Introduction

Identity theory posits that children learn what is expected of them by identifying with role performances exhibited by agents of socialization, including media (Hogg et al. 1995; Stets and Burke 2000; Stryker and Burke 2000). As will be discussed later, consistently seeing mothers in the nurturing and care-giving roles and fathers fulfilling the provider role may impress upon children what role performances are ultimately expected of them as men and women. This study examines parental role performance in children’s picture books published in the United States. All cited studies are based on U.S. samples unless otherwise stated. The purpose of the study is to assess whether parental role portrayals in picture books published between 1902 and 2000 have evolved toward greater egalitarianism. Over time is there an increase in father characters fulfilling traditional, gendered-feminine roles, such as nurturing, providing care to, and interacting as companions with their children? Similarly, are mother characters more likely to work outside the home in books published more recently?

This study builds on the findings of Anderson and Hamilton’s (2005) examination of parental gender-stereotyping in children’s books published between 1995 and 2001 and Adams et al. (2011) similar examination of parenting roles in best-selling picture books published in 2008 in the United Kingdom. Both studies found that mothers took on more domestic roles and fathers were frequently absent. However, their findings differed on the roles of present fathers. Anderson and Hamilton showed fathers were largely ineffectual in parenting, and Adams, et al. found various nurturing characteristics of fathers were present, indicating possible role evolvement. Both studies sampled books of recent popularity and did not examine time of publication as a variable.

For the purpose of this study, indicators of five separate roles (i.e. nurturer, caregiver, companion, provider and disciplinarian) were conceptualized and examined to determine if fathers are represented in egalitarian roles as co-parents with mothers (Doherty et al. 1998; Pleck and Pleck 1997). If a parental character fulfilled any of the eight observable nurturing parental behaviors, physically expressing affection, verbally expressing affection, verbally encouraging, comforting emotionally, inquiring about thoughts and feelings, praising for a completed activity, listening to problems, and teaching toward a child it was noted. The five disciplining behaviors included hitting as punishment, hitting to prevent harm, giving non-physical punishments, verbally scolding, and correcting bad behavior with a non-threatening tone. Care-giving behaviors included cooking and feeding, bathing, dressing, and an “other” category which mostly consisted of putting the child to bed. Companion behaviors included taking the child on a recreational outing, nonphysical play, and physical play. The sole providing behavior was working outside of the home. The frequency of every parental behavior associated with the five roles was initially assessed. Since parental role performances are typically limited in picture books, if a character performed one or more of the behaviors associated with a certain role than the parent would be fulfilling the role. Additional variables assessed whether parental characters behaved nontraditionally or in traditional, gender-specific ways. This was achieved by combining acts of providing, disciplining, and participating in physical play (considered traditional behaviors for fathers and non-traditional for mothers) and combining acts of nurturing, care-giving, and participating in non-physical play (considered traditional behaviors for mothers and non-traditional for fathers). To assess whether these five role portrayals and “traditional” role performances have evolved, time of publication was the independent variable.

If children’s picture books reflect societal norms of the time, U.S. books published prior to the 1960s, when the breadwinner-homemaker model was in its heyday (Bose 1987), will portray women as care givers and nurturers and men as providers. However, in subsequent decades when U.S. women have competed with men in the public sphere and men have been encouraged to participate more in the private sphere as active, involved fathers (Pleck and Pleck 1997), role portrayals are expected to become more egalitarian.

Favorite works of children’s literature have national and international appeal. These texts are translated into many languages, and the roles portrayed gain wide-reaching, global exposure (Joels 1999). Thus, scholars of gender studies worldwide should find this study of interest. Media imagery of “ideal” families that continues to promote gender-specific roles fosters gender-specific identity development and subsequent role performance, and, in turn, slows movement toward global gender equality.

Identity, Social Identity and the Role of the Media

Identity theory posits children learn what is expected of them by primarily identifying with role performances exhibited by family members, peers, media, etc. (Hogg et al. 1995; Stets and Burke 2000; Stryker and Burke 2000). For example, a young boy learns his role as a father by identifying with the performance of his own father and/or by identifying with the representations of fathers in television shows, films and books. Furthermore, parents who purchase picture books and read them to their children may reaffirm their own roles while they are seen by their children as “authorities” on acceptable role performances.

Identity theory is derived from symbolic interactionism and examines how attitudes and identities affect behavior and roles (Stryker 1968; Stryker and Burke 2000). Symbolic interactionism has been used to analyze society as well as address issues of socialization (Stryker 1980, p. 1). Identity theory is principally microsociological in that, like symbolic interactionism, it examines how internalized roles, or identities, affect an individual’s motivations and behavior (Hogg et al. 1995). According to identity theory, the self is developed in the activity of viewing oneself reflexively. “This activity is made possible by language, a system of significant symbols; language permits use of the standpoint of others in order to view oneself as an object” (Stryker 1980, p. 37). This process of classifying oneself reflexively is known as identification (McCall and Simmons 1978; Stets and Burke 2000). A young girl who reads of a mother who balances responsibilities at home with a career, for example, may interpret the language and symbols on the page, and internalize her own future role as one filled with the opportunity to pursue both options.

