space, that arise via this branched-cover construction. This paper generalises this construction from cyclic covers to Abelian covers. Let \Lambda=(n_{1},\ldots,n_{m}) be a partition of n . Consider the compact Riemann surface X_{p} which completes \{(x,y_{1},\ldots,y_{m})\in\mathbb{C}^{m+1}\mid y_{j}^{d_{j}}=(x-p_{h_{j-1}+1})\cdots(x-p_{h_{j}})\}\qquad\text{where}\qquad h_{j}=\sum_{i=1}^{j}n_{i}. Now the map X_{p}\to\mathbb{CP}^{1} , (x,y_{1},\dots,y_{m})\mapsto x is a Galois cover with abelian Galois group isomorphic to \mathbb{Z}/d_{1}\mathbb{Z}\times\cdots\times\mathbb{Z}/d_{m}\mathbb{Z} . The Galois group is generated by T_{1},\ldots,T_{m}\qquad\text{where}\qquad T_{j}(x,y_{1},\ldots,y_{m})=(x,y_{1},\ldots,\zeta_{d_{j}}y_{j},\ldots,y_{m}). Consider the variation of the point p=(p_{1},\ldots,p_{n}) within the configuration space \conf_{n,\Lambda}(\mathbb{C}) . This space is defined as the quotient: \conf_{n,\Lambda}(\mathbb{C})=\frac{\{(z_{1},\ldots,z_{n})\in\mathbb{C}^{n}\mid z_{i}\neq z_{j}\}}{S_{n,\Lambda}}. Here, S_{n,\Lambda} is the subgroup of the symmetric group S_{n} containing all permutations that preserve the partition \Lambda . This results in a fiber bundle \pi:\mathfrak{X}\to\conf_{n,\Lambda}(\mathbb{C}) , where each fiber is given by X_{p} . In this way we get a monodromy representation (1.1) \theta:\pi_{1}(\conf_{n,\Lambda}(\mathbb{C}))\to\mathrm{GL}(H^{1}(X_{p},\mathbb{C})). The fundamental group \mathcal{B}_{n,\Lambda}:=\pi_{1}(\conf_{n,\Lambda}(\mathbb{C})) is called the mixed braid group or coloured braid group for the partition \Lambda . Let us fix p\in\conf_{n,\Lambda}(\mathbb{C}) and write X:=X_{p} . The representation \theta preserves the hermitian intersection form \left\langle\xi_{1},\xi_{2}\right\rangle=\int_{X}\xi_{1}\wedge\overline{\xi}_{2},\qquad\xi_{1},\xi_{2}\in H^{1}(X) and commutes with the Galois group \mathbb{Z}/d_{1}\mathbb{Z}\times\cdots\times\mathbb{Z}/d_{m}\mathbb{Z} . Let us decompose H^{1}(X) into T_{1},\ldots,T_{m} eigenspaces H^{1}(X)=\bigoplus_{\rho=(\rho_{1},\ldots,\rho_{m})}\Big(H^{1}(X)_{\rho}:=\cap_{j}\ker(T_{j}-\rho_{j}\id)\Big) where the direct sum is over \rho=(\rho_{1},\ldots\rho_{m})\in\mathbb{C}^{m} , with \rho_{j}^{d_{j}}=1 . The representation \theta preserves each eigenspace H^{1}(X)_{\rho} . Restricting the intersection form to H^{1}(X)_{\rho} we thus get a unitary representation \theta_{\rho}:\mathcal{B}_{n,\Lambda}\to\U(H^{1}(X)_{\rho})\cong\U(r,s) where (r,s) is the signature of the hermitian form on H^{1}(X)_{\rho} . Here \U(r,s) is the pseudo-unitary group of signature (r,s) . Description of results In ( 3.2 ) we compute the signature using the Chevalley-Weil formula to find that when \rho=(\zeta_{d_{1}}^{-k_{1}},\ldots,\zeta_{d_{j}}^{-k_{m}}) and 0<k_{j}<d_{j} for each j , \displaystyle r \displaystyle=\left\lceil\sum_{j=1}^{m}\frac{n_{j}k_{j}}{d_{j}}-1\right\rceil, \displaystyle s \displaystyle=\left\lceil n-\sum_{j=1}^{m}\frac{n_{j}k_{j}}{d_{j}}-1\right\rceil and in this case, \dim H^{1}(X)_{\rho}=n-1 when \rho_{1}^{n_{1}}\cdots\rho_{m}^{n_{m}}\neq 1 , otherwise n-2 (here \zeta_{d_{j}}=e^{2\pi\sqrt{-1}/d_{j}} ). The crux of the paper is finding an explicit spanning set \omega_{1},\ldots,\omega_{n-1} for H^{1}(X)_{\rho} in which the intersection form is given by \displaystyle\left\langle\omega_{i},\omega_{i}\right\rangle \displaystyle=\sqrt{-1}(\rho_{j}+1)(\rho_{j}-1)^{-1} \displaystyle h_{j-1}<i<h_{j}, \displaystyle\left\langle\omega_{h_{j}},\omega_{h_{j}}\right\rangle \displaystyle=\sqrt{-1}(\rho_{j}\rho_{j+1}-1)(\rho_{j}-1)^{-1}(\rho_{j+1}-1)^{-1} \displaystyle 1\leq j<m, \displaystyle\left\langle\omega_{i-1},\omega_{i}\right\rangle \displaystyle=\sqrt{-1}(-\rho_{j})(1-\rho_{j})^{-1} \displaystyle h_{j-1}<i\leq h_{j}. The spanning set is obtained by taking weighted sums of Poincaré duals of lifts of certain curves in \mathbb{C}-\{p_{1},\ldots,p_{n}\} (see Figure 3 and definitions ( 3.3 )). The central result of the paper is Theorem 5.2 where we describe the representation relative to the spanning set above utilising the intersection form. This description holds for any \rho satisfying \rho_{j}\neq 1 and \rho_{j}\rho_{k}\neq 1 for any j,k\in\{1,\ldots,m\} . These conditions account for the generic case