First-Order Logical Duality

Total Page:16

File Type:pdf, Size:1020Kb

First-Order Logical Duality First-Order Logical Duality Henrik Forssell A dissertation submitted in partial ful¯llment of the requirements for the degree of PhD in Logic, Computation and Methodology Carnegie Mellon University 2008 Abstract Generalizing Stone duality for Boolean algebras, an adjunction between Bool- ean coherent categories|representing ¯rst-order syntax|and certain topo- logical groupoids|representing semantics|is constructed. The embedding of a Boolean algebra into a frame of open sets of a space of 2-valued mod- els is replaced by an embedding of a Boolean coherent category, B, into a topos of equivariant sheaves on a topological groupoid of set-valued models and isomorphisms between them. The latter is a groupoid representation of the topos of coherent sheaves on B, analogously to how the Stone space of a Boolean algebra is a spatial representation of the ideal completion of the algebra, and the category B can then be recovered from its semantical groupoid, up to pretopos completion. By equipping the groupoid of sets and bijections with a particular topology, one obtains a particular topological groupoid which plays a role analogous to that of the discrete space 2, in be- ing the dual of the object classi¯er and the object one `homs into' to recover a Boolean coherent category from its semantical groupoid. Both parts of the adjunction, then, consist of `homming into sets', similarly to how both parts of the equivalence between Boolean algebras and Stone spaces consist of `homming into 2'. By slicing over this groupoid (modi¯ed to display an alternative setup), Chapter 3 shows how the adjunction specializes to the case of ¯rst-order single sorted logic to yield an adjunction between such theories and an inde- pendently characterized slice category of topological groupoids such that the counit component at a theory is an isomorphism. Acknowledgements I would like, ¯rst and foremost, to thank my supervisor Steve Awodey. I would like to thank the members of the committee: Jeremy Avigad, Lars Birkedal, James Cummings, and Dana Scott. I would like to thank the faculty and sta® of the department of philosophy at Carnegie Mellon University. I would like to thank Michael Warren, Ko- hei Kishida, and my fellow students at the departments of philosophy and mathematics at Carnegie Mellon University. I would like to thank Elise for her patience and support. ii Contents 1 Introduction 1 1.1 Algebra, Geometry, and Logic . 1 1.2 Logical Dualities . 4 1.2.1 Propositional Logic and Stone Duality . 4 1.2.2 Lawvere's Duality for Equational Logic . 11 1.2.3 Makkai's First-Order Logical Duality . 15 1.3 A Sheaf-Theoretical Approach . 17 1.4 Representing Topoi by Groupoids . 23 2 The Syntax-Semantics Adjunction 25 2.1 Overview . 25 2.2 The Semantical Groupoid of a Decidable Coherent Category . 30 2.2.1 Spaces of Models and Isomorphisms . 31 2.2.2 Stricti¯cation: Equipping Coherent Categories with Canonical Coherent Structure . 33 2.3 Representing Decidable Coherent Topoi by Groupoids of Models 38 2.3.1 Cover-Reflecting, Set-Valued Functors . 39 2.3.2 The Semantical Groupoid of a Theory T . 41 2.3.3 Topological Groupoids and Equivariant Sheaves . 46 2.3.4 Equivariant Sheaves on the Semantical Groupoid . 50 2.4 Syntax-Semantics Adjunction . 58 2.4.1 The Representation Theorem . 58 2.4.2 The Semantical Functor . 61 2.4.3 The Syntactical Functor . 62 2.4.4 The Syntax-Semantics Adjunction . 74 2.4.5 Restricting the Adjunction . 79 2.5 The Boolean Object Classi¯er . 87 iii 3 First-Order Logical Duality 94 3.1 Sheaves on the Space of Models . 95 3.1.1 The Space of Models . 95 3.1.2 An Open Surjection . 101 3.2 Equivariant Sheaves on the Groupoid of Models . 110 3.2.1 The Groupoid of Models . 110 3.2.2 Equivariant Sheaves on the Groupoid of Models . 113 3.3 Syntax-Semantics Adjunction . 119 3.3.1 The Category of FOL . 119 3.3.2 The Object Classi¯er . 121 3.3.3 The Semantical Groupoid Functor . 125 3.3.4 The Theory Functor . 128 3.4 Stone Fibrations . 135 3.4.1 Sites for Groupoids . 135 3.4.2 Stone Fibrations over N» . 141 3.5 Groupoids and Theories . 143 3.5.1 Applications and Future Work . 143 3.5.2 Groupoids and Theory Extensions . 145 A An Argument by Descent 153 A.1 An Argument by Descent . 153 A.1.1 The Theory of Countable T-Models . 154 A.1.2 The Theory of an Isomorphism between Two Count- able T-models . 158 A.1.3 Pullback of the Open Surjection against Itself . 162 A.1.4 The Theory of Two Composable Isomorphisms and the Triple Pullback of the Open Surjection against Itself . 166 A.1.5 The Category of G-Equivariant Sheaves on X . 169 B Table of Categories 172 B.1 A Table of Categories . 