ZASPiL 61 = Proceedings of Sinn und Bedeutung 22, Vol. 2
Refine
Year of publication
- 2018 (29)
Document Type
- Part of a Book (29)
Language
- English (29)
Has Fulltext
- yes (29)
Is part of the Bibliography
- no (29)
Keywords
- Japanese (2)
- alternative semantics (2)
- comparatives (2)
- counterfactuals (2)
- kinds (2)
- lexical semantics (2)
- maximize presupposition (2)
- relative clauses (2)
- (un)conditionals (1)
- -tari (1)
This paper addresses the question of how to account for the semantic variability of weak free adjuncts. Weak free adjuncts are non-clausal adjuncts that associate with an argument of the main predicate, contribute propositional content, and can interact with temporal or modal operators, which leads to different, adverbial-clause-like interpretations. I focus on a specific type of weak adjuncts, non-clausal as-phrases, and propose a unified semantic analysis for the full range of interpretational possibilities that takes into account the interpretational contingency on different syntactic positions. I show that this analysis improves on Stump’s (1985) original analysis of weak adjuncts. I then go on to discuss the limitations of both Stump’s account and the unified account. Both accounts fail to capture that the interaction of weak adjuncts with modal operators underlies certain restrictions on the properties of the modal operators—an observation that has not been discussed in the literature so far.
'Enough'-/'too'-constructions (E/T constructions) have an implicative reading: e.g., "Mary was clever enough to leave early yesterday" entails Mary left early yesterday. I argue that this implicative reading is not due to the lexical semantics proper of 'enough'/'too', but due to its bi-clausal structure (e.g., the above-mentioned example is analyzed as "Mary left early yesterday because she was clever enough"). I analyze 'enough' and 'too' simply as degree modifiers that involve a comparison: 'enough' means reaching the lower bound of an interval, while 'too' means exceeding the upper bound of an interval. Then inspired by Schulz (2011), Baglini and Francez (2015), and Nadathur (2016), I relate the semantics of E/T constructions to causal dependence: due to some sufficiency/excess, the infinitival complement clause in E/T constructions is episodically or generically (depending on its aspect being perfective or imperfective) true/false. I also argue that this infinitive has its tense and aspect marked on the main predicate of sentences, resulting in the seeming correlation between aspect and implication in languages that overtly make a distinction between perfective and imperfective aspects (e.g., French).
Recent work by Sudo (2012) and Klinedinst (2016) proposes a new perspective on differences between classes of presupposition triggers, with an empirical split roughly mirroring Abusch’s (2002) hard vs. soft distinction and related notions. These two authors propose that triggers differ in whether or not their presuppositional content simultaneously affects the calculation of the presuppositions and of the entailments of the sentences in which they appear.
Drawing on a proposal by Glanzberg (2005) we formulate the Removability/Independence Hypothesis: triggers that do not affect entailments are triggers that can be left out of sentences without affecting interpretability. We experimentally test the hypothesis by embedding 'return', '(go) again' and '(go) back' in non-monotonic environments, which Sudo argues to elicit differences in presuppositions and entailments. Our results provide clear evidence against the RI hypothesis: whereas only the trigger 'return' is crucial for the sake of interpretability, all three triggers produced similar results. At the same time, data for the triggers 'stop' and 'also', included as controls, lend further support in favor of Sudo’s entailment-contrast proposal.
Generics and typicality
(2018)
Cimpian et al. (2010) observed that we accept generic statements of the form 'Gs are f' on relatively weak evidence, but that if we are unfamiliar with group G and we learn a generic statement about it, we still interpret it in a much stronger way: (almost) all Gs are f .
This paper makes use of notions like 'representativeness' and 'contingency' from (associative learning) psychology to provide a semantics of generics that explains why people accept generics based on weak evidence. We make use of the Heuristics and Biases approach of Tversky and Kahneman (1974) and the Associative Theory of Probability Judgements to explain pragmatically why people interpret generic statements in a much stronger way. The spirit of the approach has much in common with Leslie's (2008) cognition-based ideas about generics, but the semantics is grounded on Cohen's (1999) relative readings of generic sentences. The basic intuition is that a generic of the form 'Gs are f' is true, not because most Gs are (or tend to have) f , but because f is typical for G, which means that f is valuably associated with G.
