The Age Group Labels and Categories of Preschool Children

Questions of how young children use “age” groups to understand the social world led to 2 studies exploring the content of preschool children’s age group labels and categories. Study 1 included 32 children aged 2-4 years and determined spontaneous labels for both photographs and dolls representing the life span. Results indicated that children readily labeled all ages using a relatively limited set of terms, but showed less patterned labeling of stimuli representing adults than children. Girls’ labels were more structured than boys’. Older preschoolers showed more differentiated structures than did younger ones and used more kinship terms as labels. Study 2, on 84 children aged 3-5, was a photograph-sorting task that determined the points of transition between age categories as well as subjects’ own self-identification by age group. Results indicated that preschoolers used a nonadult method of dividing up the life span. Older children made fewer errors. Age self-identification was congruent with how children sorted photos of unfamiliar peers. However, younger boys and girls differed in their self-identification, perhaps reflecting differences in gender identification processes. ow to make sense of the immensely diverse and dynamic human world around them is a task that faces all young children as they seek to understand social interaction and to gain predictable responses from others. Recent research has shown that even infants and toddlers have more elaborate social networks than was previously realized, comprising relationships with mother, father, siblings, and peers; and certainly by preschool age most youngsters encounter on an almost daily basis a broad array of familiar and unfamiliar people in many different kinds of settings. Yet, because most of the category systems that are important to adults, such as occupation, social class, and ethnic, political, and community affiliation, are too abstract (Furth, 1980), the young child is forced to rely on social categories related to only the most static and overt cues. Hence age, gender, and familiarity are the attrib* Published in Child Development 55 (1984), pp. 440-452. Copyright © 1984 by the Society for Research in Child Development, Inc. All rights reserved. Used by permission. H T H E A G E G R O U P L A B E L S A N D C A T E G O R I E S O F P R E S C H O O L C H I L D R E N 2 utes that the child may first use to differentiate the human array (Lewis & Feiring, 1979). In fact, Kohlberg (1966) has suggested that the age-size differentiation (babies/boys and girls/grown-ups) is the earliest categorization that the child imposes on the social environment, preceding even the gender distinction. Behavioral studies indicate that infants (in one study as young as 3 months) respond and initiate quite differently to other infants than they do to either familiar or unfamiliar adults (Field, 1979; Fogel, 1979). Infants also behave differently to children than they do to adults (Brooks & Lewis, 1976). As children grow older, they develop a complex “curriculum” for social interaction in which the relative age of the other is a strong predictor of what a child will actually do (Whiting & Whiting, 1975). By age 2 or 3, children associate particular “social functions” (or types of social interaction, such as nurturance, dependency, etc.) with particular ages of “social partners” (Edwards & Lewis, 1979). Yet despite their functionality for the preschool child, the earliest category systems related to age groups have received little attention from developmental psychologists. However, we do understand the development throughout the childhood years of at least three relevant social-cognitive processes, and we can describe how preschoolers’ cognitive limitations cause them to have nonadult understandings of age. The first process concerns judging or assessing chronological age. The evident interest of preschoolers in “age” differences and comparisons between people clearly does not derive from an ability to accurately judge chronological age. That task is extremely difficult and eludes children younger than age 10 or so (Stevenson, Miller, & Hale, 1967) because it requires understanding a linear, quantitative scale of increasing years. Preschoolers’ judgments of “older” and “younger” are largely fused with “bigness” and/or “smallness” (Britton & Britton, 1969; Clark, 1972; Piaget, 1969), and their error rate is high when size cues are put into conflict (Kratochwill & Goldman, 1973; Kuczaj & Lederberg, 1977; Looft,1971; Looft, Rayman, & Rayman, 1972). The judgments of” older preschoolers may not be based entirely on size, however, because Kogan, Stephens, and Shelton (1961) held the size of stimuli constant by using passport photos and demonstrated that 50% of their 4-year-old and 90% of their 5 year-old sample showed better-than-chance agreement with adults in the age ranking of faces representing the life span. These older preschoolers had a beginning awareness of hair and facial indicators of elderly age. In general, preschoolers perform better at making relative age judgments when the stimulus materials are more realistic (Kratochwill & Goldman, 1973). They are able to judge female stimuli as accurately as male ones (see abovementioned studies). C A R O L Y N P O P E E D W A R D S 3 A second process concerns understanding of social roles. Toddlers use a few social categories such as “baby,” “kid,” and “man” as simple labels. Children do not begin to define