Liftable Mapping Class Group

• Liftable mapping class groups are finite-index subgroups defined by the condition that mapping classes admit lifts through specific covering maps of surfaces or orbifolds.
• They are characterized by explicit algebraic and homological criteria, including symplectic representations and branch-point permutations across cyclic, branched, and abelian covers.
• Techniques such as Birman–Hilden theory yield concrete generating sets and normal series that reveal the interplay between covering space geometry, topology, and symmetry.

to=arxiv_search uppernarsjson {"query":"liftable mapping class group regular cyclic covers superelliptic branched covers torus disc algebraic mapping class group rigidity", "max_results": 10} A liftable mapping class group is the subgroup of a mapping class group consisting of isotopy classes of homeomorphisms on a base surface or orbifold that admit lifts through a specified covering map. In the formulations used across recent work, if $p:S\to X$ is a finite-sheeted branched or unbranched cover, then a mapping class $f\in \mathrm{Mod}(X)$ is liftable when some representative homeomorphism of $f$ admits a lift to $S$, and the corresponding subgroup is denoted $\mathrm{LMod}_p(X)$ (Dey et al., 2021). This notion appears in several closely related settings—regular cyclic covers of closed surfaces, cyclic branched covers of spheres, balanced superelliptic covers, branched torus covers, covers of surfaces with boundary, regular abelian covers, and branched covers of the disc—where liftability is detected by covering-theoretic, homological, or permutation-theoretic constraints. The subject is also closely linked to Birman–Hilden theory, which relates liftable mapping class groups downstairs to symmetric or normalizing mapping class groups upstairs (Ghaswala et al., 2018).

1. Definition and basic framework

For a finite-sheeted regular branched cover

$p:S_g\to S_{h,n},$

the liftable mapping class group is defined by

$$\mathrm{LMod}_p(S_{h,n})=\{[f]\in \mathrm{Mod}(S_{h,n})\mid f \text{ admits a lift } \tilde f:S_g\to S_g \text{ with } p\circ \tilde f=f\circ p\}$$

(Kapari, 15 Sep 2025). In the closed-surface setting, one likewise considers $p_k:S_{k(g-1)+1}\to S_g$ and defines $\mathrm{LMod}_{p_k}(S_g)\subset \mathrm{Mod}(S_g)$ as the set of mapping classes admitting lifts through $p_k$ (Agarwal et al., 2019). For covers of compact surfaces with boundary, the boundary-fixing condition is essential: a base homeomorphism may lift, but none of its lifts need fix the boundary pointwise, so liftability must be understood in the boundary-relative sense (Ghaswala et al., 2018).

In the sphere-cover setting, the same concept is formulated for cyclic branched covers $f\in \mathrm{Mod}(X)$0 with branch set $f\in \mathrm{Mod}(X)$1, where

$f\in \mathrm{Mod}(X)$2

(Ghaswala et al., 2016). Balanced superelliptic covers provide a particularly structured family: with

$f\in \mathrm{Mod}(X)$3

the liftable groups are denoted $f\in \mathrm{Mod}(X)$4, $f\in \mathrm{Mod}(X)$5, and $f\in \mathrm{Mod}(X)$6 in the closed, marked-point, and one-boundary variants, respectively (Hirose et al., 2022, Omori, 2022).

The Birman–Hilden perspective is central. For regular covers, one studies the symmetric mapping class group upstairs, consisting of fiber-preserving or deck-normalizing mapping classes. In several settings one has an exact sequence identifying the downstairs liftable group as a quotient of the upstairs symmetric group by the deck group, while in the boundary case the identification is direct: $f\in \mathrm{Mod}(X)$7 for finite-sheeted regular covers of compact surfaces with boundary, possibly branched (Ghaswala et al., 2018). For cyclic branched covers arising from a periodic $f\in \mathrm{Mod}(X)$8, one has

$f\in \mathrm{Mod}(X)$9

and $f$0 (Kapari et al., 2023).