172 iv Chapter 1 Introduction 1.1 Algebra, Geometry, and Logic In this thesis, I present an extension of Stone Duality for Boolean Algebras from classical propositional logic to classical ¯rst-order logic. The leading idea is, in broad strokes, to take the traditional logical distinction between syntax and semantics and analyze it in terms of the classical mathematical distinction between algebra and geometry, with syntax corresponding to al- gebra and semantics to geometry. Insights from category theory allows us glean a certain duality between the two notions of algebra and geometry. We see a ¯rst glimpse of this in Stone's duality theorem for Boolean algebras, the categorical formulation of which states that a category of `algebraic' ob- jects (Boolean algebras) is the categorical dual of a category of `geometrical' objects (Stone spaces). \Categorically dual" means that the one category is opposite to the other, in that it can be obtained from the other by for- mally reversing the morphisms. In a more far reaching manner, this form of algebra-geometry duality is exhibited in modern algebraic geometry as re- formulated in the language of schemes in the Grothendieck school. E.g. in the duality between the categories of commutative rings and the category of a±ne schemes. On the other hand, we are informed by the area of category theory known as categorical logic that algebra is closely connected with logic, in the sense that logical theories can be seen as categories and suitable categories can be seen as logical theories. For instance, Boolean algebras correspond to classical propositional theories, equational theories correspond to categories 1 with ¯nite products, Boolean coherent categories correspond to classical ¯rst- order logic, and topoi|e.g. of sheaves on a space|correspond to higher-order intuitionistic logic. Thus the study of these algebraic objects has logical interpretation and, vice versa, reasoning in or about logical theories has application in their corresponding algebraic objects. With the connection between algebra and logic in hand, instances of the algebra-geometry duality can be seen to manifest a syntax-semantics duality between an algebra of syntax and a geometry of semantics. Stone duality, in its logical interpretation, manifests a syntax-semantics duality for propositional logic. As the category of Boolean algebras can be considered as the category of propositional theories modulo `algebraic' equivalence, the category of Stone spaces can be seen as the category of spaces of corresponding two-valued models. We obtain the set of models corresponding to a Boolean algebra by taking morphisms in the category of Boolean algebras from the given algebra into the two-element Boolean algebra, 2, » ModB = HomBA (B; 2) (1.1) And with suitable topologies in place, we can retrieve the Boolean algebra by taking morphisms in the category of Stone spaces from that space into the two-element Stone space, 2, » B = HomStone (ModB; 2) Here, the two-element set, 2, is in a sense living a `dual' life, and `hom- ming into 2' forms an adjunction between (the opposite of) the `syntactical' category of Boolean algebras and the category of topological spaces, which becomes an equivalence once we restrict to the `semantical' subcategory of Stone spaces. For equational (or algebraic) theories, i.e. those formulated in languages without relation symbols and with all axioms equations, an example of syntax-semantics duality occurred already in F.W. Lawvere's thesis ([16]). Such a theory A can be considered, up to `algebraic' or categorical equiva- lence, as a particular category CA with ¯nite products, and the category of set-valued models of the theory A is then the category of ¯nite product pre- / serving functors CA Sets from CA into the category of sets and functions, ModA ' HomFP (CA; Sets) (1.2) 2 Then CA can be recovered from the category ModA of models as the category / HomG(ModA; Sets) of those functors F : ModA Sets that preserve all limits, ¯ltered colimits, and regular epimorphisms, CA ' HomG(ModA; Sets) We think of those as the `continuous' maps in this context. In a wider per- spective, `homming into the dual object Sets` creates an adjunction between (the opposite of) the category of categories with ¯nite products|in which the theories live|and category of cocomplete categories|in which the cor- responding model categories live. An adjunction which can be restricted to an equivalence between the `syntactical' and `semantical' subcategories of theories and models. Full ¯rst-order theories, too, form a category when considered up to algebraic equivalence, namely the category of Boolean coherent categories and coherent functors between them. This category contains the category of sets and functions, Sets, homming into which produces the usual set- valued models that often are of particular interest for ¯rst-order theories.
