A Computational Framework

A Computational Framework One of the more articulated theories of apprehension that specifies particular processes and representations is Logan and Sadler’s (1996) computational framework. According to this framework, apprehending spatial terms involves the following processes: (a) identifying and spatially indexing the reference and located objects (finding the relevant objects); (b) imposing a reference frame with appropriate settings on the reference object (assigning directions to space); and (c) evaluating the goodness of fit of the spatial term with respect to the placement of the located object within various regions on the spatial template (computing and comparing the spatial relation). Using Spatial Language 131One important feature of this framework is that these constituent processes apply to a wide range of tasks within both production and comprehension, thereby accommodating various goals of the speaker and addressee. For example, in a relation judgment task, the speaker either informs the addressee of the whereabouts of a given object or seeks such information, as in 4a and 4b. In a cueing task, the emphasis is on the identity of one of the objects, and spatial relation is provided as a means of picking out the relevant object, as in 5a and 5b. Finally, in a verification task, the goal is to determine whether the spatial description is true of a given situation (6a), with an emphasis either on the relation (6b) or on the objects (6c). (4a) The mug is BELOW the coVee pot. (4b) Where is the mug? (5a) The MUG is below the coVee pot. (5b) What is below the coVee pot? (6a) The mug is below the coVee pot. (6b) Is the mug BELOW the coVee pot? (6c) Is the MUG below the coVee pot? Particular words are capitalized in 4–6 to indicate the emphasis that a speaker may place in the service of communicating a particular goal. This raises an important methodological point. Many previous studies examining the processes and representations involved in apprehension have relied on written comprehension in which participants read spatial descriptions (Carlson & Logan, 2001; Carlson-Radvansky & Irwin, 1993, 1994; Carlson- Radvansky & Logan, 1997; Carlson-Radvansky & Tang, 2000; Clark & Chase, 1972; Logan & Sadler, 1996) rather than oral comprehension in which they hear spoken utterances. Participants in these studies are serving as addressees, much like readers serve as addressees for authors. However, the written version does not have a traditional means for conveying emphasis, as is the case for speech in which one can emphasize a word by pronouncing it more carefully, louder, or with a longer duration. As such, the particular goal is not overtly conveyed by the spatial description, as reflected by the fact that the descriptions corresponding to the various goals in 4a, 5a, and 6a are identical. Thus, in interpreting the results from these types of studies it is particularly important that the goal for the addressee is identified