Axiom of Determinacy

Total Page:16

File Type:pdf, Size:1020Kb

Axiom of Determinacy Axiom of Determinacy William Gunther March 21, 2010 Abstract The Axiom of Determinacy (AD) states that every subset of the Baire Space, !!, is determined. It is simple to see this is inconsistent with the Axiom of Choice. But, a weaker version of choice called Dependent Choice (DC) is both consistent with AD and strong enough to prove basic theorems of analysis. We shall look at consequences of AD, with a focus on the connections with measurable cardinals. Definition 1. The Axiom of Determinacy (AD) is the statement Det(}(!!)) That is, it states all subsets of the Baire Space are determined. Proposition 1. The Axiom of Choice (AC) is inconsistent with AD. Proof. AC implies there exists a subset of !! which is not Lebesgue measurable. But AD says everything everything is Lebesgue measurable. Remark 1. As you can imagine, models of ZF+AD can be very strange because they can strongly violate choice. Because of this, we will use a weakening of the axiom of choice that is sufficent for some basic theorems of anaylsis called dependent choice (DC). To give you some intuition over the power of DC, under ZF, DC is equivilent to every pruned tree having a branch. Theorem 1 (ZFC). !1 is not measurable. Proof. !! PROOF PAGE 131 OF JECH !! 1 ⊆ 1 Lemma 1 (Σ 1 Boundedness Lemma). If B WO is Σ 1 then there is α < !1 such that type(x) < α for all x 2 B. e e Theorem 2 (ZF+AD+DC). The club filter on !1 is an σ-complete ultrafilter. In particular, !1 is measur- able. Proof. Let S ⊆ !1. Our goal is to show either it, or it's compliment contains a club. We first define the ordinal ordinal game GS . In GS , two players alternate turns playing elements of !1. I α0 α1 ··· αi; βi 2 !1; 8i 2 ! II β0 β1 Let γ = sup(f αi j i < ! g [ f βi j i < ! g). Player I wins if γ 2 S. Otherwise Player II wins. Claim 1. If S contains a club, then Player I has a winning strategy. If !1 n S is contains a club club, then Player II has a winning strategy. Proof. The strategy would be just to play on your club in the obvious way. ! Setting this game aside for a moment, we will play GS on !. Here two players alternate playing natural numbers. I x(0) x(1) ··· x(i); y(i) 2 !; 8i 2 ! II y(0) y(1) 1 And let x = hx(0); x(1);:::i and y = hy(0); y(1);:::i be the corresponding elements of the Baire Space. In each of these we will decode (in the natural way) elements of !×!×!2, which we will view as of !-many functions mapping ! × ! ! 2. That is M M x = xi y = yi i<! i<! xi; yi : ! × ! ! 2 We're viewing the xi's and yi's as characteristic functions of a binary relation on !. Then Player I wins if either of the following hold:- • There is some i < ! such that yi 2= WO • If γ = sup(f type(xi) j i < ! g [ f type(yi) j i < ! g) is an element of S. Player II wins otherwise. Let F = f S ⊆ !1 j Player I has a winning strategy in GS g Claim 2. F is a σ-complete non-principal ultrafilter. Proof. Non-principal is easy; if S is finite then it has a maximal element α Player II can ignore Player I, and just play whatever sequence codes x0 = α + 1 and xi 2 WO; 8i 2 !. Then he wins. If A 2 F then we get a strategy σ for GA. If A ⊆ B then we claim that σ also wins GB. ! Play according to the strategy, and get x; y 2 !. If i < ! such that yi 2= WO then we are done. Otherwise, look at γ = sup(f type(xi) j i < ! g [ f type(yi) j i < ! g). By definition of σ as a winning strategy we have γ 2 A ⊆ B. Thus Player I has a winning strategy for in GB, so B 2 F . To show it is ultra, look at S ⊆ !1. By AD, GS is determined. If Player I has a winning strategy, we're done. Otherwise Player II has a winning strategy, σ. Then, play the game G!1nS with auxillary game GS. In GS, play anything for Player I. Then, we get according to σ a play that Player I will play in G!1nS. Then