2. Birman–Hilden theory and symmetric groups upstairs

A recurrent structural principle is that liftability downstairs corresponds to symmetry upstairs. For balanced superelliptic coverings, if $f$1 is the balanced superelliptic rotation of order $f$2, the balanced superelliptic mapping class group consists of mapping classes represented by homeomorphisms satisfying

$f$3

(Omori, 2022). In that setting, for $f$4,

$f$5

(Omori, 2022). The analogous quotient statements also appear in the presentation-theoretic treatment of balanced superelliptic groups (Hirose et al., 2022).

For periodic mapping classes $f$6 on closed surfaces, Broughton’s characterization yields exact sequences

$f$7

and consequently

$f$8

(Kapari et al., 2023). This places liftable mapping class groups at the interface between branch-data combinatorics, periodic mapping classes, and normalizer/centralizer computations in $f$9.

In the boundary case, the role of the fundamental group must be replaced by a fundamental groupoid. Choosing a finite set of basepoints $S$0, the relevant object is

$S$1

together with a subgroupoid $S$2 coming from the cover. The subgroup $S$3 is defined by preserving $S$4 and acting trivially on the quotient groupoid $S$5, and the groupoid Birman–Hilden theorem gives

$S$6

(Ghaswala et al., 2018). This algebraic replacement is necessary precisely because boundary-fixing liftability is not detected by $S$7 alone.

3. Criteria for liftability

A major theme of the subject is that liftability admits explicit algebraic criteria in concrete covering families.

For the standard regular $S$8-sheeted cyclic cover

$S$9

liftability is detected on mod-$\mathrm{LMod}_p(X)$0 homology by the symplectic representation

$\mathrm{LMod}_p(X)$1

The following are equivalent for $\mathrm{LMod}_p(X)$2: $\mathrm{LMod}_p(X)$3

$\mathrm{LMod}_p(X)$4

and

$\mathrm{LMod}_p(X)$5

(Agarwal et al., 2019, Dey et al., 2021). Equivalently, if $\mathrm{LMod}_p(X)$6, then

$\mathrm{LMod}_p(X)$7

(Agarwal et al., 2019). This identifies $\mathrm{LMod}_p(X)$8 as the subgroup preserving the cyclic set of primitive multiples of $\mathrm{LMod}_p(X)$9 in $p:S_g\to S_{h,n},$0 (Dey et al., 2021).

For the torus-branched covers

$p:S_g\to S_{h,n},$1

the relevant representation is

$p:S_g\to S_{h,n},$2

and the homological image of the liftable subgroup is

$p:S_g\to S_{h,n},$3

(Kapari, 15 Sep 2025). Thus liftability is governed by a congruence condition on the lower-left vector, and

$p:S_g\to S_{h,n},$4

(Kapari, 15 Sep 2025).

For balanced superelliptic covers, liftability is characterized by parity of the induced permutation on branch points. If

$p:S_g\to S_{h,n},$5

is the puncture-permutation map and $p:S_g\to S_{h,n},$6 is the subgroup of permutations that are either parity-preserving or parity-reversing, then

$p:S_g\to S_{h,n},$7

(Omori, 2022, Hirose et al., 2022). In the corresponding older formulation for balanced superelliptic covers, the image is the subgroup $p:S_g\to S_{h,n},$8 consisting of permutations that either preserve parity or reverse parity, giving

$p:S_g\to S_{h,n},$9

(Ghaswala et al., 2016). A finer curve criterion is also available in this family: $$\mathrm{LMod}_p(S_{h,n})=\{[f]\in \mathrm{Mod}(S_{h,n})\mid f \text{ admits a lift } \tilde f:S_g\to S_g \text{ with } p\circ \tilde f=f\circ p\}$$0 and in homology

$$\mathrm{LMod}_p(S_{h,n})=\{[f]\in \mathrm{Mod}(S_{h,n})\mid f \text{ admits a lift } \tilde f:S_g\to S_g \text{ with } p\circ \tilde f=f\circ p\}$$1

(Ghaswala et al., 2016).

