The main focus of this research is to investigate the representation of gender in ESL textbooks at secondary school level in Punjab, Pakistan. For this purpose, a content analysis approach has been utilized. Because this study is concerned with the investigation of gender representation in textbooks, it falls within the domain of sociolinguistics. The content for this study was taken from the English language textbooks used with students of grades 9 and 10 for the academic year 2020–2021. These textbooks are prepared and published under the supervision of the PCTB. The PCTB is under the authority of the Government of Punjab, Pakistan and its textbooks are used in all public and some private sector schools in Punjab Province, and therefore have great significance from educational, social, and moral perspectives. The PCTB (2018) stresses its commitment to conducting periodic reviews and revisions of the textbooks (in the light of the comments of the review committee). However, there is no specific schedule in this regard.

Fairclough’s (1989, 2001) three-dimensional model has been adopted to investigate the representation of genders in the said textbooks and thereby reveal ideologies behind the ways in which gender is represented. Fairclough (2001) introduces three different kinds of constraints that are misused by powerful people to maintain their dominance over the non-powerful. These constraints are: (1) constraints on contents, (2) constraints on relations, and (3) constraints on subjects (p. 61). According to Fairclough (2001), these constraints are found to operate either in relatively concrete and immediate terms, or in relatively structural and long-term ways. It is however, important to mention here that power behind such discourse as marks “long term structural effects of a more general sort” (p. 61) is the concern of this research.

Fairclough (2001) further adds that to achieve commonality as well as coordination of practice in terms of beliefs, knowledge, societal identities, and relationships requires different mechanisms which vary from society to society. He classifies these mechanisms into three different types of discourse which are practiced across the world. These types are (1) inculcation mechanism, (2) communication mechanism, and (3) imposing power in a hidden fashion. This study primarily focuses on the third of these, exploring such mechanisms as are “imposed in the exercise of power in a largely hidden fashion” (Fairclough, 2001. p. 62). This is carried out using a content analysis technique to quantify the keywords related to the representation of gender in the textbooks with the help of the following factors: (1) features representing both genders; (2) tokens used for both genders; (3) inclusion or exclusion of genders; (4) titles; and (5) nouns and pronouns used for both genders. In this regard, different keywords were determined concerning both male and female genders in terms of their representation in the textbooks. Counting the frequencies of keywords involves a quantitative approach whereas defining the quantity of results involves a qualitative approach. Thus, this study utilizes both of these approaches: the quantitative approach enables the enumeration/ quantifying of keywords related with the two genders while the qualitative approach facilitates a description of which gender is most prominently represented.

To develop the corpus, the data was retrieved from the English 9 textbook available online on the PCTB official website. However, English 10 was not available online. Therefore, a hard copy was scanned and then processed using OCR to Word 7.8.1 software, thereby converting the scanned version of the textbook into MS Word. Both textbooks (English 9, and 10) contained a variety of different units, theme-based texts of varying page lengths. English 9 contained 12 units (2 poems, 10 prose lessons) while English 10 consisted of 13 units (3 poems, 10 prose lessons). Each lesson (in both textbooks) was followed by exercises based on different language skill activities. As mentioned earlier, the PCTB revises its textbooks from time to time, and these two textbooks were last revised in 2012.

This study focused on the text of the main chapters/units/lessons as well as the exercises given at the end of the chapters. Additionally, the “Dengue Fever Awareness Campaign” as well as the “Introduction” sections from English 10 were included in the study’s corpus. The Introduction section included a number of sub-topics, including grammar, guidelines for teachers, oral communication skills, pre-reading, reading comprehension, reading material, writing skills, vocabulary, while-reading, and conclusion. Similarly, the “Dengue Fever Awareness Campaign” consisted of a comprehensive essay which discussed control, dengue vectors, the dengue virus, diagnosis, government’s effort to control dengue, history, patient management, preventive measures, and symptoms, followed by the chapter review exercise.

