Pages for course material
Each course and workshop spans a whole week (5 days). Participants are free to choose the courses they want to attend.
See also the main program of Week 1.
Each course belongs to one of the standard ESSLLI streams:
Each course corresponds to one of the following three levels:
(*) The lecture room has been changed with respect to the one reported in the booklet.
Each course and workshop spans a whole week (5 days). Participants are free to choose the courses they want to attend.
See also the main program of Week 2.
Each course belongs to one of the standard ESSLLI streams:
Each course corresponds to one of the following three levels:
(*) The lecture room has been changed with respect to the one reported in the booklet.
Discourse particles and modal adverbs form a borderline case between semantics and pragmatics and so can be the source of new insights in these areas. The use of particles has been connected with discourse relations and other coherence relations, the relation between semantic content and discourse context, especially the mutual knowledge and the common discourse goals of the discourse participants, with expression of speaker beliefs, desires and intentions and with the control of interpretations that go beyond semantic content, i.e. explicatures and implicatures. And, last but not least, particles and modal adverbs can contribute to expressive and other non-truth-conditional aspects of meaning.
The workshop is the 5th in a series on formal approaches to particles and part of the ESSLLI summer school in Bolzano, taking place from 21 to 26 August 2016. The first of these took place in 2003 at a time where it was still an adventurous idea to apply formal methods to the elusive meanings of particles. The workshops however contributed to the serious progress that has been made in the area.
| Day | Time | Authors and Title |
|---|---|---|
| Mo (22-8) | 17.00 – 17.30 | Geertje van Bergen, Lotte Hogeweg & Henk Zeevat: Introduction |
| 17.30 – 18.00 | Yael Greenberg: Interaction of a discourse particle with a metalinguistic operator: The case of x but X | |
| 18.00 – 18.30 | Adriana Osa-Gomez: Discourse particles and asymmetries in knowledge: the case of two Spanish discourse markers | |
| Tue (23-8) | 17.00 – 17.30 | Mira Grubic: Additive particles, parallelity and distinctness |
| 17.30 – 18.00 | Katja Jasinskaja & Barbara Tomaszewicz: Dissociating the scalarity and additivity of EVEN – the case of ‘čak’ and ‘čak i’ in BCS | |
| 18.00 – 18.30 | Yael Greenberg & Dina Orenstein: Typologies for even-like and for only-like particles: Evidence from Modern Hebrew | |
| Wed (24-8) | 17.00 – 18.00 | Lisa Matthewson: Towards a Landscape of Discourse Particles |
| 18.00 – 18.30 | Moria Ronen, Yael Greenberg & Galit Sassoon: A study of the Hebrew hedger ‘Begadol’ in exceptive sentences | |
| Thu (25-8) | 17.00 – 17.30 | Sophia Malamud & Allyson Ettinger: Utterance modifiers ‘ba’ and reverse‐polarity tags in the conversational scoreboard |
| 17.30 – 18.00 | Junwen Lee: A modal analysis of the Colloquial Singapore English particle ‘lah’ | |
| 18.00 – 18.30 | Upsorn Tawilapakul: Discourse Particles of Thai and Their Prominent Features | |
| Fri (26-8) | 17.00 – 17.30 | Sonja Thoma: A systematic approach to DPRT function in Miesbach Bavarian |
| 17.30 – 18.00 | Anne Bertrand, Johannes Heim, Sonja Thoma & Martina Wiltschko: Let’s all be systematic (about confirmationals), eh?\ |
|
| 18.00 – 18.30 | Eva Csipak & Sarah Zobel: Discourse particles as a window into the conditional-interrogative link |
Lisa Matthewson
Henk Zeevat (chair)
Lotte Hogeweg
Geertje van Bergen
Geertje van Bergen
Liz Coppock
Lisa Matthewson
Henk Zeevat
Lotte Hogeweg (chair)Hans-Christian Schmitz
Elena Karagjosova
Andrej Malchukov
Barbara Tomasciewicz
Katja Jasinskaja
Sebastian Loebner
Regine Eckhardt
Organisational questions pertaining to the workshop can be asked to Henk Zeevat (henk.zeevat@uva.nl), questions about the programme to Lotte Hogeweg (L.Hogeweg@uva.nl) and practical ones to the ESSLLI organisation (oc.esslli2016.unibz.it).
