Coloring equilateral triangles

Jindřich Zapletal · 2025

Read Coloring equilateral triangles by Jindřich Zapletal (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-2503-18295.txt.

Introduction

Jindřich Zapletal writes a short 2025 paper in set theory, a branch of mathematics that studies the foundations of logic itself. The paper asks which patterns of colored points must appear in spaces of points, under weak added assumptions.

Zapletal shows that, if a very large kind of infinite number exists, two claims can hold at once. The pattern formed by triangles with all sides equal can be colored with only countably many colors in a tidy way. The pattern formed by triangles with two equal sides in the plane cannot. Such results map the border between what can and cannot be proved from the usual rules of mathematics.

Catalog record

CREATORJindřich Zapletal
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE45 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2503-18295

Browse the full library · About the Tower of Babel

Full text

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

Coloring equilateral triangles Jindřich Zapletal Abstract It is consistent relative to an inaccessible cardinal that ZF+DC holds, the hypergraph of equilateral triangles on a given Euclidean space has countable chromatic number, while the hypergraph of isosceles triangles on ℝ 2 \mathbb{R}^{2} does not. 2020 AMS subject classification 03E35, 14P99, 05C15. 1 Introduction This paper deals with chromatic numbers of hypergraphs of arity three on Euclidean spaces. It has been known since the 1960’s that the chromatic number of the hypergraph of equilateral triangles in \mathbb{R}^{2} is countable in ZFC [ 2 ] . Erdős asked whether the same is true for the hypergraph of isosceles triangles in \mathbb{R}^{2} . On the way towards the solution of this problem, Schmerl first [ 6 ] gave a brief argument that the chromatic number of the hypergraph of equilateral triangles in any dimension is countable, later [ 7 ] gave an affirmative answer to the question of Erdős, and eventually classified all algebraic hypergraphs of countable chromatic number in a major breakthrough [ 8 ] . This paper is a commentary on this development of events. I will show that Schmerl’s theorems appeared in order of increasing difficulty. Namely, I will prove Theorem 1.1 . If the theory ZFC+there is an inaccessible cardinal is consistent, then so is ZF+DC+for every number d\geq 1 , the hypergraph of equilateral triangles in dimension d has countable chromatic number+the hypergraph of isosceles triangles in \mathbb{R}^{2} does not have countable chromatic number. Theorem 1.1 greatly understates the understanding of the resulting models of ZF+DC. They are obtained as balanced extensions of the choiceless Solovay model. Therefore, as a matter of general properties of such models, they do not contain e.g. maximal almost disjoint families [ 4 , Theorem 14.1.1] . The posets generating them are (4,3) -balanced (Corollary 3.12 ) and as a result, the models do not contain any discontinuous homomorphisms between Polish groups [ 4 , Theorem 13.2.1] . In the model, every nonmeager (or non-null, in the sense of the usual two-dimensional Lebesgue measure) subset of \mathbb{R}^{2} contains an isosceles triangle. It is possible to achieve the independence result for all dimensions simultaneously. The posets are d -Noetherian balanced, which makes it possible to perform further dimension-specific analysis as in [ 10 ] . The inaccessible cardinal is needed only to initialize the methodology of geometric set theory independence proofs [ 4 , Part II] ; I doubt that it is necessary for the independence result. There are many open questions in the area. The first one can be asked about any algebraic hypergraph without perfect cliques. A general negative answer is known only for arity two, i.e. algebraic graphs without perfect cliques [ 9 ] . The case of equilateral triangles in dimension two is resolved in [ 11 ] . Question 1.2 . Let d\geq 3 be a number. In ZF+DC, does countable chromatic number of hypergraph of equilateral triangles in \mathbb{R}^{d} imply the existence of a Vitali set? The second