Questions embedded under responsive predicates and definite descriptions both give rise to a variety of phenomena which can be grouped under the term plurality effects: quantificational variability, cumulativity, and homogeneity effects. This similarity has not gone unnoticed, and many proposals have taken inspiration in theories of definite plurals to account for these effects with embedded questions. Recently these phenomena have received less attention, as the field has focused on the so-called intermediate exhaustive reading of embedded questions instead, after Spector called (...) into question the traditional dichotomy between weak and strong exhaustive readings. As a result, the intermediate exhaustive reading has been accounted for at the expense of empirical coverage in other areas. In this paper, I propose a modular theory which derives the currently much discussed exhaustive readings without giving up the rich semantics necessary to account for plurality effects. My account of quantificational variability, cumulativity, and homogeneity effects builds on recent work on these phenomena in the nominal domain by adopting a categorial approach to embedded questions, while the strong and intermediate exhaustive readings are implemented using an independent strengthening mechanism suggested in Klinedinst and Rothschild :1–23, 2011). The resulting theory not only recovers important results on plurality effects; it offers new, simple solutions for some puzzles presented in George :166–177, 2013) and Paillé and Schwarz Proceedings of WCCFL 36, vol 36, Cascadilla Proceedings Project, Somerville, 2018), naturally derives readings that had been postulated in previous literature, makes correct predictions in many unexplored cases, and is compatible with recent results in psycholinguistics. In the last sections I justify my assumptions and show how possible limitations I inherit from the theories I build on can be accommodated under standard assumptions. (shrink)

This paper reports an experimental investigation of presuppositions and scalar implicatures in language acquisition. Recent proposals posit the same mechanisms for generating both types of inferences, in contrast to the traditional view. We used a Covered Box picture selection task to compare the interpretations assigned by two groups of children and by adults, in response to sentences with presuppositions and ones with either ‘direct’ or ‘indirect’ scalar implicatures. The main finding was that the behavior of children and adults differed across (...) inference types. This asymmetry is consistent with the traditional perspective, but poses a challenge for the more recent uniform accounts. We discuss how the latter could be amended to account for these findings, and also relate the findings to previous results on presupposition processing. Finally, we discuss an unexpected difference found between direct and indirect scalar implicatures. (shrink)

Binary judgement on under-informative utterances is the most widely used methodology to test children’s ability to generate implicatures. Accepting under-informative utterances is considered a failure to generate implicatures. We present off-line and reaction time evidence for the Pragmatic Tolerance Hypothesis, according to which some children who accept under-informative utterances are in fact competent with implicature but do not consider pragmatic violations grave enough to reject the critical utterance. Seventy-five Dutch-speaking four to nine-year-olds completed a binary and a ternary judgement task. (...) Half of the children who accepted an utterance in Experiment A penalised it in Experiment B. Reaction times revealed that these children experienced a slow-down in the critical utterances in Experiment A, suggesting that they detected the pragmatic violation even though they did not reject it. We propose that binary judgement tasks systematically underestimate children’s competence with pragmatics. (shrink)

In a series of experiments, Bott and Noveck (2004) found that the computation of scalar inferences, a variety of conversational implicature, caused a delay in response times. In order to determine what aspect of the inferential process that underlies scalar inferences caused this delay, we extended their paradigm to three other kinds of inferences: free choice inferences, conditional perfection, and exhaustivity in “it”‐clefts. In contrast to scalar inferences, the computation of these three kinds of inferences facilitated response times. Following a (...) suggestion made by Chemla and Bott (2014), we propose that the time it takes to compute a conversational implicature depends on the structural characteristics of the required alternatives. (shrink)

The computation of scalar implicatures is sometimes costly relative to basic meanings. Among the costly computations are those that involve strengthening `some' to `not all' and strengthening inclusive disjunction to exclusive disjunction. The opposite is true for some other cases of strengthening, where the strengthened meaning is less costly than its corresponding basic meaning. These include conjunctive strengthenings of disjunctive sentences (e.g., free-choice inferences) and exactly-readings of numerals. Assuming that these are indeed all instances of strengthening via implicature/exhaustification, the puzzle (...) is to explain why strengthening sometimes increases costs while at other times it decreases costs. I develop a theory of processing costs that makes no reference to the strengthening mechanism or to other aspects of the derivation of the sentence's form/meaning. Instead, costs are determined by domain-general considerations of the grammar's output, and in particular by aspects of the meanings of ambiguous sentences and particular ways they update the context. Specifically, I propose that when the hearer has to disambiguate between a sentence's basic and strengthened meaning, the processing cost of any particular choice is a function of (i) a measure of the semantic complexity of the chosen meaning and (ii) a measure of how much relevant uncertainty it leaves behind in the context. I measure semantic complexity with Boolean Complexity in the propositional case and with semantic automata in the quantificational case, both of which give a domain-general measure of the minimal representational complexity needed to express the given meaning. I measure relevant uncertainty with the information-theoretic notion of entropy; this domain-general measure formalizes how `far' the meaning is from giving a complete answer to the question under discussion, and hence gives an indication of how much representational complexity is yet to come. Processing costs thus follow from domain-general considerations of current and anticipated representational complexity. The results might also speak to functional motivations for having strengthening mechanisms in the first place. Specifically, exhaustification allows language users to use simpler forms than would be available without it to both resolve relevant uncertainties and convey complex meanings. (shrink)

