Abstract
Second language (L2) writers often have an incomplete control of English and rely on teachers to help them develop the linguistic resources necessary for them to express themselves effectively. Genre-based pedagogies were developed to address the language needs by offering students explicit and systematic explanations of the way language functions in social context. One of these approaches, drawing on Halliday & Matthiessen’s (2004) functional grammar, identifies the discourse and grammatical structures typical in different social activities and links these linguistic choices with the social purposes and situations that the texts participate in. However, few studies have investigated the effects of genre-based pedagogy on L2 students’ writing development. The present study addressed the need by evaluating the functional approach to genre in an EFL composition course in Taiwan as college freshmen learned how to write a narrative text. Results indicated that students’ essays exhibited large pre- to posttest gains in specific rhetorical moves (i.e. character’s problem illustration and struggle identification), in content development, in textual coherence and in language usage. These results confirmed the findings of previous studies on advanced L2 learners that explicit knowledge of genre sharpens students’ awareness regarding the content, organization and language use to generate the target genre.

Key words: genre, narrative, genre pedagogy, Systemic functional linguistics

Introduction
As a paradigm shift has taken place in approaches to teaching L2 writing over the last 15 years, genre-based instruction has emerged as the most strongly advocated alternative to the widely-used process approach (Atkinson, 2003; Hyland, 2002, 2003, 2004; Hyon, 1996; Johns, 2002, 2003; Paltridge, 2001; Ramanathan & Kaplan, 2000; Cheng, 2006). A major flaw of implementing the process approach in L2 writing classroom is its overemphasis on general principles of thinking and composing with minimal attention to the language and rhetorical patterns L2 students require to express themselves effectively. Students are generally expected to discover the language they need in the process of writing itself. This is a challenging task for L2 writers, who may not have extensive exposure to the target language. A genre-based approach to L2 writing is proposed to address the language needs ofL2 learners by offering students discourse knowledge on the way the target language functions in social context.

Nevertheless, the efficacy of genre-based teaching and learning has caused great controversy between genre proponents (Hyon, 1996 & 2002; Paltridge, 2001; Carter, Ferzli, & Wiebe, 2004; Hyland, 2004). What’s at issue is that few empirical studies have been undertaken to illustrate whether a genre approach to L2 writing can achieve its potential (Cheng, 2006; Tardy, 2006). Further research is necessary to understand to what extent genre instruction benefit L2 writers. As Freedman stated (1993), “It should not be the task of the skeptics to argue against a pedagogic strategy but rather the work of the proponents to bring forward convincing research and theoretical evidence --- preferably before its wholesale introduction” (p.279). As such, the purpose of the present study is to examine the effect of explicit instruction of one type of school genre, narrative, on English freshmen in EFL context. Before reporting this empirical study, I first discussed studies on genre-based instruction in L2 context and briefly introduced the theoretical frameworks, Systematic Functional Linguistics (SFL) theory and narrative genre.

Genre-based instruction in L2 context
The existing research on genre-based L2 literacy acquisition has indicated mixed results. Henry & Roseberry (1998) examined the effect of explicit genre instruction on first-year management students’ acquisition of tourist information genre. The instruction was primarily based on analysis of genre moves by reading a variety of text models. The results indicated that the treatment group did show significant improvement in motivation and texture (cohesion and coherence) scores and their gain scores were significantly higher than the control group. Interestingly, students who have received explicit instruction on genre moves showed no significant improvements in this variable in their pre-/post-tests, although they did make certain progress. Furthermore, they also failed to gain significantly higher gain scores in the move variable than the control group who were asked to read the model texts in the absence of any explicit explanation. However, their research indicated that exposure to model texts in the absence of explicit instruction did not help students acquire genre knowledge since there were no differences in the pre-/post-tests move scores among the control group. It is still unknown why explicit instruction on genre moves failed to yield any significant progress in move scores for the experimental group.

Hyon’s study (2002) assessed the effects of an EAP genre-based reading course on building genre knowledge and developing L2 reading ability. The instructional focus was primarily a text-based analysis of rhetorical moves, language style and discourse purpose. Based on post-instructional naming tasks and interviews, the findings indicated that the course facilitated a better understanding of texts’ rhetorical elements but failed to provide all skills essential for effective L2 reading, such as reading strategies and vocabulary knowledge. In a follow-up study on the long-term effects of this course, Hyon (2001) found out that one year later after the course, the L2 graduate and undergraduate students were able to recognize specific genre features taught in the course. Yet, some participants over-generalized and misapplied some prototypical genre features. Hyon’s studies showed that genre instruction can help L2 students acquire L2 reading-related abilities to a certain extent and may need to incorporate other complementary instruction to be an effective approach to L2 reading training.