For regular abelian covers $$\mathrm{LMod}_p(S_{h,n})=\{[f]\in \mathrm{Mod}(S_{h,n})\mid f \text{ admits a lift } \tilde f:S_g\to S_g \text{ with } p\circ \tilde f=f\circ p\}$$2 with deck group $$\mathrm{LMod}_p(S_{h,n})=\{[f]\in \mathrm{Mod}(S_{h,n})\mid f \text{ admits a lift } \tilde f:S_g\to S_g \text{ with } p\circ \tilde f=f\circ p\}$$3, liftability reduces to preserving the kernel of the induced homology map

$$\mathrm{LMod}_p(S_{h,n})=\{[f]\in \mathrm{Mod}(S_{h,n})\mid f \text{ admits a lift } \tilde f:S_g\to S_g \text{ with } p\circ \tilde f=f\circ p\}$$4

namely

$$\mathrm{LMod}_p(S_{h,n})=\{[f]\in \mathrm{Mod}(S_{h,n})\mid f \text{ admits a lift } \tilde f:S_g\to S_g \text{ with } p\circ \tilde f=f\circ p\}$$5

(Dhanwani et al., 2024). This homological characterization underlies the algorithmic treatment of regular abelian covers.

4. Structural properties: index, normal series, maximality, and rigidity phenomena

Liftable mapping class groups are typically finite-index subgroups of ambient mapping class groups. For finite-sheeted branched covers of hyperbolic surfaces, $$\mathrm{LMod}_p(S_{h,n})=\{[f]\in \mathrm{Mod}(S_{h,n})\mid f \text{ admits a lift } \tilde f:S_g\to S_g \text{ with } p\circ \tilde f=f\circ p\}$$6 is finite-index in $$\mathrm{LMod}_p(S_{h,n})=\{[f]\in \mathrm{Mod}(S_{h,n})\mid f \text{ admits a lift } \tilde f:S_g\to S_g \text{ with } p\circ \tilde f=f\circ p\}$$7 (Dey et al., 2021). In the boundary setting, $$\mathrm{LMod}_p(S_{h,n})=\{[f]\in \mathrm{Mod}(S_{h,n})\mid f \text{ admits a lift } \tilde f:S_g\to S_g \text{ with } p\circ \tilde f=f\circ p\}$$8 is always finite-index in $$\mathrm{LMod}_p(S_{h,n})=\{[f]\in \mathrm{Mod}(S_{h,n})\mid f \text{ admits a lift } \tilde f:S_g\to S_g \text{ with } p\circ \tilde f=f\circ p\}$$9 (Ghaswala et al., 2018). For torus-branched covers $p_k:S_{k(g-1)+1}\to S_g$0, the exact index is $p_k:S_{k(g-1)+1}\to S_g$1 (Kapari, 15 Sep 2025). For balanced superelliptic covers with $p_k:S_{k(g-1)+1}\to S_g$2,

$p_k:S_{k(g-1)+1}\to S_g$3

(Ghaswala et al., 2016).

The cyclic closed-surface covers admit a congruence-type normal series

$p_k:S_{k(g-1)+1}\to S_g$4

with

$p_k:S_{k(g-1)+1}\to S_g$5

(Agarwal et al., 2019, Dey et al., 2021). This is presented as a higher-genus analog of

$p_k:S_{k(g-1)+1}\to S_g$6

(Agarwal et al., 2019). For the infinite ladder cover $p_k:S_{k(g-1)+1}\to S_g$7,

$p_k:S_{k(g-1)+1}\to S_g$8

(Dey et al., 2021).

Several families exhibit strong subgroup-theoretic behavior. For $p_k:S_{k(g-1)+1}\to S_g$9,

$\mathrm{LMod}_{p_k}(S_g)\subset \mathrm{Mod}(S_g)$0

so the cyclic-cover liftable groups form an infinite family of self-normalizing subgroups, and if $\mathrm{LMod}_{p_k}(S_g)\subset \mathrm{Mod}(S_g)$1 is prime they are maximal in $\mathrm{LMod}_{p_k}(S_g)\subset \mathrm{Mod}(S_g)$2 (Dey et al., 2021). The torus-branched family satisfies the parallel criterion

$\mathrm{LMod}_{p_k}(S_g)\subset \mathrm{Mod}(S_g)$3

(Kapari, 15 Sep 2025).

