English Language Teaching Textbooks

Content, Consumption, Production

N. Harwood, N. Harwood, N. Harwood

• Publisher: Palgrave Macmillan
• Year: 2013
• ISBN:
  9781137276285
• Topic: Education
• Subtopic: Teaching Languages

English language teaching textbooks (or coursebooks) play a central role in the life of a classroom. This edited volume contains research-informed chapters focusing on: analysis of textbook content; how textbooks are used in the classroom; and textbook writers' accounts of the materials writing, design, and publishing process.

1

Content, Consumption, and Production: Three Levels of Textbook Research

Nigel Harwood

Introduction: The importance of textbook research

Like my previous edited volume (Harwood, 2010a), this book is intended for teachers, teacher trainers, researchers, publishers, and materials writers who work with English Language Teaching (ELT) textbooks. The remit of the earlier book was wider, focusing on teaching materials in general rather than textbooks in particular, and so many of the chapters in the previous volume described unpublished teaching activities produced by the authors themselves. For the present purposes, in contrast, the focus is squarely on published ELT textbooks (also known as coursebooks) and, where relevant, the aids that accompany them (such as teachers’ guides, workbooks, listening exercises, etc.). More specifically, much of the focus is on ‘global’ textbooks, normally published in the West and marketed worldwide, such as well-known series like Headway, Interchange, and Cutting Edge. While it is important to analyse unpublished, teacher-/researcher-produced materials, since no textbook can ever completely meet the needs of a class and, institutional and other factors permitting, teachers will wish to supplement their textbook with other materials to cater to their learners’ needs, it is also essential to focus on the published textbook, because most teachers are required to use them to some degree. An oft-cited statistic is Tyson and Woodward’s (1989) claim that textbooks structure up to 90 per cent of what goes on in school classrooms in the US. Whatever the figure in English as a Foreign Language/English as a Second Language (EFL/ESL) contexts, textbooks are similarly important: indeed, in many contexts, textbooks constitute the syllabus, teachers being expected to follow them more or less faithfully, with end-of-course exams being based exclusively on textbook content. Furthermore, existing textbook research has been criticized by various researchers for its lack of theoretical and methodological rigour (e.g., Harwood, 2010b; Tomlinson, 2012), and there is thus a need to extend and strengthen the research base in this area.

I argue it is important to study textbooks at three different levels – the levels of content, consumption, and production – drawing on du Gay, Hall, Janes, Mackay, and Negus (1997), and Gray (2010b) in differentiating thus. At the level of content, we can investigate what textbooks include and exclude in terms of topic, linguistic information, pedagogy, and culture. Unlike studies of content, which analyse textbooks outside the classroom context, at the level of consumption we can examine how teachers and learners use textbooks. Finally, at the level of production, we can investigate the processes by which textbooks are shaped, authored, and distributed, looking at textbook writers’ design processes, the affordances and constraints placed upon them by publishers, and the norms and values of the textbook industry as a whole.

Each of these dimensions is covered in a survey of the field below. I draw on research in ELT, but, as in my previous survey (Harwood, 2010b), I argue that textbook research is more developed, rigorous, and sophisticated in mainstream education (i.e., non-ELT fields, such as mathematics), particularly regarding textbook consumption, and that we have much to learn from the work in this area. I therefore include in the discussion below work from mainstream education with which readers may be less familiar. The present survey is intended to complement my earlier piece, and so I focus for the most part here on literature I did not discuss previously.

Textbook content

The obvious way for teachers and researchers to begin an investigation into a textbook is to determine and evaluate the subject matter which is included – and omitted. Researchers may prefer to focus on one particular content-related aspect of the textbook (such as treatment of a specific grammar point) or attempt an overall analysis and evaluation using a framework such as Littlejohn’s (2011). Below I have organized my review of content analyses around the headings of language, culture, and pragmatics.

