Martin's Axiom

Total Page:16

File Type:pdf, Size:1020Kb

Martin's Axiom Chapter 14 Martin’s Axiom In this chapter, we shall introduce a set-theoretic axiom, known as Martin’s Axiom, which is independent of ZFC. In the previous chapter we have compared forcing extensions with group extensions. Similarly, we could also compare forcing exten- sions with field extensions. Now, if we start, for example, with the field of rational numbers Q and extend Q step by step with algebraic extensions, we finally obtain an algebraic closure F of Q. Since F is algebraically closed, we cannot extend F with an algebraic extension. With respect to forcing extensions, we have a somewhat similar situation: If we start, for example, with Gödel’s model L, which is a model of ZFC + CH, and extend L step by step with forcing notions of a certain type, we finally obtain a model of ZFC which cannot be extended by a forcing notion of that type. The model we obtain in this way is a model in which Martin’s Axiom holds. In other words, models in which Martin’s Axiom holds are closed under certain forcing extensions, like algebraically closed fields are closed under algebraic extensions. As a matter of fact we would like to mention that in the presence of the Continuum Hypothesis, Martin’s Axiom is vacuously true. However, if the Continuum Hypothesis fails, then Martin’s Axiom becomes an interesting combinatorial statement as well as an important tool in Combinatorics which has many applications in Topology, but also in areas like Analysis and Algebra (see RELATED RESULTS 83 & 84). Filters on Partially Ordered Sets Below, we introduce the standard terminology of partially ordered sets, as it is used in forcing constructions. In the context of forcing, a binary relation “ ” on a set P is a partial ordering of P if it is transitive (i.e., p q and q ≤r implies p r) and reflexive (i.e., p p for every p P ). Notice≤ that we do≤ not require that≤ “ ” is anti-symmetric (i.e.≤, p q and q ∈ p implies p = q), as we have done in Chapter≤ 6. If“ ”is a ≤ ≤ ≤ 341 342 14 Martin’s Axiom partial ordering on P , then (P, ) is called a partially ordered set. If P = (P, ) is a partially ordered set, then≤ the elements of P are usually called conditions≤ , since in the context of forcing, elements of partially ordered sets are conditions for sentences to be true in generic extensions. Two conditions p1 and p2 of P are called compatible, denoted p1 p2, if there exists a q P such that p1 q p2; otherwise they are called incompatible| , denoted p ∈ p . ≤ ≥ 1 ⊥ 2 A typical example of a partially ordered set is the set of finite partial functions with inclusion as their partial ordering: Let I and J be arbitrary sets. Then Fn(I,J) is the set of all functions p such that • dom(p) fin(I), i.e., dom(p) is a finite subset of I, and ∈ • ran(p) J. ⊆ For p, q Fn(I,J) define ∈ p q dom(p) dom(q) q = p. ≤ ⇐⇒ ⊆ ∧ |dom(p) If we consider functions as sets of ordered pairs, as we usually do, then p q is just p q. We leave it as an exercise to the reader to verify that (Fn(I,J), )≤is indeed a partially⊆ ordered set. ⊆ Let P = (P, ) be a partially ordered set, and for the moment let C P . ≤ ⊆ • C is called directed if for any p ,p C there is a q C such that p q p . 1 2 ∈ ∈ 1 ≤ ≥ 2 • C is called open (or upwards closed), if p C and q p implies q C. For example with respect to (Fn(I,J), ), for every∈ x I≥the set p Fn(∈ I,J): x dom(p) is open. ⊆ ∈ { ∈ ∈ } • C is called downwards closed if p C and q p implies q C. ∈ ≤ ∈ • C is called dense if for every condition p P there is a q C such that q p. For example with respect to (Fn(I,J)∈, ), for every x ∈I the set p Fn(≥I,J): x dom(p) is dense. ⊆ ∈ { ∈ ∈ } • Anon-emptyset F P is a filter (on P ) if it is directed and downwards closed. Notice that this definition⊆ of “filter” reverses the ordering from the definition given in Chapter 6. • Let D P(P ) be a set of open dense subsets of P . A filter G P is a D- generic⊆ filter on P if G D = for every set D D. As an example,⊆ consider again (Fn(I,J), ).If F∩is a filter6 ∅ on Fn(I,J), then∈ F : X J is a function, where X is some⊆ (possibly infinite) subset of I. → S