Handbook of the World Congress and School on

Universal Logic III

April 18 - 25
Lisbon - Portugal
www.uni-log.org
Edited by Jean-Yves Beziau, Carlos Caleiro, Joao Rasga e
Alexandre Costa-Leite
Contents
1 Organizers of UNILOG10 3
1.1 Scientic Committee . . . . . . . . . . . . . . . . . . . . . . . . . 3
1.2 Organizing Committee . . . . . . . . . . . . . . . . . . . . . . . . 3
2 What is Universal Logic? 4
3 World School on Universal Logic III 5
3.1 Aim of the School . . . . . . . . . . . . . . . . . . . . . . . . . . 5
3.2 Tutorials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
3.2.1 Hybrid Logic . . . . . . . . . . . . . . . . . . . . . . . . . 6
3.2.2 Logical Pluralism . . . . . . . . . . . . . . . . . . . . . . . 6
3.2.3 Truth-Values . . . . . . . . . . . . . . . . . . . . . . . . . 7
3.2.4 Refutation . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
3.2.5 Graded Consequence . . . . . . . . . . . . . . . . . . . . . 9
3.2.6 Quantum Logic . . . . . . . . . . . . . . . . . . . . . . . . 10
3.2.7 Instantiations . . . . . . . . . . . . . . . . . . . . . . . . . 10
3.2.8 Erotetic Logics . . . . . . . . . . . . . . . . . . . . . . . . 11
3.2.9 Institutions . . . . . . . . . . . . . . . . . . . . . . . . . . 12
3.2.10 Ideospheres . . . . . . . . . . . . . . . . . . . . . . . . . . 13
3.2.11 Natural Deductions . . . . . . . . . . . . . . . . . . . . . 14
3.2.12 Kripkes World . . . . . . . . . . . . . . . . . . . . . . . . 14
3.2.13 Consistency . . . . . . . . . . . . . . . . . . . . . . . . . . 15
3.2.14 Fractals, Topologies and Logic . . . . . . . . . . . . . . . 16
3.2.15 The World of Possible Logics . . . . . . . . . . . . . . . . 18
3.2.16 Geometry of Oppositions . . . . . . . . . . . . . . . . . . 19
3.2.17 Truth-Functionality . . . . . . . . . . . . . . . . . . . . . 20
3.2.18 Probabilistic Logic . . . . . . . . . . . . . . . . . . . . . . 22
1
3.2.19 Logics of Empirical Sciences . . . . . . . . . . . . . . . . . 24
3.2.20 Logics of Plurality . . . . . . . . . . . . . . . . . . . . . . 25
3.2.21 How to Cut and Paste Logical Systems . . . . . . . . . . 25
4 Congress on Universal Logic III 27
4.1 Invited Speakers . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
4.2 Sessions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
4.2.1 Logic Diagrams . . . . . . . . . . . . . . . . . . . . . . . . 32
4.2.2 Non-Classical Mathematics . . . . . . . . . . . . . . . . . 36
4.2.3 Abstract Algebraic Logic . . . . . . . . . . . . . . . . . . 42
4.2.4 Paradoxes . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
4.2.5 Substructural Logics . . . . . . . . . . . . . . . . . . . . . 55
4.2.6 Categorical Logic . . . . . . . . . . . . . . . . . . . . . . . 62
4.2.7 Negation . . . . . . . . . . . . . . . . . . . . . . . . . . . 67
4.2.8 Multimodal Logics . . . . . . . . . . . . . . . . . . . . . . 74
4.2.9 Paraconsistent Logics . . . . . . . . . . . . . . . . . . . . 81
4.2.10 Algebras for Logics . . . . . . . . . . . . . . . . . . . . . . 86
4.3 Contest . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92
4.4 Contributed Talks . . . . . . . . . . . . . . . . . . . . . . . . . . 94
5 Sponsors 141
2
1 Organizers of UNILOG10
1.1 Scientic Committee
Joham van Benthem, University of Amsterdam Stanford University, The
Netherlands and USA
Ross Brady, La Trobe University, Melbourne, Australia
Walter Carnielli, UNICAMP, Campinas, Brazil
Michael Dunn, School of Informatics, Indiana USA
Dov Gabbay, Kings College, London, UK
Huacan He, Northwestern PolytechnicalUniversity, Xian, China
Vladimir Vasyukov, Academy of Sciences, Moscow, Russia
Heinrich Wansing, Dresden University of Technology, Germany
1.2 Organizing Committee
Jean-Yves Beziau, CNPq-FUNCAP UFC, Fortaleza, Brazil
Carlos Caleiro, IST, Lisbon, Portugal
Alexandre Costa-Leite, University of Brasilia, Brazil
Katarzyna Gan-Krzywoszynska, Poznan University, Poland Archives Poincare,
University Nancy 2, France
Ricardo Gonalves, IST, Lisbon, Portugal
Paula Gouveia, IST, Lisbon, Portugal
Raja Natarajan, Tata Institute, Mumbai, India
Jaime Ramos, IST, Lisbon, Portugal
Jo ao Rasga, IST, Lisbon, Portugal
Darko Sarenac, Colorado State University, USA
3
2 What is Universal Logic?
In the same way that universal algebra is a general theory of algebraic structures,
universal logic is a general theory of logical structures. During the 20th century,
numerous logics have been created: intuitionistic logic, deontic logic, many-
valued logic, relevant logic, linear logic, non monotonic logic, etc. Universal
logic is not a new logic, it is a way of unifying this multiplicity of logics by
developing general tools and concepts that can be applied to all logics.
One aim of universal logic is to determine the domain of validity of such
and such metatheorem (e.g. the completeness theorem) and to give general
formulations of metatheorems. This is very useful for applications and helps
to make the distinction between what is really essential to a particular logic
and what is not, and thus gives a better understanding of this particular logic.
Universal logic can also be seen as a toolkit for producing a specic logic required
for a given situation, e.g. a paraconsistent deontic temporal logic.
This is the third edition of a world event dedicated to universal logic. This
event is a combination of a school and a congress. The school oers 21 tutorials
on a wide range of subjects. The congress will follow with invited talks and
contributed talks organized in many sessions including 10 special sessions. There
will also be a contest.
This event is intended to be a major event in logic, providing a platform
for future research guidelines. Such an event is of interest for all people dealing
with logic in one way or another: pure logicians, mathematicians, computer
scientists, AI researchers, linguists, psychologists, philosophers, etc.
The whole event will happen nearby Lisbon, Portugal, a place which was
the departure point of many adventures.
UNILOG2010
Join this adventure!
4
3 World School on Universal Logic III
3.1 Aim of the School
This school is on universal logic. Basically this means that tutorials will present
general techniques useful for a comprehensive study of the numerous existing
systems of logic and useful also for building and developing new ones.
For PhD students, postdoctoral students and young researchers interested
in logic, articial intelligence, mathematics, philosophy, linguistics and related
elds, this will be a unique opportunity to get a solid background for their future
researches.
The school is intended to complement some very successful interdisciplinary
summer schools which have been organized in Europe and the USA in recent
years: The ESSLLI (European Summer School on Logic, Language and Informa-
tion) in Europe and the NASSLLI (North American Summer School on Logic,
Language and Information).
The dierence is that our school will be more focused on logic, there will
be less students (these events gather several hundreds of students) and a better
interaction between advanced students and researchers through the combination
of the school and the congress (Participants of the School are strongly encour-
aged to submit a paper for the congress). We also decided to schedule our event
in Spring in order not to overlap with these big events.
5
3.2 Tutorials
3.2.1 Hybrid Logic
Patrick Blackburn
INRIA Nancy Grand-Est - France
[email protected]
These lectures introduce hybrid logic, a form of modal logic in which it is
possible to name worlds (or times, or computational states, or situations, or
nodes in parse trees, or people - indeed, whatever it is that the elements of
Kripke Models are taken to represent).
The course has three major goals. The rst is to convey, as clearly as possi-
ble, the ideas and intuitions that have guided the development of hybrid logic.
The second is to introduce a concrete skill: tableau-based hybrid deduction.
The third is to say a little about the history of the subject, and to link it to
philosophical the work of Arthur Prior. No previous knowledge of hybrid logic
is assumed.
The lecture outline is as follows:
Lecture 1: From modal logic to hybrid logic
Lecture 2: Hybrid deduction
Lecture 3: The Priorean perspective
Bibliography:
1. Arthur Prior and Hybrid Logic, by Patrick Blackburn. Synthese, 150,
329-372, 2006.
2. Representation, Reasoning, and Relational Structures: a Hybrid Logic
Manifesto, by Patrick Blackburn. Logic Journal of the IGPL, 8(3), 339-
625, 2000.
3. Chapter 7, Section 3 of Modal Logic, by Patrick Blackburn, Maarten de Ri-
jke and Yde Venema. Cambridge Tracts in Theoretical Computer Science,
53, Cambridge University Press, 2001.
3.2.2 Logical Pluralism
Catarina Dutilh-Novaes
University of Amsterdam - The Netherlands
[email protected]
Ole Hjortland
University of St. Andrews - Scotland
[email protected]
Up until the end of the 19th century, logic was typically seen as the art and
science of correct reasoning, and in this sense it was not specic to any discipline
6
or subject-matter (i.e. logic was seen as topic-neutral); moreover, even if there
might have been competing systems, the general opinion was that there should
be only one true logic. True enough, at dierent times the scope of logic went
very much beyond correct reasoning in the sense of drawing inferences strictly
speaking: in the Latin medieval period, for example, what we now call semantics
and much of what we now call epistemology also belonged to the realm of logic.
Still, even though there were dierent logical theories for dierent applications,
when it came to reasoning, syllogistic inference remained the canon of correct
reasoning for almost 2.500 years.
A bit over a century later, the status of logic as a discipline has changed
dramatically: we now have dierent logics, specially designed for certain situa-
tions, topics or tasks what one could describe as a situation of logical plurality
and the idea that there is only one correct way of reasoning and thus only one
correct logic is no longer unanimously accepted what is now referred to as the
position of logical pluralism. In this context, a universal logic seems to be a
welcome development: it should allow for the comparison between systems, and
for the arbitration of the disputes between competing systems. However, this
contemporary plethora gives rise to certain philosophical questions concerning
logical pluralism and universality, in particular the kind of universality claimed
by universal logic. The tutorial will look at logical pluralism against the back-
ground of the history of logic, and tie it to the recent debate started by Beall
and Restalls book Logical Pluralism (2006, OUP). We will then look at a num-
ber of reactions to logical pluralism in the literature, and try to connect some
of the questions to the project of Universal Logic.
Session (1) The Emergence of Logical Plurality and Logical Pluralism
Session (2) Contemporary Logical Pluralism Session (3) Reactions to Logical
Pluralism
References
1. J. Y. Beziau, Universal Logic. In T. Childers O. Majer (eds), Logica94
-Proceedings of the 8th International Symposium, Philosophia, Prague,
pp. 73-93.
2. JC Beall and G. Restall, Logical Pluralism. Oxford University Press, 2006.
3. S. Haack, Philosophy of Logics. Cambridge University Press, 2000.
4. G. Priest, Doubt Truth to Be a Liar. OUP, 2006 (ch. 12).
5. G. Restall, Carnaps Tolerance, Meaning, and Logical Pluralism, Journal
of Philosophy, 99(8): 426-443, 2002.
3.2.3 Truth-Values
Heinrich Wansing
Dresden University of Technology - Germany
7
[email protected]
Fabien Schang
Dresden University of Technology - Germany
[email protected]
The notion of a truth value has been introduced into logic and philosophy
by Gottlob Frege in two seminal papers in 1891 and 1892. Truth values play a
central role in logic and have been put to quite dierent uses. They have been
characterized as:
primitive abstract objects denoted by sentences in natural and formal
languages,
abstract entities hypostatized as the equivalence classes of sentences,
what is aimed at in judgements,
values indicating the degree of truth of sentences,
entities that can be used to explain the vagueness of concepts,
values that are preserved in valid inferences,
idealizations of basic proof-theoretical properties,
values that convey information concerning a given proposition.
Depending on their particular use, truth values have been treated as unan-
alyzed, as dened, as unstructured, or as structured entities.
In this tutorial, we will briey comment on the history of the notion of a
truth value and on the various uses of this notion in philosophy and logic. We
will then focus on some selected topics including the famous slingshot argument
and Suszkos Thesis. The slingshot argument may be interpreted to call into
question the view that sentences denote truth values, and Suszkos Thesis was
meant to prove that there are but two logical values, Freges the True and the
False. Moreover, the tutorial will deal with generalized truth values as used in
Belnaps useful four-valued logic (FDE) and extensions of FDE.
Bibliography:
1. Truth values. Part I, Studia Logica 91-3 (2009), special issue edited by Y.
Shramko and H. Wansing.
2. Truth values. Part II, Studia Logica 92-2 (2009), special issue edited by
Y. Shramko and H. Wansing
8
3.2.4 Refutation
Tom Skura
University of Zielona G ora - Poland
[email protected]
Refutation systems are inference systems, just like traditional axiomatic sys-
tems, but they generate non-valid formulas rather than valid ones. They consist
of refutation axioms (which are non-valid formulas) and refutation rules (which
are rules preserving non-validity).
In this tutorial the following topics are considered.
Examples of syntactic refutations in non-classical logics, and a general
theory of refutation systems.
Refutation systems and other standard methods (sequent systems, tableau
procedures, model building constructions, the nite model property).
Tools and techniques for proving syntactic completeness (characteristic
formulas of nite algebras, normal forms and inductive completeness proofs).
References:
1. V. Goranko, Refutation systems in modal logic. Studia Logica 53 (1994),
299-324.
2. J. Lukasiewicz, Aristotles Syllogistic from the Standpoint of Modern For-
mal Logic. Oxford, 1951.
3. D. Scott, Completeness proofs for the intuitionistic sentential calculus.
In Summaries of talks presented at the Summer Institute of Symbolic
Logic.Cornell University, 1957. second edition, Princeton, 1960, 231-241.
4. T. Skura, Refutations, proofs, and models in the modal logic K4. Studia
Logica 70 (2002), 193-204.
5. T. Skura, Maximality and refutability. Notre Dame Journal of Formal
Logic 45 (2004), 65-72.
6. T. Skura, A refutation theory. Logica Universalis (2009).
7. T. Skura, Refutation systems in propositional logic. In D. Gabbay and F.
Guenthner (eds.) Handbook of Philosophical Logic, Volume 16, 2010.
3.2.5 Graded Consequence
Mihir Chakraboty
University of Calcutta - India
9
3.2.6 Quantum Logic
Paulo Mateus
Technical University of Lisbon - Portugal
[email protected]
In this tutorial we start by reviewing the relevant concepts and results in
linear algebra and operator theory. We also present Dirac notation and the
postulates of quantum mechanics.
Then, we motivate he notion of orthocomplemented lattice and present the
main results concerning Birkho and von Neumann quantum logics.
Moved by the emergence of new practical applications on quantum informa-
tion and computation, we survey modern approaches to reason about quantum
ystems.
We focus on two approaches: exogenous quantum propositional logic (EQPL)
by Mateus and Sernadas; and probabilistic modal logic (PML) by van der Mey-
den and Patra. We present several result concerning EQPL such as: complete
Hiblert calculus; complexity of the model-checking and SAT problems; and how
to deal with dynamic extensions. We illustrate both EQPL and PML by rea-
soning about quantum security protocols.
Finally, we briey discuss how to extend a logic nto a quantum logic.
3.2.7 Instantiations
Arthur Buchsbaum
Federal University of Santa Catarina - Brazil
[email protected]
Instantiation is an operation essential for performing logical reasoning, it
is placed in the heart of its basis, when dealing with universal and existential
quantiers. In many logics it must appear explicitly in at least three of the four
laws of introduction and elimination of quantiers.
In spite of it, the study and presentation of this operation has been under-
estimated in most books on logic, usually only half of a page is dedicated to it.
Beyond that, when it is presented, it is often wrongly confused with a particular
case of another operation, named here replacement, which is used on formula-
tion of some laws related to equivalence and equality. While replacement of
a variable by a term considers all occurrences of variables which dont succeed
a quantier or a qualier, instantiation consider only free occurrences of this
variable; the former does not rename bound variables, while the latter, for a
good working, needs renaming of bound variables in many of its forms.
By consequence of lack of carefulness in dealing with instantiation, when
there is no renaming of bound variables, the presentation of the basic laws of
introduction and elimination of quantiers needs some patches, and these ones
become more dicult reasoning perform from this point on.
In this tutorial all possible alternatives of instantiation are presented, and
10
one of them, maybe the most promising one, is presented in detail, in order to
make easier reasoning perform inside a logical system.
3.2.8 Erotetic Logics
Andrzej Wisniewski
Adam Mickiewicz University in Pozna n
The term erotetic logic is often understood as synonymous to t he logic of
questions. There is no common agreement as to what erotetic logic should
be. The most developed proposals will be overviewed and compared. Then
a general setting, the Minimal Erotetic Semantics (MES), will be presented;
kinds of answers to questions, types of their presuppositions, basic relations
between questions, and certain normative concepts pertaining to questions will
be characterized in terms of MES. Next, conceptual foundations of Inferential
Erotetic Logic (IEL) will be discussed. IEL focuses its attention on erotetic
inferences, that is, roughly, inferences which have questions as conclusions. Some
of these inferences are intuitively valid; we will show how IEL explicates the
relevant concept of validity.
We will also address some more technical issues. First, we will consider
models of problem decomposition, oered by Hintikkas Interrogative Model
of Inquiry and by IEL. Second, a certain proof method grounded in IEL, the
Socratic proofs method, will be presented.
Finally, the idea of erotetic logic as a theory of internal question processing
will be discussed.
References
1. Groenendijk, J., Stokhof, M., Questions, in: J. van Benthem, and A. ter
Meulen (eds.), Handbook of Logic and Language, Elsevier, Amsterdam/
New York 1996, pp. 1055-1125.
2. Harrah, D., The Logic of Questions, in: D. Gabbay, and F. Guenth-
ner (eds.), Handbook of Philosophical Logic, Second Edition, Volume 8,
Kluwer, Dordrecht/ Boston/ London 2002, pp. 1-60.
3. Hintikka, J., Inquiry as Inquiry: A Logic of Scientic Discovery, Kluwer,
Dordrecht/ Boston/ London 1999.
4. Wisniewski, A., The Posing of Questions: Logical Foundations of Erotetic
Inferences, Kluwer, Dordrecht/ Boston/ London 1995.
5. Wisniewski, A., Questions and Inferences, Logique et Analyse 173-175,
2001, pp. 5-43.
6. Wisniewski, A., Erotetic Search Scenarios, Synthese 134, No. 3, 2003, pp.
389-427.
11
7. Wisniewski, A., Shangin, V., Socratic Proofs for Quantiers, Journal of
Philosophical Logic 35, No. 2, 2006, pp. 147-178.
3.2.9 Institutions
Razvan Diaconescu
Institute of Mathematics Simion Stoilow (IMAR) - Romania
[email protected]
Institution theory is a major model theoretic trend of universal logic that
formalizes within category theory the intuitive notion of a logical system, in-
cluding syntax, semantics, and the satisfaction relation between them. It arose
within computing science, especially specication theory [1], as a response to the
population explosion of logics there and where it has become the most impor-
tant foundational theory Later on institution theory has been succesfully used
in pure logic studies in the spirit of universal logic. This means the development
of model and proof theory in the very abstract setting of arbitrary institutions,
free of commitement to a particular logical system [2]. In this way we gain
freedom to live without concrete models, sentences satisfaction, and so on, we
gain another level of abstraction and generality and a deeper understanding of
model theoretic phenomena not hindered by the largely irrelevant details of a
particular logical system, but guided by structurally clean causality. The latter
aspect is based upon the fact that concepts come naturally as presumed features
that a logic might exhibit or not and are dened at the most appropriate level
of abstraction; hypotheses are kept as general as possible and introduced on a
by-need basis, and thus results and proofs are modular and easy to track down
regardless of their depth. The continuous interplay between the specic and
the general in institution theory brings a large array of new results for par-
ticular non-conventional, unies several known results, produces new results in
well-studied conventional areas, reveals previously unknown causality relations,
and dismantles some which are usually assumed as natural. Access to highly
non-trivial results is also considerably facilitated. The dynamic role played by
institution theory within the wider universal logic project is illustrated by the
fact that institution theory papers have come second and rst, respectively, in
the contests of the Montreux (2005) and Xiand (2007) UNILOG, respectively.
In this tutorial we will start with a brief explanation of the historical and
philosophical origins of institution theory, followed by a presentation of its basic
mathematical concepts. We will also have a trip through the rather rich body
of methods and results of the institution theoretic approach to logic and model
theory. Although institution theory is primarily a model theoretic approach
we will also discuss recent proof theoretic developments in the area. However
our real emphasis will be not on the actual mathematical developments but
on the non-substantialist way of thinking and the top-down methodologies pro-
moted by institution theory, that contrast sharply the substantialist view and
the bottom-up methodologies that pervade and underly conventional logic, this
being the most profound message of institution theory as a universal logic trend.
12
References:
1. J. Goguen, R. Burstall, Institutions: Abstract Model Theory for Speci-
cation and Programming, J.ACM 39(1) 95-146, 1992.
2. R. Diaconescu, Institution-independent Model Theory, Birkhauser, 2008.
3.2.10 Ideospheres
Dominique Luzeaux
DGA, Ministry of Defense - France
[email protected]
Jean Sallantin
Le Laboratoire dInformatique, de Robotique et de Microelectronique
de Montpellier - France
[email protected]
This tutorial will present formal structures which often structure the develop-
ment of ideospheres in Human Sciences. An ideosphere (Barthes77) is initiated
by a founding speech and helps establish correspondences to other such speeches
and take commitment and refusal positions in a system that is inachieved. In
pedopsychiatry, an ideosphere is focus on early interactions between the infant
and his environment, and examine the processes of semiotization, as well as the
use of representation abilities as a means to communicate (Golse 99,07) (Dor
02).
These structures and their organization within a general system can be for-
malized with category theory, as is done for instance when modeling compu-
tation systems and relating dierent models. We show under what conditions
they correspond to a formalization within modal logic of the system in use; at
this point we will make a comparison with what is done with categorial mod-
els which relate various logics. Finally we develop the concepts of autonomy
and learning, and use them to illustrate the presented mathematical tools and
methods; this will help model zig-zag processes between various formalizations
Bibliography:
1. Barthes R., Le Neutre, Cours au coll`ege de France 1977-1978.
2. Dor J., Introduction ` a la lecture de Lacan : linconscient structure comme
un langage, Espace analytique , Denoel 2002.
3. Golse B., Du corps ` a la pensee, Le l rouge, PUF, 1999. .
4. Golse B, letre-bebe, Le l rouge PUF 2007.
13
3.2.11 Natural Deductions
Fernando Ferreira
University of Lisbon - Portugal
[email protected]
The theory of natural deduction of pure intuitionistic logic for the (, , A)-
fragment is very elegant. It includes a strong normalization theorem and a
Church-Rosser property. This elegant treatment extends to pure second-order
intuitionistic logic (Girards polymorphic system F). The technical details are
more complicated in this case because of the impredicativity of the second-order
quantier, but strong normalization (and Church-Rosser) still holds.
The other connectives (absurdity, disjunction, existential quantication),
whose features are more typical of intuitionism, do not have such an elegant
treatment. Girard sees their elimination rules as defective. In second-order
logic, these conectives can be circumvented because they are denable in terms
of the others. It is not widely known that these conectives can also be cir-
cumvented in the treatment of rst-order logic if we embed it in a version of
predicative second-order logic.
Our tutorial will explain these issues.
Bibliography:
1. F. Ferreira. Comments on predicative logic, Journal of Philosophical Logic
35, pp. 1-8 (2006).
2. J.-Y. Girard et al. Proofs and Types, Cambridge University Press, 1989.
D. Prawitz. Natural Deduction, Dover Publications, 2006.
3. W. Tait. Intensional interpretation of functionals of nite type I, The
Journal of Symbolic Logic 32, pp. 198-212 (1967)
4. A. S. Troelstra H. Schwichtenberg. Basic Proof Theory, Cambridge Uni-
versity Press, 2000.
3.2.12 Kripkes World
Andreas Herzig
IRIT at Universite Paul Sabatier - France
[email protected]
Possible worlds models have been introduced by Saul Kripke in the early
sixties. Basically they are graphs with labelled nodes and edges. Such models
provide semantics for various modal logics such as temporal logics, logics of
knowledge and belief, logics of programs, logics of action, logics of obligation, as
well as for description logics. They also provide semantics for other nonclassical
logics such as intuitionistic logics and conditional logics. Such logics have been
studied intensively in philosophical and mathematical logic and in computer
14
science, and have been applied in various domains such as theoretical computer
science, articial intelligence, and more recently as a basis of the semantic web.
Semantic tableaux are the predominant reasoning tool for all these logics: given
a formula (or a set of formulas) they allow to check whether it has a model or not.
A tableau calculus uses a set of rules in order to build trees, and more generally
graphs. If a tableau graph is contradiction free then it can be transformed into
a model for the given formula.
The aim of the tutorial is to provide a step-by-step introduction to modal
logics, both in terms of Kripke models and in terms of semantic tableaux. The
dierent logics will be illustrated by means of examples. We will use the generic
tableaux theorem prover Lotrec (http://www.irit.fr/Lotrec), which is a piece of
software that allows to build models, check whether a given formula is true in
a model, and check whether a given formula is valid in a given logic. Lotrec
also allows to implement tableau systems for new logics by means of a sim-
ple interactive graph-based language accessible to users that are not computer
scientists. The tutorial requires basic mathematical and logic background (the
basic denitions of graph theory, and the bases of classical propositional logic).
Contents:
1. graphs, Kripke frames and Kripke models; model checking;
2. the basic modal logic K and its tableaux; soundness, completeness and
decidability;
3. description logics; - the basic modal logics (KD, KT, S4, S5,...);
4. modal logics with transitive closure (PDL, logic of common knowledge).
3.2.13 Consistency
Jui-Lin Lee
National Formosa University Taiwan
[email protected]
In classical logic the extended completeness theorem (T implies that
T for any set of sentences T and any sentence ) is frequently proved by
the following two steps:
(CME) Every consistent set has a (classical) model; and
(RAA) If T cannot derive , then T is consistent (for any set T and
any sentence ).
Sometimes the former statement, as a major step of this approach, is called
the extended completeness theorem (or the strong completeness theorem). This
is not always a correct name of it because there are some non-classical log-
ics satisfying CME, the classical model existence property (which means that
consistency implies classical satisability).
In this course we investigate this meta-logical property CME. Since there
are many dierent consistencies and in non-classical logics they are not always
15
equivalent, we will study consistency rst. Note that with dierent consistencies
the meaning of CME could be dierent.) Then we present ways to construct
non-classical logics/proof systems (by selecting axioms or rules) which satisfy
CME, and discuss how one can construct a weaker system still satisfying CME.
These (propositional or predicate) logics include some (weak extensions of) para-
consistent logics, subintuitionistic logics, or substructural logics. Applications
of CME include Glivenko-style theorem and pure implicational logic. Further-
more, we will also analyze the necessary-and-sucient condition of CME (which
is related to Left Resolutation Gentzen system in [3]) and discuss other model
existence property.
References:
1. J.-Y. Beziau, Sequents and bivaluations, Logique et Analyse, Vol. 44,
pp.373-394, 2001.
2. Jui-Lin Lee, Classical model existence theorem in propositional logics, in
Jean-Yves Beziau and Alexandre Costa-Leite (eds.),
3. Perspectives on Universal Logic, pages 179-197, Polimetrica, Monza, Italy,
2007.
4. Jui-Lin Lee, Classical Model Existence and Left Resolution, Logic and
Logical Philosophy, Vol. 16 No 4, pages 333-352, 2007.
5. Jui-Lin Lee, The Classical Model Existence Theorem in Subclassical Pred-
icate Logics I, in D. Makinson, J. Malinowski, H. Wansing (eds.), Towards
Mathematical Philosophy, Trends in Logic 28, pages 187-199, Springer,
2008.
6. Richard Zach, Proof Theory of Finite-Valued Logics, Diploma Thesis,
Technische Universitat Wien, Vienna, 1993.
3.2.14 Fractals, Topologies and Logic
Tamar Lando
University of Berkeley - USA
tlando at berkeley.edu
Darko Sarenac
Colorado University - USA
[email protected]
This tutorial explores the connection between fractal geometry and topolog-
ical intentional logic. The main cluster of results we present is a class proofs of
completeness of the modal logic S4,and Intuitionist Propositional Logics with
respect to the following fractals: Koch Snowake/Curve, Sierpinski Carpet,
Menger Cube, which all are well known classic fractals. The main corollary
16
of the paper is a new proof of the completeness of afore mentioned logics for
most signicant n-dimensional Euclidean metric topological space R, R
2
and
R
3
. The latter results were originally obtained by Tarski and MacKinsey in [7],
and much simplied and rened by Mints et al. and van Benthem et al. [exam-
ples, variations, and renements can be found in [1], [2] and [4]]. Our new proof
uses fractal techniques, that, as we will argue is the main contribution to the
topological semantics for modal logic in recent years. The completeness for both
Koch Curve and other fractals and R
n
are best seen as examples of the power of
the fractal techniques introduced here. Another somewhat original contribution
is the use of the innite binary tree with limits, Wilson tree. We prove that
such a tree is complete for S4 and related logics in several extended languages.
Such proof then facilitates topological transfer onto fractals and metric spaces.
Although this tree has not to our knowledge been used previously in the modal
logic community, and we have introduced it independently, it has since come to
our attention that the tree has been used and named in category theory by Peter
Freyd sometime in the late 1980s. [We follow his naming convention in calling
the tree Wilson tree. Wilson tree is obtained by a kind of model saturation
technique called sobering from the more usual innite binary branching tree.
The name is then derived from the fact that Wilson was the founder of the 12
step recovery program.]
In this tutorial the following topics are considered.
We begin the tutorial by placing topological techniques and the techniques
of fractal geometry in the pantheon of various spatial techniques in both uni-
versal approach to logic and in the full gamut of formal approaches to reasoning
about space, space-time, and spatiotemporal dynamics. In the spirit of the
school on universal logic, we will be emphasizing techniques that are portable
and universally applicable over particular results and systems. We continue
by introducing a class of self-similar fractals and discussing their usefulness in
various model-construction techniques in topological modal logic. We look at a
series of well-known trees, both nite and innite discussing their fractal nature
and demonstrating their connection with well known fractals, such as for in-
stance, Koch Flake/Curve, Sierpinsiki Carpet, Menger Cube, and many others.
We then explore various topological and logical properties of fractals. We will
show that each of the three fractals mentioned, Koch, Sierpinski, and Menger,
is in some sense universal for a class of model theoretic or topological objects.
A distant topological relative of this result was rst proved by Sierpinski in
the early 1920s. We will explore logical relevance of this result. In showing
the relevance we will use another three called Wilson Tree. This treelike the
innite binary branching treeought to be well known, but it is not. We will
explore the immense usefulness of the tree in reasoning about complete metric
Euclidean spaces and speculate about the reasons why the tree is not more com-
monly known among modal (and intensional) logicians. As we mention in the
introduction above, after we reinvented the tree, it turned out to be a rather
standard object in Category Theory. We will explore some connections between
results in category theory by P. Freyd and A. Scedrov and some well known
results in topological semantics.
17
We will conclude the tutorial by looking at the class of intensional logics,
(S4, S4u, S4 + time, Int. Logic, etc.) to which the fractal techniques introduced
here apply. If we have some extra time, we will look at a recently introduced
Probabilistic Modal Logics of Dana Scott.
References
1. M. Aiello, J. van Benthem, and G. Bezhanishvili, Reasoning about space:
the modal way, Journal of Logic and Computation (13)6, 889920, (2003).
2. G. Bezhanishvili, M. Gehrke, Completeness of S4 with respect to the real
line: revisited, Ann. Pure Appl. Logic 131(1-3): 287-301 (2005)
3. P. J. Freyd and A. Scedrov Allegories, Categories, North-Holland (1999)
4. P. Kremer and G. Mints, Dynamic Topological Logic, in Handbook
of Spatial Logic, M. Aiello, I. Pratt-Hartmann, J. van Benthem (Eds.),
Springer, pp 565-606, (2007).
5. T. Lando and D. Sarenac, Sierpinski Carpet, Menger Sponge, and Fractal
Techniques in Topological Modal Logic. (Forthcoming 2009.)
6. T. Lando and D. Sarenac, Fractal Completeness Techniques in Topolog-
ical Modal Logic: Koch Curve, Limit Tree, and the Real Line. (Forth-
coming 2009.)
7. J.C.C. McKinsey and A. Tarski, The algebra of topology, Ann. of Math.
(2) (45), 141191, (1944).
8. M. Schroeder, Fractals, Chaos, Power Laws: Minutes from an Innite
Paradise, New York: W. H. Freeman, pp. 41-45, (1991).
3.2.15 The World of Possible Logics
Jean-Yves Beziau
CNPq-FUNCAP/University of Cear a, Brazil
[email protected]
Many logics are possible: from a logic where nothing is a consequence of
nothing, to a logic where everything is a consequence of everything passing
through a whole spectrum including classical logic, many-valued logic, turbo
polar linear logic.
One can wonder if we are not then facing a wild jungle, from which some
monsters like anti-classical logic, the complement of classical logic where a
proposition is not a consequence of itself, have to be rejected.
In this tutorial we will discuss these questions. We will also present method-
ologies to construct logics and frameworks to compare them.
Many examples and concepts will be provided.
Bibliography:
18
1. J-.Y.Beziau, From consequence operator to universal logic: a survey of
general abstract logic in Logica Universalis; Towards a general theory of
logic, Birkhuser, Basel, 2007, pp.3-17.
2. Haskell Curry, Foundations of mathematical logic, McGraw-Hill, New
York 1963.
3. Dov Gabbay (ed), What is a logical system? Clarendon, Oxford, 1994.
4. Jean Porte, Recherches sur la theorie generale des syst`emes formels et sur
