Abstract
This paper documents an initiative in introducing a genre-based approach to EFL teaching in the Japanese university context, in a third-year writing program. Its aim is to illustrate how and why such an approach may be extended to EFL teaching. Principles based on the learning theories of Lev Vygotsky are first outlined, before the notion of a Curriculum Cycle deriving from them is introduced. Consequently, a lesson unit based on this Cycle, as well as the process and experience of designing and implementing the course, is described. Particular attention is paid to the roles played by model texts as well as contextual and textual exploration involving both teachers and students in the co-construction of knowledge and skills. The place of a functional view of language - in particular, grammar - in the course of instruction will also be briefly discussed, as will the limitations and inadequacies of more prescriptivist approaches. Although the focus is on the teaching of writing, it will be contended that such an approach and its principles are applicable to the effective learning of other language skills.

Introduction: Genre-based ELT
Genre-based approaches, where teaching and learning focuses on the understanding and production of selected genres of texts, have been identified by Rodgers (2001) as a major trend in English language teaching (ELT) in the new millennium. Such approaches are, of course, not "new". English for Specific Purposes (ESP) and English for Academic Purposes (EAP) are early examples, arising from pioneering work in genre analysis by Swales (1981, 1990) and others. However, teaching and learning around text genres has become increasingly influential in mainstream ELT in a number of situations, including "primary, secondary, tertiary, professional and community teaching contexts" involving "native speakers of English as well as ESL and EFL learners", and "in countries as diverse as Singapore, South Africa, USA, Italy, Hong Kong, Australia, UK, China, Canada, Sweden and Thailand" (Derewianka, 2003). Nevertheless, their influence in EFL in East Asian countries such as Japan and Korea still appears limited, as a trawl through ELT-related journals in the region indicates. For instance, between 2001 and the present, one finds only one article (Kim & Kim, 2005), in the Asian EFL Journal discussing or alluding in a direct way to genre-based teaching in these countries.

This paper gives a brief overview of genre-based language teaching, and then documents one initiative in extending it to the East Asian EFL context, specifically within a Japanese university context. It describes the third-year writing program developed by the author while at the Nagoya University of Commerce and Business, and explains the theoretical and practical considerations involved. In so doing, the paper aims to suggest how and why genre-based teaching may be introduced in similar contexts.

Overview of genre-based approaches
Genre-based approaches have varied theoretical bases in linguistics, such as Rhetorical Structure Theory (RST) in North America (Mann & Thompson 1988), and Generic Structure Potential (GSP) theory in Australia (Halliday & Hasan 1989), in fields such as genre analysis and discourse analysis. They also take on various forms and guises. However, some key features are common to all of them.

To begin with, genre-based approaches begin with the whole text as the unit in focus rather than the sentence. The preoccupation is thus the creation of meaning at the level of discourse and not the acquisition of syntactical forms: "(r)ather than dealing with discrete instances of language, there is recognition that meaning accumulates and evolves over a stretch of text" (Derewianka 2003). Here, a "text" refers to "a piece of language in use", which is a "harmonious collection of meanings appropriate to its context" and hence has "unity of purpose" (Butt et al. 2001, p.3). In other words, texts are stretches of language that may be considered complete in themselves as acts of social exchange. Length and mode of communication are immaterial: each text may be long or short, written or spoken. Thus, a brief exchange of greetings as two acquaintances pass each other is as much a text as is a 600-page novel.

Closely related to this, genre-based approaches are concerned with the social macro-purposes of language, and not just the semantic micro-functions of individual words and sentences: the genres in focus are generally defined according to the broad social purposes of communication. The classification and labeling of genres may vary, depending, among other things, on the theoretical influences behind each approach. For example, in some instances, writing genres are defined in terms of familiar broad categories such as 'Narratives', 'Description', 'Persuasion and Argumentation' and so on. Another approach, elaborated on later, makes a distinction between six or so text prototypes called text types, and more specific genres that employ each or combinations of these text types. Whatever the differences, categorization is based on what the discourse seeks to achieve or to do socially, for example, to tell a story ('Narratives' in many typologies) or to argue an opinion ('Argument' in some typologies, 'Exposition' in others).

