esslli95

kathrine Wenham (kathrine@gilcub.es)
Tue, 7 Feb 95 10:43:43 GMT

SEVENTH EUROPEAN SUMMER SCHOOL IN LOGIC LANGUAGE AND
INFORMATION
Barcelona, August 14-25 1995

For More Information Contact:
ESSLLI95
GILCUB
Avda. Vallvidrera 25
08017 Barcelona

Fax +43 3 2054656
Tlf +43 3 2033597
E-mail esslli95@gilcub.es

GENERAL INFORMATION

The Summer School

The Seventh European Summer School in Logic, Language and Information will be he
ld
at the Biology Faculty of Barcelona University from August 14th to 25th. The Sch
ool is
organizedunder the auspices of the European Foundation of Logic, Language and
Information (Folli), jointly by Universitat de Barcelona, Universitat Rovira i V
irgili,
Universitat Autonoma de Barcelona, and Universitat Politecnica de Catalunya.

The previous Summer Schools took place at Rijks universiteit Groningen in the
Netherlands in1989, at the Katholieke Universiteit Leuven in Belgium in 1990, at
the
Universitat Saarlandes, Saarbrucken in Germany in 1991, at the University of Ess
ex
Colchester in the United Kingdomin 1992, at the Faculdade de Letras da Universid
ade
de Lisboa in Portugal in 1993, and at the Copenhagen Business School in Denmark
in
1994. For these Schools financial support was derived from a variety of sources,
including
the Commision of the European Community, through the ERASMUS programme and
the DG XIII, Research Networks, National Research Councils, and industrial spons
ors.
The same kind of support is expected for the 1995 School.

The main focus of the Summer School is the interface between logic, linguistics
and
computation, where it concerns the modelling of human linguistic and cognitive a
bilities.
The 1995 School programme will include courses, workshops, and symposia covering
a
variety of topics within six areas of interest: Logic, Language, Computation, Lo
gic and
Computation, Computation and Language, Language and Logic.

Courses will be cast at both introductory and advanced levels. Introductory cour
ses are
designed to familiarize students with new fields and do not presuppose any backg
round
knowledge, while advanced courses are designed to allow participants to acquire
more
specialized expertise in areas they are already familiar with. Workshops will be
chaired
by an expert in ther field and will provide an opportunity for PhD students and
other
young researchers to present their work and gain informed feedback and useful co
ntacts.
Symposia will typically consist of a series of presentations on a timely topic b
y people
active in the relevant areas. Both workshops and symposia are intended to encour
age
collaboration and cross fertilization of ideas by stimulatingin-depth discussion
of issues
which are at the forefront of current research in the field. There will also be
a series of
invited evening lecturers by well-known experts in their field.

ACCOMODATION

There will be accomodation available. More information will appear in following
anouncements.

COURSE PROGRAMME

Bellow you will find the Summer School programme of courses, workshops and sympo
sia.
A few descriptions are still missing. These will, hopefully, be provided for the
final
booklet to be distributed to registered participants.

LECTURERS:

Logic and Language: G Link, J van Eijck, J Jaspars, A Visser, C.Vermeulen, M Kra
cht,
B Keller, M Kanazawa, D de Jongh, F Moltmann, C Fox, H Rott, R Cooper, M Poesio.

Logic: R Jansana, I Hodkinson, J Flum, P Blackburn, M de Rijke, Y de Venema, J F
ont,
K Devlin.

Computation and Logic: D Basin, S Mathews, F Baader, G Koestler, H Barringer, D
Gabbay, C Brink, L Wallen, M Marx, S Mikulas, F Baader, K Schlecta, S Biundo.

Computation: M Nielsen, C Paulin-Mohring, N Jones, L Meertens, S Peyton-Jones, B
Haglund.

Language and Computation: E Stabler, M Johnson, R Bod, R Scha, M Moortgat, M
Dalrymple, R Kaplan, B Carpenter, G Morrill, V Abrusci, R Dale, M Ellison, F ver
dejo,
R Kempson.

Language: P Miller, J Bresnan, H Verkuyl, H Clahsen, D Oehrle, J Doerre, S
Manandhar, A Zaenen, L Sadler.

SECTION: LOGIC AND LANGUAGE

Algebraic Semantics for Natural Language
Name of Lecturer(s)/Organizer(s): Godehard Link
Affiliation(s): Institut fur Philosophie,
Logik und Wissenschaftstheorie (IPLW)
Universitat Munchen
E-mail: glink@cis.uni-muenchen.de
Type: Introductory

Brief course description:
Language is able to refer to a wide array of objects of different sorts, which d
iffer from
each other in their characteristic structural properties. Algebraic semantics de
als with
structured domains which are designed to model those properties. The course will
give
a detailed motivation of the need to include in the domain of discourse plural e
ntities,
mass objects, and events. The irrespective axioms will be described, and applica
tions in
linguistics, philosophy, and knowledge representation (KR) will be discussed: co
llective
and distributive readings, mass relations,aspectual classes, mereology, choice o
f
ontologies in KR.
Course Prerequisites:
None, but a basic course in linguistic semantics and/or model-theoretic semantic
s of first
order predicate logic would be helpful.

Formal Forays into Language
Name of Lecturers: Jan van Eijck and Jan Jaspars
Affiliation(s): CWI, Amsterdam and OTS, Utrecht
E-mail: jve@cwi.nl /jaspars@cwi.nl
Type: Introductory

Brief course description:
The course is an introduction to the application of tools from the arsenal of lo
gic (in
particular, modal logic, tense logic, and various brands of dynamic logic) in th
e analysis
of natural language. Specific topics include: languages and grammars,the
syntax/semantics/pragmatics distinction, the logic of the major modes of English
,
variables and binding, features and their logic, the logic of syntax trees, the
composition
of meaning, and the logic of context change.
Course Prerequisites:
No specific knowledge of logic is presupposed, but very flattering assumptions a
re made
about the audience's intelligence and general interest in the subject matter.
Recommended reading:
L.T.F. Gamut, Logic, Language and Meaning, University of Chicago Press, 1991, 2
volumes.

The Dynamics of Structure
Name of Lecturer(s): C. Vermeulen, A. Visser
Affiliation(s): Department of Philosophy
E-mail: Albert.Visser@phil.ruu.nl, Kees.Vermeulen@phil.ruu.nl
Type: Advanced
Brief course description:
This course can be viewed as a variation on the broad issue of processversus str
ucture.
The idea we consider here is whether it is possible to view elements of the synt
axof a
sentence or discourse not as a syncategorematic aid to the process of interpreta
tion (as
inthe Montagovian Paradigm), but as standing for "program instructions" which ha
ve their
properhome at the (a) semantical level. An example of this plan is an ordinary b
racket
which can begiven the meaning: "introduce a new file to store local information"
. Other
examples are thedynamic interpretation of the existential quantifier and the dis
course
modifier "suppose". For somehints about the mathematical contents of the course,
see
below.
Course Prerequisites: The student should have some basic mathematical sophistica
tion.
This means that she need not necessarily know what e.g. a monoid or a morphism i
s, but
that she atleast should be able to acquire such concepts within reasonable time.
Moreover the student should be familiar with the basic definitions of DPL and DR
T, see
e.g. Kamp & Reyle, From Discourseto Logic, 1993, Kluwer, Dordrecht.

Additional Information:
The basic framework we intend to develop is meant to be consonant both with the
DRT
and the DPL tradition. To realize our aim we formulate a theory of constructing
dynamic
contexts. The mathematical tools for the construction are monoids and some low l
evel
category theory. All thenecessary mathematical machinery will be introduced duri
ng the
course.

Nearness and Syntactic Influence Spheres
Name of Lecturer: Marcus Kracht
Affiliation: Department of Mathematics, FU Berlin
E-mail: kracht@math.fu-berlin.de
Type: Advanced

Brief course description:
We will discuss ways to formalize Government and Binding theories using concepts
of
nearnessin syntactic trees.
Course Prerequisites:
Knowledge of syntax, in particular some version of GB is highly recommended. We
will
assume familiarity with the `classical' GB theory of the eighties. The formal to
ols are
mainly from formal language theory, lattice theory and logic, in particular moda
l logic.
Course material will be available. Those who wish to prepare in advance should p
erhaps
read B. A. Davey and H. A. Priestley: Introduction to Lattices and Order Cambrid
ge
University Press, 1991.

