Cleaning mums and brave policemen: gender representation and gender- inclusive language in EFL teaching materials in Lithuania

Gender representation in textbooks used for teaching English as a foreign or second language (EFL/ESL) has been analysed extensively and in various countries since the 1970s. According to most studies (e.g. Hellinger 1980; Porreca 1984; Ansary and Babaji 2003; Pihlaja 2007; Lee 2016), females tend to be underrepresented and depicted in stereotypical ways, which risks reinforcing gender clichés in students (Britton and Lumpkin 1977; Peterson and Lach 1990). To date, only a few studies (Brusokaitė and Verikaitė-Gaigalienė 2015) have focused on gender representation in teaching materials used in Lithuania; moreover, teaching materials other than textbooks remain underexplored. Using corpus analysis tools and criteria adapted from previous analyses of gender representation and language use in textbooks, this study takes a both quantitative and qualitative approach to the analysis of two EFL exercise books and two EFL test books published in Lithuania between 2005 and 2017 and currently available on the Lithuanian book market. The results show that males are more visible numerically; meanwhile, females are more likely to be represented in stereotypical ways. This article argues that neither gender representation nor language use in the books reflect developments in terms of gender equality in English-speaking societies (Pauwels 2001; Romaine 2001; Baker 2010).

In the context of this study, we need to address the question of to what extent school textbooks might affect children's gender socialisation. It has been claimed that books featuring stereotypical gender roles negatively alter children's image of themselves and others (Britton and Lumpkin 1977;Peterson and Lach 1990), which can restrict them with regard to their life styles and careers. Jones et al. (1997), however, argue that while there is a certain risk for negative effects on children's development, the exact opposite may happen as well if they challenge the sexism they are presented with by resisting to be portrayed in stereotypical ways. Despite disagreement on the negative effect of gender stereotypes in textbooks on children's social and cognitive development, the positive effect of gender-fair language, e.g. on children's perception of stereotypically male professions (Vervecken et al. 2013), has been shown. This suggests that gender-fair language does have an effect on children's socialisation with a far-reaching impact, e.g. with regard to their career choices, and, therefore, deserves our attention.
The representation of men and women in coursebooks and especially EFL/ESL teaching materials has been the focus of a great number of studies conducted in various countries since the late 1970s.
The majority of studies have found that females in EFL/ESL textbooks are underrepresented in both texts and illustrations (Ansary and Babaji 2003;Barton and Namatende Sakwa 2012); subordinated to males in at least certain contexts (Pihlaja 2007); described by stereotypically female adjectives and actions linked to their weakness (Hartman and Judd 1978;Barton and Namatende Sakwa 2012); usually named second in paired constructions (Porreca 1984;Lee and Collins 2008; Collins 2009); and ascribed social and domestic roles, while males tend to practise physically and/or intellectually demanding, more interesting and more diverse professions (Hellinger 1980;Lee and Collins 2008). It has been noted by Barton and Namatende Sakwa (2012) that research in this field has mostly been conducted from a European/Western and Asian perspective (see studies Klein A., 2019. Cleaning mums and brave policemen: gender representation and genderinclusive language in EFL teaching materials in Lithuania, 12: 75-96. https://taikomojikalbotyra.lt 78 on Australia: Lee and Collins 2009;Germany: Hellinger 1980;Hong Kong: Lee and Collins 2008;Japan: Pihlaja 2007;Lee 2016), and only recently have researchers started to include developing countries. In many of these countries English is taught as a second language and discrimination against women tends to be higher than in Western societies (Barton and Namatende Sakwa 2012).
Furthermore, the majority of researchers focus on EFL/ESL teaching materials, with only a few having analysed teaching materials for languages other than English (to name a few, see studies on Greek : Poulou 1997;Russian: Rifkin 1998). Meanwhile, most studies are synchronic in nature, presenting a snapshot of teaching materials available at one point in time. Other studies (Lee and Collins 2008) distinguish more clearly between older and more recent textbooks in order to be able to make reliable statements about changes with regard to gender representation.
Gender-representation in EFL teaching materials used in Lithuania is an underexplored topic. To date, Brusokaitė and Verikaitė-Gaigalienė (2015; based on Brusokaitė's 2013 Master's thesis) are the only ones to have done research on this topic in Lithuania, comparing one EFL textbook used in Lithuania with one each from Argentina and Great Britain. Their findings are in line with the general observation that females are likely to be depicted in stereotypical contexts and males to be named first in paired constructions.
Quantitative and qualitative content-based analysis has always been an important feature of textbook studies, with researchers counting, for instance, the number of male and female characters in a given book and analysing their depiction in gender stereotypical contexts. However, content analysis has been criticised on the basis of its static nature which does not take alternative interpretations into account (Mills 1995). Therefore, most current studies combine quantitative and qualitative research methods with elements of discourse analysis in order to explore the underlying reasons for phenomena related to gender representation in textbooks. Furthermore, some researchers such as Sunderland et al. (2001) have argued that the effect of a text can only be predicted if the teacher's mediation of the teaching materials is studied as well. Consequently, some more recent studies (Barton and Namatende Sakwa 2012) include lesson observations and interviews with the teachers concerning their discourse on the texts.
Based on Rifkin's (1998) evaluation of three different types of studies on gender representation in textbooks irrespective of the target language and the cultural and linguistic background of students, most recent studies of gender representation in teaching materials feature a list of criteria Klein A., 2019. Cleaning mums and brave policemen: gender representation and genderinclusive language in EFL teaching materials in Lithuania, 12: 75-96. https://taikomojikalbotyra.lt 79 established prior to the examination of the material. Although some authors have compiled criteria for the assessment of gender representation in foreign language teaching material in general (Hellinger 1980;Women in EFL materials 1991;Rifkin 1998), there is currently no exhaustive list of criteria that has established itself as a sort of universal compilation.
