Maria Aloni
University of Amsterdam
"Indexicals under intensional operators"
I will discuss the interpretation of indexicals within the framework of quantified modal logic.
The dominant theory on this topic is Kaplan's two-dimensional analysis. Since its appearance in 1977, Kaplan's analysis has been challenged on various grounds (e.g., Richard 1983, Nunberg 1993).
I will argue that by revising the way of modelling the objects we refer to in conversation (Aloni 2005), we can account for counterexamples to Kaplan's theory, while maintaining Kaplan's basic insights concerning the meaning of `you' and `I'.
Aloni, Maria (2005). Individual concepts in modal predicate logic. Journal of Philosophical Logic, 34(1), 1–64.
Nunberg, Geoffrey (1993). Indexicality and deixis. Linguistics and Philosophy, 16(1), pp. 1–43.
Richard, Mark (1983). Direct reference and ascriptions of belief. Journal of Philosophical Logic, 12(4), 425–452.
Paul Dekker
University of Amsterdam
"Talk about things non-existent"
Philosophers and linguists who try to make sense of talk about things non-existent construe them as things and not as non-existent. The temptation to do so is obviously strong but wrong. I want to argue that we can sensibly resist the temptation, while leaving it up for discussion whether we eventually should.
Andreas Brocher
Cologne University
Klaus von Heusinger & Andreas Brocher: "Towards a dual process model of the comprehension of definite and indefinite noun phrases"
In this presentation, we argue that the comprehension of definite and indefinite noun phrases is best described within a dual-process model of referent activation. In a first process, a comprehender accesses the concept associated with the noun phrase’s descriptive material while, in a second process, the function of the noun phrase’s article guides the comprehender to select the denoted referent(s). Importantly, definite articles signal that there is a unique element that falls under the previously activated concept. In contrast, indefinite articles signal that there are (potentially) multiple referents for the previously activated concept. The dual-process model proposed here was tested in a visual-world eye-tracking experiment, where the different functions associated with definite and indefinite noun phrases are reflected at the level of a referent’s accessibility to a comprehender.
Eliot Michaelson
King's College London
"Towards a Grand Unified Theory of Reference"
Sometime in the 20th century, philosophers largely abandoned the project of offering a unified theory of linguistic reference. This, I suggest, was partly the result of the temptation to handle a range of hard cases—involving speakers who are confused about some relevant aspect of the world—in different ways when different expressions are involved. The result has been an explosion of theories of reference: one for names, another for demonstratives, yet another for indexicals, and so on. Unfortunately, these various theories are generally highly unsatisfactory, and not just because they are disunified. The present essay investigates the prospects for reunification, arguing that they are better than is commonly assumed.
Francois Recanati
Insitut Jean Nicod, Paris
"Direct Reference"
In this talk I will contrast two notions of direct reference.
According to the first, ‘Millian’ notion, a directly referential expression is like a ‘tag’ which is directy assigned to a particular object without there being ‘a semantic mechanism to search for and determine the referent’ (Marti). Proper names are directly referential in that sense and they stand in contrast to definite descriptions, which are not.
According to another notion of direct reference, associated with David Kaplan, what characterises direct reference is just the fact that the propositional constituent corresponding to a directly referential expression is an object rather than an individual concept (a mode of presentation). No less than proper names, indexicals are directly referential in that weaker sense : they are not mere tags, since the referent is determined by a linguistic rule (e.g. the rule that ‘I’ refers to the speaker), yet their semantic value is arguably an object and they are used to express singular propositions about that object. Definite descriptions themselves have a referential use which makes them similar to indexicals. On that use, the descriptive content of the description has the purely instrumental role of helping identify the object one is talking about. The proposition expressed by the utterance is a singular proposition about that object : the mode of presentation itself is not a propositional constituent.
I will argue that the mental file framework, which I will briefly present, enables us to make sense of the Millian position even with respect to indexicals or referentially used descriptions.
