Language is a social practice that acts beyond surface linguistic features to construct identity in socio-cultural contexts (Bignold, 2015; Edwards, 2009; Harwood, 2005; Hyland, 2010; Ivanic, 1998; Norton, 2000). As a by-product of language use, writing is not quite impersonal (Hyland, 2002b; Ivanic, 1998). Authors do not only transmit knowledge. They express their experiences, beliefs, and ideas and then find answers to their queries (Hyland, 2002b, Petersen, 2003). In this sense, texts project writers' identities that are shaped in sociocultural contexts of their lives and revealed through their social activities, backgrounds, and choices of language elements, topics, and contents (Evans, 2015; Fairclough, 1993; Hyland, 2002a; Hyland, 2002b, Hyland, 2005; Ivanic, 1998; Ivanic & Camps, 2001; Norton, 2000; van Dijk, Ting-Toomey, Smitherman, & Troutman, 1997).
So far, many scholars have discussed identity in academic texts. Most of the studies have searched for ideologies in EFL and ELT textbooks. Some scholars (e.g., Amalsaleh, Javid, & Rahimi, 2010; Amerian & Esmaili, 2015; Asghari, 2011; Bahman & Rahimi, 2010; Baleghizadeh & Jamali Motahed, 2010; Basabe, 2004; Behnam & Mozaheb, 2013; Chao, 2011; Cortez, 2008; Hilliard, 2014; Khajavi & Abbasian, 2011; Roshan, 2014; Tajima, 2011) have used content analysis methods. Others (e.g., Pourhassan Moghaddam, Lotfi, & Haghverdi, 2013; Taki, 2008) have applied critical discourse analysis and few studies (e.g., Fitzgibbon, 2013; Tahririan & Sadri, 2013; Yassine, 2012) have used social semiotic approaches.
A group of scholars have conducted cross-generic studies on identity. Hyland (2011) focused on identity construction in thesis acknowledgments, doctoral prize applications, and bio statements. He found that the writers constructed their identities through goal seeking, affecting the audience's views, and learning and practicing rhetorical and academic values and features. Abdi (2009), further, conducted a cross-cultural study on English and Persian research articles to investigate how linguistic signals conveyed writers' cultural beliefs. His findings supported the influence of writers’ cultural environments on their choices of linguistic devices.
Other studies have focused on the significant roles of identity in academic contexts. Taylor, Busse, Gagova, Marsden, and Roosken (2013) investigated the relationship between identity perception and teaching and learning achievement across English as a foreign language and mathematics subjects in Bulgaria, Germany, the Netherlands, and Spain. They found positive relationships between the learners’ and the teachers’ identity perceptions and their performance levels. They emphasized the need for considering teachers’ and learners’ voices in academic contexts. In another study, Oliynyk (2013) analyzed modern reading textbooks and fictions to discover ideologies, and then focused on the teaching activities to bridge the existing ideological gaps. Gender, violence, divorce and social rejection were examples of the ideological topics discussed in that study.
Although many studies have focused on identity in academic texts, little research has investigated TEFL textbooks. The previous literature on identity has analyzed general English textbooks such as Interchange, Top Notch, Summit,and Comet in terms of the well-known social and cultural elements such as gender, social class, and geographical environment. These studies have viewed the writers as knowledge producers and neglected their role in the ideologically-driven contexts of meaning-making (Ivanic, 1998). A few studies have also investigated the importance of identity in foreign language contexts (Fujieda, 2013; Taylor
et al., 2013). More specifically, literature has paid less attention to writers’ identity. It has been because, first, when applying the social view, scholars have paid scant attention to authors’ identities. Second, identity studies have mainly focused on the readers' roles in constructing identity through writing, and they have considered the writers’ roles unproblematic. In other words, they have not viewed authors as an aspect of the context where identity is shaped (Ivanic, 1998). Finally, few studies have focused on identity development from a sociocultural perspective. This latter shortcoming is worthy of consideration because writers adopt social semiotic approaches to express their social and individual ideologies (Evans, 2015) in the broader contexts of culture and situation (Hyland, 2005).
This study, therefore, applied Systemic Functional Linguistics for text analysis to fill in the above gaps in identity studies. It was, further, a continual effort to discuss identity according to sociocultural accounts of development. The present investigation was significant as no other studies have functionally analyzed ideological identity in TEFL textbooks used in EFL contexts. With a special focus on language testing, language teaching, and linguistics textbooks, this study investigated the following question:
How authors of commonly used TEFL textbooks in EFL contexts construct and represent ideological identities?
