Home Articles Reader Opinion Editorial Book Reviews Discussion Writers Guide About TCRecord
transparent 13

Little Boy Lost

by Sara Goodman Zimet - 1970

An examination was made of the sex role models portrayed in primary reading texts during six contiguous historical periods in the United States from 1600 to 1966. (Source: ERIC)

Since those who struggle in the swamps and deserts of reading research are usually regarded by their colleagues on education's higher ground as something less than full citizens, we offer this report on a critical issue in beginning reading instruction in the hope that we may learn from each other.

The purpose of this studyi was to examine the sex role models portrayed in primary reading texts during six contiguous historical periods in the United States from 1600 to 1966.

The reading textbook is the traditional means through which reading skills are taught. But more than that, it is the first text placed in the hands of the child entering first grade or possibly even school for the first time. In addition to having an instrumental purpose, that of teaching reading, its content is intended to be motivational. In other words, what is written is intended to fit the predispositions of the child so that he will want to learn to read. Thus the motivational factors identifiable within the stories relate to the motivational factors in the learning process itself.

It is assumed then that in reading a story a child symbolically goes through the action that is described and consequently his social attitudes will be influenced by the behavior patterns portrayed in the many stories he reads. The reading text, in this sense, is one means for acculturation and socialization, a way of interpreting to the reader the peculiar characteristics of his culture. The influence of the stories upon the child will depend, however, upon the ease with which he is able to identify with the characters carrying through the action.


The need for providing more appropriate models for boys and girls in contemporary readers has been recommended by Waite, et al.ii This study investigated the content of primers from twelve publishers' series along dimensions related to motivation in learning to read. The results indicated that: (1) the stories depicted activities that are characteristic of girls of preschool age more often than not; (2) activities preferred by boys ended in failure more frequently than activities preferred by girls; and (3) a significant number of stories depicted activities that boys and girls customarily engage in, obliterating a clear differentiation in sex roles.

Both historical and contemporary studies indicate that by the age of six most children are actively involved in sex role differentiation.iii Since we know that boys show uneasiness, anxiety, and anger when they are in danger of behaving in ways regarded as characteristic of the opposite sex, then a clear sex role distinction would appear to be of vital importance to boys. We also know that girls, in addition to seeking out female adult models, are more willing, even eager, to participate with boys in play and in boy-associated activities. Therefore, a strict differentiation in sex role behavior would not be as essential for girls. The data of these studies, coupled with the growing body of research pointing to the reading difficulties experienced by boys,iv suggest a real need for a closer examination of the content of reading textbooks in terms of the sex role models they are projecting.

This investigation, therefore, was interested in finding out if the results of the Wake, et al,v study could be replicated with books used to teach reading to previous generations of American school children, and focused on the following questions: (1) Are the sex role models portrayed in accord with the behavior patterns and expectations of the time period during which the readers were being used; and (2) Are the male and female models clearly differentiated in the texts during different historical periods.

While considerable national attention has been focused on the need for providing appropriate sociocultural models in the reading materials, less attention has been given to presenting appropriate developmental models and the motivational influence of these models on the child. It is hoped that by examining the content of reading texts along the dimensions suggested by these questions that teachers, textbook authors, and textbook publishers will heighten their awareness of the nature of the content of the readers and of the possible influence of their content on the learner.


The methods devised to answer each of the above four questions involved: (1) selection of books to be coded; (2) designation of appropriate time periods; (3) development of the criterion behavior lists; (4) development of coding units; (5) categories for coding; (6) training of raters; and (7) tabulation and treatment of the data.

Selection of Books

Selection of textbooks published prior to the late nineteenth century was limited to those distributed in the Northeastern states since there was great variability of books published for the Southern and Western states. Texts published since the late nineteenth century, however, were written for national distribution and were not specifically assigned to or written for any one sector of the country. Therefore, no geographic limitation was necessary on the selection of books for this latter time period.

Only those books which met the requirements of a primary reading text were included. This term refers to those books used to formally introduce beginning reading instruction to children. During Periods I and II (1600-1835) the primary reading texts were used to introduce reading instruction to a wide age range. Grade levels in school also were not prevalent and thus there was no distinction made as to age or grade level of these early first readers. Gradually, as grade levels by age in school became conventional, a series of graded reading texts also began to appear and gain in popularity. As our knowledge of the reading process and child development grew, prefirst reading texts were written. Today it is common to find that a primary reading text consists of a series of several books, including several preprimers, primers, and a first reader. Therefore, the term primary reading text continues to take on slightly modified meaning during the progression from Period I through Period VI.