While identity theory is principally microsociological, social identity theory more closely aligns with social psychology and is concerned with “intergroup relations, group processes, and the social self” (Hogg et al. 1995, p.259). The basic premise is that an individual’s social categories (e.g., mother, Iranian, Republican) define an individual’s self concept and prescribe “what one should think and feel, and how one should behave” (p. 260). Both identity theory and social identity theory examine how self-concepts are formed, and the behaviors that evolve from these adopted identities. A reader of a picture book, therefore, may gain an understanding via language, attitudes, and performances, of her own expected behavior (as identity theory posits), or she may identify with a social category such as “mother” and define herself based on the actions of the mother characters (as social identity theory posits). Both theories indicate the potential power picture storybooks have in identity acquisition.

Likewise, symbolic interactionism asserts that an individual’s assigned social status positions have attendant behavioral expectations, which are labeled roles (LaRossa and Reitzes 1993). According to symbolic interactionism, the self is both social and active (Cooley 1902; Mead 1934). The “self” is fostered via interactions and, ultimately, “taking the role of the other” (Hogg et al. 1995, p. 256). Identity theory examines “self-defining” roles and not strictly the vast range of “attributes that can be ascribed to the self” (p. 257). Symbolic interactionists assume the self is comprised of many identities, each emerging from the interaction between individuals performing complementary statuses by carrying out roles (e.g., the relationships of mother-child, husband-wife). Thus, reading about the relationship between a parent and a child, and identifying expected behaviors of both parties, can teach a child how to behave as a child and how he is expected to behave once he becomes an adult. A father who reads such a book to his child can also gain affirmation of the expected behaviors of his own status as father. It can be argued that if books continue to promote the traditional roles of mother as homemaker and father as breadwinner, these will continue to be the statuses of greatest identity salience for the authors, publishers and ultimately the consumers.

Marsiglio et al. (2000) address the importance of determining empirically the institutional and interpersonal construction of statuses and roles. They encourage researchers to explore the media’s construction of roles, among other avenues (p. 287). Books, like other media, endorse and reflect cultural expectations. If mothers and fathers are portrayed as observing a strict, gender-driven division of labor in children’s literature, then the potential effect on early socialization must be acknowledged.

Is it possible to determine long-term effects on a child who consistently reads about characters acting out gender-stereotypic behavior? A few studies have attempted to test the immediate effects of gender roles portrayed in children’s literature. For example, Ashton (1983) examined the effects of roles on play behavior in the Human Development Laboratory School at the University of Massachusetts. Thirty-two children were allowed to play freely with female-stereotypic toys (i.e. dolls, china sets), male-stereotypic toys (i.e. trucks, guns), and neutral toys (i.e., balls, pegboards). Thereafter, they were each individually shown and read a picture book with a same-gender character playing with either a gender-role stereotypic or non-stereotypic toy and allowed to return to play. Though the girls were more influenced, both boys and girls played significantly longer with the toy indicated in the picture book. Ashton concluded that

…children’s literature has a pronounced effect on sex-role behaviors of the young child. Clearly, the picture books which children read should illustrate both male and female characters in a widened range of behaviors where conduct and assignment are not limited by stereotypic definitions of sex role (p. 46).

Jennings (1975) studied picture book preferences of 64 preschool children in Columbia, Missouri. The children were divided into same-gender small groups and read two stories about a character of their own gender wanting to be a ballet dancer in one story and a mail carrier in the other. While both boys and girls stated a preference for the books that portrayed their own gender stereotypically (i.e. boys preferred the male postal carrier story and girls preferred the female ballerina story), both boys and girls had better recall of details in the other story that showed the character in the reversed gender role.

Barclay’s (1974) study examined the effects of U.S. “cultural shaping” by reading and discussing three books and an informational pamphlet all focusing on women employed outside the home to 64 kindergarteners. It was discovered that “only three fifteen-minute lessons dealing with women’s careers …affect[ed] kindergarten children’s, particularly girls’, perceptions of women’s career roles, as reflected in increased numbers of choices of women as appropriate for career roles” (p. 13).

Picture books are clearly influential on the socialization of young children. A study by Alexander et al. (2001) of 32 European-American families found that children’s emotional attachments to books can be so strong that they may sleep with the books, act out roles, and request repetitive readings.

Through the expressions of their attachments they are also deeply engaged in the social relations found in the stories themselves. The children respond in delight and fear to the evolving relationships of the story characters. They try on the various character roles, taking as their own the relationships of those characters. Finally, by transporting familiar characters into new situations and different worlds, the children test and forge these relationships in a multitude of social spaces. (p. 392).

If imagery in children’s books has the potential to impact identity acquisition and role performance, are portrayals fostering or inhibiting egalitarianism in larger society?

Evolution of Parental Role Expectations

It is predicted that the roles portrayed in picture books will somewhat reflect role expectations of the larger society at the time of publication. Therefore, the historical evolution of U.S. parental role expectations is summarized, beginning at the turn of the 20th century when the oldest sampled text, Beatrix Potter’s The Tale of Peter Rabbit (1902), was published.