Identity is one of the fundamental concepts affecting learning (Ivanic, 2006; Lave & Wenger, 1991). It reflects the relationship between an individual and the world (Hyland, 2010). Each person owns multiple identities based on the roles they play in society (Afful & Mwinlaaru, 2010; Desrochers, Andreassi, & Thompson, 2002). One’s multifaceted self includes a collection of counter-identities, as well (Burke, 1980; Cheng, 2015; Hall, 1997; Ivanic, 1998; Ivanic & Camps, 2001; Kramsch, 1998). Therefore, Ivanic (1998) preferred to use the term “identities” over “identity” for one’s self-hood and the simultaneous roles one plays in social contexts (Ivanic, 1998).
Regarding the nature of identity, scholars are divided into three categories. The first group (e.g., Edwards, 2009) defined identity based on similarities among people. They believed in the sameness of an individual in all contexts and situations and their assumption implied that one’s sense of self is fixed and stable. Another group (e.g., Brubaker & Cooper, 2000; Grier, 2007; Hall, 2000; Hallman, 2012; Ivanic, 2006; Norris, 2011) believed that identity is a fluid and continuous process, not a state of being. It is dynamic and ever-changing rather than static. Giroux and McLaren (1994) stated that one’s identity is not permanent; it is always switching from one facet to the other and is co-constructed based on the events and actors in a social-time-place. The third group (e.g., Evans, 2015; Hyland, 2003, 2010), however, put identity on a continuum. Hyland (2003) believed that one’s identity is not always changing by time. Quite the reverse, the core of one’s identity is consistent and stable that does not change over time.
Discussing how to determine identity, three groups of studies are discovered. The first category (e.g., Giddens, 1991) views identity as what people claim to be. To put it another way, people’s narratives and recounts exactly represent who they are. The second group (e.g., Helms, 1998; Goodson, 1992) believes that one’s identity is determined not only by what one does but also by what one perceives and how one thinks. According to Goodson (1992), one’s actions do not always reflect one’s attitudes; hence, in understanding people’s identities, one should consider not only their behaviors but also how they think and feel.
The third group (Hyland, 2003, 2010) are of the opinion that identity reflects people’s performances. From this perspective, most of the time people talk about who they want to be, not who they are. Hyland (2010) stated that identity is performance, but not merely narrating something. It is manifested in what people do. Hence, recounting things that mismatch people’s experiences does not give any information about their identities (Hyland, 2010). Hyland (2003), further, emphasized that identity is a repeated performance that makes an individual align himself with others in a social community. Examples include the clothes people wear, and the language and accent they use to communicate.
Evans (2015, p.16) introduced ‘sociocultural,’ ‘objective, rational mind’, and ‘existential self’ as different explanations of cultural life. In the sociocultural account, identity is a self-concept developed based on the social use of language. The objective, rational mind refers to identity as grammatical structures resulted from human reasoning. This view is similar to what Burke (2010) calls constructivism based on which people are active constructors of their selves. The process of identity construction is ongoing and continuous in the sense that no one feels and acts the same in all situations and in each context, an individual takes a new identity (Burke, 2010).
Finally, the existential view refers to identity as human self continuously interacting with the changing meanings of language. Given these accounts, identity consists of three levels. The first level is called ‘ideological identity’ associated with sociocultural discourse analysts such as Fairclough and Halliday. The second level is 'generic identity' supporting Chomsky’s and Descartes's accounts of grammatical language development, and the third level is ‘existential identity’ attributed to Heidegger and Derrida (Evans, 2015, p. 16).
Materials of this study were nine commonly applied TEFL textbooks used by Iranian state university instructors. They included Fundamental considerations in language testing by Bachman (1990), Language testing in practice: Designing and developing useful language tests by Bachman and Palmer (1996), Principles of language learning and teaching by Brown (2007), Techniques and principles in language teaching by Larsen-Freeman (2000), Understanding language teaching: From method to post method by Kumaravadivelu (2006), Approaches and methods in language teaching by Richards and Rodgers (2001), Testing language skills: From theory to practice by Farhady, Jafarpour, and Birjandi (2007), The study of language by Yule (2014), and An introduction to language by Fromkin, Rodman, and Hyams (2014).
The study corpus covered texts from language testing, language teaching, and linguistics subfields. Systemic functional text analysis allowed the researchers to analyze linguistic features at lexical, grammatical and discourse levels. The triad roles of language including representation, exchange, and message were considered to explore depictions of ideological identities in the corpora. Examining the texts qualitatively and quantitatively, the present researchers identified several representations of identity. This task was, further, facilitated by adopting Ivanic and Camps's (2001) interpretations of Halliday's SFL.
The present researchers coded the data up to the point of saturation where they could observe no further changes in the proportions of the examined elements. After three months, one of the researchers repeated the analyses and coded the data again to check the intra-rater reliability of the study. Cohen’s Kappa for analyses of the metafunctional elements was 0.81 with 83.3% percent agreement between the first and the second coding of the data.