Finally, the focus on whole texts implies recognition that there is a higher level of order and patterning in language than just the sentence - grammar at the level of discourse organization and meta-patterning of grammatical features. Genre-based approaches emphasize that this higher order must be attended to for effective language use: "all texts conform to certain conventions, and that if a student is to be successful in joining a particular English-language discourse community, the student will need to be able to produce texts which fulfill the expectations of its readers in regards to grammar, organization, and context" (Kim & Kim 2005, citing Muncie 2002). It must be noted that sentence-level grammar is not seen as unimportant: rather, its importance is seen in terms of the part it plays in the overall patterning of whole texts (e.g. what sorts of sentence patterns tend to pre-dominate in a particular genre). Indeed, close attention is paid to sentence- and word-level grammar in many current approaches, but without such grammar being treated separately from the business of communication, unlike in older grammar-focused approaches or in many forms of communicative language teaching. Thus, genre-based approaches can be seen as being at once both whole-to-part and part to whole.

Third-year English Writing Courses at the Nagoya University of Commerce and Business
The third-year English Writing courses at the Nagoya University of Commerce and Business (NUCB), titled English Writing V and English Writing VI for the first and second semester respectively, exemplify a genre-based framework in its curricular design and pedagogical focus. Developed over the academic year April 2005 to March 2006, it is now a fully formed program, with explicit goals specified in terms of genre outcomes and fully written teaching and learning materials following a specific and consistent pedagogical pattern.

The specification of genres to be taught is based on the classification used by many systemic functional linguists, especially in applications to classroom teaching of English (e.g. Butt et al 2001, Derewianka 1990). The classification involves a distinction between text types and genres. Text types refer to text prototypes defined according to their primary social purposes, and six main text types may be identified:

Genres, on the other hand, refer to more specific classes of texts, such as newspaper reports or recipes. Texts of each genre may be purely of one text-type (for example, a bus schedule is purely an Information Report, while most recipes are purely of the text type 'Instructions') or they may be a blend (for example, sermons often include stretches of narratives or recounts, as well as explanations, while usually expository in intent). Genres may also be written or spoken.

The rationale for adopting a genre-based framework is that it facilitates clear links to the students' purposes for writing beyond the writing classroom. Thus, the primary factors in curricular selection are ensuring a balance of text types, to enable students to perform a broad range of social purposes for writing in English in future, and selection of specific genres based on the students' most immediate academic needs. With respect to the latter, the programme seeks to support the writing of a graduation essay, an extended research-based thesis that is a university-wide requirement for graduation, and that, for the English majors, is the most major piece of writing in English demanded of them. Since the research essay is a genre that potentially incorporates sections in its various sections all the main text types apart from 'Narratives', the curriculum covers all the text types except 'Narratives', which is also excluded because few students are likely to ever need to tell stories in English. Thus, the six units covered (three in each semester) comprise the following:

It might be noted that the specific genres realizing each text type are, where possible, those that approximate potential sections of the graduation essay. For example, one genre of 'Information Reports' selected, Surveys and Interview Reports, can evidently be adapted for the 'Findings' section, while the specific genre of 'Recounts', Company Histories, might be used by students doing research on a business in the 'Background' section of their essays. These possible applications to their graduation essays are made explicit to the students in the introduction to each unit.

The teaching and learning material for each unit are fully written up in a form akin to textbook units, printed, and distributed to the students. They are also made available online through the university's Blackboard Learning System. Each unit follows a similar pattern and learning cycle, and is completed over three to four 90- or 100- minute lessons. The principles on which this cycle is based, as well as a description of what a unit of learning includes, is described in the sections that follow.

To supplement the work done in the genre-based units, students are also required to submit a one-page response or comment on newspaper articles of their choice once a fortnight. This is to provide sustained opportunities for free writing by the students.