The above studies indicated that explicit instruction may have positive effects on L2 literacy acquisition but these studies have also highlighted several limitations of genre-based pedagogy. Since these limitations may derive from varying approaches to genre instruction, Pang’s study (2002) further clarifies how different approaches to genre instruction led to distinct effects for college students of varying writing proficiencies. Pang compared two different approaches to genre instruction: textual-analysis (focus on schematic structure and lexical and grammatical analysis) and contextual-awareness (types of publication, communicative purpose, writer and reader’s role) approaches. The participants under the instruction of both approaches employed the same composing strategies and produced similar quality of written texts. However, the contextual awareness group had better understanding of the discourse function while the textual analysis group performed better on the mechanistic elements associated with the target genre, such as the use of mood and person, obligatory moves and features, and format. Pang indicated that textual analysis approach worked better with students with low or high pre-test scores while contextual awareness approach helped some participants with low or medium pre-test scores. Different from the findings of Henry & Roseberry’s research (1998), Pang’s study indicated that genre instruction did enhance students’ employment of move structure.

   As shown above, the inconclusive results of the existing  research to affirm the efficacy of genre instruction led to the need for more studies. In particular, most studies centered on advanced learners in the EAP/ ESP contexts but neglect to examine the effect of the instruction in elementary genres. As Widdowson (1993) cautions, instruction in specialized genres may not provide novice L2 learners with transferable language knowledge. After all, there is no way to predict the wide range of possible genres students may need to participate in their academic and professional settings (Flowerdew, 1993). To further clarify the efficacy of genre instruction and learning, the present study has adapted SFL theory to teach novice L2 writers one type of school genre, narrative, and examined its effects on these learners. As a theoretical framework in designing the instructional content, SFL theory is briefly introduced in the following section.

SFL genre theory
ASFL approach to genre developed by linguistics and practitioners in Australia mainly draws on the linguistic theory of Michael Halliday (e.g. Halliday & Matthiessen, 2004; Halliday & Hasan, 1989). With its focus on the core educational genres and its clearly articulated social function as well as discourse and grammatical features for each genre, SFL approach is taken as the pedagogical framework for this study. This theory addresses the relationship between language and its social functions and sets out to show how language is a system from which users make choices to express meanings. It advances that the context of situation of a language event on how we use language is further divided into: field (topic or focus of the activity), tenor (the relationship between the writer and reader or the speaker and hearer), and mode (expectations for how particular text types should be organized). These three components are realized through elements in the lexico-grammatical system (Halliday & Matthiessen, 2004; Martin, 1984).That is, the way people make meaning by choosing linguistic choices varies according to the field, tenor, and mode variables that constitute the context of situation. The lexico-grammar is seen as construing three kinds of meanings, corresponding to field, tenor, and mode: with the field of discourse realized in ideational resources, the tenor of discourse realized in interpersonal resources, and the mode of discourse realized in textual resources.

According to Eggins (1994), it is the lexico-grammar level that is particularly important in understanding how the different contexts of situation are realized in lexical and grammatical choices.

SFL pedagogy is grounded in the belief that learning to write should be based on explicit awareness of language (Hyland, 2004; Hyon, 1996). As an attempt to provide a framework that will help explain genre use at all educational levels, SFL researchers characterize genres in terms of broad rhetorical patterns such as narratives, recounts, arguments, and expositions. In addition to specifying key genres that students are often asked to write, SFL researchers have also identified the purposes for communicating in each genre and the typical stages and linguistic features of these texts that can express these purposes. Furthermore, they employ analytical tools and frames from systemic functional linguistics to identify the discourse and grammatical structures needed to produce genres. By describing the typical features of valued genres, teachers can provide students with clear options for writing, both within and beyond the sentences, to help them create texts appropriate to readers (Hyland, 2004; Lin, 2006; Schleppegrell, 2004). As such, SFL’s descriptions of narrative genre will serve as the instructional focus for the present study.

Narrative genre
Narrative writing is chosen as a pedagogical genre in freshmen composition course since it acts as a fundamental genre for several advanced academic and ESP genre, such as literary response essay, news report, feature report or film review. Based upon the SFL theory, the instruction of narrative writing should focus on the social purpose realized through rhetorical moves, textual structure and language. The social purpose of narrative genre is to entertain, amuse or reflect an experience. This can be achieved by introducing the reader to a possible world in which individual characters experience a problem of some kind which they have to face and overcome (Macken & Slade, 1993; Macken-Horarik, 2002). The schematic structure of a prototypical narrative to realize the discourse purpose is represented as follows: Abstract, Orientation, Complication, Evaluation, Resolution and Coda. Abstract indicates what the story is about. Orientation provides relevant background information about the characters’ situation. Complication introduces one or more problems for characters to resolve, involving Problem and Struggle episodes. Evaluation highlights the significance of the events for characters and resolution to resolve these issues. Coda brings readers back to the present situation.

