Updating egocentric representations in humannavigationRanxiao Frances Wanga,*, Elizabeth S. SpelkebaDepartment of Psychology, University of Illinois, 603 E. Daniel Street, Champaign, IL 61820, USAbDepartment of Brain and Cognitive Sciences, Massachusetts Institute of Technology, Cambridge,MA 02139, USAReceived 25 June 1999; received in revised form 8 May 2000; accepted 25 July 2000AbstractSeven experiments tested whether human navigation depends on enduring representations,or on momentary egocentric representations that are updated as one moves. Human subjectspointed to unseen targets, either while remaining oriented or after they had been disorientedby self-rotation. Disorientation reduced not only the absolute accuracy of pointing to allobjects (`heading error') but also the relative accuracy of pointing to different objects (`con®g-uration error'). A single light providing a directional cue reduced both heading and con®g-uration errors if it was present throughout the experiment. If the light was present duringlearning and test but absent during the disorientation procedure, however, subjects showedlow heading errors (indicating that they reoriented by the light) but high con®guration errors(indicating that they failed to retrieve an accurate cognitive map of their surroundings). These®ndings provide evidence that object locations are represented egocentrically. Nevertheless,disorientation had little effect on the coherence of pointing to different room corners, suggest-ing both (a) that the disorientation effect on representations of object locations is not due to theexperimental paradigm and (b) that room geometry is captured by an enduring representation.These ®ndings cast doubt on the view that accurate navigation depends primarily on anenduring, observer-free cognitive map, for humans construct such a representation ofextended surfaces but not of objects. Like insects, humans represent the egocentric distancesand directions of objects and continuously update these representations as they move. Theprincipal evolutionary advance in animal navigation may concern the number of unseentargets whose egocentric directions and distances can be represented and updated simulta-neously, rather than a qualitative shift in navigation toward reliance on an allocentric map.q 2000 Published by Elsevier Science B.V. All rights reserved.Keywords: Updating; Egocentric representations; Human navigationR.F. Wang, E.S. Spelke / Cognition 77 (2000) 215±250 215Cognition 77 (2000) 215±250www.elsevier.com/locate/cognit0010-0277/00/$ - see front matter q 2000 Published by Elsevier Science B.V. All rights reserved.PII: S0010-0277(00)00105-0COGNITION* Corresponding author. Tel.: 11-217-244-3664; fax: 11-217-244-5876.E-mail address: [email protected] (R.F. Wang).1. IntroductionHow do people and animals represent the spatial properties of their environmentso as to locate objects and navigate effectively to signi®cant places? Research oninsects suggests that one form of navigation ± homing ± can depend on a continuousupdating process over self-motion, i.e. path integration (Srinivasan, Zhang, Lehrer,& Collett, 1996; Wehner & Srinivasan, 1981; see also Collett, 1996; Dyer, 1996).Rodents also have a path integration system that allows them to move to and fromsigni®cant locations such as the nest and the site of an enduring food source(Etienne, Maurer, & Sguinot, 1996; Mittelstaedt & Mittelstaedt, 1980; for reviewsee Gallistel, 1990). Unlike many ants and bees, however, rodents also are able tonavigate to familiar objects along novel paths from novel, arbitrary points, suggest-ing that their spatial learning involves the construction of a qualitatively differenttype of spatial representation: an enduring, observer-free `cognitive map' of theenvironment (e.g. O'Keefe & Nadel, 1978; Sutherland & Dyck, 1984; Tolman,1948; for discussion see Bennett, 1996; Gallistel, 1990). Evidence for such cognitivemaps gains much intuitive appeal from studies of humans. Although humans seem tohave a path integration mechanism that resembles that of insects and rodents (e.g.Berthoz, Israel, Francois, Grasso, & Tsuzuku, 1995; Fukusima, Loomis, & Da Silva,1997; Loomis et al., 1993), they also can perform diverse spatial tasks, such asimagining and drawing the furniture in a room, navigating through unfamiliar terri-tory by means of real maps, and even charting new territory during explorations.These latter abilities suggest that real maps have a mental counterpart, and thathumans and other mammals navigate by constructing and using enduring mentalrepresentations of the allocentric distances and directions of the objects and places intheir environment.In addition to behavioral evidence such as that cited above, evidence from neuro-physiological experiments has been interpreted as supporting the existence of one ormore cognitive maps of the environment. In particular, a variety of studies haveshown that individual neurons in the hippocampus of freely moving rats are activewhen a rat moves through a particular region of the environment (McNaughton,Knierim, & Wilson, 1995; O'Keefe & Nadel, 1978). Although vestibular, somato-sensory and visual cues are effective information for the establishment and modi®-cation of this ®ring pattern, the ®ring does not appear to rely exclusively on one oranother cue, for it persists when a rat is carried passively through the environment,when visual cues are removed, and when the visual ®eld is altered by changing therat's facing direction (e.g. Gothard, Skaggs, & McNaughton, 1996; O'Keefe &Speakman, 1987; Quirk, Muller, & Kubie, 1990; for review see McNaughton etal., 1995). Perhaps most important, the receptive ®elds of different place cells andhead direction cells in the same animal show internal coherence during cue manip-ulations (Knierim, Kudrimoti, & McNaughton, 1995; Muller & Kubie, 1987). These®ndings suggest that ensembles of hippocampal neurons serve as a cognitive map ofthe environment (O'Keefe & Nadel, 1978; Wilson & McNaughton, 1993).Nevertheless, two quite different characterizations of mammalian navigation arecompatible with the behavioral and neurophysiological evidence cited above.R.F. Wang, E.S. Spelke / Cognition 77 (2000) 215±250216According to one class of accounts (Gallistel, 1990; O'Keefe & Burgess, 1996;O'Keefe & Nadel, 1978), mammals form a representation of the allocentric loca-tions of the signi®cant objects and places in the environment. As