we continue, plugging in Player II's plays in GS for Player I's. !! COMPLETE!! THIS IS A BIT OF A PROBLEM, AS PICKING ANY RANDOM NUMBER WILL MESS THINGS UP. BUT THE IDEA IS RIGHT; IF PLAYER II HAS A W.S. IN S THEN PLAYER I PLAYING THAT STRATEGY IN !1 n S WILL WORK !! !! HAVE TO DEAL WITH SIGMA COMPLETE !! Claim 3. If a player has a winning strategy in GS then he also has a winning strategy in ordinal GS Proof. Suppose that Player I had a winning strategy σ in GS. We would like to define a strategy ordinal τ in GS . Consider the set ! ! A0 = f x0 2 ! j 9y 2 ! (x; y) = σ ∗ y g 1 1 ⊆ 1 Note A0 is Σ1(σ); in particular it is Σ 1. Also, A0 WO. Then by Lemma 1 (Σ 1 Boundedness), there is some α0 < !1 such that fore every x 2 A0, α0 > type(x). In particular,e we can assume ordinal that this is the smallest such. So, in GS we play α0, in which II answer β0. We continue. Define [ ] ! ! Ai = f xi 2 ! j 9y 2 ! 8j < i[yj = βj] and (x; y) = σ ∗ y g From this, we get αi minimal such that !1 > αi > type(x) for all x 2 Ai. We play αi. ordinal ⊆ Claim that this strategy wins GS . Suppose not; then there is some ζ !1 that wins against τ. Let f αi j i < ! g be the ordinals players by τ against ζ. Then as Player II wins γ = sup(f ζi j i < ! g [ f αi j i < ! g) 2= S L In the auxillary game GS play according to σ against y, chosen such that y = i ζi. As σ is a winning strategy for Player I on S, we have sup(f ζi j i < ! g [ f type(xi) j i < ! g) 2 S 2 Claim 4. F is the club filter on !1. Proof. 3 Let's do another proof focused more on recursion theory. To begin with, we need to know more ramifi- cations of AD. ! Lemma 2 (ZF+DC+AD). There is no injection from !1 into 2 . ! Proof. For sake of contradiction, suppose that f : !1 ! 2 was an injection, with image A. From AD, we know that A has the perfect set property. As jAj = j!1j A is not countable, and therefore has a perfect subset. Let P ⊆ A be closed and perfect. Then it contains a perfect tree, and therefore there is a homeomorphism ! π : 2 ! P . As P ⊆ im(f) there are f αi j i < !1 g such that f f(αi) j i < !1 g = P . Then define ! g : !1 ! 2 −1 i 7! π (f(αi)) Clearly this is a bijection. Then we can define the relation R ⊆ 2! × 2! by iRj () g−1(i) < g−1(j) I claim this is not Lebesgue measurable (as a subset of 2! × 2!). To see this, just note that for any fixed y 2 2! f x 2 2! j xRy g is countable, and thus has measure 0, whereas f y 2 2! j yRx g is co-countable, and thus has measure 1. This contradicts Fubini's Theorem, thus R is not measurable. Definition 2. Recall, for A; B ⊆ !, that A is recursive in B if membership of A can be decided by using B as an oracle. We also say that A is Turing Reducible to B, and denote it A ≤T B. If A ≤T B and B ≤T A then they are Turing equivilent denoted A ≡T B. This is an equivilence relation on }(!); the equivilences classes are called Turing degrees. Theorem 3. The club filter on !1 is a σ-complete ultrafilter; in particular, !1 is measurable. Alternate Proof. Let D = f [X]T j X ⊆ ! g be the set of Turing Degrees. We defined F ⊆ }(D) by F = f S ⊆ D j 9X ⊆ ! . 8Y ⊆ ! . X ≤T Y ! [Y ]T 2 S g A better definition for F is to first to define a cone. For x ⊆ !, we define the cone of x to be cone(x) = f y ⊆ ! j x ≤T y g Then F ⊆ }(D) where F 2 F () F contains a cone. Claim 1. F is a σ-complete filter. Proof. We will first show that F is a σ-complete filter. D is clearly in F , by taking X = ;, or analogously observing that cone(;) = D (or any recursive set). F is clearly upward closed, for if A 2 F we get a corresponding XA; then if A ⊆ B then for any Y ⊆ ! such that XA ≤T Y we have [Y ]T 2 A ⊆ B. F 2 f j g ⊆ F Thus we needT only show is σ-complete. Fix β !1. Take Aα α < β . We will F ⊆ ⊆ show that A = α<β Aα is in , ie. there is some x ! such that cone(x) A. For each α < β, ! we get xα ⊆ ! such that cone(xα) ⊆ Aα.