In addition to the rhetorical moves as shown above, teachers can further point out the grammatical and rhetorical features of narrative. Along the tenor dimension, writers are expected to explore the subjective world of the fictionalized characters in order to manage the reader’s responses: their needs, motives, emotional response. Overt expressions of the writer’s personal feelings and attitudes should be avoided in a narrative genre. Several linguistic resources can be utilized: “reporting and quoting the speech of imagined characters”, “revealing their thinking in mental processes”, “commenting on events”, “creating a sense of the characters’ responses and motives”, using imagery and metaphor, rhetorical questions, repetition and exclamatives in building suspense (Macken & Slade, 1993, p.217). Along the field dimension, writers are expected to sequence events properly and exploit these events appropriately. Language features in this dimension include specific, individualized characters, inclusion of relevant physical circumstances to build up the material reality of the possible world, and discriminating word choices (particularly of the thinking and doing processes). The choice of written mode requires writers to create a context-independent text in order to help the readers to reconstruct the context for the events of the narrative. Controlling this aspect of language means varying the sentence structure in terms of the use of first position in the clause, and using explicit conjunctions and various cohesive devices. By describing the typical features of a genre, teachers can provide students with clear options for writing, both within and beyond the sentences, to help them create texts that seem well-formed and appropriate to readers (Hyland, 2004; Schleppegrell, 2004).

The above descriptions of the discourse/language features in a narrative genre were adapted into instructional focus and activities presented in the following section and served as the scoring rubrics in the assessment schemes (See Appendix A).

Method
Participants
The participants were 26 English major freshmen in the Freshmen composition course taught by the researcher at a lower rank national university in Taiwan. This is the first text type taught in this composition class at the beginning of their first college year. Prior to this pedagogical intervention, the majority of the participants had no formal and systematic instruction in English writing as indicated in their responses to a pre-intervention survey on their previous English writing experience and instruction. Before entering university, all the participants had received compulsory English language education for at least six years and had practiced English paragraph writing at senior high school in preparation for the English writing test in the College Entrance Exam. At college level, English-major freshmen are also required to take other English language courses, such as grammar, reading and speaking. Novice writers were selected since the purpose of the present study is to examine the effect of genre-based instruction on less proficient L2 writers, who had received little attention in previous studies of genre-based instruction.

Writing task
All participants were required to write a narrative paragraph of about 250-300 words in response to an English prompt. Students were provided with several topics, each of which stated a hypothetical situation with various topics, audience, and purpose specification and prompted them to tell a personal experience (See Appendix B, for example)

Instructional focus
Narrative writing is considered more familiar to freshmen novice writers and is sequenced as the first genre taught at this writing class. The total instructional cycle for this genre lasted for four sessions, each of two class hours. The present approach resembled the Australian systemic functional genre application in its emphasis on the relationship between context and language use and its use of explicit discussion, modeling and analysis of genres (Flowerdew, 1993; Hyon, 2002). What follows is a lesson plan to help students develop a narrative text through a sequence of activities that move students through three different stages. Stage one of this approach focuses on understanding the contextual variables associated with narrative writing: writer’s role and communicative intent, reader’s role and intent, channel of message, and social activity. The goal of this stage is for students to understand that writers do not just respond to a topic or an assignment but take several contextual variables into consideration while generating their ideas. This contextual awareness is considered essential for students to understand how the writing context shapes how a wrier interprets task demands and influences what a writer writes.

Activity 1.1
Students were first introduced to the genre “narrative” by considering several general questions such as “Under what condition do we need to tell a story?” “What’s the purpose of doing this?” “How do we tell a story?” and based on their own personal learning experience in their L1 writing, they were required to give specific and concrete examples in support of their answers. Then, several featured questions that focused students’ attention on contextual variables in writing are given in the form of prompt as listed below.

• What is the social activity the text participates in?
• Where can you find the text?
• What purpose does it serve?
• Who is the writer? What are the writer’s qualifications for writing the text?
• What’s the writer’s purpose?
• Who are the readers? What are their concerns? What is their purpose?

Students were asked to respond to the questions as reading a narrative text while inferring the process that experienced writers engage in to produce argumentation. They are encouraged to consider these questions throughout the whole learning process and in their composing process.
           
Activity 1.2
Students studied three texts of similar genres (drawn from different publications): a modern short story, a recount and a news report. For each text, students hypothesized about the publications in which the three texts are located and deduce which types of publication the source is. They also inferred the writer roles and purposes, and the target readers, their gender, age, education, interests, and reader purposes. Students then discussed the contextual differences among the three texts. Inferring contextual components from the text is essential to contextual awareness building (Pang, 2002). By comparing three similar but distinct genres, students learned how contexts affect writer’s rhetorical considerations and subsequent writing strategies.

The main goal of stage two was to introduce the schematic moves realizing the situational context and the functions of each move. Students rearranged moves appearing as sentences in a story. They mapped some of the sentences onto the moves that are typical of stories: Abstract, Orientation, Complication, Evaluation, Resolution and Coda. They were then asked to identify the features and functions of each move, and to explore its relationship to the discourse purpose. More examples of modern short stories were provided for students to discuss which move structure was essential and which was not.

