We introduce a framework for studying repair initiation in the face of miscommunication. Our aim ... more We introduce a framework for studying repair initiation in the face of miscommunication. Our aim is to seed development of models that both predict when conversational repair is a likely communicative strategy, and explain why interlocutors would not engage in repair in the face of conversational difficulty. We identify three factors as critical to the predictability of repair: (i) the extent to which a misalignment is (un)recognised by participants (ignorance); (ii) the significance of misalignment relative to some cluster of goals (cost of misalignment); and (iii) the significance of engaging in repair relative to some cluster of goals (cost of repair). We offer a simple method for graphically depicting relevant aspects of communicative situations, and exemplify the framework with examples of non-repaired miscommunication before discussing its applicability to different empirical domains.
this paper is to argue, to the contrary, that the highly ambiguous character of natural languages... more this paper is to argue, to the contrary, that the highly ambiguous character of natural languages is surprising, and that the very existence of ambiguity calls for an explanation. Section 1 clarifies what we mean by ambiguity, discussing the distinction between vagueness and ambiguity. We go on to identify several distinct types of ambiguity. Section 2 presents evidence that English is massively ambiguous . Section 3 elaborates our central argument: if (as is widely claimed) ambiguity impedes efficient communication, then one would expect languages to evolve so as to reduce ambiguity; but this does not appear to have happened. Section 4 responds to some possible objections to the argument in Section 3. Section 5 explores some possible strategies for explaining ambiguity, concluding with pointers to our ongoing research on the subject. 1. Characterizing Ambiguity Ambiguity is a semantic property. Semanticists argue over exactly what meaning is, but it surely involves associating expr...
this paper is to argue, to the contrary, that the highly ambiguous character of natural languages... more this paper is to argue, to the contrary, that the highly ambiguous character of natural languages is surprising, and that the very existence of ambiguity calls for an explanation. Section 1 clarifies what we mean by ambiguity, discussing the distinction between vagueness and ambiguity. We go on to identify several distinct types of ambiguity. Section 2 presents evidence that English is massively ambiguous . Section 3 elaborates our central argument: if (as is widely claimed) ambiguity impedes efficient communication, then one would expect languages to evolve so as to reduce ambiguity; but this does not appear to have happened. Section 4 responds to some possible objections to the argument in Section 3. Section 5 explores some possible strategies for explaining ambiguity, concluding with pointers to our ongoing research on the subject. 1. Characterizing Ambiguity Ambiguity is a semantic property. Semanticists argue over exactly what meaning is, but it surely involves associating expr...
<p>We present approaches to the semantics and pragmatics of information structure which cen... more <p>We present approaches to the semantics and pragmatics of information structure which centre on Questions Under Discussion (QUDs). Questions, explicit or implicit, are seen as structuring discourse, and information structural marking is seen as reflecting that underlying discourse structure. Our presentation of the model is largely cast in terms of extensions of Roberts's (2012b) analysis, which is itself related to Rooth's (1985/1992) Alternative Semantics and Hamblin's (1973) approach to the semantics of questions. We present the model in terms of a range of constraints that relate information structure to discourse structure, notably constraints on the 'Relevance' of utterances, on the 'Congruence' of answers to questions, and on the 'Availability' of discourse antecedents. We discuss the application of the approach to the interpretation of focus and some cases of contrastive topics, to discourse structure, to the interpretation of focus sensitive operators, and to certain cases of presupposition projection.</p>
In this paper we offer a novel resolution of a familiar tension, that between approaches in which... more In this paper we offer a novel resolution of a familiar tension, that between approaches in which the difference between definites and indefinites is based on uniqueness, and those in which it is based on novelty/familiarity. Advocates of uniqueness-based approaches such as Horn & Abbott (2013) and Coppock & Beaver (2015) have pointed out cases where the description is unique, but not familiar. For example, Coppock & Beaver (2015) use the following example:
The smallest and most commonly used words in English are pronouns, articles, and other function w... more The smallest and most commonly used words in English are pronouns, articles, and other function words. Almost invisible to the reader or writer, function words can reveal ways people think and approach topics. A computerized text analysis of over 50,000 college admissions essays from more than 25,000 entering students found a coherent dimension of language use based on eight standard function word categories. The dimension, which reflected the degree students used categorical versus dynamic language, was analyzed to track college grades over students' four years of college. Higher grades were associated with greater article and preposition use, indicating categorical language (i.e., references to complexly organized objects and concepts). Lower grades were associated with greater use of auxiliary verbs, pronouns, adverbs, conjunctions, and negations, indicating more dynamic language (i.e., personal narratives). The links between the categorical-dynamic index (CDI) and academic performance hint at the cognitive styles rewarded by higher education institutions.