A number of studies have demonstrated pragmatic language difficulties in people with Schizophrenia Spectrum and Other Psychotic Disorders. However, research about how people with schizophrenia spectrum and other psychotic disorders understand scalar implicatures is surprisingly rare, since SIs have generated much of the most recent literature. Scalar implicatures are pragmatic inferences, based on linguistic expressions like some, must, or, which are part of a scale of informativeness. Logically, the less informative expressions imply the more informative ones, but pragmatically people usually (...) infer that the presence of a less informative term implies that the more informative term was not applicable. In one of the few existing studies with people with schizophrenia spectrum and other psychotic disorders, Wampers et al. observed that in general, people with schizophrenia spectrum and other psychotic disorders were less likely to derive SIs than controls. The current study has three main aims. First, we want to replicate the original finding with the scalar terms some-all. Second, we want to investigate how these patients deal with different scalar terms, that is, we want to investigate if scalar diversity is also observed in this clinical group. Third, we investigate the role of working memory, often seen as another important mechanism to enable inferring SIs. Twenty-one individuals with a psychotic disorder and 21 matched controls answered 54 under-informative statements, in which seven different pairs of scalar terms were used. In addition, working memory capacity was measured. Patients with schizophrenia spectrum and other psychotic disorders did not make more logical interpretations when processing quantifiers, disconfirming Wampers et al.. However, certain scalar scales elicited more pragmatic interpretations than others, which is in line with the scalar diversity hypothesis. Additionally, we observed only partial evidence for the role of working memory. Only for the scalar scale and-or, a significant effect of working memory was observed. The implications of these results for patients with schizophrenia spectrum and other psychotic disorders are discussed, but also the role of working memory for pragmatic inferences, as well as the place of SIs in experimental pragmatics. (shrink)

Experimental investigations into children’s interpretation of scalar terms show that children have difficulties with scalar implicatures in tasks. In contrast with adults, they are for instance not able deriving the pragmatic interpretation that “some” means “not all” (Noveck, 2001; Papafragou & Musolino, 2003). However, there is also substantial experimental evidence that children are not incapable of drawing scalar inferences and that they are aware of the pragmatic potential of scalar expressions. In these kinds of studies, the prime interest is to (...) discover what conditions facilitate implicature production for children. One of the factors that seem to be difficult for children is the generation of the scalar alternative. In a Felicity Judgment Task (FJT) the alternative is given. Participants are presented with a pair of utterances and were asked to choose the most felicitous description. In such a task, even five-year-old children showed a very good performance (Chierchia et al., 2001; Foppolo, Guasti, & Chierchia, 2012). Our study wants to build on this tradition, by using a FJT where not only “some-all” choices are given, but also “some-many” and “many-all”. We compared performance of 59 five-year-old children with 34 11-year-old children. The results indicated that performance of both age-groups was clearly above chance. With respect to the ‘some-all’ choice, we find in both age-groups very high accuracy, replicating Foppolo et al. (2012). However, for the five-year-old children, the critical and ambiguous items were more difficult than the control items and they also performed worse on these two items than the eleven-year-old children. Interestingly with respect to the issue of scalar diversity, the eleven-year-old children were also presented temporal items, which turned out to be more difficult than the quantitative ones. (shrink)

We evaluate the richness of the child’s input in semantics and its relation to the hypothesis space available to the child. Our case study is the acquisition of the universal quantifier every. We report two main findings regarding the acquisition of every on the basis of a corpus study of child-directed and child-ambient speech. Our first finding is that the input in semantics is rich enough to systematically eliminate instances of the subset problem of language acquisition: overly general hypotheses about (...) the meaning of every can violate pragmatic constraints, making such hypotheses incompatible with the child’s input. Our second finding is that the semantic input is too poor to eliminate instances of what we refer to as the superset problem, the mirror image of the subset problem. We argue that at least some overly specific hypotheses about the meaning of every are compatible with the child’s input, suggesting either that those hypotheses are not made available by UG or that non-trivial inductive biases are involved in children’s acquisition of every. (shrink)

One of the great challenges in word learning is that words are typically uttered in a context with many potential referents. Children's tendency to associate novel words with novel referents, which is taken to reflect a mutual exclusivity (ME) bias, forms a useful disambiguation mechanism. We study semantic learning in pragmatic agents—combining the Rational Speech Act model with gradient‐based learning—and explore the conditions under which such agents show an ME bias. This approach provides a framework for investigating a pragmatic account (...) of the ME bias in humans but also for building artificial agents that display an ME bias. A series of analyses demonstrates striking parallels between our model and human word learning regarding several aspects relevant to the ME bias phenomenon: online inference, long‐term learning, and developmental effects. By testing different implementations, we find that two components, pragmatic online inference and incremental collection of evidence for one‐to‐one correspondences between words and referents, play an important role in modeling the developmental trajectory of the ME bias. Finally, we outline an extension of our model to a deep neural network architecture that can process more naturalistic visual and linguistic input. Until now, in contrast to children, deep neural networks have needed indirect access to (supposed to be novel) test inputs during training to display an ME bias. Our model is the first one to do so without using this manipulation. (shrink)