Recommended publications
  • Games in Descriptive Set Theory, Or: It's All Fun and Games Until Someone Loses the Axiom of Choice Hugo Nobrega
    Games in Descriptive Set Theory, or: it’s all fun and games until someone loses the axiom of choice Hugo Nobrega Cool Logic 22 May 2015 Descriptive set theory and the Baire space Presentation outline [0] 1 Descriptive set theory and the Baire space Why DST, why NN? The topology of NN and its many flavors 2 Gale-Stewart games and the Axiom of Determinacy 3 Games for classes of functions The classical games The tree game Games for finite Baire classes Descriptive set theory and the Baire space Why DST, why NN? Descriptive set theory The real line R can have some pathologies (in ZFC): for example, not every set of reals is Lebesgue measurable, there may be sets of reals of cardinality strictly between |N| and |R|, etc. Descriptive set theory, the theory of definable sets of real numbers, was developed in part to try to fill in the template “No definable set of reals of complexity c can have pathology P” Descriptive set theory and the Baire space Why DST, why NN? Baire space NN For a lot of questions which interest set theorists, working with R is unnecessarily clumsy. It is often better to work with other (Cauchy-)complete topological spaces of cardinality |R| which have bases of cardinality |N| (a.k.a. Polish spaces), and this is enough (in a technically precise way). The Baire space NN is especially nice, as I hope to show you, and set theorists often (usually?) mean this when they say “real numbers”. Descriptive set theory and the Baire space The topology of NN and its many flavors The topology of NN We consider NN with the product topology of discrete N.
    [Show full text]
  • Topology and Descriptive Set Theory
    View metadata, citation and similar papers at core.ac.uk brought to you by CORE provided by Elsevier - Publisher Connector TOPOLOGY AND ITS APPLICATIONS ELSEVIER Topology and its Applications 58 (1994) 195-222 Topology and descriptive set theory Alexander S. Kechris ’ Department of Mathematics, California Institute of Technology, Pasadena, CA 91125, USA Received 28 March 1994 Abstract This paper consists essentially of the text of a series of four lectures given by the author in the Summer Conference on General Topology and Applications, Amsterdam, August 1994. Instead of attempting to give a general survey of the interrelationships between the two subjects mentioned in the title, which would be an enormous and hopeless task, we chose to illustrate them in a specific context, that of the study of Bore1 actions of Polish groups and Bore1 equivalence relations. This is a rapidly growing area of research of much current interest, which has interesting connections not only with topology and set theory (which are emphasized here), but also to ergodic theory, group representations, operator algebras and logic (particularly model theory and recursion theory). There are four parts, corresponding roughly to each one of the lectures. The first contains a brief review of some fundamental facts from descriptive set theory. In the second we discuss Polish groups, and in the third the basic theory of their Bore1 actions. The last part concentrates on Bore1 equivalence relations. The exposition is essentially self-contained, but proofs, when included at all, are often given in the barest outline. Keywords: Polish spaces; Bore1 sets; Analytic sets; Polish groups; Bore1 actions; Bore1 equivalence relations 1.
    [Show full text]
  • Are Large Cardinal Axioms Restrictive?
