Proceedings of Sinn und Bedeutung
Not a member yet
1139 research outputs found
Sort by
A propositionalist semantics for imagination and depiction reports
Get PDFWe present a formal semantics for physical and mental depiction reports (e.g. Penny is painting a penguin, Uli is imagining a unicorn) that interprets the complements of these reports as propositionally coded situations. Our semantics improves upon Zimmermann’s (2016) property-based semantics for such reports (see Zimmermann, 1993) by blocking unwarranted inferences to a common objective and by capturing the semantic interaction of DPs and CPs in depiction complements. At the same time, it preserves the merits of Zimmermann’s semantics, especially the compositional interpretation of depiction reports and the ability to account for missing de dicto-readings of reports with a strong quantificational object DP. Our semantics shows that – contrary to the received view (e.g. Forbes, 2006; Zimmermann, 2016) – depiction complements are not evidence against a propositionalist analysis of attitude complements
Expected value of conditionals and expected utility: A probabilistic account of conditional evaluative constructions
Get PDFThis paper develops a probabilistic account of the Korean conditional evaluative construction that conveys obligation, which is expressed in terms of a conditional and an evaluative predicate. Assuming that conditionals denote the degree of support for the consequent given the antecedent and that evaluative predicates are measure functions that return the utility value of a given world argument, the compositional semantics of the construction at hand gives rise to an expected utility-based semantics of deontic modality. I point out its relevance to decision theory and further offer a principled solution the problem of supererogation
A compositional semantics for free choice constituent unconditionals
Get PDFThe realization of unconditionals is subject to variation across different languages (Haspelmath and König, 1998). In this paper, we zoom in on European French, and propose a unified analysis of two types of constituent unconditionals (CUs) in the language. On the syntactic side, we propose that in both types of CUs, the adjunct clause always contains a free choice item (FCI), but this FCI can be partially elided. On the semantic side, we propose a fully compositional analysis of these two types of CUs. We further derive the characteristic properties of CUs (Haspelmath and König, 1998; Rawlins, 2013) from exhaustification (Chierchia, 2013: among others) and the principle of viability (Dayal, 2013), which have been argued to play a crucial role in the semantics of FCIs. Our approach is thus closely related to recent work on alternative unconditionals in Hungarian (Szabolcsi, 2019) and free choice in Romance (Caponigro and Falaus, 2018), and contrasts with previous work on English, where the CU adjunct clause is analyzed as a wh-question (Rawlins, 2013)
The Game of Same and Different – A framework for analyzing determiner systems
Get PDFThis paper proposes to extend the Game of Same and Different (introduced by Grønn and Szæbø, 2012), and to transform it into a general framework for reasoning about determiner systems in natural languages. The system is applied to a Latin-like language with a singular- plural system and no articles. It will be shown that the system converges under highly rational agents toward the use of the demonstrative as a definite article, and against the emergence of an indefinite article. The behavior of agents more limited in their rationality is explored, and compared with the diachronic development of definite and indefinite articles. Finally, the underlying assumptions of the model are checked against two Latin texts
A closer look at scalar diversity using contextualized semantic similarity
Get PDFWe take a closer look at van Tiel et al.’s (2016) experimental results on diversity in scalar inference rates. In contrast to their finding that semantic similarity had no significant effect on scalar inference rates, we show that a sufficiently fine-grained notion of semantic similarity does have an effect: the more similar the two terms on a scale, the lower the scalar inference rate. Moreover, we show that a context-sensitive notion of semantic similarity (in particular ELMo; Peters et al., 2018) can explain more of the variance in the data, but only modestly, only for stimuli that contain informative context words, and only when the scalar terms themselves are sufficiently context-sensitive
Generalized incrementality: The veridicality property of clause-embedding reveal-type predicates in Polish
