Bochner-type Theorems for Distributional Category
Ekansh Jauhari, John Oprea · 2025
Read Bochner-type Theorems for Distributional Category by Ekansh Jauhari, John Oprea (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-2505-21763.txt.
Introduction
Ekansh Jauhari and John Oprea use a topology measure to bound shape traits under curvature rules, in a 2025 mathematics paper. The measure, called distributional category, counts how widely a space can be covered in a certain loose sense. Bochner-type results link geometry to topology: when curvature is kind, some counts must stay low. The authors extend that spirit to this newer measure.
They show the measure caps traits like the first Betti number and macroscopic dimension from above. When the cap is hit exactly, the shape faces extra constraints, just as in the classic results. The measure also bounds the rank of a group tied to loops in the space. Settings with a symplectic flavor give sharper numbers, including for the older Lusternik-Schnirelmann category.
Catalog record
| CREATOR | Ekansh Jauhari, John Oprea |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 45 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2505-21763
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Bochner-type Theorems for Distributional Category Ekansh Jauhari, John Oprea Abstract . We show that in the presence of a geometric condition such as non-negative Ricci curvature, the distributional category of a manifold may be used to bound invariants, such as the first Betti number and macroscopic dimension, from above. Moreover, à la Bochner, when the bound is an equality, special constraints are imposed on the manifold. We show that the distributional category of a space also bounds the rank of the Gottlieb group, with equality imposing constraints on the fundamental group. These bounds are refined in the setting of cohomologically symplectic manifolds, enabling us to get specific computations for the distributional category and LS-category. 2020 Mathematics Subject Classification Primary 57N65, Secondary 55M30, 20J06, 53C23. 1. Introduction While Lusternik–Schnirelmann (LS) category was originally invented to provide an estimate for the number of critical points of a smooth function on a manifold, it soon became apparent that the concept connected in non-trivial ways with other topological invariants (see [ CLOT03 ] ). The more recent development of the idea of topological complexity and its relation to the motion planning problem (see [ Far08 ] ) spawned a re-evaluation of LS-category in that context, and this resulted in a new “categorical” notion called distributional category (see [ DJ24 ] and [ KW24 ] , where the term analog category is used in the latter). LS-category and distributional category are most certainly different. For instance, the LS-category of any real projective space {{\mathbb{R}}P^{n}} is n while the distributional category is 1 for all n . Yet, intriguingly, sometimes \dcat(X)=\cat(X) (where we have used the obvious notations). In particular, from [ KW24 ] (also see [ Dra ] ), we know that for a torsion-free discrete group \pi with finite cohomological (and homotopy) dimension k , we have \dcat(\pi)=\cat(\pi)=k . In [ Jau25 ] , the first author gave some conditions under which \dcat(M)=\cat(M) holds when M is a closed (not necessarily aspherical) manifold. In this paper, we would like to explore other situations where the above equality holds, or at least where natural topological or geometric conditions imply something interesting about \dcat . We refer to these conditions as Bochner-type conditions for reasons that will become apparent below. The results of our paper suggest close relationships between distributional category and various classical invariants of interest, such as the first Betti number, the Gottlieb group, LS-category, macroscopic dimension, etc. Furthermore, \dcat affords us an opportunity to improve upon earlier results in the literature that connect these classical invariants (see, for example, [ Opr02a ] , [ Opr02b ] , [ OS10 ] , and [ DCDJ ] ). Our results are strongest in the setting of c -symplectic manifolds and non-negatively curved manifolds, where they also give us various new estimates and computations of \dcat . In particular, unlike the results of [ DJ24 ] and [ Jau25 ] , we obtain first examples of inessential manifolds M (in the sense of [ Gro83 ] ) for which \dcat(M)=\cat(M) holds due to various topological, group-theoretical, and geometric conditions rather than purely cohomological ones. Just as LS-category has found applications in many areas of topology, geometry, and critical point theory, we hope that the results of our paper will hint at similar interconnectedness of distributional category, and at least give some intuition about its behavior on spaces with extra structures. We also mention several directions for further study that the reader may find interesting (see, for instance, Remark 5.6 ). In this paper, all topological spaces are assumed to be path-connected. 