Results and Discussion
This section presents the results obtained from examining the corpus:
Analysis of Ideational Meanings
Experiential elements including participants, processes, and circumstances were analyzed to explore representational identities in the corpus.
Analysis of the Participants
There were 11445 participants in the analyzed corpus. The textbook authors’ choice of pronouns was one way that could unfold their ideological positions. In Example 1, the author has used the plural pronoun “we” to highlight equal power division between the participants:
(1) “… so the argument goes, we could learn to operate vending machines without necessarily knowing language” (Yule, 2014, p. 19).
The pronoun ‘we’ in this sentence introduces the author and his audience as each others’ partners in performing acts of learning and knowing. The first person plural ‘we’ expresses the balance and equality of power between readers and the writer (Ivanic & Camps, 2001). Further, the use of personal pronouns may imply the author’s attempt to diminish the authorized voice and impersonal tradition of expository writing. Scholars have approved the roles of personal pronouns in constructing identity. For example, plural pronouns such as ‘we’ portray collective identity (Tornberg, 2000 cited in Pervan, 2011).
The writers’ choice among ‘she’, ‘he’ and ‘they’ for the third person singular participants, further, reflected ideological issues. In situations where the authors had solely used the pronoun ‘she’ for third person participants with unknown gender, they consciously or unconsciously expressed their bias against males. Similarly, using the pronoun ‘he’ for participants with unknown gender showed bias against females. However, the writers could use the pronoun ‘they’ to neutralize such gender-biased attitudes. Each of the above choices had roots in the writers’ membership to different social communities. The cultural resources available for the writers and their sociocultural contexts of life could determine their choices of linguistic elements (Ivanic & Camps, 2001).
In the present corpus, the authors had predominantly used gender-neutral words. Using the words “human”, “human being” and “individual” instead of “man” was indicative of Kumaravadivelu’s (2006) no belief in male gender’s superiority in (2):
(2) Individuals seek out and engage in intrinsically motivated activities in order to feel competent and self-determining. Like basic human drives, … (Kumaravadivelu, 2006, p. 40).
Introducing the teachers with female pronouns in even-numbered chapters and with male pronouns in odd-numbered chapters, Larson-Freeman (2000) also conveyed her gender-neutral attitudes towards possessing power and authority. Nevertheless, in most cases where she referred to the use of language teaching methods and approaches, she did not identify the students by names. In such cases, the pronouns and words such as “a girl” and “a boy” could help the readers identify the participants’ genders. Another example is chapter 5 of her book where Larson-Freeman (2000) identified learners, though not by names, by Portuguese as their nationality. Involving both male and female students in the reported classroom events, the author had transferred a sense of authenticity and her gender-neutral stance.
The authors’ attempts to avoid first-person pronouns showed their ideologies, as well. Apart from the dominant impersonal convention of expository writing, writers’ ideologies affected their personal pronoun use. Explicit use of first-person pronouns identified the authors as authorities in the issues they discussed. This use of pronouns was visible in many parts of the texts that transferred the attitudes of the authors. One could face many paragraphs where the authors had resorted to personal and especially plural first person pronoun ‘we’ mentioning their expert attitudes.
To reduce the subjective representation of issues, the writing experts suggest no or fewer use of personal pronouns in academic writing. In the examined texts, the authors had, at times, made choices of pronouns that complied with the academic writing conventions. Expressions such as ‘these authors’, ‘these researchers’, and ‘the present authors’ helped achieve this goal and convey authors’ attitudes and authorial presence in the texts. Additionally, authors’ bare claims explaining the concepts with no justifications or references indicated their self-assurance and certainty. Such claims connoted the objectivity of knowledge and universality of issues, as well.
There were other linguistic choices that could show divisions of power between the writers and readers. For example, citing other scholars as participants of the discourse events could hide authors’ identities and reduce their power:
(3) Celce-Murcia and Olshtain (2000) gave the definition quoted at the beginning of this section and repeated here for convenience: Discourse “is an instance of spoken or written language that has describable internal relationships of form and meaning (e.g., words, structures, cohesion) that relate coherently to an external communicative function or purpose and a given audience/interlocutor” (p. 4). Some discourse analysts (e.g., McCarthy & Carter, 1994) go beyond internal relationships of form … (p. xii), and some others (e.g., G. Cook, 1994) include “a form of knowledge of the world” (p. 24) as well. (Kumaravadivelu, 2006, p. 11)
The use of several sources in (3) has reduced the author’s authority and power. Attributing few or no sources, the authors could construct powerful images of themselves in their texts. This latter technique was frequently used by Larson-Freeman (2000). On this subject, one main reason for not projecting one’s authority in writing is feeling powerless and uncertain (Hyland, 2002a, Ivanic, 1998).