Finally, stage three focused students on analyzing lexical-grammatical features associated with English narrative writing. Such analyses intended to help students understand how the situational context (field, tenor, mode) was constructed through linguistic features.  These features included types of verbs (e.g., action, thinking, saying, being, having), use of appraisal terms, syntactic features (e.g., expanded noun phrases, various conjunctive structures), thematic choices, and cohesive ties. Learners were asked to identify how these different grammatical features realize the schematic move and the genre. For example, they were taught to analyze how to use expanded noun phrases to present a character or how to use various conjunctive structures functional in the Complication move.

In addition to these in-class activities, students analyzed instances of narrative genres on their own. For homework, they were asked to find an example of narrative genre and to describe evidence of some of the features discussed in class. Several process-based practices were also integrated into this genre-based instruction. For example, students worked on their first draft after stage one and their second draft after stage two. That is, students were asked to apply what they have learned to their own texts step by step in terms of their own goals as writers. Also, students received oral and written feedback from their peers and teachers on their multiple drafts.

Assessment scheme
Appendix A shows the assessment schemes used to score the pre- and post-test essays. It is composed of three dimensions: content, organization and language use. The content dimension was defined as the degree to which the writer’s elaboration of each rhetorical move, use of sensory details, inclusion of vivid, interesting orientation and notable turning points. The organization dimension was grounded on SFL theory and was divided into the inclusion of obligatory moves, optional moves, flexibility in move sequences, the use of thematic choices and cohesive ties. Language use dimension was defined as the range of verbs, appraisal expression, syntactic diversity and use of specific vocabulary appropriate to the experience being told. These criteria are significant in the realization of both textual and contextual meanings, as underpinned in the SFL theory. After all, the judgment whether a genre has been mastered has to be grounded on the discoursal and linguistic realizations in the text of a target genre. The scores in the rating were derived impressionistically along a scale of “not evident”, “minimum evidence of mastery”, “some evidence of mastery,” “adequate evidence of mastery,” “outstanding evidence of mastery,” and “strong evidence of mastery,” ranging from 0 to 5.

Data analysis
The pre-/post-test essays were written in response to the same topic. All texts were double rated by the researcher to check intra-rater reliability. Thirty percent of the data were randomly selected and coded by an experienced native-Taiwanese EFL writing instructor to ensure inter-rater reliability. On the analysis of various features, the intra-rater correlation coefficient ranged from 0.89 to 0.98 and the inter-rater reliability ranged from 0.82 to 0.93. Paired-samples t-tests were undertaken to examine each feature listed in content, organization and language use components in the pre-/post-test essays.

Results
The relative effectiveness of the implemented approach is reported here, though no hard and fast conclusion regarding genre acquisition can be drawn because of the limited sample size, the particular genre used, and the special environment in which the research was conducted. As shown in Table 1, the paired-samples t-test indicated that students made statistically significant progress in most features examined, as evident in the difference between the scores in the pre- & post-test essays, except the incorporation of obligatory and optional moves and flexibility in move sequence. It appears that improvement in the quality of written texts correlates highly with enhancement in participants’ awareness of discourse and language features in the taught genre.

Table 1
Means & standard deviations for functional elements at pretest and posttest essays

	Pre-test	Post-test	Gain scores
	Means    SD	Means   SD	Means    SD
Content
Amount of elaboration	1.81      .57	3.73    .96	1.92      .98
Sensory details	1.50      .86	3.69    .88	2.19     1.13
Orientation	1.46      .81	3.38    .85	1.92      .93
Turning points	1.23      .95	3.54    .90	2.31     1.56
Organization
Obligatory moves	3.93     1.43	4.69    .47	0.76*    1.56
Optional moves	1.04     1.22	1.50    .65	.46**   1.30
Flexibility	.15      .54	.73   1.37	.58***  1.07
Thematic progression	1.65      .80	3.35    .75	1.69     1.19
Cohesion/cohesiveness	2.12      .52	3.42    .86	1.31      .97
Language features
Action verbs	1.81      .75	3.69    .79	1.88     1.03
Appraisal expression	1.46      .58	3.62    .94	2.15     1.05
Syntactic diversity	1.96      .53	3.69    .79	1.73      .78
Precise vocabulary	1.54      .76	3.46    .95	1.92     1.13

N = 26, * p = .021, ** p = .083, *** p = .011

Content
Regarding the amount of elaboration in each rhetorical move (abstract, orientation, complicating action, evaluation, resolution, coda), scores for the posttest essays were significantly higher than scores for the pretest essays, t (25) = 10.04, p =.000.Also, writers in the posttest essays employed more concrete, specific sensory details as indicated in the posttest scores were significantly higher than the pretest scores, t (25) = 9.88, p =.000. In the incorporation of vivid, interesting orientation, scores for the posttest essays were significantly higher than scores for the pretest essays, t (25) = 10.49, p =.000. Further more, in the use of notable turning points, scores for the posttest essays were significantly higher than scores for the pretest essays, t (25) = 10.16, p =.000. Participants were better able to generate richer elaborations for each rhetorical move and use concrete and specific details to attract readers. Their genre knowledge helps them to make decisions about what topic would be more interesting or which episode to include.