    Are Large Cardinal Axioms Restrictive? Neil Barton∗ 24 June 2020y Abstract The independence phenomenon in set theory, while perva- sive, can be partially addressed through the use of large cardinal axioms. A commonly assumed idea is that large cardinal axioms are species of maximality principles. In this paper, I argue that whether or not large cardinal axioms count as maximality prin- ciples depends on prior commitments concerning the richness of the subset forming operation. In particular I argue that there is a conception of maximality through absoluteness, on which large cardinal axioms are restrictive. I argue, however, that large cardi- nals are still important axioms of set theory and can play many of their usual foundational roles. Introduction Large cardinal axioms are widely viewed as some of the best candi- dates for new axioms of set theory. They are (apparently) linearly ordered by consistency strength, have substantial mathematical con- sequences for questions independent from ZFC (such as consistency statements and Projective Determinacy1), and appear natural to the ∗Fachbereich Philosophie, University of Konstanz. E-mail: neil.barton@uni- konstanz.de. yI would like to thank David Aspero,´ David Fernandez-Bret´ on,´ Monroe Eskew, Sy-David Friedman, Victoria Gitman, Luca Incurvati, Michael Potter, Chris Scam- bler, Giorgio Venturi, Matteo Viale, Kameryn Williams and audiences in Cambridge, New York, Konstanz, and Sao˜ Paulo for helpful discussion. Two anonymous ref- erees also provided helpful comments, and I am grateful for their input. I am also very grateful for the generous support of the FWF (Austrian Science Fund) through Project P 28420 (The Hyperuniverse Programme) and the VolkswagenStiftung through the project Forcing: Conceptual Change in the Foundations of Mathematics.
    [Show full text]
  • Polish Spaces and Baire Spaces
    Polish spaces and Baire spaces Jordan Bell [email protected] Department of Mathematics, University of Toronto June 27, 2014 1 Introduction These notes consist of me working through those parts of the first chapter of Alexander S. Kechris, Classical Descriptive Set Theory, that I think are impor- tant in analysis. Denote by N the set of positive integers. I do not talk about universal spaces like the Cantor space 2N, the Baire space NN, and the Hilbert cube [0; 1]N, or \localization", or about Polish groups. If (X; τ) is a topological space, the Borel σ-algebra of X, denoted by BX , is the smallest σ-algebra of subsets of X that contains τ. BX contains τ, and is closed under complements and countable unions, and rather than talking merely about Borel sets (elements of the Borel σ-algebra), we can be more specific by talking about open sets, closed sets, and sets that are obtained by taking countable unions and complements. Definition 1. An Fσ set is a countable union of closed sets. A Gδ set is a complement of an Fσ set. Equivalently, it is a countable intersection of open sets. If (X; d) is a metric space, the topology induced by the metric d is the topology generated by the collection of open balls. If (X; τ) is a topological space, a metric d on the set X is said to be compatible with τ if τ is the topology induced by d.A metrizable space is a topological space whose topology is induced by some metric, and a completely metrizable space is a topological space whose topology is induced by some complete metric.
    [Show full text]
  • TOPOLOGY PRELIM REVIEW 2021: LIST THREE Topic 1: Baire Property and Gδ Sets. Definition. X Is a Baire Space If a Countable Inte
    TOPOLOGY PRELIM REVIEW 2021: LIST THREE Topic 1: Baire property and Gδ sets. Definition. X is a Baire space if a countable intersection of open, dense subsets of X is dense in X. Complete metric spaces and locally compact spaces are Baire spaces. Def. X is locally compact if for all x 2 X, and all open Ux, there exists Vx with compact closure V x ⊂ Ux. 1. X is locally compact , for all C ⊂ X compact, and all open U ⊃ C, there exists V open with compact closure, so that: C ⊂ V ⊂ V ⊂ U. 2. Def: A set E ⊂ X is nowhere dense in X if its closure E has empty interior. Show: X is a Baire space , any countable union of nowhere dense sets has empty interior. 3. A complete metric space without isolated points is uncountable. (Hint: Baire property, complements of one-point sets.) 4. Uniform boundedness principle. X complete metric, F ⊂ C(X) a family of continuous functions, bounded at each point: (8a 2 X)(9M(a) > 0)(8f 2 F)jf(a)j ≤ M(a): Then there exists a nonempty open set U ⊂ X so that F is eq¨uibounded over U{there exists a constant C > 0 so that: (8f 2 F)(8x 2 U)jf(x)j ≤ C: Hint: For n ≥ 1, consider An = fx 2 X; jf(x)j ≤ n; 8f 2 Fg. Use Baire's property. An important application of 4. is the uniform boundedness principle for families of bounded linear operators F ⊂ L(E; F ), where E; F are Banach spaces.