Feature Logics and Infinitary Descriptions
Name of Lecturer(s)/Organizer(s): Bill Keller
Affiliation(s): School of Cognitive and Computing Sciences, University of Susse
x
E-mail: billk@cogs.susx.ac.uk
Type: Advanced

Brief course description:
Constraint and unification based approaches to grammar such as FUG, GPSG, LFG,
HPSG and PATR have become increasingly popular in computational linguistics beca
use
of their flexibility and descriptive power. These approaches have developed an i
mportant
notion of feature structure that plays a key role in the representation of lingu
istic
information. This course begins with a survey and comparison of work on the
formalization of feature structures and their constraint languages in contempora
ry
approaches to grammar. The survey will include the denotational model of PATR
introduced by Pereira and Shieber, Rounds-Kasperfeature logic and its extensions
(the
treatment of negation, Carpenters logic of typed featurestructures) Johnson's
Attribute-Value languages, Smolka's language of feature terms and Blackburn's mo
dal
perspective on attribute-value logics.
The remainder of the course explores the consequences of introducing infinitary
operators ondescriptions of feature structures. Such operators include the unive
rsal
modality found in the category constraint language $L_{C}} (Category Logic) deve
loped
by Gazdar et al., and the functional uncertainty mechanism proposed by Kaplan an
d
Zaenen as a novel approach to the description of unbounded dependency constructi
ons
in LFG. The language of Regular Rounds-Kasper logic is introduced and it is show
n that
infinitary descriptions permit the representation of entire grammars within a un
iform
logical framework.
Course Prerequisites:
Propositional and predicate logic. Familiarity with unification based grammars.

A Mathematical Theory of Language Learnability
Name of Lecturer(s)/Organizer(s): Dick de Jongh and Makoto Kanazawa
Affiliation(s):Department of Mathematics and Computer Science,University of
Amsterdam
Department of Cognitive and Information Sciences, Faculty of Letters
, Chiba
University
E-mail: dickdj@fwi.uva.nl and kanazawa@CogSci.L.chiba-u.ac.jp
Type: Advanced

Brief course description:
A theory of language learnability in the tradition of Gold (1967), Blum and Blum
(1975),
Angluin(1980), Osherson, Stob, and Weinstein (1986), known as `inductive inferen
ce' or
`identificationin the limit'. The course covers: (i) Background in recursion the
ory; (ii)
Review of the main results in the general recursion-theoretic framework; (iii) R
ecent
developments concerning learnability of uniformly recursive classes of languages
; (iv)
Learnability of classes of formal grammars of various kinds, with an emphasis on
categorial grammars.
Course Prerequisites:
Rudiments of logic and formal language theory, equivalent to an undergraduate co
urse
in each.

Complex Part Structures and Natural Language
Name of Lecturer: Friederike Moltman
Affiliation: Dept of Philosophy, City University of New York (CUNY)
E-Mail: ISAGC@CUNYVM.bitnet
Type: Advanced

Brief Course Description:
Many expressions in natural language learning make reference to the part structu
re of
any entety. Among the most interesting ones are 'part-structure modifiers' such
as
'together', 'alone', 'as a whole', 'Whole/wholly', 'individual/ly', and 'separat
e(ly)', and
expressions of completion such as 'completaly', 'partly', and 'almost'. Such exp
ressions
involve part structures that are more complex than the extensional mereological
notion
of part structure traditionally assumed in semantics. Most importantly, such par
t
structures involve notions of integrated whole and may be multidimensional (i.e.
consist
of differentsub-part structures in different dimensions). Moreover, they may be
restricted
only to a particular situation.
In this course, I will present formal notions of complex part structures, integr
ated whole,
and part structure dimensions, and show how, based on a notion of complex situat
ed part
structure, the various readings that part structure modifiers and expressions of
completion display, can be derived.
I will also discuss how more complex notions of part structure correspond to the
intuitive
notion of part structure and have been proposed in the older and some of the cur
rent
philosophical literature.
Course Prerequisites:
Some familiarity with the algebraic approach to semantics for natural language,
e.g. at
the level of Link's course 'Algebraic Semantics for Natural Language' (which wou
ld be
an excelent preparation for the present course).
Background Reading:
Moltman, F: 'Parts and Wholes in Semantics'. Oxford, UP, to appear(can be obtain
ed
from the author).
Simons, P. (1987): 'Parts. A Study in Ontology'. Clarendon Press, Oxford.

Properties and Types
Name of Organizer: Chris Fox
Affiliation: Department of Computer Science University of Essex
E-mail: foxcj@essex.ac.uk
Type: Workshop
Brief course description:
This workshop will seek to discuss issues in Property Theory and TypeTheory. Bro
adly
speaking, Property Theories take properties and propositions to be basic objects
. They
can provide fine-grained intensionality, needed in the semantics of natural lang
uage,
whilst avoiding some of the problems of strongly typed semantic frameworks.By av
oiding
set-theoretic conceptions of properties and propositions,the notion of `type' ca
n be
reconsidered. For example, by moving away from the view of types offered by
Montagovian-style Intensional Logic,Type Theories may be defined that can repres
ent
the dynamic phenomena that occur in natural language semantics.
Course Prerequisites:
Knowledge of: Predicate Calculus and Montague Semantics.Suggested background
reading (one of):
Turner, R. 1990, Truth and Modality for Knowledge Representation.Pitman, Chapter
5;
Turner, R. 1992 "Properties, Propositions and Semantic Theory", in(eds) Rosner,
M. and
Johnson, R., "Computational Linguistics and Formal Semantics". Studies in Natura
l
Language Processing, Cambridge University Press, pages 159--180.

Formal epistemology
Name of Organizer: Hans Rott
Affiliation(s): University of Konstanz
E-mail: pirott@nyx.uni-konstanz.de
Type: Workshop

Brief course description:
Five talks given by people from AI and philosophy will give an overview of some
recent
stategies pursued in the formal representation of the statics and dynamics of kn
owledge,
belief and expectation.
Course Prerequisites:
Elementary logic.
Background reading (not obligatory):
Keith Lehrer, Theory of Knowledge, Westview Press 1990; Peter Gardenfors, Knowle
dge
in Flux,MIT Press 1988; Dov Gabbay et al. (eds.), Handbook of Logic in Artificia
l
Intelligence andLogic Programming, Vol. 3 and 4, Oxford UP 1994.

Underspecification in computational semantics
Name of Organizer: Robin Cooper and Massimo Poesio (FraCaS)
Affiliation(s): University of Edinburgh
E-mail: {cooper,poesio}@cogsci.ed.ac.uk
Type: Symposium

Brief course description:
This symposium will present work from the LRE project FraCaS ("A Framework for
Computational Semantics") relating to the theme of underspecification in semanti
cs. The
aim will be to provide an overview of treatments of underspecification in variou
s
approaches examined by FraCaS and to relate these to a general account of
underspecification. Planned contributionsinclude the following:
-Ambiguity, Vagueness, and Underspecification - Manfred Pinkal (Saarbrucken)
-Underspecification in Quasi-Logical Form - Dick Crouch (SRI Cambridge)
-Underspecification in DRT - Josef van Genabith/Hans Kamp (Stuttgart)
-Underspecification in situation semantics - Robin Cooper (Edinburgh)
-Underspecification in Conversation Representation Theory - Massimo Poesio
(Edinburgh)
-A Unified Look at Underspecification from the Perspective of Dynamic Logic - Ja
n
Jaspers/Janvan Eijck (CWI, Amsterdam)

SECTION: LOGIC
Introduction to Modal Logic
Name of Lecturer: Ramon Jansana
Affiliation: Department of Logic, History and Philosophy of Science, Universitat
de
Barcelona, Spain.
E-mail:jansana@cerber.ub.es
Type: Introductory

Brief course description:
The aim of the course is to introduce the student to the most important systems
of
propositional modal logic, their semantics and to the methods for studying their
properties. Special emphasis will be given to Kripke semantics and algebraic sem
antics,
and to the techniques of canonical models for proving completeness and filtratio
ns for
proving decidability. Some information on correspondence theory and duality theo
ry will
also be given.
Course prerrequisites:
Knowledge of first order logic and basic Set Theory.