les syst`emes connectifs, Gauthiers-Villars, Paris and Nauwelaerts, Lou-
vain, 1965.
3.2.16 Geometry of Oppositions
Alessio Moretti
University of Nice - France
n-opposition theory (2004) generalises Aristotles opposition theory (exem-
plied by the logical square) by introducing the notion of logical bi-simplex of
dimension n.
This theory, relevant to both quantication theory and modal logic (both
are tied to the logical square) shows that there exists a eld, between logic and
geometry, where logical-geometrical n-dimensional solids (highly symmetrical
structures whose edges are implication arrows), instead of being limited to the
square (the poorest and ugliest of them), develop into innite growing orders
according some relatively simple but generally unknown principles.
This eld is related to modal logic, in so much such n-dimensional structures
can be decorated (as can the square) with arbitrary modalities via some suited
decorating techniques (as the modal n(m)-graphs, or the setting-method, for
instance).
The theorys known applications, so far, concern mainly the study from a
new geometrical point of view of the known modal systems (normal or non-
normal, abstract or applied), but also the study of the non-logical formalisms
inspired by the logical square in psychology (cognitive science and psychoanaly-
sis), linguistics (semiotics and pragmatics), philosophy (analytical as continen-
tal) and others.
In this tutorial we will introduce to the theory of n-opposition, showing
(rapidly) its historical roots (Aristotle, Vasilev, Sesmat, Blanche), then con-
centrating both on its classical (recently discovered) tenets and on its actual
(open) research issues. At the end of the tutorial, according to the time left, we
will discuss of some of this theorys possible applications.
The theory of n-opposition
1. Beziau, J.-Y., Paraconsistent logic from a modal viewpoint, Journal of
Applied Logic, 3 (2005) 7-14.
2. Blanche,R. Sur lopposition des concepts, Theoria, 19 (1953).
19
3. Moretti, A., Geometry for Modalities? Yes: Through n-Opposition The-
ory, in Aspects of Universal Logic, Neuchtel, 2004.
4. Pellissier, R., Setting n-opposition(to appear)
3.2.17 Truth-Functionality
Joao Marcos
Federal University of Rio Grande do Norte - Brazil
[email protected]
In research areas so diverse as model theory, philosophy of language, formal
and computational linguistics, algebraic logic, and the denotational semantics
of programming languages, a common widely accepted meta-theoretical com-
positionality principle is to be found according to which the denotation of a
complex expression is built up from the denotations of its parts. In proof the-
ory and automated reasoning, also, two frequent ways in which such a structural
connection is displayed between the whole and its parts, or between the pre-
misses and their conclusions, resides in the so-called subformula property and
its connections to cut-elimination and interpolation, as well as in the notion of
analyticity of proof formalisms and strategies. One of the most straightforward
ways of realizing such principles, notions and properties is exactly through the
intuitive and well-known notion of *truth-functionality*.
The rst session of the tutorial will directly explore the characterization of
truth-functionality from the viewpoint of Universal Logic. Abstract properties
dening (single-conclusion) consequence relations that have adequate semantical
counterparts in terms of truth-tables will be surveyed, and non-truth-functional
LOGICS will be illustrated and carefully distinguished from logics that are still
truth-functional but turn out to be circumstantially characterized by way of non-
truth-functional SEMANTICS. Characterizability of logics by nite collections
of truth-values and operators will also be touched upon.
The second session of the tutorial will be devoted to a reexamination of
many-valued logics and fuzzy logics as inferential mechanisms based on truth-
functionality. The so-called Suszkos Thesis, according to which a distinction
can be made among algebraic truth-values, on the one hand, and logical
values on the other, will be explained, and a constructive approach to the result,
as applied to logics characterized by nite-valued truth-functional semantics
will be exhibited. Applications to proof theory and rewrite systems, including
generalizations of both the subformula property and the analyticity requirement,
will next be illustrated.
The nal session of the tutorial will recall some basic semantical results con-
nected to functional completeness and pre-completeness, as well as to maximal-
ity of logics with a so-called standard truth-tabular semantics. Generalizations
of truth-tabularity by the consideration of agents that behave truth-functionally
in their own quarters but that appear not to do so when combined by way of
a so-called *society semantics*, other *non-deterministic* versions of truth-
20
tabularity, and also some controlled combinations of truth-tabular scenarios by
way of the so-called *possible-translations semantics* will also be illustrated,
and shown to share some of the good computational behaviors of the standard
approach to truth-functionality, in terms of the preservation or the enjoyment
of important meta-theoretical properties such as decidability, compactness and
modularity.
References:
1. A. Avron and I. Lev, Non-deterministic multiple-valued structures, Journal
of Logic and Computation, 15 (2005), pp.241-261.
2. C. Caleiro and J. Marcos, Classic-like analytic tableaux for nite-valued
logics, in H. Ono, M. Kanazawa, and R. de Queiroz (eds.), Proceedings
of the XVI Workshop on Logic, Language, Information and Computa-
tion (WoLLIC 2009), vol. 5514 of Lecture Notes in Articial Intelligence,
Springer, 2009, pp.268-280.
3. C. Caleiro, W. Carnielli, M.E. Coniglio, and J. Marcos,Twos company:
The humbug of many logical values, in J.-Y. Beziau (ed.), Logica Univer-
salis, Birkh auser Verlag, 2005, pp.169-189.
4. W.A. Carnielli, M.E. Coniglio, D. Gabbay, P. Gouveia, and C. Sernadas,
Analysis and Synthesis of Logics: How To Cut And Paste Reasoning Sys-
tems, vol. 35 of Applied Logic, Springer, 2008.
5. W.A. Carnielli, Possible-translations semantics for paraconsistent logics,
in D. Batens, C. Mortensen, G. Priest, and J.P. Van Bendegem (eds.),
Frontiers of Paraconsistent Logic, Research Studies Press, Baldock, 2000,
pp.149-163.
6. W.A. Carnielli and M.E. Coniglio, Splitting logics, in S. Artemov, H. Bar-
ringer, A. Garcez, L. Lamb, and J. Woods (eds.), We Will Show Them! Es-
says in Honour of Dov Gabbay, vol.1, College Publications, 2005, pp.389-
414.
7. W.A. Carnielli, M.E. Coniglio, and J. Marcos, Logics of Formal Inconsis-
tency, in D. Gabbay, and F. Guenthner (eds.), Handbook of Philosophical
Logic, vol.14, 2nd edn., Springer, 2007, pp.1-93.
8. J. Dugundji, Note on a property of matrices for Lewis and Langfords
calculi of propositions, The Journal of Symbolic Logic, 5 (1940), pp.150-
151.
9. J. Marcos, Possible-translations semantics, in W.A. Carnielli, F.M. Dion-
sio, and P. Mateus, (eds.), Proceedings of the Workshop on Combina-
tion of Logics: Theory and applications (CombLog04), Departamento de
Matematica, Instituto Superior Tecnico, Lisbon, PT, 2004, pp.119-128.
21
10. J.Marcos, Possible-translations semantics for some weak classically-based
paraconsistent logics, Journal of Applied Non-Classical Logics, 18 (2008),
pp.7-28.
11. J. Marcos and D. Mendona, Towards fully automated axiom extraction for
nite-valued logics, in W. Carnielli, M.E. Coniglio, and I.M.L. DOttaviano
(eds.), The Many Sides of Logic, Studies in Logic, College Publications,
London, 2009.
12. J. Marcos, What is a non-truth-functional logic?, Studia Logica, 92 (2009),
pp.215-240.
13. D.J. Shoesmith and T.J. Smiley, Deducibility and many-valuedness, The
Journal of Symbolic Logic, 36 (1971), pp.610-622.
14. R.W ojcicki, Logical matrices strongly adequate for structural sentential
calculi, Bulletin de lAcademie Polonaise des Sciences, Serie des Sciences
Mathematiques, Astronomiques et Physiques, 17 (1969), pp.333-335.
3.2.18 Probabilistic Logic
Gregory Wheeler
Universidade Nova de Lisboa - Portugal
[email protected]
Logic is traditionally understood to be the study of what follows from what
and probabilistic logic is no exception. But whereas the entailment problem
typically asks whether a conclusion C is entailed by some premises P
1
, ..., P
n
,
with respect to some specied entailment relation , written as
(1) P
1
, ..., P
n
C?
on our approach (Haenni et al. 2009) to probabilistic logic poses a slightly
dierent question. In a probabilistic setting, attached to each premise P
i
is a
probability or set of probabilities, X
i
. But it is rare to ask whether a set of
probabilistic premises entails that a particular probability (set of probabilities)
Y is assigned to C. Instead, the entailment problem in probabilistic logic typi-
cally concerns what probability (set of probabilities) are assigned to a particular
conclusion, C, written as
(2) P
1
(X
1
), ..., P
n
(X
n
) C?
A surprisingly wide variety of probabilistic semantics can be plugged into (2),
thereby providing dierent semantics for the entailment relation and allowing
those dierent systems to be studied as bona de logics. In this respect we
view (2) as the fundamental question of probabilistic logics and view it as the
lynch pin to out proposal for unifying probabilistic logics. As for how to answer
the fundamental question, we propose a unifying approximate proof procedure
utilizing credal networks, which are probabilistic graphical models analogous
to Bayesian networks but congured to handle sets of probability functions
representing interval valued probabilities.,
22
This course introduces the progicnet framework for unifying probabilistic
logic in three parts:
Day 1. An introduction to the fundamental question for probabilistic logic
and an introduction to the most basic semantics, our generalization of the stan-
dard semantics of Bayesian probabilistic logic, as put forward in Ramsey (Ram-
sey 1926) and De Finetti (de Finetti 1937), and explicitly advocated by Howson,
(Howson 2001, 2003), Morgan (Morgan 2000), and Halpern (Halpern 2003), to
handle interval-valued probability assignments via sets of probabilities.
Day 2. An introduction to a semantics for handling relative frequency in-
formation, Evidential Probability (EP) (Kyburg 1961, Kyburg and Teng 2001,
Kyburg et al. 2007, Wheeler and Williamson 2009). EP is traditionally thought
of as a logic of probability rather than a probabilistic logic (Levi 2007), and our
framework helps to explain why this is so. Time permitting we will introduce
some extensions to EP that utilize dierent features of the progicnet framework.
Day 3. An introduction to credal networks (Levi 1980, Cozman 2000)and
approximate proof theory we develop with this machinery. Several open prob-
lems will be presented.
References:
1. Cozman, F. (2000). Credal networks. Articial Intelligence, 120(2):199-
233.
2. de Finetti, B. (1937). La prvision: ses lois logiques, ses sources subjectives.
Annales de lInstitut Henri Poincare, 7(1):1-68.
3. Haenni, R., Williamson, J., Romeyn, J.-W., and Wheeler, G. (2009).
Probabilistic Logic and Probabilistic Networks. Synthese Library. Springer,
Dordrecht.
4. Halpern, J. Y. (2003). Reasoning about Uncertainty. MIT Press, Cam-
bridge, MA.
5. Howson, C. (2001). The logic of Bayesian probability. In Coreld, D.
and Williamson, J., editors, Foundations of Bayesianism, pages 137-159.
Kluwer, Dordrecht.
6. Howson, C. (2003). Probability and logic. Journal of Applied Logic, 1(3-
4):151-165.
7. Kyburg, Jr., H. E. (1961). Probability and the Logic of Rational Belief.
Wesleyan University Press, Middletown, CT.
8. Kyburg, Jr., H. E. and Teng, C. M. (2001). Uncertain Inference. Cam-
bridge University Press, Cambridge.
9. Kyburg, Jr., H. E., Teng, C. M., and Wheeler, G. (2007). Conditionals
and consequences. Journal of Applied Logic, 5(4):638-650.
23
10. Levi, I. (1980). The Enterprise of Knowledge. MIT Press, Cambridge,
MA.
11. Levi, I. (2007). Probability logic and logical probability. In Harper, W.
and Wheeler, G., editors, Probability and Inference: Essays in Honor of
Henry E. Kyburg, Jr., pages 255-266. College Publications.
12. Morgan, C. G. (2000). The nature of nonmonotonic reasoning. Minds and
Machines, 10:321-360.
13. Ramsey, F. P. (1926). Truth and probability. In Kyburg, H. E. and
Smokler, H. E., editors, Studies in subjective probability, pages 23-52.
Robert E. Krieger Publishing Company, Huntington, New York, second
(1980) edition.
14. Wheeler, G. and Williamson, J. (2009). Evidential probability and ob-
jective bayesian epistemology. In Bandyopadhyay, P. and Forster, M.,
editors, Handbook of the Philosophy of Statistics. Elsevier Science.
3.2.19 Logics of Empirical Sciences
Decio Krause
Federal University of Santa Catarina - Brazil
[email protected]
In this tutorial, we present a general discussion on the relevance of the logi-
cal analysis of empirical theories, which is identied with the axiomatization of
the relevant theories, giving special emphasis to physics. The concept of Sup-
pes Predicate is introduced and some case studies are presented. The role of
the background set theory used for dening the predicate is discussed. In the
last part, a particular topic involving quantum mechanics plus a metaphysics of
non-individual entities is discussed.
References:
1. Browder, F. (ed.), Mathematical Problems Arising from Hilbert Problems,
Proceedings of Symposia in Pure Mathematics, Vol.XXVIII, Providence ,
American Mathematical Society, 1976.
2. French, S. and Krause, D., Identity in Physics: A Historical., Philosophi-
cal, and Formal Analysis. Oxford , Oxford Un. Press, 2006.
3. Krause, D., Introduction to the Axiomatic Foundations of Science (in
Portuguese), EPU, S.Paulo, 2002.
4. Magalh aes, J. C. M. and Krause, D., Suppes predicate for genetics and
natural selection Journal of Theoretical Biology 209 (2) 2001, pp. 141-
153.
24
5. Manin, Yu. I., Mathematical Problems, I: Foundations, in Browder 1976,
p. 36.
6. Manin, Yu. I., Georg Cantor and his heritage, arXiv:math.AG/0209244
v.1,19 Sept. 2002
7. Suppes, P., Representation and Invariance of Scientic Structures, Stan-
ford, CSLI Pu., 2002.
3.2.20 Logics of Plurality
Friederike Moltmann
CNRS/IHPST - France
[email protected]
The correct logical analysis of plural terms such as the trees in the trees
are similar or the trees are green is at the center of an important debate both
in formal semantics and in philosophical logic. Two fundamentally distinct
approaches can be distinguished, one on which the trees refers to a single collec-
tive entity, a plurality of trees, and one on which the trees refers plurally to the
various individual trees. The rst tradition is linked to the work of Link and re-
lated mereological approaches, the second to the work of Boolos and subsequent
work in that tradition (Oliver, Yi, Rayo and others). This course will give an
overview over the two kinds of approaches to the logical analysis of plural terms
with its various developments and discusses the crucial linguistic empirical and
conceptual motivations for the two kinds of approaches.
Session 1:
Reference to a plurality: The mereological approach
This session discusses the motivations and the development of the mereolog-
ical approach such as that of Link and others. It presents a range of potential
empirical and conceptual problems for that approach.
Session 2:
Plural Reference: The second-order approach
This session will discuss the seminal work of Boolos and subsequent devel-
opments such as the work of Oliver, Rayo, Yi. It focuses on the formal and
conceptual aspects of that approach.
Session 3:
This session discusses potential extensions of the second approach, such as to
to mass terms like courage, as in courage is admirable. It also discusses various
ramications of the plural reference approach and the challenges it faces from
the point of view of natural language.
3.2.21 How to Cut and Paste Logical Systems
Marcelo Coniglio
State University of Campinas - Brazil
25
[email protected]
Combination of logics is still a fairly young subject. It arose from the study
of some particular cases especially connected with modal logics, through the
techniques of fusion, product and bring. This was a rst stage of the develop-
ment of this new eld of research, and afterwards the initial techniques (mainly
bring) were generalized to general logics. The techniques for combining logics
help to the study of some fundamental phenomena of logic which are still not
properly understood, connected to what a logic is and what are the relations
between dierent formulations of a given logic. This tutorial, based on the book
[2], is mainly devoted to the study of the so-called categorial bring (or algebraic
bring), introduced in [1] and later on generalized to a wide class of logic systems
such as modal (rst-order) logics, higher-order logics and non-truth-functional
logics, among others (see [2]).
The main topics to be analyzed herein are the following:
1. Fibring syntactically: The Hilbert calculi case
2. Fibring semantically: Interpretation systems and their bring
3. Preservation results: Completenes and interpolation preservation by b-
ring
4. Heterogeneous bring: Combining abstract proof systems
5. One step ahead: Fibring rst-order (modal) logics
6. Still more generality: Fibring higher-order (modal) logics and non-truth-
functional logics
7. The future: Graph-theoretic bring
Bibliography
1. A. Sernadas, C. Sernadas and C. Caleiro. Fibring of logics as a categorial
construction. Journal of Logic and Computation, 9(2):149179, 1999.
2. W. Carnielli, M. Coniglio, D. Gabbay, P. Gouveia and C. Sernadas. Anal-
ysis and Synthesis of Logics: How to Cut and Paste Reasoning Systems.
Volume 35 of Applied Logic Series, Springer, 2008.
3. A. Sernadas, C. Sernadas, J. Rasga and M. Coniglio. On Graph-theoretic
Fibring of Logics. Journal of Logic and Computation, to appear.
26
4 Congress on Universal Logic III
4.1 Invited Speakers
Hartry Field
New York University - USA
[email protected]
Is there a problem about revising logic?
How, if at all, can one rationally change which logic one employs? Ill reply
both to those who think we cant and to those who minimize the problems of
doing so, and suggest a model for how such rational change might occur. Ill
connect the discussion with a real example: the issues surrounding what I take
to be a serious though not incontrovertible case for logical revision.
George Gr atzer
University of Manitoba - Canada
[email protected]
A Combinatorial Problem in Lattice Theory
The congruences of a nite lattice L form a nite distributive lattice. Given a
nite distributive lattice D, we can represent D as the congruence lattice of a
nite lattice L. If D has n join-irreducible elements, how small we can make
L as a function of n? The classical result of Dilworth gives an exponential up-
per bound. This was improved by several authors. The best result is Gratzer,
Lakser, Schmidt (1995): O(n2). Gratzer, Rival, Zaguia (1998) proved that this
result is best possible. While there are many results about upper bounds, this
was the only one providing a lower bound. In Gratzer, Lakser, Zaguia (1995),
we proved that a nite distributive lattice D with n join-irreducible elements
can be represented as the congruence lattice of a nite semimodular lattice L of
size O(n3). In my book, Congruence Lattices of Finite Lattices (2006), I raise
the question whether O(n3) is best possible. A lattice L is rectangular if it
is nite, semimodular, planar, has a left corner a (the only doubly irreducible
element on the left boundary), has a right corner b, and the elements a and b
are complementary. In a series of four papers with E.Knapp (2007-2010), we
prove that a nite distributive lattice D with n join-irreducible elements can be
represented as the congruence lattice of a rectangular lattice L of size O(n3)
and this result is best possible. The lower bound is kn3, where k = 1/3456.
The problem in my book remains unresolved.
References:
1. G. Gratzer, The Congruences of a Finite Lattice, A Proof-by-Picture Ap-
proach, Birkh auser Boston, 2006.
2. G. Gratzer, E.Knap, Notes on planar semimodular lattices. I. Construc-
tion, Acta Sci. Math. (Szeged) 73 (2007), 445-462.
3. G. Gratzer, E.Knap, Notes on planar semimodular lattices. II. Congru-
ences, Acta Sci. Math. (Szeged) 74 (2008), 23-36.
27
4. G. Gratzer, E.Knap, Notes on planar semimodular lattices. III. Con-
gruences of rectangular lattices, Acta Sci. Math. (Szeged) 75 (2009),
accepted for publication.
5. G. Gratzer, E.Knap, Notes on planar semimodular lattices. IV. The size
of a minimal congruence lattice representation with rectangular lattices,
Acta Sci. Math. (Szeged), 76 (2010), accepted for publication.
6. G. Gratzer, H.Lakser, E.T.Schmidt, Congruence lattices of small planar
lattices, Proc. Amer. Math. Soc. 123 (1995), 2619-2623.
7. G. Gratzer, I.Rival, N.Zaguia, Small representations of nite distributive
lattices as congruence lattices, Proc. Amer. Math. Soc. 123 (1995),
1959-1961.
Yuri Gurevich
Microsoft Research, USA
[email protected]
Algebra and Logic: Pitfalls and Potential Benets
One danger in algebra and logic is over-abstraction. You rise to a raried air
with little substance and no good theorems. Another potential problem is that
it is easy to formulate questions that are mathematically precise but uninterest-
ing. In applications though the greater problem is under-abstraction. Whether
you program or prove, it is all too easy to get bogged down with details so that
you cant see the forest for the trees. How to get the level of abstraction right?
That is where algebra and logic are indispensable, and that is the issue that we
intend to dwell upon.
Gerhard J ager
University of Bern - Switzerland
[email protected]
About the Suslin Operator in Applicative Theories
In the seventies Feferman developed his so-called explicit mathematics as a
natural formal framework for Bishop-style constructive mathematics. Explicit
mathematics is strongly inuenced by generalized recursion theory and soon
turned out to be of independent proof-theoretic interest. Since the operations
of explicit mathematics can be regarded as abstract computations, functionals
of higher types can be added in a direct and perspicuous way.
A rst important step in the proof-theoretic treatment of functionals of
higher types in the framework of explicit mathematics was the analysis of the
non-constructive minimum operator over a basic theory BON of operations and
numbers. A further interesting type two functional is the Suslin operator which
tests for well-foundedness of total binary relations.
In this talk I take up the proof-theoretic analysis of the Suslin operator in
explicit mathematics due to Strahm and myself, but present a new and concep-
tionally preferable approach. This is joint work with Dieter Probst.
28
Marcus Kracht
University of California at Los Angeles - USA
[email protected]
Judgement
The logical literature is lled with signs of judgement (typically the turnstile
and all its graphical variants). Yet a discussion of their role in logic is typically
absent. In this talk I want to rectify this imbalance by focussing on the nature
judgement. I will show that certain logics can be motivated by the character of
judgement alone. This provides a way to reconcile logical monism with logical
pluralism. For we may maintain that there is just one objective logic while there
are many subjective logics, each based on a dierent notion of judgement.
Hiroakira Ono
Japan Advanced Institute of Science and Technology - Japan
[email protected]
Regular embeddings of residuated lattices and innite distributivity
Recently there have been remarkable developments of the study of completions
of residuated lattices, in particular of canonical extensions and MacNeille com-
pletions. Embeddings associated with MacNeille completions are always regu-
lar, which means that all existing innite joins and meets are preserved, while
embeddings associated with canonical extensions are never so. Meanwhile Mac-
Neille completions do not always preserve distributivity. Here, we will consider
completions of residuated lattices with regular embeddings which preserve (in-
nite) distributivity, since such completions would be quite useful in proving
algebraic completeness of distributive substructural predicated logics. It will be
shown that the join innite distributivity (JID) will play a particularly impor-
tant role.
Giovanni Sambin
University of Padova - Italy
[email protected]
29
Jonathan Seldin
University of Lethbridge - Canada
[email protected]
Logical Algebras as Formal Systems: H. B. Currys Approach to Al-
gebraic Logic
Nowadays, the usual approach to algebras in mathematics, including algebras
of logic, is to postulate a set of objects with operations and relations on them
which satisfy certain postulates. With this approach, one uses the general prin-
ciples of logic in writing proofs, and one assumes the general properties of sets
from set theory. This was not the approach taken by H. B. Curry in [1] and
[2], Chapter 4. He took algebras to be formal systems of a certain kind, and
he did not assume either set theory or the rules of logic. I have not seen this
approach followed by anybody else. The purpose of this paper is to explain
Currys approach.
References:
1. H. B. Curry, Leons de Logique Algbrique, Paris: Gauthier-Villars and
Louvain: Nauwelaerts, 1952.
2. H. B. Curry, Foundations of Mathematical Logic, New York: McGraw-
Hill, 1963. Reprinted by Dover, 1977 and several times since.
Amilcar Sernadas
Technical University of Lisbon - Portugal
[email protected]
Parallel Compositon of Logics
The practical signicance of the problem of combining logics is widely recog-
nized, namely in knowledge representation (within articial intelligence) and
in formal specication and verication of algorithms and protocols (within soft-
ware engineering and information security). In these elds, the need for working
with several calculi at the same time is the rule rather than the exception. The
topic is also of interest on purely theoretical grounds. For instance, one might be
tempted to look at predicate temporal logic as resulting from the combination
of rst-order logic and propositional temporal logic. However, the approach
will be signicant only if general preservation results are available about the
combination mechanism at hand, namely preservation of completeness. For
these reasons, dierent forms of combining logics have been studied and several
such transference results have been reported in the literature. To name just a
few, fusion (of modal logics), temporalization and bring are now well under-
stood, although some interesting open problems remain concerning transference
results. Fibring [1] is the most general form of combination and its recent
graphic-theoretic account makes it applicable to a wide class of logics, including
substructural and non truth-functional logics. Capitalizing on these latest de-
velopments on the semantics of bring [2] and inspired by the notion of parallel
composition of processes in its most basic form (interleaving), a novel form of
combination of logics, applicable to a wide class of logics, is proposed together
30
with a couple of transference results, and compared with other combination
mechanisms, showing how they can be recovered as special cases.
References:
1. D. Gabbay. Fibred semantics and the weaving of logics: part 1. Journal
of Symbolic Logic, 61(4):10571120, 1996.
2. A. Sernadas, C. Sernadas, J. Rasga, and M. Coniglio. On graph-theoretic
bring of logics. Journal of Logic and Computation, 19:13211357, 2009.
Dana Scott
University Professor Emeritus, Carnegie Mellon University and
Visiting Scholar, University of California, Berkeley, USA
[email protected]
Mixing modality and probability
For some time at many recent workshops, the author has lectured about a
Boolean-valued model for higher-order logic (and set theory) based on the com-
plete Boolean algebra of Lebsegue measurable subsets of the unit interval mod-
ulo sets of measure zero. This algebra not only carries a probability measure,
but it also allows for a non-trivial S4-modality by using the proper subframe of
open sets modulo zero sets. This provides rich ingredients for building many
kinds of structures having non- standard random elements. The lecture will
review the basics of this type of semantics and discuss several examples and
their logical properties.
SECRET SPEAKER
University of ???
XYZ@???
???
31
4.2 Sessions
4.2.1 Logic Diagrams
This session is organized by Amirouche Mokte (IRIST, Strasbourg and LHPS,
Nancy, France) and Sun-joo Shin (Yale University, USA).
The use of diagrams in logic is old but unequal. The nineteenth century is
often said to be the golden age of diagrammatic logic, thanks to the wide use
of Euler and Venn diagrams, before a period of decline with (because of?) the
arrival of the Frege Russell tradition of mathematical logic.
One more obstacle that prevented the use of diagrams is the plurality of
logical systems and the diculty to deal with them.
However, diagram studies have known a revival in recent years. It is thus
legitimate to wonder what place diagrams hold in modern logical theory and
practice.
The aim of this session is to discuss the logical status of diagrams. Topics
may include:
What is a logic diagram?
Is diagrammatic logic one more logic?
Or are diagrams merely a notation that one can adapt to t to dierent
logics?
Do diagrams t to some logical systems better than to others?
Is there still room for the use of diagrams in modern logic?
Accepted contributed talks
Juliusz Doboszewski
Andrzej J. Nowak
Jagiellonian University - Poland
Nonlinear orthography or nonlinear reasoning: Freges Begrisschrift as seen
against Peirces existential graphs
Valeria Giardino
Institut Jean Nicod (CNRS-EHESS-ENS) - France
[email protected]
A Cognitive Approach to Diagrams in Logic
The term logocentric was chosen to dene the dogma of the standard view of
mathematics, according to which proofs are syntactic objects consisting only
of sentences arranged in a nite and inspectable way. By contrast, reasoning
is a heterogeneous activity: it is necessary to expand the territory of logic by
freeing it from a mode of representation only (Barwise and Etchemendy (1996),
Shin (2004)). In this talk, I will argue that the antidote against the logocentric
approach to diagrams consists neither in nding the right set of rules nor in
32
assuming an opticentric view, but in considering that the most relevant aspect
of diagrammatic reasoning is the way in which diagrams are manipulated to
infer some new conclusion from them; this happens in continuous interaction
with language. My view moves from a purely syntactic approach to a semantic
and indeed pragmatic approach to problem solving.
John Howse
Peter Rodgers
Gem Stapleton
Universities of Brighton and Kent - UK
Levels of Syntax for Euler Diagram Logics
Catherine Legg
[email protected]
University of Waikato - New Zealand
The Hardness of the Iconic Must: Can Peirces Existential Graphs Assist Modal
Epistemology?
The current of development in 20th century logic bypassed Peirces existential
graphs, but recently much good work has been done by formal logicians excavat-
ing the graphs from Peirces manuscripts, regularizing them and demonstrating
the soundness and completeness of the alpha and beta systems (e.g. Roberts
1973, Hammer 1998, Shin 2002). However, given that Peirce himself considered
the graphs to be his chef doeuvre in logic, and explored the distinction between
icons, indices and symbols in detail within the context of a much larger theory of
signs, much about the graphs arguably remains to be thought through from the
perspective of philosophical logic. For instance, are the graphs always merely
of heuristic value or can they convey an essential icon (analogous to the now
standardly accepted essential indexical)? This paper claims they can and do,
and suggests important consequences follow from this for the epistemology of
modality. It is boldly suggested that structural articulation, which is character-
istic of icons alone, is the source of all necessity. In other words, recognizing a
statement as necessarily true consists only in an unavoidable recognition that
a structure has the particular structure that it in fact has. (What else could it
consist in?)
Danielle Macbeth
Haverford College - USA
[email protected]
Reasoning in Diagrams: The Case of Begrisschrift
Frege designed his strange two-dimensional Begrisschrift notation as a system
of written signs within which to exhibit the contents of concepts, as those con-
tents matter to inference, and to reason in mathematics. Frege furthermore
claimed that reasoning from denitions in his language can be ampliative, a
real extension of our knowledge, and in particular, that his Begrisschrift proof
of theorem 133 is ampliative (so synthetic in Kants sense) despite being strictly
deductive (or, as Kant would think of it, analytic). The proof is, in other words,
33
constructive in something like Kants sense, despite being strictly deductive. But
if it is constructive then the notation within which the construction is made,
that is, Freges notation, must be functioning diagrammatically in something
like Kants sense (which encompasses not only Euclidean diagrams but also the
symbolic language of arithmetic and algebra). I will explain what is required of
such a notation, and show that Freges notation can be read in just this way, as a
mathematical language, a system of written signs, within which to reason from
dened concepts in mathematics. So read, Freges work, and the language he
developed within which to do it, belongs not to the tradition of mathematical
logic begun by Boole and mostly developed after Frege, but instead within the
twenty-ve hundred year long tradition of constructive paper-and-pencil math-
ematical reasoning that came before him.
Koji Mineshima
[email protected]
Mitsuhiro Okada
Ryo Takemura
[email protected]
Keio University - Japan
Reasoning with Euler diagrams: a proof-theoretical approach
This talk is concerned with a proof-theoretical investigation of Euler diagram-
matic reasoning. We introduce a novel approach to the formalization of reason-
ing with Euler diagrams, in which diagrams are dened not in terms of regions
as in the standard approach, but in terms of topological relations between dia-
grammatic objects. On this topological-relation-based approach, the unication
rule, which plays a central role in Euler diagrammatic reasoning, can be formal-
ized in a style of Gentzens natural deduction. We prove the soundness and
completeness theorems of our Euler diagrammatic inference system. We then
investigate structure of diagrammatic proofs and prove a normalization theorem.
Finally, we discuss some cognitive properties of Euler diagrammatic reasoning
in our system, in comparison to linguistic reasoning and reasoning with Venn
diagrams.
Jorgen Fischer Nilsson
Technical University of Denmark - Denmark
[email protected]
Diagrammatic Reasoning with Class Relationship Logic
We discuss diagrammatic visualization and reasoning for a class relationship
logic accomplished by extending Euler diagrams using higraphs. The considered
class relationship logic is inspired by contemporary studies of logical relations in
biomedical ontologies, and it appeals to the Closed World Assumption (CWA)
unlike e.g. Description logic. The suggested diagrams provide inference prin-
ciples inherent in the visual formalism. The considered logical forms are dealt
with at the metalogic level by variables ranging over classes and relations. There
are inference rules being formalized at the meta-level using denite clauses with-
out compound terms (Datalog).
34
Leo Olsen
Philander Smith College - USA
[email protected]
Sentential Modal Logic and the Gamma Graphs
The system of Existential Graphs is a diagrammatic system of logic that was
developed by Charles S. Peirce. This system has three main sections and Peirce
named these three sections: Alpha, Beta, and Gamma. The Alpha section of
EG is a diagrammatic account of sentential logic; the Beta section (which is
an extension of Alpha) is a diagrammatic account of rst order predicate logic
with identity; and the Gamma section (which is an extension of Beta) gives a
combined diagrammatic account of higher order predicate logic and modal logic.
In my paper I give brief account of Alpha. Once this has been done, I extend
Alpha to include some of the rules of Gamma (as presented in Peirces 1903
version of Gamma), so that a system of sentential modal logic results. After
investigating some of the properties of Gamma, I compare them to some of the
properties of contemporary accounts of sentential modal logic.
Ahti-Veikko Pietarinen
University of Helsinki - Finland
[email protected]
Tableaux for the Gammas
Tableaux methods are conveniently applied to existential graphs, because tableaux
interpret the assertions in the endoporeutic fashion. Since graphs can be read
in a number of logically equivalent ways, the number of tree rules is kept in
a minimum (negative juxtapositions go to dierent branches, positive ones to
the same branch). Already in 1885, Peirce had suggested tableaux for propo-
sitional logic. He never applied the idea to the EGs. But his semantic rules
