Adopting an SFL Approach to Teaching L2 Writing through the Teaching Learning Cycle

This study applied a Systemic Functional Linguistics (SFL) model to explore how 27 first-year university students in two different English proficiency groups improved their lexicogrammatical choices and metafunctions for writing analytical exposition essays during a 15-week course. To explore how “the teaching learning cycle” influences students’ understanding of the target genre essay, a survey was conducted; furthermore, to explore changes in students’ understanding of metafunctions (ideational, experiential, and textual meanings) of the target genre essay, students’ preand post-essays were scored by raters using the SFL framework rubric. Then, six students with lower rating scores at the pre-essay stage from both English proficiency groups were selected to explore how they progressed differently in the target linguistic resources. The results demonstrated that applying an SFL framework of writing assessment to English students’ understanding of essay writing can be used to explicitly examine their improvements.


Difficulties of Writing Essays in Higher Education in Japan
One of the issues facing Japanese university students today is that most have limited writing experience and few opportunities to write in English. English writing classes in Japan have mainly focused on writing abstracts rather than whole essay writing. Moreover, English students in Japan are exposed to a limited number of genre types and high school students have fewer learning opportunities regarding how to write. In this learning environment, Japanese university "English as a foreign language" (EFL) students struggle with English essay writing because they lack explicit writing instruction.
The concept of the Teaching and Learning Cycle (TLC) was originally developed in Australia (Martin & Rose, 2008) and later applied to writing and teaching research for both first language (L1) and second language (ESL) students of English, which demonstrated it to be an effective approach (Burns, 1990;Hammond & Hood, 1990).

Objectives of the Research
The TLC enables students to raise their awareness of both text and context (Burns, 1990;Hammond, 2001). By integrating language and content, the TLC can provide the scaffolding needed for students to communicate effectively in writing (Carlson, 2009), and can influence students' understanding of target texts' social and communicative purposes (Paltridge, 2001). The TLC can also be applied to the Genre-Based Curriculum Cycle to enable students to consider language at the text, clause, and sentence levels (Gibbons, 2002;Martin & Rose, 2008). Most research related to TLC and the genre-based approach (GBA) lessons have concluded that applying TLC and GBA facilitated improvement in students' awareness regarding their writing and the effectiveness of TLC and GBA; however, fewer studies have examined the scoring rubric and students' self-assessed areas of improvement and the usefulness of GBA writing lessons. Moreover, few studies have been conducted in Japan focusing on L2 writing and the EFL learning classroom using GBA.

Literature Review
In Japan, research has been conducted on uses of the TLC approach within the context of L2 academic writing in language classes. Explicit genre-based writing instruction can support L2 students with their writing output (Palmer, 2012). Palmer (2012) explored L2 writing using an Information and Communication Technology-mediated, Genre-based Approach (GBA) with an SFL framework for Japanese university students. Matsuo and Bevan (2002) understood that Japanese students might have a different mind-set than Australian students and feel more comfortable focusing on the grammatical features characteristic of the Sydney School approach (SFL). Thus, their work was less focused on teaching lexicogrammatical features. They applied a combined three-genre approach, including New Rhetoricians, Systemic Functional Linguistics, and English for Specific Purposes, to second-year English majors taking writing lessons at the university and implemented the following cycle: first, students were asked to write an essay on a specific topic, and teachers explained the audience and social context; second, students focused on language forms, but many previous studies did not mention the particular genre text as the model for the deconstruction stage. The results demonstrated that some students did not distinguish between the unique features of spoken language and written composition. Watanabe (2016) studied writing tasks from university entrance examinations in Japan and found an excessive focus on a limited number of genre types, most of which comprised expositions and personal reflections, with common essay questions formulated to persuade readers to think in a certain way. One of example questions is: Is eating junk food good or bad for people's lives? Present reasons for your opinion. It means that most learners are familiar with subjective and personal reflection essays; however, many of them may not have gotten enough exposure to a variety of genres such exposition, information report, and discussion genres.

