<?xml version="1.0" standalone="yes"?>
<Paper uid="A00-3008">
  <Title>Multiple Discourse Marker Occurrence: Creating Hierarchies for Natural Language Generation</Title>
  <Section position="2" start_page="0" end_page="0" type="intro">
    <SectionTitle>
1 Introduction
</SectionTitle>
    <Paragraph position="0"> The majority of studies on discourse markers implicitly assume that only one marker or discourse relation will occur in a sentence or that the presence of multiple markers will not affect the choice and placement of others. However, in reality, more than one relation may hold between text spans which may be cued by multiple markers. The available rules describing the occurrence, choice and placement of a given marker do not account for multiple marker occurrence (Grote et al., 1995; Webber and Joshi, 1998; Power et al., 1999, e.g.,). We have found that the choice and placement of discourse markers is greatly affected, not only by the presence and number of other markers, but also by the style of the text and the strength of other markers in the text span. We describe here a method for hierarchically organising discourse markers which takes account of these factors. The hierarchies are intended for use by a generation system to enable the selection and placement of multiple markers.</Paragraph>
    <Paragraph position="1"> * The author would like to thank the Engineering and</Paragraph>
  </Section>
class="xml-element"></Paper>