<?xml version="1.0" standalone="yes"?>
<Paper uid="P87-1013">
  <Title>Xerox PARC</Title>
  <Section position="7" start_page="94" end_page="94" type="concl">
    <SectionTitle>
6 Conclusion
</SectionTitle>
    <Paragraph position="0"> It would be desirable to use the notation provided by our logic to state substantive principles of particulax linguistic theories. Consider, for example, Kashket's parser for Warlpiri \[4\], which is based on GB theory. For languages like Warlpiri, we might be able to say that linear order is only explicitly represented at the morphemic level, and not at the phrase level. This would translate into a constraint on the kinds of logical formulas we could use to describe such languages: the &lt; relation could only be used as a relation between nodes of the MORPHEME type. Given such a condition on formulas, it migh t then be possible to prove complexity results which were more positive than a general undecidability theorem. Similar remarks hold for theories like relational grammar, in which many such constraints have been studied. We hope that logical tools will provide a way to classify these empirically motivated conditions.</Paragraph>
  </Section>
class="xml-element"></Paper>