    [Show full text]
  • On a Notion of Smallness for Subsets of the Baire Space Author(S): Alexander S
    On a Notion of Smallness for Subsets of the Baire Space Author(s): Alexander S. Kechris Source: Transactions of the American Mathematical Society, Vol. 229 (May, 1977), pp. 191-207 Published by: American Mathematical Society Stable URL: http://www.jstor.org/stable/1998505 . Accessed: 22/05/2013 14:14 Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at . http://www.jstor.org/page/info/about/policies/terms.jsp . JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected]. American Mathematical Society is collaborating with JSTOR to digitize, preserve and extend access to Transactions of the American Mathematical Society. http://www.jstor.org This content downloaded from 131.215.71.79 on Wed, 22 May 2013 14:14:52 PM All use subject to JSTOR Terms and Conditions TRANSACTIONS OF THE AMERICAN MATHEMATICAL SOCIETY Volume 229, 1977 ON A NOTION OF SMALLNESS FOR SUBSETS OF THE BAIRE SPACE BY ALEXANDER S. KECHRIS ABSTRACT.Let us call a set A C o' of functionsfrom X into X a-bounded if there is a countablesequence of functions(a.: n E o} C w' such that every memberof A is pointwisedominated by an elementof that sequence. We study in this paper definabilityquestions concerningthis notion of smallnessfor subsets of o'. We show that most of the usual definability results about the structureof countablesubsets of o' have corresponding versionswhich hold about a-boundedsubsets of o'.
    [Show full text]
  • BAIRE SPACES and VIETORIS HYPERSPACES 1. Introduction in This Paper, All Topological Spaces Are Assumed to Be Infinite and at Le
    PROCEEDINGS OF THE AMERICAN MATHEMATICAL SOCIETY Volume 135, Number 1, January 2007, Pages 299–303 S 0002-9939(06)08743-0 Article electronically published on August 16, 2006 BAIRE SPACES AND VIETORIS HYPERSPACES JILING CAO, SALVADOR GARC´IA-FERREIRA, AND VALENTIN GUTEV (Communicated by Jonathan M. Borwein) Abstract. We prove that if the Vietoris hyperspace CL(X) of all nonempty closed subsets of a space X is Baire, then all finite powers of X must be Baire spaces. In particular, there exists a metrizable Baire space X whose Vietoris hyperspace CL(X) is not Baire. This settles an open problem of R. A. McCoy stated in 1975. 1. Introduction In this paper, all topological spaces are assumed to be infinite and at least Haus- dorff. Also, all product spaces are endowed with the Tychonoff product topology. A topological space X is called Baire [HM] if the intersection of any sequence of dense open subsets of X is dense in X, or equivalently, if all nonempty open subsets of X are of second category. Baire spaces have numerous applications in analysis and topology, such as the open mapping and closed graph theorems, and the Banach- Steinhaus theorem [Con]. For some other applications, see [E, Za1] or [Za2]. A fundamental treatise on Baire spaces in general topology is [HM], and several open problems on Baire and related spaces are discussed in [AL]. For a space X,letCL(X) be the collection of all nonempty closed subsets of X endowed with the Vietoris topology τV . Recall that a canonical base for τV is given by all subsets of CL(X)oftheform V = F ∈ CL(X):F ⊂ V,F∩ V = ∅ for any V ∈V , where V runs over the finite families of nonempty open subsets of X.Inthesequel, any subset D⊂CL(X) will carry the relative Vietoris topology τV as a subspace of (CL(X),τV ).