Temporal Logic
Name of Lecturer(s)/Organizer(s): Ian Hodkinson
Affiliation(s): Imperial College London
E-mail: imh@doc.ic.ac.uk
Type: Advanced

Brief course description:
The course will cover aspects of two of the main areas of temporal logic researc
h,
namely axiomatisations and expressivity. I hope to cover the following:
Axiomatisations of various connectives (F, P, Until, Since) over various flows o
f time,
such as the real numbers. The use of Gabbay's "irreflexivity" rule. Possibly, Sa
hlqvist's
theorem. Expressive completeness. Examples (such as Kamp's theorem that Until &
Since are expressively complete over Dedekind complete linear time). Gabbay's
separation theorem. Connections to finite variable logics. Possibly, extended (e
.g.,
fixed-point) temporal logics, and second-order expressiveness.
Course Prerequisites:
Some sympathy with the mathematical approach to temporal logic, and familiarity
with
basic completeness theorems in modal or temporal logic, would be an advantage. T
he
last half of the survey "Finite variable logics" (Hodkinson, Bull. Europ. Assoc.
Theor.
Comp. Sci. 51 (1993) 111-140) introduces expressive completeness. Much of the co
urse
is covered in "Temporal Logic, Volume 1" (Gabbay, Hodkinson and Reynolds, Oxford
University Press, 1994).

Descriptive Complexity Theory
Name of Lecturer(s)/Organizer(s): Flum, Joerg
Affiliation(s): Mathematisches Institut, Universitaet Freiburg,
E-mail: flum@sun1.ruf.uni-freiburg.de
Type: Advanced

Brief course description:
We study logics which have been introduced to characterize complexity classes
Course Prerequisites:
Basic knowledge in logic

Algebraizations of sentential logics
Name of Lecturer: Josep Maria Font
Affiliation: Faculty of Mathematics, University of Barcelona
E-mail: font@cerber.mat.ub.es
Type: Advanced
Brief course description:
In this course, one of a more informative rather than in-depth character, we wil
l survey
some of the most prominent old or new general approaches to the issue of finding
connections between sentential logics and classes of algebras, or more generally
of
algebraic structures. We will distinguish them by the kind of algebraic objects
they
consider, by the strength of the connections they establish, and by the scope of
logics
which they apply to. Examples of logics with their algebraization, applications
to
correspondence between logical and algebraic properties, and relationships betwe
en
different approaches will be mentioned.
Course Prerequisites:
A general background on first-order logic, and some acquaintance with as many
sentential logics as possible.

Some new trends in Algebraic Logic
Name of Organizer: Josep Maria Font
Affiliation: Faculty of Mathematics, University of Barcelona
E-mail: font@cerber.mat.ub.es
Type: Workshop

Brief course description:
Several young researchers will report in some detail on their work in two (relat
ively new)
areas of Algebraic Logic. One explores the connections between AL and Proof Theo
ry,
through the algebraization of Gentzen systems of the most general kind (i.e., in
cluding
those for substructural logics and those for multiple-valued logics). The other
belongs to
Algebraic Model Theory, a field started by Czelakowski, Blok and Pigozzi, and wh
ich
develops a model theory for languages without equality (specially of their unive
rsal Horn
fragment) with a strong algebraic-logic flavour; it incorporates intuitions and/
or
techniques from traditional sentential logic (the theory of matrices), from equa
tional and
quasi-equational logic and universal algebra (the theories of varieties and
quasi-varieties), and from classical model theory of first-order languages.

Advanced Modal Logic
Name of Lecturers: Patrick Blackburn, Maarten de Rijke and Y de Venema.
Affiliation: Patrick Blackburn, (Universitat des Saarlandes, Saarbrucken),
Maarten de Rijke (CWI, Amsterdam), and Yde Venema (Vrije Universiteit,
Amsterdam).
E-mail: patrick@coli.uni-sb.de, Maarten.de.Rijke@cwi.nl and yde@cs.vu.nl
Type:Advanced

Brief Course Description:
We wish to teach a number of advanced topics in modal logic that are important i
n the
field of Logic, Language and Information, but that are not accessible to the ave
rage
student.
Concentrating on areas that have hitherto been largely inaccessible to students.
In
particular, we propose to give a detailed treatment of the Sahlqvist Correponden
ce
Theorem, the Bisimulation Theorems, the relation between completeness and canoni
city,
and the computational complexity of modal formalisms.
The course will be based around Modal Logic. A Textbook, an advanced textbook on
Modal Logic written by the course presenters. (The final Version of the book wil
l be
ready by the time of the Barcelona summer school, and will be distributed among
the
course participants.)
In various guises (for example temporal logics, propositional dynamic logic, fea
ture logic)
modal logics are proving increasingly important in applications ranging from the
analysis
of presuppositions and process theories to database theory and information retri
eval.
However, many of the techniques required to analyse such (multi)modal systems ar
e
known only to experts in the field. In particular, the important notions of expr
essivity
analysis via bisimulations, and the recent results in computational complexity,
remain
inaccessible to the average student.
We believe that the following advanced topics should be part of the tool kit of
anyone
working with modal formalisms; the aim of the course is to introduce them in det
ail.
Course prerequisites:
As should be clear from the above description, this is an advanced course. We wi
ll
assume that the participants have a good working knowledge of first-order logic.
Furthermore, although we will briefly introduce the basic ideas underlying propo
sitional
modal logic, this is essentially to fix the notation for the various generalised
formalisms
we will be working with; prior acquaintance with propositional modal loigc will
almost
certainly be necessary to follow the course.

Situation Theory with Applications
Name of Organizer: Keith Devlin
Affiliations: Saint Mary's College of California, Moraga, and CSLI, Stanford.
E-mail: devlin@stmarys-ca.edu
Type: Symposium

Brief course description:
Present-day situation theory (which differs enormously from the original Barwise
and
Perry conception of twelve years ago) provides a framework for the description a
nd
analysis of various issues concerning information, including formal and natural
language
semantics, communication (including social factors), formal and everyday reasoni
ng,
computation,and the design of information systems. The symposium will begin with
ashort introductory series of lectures outlining the basic notions ofsituation t
heory, and
end with talks on applications to natural language semantics and information-sys
tem
design.
Course Prerequisites:
The lectures and talks will assume a general background (but no particular exper
tise) in
logic, computer science, and linguistics. Attendants will gain much more from th
e
symposium if they have already read Keith Devlin's book "Logic and Information"
(Cambridge University Press, 1991), also available in German translation as "Inf
ons und
Infone" (Birkhauser, 1993). However, the introductory lectures will present the
basic
material in a slightly different way, to reflect recent developments in situatio
n theory.
BIBLIOGRAPHY
[1] Barwise, J. and Perry, J. Situations and Attitudes, Bradford Books, MIT Pres
s (1983).
[2] Devlin, K. Logic and Information, Cambridge University Press (1991).
[3] Devlin, K. and Rosenberg, D. Situation Theory and Cooperative Action, in Acz
el, P.
Israel, D, Katagiri, Y. and Peters, S. (editors), Situation Theory and its Appli
cations,
Volume 3, CSLI Lecture Notes 37 (1993), pp.213--264.
[4] Devlin, K. and Rosenberg, D. Networked Information Flow via Stylized Documen
ts,
CSLI Report 94--187 (1994).
[5] Gawron, P. and Peters, S. Anaphora and Quantification in Situation Semantics
, CSLI
Lecture Notes 19 (1990).