Pakeliui į 5 klasę (2017). For the sake of simplicity, the following abbreviations will be used to refer to the materials at hand: EB1 for Key to English. Activity Book 6 (2010); EB2 for Anglų kalbos gramatika su pratimais (2005); TB1 for Anglų kalba. Pakeliui į 5 klasę (2017); TB2 for Golden Key. Tests 3 (2008). These books were selected because all of them (i) were either exercise or test books; (ii) were written by Lithuanian authors; (iii) addressed Lithuanian students in their first and second year of studies, with units centred around topics such as holidays, school and home, i.e. contexts in which family members and friends can be expected to play a major role; and (iv) had been published within a time span of 12 years. Nevertheless, the four books differ considerably with regard to their length, ranging from 2,348 to 47,715 word tokens, and the importance they give to explaining grammar. Based on this observation, the four books were grouped into the following learning levels: Level 1: TB2; Level 2: EB1 and TB1; Level 3: EB2.
This study focuses on exercise and test books firstly because these materials have been largely ignored by researchers, and secondly because they are by nature materials that encourage independent learning. Especially when they come with an answer key as two (EB2 and TB1) of the four books do, these materials are often chosen for self-study without requiring further guidance from a teacher. If an answer key is not provided, exercises are usually discussed in class; if, however, exercises are assigned as homework, students will still be working on them independently at home, with corrections only made during the next lesson. As noted earlier, it has been argued that a teacher's mediation of a text is likely to have a greater impact on students and their social and cognitive development than the (non-)sexist nature of the text per se (Sunderland et al. 2001).
However, there is a risk that in contrast to reading texts such as dialogues from textbooks, repetitious exercises are regarded as serving grammar learning purposes only. Consequently, . Cleaning mums and brave policemen: gender representation and genderinclusive language in EFL teaching materials in Lithuania, 12: 75-96. https://taikomojikalbotyra.lt 80 teachers might not mediate or comment on their content. Therefore, lesson observations and interviews with teachers were considered irrelevant for the purpose of this study.
The four books were manually transcribed from hard copies and turned into four corpora to allow for corpus-based analysis. Exercises featuring non-open-ended questions (e.g. fill-in exercises) were completed by the researcher to imitate what exactly students would find after completion of the exercise. Vocabulary lists, if available, were not considered since they consisted mostly of words already featured in the exercises, and the few additional words were not used in example sentences.
Codes were assigned to every reference to unambiguously male and female characters, including names, all kinds of pronouns referring to a male or female character, non-gendered professions if the person could clearly be identified as being male or female, and pets if referred to as being male or female. In contrast to classical textbooks, the exercise and test books studied here feature few to no recurring characters. Therefore, it was found that counting all references to male and female characters would give a more precise picture of their respective visibility even if their numbers would be much higher and not equal to the actual number of characters in the books. For instance, in the sentence -Yesterday my sister Mary played with her friend.‖, both nouns sister and Mary and pronoun her would be coded -female‖, even if they all referred to one and the same person. By choosing a word as an analytic unit for coding rather than the whole sentence or phrase, I found that the visibility and salience of males and females could be more easily and more precisely mapped.
Instructions to the students were considered as well (in English and Lithuanian), and so were graphic materials in TB2 as they often helped to clarify whether pronouns such as I, you, we and they referred to males or females.
After the first coding cycle, a second code was added to the first one depending on whether the given character was represented in a stereotypically male or a stereotypically female context; no second code was assigned for neutral contexts. For this distinction criteria adapted from previous studies by Hellinger (1980), Porreca (1984), Women in EFL materials (1991), and Rifkin (1998) were used. Table 1 lists these criteria. Some changes were made to adapt the criteria established by the authors named above to the current study. The category -Intelligence‖ was not included as it has, arguably, lost its relevance over time.
Although it has been observed (Andrikienė and Vaičiūnienė 2016) that the gender gap in Lithuania is widening with regard to education, intelligence per se or good marks at school were not considered to be stereotypically male characteristics. Only when females were portrayed as being in need of assistance or made fun of on the basis of their ignorance or stupidity by a male was intelligence considered as a male characteristic. In this case, the category -Being led‖ for females was applied.
Afterwards, using tools for corpus analysis (AntConc) and content analysis, the extracted data were analysed according to several aspects: (1) references to females and males and the order in which they appear in paired constructions; (2) context of appearance; (3) frequency of the nouns mother/mum, father/dad, wife and husband as well as of the female address titles Miss, Mrs and Ms; (4) acceptable professions as suggested by the books; (5) references to famous women and men; interactions between males and females; (6) language use in the four books in order to find -man constructions used generically as well as gender-inclusive strategies such as he and she, he or she, he/she, she and he (also him and her, his and her, etc.), and instances of singular they (also them, their and theirs). While some of these categories had already been selected based on research previously conducted in this field (e.g. . This observation appears to support Hellinger's (1980) claim that materials targeting more advanced learners (here EB2) typically feature texts in which females are less adequately portrayed.
However, the present corpus is relatively small, and lower level of materials and higher number of female characters do not correlate: in fact, females are most visible in TB1, a book classified as Level 2, while TB2 as a Level 1 book only comes second. A significant association between the numeric differences of males and females, on the one hand, and the level of the teaching materials, on the other hand, could not be found (χ 2 (3) = 5.8839, p > 0.05).
In paired constructions, males are usually named first (TB1: 78% of all occurrences; EB1: 80%; or first names such as Amy and Andy. In TB2 only one instance of male-and female-firstness each was found; this can be explained by the book's simple language, with most sentences having only one person as their subject. These findings suggest that the risk of females' invisibility is highest in teaching materials with a high grammar content as a result of the lack of references to them and/or males being named first in paired constructions. This can be explained by the common practice of naming the male pronoun he before the female pronoun she in conjugation tables.