Vygotskian principles & pedagogical application
Genre-based ELT may be allied with a variety of pedagogical practices, and there is no single teaching/learning approach associated with it (Derewianka 2003). However, genre-based teaching runs the inherent risk of becoming (and has indeed sometimes become) overly product-focused in a prescriptive way, since the curriculum is usually defined in terms of products - texts in various genres. It may thus "undervalue skills needed to produce a text, and see the learners as largely passive" (Badger & White 2000). As Kim and Kim (2005) argue, good genre-based teaching needs to incorporate a process focus as well, so that it is at once both product- and process-oriented. The socio-cultural learning theories of Vygotsky (1978) provide a basis for this, and have been applied in a number of contexts, for example, in Australian educational contexts, as noted by Derewianka (2003). The design of the teaching material in the NUCB third-year courses is also based on Vygotskian principles.

Much of ELT in the last few decades has been based on largely cognitivist theories of Second Language Acquisition (or SLA as the field is commonly known) that view learning as pre-dominantly psychological in nature, dependant largely on what happens in the mind. For Vygotsky, however, learning is both social and psychological in nature. Mediation through the use of tools plays the key role at every point, with sign systems being the primary tools, and language being the key sign system, and hence, key tool for mediation. Vygotksy proposes the notion of Zones of Proximal Development (ZPDs), zones between what he calls 'actual' development (what the learner can do independently) and 'potential' development (what the learner can do in the future, with the help of others now). Every act of learning occurs within a ZPD, building on what the learner already knows and can do, and is first inter-psychological (social) before it is intra-psychological (psychological). Learning begins by being object-regulated, and then is others-regulated, before it is self-regulated. As the stages of object-regulation, others-regulation, and self-regulation are crucial to understanding the pedagogical design of the NUCB third year courses, they are further elaborated below.

By object-regulation, Vygotsky was referring to the role played by concrete manifestations of culture in the environment - objects and artifacts, rituals, routines and daily practices, documents and valued texts, and so on - that function as sign systems that mediate learning. The learner's starting point is thus social, in the first place, because he or she begins by taking cues from the environment. Thus, the value of some rituals, for instance, is that they enact in physical terms values of central importance to the culture, for example, filial piety through the ritual of offering tea to one's elders, and thus form a visible means of transmitting those values to succeeding generations. Children's playground activities, in the Vygotskian perspective, are also of value not so much because they provide the children opportunities to manipulate, explore and discover the environment, as in Piaget's view of human development, but more because the role-playing that often dominates such activities is a form of object-regulation of the child's understanding of his or her immediate society.

One's potential development, however, cannot be manifested, however, if learning stops at object-regulation. The key to such a manifestation is the role played by significant others in mediating learning - the stage of others-regulation. Such significant others may include, for instance, parents, elders and teachers, as well as more expert peers, who through talk and other means provide explicit or conscious as well as implicit or unconscious guidance to the learner. Returning to the examples of rituals and playground activities, this guidance may take the form, for instance, of explanations of the meanings of rituals, or of a child with wider experience telling another his or her version of how, for instance, a real doctor would act. It is at the stage of others-regulation that language becomes important, not only facilitating the transactions between 'expert' and learner, but also enabling key concepts to be captured and retained.

For the potential development manifested by what the learner is able to do with the help of others to be translated eventually into actual development, self-regulation is vital. This is the stage in which the learner processes and manipulates by himself or herself the knowledge and understanding gained, and/or begins to be capable of working independently. As opposed to the Piagetian view of the phenomenon of self-talk by young children as indicative of ego-centricity, whose reduction as the child becomes increasingly aware of others is what is responsible for its eventual disappearance, self-talk is seen by Vygotsky as a manifestation of self-regulation. Its disappearance as children grow up is simply a matter of it becoming internalized and silent: this internal self-talk, again made possible only because of language, continues to be a vital part of learning through self-regulation.