Organization
Regarding the inclusion of an obligatory move (orientation, complicating action, evaluation, resolution), scores for the posttest essays were not significantly higher than scores for the pretest essays, t (25) = 4.41, p =.000. Most of the students were able to include all the obligatory moves in their narrative paragraphs, except that 20 percent of the participants (five out of 26) interpreted narrative writing as another story genre “recount”. In recount writing, there are no problem and struggle moves but only “record of sequential events”. The missing moves in students’ first drafts are usually problem and struggle moves (Complicating action). Regarding the incorporation of optional moves, there was no statistically significant difference between pre-test and post-test scores, t (25) = 1.81, p =.083. Likewise, regarding the flexibility in move sequence, scores for the posttest essays were not significantly higher than scores for the pretest essays, t (25) = 2.76, p =.011. However, in terms of the thematic progression, scores for the posttest essays were significantly higher than scores for the pretest essays, t (25) = 7.24, p =.000. Also, scores in the use of temporal connectives and other cohesive ties for the posttest essays were significantly higher than scores for the pretest essays, t (25) = 6.88, p =.000.

Language features
Overall, participants improved their use of various language features assessed in the present study. With respect to the range of action verbs, scores for the posttest essays were significantly higher than scores for the pretest essays, t (25) = 9.31, p =.000. For the range of appraisal expression, scores for the posttest essays were significantly higher than scores for the pretest essays, t (25) = 10.49, p =.000. In terms of syntactic diversity, scores for the posttest essays were significantly higher than scores for the pretest essays, t (25) = 11.35, p =.000. Also, in the use of specific vocabulary appropriate to the experience being told, scores for the posttest essays were significantly higher than scores for the pretest essays, t (25) = 8.69, p =.000. The findings indicated that this pedagogical intervention has enhanced their awareness of language usage in terms of several important linguistic features specific to narrative writing.

Two examples
The quantitative and qualitative analyses and results are illustrated next through two sets of pre- and posttest essays selected (See Table 2) as being representative of the group’s pre- and posttest levels of achievement. These examples, showing gains of 1-3 points on various features in Content, Organization and Language, clarify and illustrate the average gain from 1.30 in the pretest to 2.19 in the posttest with respect to considered features (Table 1). The essays were written by two students who were average writers described by the instructor/researcher based on the textual quality of their pretest drafts. These examples illustrate in what ways the majority of the students improved their writing. To facilitate discussion, sections of the written texts were bracketed and labeled as Abstract, Orientation, Complicating action (Problem/ Struggle), Evaluation, Resolution and Coda (See Appendix C).

Table 2
Scores on pretest/posttest by Student A & B

	Student A Pretest  Posttest		Student B Pretest  Posttest
Content
Amount of elaboration	2	4	2	3
Sensory details	1	4	1	4
Orientation	1	3	1	4
Turning points	2	3	1	4
Organization
Obligatory moves	4	5	4	4
Optional moves	1	1	1	1
Flexibility	0	0	0	0
Thematic progression	3	4	1	3
Cohesion/cohesiveness	2	4	2	3
Language features
Action verbs	3	4	1	3
Appraisal expression	2	4	1	3
Syntactic diversity	3	4	2	3
Precise vocabulary	2	4	2	3

Pretests
Student A received scores ranging from 1 to 2 for the features on Content component because there was little effort to elaborate the physical details of her experience, to differentiate the subjective world of the characters and to build suspense or atmosphere via dialogue, questions and exclamations. Although the writer did describe some mental processes “nervous”, “glad”, “delighted”, all of her feelings and thinking were presented in a descriptive instead of a more concrete way. As such, this narrative contributes to the impression of a very “slapdash” story. The reader is left with many unanswered questions: for example, why is this “one of the most impressive experiences?” Why did the writer “make up her mind to save the dog’s life?”

The essay received 4 points on the inclusion of obligatory moves since the writer failed to draw clear lines between move boundary. The treatment of the activity sequence is believable and coherent so that it is possible to understand what goes on and to follow the sequence of events. However, the writer has been content to write a narrative event and sequence with little evidence of interest in the interpersonal world of the characters or its meaning for the reader, as discussed in the above paragraph. The evaluation move leaves the narrative rather flat due to the absence of any significant evaluative meanings, as noted by the writer at the end of the story “I think it is the most loving experience that I have ever had.” So the readers came to the end of the story with a “why” response. The text incorporated 1 optional move (Abstract) but lacks flexibility in move sequence. In terms of themes, the writer used varied and interesting sentence beginnings. The thematic choices signaled a change in the rhetorical move and helped push the narrative forward.

