<?xml version="1.0" standalone="yes"?>
<Paper uid="W06-0404">
  <Title>Pragmatic Constraints on Semantic Presupposition</Title>
  <Section position="3" start_page="0" end_page="25" type="intro">
    <SectionTitle>
1 Introduction
</SectionTitle>
    <Paragraph position="0"> Devising an appropriate theory of presupposition has been one of the main issues in semantics, pragmatics and most recently computational linguistics. Indeed, many theorists have argued extensively about the definition that captures the meaning of presupposition and whether presuppositions are a property of the utterance or of the speaker.</Paragraph>
    <Paragraph position="1"> Developments in dynamic semantics, as opposed to static semantics, resulting in DRT, have led to a framework suitable for the representation of linguistic phenomenon. Some of DRT's principal concerns are with finding the right truth conditions and interpretation for referential expressions, specifically anaphora. This is relevant for further investigation of 'pragmatic' presupposition because it has in fact been proposed by van der 1I gratefully acknowledge support from Science Foundation Ireland grant 04/IN/I527.</Paragraph>
    <Paragraph position="2"> Sandt and Geurts (1991) that it is anaphora that lies at the basis of presupposition.</Paragraph>
    <Paragraph position="3"> However, Simons (2003) has recently noted some pragmatic limitations in the present state of DRT. She refers in the following quotation to 'dynamic semantics', the field taken to include DRT.</Paragraph>
    <Paragraph position="4"> Dynamic semantics does not attempt to understand presupposition and presuppositional constraints in terms of the speaker's beliefs and intentions, or to root presuppositional constraints in terms of the broad goals of communicators (Simons 2003: 27).</Paragraph>
    <Paragraph position="5"> Indeed, Simons concludes that DRT is a theory of semantic and not pragmatic presupposition (2003). The criticism that DRT is only semantic is not wholly justified. While DRT stems from the need for appropriate semantic representation of discourse, DRT does recognize the importance of context in representing referents in discourse, which is generally taken to mark a pragmatic perspective on presupposition. Additionally, theories within DRT, such as the Binding Theory (Geurts 1999), have attempted to make DRT more pragmatic; in particular, Presupposition as Anaphora Theory's construction process of presupposition (van der Sandt and Geurts 1991). However, while this paper aims to show that DRT is not entirely devoid of pragmatics, it argues that DRT is in need of a more pragmatic treatment of presupposition, which (a) pays more attention to the beliefs and intentions of the speaker and the hearer and their relation to presupposition and (b) makes presuppositional constraints more precise. Further, other scholars have criticized DRT for being essentially truthconditional. Werth, for example, claims that DRT is essentially only about truthconditionality:  [DRT's] goal is truth-conditionality and its models are minimal worlds inhabited by predicates and variables... it does not model human understanding: there is no place in it for participant roles, setting, background knowledge, purposes, even inferences (Werth 1999: 65).</Paragraph>
    <Paragraph position="6"> Again, this criticism is not entirely just. To address the aforementioned pragmatic limitations of DRT, the aim of this paper is to address the problem of insightfully capturing pragmatic constraints on presuppositional phenomenon within the framework of DRT. To achieve this, a pragmaticallyconstrained definition of presupposition is attempted. This is followed by setting some pragmatic constraints on agents' conception of presuppositional phrases based on the agents' roles in conversation. An example of how these pragmatic constraints operate in DRT is then examined. Furthermore, an overall structure computationally encompassing these pragmatic constraints in DRT is described.</Paragraph>
  </Section>
class="xml-element"></Paper>