SECTION: COMPUTATION AND LOGIC

Logical Frameworks
Name: D. Basin, S. Mathews (Saarbruecken)
E-mail: basin@mpi-sb.mpg.de and sean@mpi-sb.mpg.de
Type: Introductory

Brief course description:
We introduce the logical foundations and practical applications of logical frame
works.
We begin by considering the question "What is a logic" and show how answers to t
his
question lead to various kinds of frameworks and theories for encoding logics. W
e
present several of the more prominent approaches including those based on type t
heory
(the Edinburgh LF), higher-order logic (Isabelle), and finitary inductive defini
tions
(Feferman's FS0). We present applications of these frameworks including encoding
standard logics, modal logics, and program verification and synthesis. We examin
e the
relationship between logical frameworks and methods for extending logics with ne
w rules
and inference procedures. We also consider limitations of the various approaches
.
Course Prerequisites:
Familiarity with basic first-order logic is required, e.g., at the level the fir
st part of Van
Dalen's "Logic and Structure" or Mendelson's book "Introduction to Mathematical
Logic".
Some familiarity with basic aspects of the lambda-calculus --- e.g., syntax, red
uction, and
typing --- would be helpful but relevant details will be reviewed. For backgroun
d reading
see either the book of Nordstrom, Petersson, and Smith on "Programming in Martin
-Lof
Type Theory" or Thompson's book "Type Theory and Functional Programming".

Knowledge Representation and Logic
Name of Lecturer(s)/Organizer(s): Franz Baader
Affiliation(s): LuFG Theoretische Informatik RWTH Aachen
E-mail: baader@informatik.rwth-aachen.de
Type: Introductory

Brief course description:
Traditionally, the role of logic in KR was often restricted to providing a forma
l
declarative semantics for the representation formalism, whereas the inference pr
ocess
was realized by (ad hoc and often incomplete) algorithms. The course will focus
on
terminological KR languages in the tradition of KL-ONE, for which this restricti
on on
the role of logic no longer holds. After an introduction into terminological KR
formalisms, we define the relevant inference problems, and show that well-known
methods and results from logic (such as tableau methods, decidable prefix-classe
s, results
for propositional modal logic) can be employed to derive sound and complete algo
rithms
for these inference problems. Subsequently, we consider the problem of integrati
ng other
means of representation (such as default rules, modal logics for belief and know
ledge)
into terminological KR formalisms.
Course Prerequisites:
Basic notions of logic; familiarity with KR formalisms such semantic networks an
d frames
are helpful (but not mandatory)

Deductive Database
Name of Lecturer(s)/Organizer(s): Gerhard Koestler
Affiliation(s): Universitaet Augsburg
E-mail: koestler@uni-augsburg.de
Type: Introductory

Brief course description:
The confluence of logic programming and databases is a general trend in computer
science combining the flexibility of logic programming with the ability of datab
ases to
store and administrate huge amount of data safely and efficiently. Deductive dat
abase
systems --- as an instantiation of this trend --- are characterized as systems w
ith the
ability to use a logic programming language called Datalog to express deductions
on the
content of a relational database. They extend the expressive power of relational
query
languages like SQL by allowing the definition of recursive views and by integrat
ing rule
knowledge in databases. The mismatch in current relational database applications
of
using a procedural host language like C and embedded SQL is diminished. The cour
se
will be organized in five sessions, keeping the balance between theory and pract
ice:
-Basics of logic programming languages and deductive databases: syntax, semantic
s and
evaluation of Datalog
-Optimization techniques: delta iteration, magic sets and other sorcery -Negatio
n in
Datalog
-Aggregation and subsumption in Datalog
-Existing deductive database systems
Course Prerequisites:
First order predicate calculus

Programming with Temporal Logic
Name: Howard Barringer
Dov Gabbay
Affiliation(s):The University of Manchester Imperial College of Science, Technol
ogy and
Medicine
E-mail:howard@man.cs.ac.uk and dg@doc.ic.ac.uk
Type: Advanced

Brief course description:
This course will introduce the rudiments and examples of programming with
connective-based temporal logic. Although a brief survey of various approaches t
o
temporal logic programming will be provided, the course will emphasise an approa
ch
based upon a synthesis of declarative and imperative readings of temporal logic
formulae, as has been embodied in the MetateM system. Examples in the areas of
software engineering and AI will be used to motivate the use of this particular
temporal
logic programming paradigm.
Course Prerequisites:
Some familiarity with connective-based temporal logic will be advantageous. Like
wise,
some knowledge of Logic Programming will be helpful. Good background in predicat
e
calculus essential.

Power Structures and Program Semantics
Name of Lecturer: Chris Brink.
Affiliation: Mathematics Department, University of Cape Town.
Email: cbrink@maths.uct.ac.za.
Type: Advanced.

Course Description
There are various approaches to program semantics. E.g: The model views a progra
m
as a binary input-output relation over a state space. Predicate transformer sema
ntics
characterises a program by its action on predicates (properties of states). Doma
in theory
structures the state space by an information ordering, and views a program as a
domain
morphism. Information systems offer a Gentzen-style inference relation on units
of
information, and characterise programs as information-preserving maps. The aim o
f this
course is to show how these various approaches relate to each other. The emphasi
s is on
fundamental ideas rather than technical details. The basic paradigm we adopt com
es
from non-classical logics, where any well-behaved logic characterises and is cha
racterised
both by an algebra and by a relational structure (its Kripke semantics). In our
case we
take as basic the relational model of program semantics, generalised to a topolo
gical
space. Then information systems correspond to a logical presentation of program
semantics, predicate transformers to an algebraic presentation, and domain theor
y to an
order-theoretic presentation. We show how to translate between these presentatio
ns
using two basic tools: power constructions, and Priestley duality.
To put all this again in a wider context, we relate these ideas to verisimilitud
e -- the
notion of one theory being closer to the truth than another.
Course prerequisites: Participants should have some background in logic, and be
familiar
with at least one of the versions of program semantics mentioned in the course
description. Basic knowledge of lattice theory and topology would be useful.

Automated Deduction in Non-Classical Logics
Name of Lecturer(s)/Organizer(s): Lincoln A. Wallen
Affiliation(s): Oxford University
E-mail: lincoln.wallen@comlab.ox.ac.uk
Type: Advanced

Brief course description:
The logical basis for modern methods of automated deduction in nonclassical logi
cs will
be developed; this includes Kripke structures and completeness proofs, Gentzen m
ethods
and their proof theory, and matrix theorems and unification. These basic results
will be
used to develop both direct computational methods such as Connection and Tableau
x
methods, useful for interactive proof-search, and various indirect methods such
as Inverse
and Resolution methods, useful for fully automated proof-search. Throughout the
course
the main examples will be drawn from intuitionistic and normal modal logics.
Course Prerequisites:
-Some familiarity with Predicate Logic: up to and including the idea of complet
eness for
some formal calculus.
-Some familiarity with Computational Logic: unification and subgoaling, as prov
ided by
limited exposure to logic programming (PROLOG) for example.
-Some familiarity with such mathematical methods as: induction, definition by r
ecursion,
and the beginnings of abstract algebra.
Suitable preparatory texts:
Logic for Computer Science, J.H. Gallier, Harper and Row, 1986. ISBN 0-06-042225
-4.
(very comprehensive)
A Mathematical Introduction to Logic, H.B. Enderton, Academic Press, 1972. (Chap
ters
1 and 2.) (comprehensive on the logical basics, no computational material)

How to tame you logic?
Maarten Marx:
University of Amsterdam
E-mail: marx@ccsom.uva.nl
Szabolcs Mikulas
Logic Graduate School
E-mail: mikulas@fwi.uva.nl
E-mail: h3762mik@ella.hu
Type: Advanced