Vygotsky's ideas on learning have been operationalized in genre-based ELT through the notion of the Curriculum Cycle, proposed by systemic-functional linguists such as Derewianka (1990) and Butt et al (2001) and influential in school settings in New South Wales and other parts of Australia, as well as in Singapore. This is a simple model for developing complete lesson units or cycles around text types or genres to be taught, and has as its ultimate aims helping learners to do things with language independently through mastery of text types and genres.

Each lesson unit or cycle has as its central focus a chosen text type or genre, and consists of a fixed sequence of stages. Descriptions of the cycle (e.g. in Derewianka, 1990 & Butt et al, 2001) vary in minor ways, but four phases essential for developing control of a genre may be identified, namely:

Figure 1: The Curriculum Cycle

Every cycle begins with context exploration, 'context' referring to the possible contexts of situation in which the chosen text-type or genre may be used. This phase resembles the pre-listening/reading/speaking/writing phase that has come to be typical in communicative language teaching, and the activities that may be carried out do indeed resemble typical pre-activities in skills-based teaching. However, where traditional genre-based activities have aims ranging from mere warming up to activation of mental schema, one primary aim of the genre-based Curriculum Cycle is to help students to become aware of and understand the social purpose of the chosen genre, as well as other contextual factors influencing the production of the texts that they will examine as models, and the texts that they may be required to produce in speech or writing. Based on Vygotskian principles, another important aim of the context exploration phase, from the teacher's point of view, is to establish the learners' 'actual development' or starting point.

The next stage, text exploration, is the first of two perhaps distinctive key phases in the Curriculum Cycle that demonstrates how a genre-based approach founded on Vygotskian principles is set apart from other forms of communicative language teaching. The aims of this phase are to familiarize the learners with the target text-type or genre, and to draw attention to organizational and linguistic features commonly found in texts belonging to it. Model texts play a crucial role in this phase, providing, in Vygotsky's terms, the necessary object-regulation. Using such model texts, pedagogical activities to make explicit the features of the text-type are carried out. These may include a gamut of established 'communicative' activities, such as the re-assembling of 'jigsaw' texts or information gap exercises, but the tasks are deliberately constructed in such a way as to highlight the salient lexical and grammatical features. Thus, the tasks aim to be at least implicitly 'analytical' in nature, and not just to facilitate interaction as an end in itself. Of course, more explicitly analytical work is also possible: for example, students may be asked to 'hunt' for and highlight all instances of a specific grammatical form. Direct teaching by the instructor is also an option, in order to make the features obvious to the learners. How the formal features work to help the text-type or genre achieve its purposes are also discussed or explored, the teacher playing a key role in others-regulation throughout this phase.

Others-regulation continues and takes centre-stage in the next stage, joint construction. Here, referring to the model text or texts, and making use of the knowledge and awareness gained from the exploration of the text, the students work with the teacher to construct their own texts (spoken or written) in the text-type or genre (or, in the case of listening and reading, to construct an understanding of the new text). This can take forms varying from teacher-fronted whole-class co-construction of a single text on the board, to small-group or pair construction with the teacher helping each group or pair by turn, to teacher conferencing with individual students. In the case of writing, as with process approaches, the texts may go through a few rounds of drafting, editing, and re-drafting. The model texts continue to provide object-regulation, while others-regulation comes from not only the teacher but also from other students, as more expert peers guide others, or as students refer each other to features in the models, and to points raised in the text exploration stage.

What is to be noted in both the text exploration and joint construction phases is that while there is much oral interaction taking place, its nature and intention is different from that of most forms of communicative language teaching. Where the interactive activities in the latter are often designed to simulate real life interaction, directed at providing opportunities for talking in the language, the talk here is about using language, and is focused on a collaborative effort to learn to accomplish a purpose in the language.

The last stage in the Cycle, individual application, as the name suggests, requires learners to work individually and independently, for example, in the case of writing, to produce individual essays. Ideally, this is carried out only after the students have successfully produced a jointly constructed text or understanding of a text. This phase then provides the opportunity for self-regulation, the crucial final stage in Vygotsky's model of learning. What each learner produces can, of course, be further re-cycled through further others-regulation (e.g. peer editing, teacher feedback), until the learner attains a desired level of attainment.