| Monday, 22.8 | ||
| 11.00 – 11.15 | Louise McNally, Carla Umbach |
Introduction |
| 11.20 – 11.40 | Alexandra Spalek (Oslo) Barbara Tomaszwicz (Cologne) |
Coercion in Polish versus English: processing complex lexical content (slides) |
| 11.45 – 12.30 | Barbara Partee (UMass Amherst) |
Lexical Semantics in Formal Semantics: History and Challenges (slides) |
| Tuesday, 23.8 | ||
| 11.00 – 11.30 | Staffan Larsson (Göteborg) | Connecting Language, Perception and Interaction using Type Theory with Records(slides) |
| 11.35 – 12.20 | Alessandro Lenci (Pisa) | “Going Dynamic” in Distributional Semantics(slides) |
| 12.20– 12.30 | Discussion: Distributional semantics | |
| Wednesday, 24.8 | ||
| 11.00 – 11.20 | Matthew Gotham (Oslo) | Conceptualization, Individuation and quantification(slides) |
| 11.25 – 11.55 | Stephanie Solt (ZAS Berlin) | Degree and Quantity – Semantics and Conceptual Representation(slides) |
| 12.00 – 12.30 | Michael Glanzberg (Northwestern) | The Cognitive Roots of Adjectival Meaning(slides) |
| Thursday, 25.8 | ||
| 11.00 – 11.30 | Antje Roßdeutscher, Tillmann Pross, Sebastian Pado, Gabriella Lapesa, Max Kisselew (Stuttgart) | ‘Over reference’: a comparative study on German prefix-verbs(handout) |
| 11.35 – 12.20 | Marcus Kracht (Bielefeld) | Simplicity of Meaning(slides) |
| 12.20– 12.30 | Discussion: Lexical semantics | |
| Friday, 26.8 | ||
| 11.00 – 11.30 | Yoad Winter (Utrecht) | A Role for Protoroles: Lexical Reciprocity and Logical Symmetry(slides) |
| 11.35 – 12.20 | Mark Steedman (Edinburgh) | A Theory of Content(slides) |
| 12.20– 12.30 | Wrap-up | |
Though referential approaches to semantics have proven very successful at providing meaningful analyses for a wide range of natural language data, some important phenomena, particularly involving the lexicon, have eluded insightful treatment. Notions going beyond reference and truth have been influencing referential semantics for years, but the interest in incorporating results and ideas from conceptually-oriented semantics into referential approaches is noticeably increasing, as seen in the recent series of workshops devoted to the issue (see also e.g. Hamm et al. 2009, Carlson 2010). In parallel, interest has also grown in bringing related insights from cognitively-informed distributional models of meaning into formal semantics (Lenci 2008, Copestake & Herbelot 2012, Baroni et al. 2014).
The aim of this workshop is to promote a 3-way dialog among these approaches in order to clarify natural points of contact and to generate specific hypotheses about how to improve the explanatory capacity of referential models in a principled and testable manner. We build from referential models given the empirical evidence that reference (whether to real or imaginary objects) is a fundamental part of linguistic communication. Crucially, however, reference makes use of complex descriptive content. Cognitive/conceptual approaches place greater emphasis precisely on the richness of descriptive content and richer theories of descriptive content clearly lead to richer accounts of compositional phenomena (see e.g. Kamp & Partee 1995, Zwarts & Winter 2000, Asher 2011, Del Pinal 2015, Gust & Umbach 2015, McNally, to appear). On the other hand, cognitive models are laborious to construct, difficult to implement/test, and face challenges in grounding. Compositional distributional models can help with the analysis of rich descriptive content but are not currently suited to dealing with reference. We therefore consider the incorporation of insights from conceptual and distributional models into referential approaches, rather than the reverse, the most viable strategy.
Barbara Partee, Marcus Kracht, Alessandro Lenci, Mark Steedman
Louise McNally (Universitat Pompeu Fabra), Carla Umbach (ZAS Berlin / University of Cologne)
Guillermo Del Pinal, Marcus Kracht, Alessandro Lenci, Emar Maier, Louise McNally, Barbara Partee, Antje Rossdeutscher, Galit W. Sassoon, Martin Schäfer, Stephanie Solt, Mark Steedman, Carla Umbach
Recent years have seen an increasing interest in ontology-mediated query answering, in which the semantic knowledge provided by an ontology is exploited when querying data. Adding an ontology has several advantages (e.g. simplifying query formulation, integrating data from different sources, providing more complete answers to queries), but it also makes the query answering task more difficult. In this course, we give a introduction to ontology-mediated query answering using description logic (DL) ontologies. Our focus is on DLs for which query answering scales polynomially in the size of the data, as these are best suited for applications requiring large amounts of data. We describe the challenges that arise when evaluating different natural types of queries in the presence of such ontologies, and we present algorithmic solutions based upon two key concepts: query rewriting and saturation. We conclude the course with an overview of recent results and active areas of ongoing research.