PROPOSITION 14.1. If (P, ) is a partially ordered set and D is a countable set of open dense subsets of P , then≤ there exists a D-generic filter on P . Moreover, for every p P there exists a D-generic filter G on P which contains p. ∈ Filters on Partially Ordered Sets 343 Proof. For D = Dn : n ω and p := p, choose for each n ω a pn Dn { ∈ } −1 ∈ ∈ such that pn pn , which is possible since Dn is dense. Then the set ≥ −1 G = q P : n ω(q pn) ∈ ∃ ∈ ≤ is a D-generic filter on P and p G. ∈ ⊣ It is natural to ask whether the restriction on the size of D can be weakened. In particular, one can ask whether D-generic filters also exist in the case when D is uncountable, or whether it is at least consistent that such fiters exist. In order to get a positive answer to these questions, we have to put a restriction on the partial ordering: Let (P, ) be a partially ordered set. A subset A P is an anti-chain in P if any two distinct≤ elements of A are incompatible. As mentioned⊆ in Chapter 6, this definition of “anti-chain” is different from the one used in Order Theory. A partially ordered set P = (P, ) satisfies the countable chain condition, denoted ccc, if every anti-chain in P is≤ at most countable (i.e., finite or countably infinite). Now we are ready to formulate Martin’s Axiom in its general form. Martin’s Axiom (MA). If P = (P, ) is a partially ordered set which satisfies ccc, and D is a set of less than c open dense≤ subsets of P , then there exists a D-generic filter on P . In other words, MA(κ) holds for each cardinal κ< c. If we assume CH, then κ < c is the same as saying κ ω, thus, by PROPO- ≤ SITION 14.1, CH implies MA. On the other hand, MA can replace the Contin- uum Hypothesis in many proofs that use CH. Furthermore, MA is consistent with ZFC + CH, as we shall see in Chapter 19. ¬ Instead of requiring that D < c, we can require that D κ for some cardinal κ. | | | |≤ MA(κ). If P = (P, ) is a partially ordered set which satisfies ccc, and D is a set of at most κ open dense≤ subsets of P , then there exists a D-generic filter on P . On the one hand, MA(ω) is just PROPOSITION 14.1, and therefore, MA(ω) is prov- able in ZFC. On the other hand, MA(c) is just false as we will see in FACT 14.5. So, we cannot generalise MA by omitting D < c. Another attempt to generalise MA would be to omit ccc. However, this attempt| | also fails. FACT 14.2. There exists a (non ccc) partially ordered set P = (P, ) and a set D ≤ of cardinality ω1 of open dense subsets of P such that no filter on P is D-generic. Proof. Consider the partially ordered set (Fn(ω,ω ), ). Notice first that, for ex- 1 ⊆ ample, 0, α : α ω1 is an uncountable anti-chain, and hence, (Fn(ω,ω1), ) does noth satisfyi ccc.∈ Now, for each α ω , the set ⊆ ∈ 1 Dα = p Fn(ω,ω ): α ran(p) ∈ 1 ∈ 344 14 Martin’s Axiom is an open dense subset of Fn(ω,ω1): Obviously, Dα is open. To see that Dα is also dense, take any p Fn(ω,ω1). If α ran(p), then p Dα and we are done. Otherwise, let n ω be∈ such that n / dom(∈ p) (notice that∈ such an n exists since ∈ ∈ dom(p) is finite). Now, let q := p n, α ; then q Dα and q p. Similarly, ∪ {h i} ∈ ≥ for each n ω, the set En = p Fn(ω,ω ): n dom(p) is open dense. ∈ { ∈ 1 ∈ } Let D = Dα : α ω En : n ω ; then D = ω . Assume that G { ∈ 1}∪{ ∈ } | | 1 ⊆ Fn(ω,ω ) is a D-generic filter on Fn(ω,ω ). Since for each n ω, G En = , 1 1 ∈ ∩ 6 ∅ fG = G is a function from ω to ω . Now, since for each α ω , G Dα = , the 1 ∈ 1 ∩ 6 ∅ function fG : ω ω1 is even surjective, which contradicts the definition of ω1. S → ⊣ In order to show that ccc cannot be omitted in MA, we first show that for countable sets J, Fn(I,J) satisfies ccc, and for this, we first prove the following powerful combinatorial result. LEMMA 14.3 (∆-SYSTEM LEMMA). Let E be an uncountable family of finite sets. Then there exists an uncountable family C E and a finite set ∆ such that for any pair of distinct elements x, y C : x y =⊆ ∆. ∈ ∩ Proof. Notice first that either there exists an uncountable E ′ E such that for every a E ′, x E ′ : a x is countable, ⊆ ∈ { ∈ ∈ } S or for every uncountable E ′ E there exists an a E ′ such that x E ′ : a x is uncountable⊆ .