    [Show full text]
  • Regularity Properties and Determinacy
    Regularity Properties and Determinacy MSc Thesis (Afstudeerscriptie) written by Yurii Khomskii (born September 5, 1980 in Moscow, Russia) under the supervision of Dr. Benedikt L¨owe, and submitted to the Board of Examiners in partial fulfillment of the requirements for the degree of MSc in Logic at the Universiteit van Amsterdam. Date of the public defense: Members of the Thesis Committee: August 14, 2007 Dr. Benedikt L¨owe Prof. Dr. Jouko V¨a¨an¨anen Prof. Dr. Joel David Hamkins Prof. Dr. Peter van Emde Boas Brian Semmes i Contents 0. Introduction............................ 1 1. Preliminaries ........................... 4 1.1 Notation. ........................... 4 1.2 The Real Numbers. ...................... 5 1.3 Trees. ............................. 6 1.4 The Forcing Notions. ..................... 7 2. ClasswiseConsequencesofDeterminacy . 11 2.1 Regularity Properties. .................... 11 2.2 Infinite Games. ........................ 14 2.3 Classwise Implications. .................... 16 3. The Marczewski-Burstin Algebra and the Baire Property . 20 3.1 MB and BP. ......................... 20 3.2 Fusion Sequences. ...................... 23 3.3 Counter-examples. ...................... 26 4. DeterminacyandtheBaireProperty.. 29 4.1 Generalized MB-algebras. .................. 29 4.2 Determinacy and BP(P). ................... 31 4.3 Determinacy and wBP(P). .................. 34 5. Determinacy andAsymmetric Properties. 39 5.1 The Asymmetric Properties. ................. 39 5.2 The General Definition of Asym(P). ............. 43 5.3 Determinacy and Asym(P). ................. 46 ii iii 0. Introduction One of the most intriguing developments of modern set theory is the investi- gation of two-player infinite games of perfect information. Of course, it is clear that applied game theory, as any other branch of mathematics, can be modeled in set theory. But we are talking about the converse: the use of infinite games as a tool to study fundamental set theoretic questions.
    [Show full text]
  • Determinacy and Large Cardinals
    Determinacy and Large Cardinals Itay Neeman∗ Abstract. The principle of determinacy has been crucial to the study of definable sets of real numbers. This paper surveys some of the uses of determinacy, concentrating specifically on the connection between determinacy and large cardinals, and takes this connection further, to the level of games of length ω1. Mathematics Subject Classification (2000). 03E55; 03E60; 03E45; 03E15. Keywords. Determinacy, iteration trees, large cardinals, long games, Woodin cardinals. 1. Determinacy Let ωω denote the set of infinite sequences of natural numbers. For A ⊂ ωω let Gω(A) denote the length ω game with payoff A. The format of Gω(A) is displayed in Diagram 1. Two players, denoted I and II, alternate playing natural numbers forming together a sequence x = hx(n) | n < ωi in ωω called a run of the game. The run is won by player I if x ∈ A, and otherwise the run is won by player II. I x(0) x(2) ...... II x(1) x(3) ...... Diagram 1. The game Gω(A). A game is determined if one of the players has a winning strategy. The set A is ω determined if Gω(A) is determined. For Γ ⊂ P(ω ), det(Γ) denotes the statement that all sets in Γ are determined. Using the axiom of choice, or more specifically using a wellordering of the reals, it is easy to construct a non-determined set A. det(P(ωω)) is therefore false. On the other hand it has become clear through research over the years that det(Γ) is true if all the sets in Γ are definable by some concrete means.
    [Show full text]
  • Dropping Polishness
    Dropping Polishness Andrea Medini Kurt Gödel Research Center University of Vienna March 6, 2015 Andrea Medini Dropping Polishness How do we “generalize” descriptive set theory? Recall that a space is Polish if it is separable and completely metrizable. I can think of three ways... 1 Dropping separability: the prototypical space is κω, where κ has the discrete topology. It is completely metrizable, but not necessarily separable. (Ask Sergey Medvedev...) 2 Dropping everything: the prototypical space is κκ with the <κ-box-topology, where κ has the discrete topology and it satisfies κ<κ = κ. (Ask Sy Friedman...) 3 Dropping Polishness: consider questions of “descriptive set-theoretic flavor” in spaces that are separable and metrizable, but not necessarily completely metrizable. (Ask Arnie Miller...) From now on, we will assume that every space is separable and metrizable, but not necessarily Polish. Andrea Medini Dropping Polishness How do you define complexity then? Γ will always be one of the following (boldface) pointclasses. 0 0 Σξ or Πξ , where ξ is an ordinal such that 1 ≤ ξ < ω1 (these are the Borel pointclasses). 1 1 Σn or Πn, where n is an ordinal such that 1 ≤ n < ω (these are the projective pointclasses). We will assume that the definition of a Γ subset of a Polish space is well-known, and recall that it can be generalized to arbitrary spaces as follows. Definition Fix a pointclass Γ. Let X be a space. We will say that A ⊆ X is a Γ subset of X if there exists a Polish space T containing X as a subspace such that A = B ∩ X for some Γ subset B of T .