Although there are no formal prerequisites, some familiarity with description logics, knowledge representation and reasoning, and/or databases would be helpful. Students who are not already familiar with description logics are strongly encouraged to attend the foundational course “Description Logics: a Nice Family of Logics” offered during Week 1.
Unit 1: Introduction
Unit 2: Instance Queries
Unit 3: Conjunctive Queries
Unit 4: Navigational Queries
Unit 5: Queries with Negation and Other Forms of Recursion
Unit 6: Research Topics in OMQA
Unit 7: Ontology-Based Data Access with Ontop
This course is intended to expose the audience to the following topics,
from a computational point of view: mean payoff games, max-plus algebra,
constraint satisfaction problems. Each of the topics is vast, and would
easily require a course by itself, were all the major ramifications to be
presented. Our focus will be on presenting well known and hopefully
inspiring facts in each domain individually, and exposing a network of
inter-domain relations which have been the subject of recent research.
This course will present several results linking logics for semistructured data (such as XML) with counter systems. The course will focus on two well-studied data models: data words, and data trees. These are words and trees whose every element carries a label from a finite alphabet and a data value from an infinite domain; indeed these are standard abstractions for semi-structured documents. The focus is on the complexity and decidability of reasoning on these structures. The plan is to show three groups of logics with very differing expressive power and capabilities, in order to give an overall idea of the state of the art and different powerful techniques for proving decidability in the area. These logics are divided into: first-order logics, temporal logics, and path logics.
The course material should be useful to anyone with an interest in query languages for semi-structured data, counter systems or more generally on verification of infinite-state systems. This course has some technically demanding parts, and should appeal mainly to an audience from Logic, Verification and Theoretical Computer Science.
Well-quasi-orderings (wqos) are a fundamental tool in logic and computer science. They provide termination arguments in a large number of decidability (or finiteness, regularity, …) results. In constraint solving, automated deduction, program analysis, and many more fields, wqos usually appear under the guise of specific tools, like Dickson’s Lemma (for tuples of integers), Higman’s Lemma (for words and their subwords), Kruskal’s Tree Theorem and its variants (for finite trees with embeddings), and recently the Robertson-Seymour Theorem (for graphs and their minors). What is not very well known is that wqo-based proofs have an algorithmic content.
The purpose of this course is to provide an introduction to the algorithmic aspects of wqos: to present generic algorithms working on large classes of problems, to introduce the techniques used to prove complexity upper bounds and lower bounds, to explain the use of wqo ideals in algorithms, and provide several applications in logics (e.g. data logics, relevance logic), verification (prominently for well-structured transition systems), and formal languages. Because wqos are in such wide use, we believe this topic to be of relevance to a broad community with interests in complexity theory and decision procedures for logical theories.
The proposed course will present several fundamental results linking logics with database query languages. The course can thus serve as an introduction to finite model theory and database theory. More specifically, it will cover basic knowledge about first-order logic, Relational Algebra and Conjunctive Queries; complexity of fundamental decision problems, such as evaluation and satisfiability; as well as results about expressiveness of specification languages. Regarding the logical formalisms, we will focus mainly on two important query languages for relational databases, namely, Relational Algebra and Conjunctive Queries. We will not assume any prior knowledge on either logic or databases, but some familiarity with basic results in complexity theory will be required. The goal of the course is to give the fundamental tools that enable a deeper understanding of database query languages.
The course will be given in English; however, if needed, questions can be posed and answered in French, Spanish, or Italian. The course will be structured into five lessons of about 90 minutes each.
Below is a tentative schedule of the covered subjects.
[1] Serge Abiteboul, Richard Hull, Victor Vianu, Foundations of Databases. Addison Wesley.
Available online at http://webdam.inria.fr/Alice/
[2] Leonid Libkin. Finite Model Theory. Springer, 2004.
This introductory course will be concerned with the probabilistic analysis of deterministic and probabilistic programs, that is on how to obtain, estimate or approximate statistical or probabilistic judgements about certain (quantitative) properties of such programs.
In order to allow for a formal investigation of such features we will first discuss ways to specify or define the semantics of probabilistic programs; in particular to this purpose we will consider Discrete Time Markov Chains. Our focus will then be on Probabilistic Abstract Interpretation (PAI) as a way to obtain simplified or abstract versions of the concrete semantics of programs.
We will introduce the mathematical theory of PAI which is based on linear operators and a notion of pseudo-inverse generalising the one of a Galois connection.
Various static and dynamic analysis techniques based on PAI will be demonstrated by presenting their applications, e.g. in the context of information security, and concrete calculations based on an experimental tool.
We will also discuss how this framework compares with other approaches towards probabilistic semantics (e.g. probabilistic traces and weakest preconditions) and program analysis (such as probabilistic model checking and verification).