Recommended publications
  • The Dividing Line Methodology: Model Theory Motivating Set Theory
    The dividing line methodology: Model theory motivating set theory John T. Baldwin University of Illinois at Chicago∗ January 6, 2020 The 1960’s produced technical revolutions in both set theory and model theory. Researchers such as Martin, Solovay, and Moschovakis kept the central philosophical importance of the set theoretic work for the foundations of mathematics in full view. In contrast the model theoretic shift is often seen as ‘ technical’ or at least ‘merely mathematical’. Although the shift is productive in multiple senses, is a rich mathematical subject that provides a metatheory in which to investigate many problems of traditional mathematics: the profound change in viewpoint of the nature of model theory is overlooked. We will discuss the effect of Shelah’s dividing line methodology in shaping the last half century of model theory. This description will provide some background, definitions, and context for [She19]. In this introduction we briefly describe the paradigm shift in first order1 model theory that is laid out in more detail in [Bal18]. We outline some of its philosophical consequences, in particular concerning the role of model theory in mathematics. We expound in Section 1 the classification of theories which is the heart of the shift and the effect of this division of all theories into a finite number of classes on the development of first order model theory, its role in other areas of mathematics, and on its connections with set theory in the last third of the 20th century. We emphasize that for most practitioners of late 20th century model theory and especially for applications in traditional mathematics the effect of this shift was to lessen the links with set theory that had seemed evident in the 1960’s.
    [Show full text]
  • Mathematical Society April 2003 Volume 50, Number 4
    ISS N 0002 -9920 of the American Mathematical Society April 2003 Volume 50, Number 4 An Introduction to Analysis on Metric Spaces page 438 Artful Mathematics: The Heritage of M. C. Escher page 446 Filling in Escher's blank space (see page 457) The Open Computer Algebra System ,.;' MuPAD Pro - [DtffEq.mnb] fo:B. MuPAD Pro is a full-featured computer algebra system in an integrated and 0 30 Vtewer - [VCamKiem2.vca] < Edit View Tools Animation Window Help open environment for symbolic and numeric computing . The MuPAD language has a Pascal-like syntax and allows imperative, functional, and object­ oriented programming . Its domains and categories are like object-oriented MuP .AJ) recognized an inhomogenous linear differential equation classes that allow overriding and corresponding homogenous system is expresse~ using a new ide overloading methods and operators, The differential equation y'"'(x) - 5y"(x) + 4 y(x) = 2 cos(x) o inheritance, and generic algorithms. A • solve( ode ( y''''(x)- S*y''(x) + 4*y comfortable notebook interface includes a graphics tool for visualization, an integrated source-level debugger, a profiler, and hypertext help. It is also possible to solve systems of differential equations. Consi Some of the new features in version 2.5 include: f(x) - j(x) + g'(x) + 2 g(x) =l+eA g'(x) + 2 g(x) + h'(x) + h(x) =2 +ex X • High quality 3-D graphics • HTML export Ready • Increased computational capability • Available with Scilab for very fast numerical computations ~MacKichan SOFTWAR E, INC . Tools for Scientific Creativity Since
    [Show full text]
  • How to Choose New Axioms for Set Theory?