A list of those textbooks in wide use during the first five periods was provided by Carpenter,vi Johnson,vii Nietzviii and Smith.ix Final selection was based on their availability through interlibrary loan. The choice of modern primers was based on a national survey by Hollinsx of the three most frequently used contemporary reading series. The criterion of wide use reflected best the acceptance of their content by society. Thus three primary readers representative of each of the six historical periods were selected for coding, a total of eighteen readers in all.

All the stories were coded in all the books with the exception of two of the three books from Period II (1776-1835). Both of these were compiled by the same author and contained lengthy speeches and dialogues dealing with moral issues. In one book only the dialogues were selected, while in the other book ten speeches were chosen by using a table of random numbers.

Designation of Appropriate Time Periods

Sex role behavior expectancies were described within the context of the following time periods: Period I, 1600-1776; Period II, 1776-1835; Period III, 1835-1898; Period IV, 1898-1921; Period V, 1921-1940; and Period VI, 1940-1966. These time divisions as well as the criterion lists were established after a careful examination of the literature from: (1) social histories and historical studies of family life;xi (2) studies of children's behavior and activity preferences;xii and (3) studies analyzing child-rearing manuals and juvenile literature.xiii Most of the data reviewed were rich in descriptive content of the life and times of each of the periods covered in this study.

The significant changes in the behavior expectancies for adults and children over the past 350 years were identified. Both predominant behavior standards and trends toward changes in these standards were noted, and time boundaries were drawn to include trends away from a previous behavior standard, the predominant behavior standard, and trends toward a new standard of behavior.

Although these periods were arrived at empirically, similar time divisions have been used by historians of various aspects of American life,xiv and give credence to the time designations employed by this investigator. The major differences in time designations are found during 1776 to 1860. Usually this span is covered as a single unit. To meet the purpose of this study, the recognition of childhood as a special period of life and the introduction of economic and social changes which affected the status and role of women caused this investigator to extend this period to 1898 and to divide it into two separate units, 1776-1835 and 1835-1898.

The Criterion Behavior Lists

The age and sex role behavior expectancies for each of the six time periods were identified to include a list of: (1) play activities participated in by boys, girls, and both boys and girls; (2) behavior expectancy according to age (under and over five years of age) for boys, girls, and both boys and girls; (3) behavior expectancy for adults (men and women); and (4) a summary statement on children's behavior expectancy. By comparing the age and sex role behavior of the characters in the stories with the age and sex role behavior criterion lists, it was possible to answer, in part, the questions of primary concern to this investigation.

The Coding Units

This writer, in collaboration with other investigators in the content analysis of motivational dimensions in contemporary first-grade reading textbooks, had ascertained the workability of using the individual stories as units to be coded in the readers.xv Each story was numbered and dealt with as a complete unit, even though the succeeding story might be a continuation of the activities already begun. This was done because in the process of teaching the child to read these stories are usually presented as single, complete units.

Categories for Coding

A coding manual was developed for use in rating the sex role behavior of adult and children story characters in the primary reading texts. Items related to sex role behavior were listed by the author and subsequently discussed with two other research colleagues, a child psychiatrist and a child psychologist. Only those items which received three checks were included in the coding manual. Each story was then coded along the following dimensions:

1. Characters—Each story was coded according to the presence of its main characters. Ten categories were used, such as "children only," "children and mother/1 "adults only," "animals only," etc.

2. Theme—The predominant theme (activity) of each story was coded. Seventeen possible choices were available to the raters and included "active play," "pets," "lessons from life," "work projects," etc.

3. Age and Sex—Each story was coded according to: (a) the age at which children predominantly engage in or would be interested in the main activity depicted (under five-years old and over five-years old), and (b) whether this main activity would be performed primarily by boys, girls, or both boys and girls regardless of the sex of the character(s) carrying out the activity in the story.

4. Role of the Adult—Nine categories representative of adult roles, i.e., "protector," "nurturer," etc., were listed in the manual. The sex and role of the adult characters in each of the stories were thus identified and tabulated accordingly.

5. Outcome—The main theme (activity) of each story was coded as to the nature of its conclusion. Five categories were' available to the raters: (a) Success, when the activity or purpose was accomplished; (b) Failure, when the activity or purpose was frustrated, failed, or not achieved; (c) Help, when the activity or purpose was frustrated but then accomplished through the efforts of someone else; (d) Self-help, when the activity or purpose was frustrated but finally accomplished through the individual's own efforts; and (e) Uncertain, when the rater was unable to determine the outcome.

