Four-Month-Old Infants' Categorization of Animals: Does Any Body Part Hold Privileged Status?

Article excerpt

How humans form categories and why they form the categories they do has long been the topic of considerable interest and debate in psychology (see Medin, 1989, for a review). Categorization is an important topic to study and understand because the process itself is so fundamental to human existence and seems to contribute to human survival (Wisniewski & Medin, 1991). Humans must frequently decide whether one species of animal is likely to be harmless or dangerous, whether one type of food is a fruit or a vegetable, whether an object is a chair or a table, and so on. This type of decision making helps humans make sense of and predictions about the environment (Wisniewski & Medin, 1991).

Of particular interest recently are the category decision-making processes of human infants. Early research in the area assessed the ability of infants under 1 year of age to form perceptual categories (e.g., Cohen & Strauss, 1979). More recent recent research has focused on the types of categories formed and the features used in categorizing. For example, research (e.g., Rakison & Butterworth, 1998) has examined the features used by infants to distinguish between animate and inanimate objects. Other research (e.g., Quinn & Eimas, 1996a) has focused on the features used by infants to categorize similar-looking species of animals such as cats and dogs, and has concluded that 3-and 4-month-old infants use the face as the primary feature distinguishing cats from dogs.

Previous research on face processing had examined adults' (e.g., Bruce & Humphreys, 1994) and infants' responses (e.g., Mondloch et al., 1999; Morton & Johnson, 1991) to stimuli resembling human faces. However, as mentioned, Quinn and Eimas (1996a) have recently emphasized the role of the face in young infants' perceptual categorization of basic-level animal categories, whose members are frequently similar in size and shape. According to Quinn and Eimas (1996a), facial information, internal features or external contour, provides infants with "a necessary and sufficient basis for distinguishing cats and dogs" (p. 200).

In arriving at their conclusion about the importance of facial information, Quinn and Eimas (1996a) employed a visual preference procedure in their experiments. In this procedure, infants are shown visual stimuli during a familiarization phase. In a subsequent test phase, infants are presented with the familiar visual stimulus and a novel stimulus. Looking times to both novel and familiar stimuli are compared. Infants' preference for the novel stimulus is taken to indicate they have discriminated between the novel and familiar stimuli. Lack of preference is assumed to indicate, under appropriate experimental control, categorical similarity of the stimuli. Results of the Quinn and Eimas (1996a) series of experiments indicated that 3- and 4-month-old infants are capable of using the face/head region to differentiate dogs from cats.

In Quinn and Eimas' (1996a) initial experiment, infants were randomly assigned to one of three conditions in which they were familiarized with multiple exemplars of cats or parts of cats. In the whole animal condition, infants were familiarized with whole cats, for example, and then tested with a novel whole cat versus a whole dog. Infants in the remaining conditions--face only and body only--were familiarized with cat exemplars whose other body parts (e.g., face, head, torso) were occluded by gray construction paper and tested with a novel cat exemplar versus a dog exemplar, each having its body parts occluded to resemble the occluded familiarization stimuli. The dependent measure was a novelty preference score based on test trials, calculated by dividing total looking time to the out-of-category exemplar by total looking time to the out-of-category exemplar plus total looking time to the novel within-category exemplar. Infants in the whole animal and face only conditions treated the novel cat as familiar (i. …