Context of appearance
While both males and females tend to be portrayed in stereotypical ways, the overall observation is that females are subjected to stereotypical representation to a greater extent than men. A comparison of the four books shows that while there are important differences between the books  Although marriage rates today are not as high as they were until the 1990s, these numbers contrast with the trend observed in most other European countries, i.e. decreasing marriage and increasing divorce rates (Eurostat 2018

Family roles and address titles
The corpus search conducted for the terms referring to family roles (mother/mum, father/dad, wife and husband) shows that despite their lower numerical visibility females are more present in the home domain than males. In all four books there are more mothers and mums than fathers or dads

Range of professions
A close analysis of the professions seemingly available or considered desirable for males and females further supports the argument that in the teaching materials at hand both genders tend to be portrayed in stereotypical ways. The books suggest not only that males can choose from a greater variety of professions than females, but also that females should preferably stick with stereotypically female professions related to teaching, care-taking, fashion and beauty, or the household. Males, on the other hand, can be found in leading positions, e.g. as headteachers, coaches, professors, managers, spokesmen, governors and bosses. Moreover, all criminals (thieves, robbers, murderers) in EB2 were male, which was also observed by Brusokaitė and Verikaitė-Gaigalienė (2015). While some professions such as painters, singers, butchers, doctors or pilots are more or less explicitly available for both males and females, there is clear preference for gendered terms such as policeman, postman, fireman or spokesman over their gender-neutral alternatives

Famous men and women
Differences between the respective numbers of references to famous males and females further contribute to women's invisibility. References to famous individuals can only be found in EB1 and

EB2.
In EB1 there is one mentioning each of Karl Benz, the Queen (referred to without her name) and the female singer Pink, whereas in EB2 mentions of male and female celebrities abound.
However, the total number of famous males (N=28) including multiple mention is much higher than the number of famous females (N=8 Finally, this analysis reveals a certain degree of negligence on the part of the authors who did not check the spelling of a name or referred with her actual name to a woman who did achieve something that was rather uncommon for females, especially in 1926.