Keywords:
Classical and modal logic, algebraic logic, universal algebra, algebras of relat
ions,
completeness, decidability
Target group: anybody who is interested in logic and/or its applications, logici
ans,
linguists, theoretical computer scientists
Prerequisits: basics of classical first-order and modal logic, some familiarity
with
(universal) algebra is useful
Course material: collection of related papers
Course description: Our aim is to introduce a purely logical problem with comput
ational
and linguistic relevance, and to provide several ways to handle it. Doing this,
the efficient
tools of algebraic logic are widely used; in this sense, this course would be a
continuation
of the course on algebraic logic and its application to logic given by Andreka,
Nemeti
and Sain at ESSLLI '93. There are many well-investigated and useful logics that
do not
behave nicely in certain respects. Such examples are the undecidability of class
ical
first-order logic, FOL, incompleteness of its finite variable fragment, Ln, and
the
incompleteness and undecidability of the strongest version of arrow logic invent
ed by van
Benthem and Venema. Following the ideas of van Benthem and Nemeti, we will discu
ss
how to "tame" logics so that they behave in a nice way. That is, we will reformu
late
several logics and show that these new versions have desirable properties such a
s
completeness, decidability, Beth definability property, Craig interpolation prop
erty,
deduction theorem etc. One way to achieve this is to find out the "computational
core"
of a logic and forget about some inessential principles. In this way, van Benthe
m found
a weakened version of FOL that is complete and decidable by a result of Nemeti.
Another, more semantical, approach is to widen the class of models for a logic.
In this
way, called relativization, Ln (Nemeti) and arrow logic (Marx, Mikulas, Nemeti,
Sain,
Simon) are proved to have nice (complete and decidable) versions. Although the a
bove
mentioned weakened versions are interesting and applicable, usually the price of
the nice
behaviour is the loss of the expressive power of the logic. Fortunately, algebra
ic logic,
where the relativization idea is coming from, provide us powerful methods to str
engthen
logics. For instance, we reintroduce connectives that are no more definable afte
r
weakening without losing the nice properties. Such example is arrow logic: weake
ning
the associativity axiom causes that the universal modality is not definable, but
it can be
added as a primitive connective without loosing completeness and decidability. T
he
above list is only a sampler. Both of us (in cooperation with Dutch and Hungaria
n
logicians) are working in this field; we investigate more logics and more proper
ties.

Expressive Nonmonotonic Reasoning (NMR)
Name: Franz Baader
Affiliation(s): LuFG Theoretische Informatik, RWTH Aachen
E-mail: baader@informatik.rwth-aachen.de
Name: Karl Schlechta
Affiliation(s): Laboratoire d'Informatique de Marseille, URA CNRS 178 Universite
de
Provence.
E-mail: ks@gyptis.univ-mrs.fr
Type: workshop

Brief course description:
Unlike classical logical reasoning, commonsense reasoning is often based on assu
mptions
that may ultimately be shown to be false. This explains the need for nonmonotoni
c
formalisms, which allow to withdraw conclusions when additional (contradicting)
information is acquired. In spite of this demand for NMR formalisms, and even th
ough
there is abundant formal work on NMR, fairly few NMR systems are applied in prac
tice.
One reason is that their expressive power is often rather restricted; in particu
lar, much
of the formal work concentrated on propositional logic as base logic. One goal o
f the
workshop is to discuss the problems that occur when more expressive formalisms (
such
as full first-order logic) are used as base logic. Another interesting topic cou
ld be the
question whether expressive nonmonotonic reasoning can be tractable.
Course Prerequisites:
Basic notions of logic; familiarity with nonmonotonic reasoning formalisms such
as
Default Logic are helpful

Logical Formalisms for Planning, Plan Recognition, and Plan Modification
Dr. Susanne Biundo
German Research Center for Artificial Intelligence
email: biundo@dfki.uni-sb.de
Type: Workshop

Description:
The workshop will be divided into five sections. Each section starts by an invit
ed talk
presenting current but completely elaborated work in the field.
1. Knowledge Representation
This part is devoted to the use of various logical formalisms for knowledge
representattion in planning. In particular, temporal and modal logics are consid
ered.
2. Plan Formation
Planning methods based on logical inference procedures are presented, in partic
ular
deductive planning approaches.
3. Plan Modification
This section considers logic-based plan modification techniques and the formal
generation and maintenance of plan libraries.
4. Plan Recognition
This section is devoted to plan recognition based on logic. Approaches to forma
l user
modeling are also considered.
5. Tactical Planning
In this section approaches to the use of tactical theorem proving techniques fo
r
reasoning about plans are introduced.

Methods for Constructing and Manipulating Logical Systems
Prof. Dov M GABBAY
Imperial College of Science, Technology and Medicine
EMAIL: dg@doc.ic.ac.uk
SUBJECT Logic Engineering : Labelled Deductive Systems,and computer support for
studying logics
Type: Symposium

Course Description:

Course Prerequisites:

SECTION: COMPUTATION

Partial evaluation
Name of Lecturer(s): Neil Jones
Affiliation(s): DIKU, University of Copenhagen
E-Mail: neil@diku.dk
Type: Introductory

Course Description
A partial evaluator is an algorithm which, when given a program and some of its
input
data, produces a so-called specialized program. Running the specialized program
on the
remaining input data will yield the same result as running the original program
on all of
its input data.
These lectures concern the practical realization and application of partial eval
uation. It
can be thought of as a special case of program transformation, but emphasizes fu
ll
automation and generation of program generators as well as transformation of sin
gle
programs.
It is well known that a one-argument function can be obtained from a two-argumen
t
function by specialization, i.e.\ by fixing one input to a particular value. In
logic this is
called `currying', in analysis it is called `restriction' or `projection'. Parti
al evaluation,
however, works with program texts rather than mathematical functions, and a more
precise analogy is with Kleene's `s-m-n theorem' in recursive function theory.
Partial evaluation sheds new light on techniques for program optimization, compi
lation,
interpretation, and program generators including the automatic generation of com
piler
generators by `self-application'.
The application of partial evaluation is not restricted to compiling and compile
r
generation. The range of potential applications is extremely large. For example
it has
been seen to yield significant speedups in pattern recognition, computer graphic
s by `ray
tracing,' and scientific computing. An overview will be given of characteristics
that make
a problem susceptible to improvement by partial evaluation, and delineate certai
n
pitfalls, e.g.\ loss of efficiency, or code explosion.
Course Prerequisites:
Maturity in programming languages, eg with compiler construction and/or semantic
s.
Background reading
The lectures will cover material from the book Partial Evaluation and Automatic
Program
Generation by Neil D.\ Jones, Carsten Gomard and Peter Sestoft, Prentice Hall
International, 1993.

Interactive development of proofs and programs
Name of Lecturer(s): Christine Paulin-Mohring
Affiliation(s): LIP-CNRS URA 1398-Ecole Normale Superieure de Lyon
E-mail: Christine.Paulin@lip.ens-lyon.fr
Type: Advanced

Brief course description:
The course will describe the constructive interpretation of intuitionnistic logi
c and its
application to the development of programs which are certified to be correct. Th
is
methodology will be illustrated with examples developped in the Coq Proof Assist
ant.
Course Prerequisites:
Some knowledge about logic (first-order arithmetic, natural deduction)
Some practice of functional programming languages such as ML or some theoretical
knowledge about lambda-calculus (substitution, reduction).

Implementation of functional programming languages
Name of Lecturer(s): Simon Peyton Jones
Affiliation(s): Department of Computing Science, University of Glasgow
E-mail: simonpj@dcs.gla.ac.uk
Type: Advanced

Brief course description:
The functional-language implementation area has developed very rapidly over the
last
five or ten years. The net result is that our implementations are at last becomi
ng
competitive; that is, usable for real applications. This is not just a quantitat
ive
improvement; it is now leading to a qualitative shift in culture and priorities.
Instead of
counting combinator reductions we now just beginning to count machine instructio
ns,
memory loads and stores, and cache misses. Instead of measuring nfib we now meas
ure
real applications. Instead of comparing our implementations with themselves we n
ow
compare them with C. Compilation time has become a more important issue than run
time for many applications. Profiling tools have become a priority. So has decen
t
input/output.
The course will explore some of ideas which have made this transition possible.
Rather
than give a waffly high-level survey, I will focus on the insights and practical
lessons I
have learned from the design and implementation of a particular Haskell compiler
. At
the same time, I will try to avoid being too parochial, and will draw out genera
l lessons.
I will use two unifying themes. The first is the interplay of theory and practic
e: it is
suprisingly often possible to exploit apparently abstract theory to give direct
implementation pay-offs. The second is the simple but powerful idea of em compil
ation
by transformation: the idea that (a large part of) the compilation process can b
e
regarded as the composition of a series of small transformations.
The course will not be a re-hash of either of my books on the subject.
I'll cover at least some of these topics: general issues and intermediate langua
ge;
second-order lambda calculus; unboxed types; useful transformations; strictness
analysis;
deforestation; state; storage management; the run-time system; profiling. The ma
in
omission is that I will not cover the Haskell type class system and its various
implementation requirements.
Course Prerequisites:
I will assume that you are familiar with functional programming. It would be a h
elp if
you had read some papers or books on implementation of functional languages (for
example [1], [2] or [3])
[1] Peyton Jones "The implementation of functional programming languages" Prenti
ce
Hall 1987
[2] Peyton Jones and Lester "Implementing functional languages: a tutorial" Pren
tice Hall
1992
[3] Plasmeijer and van Eekelen "Functional programming and parallel graph reduct
ion"
Addison Wesley 1993.

