Kulkarni limit sets for cyclic quaternionic projective groups
Sandipan Dutta, Krishnendu Gongopadhyay and Rahul Mondal · 2025
Read Kulkarni limit sets for cyclic quaternionic projective groups by Sandipan Dutta, Krishnendu Gongopadhyay and Rahul Mondal (2025) 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-2505-11968.txt.
Introduction
Dutta, Gongopadhyay, and Mondal worked out limit sets for quaternionic groups in this math paper from twenty twenty-five. The quaternionic projective linear group acts in a natural way on quaternionic projective space. That space is built from quaternion coordinates. The team took each cyclic subgroup, a subgroup made by one map used again and again. For each one, they found its Kulkarni limit set. That set holds the points where the group's action piles up and turns wild. The work brings classic complex dynamics into the quaternionic case.
Catalog record
| CREATOR | Sandipan Dutta, Krishnendu Gongopadhyay and Rahul Mondal |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 66 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2505-11968
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Kulkarni limit sets for cyclic quaternionic projective groups Sandipan Dutta, Krishnendu Gongopadhyay and Rahul Mondal Abstract . We consider the natural action of the quaternionic projective linear group PSL ( n + 1 , ℍ ) \displaystyle\mathrm{PSL}(n+1,\mathbb{H}) on the quaternionic projective space ℙ ℍ n \displaystyle\mathbb{P}^{n}_{\mathbb{H}} . We compute the Kulkarni limit sets for the cyclic subgroups of PSL ( n + 1 , ℍ ) \displaystyle\mathrm{PSL}(n+1,\mathbb{H}) . 2010 Mathematics Subject Classification Primary 20H10; Secondary 15B33, 22E40 1. Introduction One way to study discrete subgroups of Lie groups is through their dynamical properties. A notable example involves Kleinian groups—discrete subgroups of \displaystyle\mathrm{PSL}(2,\mathbb{C}) —which act on the Riemann sphere via Möbius transformations. Their orbits, if infinite, accumulate on a boundary-invariant subset called the limit set [ CG , Mas ] , with the group acting properly discontinuously on its complement. The properties of the limit sets carry a striking analogy with the behavior of rational maps of the Riemann sphere; cf. [ Mc ] . A natural question in higher-dimensional dynamics concerns the behavior of discrete subgroups of \displaystyle{\rm PSL}(n,\mathbb{C}) . With the motivation to build a connection between the dynamics of projective automorphisms of \displaystyle\mathbb{C}\mathrm{P}^{n} and the discrete subgroups of \displaystyle{\rm PSL}(n,\mathbb{C}) , Seade and Verjovsky [ SV99 ] , [ SV01 ] have initiated investigation of the ‘Complex Kleinian groups’, that is, discrete subgroups of complex projective linear transformations. A central object in this study is the Kulkarni limit set, introduced by Kulkarni in [ Ku ] , who proposed a domain of discontinuity for any locally compact group acting on a locally compact Hausdorff space. Following the foundational work of Seade and Verjovsky, the study of Kulkarni limit sets for discrete subgroups of \displaystyle{\rm PSL}(n,\mathbb{C}) , particularly in the case \displaystyle n=3 , has been further developed by Barrera, Cano, Navarrete, and others; see, for instance, the monograph [ CNS ] and the articles [ BCNS , CS14 , CS10 , Na , Na06 ] . To generalize this framework to the quaternionic setting, we consider discrete subgroups of \displaystyle{\rm PSL}(n,\mathbb{H}) . The group \displaystyle{\rm PSL}(2,\mathbb{H}) corresponds to the isometry group of five-dimensional real hyperbolic space. This naturally leads to the intriguing question of understanding higher-rank projective linear groups over the quaternions. In [ DGL ] , Kulkarni limit sets for cyclic subgroups of \displaystyle{\rm PSL}(3,\mathbb{H}) were classified. The goal of this paper is to extend that classification to Kulkarni limit sets of cyclic subgroups of \displaystyle{\rm PSL}(n,\mathbb{H}) for arbitrary \displaystyle n , thereby generalizing the results of [ DGL ] . It is worth noting that Kulkarni limit sets for cyclic subgroups of complex projective transformations in \displaystyle{\rm PSL}(n,\mathbb{C}) were computed by Cano et al. in [ CLU ] . Inspired by their approach, we adopt a similar methodology in this work. However, due to the non-commutativity of the quaternions, the computations require a different treatment. Since \displaystyle{\rm SL}(n+1,\mathbb{H}) is a double cover of \displaystyle{\rm PSL}(n+1,\mathbb{H}) by \displaystyle\{\pm\mathrm{I}_{n}\} , we often lift elements from the projective group to \displaystyle{\rm PSL}(n+1,\mathbb{H}) and consider their matrix representations. The limit sets will be described in the following table. Before that, we first recall the definition of the Kulkarni limit sets. Definition 1.1 . [ Ku ] Let \displaystyle P_{X}=\{A_{\beta}:\;\beta\in B\} be a family of subsets of \displaystyle X (where \displaystyle B is an infinite indexing set). A point \displaystyle p is called a cluster point of \displaystyle P_{X} if for every neighbourhood \displaystyle N(p) of \displaystyle p , the subset of the indexing set \displaystyle B defined as \displaystyle\{\beta\in B:\;N(p)\cap A_{\beta}\neq\emptyset\} is infinite. Consider the natural action of a subgroup \displaystyle G of \displaystyle{\rm SL}(n+1,\mathbb{H}) on the \displaystyle n -dimensional quaternionic projective space \displaystyle X=\mathbb{P}_{\mathbb{H}}^{n} . Consider the following three sets (a) \displaystyle L_{0}(G):= the closure of the set of points of \displaystyle X which have an infinite isotropy group, (b) \displaystyle L_{1}(G):= the closure of the cluster points of orbits of points in \displaystyle X\setminus L_{0}(G) , and (c) \displaystyle L_{2}(G):= the closure of the cluster points of \displaystyle\{g(K)\}_{g\in G} , where \displaystyle K runs over all the compact subsets of \displaystyle X\setminus\{L_{0}(G)\cup L_{1}(G)\} . These sets are called the Kulkarni sets. Definition 1.2 . [ Ku ] With \displaystyle G and \displaystyle X as defined above, the Kulkarni limit set of \displaystyle G is (1.1) \Lambda(G):=L_{0}(G)\cup L_{1}(G)\cup L_{2}(G). The Kulkarni domain of discontinuity of \displaystyle G is defined as \displaystyle\Omega(G)=X\setminus\Lambda(G) . We remark that the Definition 1.2 is valid for any group \displaystyle G acting on a locally compact Hausdorff space \displaystyle X and Kulkarni proved the following proposition in [ Ku ] . Proposition 1.3 (cf. [ Ku ] ) . Let \displaystyle X be a locally compact Hausdorff space and \displaystyle G be a group acting on \displaystyle X , then \displaystyle G acts properly discontinuously on \displaystyle\Omega(G) . In addition, \displaystyle\Omega(G) is an open subset of \displaystyle X . Further, if \displaystyle\Omega(G)\neq\emptyset , then \displaystyle G is discrete. To classify the limit sets, we pick up Jordan forms in \displaystyle\mathrm{SL}(n,\mathbb{H}) and compute the limit