Penetrating the Geometric Module: Catalyzing Children's Use of Landmarks

Penetrating the Geometric Module: Catalyzing Children’s Use of Landmarks Alexandra Twyman (atwyman@temple.edu) Department of Psychology, 1701 N. 13 th Street Philadelphia, PA 19122 USA Alinda Friedman (alinda@ualberta.ca) Department of Psychology, 217 Biological Sciences Bldg, University of Alberta Edmonton, AB T6G 2E9 CANADA Marcia Spetch (mspetch@ualberta.ca) Department of Psychology, 217 Biological Sciences Bldg, University of Alberta Edmonton, AB T6G 2E9 CANADA Abstract We examined whether 4­5 year old children could be trained to use landmark features to relocate goals after disorientation. In Experiment 1, half of the children were pretrained in a small equilateral triangle shaped room with different colored walls. These children and a control group were tested in a small rectangular room with a feature wall. Children with pretraining responded more frequently to the correct corner than to the diagonally congruent corner on their first set of four trials in the rectangular room whereas the children in the control group used geometric cues exclusively. Three additional groups of children (Experiment 2) showed that the use of landmark features – both salient and subtle – can be learned in as few as four practice trials in a small rectangular room. The data support the view that both geometry and landmark features are combined in the same representation. Key Words: children; feature; learning; single representation Introduction The ability to orient oneself in space is a fundamental skill that has evolutionary and ecological significance for all mobile organisms. Virtually all species tested encode geometric properties of the environment (see Cheng & Newcombe, 2005, for review). While children under 6 years of age (Hermer­Vasquez, Moffet, & Munkholm, 2001; Hermer & Spelke, 1996) appear to use geometric cues exclusively, many species are able to conjoin geometric and feature cues, including pigeons, (Kelly, Spetch, & Heth, 1998) rhesus monkeys (Gouteux, Thinus­Blanc, & Vauclair, 2001) chicks, (Vallortigara, Zanforlin, & Pasti, 1990) and fish (Sovrano, Bisazza, & Vallortigara, 2002). Cheng (1986) first proposed the existence of an encapsulated geometric module devoted to the task of orienting (and reorienting) in space. In a working memory task, rats used geometric information exclusively even when the corners of the space were completely disambiguated by various non­geometric landmarks. In a reference memory task, the rats learned to use featural information to locate the target within 30 trials. However, when the featural information was removed from the target corner, the rats were unable to use the remaining cues and reverted to using geometric cues exclusively. Cheng hypothesized that rats reoriented by using representations encoded in a geometric module, which essentially records metric information about the shape of the environment. Information about landmarks near the target was hypothesized to be “glued” onto the metric frame provided by the geometric module. Subsequent researchers have assumed that such nongeometric features are encoded and represented in separate modules (see Cheng & Newcombe, 2005, for review). Hermer and Spelke (1994, 1996) extended Cheng’s (1986) work to human children and found that younger children (between 2 and 4 years of age) behaved identically to the rats. Choices were divided evenly between the rotationally equivalent corners of a small rectangular room, even though a blue wall – a landmark – could have been used to improve accuracy. Hermer­Vasquez, Moffet, and Munkholm (2001) demonstrated that children are able to integrate landmark information with geometric information at about the same age as they are mastering spatial language, particularly “left” and “right”. However, Learmonth, Nadel, and Newcombe (2002) challenged the modular view by contrasting performance of 5­year­old children in small (4 x 6ft.) and large (8 x 12ft.) rooms. The children in the small room replicated Hermer and Spelke’s (1996) findings: they failed to use the blue feature wall. In contrast, children were able to integrate both feature and landmark information successfully in the large room. Thus, children are able to use feature information before the mastery of spatial language in some cases. By age 6, children were able to succeed regardless of room size. Learmounth et al. (2002) proposed three explanations for why room size would matter. First, younger children may fail to generalize what they know about navigation in large spaces to the small experimental space. With more experience, older children may learn that color can be a useful cue regardless of the size of the search space. Second, the task demands of the two spaces may have been different. The last, and related, possibility is that the blue wall was used as a heading cue in the large room, but may not have been perceived as a useful heading cue in the small room. The fragility and variability of landmark use with room size argues against an encapsulated representation of geometric information. It is difficult to

Récupéré en direct depuis OpenAlex et désinversé. Les résumés ne sont pas conservés dans cette base de données : les index inversés représentent 8,6 Go des 9,3 Go de texte de la base, et le serveur dispose de 13 Go libres.