8:30 — 9 am: Breakfast and coffee
9 — 9: 35 am: Chris Kennedy (joint work with Malte Willer), Numerical free choice
There is by now a substantial literature on the derivation of so-called “free choice” inferences with disjunction, geared towards explaining why (1a) is heard as communicating (1c) and (1d), and how the derivation of these inferences relates to the derivation of the upper bounding inference in (1b).
(1) a. Kim is allowed to have cookies or ice cream.
b. Kim is not allowed to have cookies and ice cream.
c. Kim is allowed to have only cookies.
d. Kim is allowed to have only ice cream.
A fact that does not appear to have been noticed in this literature is that free choice interpretations also arise with numerals. Consider a context in which I am giving my babysitter instructions about what my child Kim is allowed to eat while I am out. My utterance of (2a) is heard to communicate not only that eating more than three cookies is not allowed (the upper bounding inference in (2b)), but also that eating fewer than three cookies, and even eating no cookies, is allowed (the free choice inferences in (2c-e)).
(2) a. Kim is allowed to have three cookies.
b. Kim is not allowed to have more than three cookies.
c. Kim is allowed to have only two cookies.
d. Kim is allowed to have only one cookie.
e. Kim is allowed to have no cookies.
In this talk, we will show that: 1) free choice inferences with numerals are not entailments; 2) free choice inferences with numerals and free choice inferences with disjunction do not have the same distributions; and 3) the most common analyses of free choice inferences with disjunction do not extend to an account of free choice inferences with numerals. With any luck, we will conclude with an analysis of numerical free choice and an explanation of these differences.
9:35 — 10:10 am: Michela Ippolito, Varieties of Sobel Sequences
The goal of this talk is to provide a unified analysis of a number of pragmatic anomalies that have been discussed in the literature. We will look at Reverse Sobel sequences of conditionals, sequences of disjunctive sentences, sequences of sentences with definite phrases and superlatives. The starting point is the observation that, while all these sequences are felicitous in one order, they become infelicitous when the order is reversed. Previous proposals have focused on particular types of infelicitous sequences (e.g. von Fintel (2001), Moss (2012), Lewis (2017) a.o.), or a subset of all the phenomena cited above (e.g. Singh (2008c), Singh (2008b), Dohrn (2017), a.o.).
10:10 — 10:30 am: Coffee break
10:30 — 11:05 am: Scott AnderBois, At-issueness in direct quotation: the case of Mayan quotatives
Farkas and Bruce (2010) develop an extremely influential model of the fine-grained ways in which at-issue content is negotiated in discourse. In addition to paving the way for much subsequent research on how such negotiations play out for different sentence types, constructions, and particles, this work has also led to fine-grained analyses of how different kinds of not-at-issue contents update or otherwise make reference to the discourse scoreboard. For example, a number of recent works argue that certain indirect speech report constructions, including reportative evidentials, conventionally encode that only the content of the report may behave as at-issue content (e.g. being the target of unmarked responses like ‘yes’ and ‘no’). In this talk, I argue that quotative morphemes in Mayan languages make an analogous at-issueness distinction in the domain of direct quotation, marking the quoted material itself as a sentence’s at-issue content. In addition to arguing for the need for such a distinction, I explore preliminarily an implementation of this idea extending Eckhardt (2014)’s bicontextualist framework.
11:05 — 11:40 am: Robert Henderson, Donkeys under Discussion
Donkey sentences have existential and universal readings, but they are not often perceived as ambiguous. We extend the pragmatic theory of non-maximality in plural definites by Križ (2016) to explain how hearers use Questions under Discussion to fix the interpretation of donkey sentences in context. We propose that the denotations of such sentences involve truth-value gaps — in certain scenarios the sentences are neither true nor false — and demonstrate that Križ’s pragmatic theory fills these gaps to generate the standard judgments of the literature. Building on Muskens’s (1996) Compositional Discourse Representation Theory and on ideas from supervaluation semantics, we define a general schema for dynamic quantification that delivers the required truth-value gaps.