for EGs are equivalent to those of the game-theoretic semantics, and thus are
naturally amenable also to semantic tableaux systems. The gamma part of EGs
was Peirces boutique of modal and higher-order logics, metagraphs, and many
others. I will dene semantic tableaux-type proofs for the modal gammas, and
propose such transformation rules for the broken-cut gamma that yield better
correspondences with modern characterisations of modal logical systems.
Denis I. Saveliev
Moscow State University of Russian Academy of Sciences - Russia
[email protected]
Diagrams in the Membership Game
There is a natural way to represent sets by certain diagrams; it has grown
customary in modern studies of ill-founded sets. We use this way to study
a set-theoretic game, the membership game. This game, introduced probably
by T.Forster in [1] (see also [2]) and originally related to NF-like set theories,
is a perfect information game of two persons played on a given set along its
membership: The players choose in turn an element of the set, an element of
this element, etc.; a player wins if its adversary cannot make any following move,
i.e. if he could choose the empty set. Sets that are winning, i.e. have a winning
35
strategy for some of the players, form an ordinal hierarchy easily visualized by
diagrams.
We show that all levels of the hierarchy are nonempty and the class of hered-
itarily winning sets is a full model of set theory containing all well-founded sets.
Then we show that each of four possible relationships between the universe, the
class of hereditarily winning sets, and the class of well-founded sets is consistent.
For consistency results, we propose a new method to get models with ill-founded
sets. Its main feature is that such models are stratied like the cumulative hier-
archy and reect certain formulae at own lowest layers. Thus to know properties
of whole models it suces to observe diagrams of their lowest layers, usually
very simple objects. We apply this method to establish various ne results,
some of which display a deep dierence between odd- and even-winning sets.
Our results are proved in a weak set theory, ZF minus both choice and reg-
ularity axioms. Although they can be established without using of diagrams,
diagrams make the constructions much more clear and easily observable. These
results were announced in [3] and appeared with detailed proofs in [4].
References
1. FORSTER, Thomas E. Axiomatising Set Theory with a Universal Set.
In: La Theorie des ensembles de Quine. Cahiers du Centre de Logique 4,
CABAY Louvain-La-neuve, 1983.
2. FORSTER, Thomas E. Games Played on an Illfounded Membership Rela-
tion. In: A Tribute to Maurice Boa. Eds: Crabbe, Point, and Michaux.
(Supplement to the December 2001 number of the Bulletin of the Belgian
Mathematical Society.
3. SAVELIEV, Denis I. A report on a game on the universe of sets. Mathe-
matical Notes, vol. 81, no. 5-6 (2007), pp. 716-719.
4. SAVELIEV, Denis I. A game on the universe of sets. Izvestiya: Mathe-
matics (Proc. Russ. Acad. Sci., Ser. Math.), vol. 72, no. 3 (2008), pp.
581-625.
4.2.2 Non-Classical Mathematics
The 20th century has witnessed several attempts to build parts of mathematics
on grounds other than those provided by classical logic. The original intuitionist
renderings of set theory, arithmetic, analysis, etc. were later accompanied by
those based on relevant, paraconsistent, contraction-free, modal, and other non-
classical logics. The subject studying such theories can be called non-classical
mathematics, i.e., the study of any part of mathematics that is, or can in princi-
ple be, formalized in some logic other than classical. This special session at the
2010 UniLog World Congress is a follow-up to the conference on Non-Classical
Mathematics that was held in Hejnice, Czech Republic, in June 2009.
The scope of interest of this special session contains, but is not limited to
the following topics:
36
* Intuitionistic, constructive, and predicative mathematics: Heyting arith-
metic, intuitionistic set theory, topos-theoretical foundations of mathematics,
constructive or predicative set and type theories, pointfree topology, etc.
* Substructural and fuzzy mathematics: relevant arithmetic, contraction-
free nave set theories, axiomatic fuzzy set theories, fuzzy arithmetic, etc.
* Inconsistent mathematics: calculi of innitesimals, inconsistent set theo-
ries, etc.
* Modal mathematics: arithmetic or set theory with epistemic, alethic, or
other modalities, modal comprehension principles, modal treatments of vague
objects, modal structuralism, etc.
* Alternative classical mathematics: alternative foundational theories over
classical logic, non-standard analysis, etc.
* Topics related to non-classical mathematics: metamathematics of non-
classical or alternative mathematical theories, their relative interpretability, etc.
Non-Classical Mathematics
This session is organized by Libor Behounek and Petr Cintula from the Czech
Academy of Sciences.
Accepted contributed talks
Arnon Avron
Tel-Aviv University - Israel
[email protected]
A New Approach to Predicative Set Theory
The Poincare-Weyl-Feferman predicativist program for the foundations of math-
ematics seeks to establish certainty in mathematics without necessarily revolu-
tionizing it (as the intuitionistic program does). The main goal of this paper
is to suggest a new framework for this program by constructing an absolutely
reliable predicative pure set theory PZF whose language is type-free, and (from
a platonic point of view) the universe ZF is a model of it. Our basic idea is that
the predicatively acceptable instances of the comprehension schema are those
which determine the collections they dene in an absolute way, independent of
the extension of the surrounding universe. This idea is implemented using a
syntactic safety relation between formulas and sets of variables. This relation is
obtained as a common generalization of syntactic approximations of the notion
of domain-independence used in database theory, and syntactic approximations
of Godels notion of absoluteness used in set theory. Two important features of
our framework is that it makes an extensive use of abstraction terms, and its
underlying language is that of ancestral logic (which is strictly stronger than
rst-order languages, but much weaker then full second-order languages). An-
other crucial feature is that it is possible to use it together with classical logic,
but it equally makes sense to use it in combination with some non-classical logic,
especially intuitionistic logic.
Libor Behounek
[email protected]
Petr Cintula
37
[email protected]
Academy of Sciences of the Czech Republic
Features of Mathematical Theories in Contraction-Free Logics
First-order contraction-free substructural logics are strong enough to support
non-trivial mathematical theories. However, the absence of the law of contrac-
tion makes certain practices that are commonly used in classical mathematics
meaningless. This talk summarizes the most important dierences between
classical and contraction-free mathematics and suggests several guidelines for
developing contraction-free theories.
In particular, in contraction-free mathematics, dened notions need be pa-
rameterized by multiplicities of subformulae of the dening formula; free combi-
nations of subconditions replace conjunctive compound notions; non-contractive
preconditions need occur as premises of theorems (with variable multiplicities)
rather than in denitions; non-contractive subsets can only be viewed as an
additional structure rather than universes for substructures; and instead of
equivalences, theorems indicating mutual bounds for truth-values are regularly
obtained in contraction-free mathematics.
Some of these features were rst observed in our paper Features of math-
ematical theories in formal fuzzy logic (LNAI 4529:523-532, 2007), but they
actually apply to all contraction-free substructural mathematics. Further fea-
tures, such as the splitting of classically equivalent notions into several variants,
are common to all branches of non-classical mathematics.
Frode Bjordal
University of Oslo - Norway
[email protected]
Minimalistic Liberalism a contradictory semi-formal set theory respecting clas-
sical logic, and with mathematical power beyond PI(1, 1) CA(0)
We present the semantics and some salient partial axiomatic and inferential
principles for a semi formal set, or predication theoretic framework which we
call Minimalistic Liberalism (ML). ML deals with paradoxes in a way which is
akin to paraconsistent approaches, though diers in that classical logical princi-
ples are always theorems and their negations always fail to be theorems, even in
the presence of what is called liberal comprehension. E.g., if R is Russells set,
it will be a theorem of ML that R is in R, and it will be a theorem in ML that R
is not in R, but it will not be a theorem of ML that R is in R and R is not in R.
So ML is non-adjunctive. One focus of the presentation will be upon explaining
the semantical framework, which relies upon an additional twist upon revision-
ary types of semantics in the tradition from Herzberger, Gupta and Belnap, and
to isolate some of the salient axiomatic and inferential principles which hold. It
is a fact that ML is strong enoug to interpret ACA. We show how we may make
use of a xed point construction going back to Cantini and Visser in order to
show that a set of hereditarily non-paradoxical iterative sets in ML interprets
the theory of nitely iterated inductive denitions ID < (omega which has the
same proof theoretic strenth as PI(1, 1) CA(0). So this is a lower limit for
the kind of proof theoretic strength at stake.
38
Piotr Blaszczyk
Pedagogical University of Cracow - Poland
[email protected]
Nonstandard Analysis in action. Zeno of Elea and his modern rivals revisited
In my talk I consider The Achilles Paradox with regard to the structure of
continuum. To compare Zenos arguments with the modern solutions to the
paradox I suppose that Zenos space is that of Euclid. As a result, provided that
(F, +, ., 0, 1, <) is a real closed eld, F F F is a model of Zenos space, so it
does not have to be continuous (in Dedekind sense). In the classical resolution
to the paradox (Ajdukiewicz, Grnbaum) it is supposed that Zenos arguments
can be represented in the arithmetic of reals and R R R is the only model
of Zenos space. There is an implicit supposition both in Achilles paradox it-
self and its classical resolution: to overtake the Tortoise, Achilles must rst
reach the point where he catches up with the Tortoise. I show that Achilles
can overtake the Tortoise and there may be no point where he catches up with
the Tortoise. To this end an example is given: a continuous map dened on
the set of hyperreals that does not has the intermediate value property. Next, I
provide such a model (a hypernite time line that is a subset of hyperreal line)
that Achilles overtakes the Tortoise, there is no point where he catches up with
it, there is the last point where the Tortoise is ahead of Achilles, and the rst
point where Achilles is ahead of the Tortoise.
Ross T. Brady
La Trobe University - Australia
[email protected]
The Simple Consistency of Arithmetic, for Metacomplete Logics without a Quan-
tied Form of Distribution
We rst prove the simple consistency of arithmetic by nitary methods, where
the arithmetic is based on a logic MC of meaning containment. The essen-
tial dierence between MC and classical logic is that it is conceptualized in
terms of meanings rather than truth and falsity. This idea will permeate the
axiomatization of the logic, and its quanticational and arithmetic extensions.
The sentential logic MC is a weak relevant logic containing neither of the key
classical principles: the Law of Excluded Middle (LEM) and the Disjunctive
Syllogism (DS). It also does not include the sentential distribution properties.
Further, the quanticational extension MCQ does not include the corresponding
two distribution properties even in their rule forms. The reasons for dropping
all these principles is that conjunction and disjunction are extensional whilst
the entailment and the quantiers are intensional, the latter possibly applying
to a non-recursive property. We then set up the axiomatization of arithmetic
capturing the spirit of Peanos axioms in the form of rules. This is because
the entailment is an inappropriate relationship between statements involving
distinct natural numbers. We do, however, take all the identity statements as
classical, i.e. the LEM and the DS both apply to them. However, due to con-
sistency, we can add the full DS admissibly. Since metavaluations are used for
39
modelling purposes, this result will extend to metacomplete logics. Due to the
recursiveness in the proof, we can add back in the existential distribution rule.
Lastly, we will examine the development of arithmetic in Mendelsons Introduc-
tion to Mathematical Logic [1964] to compare what our system can and cannot
do in relation to the classical Peano arithmetic.
Michal Hol capek
University of Ostrava - Czech Republic
[email protected]
Fuzzy Sets and Fuzzy Classes in Universes of Sets
Traditionally, the concept of fuzzy sets is closely related to a x set of interesting
objects called the universe of discourse. This restriction, however, seems
to have several disadvantages. We can recognize a natural need to deal with
fuzzy sets over dierent universes as in the case of fuzzy sets of fresh apples
in the rst basket and fresh pears in the second basket, where the sets of
all apples and pears in the baskets are the corresponding universes. Further,
the presumption of a x set as a universe for fuzzy sets has some fuzzy sets
construction limitation. Practically, a fuzzy set theory cannot be introduced on
a x universe. An analogical disadvantage was also recognized by S. Gottwald
and, therefore, he proposed a cumulative system of fuzzy sets [1].
In the presentation, we will introduce a universe of sets over which fuzzy sets
are dened. The denition is based on the axioms of Grothendieck universe, i.e.,
on a set in which the whole set theory may be formed (see e.g. [2]), where an
axiom ensuring the existence of fuzzy sets with membership degrees interpreted
in a complete residuated lattice is added. Some of the examples and properties
of the universe of sets will be demonstrated. Further, we will establish the
concept of fuzzy set in a universe of sets and show several constructions of fuzzy
objects and fuzzy relations that are well known in the fuzzy set theory. Finally,
we will dene the informal but very useful notion of fuzzy class in a universe of
sets which generalizes the concept of fuzzy set. Some properties of fuzzy classes
will be also presented.
1. S. Gottwald. Set theory for fuzzy sets of higher level. Fuzzy Sets and
Systems , 2: 125 151, 1979.
2. C. McLarty. Elementary Categories, Elementary Toposes. Oxford Uni-
versity Press, 1992.
Valery Khakhanian
Moscow State University of Railway Communications - Russia
[email protected]
Intuitionistic Approach for Justication of a Set Theory
The crisis in foundation of mathematics in the end of XIX - beginning of XX
centuries initiated a number of axiomatic set theoretical systems during the rst
half of XX century. These systems were the result of dierent philosophical ap-
proaches (in view of second Godels Theorem) which were aimed to overcome
the above crisis. But the way out of this situation has never been found. In
40
my report I will try to give a new approach to solve this problem using a basic
axiomatic system of a set theory with intuitionistic logic. I will present many
mathematical results which were obtained during last thirty years. We will
survey more than thirty years development of the set theory with intuitionistic
logic underlining main points and formulating unsolved problems and describe
the basic system of intuitionistic set theory.
Vilem Novak
University of Ostrava - Czech Republic
[email protected]
Many-valued (Fuzzy) Type Theories
Mathematical fuzzy logic is a well established formal tool for modeling of human
reasoning aected by the vagueness phenomenon and captured via degree theo-
retical approach. Besides various kinds of propositional and rst-order calculi,
also higher-order fuzzy logic calculi have been developed that are in analogy
with classical logic called fuzzy type theories (FTT). These are generalization
of classical type theory presented, e.g., in [1]. The generalization consists es-
pecially in replacement of the axiom stating there are just two truth values by
a sequence of axioms characterizing structure of the algebra of truth values.
The truth values form either an IMTL-algebra (a prelinear residuated lattices
with double negation) or an EQ-algebra in which the main operation is a fuzzy
equality equivalence) and, unlike residuated lattices, implication is a derived
operation. The syntax of FTT is a generalization of the lambda-calculus con-
structed in a classical way, but diering from the classical one by denition of
additional special connectives, and by logical axioms. The fundamental connec-
tive in FTT is that of a fuzzy equality interpreted by a reexive, symmetric and
weakly transitive binary fuzzy relation. This paper provides an overview of the
main calculi of FTT.
1. P. Andrews, An Introduction to Mathematical Logic and Type Theory:
To Truth Through Proof, Kluwer, Dordrecht, 2002
2. V. Novak, On fuzzy type theory, Fuzzy Sets and Systems 149 (2005) 235
273.
3. V. Novak, EQ-algebra-based fuzzy type theory and its extensions, Logic
Journal of the IGPL (to appear).
Graham Priest
Universities of Melbourne (Australia) and St Andrews (UK), and
the Graduate Center, City University of New York (USA)
[email protected]
Mathematical Pluralism
There is a plurality of mathematical investigations. These cannot all be re-
duced to proofs within the framework of Zermelo-Fraenkel set theory, if only
because some of them use non-classical logic (such as the various branches of
41
intuitionist mathematics). How is one to understand this situation? In this
paper, I suggest that one should see this plurality as analogous the plurality
of games, any of which may be played. Various objections are considered and
rejected, including the charge that the picture engenders a pernicious relativism.
Itala M.L. DOttaviano
State University of Campinas - Brazil
[email protected]
Tadeu Fernandes de Carvalho
Pontifical University of Campinas - Brazil
[email protected]
Relations between the classical dierential calculus and da Costas paraconsis-
tent dierential calculus
Da Costa(2000) introduces a paraconsistent dierential calculus, whose under-
lying logic and set theory are, respectively, his known paraconsistent predicate
calculus C=1 and his paraconsistent set theory CHU1, introduced in 1986; two
special algebraic structures are constructed, the hyper-ring A and the quasi-
ring A*, which extend the eld of the real numbers and whose elements are
called hyper-real numbers. From A*, da Costa proposes the construction of
a paraconsistent dierential calculus, whose language is the language of C=1,
extended to the language of CHU1, in which we deal with the elements of A*.
Carvalho (2004), by using the paraconsistent apparatus, studies and improves
the calculus proposed by da Costa, obtaining extensions of several fundamental
theorems of the classical dierential calculus. In this paper, we introduce the
concepts of paraconsistent super-structure and monomorphism between para-
consistent super-structures. By considering the paraconsistent super-structure
R constructed over the set of standard real numbers, and mixing the structures
A and A* as basic sets for the construction of extensions of R, we prove a
Transference Theorem, that translates the classical dierential calculus into da
Costas paraconsistent calculus.
References
1. CARVALHO, T.F. On da Costas paraconsistent dierential calculus, 2004.
200p. (Doctorate Thesis). Campinas, State University of Campinas, 2004.
2. Da COSTA, N.C.A. Paraconsistent Mathematics. In: I WORLDCONGRESS
ON PARACONSISTENCY, 1998, Ghent, Belgium. Frontiers in paracon-
sistent logic: proceedings. Edited by D. Batens, C. Mortensen, G. Priest,
J.P. van Bendegen. London: Kings College Publications, 2000, p. 165-179.
4.2.3 Abstract Algebraic Logic
This session is organized by Josep Maria Font and Ramon Jansana from the
University of Barcelona, Spain.
This discipline can be described as Algebraic Logic for the XXIst century. It
gathers all mathematical studies of the process of algebraization of logic in its
42
most abstract and general aspects. In particular it provides frameworks where
statements such as A logic satises (some form of) the interpolation theorem
if and only if the class of its algebraic counterparts satises (some form of)
amalgamation become meaningful; then one may be able to prove them in
total generality, or one may investigate their scope, or prove them after adding
some restrictions, etc.
The term appeared for the rst time in Volume II of Henkin-Monk- Tarskis
Cylindric Algebras, referring to the algebraization of rst-order logics, but
after the Workshop on Abstract Algebraic Logic (Barcelona, 1997) it has been
adopted to denote all the ramications in the studies of sentential-like logics
that have ourished following Blok, Pigozzi and Czelakowskis pioneering works
in the 1980s. Abstract Algebraic Logic has been considered as the natural
evolution of the traditional works in Algebraic Logic in the style of Rasiowa,
Sikorski, Wjcicki, etc., and integrates the theory of logical matrices into a more
general framework.
The 2010 version of the Mathematics Subject Classication will incorporate
Abstract Algebraic Logic as entry 03G27, which witnesses the well-delimited,
qualitatively distinctive character of this discipline and its quantitative growth.
Topics that can t this Special Session include, but are not limited to, the
following ones:
1. Studies of the Leibniz hierarchy, the Frege hierarchy and their renements,
and relations between them.
2. Lattice-theoretic and category-theoretic approaches to representability and
equivalence of logical systems.
3. Use of algebraic tools to study aspects of the interplay between senten-
tial logics and Gentzen systems, hypersequent systems and other kinds of
calculi and logical formalisms
4. Formulation of abstract versions of well-known algebraic procedures such
as completions, representation theory and duality.
5. Study of the algebraization process for logics where order, besides equality,
is the main relation to be considered in the algebraic counterparts.
6. Extensions to other frameworks motivated by applications to computer
science, such as institutions, behavioural logics, secrecy logic, etc.
7. Study of algebra-based semantics of rst-order logics.
Accepted contributed talks
Felix Bou
IIIA - CSIC - Spain
[email protected]
Umberto Rivieccio
43
University of Genoa - Italy
[email protected]
Implicative Bilattices
Bilattices are algebraic structures introduced by Matt Ginsberg in the context of
A.I. In recent years, Arieli and Avron introduced a logic dened from matrices
called logical bilattices. In a previous work we studied, from the perspective
of Abstract Algebraic Logic, the implicationless fragment of Arieli and Avrons
logic. Here we complete this study considering the full system. We prove that
this logic is strongly (but not regularly) algebraizable and dene its equivalent
algebraic semantics through an equational presentation. We call the algebras in
this variety implicative bilattices. We obtain several results on this class, in
particular that it is a discriminator variety (hence arithmetical), generated by
a single nite algebra, and characterize its members as certain bilattice prod-
ucts of two copies of a generalized Boolean algebra. We also characterize some
subreducts of implicative bilattices that have a particular logical interest.
Carlos Caleiro
Technical University of Lisbon - Portugal
[email protected]
Ricardo Gonalves
Technical University of Lisbon - Portugal
[email protected]
Behavioral algebraization of logics (I)
The theory of Abstract Algebraic Logic (AAL) aims at drawing a strong bridge
between logic and algebra. It can be seen as a generalization of the well known
Lindenbaum-Tarski method. Although the enormous success of the theory we
can point out some drawbacks. An evident one is the inability of the theory
to deal with logics with a many-sorted language. Even if one restricts to the
study of propositional based logics, there are some logics that simply fall out of
the scope of this theory. One paradigmatic example is the case of the so-called
non-truth-functional logics that lack of congruence of some of its connectives,
a key ingredient in the algebraization process. The quest for a more general
framework to the deal with these kinds of logics is the subject of our work.
In this two-sessions talk we will present a generalization of AAL obtained by
substituting the role of unsorted equational logic with (many-sorted) behavioral
logic. The incorporation of behavioral reasoning in the algebraization process
will allow to amenably deal with connectives that are not congruent, while the
many sorted framework will allow to reect the many sorted character of a given
logic to its algebraic counterpart. In this rst part of the talk we focus on syn-
tactical issues, leaving the semantical issues for the second part of the talk. We
illustrate theses ideas by exploring some examples, namely, paraconsistent logic
C1 of da Costa.
Petr Cintula
Academy of Sciences of the Czech Rrepublic
[email protected] Carles Noguera
44
Artificial Intelligence Research Institute (IIIA - CSIC) - Spain
[email protected]
A general approach to non-classical rst-order logics
The goal of this talk is to present a general theory of rst-order non-classical
logics. Our approach generalizes the tradition of Rasiowas implicative logics,
Godel-Dummett rst-order logic, and H ajeks rst-order fuzzy logic, i.e. starting
from a propositional non-classical logic we add quantiers in the same way as in
rst-order intuitionistic logic. The unifying idea of this treatment of rst-order
logics can be formulated simply as:
the truth value of a universally (resp. existentially) quantied formula is the
inmum (resp. supremum) of all instances of that formula w.r.t. the existing
matrix order.
To do so, one needs a good notion of order in the semantics which is typ-
ically obtained from a suitable implication in the syntax, thus our underlying
propositional logics are the so called weakly p-implicational logics previously
studied by the authors this diers from other approaches where the order is
extralogical such as a recent paper by James Raftery). Given a propositional
logic L we present a rst-order Hilbert-style calculus extending it, and prove its
completeness w.r.t. the class of all rst-order structures based upon the matrix
semantics of L and hence, it turns out to be the minimal rst-order logic over L.
Having this suitable minimal logic we can study several of its extensions in order
to cope with important examples of variants of non-classical rst-order logics in
the literature. For instance, we nd a uniform axiomatization for logics com-
plete with respect to linearly ordered or witnessed semantics and characterize
logics enjoying Skolemization for their prenex fragment.
Janusz Czelakowski
Opole University - Poland
[email protected]
Structural Theorems on Congruence Modular Quasivarieties of Algebras
The focus of the talk is on applying the methods worked out by Abstract
Algebraic Logic (AAL) to the problem of nite axiomatizabilty of classes of
algebras. AAL oers here convenient tools based on the notion of a commu-
tator equation. This notion behaves pretty well in the context of relatively
congruence-distributive (RCD) quasivarieties of algebras and yields an elegant
proof of the theorem stating that every nitely generated RCD quasivariety is
nitely based. For relatively congruence modular (RCM) quasivarieties the anal-
ogous problem is much harder. In the talk a number of observations concerning
the structure of quasivarieties possessing the additive equational commutator is
presented. The class of quasivarieties with the additive equational commutator
encompasses RCM quasivarieties.
References
1. J. Czelakowski [2006] General theory of the commutator for deductive
45
systems. Part I. Basic facts, Studia Logica 83, 183-214.
2. K. Kearnes and R. McKenzie [1992] Commutator theory for relatively
modular quasivarieties, Transactions of the American Mathematical Soci-
ety 331, No. 2, 465 502.
3. D. Pigozzi
[1988] Finite basis theorems for relatively congruence-distributive quasi-
varieties, Transactions of
the American Mathematical Society 310, No. 2, 499-533.
Lisa Fulford
[email protected]
Alessandra Palmigiano
[email protected]
Universiteit van Amsterdam - The Netherlands
Modular canonicity for bi-implicative algebras
We will report on the canonicity of certain identities and inequalities in a sig-
nature consisting of constants , and implications , , which relates to
the canonicity of logics associated with certain distinguished sub-quasivarieties
of (bi-)implicative algebras, the best known of which are the varieties of (bi-
)Hilbert algebras and (bi-)Tarski algebras. These results are instances of a re-
search program connecting canonical extensions and Abstract Algebraic Logic
(see [?]). Previous results of this kind were obtained in [?]. Our basic setting is
the quasi-axiomatization of implicative algebras and expand it symmetrically
with a subtraction operator and a bottom. Within this setting we analyse the
independence and interdependence of certain axioms w.r.t canonicity, which
yields a better, more modular understanding of the canonicity of Hilbert alge-
bras.
Nikolaos Galatos
University of Denver - USA
[email protected] Jos Gil-Ferez
Japan Advanced Institute of Science and Technology - Japan
The Isomorphism Problem for modules over quantaloids, Part I
Given a structural consequence relation, the lattice of theories can be expanded
to include the action of substitutions (or equivalently inverse substitutions).
Blok and Pigozzi, in their monograph, showed that the expanded lattice of the-
ories of an algebraizable sentential logic is isomorphic to the expanded lattice
of theories of the corresponding algebraic consequence relation, and conversely
any such isomorphism comes from an algebraizable sentential logic. The result
was further extended from sentential logics to k-deductive systems. Blok and
Jonsson, realizing that the action of the monoid of substitutions to the set of
formulas plays a crucial role, developed a general framework, where one consid-
ers the action of an arbitrary monoid M on a set, yielding an M-set. Considering
46
logics on (equivalently, closure operators over) M-sets, one can easily prove that
every bidirectional syntactic translation between two such logics (a notion that
specializes to algebraizability, if one of the two logics comes from a class of al-
gebras) yields an isomorphism between the expanded lattices of theories. The
converse (unlike in the case of sentential and k-deductive systems) is not always
true, and determining when it holds is known as the Isomorphism Problem.
Blok and Jonsson gave a sucient condition (existence of a basis) for the Iso-
morphism Problem and Gil-Ferez provided a more general sucient condition
(existence of a variable), while a necessary and sucient condition (a character-
ization of cyclic projective modules) was given by Galatos and Tsinakis. Their
proof puts the problem in the correct level of abstraction, by extending the
framework even further to modules (join-complete lattices) over complete resid-
uated lattices (or quantales); this corresponds to passing to the action of the
powerset of M to the powerset of the M-set. In particular, both the syntactic
translation and the semantic isomorphism become morphisms at the same level,
namely between modules.
Sentential logics and k-deductive systems are deductive systems where the
syntactic objects involved all have a xed length, while this is not the case
for sequent and hypersequent deductive systems. Although the latter are also
examples of M-sets, the sucient condition for the Isomorphism Problem of
Blok and Jonsson does not apply. The Isomorphism Problem for associative
sequent systems was addressed by Rebagliato and Verdu, Gil-Ferez gave a su-
cient condition (existence of a multi-variable) in the setting of M-sets, while the
general solution (in the context of modules) follows from the work of Galatos
and Tsinakis.
The Isomorphism Problem was also considered in the context of pi-institutions
by Voutsadakis, who provided a sucient condition (the term condition). The
context of pi-institution extends that of M-sets in a dierent direction than
its extension to modules and Voutsadakis condition covers extensions of xed
length deductive systems. A sucient condition (the multi-term condition) for
variable length pi-institutions was provided by Gil-Ferez. The problem of an
exact solution of the Isomorphism Problem for pi-institutions was open.
Our work provides, in particular, a solution to the Isomorphism Problem
for pi-institutions. We rst provide a general categorical context that encom-
passes pi-institutions and modules over quantales and we solve the Isomorphism
Problem in its full generality.
More specically, in the rst of the two talks, we consider the category
of modules over quantaloids. A quantaloid is an enriched category over the
category of join-complete lattices. A one-object quantaloid is coextensive with
a quantale, so our theory can be viewed as a categorical extension of the work
of Galatos and Tsinakis. A module over a quantaloid is dened as an enriched
functor from the quantaloid to the category of join-complete lattices; it is the
natural generalization of an quantale module and of an M-set, where the action
is identied with a homomorphism to the endomorphisms of the join-complete
lattice or to the endomaps of the M-set, respectively.
We solve the Isomorphism Problem by characterizing the modules over quan-
47
taloids for which the theorem holds (every isomorphism is induced by syntactic
translators), which end up being the the projective objects in the category. The
characterization for the cyclic projective modules is given by the existence of a
generalized variable.
`
Angel Gil
Universitat Pompeu Fabra - Spain
[email protected]
On Hilbertizable Gentzen systems associated with nite valued logics
According to J.G. Rafterys [2] denition, a Gentzen relation is Hilbertizable
if it is equivalent to some Hilbert relation. This is the strongest among sev-
eral relations that have been considered in the literature between a Gentzen
system and a Hilbert system. In this work we show that when we consider an
m-dimensional sequent calculus associated with a nite algebra L (in the sense
of M. Baaz et al. [1]) and its corresponding Gentzen system G
L
, then G
L
is
Hilbertizable if and only if the elements of L can can be expressed by terms.
This means that for every element l, there exists a term function p
l
(x) that
always takes the value l when evaluated in L. This will be the case, for instance,
if L is a nite MV-algebra.
References
1. M. Baaz et al,.Elimination of cuts in rst-order nite-valued logics, Jour-
nal of Information Processing and Cybernetics EIK, 29. 1994
2. J.G. Rafterys, Correspondences between Gentzen and Hilbert Systems.
Journal of Symbolic Logic 71(3), 2006.
Nikolaos Galatos
University of Denver - USA
[email protected] Jos Gil-Ferez
Japan Advanced Institute of Science and Technology - Japan
The Isomorphism Problem for modules over quantaloids, Part II
Carlos Caleiro
Technical University of Lisbon - Portugal
[email protected]
Ricardo Gonalves
Technical University of Lisbon - Portugal
[email protected]
Behavioral algebraization of logics (I)
In this second part of the talk we focus on the semantical issues of the theory of
behavioral algebraization of logics. This newly developed behavioral approach to
the algebraization of logics extends the applicability of the methods of algebraic
logic to a wider range of logical systems, namely encompassing many-sorted
languages and non-truth-functionality. However, where a logician adopting the
traditional approach to algebraic logic nds in the notion of a logical matrix the
48
most natural semantic companion, a correspondingly suitable tool is still lacking
in the behavioral setting. Herein, we analyze this question and set the ground
towards adopting an algebraic formulation of valuation semantics as the natural
generalization of logical matrices to the behavioral setting, by establishing some
promising results. or illustration, we use again da Costas paraconsistent logic
C1.
Ramon Jansana
Universitat de Barcelona - Spain
[email protected]
Equationaly orderable quasivarieties and sequent calculi
A quasivariety is equationally orderable if there is a nite set of equations in two
variables that in every member of the quasivariety denes a partial order. We
characterize the equationally orderable quasivarieties as the equivalent algebraic
semantics of the sequent calculi with the binary cut rule which are algebraiz-
able with the translation from equations to sequents performed by the map that
sends an equation t = t