The essay is only moderately well handled in terms of language use. The writer has attempted to employ a range of action verbs and was able to integrate a range of clause types in her text, for example, “Although knowing that I would be late to school if I took it to the hospital, I made up my mind to save its life”; “About five minutes later, an ambulance coming, the nurse took it to the hospital, telling me that everything would be fine and did not be so nervous.” Yet, the text has few appraisal expressions except “the most impressive experience happened” and “the most loving experience that I have ever had.” Also, much of the word usage was not appropriate to the experience being told, for example, “…but the hurter was not a person…” and “…appealed to others for not to clash or tread on the dog anymore.”

Text B breaks out of the traditional narrative mode and is characteristic of what Gee (1989) calls essayist discourse, associated with expository prose in its lack of any emotive, performative content. Without any dramatic and expressive devices to create a dramatic performance, it is impossible for readers to be caught up in the story. As such, the text scored 2 in terms of the elaboration of each rhetorical move, indicating that the writer understands what to be included in each move without providing sufficient details to engage readers. More information about the participants, processes and circumstances needs to be supplied in the text. Similar to the pretest essay by student A, the subjectivity of the characters is left undeveloped. Student B needs to make more use of devices like imagery, exclamatives, more descriptions of the characters’ responses and motives or connecting on events for the purpose of building suspense, atmosphere and characterization. Consequently, the text failed to give concrete sensory details and create vivid orientation and notable turning points. In terms of these features, the text scored only 1 point, indicating minimal evidence of mastery.
In terms of organization, the essay received 4 points in the inclusion of obligatory moves because it incorporated all the obligatory moves but the change of the narrative stage is not evident. It incorporated one optional move (abstract) and all the moves were sequenced without any flexibility. The use of narrative theme choices fails to effectively signal changes in the temporal sequence of events. Also, the variety of sentence beginnings could be diversified. Logical relations between events or sentences are moderately established. Her use of cohesive ties assists the readers to recover referents, although some explicit connections are needed to mark the logical relations between sentences. For example, student B wrote, “She told me all about this situation. I worried about Annie very much.”

On language use, text B receives 2 points in syntactic diversity and the use of vocabulary appropriate to the experience being told. However, it scored only 1 point for the range of action verbs and appraisal expression, indicating the writer showed minimal control of action verbs (for example, repetitive use of “eat”), and made little attempt to evaluate the events except “difficult”, “healthiest” and “ not a good way”. Her control of the sentence is rather uneasy throughout the whole text with several choppy sentences; for example, “I believe everyone had heard about anorexia. For time goes by, she ate less and less.”

Posttests
On the posttest essay, student A began to experiment with the subjectivity of the characters, making use of devices like interior dialogue (for example, “If I took this dog to a hospital, I would be late for school. Even worse, that day was our final exam”), imagery (“I was a killer!”) and exclamatives (“How terrible it is! I don’t want to get involved in their incident!”) in building atmosphere, suspense and character. The physical events are also specified - usually in the thinking and doing processes. Instead of simply telling “I flagged a taxi and took the dog to animal hospital immediately,” the writer further elaborated with “By the time I arrived at hospital, owing to the fact that I feared the dog might die, I was sweating, crying and pale with fear.” As such, the text scored 4 for its rich elaboration of each rhetorical move and generation of concrete, specific sensory details while it scored 3 for an even and appropriate orientation and 3 points for a moderately effective suspense.

In terms of organization, the text improved by generating a more discriminating move boundary (5 points), more diversified thematic choices (4 points) and better established logical relations between events and sentences with appropriate use of conjunctions and varying cohesive ties (4 points). Overall, the text does hang together very well.
The text was well handled in terms of language features, scoring 4 for all rubrics, although some grammatical errors should be avoided. The writer made more discriminating and evocative word choices (for example, “several thought flashing through my mind” instead of “I think”; “whispering in a low voice” instead of “speaking quietly”). The use of action verbs, appraisal, sentence diversity or vocabulary was appropriate to the experience being told and subsequently contributed to the overall development of the narrative.

On the posttest essay, student B switched the topic to a personal experience, contending that the present story was easier to project the suspense. This text scored 3 by generating an adequate elaboration of each rhetorical move, while scoring 4 with regards to its strong evidence of mastery in providing sensory details, orientation and building suspense. Most entertaining part of this story came from its employment of interior dialogue (for example, “What? It’s the railroad ring! God! How can I do?” I thought), exclamatives (“Am I still alive? Am I out of hell?”), descriptions of sounds (“Dung-dung! Dung-dung! Something is ringing.”), and imagery (“angry-looking clouds”, “She is like a living Jesus”). The major weakness in terms of content was its lack of any significant evaluation. As noted in Macken & Slade (1993), the events of any narrative are given their significance through evaluation.