The distribution and length of the corpora is given in Table 1. It should be noted that this study sets itself apart in considering certain sections (e.g., index, introduction, and exercises) for analysis purpose, sections left unexamined by all previous studies. This study is, therefore, unique. However, it was difficult for the researchers to process the images contained in the textbooks. Therefore, the images were omitted from this study.

The corpora arrangement involved five stages: (i) it was collected in electronic version, (ii) it was recorded in Microsoft Excel, marking file number, word types, and token types; (iii) the files were renamed; (iv) the corpora files were converted into Notepad for processing; and (v) the corpora were arranged in two separate folders named as “Compiled up Files” that were further arranged into “One File.” The corpora saved in Notepad files were later analyzed using AntConc 3.4.4.0 by (a) tagging through a CLAWS tagger to obtain grammatical terms via tags i.e. proper nouns (Car, Ali, London), reflexive nouns (herself, himself), and personal pronouns (he, his, him, she, her, hers); and (b) observing grammatical terms through the tags like reflexive pronouns (PNX), proper nouns (NP0), and personal pronouns (PNP). Finally, the frequencies were counted manually and interpreted, as described in the Results section.

One limitation of the study is that the analysis is based on only two English language textbooks taught at secondary level in Punjab province. Therefore, its results are not necessarily generalizable to the entirety of teaching/learning materials in Pakistan.

After the analysis, it was observed that both of the textbooks, i.e. English 9 and English 10, contained features representing both genders. However, there were considerable differences in their representation in that male references exceeded female references. English 9 used 17 male and 13 female features. Similarly, English 10 used 15 male and only 8 female features. As a whole, both of the textbooks represented women less than men through the features representing both genders, though the difference was significantly greater in English 10. Table 2 presents the clear picture of the representation of both genders through features.

English 9 used 17 cluster types and 264 cluster tokens for the male gender: he, his, man, him, grandfather, son, brother, uncle, boy, father, male, men, Mr., nephew, sir, and husband. Of these, the male token he was observed to have the maximum frequency, i.e. 123, whereas himself and husband had the lowest frequency, i.e. 1. His and man came second and third in terms of maximum frequency occurrence (66 and 20 respectively). For details about the frequencies of other tokens, see Table 2.

English 9 used 13 cluster types and 220 cluster tokens for the female gender: her, she, female, sister, mother, woman, women, daughter, madam, girl, aunt, herself, and wife. Of these, the female token her was observed to have the maximum frequency, i.e., 87, whereas herself, girl, aunt, and wife had the lowest frequency, i.e., 1. She and female came second and third in terms of maximum frequency occurrence (85 and 11 respectively). For details about the frequencies of other tokens, see Table 2.

English 10 used 17 cluster types and 264 cluster tokens for the male gender: he, his, him, man, father, men, Mr., son, brother, himself, boy, boys, husband, sir, and uncle. Of these, he, as was the case for English 9, was observed to have the maximum frequency, i.e., 132, whereas sir, husband, and uncle had the lowest frequency, i.e., 1. His and him came second and third in terms of maximum frequency occurrence (84 and 31 respectively). For details about the frequencies of other tokens, see Table 2.

English 10 used 8 cluster types and 83 cluster tokens for the female gender: her, she, sister, mother, woman, aunt, Mrs., and women. Of these, she was observed to have the maximum frequency, i.e. 42, whereas aunt, Mrs., and women had the lowest frequency, i.e., 1. Her and sister came second and third in terms of maximum frequency occurrence (20 and 9 respectively). For details about the frequencies of other tokens, see Table 2.

Comparative analysis shows that both of the textbooks i.e. English 9 and 10, used features/tokens for female and male genders. However, a considerable difference between the frequencies of the features of both genders was seen.