Content analyses of language

Numerous studies evaluate the linguistic syllabus of textbooks by assessing the closeness of fit (or more commonly, lack of fit) between textbook language and the language of real life, as attested by corpora (e.g., Biber and Reppen, 2002; Conrad, 2004; Holmes, 1988; Lee, 2006; Miller, 2011; Mukundan and Khojasteh, 2011; Römer, 2005). One such study by Rühlemann (2009) analysed the treatment of reported speech in seven intermediate-level textbooks compared with British National Corpus (BNC) data, finding that a number of the most frequently used reporting verbs in real-life data were omitted from some of the textbooks, and that those verbs which were included differed widely between the books, suggesting that corpus information on frequency was not used by the writers when deciding what to include in their syllabus.

The textbook vocabulary syllabus has also been found wanting. Koprowski (2005) compared three textbooks’ treatment of lexical bundles in terms of frequency and range with data in the COBUILD Corpus, finding that more than 14 per cent (118) of the 822 bundles in the textbooks were absent from the corpus. Furthermore, not one bundle featured in all three textbooks. Gouverneur’s (2008) results were similar: analysing the phraseological treatment of the high-frequency verbs make and take in three intermediate and advanced textbooks, she found the books covered a varied selection of lexical phrases, with only 7 per cent and 15 per cent of make patterns appearing in all three advanced and intermediate textbooks respectively, and with not a single take collocation appearing in all advanced books. Findings such as these cause Gouverneur and Koprowski to question the criteria the textbook writers used to compile their vocabulary syllabuses. Koprowski argues that, although it may be an onerous task for the textbook writer to begin the development of lexical phrase textbook materials by consulting corpora, it should not be too much to expect textbook writers to check the frequency and range of the lexical phrases they are teaching when the materials are in draft form, refining their choice based on corpus evidence.

Other studies finding patchy treatment of vocabulary include Brown (2011). Whereas Nation (2001) describes nine different aspects of word knowledge, Brown found that ‘only three aspects consistently receive attention’ (p.88) in the textbook sample examined. Similarly, the textbook in focus in Criado (2009) was judged unsatisfactory in terms of the items included, with many of the most frequent words in English being absent, in terms of the frequency with which words are recycled, being too low to suggest acquisition would be likely, and in terms of the amount of words it is assumed students will learn as they progress through the book, which is far higher than research predicts. For their part, Matsuoka and Hirsh (2010) found ‘few opportunities’ to acquire vocabulary knowledge beyond the 2,000-word range in a best-selling textbook because of a lack of recycling (p.67); and Miller’s (2011) recent study of advanced-level, academically focused ESL reading textbooks concluded that the books’ vocabulary content was wanting.

An example of a content analysis focusing on a specific linguistic item is Lam’s (2009) comparison of 15 textbooks’ treatment of the discourse marker well with the use of well in a spoken corpus. Quantitative comparison revealed ‘major discrepancies’ (p.275) between textbook and corpus frequencies. And while well commonly occurs in either utterance-initial or medial position in the corpus, the textbooks give the impression that it occurs almost exclusively in utterance-initial position. Information is lacking in the textbooks about the various discourse functions of well, as are substantial, context-embedded examples.

An under-researched aspect of textbook content is pronunciation (but see Jones, 1997; Levis, 1999). Kopperoinen (2011) is a recent analysis of English as a Lingua Franca (ELF) pronunciation in two best-selling Finnish ELT textbook series. Kopperoinen studied all recordings and listening exercises, finding that outer/expanding circle accents accounted for between only 1–3 per cent of accents. In commenting on these results, Kopperoinen quotes Seidlhofer (2003:13), arguing that the outer/expanding circle speakers currently play a role of ‘exotic optional extras’ in the materials (p.84), despite the fact that most communication in English now takes place between second language (L2) speakers.

Content analyses of culture¹

There have been calls for textbook evaluation checklists to make cultural concerns more prominent (Cortazzi and Jin, 1999; Feng and Byram, 2002; Kullman, 2003): as the number of culturally focused content analyses has grown, so these analyses have criticized global textbooks for cultural inappropriacy, or at least inappropriacy when the materials are used in certain contexts (e.g., Canagarajah, 1993a, 1993b; Sokolik, 2007; Suaysuwan and Kapitzke, 2005; Yuen, 2011). For instance, ‘buying by credit card, ordering meals for delivery, and finding out snow conditions for skiing’ are seen by Auerbach and Burgess (1985: 479) as inappropriate lifestyle-related content in textbooks used by immigrants to the US and Canada. Auerbach and Burgess also point to what is absent, with no mention of typical issues likely to be experienced by immigrants, such as communication problems and difficulties finding employment, tying in with Gulliver’s (2010) analysis of textbook accounts of Canadian immigrants’ lives, which found the risk of failure is underplayed.