Intuitively, "before" and "after" are converses. However, the received view among philosophers an... more Intuitively, "before" and "after" are converses. However, the received view among philosophers and linguists since Anscombe (1964) has been that they are not converses, even when the complement is veridical. Based on a distinction between lexical and sentential meanings, we argue for a restoration of intuitions, presenting a simple compositional system that accounts for a range of problematic data. On Anscombe's view, "A before B" means ∃tA(t) ∧ ∀t B(t) → t < t , while "A after B" means (something logically equivalent to) ∃t, t A(t)∧B(t)∧t > t (also Landman (1991), Ogihara (1995), Sánchez-Valencia, van der Wouden and Zwarts (1992)). In Heinämäki's (1974) influential alternative approach, the meanings of the temporal connectives rely on independent and unrelated methods of identifying, on the basis of the intervals at which the sentential arguments hold, the time points to be ordered. As with Anscombe, the connectives are not analyzed as converses, and there is no obvious way of deriving the meaning of one from that of the other. Two phenomena militate, on the face of it, against a converse analysis: inferential properties, studied by Anscombe, and polarity item licensing, discussed by Landman and Sanchez-Valencia et al. among others. With regard to the first, Anscombe argues that "before" is asymmetric and transitive, while "after" is neither, at least for stative sentential arguments. With regard to the second, "before" licenses NPIs, while "after" does not. The well-known non-veridical readings of "before" (Heinämäki 1972; 1974) could also be taken as evidence for the presence of a universal. Building universal quantification into the lexical meaning of "before" and existential quantification into the lexical meaning of "after" accounts for the asymmetries, albeit in a brute force way. A direct argument against universal quantification being part of the lexical meaning of "before" can be made using measure phrases. For example, on an Anscombe-type analysis, (1) "She left exactly 5 seconds before he sang" would be incorrectly predicted to imply that she left exactly 5 seconds before every time point at which he was singing. We show that the data can be accounted for within a theory that preserves the intuition that the connectives are converses, and in fact analyzes them as, lexically, no more than the standard orderings across time points, earlier-than (<) and later-than (>). This leads to a uniform treatment of temporal modifiers that makes no extra assumptions distinguishing between "before" and "after" clauses. The observed asymmetries arise as byproducts of the compositional build-up of sentence meanings. The main ingredients of our analysis are as follows: (i) there is a single semantic tense, taking wide scope, (ii) untensed sentences denote sets of times, (iii) existential quantification over the main clause is standard existential closure occurring when tense is applied, not part of the semantics of the connectives, (iv) temporal modifier clauses restrict the set of times denoted by the main clause, (v) "before" and "after" require their complement to pick out a unique time point, but since the untensed complement clause denotes a set of times, a coercion operation must apply, (vi) nonveridicality of "before" arises from geometric considerations in a branching time model. We show that all the positive results of Anscombe and Heinämäki can be achieved merely by assuming that the coercion operation maps any set of times onto the first time in the set (the function earliest). After showing that for the crucial cases Heinämäki's system reduces to Anscombe's, we prove the following proposition: if the coercion operator is earliest, then for any sentence "A before/after B" with achievement, activity or stative clauses and such that earliest B is defined, the above assumptions yield identical truth conditions to those in Anscombe's proposal. This proposition is
Within the context of a broader project concerning what speakers take to be at-issue, I will pres... more Within the context of a broader project concerning what speakers take to be at-issue, I will present an analysis of the semantics and discourse function of a large range of constructions that I refer to as Inquiry Terminating (IT) constructions. In English, these include it-clefts and exclusives such as" only"," just", and" mere (ly)". I claim that despite their many differences, such constructions cross-linguistically have much in common: they are always focus-sensitive or focus markers, they have a uniform semantics (modulo differences related to ...