Get PDFIt is a common observation that the so-called incremental theme verbs like ‘eat’, ‘drink’ or ‘build’ enforce a gradual affectedness of a direct object by the verbal process (Dowty, 1991; Krifka, 1992 among others). In aspect languages like Polish, perfective incremental theme verbs imply a total affectedness: The object vanishes (in the case of ‘eat’ or ‘drink’), it appears in its totality (in the case of ‘build’) or is fully involved in any other way (for example in the case of ‘read’, where there are no unread pages left in a single reading event). In contrast, imperfective counterparts only imply partial affectedness of their objects (Wierzbicka, 1967; Filip, 1985, 1997, 1999; Krifka, 1989a, 1989b, 1989c, 1992). Crucially, the gradual/total affectedness relation between the verb and its object only holds for incremental theme verbs. For instance, seeing a sandwich, in contrast to eating a sandwich, neither changes anything in the structure of a sandwich nor does it necessarily relate to its parts. This paper shows that propositional objects are also subjected to verbal events in a gradual manner if a clause- embedding verb is an incremental theme verb. In this case, the counterpart to total affectedness is veridicality, i.e. the revelation of a truth-conditional object. Building upon Zuchewicz (2020), we propose an account of generalized incrementality. In this account, incrementality is defined on the level of events and does not enforce the object to be divided into parts (compared to Krifka, 1989a, 1989b, 1989c, 1992 among others). As a result, the analysis holds for all incremental theme verbs, regardless of the type of a complement they combine with. The justification for this line of reasoning comes from a diverse nature of incrementality. Our object of investigation are transitive verbs which, if realized by clause-embedding predicates, take a that-clause as a complement. Their incremental character cannot be captured by dividing a proposition expressed by a that-clause into parts, but by a gradual creation of a proof for an embedded proposition
The temporal perspective of epistemics in Dutch
Get PDFA series of experiments is conducted on naïve native speakers of Dutch and English to study the scope relation between tense and epistemic modality. The results are consistent with the claim that epistemics scope over tense (Stowell 2004, Hacquard 2006, a.o.), and challenge recent research that states that epistemics can, or must, scope under tense (von Fintel and Gillies 2007, Rullmann & Matthewson 2018): Dutch and English participants in a Truth Value Judgment Task judge sentences to be false when the past tense forms of the modals have to and moeten \u27have to\u27 are used to make an epistemic claim that held at a time before speech time, and true when they are used to make an epistemic claim that holds at speech time. Moreover, English participants in an Acceptability Judgment Task judge sentences to be infelicitous when the same past tense form of have to is used to make an epistemic claim that held at a time before speech time. Besides these general patterns, the results show variation within and across the two languages, which leads to interesting new questions about the interaction between tense and (epistemic) modality
Informational Object Nouns and the mass/count distinction
Get PDFMost theories of the count/mass distinction analyse concrete nouns (denoting physical entities or stuff), and so, implicitly or explicitly, set abstract nouns to one side. We build on a growing number of recent works that address this gap with our analysis of a class of abstract nouns, Informational Object Nouns (IONs), such as information, evidence, belief, and statement. We argue that by incorporating recent work done by Schmitt (2013, 2017) on the development of a domain general mereological sum operation, we can modify theories of the mass/count distinction for concrete nouns and extend their coverage to the set of IONs. As we also argue, an important factor in extending such theories is that they are grounded in the notion that count nouns are interpreted under individuation schemas, relative to a context (Sutton and Filip 2016b; Filip and Sutton 2017, amongst others)
The Finnish partitive in counting and measuring constructions
Get PDFThis paper proposes a compositional semantic analysis for the use of the partitive case in counting and measuring constructions in Finnish. Count nouns in counting construc- tions are partitive singular but partitive plural in measure constructions. Mass nouns are infelicitous in counting constructions but are partitive singular in measure constructions. We propose an analysis for this pattern by making the semantics of the partitive morpheme both (i) derived from the notion of mereological parthood and (ii) sensitive to quantization. Finally, we show how this analysis extends to account for why mass nouns and plural count nouns in partitive case are felicitous as partitive subjects, when singular count nouns in partitive case are not
Degrees as kinds vs. degrees as numbers: Evidence from equatives
Get PDFWithin formal semantics, there are two views with regard to the ontological status of degrees: the ‘degree-as-number’ view (e.g., Seuren, 1973; Hellan, 1981; von Stechow, 1984) and the ‘degree-as-kind’ view (e.g., Anderson and Morzycki, 2015). Based on (i) empirical distinctions between comparatives and equatives and (ii) Stevens’s (1946) theory on the four levels of measurements, I argue that both views are motivated and needed in accounting for measurement- and comparison-related meanings in natural language. Specifically, I argue that since the semantics of comparatives potentially involves measurable differences, comparatives need to be analyzed based on scales with units, on which degrees are like (real) numbers. In contrast, since equatives are typically used to convey the non-existence of differences, equatives can be based on scales without units, on which degrees can be considered kinds