In addition tabulations were made of Outcome according to: (1) the age and sex of the children involved in the activity, and (2) the role performed by the male and/or female adult.

6. Dependency Behavior and Aggressive Behavior—Each of the stories was coded for Dependency and Aggression according to: (a) the age and sex of the person behaving dependently or aggressively, and (b) the reward or punishment response of adults and children (male or female) either as observers or recipients of the aggressive or dependent behavior. A special rating was also devised to record Dependency and Aggressive behavior in animals and the reward or punishment response to this behavior,

7. Behavior Expectancy—Stories that communicated by lecture, preaching, example, or implication how people were expected to behave were coded as to the age and sex of the person (s) at which the statement was directed and the valence (positive or negative example) of the statement.

8. Environmental Setting—The geographic location of each story was coded using the classification "Urban," "Suburban," "Rural," "Unrealistic or Imaginary," "Just Verbal Message," and "Not Clear."

9. Occupations—Any references or depictions to specific work roles and the sex of the character (s) doing the work were recorded for each story.

10. Illustrations—Notations were made on the story illustrations as to the physical characteristics of the characters depicted and their style of dress.

11. Appropriateness Ratings—Appropriateness ratings were made for: (a) the time period for which the activity was appropriate; (b) the sex and age of the children for whom the activity was appropriate; (c) the sex of the adult for whom the activity was appropriate; and (d) the time period for which the sex of the adult role was appropriate.

Training the Raters

Since sex role behavior was being coded, a male and female graduate student rated the stories. After a training period, independent interrater agreements were calculated to be at a mean of 95 percent over the 31 categories. The raters divided the books between them and proceeded to code them independently.

Tabulation and Treatment of the Data

Frequency distributions and percentages were tabulated for each of the dimensions described in the coding manual, and tables were drawn describing these data. Thus these data provide a statistical description of the sex role models portrayed in primary reading textbooks used by most American children between the years 1600 and 1966.


A diffuse sex role model was presented in varying and increasing degrees from colonial days to the present. The lack of specificity in sex role was consistent with the diffuse model described in the behavior criterion lists for each of the six periods. To this extent, the models presented were in accord with the times. However, it should be noted that a sex-differentiated model was also described on this list, but the portrayal of adult males and females performing similar roles and of boys and girls playing at the same activities was the model selected for presentation in these textbooks. This lack of sex specificity also presented a model of behavior more appropriate to children under five years than to children over five years old for whom they were designed. In presenting less mature models of behavior, we have ignored the tendency of older children to revert to behavior that was appropriate the year before, A prime insult is fro be accused of acting like a baby.

In addition to the consistent pattern of sex role diffusion which shows up from 1600 to the present, another consistent and complementary pattern appears to evolve. From a character count it was noted that textbook authors began to increase the number of female characters in the stories as formal education was opened to girls (between 1776 and 1835). This trend continued so that by 1898 and up through 1966 girl characters actually outnumbered boy characters in the texts. Despite this greater frequency of females in the stories, a distinct female behavior identity was avoided. A possible explanation for the minimizing of sex differences may be found in the desire to present materials that would be acceptable to a classroom grouping of boys and girls. It remains a curious matter, however, that other alternatives were not attempted. Thus one might also speculate that the neutral, non-sex-linked male and female behavior described in the stories was an unconscious effort to deny the existence of sexuality in children.

Another developmentally inappropriate model, that of dependency, was consistently sustained throughout the six periods. This behavior was particularly prominent in the books coded from 1921 through 4966, and was rewarded overwhelmingly for both sexes and for all age levels. Here again we have an example of a regressive pull in the presentation of a model of behavior which is more appropriate to children younger than the beginning reader. Since dependency is characteristically associated with females, a feminizing quality also was present in the positive characterization of the male dependent model.

Since it is primarily from adult models that sex role behavior is learned, it was important to examine the characterizations of adults in these primers within the family structure described in the behavior criterion lists. In Periods I and II (1600-1835), the American family was essentially a productive, functional unit. As soon as the young were able to cope with the physical demands of the job, they carried out the same tasks as the adults in their lives. Boys emulated their fathers, and girls their mothers, and they were treated as adults. In the textbooks coded for these two periods, adult characters predominated. They were portrayed as idealized models of religious and ethical behavior, but they were not participating in distinctly male or female roles.

As the productivity function was removed from the home, children no longer had visible work models to emulate and childhood came to be regarded as a period of carefree play. This more affluent America discovered childhood, and with the discovery came the separation of children from the adult world. In the books representative of Periods III and IV (1835-1921), adult characters practically disappeared from the texts and were displaced by children and animals. When adults were present, they served as backdrops to the actions going on among the children. Their very absence suggested that the interests and activities of children were different from those of adults, despite the fact that these differences were neither spelled out nor were they clearly visible. Thus these books carried no specific description of male or female adult roles.