Recommended publications
  • Topological Duality and Lattice Expansions Part I: a Topological Construction of Canonical Extensions
    TOPOLOGICAL DUALITY AND LATTICE EXPANSIONS PART I: A TOPOLOGICAL CONSTRUCTION OF CANONICAL EXTENSIONS M. ANDREW MOSHIER AND PETER JIPSEN 1. INTRODUCTION The two main objectives of this paper are (a) to prove topological duality theorems for semilattices and bounded lattices, and (b) to show that the topological duality from (a) provides a construction of canonical extensions of bounded lattices. The paper is first of two parts. The main objective of the sequel is to establish a characterization of lattice expansions, i.e., lattices with additional operations, in the topological setting built in this paper. Regarding objective (a), consider the following simple question: Is there a subcategory of Top that is dually equivalent to Lat? Here, Top is the category of topological spaces and continuous maps and Lat is the category of bounded lattices and lattice homomorphisms. To date, the question has been answered positively either by specializing Lat or by generalizing Top. The earliest examples are of the former sort. Tarski [Tar29] (treated in English, e.g., in [BD74]) showed that every complete atomic Boolean lattice is represented by a powerset. Taking some historical license, we can say this result shows that the category of complete atomic Boolean lattices with complete lat- tice homomorphisms is dually equivalent to the category of discrete topological spaces. Birkhoff [Bir37] showed that every finite distributive lattice is represented by the lower sets of a finite partial order. Again, we can now say that this shows that the category of finite distributive lattices is dually equivalent to the category of finite T0 spaces and con- tinuous maps.
    [Show full text]
  • Dynamic Algebras: Examples, Constructions, Applications
    Dynamic Algebras: Examples, Constructions, Applications Vaughan Pratt∗ Dept. of Computer Science, Stanford, CA 94305 Abstract Dynamic algebras combine the classes of Boolean (B ∨ 0 0) and regu- lar (R ∪ ; ∗) algebras into a single finitely axiomatized variety (BR 3) resembling an R-module with “scalar” multiplication 3. The basic result is that ∗ is reflexive transitive closure, contrary to the intuition that this concept should require quantifiers for its definition. Using this result we give several examples of dynamic algebras arising naturally in connection with additive functions, binary relations, state trajectories, languages, and flowcharts. The main result is that free dynamic algebras are residu- ally finite (i.e. factor as a subdirect product of finite dynamic algebras), important because finite separable dynamic algebras are isomorphic to Kripke structures. Applications include a new completeness proof for the Segerberg axiomatization of propositional dynamic logic, and yet another notion of regular algebra. Key words: Dynamic algebra, logic, program verification, regular algebra. This paper originally appeared as Technical Memo #138, Laboratory for Computer Science, MIT, July 1979, and was subsequently revised, shortened, retitled, and published as [Pra80b]. After more than a decade of armtwisting I. N´emeti and H. Andr´eka finally persuaded the author to publish TM#138 itself on the ground that it contained a substantial body of interesting material that for the sake of brevity had been deleted from the STOC-80 version. The principal changes here to TM#138 are the addition of footnotes and the sections at the end on retrospective citations and reflections. ∗This research was supported by the National Science Foundation under NSF grant no.