The textbook authors had also employed Halliday’s specified SFL elements to establish ‘authorial’, ‘discoursal’ and ‘autobiographical’ selves, the terms coined by Ivanic (1998). The discoursal self, communicating the authors’ shared experiences and activities with the readers, was discovered when the authors had used the pronoun ‘we’ in the subject position. At times, the authors had represented their authorial self by expressing their personal beliefs and opinions and ignoring what the literature stated. An example is (4):
(4) I began the Preface to this book with a Batesonian observation. I would like to end the Postscript with another.… This book represents my attempt to throw light upon the nature of the pattern…, and, as such, I may be pardoned for believing that it is “surely nontrivial” (Kumaravadivelu, 2006, p. 226).
Here, the author has used “This book represents my attempts…” and “I may be pardoned for believing that it is surely nontrivial” as signals showing his authorial self. In both of these expressions, personal pronouns ‘I’ and ‘my’ as the participants assisted the author to express his beliefs and accept the responsibility for the contents. Authorial self was further developed through authors’ cited sources as they could show those with whom the authors agreed and disagreed, thereby, specifying the authors’ lines of thought. These findings supported Ivanic’s (1998) belief that texts are not devoid of their authors’ beliefs and identities. Similarly, Hyland (2002b) maintained that writers can use personal pronouns to announce their authorial presence in the texts.
The writers can also use different lexical, grammatical and discursive signals to introduce their various roles such as a colleague, a student, a researcher, and a teacher which are, in fact, their social identities (Ivanic, 1998). Generic references identifying the individuals by their roles in the discourse events displayed writers’ attitudes, as well. An example is (5) where roles such as teacher or student helped recognize individuals:
(5) “… Student 11: Ten thirty p.m.
Student: Leaves Jolarpet at ten thirty . . .
Student 11: a.m.
Teacher: a.m. yes. Ten thirty a.m. correct . . . Now you have to listen carefully. For how long . . . for how long does it stop at Katpadi? How long is the stop in Katpadi . . . [indicates student 4].
Student 4: Five minutes.
Teacher: Five minutes, yes. How do you know?
Student X: Twenty . . .” (Kumaravadivelu, 2006, p. 151)
There were other instances where the lexical choice could show the textbook writers’ attitudes. Fromkin et al. (2014) had used this technique to divide participants into black and white. Their classification connoted racial attitudes as they used the terms ‘white child’ and ‘black child’ instead of neutral terms such as ‘human’ or ‘child’. Fewer instances of classificatory lexis, however, were present in Yule’s (2014) book. When referring to human beings, Yule (2014) had used the term ‘human’ instead of the generic male-specific word ‘man’.
The other technique reflecting writers’ opinions was depersonalization. In the previous example, the students were identified by numbers. In (6), as another example, alphabetical letters represent the participants:
(6) “A: Got the time?
B: Ten-fifteen.” (Brown, 2007, p. 226)
Using these techniques to distinguish the participants, the author could create a sense of de-realization in the readers’ minds. The readers, thus, may accept dialogues not as instances of authentic experiences, but rather as instruments mediating their understanding of the authors’ messages. These practices of identity establish the authorial self at the expense of hiding the participants’ identities. Excluding the participants’ names, the authors could hide aspects of identity such as nationality, culture, religion, age, attitudes, interests, and gender in the language exchanges. Further, linguistic choices established the writers’ authorial selves as the experiential processes that they frequently used revealed writers’ voices (Hyland, 2015).
Personal pronouns such as “we”, “you”, “your”, and “our” transfer the sense of personalization as in (7) and (8):
(7) … the next chapter, in which we present a model for the description, analysis, and comparison of methods. This model will be used as a framework for our subsequent discussions …(Richards & Rogers, 2001, p. 12).
(8) “You can now begin taking objects out of your box …” (Richards & Rogers, 2001, p. 40).
In these cases, first-person pronouns ‘I’ and ‘my’ showed authorized voice and personal attitude in writing (Hyland, 2010).The findings also confirmed Hyland’s (2002b) claims that using passive structures, and avoiding personal pronouns, writers can disguise their identities.
Many textbooks prescribe an objective and dry writing style with no signs of writer’s subjectivity in the text (Hyland, 2002b). One way to establish the impersonal academic stance is using passive voice (Ivanic & Camps, 2001) in which at least the authors do not refer to some of the participants. For example, in this study, the writers had used expressions such as ‘it can be concluded that’ instead of ‘I conclude that’.