Shared mechanism underlying unembedded and embedded enrichments:
evidence from enrichment priming
(2018)
In this paper, we use a priming paradigm to explore the mechanisms underlying unembedded and embedded scalar enrichments. In particular, the aim is to see if local pragmatic enrichment could be a shared mechanism, involved in both. The two experiments presented adopt Bott & Chemla's (2016) enrichment priming paradigm and test whether unembedded and embedded enrichments could prime each other. The goal is to investigate whether local pragmatic enrichment is indeed being accessed for the interpretation of the unembedded scalar and whether local enrichments, like other lexical semantic phenomena, are susceptible to priming.
The paper proposes a new semantics for good-predications involving finite if -and that-clauses. The proposal combines a standard semantics for conditionals with a standard semantics for the positive form of gradable adjectives and a minimal semantics for modal good. The predicted truth-conditions and conditions of use solve the mood puzzle presented in the first part of the paper. The remainder of the paper defends the classical notion of comparative goodness in terms of a comparison between possible worlds against Lassiter (2017)’s challenge.
This paper investigates the interpretation of Japanese -toka and -tari, two nonexhaustive particles that receive conjunctive interpretations in upward-entailing environments, but disjunctive interpretations in downward-entailing and question contexts.
We analyze -toka and -tari as items that introduce unstructured sets of alternatives in a Hamblin-style alternative semantics (Hamblin, 1973; Kratzer and Shimoyama, 2002), and derive their conjunctive and disjunctive readings via an interaction between these sets and the semantics of the environment containing them.
Based on a sample of seven languages, I show that the so-called modal inferences in ever free relatives (ignorance and indifference) are not universally available. The primary reading of ever free relatives crosslinguistically turns out to be a “non-modal” one, which is available to all languages under investigation. The implication is that if there is a modal inference triggered by the use of the ever-morpheme in FRs, the inference is likely to have a source external to the ever free relative (Lauer, 2009; Condoravdi, 2015; Hirsch, 2016). In line with this conclusion, I propose to generalize Hirsch’s (2016) analysis of ignorance ever free relatives, suggesting that all ever free relatives, no matter how they are ultimately interpreted, are instances of (un)conditionals + donkey-anaphoric definite descriptions.
Revising a proposal by Guerzoni (2003), we propose to derive universal projection of presuppositions in wh-questions, where attested, from a family of three felicity conditions on question use. Assuming that these felicity conditions can be violated under certain conditions, this proposal predicts a typology of contexts where universal projection can exceptionally be unattested. We propose that this prediction is correct, presenting a family of scenarios where the expected absence of universal projection is observed.
The meaning of counterfactual conditionals is standardly described using the similarity approach (Stalnaker, 1968; Lewis, 1973). This approach has recently been challenged by Ciardelli et al. (2018). They argue that the similarity approach is in principle unable to account for the meaning of counterfactuals with an antecedent consisting of a conjunction embedded under a negation (¬(p^q)). Ciardelli et al. (2018) dismiss the approach on these grounds and offer an alternative. The main goal of the present paper is to defend the similarity approach against this attack. I will argue that the problem that underlies the observations in Ciardelli et al. 2018 is more general and not solved by the solution they offer. I will furthermore argue, against Ciardelli et al. (2018), that the cause of the problem is not the similarity approach, but the interaction of negation with the meaning of counterfactual conditionals. The paper will conclude with a first outline of a solution for the problem, which still uses the similarity approach, but combines it with an alternative semantics for negation.
This paper deals with topic markers interacting with discourse information in imperatives. It compares two topic markers from Slovenian (‘pa’) and Japanese (‘-wa’) and shows that while they mostly match in terms of the foci they associate with, their functions differ in imperatives: only ‘pa’ may yield a concessive imperative reading. It is shown that this reading can be derived while keeping a single entry for ‘pa’ by making attitudes of discourse participants part of the focus ‘pa’ associates with. The split between Slovenian and Japanese can then be attributed to minor differences in terms of which foci ‘pa’ and ‘-wa’ may associate with.