    [Show full text]
  • Distributive Envelopes and Topological Duality for Lattices Via Canonical Extensions
    DISTRIBUTIVE ENVELOPES AND TOPOLOGICAL DUALITY FOR LATTICES VIA CANONICAL EXTENSIONS MAI GEHRKE AND SAM VAN GOOL Abstract. We establish a topological duality for bounded lattices. The two main features of our duality are that it generalizes Stone duality for bounded distributive lattices, and that the morphisms on either side are not the stan- dard ones. A positive consequence of the choice of morphisms is that those on the topological side are functional. In the course of the paper, we obtain the following results: (1) canoni- cal extensions of bounded lattices are the algebraic versions of the existing dualities for bounded lattices by Urquhart and Hartung, (2) there is a univer- sal construction which associates to an arbitrary lattice two distributive lattice envelopes with an adjoint pair between them, and (3) we give a topological du- ality for bounded lattices which generalizes Priestley duality and which shows precisely which maps between bounded lattices admit functional duals. For the result in (1), we rely on previous work of Gehrke, J´onssonand Harding. For the universal construction in (2), we modify a construction of the injective hull of a semilattice by Bruns and Lakser, adjusting their concept of `admis- sibility' to the finitary case. For (3), we use Priestley duality for distributive lattices and our universal characterization of the distributive envelopes. 1. Introduction Topological duality for Boolean algebras [22] and distributive lattices [23] is a useful tool for studying relational semantics for propositional logics. Canonical extensions [16, 17, 11, 10] provide a way of looking at these semantics algebraically. In the absence of a satisfactory topological duality, canonical extensions have been used [3] to treat relational semantics for substructural logics.
    [Show full text]
  • On Families of Mutually Exclusive Sets
    ANNALS OF MATHEMATICS Vol. 44, No . 2, April, 1943 ON FAMILIES OF MUTUALLY EXCLUSIVE SETS BY P . ERDÖS AND A. TARSKI (Received August 11, 1942) In this paper we shall be concerned with a certain particular problem from the general theory of sets, namely with the problem of the existence of families of mutually exclusive sets with a maximal power . It will turn out-in a rather unexpected way that the solution of these problems essentially involves the notion of the so-called "inaccessible numbers ." In this connection we shall make some general remarks regarding inaccessible numbers in the last section of our paper . §1. FORMULATION OF THE PROBLEM . TERMINOLOGY' The problem in which we are interested can be stated as follows : Is it true that every field F of sets contains a family of mutually exclusive sets with a maximal power, i .e . a family O whose cardinal number is not smaller than the cardinal number of any other family of mutually exclusive sets contained in F . By a field of sets we understand here as usual a family F of sets which to- gether with every two sets X and Y contains also their union X U Y and their difference X - Y (i.e. the set of those elements of X which do not belong to Y) among its elements . A family O is called a family of mutually exclusive sets if no set X of X of O is empty and if any two different sets of O have an empty inter- section. A similar problem can be formulated for other families e .g .
    [Show full text]
  • (Aka “Geometric”) Iff It Is Axiomatised by “Coherent Implications”
    GEOMETRISATION OF FIRST-ORDER LOGIC ROY DYCKHOFF AND SARA NEGRI Abstract. That every first-order theory has a coherent conservative extension is regarded by some as obvious, even trivial, and by others as not at all obvious, but instead remarkable and valuable; the result is in any case neither sufficiently well-known nor easily found in the literature. Various approaches to the result are presented and discussed in detail, including one inspired by a problem in the proof theory of intermediate logics that led us to the proof of the present paper. It can be seen as a modification of Skolem's argument from 1920 for his \Normal Form" theorem. \Geometric" being the infinitary version of \coherent", it is further shown that every infinitary first-order theory, suitably restricted, has a geometric conservative extension, hence the title. The results are applied to simplify methods used in reasoning in and about modal and intermediate logics. We include also a new algorithm to generate special coherent implications from an axiom, designed to preserve the structure of formulae with relatively little use of normal forms. x1. Introduction. A theory is \coherent" (aka \geometric") iff it is axiomatised by \coherent implications", i.e. formulae of a certain simple syntactic form (given in Defini- tion 2.4). That every first-order theory has a coherent conservative extension is regarded by some as obvious (and trivial) and by others (including ourselves) as non- obvious, remarkable and valuable; it is neither well-enough known nor easily found in the literature. We came upon the result (and our first proof) while clarifying an argument from our paper [17].