An important issue is that first-person pronouns can show author identity only when they refer to the authors’ personal opinions not what is in the text (Hyland, 2002b). Author identity is on a continuum. The strongest form of author identity is when the author uses personal pronouns to talk about their own beliefs, and claims (Hyland, 2002b). When a personal pronoun is used to express the writer’s opinion, it has the highest authoritative identity than when, for example, it is used to guide the readers through the text (Hyland, 2002a).
Analysis of Processes
Verbal processes together with generic moves convey ideologies in an academic genre (Hyland, 2015). In Example (9), the verbs “perform”, “do”, and “ordering” are material processes:
(9) What do we do with language? and answered, simply: We perform speech acts. By speech acts, he refers to the everyday activity of informing, instructing, ordering, threatening, complaining, describing, and scores of other such activities for which we use our language (Kumaravadivelu, 2006, p. 10).
Here, “answered”, “refers”, “informing”, “instructing”, “threatening”, “complaining” and “describing” are typical examples of verbal processes as acts fulfilled only through expression.
The analyses showed that relational and behavioral processes were the most preferred ways to establish identity. According to Halliday and Matthiessen (2004), highly frequent use of relational and behavioral verbs in a text highlights the text author’s interest in psychological and physiological phenomena. Ivanic (1998) put this claim into other words and believed that dominant use of relational and behavioral processes shows author’s careful attention to the relationships between animate and inanimate entities.
The process types reflected the textbook authors’ views on knowledge making. Acts of having and being, expressed through relational verbs with high frequency, showed authors’ preferences for mentioning undeniable facts. Using these verbs, the authors had written with self-assurance and certainty (Ivanic & Camps, 2001). The knowledge-making processes differed across the analyzed texts. Example (10) shows Yule’s view on knowledge making:
(10) The word “discourse” is usually deﬁned as “language beyond the sentence” and so the analysis of discourse is typically concerned with the study of language in texts and conversation. In many of the preceding chapters, when we were concentrating on linguistic description, we were concerned with the accurate representation of the forms and structures. However, as language users, we are capable of more than simply recognizing correct versus incorrect forms and structures. We can cope with fragments in newspaper headlines such as Trains collide, two die, and know that what happened in the ﬁrst part was the cause of what happened in the second part. We can also make sense of notices like No shoes, no service, on shop windows in summer understanding that a conditional relation exists between the two parts (“If you are wearing no shoes, you will receive no service”). We have the ability to create complex discourse interpretations of fragmentary linguistic messages (Yule, 2014, p. 140).
In this extract, the present tense verbs “is”, “know”, “is defined”, “exist”, and “have”, express objective facts. The author has used categorical verbs with no auxiliaries to express his certainty and assurance toward the issues. Verbal groups such as “can cope with”, “can … make sense”, however; represent the author’s tentative stance (Ivanic & Camps, 2001). Further, the author has represented objective and universally accepted knowledge through relational verbs like “is” and “have”. By contrast, mental verbs such as “concentrate”, “make sense”, “recognize” and “concerned” show the author’s subjective position toward the stated facts. In this example, except for where the author has quoted the definition of discourse from someone else, in the other parts he dominates the discourse to establish his authorial voice. In the present corpus, there were many cases where the writers had communicated contradictory ideologies. In (10), for example, Yule (2014) has used mental processes to express his subjective ideas. Nevertheless, using passive verbs, with their special focus on action at the expense of identifying agents, conveys author’s view of knowledge as an objective enterprise.
The use of colloquial verbs and contractions was another way to express ideational attitudes that are usually found in the dialogues presented in the textbooks. Brown (2007), for example, had used informal conversation characterized by contractions:
(10) “He's following us.- I think you're right.
This is good coffee.- It's very good.” (Brown, 2007, p.54)
(11) “It’s the conversation that has to be got. However, if it can't be, I find that whatever a child chooses to make in the creative period may quite likely be such a word.” (Brown, 2007, p. 162)
In example 10, Brown (2007) has used “got” instead of “understand” accompanied by some contractions as a way to show the authenticity of the contents and intimacy with the readers.
In this study corpus, the writers had frequently resorted to relational verbs to explain the contents. This finding suggested the textbook authors’ more interest in general facts than subjective issues (Ivanic, 1998; Ivanic & Camps, 2001). Relational processes were also indicative of what authors claimed to be to identify themselves intentionally (Hyland & Tse, 2012). As Hyland and Tse (2012, p.157) maintained:
Such choices matter in identity performance. So for example, ‘she is interested in’ (a mental process), constructs the author as an active, thinking being exercising conscious choice in a research interest, whereas ‘her research interests are’ (a relational process) is more impersonal, downplaying the author’s role to highlight something that belongs to her.