Bose (1987) examined the origins of the dual spheres in U.S. history. Beginning in the late 1800s the dual spheres of men in the labor force and women in the home emerged as it became difficult for families to choose agriculture over wage labor. At the turn of the century leading up to World War I a slight rise in women’s employment occurred, but the fear of job competition with men, which intensified during the Great Depression, sent many employed women back to their homes. As most took on domestic odd jobs (e.g., laundry, sewing) their expected role within the home was confirmed. World War II created a boost in female employment, but when soldiers returned to the factories, most women returned to their homes. Mothers in picture books published prior to the 1950s are expected, therefore, to fulfill the domestic role of nurturer and care giver and fathers the role of provider, if as Bose (1987) contends, the dual spheres were well established.

Around this time, Parsons (1943) posited that modern industrialized society was best served by a gender-based division of labor. This division of labor assigned the economic provider role to the father to meet the instrumental needs of the family and the emotional role to the non-employed, homemaker mother to meet the expressive needs of the family (Parsons and Bales 1955). Thus, the 1950s ideal version of the father continued to be seen as the “good provider” who “provided a decent home, paid the mortgage, bought the shoes, and kept his children warmly clothed” while mothers were inundated with television shows and advertisements hailing their ideal role as homemaker (Bernard 1981 p. 3–4). It can be expected that children’s literature published prior to 1960 will reflect these traditional parental roles.

Research in the 1960s reported that in U.S. working-class families the provider role was reserved for the father and the caregiving role was to be adopted by the mother (Aldous 1969), and mothers who were employed outside of the home viewed employment as a supplementary and not a primary role (Hartley 1969). Thus, U.S. picture books of the 1960s should show an increase in mothers employed outside the home, but the gendered division of labor should be maintained.

In the 1970s, research showed that U.S. mothers engaged more in feeding and caretaking activities (Kotelchuck 1976; Parke and O’Leary 1976) but fathers spent a large percentage of time in recreational play as part of the companion role (Kotelchuck 1976; Lamb 1977). Thus, if media are a reflection of society, traditional parental roles portrayed in books published in the 1970s will continue to dominate, but a slight increase of women in the provider role and men in the companion role should be noted.

More recent research maintains that a man’s status as a father, involving adopting the roles of caregiver, companion, provider, nurturer and disciplinarian has increased in identity salience for men over time (Ihinger-Tallman et al. 1995, p.560). Hence, it might be expected that this evolution is reflected in children’s books. We may also expect the shift in mothers’ status as they have increasingly entered the labor force to be represented in children’s literature.

However, some researchers suggest that although women in the United States have increasingly entered the workforce since World War II, they have been unable to escape their traditional roles involving homemaking and childcare (Bose 1987; Hochschild 1989). Pleck and Pleck (1997, p. 47) assert that since the 1970s the emergence of the “co-parent father” was a significant development in the United States, encouraging men to be more active caregivers. At the same time, however, discourse about “deadbeat dads” encouraged women to adopt more of a provider role. Regardless, parental roles have experienced an evolution in society over the past several decades. Whether children’s literature will support traditional roles or reflect evolution is under investigation in this study.

Gender studies on children’s literature suggest the persistence of stereotypical roles. Several of these studies have used award-winning texts as their samples, including recipients of the Caldecott Medal (for distinguished American picture book), Newbery Medal (for distinguished American literature for children), and Coretta Scott King Award (to outstanding African American authors and illustrators of books that represent African American culture). One study of 30 Caldecott-awarded texts published between 1984 and 1994 found that while women are moving beyond former stereotypes in society, this evolvement has not been reflected in such texts (Turner-Bowker 1996). Both Gooden and Gooden’s (2001) study of 83 picture books published between 1995 and 1999 and designated “notable” by the American Library Association and also Hillman’s (1974) study of children’s books selected from a list of book reviews across two time periods (i.e., 1930s versus mid-1960s and mid-1970s) found that males in children’s literature have access to a broader range of occupations and are rarely seen in housework roles. Crabb and Bielawski (1994) examined illustrations in picture books that had won the Caldecott Medal or Honor between 1938 and 1989. Proportionally more women than men were depicted employing household artifacts used in cleaning, cooking, etc. and more men than women were depicted utilizing production artifacts used in outside occupations such as construction and agriculture. Crabb and Marciano (2011) followed up with a study on Caldecott award-winning texts from 1990 to 2009 and found little had changed. As predicted, females were still more likely to use household artifacts and males were more likely to use production artifacts.

A content analysis of Little Golden Books, etiquette books, and Caldecott and Newbery award-winning books by Weitzman et al. (1972) found that male characters greatly outnumber female characters and females are most often portrayed in stereotypical and traditional gender roles. They are passive, usually shown indoors, and do not work outside of the home. Kolbe and LaVoie (1981) examined Caldecott Medal winners (1972–1979) since the publication of the Weitzman et al. study and found little had changed. Female characters were still underrepresented (though the ratios were less extreme) and were still more likely to be portrayed in expressive, insignificant, stereotyped and traditional roles. Male characters were more likely to be portrayed in instrumental and significant roles, and their roles were also stereotypical. Clark et al. (1993) examined gendered images in Caldecott and Coretta Scott King Award-winning books and found that behavioral traits of men and women have slowly become more egalitarian. Allen et al. (1993) compared Caldecott Medal-winning picture books from two time periods: 1938–1940 versus 1986–1988. They found minimal change to the gender-role stereotyping and noted some role regression. For example, all female characters in the more recent time period were in traditional roles (versus 69 % in the earlier works) and females experienced a decline in the variety of occupations (23 to 20). Portrayals of male characters showed some progression as the number of traditional roles declined. Nonetheless, males were more likely to be active, central characters in a wider array of occupations.