    [Show full text]
  • Bitopological Duality for Distributive Lattices and Heyting Algebras
    BITOPOLOGICAL DUALITY FOR DISTRIBUTIVE LATTICES AND HEYTING ALGEBRAS GURAM BEZHANISHVILI, NICK BEZHANISHVILI, DAVID GABELAIA, ALEXANDER KURZ Abstract. We introduce pairwise Stone spaces as a natural bitopological generalization of Stone spaces—the duals of Boolean algebras—and show that they are exactly the bitopolog- ical duals of bounded distributive lattices. The category PStone of pairwise Stone spaces is isomorphic to the category Spec of spectral spaces and to the category Pries of Priestley spaces. In fact, the isomorphism of Spec and Pries is most naturally seen through PStone by first establishing that Pries is isomorphic to PStone, and then showing that PStone is isomorphic to Spec. We provide the bitopological and spectral descriptions of many alge- braic concepts important for the study of distributive lattices. We also give new bitopological and spectral dualities for Heyting algebras, co-Heyting algebras, and bi-Heyting algebras, thus providing two new alternatives of Esakia’s duality. 1. Introduction It is widely considered that the beginning of duality theory was Stone’s groundbreaking work in the mid 30ies on the dual equivalence of the category Bool of Boolean algebras and Boolean algebra homomorphism and the category Stone of compact Hausdorff zero- dimensional spaces, which became known as Stone spaces, and continuous functions. In 1937 Stone [28] extended this to the dual equivalence of the category DLat of bounded distributive lattices and bounded lattice homomorphisms and the category Spec of what later became known as spectral spaces and spectral maps. Spectral spaces provide a generalization of Stone 1 spaces. Unlike Stone spaces, spectral spaces are not Hausdorff (not even T1) , and as a result, are more difficult to work with.
    [Show full text]
  • Part II Topological Dualities
    Part II Top ological dualities Chapter Top ology and armative predicates In the rst part of this monograph we considered predicates to be subsets of an abstract set of states If we think of the states as the denotations of results of computations of programs then predicates b ecome computationally mean ingful in the sense that we can use partial information ab out a computation to tell whether or not a predicate holds for that computation A predicate for which only nite information ab out a computation is needed to arm whether it holds is called an armative predicate The set of armative predicates is closed under nite intersections and ar bitrary unions Hence armative predicates can be identied with the op en sets of a top ological space The idea that op en sets are observable predi cates was prop osed by Smyth in although it is also brie y mentioned in Smyth interprets op en sets as semidecidable prop erties in some eectively given top ological space More generally op en sets can be inter preted as nitely observable predicates Alp ern and Schneider and Kwiatkowska use op en sets as nite liveness predicates and closed sets as safety predicates to formalize the informal characterization of liveness and safety prop erties of Lamp ort The name armative predicates has b een intro duced by Vickers for denoting the abstract op en sets of a frame Armative predicates are also called veriable predicates by Rewitzky who uses the term observable for predicates which are b oth armative and refutative Bonsangue In this chapter we intro
    [Show full text]
  • Spatio-Temporal Domains: an Overview David Janin
    Spatio-temporal domains: an overview David Janin To cite this version: David Janin. Spatio-temporal domains: an overview. 2018. hal-01634897v2 HAL Id: hal-01634897 https://hal.archives-ouvertes.fr/hal-01634897v2 Preprint submitted on 17 Jul 2018 HAL is a multi-disciplinary open access L’archive ouverte pluridisciplinaire HAL, est archive for the deposit and dissemination of sci- destinée au dépôt et à la diffusion de documents entific research documents, whether they are pub- scientifiques de niveau recherche, publiés ou non, lished or not. The documents may come from émanant des établissements d’enseignement et de teaching and research institutions in France or recherche français ou étrangers, des laboratoires abroad, or from public or private research centers. publics ou privés. Spatio-temporal domains: an overview David Janin? UMR LaBRI, Bordeaux INP Université de Bordeaux [email protected] Abstract. We consider the possibility of defining a general mathemat- ical framework for the homogeneous modeling and analysis of hetero- geneous spatio-temporal computations as they occur more and more in modern computerized systems of systems. It appears that certain fibra- tions of posets into posets, called here spatio-temporal domains, eventu- ally provide a fully featured category that extends to space and time the category of cpos and continuous functions, aka Scott Domains, used in classical denotational semantics. 1 Introduction Research context. Program semantics is classically divided between two com- plementary approaches : denotational semantics and operational semantics. De- notational semantics generally refers to what the partial functions encoded by programs are : what is the relationship between (models of) their input val- ues (or input memory state) and their output values (or output memory state).
