Symposium
Chair: Maya G. Sen
Discussant: Patricia J. Bauer
The categories of gender and race feature prominently in our society. These characteristics are salient, stable features of people. Thus, the ability to categorize oneself or others as female or male, or by race, is generally an effortless and automatic process. This categorical knowledge, combined with people's beliefs about differences between gender or racial groups (i.e., stereotypes) and their attitudes about these social categories are known as schemas. Schemas can be used to interpret the behaviors and characteristics of others, and to guide behavior, a process known as schematic processing. The purpose of this symposium is to provide an overview of the recent research on infants' social categorization and early schematic processing, and on the relations between these areas. Beginning in the first year of life, infants exhibit social category knowledge. For example, in the area of gender, nine-month-olds are able to discriminate between female and male faces (Leinbach & Fagot, 1993). Between 18 and 24 months of age, gender stereotype knowledge begins to emerge (Serbin et al., in press). Also, infants' gender schemas can affect their memory for gender-stereotyped information. For example, 25-month-old boys show better memory for masculine- than for feminine-stereotyped information (Bauer, 1993). However, the precursors and the extent of schematic processing in infancy are still unknown. In order for stereotypes about a particular category to develop, certain skills and knowledge (e.g., social categorization) must first be in place. Schema theorists emphasize cognitive maturity as the catalyst for schematic processing. For example, according to gender schema theories, before children begin to act in a gender-stereotypical manner, gender must first be a salient distinction for them. The type of knowledge that is critical is still at issue. Traditionally, knowledge of one's own gender was thought to be crucial. Recently, researchers have argued that only rudimentary, implicit gender concepts (e.g., discrimination between females and males) may be needed (Serbin et al., in press). Despite the theories positing its early onset, most research on gender-based schematic processing has focused on children ages 3 years and older, because traditional measures of schematic processing require verbal responses that are too complex for children under three years.The research in this symposium examines the development of social category understanding across infancy. The first talk explores 9- and 14-month-olds' racial categorization, and finds that infants can categorize faces by race and by gender within race. The second talk is on infants' attention to gender categories in the second year, and finds that infants begin to attend to the categories of female and male between 18 and 22 months, although they can discriminate between these categories earlier. The third and fourth talks in this symposium use new methodologies to find early evidence of gender-based schematic processing. The third talk is a study of gender-based schematic processing in 18- to 36-month-old children, using a measure of vocabulary development, and finds greater same-gender than other-gender productive vocabularies. The fourth uses a measure of elicited imitation to find gender-based schematic processing in 24- and 30-month-olds, and finds that gender salience and various types of gender knowledge relate to differential memory for own-gender information. These findings will be discussed in the context of schematic processing theories.
Details of individual items:
paper
Although it is known that attention to gender categories emerges during infancy, little research has considered the early development of attention to racial categories. In the current studies, visual habituation was used to examine the development of racial categorization and identification during infancy. The goal of Study 1 was to determine whether, and when, infants attend to the categorical distinction between African-American and Euro-American faces. Participants (9- and 14-month-old infants) were Euro-American and viewed photographs of African-American and Euro-American adults. During familiarization, infants were presented with 4 different faces of the same race. Following familiarization, infants were tested with a familiar face, a new face of the familiarized race (within-category face) and a new face of the other race (out-of-category face). Infants showed clear categorical responding to these faces. They increased attention to the out-of-category face but not to the within-category face. Categorization on the basis of race could lead to reduced discriminability within race on the basis of other features. Adults have difficulty identifying the sex of faces of members of races other than their own (O'Toole, Peterson, & Deffenbacher, 1996). This 'other-race effect' may be present as early as infants categorize faces on the basis of race. Although Leinbach and Fagot (1993) established that infants categorize faces on the basis of sex by 9 months of age, it is not known whether this result generalizes to infants' viewing other-race faces. The goal of Study 2 was to determine whether 9- and 14-month old infants categorize 'other-race' faces on the basis of sex. The procedure and design were similar to Study 1 but infants (all Euro-American) were familiarized with four African-American faces of the same sex. Following familiarization, infants were tested with a familiar face, a new face of the familiarized sex (within-category face) and a new face of the other sex (out-of-category face). Again, infants showed clear categorical responding, increasing attention to the out-of-category face, but not to the within-category face. Thus, it seems clear that 9-month-old infants (unlike the adults in O'Toole, et al.'s research) have no difficulty in categorizing other-race faces on the basis of sex. Nonetheless, it is possible that an 'other-race effect' might emerge early in life. The goal of Study 3 (currently in progress) is to more directly assess infants' visual recognition memory for same- and other-race faces. Nine-month-old infants are briefly familiarized with a face (either Euro-American or African-American) and then presented with the familiar face paired with a novel face of the same race. Visual recognition memory is assessed by calculating a novelty-preference score. Better memory for the same-race faces (as is found in adults) would be indicated by higher novelty-preference scores for the same-race faces than for the other-race faces. The results of these studies will tell us whether infants respond to the same kinds of distinctions that are meaningful to adults. To the extent that these categorical distinctions support inferences about other characteristics, the roots of social stereotyping may be present from early in life. Portions of Study 1 and Study 2 were presented at the Society for Research in Child Development, 1999.