for large d_{j} ; indeed, if the d_{j} are pairwise co-prime, the second condition holds by default. Furthermore, in Proposition 5.4 , we demonstrate that \theta_{\rho} is irreducible under these assumptions . The representation \theta_{\rho} is not faithful, in fact in Proposition 5.1 , we show that if \tau=(\sigma_{1}\cdots\sigma_{n-1})^{n} , that is the generator of the center of \mathcal{B}_{n} , (where \sigma_{i} are the standard generators of the braid group), then \tau^{f} is always in the kernel of \theta_{\rho} where f=\ord(n_{1}T_{1}+\cdots+n_{m}T_{m}) . We show that the representation \theta_{\rho} is isomorphic to the dual of the reduced multivariate Burau representation evaluated at \overline{\rho} (see Proposition 5.5 ). When \Lambda=(n) the representation \theta_{\rho} is the reduced Burau representation for the braid group \mathcal{B}_{n} evaluated at \overline{\rho} (see Theorem 5.5 of [ 14 ] ). At the other extreme for \Lambda=(1,\ldots,1) we get the dual of reduced Gassner representation evaluated at \overline{\rho} . Thus our family of representations interpolate between the reduced Burau and the reduced Gassner representations. Finally we observe that \theta_{\rho}(\mathcal{B}_{n,\Lambda}) is infinite whenever for some j , n_{j}\geq 3 , \rho_{j} is a primitive d_{j} -th root of unity and (n_{j},d_{j})\not\in\{(3,3),(3,4),(3,6),(3,10),(4,4),(4,6),(5,6),(6,6)\}. Connections with related literature It has been shown by Murakami [ 16 ] , Morton [ 15 ] , and Gabrovšek-Horvat [ 8 ] that the multivariate Burau representation can be used to calculate the multivariate Alexander polynomial of links. This establishes a direct connection between the representations investigated in this paper and knot theory, suggesting significant potential for future research into these connections. The topological framework of this paper is deeply tied to the classical Deligne-Mostow theory of hyper-geometric monodromy [ 5 ] . In Section 12, Deligne-Mostow study monodromy representation of cyclic covers of \mathbb{CP}^{1} of the form y^{d}=(x-a_{1})^{k_{1}}\cdots(x-a_{n})^{k_{n}} primarily to classify which rational weights (k_{1}/d,\ldots,k_{n}/d) yield discrete arithmetic or non-arithmetic lattices in the complex hyperbolic isometry group \mathrm{PU}(1,N) . This is related to the image of \theta_{\rho} when when \rho=(\zeta_{d}^{k_{1}},\ldots,\zeta_{d}^{k_{m}}) and \lceil n-\sum_{i}k_{i}/d-1\rceil=1 . The images of the Burau and Gassner representations were studied extensively by Venkataramana in [ 18 ] and [ 17 ] , where he demonstrated that these images are arithmetic in a wide range of cases. In a related direction, Looijenga [ 10 ] investigated Abelian covers of surfaces of genus at least 2. He demonstrated that the image of the representation of the liftable mapping class group on the cohomology of the cover is an arithmetic group in all but a few cases in genus 2. Future directions It will be very interesting to investigate the image and kernel of \theta_{\rho} for different partitions \Lambda and different roots of unity \rho . The following question is open but notoriously difficult: Question 1.1 . What is the kernel of \theta_{\rho} ? What is the image \theta_{\rho}(\mathcal{B}_{n,\Lambda}) in \U(r,s) ? An exhaustive account seems out of reach right now but small cases can be analysed to start with. We plan to take this up in a subsequent paper. In a different direction it will be worthwhile to explore the structures on the moduli space M_{0,n} of genus zero Riemann surfaces with n marked points that are induced by \theta_{\rho} . This was explored in the cyclic case by McMullen in sections 7 and 8 of [ 14 ] . Finally in the future we would like to generalise our work to the most general Abelian covers \Big\{(x,y_{1},\ldots,y_{m})\in\mathbb{C}^{m+1}\mid y_{j}^{d_{j}}=(x-p_{h_{j-1}+1})^{a_{h_{j-1}+1}}\cdots(x-p_{h_{j}})^{a_{h_{j}}}\Big\}. 2. Topological Construction of the Monodromy Mixed braid groups Let \mathcal{B}_{n} be Artin’s braid group on n strands defined as, \mathcal{B}_{n}:=\Pi_{1}((\mathbb{C}^{n}\setminus\Delta)/S_{n}) where \Delta=\{(z_{1},...,z_{n})\in\mathbb{C}^{n}\mid z_{i}=z_{j}\text{ for some }i\neq j\} is the big diagonal and S_{n} is the symmetric group