Lesson unit example
Each lesson unit in the NUCB third-year courses is constructed based on the Curriculum Cycle model. As an example, the first unit in the first semester, 'Survey and Interview Reports', is described.

The entire lesson unit is completed over the course of three to four 90-minute lessons, the pace depending on the students' progress, and was designed to serve secondary purposes of allowing the instructors insights into the students' attitudes to and extent of writing in English, and getting the students to reflect on their own attitudes and learning strategies.

The Context Exploration stage that begins the unit opens with the students working in small groups to simulate a survey. Using a prepared questionnaire on 'Attitudes to Writing' (see Figure 2 below) provided to them, the students interview each other, collate the results in their groups (calculating the percentages for each response), and then report what they think are the most significant figures, through appointed group representatives. Still working in their groups, the students then discuss and write down answers to the following questions: Why did your teacher ask you to do this survey? Who would be interested in knowing the results? Why? Where are some places where the results can be reported? What are the most important points to include? On the basis of their answers the instructor then explores with the class why surveys are carried out, draws attention to the fact that the results need usually to be reported in writing, and then discusses possible purposes and audiences that survey reports may be written for.

Survey: ATTITUDES TO WRITING

Name of Interviewer: ______________________________________

Name of Respondent: ______________________________________

Date of Interview: _________________________________________

Q 1: When do you write in English?

Figure 2: Extract from Survey Questionnaire Used by Students

Now prepared for the Text Exploration phase, the students are directed to read the model text 'Reading Materials in the SAC (Self-Access Centre)' found in Appendix A. Salient features of text organization, grammar and vocabulary are highlighted in the margins of the model text in this unit, but in later units, students are gradually guided to make similar notes on the model texts. This is to encourage students to learn to observe texts and learn on their own from their reading. Before the students proceed to examine the textual features more closely, the instructor ensures that they understand the content of the text, and the students then write answers to questions in the unit to reflect on the purpose and audience of the report, discussing this subsequently as a class.

Text exploration begins in earnest with the organization of the text. A jigsaw activity on another survey report similar to the model text is used to heighten the students' awareness of the possible parts of survey reports, and how they may be ordered. Once the students have re-assembled the new report, they are asked to compare it with the model text, and identify the organizational features common to both. These features include both the macro-structure of the texts, as well as organization of content at the level of the paragraph.

Exploration then moves downwards to the lexico-grammar, that is, both grammar at the level of cohesive devices, the sentence, and the word, as well as vocabulary central to the text. One of the features explored in this unit is the use of reported speech, which is often prominent in survey reports. The instructor gives examples of reported speech sentences, and teaches the basic syntax of reported speech sentences explicitly. Following an example by the instructor, the students are then directed to look for and highlight all instances of reported speech in the model text, marking out the 'reporting' clauses (e.g. "80% of the students said…") and the 'reported' clauses (e.g. "…that they could not understand most of the magazines and articles") clearly. This draws their attention to the common use of reported speech in the genre, as well as provides them further examples of the syntax of reported speech sentences. The students are also further asked to circle all the reporting verbs (i.e. verbs used in the 'reporting' clause) and classify them as either 'Saying' (e.g. "said", "explained") or 'Thinking/Feeling' (e.g. "felt", "thought") verbs. Thus, the vocabulary explored here is that of possible reporting verbs.