2. Preliminaries In order to make this paper somewhat self-contained, we must review some foundational topics. 2.A. Topological basics Let R be a commutative ring with unity. Then the cup-length of the cohomology ring H^{*}(X;R) , denoted c\ell_{R}(X) , is defined to be the largest integer k for which there exist \alpha_{i}\in H^{s_{i}}(X;R) , 1\leq i\leq k , s_{i}>0 , such that \alpha_{1}\alpha_{2}\cdots\alpha_{k}\neq 0 . In this paper, we reserve the notation \cu(X) for the rational cup-length c\ell_{{\mathbb{Q}}}(X) . Let \Gamma be a discrete group. Then its cohomological dimension , denoted \cd(\Gamma) , is defined to be the largest integer k such that H^{k}(\Gamma,A)\neq 0 for some {\mathbb{Z}}\Gamma -module A . The Eilenberg–Ganea theorem and the Stallings–Swan theorem (see [ Bro82 ] ) together say that for \cd(\Gamma)=k\neq 2 , there is an Eilenberg–Mac Lane space K(\Gamma,1) of dimension k , and this is the smallest such dimension. 2.B. LS-category The Lusternik–Schnirelmann category (LS-category) of a space X , denoted \cat(X) , is defined to be the smallest integer n for which there is a covering \{U_{i}\} of X by n+1 open sets that are contractible in X . In particular, \cat(X)=0 if and only if X is contractible. The following are some well-known facts about Lusternik–Schnirelmann category, whose proofs can be found in [ CLOT03 ] . 2.1 Theorem . Let X be CW a complex. (1) If Y is homotopy equivalent to X (i.e., Y\simeq X ), then \cat(Y)=\cat(X). (2) If X is r -connected for r\geq 0 , then \cat(X)\leq\tfrac{\dim(X)}{r+1} . (3) For any CW complex Y , \cat(X\times Y)\leq\cat(X)+\cat(Y) . (4) If X^{\prime} is a covering space of X , then \cat(X^{\prime})\leq\cat(X) . (5) For any ring R , c\ell_{R}(X)\leq\cat(X) . In particular, \cu(X)\leq\cat(X) . These properties of \cat will be quite useful for our purpose. In particular, Properties (2) and (5) tell us immediately that for each n , \cat({\mathbb{R}}P^{n})=n=\cat({\mathbb{C}}P^{n}) . A harder fact is that \cat(X)=1 if and only if X is a co- H -space, see [ CLOT03 ] . 2.C. Distributional category Recently, a probabilistic variant of LS-category was introduced in [ DJ24 ] and [ KW24 ] . This is the distributional category (or analog category in [ KW24 ] ). We refer to [ KW24 ] and [ Jau25 ] for motivations behind this new invariant. Given a pointed metric space (X,x_{0}) , the distributional category of X , denoted \dcat(X) , is defined to be the smallest integer n for which there exists a map H:X\to{\mathcal{B}}_{n+1}(P_{0}(X)) such that H(x)(0)\in{\mathcal{B}}_{n+1}(P(x,x_{0})) for each x\in X . Here, P_{0}(X)=\{\phi\in X^{[0,1]}\mid\phi(1)=x_{0}\} and P(x,x_{0})=\{\phi\in P_{0}(X)\mid\phi(0)=x\} have the (metrizable) subspace topology coming from the compact-open topology of X^{[0,1]} , and {\mathcal{B}}_{n+1}(Z) is the space of probability measures on Z supported by at most n+1 points, equipped with the Lévy–Prokhorov metric, [ Pro56 ] . We note that like \cat(X) (see [ CLOT03 , Lemma 1.25] ), \dcat(X) is also independent of the choice of the basepoint x_{0}\in X , see also [ DJ24 , Definition 3.1] . For a proof of the following, we refer to [ DJ24 ] . 2.2 Theorem (Dranishnikov–Jauhari) . Let X be a CW complex. (1) \dcat(X)\leq\cat(X) . (2) If Y\simeq X , then \dcat(Y)=\dcat(X) . (3) If X^{\prime} is a covering space of X , then \dcat(X^{\prime})\leq\dcat(X) . (4) If X is a finite CW complex, then \cu(X)\leq\dcat(X) . 2.3 Example . For each n , \dcat(T^{n})=\cat(T^{n})=n . If X is a closed surface, then \dcat(X)=1 if X\simeq S^{2} or X\simeq{\mathbb{R}}P^{2} , and \dcat(X)=\cat(X)=2 otherwise. In general, the distributional category and the LS-category of a space may differ. Indeed, \dcat({\mathbb{R}}P^{n})=1<n=\cat({\mathbb{R}}P^{n}) for each n\geq 2