Relationships between males and females
A qualitative analysis of the relationships between males and females outside the core family reveals that both males and females are overall more likely to interact with individuals of the same sex than of the opposite sex. While boys and girls play with or take care of their siblings or cousins of the opposite sex, opposite-sex interactions are less common when individuals other than family members are involved. Thus, male characters tend to be named together with other males, and females with other females. Moreover, same-sex pairs or groups of people are likely to pursue stereotypically male or female activities, respectively. Judging from their interests and types of activities, we can assume that most characters in the books are of roughly the same age as the students (i.e. age 11 to 14) unless stated otherwise. Hence, same-sex relationships at this age are portrayed as the norm, and opposite-sex relationships as the exception.
Once the characters become older, interactions with the opposite sex in the context of romantic relationships are featured as well. It is, for instance, mentioned that older brothers and sisters are dating people of the opposite sex, or stories about how elderly people got to know each other are told. In this context, however, one can observe a higher degree of male activity and female passivity: in EB1, the narrator's brother buys a diamond ring for his girlfriend, thereby complying with traditional gender roles. The same pattern can be found in EB2, e.g. in -He met his first wife during the war and married her in 1945.‖ (EB2: 105) and -He fell in love with her at first sight‖ (Ibid.). Females in romantic relationships are here portrayed as objects rather than partners at eyelevel. Another problematic example from EB2 is a short dialogue between a husband and his wife about the man they would like to become the husband of one of their daughters as he is a -single man and very rich‖ and, therefore, -a fine thing for our girls‖ (EB2: 115). The dialogue is, judging from the characters' names and the style of writing, based on Jane Austen's Pride and Prejudice.
Klein A., 2019. Cleaning mums and brave policemen: gender representation and genderinclusive language in EFL teaching materials in Lithuania, 12: 75-96. https://taikomojikalbotyra.lt 88 However, neither the title nor the original author's name are given. As it cannot be assumed that young learners of English are familiar with Austen's work, there is a high risk that they accept the presented role models as a social norm rather than discuss them in their historical context.
The results of this analysis show that in these books opposite-sex interactions are less frequent than interactions between people of the same sex. Moreover, in romantic relationships, females are often patronised and depicted as objects rather than subjects who can raise their voice. Last but not least, a certain negligence with regard to references not only results in a missed opportunity to discuss traditional gender roles, but also entails the risk of consolidating them.

Gender-inclusive language: paired constructions, singular they and man used generically
The language used in the teaching materials is to some extent gender-inclusive; however, the strategies used by the authors of the books do not reflect current developments in English-speaking countries (cf. Pauwels 2001;Romaine 2001;Baker 2010) but rather show that the authors were influenced by linguistic features of Lithuanian, i.e. their own mother tongue and that of the students.
While some split forms such as he/she or his/her can be found, e.g. in instructions to the reader, strategies aiming to neutralise the language by means of the singular they are less common. As a matter of fact, only two examples of they in singular use can be found, both of which in EB1 and in connection with everybody. This result can be explained by the fact that while in Lithuanian it is possible to name both male and female forms of a term to make the language more genderinclusive, the language does not allow for neutralisation due to the lack of neutral grammatical gender. Instead, all nouns and pronouns are either masculine or feminine, with grammatical gender also having major influence on other parts of speech, e.g. adjectives and participles (Ambrazas et al. 1997). As a consequence, they in singular use might have been considered confusing for young learners of English who are likely to be more used to the naming of both male and female terms.
This suggests that the findings of this study do not contradict Hellinger's (1990) claim that English speakers tend to prefer neutralisation strategies over splitting; they rather point out the extent to which the authors were influenced by their mother tongue rather than by developments in the target language.
In addition to the authors' preference for gendered professional terms such as postman or policeman over their gender-neutral alternatives discussed above, some uses of man were found that are Klein A., 2019. Cleaning mums and brave policemen: gender representation and genderinclusive language in EFL teaching materials in Lithuania, 12: 75-96 with person in order to include the possibility of a female narrator; by opting for man, however, the authors fail to acknowledge the fact that females can also succeed in life. Man is also used generically in two exercises in EB2. The first exercise asks students to choose the correct tense form of verbs in a short text dealing with climate change. The text concludes that -man is responsible for the problems Earth faces‖ (EB2: 99). In the second exercise, students have to read a poem based on the Native American saying that humankind will realise too late that one cannot eat money, and then make suggestions to save the planet using to, in order to, and so that. The poem, however, reads that -the white man‖ (EB2: 121) will understand too late that Earth cannot be saved anymore.
Although both texts are rather critical of the way humankind has been treating nature especially in the twentieth century, the authors here erase (cf. Irvin and Gal 2000) females' contribution to both the destruction of the environment and the realisation that we need to change our behaviour.

