On Distributional One-Category, diagonal distributional complexity, and related invariants
Ekansh Jauhari, John Oprea · 2025
Read On Distributional One-Category, diagonal distributional complexity, and related invariants 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-2508-06973.txt.
Introduction
Ekansh Jauhari and John Oprea build random forms of two classic ways to gauge how complex a space is. The two ways are Lusternik-Schnirelmann category and topological complexity. They count how hard it is to split up a space or to move through it. The random forms can set lower bounds and add new insight, but how they act was only partly known.
The new measures stay rigid on spaces whose fundamental group is finite, unlike past forms. On a key class of spaces they match known distributional invariants. That match makes clear how those invariants act on aspherical spaces and on products. The paper also gives results on covering maps and linked constructions. The authors give the first closed manifolds past projective spaces where the distributional and classical theories disagree.
Catalog record
| CREATOR | Ekansh Jauhari, John Oprea |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 99 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2508-06973
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
On Distributional One-Category, diagonal distributional complexity, and related invariants Ekansh Jauhari, John Oprea Abstract . We develop the theory of probabilistic variants of the one-category and diagonal topological complexity, which bound the classical LS-category and topological complexity from below. Unlike any other classical or probabilistic invariants, these invariants are rigid on spaces with finite fundamental group. On Eilenberg–Mac Lane spaces, we identify these new invariants with distributional category and complexity, respectively, and use them to illuminate aspects of the behavior of the latter invariants on aspherical spaces and products of spaces. We also study their properties on covering maps, π 1 \pi_{1} -isomorphisms, H H -spaces, and closed essential manifolds, and consequently, obtain the first examples of closed manifolds beyond the real projective spaces on which the distributional theory disagrees with the classical one. 2020 Mathematics Subject Classification Primary 55M30, 57N65, Secondary 60B05, 55R05, 57M10. 1. Introduction The subject of Lusternik–Schnirelmann (LS) category goes back almost 100 years. The original idea was to understand the complexity of a space (such as a manifold) by determining the least number of simple pieces that were required to assemble it. This number (or nowadays one less than this number) is then an invariant of (the homotopy type of) the space and serves the usual algebraic topological function of distinguishing homotopically different spaces. Of course, Lusternik and Schnirelmann defined category in order to bound the number of critical points of any smooth function on a manifold from below, and this was essential for their proof of the fact that the topological sphere with any metric possesses at least 3 closed geodesics. Fox is often credited with bringing LS-category into the realm of algebraic topology and, in fact, here we will use a variant of his so-called one-category . From the description above, we see that LS-category is a “molecular theory” in the best reductionist tradition of classical 19 -th century science — see [ CLOT03 ] for a modern approach to LS-category. In the last 20 years or so, a distinct, but intertwined, theory of topological complexity has been developed by Michael Farber [ Fa03 ] and (now) a host of others. This theory has its roots in understanding how difficult the motion planning problem in robotics is for various configuration spaces. This, too, is a molecular theory. Even from the vague descriptions we have just given, it should be apparent that calculating LS-category and topological complexity is a very hard thing to do for most topological spaces, so any approximating invariants we can find for them prove to be useful (and sometimes essential). Recently, some new homotopy invariants of Lusternik–Schnirelmann type have been introduced in [ DJ24 , KW24 ] . They are called distributional category in [ DJ24 ] and analog category in [ KW24 ] , and they provide new lower bounds for LS-category. Both approaches define invariants in terms of probability measures on spaces, but different topologies are used in the definitions. In many cases, it has now been shown that the two notions agree. Here, we shall always refer to distributional category and use the notation \dcat(X) for the distributional category of a space X , but we will also refer to [ KW24 ] for various results when they apply. Distributional category considers probability measures on the based path space P_{0}(X) which lie in a single fiber of the evaluation map to X . Of course, the fiber of this map is the loop space \Omega X , and a probability measure here is a relatively complicated thing. Especially, in the original formulation of distributional category (and topological complexity) [ DJ24 , Ja25a ] , we see a collection of dynamical states (i.e., paths) with associated probabilities (adding to one, of course) which all collapse to a designated final state, and without carrying the analogy too far (we hope), this is reminiscent of the Feynman path-integral approach to Quantum Mechanics. Certainly, we have gone too far here, but the idea of a kind of superposition of categorical states is not too outlandish. Furthermore, at present, there does not seem to be a way of molecularizing the theory by cutting up a given space into “atoms”. In this sense, the theories of distributional category and distributional topological complexity are holistic theories rather than reductionist ones. In this paper, we go a step further by “discretizing” in a well-defined sense. For a space X , we define a new invariant called the distributional one-category , denoted \dcat_{1}(X) , that instead considers probability measures on the universal cover q\colon\widetilde{X}\to X that lie in a single fiber of q . Of course, the advantage here is that the fiber of a covering is discrete, and this entails all sorts of simplifying factors. We prove that \dcat_{1} is a lower bound for \dcat and, in various cases, is equal to \dcat . Because we work with discrete fibers, we can be rather explicit in our proofs, and we hope this makes the subject more transparent. As we show in this paper, \dcat_{1} turns out to be a very useful tool in understanding other distributional invariants. For instance, \dcat_{1} helps us obtain new examples of manifolds (namely, some Lens spaces and projective product spaces) on which the distributional invariants differ from their classical counterparts. Also, it helps disprove a product inequality for the distributional invariants; the question of the existence of such an inequality (whose analogue in the classical theory exists and is very useful for computations) has been open since the inception of these new invariants. Furthermore, owing to its discrete nature, \dcat_{1} offers relatively simpler proofs to some key results in the theory of distributional category of aspherical spaces. These features of \dcat_{1} provide motivation to develop further the theory of distributional one-category and related complexity-type invariants. Structure of the paper This paper is organized as follows. After some background material in Section 2 , we formally define distributional one-category ( \dcat_{1} ) in Section 3 and elucidate its essential properties. In particular, we show that for a covering map f\colon Y\to X , we have \dcat_{1}(Y)\leq\dcat_{1}(X) . In addition, we show that for a discrete group \pi , we have \dcat_{1}(\pi)=\dcat(\pi) , where \dcat is the previously defined distributional category of [ DJ24 ] . This then allows us to analyze \dcat(\pi) from a simpler viewpoint. In Section 4 , we describe cases where \dcat_{1}(X)=\dcat(K(\pi_{1}(X),1)) , and look at possibilities for \dcat_{1}(X) and \dcat_{1}(\pi_{1}(X)) to be arbitrarily far apart. Moreover, we extend a result of [ KW24 ] by showing that if \pi_{1}(X) is finite for a space X , then \dcat_{1}(X)\leq|\pi_{1}(X)|-1 . The powerful influence of the fundamental group on \dcat_{1} is also displayed by the fact that \dcat_{1}(X)=2 for any space X whose fundamental group has a single relator (that is not a proper power, of course). In Section 5 , we use \dcat_{1} to revisit results about \dcat(\pi) for \pi either a finite group or a torsion-free discrete group. In the latter case, we show an Eilenberg–Ganea-type equality \dcat_{1}(\pi)=\cd(\pi) in the spirit of [ KW24 ] . While these results are known, we believe that the \dcat_{1} -approach offers an enlightening perspective. In Section 6 , we study