to the pair of sequents t t

, t

t. Sequents are
taken as a pair of a nite sequence of formulas and a formula. We will discuss
characterizations of these sequent calculi and the relations they may have with
their external deductive systems.
Tomasz Kowalski
Francesco Paoli
niversit` a di Cagliari - Italy
Matthew Spinks
Quasi-subtractive varieties
Manuel A. Martins
Universidade de Aveiro - Portugal
[email protected]
Abstract Algebraic Logic approach to Algebraic Specication
Standard abstract algebraic logic (AAL) cannot be straightforwardly applied
to the theory of specication of abstract data types. Specication logic must
be seen as a deductive system (i.e., as a substitution-invariant consequence
relation on an appropriate set of formulas) and behavioral equivalence as some
generalized notion of Leibniz congruence. The class of deductive systems has
to be expanded in order to include multisorted as well as one-sorted systems.
The notion of Leibniz congruence has to be considered in the context of the
dichotomy of visible vs. hidden. In our approach ([MP07, Mar07]), the standard
AAL theory of deductive systems is generalized to the hidden heterogeneous
case. Data structures are viewed as sorted algebras endowed with a designated
subset of the visible part of the algebra, called a lter, which represents the
set of truth values. This new perspective helps to provide a better insight on
the properties of the behavioral equivalence, the key concept in the behavioral
algebraic specication theory ([Hen97]).
In another direction, recently in [MMB09a, MMB09b], the authors intro-
49
duced an alternative approach to renement of specications in which signature
morphisms are replaced by logic interpretations. Intuitively, an interpretation
is a logic translation which preserves meaning. Originally dened in the area
of algebraic logic, in particular as a tool for studying equivalent algebraic se-
mantics ([BP89]), the notion has proved to be an eective tool to capture a
number of transformations dicult to deal with in classical terms, such as data
encapsulation and the decomposition of operations into atomic transactions.
Keywords: Behavioral Equivalence, Behavioral Specication, Renement,
Hidden Logic, Leibniz congruence, Interpretation.
References
BP89 W. J. Blok and D. Pigozzi. Algebraizable logics. Mem. Am. Math. Soc.,
396, 1989.
Hen97 R. Hennicker. Structural specications with behavioural operators: se-
mantics, proof methods and applications. Habilitationsschrift.1997.
Mar07 M. A. Martins. Closure properties for the class of behavioral models.
Theor. Comput. Sci., 379(1-2):53-83, 2007.
MMB09a M. A. Martins, A. Madeira, and L. S. Barbosa. Renement via inter-
pretation. In Proc. of 7th IEEE Int. Conf. on Software Engineering
and Formal Methods, Hanoi, Vietnam, November 2009. IEEE Computer
Society Press.
MMB09b M. A. Martins, A. Madeira, and L. S. Barbosa. Renement via inter-
pretation in a general setting. In Proc. of the Renement Workshop
REFINE09, Eindhoven, Netherlands (co-located with Formal Methods
2009). Electr. Notes Theor. Comput. Sci., 2009.
MP07 M. A. Martins and D. Pigozzi. Behavioural reasoning for conditional equa-
tions. Math. Struct. Comput. Sci., 17(5):1075-1113, 2007.
Micha l Stronkowski
Warsaw University of Technology - Poland Charles University -
Czech Republic
[email protected]
Finite axiomatizability theorems and sub-technology
In 2002 Baker and Wang gave a very clear proof of Bakers theorem: Finitely
generated congruence-distributive varieties are nitely axiomatizable. For this
purpose they introduced denable principal SUBcongruences. In the talk we
would like to notice that the sub-technology of Baker and Wang may be adapted
to
Quasivarieties: classes of algebras dened by quasi-identities, i.e. sentences of
the form ( x)[t
1
( x) = s
1
( x) t
n
( x) = s
n
( x) t( x) = s( x)];
Strict universal Horn classes: classes of models dened by sentences of the
50
form ( x)[
1
( x)
n
( x) ( x)], where ( x),
1
( x), . . . ,
n
( x) are atomic for-
mulas dierent from equations.
The last type of classes is especially interesting from abstract algebraic logic
perspective. Indeed, each sentential logic corresponds to a strict universal Horn
class of logical matrices (algebras endowed with one unary predicate).
We obtained new proofs of
Pigozzis theorem: Finitely generated relative congruence-distributive quasi-
varieties are nitely axiomatizable;
Palasi nskas theorem: Finitely generated lter-distributive protoalgebraic strict
universal Horn classes are nitely axiomatizable.
To this end we introduced denable relative principal SUBcongruences and de-
nable principal SUBlters.
4.2.4 Paradoxes
This session is organized by Andrea Cantini and Pierluigi Minari from the Uni-
versity of Florence - Italy.
Between the end of the 19th century and the beginning of the 20th century,
the foundations of logic and mathematics were aected by the discovery of
a number of paradoxes, involving fundamental notions and basic methods of
denition and inference, which were usually accepted as unproblematic. Since
then paradoxes have acquired a new role in contemporary logic: basic notions
of logic, as it is presently taught, and important metatheorems have reached
their present shape at the end of a process which has been often triggered by
various attempts to solve paradoxes.
The broad aim of this session is to compare and evaluate dierent logical
systems able to solve paradoxes, which involve the notions of truth, set, opera-
tion, abstraction, vagueness. More precisely, the idea is to discuss most recent
proposals which bring about new logical ideas and methods.
We encourage submissions of papers which use proof-theoretic as well model-
theoretic methods, and deal with topics from the following list:
1. Paradoxes from the viewpoint of (the whole spectrum of) substructural
logics, fuzzy logics included;
2. Theories of predication and truth;
3. Paradoxes in type theories and theories of operations;
4. Recent developments in the Neofregean approach;
5. Theories of predication and truth;
6. Epistemic paradoxes.
Papers dealing with the history of paradoxes are also welcome.
51
Accepted contributed talks
Reinhard Kahle
Universidade Nova de Lisboa - Portugal
[email protected]
David Hilbert and the set-theoretic paradoxes
In this talk we present the discussion of dierent set-theoretic paradoxes by
David Hilbert in a series of lectures given at the Department of Mathematics
at the University of Gottingen between 1905 and 1920. This discussion, pre-
served in unpublished lecture notes available at the library of the Mathematical
Institute in Gottingen, shows that the paradoxes played an important r ole in
Hilberts elaboration of what is today called Hilberts programme.
Michal Walicki
University of Bergen - Norway
Paradoxes - propositionally
A feature of natural discourse, not represented by a straightforward propo-
sitional model, is that every pronounced statement has, in addition to its actual
content, also its unique identity (as witnessed, for instance, by cataphoric and
anaphoric references.) The following discourse: (*) The next statement is not
false. The previous statement is false. is purely propositional and, giving an
explicit identier to every statement, it can be represented by the two propo-
sitions: (**) x1 - not(not(x2)) x2 - not(x1). Simplicity of this representa-
tion notwithstanding, it gives a denite account of propositional paradoxes of so
called self-reference, determining uniquely the paradoxical or non-paradoxical
status of every discourse. It does it without any use of higher-order logic or
arithmetics and without recourse to any innitary constructions. Still, it is not
limited to nite, circular paradoxes and applies unmodifed to the more recent,
innitary paradoxes without self-reference.
The model leads also to an interesting equivalence between consistent propo-
sitional theories and directed graphs possessing kernels (independent subset of
nodes with an edge from every outside node to some node in the set). One central
result guarantees the existence of a kernel in graphs without odd cycles, and this
nds its counterpart in the consistency of our theories without self-referential
negation. (In (**) such a negation obtains, e.g.: x1 - not(not(not(x1)))), and
simplies to the Liar: x1 - not(x1).) Thus, besides illuminating the relation
between paradoxes, consistency and circularity, as a more practical consequence,
we can also envision transfer of the results and algorithms between kernel theory
and the research on satisability.
Vilem Novak
University of Ostrava - Czech Republic
[email protected]
Coping With Falakros Paradox in the Meaning of Specials Linguistic Expres-
52
sions
A special but very important class of linguistic expressions is formed by the,
so called evaluative linguistic expressions (e.g., small, medium, big, very weak,
medium strong, extrmely strong, about 1 million, etc.). Logical analysis of their
meaning reveals that the falakros (sorites) paradox is hidden in it. Moreover,
the paradox extends even to real numbers ((very) small amount of water, an
extremely strong pressure, etc.). We present a formal theory, the interpretation
of which provides a model of their meaning and demonstrate that the theory
copes well with the falakros paradox. The main assumption is that vagueness
of the meaning of these expressions is a consequence of the indiscernibility re-
lation between objects. Our main formal tool is the fuzzy type theory (FTT)
introduced in [1]. We developed a formal theory T of the meaning of evaluative
expressions.
Theorem 1: Let an evaluative linguistic expressions Sm be given (for example
very small, extremely weak, etc.). Then in each context the following is
provable in T: 1. Sm(0) 2. There exists n such that surely not Sm(n). 3. There
does not exist x such that surely Sm(n) and surely not Sm(n+1). 4. For all n,
if Sm(n) then it is almost true that Sm(n+1).
Theorem 2: In arbitrary context, there is no last surely small x and no rst
surely big x .
The proofs of these theorems are syntactical and so, they can have various
kinds of interpretation. Full formalism, proofs and other details can be found
in [2].
References
1. V. Novak, On fuzzy type theory, Fuzzy Sets and Systems 149 (2005) 235-
273.
2. V. Novak, A comprehensive theory of trichotomous evaluative linguistic
expressions, Fuzzy Sets and Systems 159 (22) (2008) 2939-2969.
Frode Bjordal
University of Oslo - Norway
A contradictory semi formal resolution respecting classical logic.
We undertake to show that there is a semi formal approach honouring classi-
cal logic which shares features with paraconsistent approaches and also retains
virtually all nave principles of truth and comprehension. The semantics and
some salient partial axiomatic and inferential principles of the system, called
Minimalistic Liberalism (ML), will be presented. ML diers from paraconsis-
tent approaches in that classical logical principles are always theorems and their
negations never are, even in the presence of paradoxical phenomena generated
by the liberal comprehension principle implicitly dened by the theory; the
latter happens by means of principles governing a truth operator T and by pre-
supposing that it is, and never fails to be a theorem that s is an element of
x:F(x) i TF(s). E.g., if R is Russells set, it will be a theorem of ML that
53
R is in R, and it will as well be a theorem in ML that R is not in R. But
it will not be a theorem of ML that R is in R and R is not in R. So ML is
non-adjunctive; as like magnetic poles, contradictory theorems repel. One focus
of the presentation will be upon explaining the semantical framework, which
relies upon some additional twists upon revisionary types of semantics in the
tradition from Herzberger, Gupta and Belnap, and to isolate some of the salient
axiomatic and inferential principles which hold. Importantly, modus ponens is
not generally valid. There are, however, a row of other inference rules which
compensate for this. We will focus upon how paradoxes are resolved, and pay
particular attention to the recalcitrant Curry-paradoxes, and how their resolu-
tion crucially hinges upon the fact that ML is, and must needs be a semi formal
system.
Riccardo Bruni
Universit` a di Firenze - Italy
[email protected]
An approach to paradoxes by Beppo Levi
In a 1908 note of his [2], Beppo Levi argued that some of the most famous
among the logical antinomies of mathematics, could be shown to be based on
some fundamental misunderstanding of the meaning of the terms involved in
their formulation. The aim of this talk is to introduce the audience to Levis
approach to the issue, and to discuss some consequences of it which might be
useful to consider from the metamathematical viewpoint. Though this paper of
Levis carries no original methodology for dealing with the paradoxes, we will
present reasons of a dierent sort for re-considering this contribution from a
contemporary perspective: rst of all, the similarities between Levis proposed
analysis of the logical antinomies and those provided elsewhere (as within the
French group of so-called semi-predicativists, or even those given by Russell and
Poincare); relatedly, Levis peculiarity of not making the choice between what
are commonly considered as alternative views in the debate on the foundations
of mathematics; nally, the relation with a slightly more popular contribution
of Levis in the metamathematical eld, the one on Zermelos axiom of choice
naturality as an assumption in the settheoretic setting, and his proposed sub-
stitute for that assumption, his approximation principle, which, although so
far remained unclear and unexplored, has been recently undergone a process of
reformulation (in [1]) that seems capable of making it intelligible.
1. R. Bruni and P. Schuster. On Beppo Levis approximation principle. In
preparation, 2009.
2. B. Levi. Antinomie logiche. Annali di Matematica Pura ed Applicata,
(III), Tomo XV:187216, 1908.
Riccardo Strobino
Scuola Normale Superiore - Italy
[email protected]
54
Medieval obligations and paradoxes
The medieval theory of obligations can be regarded as an attempt to codify in
a highly regimented way the structure of a dialectical disputation. Consider a
situation in which two disputants perform a series of actions, governed by rules.
The Opponent is supposed to lay down an initial thesis and then, in case the
Respondent admits it, to put forward further sentences with the aim of forcing
the latter into a contradiction. The Respondent, on the other hand, should
try to maintain consistency by conceding, denying or doubting the sentences
proposed during the disputation according to certain criteria. Such criteria are
precisely the object of a number of rules that constitute the theoretical bulk of
medieval treatises on the topic.
One interesting feature of this literature is that, after setting up the stage
at an abstract level, the focus shifts on the application of rules in the context of
particular examples which are supposed to count as limit-cases devised to test
the theory and see it at work. Now, such limit-cases often come in the form of
paradoxes. Much in this game seems to depend on the Respondents ability to
detect the paradox in advance and choose either to reject it or, if it is possible,
to work out a solution, i.e. accept the paradox as an initial thesis and nd out
a way to defend it.
I will examine a number of examples in order to clarify how the theory was
actually employed in such contexts. The focus will be especially on epistemic
paradoxes of the form: if p is the case, then you know that p is the case, and if
not-p is the case, then you know that not-p is the case.
4.2.5 Substructural Logics
This session is organized by Francesco Paoli and Tomasz Kowalski fro the Uni-
versity of Cagliari, Italy.
Substructural logics are usually described as logics that lack some members
of the usual triple of structural rules: contraction, weakening, or exchange.
From this descripion alone it is clear that substructural logics are intimately
connected with sequent calculi. Indeed their origin is rooted in proof theory
and Gentzen-style systems. Four broad families of logics immediately answer
the description:
relevant logics
BCK logics, or logics without contraction
linear logic and its extensions
Lambek calculus
It was realised early on that substructural logics share a common algebraic
characteristic. Namely, all the algebraic semantics for substructural logics are
(embeddable in) residuated structures. Hence the slogan Substructural logics
are logics of residuated structures. This shift of focus brings forth a fruitful
55
connection with more traditional areas of mathematical research, such as lattice-
ordered groups, as well as encompassing two families of logics that the proof-
theoretical description misses:
fuzzy logics (where the sequent presentation is not obvious)
intuitionistic and intermediate logics, including classical logic (where all
structural rules are present)
Thus, a carefully stated proof-theoretical description of substructural logics
could perhaps read: axiomatic extensions of any logic that, if presented as a
sequent calculus, lacks zero or more structural rules. A phrase worthy of a
logician, without a doubt.
We welcome submissions of papers on topics from (but possibly also outside
of) the following list:
Proof theory of substructural logics;
Substructural logics from the viewpoint of abstract algebraic logic;
Residuated lattices;
Individual classes of residuated lattices (l-groups, MV algebras, etc.);
Reducts and expansions of residuated lattices (BCK algebras, equivalential
algebras, modal FL-algebras, etc.)
Relationships between substructural logics and other non-classical logics
(modal, paraconsistent, quantum logics, etc.);
Applications of substructural logic
Accepted contributed talks
Arnon Avron
Tel Aviv University - Israel
[email protected]
Substructural Logics are not Logics of Residuated Structures
RMI
m
is the logic which is obtained from the sequent calculus for R
m
, the
purely intensional (or multiplicative) fragment of the relevant logic R, by
viewing sequents as consisting of nite sets of formulas on both of their sides.
This is equivalent to adding the converse of contraction to the usual sequential
formulation of R
m
. Now it is well known that RMI
m
is a relevant logic, enjoying
major properties like the variable-sharing property, cut-elimination, and a very
natural version of the relevant deduction theorem. However, by enriching it with
an additive conjunction, one gets an unconservative extension (with the full
power of the intensional fragment of the semi-relevant system RM, for which the
variable-sharing property fails). It follows that the substructural logic RMI
m
56
is not the logic of any class of residuated structures. Despite this observation,
RMI
m
has a very useful and eective semantics. In this paper we construct an
eective class S of nite-valued matrices, for which RMI
m
is nitely strongly
sound and complete. Moreover: all the matrices in S can be embedded in one
eective innite-valued matrix for which RMI
m
is strongly sound and complete.
We also show that the last result cannot be improved: there is no nite-valued
matrix for which RMI
m
is even weakly sound and complete.
Arno Bastenhof
Utrecht University - The Netherlands
[email protected]
An algebraic semantics for type similarity in Symmetric Categorial Grammar
Starting with (unit-free) intuitionistic multiplicative linear logic, one descends
the Lambek categorial hierarchy by gradually removing the structural rules
of exchange and even associativity. The increase in structural discrimination
so achieved is coupled with a decrease in the expressivity of type similarity:
the reexive, transitive and, crucially, the symmetric closure of derivability.
Pentus and Moortgat found a counterexample to this pattern in the Lambek-
Grishin calculus (LG), where associativity and commutativity remain underiv-
able while at the same time being validated under type-similarity. Compared
to the intuitionistic bias underlying the Lambek hierarchy, LG manifests an
arrow-reversing duality by adding a co-residuated family of connectives (headed
by the par), thus achieving full symmetry. In this talk, we revisit Pentus and
Moortgats results by showing type-similarity in LG sound and complete w.r.t.
an algebraic semantics featuring two associative and commutative binary op-
erations, related by linear distributivity. In addition, we show that this result
extends to a recent suggestion of Moortgat to augment LG with families of (co-
)Galois connected operators.
Giulia Battilotti
University of Padua - Italy
Structural rules and implication in a sequent calculus for quantum computa-
tion
Basic logic is a core for sequent calculi of several substructural extensional log-
ics, including intuitionistic, linear and quantum logics. It shows how to obtain
such extensions by the addition of structural rules to the basic calculus, that is
obtained translating metalinguistic links into logical connectives. The critical
point is represented by the implication. In the framework of basic logic, we have
recently introduced a paraconsistent and predicative interpretation of quantum
parallelism in terms of sequents. It allows to consider logical implication as a
causal link, in opposition to a semi-predicative associative link, formalizing a
quantum link, which can have a meaning only in a paraconsistent setting, where
the variable has the role of a random variable. Substitution of variables by closed
terms is the structural rule which makes the quantum link collapse. The cal-
culus exploits the limitation of contexts in the rules for quantiers. One could
57
argue a computational interpretation of quantum contextuality, in terms of
sequents.
James Brotherston
Imperial College London - UK
[email protected]
Bunched logics displayed
*Display calculi*, as formulated by Belnap in the 1980s, can be seen as natural
successors of Gentzens sequent calculi, suitable for the proof-theoretic analysis
of substructural logics. They are characterised by the display property, which
essentially states that proof judgments may be rearranged so that any chosen
part appears alone on one side of the proof turnstile.
In this talk, we apply display calculus techniques to obtain a unied proof
theory for *bunched logics*, which originate in OHearn and Pyms BI and
can be seen as the result of freely combining a standard (additive) propositional
logic with a (multiplicative) linear logic. The practical interest in bunched logics
stems from their Tarskian resource interpretation of formulas, as used e.g. in
the heap model of separation logic. However, the cut-free sequent calculus for
BI does not extend naturally to its important variants such as Boolean BI. We
show how cut-eliminating display calculi may be uniformly obtained for all the
principal varieties of bunched logic, and incidentally provide an explanation as
to why well-behaved sequent calculi seem very unlikely to exist for most of these
varieties.
This talk is based upon a related paper by the speaker A unied display
proof theory for bunched logic. Submitted, 2009.
Anatolij Dvure censkij
Slovak Academy of Sciences - Slovakia
[email protected]
State-morphism MV-algebras
In the last decade, the interest to probabilistic uncertainty in many valued logic
increased. A new approach to states on MV-algebras was recently presented
by T. Flaminio and F. Montagna; they added a unary operation, , (called
as an inner state or a state-operator) to the language of MV-algebras, which
preserves the usual properties of states. It presents a unied approach to states
and probabilistic many valued logic in a logical and algebraic settings.
In the talk, we show how subdirectly irreducible elements can be described,
we show that any state-operator on the variety V(S
1
, . . . , S
n
) is a state-morphism-
operator. We describe an analogue of the LoomisSikorski theorem for a state-
morphism MV-algebra (A, ), where A is a -complete MV-algebra and is a
-endomorphism: We show that any such state-morphism MV-algebra is a -
epimorphic image of (T ,
T
), where T is a tribe dened on a totally disconnected
compact Hausdor topological space and
T
is a -endomorphism generated by
a continuous function.
References:
58
1. A. Di Nola, A. Dvurecenskij, State-morphism MV-algebras, Ann. Pure
Appl. Logic 161 (2009), 161173.
2. A. Di Nola, A. Dvurecenskij, On some classes of state-morphism MV-
algebras, Math. Slovaca 59 (2009), 517-534.
3. A. Di Nola, A. Dvurecenskij, A. Lettieri, On varieties of MV-algebras with
internal states, submitted.
4. T. Flaminio, F. Montagna, MV-algebras with internal states and proba-
bilistic fuzzy logic, Inter. J. Approx. Reasoning 50 (2009), 138-152.
Rostislav Horcik
Czech Technical University - Czech Republic
[email protected]
Disjunction Property and Complexity of Substructural Logics
In this talk we are going to present an algebraic method for proving PSPACE-
hardness of a substructural logic which is less dependent on the sequent calculus.
More precisely, we will prove by algebraic means that each substructural logic
satisfying a stronger version of the disjunction property (SDP) is PSPACE-
hard. This gives us a simpler method since it is usually easy to prove SDP from
a cut-free sequent calculus. We demonstrate it by showing that the basic sub-
structural logics (i.e., Full Lambek calculus and its extensions by the structural
rules of exchange, contraction, left and right weakening) have SDP. Thus, as a
corollary, we obtain PSPACE-hardness for these logics.
George Metcalfe
University of Bern - Switzerland
[email protected]
Admissible Rules for Substructural Logics
The study of logical systems usually focuses on derivability: whether or not a
rule, understood as a set of premises and a conclusion, belongs to the conse-
quence relation of the system. Such rules may be thought of as providing an
internal description of the system. However, an external perspective, de-
scribing properties of the system, can also be valuable. Following Lorenzen, a
rule is said to be admissible in a system if the set of derivable structures is closed
under the rule; that is, adding the rule to the system does not give any new
derivable structures. In algebra, such rules correspond to quasi-equations hold-
ing in free algebras, while from a computer science perspective, admissibility is
intimately related to equational unication.
For classical logic, derivability and admissibility coincide: the logic is struc-
turally complete. However, for many non-classical logics, in particular, core
modal, many-valued, intermediate, and substructural logics, this is no longer
the case. In this work, we consider a selection of open problems for characteriz-
ing admissible rules. In particular, while the admissible rules of most fragments
of intermediate logics have been characterized, we provide rst bases for the
implication-negation fragments. We also provide bases for the admissible rules
59
of fragments of the relevant logic RM and identify some interesting research
challenges in the area.
Szabolcs Mikulas
University of London - UK
[email protected]
On representable distribtive lattice-ordered residuated semigroups
Residuated algebras have been extensively investigated in the literature, partly
because of their connection to substructural logics. We call a residuated algebra
representable if it is isomorphic to a family of binary relations. They provide
sound semantics for substructural logics like the Lambek calculus (LC) and
relevance logics. Cases of completeness include the LC (Andreka and Mikulas)
and relevance logic with mingle RM (Maddux).
In this talk, I address the problem whether the completeness results above
can be extended. In the case of LC, we look at the similarity type expanded with
join, and we also consider relevance logics without the mingle axiom. We will
look at the corresponding classes of representable residuated algebras and see
that they have nonnitely axiomatizable (quasi)equational theories. Applying
this result to logic, we get that it is impossible to get completeness with nitely
many axioms and standard derivation rules for a variety of substructural logics.
Hiroakira Ono
JAIST - Japan
[email protected]
William Young
Vanderbilt University - USA
[email protected]
Extending the G odel-McKinsey-Tarski Translation to Substructural Logics
In 1933, Godel suggested a way of translating formulas in the language of intu-
itionistic logic into the language of classical modal logic. He conjectured that
for any intuitionistic formula ;, ; is provable in the intuitionistic logic LJ i
its translation G(66;) is provable in the modal logic S4, making S4 a modal
counterpart of LJ. This conjecture was later proven by McKinsey and Tarski,
using algebraic methods. Then, Dummett and Lemmon extended this result
to all superintuitionistic logics, showing that for any superintuitionistic logic L,
there exists a smallest modal counterpart of L, among the normal modal logics
over S4. Maksimova and Rybakov showed that there also is a greatest modal
counterpart of L, and Blok proved a conjecture of Esakias concerning an axiom-
atization for this greatest modal counterpart. In this talk, the sequent system
S4FL for substructural modal logics (i.e., modal logics with all of the theorems
of the substructural logic FL) will be introduced, corresponding to the classical
modal logic S4. The Godel-McKinsey-Tarski translation will then be extended
to take formulas in substructural logics into formulas in substructural modal
logics, and analogues of the aforementioned theorems will be considered for the
class of substructural logics over FL with respect to the class of substructural
modal logics over S4FL. In particular, we will show that there is a least element
60
and a maximal element (with a specic axiomatization) among the modal coun-
terparts of a given substructural logic.
James Raftery
University of Kwazulu-Natal - South Africa
[email protected]
Finiteness properties for idempotent residuated structures
An idempotent residuated po-monoid is said to be semiconic if it is a subdirect
product of algebras in which the monoid identity is comparable with all other
elements. It is proved that the quasivariety SCIP of all semiconic idempotent
commutative residuated po-monoids is locally nite. The lattice-ordered mem-
bers of this class form a variety SCIL, provided that we add the lattice operations
to the signature. This variety is not locally nite, but it is proved that SCIL
has the nite embeddability property (FEP). More generally, for every relative
subvariety K of SCIP, the lattice-ordered members of K are shown to have the
FEP. This gives a unied explanation of the strong nite model property for
a range of logical systems. It is also proved that SCIL has continuously many
semisimple subvarieties, and that the involutive algebras in SCIL are subdirect
products of chains.
Susan Rogerson
Sam Butchart
Monash University - Australia
A, more adorable
Sebastian Sequoiah-Grayson
University of Leuven - Belgium
[email protected]
Permuting Nonassociating Lambek Calculus and Cognitive Grammars
We show how the residuation structure that encodes the permuting, nonassociat-
ing Lambek calculus is a tempting basestructure for modeling the information
ow in monoagent deductive reasoning scenarios. The resulting model has a
straightforward interpretation in terms of the database structure, or grammar,
of the cognitive langauge of deductive reasoning.
Katarzyna Slomczynska
Pedagogical University - Poland
[email protected]
Projective algebras in Fregean varieties
Projective algebras in Fregean varieties A variety of algebras with a distin-
guished constant 1 is called Fregean if it is 1-regular and congruence orderable.
The subdirectly irreducible algebras in a Fregean variety can be characterized as
those which have the largest non-unit element, traditionally denoted by *. Ev-
ery congruence permutable Fregean variety has a binary term that turns each of
its algebras into an equivalential algebra, where by an equivalential algebra we
mean a grupoid that is a subreduct of a Heyting algebra with the naturally given
61
equivalence operation. We show that for a congruence permutable Fregean va-
riety V of nite type L the following conditions are equivalent: (1) every nitely
generated algebra from V is projective; (2) A-* forms a subalgebra of A for every
subdirectly irreducible algebra A from V; (3) V fullls the identities: t(1,,1) =
1 and t(x
1
, , x
n
)yy = t((x
1
)yy, , (x
n
)yy) for every t from L. The equivalence of
(1) and (2) is also true under the weaker assumption that V is a substractive
Fregean variety. The case where the language of V contains more than one con-
stant is also discussed. In particular, we show that in the variety of equivalential
algebras with 0, that gives the algebraic semantics for the equivalence-negation
fragment of IPC, a nitely generated algebra is projective, i 0 is not equal to 1.
1. S. Ghilardi, Unication, nite duality and projectivity in varieties of Heyt-
ing algebras, Ann. Pure Appl. Logic 127 (2004), 99-115.
2. P.M. Idziak, K. Slomczynska, A. Wronski, Fregean Varieties, Internat. J.
Algebra Comput. 19 (2009), 595-645.
Shawn Standefer
University of Pittsburgh - USA
[email protected]
Philosophical Aspects of Display Logic
Many discussions of logical inferentialism take place against a backdrop of
natural deduction systems. Against this, we argue that the proper setting for
philosophical re?ection on the meaning of connectives for the inferentialist is the
consecution calculus. The arguments for this are based on the role of structure in
natural deduction systems and in consecution calculuses. We present examples
of elements of natural deduction that play a structural role, arguing that their
in?uence on the meaning of the connectives is obscured in natural deduction
systems. This role is made explicit in consecution calculuses, which present a
clearer picture of the context of deducibility.
Once the inferentialist has shifted her attention to consecution calculuses,
similar considerations seem to favor adopting the display logic generalization
of consecution calculuses. This is because display logic makes the distinction
between structure and non-structure sharper. There is a natural question of
what constitutes harmony in this setting. We provide a novel characterization
of harmony, suited to display logic, and close by presenting two applications of it.
4.2.6 Categorical Logic
This session is organized by Valeria de Paiva (Cuil. Inc, USA) and Andrei Rodin
(University of Paris 7, France).
Categorical logic is a branch of mathematical logic that uses category the-
ory as its principal mathematical tool and as mathematical foundation. This
mathematical setting profoundly changes the conception of logic put forward
62
by Frege and Russell in the beginning of 20th century both in its technical
and philosophical aspects. On the technical side categorical logic inherits fea-
tures from earlier constructive and algorithmic approaches to logic, in particular
from realizability, lambda-calculus, intuitionistic logic and type theory. (In fact
a typed intuitionistic logical calculus in the categorical setting appears to be the
most natural system of logic while classical logic turns out to be a very special
case that requires strong additional conditions.). This is one of the reasons why
categorical logic is so successfully used in computer science. On its philosophical
side categorical logic suggests a new notion of intrinsic logic that is analogous
to the notion of intrinsic geometry that made a revolution in this mathematical
discipline in 19th century. Frege and Russell after Aristotle conceived of logic as
a system of universal rules of reasoning independent of any particular subject
domain. Categorical logic not only diversies the notion of logic by giving a
space for dierent systems of logic, but also provides a mechanism of adjust-
ment of a system of logic (i.e. a formal language) to a given domain of study
and thought.
Some of the research in categorical logic sees a great dichotomy between
categorical proof theory and categorical model theory. Categorical proof
theory is able to model dierent proofs of a given theorem, and compares these
dierent proofs, using categorical concepts. Categorical model theory is an
extension of traditional model theory, where models are categories. We see this
meeting as encompassing both aspects of categorical logic.
Topics that t this Special Session include, but are not limited to, the fol-
lowing:
1. Relationships between logic and geometry in a topos-theoretic setting
2. Categorical logic and Categorical foundations of mathematics
3. Sketch theory; diagrammatic syntax
4. Functorial semantics and Categorical Model theory
5. Quantum logic categorically
6. Extensions of categorical semantics to dierent kinds of logics, such as
modal and substructural logics
7. Comparison of dierent categorical frameworks
Invited speaker
Maria Emilia Maietti
University of Padova - Italy
[email protected]
The role of the quotient completion for the foundations of constructive mathe-
matics
A key characteristic of the foundations for constructive mathematics is that
they should enjoy a computational interpretation.
63
In joint work with G. Sambin [2] we argued that a foundation for constructive
mathematics should have two levels: an intensional one acting as a programming
language and an extensional one in which to develop mathematics. The link
between the two levels should guarantee the extraction of programs from proofs.
Category theory oers a tool to characterize such a link in terms of quotient
completion with respect to a suitable bration developed in joint work with G.
Rosolini.
Key examples of two-level foundations are available based on Martin-Lofs
type theory and the minimalist type theory in [1] following [2].
1. M.E. Maietti A minimalist two-level foundation for constructive mathe-
matics APAL, 160(3):319354,2009
2. M.E. Maietti, G. Sambin Toward a minimalist foundation for construc-
tive mathematics in From Sets and Types to Topology and Analysis:
Practicable Foundations for Constructive Mathematics, (L. Crosilla and
P. Schuster eds.) OUP, 2005.
Acepted contributed talks
Olivia Caramello
Scuola Normale Superiore - Italy
[email protected]
Frasses construction from a topos-theoretic perspective
We present a topos-theoretic interpretation of (a categorical generalization of)
Frasses construction in Model Theory, with applications to countably categor-
ical theories.
The proof of our main theorem represents an instance of exploiting the inter-
play of syntactic, semantic and geometric ideas in the foundations of Topos The-
ory; specically, the three concepts involved in the classical Frasss construction
(i.e. amalgamation and joint embedding properties, homogeneous structures,
atomicity of the resulting theory) are seen to correspond precisely to three dif-
ferent ways (resp. of geometric, semantic and syntactic nature) of looking at
the same classifying topos.
References:
1. Olivia Caramello, Frasses construction from a topos-theoretic perspec-
tive, arXiv:math.CT/0805.2778.
Dominique Duval
Universite de Grenoble - France
[email protected]
Deduction rules are fractions
A deduction rule is usually written as a fraction H/C. The aim of the talk is to
actually dene a deduction rule as a fraction in the categorical sense of Gabriel
64
and Zisman. However, then it is rather written as C/H, with the hypotheses as
denominator and the conclusion as numerator. This point of view relies on the
denition of categorical entailments, which might be called potential isomor-
phisms. In terms of logic, as long as models are concerned the entailments may
be seen as isomorphisms, but for dealing with proofs it is essential to consider
that the entailments are not invertible.
Luis Estrada-Gonz alez
State University of Morelos - Mexico
[email protected]
The other topos theory
In this talk I will argue that topoi as studied up to now, standard topoi, are
just part of the concept of topos and therefore common theorems on topos logic
tell just part of the relevant story. As I shall explain, that incomplete picture is
due to a prejudice towards truth which permeates from the denition of a topos
(via the use of the morphism true in dening the subobject classier) to the
denition of validity in the internal logic and then to theorems and proofs about
topoi and their internal logic. All this nally leads to distorted philosophical
claims made on base of those results, like The internal logic of a topos is in
general intuitionistic, Intuitionistic logic is the objective logic of variable sets,
The universal laws of mathematics are intuitionistic, A subobject classier
is a truth-values object or The internal logic of a topos is in general many-
valued. Chris Mortensen speaks of a considerable Public Relations Exercise
done on behalf of intuitionistic logic in topos theory, but as I see it it is derived
from the prejudice towards truth, in starting from true in the classier instead
from false, as Mortensen does in dening complement-topoi, or from none of
them in particular, as I will suggest. I will sketch how topos theory looks like
once more appropriate levels of abstraction are introduced.
References
1. Bell, J.L. (1986): From absolute to local mathematics, Synthese, 69 (3),
pp. 409-426.
2. Estrada-Gonzlez, L. (forthcoming): Complement-topoi and dual intuition-
istic logic, to appear in the Australasian Journal of Logic, vol. 8.
3. McLarty, C. (2006): Two constructivist aspects of category theory, Philosophia
Scienti, Cahier spcial 6, 2006, pp. 95-114.
4. Mortensen, C. (1995): Inconsistent Mathematics, Dordrecht: Kluwer.
5. Mortensen, C. (2003): Closed set logic, in Brady, R. T. (ed.), Relevant
Logics and Their Rivals, Vol. II, pp. 254-262. Ashgate Publishing, Alder-
shot, 2003.
Rene Guitart
University Paris-Diderot 7 - France
[email protected]
65
The link between logic and geometry in the mathematical pulsation between 3-
ary and 2-ary laws
Michael Lieberman
University of Pennsylvania - USA
[email protected]
Accessible Categories and Abstract Elementary Classes
We present a family of rank functionscomplete with topological motivation
for use in the analysis of stability in abstract elementary classes with amal-
gamation, and derive a partial stability spectrum result for tame classes that
generalizes a result of the seminal paper of Baldwin, Kueker, and VanDieren.
We also extract a partial spectrum result for weakly tame AECs, thanks to the
surprise appearance of a notion from the theory of accessible categories. We
highlight the connections between these two elds (whose deep anities have
yet to be fully appreciated) and distill AECs down to their category-theoretic
essence. Once we begin looking at things through the eyes of a category theorist,
some very surprising results appear, seemingly out of nowhere. In particular,
using nothing more than the Yoneda embedding, we obtain a peculiar structure
theorem for categorical AECs, an equivalence of categories that identies the
large structures in a -categorical AEC with sets equipped with an action of
the monoid of endomorphisms of the unique structure of cardinality .
Eduardo Ochs
Federal Fluminense University - Brazil
[email protected]
Downcasing Types
When we represent a category C in a type system it becomes a 7-uple, whose
rst four components - class of objects, Hom, id, composition - are structure;
the other three components are properties, and only these last three involve
equalities of morphisms.
We can dene a projection that keeps the structure and drops the proper-
ties part; it takes a category and returns a proto-category, and it also acts on
functors, isos, adjunctions, proofs, etc, producing proto-functors, proto-proofs,
and so on.
We say that this projection goes from the real world to the syntactical
world; and that it takes a real proof, P, of some categorical fact, and returns
its syntactical skeleton, P