Upon hearing “Some of Michelangelo’s sculptures are in Rome,” adults can easily generate a scalar implicature and infer that the intended meaning of the utterance corresponds to “Some but not all Michelangelo’s sculptures are in Rome.” Comprehension experiments show that preschoolers struggle with this kind of inference until at least 5 years of age. Surprisingly, the few studies having investigated children’s production of scalar expressions like some and all suggest that production is adult-like already in their third year of life. (...) Thus, children’s production of implicatures seems to develop at least 2 years before their comprehension of implicatures. In this paper, we present a novel account of scalar implicature generation in the framework of Bidirectional Optimality Theory: the Asymmetry Account. We show that the production–comprehension asymmetry is predicted to emerge because the comprehension of some requires the hearer to consider the speaker’s perspective, but the production of some does not require the speaker to consider the hearer’s perspective. Hence, children’s comprehension of scalar expressions, but not their production of scalar expressions, is predicted to be related to their theory of mind development. Not possessing fully developed theory of mind abilities yet, children thus have difficulty in comprehending scalar expressions such as some in an adult-like way. Our account also explains why variable performance is found in experimental studies testing children’s ability to generate scalar implicatures; moreover, it describes the differences between children’s and adults’ implicature generation in terms of their ability to recursively apply theory of mind; finally, it sheds new light on the question why the interpretation of numerals does not require implicature generation. (shrink)

Starting in early childhood, children are socialized to be honest. However, they are also expected to avoid telling the truth in sensitive situations if doing so could be seen as inappropriate or impolite. Across two studies (total N = 358), the reasoning of 3‐ to 5‐year‐old children in such a scenario was investigated by manipulating whether the information in question would be helpful to the recipient. The studies used a reverse rouge paradigm, in which a confederate with a highly salient (...) red mark on her nose asked children whether she looked okay prior to having her picture taken. In Study 1, children tended to tell the truth only if they were able to observe that the mark was temporary and the confederate did not know it was there. In Study 2, children tended to tell the truth only if they were able to observe that the mark could be concealed with makeup. These findings show that for children as young as age 3, decisions about whether to tell the truth are influenced by the likelihood that the information would be helpful to the recipient. (shrink)

Franke, Dulcinati and Pouscoulous also examine a form of covert lying, by considering to what extent speakers use implicatures to deceive their addressee. The participants in their online signaling game had to describe a card, which a virtual coplayer then had to select. When the goal was to deceive rather than help the coplayer, participants produced more false descriptions (overt lies), but also more uninformative descriptions (covert lies by means of an implicature). [73].

I provide arguments in favor of an implicature approach to Homogeneity where the basic meaning of the kids laughed is some of the kids laughed, and its strengthened meaning is all of the kids laughed. The arguments come from asymmetries between positive and negatives sentences containing definite plurals with respect to children’s behavior, the availability of Non-maximal readings, and the robustness of neither-true-nor-false judgments :205–248, 2015). I propose to avoid some problems of Magri’s analysis by modeling the Implicature account of (...) Homogeneity after the Implicature account of Free Choice, based on a hitherto unnoticed analogy between the two phenomena. The approach that emerges has the advantages of Magri’s implicature account of Homogeneity, while at the same time bears a close resemblance to recent approaches to Non-maximality :1–58. https://doi.org/10.3765/sp.5.3, 2012; Križ and Spector in Interpreting plural predication: homogeneity and non-maximality, Ms., Institut Jean Nicod, 2017), which enables restating their account of Non-maximality as following from the context-sensitivity of implicature calculation. (shrink)

This paper investigates when and how children figure out the force of modals: that possibility modals (e.g., _can_/_might_) express possibility, and necessity modals (e.g., _must_/_have to_) express necessity. Modals raise a classic subset problem: given that necessity entails possibility, what prevents learners from hypothesizing possibility meanings for necessity modals? Three solutions to such subset problems can be found in the literature: the first is for learners to rely on downward-entailing (DE) environments (Gualmini and Schwarz in J. Semant. 26(2):185–215, 2009 ); (...) the second is a bias for strong (here, necessity) meanings; the third is for learners to rely on pragmatic cues stemming from the conversational context (Dieuleveut et al. in Proceedings of the 2019 Amsterdam colloqnium, pp. 111–122, 2019a ; Rasin and Aravind in Nat. Lang. Semant. 29:339–375, 2020 ). This paper assesses the viability of each of these solutions by examining the modals used in speech to and by 2-year-old children, through a combination of corpus studies and experiments testing the guessability of modal force based on their context of use. Our results suggest that, given the way modals are used in speech to children, the first solution is not viable and the second is unnecessary. Instead, we argue that the conversational context in which modals occur is highly informative as to their force and sufficient, in principle, to sidestep the subset problem. Our child results further suggest an early mastery of possibility—but not necessity—modals and show no evidence for a necessity bias. (shrink)