paper
Research has demonstrated that infants can distinguish between female and male faces by nine months of age (Leinbach & Fagot, 1993). However, despite the very early emergence of gender category recognition, awareness of gender stereotypes is not found until relatively later. Serbin et al. (1999) found stereotype knowledge about toys in 18-month-old girls; however, most studies have not consistently found gender stereotype knowledge until about two years of age, and this knowledge improves rapidly between two and three years of age (Thompson, 1975; Weinraub et al., 1984). The current studies examined the emergence of attention to gender categories, which may be a mediating step between gender category recognition and gender stereotype formation. That is, attention to the particular categories that support stereotypes must be in place before children learn associations between a category and a behavior or object. The very early emergence of recognition of gender categories, and the relatively later acquisition of gender stereotype knowledge suggests that the ability to distinguish between women and men appears to be a necessary but not a sufficient condition for the formation of stereotypes. Rather, children may need to spontaneously divide the social world into categories of women and men, and girls and boys, before they can begin to notice associations between gender and behaviors or attributes. Experiment 1 examined 18-month-olds' attention to gender categories using a sequential touching task. Infants were given a set of four female and four male dolls to manipulate for two minutes, and their categorical touching patterns were examined. Results suggested that 18-month-olds showed little attention to gender. Experiment 2 tested the possibility that the infants could not discriminate the dolls by gender. Fourteen- and 18-month-olds participated in an object examining task to assess their ability to distinguish the male and female dolls used in Experiment 1. Both 14- and 18-month olds were able to discriminate the gender categories, suggesting that 18-month-olds' lack of attention to gender in Experiment 1 was not due to an inability to distinguish male and female dolls. Experiment 3 investigated developmental changes in attention to gender categories in 18-, 22-, and 30-month-olds. Children's categorical manipulation of a different set of male and female dolls was examined using a sequential touching task. A sharp increase in attention to categories was found between 18 and 22 months, but no further increase was found between 22 and 30 months. Given these findings, it is perhaps not surprising that gender stereotypes are not found consistently until around two years of age (Serbin et al2E, 1999; Thompson, 1975; Weinraub et al., 1984), while recognition of male and female faces emerges much earlier, around nine months (Leinbach & Fagot, 1993). Until children actively attend to gender in the social environment, they may not be ready to use gender categories to derive associations between gender categories and behaviors or objects, and to support inferences about behaviors and object preferences. These findings also point to the importance of distinguishing between gender category recognition, as measured by visual habituation or object examining tasks, and attention to gender categories as measured by more cognitively demanding tasks, such as the sequential touching task. Portions of this research were presented at the Society for Research in Child Development, 1997 and at the International Society for the Study of Behavioural Development, 1996.