. This P

is especially amenable to diagrammatic

representations, because it has only the constructions from the original P the
diagram chasings have been dropped.
We will show how to lift the proto-proofs of the Yoneda Lemma and of
some facts about monads and about hyperdoctrines from the syntactical world
to the real world. Also, we will show how each arrow in our diagrams is a term
in a precise diagrammatic language, and how these diagrams can be read out as
denitions. The downcased diagrams for hyperdoctrines, in particular, look
as diagrams about Set (the archetypical hyperdoctrine), yet they state the def-
inition of an arbitrary hyperdoctrine, plus (proto-)theorems.
66
Alain Proute
University Paris-Diderot 7 - France
[email protected]
On the link between topoi and the vernacular of mathematics
Eike Ritter
University of Birmingham, UK
[email protected]
Valeria de Paiva
Cuil, Inc. USA
[email protected]
A brational semantics for System K
The Curry-Howard isomorphism for intuitionistic modal logic S4 has been well
established - there is both a well-established type theory and a well-established
categorical semantics. In particular, brations can be used to model the dis-
tinction between modal and intuititonistic formulae.
For the weaker System K the situation is more complicated. There have
been denitions of a suitable type theories and also categorical semantics, but a
categorical semantics using brations in the same way as the one for intuition-
istic S4 has not been given. We give such a categorical semantics and show that
it also links in with the already existing type theories.
Jean-Jacques Szczeciniarz
University Paris-Diderot 7 - France
Categories and diagrammatic proofs
Vladimir L. Vasyukov
Institute of Philosophy RAS - Russia
[email protected]
Quantum Logics and Categories: Localism vs. Globalism
4.2.7 Negation
This session is organized by Sergei Odintsov (Sobolev Institute of Mathematics
- Russia) and Heinrich Wansing (Dresden University of Technology - Germany)
Dealing with a certain polarity of thought, negation is, perhaps, the most
crucial among the logical connectives. It has been studied since antiquity and
has been subject to thorough investigations in the development of philosophi-
cal logic, linguistics, articial intelligence and logic programming. This devel-
opment shows that bringing into play various types of negation may produce
highly fruitful and promising results in many areas, such as paraconsistent logic,
non-monotonic reasoning, the theory of data bases and logic programming.
67
The properties of negation - in combination with those of other logical op-
erations and structural features of the deductibility relation - serve as gateways
among logical systems. Moreover, a dierence between various logical systems
can often be reconstructed as a dierence of certain features of negation opera-
tors used in these systems.
Notwithstanding the importance of negation, the immense literature on
negation is full of disagreements concerning at least necessary conditions un-
der which a unary connective ought to be regarded as a negation operation, the
syntactical type to which a negation operator should belong , etc. We hope
that this session will contribute to comparing dierent kinds of negation, devel-
oping a general theory of negation, and investigating the scope and validity of
principles about negation.
Topics suitable for this Special Session include, but are not limited to, the
following ones:
1. proof-theoretical versus semantical treatments of negation
2. negation, consistency, and inconsistency; interrelations between these no-
tions
3. negation and Galois connections; correspondence theory for negation
4. negation in the light of modal logic
5. negation in relevant and substructural logics
6. constructive treatments of negation
7. negation in belief revision
8. negation in logic programming and non-monotonic reasoning
9. negation in adaptive logics
10. negation in paraconsistent logics
11. negation in categorical grammar
12. negation in concept analysis
Accepted contributed talks
Ofer Arieli
The Academic College of Tel-Aviv - Israel
Arnon Avron
Tel-Aviv University - Israel
[email protected]
Anna Zamansky
Jerusalem College of Engineering - Israel
All Natural Three-valued Paraconsistent Logics are Maximal
68
Maximality is a desirable property of paraconsistent logics, motivated by the
aspiration to tolerate inconsistencies, but at the same time retain from classical
logic as much as possible. This notion of maximality, as well as other notions
considered in the literature, is based on extending the set of theorems of a logic.
In this paper, we use a strictly stronger notion of maximality: a paraconsistent
logic L is maximally paraconsistent (in the strong sense) if every logic L that
properly extends L is not paraconsistent. We show that all the three-valued
paraconsistent logics, satisfying some very natural conditions, are maximal in
the strong sense. This includes well-known three-valued paraconsistent logics
like Settes logic P1, Priests logic LP, the semi-relevant logic SRM3, the logics
PAC and J3, as well as any extension of them obtained by enriching their lan-
guages with extra three-valued connectives.
Diderik Batens
Gent University - Belgium
[email protected]
Negations with a Contextual Meaning
Some adaptive logics, including all inconsistency-adaptive logics, assign to nega-
tion a meaning which is contextual: the premise set determines which occur-
rences of negation have one meaning, say the classical one, and which the other,
say a paraconsistent one. Making the meaning of negation contingent on the
premise set results in sensible applications with respect to interpreting texts,
but also with respect to the mathematical and empirical theories.
The technicalities will be illustrated in terms of a specic inconsistency-
adaptive logic. Several variant logics will be presented and their use and eect
with respect to negation discussed. Some of these logics are not inconsistency-
adaptive but realize an adaptation with respect to other properties of negation,
for example its completeness.
A dierent point concerns the way in which logical systems dene the mean-
ing of logical symbols and so determine theories from sets of non-logical axioms.
Such questions are raised by adaptive logics in that, for example, the meaning
assigned to all negations in consistent premise sets by inconsistency-adaptive
logics is identical to the meaning assigned to them by Classical Logic.
Giulia Battilotti
University of Padua - Italy
[email protected]
Negation as a Primitive Duality in a Model for Quantum Computation
We dene the random rst order domain of the set of outcomes of a mea-
surement experiment on a quantum system. We consider a primitive negation
of assertions originated by Girards duality and extend it to random rst order
domains. So we obtain a representation of quantum states in dual couples, that
allows to represent the NOT of quantum states. In this setting, we nd a char-
acterization of the eigenstates of the NOT matrix (the NOT gate in a quantum
computer). As is well known, such eigenstates, including Bells states, are very
signicant in quantum computation. In our representation, the eigenstates are
69
xed points in a cut elimination procedure exploiting the duality.
1. Battilotti, G.: Interpreting quantum parallelism by sequents, Interna-
tional Journal of Theoretical Physics, to appear, online publication 31th
july 2009.
2. Dalla Chiara M.L., Giuntini R., Leporini R., (2003) Quantum Computa-
tional Logics: a survey. in V. F. Hendricks, J. Malinowski (eds.), Trends
in Logic: 50 Years of Studia Logica, Kluwer Academic Publishers, 213-255
3. Sambin, G., Battilotti, G. and Faggian, C.: Basic Logic: Reection, Sym-
metry, Visibility, The Journal of Symbolic Logic 65 (2000) 979-1013.
Jose Martinez Fernandez
University of Barcelona - Spain
[email protected]
Negation and the Fixed-point Property
k-valued clone with constants is a set of k-valued functions that is closed with
respect to composition and includes the projections and the constant functions.
By denition, a clone F has the (Gupta-Belnap) xed-point property if, and
only if, every system of equations x = f(x,y,...), y = g(x,y,...),... (for functions
f,g,... in F) has a solution. Clones will be understood as the interpretation of
propositional languages and systems of equations are meant to represent self-
referential nets of sentences (for instance, the Liar paradox is represented as x =
x). By well-known xed-point theorems, the three-valued clones with constants
generated by Kleene logics (strong and weak) have the xed-point property. In
this talk we will characterize the xed-point property for 2- and 3-valued clones
as depending on whether the clones are able to express a generalized form of
negation. We will also present some general results that specify classes of k-
valued clones that have the xed-point property. Finally those results will be
applied to show that certain clones that generalize the weak Kleene operators
and that include a pseudo-operator of strong negation and pathologicality op-
erators (i.e. operators meaning in the intended philosophical interpretation of
the truth values- p is a pathological sentence, p is neither true nor false...) have
the xed-point property.
Reinhard Kahle
Universidade Nova de Lisboa - Portugal
[email protected]
Negation: Default is Explicit + Update
Negation is still one of the controversial concepts underlying logic programming.
While the negation-as-failure interpretation is operationally well-understood,
the logical interpretation of negation is less clear. The two most prominent
approachesclosed world assumption and completionhave helped us to un-
derstand the logic that underlies negation as failure. But, as Shepherdson re-
marks, [t]heir disadvantage is that the logics involved are more complicated
70
and less familiar than classical logic so that they are not likely to help the naive
programmer express his problem by means of a logic program, or to check the
correctness of a program. We will show that default negation in (normal, gener-
alized, and dynamic) logic programming can be understand as explicit negation
in an update framework, as long as we consider stable model semantics. The
updates are taken from dynamic logic programming. The technical result is not
particularly complicated, but it has some interesting conceptual consequences.
Default negation is a prime example in non-monotonic reasoning. Our result
questionson philosophical groundsthe status of default negation as a spe-
cial, non-monotonic form of negation. Thus, default negation can be seen as
explicit negation just involving an update aspect (which could also be consid-
ered temporally). Our analysis suggests that non-monotonic reasoning can be
recast as classical reasoning incorporating forms of meta reasoning.
Hitoshi Omori
[email protected]
Toshiharu Waragai
[email protected]
Tokio Institute of Technology - Japan Some Modal Logics Seen as Sub-
systems of Classical Propositional Calculus
One of the well-known and widely studied non-classical logics is the so-called
Modal Logic, which is nowadays often regarded as extension of classical proposi-
tional calculus(CPC). However, there are other ways to formulate systems which
have the same theorems as that of well-known modal logics. Namely, we can
actually obtain such systems by rst regulating CPC and second giving some
appropriate denitions. This kind of results, recovering the original system
after restricting the system , are not new. Indeed, Lukasiewicz showed that
intuitionistic propositional calculus contains CPC and da Costa showed that
systems of paraconsistent logics C
n
(1 n < ) also contain CPC (cf. [3], [2]).
Now, the fact that modal logic S5 can be formulated by using weak nega-
tion instead of employing the classical negation and the necessity operator was
rst proved by Beziau and also by Waragai and Shidori independently(cf. [1],
[6]). As for the problem of formulating other systems from S5, it remained
open for a while until Marcos made an important contribution showing some
formulations for systems such as K, T, KB, K5, B, etc (cf. [4]).
Based on these studies, the purpose of the present paper is twofold. Firstly,
we extend and rene the results obtained by Marcos. This will include (1)
formulations of some systems, including S4, which have been open after the work
of Marcos and (2) an alternative economic formulation of systems treated by
Marcos in [4]. Secondly, we emphasize that the systems developed by Beziau and
Marcos can be seen as restriction of CPC. In order to make this viewpoint clear,
a framework which is sketched roughly in [5] will be introduced and discussed
in detail.
References:
1. Beziau, J. -Y., The Paraconsistent Logic ZA possible solution to Jaskowskis
71
problem, Logic and Logical Philosophy, vol.15, 99-111, 2006.
2. da Costa, N. C. A., On the Theory of Inconsistent Formal Systems, Notre
Dame Journal of Formal Logic, vol.15, No.4, 497-510, 1974.
3. Lukasiewicz, J., On the Intuitionistic Theory of Deduction, in L. Borkowski
(ed.) Jan Lukasiewicz Selected Works, North-Holland, 325-340, 1970.
4. Marcos, J., Nearly Every Normal Modal Logic is Paranormal, Logique et
Analyse, 48, 279-300, 2005.
5. Omori, H. & Waragai, T., On Beziaus Logic Z, Logic and Logical Philos-
ophy, vol.17, 305-320, 2008.
6. Waragai, T. & Shidori, T., A system of paraconsistent logic that has the
notion of behaving classically in terms of the law of double negation and
its relation to S5, in J. -Y. Beziau, W. A. Carnielli and D. Gabbay (eds.)
Handbook of Paraconsistency, College Publications, 177-187, 2007.
Andreas Pietz
University of Barcelona - Spain
[email protected]
Negation in Dual Intuitionistic Logic
Dave Ripley
Institut Jean-Nicod - France
[email protected]
Weak Negations and Neighborhood Semantics
In this talk, I explore the logic and semantics of weak negations: unary operators
validating only some of the inferences usually associated with negation. In par-
ticular, I am concerned with operators that are downward entailing (operators
N such that if A entails B, then NB entails NA); these seem to be particularly
important for linguistic theory. As such, I focus on a base system with a single
downward entailing operator and various strengthenings (requiring the operator
to validate more inferences tradionally associated with negation).
In the course of this investigation, I adapt model-theoretic techniques from
modal and substructural logic to provide a neighborhood semantics for these
systems. The systems are shown to be sound and complete with respect to
classes of frames, and inferences are shown to correspond to certain conditions on
frames. Neighborhood semantics allows for the exploration of weaker negations
than can be captured in a relational semantics like that pursued for negations
by Dunn and others. Thus, the neighborhood approach allows for an expansion
of Dunn and Zhou (2005)s map of various negations and negation-like operators.
Sebastian Sequoiah-Grayson
University of Leuven - Belgium
[email protected]
72
Negation as Test-Failure in DPL and Negation as Process Exclusion in Cate-
gorical Grammar
We give a dynamic semantics for the Lambek Calculi, to which a dynamic nega-
tion is added. This dynamic negation is interpreted as procedural prohibition,
or process exclusion. The resulting framework suggests connections with the
analysis of negation as testfailure in Dynamic Predicate Logic (DPL). The aim
of the present article is to explore this connection in detail.
Luca Tranchini
Siena University - Italy Tuebingen University - Germany
[email protected]
Negation and Refutation in proof-theoretic semantics
The intuitionistic informal account of negation, according to which a construc-
tion for the negation of A is a function mapping constructions of A onto con-
structions of the absurdity is somewhat unsatisfactory. This emerges in partic-
ular within the so-called proof-theoretic semantic framework, a development of
intuitionism on the basis of Gentzens work in proof-theory. In a truth-theoretic
semantics, truth values are assigned to formulas: in the case of closed formulas
(i.e. sentences) the semantics directly maps the sentence onto a truth value. In
the case of open formulas, this happens indirectly, i.e. modulo an assignment
of values to the open variables. A proof-theoretic semantics, on the other hand,
takes proofs as the semantics values to be assigned to valid derivations in a
deductive system. In a natural deduction setting, closed derivations are directly
mapped onto proofs while open derivations are mapped onto them given an
assignment of values for the open assumptions. When we come to negation,
we have that a closed derivation of the negation of A is constituted by an open
derivations having the absurdity as conclusion and A as only assumption, where
the absrudity is what there is no proof of. The problem lays in the fact that the
semantics is expected to map the open derivation onto a proof of the absurdity
given an assignment of proofs of the assumption A. But if the negation of A is
provable then of course there is no proof of A. The problem resembles the ones
that led to the introduction of free logics: an analogous solution in this case
would be a distinction between actual proofs and possible proofs. To avoid the
ad hoc character of this strategy we consider an alternative one: the introduc-
tion of refutations alongside proofs as possible semantics values of derivations.
We discuss a few alternative ways of doing this.
Susumu Yamasaki
Okayama University - Japan
[email protected]
Nonmonotonic Functions Caused by Distributive Negatives
A distributed system consisting of deductive processes with treatments of neg-
atives conceives a nonmonotonic function over set variables (denoting true and
false sets of atomic formulas). From model theory views, xpoint and proce-
dural semantics are well dened for logic programs with negation, as discussed
in Pereira,P.M. et al. (1991), Shepherdson,J.C. (1998) and Yamasaki,S. et al.
73
(2001), where the 3-valued logic is applied to the denotations of true and false
sets. The procedural semantics is closely related to computing based on the
rst-order logic such that the xpoint semantics is abstract enough to capture
the computational behaviour of programs. So far varieties of negations are si-
multaneously treated for logic programming, where multi-negation induces a
nonmonotonic function as in Yamasaki,S. (2006). In this talk, a general frame-
work is dened for a nonmonotonic function, caused by a distributed system
containing deductive derivations with negatives. The nonmonotonic function is
not so easy to analyze, when we want to have a mathematical representation of
the whole behaviour of the system in which the atomic formulas with and with-
out negation are both transferred from one deductive process to others through
communication. It is because the function cannot be in general monotonic even
by means of the 3-valued logic. For a general theory, a xpoint semantics can be
considered as model theory, however, a pragmatic, procedural semantics may be
reasonable with reference to model theory, in concretization of the framework
applied to a program.
4.2.8 Multimodal Logics
This session is organized by Walter Carnielli (CLE-UNICAMP, Brazil) and
Claudio Pizzi (University of Siena , Italy).
Contemporary modal logic, ocially born in 1932, received a powerful im-
pulse in the Sixties with the development of so-called relational semantics. After
this important turn modal logic underwent a constant, and indeed impressive,
progress passing through a specialized analysis of dierent concepts of necessity
and possibility and giving rise to such branches as tense logic, epistemic logic,
deontic logic, dynamic logic and so on.
The last step of this development has been provided by the growth of mul-
timodal logics, i.e. of logics whose language contains more than one primitive
modal operator and whose axioms dene the logical properties of each one of
them along with their interaction. Multimodal logic has already reached inter-
esting results in the abstract analysis of the properties of multimodal systems.
As a matter of fact, multimodal logic is not a new branch of modal logic but
rather a new way to study modal notions by using a more general and deep
approach, akin to the spirit of Universal Logic.
The aim of this session is to collect contributions to the eld of multimodal
logic. A wide number of subjects may be treated in this realm. Topics regarded
as being of special interest are the following:
1. Temporal logics
2. Logics of physical modalities
3. Epistemic-doxastic logics
4. Multimodal analysis of conditionality
5. Topological logics
74
6. Multimodal systems of mathematical provability
7. Multimodal systems with non-classical propositional basis
8. Combinations of (multi)modal systems
9. Incompleteness of multimodal systems
10. Decision procedures for multimodal systems
11. New semantics and proof methods for (multi)modal systems
12. Multimodal quantied logics
13. Modal treatments of quantication
14. Abstract properties of multimodal systems
Accepted contributed talks
Leandro Oliva Suguitani
Itala M.L. DOttaviano
State University of Campinas - Brazil
Hybrid Language and Relation Algebra
We shall take the development of modal logic and the theory of binary relations
up to a meeting point at which they are combined (by means of an hybrid
language and relation algebra (RA)) in a formalism which has full rst order
logic expressiveness. Our aim is discussing advantages and drawbacks of such
formalism for both as a logical system equipolent to rst order logic and as a
formalism with which one can formalize set theory and work on foundational
issues of mathematics.
1. Blackburn, P. and Seligman, J., 1995, Hybrid language, Journal of
Logic, Language and Information, 4, 252-272.
2. Marx, M., 2001, Relation Algebra with Binders, J. Logic Comp., v 11,
pp. 691-700.
3. Prior, A., 1967, Past, Present and Future. Oxford University Press, Ox-
ford.
4. Tarski, A. and Givant, S., 1987, A Formalization of Set Theory without
Variables, volume 41, AMS Colloquium publications, Providence, Rhode
Island.
Mario Benevides
Federal University of Rio de Janeiro - Brazil,
[email protected]
Propositional Dynamic Logic, Lb Axiom and Total Correctness
75
This work extends Propositional Dynamic Logic for Regular Programs with Lob
axiom in order to express total correctness. First, we extend PDL for regular
programs, without test, with Lob axiom and prove completeness with respect
to the class of well-founded programs. This ensures that all programs are well-
founded and consequently all computations sequences halt. Total correctness,
which could not be expressed in PDL for regular programs, can be expressed
by a PDL formula. Second, we add test and the predicate wf to PDL. We pro-
vide an axiomatization of PDL with test and the predicate wf and a proof of
completeness using ltration. Finally, properties about halting programs can
be expressed in PDL for regular programs in a neat way.
Juliana Bueno-Soler
University of S ao Paulo - Brazil
[email protected]
Description logics from a paraconsistent viewpoint
Walter Carnielli
State University of Campinas - Brazil
[email protected]
Juan C. Agudelo
Eafit University - Colombia
[email protected]
Polynomial ring calculus for S4, intuitionistic logic and multimodal logics
Polynomial ring calculi (PRC) consist in translating logic sentences into polyno-
mials over algebraically closed elds (usually nite elds) and into performing
deductions by means of polynomial operations. Elements of the eld represent
truth-values and polynomial equations express truth-conditions (in a similar
way as specifying conditions in valuation semantics). In this way PRC can be
legitimately considered a semantics, as well as a tool for performing deductions
by means of polynomial operations, establishing a proof method appropriated
for automation.
PRC were introduced in [2], where the method is applied to the classical
propositional calculus, many-valued logics and paraconsistent logics. In [1] a
PRC for the modal logic S5 is dened, and it is conjectured that such PRC
can be adapted to a large class of modal and multimodal logics. Our purpose
here is to adapt the PRC dened in [1] to the modal logic S4, providing a new
semantics and proof method not only for S4, but also for intuitionistic logic,
using the well-known Gdels translation (1933) of Heytings intuitionistic logic
into S4. We discuss some advantages of the PRC for modal and intuitionistic
logics in contrast with the traditional Kripke semantics and tableau proof meth-
ods. We also show how PRC can be extended to multimodal logics and to the
provability logic (called KGL in [3]). KGL has strong connections with Gdels
incompleteness theorems of 1931 and Lbs theorem of 1953. We conjecture that
the celebrated xed point theorem of KGL (D. de Jongh and G. Sambin, in-
dependently, in 1975) can even be proven by means of formal polynomials via
PRC.
76
1. Juan C. Agudelo and Walter Carnielli. Polynomial ring calculus for modal
logics: a new semantics and proof method for modalities. Available at
CLE e-Prints vol 9, n. 4, 2009. http://www.cle.unicamp.br/e-prints/vol
9,n 4,2009.html, 2009.
2. Walter A. Carnielli. Polynomial ring calculus for many-valued logics. In
B. Werner, editor, Proceedings of the 35th International Symposium on
Multiple-Valued Logic, pages 2025. IEEE Computer Society, 2005.
3. Walter A. Carnielli and Claudio Pizzi. Modalities and Multimodalities.
Amsterdam: Sringer, 2008.
Roberto Ciuni
Delft University of Technology - The Netherlands
[email protected]
Alberto Zanardo
University of Padova - Italy
[email protected]
A tense-modal logic with choices for agency and temporal distinctions
Truth in BTC logic (Branching Time logic with Choices) is relative to a moment
and to a particular choice (of a single agent) at that moment. BTC+ is an ex-
tension of BTC in which we can consider simultaneous choices of a set of agents.
According to [1] choices are viewed as sets of histories (paths) and, according
to [2] BTC and BTC+ semantics have both an Ockhamist and a Peircean di-
mension: a modal operator quanties over choices and tense operators quantify
within a given choice. As we are mainly interested in Computer Science appli-
cations, histories are assumed to be isomorphic to the set of natural numbers.
The paper is mainly concerned with the expressive power of BTC+, and shows
that CTL and a fragment of CTL* are expressed in the new logic.
1. Belnap Nuel and Perlo Michael and Xu Ming (2001) Facing the Future:
Agents and Choices in our Indeterminist World, Oxford, Oxford University
Press.
2. Zanardo Alberto (1998) Undivided and Indistinguishable Histories in Branching-
Time Logics. Journal of Logic, Language and Information, 7:297-315.
Katsuhiko Sano
Kyoto University - Japan
[email protected]
Axiomatizing hybrid products of topologies
The main aim of this paper is to propose a robust way to combine two topo-
logical hybrid logics. First, we introduce two kinds of: nominals: i (e.g. for
a moment of time) and a (e.g. for a spatial point), and satisfaction operators:
@
i
and @
a
to the bimodal logic of products of topologies [1] (note: product of
topologies is a dierent notion from product topology). Second, we give ve
77
interaction axioms and establish a general completeness result called pure com-
pleteness of bi -hybrid logic of products of topologies. Finally, we explain how to
capture the dependence of one dimension (e.g. space) to the other (e.g. time) in
a generalization of our setting. Our enrichment of the language was a suggested
further direction in [1]. This work can also be regarded as a further extension
of both hybrid logic of topological spaces [3] and bi-hybrid logic of products of
Kripke frames [2].
References:
1. Bezhanishvili G. Ten Cate B. Benthem, J. van and D. Sarenac. Mul-
timodal logics of products of topologies. Studia Logica, 84(3):369392,
2006.
2. K. Sano. Axiomatizing hybrid products. Submitted to Logic Journal of
the IGPL, Special Issue on Hybrid Logics. Available from http://www.
k2sn.com/hybrid_product2.pdf, 2009.
3. B. ten Cate and T. Litak. Topological perspective on the hybrid proof
rules. Electronic Notes in Theoretical Computer Science, 174(6):7994,
2007.
Newton M.Peron
State University of Campinas - Brazil
[email protected]
Marcelo E. Coniglio
State University of Campinas - Brazil
[email protected]
On necessary and possible contradictions
Despite several non-classical modal systems were proposed in the literature,
the method for obtaining them is not too much divergent: most of them are
the result of combining a propositional non-classical logic with a hierarchy of
(multi)modal systems. On the other hand, the denition of non-classical prim-
itive modal operator was investigated to a lesser extent.
Departing from paraconsistent logics (more precisely, from Logics of Formal
Inconsistency LFIs), and by analogy with the inconsistency operator, we
propose here a new unary modality, reading as necessary contradictory, and
its dual, reading as possible contradictory. From the semantical point of
view, we consider paraconsistent Kripke structures such that necessary con-
tradictory alpha holds in a state w whenever both alpha and not alpha hold
in every state w such that wRw. Analogously, possible contradictory holds
in w whenever there is some w such that wRw and both alpha and not alpha
hold in w.
We introduce a minimal normal system analogous to K with the new op-
erators. After this, we analyse the new version of axioms D, T, B, 4 and 5
suitable to the new modalities. For instance, by considering paraconsistent
78
Kripke structures where R is reexive, we shown that the suitable version of
axiom T is necessary contradictory alpha implies alpha and not alpha
The subtle discrepancies with the classical versions of such axioms can be
explained by translating the propositional modal logic into a rst-order LFI.
This shows the important role that an appropriate non-classical Correspondence
Theory would have for non-classical (multi)modal logics, mainly for systems
with non-classical primitive modalities.
Finally, we shown how to combine these systems with the classical modali-
ties necessary and possible and outline possible applications of such systems.
Claudio Pizzi,
University of Sienna - Italy
[email protected]
Contingency operators and multimodal logics
Fabien Schang
Dresden University of Technology - Germany
[email protected]
A structural semantics for multimodalities
A setting of multiple-valued logics is set forth within the general framework of
QAS (Question-Answer Semantics). Starting from the view that any formula is
a structured object whose components are a function symbol and its argument,
QAS is used in the following to reformulate two algebraic theories: Gottschalks
theory of quaternality, and Piagets theory of reversibility (INRC Group).
We propose to extend the process to a semantics for multimodalities, where
the arguments include at least one further function symbol. This will be done
with either of Gottschalks and Piagets questions, the content of which is not
the same. A structural semantics relies upon the use of two basic answer-values,
namely: armation (1), and denial (0, or -1). This general setting result gives
rise to a family of logics, where logical pluralism lies in the plurality of questions
Q to be asked among the variety of structured formulas.
Finally, the aim of our conceptual framework is to replace the basic notions
of truth and consequence by those of negation and opposition; but some chal-
lenges are still to be overcome in order to conrm its relevance, including the
reformulation of modal logics within QAS. It remains to see whether a multiple-
valued family of systems could throw some new light upon the awkward relation
between many-valued and modal systems: the latter challenge is reminiscent of
Dugundjis negative theorem, according to which no nitely many-valued ma-
trix can characterize the Lewis modal systems S1-S5.
References
1. Blanch, Robert: Sur lopposition des concepts, Theoria 19(1953), pp.
89-130.
2. Dugundji, James: Note on a property of matrices for Lewis and Lang-
79
fords Calculi of Propositions, The Journal of Symbolic Logic 5, pp. 150-
1.
3. Englebretsen, George: Knowledge, negation, and incompatibility, The
Journal of Philosophy 66 (1969), pp. 580-5.
4. Gottschalk, Wilhelm H.: The theory of quaternality, The Journal of
Symbolic Logic 18 (1953), pp. 193-6.
5. Piaget, Jean: Trait de logique (Essai de logistique opratoire), A. Colin
(2nd edition), 1972.
6. Smessaert, Hans: On the 3D visualisation of logical relations, Logica
Universalis 3, pp. 303-332.
Sara Negri
University of Helsinki - Finland
[email protected]
Laura Semini
University of Pisa - Italy
[email protected]
Specifying and reasoning with asynchrony: proof theory of DSL
To specify and reason on distributed system with asynchronous communications,
the use of a many-dimensional modal logic combining spatial and temporal op-
erators is fostered. This combination has an adequate expressive power, but
suers from the lack of a complete proof system, direct standard proof tech-
niques having failed, as well as those based on compositionality results.
The Labelled Sequent Calculi approach, introduced in [1], is proposed here
to dene a complete proof system for the logic. This calculus is a good candidate
since both admissibility of the structural rules and completeness are obtained in
a modular way: The same proof method works for all extensions and the addi-
tion of other modalities or frame properties requires only small local additions
to the proofs. As a rst step, the formalization of the spatial fragment in the
labelled sequent calculus is addressed.
1. S. Negri. Proof analysis in modal logic. Journal of Philosophical Logic,
vol.34, pp.507-544, 2005
Caroline Semmling
Dresden University of Technology - Germany
[email protected]
Interplay of Beliefs, Desires, and Actions
In the bdi-stit logic of beliefs, desires, intentions, and actions, we combine nor-
mal modal operators as in Stit Theory with standard monotonic, regular op-
erators for pro-attitudes and cognitive states. The emphasis of the talk is on
postulating axioms describing a philosophically interesting interplay between
80
beliefs, desires and actions of one and the same agent. On the other hand, we
suggest an independence condition for neighbourhood semantics to ensure the
independence of dierent agents concerning belief attitudes. Therefore, we re-
late properties of neighbourhood functions with the underlying Branching Time
structure. The completeness of the bdi-stit logic is shown by expanding the
canonical models of Xu for Stit Theory.
1. C. Semmling, H. Wansing, From BDI and stit to bdi-stit logic, Logic and
Logical Philosophy, 185-207, 2008.
2. M.Xu, Decidability of Deliberative Stit Theories with Multiple Agents, in:
D.M. Gabbay and H.J. Ohlbach (eds.), Temporal logic, First International
Conference, ICTL94, Springer-Verlag, Berlin, 332-348, 1994.
Dimiter Vakarelov
Sofia University - Bulgaria
[email protected]
Whiteheads theory of space and time
Whitehead is known not only as one of the authors of the famous book Prin-
cipia Mathematica. He also was an initiator of a new approach to the theory of
space and time based on some simple spatio-temporal relations between mate-
rial things taking as primitives [2]. Whitehead presented in [2] a quite detailed
program how to build a mathematical formalization of the spatial part of his
theory, known now as Region Based Theory of Space (RBTS) (see [1]). Unfortu-
nately, the part of his theory dealing with time (called in [2] Epochal Theory of
Time (ETT)) is developed only in an informal way, which makes quite dicult
to extract a good mathematical theory of space and time. In our talk we will
present an integrated mathematical formalization of a fragment of Whiteheads
theory of space and time incorporating in it some modern versions of RBTS
and some formal counterpart of ETT. The main result is a Stone-like represen-
tation theorem stating that any abstract system is representable in the class of
standard models.
1. D. Vakarelov, Region-Based Theory of Space: Algebras of Regions, Rep-
resentation Theory and Logics. In: Dov Gabbay et al.(Eds.) Mathemat-
ical Problems from Applied Logics. New Logics for the XXIst Century.
II. Springer, 2007, 267-348.
2. A. N. Whitehead, Process and Reality. New York, MacMillan, 1929.
4.2.9 Paraconsistent Logics
This session is organized by Alexandre Costa-Leite from the University of Brasilia,
Brazil. It is a session in honor of Casey Neil McGinnis (USA) (in memoriam).
81
Graham Priest - (Invited Speaker)
Universities of Melbourne (Australia) and St Andrews (UK), and
the Graduate Center, City University of New York (USA)
[email protected]
The catuskoti
In early Buddhist logic, it was standard to assume that for any state of aairs
there were four possibilities: that it held, that it did not, both, or neither. This
is the catuskoti (or tetralemma). Classical logicians have had a hard time mak-
ing sense of this, but it makes perfectly good sense in the semantics of various
paraconsistent logics, such as First Degree Entailment. Matters are more com-
plicated for later Buddhist thinkers, such as Nagarjuna, who appear to suggest
that none or these options, or more than one, may hold. The point of this talk
to to examine the matter, including the formal logical machinery that may be
appropriate.
Accepted contributed talks
Alexandre Costa-Leite
University of Brasilia - Brazil
[email protected]
Paraconsistentization: the general theory of paraconsistent logics
This talk describes a method to convert any logic into a paraconsistent logic.
This is a universal approach to the concept of paraconsistency. Instead of in-
vestigating particular systems of paraconsistent logics, we explore general and
abstract properties of these systems inspired by the idea of universal logic.
Soma Dutta
Indian Institute of Biosocial Research and Development - India
[email protected]
Negation in Paraconsistent Logic
Classically, the notion of consequence and inconsistency are interwoven. Could
any such connection be found in case of paraconsistent consequence? This is the
key motivator behind this study.In classical context, a formula and its negation
together yield every formula. This property of classical consequence is known
as the explosiveness condition. There is a general discord in accepting the idea
that without any relevancy an arbitrary formula can follow from a, a or
more generally, from an inconsistent set. Paraconsistent notion of consequence,
by denition, negates this explosiveness condition. That is, in paraconsistent
logics a set of formulae yielding a formula a and its negation a, does not nec-
essarily yield every . So what should be the proper way of addressing such a
set X which yields a pair of formulas a, a and does not yield at least one of
, . Intuitively, it seems to be plausible to say, X behaves inconsistently if one
considers a and behaves consistently if one considers . With this intuition, we
looked for sets of axioms characterizing the notion of consequence as well as
the notion of inconsistency in paraconsistent logic. While axiomatizing sets of
axioms for the notion of consequence as well as the notion of inconsistency of
82
paraconsistent logic by modifying already existing classical consequence axioms
as minimum as possible, a strange connection of double negation has come out.
In this paper we will propose this new notion viz. inconsistency with respect to
formula with a plausible axiomatization and explore the connection of double
negation (which emerged from the proposed axioms) with paraconsistent logics
in the existing literature.
References
1. Priest, Beall, Armour-Garb, The Law of Non-Contradiction, New Philo-
sophical Essays
2. Priest G., Routley R., A preliminary History of Paraconsistent and Di-
alethic Approaches
3. Carnielli W. A., Marcos J., A Taxonomy of C-systems
4. Chakraborty M. K., Dutta S., Axioms of paraconsistency and Paraconsis-
tent Consequence Relation, L. Magdalena, M. Ojeda-Aciego, J. L. Varde-
gay (eds.), Proceedings of IPMU08, pp 23 - 30, Torremolinos, Malaga,
Spain, June 22 - 27, 2008
Jean-Louis Hudry
University of Tartu - Estonia
[email protected]
Aristotle and the Principle of Non-Contradiction: Where are the Proofs?
If PNC is so strong a principle, why is it that Aristotles justications are so
weak? Priest (2006) claims that the weak justications illustrate the failure
of PNC as a logical principle. My paper aims to show that Aristotle would
have had no problem to agree with such a view. Lukasiewicz (1910) is the rst
to suggest a logical interpretation of Aristotles PNC (cf. Metaphysics). Yet,
it would have been more exact to restrict the interpretation to an ontological
PNC and an epistemic PNC. The ontological PNC asserts that it is impossible
for something to be both predicated and not predicated of a same thing at the
same time and in the same respect. As for the epistemic PNC, it amounts to
a belief in the ontological PNC, such that speakers believe that it is impossible
for a predication of things to be both true and false at the same time and in
the same respect. The necessary truth of the ontological PNC does not apply
to the belief itself, since a belief may be false. Aristotle is also aware that a
general proof of the epistemic PNC is circular: to prove that every belief in
the ontological PNC is true requires that speakers already accept the epistemic
PNC as true. Yet, it is possible to show that a belief rejecting the ontological
PNC is false. Aristotle speaks of refutation (elenchos), but adds that it is not a
proper proof, as it cannot be generalised to every belief. Indeed, the refutation
of a false belief involves a dialectical condition, namely a discussion between
at least two speakers. Therefore, Aristotle is ready to assert that meaningful
83
discourse requires an epistemic PNC; but he is quick to acknowledge that the
justications of the epistemic PNC are intrinsically weak, as they are nothing
more than individual refutations of false beliefs rejecting the ontological PNC.
Srecko Kovac
Institute of Philosophy - Croatia
[email protected]
Paraconsistent identities and modalities
The idea of modalizing the satisfaction of formulas is familiar in paraconsistent
logic. For instance, was interpreted by S. Jaskowski in his discussive logic
[?] (see also [?]) as 3. In J.-Y. Beziau [?, ?] the approach is generalized to a
specic four-valued logic, where the four values 0

, 0
+
, 1

and 1
+
are conceived
as necessarily false, possibly false, possibly true, and necessarily true,
respectively. Both approaches can be extended to a quantication theory with
identity.
Those theories are compared with a theory proposed in a continuation of [?],
which denes a class of paraconsistent rst-order models with an equivalence
relation as a paraconsistent identity, the classical identity being a subset
of . An equivalence class of objects at a world, [u]
w
, is dened as the set
u

[ u

w
u. The twofold satisfaction relation (positive, [=, and negative,
[=) is based on modally dened satisfaction of atomic formulas. Regarding the
satisfaction of identity formulas, u
1
and u
2
are identical at w i their respective
-counterparts at w are each others -counterparts at an accessible world w