SECTION: LANGUAGE AND COMPUTATION

Formalizing and implementing syntactic theories
Name of Lecturer(s)/Organizer(s): Edward Stabler and Mark Johnson
Affiliation(s): Edward Stabler, UCLA and Mark Johnson, Brown University
E-mail: stabler@cognet.ucla.edu and mj@cs.brown.edu
Type: Introductory

Brief course description:
This class will briefly present some of the ideas and problems in formalizing an
d
implementing syntactic theories. The class will focus on examples from the "Gove
rnment
and Binding" (GB) tradition in syntax, but the basic problems facing this theory
face all
constraint-based theories. We show why, even with purist approaches to parsing w
ith a
collection of GB principles and parameters (or any sort of constraints), process
ing should
be organized around phrase structure constituency (or some similar local domains
). In
light of this perspective, we show how some generalizations of traditional parsi
ng theory
extend to constraint-based parsing strategies, providing a wide range of process
ing
options which has not previously been explored.
Course Prerequisites:
Acquaintance with basic GB theory, logic and parsing theory will be helpful.
Background reading:
Stabler "The Logical Approach to Syntax",
Pereira & Shieber "Prolog and Natural Language Analysis".

Corpus-Based Models of Language Processing
Name of Lecturer(s)/Organizer(s): Rens Bod and Remko Scha
Affiliation(s): University of Amsterdam Institute for Logic, Language and Comput
ation
Department of Computational Linguistics
E-mail: rens@alf.let.uva.nl
scha@alf.let.uva.nl
Type: Introductory

Brief course description:
When computational models of language processing are not constructed in a purely
linguistic context, but aim at being relevant for practical applications or for
psychological
theory, they ought to be able to perform tasks like disambiguation and predictio
n. For
this reason, increasingly many models take statistical properties of a sample co
rpus into
account when they process new input. This course will give an overview of the mo
st
important approaches that are used in such corpus-based language models.
The course starts out with a short introduction to probability theory, and conti
nues with
the following topics:
- Markov Models
- The challenge of integrating structural and statistical information
- Stochastic Grammars
- Annotating corpora
- Data-driven parsing
- A corpus as a sample of a larger population: the problem of unknown words
- Statistical semantic interpretation
- Interfacing language models and speech understanding
Course Prerequisites:
Basic linguistic background will be assumed.
Previous knowledge of statistics or corpus techniques will not be assumed.

Automated deduction for logics of linguistic resources
Name of Lecturer(s)/Organizer(s): Michael Moortgat
Affiliation(s): Research Institute for Language and Speech (OTS)
E-mail: moortgat@let.ruu.nl
Type: Advanced

Brief course description:
This course offers a computational perspective on Lambek-style categorial type l
ogics.
We start with a discussion of the problem of spurious ambiguity for various
proof-theoretic presentations (combinator, natural deduction, Gentzen-style) of
standard
systems in the categorial hierachy and present efficient theorem proving techniq
ues. Then
we move to the parsing-as-deduction problem in current `mixed' or `multimodal'
categorial frameworks. We develop uniform algorithmic proof theory for these
linguistically more discriminating systems, combining insights from the Proof Ne
t
(Girard) and Labelled Deduction (Gabbay) approaches.
Course Prerequisites:

Familiarity with the logical and linguistic aspects of categorial type logics.
Background Reading:
Michael Moortgat (1994) `Categorial Grammar'. In Van Benthem and ter Meulen (eds
)
Handbook of Logic and Language. Elsevier Science Publishers.
Glyn Morrill (1994) Type Logical Grammar. Categorial Logic of Signs. Kluwer.

Topics in Lexical-Functional Grammar
Name of Lecturer(s)/Organizer(s): Ronald Kaplan and Mary Dalrymple
Affiliation(s): Xerox PARC
E-mail: kaplan@parc.xerox.com ; dalrymple@parc.xerox.com
Type: Advanced

Brief course description
This course will explore the descriptive mechanisms of LFG for solutions to part
icular
linguistic problems. LFG resembles other unification-based grammar formalisms in
its
use of equations over attribute-value structures, but has introduced additional
extensions
and modifications to capture formal generalizations about particular linguistic
phenomena. Among other issues, we will examine the use of functional uncertainty
to
encode long-distance dependencies and anaphor-antecedent relations, the use of o
ff-path
constraints to state restrictions on these long-distance relations, the treatmen
t of complex
predicates and modification by means of the restriction operator, and the treatm
ent of
word order restrictions by means of f-precedence.
Course Prerequisites:
No prerequisites, but the course will complement the one taught by Joan Bresnan
earlier
in the institute.

Categorial Syntax and Semantics
Name of Lecturer(s)/Organizer(s): Bob Carpenter* and Glyn Morrill**
Affiliation(s): Philosophy Department, Carnegie Mellon University*
Department of Computer Systems and Languages, Polytechnic University of Cataluny
a**
E-mail: carp@lcl.cmu.edu* and morrill@lsi.upc.es**
Type: Advanced

Brief course description:

We intend to offer a broad but self-contained view of the state of the art in ty
pe logical
categorial grammar as regards the development of formal fragments with model the
oretic
semantics. Emphasis will be on linguistic applications of the general techniques
to
phenomena such as coordination, quantifier scope, plurals, extraction, bound ana
phora
and constraints. Starting from the classical categorial calculi of Lambek we sho
w what
properties are suitable for linguistic formalism, what limitations exist, and ho
w we can
proceed to tackle significant linguistic issues while preserving a clean theoret
ical
perspective. Course material will be drawn from sources including Type Logical
Grammar (Morrill, 1994, Kluwer) and Lectures on Type Logical Semantics (Carpente
r,
1995, MIT Press).
Course Prerequisites:
Familiarity with formal syntax and semantics and the puzzles these present.

Lambek Calculus and Linear Logic
Name of Lecturer(s)/Organizer(s): V. Michele ABRUSCI
Affiliation(s): Universitat di Roma La Sapienza
Dipartimento di Studi Filosofici ed Epistemologici
E-mail: abrusci@sci.uniroma1.it
Type: Advanced

Brief course description:
1. Linear Logic : sequent calculus, natural deduction (proof nets), phase semant
ics,
semantics of proofs, semantics of cut-elimination.
2. Lambek Calculus as a fragment of intuitionistic noncommutative linear logic.
3. Use of full noncommutative linear logic (extension of Lambek calculus) in lin
guistics.
Course Prerequisites:
Classical and intuitionistic logic.
Categorial grammar.

Natural Language Generation
Name of Lecturer(s)/Organizer(s): Robert Dale
Affiliation(s): Microsoft Institute of Advanced Software Technology
E-mail: rdale@microsoft.com
Type: Advanced

Brief course description:
Researchers in computational linguistics and natural language processing often f
orget
that we need theories and models of how language can be produced as well as how
it can
be understood. This course provides an introduction to research in natural langu
age
generation: the course will cover what are generally seen as the two main subare
as of the
field, text planning and linguistic realisation; we'll look closely at the probl
ems that arise
in these areas and provide an overview of the solutions that have been proposed
in the
literature. We'll also look at the problem of referring expression generation in
some
detail.
Course Prerequisites:
Some prior knowledge to the concepts that arise in natural language processing w
ill be
assumed. Some previous exposure to NLP through an undergraduate AI course, for
example, will be very useful; the course will also be appropriate for those who
have had
a master's level course in computational linguistics or NLP.