We introduce a framework for studying repair initiation in the face of miscommunication. Our aim ... more We introduce a framework for studying repair initiation in the face of miscommunication. Our aim is to seed development of models that both predict when conversational repair is a likely communicative strategy, and explain why interlocutors would not engage in repair in the face of conversational difficulty. We identify three factors as critical to the predictability of repair: (i) the extent to which a misalignment is (un)recognised by participants (ignorance); (ii) the significance of misalignment relative to some cluster of goals (cost of misalignment); and (iii) the significance of engaging in repair relative to some cluster of goals (cost of repair). We offer a simple method for graphically depicting relevant aspects of communicative situations, and exemplify the framework with examples of non-repaired miscommunication before discussing its applicability to different empirical domains.
this paper is to argue, to the contrary, that the highly ambiguous character of natural languages... more this paper is to argue, to the contrary, that the highly ambiguous character of natural languages is surprising, and that the very existence of ambiguity calls for an explanation. Section 1 clarifies what we mean by ambiguity, discussing the distinction between vagueness and ambiguity. We go on to identify several distinct types of ambiguity. Section 2 presents evidence that English is massively ambiguous . Section 3 elaborates our central argument: if (as is widely claimed) ambiguity impedes efficient communication, then one would expect languages to evolve so as to reduce ambiguity; but this does not appear to have happened. Section 4 responds to some possible objections to the argument in Section 3. Section 5 explores some possible strategies for explaining ambiguity, concluding with pointers to our ongoing research on the subject. 1. Characterizing Ambiguity Ambiguity is a semantic property. Semanticists argue over exactly what meaning is, but it surely involves associating expr...
this paper is to argue, to the contrary, that the highly ambiguous character of natural languages... more this paper is to argue, to the contrary, that the highly ambiguous character of natural languages is surprising, and that the very existence of ambiguity calls for an explanation. Section 1 clarifies what we mean by ambiguity, discussing the distinction between vagueness and ambiguity. We go on to identify several distinct types of ambiguity. Section 2 presents evidence that English is massively ambiguous . Section 3 elaborates our central argument: if (as is widely claimed) ambiguity impedes efficient communication, then one would expect languages to evolve so as to reduce ambiguity; but this does not appear to have happened. Section 4 responds to some possible objections to the argument in Section 3. Section 5 explores some possible strategies for explaining ambiguity, concluding with pointers to our ongoing research on the subject. 1. Characterizing Ambiguity Ambiguity is a semantic property. Semanticists argue over exactly what meaning is, but it surely involves associating expr...
<p>We present approaches to the semantics and pragmatics of information structure which cen... more <p>We present approaches to the semantics and pragmatics of information structure which centre on Questions Under Discussion (QUDs). Questions, explicit or implicit, are seen as structuring discourse, and information structural marking is seen as reflecting that underlying discourse structure. Our presentation of the model is largely cast in terms of extensions of Roberts's (2012b) analysis, which is itself related to Rooth's (1985/1992) Alternative Semantics and Hamblin's (1973) approach to the semantics of questions. We present the model in terms of a range of constraints that relate information structure to discourse structure, notably constraints on the 'Relevance' of utterances, on the 'Congruence' of answers to questions, and on the 'Availability' of discourse antecedents. We discuss the application of the approach to the interpretation of focus and some cases of contrastive topics, to discourse structure, to the interpretation of focus sensitive operators, and to certain cases of presupposition projection.</p>
In this paper we offer a novel resolution of a familiar tension, that between approaches in which... more In this paper we offer a novel resolution of a familiar tension, that between approaches in which the difference between definites and indefinites is based on uniqueness, and those in which it is based on novelty/familiarity. Advocates of uniqueness-based approaches such as Horn & Abbott (2013) and Coppock & Beaver (2015) have pointed out cases where the description is unique, but not familiar. For example, Coppock & Beaver (2015) use the following example:
The smallest and most commonly used words in English are pronouns, articles, and other function w... more The smallest and most commonly used words in English are pronouns, articles, and other function words. Almost invisible to the reader or writer, function words can reveal ways people think and approach topics. A computerized text analysis of over 50,000 college admissions essays from more than 25,000 entering students found a coherent dimension of language use based on eight standard function word categories. The dimension, which reflected the degree students used categorical versus dynamic language, was analyzed to track college grades over students' four years of college. Higher grades were associated with greater article and preposition use, indicating categorical language (i.e., references to complexly organized objects and concepts). Lower grades were associated with greater use of auxiliary verbs, pronouns, adverbs, conjunctions, and negations, indicating more dynamic language (i.e., personal narratives). The links between the categorical-dynamic index (CDI) and academic performance hint at the cognitive styles rewarded by higher education institutions.