:
M, w [=
v
t
1
= t
n
i (w

wRw

)(u

1
[u
1
]
w
)(u

2
[u
2
]
w
) u

1

w
u

2
We obtain as one of the consequences that identical thing(s) do not have to
share all their properties.
Hans Lycke
Gent University - Belgium
[email protected]
Paraconsistent Intuitionistic Logic
In standard intuitionistic logic INT, negation is proof theoretically characterized
by the inference rules reductio (RED) and ex falso quodlibet (EFQ). Because
of EFQ, the logic INT is explosive, which means that anything follows from a
contradiction. However, in view of the intuitive interpretation of intuitionistic
negation, i.e. the possibility to derive a contradiction, it is hard to see why
the logic INT should validate the inference rule EFQ. For, the construction of a
contradiction certainly doesn?t guarantee the construction of any formula what-
soever. An obvious solution to this problem consists in the overall rejection of
the inference rule EFQ. The logic resulting from this move is the paraconsistent
logic INTuN. However, the rejection of EFQ comes with a serious disadvantage,
for most applications of the INT?derivable inference rule reductio ad absurdum
(RAA) aren?t valid either. In view of the intuitive interpretation of intuition-
istic negation, most applications of RAA should be valid. In this paper, I will
84
characterize the adaptive logic INTuNr. The latter adds to the logic INTuN
all unproblematic applications of the inference rule EFQ (e.g. the applications
that lead to unproblematic applications of RAA). In this way, the logic INTuNr
doesn?t only avoid explosion, but also captures the intuitive meaning of intu-
itionistic negation.
Jos M. Mendez
University of Salamanca
[email protected]
Gemma Robles
Universidad de La Laguna - Spain
[email protected]
Francisco Salto
Universidad de Le on - Spain
[email protected]
Paraconsistent logics adequate to consistency understood as the absence of the
negation of any implicative theorem
In [1] two senses of a so-called weak-consistency are dened: consistency as
the absence of the negation of any theorem; and consistency as the absence of
the argument of any negation theorem.
Let us now dene a third sense of weak Consistency, which is the concept
of consistency the title of this paper refers to: consistency understood as the
absence of the negation of any implicative theorem
[Weak consistency in a third sense] Let S be a logic and T be a theory
built upon S. Then, T is w3-inconsistent (weak inconsistent in a third sense) i
(A B) T, A B being a theorem of S (a theory is w3-consistent weak
consistent in a third sense i it is not w3-inconsistent).
The aim of this paper is to dene a series of logics adequate to this sense of
consistency in the ternary relational semantics with a set of designated points.
These logics are said to be adequate to the concept of consistency in Denition
1 in the sense that the completeness proof can be carried out if consistency
is understood as stated in this denition. If consistency were understood in
the standard sense, the completeness proof would fail, at least in the present
semantical context, i.e., the ternary relational semantics with a set of designated
points. Now, let us dene:
[w3-paraconsistency] A logic S is w3-paraconsistent i the rule If A B,
then (A B) B is not a rule of S.
It will be proved that all logics in this paper are paraconsistent in the stan-
dard sense, but that none of them is w3-paraconsistent.
All logics are included in Lewis S4, some of them include classical proposi-
tional logic, but none of them is relevant.
A Routley-Meyer type ternary relational semantics, negation being modelled
with the Routley operator is provided for each one of these logics. Soundness
and completeness theorems are proved. In some cases, strong i.e., in respect
of deducibility soundness and completeness theorems are also proven.
85
Acknowledgements: Work supported by research projects FFI2008-05859/FISO
and FFI2008-01205/FISO, nanced by the MICINN (Spanish Ministry of Sci-
ence and Innovation). G. Robles is supported by Juan de la Cierva Program of
the MICINN.
References:
1. G. Robles, J. M. Mndez, The basic constructive logic for a weak sense
of consistency, Journal of Logic Language and Information, 17/1, pp.
89-107, 2008.
Dmitry Zaitsev
Moscow State University - Russia
[email protected]
Paraconsistent classical logic
Any system of relevant logic is known to be paraconsistent in a strict sense, that
is to be both atomic and molecular paraconsistent. In my talk, I will show how
paraconsistent hybrid logic can be generated by spoiling the system of the rst
degree relevant entailment (FDE).
Starting with Belnaps matrixes for negation, conjunction and disjunction I
spoil the semantical logic with classical implication. The resulting logic, call it
FDEP, is a hybrid of classical positive logic (TV+) with DeMorgan logic. An
adequate axiomatization for this logic will be presented.
4.2.10 Algebras for Logics
This session is organized by Joanna Grygiel from the University of Czestochowa,
Poland.
The use of algebra for the theory of reasoning was a fundamental turn in
the development of logic. It was the rst way to use mathematics to deal with
logic, a fundamental step towards mathematical logic.
In this special session dierent algebraic structures and algebraic operators
useful for the understanding of logic will be presented and discussed.
Accepted contributed talks
Carlos Gallardo
Universidad Nacional del Sur - Argentina
Alicia Ziliani
Universidad Nacional de San Juan - Argentina
Weak implication in the discriminator variety of m-generalized Lukasiweicz al-
gebras of order n
In 1940, Gr C Moisil introduced Lukasiewicz algebras of order 3 and 4 in or-
der to obtain the algebraic counterpart of the corresponding Lukasiewicz logics.
A year later, this author generalized these notions by dening Lukasiewicz al-
gebras of order n, and he studied them from an algebraic point of view. On
86
the other hand, in 1969 R. Cignoli indicated an equational denition of these
algebras equivalent to that given by Moisil. Ockham algebras wich are more
closed related to De Morgan algebras are the ones that satisfy the identity
f
2m
x = x, for some 1 m.The variety of these algebras will be denoted by
K
m,0
.As Lukasiewicz algebras of order n have a reduct which is a De Morgan
algebra J. Vaz De Carvalho and T. Almada generalized them by considering
algebras of the same type which have a reduct in K
m,0
. Hence, they introduced
the variety L
m
n
, 1 m, 2 n, of m generalized Lukasiewicz algebras of order
n(A generalization of the Lukasiewicz algebras, Studia Logica 69 (2001), 329-
338). We dene an implication operation on m generalized Lukasiewicz algebras
of order n, called weak implication, from which we obtain a new characteriza-
tion of the congruences on these algebras by means of certain special subsets of
them. Besides, we describe the principal congruences in a dierent way from
that indicated in the above mentioned paper. Finally, we prove that L
m
n
is a
discriminator variety.
1. J. Bermann, Distributive Lattices with an additional unary operation, Ae-
quationes Math. 16 (1977), 165-171.
2. V. Boisescu, A. Filipoiu, G. Georgescu and S. Rudeanu, Lukasiewicz-
Moisil Algebras, Annals of Discrete Mathematics 49, North-Holland, 1991.
3. T. Blyth and J. Varlet, Ockham Algebras, Oxford University Press, New
York, 1994.
4. R. Cignoli, Moisil Algebras, Notas de L ogica Matematica 27, Univ. Na-
cional del Sur, Baha Blanca, Argentina, 1970.
5. Gr. C. Moisil, Notes sur les logiques nonchrysippiennes, Ann. Sci. Univ.
Jassy 27 (1941), 8698.
6. Gr. C. Moisil, Sur les logiques de Lukasiewicz a un nombre ni de valeurs,
Rev. Roum. Math. Pures Appl. 9 (1964), 905920.
7. A. Monteiro, La semi-simplicite des algebres de Boole topologiques et les
systemes deductifs, Revista de la Uni on Matematica Argentina, Vol. 25
(1975), 417-448.
8. A. Urquhart, Distributive lattices with a dual homomorphic operation, Stu-
dia Logica 38 (1979), 201-209.
9. J. Vaz De Carvalho and T. Almada, A generalization of the Lukasiewicz
algebras, Studia Logica 69 (2001), 329-338.
Jan Ga luszka
Silesian University of Technology - Poland
[email protected]
Lattice of quasigroup formulas
87
The family of power terms xy
n
where n is a positive integer is inductively
dened as follows: xy
1
:= xy, xy
n
:= (xy
n1
)y. With these terms the following
family of formulas are associated: x, y xy
n
= x where n is a positive integer.
As a generalization of these formulas we propose the following torsion formula:
x, y n xy
n
= x. Using Steinitz numbers (for the original construction of these
numbers see [6], a slightly reformulated construction is proposed in [3] and in
[4]) we can introduce a family of new formulas as follows: n x, y n[s, xy
n
= x
where s is a Steinitz number. As associated with the torsion formula we have
the following family of formulas: x, y n n[s, xy
n
= x where s is a Steinitz
number.
Formulas described above are named one-sided quasigroup formulas (groupoids
satisfying at least one of them are one-sided quasigroups (see [2] and [4])). Evi-
dently there are groupoids being one-sided quasigroups satisfying none of them.
On the set of all formulas dening one-sided quasigroups we introduce a lattice
structure and prove the following theorem:
Theorem One-sided quasigroup formulas form a lattice isomorphic to the
lattice of closed Steinitz numbers with the divisibility relation.
References
[1] V.D. Belousov, Foundations of the Theory of Quasigroups and Loops
(Russian), Izdat. Nauka, Moscow, 1967.
[2] J. Galuszka, Groupoids with quasigroup and Latin square properties,
Discrete Math. 308 (2008), no. 24, 64146425.
[3] J. Galuszka, Codes of groupoids with one-sided quasigroup conditions,
Algebra Discrete Math. (2009), no. 2, 2744.
[4] J. Galuszka, Lattices of classes of groupoids with one-sided quasigroup
conditions, Algebra Discrete Math. (2010), no. 1.
[5] Quasigroups and loops: theory and applications, Heldermann, Berlin,
1990.
[6] E. Steinitz, Algebraische Theorie der K orper, J. reine angew. Math. 137
(1910), 167309.
Majid Alizadeh
University of Teheran - Iran
Hector Freytes
Argentinian Institute of Mathematics - Argentina
Antonio Ledda
University of Cagliari - Italy
[email protected]
Completion and amalgamation of bounded distributive quasi lattices
One of the basic motivations for studying the completion of a certain structure
is due to the need of lling the gaps of the original one; a leading example could
be turning a partial algebra into a total one. In the case of lattice ordered
structures, the most relevant examples are represented by canonical extensions
and Dedekind-MacNeille completions, see e.g. [2]. Nevertheless, once we move
from lattice ordered structures to quasi ordered ones, the classical approach
does not work. This observation motivated us in investigating a generalization
88
of the classical lter-based approach to the case of bounded distributive quasi
lattices (bdq-lattices), introduced by I. Chajda in [1]. The main problem in the
completion of bdq-lattices lies in the fact that it may happen for elements x, y
in a bdq-lattice L that x y, y x but x ,= y. In this case, the usual notion of
lattice lter is no longer useful to distinguish x and y. Thus, we use a particular
system of congruences which allows to obtain, out of any bdq-lattice L, a quasi
ordered space of functions E(L), , _), where is a topology on E(L) admitting
as a quotient the Priestley topology [5]. By this construction, we can embed the
original bdq-lattice L into a (functionally) complete one. As an application
of the previous results, we close the paper by proving, along the style of [4], the
amalgamation property for the class of bdq-lattices.
1. Chajda I., Lattices in quasiordered sets, Acta Univ. Palack Olumuc, 31,
1992, pp. 6-12.
2. Galatos N., Jipsen P., Kowalski T., Ono H., Residuated Lattices: An Al-
gebraic Glimpse at Substructural Logics, Volume 151, Elsevier, 2007.
3. Gumm H.P., Ursini A., Ideals in universal algebra, Algebra Universalis,
19, 1984, pp. 45-54.
4. Maksimova L. L., Craigs theorem in superintuitionistic logics and amal-
gamable varieties of pseudo-Boolean algebras, Algebra and Logic, 16,
1977, pp. 427-455.
5. Priestley H., Ordered sets and duality for distributive lattices, Annals
of Discrete Mathematics, 23, 1984, pp. 39-60.
Blaise Bleriot
Koguep Njionou
University of Yaounde 1 - Cameroon
[email protected] Some properties of fuzzy ideal of hyperlattice
The concept of fuzzy ideal on an algebraic structure is well-known in the liter-
ature, but so far mostly fuzzy ideals on hypergroups and hyperrings have been
studied, while the study of fuzzy ideals on hyperlattice has been neglected. In
[6] we have introduced the notion of fuzzy ideal of hyperlattice and established
some important properties. The notion of fuzzy prime ideal of a lattice has been
studied in [4] and the notion of prime fuzzy ideal of a lattice was studied in [8].
In this talk the following topics are considered:
1) Examples of fuzzy ideals of a hyperlattice. 2) Construction of fuzzy ideal
of a hyperlattice L induced by a fuzzy set of L. 3) Dierence between fuzzy
prime ideal and prime fuzzy ideal of hyperlattice and some properties of. 4)
Prove of fuzzy prime ideal theorem for hyperlattice.
References
[1] M. Attallah, Completely Fuzzy Prime Ideals of distributive Lattices, The
journal of Fuzzy Mathematics vol. 8, No. 1, 2000, Los Angeles, 151-156.
[2] P. Corsini, V. Leoreanu, Applications of Hyperstructure Theory, Dor-
drecht, Kluwer Academic Publisher, 2003.
89
[3] B.A. Davey and H.A. Priestley, Introduction to Lattices and order,
Second edition Cambridge (2002).
[4] T.K. Dutta and B.K. Biswas, On completely Fuzzy Semiprime ideals of
a Semiring, The journal of Fuzzy Mathematics, Vol 8, No. 3, 2000, 577-581.
[5] B.B. N. Koguep, C. Nkuimi and C. Lele, On fuzzy prime ideal of lattice,
SAMSA Journal of pure and applicable mathematics, Vol. 3, pp 1-11, 2008.
[6] B.B. N. Koguep, C. Nkuimi and C. Lele, On fuzzy prime ideal of hyper-
lattice, International Journal of Algebra, Vol. 2, no. 15, (2008) 739-750.
[7] A. Rahnamai-Barghi, The prime ideal theorem for distributive hyperlat-
tices, Ital. Journal of Pure and Applied Math., vol. 10, pp.75-78, 2001.
[8] U.M. Swany, D. Viswanadha Raju, Fuzzy ideals and congruences of Lat-
tices, Fuzzy Sets and Systemes 95 (1998) 249-253.
Joanna Grygiel
Jan Dlugosz University - Poland
[email protected]
On some characterization of distributive lattice
Aldo V.Figallo
Universidad Nacional del Sur - Argentina
[email protected]
Gustavo Pelaitay
Universidad Nacional de San Juan - Argentina
[email protected]
Claudia Sanza
Universidad Nacional de San Juan
[email protected]
Tense operators on symmetrical Heyting algebras
In 1942, Gr. C. Moisil introduced symmetrical Heyting algebras as Heyting
algebras with a dual involutive endomorphism. These algebras were investigated
by A. Monteiro [4] and later on, by H. P. Sankappanavar [5] and L. Iturrioz
[1]. In this paper, we dene and study tense symmetrical Heyting algebras (or
TSH-algebras) namely, symmetrical Heyting algebras endowed with two unary
operators. These algebras constitute a generalization of tense Boolean algebras
[3]. In particular, we obtain a topological duality for TSH-algebras. Besides, we
present two characterizations of the TSH-congruences, one of them by means
of the duality above mentioned.
References
1. L. Iturrioz, Alg`ebres de Heyting Involutives, Notas de L ogica Matematica
31, UNS, 1974.
2. L. Iturrioz, Symmetrical Heyting algebras with operators, Zeitschr. F.
Math. Logik und Grundlagen. D. Math. 29 (1983), 3370.
3. T. Kowalski, Varieties of tense algebras, Reports on Mathematical Logic
32(1998), 5396.
90
4. A. Monteiro, Sur les alg`ebres de Heyting Simetriques, Portugaliae Math.
39, 14(1980), 1237.
5. H.P. Sankappanavar, Heyting algebras with a dual endomorphism, Zeitschr.
F. Math. Logik und Grundlagen. D. Math. 33(1987) , 563-573.
Tomoyuki Suzuki
University of Leicester - UK
[email protected]
Canonical inequalities on FL-algebras
Canonicity of substructural logic, or more generally the canonicity of lattice ex-
pansions with non-smooth operations, has seen many contributions over the last
decades. However, when we consider non-smooth operations, e.g FL-algebras,
the canonicity results obtained from generalizing Jnsson-Tarskis methods run
into the problem of the existence of two types of extensions (sigma-extensions
and pi-extensions), which do not coincide in general. In this talk, we will show
how to harness Ghilardi and Melonis technique of parallel calculation (Ghi-
lardi and Meloni, 1997) to obtain new canonicity results for substructural logic.
The method will be presented in the light of the recent work (Dunn, Gehrke
and Palmigiano, 2005).
Rodolfo Ertola
Universidad Nacional del Sur - Argentina
Adriana Galli
Universidad Nacional de la Plata
Hern an San Martn
[email protected]
Non-compatible Operations in Heyting Algebras
Abstract: We study some non-compatible operations that can be dened using
the min operator in the context of a Heyting algebra. One example is the
minimum x-dense, that has been studied from a logical point of view in [2]. We
are interested, for instance, in interdenability and equationality. Regarding
logical questions, we focus on dierent axiomatizations and we consider if the
corresponding extensions of intuitionistic logic are conservative. Finally, we
study the relationship with the successor (see [1]).
References
[1] Caicedo, X. and Cignoli, R. An algebraic Approach to Intuitionistic Con-
nectives. Journal of Symbolic Logic. vol 66 (2001). pp. 1620-1636.
[2] Humberstone, L. The Pleasures of Anticipation: Enriching Intuitionistic
Logic. Journal of Philosophical Logic. vol. 30 (2001). pp. 395-438.
91
4.3 Contest
Contest: How to combine logics?
When we have two logics we may want to put them together. For example on
the one hand we have a temporal logic, on the other hand a deontic logic, how
to produce then a temporal deontic logic? This is a very interesting question in
the engineering of logic. People have been working in the subject since about
15 years. But there are still some fundamental problems not completely solved,
such as the collapsing problem. These problems are connected to the very
nature of what a logical system is. One may wonder if the intuitive denition of
combination of logic as the smallest conservative extension of two given logics
really works, and also if it is always possible to combine two logics.
References
1. J.-Y.beziau, A paradox in the combination of logics, in Workshop on Com-
bination of Logics: Theory and Applications, W.A.Carnielli, F.M.Dionisio
and P.Mateus (ed), IST, Lisbon, 2004, pp.75-78.
2. C. Caleiro and J. Ramos. From bring to cryptobring: a solution to the
collapsing problem. Logica Universalis, 1(1):7192, 2007.
3. W.A. Carnielli; M.E. Coniglio; D. Gabbay; P. Gouveia; C. Sernadas, Anal-
ysis and Synthesis of Logics. How to Cut and Paste Reasoning Systems,
Springer, 2008.
4. D.Gabbay, Fibring Logics, Oxford University Press, 1999.
How to take part in the contest?
All participants of the school and the congress are welcome to take part
in the contest. Send a short paper (10 to 15 pages) to [email protected]
before January 15th, 2010. The best ones will be selected for presentation at a
special session during the congress and the jury will then decide which, if any,
is the winner.
The jury is composed by Marcus Kracht, Dana Scott and Cristina Sernadas.
The price will be oered by Birkh auser.
Christoph Benzm uller
Articulate Software - USA
[email protected]
Simple type theory is well suited as framework for combining
Simple type theory is well suited as framework for combining classical and non-
classical logics. This claim is based on the observation that various prominent
logics, including (quantied) multimodal logics and intuitionistic logics, can be
elegantly embedded in simple type theory. Furthermore, simple type theory
is suciently expressive to model combinations of embedded logics and it has
a well understood semantics. O-the-shelf reasoning systems for simple type
theory exist that can be uniformly employed for reasoning within and about
92
combinations of logics.
Dominique Duval
University of Grenoble - France
[email protected]
How to combine diagrammatic logics
We claim that combining things, whatever these things are, is made easier if
these things can be seen as the objects of a category. We dene the category
of diagrammatic logics, so that categorical constructions can be used for com-
bining diagrammatic logics. As an example, a combination of logics using an
opbration is presented, in order to study computational side-eects due to the
evolution of the state during the execution of an imperative program.
Vladimir L. Vasyukov
Institute of Philosophy RAS - Russia
[email protected]
Combining Logics from the Point of View of Universal Logic
In number of papers concerning the issue of combining logics, in fact, one and
the same type of combination is exploited. But there are another types of com-
bination which as a rule lay beyond the scope of scholars. In the paper all these
types are considered in categorical setting. The categorical constructions in-
troduced allow to describe the nature and the structure of the general universe
of possible combinations of logical systems. It is shown that categorically such
universe turns out to be both a topos and a paraconsistent complement topos.
93
4.4 Contributed Talks
Musa Akrami
Islamic Azad University - Iran
[email protected]
Avicennian ontological reading of the principle of non-contradiction
The main position of the logician as such may be the belief that principle of non-
contradiction (PNC) refers in principle to propositions. Such an interpretation
is a minimal reading of non-contradiction. There are also positions that deal
with NCP metaphysically rather than semantically or logically.
Avicenna, as a peripatetic metaphysician-logician of Islamic era, gives some
remarks on NCP in his books on general metaphysics rather than in books on
logic. He follows Aristotles approach to show that non-contradiction is among
the occurrences (avarez) to the existent inasmuch as it exists (or existence qua
existence). According to Avicenna, truth (haqq) has several meanings, including
existence in external things, permanent existence, and the state of the verbal
statements or the state of the belief indicating the state of the external thing.
This Aristotelian metaphysical interpretation of NCP, accepted by Avicenna,
allows one to regard it as the most primary of all true statements. One should
accept it on the basis of some metaphysical analysis and insight, so that it is
the most evident of the self-evidents as a feature of existence. If existence qua
existence is the subject matter of metaphysics, it exhibits non-contradiction as
its most evident metaphysical feature.
In such an ontological reading of (PNC) in particular, and of logic in general,
logic and its classical foundations manifest a metaphysical feature, showing a
deep relation and correspondence between logic and ontology: logical entities,
relations and events exhibit a correspondence to ontological entities, relations
and events. As a corollary result of the research reected in the present paper,
any attempt to go toward universal logic must have an overlap with understand-
ing formal ontology as the science of all formal principles governing existence
qua existence and all entities belonging to various kinds of existence.
References
Aristotle (1984) Metaphysics, trans. W. D. Ross, revised by J. Barnes
(Princeton, NJ: Princeton University Press).
Avicenna, The Metaphysics of the Healing, Translated by Michael E. Mar-
mura, Utah: Brigham Young University Press, 2005.
Priest, G. et al. (2004) The Law of Non-contradiction Principle, Oxford
University Press.
Bernardo Alonso
Federal University of Rio de Janeiro - Brazil
University of Hertfordshire - UK
[email protected]
Information Overload and The Entailment Property in The Logic of Being In-
formed
Informational reading of KTB (Browers System), as proposed in Floridi 2006,
94
formalizes the relation of being informed. To be more precise, in that paper
it is argued that there is an information logic (IL), dierent from epistemic logic
(EL) and from doxastic logic (DL) (Hintikka, 1962), that formalizes the relation
a is informed that p (holds the information). This sense of being informed is
related to cognitive issues and to the logical analysis of an agents possession of a
belief or a piece of knowledge. In this paper we examine the process of combin-
ing a number of cognitively interpretable normal modal logics (NML) through
the combination of axioms that satises IL. The motivation behind rebuilding
Floridis task is to analyse in detail two main philosophical consequences: In-
formation overload and the dethroning of the Kp Bp principle (Girle, 2000).
We argue that these two consequences of IL pose a risk to theories about active
externalism, such as The Extended Mind (Chalmers and Clark, 1998), bringing
to light a weird kind of factual omniscience, with no logics that attribute such
epistemic divinity to their agents. It is claimed that, to avoid this unexpected
consequence, contents that were coupled to a cognitive system must have pre-
viously been consciously endorsed.
Doroteya Angelova
Bulgarian Academy of Sciences - Bulgaria
[email protected]
Paraconsistent and paracomplete logics: or how to treat logical pluralism
The aim of the talk is to be investigated some systems of paraconsistent and
paracomplete logics and their role for the defence of Beall and Restalls interpre-
tation of logical pluralism, namely that logic is both one and many [1:17]. This
standpoint is strongly embraced with the view that the core notion for logic is
logical consequence, i.e. the valid argument is one whose conclusion is true in
every case in which all its premises are true [2:23]. The basic idea is that logical
consequence is the stable kernel in the logical systems which entitle them to be
called logic (this circumstance denes the uniformity of logic, it is a reason to
be said that logic is one) and in the same time is a reason for their dierence
because the dierent logics present dierent explications of the mentioned cases.
In this regard I will try to defend Restall and Bealls position that the specic
cases, expressed by some paraconsistent and paracomplete logics, are the main
characteristics which present some of the most essential and fruitful variety of
logics. The reason is that they are connected with the precisication of conse-
quence the necessary requirement to form logic and are basic for the models the
sucient condition to work this logic. I will illustrate the above circumstances
regarding relevant logic and a couple of other paraconsistent and paracomplete
systems. In the same time I will argue against the critiques towards the above
view of logical pluralism.
References:
1. Beall, JC and Restall, G. (1999). Defending Logical Pluralism
in: http://consequently.org/writing/defplur/
2. Beall, JC and Restall, G. (2006). Logical Pluralism. Oxford: Oxford Uni-
versity Press.
95
Arnon Avron
Tel-Aviv University
[email protected]
Jonathan Ben-Naim
University Paul Sabatier - France
[email protected]
Beata Konikowska
Polish Academy of Sciences - Poland
[email protected]
Logics of Reasonable Information Sources
Extending the work of Belnap [1], we study structures which consist of a set of
information sources providing information about formulas of classical logic and
a processor collecting information from the sources and extending it using cer-
tain rules compliant with the truth tables of classical logic, which we call ESP
structures. In continuation of our previous work [2], characterizing general ESP
structures and the source-processor logics they generate, we now examine ESP
structures with reasonable information sources which provide coherent informa-
tion about formulas. We characterize the logic of a single reasonable source,
and prove that the logics generated by ESP structures with reasonable sources
coincide with the general source-processor logics described in [2]. However, we
show that there are processor valuations in general ESP structures which cannot
be obtained from any nite number of valuations dened by reasonable sources.
[1] A. Avron, J. Ben-Naim, and B. Konikowska, Processing Information from
a Set of Sources, in: D. Makinson, J. Malinowski, and H. Wansing eds., Towards
Mathematical Philosophy, Trends in Logic 28, 165 -186, Springer, 2009.
[2] N.D. Belnap, A useful four-valued logic, in: G.Epstein and J.M. Dunn
eds., Modern Uses of Multiple-Valued Logic, 7-37. Reidel, Dordrecht, 1977.
Hilan Bensusan
University of Brasilia - Brazil
[email protected]
Towards a universal logic of singularity
There are many dierent logical resources to express dierent kinds of singu-
larity: those that satisfy descriptions, those that are demonstrated, those that
are rigidly designated. Inspired by the approach taken by Universal Logic, we
can develop a general framework for all encounters with singular items of all
sorts a machinery that would enable us to express items in dierent patterns
of singularity. In fact, singularities can be expressed by several kinds of logi-
cal resources, but there are arguably common elements to those resources. In
particular, they work as some kind of glue between an expression and an item
capable either to bear it or to be specied by it. These common elements to
all kinds could start out with a break with all sorts of any: whatever is singular
contrasts with what is no more than an example at least by being an example
of too many things. In this work, Ill begin to explore what is common between
96
the various forms of expressing what is singular.
Mathieu Beirlaen
Ghent University - Belgium
[email protected]
Some Qualitative Adaptive Logics of Conrmation
Due to its denition in terms of classical logic, Hempels satisfaction criterion of
conrmation is unable to handle both incomplete data sets and data sets con-
taining irrelevant information. In order to adequately handle these problems,
a logic of conrmation must be non-monotonic. As an alternative to the sat-
isfaction criterion of conrmation, I will present an adaptive logic of induction
(devised by Diderik Batens) of which the consequence relation also serves as a
criterion of conrmation. A variant of this logic will allow us to conrm only
those hypotheses of which there is a positive instance in our data set (the notion
of positive instance can be traced back to the work of Nelson Goodman, and
provides us with an adequate solution for Hempels Raven Paradox). Each of
these logics can then be further extended to an even stronger logic of conrma-
tion by performing various minor operations on its respective denition. Thus
we become a whole family of logics of conrmation, each of which naturally
characterizes the dynamics of conrmation and provides us with new heuristic
tools for further research.
Piotr Blaszczyk
Pedagogical Universiyt of Krakow - Poland
About the point where Achilles catches up with the Tortoise
Francesca Boccuni
University of Padua - Italy
[email protected]
Plural logicism
PG (Plural Grundgesetze) is a predicative monadic second-order system which
is aimed to derive second-order Peano arithmetic. It exploits the notion of plural
quantication and a few Fregean devices, among which the infamous Basic Law
V. George Boolos plural semantics is replaced with Martinos Acts of Choice
Semantics (ACS), which is developed from the notion of arbitrary reference
in mathematical reasoning. Also, substitutional quantication is exploited to
interpret quantication into predicate position. ACS provides a form of logicism
which is radically alternative to Freges and which is grounded on the existence
of individuals rather than on the existence of concepts.
References
1. Martino, E. (2001). Arbitrary Reference in Mathematical Reasoning.
Topoi, 20, 65-77
2. Martino, E. (2004). Lupi, pecore e logica. (In M. Carrara and P. Giaretta
97
(Eds.), Filosoa e logica (pp.103-33). Catanzaro: Rubettino.)
Zbigniew Bonikowski
Opole University
[email protected]
Urszula Wybraniec-Skardowska
Poznan School of Banking - Poland
[email protected]
A Generalization of Certain Set-theoretical Operations
Rainhard Z. Benguez
Technology University of Munich
[email protected]
On G odels completeness theorem
In this contribution we will give a (constrained) constructive proof of Godels
Completeness-Theorem. For this we will dene a certain set of formulae and
derive a model which means that we are taking a semantic point of view (in
the tradition of Tarski). Additionally to our main goal we will gain a didactic
method of introducing, using and, therefore, understanding the semantic point
of view (terms, manner of speaking).
Selmer Bringsjord
Rensselaer Polytechnic Institute (RPI) - USA
[email protected]
Joshua Taylor
Rensselaer Polytechnic Institute (RPI) - USA
[email protected]
The cognitive path to universal Logic
One path to universal logic (and aspects and components thereof) is to focus
on logics themselves. We give specically the metaphor of the universal logi-
cian working in a library whose volumes are logics, where he or she reads and
processes the content in these books in various ways (merging them, etc.). The
cognitive path is dierent. It is based on an attempt to gure out how human
beings populated the library to begin with: How do we create logics in the
rst place? In words some may nd a bit more helpful: What is background
logic, formally speaking, and how do human cognizers problem-solve, reason,
and make decisions in and on the basis of background logic? An answer to
this question would, we claim, enable signicant progress in universal logic. We
report on our research intended to answer this question, and on a working com-
putational system - Slate - designed to assist humans reasoning in background
logic.
Ramon S. Capelle de Andrade
Itala M.L. DOttaviano
State University of Campinas - Brazil
98
[email protected]
[email protected]
Implication and Causation: a realistic perspective
According to Peirce (1958), regularities result from conditional nal causes un-
derstood as crystallised habits. A conditional sentence If A, then B is composed
of two clauses, the antecedent, or the if part, and the consequent, or the then
part. We understand the conditional sentence if A, then B in terms of a nal
cause. The antecedent A will be understood in terms of an ecient cause. So,
for example, the consequent activate the production of histidine will be nomo-
logically determined, given the presence of the state of aairs histidine less than
X, by the nal cause if the quantity of histidine is less than X, then activate the
production of histidine. The aim of this work is to defend the hypothesis accord-
ing to which the logical form If A, then B is present in the mode of expression
of a physico-chemical law, of a biological conditional and of a habit. But the
nomological connection between the antecedent A and the consequent B is not
the same, which leads us to relate dierent pairs of antecedents and consequents
with dierent types of implication. We will argue that physico-chemical laws,
which have strong nomological power, are compatible with strict implication
(the antecedent implies the consequent in all possible states of aairs). Biologi-
cal conditionals, which have moderate nomological power, are compatible with
material implication (the antecedent may be true and the consequent false).
Habits, which have weak nomological power, are compatible with relevant im-
plication (the antecedent may be true and the consequent false, but we need to
avoid vacuously true consequents to suppose that there is a causal connection
between the antecedent and the consequent).
References
1. ANDERSON, A.R.A.; BELNAP, N. D. Entailment: the logic of relevance
and necessity. Princeton: Princeton University Press, 1975.
2. ARMSTRONG, D. M. What is a law of nature? Cambridge: Cambridge
University Press, 1991.
3. DOTTAVIANO, I. M. L On the development of paraconsistent logic and
Da Costas work. The Journal of Non Classical Logic. V.7, 1990, p. 89-152.
4. PEIRCE, C. S. Collected Papers 6 vol. org. Harttshorne, c; Weiss,
Cambridge, Ma: Harvard University Press, 1958.
5. SHOENFIELD. J. R. Mathematical logic. Readings: Addison-Wesley,
1967.
6. WIENER, N. Cybernetics, or control and communication in animals and
machines. New York: Wiley, 1948.
Saloua Chatti
University of Tunis - Tunisia
99
[email protected]
Referential opacity and epistemic logic
Referential opacity is the failure of substitutivity of identity and in Quines view
of existential generalization as well. Quine thinks that no solution of opacity
is available in belief or modal contexts. But epistemic logicians like Hintikka
and Lenzen think that referential opacity can be solved since it is due not to
absence of reference but to plurality of references: its solution is provided by
stabilizing the reference in belief and knowledge contexts. However, stabilizing
the reference needs in Hintikkas frame pragmatic methods of identication while
Lenzen prefers using Kripkes frame and rigid designation. My aim in this paper
is to analyze these solutions in order to show if and how they provide answers
to Quines criticisms. I will then compare between these options and answer
these questions (among others): How can one identify individuals? Under what
conditions may one use exportation ? How is the De Re / De Dicto distinction
treated? [De Re: (Ex) Ka (x=b); De Dicto: Ka (Ex)(x=b)]. By answering these
questions, I will show that Hintikkas approach keeps all its actuality since rigid
designation is not always warranted and that we can preserve the De Re/De
Dicto distinction, i.e. the singular/general distinction which is crucial in belief
and knowledge contexts.
Bibliography:
1. Hintikka, J: Knowledge and belief: An introduction to the logic of the
two notions, prepared by V.Hendricks and J.Symons, Cornell University
press, 2005
2. Lenzen, W: Epistemic logic, in Handbook of epistemology, I. Niniiluoto,
M. Sintonen,and J Wolenski (eds), Kluwer 2002
3. Quine: Quantiers and propositional attitudes, in The Ways of paradox
and other essays, Harvard University Press,,revised edition 1976
Shih-Hsun Chen
National Chung-Cheng University - Taiwan
On the Appropriateness of Taking RCS(R3) as Domain of Region Why ROS(R3)
is a good model of regions?
This paper is a note on treating ROS(R3) as a model of regions. Ian Pratt-
Hartmann proposes taking ROS(R3) as a good model of regions in[1]. First of
all, there are two meanings of mereotopology dened by Ian Pratt-Hartmann:
(1) we can see mereotopology as a model of formal language of Boolean alge-
bra, and (2) we can see mereotopology as a model of rst-order language which
describing regions. I try to point out the dierence here, since the existence of re-
gion and that of boundary are dierent.We can easily construct another Boolean
algebraic structure (in this structure, region contains its boundary) which is iso-
morphic to mereotopology, however, we could also nd that this structure will
confront some problems as being a model of rst-order language which describes
regions. Second, I will explain the reason why Ian Pratt-Hartmann thinks that
100
ROS(R3)(which is a mereotopology) is a good model of regions, moreover, I sup-
ply some proofs which Ian Pratt-Hartmann omits, such that readers interested
in mereotopoloty can realize easily.
References:
1. Pratt-Hartmann, I., Fisrt-Order Mereotopology: in Handbook of Spatial
Logics Aiello, M, I. Pratt-Hartmann and J. van Benthem ed. Dordrecht:
Springer. pp.13-97, (2007)
2. van den Dries, L., Tame Topology and O-Minimal Structures, volume 248
of London Mathematical Society Lecture Note Series. Cambridge Univer-
sity Press, Cambridge. (1998)
Yue Chen
Ray Jennings
J.Sahasrabudhe
Simon Fraser University - Canada
[email protected]
[email protected]
[email protected]
Hypergraph as a semantic idiom
A new semantic idiom is presented in the paper, one that uses hypergraphs
as the semantic representatives of sentences of a propositional language. A
standard hypergraph H on a base set X is a collection of hyperedges that are
subsets of X. We trace a short history of this idiom, one that focuses on a spe-
cial kind of hypergraph whose base set is the power set p(U) of the universe
of a model. Power-set hypergraphs serve as a unifying semantic idiom that
combines various previously known semantics for the rst degree entailment of
the system E of entailment. In this idiom, an ordering relation between simple
(that is, inclusion-disordered) hypergraphs models the entailment relation. Such
an entailment relation preserves chromaticity of the hypergraphs. Hypergraph
therefore aords a richer idiom for semantic investigation than that of truth and
meaning in standard semantics. As the ordering relation between simple hyper-
graphs varies, the comparatively rich hypergraph-theoretic idiom also reveals
other, hitherto unstudied, systems. Furthermore, we show that any distribu-
tive lattice can be represented as a hypergraph lattice, whose expressive power
therefore can hardly be exaggerated. These hypergraph structures are instances
of more abstract structures amenable to algebraic methods. We can adopt as
base structures, abstract algebraic structures more general than the Boolean al-
gebra of subsets, for example, distributive lattices of various kinds, and usefully
generalize the notion of simple hypergraph itself, as an antichain on a lattice.
Taking this lead, in the nal portion of the paper we generalize hypergraphs to
a more abstract setting. Here we show that the language of hypergraphs is quite
powerful and give a characterization of boolean lattices in this language. An
examination of the properties of lattice hypergraph not only generates familiar
101
logic systems previously interpreted by more cumbrous semantic structures, but
also aords new insights into connections between hypergraphs and lattices.
Few bibliographical references:
[1] A. R. Anderson and N. D. Belnap. Tautological entailments. Philosoph-
ical Studies, 13:924, 1961. [2] N. Belnap. How a computer should think. In
Contemporary Aspects of Philosophy, pages 3056, 1977. [3] N. D. Belnap. In-
tensional models for rst degree formulas. Journal of Symbolic Logic, 32:122,
1967. [4] B. Brown. Lp, fde and ambiguity. In Proceedings of the 2001 Internal
Conference on Articial Intelligence, 2001. [5] R. E. Jennings and Y. Chen. Ar-
ticular models for rst degree entailment. forthcoming, 2009. [6] R. E. Jennings
and Y. Chen. Articular models for rst degree paraconsistent systems. forth-
coming, 2009. [7] J. M. Dunn. Intuitive semantics for rst-degree entailments
and coupled trees. Philosophical Studies, 29:149168, 1976. [8] J. Hintikka. Ex-
istential presuppositions and existential commitments. Journal of Philosophy,
56:12537, 1959. [9] R. C. Je rey. Formal Logic: Its Scope and Limits. New
York: McGraw-Hill Book Company, 1967. [10] W. T. Parry. Ein axiomsystem
fur eine neue art von implication (analytische implication). Ergebrisse eines
Mathematischen Colloquiums, 4:56, 1933. [11] R. Routley and V. Routley. The
semantics of rst degree entailment. nous, 6:335359, 1972.
Leonardo Cabrer
Norbert Preining
Japanese Advanced Institute of Science and Technology - Japan
[email protected]
[email protected]
The semantic hierarchies of First Order Gdel Logics - algebra versus Kripke
frames
The class of (rst order) Godel logics admits several layers of semantics. The
most restrictive being semantics based on the Reals, the most permissive being
Godel (L-) algebras, with linear Kripke frames with constant domains between
those two.
It has been shown that if one considers the generated logics as comparing
factors, the subset of countable Kripke frames corresponds exactely to the real
valued semantics, i.e., for each countable linear Kripke frame with constant
domains there is a truth value set such that the respective logics coincide, and
for each truth value set there is a countable linear Kripke frame with constant
domains such that the respective logics coincide.
While both the study of logics based on a single truth value set or Kripke
frame has been undertaken since about 20 years by various groups (e.g., [1,2]), in
the algebraic case generally only logics dened by all or a large class of algebras
has been studied.
The current work is targetted at pinpointing the sub-class of Godel alge-
bras that allow a similar representation result w.r.t. the set of linearly ordered
Kripke frames with constant domains. Furthermore, we discuss the possibility
to represent logics of single Godel algebras as intersections of logics dened by
Kripke frames.
102
[1] M. Takano, Ordered sets R and Q as bases of Kripke models. Studia
Logica 46:137-148, 1987. [2] N. Preining, Godel logics and Cantor-Bendixon
Analysis. Proceedings of LPAR 2002, LNAI 2514, 327-336.
Jean-Pierre Descl`es
University of Paris-Sorbonne - France
[email protected]
Anca Christine Pascu
Universit de Bretagne Occidentale - France
[email protected]
How to articulate extension with intension and objects with concepts
From a logical viewpoint, object is never dened, even by a negative deni-
tion. This paper is a theoretical contribution about object by the way of a new
constructivist logical approach called Logic of Determination of Objects (LDO)
founded on a basic operation, called determination. This new logics takes in
account cognitive problems as heritage of properties by non typical occurrences
or by indeterminate atypical objects in opposition to prototypes that are typ-
ical completely determinate objects. We show how are dened and organized
extensional classes, intensions, more and less determined objects, more or less
typical representative of a concept, prototype, using a determination operation
that constructs a class of indeterminate objects from an object representation
of a concept called typical object.
The aim of the LDO is to provide conceptual and logical answers to the
following questions and issues: What is a typical instance? What is an atypical
instance? What is a prototypical instance? How does one dene a typicality
relation between instances without using a degree of membership? Articulate
intension with extension? Dene family resemblances? Manage exceptions?
Manage prototypical properties? Manage multiple inheritances? To study these
problems, we adopt the general framework of Combinatory Logic (CL) (Curry,
Combinatory Logic, 1958) with types in which it becomes possible to dene
not only operators of predication (predicates), but also operators of determina-
tion of objects, and operators for building indeterminate objects from concepts,
as well as explicitly articulating, for given concepts, the structured intension
of concepts with the structured extension of concepts without taking into
consideration that intension is dened only by duality with the extension.
The LDO is a non-classical logic of construction of objects. It contains a
theory of typicality and a extended system of quantication.
Elena Dragalina-Chernaya
Moscow Higher School of Economics - Russia
The binary quantiers perspective on logicality
My purpose is to interpret abstract logics as formal ontologies, i.e. as genuine
logics at least in phenomenological sense. My proposal is to consider classes
of isomorphism as model-theoretic analogues of categorical objects of Husserls
formal region. Logic has no ontology, but logic is formal ontology. Some prin-
ciples of demarcation of the bounds of logic as formal ontology are discussed.
103
Although Tarskis philosophical generalization of his permutation invariance cri-
terion - our logic is logic of cardinality - appeared to be justied by the theory
of monadic quantication (logic of properties of classes of individuals), it is not
correct for the theory of binary quantication (logic of properties of classes of
pairs of individuals). The point is that heterogeneous quantier prexes con-
sidered as binary quantiers distinguish equicardinal relations. Thus not only
cardinalities, but also patterns of ordering of the universe have to be taken into
account by logic with binary quantiers.
Catarina Dutilh-Novaes
University of Amsterdam - The Netherlands
[email protected]
The undergeneration of permutation invariance as a criterion for logicality
Permutation-invariance as a criterion for logicality is a much-discussed topic in
the recent and not-so-recent literature. It has its enthusiastic defenders (Sher,
Bonnay), but it has also met with considerable resistance. The most frequent
charge of inadequacy is based on the claim that the criterion overgenerates, i.e.
that it counts as logical notions that are arguably non-logical, in particular nu-
merical notions. In fact, in his 1966 lecture What are logical notions? Tarski had
already observed that it turns out that our logic [based on permutation invari-
ance] is even less than a logic of extension, it is a logic of number, of numerical
relations. (p.151) But what seems to be equally signicant, and yet scarcely