This paper explores Turkish numeral constructions, which have typologically two interesting properties: (i) the existence of an optional classifier, (ii) the incompatibility of plurals with them. I argue that numerals are modifiers of type <<e,t>,<e,t>> defined only for atomic properties (Ionin and Matushansky 2006). The explanation rests on the semantics of bare singulars proposed to denote sets of atoms (contra Bale et al. 2010), and the semantics of the classifier claimed to be a partial identity function presupposing atomic properties.
Schlenker (2012) proposes that when framed within a modern Stalnakerian view of presupposition and common ground (Stalnaker, 1998, 2002), Maximize Presupposition! (Heim, 1991; Sauerland, 2008) can be viewed as a special case of the maxim of Quantity (Grice, 1975).
We provide data suggesting that in some cases, Maximize Presupposition! applies even when speakers are not expected to use a presupposition as vectors of new information. We argue that these data support the view that Maximize Presupposition! is an independent pragmatic principle, distinct from Quantity.
The proper semantic treatment of the complements of Responsive Predicates (ResPs), those predicates which may embed either declarative or interrogative clauses, is a longstanding puzzle, given standard assumptions about complement selection. In order to avoid positing systematic polysemy for ResPs, typical treatments of ResP complements treat their arguments either as uniformly declarative-like (propositional) or interrogative-like (question).
I shed new light on this question with novel data from Estonian, in which there are verbs think-like meanings with declarative complements and wonder-like meanings with interrogative complements. I argue that these verbs’ meaning is fundamentally incompatible with a proposition-taking semantics for ResPs, and therefore a question-taking semantics is to be preferred.
This paper presents the results of two experiments in German testing the acceptability of (non-)restrictive relative clauses (NRCs/RRCs) with split antecedents (SpAs). According to Moltmann (1992), SpAs are only grammatical if their parts occur within the conjuncts of a coordinate structure and if they have identical grammatical functions. Non-conjoined SpAs that form the subject and the object of a transitive verb are predicted to be ungrammatical. Our study shows that the acceptability of such examples improves significantly if the predicate that relates the parts of the SpA is symmetric. Moreover, it suggests that NRCs and RRCs behave differently in these cases with respect to the SpA-construal. We can make sense of this observation if we follow Winter (2016) in assuming that transitive symmetric predicates have to be analyzed as unary collective predicates and thus provide a collective antecedent for the RC at the semantic (not the syntactic) level. As we will argue, this accounts for some of the disagreement we found in the literature and gives us new insights into both the semantics of symmetric predicates and the semantics of NRCs.
Schlenker (2010) recently provided data from English and French suggesting that, contrary to standard assumptions (McCawley, 1982; Potts, 2005; Arnold, 2007; AnderBois et al., 2011), non-restrictive relative clauses (NRCs) can take narrow scope under operators of the sentence within which they are embedded. This paper presents three experiments in German confirming this claim. The results show that embedded readings are available with NRCs in German and give first insights into the puzzle under which conditions these embedded readings do or do not show up.
This paper argues that traces only range over individual semantic types and cannot be type shifted into higher types to circumvent this restriction. The evidence comes from movement targeting positions where DPs must denote properties and the behavior of definite descriptions in these positions. These constraints on possible traces demonstrate that syntactic operations impose active restrictions on permissible semantic types in natural language.
This paper presents an exploratory production study of Bharatanatyam, a figurative (narrative) dance. We investigate the encoding of coreference vs. disjoint reference in this dance and argue that a formal semantics of narrative dance can be modeled in line with Abusch’s (2013, 2014, 2015) semantics of visual narrative (drawing also on Schlenker’s, 2017a, approach to music semantics). A main finding of our investigation is that larger-level group-boundaries (Charnavel, 2016) can be seen as triggers for discontinuity inferences (possibly involving the dynamic shift from one salient entity to another).
