File Information

File: 05-lr/acl_arc_1_sum/cleansed_text/xml_by_section/intro/06/w06-3506_intro.xml

Size: 4,041 bytes

Last Modified: 2025-10-06 14:04:15

<?xml version="1.0" standalone="yes"?>
<Paper uid="W06-3506">
  <Title>Catching Metaphors</Title>
  <Section position="3" start_page="0" end_page="41" type="intro">
    <SectionTitle>
2 Metaphor
</SectionTitle>
    <Paragraph position="0"> Work in Cognitive Semantics (Lakoff and Johnson, 1980; Johnson, 1987; Langacker, 1987; Lakoff, 1994) suggests that the structure of abstract actions (such as states, causes, purposes, and means) are characterized cognitively in terms of image schemas which are schematized recurring patterns from the embodied domains of force, motion, and space.</Paragraph>
    <Paragraph position="1"> Consider our conceptualization of events as exemplified in the mapping called the Event Structure Metaphor.</Paragraph>
    <Paragraph position="2">  * States are locations (bounded regions in space). * Changes are movements (into or out of bounded regions).</Paragraph>
    <Paragraph position="3">  * Causes are forces.</Paragraph>
    <Paragraph position="4"> * Actions are self-propelled movements.</Paragraph>
    <Paragraph position="5"> * Purposes are destinations.</Paragraph>
    <Paragraph position="6"> * Difficulties are impediments to motion.  This mapping generalizes over an extremely wide rangeofexpressionsforoneormoreaspectsofevent structure. For example, take states and changes. We speak of being in or out of a state, of entering or leaving it, of getting to a state or emerging from it. This is a rich and complex metaphor whose parts interact in complex ways. To get an idea of how it works, consider the submapping Difficulties are impediments to motion. In the metaphor, purposeful action is self-propelled motion toward a destination. A difficulty is something that impedes such motion. Metaphorical difficulties of this sort come in five types: blockages; features of the terrain; burdens; counterforces; lack of an energy source. Here are examples of each: Blockages: He's trying to get around the regulations. We've got him boxed into a corner. Features of the terrain: It's been uphill all the way. We've been hacking our way through a jungle of regulations. Burdens: He's carrying quite a load. Get off my back! Counterforces: Quit pushing me around. She's leading him around by the nose. Lack of an energy source: I'm out of gas. We're running out of steam.</Paragraph>
    <Paragraph position="7"> In summary, these metaphors are ontological mappings across conceptual domains, from the source domain of motion and forces to the target domain of abstract actions. The mapping is conventional, that is, it is a fixed part of our conceptual system, one of our conventional ways of conceptualizing actions. Conventional metaphors capture generalizations governing polysemy, over inference patterns, and governing novel metaphorical language (Lakoff and Turner, 1989).</Paragraph>
    <Section position="1" start_page="41" end_page="41" type="sub_section">
      <SectionTitle>
2.1 Metaphors vs. Different Word Senses
</SectionTitle>
      <Paragraph position="0"> Presumably, one could treat the metaphoric usage of run as a different sense, much in the same way that move forward on a business plan is treated as a different sense from literal move forward. From a parsing/information extraction point of view, these two approaches are equivalent in terms of their representational requirements.</Paragraph>
      <Paragraph position="1"> The benefit of employing the metaphor-based approach, as suggested in the introduction, comes when performing inference. As shown by (Narayanan, 1997), a metaphorical usage and a literal usage share inferential structure. For example, the aspectual structure of run is the same in either domain whether it is literal or metaphorical usage.</Paragraph>
      <Paragraph position="2"> Further, this sharing of inferential structure between the source and target domains simplifies the representational mechanisms used for inference making it easier to build the world models necessary for knowledge-intensive tasks like question answering (Sinha and Narayanan, 2005).</Paragraph>
    </Section>
  </Section>
class="xml-element"></Paper>
Download Original XML