Jackson and Gee (2005) analyzed parental role portrayals in popular texts used in New Zealand schools over the previous five decades to teach students to read. They found little has changed over the past five decades. “Across all decades mothers were positioned as nurturers, providing care, food and supplies for their children and husbands” (p. 123–124). Fathers were rarely shown in domestic labor, and when interacting with children it was most often in outdoor activities. The evolution in these texts was not a change in roles, but a diminishing of roles as parents appeared less often in recent editions.

While research using relatively selective samples of children’s books overwhelmingly shows that traditional gender roles continue to dominate, this study examines gendered parental role evolvement over time using a broader sample of texts.

Five Parental Role Categories

Five role categories tend to be used in studies on parental role performance. These are parent as companion, parent as disciplinarian, parent as caregiver, parent as nurturer, and parent as provider (Doherty et al. 1998; Ihinger-Tallman et al. 1995). Most researchers have argued that society assigns the role of nurturer to mothers (Chodorow 1978; Simon 1995; Thurer 1994). For example, Anderson and Hamilton (2005) examined parental nurturing behaviors as portrayed in both Caldecott-winning and best-selling children’s picture books published between 1995 and 2001 and found mothers were 10 times more likely than fathers to nurture babies and twice as likely to nurture older children. Their follow up study found that females portrayed in picture books were over three times more likely to provide nurturance and care than male counterparts (Hamilton et al. 2006). Adams et al. (2011) examined picture books in the United Kingdom and found that fathers were portrayed as more emotionally detached and were “significantly less likely to be depicted in physical contact with their children” (p. 264).

Other research suggests that fathers act more as companions with their children, involving themselves more in physical play, than they do in being nurturers (Marsiglio 1991; Minton and Palsey 1996). According to identity theorists, if the mother were to identify herself as the child’s nurturer, and the father were to identify himself as the child’s companion these are the roles they would most likely play. If “mothers’ interactions with their children are dominated by care-taking whereas fathers are behaviorally defined as playmates” (Lamb 1987 p. 10), will parental portrayals in children’s literature support these perceptions? Some research suggests both parents act as companions, but the type of play in which mothers and fathers participate differs. Lamb (1977, 1997) asserts fathers participate in more physically stimulating play whereas mothers may be more likely to participate in non-physical play.

The disciplinarian role is based on incidences of a parent correcting a child’s behavior. This may include scolding, physical punishment, and non-physical punishment. The caregiver role, involving feeding, dressing, and cleaning the child, has historically been an expected role for mothers. Depictions of fathers in care-giving roles have provided comic relief in various media, including movies and television shows. LaRossa et al. (2000) found evidence that fathers in 490 Father’s Day and Mother’s Day comic strips published between 1940 and 1999 were often shown as incompetent in their care-giving responsibilities.

The provider role is usually measured by whether the parent works for wages or exchange value outside of the home. Do fathers continue to be portrayed as breadwinners, or have the themes of children’s literature adapted to account for the increasing number of mothers in the labor force? A study of early Caldecott Award-winning books found that not one female adult character had a job or profession other than that of mother and homemaker (Weitzman et al. 1972). Hamilton et al. (2006) found the likelihood of female characters to work outside the home remains minimal and their range of occupations when compared to males remains limited.

In the current study of parental portrayals in children’s picture books over time, performance of these five roles by mother and father characters was examined. Based on findings from the extensive review of the literature on parental role performance, the mother’s role as primary nurturer, companion, and caregiver was not expected to change over time. However, some evolvement toward egalitarianism, even if minimal, was expected. Therefore, the following hypotheses were developed for testing:

  1. 1.

    The likelihood of fathers performing the roles of nurturer, companion, and care-giver would increase.

Western culture is embracing greater egalitarianism so it is expected that fathers in recent texts will be more likely to participate in once gendered-feminine parenting behaviors than fathers of earlier texts.

  1. 2.

    The likelihood of mothers acting as providers would increase.

As more U.S. women entered the labor force, particularly after the 1970s, it is expected that an increasing number of women in more recent texts will act as providers.

  1. 3.

    The likelihood of mothers acting as disciplinarian would also increase.

While discipline may traditionally be considered gender-masculine, there is more mother-child interaction versus father-child interaction in children’s books. Since extreme forms of discipline are virtually nonexistent in children’s books, particularly in picture books with more limited storylines, any scolding and/or correcting would likely be done by the present parent, which, as indicated, is likely to be the mother. This tendency may affect the trends observed.

  1. 4.

    Both the mother and the father roles were expected to behave less traditionally over time.

  2. 5.

    Both the mother and the father roles were expected to behave more nontraditionally over time.