11:40 — 11:55 am: Coffee break
11:55 am — 12:30 pm: Cleo Condoravdi, Identity and Other Trivialities in Modal Semantics
The classical semantics for conditional obligation validates the conditional obligation for A given fact A, and hence satisfies the principle of identity, whose intuitive status is questionable. The same problem arises for natural language conditionals with a deontic modal in the consequent. On the standard modal restrictor analysis of if-clauses due to Kratzer (1979, 1981), `If A, then must A’ cannot fail to be true. Frank (1997), Zvolenszky (2002), Geurts (2004), Nauze (2008) have discussed the problem from a linguistic perspective, reaching different conclusions about its significance for the semantics of modals and conditionals. This talk revisits this basic problem as it pertains to priority modals within and outside of conditionals. A major challenge is how to square the trivial truth that the semantics delivers with the intuitive judgment of falsity. The key to addressing this challenge is the role of pragmatic reasoning about the information a modal statement conveys regarding the premise sets constituting the modal’s modal base and ordering source.
12:30 — 2 pm: Lunch
2 — 2:35 pm: Floris Roelofsen, Negative polarity items in questions
The distribution of negative polarity items (NPIs) can be characterised, at least to a large extent, in terms of the logical notion of entailment. This discovery has been seen as an important argument for the relevance of logic for natural language semantics. However, entailment-based theories have failed so far to explain the distribution of NPIs in questions. This talk is an initial exploration of the extent to which this can be remedied by means of the notion of entailment offered by inquisitive semantics.
2:35 — 3:10 pm: Peter Alrenga (joint work with Daniel Hardt, Line Mikkelsen and Yenan Sun), More of the same
Anaphoric same can appear in argument (1) or predicative (2) NPs, with interpretive differences:
(1) I saw the rose on the table. Later I saw the same flower on the floor.
(2) The rose was red and beautiful. The carnation was *(the) same.
In (1), the same flower expresses simple identity of individuals, while in (2), the same expresses similarity of properties–like the rose, the carnation was red and beautiful. Previous work (e.g., Alrenga 2007, Matushansky & Ruys 2007, Sun 2018) has accounted for this difference by positing an ambiguity (or at least, an indeterminacy) in same’s interpretation. However, such accounts cannot easily explain obligatory presence of the in (2). We propose instead that the same (N) is always an ordinary NP, and that same always means simple identity–in (2), the same refers to a nominalized property, but is otherwise an ordinary anaphoric definite. But there is still a puzzle here. In some ways, predicative same is sharply distinguished from argument same–in English, predicative same-NPs freely allow N-ellipsis, while argument same-NPs do not. We observe that similar distinctions between same’s two uses are attested in other languages, and conclude with some speculations on why that might be.
3:10 — 3:30 pm: Coffee break
3:30 — 4:45 pm: Plenary talk — Donka Farkas, Non-intrusive questions: the case of oare-marked interrogatives in Romanian
This talk is part of a larger project whose aims are to understand the difference between canonical and non-canonical questions, and to draw a typology of the latter. The first part of the talk sets up a series of pragmatic assumptions present in canonical questions, assumptions that follow from the semantics and conventional discourse effects of canonical interrogatives. The second part presents a particular type of non-canonical questions, called non-intrusive. Non-intrusive questions signal the absence of one of the pragmatic assumptions associated with canonical questions, namely the assumption of Addressee Compliance, i.e, they signal that the Speaker does not assume that the Addressee will settle the issue in the next move. The case study discussed in detail is that of a special interrogative form in Romanian,
namely an interrogative marked by the morpheme oare. In the account to be worked out in the talk, the role of this morpheme is to mark a special conventional discourse effect that affects the projected Addressee responses.
5:30 — 7:30 pm: Dinner at Cowell Provost house.
Organizers: Adrian Brasoveanu, Floris Roelofsen