    [Show full text]
  • Geometry and Categoricity∗
    Geometry and Categoricity∗ John T. Baldwiny University of Illinois at Chicago July 5, 2010 1 Introduction The introduction of strongly minimal sets [BL71, Mar66] began the idea of the analysis of models of categorical first order theories in terms of combinatorial geometries. This analysis was made much more precise in Zilber’s early work (collected in [Zil91]). Shelah introduced the idea of studying certain classes of stable theories by a notion of independence, which generalizes a combinatorial geometry, and characterizes models as being prime over certain independent trees of elements. Zilber’s work on finite axiomatizability of totally categorical first order theories led to the development of geometric stability theory. We discuss some of the many applications of stability theory to algebraic geometry (focusing on the role of infinitary logic). And we conclude by noting the connections with non-commutative geometry. This paper is a kind of Whig history- tying into a (I hope) coherent and apparently forward moving narrative what were in fact a number of independent and sometimes conflicting themes. This paper developed from talk at the Boris-fest in 2010. But I have tried here to show how the ideas of Shelah and Zilber complement each other in the development of model theory. Their analysis led to frameworks which generalize first order logic in several ways. First they are led to consider more powerful logics and then to more ‘mathematical’ investigations of classes of structures satisfying appropriate properties. We refer to [Bal09] for expositions of many of the results; that’s why that book was written. It contains full historical references.
    [Show full text]
  • On the Categorical Semantics of Elementary Linear Logic
    Theory and Applications of Categories, Vol. 22, No. 10, 2009, pp. 269{301. ON THE CATEGORICAL SEMANTICS OF ELEMENTARY LINEAR LOGIC OLIVIER LAURENT Abstract. We introduce the notion of elementary Seely category as a notion of cate- gorical model of Elementary Linear Logic (ELL) inspired from Seely's de¯nition of models of Linear Logic (LL). In order to deal with additive connectives in ELL, we use the ap- proach of Danos and Joinet [DJ03]. From the categorical point of view, this requires us to go outside the usual interpretation of connectives by functors. The ! connective is decomposed into a pre-connective ] which is interpreted by a whole family of functors (generated by id, ­ and &). As an application, we prove the strati¯ed coherent model and the obsessional coherent model to be elementary Seely categories and thus models of ELL. Introduction The goal of implicit computational complexity is to give characterizations of complexity classes which rely neither on a particular computation model nor on explicit bounds. In linear logic (LL) [Gir87], the introduction of the exponential connectives gives a precise status to duplication and erasure of formulas (the qualitative analysis). It has been shown that putting constraints on the use of exponentials permits one to give a quantitative analysis of the cut elimination procedure of LL and to de¯ne light sub-systems of LL characterizing complexity classes (for example BLL [GSS92], LLL [Gir98] or SLL [Laf04] for polynomial time and ELL [Gir98, DJ03] for elementary time). In order to have a better understanding of the mathematical structures underlying these systems, various proposals have been made in the last years with the common goal of de¯ning denotational models of light systems [MO00, Bai04, DLH05, LTdF06, Red07].