If family dynamics portrayed in picture books are a reflection of norms and values in Western cultures, then evolvement toward greater egalitarianism and the challenging of traditional gendered expectations should be noted.

Method

Sample

The population of the study was all children’s picture books published between 1900 and 2000. The sampling frame consisted of the more than 1,400 “easy” children’s books listed in the Children’s Catalog (H.W. Wilson Co., 2001). The books listed in the Children’s Catalog were all selected by an advisory committee of distinguished librarians and then re-evaluated by librarians around the United States. This comprehensive list of books is used to aid school and community libraries in selecting quality books for collection maintenance. As such, all of the randomly sampled texts used in this study were easily found in local public libraries. The population of “easy books” (N = 1,448) from which the sample was obtained includes fiction and nonfiction picture books and short stories all suitable for children at a preschool to third-grade reading level.

Books categorized as counting (n = 52), vocabulary (n = 28), and alphabet (n = 41) were eliminated as they would lack narratives for analysis. Informational books (e.g., books on anatomy, dinosaurs, seasons of the year) were eliminated for the same reason (n = 57). Books of collections (e.g., short stories, puzzles, songs, nursery rhymes, and prayers) were eliminated for the sheer number of characters they could potentially introduce (n = 9). Finally, adaptations (n = 9) were removed because it was unclear whether they should be analyzed based on their most recent year of publication or on the year when the original story was first published. As a result, 196 books were removed from consideration. Therefore, the sample of books was selected from the remaining 1,252 storybooks categorized as fiction, stories in rhyme, fairy tales, fables and tall tales (See Appendix Table 5).

Due to the larger population of books in some time periods and the limited number of books in others, the list of the remaining 1,252 books was stratified based on the time periods under investigation before a computer-generated random sample of 300 books was selected. The first 50 books were randomly selected from those books first published between 1900 and 1959. The collapsing of the first five decades into one stratum was done for two reasons: the first is the prediction that traditional parental roles will be depicted in books prior to the 1960s, and the second is the smaller number of children’s books that were published in earlier decades. An additional 50 books were then randomly selected from each of the four subsequent decades (1960–69; 1970–79; 1980–89; 1990–1999), and a final 50 books were selected from those published in the year 2000. Thus, all books in the population did not have the same probability of being selected. For example, if 50 books were chosen from the 63 books published in the 1960s and another 50 were chosen from the 637 books published in the 1990s, clearly a book published in the 1960s had a far greater chance of being selected. If relationships within the sample are to reflect those in the population, the disproportionate probabilities of books being selected into the sample and the stratifying of the sample must be considered in the analyses. Therefore, analyses conducted to test the study hypotheses used procedures that take into account complex sample design. Forty-nine of the 300 books sampled lacked a child character and were, consequently, removed from the analyses. This reduced the unweighted sample size to 251 and the weighted sample size to 1072.

The Survey Instrument

A survey instrument was designed to use in gathering information from both the illustrations and the narrative text in each of the 300 books on the five parental roles of companion, disciplinarian, caregiver, nurturer and provider. One of the challenges in constructing the instrument was determining which observed actions in the texts were evidence of each of the five parental roles examined. Defining the five roles (the dependent variables) and delineating their attendant behaviors was essential before data collection could begin.

The nurturer role was defined by adapting the variables used in LaRossa et al. (2000) analysis of fathering portrayals in comic strips. They used Coltrane and Allan’s (1994) nurturing behaviors of physical and verbal expressions of affection, verbal encouragement, emotional comforting, inquiries about thoughts and feelings, and service or care for a child and added three of their own (i.e., praise for a completed task or activity, listening to problems, and direct teaching). For this current study, the behavior “serving or caring for a child” was removed as that is included in the caregiver analysis. In addition, LaRossa et al. (2000) examined “physical and verbal expressions of affection” as one variable, but physical expressions and verbal expressions of affection were considered separately for the purpose of this study.

The disciplinarian role was assumed by any parent who physically punished, hit to prevent the child’s harm, punished non-physically (e.g., sending to room), scolded, or corrected bad behavior. The parental role of caregiver is comprised of actions toward the child related to hygiene (i.e. bathing, cleaning, changing diapers), personal attire (i.e., picking out clothes, dressing), and sustenance (i.e., preparing meals, feeding). The role of companion is comprised of three observational variables: taking a child on a recreational outing, physical play, and nonphysical play. Finally, the provider role is assumed by any parental character who works outside of the home.

Some of the behaviors associated with the roles were more readily observed in illustrations (e.g., “physically express affection for the child” and “prepare meals for and/or feed the child”) and some were more readily found within the written text (e.g., “verbally express affection for the child” and “verbally scold the child”). Any time a behavior was performed, it was noted.