Conclusion
The aim of this study was to investigate the representation of males and females as well as the use of gender-inclusive language in four EFL exercise and test books published in Lithuania since 2005.
Overall, our findings are in line with the results of previous studies on gender representation and language use in teaching materials used in various countries all over the world: in terms of numbers females are less visible than males, and they are more often represented in stereotypical contexts.

This answers Research Question 1 on the numerical visibility of females and males, and Research
Question 2 on the contexts in which females and males are portrayed. As for Research Question 3 on the use of gender-inclusive language, the use of certain, yet not all available strategies could be observed: while the analysis of the authors' language use indicates that they were aware of issues related to gender-exclusive language and tried to avoid them by using certain strategies to make the language more gender-inclusive, my findings suggest that they were heavily influenced by the strategies that exist for the Lithuanian language, for instance, by the practice of naming both male and female pronouns when referring to an individual of unknown gender. The language used in the four exercise and test books analysed in this study can, therefore, be said to differ to some extent language. In his study on Singapore junior college students' opinion on and use of gender-inclusive language Jacobs found an -apparent presence of a ripple effect‖ (1999: 28) as a consequence of the spread of developments in the English language in countries of the so-called Inner Circle (cf. Kachru 1982) to countries of the Outer Circle; however, this study shows that the same cannot necessarily be observed with regard to countries of the Expanding Circle, e.g. Lithuania.
When it comes to the representation of males and females in teaching materials, it should not be ignored that the picture conveyed in the materials depends to a great extent on the choices of the authors. Authors have to weigh their options and then decide in how far the materials should be based upon generalisations and how much importance should be given to deviation from what is generally seen as the norm. Considering the young age of the learners, the authors of the four books analysed in this study clearly opted for the role models students at large would be most familiar and, therefore, most likely to identify with. At this point in time, we also need to distinguish between clearly outdated role models on the one hand, and ways of living less featured in society (e.g. homosexual parents) on the other hand. While authors of teaching materials should, by now, be aware of the issues related to clichéd representations of males and females, the question remains of to what extent less common alternative ways of living should be featured in teaching materials.
Arguably, there is a certain risk of over-emphasis of variation with the result that children might not recognise themselves in the characters. Therefore, authors of teaching materials should keep in mind recent changes in society and regularly update the picture they convey to the student. While, all in all, they should aim at a more balanced and, therefore, more realistic representation of males and females, they should also not ignore the importance of teaching children that difference is a positive feature.
This study set out to investigate gender representation not only in the underexplored context of Lithuania, but also in a type of teaching materials still largely ignored by researchers. The latter had a great impact on the design of this study. Counting all references to males and females rather than different characters proved to be the best way to map the visibility of males and females in numbers. With regard to the question of whether the teacher's mediation of the materials needs to be studied as well, I maintain that for exercise and test books, especially those that are accompanied by an answer key, the need to conduct classroom observations is lower than for textbooks. . Cleaning mums and brave policemen: gender representation and genderinclusive language in EFL teaching materials in Lithuania, 12: 75-96. https://taikomojikalbotyra.lt 91 However, the lack of classroom observation certainly remains a limitation to this study. In addition, students are increasingly exposed to books, movies, series and games in English outside the classroom. The influence of these media on the social and cognitive development of children should not be underestimated. Consequently, researchers interested in gender representation in teaching materials should keep in mind that these materials are just one of many resources of children's knowledge of the target language and culture.
Further research in the field of language and gender could be conducted in other countries belonging to the so-called Expanding Circle (cf. Kachru 1982). In this context, researchers could investigate if there is a correlation between the use of gender-inclusive English and the importance of grammatical gender in the respective local language: researchers might wish to focus on languages in which gender either plays a major role, e.g. Modern Hebrew, or a minor role, e.g. in genderless languages such as Finnish. Interesting results might also be yielded in countries in which there is already a debate on the gender-inclusiveness of the local language, e.g. in Sweden, since people will be more likely to be sensitised for the topic.