    [Show full text]
  • The Development of Mathematical Logic from Russell to Tarski: 1900–1935
    The Development of Mathematical Logic from Russell to Tarski: 1900–1935 Paolo Mancosu Richard Zach Calixto Badesa The Development of Mathematical Logic from Russell to Tarski: 1900–1935 Paolo Mancosu (University of California, Berkeley) Richard Zach (University of Calgary) Calixto Badesa (Universitat de Barcelona) Final Draft—May 2004 To appear in: Leila Haaparanta, ed., The Development of Modern Logic. New York and Oxford: Oxford University Press, 2004 Contents Contents i Introduction 1 1 Itinerary I: Metatheoretical Properties of Axiomatic Systems 3 1.1 Introduction . 3 1.2 Peano’s school on the logical structure of theories . 4 1.3 Hilbert on axiomatization . 8 1.4 Completeness and categoricity in the work of Veblen and Huntington . 10 1.5 Truth in a structure . 12 2 Itinerary II: Bertrand Russell’s Mathematical Logic 15 2.1 From the Paris congress to the Principles of Mathematics 1900–1903 . 15 2.2 Russell and Poincar´e on predicativity . 19 2.3 On Denoting . 21 2.4 Russell’s ramified type theory . 22 2.5 The logic of Principia ......................... 25 2.6 Further developments . 26 3 Itinerary III: Zermelo’s Axiomatization of Set Theory and Re- lated Foundational Issues 29 3.1 The debate on the axiom of choice . 29 3.2 Zermelo’s axiomatization of set theory . 32 3.3 The discussion on the notion of “definit” . 35 3.4 Metatheoretical studies of Zermelo’s axiomatization . 38 4 Itinerary IV: The Theory of Relatives and Lowenheim’s¨ Theorem 41 4.1 Theory of relatives and model theory . 41 4.2 The logic of relatives .
    [Show full text]
  • Heyting Duality
    Heyting Duality Vikraman Choudhury April, 2018 1 Heyting Duality Pairs of “equivalent” concepts or phenomena are ubiquitous in mathematics, and dualities relate such two different, even opposing concepts. Stone’s Representa- tion theorem (1936), due to Marshall Stone, states the duality between Boolean algebras and topological spaces. It shows that, for classical propositional logic, the Lindenbaum-Tarski algebra of a set of propositions is isomorphic to the clopen subsets of the set of its valuations, thereby exposing an algebraic viewpoint on logic. In this essay, we consider the case for intuitionistic propositional logic, that is, a duality result for Heyting algebras. It borrows heavily from an exposition of Stone and Heyting duality by van Schijndel and Landsman [2017]. 1.1 Preliminaries Definition (Lattice). A lattice is a poset which admits all finite meets and joins. Categorically, it is a (0, 1) − 푐푎푡푒푔표푟푦 (or a thin category) with all finite limits and finite colimits. Alternatively, a lattice is an algebraic structure in the signature (∧, ∨, 0, 1) that satisfies the following axioms. • ∧ and ∨ are each idempotent, commutative, and associative with respective identities 1 and 0. • the absorption laws, 푥 ∨ (푥 ∧ 푦) = 푥, and 푥 ∧ (푥 ∨ 푦) = 푥. Definition (Distributive lattice). A distributive lattice is a lattice in which ∧and ∨ distribute over each other, that is, the following distributivity axioms are satisfied. Categorically, this makes it a distributive category. 1 • 푥 ∨ (푦 ∧ 푧) = (푥 ∨ 푦) ∧ (푥 ∨ 푧) • 푥 ∧ (푦 ∨ 푧) = (푥 ∧ 푦) ∨ (푥 ∧ 푧) Definition (Complements in a lattice). A complement of an element 푥 of a lattice is an element 푦 such that, 푥 ∧ 푦 = 0 and 푥 ∨ 푦 = 1.
    [Show full text]