To test for reliability of the researcher’s designated parenting attributes, 38 student volunteers from a junior-level university course entitled Sociology of Marriage and Family were enlisted to receive training and complete the survey instrument for a subset of 84 books in the sample. Training involved both reading every question out loud for clarity and explaining meaning. Each of the students completed a survey instrument on up to three different books in the subset. Consistency between the responses generated by the 84 student surveys and the original researcher’s responses for the subset of 84 books was assessed by an inter-rater reliability analysis using Cohen’s Kappa statistic for each of the behavior variables for both the mother and father character (See Appendix Table 6). Kappa measured near-perfect or perfect agreement on each of the mother’s observed disciplining behaviors (Kappa = .97–1.0), nurturing behaviors (.94–1.0), care-giving behaviors (Kappa = 1.0), companion behaviors (Kappa = .976–.979), and providing behavior (Kappa = .977). Perfect agreement between raters (Kappa = 1.0) was found across all of the father’s disciplining, nurturing, providing and companion behaviors and near-perfect agreement was achieved for observed care-giving behaviors (Kappa = .972–1.0). Part of the strong agreement can be explained by the fact that only 25 fathers and 42 mothers were found in the subset of 84 books included in the inter-rater reliability analyses, and while the mothers’ roles were often brief, the fathers were largely inactive. Nonetheless, all of the 84 student-generated survey instruments were above 90 % agreement with the original researcher-generated data, representing high reproducibility (Krippendorff 2004). Any discrepancies required rereading the text in dispute and mutually deciding upon the most appropriate response.

Data Coding and Analysis

Table 1 delineates the attendant behaviors gathered via the survey instrument for each of the five role performances for the mother and father characters in the sampled books. Approximately 70.5 % (177) of the 251 books with a child character had at least one mother character. Only about one-half (50.9 % or 128) had at least one father character. Only books with at least one parental character are represented in Table 1, which also gives the percentages of the books whose parental characters perform each of the various behaviors. Because parental characters in a book could perform one, multiple, or no behaviors within each role, the percentage totals within roles in the table do not equal 100 %.

Table 1 Percentages of nurturing, disciplining, care-giving, and companion behaviors toward children for mothers and fathers presented in sampled books
Full size table

To test the hypotheses, the behaviors in Table 1 were recoded to construct the dependent variables. For each of the five parental roles, a variable was constructed for each book, which had three attributes: (1) mother character in the book fulfilled any of the attendant behaviors associated with the role, (2) the mother character fulfilled none of the attendant behaviors associated with the role, or (3) no mother character was in the book. A similar set of five variables was constructed with reference to father characters in the books. Thus, a parental character was considered a “caregiver,” for example, whether she or he fulfilled only one or multiple behaviors associated with that role. For those books in the sample with at least one mother character, the first five rows of Table 2 present the percentages in which mother character(s) did (column 1) or did not (column 2) demonstrate behaviors associated with a given role. Corresponding percentages for books with at least one father character are provided in columns 3 and 4, respectively.

Table 2 Percentages of role performance by mothers and fathers presented in sampled books
Full size table

Further, to assess traditional and non-traditional behaviors, four additional variables based on a combination of role performances were constructed (see last two rows in Table 2). For these variables, any acts of providing, disciplining a child, and participating in physical play with a child (i.e., an individual behavior associated with the companion role) were considered traditional behaviors for fathers and non-traditional for mothers. Any acts of nurturing, care-giving, and participating in non-physical play were considered traditional behaviors for mothers and non-traditional for fathers.

Cross-tabulation analysis of each of the 14 dependent variables with the independent variable time of publication was used to test the study hypotheses (see Tables 3 and 4). Unlike previous tables all books in the final sample (unweighted n = 251, weighted N = 1,072) were included. The complex survey design required the conversion of the Pearson chi-squared test statistic, using Rao and Scott’s (1984) second-order correction, into a design-based F test statistic.

Table 3 Summary of cross-tabulations of mothers’ behaviors and presence by time period of publication of books (Weighted N = 1,072)
Full size table
Table 4 Summary of cross-tabulations of fathers’ behaviors and presence by time period of publication of books with child characters (Weighted N = 1,072)
Full size table

Results

As Table 1 shows, mothers in the books were more likely than fathers to perform almost every nurturing behavior, including verbal and physical expressions of love, encouraging, praising and listening. A slightly greater percentage of fathers comforted (11.7 %, n = 15 versus 11.3 %, n = 20) and purposefully taught (9.4 %, n = 12 versus 8.5 %, n = 15) the child, but it is fairly evident that, in general, the role of nurturer is gender-determined, especially when emotional expression is involved. Indeed, as Table 2 shows, almost 51 % (n = 90) of the mothers present in the stories demonstrated at least one of the nurturing behaviors, whereas about 42 % (n = 54) of present fathers demonstrated some form of nurturance.

Across every care-giving behavior, mothers outperformed fathers (Table 1). Table 2 shows that 35.6 % (n = 63) of mothers versus 10.9 % (n = 14) of fathers demonstrated any care-giving behavior. Thus, mothers more so than fathers are depicted cooking, feeding, cleaning and dressing children. Disciplining behaviors, while also more likely to be performed by mothers, are more equally distributed.

Behaviors associated with both the companion and provider role appear to be gender-masculine. Fathers were only slightly more likely to take a child on a recreational outing, but much more likely than mothers to participate in both physical and nonphysical play (Table 1). Table 2 illustrates that 32.0 % (n = 41) of fathers versus 24.9 % (n = 44) of mothers participated in any companion behavior. Fathers were far more likely to be portrayed in the provider role, as 26.6 % (n = 34) of fathers versus 5.6 % (n = 10) of mothers worked outside the home.

