<?xml version="1.0" standalone="yes"?>
<Paper uid="P88-1005">
  <Title>QUANTIFIER SCOPING IN THE SRI CORE LANGUAGE ENGINE</Title>
  <Section position="6" start_page="38" end_page="39" type="concl">
    <SectionTitle>
CONCLUSION
</SectionTitle>
    <Paragraph position="0"> For lack of a reasonable corpus of human quantifier scoping preferences, the testing of'this system has been limited to checking conformance to  the stated rules, s The semantic component of the CLE does not produce logical forms with mass or count NPs or collective readings, but that capability is currently being developed. The foregoing description of qterms is a slight simplification; an extended form is now being used to support generalized quantifiers in the new semantic rules. Examples offered by VanLehn (1978) indicate that dative movement affects quantifier scoping, but the cause may actually be domain or discourse information. Our examples show that passivization affects quantifier scoping, but we have not yet found a means of determining whether the effect is due solely to the cost of raising out of the PP.</Paragraph>
    <Paragraph position="1"> The algorithm does not handle &amp;quot;donkey sentences,&amp;quot; nor is it intended to. A scheme for handling such sentences is being explored as part of the continuing development of the CLE (Fernando Pereira, personal communication). This would be a separate mechanism, rather than an extension of quantifier scoping.</Paragraph>
  </Section>
class="xml-element"></Paper>