Using Erdős’s methods to study Yorke’s problems

Eran Igra and Valerii Sopin · 2025

Read Using Erdős’s methods to study Yorke’s problems by Eran Igra and Valerii Sopin (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-2509-06852.txt.

Introduction

Routes to chaos in high-dimensional systems may have no bound on their complexity, argue Eran Igra and one co-author in a 2025 dynamical systems paper. The authors revisit problems posed by Yorke using methods in the spirit of Erdős, the famously probabilistic mathematician. Periodic orbits and their bifurcations are the objects of study.

Bifurcations are analyzed as graphs: the authors build collections of graphs that idealize bifurcation diagrams, then color them using the Mallet-Yorke orbit index together with the Lefschetz fixed point theorem. That machinery lets them study how generic routes to chaos are and how complex those routes can become. The headline reading is stark: in high-dimensional systems, a route to chaos can be arbitrarily complex, with no ceiling on that complexity. Graph-based idealizations thus expose wild behavior that direct simulation might never map.

Catalog record

CREATOREran Igra and Valerii Sopin
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE90 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2509-06852

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Using Erdős’s methods to study Yorke’s problems Eran Igra and Valerii Sopin Abstract . In this paper, we study the possible bifurcations of periodic orbits by reducing them to graphs. The aforementioned allows to study the genericity of routes to chaos, as well as to analyze their possible complexity. In particular, our results show that there is no upper bound on the possible complexity for dimension d > 3 d>3 . Moreover, it suggests general fermionic description via the virial expansion and universal description via the Rado graph. 1. Introduction Let S be a closed, smooth, orientable manifold of dimension d>1 , and let f_{t}:S\to S , t\in[0,1] , be an orientation-preserving smooth isotopy. Moreover, assume the dynamics of f_{0} is ”simple”, while the dynamics of f_{1} is ”complex” - for example, a.e. initial condition on S is attracted to a stable periodic orbit under f_{0} , while f_{1} includes infinitely many sinks, as per the Newhouse Phenomenon [ 15 ] . Can we characterize the bifurcations that lead from the orderly states of f_{0} to the complex behavior of f_{1} ? Or, more precisely, can we tell which bifurcations we should expect for a generic choice of such an isotopy? One famous approach to studying this problem is that of the Orbit Index, originally introduced in [ 20 ] , inspired by ideas from [ 10 ] , [ 11 ] , [ 12 ] , and [ 13 ] (among others). This notion, as developed further in [ 21 ] , [ 22 ] , and [ 24 ] , was well suited to the study of such questions. One important application is [ 26 ] , where it was proven that when f_{1} is a Smale Horseshoe map, one should generically expect the bifurcations from f_{0} to f_{1} to include infinitely many period-doubling cascades. Other interesting examples are given in works [ 27 ] , [ 30 ] and [ 32 ] , where the Orbit Index was generalized from periodic orbits to homoclinic trajectories and used to study the bifurcations of flows. \begin{overpic}[images/graph1.png] \end{overpic} Figure 1. A (partial) illustration of a period-doubling route to chaos as a collection infinitely many trees, whose edges are colored w.r.t. the Orbit Index (see Section 2 and [ 26 ] ). Based on the Orbit Index Theory (and in particular by the results of [ 26 ] ), in this paper we tackle these problems using Graph Theory. In detail, given any dimension d>1 and k>0 we construct G^{A}_{k,d} , a collection of colored acyclic graphs - i.e., colored trees - which encode the transition from order to chaos (where the number k describes the maximal number of distinct periodic orbits every periodic orbits is allowed to split into along the isotopy). Following that, we consider the relative dimension of G^{A}_{k,d} within G^{A}_{k+1,d} , which we denote by \frac{|G^{A}_{k+1,d}|}{|G^{A}_{k,d}|} and treat as direct limit. As we will prove, for all d>3,k>1 we have \frac{|G^{A}_{k+1,d}|}{|G^{A}_{k,d}|}=\frac{|G^{A}_{k+1,d}|}{|G^{A}_{2,d}|}=\frac{|G^{A}_{k,d}|}{|G^{A}_{1,d}|}=\infty , see Theorem 3.3 for the precise details. The meaning of this result is that for all closed, smooth manifolds S of dimension 4 and higher and all k>0 , there are much more routes to chaos involving ”degenerate” bifurcations where periodic orbits split into arbitrarily many distinct orbits than there are routes to chaos involving only saddle-node and period-doubling bifurcations. In other words, in higher dimensions the period-doubling route to chaos is just one out of many, many others. As will be made clear, our results apply only to four- and higher-dimensional systems. In detail, we show that in dimensions 2 and 3 the arguments used to prove Theorem 3.3 no longer apply. Nevertheless, as we will prove, the relative dimension in these cases strongly depends on how much the two- and three-dimensional bifurcations ”fill up” G^{A}_{k,4} , k>0 (see Corollary 3.4 for the precise formulation). Intuitively, these results should be understood as a quantification of how much three-dimensional bifurcations are an intermediate stage between two- and four-dimensional bifurcation scenarios. At this point we remark our results correlate well with observations made in several numerical and theoretical studies. For example, in [ 36 ] it was numerically observed how many routes to chaos need not be the classical period-doubling one. Similarly, in [ 57 ] it was observed how the introduction to mixed convection in a flow could alter the type of route to chaos. In a more theoretical direction, in [ 33 ] it was concluded that there exist routes to chaos whose dynamics can be much more complex than the regular period-doubling route. Finally, as noted in [ 43 ] , as far as one can talk of a ”likely” route to chaos, in higher-dimensional systems these routes to chaos appear to be based on a Neimark-Sacker bifurcations, as opposed to the usual period-doubling one. This paper is organized as follows. In Section 2 we remind several basic facts about the Orbit Index, use it to translate families of isotopies into colored graphs (see Propositions 2.3 - 2.5 ), and rigorously define bifurcation diagrams (see Definition 2.1 ). In Subsection 3.1 we study the relative size of G^{A}_{k,d} in G^{A}_{r,d} , k<r , proving Theorem 3.3 , and discussing what happens in dimensions 2 and 3 . Following that, in Subsection 3.2 we consider an intermediate scenario between general graphs and trees. Finally, we discuss the wider context of our results, both in Topological Dynamics and possibly also in Statistical Mechanics. Before we begin, we would like to state that even though it is not clear from the text, our ideas were very much inspired by the results of [ 35 ] . To elaborate, recall the problem studied in [ 35 ] was constructing a measure that quantifies the size of generic families of diffeomorphisms on manifolds. As proven in [ 35 ] , even though there is no complete answer to this question, one can, in general, check when a certain collection of such diffeomorphisms is negligible. These ideas inspired to study the analogous question for generic bifurcations, which eventually led to Theorem 3.3 . A final source of inspiration comes from the ways graphs are used in tropical geometry [ 51 ] , which gave us the motivation to apply these methods to bifurcation diagrams. 2. Bifurcation Laws for Diffeomorphisms From now on, S will always denote a closed orientable manifold of dimension d>1 . Now, consider an orientation preserving diffeomorphism f:S\to S . Let x\in S be a periodic orbit of minimal period n , and denote by D_{f^{n}}(x) the differential of f^{n} at some point of the periodic orbit x . Let \sigma^{+} and \sigma^{-} denote the number of eigenvalues of D_{f^{n}}(x) at (1,\infty) and (-\infty,-1) , respectively - it is easy to see these quantities are independent of the choice of a point in the periodic orbit x . We define \phi(x) , the Orbit Index of x , as follows: \phi(x)=\begin{cases}(-1)^{\sigma^{+}},&\text{if $\sigma^{-}$ is even}\\ 0,&\text{if $\sigma^{-}$ is odd}\end{cases} \begin{overpic}[images/fig2.png] \end{overpic} Figure 2. From left to right - a type 0 orbit, type 1 , type 2 , and type m\geq 3 , represented by the black dot. The orbit index is red when it is -1

Read the complete text