Further Evidence That Rats Rely on Local Rather Than Global Spatial Information to Locate a Hidden Goal: Reply to Cheng and Gallistel (2005) Anthony McGregor, Peter M. Jones, Mark A. Good, and John M. Pearce Cardiff University Naive male Hooded Lister rats (Rattus norvegicus) were required to find a submerged platform in a right-angled corner between a long and a short wall of a pool in the shape of an irregular pentagon. Tests in a rectangular pool revealed a preference for the corners that corresponded with the correct corner in the pentagon. These findings indicate that rats identified the correct corner in the pentagon by local cues. They contradict the suggestion that rats navigate by moving in a particular direction relative to the principal axis of the shape of their environment. Keywords: spatial learning, geometric module, navigation Considerable theoretical significance has been attached to the discovery that animals can find a hidden goal by reference to the shape of their environment. The first experiment to demonstrate this ability was by Cheng (1986). Hungry rats were placed in a rectangular arena with food hidden in one corner and with a distinctive landmark in each corner. The rats soon learned to search in the correct corner of the arena, but initially they made rotational errors of searching in the corner that was diametrically opposite to this corner. Furthermore, when they were tested in the arena in the absence of the landmarks and food, rats showed a strong preference for the two corners that were geometrically equivalent to the corner that had previously contained food. Cheng (1986; see also Gallistel, 1990) proposed that these findings were a consequence of rats using the overall shape of the arena to identify where food was located. Learning about the shape of the environment was said to take place in a dedicated module, the exclusive purpose of which was to encode information about the geometric properties of the environment. Since the report by Cheng (1986), a number of similar experiments have been con- ducted and with a variety of species, including fish (Sovrano, Bisazza, & Vallortigara, 2002), chicks (Vallortigara, Zanforlin, & Pasti, 1990), rhesus monkeys (Gouteux, Thinus-Blanc, & Vauclair, 2001), and humans (Hermer & Spelke, 1994). The results from these experiments have been consistent with those reported by Cheng (1986), and they have been interpreted in much the same way (see Cheng & Newcombe, 2005, for a review). See also Cheng (2005a, 2005b) for accounts of navigation in a rectangular arena in which geometric and nongeometric information may interact. There is, therefore, a widespread belief that many different species navigate with reference to the shape of the environment and that this ability reflects the existence of cognitive and neural mechanisms that are common to species as diverse as fish and humans. Set against this consensus are two reports of experiments that have been said to show that animals might not use the overall shape of the environment to find a hidden goal. If this is correct, then the justification for postulating the existence of a geometric module that processes global information based on the shape of the environment is seriously undermined. Pearce, Good, Jones, and McGregor (2004) trained rats to find a submerged platform in one corner of a rectangular pool before the rats were tested in a pool in the shape of a kite. These shapes are shown in the top left and center panels of Figure 1, where the platform is located in a corner in the rectangle with a long wall to the left of a short wall. The kite-shaped pool was constructed from the same walls as the rectangular pool, and the corners between the short and long sides were both right angled. As a consequence, one of the right-angled corners (the correct corner) in the kite corre- sponded with the corner containing the platform in the rectangle, and the other right-angled corner in the kite (the incorrect corner) corresponded with the corners where the platform was never located. Pearce et al. (2004) argued that if rats found the platform in the rectangular pool by referring to its overall shape, then they should effectively be lost in the kite because of the lack of congruence between the shapes of the two environments. In con- trast to this prediction, rats showed a clear preference for searching in the correct, rather than the incorrect, corner of the kite. Pearce et al. (2004) explained this outcome by proposing that during their training in the rectangular pool, rats identified the location of the platform by means of local information, rather than global infor- mation, such as the overall shape of the pool. For instance, the platform in the rectangle in Figure 1 could be found by swimming toward a corner with a long wall to the left of a short wall. Alternatively, it could be found by swimming to the right-hand end of a long wall. Either of these strategies would result in rats heading for the corner with the platform in the rectangle or to the Anthony McGregor, Peter M. Jones, Mark A. Good, and John M. Pearce, School of Psychology, Cardiff University, Cardiff, United Kingdom. This research was supported by a grant from the United Kingdom Medical Research Council. We thank Laura Bunting for her assistance with the experiment. Correspondence concerning this article should be addressed to Anthony McGregor or John M. Pearce, School of Psychology, Cardiff University, Cardiff CF10 3YG, United Kingdom. E-mail: mcgregora@cardiff.ac.uk or pearcejm@cardiff.ac.uk Journal of Experimental Psychology: Copyright 2006 by the American Psychological Association Animal Behavior Processes 2006, Vol. 32, No. 3, 314 –321 0097-7403/06/$12.00 DOI: 10.1037/0097-7403.32.3.314 314