set for the cyclic group generated by that Jordan form. The elements of \displaystyle{\rm PSL}(n+1,\mathbb{H}) can be divided into three mutually exclusive classes as follows. Definition 1.4 . Let \displaystyle g be an element in \displaystyle{\rm PSL}(n+1,\mathbb{H}) . (i) \displaystyle g is called elliptic if it is semisimple and the eigenvalue classes are represented by unit modulus quaternionic numbers. (ii) \displaystyle g is called loxodromic if it is semi-simple and not all the eigenvalue classes are represented by unit modulus quaternionic numbers. (iii) \displaystyle g is called parabolic if it is not semi-simple and eigenvalues have unit modulus quaternionic numbers. (iv) \displaystyle g is called loxoparabolic if it is not semi-simple and at least one of the eigenvalues do not have unit modulus. These classes can be further refined. An elliptic is rational if all of \displaystyle\alpha_{i} ’s are rational, where \displaystyle e^{2\pi i\alpha_{i}} ’s are representative eigenvalues of \displaystyle g . Finally, we remark that the results in the quaternionic setting differ in several instances from those in the complex setting, as presented in [ CLU ] . Even when the outcomes coincide, the computational methods diverge due to the non-commutative structure of the quaternions. Notably, the computation of the first Kulkarni limit set \displaystyle L_{0}(G) often deviates from the corresponding results in \displaystyle{\rm PSL}(n+1,\mathbb{C}) (cf. [ CLU ] ). Table 1 summarizes the Kulkarni limit sets for the cyclic subgroups of \displaystyle{\rm PSL}(n+1,\mathbb{H}) . Kulkarni Limit Sets for Cyclic Subgroups of \displaystyle\mathrm{PSL}(n+1,\mathbb{H}) Conjugacy Type of Matrix Jordan Form \displaystyle\mathbf{\Lambda(G)} Rational Elliptic , Th. 3.2 \displaystyle\mathrm{D}(\lambda_{1},\ldots,\lambda_{n+1}) \displaystyle\emptyset Simple Irrational Elliptic , Th. 3.2 \displaystyle\mathrm{D}(\lambda_{1},\ldots,\lambda_{n+1}) \displaystyle\mathbb{P}^{n}_{\mathbb{H}} Compound Elliptic , Th. 3.2 \displaystyle\mathrm{D}(\lambda_{1},\ldots,\lambda_{n+1}) \displaystyle\mathbb{P}^{n}_{\mathbb{H}} Parabolic (Type 1) , Th. 4.1 \displaystyle\mathrm{J}(\lambda,n+1) \displaystyle L\{e_{1},\ldots,e_{n}\} Parabolic (Type 2) , Th. 4.2 \displaystyle\begin{bmatrix}\mathrm{D}(\lambda_{1},\ldots,\lambda_{k})&\\ &\mathrm{J}(1,l)\end{bmatrix} \displaystyle L\{e_{1},\ldots,e_{k+l-1}\} Parabolic (Type 3) , Th. 4.3 \displaystyle\begin{bmatrix}\mathrm{J}(1,l)&\\ &\mathrm{J}(1,l)\end{bmatrix} \displaystyle L\{e_{1},\ldots,e_{l-1},e_{l+1}\ldots,e_{2l-1}\} Parabolic (Type 4) , Th. 4.5 \displaystyle\begin{bmatrix}\mathrm{J}(1,k)&\\ &\mathrm{J}(1,l)\end{bmatrix} \displaystyle L\{e_{1},\ldots,e_{k-1},e_{k+1},\ldots,e_{k+l-1}\} Loxodromic (Type 1) , Th. 5.1 \displaystyle\mathrm{D}(\lambda_{1},\lambda_{2},\ldots,\lambda_{n+1}) \displaystyle L\{e_{1},e_{2},\ldots,e_{n}\}\cup \displaystyle L\{e_{2},\ldots,e_{n+1}\} Loxodromic (Type 2) , Th. 5.2 \displaystyle\mathrm{D}(\lambda_{1},\ldots,\lambda_{m},\mu_{1},\ldots,\mu_{p}) \displaystyle L\{e_{1},e_{2},\ldots,e_{n}\}\cup \displaystyle L\{e_{m+1},e_{m+2},\ldots,e_{n+1}\} Loxoparabolic , Th. 6.1 \displaystyle\begin{bmatrix}\lambda_{1}\mathrm{J}(1,k)&\\ &\lambda_{2}\mathrm{J}(1,l)\end{bmatrix} \displaystyle L\{e_{1},\ldots,e_{k-1},e_{k+1},\ldots,e_{k+l}\}\cup \displaystyle L\{e_{1},e_{2},\ldots,e_{k+l-1}\} Table 1. Kulkarni sets for cyclic subgroups of elliptic and other elements in \displaystyle\mathrm{PSL}(n+1,\mathbb{H}) 1.1.