After this is done, the instructor leads a discussion on the functions of the observed features - that is, how do they help the writer to achieve his or her purposes in the text? For example, why do writers of survey reports use reported speech? Why are some findings reported using 'Saying' verbs, and others using 'Thinking/Feeling' verbs? It is important to note that the discussion should explore possibilities rather than make dogmatic interpretations as to the writer's intentions. What it seeks is to show that the grammar used is a motivated choice to help achieve communicative purposes, rather than a matter of prescription for the genre. Reported speech, and the range of reporting verbs, are, in other words, a set of resources that writers of survey reports can use. To help the students master this possible resource, they are then give sentence writing practice, using concocted figures for some of the items on the 'Attitudes to Writing' questionnaire used in the context exploration phase. Figure 3 shows part of this exercise:

Apart from reported speech and reporting verbs, quantifiers for reporting numbers (e.g. "Most of the students", "80% of the students") as well as vocabulary often found in most survey reports are also explicitly highlighted and taught. The instructor can, of course, employ a variety of different activities and tasks for this, and has discretion as to the range of lexico-grammatical features to cover, but always, the teaching of the grammar and vocabulary is related to their function and use in the genre. In addition, the meta-language or grammatical terminology is also always introduced and used, as this facilitates discussion both during this phase and during the collaborative work in the joint construction stage. The students can also be asked, as homework, to look at more examples of survey reports (either provided by the instructor or to be found by themselves), and take down more examples of the various lexico-grammatical features at work.

Equipped now with an awareness of the organizational possibilities and lexico-grammatical resources at their disposal, the students are now prepared to carry out joint construction of their own survey reports. This they do in pairs or small groups, using their findings from the mini-survey in the context exploration stage. Alternatively, each student may write his or her first draft individually, then work with a partner to improve this draft. To provide scaffolding, the students are encouraged to constantly refer to the model text and their grammar work as they write, and the instructor plays an active role by circulating around the classroom and guiding the students in turn or when they are in need, reminding them constantly about the text organization, appropriate use of reported speech, how they can vary the reporting verbs, and so on. Each piece of writing goes through more than one draft, before it is submitted to the instructor, who then provides further feedback for a final draft to be submitted for grading.

Following the Curriculum Cycle, the students should then be given an opportunity for individual application. This can be done through students carrying their individual mini-surveys on a topic of their choice, using their own questionnaires, then writing individual reports. However, due to time constraints, the individual application is not carried out within the writing course. Instead, the course assumes that this is carried out in the students' writing of their graduation essay, which may include a section involving a survey. As mentioned, the possible application to their graduation essays is made explicit by the instructor, who may, in addition, advise individual students how the learning in the unit may be used. Efforts are also made to inform instructors supervising the graduation essays of the work done in this unit in the writing class.

A functional view of grammar
As can be from the description above, grammar instruction is an integral part of the teaching of the genre. The instruction not only is explicit, but also involves some degree of student analysis. However, it differs from more traditional forms of grammar instruction in some very fundamental ways.

To begin with, grammar is not taught in isolation from the communicative functions of the language. It is also neither taught as an end in itself, nor as the central focus of instruction, but as a means to an end - achieving the social purposes of the genre that is the overarching focus of teaching. The most traditional forms of grammar teaching focus on sentence analysis and the learning of prescriptive grammar rules, with the emphasis on producing 'correct' form at the level of the sentence, and little reference to the communicative functions of language. The structuralist approaches that developed in the late 1950's and early 1960's, emphasized sentence pattern drills, with less or no recourse to explicit teaching of 'rules'. However, while based on more descriptive linguistics, they continue this formal focus and emphasis on accuracy, although later forms of these approaches include attempts at fairly artificial 'communicative' practice around the chosen patterns. While of a broad church, the communicative approaches that followed, on the other hand, generally focus on the use of language, emphasizing communicative practice, with grammar either believed to be just learned implicitly in the course of receiving language 'input' or through interaction, or taught as a separate matter. This is reflected in the ongoing debates that still persist today in ELT and SLA (second language acquisition) literature about what the balance should be between 'focus on form' and 'focus on function'. The genre-based material in the NUCB course, on the other hand, views grammatical form as intrinsically linked to communicative function, and seeks to help students understand the links between particular aspects of English grammar and their functions. Moreover, the focus does not remain focused at sentence level, but moves beyond to how sentence-level patterns fit into natural whole texts in communication. In all these senses, the teaching of grammar is 'functional'.