    HOW TO CHOOSE NEW AXIOMS FOR SET THEORY? LAURA FONTANELLA 1. Introduction The development of axiomatic set theory originated from the need for a rigorous investigation of the basic principles at the foundations of mathematics. The classical theory of sets ZFC offers a rich framework, nevertheless many important mathemati- cal problems (such as the famous continuum hypothesis) cannot be solved within this theory. Set theorists have been exploring new axioms that would allow one to answer such fundamental questions that are independent from ZFC. Research in this area has led to consider several candidates for a new axiomatisation such as Large Cardi- nal Axioms, Forcing Axioms, Projective Determinacy and others. The legitimacy of these new axioms is, however, heavily debated and gave rise to extensive discussions around an intriguing philosophical problem: what criteria should be satisfied by ax- ioms? What aspects would distinguish an axiom from a hypothesis, a conjecture and other mathematical statements? What is an axiom after all? The future of set theory very much depends on how we answer such questions. Self-evidence, intuitive appeal, fruitfulness are some of the many criteria that have been proposed. In the first part of this paper, we illustrate some classical views about the nature of axioms and the main challenges associated with these positions. In the second part, we outline a survey of the most promising candidates for a new axiomatization for set theory and we discuss to what extent those criteria are met. We assume basic knowledge of the theory ZFC. 2. Ordinary mathematics Before we start our analysis of the axioms of set theory and the discussion about what criteria can legitimate those axioms, we should address a quite radical view based on the belief that `ordinary mathematics needs much less than ZFC or ZF '.
    [Show full text]
  • Whitehead Modules Over Domains Thomas Becker, Laszlo Fuchs1, and Saharon Shelah2 (Communicated by Rüdiger Göbel)
    Sh:317 Forum Math, l (1989), 53-68 Forum Mathematicum © de Gruyter 1989 Whitehead Modules over Domains Thomas Becker, Laszlo Fuchs1, and Saharon Shelah2 (Communicated by Rüdiger Göbel) Abstract. Let R be a commutative domain with 1. By a Whitehead module is meant an R- module M satisfying Ext^(M, R) = 0. If R is such that .RD-submodules of torsion-free Whitehead modules are again Whitehead, then the hypothesis V = L makes it possible to reduce the problem of characterizing torsion-free Whitehead modules to Whitehead modules of cardinality < \R . Proper Forcing is used to show that this criterion fails in ZFC. Applications are given to P.I.D.s of cardinality K1? countable valuation domains and almost maximal valuation domains of cardinality Kt. 1980 Mathematics Subject Classification (1985 Revision): 13C05; 16A53, 03C25. For about two decades, the Whitehead problem was one of the central open problems in abelian group theory. In 1952, J. H. C. Whitehead asked if an abelian group A with Ext1 (A,Z) = Q (now called a Whitehead group) had to be free. (For the connection with other problems, see e. g. Nunke [10].) For countable A, the answer was already available in a paper by Stein [15]; unaware of this, Ehrenfeucht [1] published the (affirmative) solution for countable groups. Several authors were attracted by this problem, but could obtain only fragmentary results. The füll answer was given by Shelah [11] in an unexpected claim: Whitehead's problem is undecidable in ZFC. More precisely, he proved that in L (the constructible universe) all Whitehead groups of cardinality K! are free, while in the presence of Martin's Axiom and the denial of the CH (Continuum Hypothesis), there do exist non-free Whitehead groups of cardinality Kt.