discussed, is the fact that the criterion also appears to undergenerate in light of
the developments in logic of the last decades. At the time of Tarskis lecture, the
logical systems being studied were still those which had been developed against
the background of the logicist program, and as the purpose of logic according
to the logicist program was to provide foundations to arithmetic, it was to be
expected that these logical systems would be particularly sensitive to matters
concerning cardinality and numbers. However, since then logic has developed
in a variety of new directions, and its interface with computer science is partic-
ularly signicant. Indeed, many of the notions and operators that are currently
considered to be logical do not satisfy Tarskis permutation invariance criterion;
this is the case for example of any non-S5 modal operator if interpreted on
a Kripke-semantics (only the S5 modal operators satisfy the criterion). In my
talk, I look into cases of undergeneration of the permutation-invariance criterion
and discuss their signicance for the matter of the true nature and scope of logic.
Oscar Esquisabel
National University of La Plata - Argentina
[email protected]
Leibniz: mathesis universalis, logic and science of forms
There is a received talk about the Leibnizian Mathesis Universalis as a kind
of general or universal calculus. Our intention is not to contest or to reverse
this interpretation, which has a long tradition imposed by the use, but rather
to analyze more in detail the problems that are connected with the Leibnizs
own conception about a Mathesis Universalis. Our objective here is not to get
104
into the details of this Leibnizian conception, but to characterize the status
that Leibniz gives to the Mathesis Universalis, and, if possible, to delimitate
the reach of such discipline.
We will try to defend the idea that the Mathesis Universalis has a variable
status. In some point of its intellectual development, Leibniz conceived it as
a structural mathematical science not limited to quantity exclusively, therefore
he included the science of forms (named too the science of similar and dis-
similar) within its domain. Later, Leibniz retracted from that inclusion and
limited the Mathesis Universalis to the domain of quantity. In this way, the
Mathesis Universalis became practically identical to algebra, although it was
an extended algebra, as we have said beforehand. On the other hand, the
structural-qualitative aspect became the object of the sciences of forms, which
turned into a subordinating science in relation to the Mathesis Universalis. The
reasons for these variations are probably due to changes in points of view about
the reach of the science of forms, the conception of the concept of number and
the way of understanding the relation between arithmetic and geometry.
1. Sasaki, Chikara, Descartes Mathematical Thought, Dordrecht, Kluwer,
2003
2. Van Schooten, Principia Matheseos Universalis, seu Introductio ad Ge-
ometriae Methodum Renati Des Cartes, 1651.
3. Wallis, Mathesis Universalis, sive Arithmeticum Opus Intgrum, 1649; New-
ton, Arithmetica Universalis and Mathesis Universalis Specimina, manuscripts
from 1648.
4. Leibniz, G.W. Mathematische Schriften, ed. por C.I. Gerhardt, 7 vols.,
Berlin-Londres 1849-1863 (GM)
5. Leibniz, G.W., Smtliche Schriften und Briefe. Herausgegeben von der
Deutschen Akademie der Wissenschaften, Darmstadt, 1923; Leipzig, 1938;
Berlin, 1950 = A.
6. Rabouin, David, Mathesis Universalis. Lide de mathmatique universelle
dAristote Descartes, Paris, PUF, 2009.
Daniel G ain a
University of Bucharest
[email protected]
Omitting Types Theorem Institutionally
We present a generalization of forcing in institution-independent model theory
which is used to prove an abstract omitting types theorem applicable to many
rst-order logics, which are, roughly speaking, logics whose sentences can be
constructed from atomic formulae by means of boolean connectives and nitary
rst-order quantiers. These include rst-order logic (FOL), logic of order-
sorted algebra (OSA), pre-order algebra (POA), partial algebras (PA)as well
105
as their innitary variants FOL
1,
, OSA
1,
, POA
1,
, PA
1,
.
Luiz Estrada-Gonzalez
Autonomous University of Morelos State - Mexico,
[email protected]
On the necessary properties of logical consequence
The question I am concerned with in this talk is whether the logical consequence
relation (henceforth logical consequence for short) must satisfy necessarily some
properties. In several papers Jean-Yves beziau has put forward an argument
to show that logical consequence needs no satisfy any principles. I will call
this the argument of induction-practice-analogy and can be summarized as
follows. Over the last one hundred years virtually every theorem, principle for
connectives, principle for the consequence relation, etc. has been thrown out.
[Now nearly every structure resulting for dropping such principles is accepted
as logic.] This suggests that, in spite of the appearances, logic is not grounded
on any principles or laws. This is much like the situation in algebra, where an
algebra is dened as a collection of operations on a structure with no additional
constraints on those operations. The shortcomings of this argument are those of
any inductive and analogy argument. On one hand one has to be careful about
how close the analogy between logic and algebra is. On the other hand, we
have succeeded in dropping certain principles, theorems, rules, etc. but there
are problematic cases. From the fact that we can do without monotonicity of
logical consequence it does not follow that the result will be the same if we
drop reexivity or transitivity. The last contention is exactly Beall and Re-
stalls in their book Logical Pluralism, since for them, from the analysis of the
very notion of logical consequence, being it in the business of truth preservation
(from premises to conclusions), logical consequence is transitive and reexive,
for truth preservation is indeed a reexive and transitive relation. I am ready to
grant that truth preservation is reexive and transitive, but what I contend is
Beall and Restalls characterization of the pretheoretical notion of logical con-
sequence. More accurately, I contend the uniqueness of such a characterization.
There are other characterizations of logical consequence, like the well-known
q-consequence, but there are at least other two notions of logical consequence,
one of them less known but already present in the literature, pconsequence,
and other of which I have no notice and that I will dub r-consequence. Under
those alternative characterizations logical consequence might not be reexive or
transitive, and this supports beziaus intuition that logicality lies beyond any
principles or laws.
Katarzyna Gan-Krzywoszynska
University of Nancy 2 - France
Piotr Lesniewski
Adam Mickiewicz University - Poland
[email protected]
[email protected]
Vuillemin between two Lukasiewiczs answers
106
Hao Wang said once There is more philosophical value in placing things in
their right perspective than in solving specic problems?. Let us emphasize the
phrase more philosophical value. It is taken for granted that the fundamental
(philosophical) question What is logic? (the question Q for short) is a prop-
erly (i.e. well-posed) question. Is that so? Following Ajdukiewiczs pragmatic
logic we introduce some modied concepts of a positive (negative) assumption
of a given question. (Although the so called an assumption based conception of
presupposition is no longer apt. We are aware well of that David I. Beaver?s
belief.) We use the term assumption as more neutral than the term presupposi-
tion; obviously, the latter is worked out with great care and nicety of detail not
only in basic systems of erotetic logic. A question whose positive or negative
assumption is not true is called an improperly posed question. It is our consid-
ered opinion that the question Q could be posed in the contexts of progressive
structures based on Roman Suszkos diachronic logic in particular. For the sake
of brevity we call them progressive contexts. It goes without saying that the
non-reductionistic approach to questions is adopted. Two dierent (and both
famous) Lukasiewiczs answers to the question Q are expanded within broader
scope of philosophical systems classication by Jules Vuillemin. Such an inter-
action between the principal trend in universal logic (that leads from logical
matrices to abstract logics) and philosophical studies of assumptions goes far
beyond the very idea of practical turn in logic (in the Gabbay/Woods sense).
Juan Luis Gastaldi
University of Bordeaux 3 - France
[email protected]
Logic and Philosophy in the work of Kant and Hegel: The philosophical condi-
tions for a general logic
German Idealism, which dominated the philosophical scene in the 19th century,
appears in the work of its two major gures, Kant and Hegel, as the endeav-
our to build and establish a general logic, to the determinations of which the
classical metaphysical and ontological problems should be reduced. Although
these attempts have scarcely contributed to the positive development of formal
logic, it is possible to argue that they created the conditions that made pos-
sible the expansion of Freges works impact, thus allowing his essential logical
achievement to also become a major philosophical event that played a critical
role in the birth of contemporary philosophy. This event should be seen as
the expression of the philosophical generality or universality that the new log-
ical achievements were able to claim. From a detailed analysis of both Kants
Transcendental Logic and Hegels Dialect, focused on the distinction between
conceptual determination and formal determination, this contribution will try
to characterize this particular kind of universality that since the 19th century
philosophy has expected from logic. Its aim is to propose a discussion on the
conditions a logic should satisfy to be considered as general or universal from a
philosophical point of view, as well as on the conditions that logic is, in return,
allowed to impose on the philosophical discourse.
107
Yvon Gauthier
University of Montreal - Canada
[email protected]
Modular polynomial logic
Modular polynomial logic is an arithmetical logic designed to produce a syn-
tactic proof of the self-consistency of arithmetic. Arithmetic here is not the
Dedekind-Peano set-theoretic arithmetic, but Fermat-Kronecker classical arith-
metic or number theory with Fermats innite descent replacing Peanos induction
postulate and Kroneckers forms (homogeneous polynomials) playing the rle of
generalized integers. Fermats method of innite descent following Andr Weil
it is really a nite descent in nite elds is a constructive method in number
theory and is not equivalent to complete or transnite induction from a construc-
tivist point of view (see 1,2,3) The formalization of innite descent introduces
non-classical or non-standard logical concepts that challenge the consistency of
Peano arithmetic. According to Greg Restalls characterization, a paraconsis-
tent logic is an inconsistency-tolerant logic; provided that the inconsistencies
in question are located on the side of classical logic and standard arthmetic,
modular polynomial logic is an absolutely paraconsitent logic in the sense that
inconsistencies are everywhere else.
References
1. Ga Gauthier, Y., Internal. Logic. Foundations of Mathematics from Kro-
necker to Hilbert, Kluwer, Synthese Library, Dordrecht/Boston/London,
2002.
2. Gauthier, Y. Classical Function Theory and Applied Proof Theory ,
International Journal of Pure and Applied Mathematics, Volume 56, no.
2 (2009), 223-233.
Emmanuel Genot
Shahid Rahman
University of Lille 3 - France
[email protected]
[email protected]
A Dialogical Semantics for Interrogative Consequence
Interrogative consequence has been presented by (Hintikka et al. 1999), as an
extension of classical consequence, in order to account for empirical reason-
ing as a variety of deductive reasoning with interrogative steps. Interrogative
tableaux systems have also been used, and were given a game-theoretic seman-
tics, reconstructing interrogative reasoning as a game against Nature. However,
their puzzling proof theoretic aspects have never been thoroughly discussed, and
the semantics oered is unsatisfactory on many aspects. We discuss the seman-
tics oered by (Harris 1994), and show why it has to be improved. We then
propose a dialogical reconstruction of interrogative logic, and we establish that
proof-theoretically, the core system captures Anderson and Belnaps rst degree
entailment, recovering classical logic by ad hoc closure rules. We show that the
108
dialogical semantics improves upon Harris game-theoretic interpretation, and
discuss the relation between this dialogic, classical game-theoretic semantics (for
non-interrogative consequence) and dialogical semantics in general, with a par-
ticular attention to the interpretation of negation. We conclude by oering some
comments on the modal interpretation of interrogative consequencesince the
conclusion of an interrogative argument is, according to Hintikka, known
and its extension to cover nonmonotonic reasoning.
Bibliography
1. Genot, E.J., 2009, The Game of Inquiry: the interrogative approach of
inquiry and belief revision theory, Synthese, 171.
2. Harris, S, 1994, GTS and Interrogative Tableaux, Synthese, 99
3. Hintikka, J, Halonen, and A Mutanen, 1999, Interrogative Logic as a
general theory of reasoning, in J. Hintikka, 1999, Inquiry as Inquiry,
Kluwer.
Anderson, A. R. and Belnap, N. D., 1975, Entailment: The Logic of Rel-
evance and Necessity, Princeton University Press, vol. I
Katarzyna Grygiel
Jagiellonian University - Poland
[email protected]
Lambda calculus vs combinatory logic: quantitative approach
In the beginning of the 20th century two new models of computations were in-
troduced: lambda calculus and combinatory logic. It is a well known fact that
they are equivalent in question of expression power. It turns out, however, that
a typical lambda calculus term and a typical combinatory logic term are very
unlike. Due to the quantitative appoach towards those models we have shown
that not only the structue, but also the semantics of typical terms dier radi-
cally. Using the notion of density we can dene typical terms. By looking at
them as at some special graph theoretical objects and appylying combinatorial
methods we can say how typical terms look like and on that basis claim some
facts about their semantics (e.g. strong normalisation). Joint work with: Rene
David, Jakub Kozik, Christophe Raalli, Guillaume Theyssier, Marek Zaionc.
Raffaela Giovagnoli
University of Rome 2 - Italy
[email protected]
From Single to Relational Scoreboards
I move from some problems concerning the conception of single scoreboard in
conversation grounded on a shared epistemic standard (David Lewis). Score
evolves, like baseball, in a more-or-less rule-governed way: If at time t the con-
versational score is s, and if between time t and time t the course of conversation
is c, then at time t the score is s, where s is determined in a certain way by s
and c. Robert Brandom modies Lewis single scoreboard: scorekeeping entails
109
that each interlocutor is assigned a dierent score. For to each, at each stage
of conversation, dierent commitments and dierent entitlements are assigned.
Ill describe some formal steps moving also from Mark Lances and McFarlanes
views of scoreboard in conversation to arrive at a plausible scheme that consid-
ers deontic statuses and deontic attitudes.
Paul Healey
London - UK
[email protected]
Understandings of logic sublated by the dialectic
Empirical, Formal and Speculative understandings of logic as identied by Hegel
will be given an axiomatic interpretation. The domain of their values will be de-
cided by a three-valued modal relation for the probability mass function. Keynes
and Poppers solution to the problem of unknown proportions will be challenged
by an understanding of logic that puts the content of what the axioms mean for
making rational decisions before their mere being. What is true for inferences
will then work for the principle of a dialectic function in contrast to the two
proposed by Hume and the one proposed by David Lewis . In this way, it will
be demonstrated that Hegels understanding of logic is still more advanced than
ones that fails to recognise they function within the dialectic.
Ole Hjortland
The University of St Andrews - Scotland
[email protected]
Categoricity and Non-Classical Logics
We explore Carnaps so-called categoricity problem. Smiley (1996) solved the
problem for classical logic, and Rumtt (1997) follows up with a solution for
Strong Kleene. Using n-sided sequent calculus, we oer a proof-theoretic frame-
work to handle the categoricity problem for a range of many-valued logics. In
conclusion, we discuss some connections with proof-theoretic semantics and the
semantic role of proof-conditions.
1. Rumtt, I.: 1997, The categoricity problem and truth-value gaps, Analysis
57(4), 223236.
2. Smiley, T.: 1996, Rejection, Analysis 56(1), 19.
Dale Jacquette
University of Bern - Switzerland
[email protected]
An argument for universal logic: troubles relativizing truth-functions in Tarskis
truth-schemata hierarchy
Tarski avoids the liar paradox by relativizing truth and falsehood to particular
languages and forbidding the predication to sentences in a language of truth or
falsehood by any sentences belonging to the same language. The Tarski truth-
schemata stratify an object language and indenitely ascending hierarchy of
110
meta-languages in which the truth or falsehood of sentences in a language can
only be asserted or denied in a higher-order meta-language. However, Tarski?s
statement of the truth-schemata themselves involve general truth functions, and
in particular the biconditional, dened in terms of truth conditions involving
truth values standardly displayed in a truth table. Consistently with his se-
mantic program, all such truth values should also be relativized to particular
languages for Tarski. The objection thus points toward the more interesting
problem of Tarski?s concept of the exact status of truth predications in a gen-
eral logic of sentential connectives. Tarski?s three-part solution to the circularity
objection which he anticipates is discussed and refuted in detail. The upshot is
to support a universal logic of propositional connectives whose truth values do
not need to be relativized to a Tarskian hierarchy of linguistically relativized
truth values.
Colin James III
CEC Services - USA
[email protected]
4vbc is proved to be a group, ring, and module
Four valued bit code (4vbc) consists of the elements 0, 0, 0, 1, 1, 0, and
1, 1. The motivation is to show that 4vbc is a vector space, or if not then to
show what category 4vbc is. This paper proves that 4vbc is not a vector space
but is an Abelian group, a ring, and a module.
Amy D. Karofsky
Hofstra University - USA
[email protected]
Conditionals and contingentarianism
Conditional statements are known to be problematic. Its not clear whether a
conditional should be treated truth-functionally and, if so, whether it should
be treated as a single proposition or as a relation between two propositions.
Furthermore, treating conditionals truth-functionally results in some strange
consequences: It seems counterintuitive that anything can follow from a false
antecedent and that anything can lead to a necessarily true consequent. And,
in general, its not clear how to treat counterfactual conditionals. All of these
problems are compounded when dealing with nested conditionals. I show that
these and some other problems involving conditionals are due to the assump-
tion that there is genuine contingency in the world. I argue that a rejection
of contingentarianism in favor of necessitarianism results in the dissolution of
all of these problems. I provide a necessitarian account of (so-called) condi-
tional statements, according to which the only appropriate use of a conditional
is when the truth value of the antecedent is unknown, and according to which
(so-called) true conditional statements are more accurately expressed as noncon-
ditional, universal or general propositions. Necessitarianism entails the collapse
of modal logic (since the study of modal propositions will be the study of nec-
essary propositions only (i.e. all propositions)). I argue that this consequence
is both desirable and true.
111
Yaroslav Kokhan
Academy of Sciences of Ukraine
[email protected]
Semantical Presuppositions in Logical Syntax
Logical syntax itself depends on semantics because even principles the modern
formal languages based on them implicitly contain two following semantical pre-
suppositions: (I) all names of individuals are proper names i. e. everyone of
them has the only denotation (not more or less); (II) model objects which are
interpretants of linguistic expressions are all the denotations but not the senses.
Let us call semantics based on these two principles the standard semantics. One
can take dierent nonstandard semantics and build formal languages and cal-
culi on them as on the basis. The author proposes to replace predicate notion
with notion of generalized function for this purpose. This means that any such
a function (a) can have any number of values as long as its arguments have
xed and (b) can have no arguments at all. Formal languages based on such
notion of function require to generalize the equality relation and to introduce
the special logical function of choice because of many-valuedness of functions in
general case. Equality is denable in new languages. We can construct calculi
with any given semantics in mentioned function languages.
Zofia Kostrzycka
University of Technology - Poland
[email protected]
Locally nite logics have the density
Our aim is to give more general conditions on the existence of the density of
truth. The notion of local niteness (for propositional logics) turns out to be
very helpful in this task. We do not restrict our attention to logical systems
dened in the standard propositional language such as classical or intuitionistic
logic. We also take into consideration a family of modal logics which obey
our criterium. We prove that the density of truth exists for a large class of
locally nite (locally tabular) propositional logics. We are primarily interested
in classical and intuitionistic logic and show that their implicational fragments
have the same density. There are also given some locally nite logics without
the density of truth.
[1] Kostrzycka Z. On the Density of Truth of Locally Finite Logics, Journal
of Logic and Computation, Advance Access, June 26, 2009.
Boris Kovalerchuk
Central Washington University - USA
[email protected]
Leonid Perlovsky
Harvard University and AFRL - USA
[email protected]
On Dynamic Logic for Phenomena Modeling
This presentation provides a formal description of the Dynamic Logic for Phe-
112
nomena modeling (DLP) along with the comparison of the goals of DLP and
some other dynamic logics. DLP is presented with its syntactic, reasoning, and
semantic parts in terms of the model theory. Computational complexity issues
that motivate this work are presented using an example of polynomial models.
Modeling of complex real-world phenomena such as the mind presents tremen-
dous computational complexity challenges. DLP addresses these challenges in a
non-traditional way. The main idea behind its success in applications is match-
ing the levels of uncertainty of the problem/model and the levels of uncertainty
of the evaluation criterion used to identify the model. When a model becomes
more certain then the evaluation criterion is also adjusted dynamically to match
the adjusted model. This process is called Dynamic Logic of Phenomena mod-
eling, which mimics processes of the mind and natural evolution. There are two
complimentary trends in modeling physical phenomena and logic. In the rst
one, it is adding more logic structures to classical mathematical techniques. In
logic, it is ?dynamication? of logic with various new non-classical logics about
actions rather than about propositions as well as about making logic operations
dynamic. To actually benet each other these areas need to be close enough in
specic tasks and goals. We use the comparison of goals to establish initial links
of DLP with some dynamic logics and to facilitate further studies of linkages in
both communities.
Tamar Lando
UC Berkeley College of Letters and Science - USA
[email protected]
Completeness of S4 with respect to the measure algebra
Jui-Lin Lee
Che-Ping Su
National Formosa University - Taiwan
[email protected]
[email protected]
A Remark on the minimality of
In this talk we will present the fact that the axioms of comprehension scheme
and innity in ZF do not exactly oer us the minimal collection of all natural
numbers, as we usually expect or what we have learned from elementary set
theory course. Since this is also true for any inductively dened set, inductive
sets like the set of all terms (for a xed rst order language) or the quantier-
free part of a theory extension (constructed in [1] for a possible model of any
given theory) are not as concrete as we think (though no paradox is generated
and no harm is really made in the sense of formalism).
Reference:
[1] Jui-Lin Lee, The classical model existence theorem in subclassical pred-
icate logics. I. Towards mathematical philosophy, pages 187-199, Trends Log.
Stud. Log. Libr., 28, Springer, Dordrecht, 2009.
113
Javier Legris
National Academy of Sciences - Argentina
[email protected]
Universality as generality: the case of Ernst Schroders pasygraphy
According to Ernst Schr oders original idea of formal algebra, a formal theory
is a set of basic symbols with certain combination rules. For these operations
and relations certain postulates hold, from which theorems are derived. In this
conception, the structural side of formal algebra is implicit. A formal theory is
merely schematic, including the extreme case of schemata consisting exclusively
of blanks. The applications of it to dierent domains can lead to formal systems
with more meaningful symbols and less blanks. This general applicability of
formal algebra suggested an idea of universality. It was implicit in Schr oders
Handbook of algebra and arithmetic (1873) and underlies his Lectures on the
Algebra of Logic. In volumen III of this work (1895), he extended this program,
adopting the algebra of relatives of Peirce. Now, in his paper On Pasigraphy
(1898), Schrder considered the algebra of relatives as a universal scientic lan-
guage. At the same time, this pasigraphy was a foundational theory for prima
facie every scientic domain. In this theory the basic or fundamental notions of
pure mathematics (logic, arithmetic and geometry) are introduced. From these
notions further theories could be formulated. This is a dierent conception of
universality, resembling previous ideas of Frege, concerning a contentual univer-
sality. The aim of this paper is to analyze Schrders conception of universality
underlying his idea of a pasigraphy in its historical context, that is, as a chapter
in the history of universal logic. Moreover, it will be discuss to what extent
a general symbolic structure can be used as a basis for a universal language.
It will be argued that the distinction between alternative senses of universality
rests on alternative ideas of formal entities, whose ontological presuppositions
should be elucidated. If a theory can be formal as far as it represents alter-
natively formal objects, on one side, and properties or structures, on the other
side, then these two categories have to be clearly distinguished.
Domenico Lenzi
University of Salento - Italy
[email protected]
On a particular axiomatization of Propositional Calculus with the Negation and
the Implication as the unique connectives
In a previous work we presented a complete axiomatization of the implicational
fragment of the classical Propositional Calculus and gave a constructive proof
of its weak completeness. Here we extend our results to the classical Propo-
sitional Calculus, with the negation and implication connectives. We will use
only constructive proofs.
Maria Manzano
University of Salamanca - Spain
[email protected]
Enrique Alonso
114
Autonomous University of Madrid - Spain
[email protected]
Completeness: actors, history, and meaning
The completeness theorem is one of the best pointers of the balance, health and
goodness of a formal system. A calculus without a semantic counterpart is a
fruitless set of rules without a proper logical purpose. When and how is the
necessity of a completeness proof born? When does it separate itself from a
theorem concerning the decidability of satisability for a given logic? Russell
and Whitehead were not in the possition to properly distinguish and separate
formal calculus and semantic issues, they do not believe that the semantical
level constitute in itself a land for formal analysis. The early completeness
demonstrations for Propositional Logics are intimately related to decidability
and representation in terms of nite algebras. The original publications of Post,
Stone, Quine, Tarski, and Godel are relevant for this line. It was Godels idea
not to presupose decidability of the calculus and to search for completeness
using a mecanism previously employed by L owenheim and Skolem.
Henkins completeness proof came two decades after, but it was soon adopted
as the method, since it was rather exible and could be easily applied to other
logical systems. In fact, Henkin himself created it for Type Theory, but shortly
realized that it could also be used in rst order logic. This universal method
is at present our main research interest. Its genesis is what we are trying to
clarify.
This historical overview help us to analyse the present equilibrium between
models and calculi of a given logical system. Can we use as well Henkins method
in Universal Logic?
Maria Manzano
University of Salamanca - Spain
[email protected]
Luis Urtubey
National University of Cordoba - Argentina
[email protected]
Completeness and interpolation with a standard abstract consequence relation.
Henkins Completeness Proof deserves a place among those specic results lifted
to broader logical contexts. Its advantages concerning exibility and broad range
of application have been widely appreciated. In the same vain, Robert Gold-
blatt has accomplished a sort of generalization of the method of Henkins com-
pleteness proof, more recently. After having developed his nowadays-famous
proofs, Henkin adapted his method to apply it to other targets, such as the
Craig-Lyndon interpolation. Incidentally, potential generalizations of Henkins
proper extension of the Craig-Lyndon interpolation theorem using his method of
constants deserve further consideration. We expect that such a generalization
would shed light on the epistemological signicance of conditions concerning
completeness. This paper aims to stress the signicance of a generalization of
Henkins method, similar to the generalization accomplished by Goldblatt some
years ago, to understand certain facts relevant to seek for some logical universals.
115
Federico Marulanda
Universidad Michoacana de San Nicol as de Hidalgo - Mexico
[email protected]
Limitations of axiomatic dialetheic truth theory
The best-known application of dialetheism is to semantic paradoxes such as the
Liar. In particular, Graham Priest has advocated the adoption of an arithme-
tized axiomatic truth theory, which I will call PA*, in which contradictions aris-
ing from the Liar paradox can be accepted as theorems, thanks to the adoption
of an underlying paraconsistent logic. It would be remarkable if the soundness
of PA* could be proved from within PA*, but this has not been accomplished
(and Hartry Field has shown that the obvious proof strategy cannot succeed). I
argue that if such a proof were forthcoming, certain facts about the behavior of
the provability predicate for PA* could not be established, whether in PA* or
outside of it, on pain of trivializing the theory. I discuss drawbacks associated
with this fact, while raising several other criticisms of the dialetheic strategy
along the way. I conclude that it is not in virtue of its truth theory that di-
aletheism is to be considered an attractive position.
Joke Meheus
Ghent University
[email protected]
Adaptive Deontic Logics for Various Types of Normative Conicts
It is commonly known that Standard Deontic Logic leads to triviality when
applied to normative conicts. Over the past three decades, considerable atten-
tion has been paid to normative conicts of the form A is obliged and not-A
is obliged. At this moment, a plethora of systems is available that can handle
this particular form of deontic conict. Other forms of normative conicts have
been largely (and in my opinion, unjustly) ignored.
The aim of this paper is threefold. First, I shall present a taxonomy of the
dierent kinds of normative conicts. Attention will be paid to the origin of
the conict as well as to the dierent kinds of impossibility that are involved.
Next, I shall discuss dierent strategies to deal with the dierent kinds of nor-
mative conicts and present general procedures to characterize the (monotonic)
logics that are obtained by means of these strategies. Finally, I shall argue that
these monotonic systems are too poor to deal with the dierent kinds of norma-
tive conicts, but that the adaptive versions based on them lead to satisfactory
results.
Ricardo Mendes Grande
State University of Campinas - Brazil
[email protected]
On the applicability of mathematics
The purpose of this talk is to show how the concepts of formal content and
duality due to Granger [1,2] can help us to explain why mathematics ts so ad-
equately to the description of empirical reality. We take quantum mechanics as
116
a study of case to understand such a success of the applicability of mathematics.
We try to give a dierent solution (from the well known ones) to the problem
of the applicability of mathematics to empirical sciences. By those known solu-
tions we mean the realist approaches derived from Plato, developed essentially
by Frege [4], Steiners [5] anthropocentric account and few other approac hes
(apud[4]) . We also try to account for another special feature of mathematics
and that such a few philosophers have tried to understand; thats the role of the
symbolic reasoning in helping scientists to make new discoveries, such as the
ones we nd in Diracs work. We mean by that the prediction of anti-matter
(see [3]). From a reasoning that seemed to be only a free symbol manipulation,
Dirac was able to make these predictions that were conrmed in a short period
of time; so we think that its necessary to give an epistemological solution to
this subject and that does not take into account the existence of mathemat-
ical objects in the sense the realists approaches believe they exist, i.e., in an
analogous sense of the physical existence.
References
[1] Granger, G.G. The notion of formal content, Social research, pag. 359-
382, 49(2) 1982
[2] Granger, G.G. Contenus formels et dualite, Manuscrito-revista interna-
cional de losoa. Vol.30, n.2, jul.dez.2007. Centro de lgica, epistemologia e
histria da cincia-Unicamp
[3] Lindsay, R. and B. Margenau, H. Foundations of physics, Chapter 1,
Dover 1957
[4] Da Silva, J.J. Filosoas da matemtica, Edunesp 2007.
Michael Moortgat
Utrecht Institute of Linguistics - The Netherlands
[email protected]
Tertium non datur in the Lambek-Grishin calculus
The Lambek-Grishin calculus (LG) is a bilinear extension of the Syntactic Cal-
culus obtained by complementing Lambeks multiplicative conjunction and the
residual left and right implications with a multiplicative disjunction and residual
left and right coimplications (dierence operations). Interaction between these
two families is governed by linear distributivity principles. These principles are
structure-preserving: they respect the non-commutativity and non-associativity
of the multiplicative operationsessential properties for a logic that wants to
capture word order and phrase structure in natural language. In this talk, we
add two pairs of negations to LG: a Galois connected pair, and by arrow re-
versal, a dual Galois connected pair. The composition of the Galois connected
operations yields a closure operation (being expanding, monotone, and idempo-
tent); from the composition of the dual Galois connected operations, one obtains
an interior operation (contracting, monotone, idempotent). We generalize the
distributivity principles relating the product and coproduct operations to the
unary negations and show that (directional versions of) the law of the excluded
middle are theorems in this setting. Because the distributivity principles have
the status of irreversible rules, collapse into a one-negation system with an in-
117
volutive negation and de Morgan interdenability of the product and coproduct
families is avoided. Instead, we nd half of the de Morgan laws. We show that
LG extended with these negations has particularly simple analyses for a num-
ber of syntactic and semantic phenomena that are problematic for the original
Lambek calculi.
References
1. Moortgat, M. (2009) Symmetric categorial grammar. J Philosophical
Logic 38(6), 681-710.
2. Bernardi, R. and Moortgat, M. (to appear) Continuation
3. semantics for the Lambek-Grishin calculus.
Information and Computation.
Alessio Moretti
University of Nice - France
[email protected]
The Aristotelian p
q
-semantics and their p
q
-lattices
Universal logic (UL) is said to deal essentially with the most abstract proper-
ties of consequence. But if one acknowledges that logic is also concerned with
negation, things get more complex: since 2004 it has turned out that the pure
formal study on negation opens to a whole new branch of mathematics, at the
intersection of logic and geometry: n-opposition theory (NOT), the geometry
of oppositions (negation being only a special kind of opposition). NOT teaches
that the elementary oppositional structures (the logical square, hexagon,
cube,...) are just instances of logical bi-simplexes of dim. m. And these, in
turn, generalised by a game-theoretical Aristotelian pQ-semantics (generating
pQ-lattices), are a particular case of logical poly-simplexes of dim m. So,
if NOT conrms the mathematical specicity of UL, its hyper-geometric nature
changes ULs very architecture.
Ional Narita
West University of Timisoara - Romania
[email protected]
Decision method for the paradoxes of material implication
The aim of this study is to develop a method to select the occurrences of mate-
rial implication in a formula of propositional logic which can be interpreted by
the consequential relationship. It is well known that there are a lot of formulas
of propositional logic, so that, despite their validity, they cannot be interpreted
by correct arguments, because the material implication symbol cannot be in-
terpreted by consequence (or deduction) relationship. For instance, though the
formula A(BA) is logically valid, there are incorrect arguments having the
same logical form. These situations are known as paradoxes of material impli-
cation. On the other hand, there are formulas, as modus ponens, where the in-
terpretation of material implication by consequence relation is possible without
118
paradox. It results that the symbol of material implication can be interpreted
by consequence relationship only conditioned. In this study, a test to distinguish
between paradoxical and unparadoxical occurrences of material implication is
presented. By several examples, it is shown that in the same formula, dier-
ent types of material implication occurrences (paradoxical or nonparadoxical)
can exist. For example, the transitivity law, ((AB) and (BC)) (AC), ad-
mits both paradoxical and nonparadoxical interpretations. While the argument
((P-Q) and (Q-R)) - (P-R) is correct, an interpretation like ((PQ)
and (Q-R)) - (P-R) is an incorrect argument. ( = material implica-
tion symbol; - = the relation of consequence).
Raja Natarajan
Tata Institute of Fundamental Research - India
[email protected]
Computer Aided Proof
We examine the impact of computers on the notion of mathematcal proof. Re-
cent advances in the eld of Interactive Proof Checking with the associated
development of powerful tools such as Proof Assistants have given rise to an
interesting consequence viz. the practical feasibility of importing techniques
developed in the computer science community and redeploying them to improve
the main activity of the working mathematician, namely the process of proof
development. At the core of such redeployed techniques lie the notions of formal
systems, formal reasoning, and formal proofs. However the process of formal-
izing mathematics is a highly non-trivial task, and gives rise to a number of
challenging and interesting issues which need to be addressed in order to make
the discipline of computer assisted mathematics more prevalent in the future.
References:
Asperti, A., Geuvers, H., and Natarajan R.: Social processes, program verica-
tion and all that, Mathematical Structures in Computer Science, Vol. 19, No.
5 (October 2009) pp. 877896.
Grigory K. Olkhovikov
Urals State University - Russia
[email protected]
Rosss Paradox: A New Approach
The report dwells upon a well-known paradox of deontic logic, namely, the
Rosss paradox. It demonstrates how the paradox uncovers an important princi-
ple about norms. This principle is analyzed against the following backgrounds:
(a) standard deontic logic; (b) propositional dynamic deontic logic; (c) proposi-
tional dynamic deontic logic of long-term obligations; (d) a mathematical theory
of norms, devised by the author not long ago. Within the rst three of the above-
mentioned systems, the principle either cannot be presented in its true form, or
can be presented and is invalid, which leads to obvious inadequacies in presen-
tation of reasoning about norms. Within the fourth system, some versions of
the principle can be presented and can be proven valid as well.
119
Fabrice Pataut
University of Paris 1 - France
[email protected]
Anti-realism, strict nitisn and structural rules
The conclusion of Michael Dummettss manifestability argument is that intu-
itionistic logic satises the semantic requirements of antirealism. Some philoso-
phers have argued that the traditional antirealist desideratum of decidability in
principle is too weak. Semantic antirealism properly construed must be commit-
ted to eective decidability. As such, it either leads to strict nitism (Wright
[1982] 1987),or to a much stronger kind of logical revisionism: substructural
logics,and in particular linear logics, rather than intuitionistic logic,satisfy the
semantic requirements of strict antirealism (Dubucs and Marion 2003). I shall
develop two kinds of replies. The rst kind of reply is that if we jettison the
eectively vs.in principle distinction, as applied to manifestability-type argu-
ments, we end up with an unsatisfactory explanation of how the meaning of
statements covering the practically unsurveyable cases is xed. I shall then
look at two radical antirealist principles disqualifying structural rules: Token
Preservation and Preservation of Local Feasibility. Against Bonnay and Cozics
criticisms (Bonnay and Cozic, forthcoming), I shall argue (i) that some concep-
tual support may be provided for Token Preservation, which doesnt rely on a
causal misreading of the turnstile, and (ii) that the appeal to non feasible ways
of doing feasible things is not a good way to argue for Preservation of Local
Feasibility.
References
1. BONNAY (Denis) and COZIK (Michael),forthcoming, Which Logic for
the Radical Antirealist?, Logic, Epistemology and the Unity of Science,
Springer Verlag, Berlin and Heidelberg [24 pages, in print].
2. DUBUCS (Jacques) and MARION (Mathieu), 2004, Strict Antirealism
and Substructural Logics, Philosophical Dimensions of Logic and Sci-
ence, J. Cachro, G. Kurezewski and A. Rojsczak, eds., Kluwer Academic
Publishers, Dordrecht, pp. 235-249.
3. WRIGHT (Crispin), [1982] 1987, Strict Finitism, Realism, Meaning and
Truth, Basil Blackwell, Oxford, 1st ed., pp. 107-175.
Oxana Pavlenko
Southen Federal University - Russia
[email protected]
The place of logic structure question-answer in dialogue theory of under-
standin
Understanding question as a particular form of conclusion appeared in philoso-
phy discussions of Antiquity. According to Socrates a dialogue is joint search for
truth and is formed from questions (problems and contradictions) and answers
120
(their solutions). Questioner and respondent in the process of dialogue exchang-
ing thoughts express dierent views on the world. None of them is not true,
but each of them can express his own understanding because the truth cannot
be known by one person. In philosophy the hermeneutic method of understand-
ing the sense of the text is the same as dialogue method where the structure
question-answer is used. In his turn in informal logic D.Walton presenting a
dialectical theory of explanation dened a successful explanation as a transfer
of understanding from a respondent to a questioner in a dialogue. This theory
combines two views on the nature of explanation: an explanation is seen as
search for an answer to a question and is dened in terms of a concept called
understanding. Such combination of dierent points of view is eectively for
dening scientic explanations. Thus, logic of question is the main component
in dialogue theory of understanding. Such coincidence of approaches to research
of the procedure of understanding in philosophy and informal logic shows the
importance of dialogue and its logic structure question-answer in developing
of modern methodology of scientic cognition in whole. References: 1. Walton,
Douglas. 2000. The place of dialogue theory in logic, computer science and
communication studies. In International Journal for Epistemology, logic and
Philosophy of Science, vol. 123, 327-346. 2. Walton, Douglas. 2004. A new
dialectical theory of explanation. In Philosophical Explanations, vol.7, No 1,
71-89.
Sergey Pavlov
Russian Academy of Sciences - Russia
[email protected]
Truth and Falsehood Operators Logic from Point of View of Universal Logic
This paper proposes the truth and falsehood operators logic. These operators
are included in language of this logic which allows the iteration of such operators.
The truth and falsehood operators (T, F) are logically independent. It give us
an opportunity to consider class of sentential logics in the language of this logic
which have 2, 3, 4-valued interpretations. So this logic can treated as universal
logic relatively to this class of sentential logics.
The truth and falsehood tetralemma [1]:
either (TS and FS), or (TS and FS), or (TS and FS), or (TS and FS).
Now let us consider language of logic with negation and implication ?.
Truth and falsehood conditions for negation and implication are standard.
In spite of above laws of classical logic are not valid.
So we can assert axiomatic emptiness relative to languages of logics with
negation and implication.
Finally, bivalence principle is equivalent to T-biconditional.
Reference
[1] S.A. Pavlov. Logic FL4 with Falsehood Operator, MultiValued Logics,
V. 5, 2000, pp. 125-138.
Gillman Payette
University of Calgary - Canada
121
[email protected]
Preservationism and Truth
This paper rst introduces preservationism and give it a precise formulation.
Second, Fields discussion of Kreisels squeeze argument for using model-
theoretic consequence as capturing the intuitive notion of validity is exposed.
Discussion of Fields analysis of validity will lead to a discussion of Fields own
view on the matter of soundness and his thesis that validity should be considered
a basic, intuitive notion, given that validity cant be captured, even extension-
ally, by necessary truth preservation. I call his position sociologism-to be
reminiscent of psychologismsince it relies on what we think to be valid reason-
ing, but it has less of a psychological bent. It is shown how the sociologism
ts with the preservationist program. Finally, the matter of pluralism will be
addressed. The end of the paper ties the preservationist program to the theme
of universal logic.
Andrea Pedeferri
University of Milan - Italy
[email protected]
Some remarks about the status of second order logic
Andrea Pedeferri, Dpt of Philosophy, University of Milan, Italy, Some remarks
about the status of second order logic Second order logic has been always con-
sidered problematic by modern logicians: so problematic that some of them
refuse to call it logic at all. Lindstr om Theorem sets out a boundary between
the pure logicality of rst order logic and the mathematicality of second or-
der logic: is the validity of completeness, compactness and L owenheim-Skolem
Theorem the only qualication to call a formal system logic? After all the
lacking of expressive power of rst order represented by the lacking of categoric-
