<?xml version="1.0" standalone="yes"?>
<Paper uid="E87-1035">
  <Title>DETERMINISTIC PARSING AND UNBOUNDED DEPENDENCIES</Title>
  <Section position="3" start_page="0" end_page="0" type="intro">
    <SectionTitle>
INTRODUCTION
</SectionTitle>
    <Paragraph position="0"> Church (1980:117) and Johnson-Laird (1983:313) have argued that the high degree of ambiguity in unbounded dependencies undermines the programme of deterministic parsing. Their conclusion is based on critiques of various versions of the Marcus parser (Marcus, 1980; Berwick &amp; Weinberg, 1984). This parser represents only one of many possible approaches to deterministic parsing.. Therefore, the conclusion that deterministic parsing, m general, is impractical or psychologically implausible may be premature.</Paragraph>
    <Paragraph position="1"> In the next section, we outline the problems for the deterministic analysis of unbounded dependencies. In the succeeding sections, we present two alternative parsing techniques (and associated grammars) which make differing predictions concerning the onset and location of indeterminacy in the analysis of unbounded dependencies.</Paragraph>
    <Paragraph position="2"> We argue that the LR(1) parser is capable of deterministically resolving the local ambiguities which occur in these constructions, whilst the Lexicat parser is not. In the final section, we evaluate these predictions in the light of the Determinism Hypothesis (Marcus, 1980) and the Interactive Determinism Hypothesis (Briscoe &amp; Boguraev, 1984; Briscoe, in press) and argue that the Lexicat parser in conjunction with the Interactive Determinism Hypothesis provides the most psychologically plausible and unified account of the parsing of unbounded dependencies.</Paragraph>
  </Section>
class="xml-element"></Paper>