    [Show full text]
  • Martin's Maximum Implies Woodin's Axiom
    Martin's Maximum++ implies Woodin's Axiom (∗) David Asper´o* and Ralf Schindler March 18, 2021 Abstract ++ We show that Martin's Maximum implies Woodin's Pmax axiom (∗). This answers a question from the 1990's and amalgamates two prominent ax- ioms of set theory which were both known to imply that there are @2 many real numbers. 1 Introduction. Cantor's Continuum Problem, which later became Hilbert's first Problem (see [18]), asks how many real numbers there are. After having proved his celebrated theorem @0 according to which R is uncountable, i.e., 2 > @0, see [4], Cantor conjectured that @0 every uncountable set of reals has the same size as R, i.e., 2 = @1. This statement is known as Cantor's Continuum Hypothesis (CH). G¨odel[14] proved in the 1930's that CH is consistent with the standard axiom system for set theory, ZFC, by showing that CH holds in his constructible universe L, the minimal transitive model of ZFC containing all the ordinals. The axiom V = L, saying that the universe V of all sets is simply identical with L, has often been rejected, however, as an undesirable minimalistic assumption about V . For instance, L cannot have measurable cardinals by a result of Scott [43]. G¨odelhimself believed that CH would be shown not to *School of Mathematics, University of East Anglia, Norwich Research Park, Norwich, Norfolk, NR4 7TJ, UK. Funded by EPSRC Grant EP/N032160/1. Institut f¨urMathematische Logik und Grundlagenforschung, Universit¨atM¨unster,Einsteinstr. 62, 48149 M¨unster,FRG. Funded by the Deutsche Forschungsgemeinschaft (DFG, German Research Foundation) under Germany's Excellence Strategy EXC 2044 390685587, Mathematics M¨unster: Dynamics - Geometry - Structure.
    [Show full text]
  • Jensen's Diamond Principle and Its Relatives
    JENSEN’S DIAMOND PRINCIPLE AND ITS RELATIVES ASSAF RINOT Abstract. We survey some recent results on the validity of Jensen’s diamond principle at successor cardinals. We also discuss weakening of this principle such as club guessing, and anti-diamond principles such as uniformization. A collection of open problems is included. Contents Introduction 1 Organization of this paper 2 Notation and conventions 2 1. Diamond 3 2. Weak Diamond and the Uniformization Property 12 3. The Souslin Hypothesis and Club Guessing 23 4. SaturationoftheNonstationaryIdeal 29 References 34 Introduction. Cantor’s continuum hypothesis has many equivalent formu- lations in the context of ZFC. One of the standard formulations asserts the existence of an enumeration {Aα | α<ω1} of the set P(ω). A non-standard, twisted, formulation of CH is as follows: arXiv:0911.2151v2 [math.LO] 22 Jun 2010 (∃)2 there exists a sequence, Aα | α<ω1, such that for every subset Z ⊆ ω1, there exist two infinite ordinals α,β <ω1 such that Z ∩β = Aα. By omitting one of the two closing quantifiers in the above statement, we arrive to the following enumeration principle: (∃)1 there exists a sequence, Aα | α<ω1, such that for every subset Z ⊆ ω1, there exists an infinite ordinal α<ω1 such that Z ∩α = Aα. Jensen discovered this last principle during his analysis of G¨odel’s con- structible universe, and gave it the name of diamond, ♦. In [28], Jensen 2000 Mathematics Subject Classification. Primary 03E05; Secondary 03E35, 03E50. Key words and phrases. Diamond, Uniformization, Club Guessing, Stationary Hitting, Souslin Trees, Saturation, Square, Approachability, SAP.