Male and female adult characters entered the book again in sizable numbers during the last two periods (1921-1966). During that time they were presented as facilitators of their children's wishes, interests, and needs but without distinct interests and needs of their own. This characterization was consistent with the child-rearing values of family togetherness espoused during these periods. Nevertheless, clearly defined adult male and female behavior was avoided in this portrayal of a child-centered adult model. There also was no reference to the world outside this close-knit nuclear family. Thus from this overall review of the presentation of the adult models in all the books coded for all the six periods, it can be seen that a very limited and restricted prototype of sex role behavior has been portrayed. To this extent, the primary reading texts fell far short of fulfilling their role of an acculturation medium.

By comparing the Adult Roles, Illustrations, and Behavior Messages to the Behavior Criterion Lists for each period, and by reviewing the Environmental Place Settings of the stories coded in all six periods, it was quite apparent that only one socioeconomic and cultural group was represented in the total sample of texts examined, and thus only one possible social class model of sex role behavior was presented.

Perhaps in this sense what was left out of the content of these primary reading texts is as important to examine as what was put in. The exclusion of the plurality of sex role models that exists in American society suggests that these texts ignored the differences in cultural backgrounds and socioeconomic conditions that account for these differences. It is interesting to speculate whether this was an attempt to unify a diverse people under the white, Anglo-Saxon, middle-class model in the spirit of egalitarianism, or if this was a reflection of the attitude toward the role of education as a selector and sustainer of tradition. The avoidance of socioeconomic and cultural differences is similar in a sense to the avoidance of sex differences. We are saying, in essence, that by ignoring them or diffusing them we are doing away with the evils and inequities associated with them. This is the old story of treating the symptom rather than the cause. To the extent that sex labels, cultural labels, and socioeconomic labels produce inequities in our society, the inequities should be eliminated, not the differences.


The results of this study point to the need for a more pluralistic depiction of American life in reading textbooks. Other studies that have analyzed the cognitive,xvi affective,xvii and vocationalxviii behavior in reading texts at other grade levels have come to this conclusion, pointing to a consistent pattern in the writing of these texts.

The present investigation also suggests that there is a vital role that these books can perform in overcoming the cultural discontinuities between childhood and adulthood, by helping the child to get a clearer understanding of his present role and by making visible the actual behavior expectations that lie ahead. What is needed next is an investigation into the actual effect of content on children's attitudes and reading acquisition. It may very well be that the degree of success these texts achieve in teaching reading to children may depend upon their success as a socializing agent. Once we gain more insight and understanding in this area, it may be possible to combine our knowledge of child development with the talent of the artist and write textbooks that are equally effective as instruments of acculturation and of teaching the reading skill.


i This investigation was supported by a research grant from the Office of Education of the Department of Health, Education, and Welfare, Project No. 3094.

ii Richard R. Waite, Gaston E. Blom, Sara G. Zimet, and Stella Edge, "First Grade Reading Textbooks," Elementary School Journal, Vol. 67, 1967, pp. 366-74.

iii Alice Earle. Child Life in Colonial Days. New York: Macmillan, 1899. Jerome Kagan, "Acquisition and Significance of Sex-typing and Sex Role Identity," in M. L. Hoffman and L. W. Hoffman (eds.). Review of Child Development Research. New York: Russell Sage Foundation, 1964, Vol. 1. Roberta Oetzel, "Annotated Bibliography," in E. E. Maccoby (ed.), The Development of Sex Differences. Stanford, California: Stanford University Press, 1966. G, H. Steere, "Changing Values in Child Socialization: A Study of U.S. Child Rearing Literature 1865-1929" (Unpublished doctoral dissertation, University of Pennsylvania, Philadelphia, Pennsylvania, 1964). B. Sutton-Smith and B. G. Rosenberg, "Sixty Years of Historical Change in the Game Preferences of American Children," Journal of American Folklore, Vol. 74, 1961, pp. 17-46.

iv J. Money, "On Learning and Not Learning to Read," in J. Money (ed.). The Disabled Reader. Baltimore: Johns Hopkins Press, 1966.

v Wake, Blom, Zimet, and Edge, op. cit.

vi C. Carpenter. History of American Schoolbooks. Philadelphia: University of Pennsylvania Press, 1963.

