Quinn (2004) reported that the asymmetry in categorization of humans versus nonhuman animals by human infants (i.e., a representation for humans that includes nonhuman animals vs. a representation for nonhuman animals that excludes humans) was based on holistic information. The current research investigated how this asymmetry was affected by stimulus inversion. Three- to 4-month-olds were familiarized with humans or cats and then tested with a novel cat versus a novel human in upright and inverted presentation conditions. Only the human portion of the asymmetry was affected by inversion, implying that only the upright human familiarization condition engaged configural processing. The findings are consistent with the idea that differences in the way that infants process humans versus nonhuman animals can be likened to an expert-novice distinction in the early development of perceptual category representations.
KEYWORDS: categorization, perceptual expertise, configural processing, infants.
The question of how humans are conceptualized among other animals has met with different answers in the cognitive literature. For example, Carey (1985) has proposed that humans are the prototype for other animals, whereas Medin and Waxman (in press) have argued that humans are represented as having a dual status inclusive of a general representation in which humans are represented as being like other animals and a more specific representation in which humans are represented as being distinct from other animals. The question of how infants represent humans has also drawn different responses, including arguments that infants represent humans as intentional agents that are not constrained by physical principles (Kuhlmeier, Bloom, & Wynn, 2004), or as solid, material objects (Saxe, Tzelnic, & Carey, 2006). There has also been the suggestion that infants represent the configuration of parts that make up the human body (Gliga & Dehaene-Lambertz, 2005), although with debate regarding when during early development this configural representation emerges (Slaughter & Heron, 2004). In addition, there has been the report that humans are the first class of objects to give rise to processes of object individuation (Bonatti, Frot, Zangl, & Mehler, 2002). Finally, Quinn (2005, in press) has reported differences in the ways that young infants, 3 to 4 months of age, represent categories of humans and nonhuman animals, leading to the argument that young infants' considerable familiarity with humans may generate the first domain of perceptual expertise. Here we report evidence that infants' generalization from human to nonhuman animal images occurs only with upright whole images, which is consistent with the idea that infants represent humans on a configural expert-like basis.
Do Infants Represent Humans as a Category?
If one is interested in the question of how infants conceptualize humans, an important issue to address is whether infants categorize humans as similar to or different from categories of nonhuman animals. Given the perceptual differences between humans and nonhuman animals (i.e., humans wear clothes, stand upright, and have arms) and given the abilities of young infants to separate various nonhuman animal species into different categories (Eimas & Quinn, 1994), one might expect that infants would represent humans as distinct from categories of nonhuman animals. To determine whether this was the case, Quinn and Eimas (1998) presented 3- to 4-month-olds with a dozen photographic images of clothed humans (males and females) in standing, walking, and running poses followed by a preference test consisting of a novel human paired with a horse or cat image. Counter to expectation, the infants generalized their familiarization to the novel humans, cats, and horses; they did not exhibit novel category preferences for the cats or horses. In addition, a control experiment assessing a priori preference showed that infants did not have a spontaneous preference for looking at humans over cats or horses. …