The approach to grammar is influenced by the systemic functional linguistics of Michael Halliday and his associates, whose model of grammar is most definitively described in Halliday (1994) and Halliday and Matthiesen (2004). Butt et al (2000), amongst others, have also produced more teacher-oriented descriptions of the grammar. Essentially, the theory sees language as a system of resources for making meaning in context. Thus, the grammar of a language, or what systemic functional linguists call the 'lexico-grammar', is seen not merely as a set of rules or patterns of syntax, but a set of agreed-on conventions intimately related to possible meanings, determined by contexts of situation embedded in contexts of culture in which language is used.

To know a language then, is to know the set of grammatical resources - e.g. the range of sentence patterns or word-level grammatical features - that one may choose from, in a particular situation, to achieve a particular purpose. Thus, what is most important to know and teach about grammatical patterns or features are their functions and use in discourse - their meaning potential, what each can achieve in communication, in what contexts, both situational and cultural, and for what purposes, and, conversely, when and where each might be used. While accuracy of form remains a legitimate concern, the more critical focus then is to help learners develop knowledge of the range of specific grammatical resources that they may employ for particular tasks. This is illustrated in the focus on reported speech in the lesson unit described earlier: reported speech is often a resource used in such reports to attribute opinions to others (the respondents to the survey) in order to emphasize objectivity.

It should also be noted that the grammar instruction within the approach described does not seek to be comprehensive - what is taught is dependent on what is needed for the genres selected. Instructors are encouraged in their feedback to students' writing to focus mainly on the use of the features highlighted in each lesson unit, and aim at helping students master these aspects, rather than try to attend to all aspects of the students' grammar in their writing. Assessment criteria in mid- and end-of-semester examinations also focus on these as well as the larger organizational aspects taught, and this is made known to the students.

The tasks and exercises for grammar learning draw on established methods and techniques, in recognition that they have their place and worth. For instance, in some units, substitution tables to practice particular sentence patterns are used. This, of course, draws on a technique influenced by structuralist approaches, the difference being that the sentence patterns are now learned in the contexts of their natural use in particular genres, and the aim is for the students to put them into immediate use in producing writing in the selected genres. In addition, teachers also explain the functions of such patterns in achieving the purposes of the genres.

Finally, the instruction ideally aims at emphasizing that most of the grammatical features taught are likely and very useful, but not mandatory, resources for each genre. However, given that the students are relatively low-level EFL learners, there is insistence that they use the features taught in each unit in the production of their own texts. This is less out of dogmatic prescriptiveness, then of pragmatic consideration that the learners would otherwise not be in possession of appropriate grammatical resources. Moreover, for those who do possess other grammatical resources for achieving the genre's purposes, the insistence extends the range of resources they are capable of using. Nevertheless, what is emphasized in the course of instruction is that learning the grammatical features will help them produce writing in the target genre, rather than that failure to use the features constitutes wrong writing in itself.

Conclusion
Thus far, the courses have affirmed many of the claims of the merits of genre-based approaches, found, for example, in Paltridge (2001: 7-10). The students, in their assignments and examinations, produce good coherent writing in the genres taught, usually relying on the model texts provided. The writing of the best students even under examination conditions is impressive, showing ability to adapt the features learnt for their own communicative purposes. Informal feedback from the students and instructors involved has been very positive. Students have reported that the third-year programme has given them more confidence and security in their own abilities to produce independent writing than their previous writing courses have. In particular, they point to the use of model texts and the teaching of particular grammatical features as very helpful in knowing how to go about writing texts of similar genres. Some have reported that specific learning units have helped them in writing their graduation essays, showing their ability to transfer the learning to other contexts. The instructors, on the other hand, appreciated the consistency of the pedagogical cycle, which at the same time allowed for a variety of tasks and activities. They also expressed much satisfaction at the visible progress of the students towards producing relatively competent writing in the genres taught.