In addition to reflecting who the participants claimed to be, relational processes showed what the participants possessed (Halliday & Matthiessen, 2004). The writers’ possessions in the analyzed corpus, based on Halliday and Matthiessen’s (2004) accounts, were their interests, experiences, and affiliations.
According to Ivanic and Camps (2001), writers’ linguistic choices shift them along the objective-subjective knowledge making continuum. The same was true in the analyzed corpus. As an indication of objectivity, present tense was the dominant verb tense in the corpus. Mental verbs indicating the processes of thinking and believing were fewer than relational processes. The authors had modalized some verbs (Ivanic & Camps, 2001) and they had sometimes made references to other scholars. On the other hand, to show knowledge was the product of human mental and empirical activities, the authors had used passive verbs without referring to themselves as agents of the processes.
As Ivanic and Camps (2001) and Hyland (2015) stated, voice and word types are an asset to discover identity in the texts. The analyses indicated the dominance of active voice representing personalization and inter-relational identity. There were, though, passive verbs and nominalizations manifesting impersonality that along the formal mood tag could indicate authors’ conservative sides towards the addressed issues. A typical example is below:
(11) We believe that the activities involved in design, operationalization, and administration need to be carried out for every test we develop. What differs from situation to situation is the amount of detail, resources, etc. involved. For some classroom tests, relatively few resources may be required. For example, suppose a course teacher wanted to prepare a short vocabulary quiz whose primary measurement use was for assigning course grades and whose instructional use was to encourage students to do their homework. It might be possible to plan the design statement rather quickly because at the outset there would be a fairly clear idea of the purpose, the TLU domain and tasks, the test takers, the construct to be measured, as well as how to evaluate the qualities of usefulness. Moreover, the teacher would probably do all of the actual writing, administration, analysis of results, and achieving her- or himself (Bachman & Palmer, 1996, p. 164).
In the excerpt (11), passive verbs and nominalizations show the authors’ impersonal style and their attempts to hide identities of the involved participants in the discourse events. A great number of the verbs are in passive voice. Examples include “involved”, “be carried out”, “be required”, and “be measured”. “Operationalization”, “administration”, “measurement”, and “analysis” are examples of nominalizations. By contrast, active verbs like “believe”, “develop”, “differs”, “suppose”, “wanted”, “was”, and “do” indicate the authors’ interest in identifying participants who had experienced different Hallidayan processes.
Analysis of Circumstances
Circumstances governing the textbook contents were varied. The evaluative words were usually used for describing discourse circumstances (Ivanic & Camps, 2001; Hyland, 2010):
(12) “There are several factors that contributed to this less-than-desirable outcome. First, language-centered pedagogists failed to recognize that superficial linguistic behavior...” (Kumaravadivelu, 2006, p. 110)
Here, the writer has used the verb “failed”, and the adjective “less-than-desirable” to personally evaluate the learning outcomes of the language-centered pedagogies and express his attitudes on them explicitly. The evaluative words could also construct authors’ authorial self in the texts:
(13) It is highly probable that, given his disappointment with the top–down concept of method, Stern had planned to delve more deeply into the role of the teacher visà vis his multidimensional framework. It is, indeed, a great loss to the profession that he could not complete his planned mission (Kumaravadivelu, 2006, p. 193)
In this example, the evaluative adverbs including “highly” and “deeply” and the adjective “great” show the author’s self-assurance and knowledge about the discussed issues.
Analysis of Interpersonal Meanings
Systems of mood and modality were analyzed to discover the projected identities in the texts.
The dominant mood in the analyzed corpus was declarative (90.8%) indicating some extents of writers’ authority in McCabe’s (2004) opinion. The imperative mood, further, personalized the social actions as unlike the rigid tradition of expository writing the texts addressed the readers. The interrogative and imperative sentences signified power relations between the writers and the readers, too (Ivanic & Camps, 2001).
Ivanic and Camps (2001) stated that interpersonal positioning indicates degrees of self-confidence and assurance, people’s perceptions about their authority, and their relations with their interlocutors. They believed that when writers express their views with self-assurance, they show their higher power than their readers. In the present corpus, declaratives expressing information and asserting concepts showed mild levels of authority, while imperatives signaled higher levels of authority. Interrogatives involving the readers in the discourse reflected equal levels of power between writers and readers. Also, mood and first person plural ‘we’ in the present corpus established power relations between the authors and readers (Ivanic & Camps, 2001). These results echoed scholars’ (Norris, 2011; Wodak, 1989) accounts concerning close connections between identity and power relations.
Analysis of Modality
Modality showed different levels of writers’ self-assurance and certainty (Ivanic & Camps, 2001). In line with Halliday and Matthiessen’s (2004) belief, frequent use of epistemic and deontic modalities legitimized author’s authority. For example, in (14), the author has communicated his certainty using the epistemic ‘must’:
(14) “… we must take into account considerations and follow procedures that are characteristic of tests and measurement in the social sciences in general” (Bachman, 1990, p. 18).