Formal and Computational Phonology
Name of Lecturer(s)/Organizer(s): T. Mark Ellison
Affiliation(s): University of Edinburgh
(till January, then INESC, Lisbon)
E-mail: marke@cogsci.ed.ac.uk
Type: Workshop

Brief course description:
Phonology, like many other fields in linguistics, often suffers from the problem
of
vagueness. Analyses may not be empirically verifiable, because they are inexplic
it in the
relationship they establish between description and object. Currently, there are
two
approaches to redressing this lack of exactness: providing a computational basis
to
phonology by implementation, or finding a formal basis for analyses in logic or
mathematics. This workshop will include an introduction to these two approaches,
as well
as presentations of on-going work situated within them.
Course Prerequisites:
A modicum of phonology as well as computation, mathematics and/or logic

The Computational Lexicon
Name Organizer: M.Felisa Verdejo
Affiliation(s): Universidad Nacional Educacion a Distancia
E-mail: felisa@horacio.dieec.uned.es
Type: Workshop

Brief course description:
Research on representing and acquiring multilingual lexical knowledge for natura
l
language processing systems is becoming increasingly important. This workshop wi
ll
address a number of theoretical aspects as well as practical results. Relevant t
opics
include, but are not limited to the following:
-structure and content of a lexical entry for a multilingual computational lexic
on
-lexical representation language and operations provided to support lexical proc
esses
-sources of lexical information and semi-automatic extraction techniques to buil
d
computational lexicons
In addition to invited talks, a limited number of other parties will be allowed
to
contribute.
Those wishing to make presentations should submit abstracts (up to 6 pages, 12 p
oint
font) to the address above, to arrive by April 15,1995.
Theoretical papers addressing the basic issues, practical papers on relevant
experience and software demos are equally welcome.

Proof Theory, Labelled Deduction and Natural Language
Name of Lecturer(s)/Organizer(s): Professor Ruth Kempson
Affiliation(s): School of Oriental & African Studies, University of London
E-mail:kempson@clus1.ulcc.ac.uk
Type: Symposium

Course Description
There has been use of natural deduction systems as a technique for Natural Langu
age
(NL) parsing for some time now. But this interest has spread to all areas of gra
mmar
articulation and natural language understanding. In syntax, there is the develop
ment of
categorial grammars as logics; in semantics, natural deduction systems are bieng
developed for generalised quantifiers; in pragmatics a labelled deductive system
for
natural language understanding is being developed; Discourse Representation Theo
ry has
an associated natural deduction system and new work addresses problems of reason
ing
with under-determined objects. The philosophical interest behind this growth of
research
activity in the link between proof systems and NL structure/understanding is the
new
perspective it provides for reconsidering the relation between formal systems
underpinning knowledge of language and and formal systems underpinning the human
capacity for reasoning. The prupose of this symposium is to bring together these
various
strands of research activity. A particular focus of interest will be the use of
the Labelled
Deductive Systems framework as a unifying methodology for the application of ded
uction
techniques to natural language.
Prerequisites:
Basic knowledge of proof systems for propositional and predicate logic (e.g. na
tural
deduction), syntax (preferably categorial grammar), model-theoretic semantics.

SECTION: LANGUAGE

Head-Driven Phrase Structure Grammar
Name of Lecturer(s)/Organizer(s): Philip H. Miller
Affiliation(s): Universite' de Lille 3
E-mail: pmiller@ulb.ac.be
Type: Introductory

Brief course description:
This intensive course will provide a general introduction to the theory of Head-
Driven
Phrase Structure Grammar (HPSG), as presented in Pollard and Sag 1994. We will c
over
the main topics of the book, and various further readings will be proposed. Deta
iled
problem sheets and solutions (based mainly on English but also on other European
languages) will be provided daily.
Course Prerequisites:
Familiarity with basic syntactic phenomena (e.g. subcategorization, unbounded
dependencies) and/or with elementary principles of set theory will be useful.

Phrase Structure Typology and LFG
Name of Lecturer(s)/Organizer(s): Joan Bresnan
Affiliation(s): Stanford University
E-mail: bresnan@csli.stanford.edu
Type: Advanced

Brief course description:
This course will focus on phrase structure typology within LFG. Topics will incl
ude
category mismatches and movement paradoxes; weak crossover, syntactic rank, and
f-precedence; apparent movement such as head movement; lexical integrity and
nonconfigurationality.
Course Prerequisites:
Some previous knowledge of syntax. In particulat, some familiarity with LFG or o
ther
non-transformational grammatical formalisms would be useful

Aspectuality and Event Structure
Name of Lecturer(s)/Organizer(s): Henk Verkuyl
Affiliation(s): Utrecht University OTS
E-mail: verkuyl@let.ruu.nl
Type: Introductory

Brief course description:
The course will start by surveying the theory on aspectuality developed in Verku
yl (1993).
This theory focusses on two things: (a) compositionality of aspectual informatio
n; (b) the
interaction between the information conveyed by the verb and by its arguments. A
s the
book discusses a lot of alternative views, these will be discussed in order to a
chieve a
broad perspective on the issues involved (Davidson, Krifka, Pustejovsky, Jackend
off,
among others.). In particular, the formalization of the (generative) Gruber/Jack
endoff
localistic perspective on change and state into model-theory will be explained a
nd be
compared with event-semantic approaches

The Acquisition of Syntax and Morphology
Name of Lecturer: Harald Clahsen
Affiliation: Department of Language and Linguistics, University of Essex
E-mail: harald@essex.ac.uk
Type: Advanced

Brief course description:

Prosody, information, and grammatical architecture
Name of Lecturer(s)/Organizer(s): Dick Oehrle
Affiliation(s): Department of Linguistics, University of Arizona
E-mail: oehrle@convx1.ccit.arizona.edu
Type: Advanced

Brief course description:
We regard the sound structure of spoken discourse as reflecting the phonological
properties of the lexical elements it contains as modulated by prosodic properti
es
typically acting on the continuous dimensions of duration, pitch, and loudness.
This
course provides an introduction to the study of prosody. The emphasis is on the
integration of prosodic properties with other aspects of sound structure, and th
e
pragmatic interpretation of prosodic properties---particularly, the placement of
pitch
accents, the choice of pitch accents, and the role of prosodic phrasing. This ph
onological
/ pragmatic duality of prosodic interpretation and the independence of prosody f
rom the
representation of lexical elements have important consequences for the study of
grammatical architecture.
We will consider a range of proposals for integrating prosody with other aspects
of
grammatical composition and examine in detail some of the proposals within categ
orial
grammar which are especially amenable to logical scrutiny and computational
implementation.
Course Prerequisites:
This course studies the phonological and pragmatic interpretation of prosody fro
m the
point of view of questions of compositionality and grammatical architecture; it
is useful
to have some background in the study of sound structure or the pragmatic interpr
etation
of prosody; more important is an appreciation of how problems of compositionalit
y arise
across different dimensions of linguistic structure, such as in the study of syn
tax and
semantics.

Constraint Based Formalisms and Grammar Writing
Name of Lecturer(s)/Organizer(s): Jochen Doerre and Suresh Manandhar
Affiliation(s): University of Stuttgart and University of Edinburgh
E-mail: Suresh.Manandhar@ed.ac.uk
Jochen.Doerre@ims.uni-stuttgart.de
Type: WORKSHOP

Brief course description:
The course will take you through a guided tour of current typed feature formalis
ms and
illustrate how HPSG grammars can be formulated in different formalisms. The firs
t part
of the course will introduce typed feature formalisms covering relational depend
encies,
functional dependencies and appropriateness conditions. Various freely available
typed
feature formalisms will be compared illustrating the different design choices th
at has
been taken and their implications for the grammar writer. The second part of the
course
will start with some basic HPSG. It will then cover how various HPSG types, prin
ciples
and schemas need to be encoded in different formalisms. The course will conclude
by
discussing current developments in typed feature formalisms.
Course Prerequisites:
Only a minimal background will be assumed. Basic familiarity with unification-ba
sed
grammars such as Shieber's CSLI lecture notes - "An Introduction to Unification
Based
Approaches to Grammar" will be assumed. Some familiarity with logic programming
will
be useful but not essential.