ity, could well be considered an important aw too. Moreover, it could sound
odd that, on the one hand we do not call second order a proper logic due to
its beeing uncontrollable, and on the other hand we state, as a corner stone
of the controllable rst order, the L owenheim-Skolem Theorem, a theorem
which states the incapability of a theory to control its models.
Denis Ponomaryov
Institute of Informatics Systems - Russia
[email protected]
On decomposability in logical calculi
In Computer Science, decomposition is a standard technique to reduce com-
plexity of problems. In Logic, the notion of decomposition appears in numerous
applications including the important eld of automated reasoning over theories.
The main idea is to identify those fragments of a theory that are necessary and
sucient for testing a given property, thus reducing the search space and com-
plexity of reasoning. There is a number of papers, in which modularization and
decomposition methods for logical theories are considered.
In presence of interpolation, a natural approach is to consider signature
122
partitions of theories.
Assume T is a theory, sig(T) is the signature of T, and is a subset of
sig(T). We call the theory T -decomposable, if it is equivalent to a union of
theories T
1
and T
2
such that the union of sig(T
1
) and sig(T
2
) equals sig(T) and
the intersection of sig(T
1
) and sig(T
2
) is exactly . In the case is empty,
we speak of a pure decomposition of T into signature-disjoint theories and if
sig(T
1
) or sig(T
2
) equals , we say that T is trivially -decomposable. In fact,
the components T
1
and T
2
induce a partition of the signature sig(T) which we
call a signature -decomposition of T. Note that T
1
or T
2
may happen to be non-
trivially -decomposable and thus, give a ner signature Delta-decomposition
of T.
We consider the problem of deciding whether a given nite set of formu-
las is non-trivially -decomposable for a given subsignature . Which logical
calculi allow for an algorithm to compute signature -decompositions for an
arbitrary given nite set of formulas T and a sig(T)? What properties
should a calculus satisfy for every its set of formulas to have a unique nest
signature -decomposition for each subsignature ? In the talk, we give a
partial answer to these questions on the example of a broad class of logical
calculi. Unsurprisingly, the considered problems are closely related to impor-
tant interpolation properties studied in logics: the Craig interpolation property
and uniform interpolation property. The results covered in the talk are directly
transferred to the classical, intuitionistic logic, and a wide range of modal logics.
Frederik Van de Putte
Ghent University - Belgium
[email protected]
Three Adaptive Logics of Induction
In this talk I want to present three basic adaptive logics of induction: LI, IL and
G, dened within the standard format for adaptive logics from [2]. They provide
three dierent qualitative accounts of how people might derive generalizations
from a set of data. As expected, and as usual for most adaptive logics, they are
nonmonotonic and have a dynamic proof theory.
The consequence sets of the three basic logics are closed under classical logic.
For so-called complete data, the basic logics all lead to the same consequence
set. The dierence lies in the conditional assumption each of them makes,
which results in a dierent outlook of the proofs, and in signicantly dierent
consequences for some canonical incomplete premise sets. The logics can be
ordered according to their strength (proofs have been provided). This leads to
unexpected results when it comes to the need of instances or even positive in-
stances to derive a generalization. It turns out that this extra condition actually
strengthens the consequence relation, rather then weakening it.
Starting from some observations about G, I will show that yet another
strengthening of this logic is both plausible and possible. I will rely on the Pop-
perian idea that stronger generalizations should be priviliged over less stronger
ones, and apply this to G. This will result in a denition of SG, the strongest
logic within the whole range of adaptive logics of induction studied so far.
123
References
[1] D. Batens and L. Haesaert. On classical adaptive logics of induction.
Logique et Analyse, 46:225-290, 2003.
[2] D. Batens. A universal logic approach to adaptative logics. Logica Uni-
versalis, 1:221-242, 2007.
A.V. Ravishankar Sarma
Institute of Technology - India
[email protected]
Role of Abductive Reasoning in Belief Revision
Belief revision is concerned with how we modify beliefs when we receive new
evidential information. The rational behind such modication is to x some
of the inconsistencies arising out of a conict between accepting new and old
information together. Abductive reasoning on the other hand starts from a set
of accepted facts and infers to their most likely or best explanation. To explain
an event is to give a causal history, in many cases, causal explanation can be
taken as best explanation. Suppose I observe that my car wont start and makes
icking noise. Then I look for potential explanations (empty fuel tank, bad
weather) which deals with the problem. The causal explanation which invokes
sucient conditions of why my car wont start is that the battery is dead. We
revise beliefs accordingly and accept the inference to the best explanation, i.e.,
causal explanation. Despite immense importance of causal explanation in the
process of belief change, we come across a very little literature on how we change
beliefs in situations where causal explanations are important.
In this paper we present a theory of belief revision while extending Pag-
nucco, Nayak and Foos approach of belief revision in the context of abduction,
and propose an abductive entrenchment ordering for generating and evaluat-
ing the potential explanations. Our basic hypothesis is that an agent seeks
explanation, before adding the new evidential information with the old. These
potential explanations are then ordered based on the informational value and
background context and eventually lead to a best explanation which explains
the data.The best or adequate explanation is the one which gives necessary and
sucient conditions for phenomenon under observation. In other words, not all
explanations are acceptable for abduction, but only the best or at least good
ones as proposed by Peter Lipton, which is popularly known as inference to
the loveliest explanation. We provide a criteria of such preferential ordering of
explanations and provide sphere semantics for the resulting abductive entrench-
ment ordering. This approach to belief revision may be called causal approach
to belief revision. One limitation and diculty with causal approach is to deal
with belief revision due to non-causal explanations. .
1. Peter Gardenfors, Knowledge in Flux: Modeling the Dynamics of Epis-
temic States, Bradford Books, MIT Press, 1988.
2. Lipton, Peter. (2001). Inference to the Best Explanation, London: Rout-
ledge.
124
3. Maurice Pagnucco, Abhaya C. Nayak, Norman Y. Foo: Abductive Reason-
ing, Belief Expansion and Nonmonotonic Cesequence. Deductive Databases
/ Abduction in Deductive Databases 1995: 143-158
Helio Rebello Cardoso Jr
State University of S ao Paulo - Brazil
[email protected]
Peirces logic of relatives and the pluralism of continuity relative and continu-
ous predicates
One of the most important Peircean philosophical problems is continuity. Its
mathematical and metaphysical problems had received a logical reasoning. In
fact, Peirce developed the logic of relatives, in order to study pluralistic change
as continuitys main character. He states: continuity is simply what generality
becomes in the logic of relatives (CP 5.436, 1905). It means that the general
feature of continuity can be mostly spelled out in terms of relations, since they
are the proper predicate to dene continuity. Therefore, a pluralistic changing
reality could be expressed according to relative predicates (CP, 3.638, 1901) and
these ones are supposed to be ruled by the utmost form of pure or continuous
predicate (New Elements of Math., 1908). According to Peirce, the relative
predicate provides the kind of predication in which dierent relates belonging
to the predicate stand generally for a relation or relative character and the
reality which corresponds to a proposition with a relative predicate is called
fundamentum relationis. Moreover, dierent relations might gather together so
as to form a relationship system. Being the relative predicate a suitable way to
dene continuous reality, it would be appropriate that the system of continua
would be also dened in terms of a conceivable predication. As a matter of
fact, Peirce called continuous/pure predicate the self-containing character com-
ing up through a series of relations so as to make its reality perfectly continuous.
In this paper I shall consider, rst, the vital link between continuity and the
philosophical problem of relationship in some of Peirces writings. Accordingly,
second, I shall take into account that the development of the Peircean continuity
has two parts. The rst part envelops the typology of discrete collections and
multitudes; the second one the typology of continuous multiplicities. Roughly,
discrete collections are formed by units that could be individually assigned, while
the unities of continuous multiplicities could not. Indeed, the line between dis-
crete collections and continuous multiplicities is harder to draw than it appear
at a rst glance, because there are discrete collections that include individually
or/and generally designated units, as they hold the premonition of continuity.
This challenge is what makes the Peircean logic of relatives worthwhile, for
it remains to be inspected if the implications between discrete collections and
continuous multiplicities also involve the very denition of continuity so as to
sustain relations as its proper predicate and not merely as a metaphorical way
of speaking. An it fullls my third goal.
Gemma Robles
125
University of La Laguna - Spain
[email protected]
Disjunctive Syllogism, Lewis modal logics and paraconsistency
In logics where rules E (A B A, B) and I ( A, B A B hold, ECQ (E
contradictione quodlibet: A, A B) is equivalently formulated in the form
A A B.
As E and I hold in all logics dened in this paper, by ECQ we shall refer
to this second version of the rule.
As it is well-known (cf. [4]), a logic S is paraconsistent if ECQ is not a rule
of S. Not less well-known is the fact that Lewis logics are not paraconsistent.
But this fact is, indeed, no accident. Actually, Lewis holds that B is deducible
from A A (equivalently, (A A) ; B ; stands for strict implication),
according to the argument in [3], p. 250, currently known as Lewis argument
(cf. [1] 16.1), which relies on the disjunctive syllogism (d.s) understood as a
rule of inference.
The aim of this paper is not to discuss this challenged argument again (cf.
e.g. [1] and [2]) but to dene a series of paraconsistent logics included in Lewis
S4.
In all logics in this paper, d.s is understood as a rule of proof, not as a rule
of inference. Therefore, ECQ does not hold in any of them. So, theories built
upon these logics are not necessarily closed by d.s, and consequently, in case of
inconsistency, triviality does not follow automatically. Nevertheless, consistent,
prime theories are of course always closed by d.s.
All logics in this paper contain classical logic in the sense that all tautologies
in , and are provable. Moreover, as d.s is a primitive rule in each one of
them, Modus Ponens for is available. A Routley-Meyer type ternary relational
semantics is provided for each one of these logics. Soundness and completeness
theorems are proved.
Acknowledgements: Work supported by research projects FFI2008-05859/FISO
and FFI2008-01205/ FISO, nanced by the Spanish Ministry of Science and In-
novation. G. Robles is supported by Juan de la Cierva Program of the Spanish
Ministry of Science and Innovation.
References:
[1] Anderson, A. R., Belnap, N. D. Jr. (1975) Entailment. The Logic of
Relevance and Necessity, vol I. (Princeton University Press).
[2] Anderson, A. R., Belnap, N. D. Jr., Dunn, J. M. (1992) Entailment. The
Logic of Relevance and Necessity, vol II. (Princeton University Press).
[3] Lewis, C. I., Langford, H. (1932) Symbolic Logic, Dover, New York, 1959.
[4] Priest, G., Tanaka, K. (2004) Paraconsistent Logic. (In E. N. Zalta
(Ed.), The Standford Encyclopedia of Philosophy. Winter 2004 Edition). URL:
http://plato.stanford.edu/archives/win2004/entries/logic-paraconsistent.
Cassiano Terra Rodrigues
Pontifical Catholic University of S ao Paulo - Brazil
[email protected]
Deduction, Induction, and Abduction according to C.S.Peirce: Necessity, Prob-
126
ability, Discovery
According to C.S. Peirce, the essence of reasoning lies in the fact that where
certain relations are observed to occur, certain others also obtain. So, there
are three basic ways in which we can ascertain the logical connexion between
premises and conclusions. Deduction is in rst place the mode of necessary
logical reasoning. It is the only mode of drawing necessary conclusions, for the
truth of the premises grounds the truth of the conclusions. It can be described
in its logical form as the well-known modus ponens or modus tollens. Induction
is the type of reasoning that contrary to deduction doesnt draw necessary con-
clusions, for the truth of the premises doesnt necessarily warrant the truth of
the conclusions, but only probably state it. Induction, therefore, doesnt allow
for discovery, only for testing the conclusions we draw by deduction. Its specic
feature is to allow us to see that certain characters belong to certain objects.
Abduction is the type of reasoning that doesnt have any logical necessity and
has the least probability of establishing a true relation between premises and
conclusions, but is the only one with heuristic power. Its distinctive feature is
that it tells us that certain objects might have certain characters. It can be
described nonetheless as the logical fallacy of arming the consequent, or Post
hoc ergo propter hoc.
Francisco J.Salguero-Lamillar
Angel Nepomuceno-Fernandez
Fernando Soler-Toscano
Emilio G omez-Caminero
Universidad de Sevilla - Spain
[email protected]
[email protected]
[email protected]
[email protected]
Knowing and applying rules: tableaux for mixed modal deontic-epistemic logics
A game can be formalized in a multimodal logic with epistemic and deontic op-
erators. Then, the game rules must be dened by means of associated epistemic
and deontic actions, what means that we need a mixed modal deontic-epistemic
logic to dene a game as a set of rules in this way. Given a set of agents, models
are constructed with two kinds of accessibility relations: equivalence relations
for epistemic accessibility, and serial relations for deontic accessibility.
We will use labelled tableaux for dealing with this logic. Dierent labels are
used according to the accessibility relation and the agent considered. Two kinds
of rules are also used in this tableaux method that can be called common rules
and inheritance rules. The former are used only once in the execution of the
tableaux; the latter may be used so many times as we need and they guarantee
that the accessibility relations have the desired properties. Indeed, we can deal
with a S4 or a T system for epistemic operators just by changing the inheritance
rules. Therefore, soundness is guaranteed by means of these inheritance rules.
In such a system the operators iteration can make the tableaux innite. But
this tableaux method solves the problem by allowing us to get nite models
127
for interpreting mixed modal deontic-epistemic sentences that are satisable.
To achieve this result we dene accessibility relations for any deontic and epis-
temic interpretation of our modal opera tors and use this denition to develop
construction rules for the tableaux with certain restrictions on world indices.
This logic is very close to a dynamic interpretation of modal operators, since
we can establish a relation between sentences that express strategic knowledge
and some formulas of this logic.
Darko Sarenac
Colorado State University - USA
darko.sarenac@colostate dot edu
Modal logic for qualitative dynamics
Denis I. Saveliev
Moscow State University - Russia
[email protected]
Completeness in innitary logi
We investigate innitary logics L
,
and some more abstract logics, in view of
the following general question: Given a theorem of the usual nitary logic L
,
,
as L owenheimSkolem, Compactness, Completeness, Ultraproduct, Omitting
Type, etc., for which , the logic L
,
satises an appropriate analog of that
theorem? Typically, the answer leads to large cardinals. A particular emphasis
of the talk will be to our recent result on the Innitary Completeness Theorem.
We show that in fact Completeness is equivalent to Compactness: any consis-
tent theory in L
,
has a model if and only if is strongly compact, and any
consistent theory in L
,
using at most non-logical symbols has a model if and
only if is weakly compact. Actually, this result is applicable to a wider class
of innitary logics, e.g., this holds for higher-order ones: any consistent theory
in L
n
,
has a model if and only if is extendible. We discuss also innitary
analogs of other logical theorems.
References:
[1] J. Barwise and S. Feferman (ed.). Model-Theoretic Logics. Springer,
1985, xviii + 893 pp.
[2] M. A. Dickmann. Large Innitary Languages: Model Theory. North-
Holland, 1975, xv + 464 pp.
[3] A. Kanamori. The Higher Innite. Springer, 2005, xxii + 536 pp.
[4] H. J. Keisler. Model Theory for Innitary Logic: Logic with Countable
Conjunctions and Finite Quantiers. North-Holland, 1971, x + 208 pp.
[5] H. J. Keisler and A. Tarski. From accessible to inaccessible cardinals.
Fundam. Math., vol. 53 (1964), pp. 225308.
[6] M. Magidor. On the role of supercompact and extendible cardinals in
logic. Israel J. Math., vol. 10 (1971), pp. 147157.
Marcos Silva
Leipzig University - Germany
128
Pontifical Catholic University of Rio de Janeiro - Brazi
[email protected] Truth Tables as Metaphysics: sense, calculus
and limits of expression in Wittgensteins Tractatus
The way of this research is to investigate how Wittgensteins Tractatus can be
read from its account of truth tables as a more adequate notational system
for prositional expression, in comparison with the notation used in Principia
Mathematica, for instance. In this way, I hold its truth table metaphysics
as an specially illuminating key concept or reading strategy for understanding
Tractatus. I defend that its truth table account is rich enough to allow a global
reading of central tractarian theses about modeling paradigms in the nature of
language.
Here follows an enumeration to these positive tractarian theses: 1) a strong
principle of Frege or truthfunctional principle (every proposition is elementary
or can be analysed in terms of elementary propositions); 2) a postulating of
a complete propositional analysis; 3) sense propositional determinateness; 4)
essential propositional bipolarity as a criterion for sense (a legitimate proposition
should be able to be true and to be false); 5) a possible fully expression of
reality. Besides, there are some specic theses about logic, namely: 6) logical
propositions are tautologies; 7) they can be recognized by the symbol itself
(where truth table could be a decision algorithm); 8) they are complex, i.e.,
a special articulation of elementary proposition; 9) logical propositions make
manifest (zeigen) the inner structure of language; 10) logical operators do not
denote anything in reality (its Grundgedanke!). This tentative aproach shed
light indireticly on the strong original relationship between tractarian theses
and its specic account of logic. In this sense, it is easier to acknowledge that
one of the major eorts made in Tractatus was the defence of an account of
logic that could make logical propositions categorically dierent from scientifc
ones.
My point here is to show that when truth table is understood as an spe-
cial notation, some theses in Tractatus can be elucidated both positively and
negatively. In a positive way, as the topics above suggest. And in a negative
way: revealing where and when these tractarian theses fail and bring down the
whole tractarian project. We can read Tractatus from its truth table account
emphazing what Wittgenstein was trying to perform, i.e., to determinate an
exhaustive horizon to propositional sense. This suggestion of reading can also
exhibit tractarian failure, namely, the expressive incapacity of truthfunctional
analysis deals with all empirical propositions, specially the ones which convey
generality or hold any kind of gradation or series. In fact, Tractatus fails to deal
satisfactorially with generalities (e.g. quantication in innit domains), it lacks
also subtleness in comparison to predicative negation (Tractatus holds negation
as a propositional operator) and the exclusion by contrariety (e.g. ascription of
collor to visual points).
Tractatus fails where the truth table notation fails. It cannot deals with
innite generalities, predicative negation and exclusion by contrariety (only by
contradiction). A fortiori, I defend that the failure ground to Tractatus is its
strong compromise with the truth table as a more adequate notation to express
129
propositions. Tractatus has bet too much on a technique which has a very short
scope of expressivity because of its high level of abstraction in analysis. As a
result, if we hold the truth table notation as a conceptual key to Tractatus we
can fully understand its project and failure. Moreover, we can also explicit the
relevant metaphysical compromisses of a truth table technique in its own origin
(greatly forgotten nowadays).
Luiz Henrique da Cruz Silvestrini
Marcelo Esteban Coniglio
State University of Campinas - Brazil
[email protected]
[email protected]
Pragmatic satisfaction and quasi-truth
Newton da Costa and his collaborators (cf. [1]) have introduced the notion
of quasi-truth by means of partial structures, where the relations within the
structure are partial. Thus, the membership (or not) of a given tuple of the
domain in such a relation is not always dened, and so any partial relation
R is a triple of sets (P, M, D) where P is the set of tuples which eectively
belong to R, M is the set of tuples which eectively do not belong to R, and
D is the set of tuples whose membership to R is (still) undetermined. In this
way, the predicates as triples approach provides a conceptual framework to
analyse the use of (rst-order) structures in science in contexts of informational
incompleteness.
In this paper the notion of predicates as triples is extended recursively to any
complex (i.e., non-atomic) formula of the rst-order object language. Thus, the
interpretation of any formula j in a partial structure A inductively originates a
triple, generalizing da Costas approach to atomic formulas.
Moreover, this proposal generalizes the usual perspective of a given rst-
order formula j (with at most n free variables) within a structure A seen as a
relation R, which is dened inductively. From this, a new denition of quasi-
truth via the notion of pragmatic satisfaction is obtained.
References
[1] da Costa, N.C.A.,and French, S. (2003). Science and partial truth: A
unitary approach to models and scientic reasoning. Oxford: Oxford University
Press.
Edelcio Gonalves de Souza
Pontifical Catholic University of S ao Paulo - Brazil
[email protected]
Alexandre Augusto Martins Rodrigues
University of S ao Paulo - Brazil
[email protected]
Quantier Elimination and Strong Isomorphism
We generalize the classical theorem on quantier elimination for innitary lan-
guages, studying its consequences in respect to the notion of strong isomorphism
130
of substructures of a given structure that is a model of our initial theory. There-
fore, we obtain a kind of abstract version of Steinitz theorem of the classical
eld theory.
Shashi M. Srivastava
Indian Statistical Institute - India
[email protected]
Some parametrization theorems for measurable sets with uncountable sections
Every analytic set in the product of two Polish spaces with uncountable vertical
sections contains a set in the product of Selivanowski sigma-algebra and Borel
sigma-algebra with sections perfect. This generalizes similar result proved by
Wesley (using forcing) and Cenzer-Mauldin. We use it to prove some parametriza-
tion theorems for measurable sets with uncountable sections.
References:
1. D. Cenzer and R. D. Mauldin, Measurable parametrizations and selec-
tions, Trans. Amer. Math. Soc., 245, 1978, 309-408.
2. E. Wesley, Extensions of measurable choice theorems by means of forcing,
Israel J. of Math., 14, 1973, 104-114.
Petros Stefaneas
National Technical University of Athens
[email protected]
On institutions and proof events
Jendrik Stelling
Leipzig University - Germany
[email protected]
Are Truth Values Algebraic?
Logicians talk about truth values all the time. When asked how to locate them,
or how to decide how many truth values a given logic has, the orthodox an-
swer has to to with algebraic semantics. The formula is simple: Construct a
certain kind of minimal matrix semantics, look at the algebraic elements of the
matrixs algebraic component, and voil: Youve located the truth values. This
view seems so plausible that it went virtually unchallenged until Polish logician
Roman Suszko wrote a biting attack against the people advocating many-valued
logics, chiey among them his fellow Pole Lukasiewicz. Suszko argued that all
many-valued semantics could be reduced to ordinary two-valued ones (albeit
with some losses on the side). Since then, the nature of truth-values has en-
tered the publications again. In this talk, I will ultimately argue that the whole
idea of equating truth-values with algebraic elements is mistaken. I will present
a dierent analysis, one that takes notions like truth and falsity to be secondary
to the related notions of being truth-preserving and being falsity-preserving.
This conception shifts the focus from the algebraic to the relational part of a
given logical system. Related issues such as many-valuedness are also discussed.
Gabriela Steren
131
Ariel Arbiser
University of Buenos Aires - Argentina
[email protected]
[email protected]
Combinatory Logics for Lambda Calculi with Patterns
We introduce CLp, a combinatory logic system for a lambda-calculus with pat-
terns (namely lambda-P), obtaining a consistent extension of classical com-
binatory logic (CL). Our goal is to nd an appropriate bridge between the
two formalisms, and take advantage of some of the positive aspects of each.
As in classical CL, our system will avoid dealing with abstractions and bound
variables, while allowing functions to impose restrictions over their arguments
through pattern matching, in the same spirit as in lambda-P. We introduce
back and forth translation rules which allow us to represent this pattern calcu-
lus within the language of combinators, simulating the abstraction mechanism
and achieving combinatorial completeness. Since the full language becomes non-
conuent - as does the unrestricted lambda-P - we provide a restriction to the
set of patterns (based on the revised Rigid Pattern Condition of lambda-P) so
that the whole system satises conuence and ensures the consistency of the
underlying logical theory. We propose and study other variants of interest such
as the introduction of both curried and rst-order constructors for modeling
data structures and the generalization of the matching mechanism. We also
introduce a more general denition of pattern matching, characterizing a fam-
ily of conuent variants. Finally, we propose two type systems and prove the
fundamental properties.
Bibliographical references:
1. Barendregt, H. P.: The Lambda Calculus: its Syntax and Semantics.
Studies in Logic and the Foundations of Mathematics 103. North-Holland, Am-
sterdam, revised edition, 1984.
2. Klop, J. W., van Oostrom, V., de Vrijer, R.: Lambda Calculus with
Patterns. Theoretical Computer Science, 398(1-3):16-31, 2008.
3. Schnnkel, M.: ber die Bausteinen der matematische Logik. Math. Ann.
92:305-316, 1924.
4. van Oostrom, V.: Lambda Calculus with Patterns. Technical report,
Vrije Univ., 1990.
Christian Strasser
Ghent University - Belgium
[email protected]
Adaptively Applying Modus Ponens in Conditional Logics of Normality
Since the early eighties default reasoning, i.e., reasoning on the basis of what
is normally or typically the case, has drawn much attention from philosophi-
cal logicians as well as scholars working in articial intelligence. A promising
logical representation has been given in form of conditional logics of normal-
ity (see e.g. [2], [3]). Compared to classical approaches such as default logic,
circumscription, autoepistemic logic, these logics benet from a natural, simple
132
and unifying representation of default knowledge in form of conditionals A?B:
From A normally follows B. While these logics oer promising ways to reason
about conditionals, what is missing is the ability to perform default inferencing
itself. This talk presents a generic way to enhance given conditional logics of
normality in order to allow for defeasible applications of modus ponens (MP)
to conditionals. For that purpose an adaptive logic framework is introduced.
Adaptive logics (see [1]) allow us to apply MP to a conditional A?B and a fact A
on the condition that it is safe to do so, concerning the factual and conditional
knowledge at hand. It is unsafe, for instance, if the factual information describes
exceptional circumstances to a given rule. The two adaptive standard strategies
are shown to correspond to dierent intuitions, a skeptical and a credulous one,
that manifest themselves in the handling of so-called oating conclusions.
References:
[1] Diderik Batens. A universal logic approach to adaptive logics. Logica
Universalis, 1:221242, 2007.
[2] Craig Boutilier. Unifying default reasoning and belief revision in a modal
framework. Articial Intelligence, 68(1):3385, 1994.
[3] Philippe Lamarre. S4 as the conditional logic of nonmonotonicity. In
KR, pages 357367, 1991.
Thomas Studer
University of Bern - Switzerland
[email protected]
Syntactic cut-elimination for modal xed point logics
Fixed point extensions of propositional modal logics occur naturally in many
dierent contexts. In epistemic logic, for instance, common knowledge of a
proposition is dened as the greatest xed point of a certain positive operator.
Also in temporal logics safety and liveness properties are formalized by least,
respectively greatest, xed points. Kozens mu-calculus provides the general
extension of modal logic with xed points for arbitrary positive operators.
From a proof-theoretic perspective, syntactic cut-elimination is one of the
major open problems for these logics. In this talk we will survey the problem of
cut-elimination for modal xed point logics. We will present a solution for the
case of common knowledge which makes essential use of deep inference. More-
over, we discuss whether and how far this approach can be extended towards a
cut-elimination procedure for the modal mu-calculus. We will show new results
characterizing exactly the fragment of the mu-calculus for which the deep infer-
ence approach to cut-elimination works.
Kordula Swietorzecka
Johannes Czermak
University of Cardinal St. Wyszynski in Warsaw, Poland
Some calculus for a logic of change
Tinko Tinchev
133
Sofia University St Kliment Ohridski - Bulgaria
[email protected]
Universal fragments of the logics of the strong contact and the strong connect-
edness
Traditionally, the classical geometry and topology as an abstract kind of ge-
ometry are point-based in a sense that they take the notion of a point as one
of the basic primitive notions. An alternative approach to the theories of the
space goes back to Whitehead. It is based on the notion region as primitive and
the binary predicate contact between regions. Usually the regions are regular
closed sets in a given topological space (which form a Boolean algebra), the
binary contact relation between regions is the relation non-empty intersection
and the unary relation connectedness is connectedness in topological sense. The
binary relation strong contact between regions is dened as topological connect-
edness of the interior of the meet of non-empty subregions. The unary relation
strong connectedness is dened as topological connectedness of the interior.
In the present talk axiomatizations of the universal fragment of the rst-order
theory of the Boolean algebra of regular closed subsets with the relations strong
contact and strong connectedness for dierent classes of topological spaces are
given. In particular we consider the so-called polytops in Euclidean plane.
Ivan Varzinczak
Johannes Heidema
Katarina Britzadimir
Meraka Institute and University of South Africa
[email protected]
[email protected]
[email protected]
Pertinent Entailment
We investigate infra-classical entailment relations that we call pertinent entail-
ments. The notion of pertinence proposed here is induced by a binary accessibil-
ity relation on worlds establishing a link (representing some form of pertinence)
between premiss and consequence. We show that this notion can be captured
elegantly using a simple modal logic without nested modalities. One road to
infra-classicality has been studied extensively, that of substructural logics, which
weaken the generating engine of axioms and inference rules for producing en-
tailment pairs (X,Y). Here we follow an alternative (not antagonistic) strategy:
we rst demand that X entails Y classically, and then, with supplementary in-
formation provided by an accessibility relation, more, trimming down the set of
entailment pairs to infra-classicality. It turns out that our pertinent entailment
relations restrict well-known paradoxes avoided by relevance/relevant logic in
an interesting way. We also show that they possess other non-classical proper-
ties, like paraconsistency. Moreover, we investigate how a notion of obligation
can be captured elegantly and simply by our formalism. We also discuss the
properties of these pertinent entailment relations with respect to inference rules
traditionally considered in the literature.
134
Vladimir Vasyukov
Russian Academy of Sciences - Russia
[email protected]
From logical to metalogical pluralism
The recent situation in logics features the stable proliferation of non-classical
logical systems and this process, to all appearances, is irreversible by its nature.
During more than two thousand years the scholars considered Aristotelian and
Stoic logics solely; modern classical logic is a continuation of this tradition being
dierent just by its means. The emerging of non-classical logics stroke seriously
the logical investigations compelling to revaluate and cast doubt on many re-
sults which were took for granted so far. Though such a situation would be
easily methodologically diagnosed - as the well-known monism vs. pluralism
dilemma - but this by no means can help us to resolve a number of fundamental
questions concerning the problem of plurality of non-classical logic responses on
eternal disciplinary requests. This plurality makes important an issue of choice
of the uniquely true among them while taking into account that there are no
recipes and prescriptions for to-date. The strategy of overcoming this problem
situation comes either to the quest for paradise lost (classical or other logic
as the only recipe) or to the decisive acceptation of the point of view of the
principal plurality of logical systems as the future prognostic perspective. An
interesting aspect of the confrontation considered is an issue of non-classical
metalogics arising within the metalanguage formulation of logical consequence:
from ? follows ? if and only if from A is true follows ? is true. The second
word follows points to the metalogical consequence and then a question arises:
would this consequence be necessary the classical one? In fact, this condition
is not necessary in many cases e.g. in relevant one but could we in this case
reformulate the denition as from ? relevantly follows ? if and only if from A
is true relevantly follows ? is true? Here we have a relevant logic in metalevel.
But then there arises a temptation to introduce formulations of such a kind:
from ? intuitionistically follows ? if and only if from A is true intuitionistically
follows ? is true, from ? quantum logically follows ? if and only if from A
is true quantum logically follows ? is true etc. If we will try to identify some
formulation of type from ? in logic X1 follows ? if and only if from A is true in
logic Y1 follows ? is true and from ? in logic X2 follows ? if and only if from
A is true in logic Y2 follows ? is true then we will need a meta-metatheory for
dening criteria of such an identication. The universal logic seems to be the
good one for this aim since within it the cross mutual translatability of logical
systems is considered.
Sheila Veloso
State University of Rio de janeiro UERJ
[email protected]
Paulo Veloso
Federal University of Rio de Janeiro, UFRJ
Paula Veloso
University of S ao Paulo
135
Piaget-like transformation in universal logic
Piaget examined the behavior of classical propositional negation in the 50s.
Examining the behavior of some negation induced transformations under com-
position, he found that they form a familiar 4-element group: the Klein group
of symmetries. Piagets analysis can be explained by the behavior of classical
negation on a propositional letter: the integers modulo 2. This approach also
works for other kinds of logics: e.g., in the case of intuitionistic logic, one obtains
a monoid of transformations.
We extend these ideas to the context of universal logic and other kinds of
negation-like unary symbols. We consider two unary symbols and examine the
eect of applying one externally and the other internally, up to equivalence.
These transformations under composition form a Piaget-like monoid. The be-
havior of the underlying unary symbols under composition form (cyclic) monoids
of transformations. These underlying monoids of transformations impose con-
straints on the corresponding Piaget-like monoid, which determine, to a large
extent, its structure. Piaget-like monoids provide useful information on their
logics. Much as the determinant provides some information about a matrix,
the Piaget-like monoid gives a condensed view of the logic, which may be easier
to handle. Distinct logics may present the same Piaget-like monoid, but logics
with non-isomorphic Piaget-like monoids are non-isomorphic as well.
Piaget-like monoids are useful tools for analyzing and comparing logics.
Peter Verdee
Ghent University - Belgium
[email protected]
Rich Set Theory using Adaptive Logic
In this talk I present two solutions for the paradoxes of naive set theory. Naive
set theory is based on the Abstraction axiom schema and the Extensionality
axiom and leads to dierent paradoxes among which Currys paradox, Russells
paradox and Cantors paradox. These paradoxes and classical logic (CL) to-
gether trivialize naive set theory. The solutions presented here make use of so
called adaptive logic (AL). This is a class of logics with a dynamic proof theory
that elegantly formalize various kinds of complex defeasible reasoning forms.
I discuss two classes of ALs as the underlying logic of a set theory based on
the axioms of naive set theory. In this way, it is possible to block problematic
(paradoxical) consequences of the axioms but allow for the unproblematic ones.
The downside of going adaptive is a substantial increase in computational com-
plexity. The philosophical purpose of this project is twofold. First, the process
of searching for an appropriate AL solution might be seen as an intuitive ex-
plication for the defeasible reasoning process of overcoming the paradoxes of
set theory. Secondly, the here presented adaptive set theories are interesting
theories in their own respect. Most useful theorems in Zermelo Fraenkel set
theory (ZF) are provable, but the adaptive theories are evidently non-trivial, in
contrast to set theories like ZF. Moreover, the adaptive set theories start from
no other than intuitive axioms on sets, unlike the rather counterintuitive set of
axioms for ZF. One could argue that the adaptive set theories rely on weaker
136
philosophical assumptions and thus make fewer metaphysical claims.
Jorgen Villadsen
Technical University of Denmark - Denmark
[email protected]
Nominalistic logic
Nominalistic Logic (Jrgen Villadsen 2008) is a presentation of Intensional Type
Theory (Paul Gilmore 2001) as a sequent calculus together with a succinct nom-
inalization axiom (N) that permits names of predicates as individuals in certain
cases. The logic has a exible comprehension axiom, but no extensionality ax-
iom and no innity axiom, although axiom N is the key to the derivation of
Peanos postulates for the natural numbers. We present a revised Nominalistic
Logic with a new rule for application in the type theory such that each term
has a unique type. We also add a choice axiom. The resulting logic provides a
very concise foundation of mathematics.
References
Paul Gilmore 2001. An Intensional Type Theory: Motivation and Cut-
Elimination. Journal of Symbolic Logic p. 383-400.
Jrgen Villadsen 2008. Nominalistic Logic: From Naive Set Theory to Inten-
sional Type Theory. In Klaus Robering (Editor): New Approaches to Classes
and Concepts. Studies in Logic: Volume 14 p. 57-85 (College Publications).
Alexey G. Vladimirov
Maymonid State Classical Academy - Russia
[email protected]
Completeness in innitary modal logic
We extend some results of Savelievs Completeness in Innitary Logic to in-
nitary modal logics. We show that the Canonical Model Theorem holds for
all innitary modal logics in the language L
,1,
if and only if is a weakly
compact cardinal. We show that some modal logics that are complete in L
,1,
,
e.g., K, K4, S4, remain complete in L
,1,
for weakly compact . We discuss an
analog of Salqvists theorem. We discuss also Completeness of predicate modal
logics in L
,,
. This is a joint work with Denis I. Saveliev.
References
[1] J. Barwise and S. Feferman (ed.). Model-Theoretic Logics. Springer,
1985, xviii + 893 pp.
[2] P. Blackburn, J. van Benthem and F. Wolter (ed.). Handbook for Modal
Logic. Elsevier, 2007, xii + 1231 pp.
[3] A. Chagrov and M. Zakharyaschev. Modal logic. Oxford, 1997, xvi +
605 pp.
[4] M. A. Dickmann. Large Innitary Languages: Model Theory. North-
Holland, 1975, xv + 464 pp.
[5] A. Kanamori. The Higher Innite. Springer, 2005, xxii + 536 pp.
[6] G. Sambin and V. Vaccaro. A new proof of Sahlqvists theorem on modal
denability and completeness. J. Symb. Logic, vol. 54 (1989), pp. 992999.
137
Urszula Wybraniec-Skardowska
Poznan School of Banking, Poland
[email protected]
Jacek Waldmajer
University of Opole, Poland
[email protected]
On consequence operations
Stephan van der Waart van Gulik
Peter Verdee
Ghent University
[email protected]
[email protected]
Gaining insight: towards a functional characterization of information in formal
logics
It is quite common for logicians to accept the idea that deductive inferences
are non-informative. Yet, this idea clearly conicts with our everyday reasoning
experiences: most of us will acknowledge that some inferences bring along new
insights and thus seem to contain some kind of information (for example, it is
quite hard to argue that Euclids theorem contains no new information). In this
talk, we shall present various conceptual and technical aspects of our project
aimed at developing a formal framework that allows classical and non-classical
logics to deal with the notion information in an intuitive way. The formal
core of the framework is based on a slightly modied version of the so-called
block semantics, developed by Diderik Batens, cf. [1], and encompasses two
new quantitative information measure functions that enable us to express how
much information is present (at a certain line) in a formal logic proof. The
functions are based on the formal notions of entropy and self-information as
dened in standard information theory. More specically, the rst function is
able to express a precise distance between the amount of information at a line
of a proof and the total amount of information present in the premises. The
second function is a goal-directed variant of the rst function and allows us to
formalize goal dependent accounts of information.
References
[1] D. Batens, Blocks. The clue to dynamic aspects of logic, Logique et
Analyse 150-152 (1995) 285-328, appeared 1997.
Tomoyuki Yamada
Hokkaido University - Japan
[email protected]
Acts of Withdrawing in Dynamic Logic of Propositional Commitments
We usually withdraw our own assertions or concessions when our beliefs are
revised. The eects of acts of withdrawing assertions or concessions, however,
are signicantly dierent from the eects of retractions studied in the AGM
approach to belief revision; they do not satisfy AGM postulates for belief re-
tractions. We develop a dynamic modal logic that can deal with eects of acts of
138
withdrawing of assertions and concessions by extending a dynamic logic called
DMPCL (Dynamied Multi-agent Propositional Commitment Logic), which is
developed according to the same strategy that leads to the development of dy-
namic epistemic logics in the last two decades. Thus we rst develop a propo-
sitional modal logic, MPCL, which deals with static structures of multi-agent
propositional commitments, and then dynamify it by adding dynamic modal-
ities that represent acts of asserting and conceding to MPCL. In the resulting
logic DMPCL, acts of asserting and conceding are modeled as events that update
propositional commitments borne by individual agents involved in a discourse.
MPCL is axiomatized in a completely standard way, and DMPCL is axioma-
tized by adding a set of so-called reduction axioms to the proof system of
MPCL. Acts of withdrawing assertions and withdrawing concessions can then
be modeled as yet another kind of events that update agents propositional
commitments by adding dynamic modalities that represent acts of withdrawing
assertions and withdrawing concessions to DMPCL. Unsurprisingly, the eects
of acts of withdrawing are very dicult to capture, and the completeness prob-
lem for the extended logic, called DMPCL+, is still open. Yet the possibility of
withdrawal seems to be a distinguishing characteristic common to a wide range
of acts whose eects are conventional or institutional, and thus the logical dy-
namics of acts of withdrawing seems to be of great signicance to the study of
social interactions among rational agents.
Aziz F. Zambak
Yeditepe University - Turkey
[email protected]
A Trans-logic system for machine intelligence: the transformation of data within
various logical systems
We attribute a constructive and regulative role to logic in AI to nd a proper
(ideal) way of reasoning for machine agency. For the realization of such roles,
a logical model that can operate in complex situations and overcome the frame
problem should be developed. In this presentation, some basic principles for a
logical model in AI will be proposed. Our aim is to present the general skeleton
of the logical model called the trans-logic system. The main idea behind the
trans-logic system is that in AI, reasoning is based on the idea of using data
(S-units and M-sets) and operating successive processes until the nal informa-
tion is achieved (realized). The trans-logic system includes concomitant logics
which have various functions for reasoning processes in machine intelligence. In
AI, we propose to use groups of programs, each of which are based on dierent
logical systems that allow a machine intelligence process to handle particular
data in a large set of functional analysis. In the trans-logic model, we give fuzzy
logic a regulative and transformational role in logic programming because fuzzy
logic can regulate sequences of information processing, permitting passage from
one stage (for example, deductive reasoning system) to another stage (for exam-
ple, paraconsistent systems). Fuzzy logic is also important for idealization and
appropriation because appropriation is a signicant criterion for understand-
ing whether a piece of data is suitable, proper, and relevant to the agent or
not. Fuzzy logic is not an eective reasoning model; but rather an eective
139
regulative model for constructing an interactional and transformational system
between dierent reasoning models.
140
5 Sponsors
1. FCT - Fundac ao para a Ciencia e Tecnologia - Portugal
2. Instituto de Telecomunicac oes - Portugal
3. Fundacao Calouste Gulbenkian - Portugal
4. Universidade Tecnica de Lisboa - Portugal
5. Centro Internacional de Matematica - Portugal
6. Redes Energeticas Nacionais - Portugal
7. Association for Symbolic Logic - United States of America
8. Sociedade Brasileira de Lgica
141