Modalities showed authors’ subjective ideas and their degrees of certainty (Karimi & Asadnia, 2014). Lack of modal expressions, however, indicated assertiveness (Rahimivand & Kuhi, 2014). To show their impersonal positions, the authors had sometimes modalized their language:
(15) “It is argued that communicative competence is so global that typical pragmatic tests cannot adequately gauge it” (Farhady et al., 2007, p. 169).
(16) “To select a passage for non-native speakers, it is recommended to first determine the content of …” (Farhady et al., 2007, p. 282).
In (15) and (16), using “it is argued that” and “it is recommended” instead of “we argue that” and “we recommend”, the authors have impersonalized and objectified their language through modalization. In some other instances, however, the authors had used opinion holders such as “I” and “we” to emphasize that the provided information came from their efforts.
Analysis of Textual Meanings
Analysis of Thematic Structures
One way to identify textbook authors was through their writing styles. Elements of Halliday’s textual metafunction unveiled writers’ stylistic preferences and dispositions. Through the analyses, the present researchers identified Hallidayan experiential, textual and interpersonal Themes in the texts. As Halliday and Matthiessen (2004) argued, simple and complex nominal groups, embedded clauses, adverbials, prepositional phrases, and complements function as the experiential or topical Theme. Example (17) shows embedded clause as Theme:
(17) “What it lacks is any structural organization” (Yule, 2014, p. 6).
The continuatives and connectives in the texts represented the textual Theme:
(18) “On the other hand, the illocutionary act is performed by or in rather than merely through using those words.” (Connective as Theme; Kumaravadivelu, 2006, p. 11).
Further, the vocatives and processes used at the beginning of the clauses expressed the interpersonal Theme:
(19) “Point to a mountain range in the West. What mountains are they?” (Process as Theme; Larson-Freeman, 2000, p. 25)
Analysis of Cohesive Elements
One way to represent the textual positioning was through cohesive devices (Ivanic & Camps, 2001). The writers established cohesion to guide the readers through the texts (Hyland, 2005). In so doing, they considered Given and New information, reference, substitution, ellipsis, and conjunction. The underlined elements in the following examples represent cohesive markers:
(20) “My own bias ... is to avoid use of the terms conscious and unconscious in second language theory.” (Establishing cohesion with ellipsis; Brown, 2007, p. 281)
(21) “Evaluation, therefore, does not necessarily entail testing.” (Establishing cohesion with conjunctions; Bachman, 1990, p. 22)
Additionally, anaphoric, cataphoric and exophoric references were elements establishing cohesion in the textbooks (Halliday & Matthiessen, 2004):
(22) “… I, as a nonnative speaker of English, do not have “permission” to coin a new word, and if I had coined one, it might have been corrected. It is unfair, but nevertheless true!” (Kumaravadivelu, 2006, p. 14)
In (22), “I” is a cataphoric reference co-referring with “a nonnative speaker”. Besides, showing the author’s preference for the personal writing style, the pronoun identifies the author in the community of non-native English writers. Another cataphoric reference is ‘one’ which co-refers with ‘it’ in the second sentence.
Within the textual realm, identity issues were expressed through Given and New information, as well. Given information refers to the early part of the sentence that provides common or already known knowledge. By contrast, the part that goes after and gives unknown information is called New information (Halliday & Matthiessen, 2004). The choice of which part goes first in a sentence depends on the writer’s prediction of the readers’ background and knowledge:
(23) “In Charles Darwin's vision of the origins of language, early humans had already developed musical ability prior to language and were using it "to charm each other”.” (Yule, 2014, p. 1)
The Given information, already conveyed in the discourse in (23), is “In Charles Darwin's vision of the origins of language.” When making this sentence, the author may have assumed that the readers already knew Charles Darwin as a scholar proposing the theory of evolution. As no previous text has supported this assumption, the writer himself may have chosen this structure based on his predictions of the readers’ familiarity with Charles Darwin. The parts “early humans had already developed musical ability prior to language” and “were using it “to charm each other” provided New information implying that the writer might have assumed that the readers had no or less knowledge of the presented issues.
Writers successfully guided the readers when they could identify who their readers were. This process incorporated many identification clues such as readers’ ages, backgrounds, expertise, and experiences. One manifestation of the identification process was organizing New and Given information. The author can guide and control the readers’ understanding by New information which is in the Rheme position where it is easier for the readers to process and understand it (Gosden, 1993; Hyland, 2002a). The Theme, however, is occupied by elements expressing the writers’ attitudes and attempts that prepare the readers for New information. Accordingly, evaluative expressions such as “it is important that…” are found in the thematic position (Hyland, 2010; Hyland, 2015). Hyland (2002a) stated that since authors typically use first-person pronouns in the thematic place, the source of the information in the Rheme, as the main part of an utterance, is the author himself. An example was example (22) where the thematic ‘I’ implied the author’s identification of himself in the Rheme position.