At the opposite extreme, some covers have every mapping class lift. For finite-sheeted regular covers of compact surfaces with boundary, possibly branched, one has

$\mathrm{LMod}_{p_k}(S_g)\subset \mathrm{Mod}(S_g)$4

if and only if the cover is a Burau cover (Ghaswala et al., 2018). In the hyperelliptic case $\mathrm{LMod}_{p_k}(S_g)\subset \mathrm{Mod}(S_g)$5, for cyclic branched sphere covers,

$\mathrm{LMod}_{p_k}(S_g)\subset \mathrm{Mod}(S_g)$6

(Ghaswala et al., 2016). More generally, for spherical cyclic data, one recovers the criterion

$\mathrm{LMod}_{p_k}(S_g)\subset \mathrm{Mod}(S_g)$7

(Kapari et al., 2023).

5. Generating sets and presentations

A prominent development in the subject is the derivation of explicit finite generating sets.

For regular cyclic covers of closed surfaces with $\mathrm{LMod}_{p_k}(S_g)\subset \mathrm{Mod}(S_g)$8, $\mathrm{LMod}_{p_k}(S_g)\subset \mathrm{Mod}(S_g)$9,

$p_k$0

(Dey et al., 2021). Here $p_k$1 consists of explicit coset representatives, $p_k$2 is a lift of a finite generating set for $p_k$3, and $p_k$4 is a collection of bounding pair maps (Dey et al., 2021). The genus-$p_k$5, degree-$p_k$6 exceptional case is

$p_k$7

(Dey et al., 2021).

For cyclic branched covers of $p_k$8, the uniform theorem is

$p_k$9

for every $f\in \mathrm{Mod}(X)$00 (Kapari, 15 Sep 2025). For $f\in \mathrm{Mod}(X)$01, this simplifies to

$f\in \mathrm{Mod}(X)$02

(Kapari, 15 Sep 2025).

For balanced superelliptic covers, a three-generator phenomenon occurs. For $f\in \mathrm{Mod}(X)$03,

$f\in \mathrm{Mod}(X)$04

and for $f\in \mathrm{Mod}(X)$05,

$f\in \mathrm{Mod}(X)$06

are generated by

$f\in \mathrm{Mod}(X)$07

(Omori, 2022). The low-complexity exceptions satisfy

$f\in \mathrm{Mod}(X)$08

(Omori, 2022).

The presentation theory for balanced superelliptic liftable groups is also explicit. In the closed case, generators are

$f\in \mathrm{Mod}(X)$09

subject to commutative relations, conjugation relations, pentagonal relations, and formulas expressing longer twists in terms of $f\in \mathrm{Mod}(X)$10 and adjacent $f\in \mathrm{Mod}(X)$11 (Hirose et al., 2022). The marked-point case adds the relation

$f\in \mathrm{Mod}(X)$12

and the one-boundary case includes $f\in \mathrm{Mod}(X)$13, the parity-reversing generator $f\in \mathrm{Mod}(X)$14, and the boundary/capping relation

$f\in \mathrm{Mod}(X)$15

(Hirose et al., 2022).

For cyclic branched covers of spheres arising from periodic mapping classes, an algorithmic generating theorem is given in terms of three sets of generators. Writing $f\in \mathrm{Mod}(X)$16 for the pure braid generators $f\in \mathrm{Mod}(X)$17, $f\in \mathrm{Mod}(X)$18 for half-twists interchanging branch points with identical local data, and $f\in \mathrm{Mod}(X)$19 for lifts of units in $f\in \mathrm{Mod}(X)$20, one has

$f\in \mathrm{Mod}(X)$21

(Kapari et al., 2023). This recovers the hyperelliptic and balanced superelliptic generating sets as special cases (Kapari et al., 2023).

For regular abelian covers, the emphasis shifts from closed-form formulas to a general generation algorithm using an action on the nonseparating curve graph and a graph-action theorem. This yields explicit generators in concrete cases such as

$f\in \mathrm{Mod}(X)$22

and for a $f\in \mathrm{Mod}(X)$23-cover $f\in \mathrm{Mod}(X)$24,

$f\in \mathrm{Mod}(X)$25

(Dhanwani et al., 2024).

6. Variants, applications, and related notions

The liftable mapping class group appears in several geometrically distinct but structurally related contexts.