a social category (e.g., “child”) in relation to its complement (e.g., “adult”) until age 31⁄2–4 (Watson & Fischer, 1980). Then, in terms of thinking about transformations, they can mentally operate on only one dimension at a time (such as age or gender) until age 6 or 7, causing interesting distortions in reasoning about parent and child roles (Fischer & Watson, 1981). The third process concerns children’s kinship knowledge. Two of children’s earliest social labels, “mommy” and “daddy,” are sometimes overextended to include people other than parents (Brooks-Gunn & Lewis, 1979; Greenfield, 1973; Thomson & Chapman, 1977). Research on children’s acquisition of kinship knowledge has established that preoperational children define kinship terms categorically, that is, in terms of concrete attributes related to age and sex, rather than relationally, that is, in terms of descent or marriage links (Chambers & Tavuchis, 1976; Danziger, 1957; Elkind, 1962; Haviland & Clark, 1974; Jordan, 1980; Piaget, 1928). For example, young children insist that old men cannot be “fathers” because they must be “grandfathers.” However, while stressing preschoolers’ conceptual limitations, at least the literature indirectly implies that preschoolers sometimes use kinship terms as labels to designate sex/age groups in society. The relative neglect of research on preschool children’s social categories suggests the need for a close study of this age group. First, a study was conducted to elicit from children aged 2–4 years terms of reference for stimuli representing the life span. This study used both doll and photographic stimuli in order to make it likely that results were not idiosyncratic to one task. Following Daum (1978), the age labels produced by children are termed “age ascriptions,” but it is not presumed that the labels necessarily correspond one-to-one to underlying concepts. Young children possess nonlinguistic concepts onto which words have not yet been mapped (Smith, 1978); they also may respond appropriately to a word on a comprehension task that they do not produce on a labeling task (Anglin, 1978). Nevertheless, it is presumed that children learn terms of reference for those categories that are salient and significant in their dealings with the world (Anglin, 1977), and age group terms appear to be socially useful and highly interesting to children. This study determines children’s working vocabulary—what terms of reference T H E A G E G R O U P L A B E L S A N D C A T E G O R I E S O F P R E S C H O O L C H I L D R E N 4 they most use—as they attempt to label a social array varying in age and gender. It is claimed that their use of labels indicates which age groups are most noticeable to them and also what kinds of people they find to be perceptually or functionally similar (Nelson, Rescorla, Gruendal, & Benedict, 1978). Brooks-Gunn and Lewis (1979) found that toddlers aged 19-22 months could readily label baby and adult photographs (usually using terms, “baby,” “mommy,” “daddy,” “lady,” and “man”), but they had much more difficulty with photographs of children. The present study uses an older age group of subjects and stimuli representative of a much greater array of target ages. The format of the task was designed to highlight the dimension of age for the children in that all doll or photo stimuli were spread out at once on a table before the child so that they could be easily contrasted. Three questions were addressed. The first hypothesis was that the stimuli that would be the most difficult for children to label would be the adult stimuli. This hypothesis follows from Britton and Britton’s(1969) conclusion that 2–4-year-olds cannot age order drawings representing adults and merge “all persons past the young adult level in one big category” (p. 459). The alternate hypothesis was that the child stimuli would present the greatest difficulty based on Brooks-Gunn and Lewis’s (1979) above-mentioned finding for toddlers. The second question concerned age differences within the sample. Older preschoolers were hypothesized to respond, relative to younger ones, with (1) more structure or pattern (vs. nonpatterned responding), and (2) more differentiations (age group distinctions) made between stimuli, The alternative hypothesis, that subject age differences do not occur, would emerge if children’s use of age group terminology undergoes few developments during years 2-4. The alternate hypothesis seemed unlikely but worth examining because processes of chronological age assessment and kinshipterm definition undergo their major transformations during the elementary rather than preschool years. The third question concerned sex differences. Girls were hypothesized to perform more maturely than boys. Many prior researchers studying age and kinship knowledge have mentioned significant or nonsignificant trends to favor young girls(Britton & Britton, 1969; Chambers & Tavuchis, 1976; Danziger, 1957; Jordan, 1980; Kogan et al., 1961; Kratochwill & Goldman, 1973; LeVine & Price-Williams, 1974). The alternative hypothesis, that no sex differences would be found, would be inline with Schantz’s (1975) general conclusion concerning social-cognitive tasks. C A R O L Y N P O P E E D W A R D S 5 Because age labels were so easily elicited by Study 1 and because there was so much consensus in their application, a further study was conducted to find out to what chronological age breaks the most common labels actually corresponded, and to investigate children’s own age group self-concept.