<?xml version="1.0" standalone="yes"?>
<Paper uid="P94-1009">
  <Title>A HYBRID REASONING MODEL FOR INDIRECT ANSWERS</Title>
  <Section position="6" start_page="63" end_page="63" type="relat">
    <SectionTitle>
5. RELATED RESEARCH
</SectionTitle>
    <Paragraph position="0"> It has been noted \[Diller, 1989, Hirsehberg, 1985, Lakoff, 1973\] that indirect answers conversationally implicale \[Grice, 1975\] direct answers. Recently, philosophers \[Thomason, 1990, MeCafferty, 1987\] have argued for a plan-based approach to conversational implicature. Plan-based computational models have been proposed for similar discourse interpretation problems, e.g.</Paragraph>
    <Paragraph position="1"> indirect speech acts \[Perrault and Allen, 1980, Hinkelman, 1989\], but none of these models address the interpretation of indirect answers. Also, our use of coherence relations, both 1) as constraints on the relevance of indirect answers, and 2) in our hypothesis generation algorithm, is unique in plan-based interpretation models.</Paragraph>
    <Paragraph position="2"> In addition to RST, a number of theories of text coherence have been proposed \[Grimes, 1975, Halliday, 1976, Hobbs, 1979, Polanyi, 1986, Reiehman, 1984\]. Coherence relations have been used in interpretation \[Dahlgren, 1989, Wu and Lytinen, 1990\]. However, inference of coherence relations alone is insufficient for interpreting indirect answers, since additional pragmatic knowledge (what we represent as discourse plan operators) and discourse expectations are necessary also. Coherence relations have been used in generation \[MeKeown, 1985, Hovy, 1988, Moore and Paris, 1988, Horacek, 1992\] but none of these models generate indirect answers. Also, our use of stimulus conditions is unique in generation models.</Paragraph>
    <Paragraph position="3"> Most previous formal and computational models of conversational implicature \[Gazdar, 1979, Green, 1990, Hirschberg, 1985, Lasearides and Asher, 1991\] derive implieatures by classical or nonclassical logical inference with one or more licensing rules defining a class of implicatures. Our coherence rules are similar conceptually to the licensing rules in Lascarides et al.'s model of temporal implicature. (However, different coherence relations play a role in indirect answers.) While Lascarides et al. model temporal implicatures as defeasible inferences, such an approach to indirect answers would fail to distinguish what R intends to convey by his response from other default inferences. We claim that R's response in (1), for example, does not warrant the attribution to R of the intention to convey that the rear axle of R's car is made of metal. Hirsehberg's model for deriving scalar implicatures addresses only a few of the types of indirect answers that our model does. Furthermore, our discourse-plan-based approach avoids problems faced by licensing-rule-based approaches in handling backward cancellation and multipleutterance responses \[Green and Carberry, 1992\]. Also, a potential problem faced by those approaches is scalability, i.e., as licensing rules for handling more types of implieature are added, rule conflicts may arise and tractability may decrease.</Paragraph>
    <Paragraph position="4"> In contrast, our approach avoids such problems by restricting the use of logical inference.</Paragraph>
  </Section>
class="xml-element"></Paper>