The verb ‘rise’ can be used both with property-denoting nouns like ‘temperature’ but also with NPs like ‘a Titan’ or ‘China’. Whereas in the former case the change triggered by a rising event is directly related to the subject (its current value increases), this does not hold for ‘a titan’ or ‘China’. In this case it is a property of these objects, say their height or their political power, which increases in value. Furthermore, ‘rise’ does not target a particular property as the examples above show. This data has led Cooper (2010) to the conclusion that it is presumably not possible (i) “to extract a single general meaning of words which covers all the particular meanings of the word in context”, and (ii) “to determine once and for all the set of particular contextually determined meanings of a word”. In this article we present a solution to the two problems raised by ‘rise’ in a frame theory. ‘Rise’ is analyzed as a scalar verb which does not lexicalize a complete scale in its meaning. Rather, it shows underspecification relative to the dimension (property) parameter of a scale. The set of admissible properties is determined by a constraint on the value ranges of properties. If the property is not uniquely determined by the subject, the comprehender uses probabilistic reasoning based on world knowledge and discourse information to defeasibly infer the most likely candidates from this set (2nd problem).
The first problem is solved not by simply introducing objects into the representation of a discourse but instead by pairs consisting of an object and an associated frame component which collects the object information contributed by the discourse. Changes triggered by events like the one denoted by ‘rise’ are modelled as update operations on the frame component while the object component is left unchanged.
Some kind of relative clause
(2018)
Amount Relatives (ARs) differ from restrictive relative clauses in that they do not refer to a particular object denoted by the head of the relative clause, but to an amount of such objects (Carlson, 1977a; Heim, 1987). Traditionally, ARs have been regarded as degree expressions.
In this paper I argue against this view and propose instead that amount interpretations of relative clauses are in fact a special case of kind interpretation.
Extreme nouns and maximizers
(2018)
Maximizers (completamente ‘completely’, totalmente ‘totally’) are degree modifiers restricted to maximum standard adjectives. Spanish adjectives of completeness [ACs] (completo ‘complete’, total ‘total’) display a behavior similar to that of their adverbial counterparts when they combine with nouns like idiot. This paper argues that ACs are maximality modifiers of idiot-like nouns, which are defended to be gradable and denote extreme degrees of properties.
Establishing a parallelism between adverbs and adjectives of completeness allows us to explore scalarity across categories and the relevance of scale structure in the nominal domain.
In this paper, I address verbal predicates of change in Southern Aymara, an understudied Andean language. I concentrate on verbs that are derived with the suffix -cha. This suffix derives degree achievements and creation predicates. I propose that they should be analyzed uniformly as degree achievements. The main empirical point of this paper is that there are two degree morphemes that combine with verbs with -cha, namely, a covert positive morpheme v.POS and an overt suffix -su. The latter is a degree morpheme that restricts the standard of comparison to lexical or contextual maximal degrees. I propose an analysis in terms of Maximize Presupposition: v.POS and -su constitute lexical alternatives where the latter is preferred over the former when maximal values are reached. v.POS is thus felicitous when no maximum is reached. The discussion bears on how telicity is achieved cross-linguistically when degree achievements are considered, thus enriching our typologies on the topic.
It is traditionally assumed that lexical causative verbs (e.g. kill) express direct causation only, while periphrastic (bi-clausal) causatives (e.g. cause to die) may also express indirect causation. In favour of this constraint, Fodor famously observed that the (change of) state introduced by lexical causative verbs is not accessible for separate adverbial modification by temporal (or manner) adverbials. In this paper, I present old and new arguments against the direct causation constraint under the definitions of directness of Fodor and Wolff. I then propose a new definition of directness in terms of ab-initio causal sufficiency framed in Kvart’s probabilistic account of singular causation. I argue that directness so redefined is an implicature rather than an entailment of lexical causative verbs, which enables me to account for old and new data. Furthermore, I account for why the constraint on separate modification by temporal adverbials can be relaxed with eventuality-denoting subjects.
Whether degrees should be modeled as simple semantic primitives or ontologically complex entities has been an issue in recent formal semantic research. This article aims to make a contribution to this scholarly enterprise by investigating the Differential Verbal Comparative (DVC) construction in Chinese. DVCs exhibit peculiar properties : (i) obligatory differentials, and (ii) DPs as differentials(e.g., liang ben xiaoshuo ‘two CL novel’).
We propose that a degree is the entity correlate of a property that is formed on the basis of a measure, akin to Chierchia-style kind. This new kind of degree, coupled with a difference function-based semantics for comparatives, correctly predicts the behaviors of DVCs which would otherwise remain formally inscrutable. This article’s contributions are twofold: (i) it provides direct support for the degree-as-kind analysis by extending its empirical scope; and (ii) by combining degrees as kinds with a difference function-based semantics, it represents an improvement over the previous degree-as-kind analysis based on linear ordering.