In terms of organization, the writer of text B made little progress on the use of obligatory moves and optional moves but developed diversified thematic choices and established better temporal and logical relations between events. The themes for each new narrative stage moderately heralded a change in the temporal setting to push the event forward (Problem: “Suddenly”; Struggle: “Although I still can think but my body is freezing,…”; Resolution: “After arriving the safe road,…”; Coda: “now,…”, but not in the Orientation move). 
                                    
With respect to language, the writer of text B made notable progress from 1 or 2 points in the pretest essay to 3 points in posttest essay. The writer was able to integrate and embed a range of clause types within sentences and employed specific vocabulary to describe her experience (for example, “accelerator”, “crash”, “burst into tears”), although some of her expressions were not native-like and there were several grammatical errors.

Discussion

The results in the previous section suggested that a genre approach to teaching writing can produce some positive results. For these students explicit teaching of genre elements improved the quality of their narrative paragraphs in a relatively short period of time -four weeks. Prior to instruction, a large number of the participants (except five students) were quite familiar with the structural elements of a narrative as indicated in their pretest essays, although most of them failed to succeed in using language to achieve the purpose of each rhetorical move. In the pretest, participants wrote a skeletal narrative in the absence of any attempt to elaborate the physical details of their story, the subjective world of the characters and the atmospheric feelings of the text. Explicit instruction on how to use language to realize each move appears to help enhance participants’ awareness and develop each rhetorical move. Students are better capable of presenting more sensory and performative details, building more effective suspense and atmosphere and adding richer exploration of the story characters. In line with Pang’s (2002) study, explicit instruction contributes to the content development and linguistic expressions of students’ text in the target genre.

However, similar to Henry and Roseberry’s (1998) findings, explicit instruction on genre moves failed to produce any significant progress in move scores from pretest to posttest, although they did make certain progress. Narrative genre is one of the elementary genres that students have acquired through their L1 literacy context. Participants in the present study are capable of transferring their L1 literacy knowledge to L2 composition. Their major difficulty in composing an English narrative is their lack of linguistic knowledge to realize the discourse function of each rhetorical move. This finding has implication for pedagogical priority in designing the classroom activity as teaching fundamental genres to L2 novice writers.
Moreover, explicit instruction on generic structures did not significantly lead students to use the rhetorical move in a flexible way. This does not imply that students necessarily learned rote procedures rather than principles as contended by several New Rhetoric approach researchers. One possible explanation for this failure is that the model essays selected for in-class readings followed the fixed sequence of genre patterns so that novice writers may not be offered ways to diversify their organization patterns. The second explanation is that flexibility to use these structures may come at the later stage of acquisition as these writers placed more emphasis on determining appropriate language to realize the rhetorical moves. Another explanation is that the assigned topics required students to describe their own personal experience, which may in consequence urge them to report their story in a chronological order.

Conclusions

Researchers have advocated that teaching/learning needs to be put into a more focal position in the ESP/EAP genre-based approach (Hyon, 2002; Cheng, 2006), given our relatively less developed understanding of the efficacy of this approach. The present study adds to our knowledge about what L2 students in EFL context learn in the genre-based pedagogical context. The findings also help us to clarify this approach to L2 literacy learning and teaching and measure its effect for EFL novice college writers.

There are several implications for teaching to be drawn from the mixed results of this study. Due to students’ previous acquisition of narrative genre moves in their L1 literacy context, the instructional focus can be placed less on the introduction of move structures but more on the realization of the functions of each move by providing discourse and language features to achieve the intended goals. For example, comparing the effects of varying organization patterns of obligatory moves to achieve a certain purpose can facilitate students’ understanding of potential ways to achieve discourse goal and such awareness may lead to subsequent transference of their knowledge to their texts. Furthermore, although the present study did not analyze the ways students employ each rhetorical move in great details, the participants appeared to be less capable of representing evaluation move even after the intervention. As Cortazzi & Jin (2000) noted, one Chinese discourse principle rests on the belief that the primary evaluation in a narrative does not need to be explicitly mentioned because it is the reader’s responsibility to decode the evaluative intention of the storyteller. This pinpoints the need for the EFL teachers to place emphasis on how the evaluation is used in an English narrative. For instance, more practice on specifying the language features of evaluative expression and its functions in narrative genre can facilitate students’ acquisition of the evaluation move and enhance their ability to use it more effectively.

One major shortcoming of the present study is the lack of a control group to detect specifically in what ways and to what extent genre approach benefits these novice writers. Replication of the present study is needed with quasi-experimental method to directly address this issue. In addition, more research is required to further explore learner dynamics as receiving genre instruction and to understand the complexities of genre-based learning. One area of study as suggested by Cheng (2006) is to examine how learners of various proficiency levels interact with a genre approach. Although participants in the study were all categorized as novice L2 writers due to their freshmen status, their L1 writing ability and English proficiency may vary. It would be useful to gather more information from them, possibly through case studies or interviews, to find out how this approach benefits them and what other skills they require to improve their L2 writing.
           
