<?xml version="1.0" standalone="yes"?>
<Paper uid="H93-1027">
  <Title>INTERPRETING TEMPORAL ADVERBIALS*</Title>
  <Section position="6" start_page="142" end_page="142" type="concl">
    <SectionTitle>
4. CONCLUSION
</SectionTitle>
    <Paragraph position="0"> Much theoretical work has been done on temporal adverbials (e.g., \[4, 5, 7, 14, 16, 19\]). There is also some computationally oriented work. For instance, H0bbs \[8\] provided simple rules for some temporal adverbials, including frequency ones. Moens and Steedman \[15\], among others, discussed the interaction of adverbials and aspectual categories. Our work goes further, in terms of (1) the scope of syntactic coverage, (2) interaction of adverbials with each other and with tense and aspect, (3) systematic (and compositional) transduction from syntax to logical form (with logical-form deindexing),  (4) formal interpretability of the resulting logical forms, and (5) demonstrable use of the resulting logical forms for infer null ence.</Paragraph>
    <Paragraph position="1"> Our initial focus in the analysis of temporal adverbials has been PP-adverbials. Remaining work includes the analysis of clausal adverbials. Also, interactions with negation and aspect (perfect and progressive) have not been completely worked out. Negations of statives are statives, but negations of bounded sentences may be either bounded or unbounded (cf., &amp;quot;We haven't met for thre e years&amp;quot; versus &amp;quot;I have friends I haven't met in three years&amp;quot;). The interaction between present perfect and multiple adverbials of temporal location also creates some subtle difficulties. E.g., in &amp;quot;Mary has jogged {at dawn} {this month},&amp;quot; the inner time adverbial modifies the &amp;quot;jogging&amp;quot; episode, while the outer one modifies the interval that contains the &amp;quot;jogging&amp;quot; episode as well as the utterance time. See \[11\] for some relevant points. Another issue that requires thought is adverbials involving implicit anaphoric referents. Consider, e.g., &amp;quot;Shortly, Mary came in,&amp;quot; &amp;quot;John came back in ten minutes,&amp;quot; and &amp;quot;After three years, John proposed to Mary.&amp;quot; These adverbials involve an implicit reference episode. Such implicit referents may often be identified from our tense trees, but at other times require inference. Another important remaining issue is the interaction between event nominals and frequency adjectives (along the lines of \[25\]).</Paragraph>
    <Paragraph position="2"> 1degHere, iterated sleep is understood as daily sleep--something that must be determined by pragmaties.</Paragraph>
  </Section>
class="xml-element"></Paper>