Analysis of the ESL materials for Grades 9 and 10 reveals that gender discrimination is deeply rooted in them. The representation of men and women is not equal i.e., women are far less represented. Inequality has been observed through all indicators, whether it is features, titles, nouns, pronouns, inclusion/exclusion, or tokens. All these indicators showed male over-representation and female under-representation. In the view of Amerian and Esmaili (2015), this represents an organized and systematic manipulation of the materials, in which both sexes, particularly women, have been presented in a certain way and one which makes women appear inferior. The results show that gender bigotry is quite evident in the textbooks. Thus, not only can it be said that men and women are not equally been represented in the textbooks, in fact the textbooks seem replete with examples of male dominance (Foroutan, 2012). This situation conforms to the belief of Bahman and Rahimi (2010) that educational materials are manipulated to exalt men and present women as derogatory creatures, and also validates the findings of Masud (2017) that the textbooks used in Pakistani schools are not presenting men and women equally. Rather, the textbooks (see Azhar et al., 2014) are based on notions of male superiority and female inferiority.

In Pakistan, gender discrimination in textbooks is prevalent (Durrani, 2008; Masud, 2017; UNESCO, 2004) and national identities (as presented in the textbooks) create an “understanding of the relative positioning of religion and gender in relation to nationhood” among the children (Durrani, 2008, p. 599). A study on 194 textbooks taken from all four provinces of Pakistan found that the national curriculum at school level mirrored gender to a great degree (UNESCO, 2004). The study also revealed that textbooks contained only 7.7% representation of women, most of which were taken from Islamic history. As a whole, according to Shah (2012), there is significantly less representation of women in textbooks than men.

The contexts in which women are generally depicted in Pakistani textbooks is also gender biased. Women are usually represented as being powerless, tolerant, dependent, and pious, caring for and nurturing children and husbands (Durrani, 2008; Ullah & Skelton, 2013). In addition, they are depicted in typical womanistic roles like cooking, washing dresses, cleaning, and raising the children. Similarly, in the professional sphere, the role of a woman is restricted to limited fields e.g., teaching (Ahmad & Shah, 2019; Durrani, 2008; Masud, 2017; Ullah & Skelton, 2013; UNESCO, 2004). Ram (2008) adds that the occupational roles of women are restricted to traditional “women’s professions,” such as nursing, and household work is typically shown to be a woman’s only domain.

According to Masud (2017) and Smith (1991), education has the power to structure the behavior, attitudes, and opinions of human beings. It is used as an instrument to promote national identity and boost the privilege of some social groups over others, including that of men over women. These privileges, particularly inequalities, are defined, reproduced, shaped, reinforced, and promoted through education, both implicitly and explicitly. Materials used for educational purposes, say Britton and Lumpkin (1977), not only provide learners with information, but also inject negative or positive attitudes into the children’s minds about different aspects of life, such as life chances, life expectations, occupations, religion, race, and sex. And gender bias, according to Rahimi and Sahragard (2006, p. 29), is “rife in the ideological manipulations of the texts.” In addition, educational materials, according to Thorne et al. (1983) and Weatherall (2002), contain linguistic terms (like lady, mistress, and spinster) that ignore and belittle females, while Baskan and Erduran (2009) add that the said terms represent women as inferior creatures. And according to some other researchers, like Bauer et al. (2006), Kramer (1974), and Parks and Roberton (2004), male generic items (like man, boy, mankind) apparently seem to refer to both genders but in fact, such terms make women indiscernible. As education plays a vital role in mapping young learners’ minds (Biswas, 2017), such texts with their biased projections of genders might negatively affect the minds of the learners (Ullah & Haque, 2016).

The discriminatory positioning of women and privileged positioning of men, both implicitly and explicitly, can raise them as gendered subjects (Ahmad & Shah, 2019; Durrani, 2008). Such depictions in textbooks indicate what it means to be a child in a particular context, which includes the learning of gender identities in the process of socialization (Kereszty, 2009). Textbooks are specifically more critical as they present themselves as embodying a realistic account of things it is assumed children need to learn; such unequal and discriminated representations of both genders can result in gender stereotyping in young minds (Kereszty, 2009).

Such a discriminatory representation of genders might condition young children to grow up as gender-conscious individuals. It is interesting to note Durrani’s (2008) study on children in this context. She asked some boys and girls to draw an image of us. She found that none of the male participants drew female images. Female participants, on the other hand, drew female images but they were shown carrying out domestic activities. Then, she asked the participants to pick icons of their choice from the textbooks: Most of the boys selected male icons and girls opted for female icons. When she asked about the reason for their choosing the specific female icons, female participants replied that they opted for “good wives or good mothers.”