paper
Gender-based schematic processing (i.e., using gender knowledge as a basis for attending to, interpreting, and structuring information) is evident in 25-month-old boys' memory for gender-stereotyped events (Bauer, 1993), and in 3 1/2-year-old girls' and boys' reasoning about gender-typed properties (Bauer & Coyne, 1997). Evidence for schematic processing has not been established for children younger than 2 years of age. In the present research, we used the unobtrusive measure of children's vocabulary development to examine gender-based schematic processing in the 2nd and 3rd years of life. During this period, children's vocabularies develop very rapidly (Fenson et al., 1994). Developments in language are related to development in other cognitive, social, and cultural domains (Bloom, 1998), making vocabulary a potentially useful measure of children's gender knowledge. In the context of a longitudinal study, the mothers of 50 children (1/2 girls) completed MacArthur Communicative Development Inventories (MCDI). At 18 months, measures were obtained of word comprehension and production as well as production of a number of communicative gestures, some of which are consistent with culturally prescribed gender-roles. At 24, 30 and 36 months of age, measures of productive vocabulary are available. An independent sample of twenty-four adults (1/2 women) rated the vocabulary words appearing on the MCDI as associated with feminine or masculine gender concepts, or as gender-neutral. These data were used to generate lists of feminine and masculine gender-typed words (see Table 1).Consistent with other reports of gender-based schematic processing in 2- to 3-year-olds, at the 24-, 30-, and 36-month assessments, boys' vocabularies contained significantly more same-gender than other-gender words. The difference in girls' vocabularies was in the same direction although it was not statistically significant. However, when the variance due to the large individual differences in overall vocabulary size was controlled, the pattern of greater same-gender than other-gender vocabularies was found to be significant for both boys and girls. Thus, using a gender-typed productive vocabulary measure with 2- and 3-year-olds yields results consistent with research utilizing other techniques at this age. Use of a vocabulary measure also enables us to study gender-based schematic processing in children younger than two years of age. Because very young children comprehend many more words than they produce (Fenson et al., 1994), comprehension is a developmentally appropriate gauge of language abilities at this age (Bloom, 1998). Examination of the children's receptive vocabularies at 18 months, revealed that comprehension of masculine-typed words did not differ by sex. However, boys comprehended fewer feminine-typed words than did girls. Moreover, 18-month-old boys also produced a smaller number of seemingly feminine-typed pretend play behaviors (e.g., putting a doll to bed, feeding a doll with a bottle) than did girls. Performance of feminine-typed play behaviors at 18 months is related to girls' and boys' production of feminine-typed words at 24 months. Thus, over the 2nd and 3rd years we find continuity in children's gender-knowledge expressed early in development through the gestures they produce and later through the words that they say.This research has provided evidence of gender-based schematic processing using developmentally appropriate measures of children's vocabularies and communicative behaviors. During the 3rd year of life, productive vocabulary for gender-typed words mirrors the effects of gender-based schematic processing found using laboratory procedures. Extending the investigation down to the 2nd year, we find interesting new effects of gender-based schematic processing expressed through developmentally earlier forms of communication.
paper
People use their gender knowledge and attitudes (i.e., gender schemas) when interacting with their environments. For example, gender schemas can affect people's memory of gender-related material. When remembering gender-stereotyped information, children and adults show differential memory (DM). That is, they remember own-gender information better than other-gender information. Several studies have found that verbal children show DM, using measures requiring fairly advanced verbal skills, which are difficult for children under three years. Only one previous study has shown evidence of such processing in children with less advanced verbal skills (Bauer, 1993). Using an age-appropriate measure (elicited imitation; EI), this study found that at 25 months, boys showed lower levels of recall for feminine events than for masculine and gender-neutral events. In contrast, girls performed at equivalent levels on all three types of events. The reason for this gender difference in younger children is unclear. The purpose of these studies is to identify the factors that are related to this asymmetrical pattern.Two possible explanations were examined. First, younger boys could have better stereotype knowledge than girls of the same age. Second, gender could be more salient to boys than it is to girls. Explicitly labeling the stereotyped nature of events can also affect children's DM. Labeling feminine and masculine events as 'for girls' and 'for boys,' respectively, equates knowledge of the events' stereotyped nature. In addition, labeling increases the salience of gender by emphasizing the stereotype. For example, preschool-aged boys showed DM for unlabeled gender-related scripts, but girls did not (Boston & Levy, 1991). However, in another study, when objects were verbally labeled as gender-stereotyped, both preschool-aged girls and boys showed DM (Bradbard & Endsley, 1983). To determine what factors are related to infants' DM, infants participated in a memory task (EI) and several gender-related cognition tasks: group membership of self and others, receptive labeling of self and others, stereotype knowledge, and gender salience. The correlations between infants' performance on the gender-related cognition tasks and on the memory task were calculated to determine which tasks were related to children's DM. Labeling the events by gender was also manipulated to determine its effect on younger girls' and boys' DM. In Experiment 1, 24- and 30-month-old children were presented with feminine, masculine, and gender-neutral events. For half of the children, the stereotyped nature of the events was verbally labeled (e.g., 'for girls'). For the others, the events were unlabeled. Boys showed DM for masculine events in both conditions. At 30 months, their masculine stereotype knowledge and their salience levels were related to their DM. At 24 months, no gender-related task was related to their DM. As a group, girls showed no evidence of DM. On an individual level, their knowledge levels predicted their DM. In the memory task in Experiment 2, gender-neutral tasks were labeled as being feminine and masculine. Although preschool-aged girls and boys showed evidence of DM on a similar task in previous research, 30-month-old girls and boys did not. Possible explanations for the gender difference and implications for gender schema theory will be discussed. A portion of Experiment 1 was presented at the International Conference on Infant Studies, 1998.