Mothers (65.0 %, n = 115) were more likely than fathers (39.1 %, n = 50) to perform at least one traditional, gender-stereotyped behavior (Table 2). Fathers (48.4 %, n = 62) were more likely than mothers (23.7 %, n = 42) to perform at least one non-traditional, gender-stereotyped behavior. Mother characters are less likely, therefore, to be portrayed outside of their traditional domestic roles. It appears that father characters, while frequently inactive when present in these texts, are more likely to nurture than mothers are to provide. Thus initial frequencies suggested that gender-stereotyped parental roles dominate children’s picture books.

The cross tabulations revealed, as expected, that the mothers’ roles as nurturer, caregiver, and companion over time did not evolve (Table 3). The lack of statistical significance suggests these domestic roles remain gendered-feminine. The testing of hypotheses also indicated a lack of movement toward egalitarianism.

Hypothesis 1

The first hypothesis, that the likelihood of fathers performing the roles of nurturer, companion, and care-giver would increase, was not supported. The results of the cross-tabulation analyses of each of these three father roles by time of publication are summarized in Table 4. The associated design-based F statistics reveal that none of these cross-tabulation analyses is statistically significant. While fathers’ portrayals as care-giving, nurturing, companions in children’s picture books peaked in the 1970s (10.8 %, 32.5 %, and 35.2 % respectively), these portrayals had essentially leveled off in later decades.

Hypothesis 2

The second hypothesis, that the likelihood of mothers acting as a provider would increase, also lacked support. While the largest percentage of employed mothers (i.e. 7.9 %) was in the year 2000, nearly double the percentage pre-1960 (i.e., 4.4 %), this trend toward egalitarianism was not statistically significant (Table 3).

Hypothesis 3

The third hypothesis, that the likelihood of mothers acting as disciplinarians would increase, was not supported. While mother characters were unexpectedly more likely to discipline than father characters in every time period except both the 1970s, when they were equally likely to discipline (i.e., 10.8 % of both mothers and fathers), and also 2000, when the percentage of mothers disciplining sharply dropped (i.e., 2.6 %), mother characters did not increasingly discipline.

Hypothesis 4

The hypothesis that both parental roles were expected to behave less traditionally was not supported. At first glance it appears that mother and father characters post 1969 are portrayed more traditionally than mother and father characters in books published before 1970. However, both of these trends lack statistical significance.

Hypothesis 5

The hypothesis that both parental roles were expected to behave more nontraditionally was also not supported. Mothers and fathers were both most likely to behave in nontraditional roles in books published in the 1970s than any other time period; however, this too lacked significance.

Discussion

While opportunities in the public sphere have increased for U.S. women, and men are being encouraged to actively participate in the private sphere, children’s picture books embrace tradition. Mothers are much more likely to be portrayed nurturing and caring for children, and men are more likely to work outside of the home. These depictions have not significantly changed over time, so that these storybook characters often inhabit a bygone, male breadwinner-female homemaker era. Still, some evidence points to increased egalitarian portrayals during the last three decades of the Twentieth Century. Fathers in books published in 2000 exhibited increased care-giving and nurturing from previous time periods, and mothers exhibited increased work outside of the home. But the latest trends lack statistical significance because similar performance peaks occurred in the 1970s depictions only to drop in subsequent periods. These two peaks and the extended valley in between question whether the changing roles represented a substantial increase in egalitarianism. Many traditional gender distinctions remain. Identity theory posits that an individual develops a sense of self, including the roles and identities he or she will assume, by identifying with the language and symbols associated with the roles of others (Stryker 1980). If children, especially girls, continue to be exposed to portrayals that suggest opportunities for women are limited to the home and that men provide, their aspirations and independence will be muted.

The nurturer role is gender-feminine in children’s picture books. These findings lend support to research that suggests society assigns the role of nurturer to mothers (Chodorow 1978; Simon 1995; Thurer 1994). Behaviors associated with the nurturing role that involved intimate interaction such as physical and verbal expressions of affection, encouragement, praising, inquiring about feelings and listening were more likely performed by mothers. She is the socio-emotional leader of the family. The caregiver role is also gender-feminine within this sample of texts. All care-giving behaviors including feeding, cleaning, and dressing children were more often performed by mothers. Since neither the performance of the nurturer nor the caregiver role significantly changed over time for fathers or mothers, it seems that tradition is still being embraced. Whether the publishers, authors, or adult consumers are driving this adherence to gender-specific roles is unclear. However, why in Westernized society that values gender equality are not more people demanding egalitarian portrayals?

The provider role, on the other hand, is gender-masculine in these picture books. Among present parents, fathers were almost five times more likely to be employed (26.6 % versus 5.6 %). It must be noted that among “employed” mothers a couple outlandish occupations (i.e., the Easter bunny-mother; an artist-mother who painted Easter eggs) were classified as fulfilling the provider role. Thus, if only realistic occupations had been noted, perhaps the discrepancy would be even more dramatic. Employed mothers in children’s literature with occupations that depict greater independence and diversity would be stronger role models for young readers.