Unequal gender representation might create a limited understanding of the societal role a woman plays and reduce the opportunities available to female learners to practice the target language compared to male learners (see Amerian & Esmaili, 2015; Brusokaitė, 2013; Ebadi, Salman, & Marjal, 2015; Gershuny, 1977; Nagatomo, 2010; Nofal & Qawar, 2015; Siren, 2018). In addition, tolerating unequal gender representation and gender-specific conditioned learning and development of children in our country where women form more than half of the total population cannot be justified: It may lead to the children growing up as gendered subjects (Durrani, 2008) and have a negative effect on the personalities of the female students. As a result, they may fail to realize their potential (Amerian & Esmaili, 2015). Moreover, it may also adversely affect the children’s identity formation process (Ullah & Haque, 2016), shaping gender identities that will negatively affect the achievement of educational goals regarding gender equality in education, and women’s empowerment in terms of economic, national, and social development (Mustapha & Mills, 2015).

According to Azhar et al. (2014), the depiction of gender in learning materials is characteristic of the culture it is related to. They confirmed this notion through a comparative analysis of British and Pakistani works of fiction in English which indicated a balanced representation of both genders in British English fiction but quite the contrary for Pakistan. This imbalance is not confined to works of fiction. The reality is that men are regarded as superior to women in Pakistani culture, and this is clearly reflected in textbooks. In particular, the appearances of he and she in textbooks can be based on the notions of the inferiority and superiority of the genders in Pakistani culture.

Poor economic conditions can be taken as another factor in the biased treatment of women. In Pakistani culture, men are responsible for social and domestic expenditures, and women depend on men for their economic needs. That is why the participation of women in the Pakistani labor market is low (55: 100 is the female: male ratio) as compared to men (Islam & Asadullah, 2018). In addition, women in Pakistan tend to be restricted to limited jobs like housework. This restriction can be seen as limiting their representation in the learning materials.

A patriarchal mindset is another factor that can cause biased representation of women in the learning materials. Such a mindset has, in fact, already influenced pro-women policies in many countries e.g. denunciation of women as second-class citizens in Malaysian family law (Kent, 2006), the practice of honor killing in Pakistan (Londergan, 2013), virginity tests for Indonesian female police recruits (Watch, 2014, and reforms in marriage laws in Bangladesh and Malaysia in favor of child marriage (Amin, Asadullah, Hossain, & Wahhaj, 2017; Chowdhury, 2010). The patriarchal mindset (particularly in Pakistan) has resulted in limited women’s participation in the economic and political spheres, resulting in men being shown as dominant over women in learning materials. This dominance of men over women is quite characteristic of, and grounded in, Pakistan’s culture (Islam & Asadullah, 2018).

The World Bank Development Report (2012) highlighted gender equality as its core objective, stressing that gender equality can enhance productivity and economic growth for the current generation, and improve development outcomes for the next. Therefore, efforts should be made to represent both genders equally. It must be ensured, as proposed by Britton and Lumpkin (1977, p. 41), that “the values and societal roles suggested in instructional materials be positive ones, and be as free as possible from bias, stereotypes, and career-role restriction.” This can be achieved, if male and female distinguishing terms are replaced by human beings (Kramer, 1974); man is replaced by person (Graham, 1974); and gender-neutral terms such as people are used (Sunderland & Litosseliti, 2002). In addition, effective changes are required at institutional and personal levels (see Litosseliti, 2014), particularly an increase in the number of female writers; the sensitization of teachers as well as writers regarding gender bias; reflection of dynamic female roles such as professionals and income earners as well as caregivers and mothers (Islam & Asadullah, 2018); and the inclusion of more female characters (Mahmood & Kausar, 2019) which would be helpful in balancing gender representation in school textbooks (Ahmad & Shah, 2019; Durrani, 2008; Masud, 2017). According to the findings of Ullah and Skelton (2013), current textbook design exhibits all the signs of gender blindness.