    [Show full text]
  • Homological Algebra and SET THEORY 211
    TRANSACTIONSOF THE AMERICANMATHEMATICAL SOCIETY Volume 227, 1977 HOMOLOGICALALGEBRA AND SET THEORY BY PAUL C. EKLOF(') Abstract. Assuming the Axiom of Constructibility, necessary and sufficient conditions are given for the vanishing of Ext\ for rings A of global dimension 1. Using Martin's Axiom, the necessity of these conditions is shown not to be a theorem of ZFC. Applications are given to abelian group theory, including a partial solution (assuming V = L) to a problem of Baer on the splitting of abelian groups. Some independence results in abelian group theory are also proved. Introduction. In a remarkable recent paper [22], S. Shelah proved that Whitehead's problem is unsolvable in ordinary set theory (Zermelo-Frankel set theory with the Axiom of Choice). More precisely, he proved that two different answers to the problem are obtained when set theory is extended by adding two different axioms: the Axiom of Constructibility and Martin's Axiom. (See Theorem 4.1.) Shelah's arguments were specific to Whitehead's problem and nonhomological in character. In this paper we extend Shelah's techniques in order to obtain some general results about the vanishing of ExtA (for modules over a ring A of global dimension 1) in different models of set theory, and we apply them to some specific problems in abelian group theory. Assuming the Axiom of Constructibility we obtain necessary and sufficient conditions for the vanishing of ExtA (Theorems 1.2 and 1.5). To state the result in. a special case: if C and A are abelian groups of cardinality N,, Ext(C, A) — 0 if and only if C is the union of an increasing chain of countable subgroups [Cv\v < «,} such that the chain is smooth (i.e., C0 — U„<oC, for aii iimit ordinals a) and satisfies Ext(C0, A) = 0 and E\i(Cv+x/Cv,A) = 0 for all v < K,.
    [Show full text]
  • And Whitehead's Problem
    Universit`adegli studi di Pisa Facolta` di Scienze Matematiche, Fisiche e Naturali Corso di Laurea Triennale in Matematica Tesi di laurea triennale The ranks of Ext1(A, Z) and Whitehead’s Problem Relatore: Correlatore: Professor Matteo Viale Professor Giovanni Gaiffi Candidato: Morena Porzio Anno accademico 2017/2018 ii iii Abstract In this thesis we analyse Whitehead's conjecture, stating that a commutative group 1 A is free if and only if the associated group Ext (A; Z) is trivial. First we introduce the necessary background in homological algebra and logic needed to properly for- mulate and tackle the conjecture. Then we present a characterization of the ranks 1 of Ext (A; Z) for countable abelian groups A due to Chase (1963), which improves the classical theorem of Stein (1951), and solves positively Whitehead's conjecture for the case of countable abelian groups. Finally we discuss the undecidability of the conjecture with respect to the standard axioms of mathematics, focusing on its instantiation for abelian groups of cardinality @1. iv Contents Introduction1 1 Background3 1.1 Algebraic Prerequisites..........................3 1.1.1 Rings and modules........................3 1.1.2 Homological Algebra.......................6 1.2 Logical Prerequisites........................... 13 1.2.1 Basic set theoretic notions.................... 13 1.2.2 Diamond............................. 16 1.2.3 Martin's axiom.......................... 19 2 ZFC results 23 2.1 Ext1(A; Z) for torsion groups A ..................... 24 2.2 The ranks of a group........................... 24 2.3 Classification of divisible groups..................... 28 2.4 The torsion-free rank of Ext1(A; Z).................. 29 2.5 The p-rank of Ext1(A; Z).......................
    [Show full text]
  • Abelian Group 1 Abelian Group
    Abelian group 1 Abelian group In abstract algebra, an abelian group, also called a commutative group, is a group in which the result of applying the group operation to two group elements does not depend on their order (the axiom of commutativity). Abelian groups generalize the arithmetic of addition of integers. They are named after Niels Henrik Abel.[1] The concept of an abelian group is one of the first concepts encountered in undergraduate abstract algebra, with many other basic objects, such as a module and a vector space, being its refinements. The theory of abelian groups is generally simpler than that of their non-abelian counterparts, and finite abelian groups are very well understood. On the other hand, the theory of infinite abelian groups is an area of current research. Definition An abelian group is a set, A, together with an operation "•" that combines any two elements a and b to form another element denoted a • b. The symbol "•" is a general placeholder for a concretely given operation. To qualify as an abelian group, the set and operation, (A, •), must satisfy five requirements known as the abelian group axioms: Closure For all a, b in A, the result of the operation a • b is also in A. Associativity For all a, b and c in A, the equation (a • b) • c = a • (b • c) holds. Identity element There exists an element e in A, such that for all elements a in A, the equation e • a = a • e = a holds. Inverse element For each a in A, there exists an element b in A such that a • b = b • a = e, where e is the identity element.
    [Show full text]