Focusing specifically on grammar textbooks, Sokolik (2007) concludes that these are helping to transmit, and tacitly approve, a culture of consumerism, as evidenced by example sentences from the books such as the following:

They go to Florida every summer.

My watch is new.

Maria wears a lot of jewelry.

I shelled out a lot of money on the diamond engagement ring that I bought for her.

I bought a new car last month.

She’s thinking about buying a new house.

And Boriboon (2004) points out that the provincial Thai learners he works with have very different social and cultural lifeworlds from those contemporary textbook characters cited by Sokolik, arguing that this may adversely affect the learners’ motivation and willingness to communicate. Boriboon illustrates his argument by taking a sample communicative activity from New Headway Intermediate (Soars and Soars, 1996: 45) which focuses on shopping and has the learners buy petrol, pay an electricity bill, and collect plane tickets, none of which his learners are likely to have experienced.

Two particularly detailed studies of global textbook cultural content are a PhD thesis by Kullman (2003) and a book by Gray (2010b). Focusing on 12 UK-published global textbooks written in the 1970s, 1980s, and 1990s and featuring interviews with eight British textbook writers (although not the authors of the textbooks analysed), Kullman found textbooks have become more ‘international’ in flavour, featuring characters and settings outside the UK. Rather than focus on other people’s lives, more recent books ask learners to talk about themselves, the tenor having shifted from the more ‘educational’ and ‘serious’ towards the more consumerist, emotional, and aspirational. Kullman also argues that some contemporary textbook topics will likely prove culturally inappropriate in certain contexts, singling out the treatment of ‘assertiveness’ in one book, which seemingly gives a straightforward message to learners that assertiveness is a useful attribute.

Examining four recent and not so recent best-selling intermediate-level textbooks, Gray (2010b) studies how textbooks have evolved in the cultural messages they transmit. He shows how the range of accents learners are exposed to moves away from mainly received pronunciation (RP) or modified RP in the older material to a more diverse range in the newer textbooks to include non-UK inner and outer/expanding circle speakers, but how all four textbooks tend to associate regional accents with characters in lower-status employment. With regards to the depiction of race and ethnicity, Gray finds a gradual progression towards multiculturalism and cosmopolitanism. He also notes the spectacular success of the fictional characters that learners encounter in the newer materials – characters who supposedly juggle jobs such as UN ambassador, film star, and best-selling author. Gray claims the discourse is consumerist and aspirational: characters are defined by freely available choices, and succeed in their choices apparently unproblematically. In subsequent research, Gray (2010a) analysed representations of the world of work, again finding that discourses of aspiration, success, and individual choice pervade the materials. Finally, Gray (2012) showed that the theme of celebrity was absent from materials until the late 1970s, since when it has become more and more apparent. The focus has shifted from an emphasis on celebrities’ achievements to their wealth, and the textbook activities in Gray’s dataset apparently hold up these celebrities for learners’ approval.

Gender and sexist bias have been studied extensively in ELT textbooks (e.g., Carroll and Kowitz, 1994; Jones, Kitetu, and Sunderland, 1997; Matsuno, 2002; McGrath, 2004; Ndura, 2004; Sherman, 2010; Sunderland, 2000). Sunderland (2000) and Matsuno (2002) specify how and where sexist content may occur, Matsuno’s classification being as follows: (i) in the omission or under-representation of females; (ii) in the depiction of females’ occupations; (iii) in stereotypical gender identities; and (iv) in sexist language items (chairman, houseman, etc.) (pp.84–85). Sunderland suggests that there are fewer content analyses of gender nowadays, perhaps because bias is less evident in contemporary materials, and Gray’s (2010b) study of textbooks’ treatment of gender supports Sunderland’s claims: in the older materials, men have a greater presence in the artwork and the listening tapescripts as well as in the textbook readings and practice dialogues. Women occupy subordinate positions, and are sometimes depicted as dependent on men and incapable of making decisions alone. In contrast, Gray finds the recent textbooks portray men and women more equally. Nonetheless, in their analysis of six textbooks, Carroll and Kowitz (1994) found that ‘the most important adjectives used to describe women are busy, beautiful, pretty and tall’ and that women are associated with ‘passivity, physical characteristics, menial roles, irrational worries, [and] constant undemanding activities’ (pp.79, 82). Other accounts of locally produced textbooks which have identified gender bias include Matsuno (2002), Sherman (2010), and McGrath (2004), who reports on a large study of 289 textbooks conducted in Hong Kong, where 71 per cent of some 32,000 gender-specific references were to males, and where women were sometimes stereotyped as weak and emotional.