This paper investigates the meaning adaptability of change of state (CoS) verbs. It
argues that both coercion and underspecification are necessary mechanisms in order to properly
account for the semantic adaptability observable for CoS verbs in combination with their
complements. This type of meaning adaptability has received little formal attention to date,
although some recent work has already led the way on this topic (Spalek, 2014; Lukassek and
Spalek, 2016; Asher et al., 2017). Our paper is part of a cross-linguistic case study of German
einfrieren and Spanish congelar (‘freeze’). We model the meaning adaptability of this test case
within Type Composition Logic (TCL) (Asher, 2011). We build on Asher’s coercion mechanism
and introduce an additional mechanism for underspecification that exploits the fine-grained type
system in TCL.
Questioning speech acts
(2018)
The function of mapping from the semantic content of an utterance to its convention of use (a division of labor first made by Frege (1956)) has been attributed to abstract speech act operators (also known as force operators), such as ASSERT, QUESTION, and COMMAND. These operators have been traditionally assumed to occupy the highest echelons of the clausal periphery. The precise formulation of these operators has attracted a lot of attention from semanticists, as they are crucial for formalizing the diverse discourse functions of speech acts (Farkas and Bruce, 2009; Farkas and Roelofsen, 2017; Malamud and Stepheson, 2015; Krifka, 2015). These high operators usually come packaged with two assumptions: i) they are not embeddable under other elements, and ii) they belong to the realm of pure pragmatics and not compositional semantics. Recent research in both semantics and syntax have challenged these assumptions (Krifka, 2015; Davis, 2011; Wiltschko, 2017; Heim et al., 2016). Based on evidence from a language with a rich array of sentence-final particles (SFPs), Cantonese, we argue in this paper that not only are abstract speech operators embeddable, it is also the case that we need compositional mechanisms in these high regions of the clause. We will investigate the SFP stacking phenomenon, and argue that such grammaticalized operations on speech act operators reveal the need for a system that can compose the content of an utterance with multiple particles that update the discourse in a number of different, non-trivial ways.
Schwager (2011) and Sudo (2014) argued that there are cases of the so-called third readings of attitude reports, initially discovered by Fodor (1970), that cannot be accounted for in terms of a theory of indexed world variables (Percus, 2000), which is often referred to as the Standard Solution. More complicated alternatives to the Standard Solution have been recently formulated in the literature in a number of papers. We argue that all the seemingly problematic cases can be naturally accounted for in terms of the Standard Solution, if we take into account the existence of previously unrecognized elided material in these reports.
This paper argues that Double Access sentences in English (Smith, 1978) are a kind of loose talk. When the meaning of a Double Access sentence is computed literally, the result is infelicity. Double Access sentences can be used meaningfully only when rescued by pragmatics which intervenes to interpret the embedded clause loosely. A formal model for loose interpretation, building on Klecha (2018), is provided.
Counteridenticals are counterfactual conditional sentences whose antecedent clauses contain an identity statement, e.g. "If I were you, I’d buy the blue dress". Here, we argue that counteridenticals are best analyzed along the lines of dream reports. After showing that counteridenticals and dream reports exhibit striking grammatical and perceptual parallels, we suggest an analysis of counteridenticals with Percus and Sauerland’s (2003) analysis of dream reports. Following their proposal, we propose to make use of concept generators, realized as centered worlds. To this end, we argue that the presence of 'if' licenses the presence of an 'imagine'-operator, which constitutes the attitude the antecedent clause "x be-PAST y" is taken under. The speaker predicates, in the imagine mode, the consequent property to his/her imagined self.
To capture the different degrees of identification between the subject and the predicate of the identity statement of counteridenticals’ antecedents observed in the literature, we incorporate Percus and Sharvit’s (2014) notion of asymmetric be into the analysis. This proposal has several advantages over existing analyses (Lakoff, 1996; Kocurek, 2016) of counteridentical meaning, as it both explains the different degrees of identification observed for counteridenticals and correctly predicts the parallels between counteridenticals and dream reports.