vii C. Johnson. Old-time Schools and Schoolbooks. New York: Dover Publications, 1963.

viii J. A. Nietz. Old Textbooks. Pittsburgh: University of Pittsburgh Press, 1961.

ix Nila B. Smith. American Reading Instruction. Newark, Delaware: International Reading Association, 1965.

x W. H. Hollins, "A National Survey of Commonly Used First Grade Readers" (Unpublished data, Alabama A & M College, 1955).

xi P. Aries. Centuries of Childhood. New York: Alfred A. Knopf, 1962. D. G. Brown, "Sex Role Development in a Changing Culture," Psychological Bulletin, Vol. 60, 1958, pp. 232-42. A. A. Calhoun. A Social History of the American Family 1601-1919. (3 vol.), New York: Barnes and Noble, 1919. Earle, op. cit. Rachel Elder, "Traditional and Developmental Conceptions of Fatherhood," Marriage and Family Living, Vol. 11, 1949, pp. 98-106. A. L. Gesell (ed.). The First Five Years of Life. New York: Harper Brothers, 1946. A, Gesell and Frances Ilg. The Child from Five to Ten, New York: Harper Brothers, 1946. J. Kagan, op. cit. Margaret Mead. Male and Female. New York: William Morrow, 1949. Roberta Oetzel, op. cit. H. E. Scudder (ed,). Men and Manners in America One Hundred Years Ago. New York: Scribner, Armstrong and Co., 1876. Anna Spencer. The Family and Its Members. Philadelphia: J. B. Lippincott, 1923. J. H. van den Berg. The Changing Nature of Man. New York: W. W. Norton and Co., 1961. W. E. Woodward. The Way Our People Lived. New York: E, P. Dutton and Co., 1946.

xii T. R. Croswell, "Amusements of Worcester School Children," The Pedagogical Seminary, Vol. 6, 1898-99, pp. 314-71. W. W. Newell. Games and Songs of American Children. New York: Harper Brothers, 1883. Button-Smith and Rosenberg, op. cit.

xiii Monica Kiefer. American Children through Their Books 1100-183$. Philadelphia: University of Pennsylvania Press, 1948. Clara D. Riley, "Perceptions Concerning Children as Revealed through Poetry for Children 1853-1850, 1875-1890" (Unpublished doctoral dissertation, Florida State University, Tallahassee, Florida, 1963). Alice Ryerson, "Medical Advice on Child Rearing, 1550-1890" (Unpublished doctoral dissertation, Harvard University, Cambridge, Massachusetts, 1960). G. H. Steere, "Changing Values in Child Socialization: A Study of U.S. Child Rearing Literature 1865-1929" (Unpublished doctoral dissertation, University of Pennsylvania, Philadelphia, Pennsylvania, 1964).

xiv R. F. Butts and L. A. Cremin. A History of Education in American Culture. New York: Holt, Rinehart & Winston, 1953. A. Calhoun, op. cit. L. V. Parrington. Main Currents of American Thought. New York: Harcourt, Brace and Co., 1930. Nila B. Smith, op, cit.

xv G. E. Blom, R. R. Waite, and S. G. Zimet, "Ethnic Integration and Urbanization of a First Grade Reading Textbook: A Research Study," Psychology in the Schools, Vol. 4, 1967, pp. 176-81. G. E. Blom, R, R. Waite, and S, G. Zimet, "A Content Analysis of First Grade Reading Textbooks," The Reading Teacher, Vol. 21, 1968, pp. 317-23. R. R. Waite, op. cit.

xvi I. L. Child, E. H. Potter, and E. M. Levine, "Children's Textbooks and Personality Development: An Exploration in the Social Psychology of Education," Psychological Monographs, Vol. 60, 1946, pp. 1-54.

xvii Monica B. Holmes, "A Cross-Cultural Study of the Relationship between Values and Modal Conscience," in W, Muensterberger and S. Axelrad (eds.). The Psychoanalytic Study of Society. New York: International Universities Press, 1960, Vol. I.

xviii Juanita Clyse, "What Do Basic Readers Teach About Jobs?" Elementary School journal, Vol. 59, 1959, pp. 456-60. W. W, Tennyson and L. P. Monnens, "The World of Work through Elementary Readers," Vocational Guidance Quarterly, Winter 1963-64, pp. 85-88.

Cite This Article as: Teachers College Record Volume 72 Number 1, 1970, p. 31-40
https://www.tcrecord.org ID Number: 1662, Date Accessed: 1/25/2022 6:12:28 PM

Purchase Reprint Rights for this article or review
Member Center
In Print
This Month's Issue