question deals with a specific type of coloring. Given a hypergraph \Gamma on a topological space X , a \Gamma - free neighborhood assignment is a map c assigning to each point x\in X an open neighborhood of x in such a way that for every hyperedge e\in\Gamma there is a vertex x\in e such that e\not\subseteq c(x) . Question 1.3 . Let d\geq 2 be a number. Is there a balanced extension of the Solovay model in which there is an equilateral-triangle-free neighborhood assignment? As for the architecture of the paper, Section 2 collects the necessary concepts from real algebraic geometry, with the central Definition 2.3 (2). Section 3 constructs the coloring poset generating the model for Theorem 1.1 . Finally, Section 4 shows how to prove that in the resulting model, there is no coloring of isosceles triangles in \mathbb{R}^{2} . The main point is very easy to describe: if \{x_{0},x_{1},x_{2}\} is an isosceles triangle in \mathbb{R}^{2} which is in a suitable sense generic, then V[x_{0},x_{1}]\cap V[x_{1},x_{2}]\cap V[x_{0},x_{2}]=V . On the other hand, if this is an equilateral triangle in any dimension, then the intersection of the three models contains the length of the side of the triangle. The whole argument turns on this simple geometric point. The terminology of the paper sticks to the set theoretic standard of [ 3 ] . For a bounded nonempty subset O of a Euclidean space, I write \mathrm{diam}(O) for the supremum of all Euclidean distances between the points in the set O . 2 Algebraic geometry The proof uses several standard concepts from real algebraic geometry. To begin, recall [ 5 , Section 3.3] that the language of real closed fields includes symbols for addition, multiplication and inequality. The theory of real closed fields includes the axioms of an ordered field, the statement that squares are exactly the non-negative elements, and the statements saying that every nontrivial polynomial of odd degree has a root. The theory of real closed fields admits elimination of quantifiers [ 5 , Theorem 3.3.15] . For algebraic sets, I will need the following. Definition 2.1 . Let n\geq 1 be a number, and let A\subseteq\mathbb{R}^{d} be a set. 1. The set A is algebraic if it is of the form \{z\in\mathbb{R}^{d}\colon p(z)=0\} where p is a polynomial with real coefficients; 2. the set A is irreducible if A\subset\mathbb{R}^{d} is algebraic and whenever A=B\cup C is a union of two algebraic sets, then either B=A or C=A ; 3. [ 1 , Definition 3.3.3] a point x\in A is nonsingular in A if the dimension of the Zariski tangent space to A at x has the smallest possible value \dim(A) , where the dimension is as defined in [ 1 , Definition 2.8.1] . The following fact records the basic results about algebraic sets and their nonsingular points. Fact 2.2 . Let n\geq 1 be a number. 1. (Hilbert basis theorem) There is no infinite sequence of algebraic sets which is strictly decreasing with respect to inclusion; 2. [ 1 , Theorem 2.8.3] every algebraic set A\subseteq\mathbb{R}^{d} has a unique decomposition into irreducible components: A=\bigcup\{A_{i}\colon i\in j\} where j\geq 1 is a number and each sets A_{i} is algebraic, irreducible, and not included in the union of the others; 3. [ 10 , Claim 4.9] if A,B\subseteq\mathbb{R}^{d} are algebraic sets and A is irreducible, then either A\subseteq B or A\cap B is nowhere dense in the set of nonsingular points of A . Coming to the original contents of this paper, the following concept is key. Definition 2.3 . Let n\geq 1 be a number, x\in\mathbb{R}^{d} , and F\subseteq\mathbb{R} . 1. A set A\subset\mathbb{R}^{d} is F -semialgebraic if it is of the form \{z\in\mathbb{R}^{d}\colon\mathbb{R}\models\psi(z)\} where \psi is a a formula of the language of real closed fields with parameters in F ; 2. D(F,x) is the smallest set among the collection of F -semialgebraic sets containing x , which happen to be algebraic. Example 2.4 . Let d=1 , F=\mathbb{Q} , and x=\sqrt{2} . Then D(F,x)=\{\sqrt{2}\} , even though the smallest zero set of a polynomial with rational coefficients containing x is \{-\sqrt{2},\sqrt{2}\} . If the set F\subset\mathbb{R} is a real closed subfield, then the set D(F,x)

Read the complete text