When it comes to cultural representations in textbooks, however, McGrath (2004) raises the difficult issue of how ‘real’ textbook writers’ portrayals should be: ‘Is it the role of textbooks simply to reflect reality or to change it for (what we think of as) the better?’ (p.357). And, as we shall see below, when we focus on studies of textbook production, these choices are not always in the writers’ hands: publishers avoid including materials which may provoke controversy since this can impact upon sales figures or even result in a textbook being excluded from a state-approved list. In the end, though, however well intentioned or politically correct the message of the textbook content, there is no guarantee this message will be taken up: as Gray (2010b) puts it, teachers and students may ‘read against’ or resist the intended meaning or message of the textbook (p.26).

Content analyses of pragmatics

Attention to the pragmatics of English should not be regarded as an optional extra in a textbook syllabus, since pragmatic norms in one language or culture do not always transfer straightforwardly to the target language:

What is considered in one culture to be a normal amount of complimenting may seem excessive in another. What may be viewed as accepted topics of phatic communion (i.e., small talk) in one culture may be perceived negatively in another.

(Meier, 1997: 24)

The potential dangers of miscommunication because of pragmatic failure (see Thomas, 1983) are very real, then. However, some studies of textbooks’ handling of pragmatics conclude that treatment is ‘arbitrary’ and ‘oversimplistic’ (Meier, 1997: 24; see also Bardovi-Harlig, Hartford, Mahan-Taylor, Morgan, and Reynolds, 1991; Boxer and Pickering, 1995; Lee and Park, 2008; Millard, 2000; Nguyen, 2011; Wong, 2002), and that acquisition of pragmatic competence is ‘highly unlikely’ (Vellenga, 2004: 1) on the basis of the inadequate information textbooks provide.

One of the problems researchers find with textbooks’ treatment of pragmatics is that learners are often presented with insufficient context when the target language is introduced; hence it is difficult to appreciate how factors such as the relationship between speakers in a dialogue would influence what interlocutors say. Another problem concerns the choice of speech acts focused on: these can appear idiosyncratic, with two of the textbooks in Vellenga’s (2004) study teaching learners how to threaten, for instance, but not how to apologize. And a highly restricted set of linguistic items may be associated with a given speech act: in the same study by Vellenga, the only means of expressing making suggestions and giving advice is should. Vellenga’s study is particularly noteworthy in that it includes an examination of both teachers’ and students’ versions of the materials, with the teachers’ material also found wanting, containing ‘no metapragmatic information or extensions beyond what was provided in the textbook’ (p.14).

Nguyen’s (2011) study of the presentation of speech acts in EFL textbooks produced in Vietnam identifies problems with the type of language presented and how it is taught. The books teach bald on record language of disagreement (I completely disagree; That’s wrong, etc.: see Brown and Levinson 1987), which corpora suggest speakers largely avoid. It is therefore possible that

textbooks might mislead learners to falsely believe that English NSs [native speakers] tend to disagree more frequently and more directly than is the case, and that it is appropriate to use these unmitigated fo...

Table of contents

• Cover
• Title Page
• Copyright
• Contents
• List of Tables and Figures
• Acknowledgements
• Notes on Contributors
• 1. Content, Consumption, and Production: Three Levels of Textbook Research
• Part I: Studies of Textbook Content
• Part II: Studies of Textbook Consumption
• Part III: Studies of Textbook Production
• Author Index
• Subject Index