The Use of Categorical Features in Adult Spatial Reorientation

The Use of Categorical Features in Adult Spatial Reorientation Daniel J. Grodner (dgrodne1@swarthmore.edu) Department of Psychology, 500 College Avenue Swarthmore, PA 19081 USA Carey Pietsch (cpietsc1@swarthmore.edu) Department of Psychology, 500 College Avenue Swarthmore, PA 19081 USA Frank H. Durgin (fdurgin1@swarthmore.edu) Department of Psychology, 500 College Avenue Swarthmore, PA 19081 USA Abstract (Huttenlocher & Vasilyeva 2003), a rhombus (Hupbach & Nadel 2005), and an octagon (Newcombe & Ratcliff 2006). It has been proposed that geometric information might play a privileged role in guiding reorientation. Gallistel (1990) points out that, in a natural environment, the geometric layout of a landscape is relatively stable. Nongeometric features such as the colors or textures of landmarks are subject to seasonal and other types of variation. For this reason, organisms may have evolved to be especially sensitive to geometric information when reorienting. Support for the special status of geometric cues comes from the fact that non-geometric information is often ignored in this task. For instance, Cheng found that rats searched rotationally equivalent corners of a rectangular room even when provided with a landmark feature, such as a differently colored wall, that could specify the correct location (1986). Children up to five years-old also fail to use nongeometric information in certain circumstances (Hermer & Spelke 1994, 1996; Hermer-Velasquez, et al 2001; Learmonth, Nadel & Newcombe 2002). These results have been adduced as support for an encapsulated module that considers only geometric information in reorienting (Cheng 1986; Gallistel 1990; Hermer & Spelke 1986; Wang & Spelke, 2002, 2003) Despite its theoretical appeal, recent results cast doubt on the existence of a purely geometric module for reorienting. One issue is that the size of the enclosure influences sensitivity to nongeometric information. For instance children will ignore featural information if a rectangular enclosure is four feet by six feet, but become more likely to use featural information if the dimensions increase to eight feet by twelve feet (Learmonth, Newcombe & Huttenlocher, 2001; Learmonth, Nadel & Newcombe 2002). This demonstrates that nongeometric information can be used in certain ordinary environments, reducing somewhat the scope of a potential module. Another challenge for a modular architecture is that some types of non-geometric information can be used by children as young as 18-24 months. Huttenlocher and Laurenco (2007) used a square enclosure with circles on the walls. Opposite walls had identically sized circles, but the circles for adjacent walls differed in size. Though all corners were geometrically equivalent, toddlers reliably selected the Children often fail to use informative non-geometric features to recover orientation in cases where adults succeed. The present work demonstrates that adults can also fail to use distinguishing information if that information is not encoded categorically. Adults were shown an object being hidden in a corner of a rectangular room rendered with an immersive virtual display. They were then disoriented and asked to locate the object. The short walls of the room were discriminably different colors so that they could be used to uniquely specify the orientation of the viewer with respect to the enclosure. When the colors were members of different major color categories, participants were more likely to succeed in the task. In fact, participants succeeded in the task if and only if they encoded the color difference lexically. This indicates that categorical encoding plays an important part in reorienting with non-geometric features and implicates language as a default medium of categorical encoding. Keywords: Spatial Memory; Navigation; Language; Reorientation Introduction All mobile animals appear to possess mechanisms for reestablishing orientation in a familiar environment after disorientation (Cheng & Newcombe 2005). To investigate these mechanisms, researchers commonly employ a variant of the disorientation procedure. In this task, the animal or human participant observes while an object is hidden in an enclosed space. The animal is then disoriented before being allowed to search for the object. By manipulating the environmental cues available in the enclosure, a researcher can identify the allocentric features used in this process. For instance, Cheng and Gallistel determined that rats can reorient using the relative lengths of adjacent walls in a rectangular enclosure. If food is concealed in a corner of the room (e.g., where a long wall is to the left of the short wall), a disoriented rat will search both in the correct corner and the opposite corner, which is geometrically equivalent (Cheng & Gallistel 1984; Cheng 1986). Subsequent research has established that children and many animals perform similarly (see Ratcliff & Newcombe 2008 for a review). Children have also been shown to use the geometric arrangement of extended surfaces when reorienting in an enclosure shaped like an isosceles triangle