Petra Schumacher
Cologne University
"Referential Processing electrified"
Referential expressions are essential ingredients for information processing. Speakers use different referential forms to convey different discourse functions, such as backward oriented linking up with an antecedent or forward directed cues of discourse progression. In the past, I have investigated different types of referential expressions with a particular focus on the temporal dynamics of reference processing. The talk will present evidence from event-related potential (ERP) studies that indicate that referential processing is guided by two core mechanisms associated with expectation-based processing (N400) on the one hand and mental model updating (Late Positivity) on the other hand. The talk will further discuss which factors contribute to expectation-based parsing (e.g., prominence cues) and which phenomena trigger updating processes (e.g., demonstrative pronouns).
Kees van Deemter
University of Aberdeen
"Computational models of referring: advancing the state of the Art"
This talk will introduce the idea that computer algorithms can be interesting models of human language use, which are able to embody insights that elude other approaches to the study of language. Focussing specifically on models of human referring, I will introduce some classic algorithms, discussing what these algorithms are able to do well and what it is that they still struggle to do. Next, I will use evidence from experiments with human speakers and hearers to argue that the most difficult problems in this area arise from situations in which reference is something other than the "simple" identification of a referent through shared knowledge; I will give examples of these problematic situations, and of generation algorithms that address them.
This talk includes themes from my book “Computational Models of
Referring: A Study in Cognitive Science”, MIT Press, June 2016.
Remi van Trijp
Sony Computer Science Laboratories Paris
"Grounding Referential Expressions: A Computational Model in Fluid Construction Grammar"
Computer scientists and AI researchers typically take a “systems perspective” when they want to understand certain phenomena (Steels 2008). System explanations focus on the elements of these phenomena and particularly the processes that exist between these elements. The goal of a computational model is therefore to seek algorithms (or the
systems) that underlie cognitive activities like reference and symbol grounding.
In this presentation, I will present an overview of how this approach has successfully been applied to the symbol grounding problem, and more specifically how a population of embodied artificial agents can self-organize vocabularies for referring to objects in their environment (Steels 2015). In the second part of the presentation, I will then show how grammatical referential expressions (i.e. noun
phrases) can be grounded through a computational model in Fluid Construction Grammar (www.fcg-net.org).
References
Steels, Luc (2008). “The symbol grounding problem has been solved. So what’s next.” In: Manuel de Vega, Arthur M. Glenberg & Arthur C.
Graesser (eds.), Symbols and Embodiment: Debates on Meaning and Cognition. Oxford University Press. pp. 223—244.
Steels, Luc (2015). The Talking Heads Experiment: Origins of Words and Meanings. Berlin: Language Science Press. Available as Open Access at:
langsci-press.org/catalog/book/49
Anna-Lisa Vollmer
CITEC, Bielefeld
"Co-constructing Meaning: Pragmatic Frames in teaching/learning interaction"
In interaction, humans align and effortlessly create common ground in communication, allowing efficient collaboration in widely diverse contexts. Robots are still far away from being able to adapt in such a flexible manner with non-expert humans to work toward joint goals.
Challenges include the capability to understand unknown feedback or guidance signals, to make sense of what they refer to depending on their timing and context, and to agree on how to organize the interaction into roles and turns. In this talk, I will present an experimental setup we invented, in which two humans in an unfamiliar situation similar to human-robot interaction (HRI) have to collaborate to solve a task. This will lead us to the concept of pragmatic frames and how it could be leveraged for HRI.
Robert Van Valin
University at Buffalo, The State University of New York
& Heinrich Heine University Düsseldorf
"Coding reference in articles: Siouan article systems"
Many languages have grammatical means for indicating whether a referring expression actually refers to a referent or not or whether the speaker believes the addressee to be familiar with the referent or not. These are normally known as ‘articles’, which in familiar European languages come in two types, definite (e.g. English the) and indefinite (e.g. English a(n)). Siouan languages typically have much richer article systems than European languages, and they make explicit distinctions which are either implicit or not entailed at all in familiar systems. Lakhota, for example, has two definite articles, and nine, possibly ten, indefinite articles. This talk will present the Lakhota article system in some detail and then contrast it briefly with some other Siouan languages, namely Crow, Osage and Omaha-Ponca, which have rich article systems which don’t express the same distinctions as the Lakhota system. Siouan articles thus afford insights into different properties of reference as revealed by the grammatical markers that express them.