While genre-based teaching is most readily applicable to writing instruction, and has most extensively been used for it, it may be and has been applied to the teaching of listening, speaking and reading, following a similar pattern to that described in this article. Paltridge (2001) discusses and demonstrates the ways in which this may done. In terms of teaching conversation, for example (ibid, pp.33-40), he suggests that it may be based on work done by Eggins and Slade (1997), which demonstrates that casual conversation in English does follow schematic structures and generic patterns of openings and closings, turn taking, topic management and such, realized through particular grammatical features in the same way as written genres. Through using the Curriculum Cycle model, students may be helped in a similar way to manage particular genres of spoken conversation.

There are, of course, potential difficulties and limitations associated with genre-based approaches. Some of these are highlighted in various chapters in Paltridge (2001), and illustrated in Lin (2003), in relation to the implementation of the genre-based English Language syllabus in schools in Singapore. Among these is the potential for instructors to extend rigid prescriptivism to the level of the text, insisting for example, that all instances of writing in a particular genre must conform to one specific structure and set of grammatical features. This can tend to stifle individual expression and possible linguistic innovation. However, it has not been the intention of this article to explore genre-based teaching in full: hence, these issues will not be explored, although readers are reminded that like all approaches to ELT, genre-based approaches are not a panacea or final revelation, and the article does not pretend that they are. Rather, it is hoped that through its description of a writing programme and the considerations involved in its construction, the article has pointed a way towards how effective language instruction may designed through application of what is believed to be sound theory.

References
Butt, D; Fahey, R; Feez, S; Spinks, S; & Yallop, C. (2000). Using functional grammar: An explorer's guide. Macquarie University: NCELTR

Curriculum Planning and Development Division, Ministry of Education, Singapore. (2001). English Language Syllabus 2001 for Primary & Secondary Schools. Singapore: Ministry of Education. (Also available on the Internet at the following url address: http://www1.moe.edu.sg/syllabuses/doc/English.pdf)

Derewianka, B. (1990). Exploring how texts work. NSW: Primary English Teaching Association.

Derewianka, B. (2003). Trends and issues in genre-based approaches. RELC Journal. 34(2), 133-54.

Eggins, S. & Slade, D. (1997). Analysing casual conversation. London: Cassell.

Foley, J. & Thompson, L. (2003). Language learning: A lifelong process. London: Arnold.

Halliday, M A K. (1994). An introduction to functional grammar. (2nd edition). London: Edward Arnold.

Halliday, M. A. K., & Hasan, R. (1989). Language, context, and text: Aspects of language in a social-semiotic perspective (2nd edition.). Oxford: Oxford University Press.

Halliday, M A K & Matthiessen, C. (2004). An introduction to functional grammar. 3rd Edition. London: Arnold.

Kim, Y. & Kim, J. (2005). Teaching Korean university writing class: Balancing the process and the genre approach. Asian EFL Journal Quarterly, 7(2), 69-90.

Ladbrook (2005). Writing recipes. The Language Teacher, 29(4), 23-27

Lantolf, J.P. (2000). Sociocultural theory and second language learning. Oxford: Oxford University Press.

Lin, B. (2003). English in Singapore: An insider's perspective of syllabus renewal through a genre-based approach. RELC Journal, 34(2), 223-46.

Mann, W.C., & Thompson, S.A. (1988). Rhetorical structure theory: Toward a functional theory of text organization. Text, 8(3), 243-281.

Paltridge, B. (2001). Genre and the language learning classroom. Ann Arbor: The University of Michigan Press.

Rodgers, T. (2001). Language teaching methodology (http://www.cal.org/ericell/digest/rodgers.html, ERIC Digest, September 2001)

Swales, J. (1981). Aspects of article introductions. Birmingham, UK: University of Aston Language Studies Unit.

Swales, J. (1990). Genre analysis: English in academic and research settings. Cambridge: Cambridge University Press.

Vygotsky, L. (1978). Mind in society. Cambridge, Mass: Harvard University Press.

APPENDIX A

II. LET'S LOOK AT A MODEL:

a. A possible article in a student magazine