The companion role is also more likely assumed by father characters. While taking a child on a recreational outing is a fairly equally performed behavior by either parent, playing with the child in both physical and nonphysical ways was done most often by fathers. The greatest difference between mothers and fathers was in the area of physical play; about 9.4 % of fathers, compared to 4.5 % of mothers, participated in physical play. Studies show that fathers act more as companions in recreational play (Kotelchuck 1976; Lamb 1977; Marsiglio 1991; Minton and Palsey 1996). These gender-masculine companion role portrayals in children’s books lend support to Lamb’s assertions that fathers are often “behaviorally defined as playmates” (1987, p. 10).

The role of disciplinarian was shared, without statistically significant change, between mothers and fathers, though overall acts of any discipline were relatively rare. None of the parents in any of the books physically punished a child. A few child characters received non-physical punishments like being sent to their room without dinner, a punishment Max received for wearing a wolf suit and causing mischief in Maurice Sendak’s Where the Wild Things Are (1963).

Findings from this study add to the body of knowledge about trends in parental role portrayals and the identity theoretical framework through its application to the medium of a century of children’s picture books. If children develop their identities by adopting observed behaviors of others, then examining those portrayals that children are often exposed to is essential. Acknowledging parental role performances in children’s media, particularly if they are consistently absent, negative, stereotypical, or limited in scope and opportunity is warranted.

Interestingly, while female characters are underrepresented in children’s literature, mother characters greatly outnumber father characters. Parental portrayals, in general, are perhaps rarely of primary concern to authors of children’s picture books, as mother and fathers often take second stage to the plot, if even permitted to make an appearance in the text. Nonetheless, if those albeit brief performances consistently reinforce traditional, gender-specific roles, the potential effects on a young reader’s gender identity acquisition must be acknowledged.

This study was not without limitations. The data set of books obtained from the Children’s Catalog (H.W. Wilson Co., 2001) includes texts recommended by librarians and may not represent the most widely read, and therefore the most influential, children’s literature. Furthermore, acclaimed books may exclude many books with negative portrayals. Perhaps best-selling texts would prove more revealing, but since the sample used in this analysis was stratified by time of publication, and since best-selling children’s literature lists have been historically inconsistent in their creation, using solely bestsellers was not a possibility.

Almost one-sixth of the sample was immediately lost when 49 of the original 300 books were removed from the analysis because they lacked a child character. Furthermore, of the 251 texts that remained only 177 had mother characters and 128 had father characters. Perhaps a way of sampling only from books that included a parental character would have garnered greater evidence of role evolvement.

Explaining the purpose of this study to colleagues and friends often resulted in recommendation of picture books representing positive and diverse parental portrayals. These recommended books were invariably part of a relatively new crop of books targeting schools and early childhood educators. Books presenting various issues, including divorce, same-gender parents, adoption and occupational opportunities have become increasingly popular in preschool and elementary school classrooms. Sampling and evaluating parental roles in such books should garner different results.

The application of identity theory to interpreting the findings raises a number of concerns. These concerns suggest a number of recommendations directed toward the authors and publishers of these books and the consumers, especially parents, who purchase them. Identity theory posits people develop their identities through the observation and adopting of behaviors associated with the role performances of others. Hence authors and publishers of children’s literature need to be more sensitive to the portrayal of parents in their works, as these books act as a major agent of early socialization and can reinforce gender-specific behaviors even for the adults who read these books to children. If parents are consistently portrayed in stereotyped roles, then readers who identify with the mother or father role will be limited by what they consider their “expected” behaviors.

The sheer absence of parents in picture books also needs to be addressed. Children’s picture books have the potential to be powerful and positive influences on identity acquisition, but with the limited representation of parents, that potential is neglected. While children’s book publishing today is a business, and while publishers and authors may realize a giant, talking dog sells more than a strong father-son relationship, strong parental portrayals need to be woven into more plots.

Additionally, these parental portrayals need to reflect active and quality parenting. When present in the texts, too many parental characters are mere props in illustrations and are not actively involved in the child characters’ lives. This is represented in the number of present parents who did not act out a role performance (see Tables 3 and 4). If an individual develops his identity through the observation of language and behavior, based on many of these parental portrayals he may conclude a parent’s role is inactive and unimportant. Attentive parents of both genders, portrayed as caregivers, nurturers, and companions, need greater representation.

Furthermore, employed mothers in children’s literature with occupations that depict greater independence and diversity may be stronger role models for young readers. This is particularly important for the identity acquisition of young girls. Books rarely depict the vast options available to women in both the public and private spheres.

If parental role expectations in the United States have evolved from a somewhat strict dual sphere prior to the 1960s, to a limited entry into the public sphere for women with continued traditional private sphere roles in the 1960s and 1970s, to a progressive melding of spheres for mothers and fathers after the 1970s, then it might be expected that similar progress is reflected in children’s literature. This is not the case. Traditional roles that involve nurturance and care-giving from mothers and providing from fathers continue to dominate children’s literature. Since little significant variance was found between any of the mother and father role variables across time, traditional role performance remains the norm. Little boys learn that nurturing and caring for children are not masculine traits. And little girls continue to learn that the broom, not the scalpel, is a woman’s future tool for success.