<?xml version="1.0" standalone="yes"?>
<Paper uid="E87-1035">
  <Title>DETERMINISTIC PARSING AND UNBOUNDED DEPENDENCIES</Title>
  <Section position="7" start_page="215" end_page="216" type="concl">
    <SectionTitle>
CONCLUSION
</SectionTitle>
    <Paragraph position="0"> The paper has presented two approaches to the deterministic analysis of unbounded dependencies. The LR(1) technique is capable of resolving the type of local ambiguities which appear to occur in these constructions, suggesting that Church and Johnson-Laird were wrong to reject deterministic parsing on the basis of this data.</Paragraph>
    <Paragraph position="1"> However, we have argued that the Lexicat parser provides a better psychological model of the parsing of unbounded dependencies because a) it predicts the garden path data and b), in conjunction with the IDH, it predicts the apparent distribution Of prosodic boundaries in these constructions more successfully, and c) it provides a unified account of the resolution of local and global  ambiguities, and d) it is a simpler model of deterministc parsing which does not require preprocessing the grammar or maintaining state information concerning the left context*</Paragraph>
  </Section>
class="xml-element"></Paper>