In the hyperelliptic setting, the intersection with the hyperelliptic mapping class group can be computed explicitly. For $f\in \mathrm{Mod}(X)$26,

$f\in \mathrm{Mod}(X)$27

(Dey et al., 2021). This uses the hyperelliptic quotient

$f\in \mathrm{Mod}(X)$28

and the Birman–Hilden epimorphism

$f\in \mathrm{Mod}(X)$29

(Dey et al., 2021).

The theory also provides generators for normalizers of deck groups upstairs. For $f\in \mathrm{Mod}(X)$30 in the torus-branched family,

$f\in \mathrm{Mod}(X)$31

and explicit generating sets for these normalizers are obtained (Kapari, 15 Sep 2025). Analogous normalizer-generation results are derived for cyclic and non-cyclic regular abelian covers (Dhanwani et al., 2024).

For irreducible periodic mapping classes, the liftable group on the spherical quotient can be trivial, cyclic of order $f\in \mathrm{Mod}(X)$32, or cyclic of order $f\in \mathrm{Mod}(X)$33, depending on the stabilizer of the generating $f\in \mathrm{Mod}(X)$34-vector under multiplication by units in $f\in \mathrm{Mod}(X)$35 (Kapari et al., 2023). This classification feeds directly into explicit descriptions of $f\in \mathrm{Mod}(X)$36 and $f\in \mathrm{Mod}(X)$37 (Kapari et al., 2023).

The disc-braid setting supplies a related but not identical notion. For a branched cover of the disc

$f\in \mathrm{Mod}(X)$38

a braid $f\in \mathrm{Mod}(X)$39 is liftable if there exists $f\in \mathrm{Mod}(X)$40 such that

$f\in \mathrm{Mod}(X)$41

(Licata et al., 7 Aug 2025). The classical liftable braid group $f\in \mathrm{Mod}(X)$42 is the subgroup of braid group elements liftable with respect to a fixed cover, and the paper extends the classical lifting homomorphism to a groupoid homomorphism

$f\in \mathrm{Mod}(X)$43

for all simple covers of the disc (Licata et al., 7 Aug 2025). A plausible implication is that the braid-theoretic lifting formalism provides a groupoid-level analogue of liftable mapping class groups when the base is a disc with marked branch values.

A useful caution concerns terminology. The automorphism group of a character variety studied in “Algebraic Mapping Class Group Rigidity” is not defined by liftability. There the relevant group is the relative automorphism group $f\in \mathrm{Mod}(X)$44, defined by preserving puncture monodromy traces, and the main result is that it is generated by the image of the mapping class group together with central representations fixing the boundary monodromies (Kim, 13 Aug 2025). This is a rigidity statement for character-variety automorphisms, not a characterization of a liftable subgroup of $f\in \mathrm{Mod}(X)$45 (Kim, 13 Aug 2025).

7. Conceptual summary

Across these works, the liftable mapping class group is a finite-index subgroup singled out by compatibility with a cover, but the mechanism of compatibility depends sharply on the covering family. In regular cyclic covers of closed surfaces, liftability is a symplectic stabilizer condition in $f\in \mathrm{Mod}(X)$46 (Agarwal et al., 2019, Dey et al., 2021). In torus-branched covers, it is a congruence condition in a $f\in \mathrm{Mod}(X)$47-representation (Kapari, 15 Sep 2025). In balanced superelliptic and more general sphere covers, it is controlled by branch-point permutations, parity constraints, and branch-data stabilizers (Ghaswala et al., 2016, Hirose et al., 2022, Omori, 2022, Kapari et al., 2023). In regular abelian covers, it is the subgroup preserving $f\in \mathrm{Mod}(X)$48 in homology, and this admits an algorithmic treatment via finite quotient graphs of the nonseparating curve graph (Dhanwani et al., 2024). In surfaces with boundary, the correct formalism is groupoidal, and the distinction between ordinary lifting and boundary-fixing lifting becomes essential (Ghaswala et al., 2018).

The common structural pattern is that liftable mapping class groups mediate between covering-space topology, algebraic representations of mapping class groups, and the normalizer or symmetric subgroup upstairs. This suggests that the subject is best understood not as a single construction with one universal criterion, but as a family of subgroup-identification problems whose solutions reflect the geometry of the cover and the algebraic structure visible downstairs.