Intuitively, "before" and "after" are converses. However, the received view among philosophers an... more Intuitively, "before" and "after" are converses. However, the received view among philosophers and linguists since Anscombe (1964) has been that they are not converses, even when the complement is veridical. Based on a distinction between lexical and sentential meanings, we argue for a restoration of intuitions, presenting a simple compositional system that accounts for a range of problematic data. On Anscombe's view, "A before B" means ∃tA(t) ∧ ∀t B(t) → t < t , while "A after B" means (something logically equivalent to) ∃t, t A(t)∧B(t)∧t > t (also Landman (1991), Ogihara (1995), Sánchez-Valencia, van der Wouden and Zwarts (1992)). In Heinämäki's (1974) influential alternative approach, the meanings of the temporal connectives rely on independent and unrelated methods of identifying, on the basis of the intervals at which the sentential arguments hold, the time points to be ordered. As with Anscombe, the connectives are not analyzed as converses, and there is no obvious way of deriving the meaning of one from that of the other. Two phenomena militate, on the face of it, against a converse analysis: inferential properties, studied by Anscombe, and polarity item licensing, discussed by Landman and Sanchez-Valencia et al. among others. With regard to the first, Anscombe argues that "before" is asymmetric and transitive, while "after" is neither, at least for stative sentential arguments. With regard to the second, "before" licenses NPIs, while "after" does not. The well-known non-veridical readings of "before" (Heinämäki 1972; 1974) could also be taken as evidence for the presence of a universal. Building universal quantification into the lexical meaning of "before" and existential quantification into the lexical meaning of "after" accounts for the asymmetries, albeit in a brute force way. A direct argument against universal quantification being part of the lexical meaning of "before" can be made using measure phrases. For example, on an Anscombe-type analysis, (1) "She left exactly 5 seconds before he sang" would be incorrectly predicted to imply that she left exactly 5 seconds before every time point at which he was singing. We show that the data can be accounted for within a theory that preserves the intuition that the connectives are converses, and in fact analyzes them as, lexically, no more than the standard orderings across time points, earlier-than (<) and later-than (>). This leads to a uniform treatment of temporal modifiers that makes no extra assumptions distinguishing between "before" and "after" clauses. The observed asymmetries arise as byproducts of the compositional build-up of sentence meanings. The main ingredients of our analysis are as follows: (i) there is a single semantic tense, taking wide scope, (ii) untensed sentences denote sets of times, (iii) existential quantification over the main clause is standard existential closure occurring when tense is applied, not part of the semantics of the connectives, (iv) temporal modifier clauses restrict the set of times denoted by the main clause, (v) "before" and "after" require their complement to pick out a unique time point, but since the untensed complement clause denotes a set of times, a coercion operation must apply, (vi) nonveridicality of "before" arises from geometric considerations in a branching time model. We show that all the positive results of Anscombe and Heinämäki can be achieved merely by assuming that the coercion operation maps any set of times onto the first time in the set (the function earliest). After showing that for the crucial cases Heinämäki's system reduces to Anscombe's, we prove the following proposition: if the coercion operator is earliest, then for any sentence "A before/after B" with achievement, activity or stative clauses and such that earliest B is defined, the above assumptions yield identical truth conditions to those in Anscombe's proposal. This proposition is
Within the context of a broader project concerning what speakers take to be at-issue, I will pres... more Within the context of a broader project concerning what speakers take to be at-issue, I will present an analysis of the semantics and discourse function of a large range of constructions that I refer to as Inquiry Terminating (IT) constructions. In English, these include it-clefts and exclusives such as" only"," just", and" mere (ly)". I claim that despite their many differences, such constructions cross-linguistically have much in common: they are always focus-sensitive or focus markers, they have a uniform semantics (modulo differences related to ...
Uploads
Papers by David Beaver