<?xml version="1.0" standalone="yes"?>
<Paper uid="P06-1094">
  <Title/>
  <Section position="4" start_page="745" end_page="746" type="intro">
    <SectionTitle>
2 Data
</SectionTitle>
    <Paragraph position="0"> Below we discuss previous psycholinguistic experients, focusing on how contextual factors such as distance, size, and salience may affect proximity.</Paragraph>
    <Paragraph position="1"> We also present novel examples, showing that the location of other objects in a scene may interfere with the acceptability of a proximal description to locate a target relative to a landmark. These examples motivate the model in SS3.</Paragraph>
    <Paragraph position="2">  the relation the X is near O as a function of the position occupied by X.</Paragraph>
    <Paragraph position="3"> Spatial reasoning is a complex activity that involves at least two levels of processing: a geometric level where metric, topological, and projective properties are handled, (Herskovits, 1986); and a functional level where the normal function of an entity affects the spatial relationships attributed to it in a context, cf. (Coventry and Garrod, 2004).</Paragraph>
    <Paragraph position="4"> We focus on geometric factors.</Paragraph>
    <Paragraph position="5"> Although a lot of experimental work has been done on spatial reasoning and language (cf.</Paragraph>
    <Paragraph position="6"> (Coventry and Garrod, 2004)), only Logan and Sadler (1996) examined topological prepositions in a context where functional factors were excluded. They introduced the notion of a spatial template. The template is centred on the landmark and identifies for each point in its space the acceptability of the spatial relationship between the landmark and the target appearing at that point being described by the preposition. Logan &amp; Sadler examined various spatial prepositions this way. In their experiments, a human subject was shown sentences of the form &amp;quot;the X is [relation] the O&amp;quot;, each with a picture of a spatial configuration of an O in the center of an invisible 7-by-7 cell grid, and an X in one of the 48 surrounding positions. The subject then had to rate how well the sentence described the picture, on a scale from 1(bad) to 9(good). Figure 2 gives the mean goodness rating for the relation &amp;quot;near to&amp;quot; as a function of the position occupied by X (Logan and Sadler, 1996). It is clear from Figure 2 that ratings diminish as the distance between X and O increases, but also that even at the extremes of the grid the ratings were still above 1 (min. rating).</Paragraph>
    <Paragraph position="7"> Besides distance there are also other factors that determine the applicability of a proximal relation.</Paragraph>
    <Paragraph position="8"> For example, given prototypical size, the region denoted by &amp;quot;near the building&amp;quot; is larger than that of &amp;quot;near the apple&amp;quot; (Gapp, 1994). Moreover, an object's salience influences the determination of the proximal region associated with it (Regier and Carlson, 2001; Roy, 2002).</Paragraph>
    <Paragraph position="9"> Finally, the two scenes in Figure 3 show interference as a contextual factor. For the scene on the left we can use &amp;quot;the blue box is near the black box&amp;quot; to describe object (c). This seems inappropriate in the scene on the right. Placing an object (d) beside (b) appears to interfere with the appropriateness of using a proximal relation to locate (c) relative to (b), even though the absolute distance between (c) and (b) has not changed.</Paragraph>
    <Paragraph position="10"> Thus, there is empirical evidence for several  contextual factors determining the applicability of a proximal description. We argued that the loca-tion of other distractor objects in context may also interfere with this applicability. The model in SS3 captures all these factors, and is evaluated in SS4.</Paragraph>
  </Section>
class="xml-element"></Paper>