Acknowledgements
An earlier version of this article was presented at the 22nd Conference on English Teaching and Learning in the Republic of China on June 4, 2005 at National Taiwan Normal University, Taipei, Taiwan. 

References
Atkinson, D. (2003). L2 writing in the post-process era: Introduction. Journal of Second Language, 12, 3-15 

Berkenkotter, C., & Huckin, T.N. (1995). Genre knowledge in disciplinary communication: Cognition, culture, power. Hillsdale, NJ: Lawrence Erlbaum.

Carter, A., Ferzli, M., & Wiebe, E. (2004). Teaching genre to English first-language adults: A study of the laboratory report. Research in the Teaching of English, 38, 4, 395-419.

Cheng, A. (2006). Understanding learners and learning in ESP genre-based writing instruction. English for Specific Purpose, 25, 76-89.

Cortazzi, M., & Jin, L. (2000). Evaluating evaluation in narrative. In S. Hunston & G. Thompson (Ed.), Evaluation in text (pp.102-120). Oxford: Oxford University Press.

Eggins, S. (1994). An introduction to systemic functional linguistics. London: Pinter Publishing.

Flowerdew, J. (1993). An educational, or process, approach to the teaching of professional genres. ELT Journal, 47, 305-316.

Freedman, A. (1993). Show and tell? The role of explicit teaching in the learning of new genres. Research in the Teaching of English, 27, 222-251.

Gee, J. P. (1989). Two styles of narrative construction and their linguistic and educational implications. Journal of Education, 171, 1, 97-117.

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

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

Henry, A., & Roseberry, R. L. (1998). An evaluation of a genre-based approach to the teaching of EAP/ESP writing. TESOL Quarterly, 32, 147-156

Hyland, K. (2002). Teaching and researching writing. London: Longman. Hyland, K. (2003). Second language writing. Cambridge: Cambridge University Press.

Hyland, K. (2004). Genre and second language writing. Ann Arbor, MI: The University of Michigan Press.

Hyon, S. (1996). Genre in three traditions: Implications for ESL. TESOL Quarterly,30, 693-722.

Hyon, S. (2001). Long-term effects of genre-based instruction: a follow-up study of an EAP reading courses. English for Specific Purposes, 20, 417-438.

Hyon, S. (2002). Genre and ESL reading: a classroom study. In A. M. Johns (Ed.), Genre in the classroom: Multiple perspectives (pp. 121-141). Mahwah, NJ: Lawrence Erlbaum.

Johns, A. M. (2002). Genre in the classroom: Multiple perspectives. Mahwah, NJ: Lawrence Erlbaum.

Johns, A. M. (2003). Genre and ESL/EFL composition instruction. In B. Kroll (Ed.),Exploring the dynamics of second language writing (pp. 195-217). Cambridge, UK: Cambridge University Press.

Lin, B. (2006). Genre-based teaching and Vygotskian principles in EFL: The case of a university writing course. Asian EFL journal, 8, 3. (Article retrieved from
http://www.asian-efl-journal.com/Sept_06_bl.php)

Macken, M., & Slade, D. (1993). Assessment: A foundation for effective learning in the school context. In B. Cope and M. Kalantzis (Eds.), The powers of literacy: A genre approach to the teaching of writing (pp. 203-230). London: Falmer Press.

Macken-Horarik, M. (2002). “Something to shoot for”: A systemic functional approach to Teaching genre in secondary school science. In A. M. Johns. (Ed.). (2002). Genre in the classroom: Multiple perspectives (p.17-42). Mahwah, NJ: Lawrence Erlbaum.

Martin, J. R. (1984). Language, register, and genre. In F. Christie (Ed.), Language studies: Children writing leader (pp. 21-30). Geelong, Victoria: Deakins University Press.

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

Pang, T. (2002). Textual analysis and contextual awareness building: a comparison of two approaches to teaching genre. In A. M. Johns (Ed.), Genre in the classroom: Multiple perspectives (pp. 145-161). Mahwah, NJ: Lawrence Erlbaum.

Ramanathan, V., & Kaplan, R. B. (2000). Genres, authors, discourse communities: theory and application for (L1 and) L2 writing instructors. Journal of Second language writing, 9, 171-191.

Schleppegrell, M. J. (2004). The language of schooling: A functional linguistics perspective. Mahwah, NJ: Lawrence Erlbaum.

Tardy, C. M. (2006). Researching first and second language genre learning: A comparative review and a look ahead. Journal of second language writing, 15, 2,79-101.

Widdowson, H. G. (1993). The relevant conditions of language use and learning. In M. Krueger & F. Ryan (Eds.), Language and content: Discipline and content-based approaches to language study (pp.27-360). Lexington, MA: D. C. Heath.

Appendix See PDF