    [Show full text]
  • Effective Descriptive Set Theory
    Effective Descriptive Set Theory Andrew Marks December 14, 2019 1 1 These notes introduce the effective (lightface) Borel, Σ1 and Π1 sets. This study uses ideas and tools from descriptive set theory and computability theory. Our central motivation is in applications of the effective theory to theorems of classical (boldface) descriptive set theory, especially techniques which have no classical analogues. These notes have many errors and are very incomplete. Some important topics not covered include: • The Harrington-Shore-Slaman theorem [HSS] which implies many of the theorems of Section 3. • Steel forcing (see [BD, N, Mo, St78]) • Nonstandard model arguments • Barwise compactness, Jensen's model existence theorem • α-recursion theory • Recent beautiful work of the \French School": Debs, Saint-Raymond, Lecompte, Louveau, etc. These notes are from a class I taught in spring 2019. Thanks to Adam Day, Thomas Gilton, Kirill Gura, Alexander Kastner, Alexander Kechris, Derek Levinson, Antonio Montalb´an,Dean Menezes and Riley Thornton, for helpful conversations and comments on earlier versions of these notes. 1 Contents 1 1 1 1 Characterizing Σ1, ∆1, and Π1 sets 4 1 1.1 Σn formulas, closure properties, and universal sets . .4 1.2 Boldface vs lightface sets and relativization . .5 1 1.3 Normal forms for Σ1 formulas . .5 1.4 Ranking trees and Spector boundedness . .7 1 1.5 ∆1 = effectively Borel . .9 1.6 Computable ordinals, hyperarithmetic sets . 11 1 1.7 ∆1 = hyperarithmetic . 14 x 1 1.8 The hyperjump, !1 , and the analogy between c.e. and Π1 .... 15 2 Basic tools 18 2.1 Existence proofs via completeness results .
    [Show full text]
  • NONMEASURABLE ALGEBRAIC SUMS of SETS of REALS Sierpinski Showed in [14] That There Exist Two Sets X, Y ⊆ R of Lebesgue Measure
    NONMEASURABLE ALGEBRAIC SUMS OF SETS OF REALS MARCIN KYSIAK Abstract. We present a theorem which generalizes some known theorems on the existence of nonmeasurable (in various senses) sets of the form X+Y . Some additional related questions concerning measure, category and the algebra of Borel sets are also studied. Sierpi´nskishowed in [14] that there exist two sets X; Y ⊆ R of Lebesgue measure zero such that their algebraic sum, i.e. the set X + Y = fx + y : x 2 X; y 2 Y g is nonmeasurable. The analogous result is also true for the Baire property. Sierpi´nski’sconstruction has been generalized to other σ-algebras and σ-ideals of subsets of R. Kharazishvili proves in [10] that for every σ-ideal I which is not closed under algebraic sums and every σ-algebra A such that the quotient algebra A=I satisfies the countable chain condition, there exist sets X; Y 2 I such that X + Y 62 A. A similar result was proved by Cicho´nand Jasi´nskiin [3] for every σ-ideal I with coanalytic base and the algebra Bor[I] (i.e. the smallest algebra containing I and Bor). Ciesielski, Fejzi´cand Freiling prove in [4] a stronger version of Sierpi´nski’sthe- orem. They show that for every set C ⊆ R such that C + C has positive outer measure there exists X ⊆ C such that X + X is not Lebesgue measurable. In particular, starting with such a set C of measure zero (the “middle third” Cantor set in [0; 1] for example), we obtain Sierpi´nski’sexample as a corollary.
    [Show full text]