Argument Structure and Linking Theory
Name of Lecturer(s)/Organizer(s): Annie Zaenen (*) and Louisa Sadler (**)
Affiliation(s):(*) Rank Xerox Research Centre and (**) University of Essex
E-mail:(*) Annie.Zaenen@xerox.fr and (**) louisa@essex.ac.uk
Type: Symposium

Brief course description:
This symposium will provide an opportunity for researchers working in the curren
tly very
active field of argument structure to present some of their recent work on the
lexico-semantic structure of predicates and how these structures are mapped into
the
syntax.
Course Prerequisites:
None.

How to raise funds for coming to the European Summer Schoool in Logic,Language a
nd
Information.

The organization of the FoLLI Summer School operates on a tight budget. They wou
ld
like, but cannot, give grants to people that would like to visit the Summer Scho
ol.
Participants have to obtain funding to pay for travel, registration and accommod
ation.
This document contains a number of suggestions. The information below is 'mixed'
and
is meant for students ofall categories as well as for scholars in the area. Plea
se send us
your suggestions and experiences (folli@fwi.uva.nl), so we can set up a database
of
funding options per country. This would be a great help for your fellow students
and/or
colleagues that want to attend future editions of the Summer School.

GENERAL

- Before you start raising funds ask yourself why you want to go to Barcelona. T
he
stronger the argument for doing so, the greater the chance that you are succesfu
l in
raising funds. Also if you talk to people informally, they will ask you for your
reasons,
so be prepared for that.

- Please take into account that you may have to raise the funding from various s
ources:
sometimes you would find an organization that supplies all funds, but this is no
t the rule.
Mostly, one source will supply money IFF some other source does the same. So, yo
ur
faculty will pay 50 percent IFF the research council does the same. To bootstrap
this
so-called matched funding, you could ask your department or faculty to come up w
ith the
first (small) promise and/or you could promise to come up with some funds yourse
lf.
You can use that to persuade others to supply funds.

- Be selective in the funding organizations your approach. Choose those organiza
tions
were you have a good chance.

- Start early. Formal applications, for instance to the research council, will h
ave to be
made well in advance. Also, most organizations have restricted budgets, and you
are not
the only one who applies: all funds could already have been allocated!

- After informal consultation with the funding sources we mention below, they wi
ll mostly
ask you to write a formal application. We provide you with an example (in Englis
h)
below. If you have been invited to speak in a Workshop or Symposium, you can ask
the
organiser of that event, or the Organising Committee, to send you an official in
vitation
which uyou can use to support your application.You will most likely have to prov
ide
justification of the amopunt requested. A sample budget is given below. Furtherm
ore,
you may need letters of reference.

- A good source of information are of course people that have been to previous e
ditions
of the Summer School. Ask them how they raised their funds.

THE SOURCES
- Your own department or research institute may have a budget to fund your visit
. To
find our about this, contact the staff of your department/research institute, or
ask the
person who is in charge of international contacts. Check whether your department
or
faculty is involved in an exchange programme (e.g. TEMPUS, SOCRATES, ERASMUS).
These programs may have possibilities for visits to Summer Schools. Also, your
department might be involved in projects or networks that have funds that allow
you to
travel to the Summer School.

- The faculty to which your department belongs may have funds as well. Most facu
lties
have an employee in their office who is responsible for international affairs.

- Check with the International Office at your university whether they have funds
for your
trip. Also check with them whether they have special agreements with Spain, Cata
lonia,
or some university in the relevant region.

Furthermore, the International Office or the Students Affairs Office may have
information about the national funding possibilities in your country (research c
ouncil,
academy of sciences, private foundations). In at least some countries, this info
rmation
is available and can be consulted in electronic form.

- Most national research councils and/or national academies of science have prog
rams
for international travel. Please consult them for the procedure to follow. Mostl
y, they will
send you a form, and require you to specify a budget (see below). Note that you
will
have to apply for funds well before the event.

- In some countries, private foundations exist that might be willing to fund you
r trip.
Sometimes, these funds have specific conditions attached (i.e. you must be an un
married
(fe)male pursuing a degree in Philosophy).

- Some companies have student awards, travel money etc. You can ask them to spon
sor
your trip as well.

- If in the end you have to pay (part) of the school yourself, note that the mon
ey you
would have to pay yourself may be tax-deductible in your country.

- Below we present a sample request. You can translate this, and check the resul
t with
an experienced fund-raiser in your department.

[your address]

[their address]

[place], [date]

Concerns: sponsoring of visit to Summer School

Dear [name as you know it],

I am a (under)graduate/Ph.D. student/staff member at the department of [name] at
[university]/the faculty of [name] at [university]/ the institute [name] at [uni
versity] in
[place].

My main interest is [discipline]. This discipline is part of the larger field of
Logic,
Language and Information. In this area a large annual European Summer School in
Logic, Language and Information in organized. In 1995/1996 the School will take
place
in Barcelona/Prague. Some 400-600 persons are expected to attend the School each
year.
The school offers an excellent two-week program with some 50 introductory and
advanced courses, workshops and symposia. A visit to the school enables students
to
meet leading scientists in the field, to take courses that are not available at
their home
university, and to make important contacts for their own work.

For me personally, the school is vital since attending this school is an obligat
ory part of
my training / I can attend courses on [discipline] and [discipline] this year, w
hich are
vital to my research/training / I am invited to lecture in a workshop/symposium.
Presenting my work at this workshop presents an excellent opportunity to obtain
the view
of the international community on my work. / I will have the opportunity to disc
uss my
work with colleagues that do not gather at other occasions./ In particular, [nam
e] from
[place] who is an international expert on [discipline] will be at the Summer Sch
ool. A
visit would be an excellent opportunity to follow his/her course and discuss my
research
with him/her.
..........

Enclosed please find a budget which details my expenses and a proposal to cover
these.

I hope to have made clear to you the importance of this Summer School to my
studies/research/academic career and that you will be in a position to sponsor m
y visit.

[Signature]

- Sample budget. Please note that this is just an rough division of the costs

Budget for the visit of [name] to the 7th edition of the European Summer School
in
Logic, Language and Information, to be held in Barcelona, from till.

All figures in [my currency]
Exchange rate between Spanish Pesetas and [my currency]: [number] Pesetas =
[amount] [my currency]

Travel
- [place]-Barcelona vice versa, train/flight/buss* [amount in my currency]

Accomodation
- [number] nights times [amount per night]** = [amount in
Pesetas] =
[amount in my currency]

Registration fee
- [Pesetas]** = [amount in my currency
]

Daily expenses***
- [number] days times [ammount in my currenty]= [ammount in my currency]

Lecture Material****
- [ammount in Pesetas] = [ammount in my currency]

TOTAL EXPENDITURE [ammount in my currenc
y]

Income:

[ Sponsoring organization] [ ammount in my currency]
[confirmed/applied for]*****
[ Sponsoring organization] [ ammount in my currency]
[confirmed/applied for]
..........
TOTAL INCOME [ammount in my currenc
y]

* [consult the railway company to obtain a good estimate; there are very inexpen
sive buss
tickets to the Costa Brava; ask a travel agent for an inexpensive ticket]

** [you can find this ammount in this brochure]

***[if you have full board, you could include this figure in the accomodation. S
ome
organizations have standards per country per day. If so, us these. If not, you w
ill have to
make an estimate. [ammount in Pesetas] is realistic from the point of view of th
e costs
you will actually make in Barcelone. However, the sponsoring organization may no
t feel
the same way.

**** If you do not have an exact figure, [ammount in Pesetas] is a good guess.

***** Please indicate whether you have applied for these funds, or whether you a
re sure
that you will receive them.