Apart from thematic and cohesive elements, some other textual components were further at work to unveil the textbook authors’ ideological identities. Textual ideologies refer to how the author develops a text. This type of positioning was determined based on the length of the phrases, word syllables, linking tools, and semiotic values (Ivanic & Camps, 2001). Kress (1993) and Starfield (2004) stated that adopting standard academic writing styles characterized by impersonal writing which, in turn, is represented by passive structures, nominalizations, modalizations, and suppressing personal pronouns indicates unequal power relations. The texts in this study belonged to different discourse types each specifically representing the authors’ voices. In (24), elements characterizing academic writing were to some extent present in each paragraph.
(24) This type of emotional reaction, or “affect,” may also be caused by dull textbooks, unpleasant classroom surroundings or an exhausting schedule of study and/or work. All these negative feelings or experiences are affective factors that can create a barrier to acquisition. Basically, if we are stressed, uncomfortable, self- conscious or unmotivated, we are unlikely to learn very much. In contrast, learners who have other personality traits, such as self-conﬁdence, low anxiety and a positive self-image, seem better able to overcome difﬁculties encountered in the learning space.
Children are generally less constrained by affective factors. Descriptions of L2 acquisition in childhood are full of instances where young children quickly overcome their inhibitions as they try to use new words and phrases. Adults can sometimes overcome their inhibitions too. In one intriguing study, a group of adult L2 learners volunteered to have their self-consciousness levels reduced by having their alcohol levels gradually increased. Up to a certain point, the pronunciation of the L2 noticeably improved, but after a certain number of drinks, as we might expect, pronunciations deteriorated rapidly. Courses introducing “French with cognac” or “Russian with vodka” may provide a partial solution, but the inhibitions are likely to return with sobriety. (Yule, 2014, p. 189)
In the first paragraph, the sentences are neither too long nor too short. There are few nominalizations. Functional items are three times more frequent than lexical ones (65 to 21). The cohesive devices, such as “or”, “and” and “in contrast”, have made the paragraph reader-oriented and have guided the readers’ understanding. The reader-considerate style of the writer is also felt because “affective factors”, as a new expression, is written in bold. In the second paragraph, however, the sentences are relatively long. There are some embedded clauses within the main clauses. Compared to the first paragraph, more nominalizations such as “acquisition”, “inhibition” and “pronunciation” are used. Cohesive devices such as “but” and “or” are used to orient the readers. Nevertheless, the text is writer-oriented as it does not include many cohesive devices.
Though generally expected to be ideologically rich due to the functional and meaning-based nature of language, the question of whether TEFL textbooks could impose any ideologies needed to be addressed. This study provided a thick description of types, forms, and functions of the communicated ideologies in the commonly applied textbooks in Iranian TEFL programs. The analysis was organized based on the major elements of each Hallidyan metafunction and unveiled the dominant ideological identities in the corpus. The combined results from discovering each of the metafunctions revealed that TEFL texts are not completely innocent. Whether intentionally or unintentionally, each text may suppress or promote some ideologies and control its readers’ mental behaviors.
The findings of this study will redound to the benefit of materials developers and textbook users considering that ideologies shape people’s understanding of the world. The ever developing world of educational arena calls for textbooks whereby learners’ cognitive and ideological development is ensured. Thus, the university professors who consider the hidden voices of the textbooks can be great assets to their students in liberating them and increasing their agency and power. Materials developers can also be guided on what should be transferred to the students and what is consistent with the demands of global and local contexts of learning. The results can also increase the students’ awareness of the existing hegemonies which are delicately conveyed through learning materials.
This study had some inevitable limitations. First, because of the multilayered and time-consuming nature of textual analysis, this study was narrowed down and conducted on the most commonly applied TEFL textbooks. Second, the textbooks were analyzed up to the point of data saturation; accordingly, not all textbook sections were scrutinized. Also, the time lapse between the first and second coding of the data might have affected the findings and interpretations.
Future studies can triangulate the present findings through questionnaires and interviews checking textbook users’ perceptions about the transferred ideologies. The researchers are also suggested to conduct cross-generic studies to look at the academic writers’ occluded and visible trends for promoting ideological identities. Another possible line of research is investigating the students’ and instructors’ disciplinary needs for increasing their awareness of the ideologies that surround them and influence the quality of their performance.