Systemic Functional Linguistics and the Genre-Based Approach to Language Teaching
Within the SFL framework, written and spoken genre texts involve several states (structures) to achieve goals (e.g., discussion essays comprise four paragraphs to introduce both sides: introduction, pros, cons, and writer's opinion). Writers' goals and appropriate lexicogrammatical choices in written and spoken texts, such as procedure, recount, and description genre texts, are influenced by the writer's understanding of (a) what the text is about, (b) who the audience is, and (c) whether it is spoken or written, which is known as the "staged, goal-oriented, social process" (Martin & Rose, 2008). Martin and Rose's (2008) research on SFL genre theory in classroom contexts has influenced current generic structures and lexicogrammatical features of various genres. During the 1980s in Australia, Systemic Functional Linguistic Genre Pedagogy (SFL-GP) was developed (Martin & Rothery, 1980) and applied in various educational contexts by David Rose. In the SFL genre framework, students understand the literacy of a target genre in context and learn how to read and write through explicit scaffold writing instruction, in what is known as the teaching-learning cycle (TLC). However, several criticisms have been outlined in relation to the teaching of genre in education: for instance, since the SFL genre-based approach is explicit in nature, some believe it to be unnecessary except when editing texts, and that such explicit teaching can lead to overlearning and misapplications among students (Freedman, 1993). Moreover, practitioners may not understand a genre well enough to teach it, and if they are not sufficiently familiar with the target genre texts, they may not be able to fully understand the complex theory behind them. Finally, students who are taught to write using the explicit method may overgeneralize or misapply the rules, potentially leading to misconceptions and poor writing performance. However, a number of studies have addressed these critiques, and both qualitative and quantitative investigations have demonstrated the viability of explicit genre instruction as a teaching method (Chivizhe, McKnight, & Smith, 2003;Schleppegrell, 2010;Chen & Su, 2011;Yasuda, 2011Yasuda, , 2012Yasuda, , 2015Yasuda, , 2017).

Teaching Genre in Education
SFL is concerned with the relationship between language and society in a motived way and suggests that the social (genres) can be naturally related to language (meanings) (Martin, 2000). An SFL-GBA (genre-based approach) allowed participants to focus more on a contextual, purposeful, and reader-oriented view of writing. Tshotsho (2014) found it important to explicitly teach textual features in academic English writing. Gómez Burgos (2017) examined the expository essays of undergraduate students who had experienced a genre-based approach to writing in the TLC. Following the intervention period, their understanding of compositions had improved. Thus, many SFL genre-based writing studies have been conducted, most focusing on changes in students' understanding of the generic structure and lexicogrammatical features of target genre essays. However, few studies have tracked such features using the SFL framework of metafunctional features, that is, Ideational (experiential and logical), Interpersonal, and Textual meaning, for assessment.
Purposes of GBA by Swales (1990), Systemic Functional Linguistics (Sydney School) by Halliday (1994), and Rhetorical Genre Studies by Miller (1984Miller ( , 1994. Regarding the SFL framework in the genre-based approach, one of the unique features of GBA is to analyze multiple layers of one written or spoken sentence. Halliday (1994) developed a theory of the fundamental functions of language in which he categorized lexicogrammar into three broad metafunctions: ideational, interpersonal, and textual, each of which regard a different aspect of the world. Ideational meanings refer to what is going on, the participants, circumstances, and surrounding events, or in other words, the when, what, who, and background information (Droga & Humphrey, 2002;Kobayashi, 2017, p. 12); Interpersonal meanings refer to reader-writer relationships or ways of instituting relationships with others (Halliday, 1994;Oliveria, 2015, p. 231) and also "create impersonality, more subjective and involved style" (Kobayashi, 2017, p. 12;Oliveria, 2015, p. 233). "The interpersonal metafunction uses language to encode interaction and to show how defensible or binding we find our proposition or proposal" (Butt, Fahey, Feez, Spinks, & Yallop, 2000), which expresses phenomena such as living things and events (Gerot & Wignell, 1994). Finally, textual meaning is "language-oriented and deals with cohesive and coherent text production by organizing and structuring the linguistic information in the clause" (Halliday & Martin, 1981, p. 328;Haratyan, 2011, p. 261;Kobayashi, 2017, p. 12). Textual meaning conveys the message related with the text's coherence and cohesion (Gerot & Wignell, 1994).
In summary, when writers and speakers attempt to convey their intended meaning, these metafunctions can reflect the social roles of their texts through their choice of lexicogrammatical features. In the SFL framework, the three features of ideational, interpersonal, and textual meaning serve as a system network of meaning potentials with a set of semantic features. To fully understand a text, it is necessary to consider all these metafunctions simultaneously (Butt et al., 2000). Therefore, in this study, by analyzing or identifying lexicogrammatical features, including their ideational meaning, we can understand what picture of reality writers seek to convey and encode their experience in.

Research Questions
The current study was conducted in a first-year Pre-Enrollment English Program (PEP) at a Japanese university in which most English classes were comprised of novice students with limited previous experience of genre essay writing in English. In this study, analytical exposition essays were selected for the pre-and post-essays, as argumentation is a necessary skill for university courses and students are more likely to encounter this genre in their future study abroad experiences. The exposition essay's generic structure and language features were adopted from Martin and Rose (2008), Rose and Martin (2012), Butt et al. (2000), and Knapp and Watkins (2005). In contrast to many Japanese-based genre-writing studies, which have focused solely on assessing writing using SFL-GP (genre pedagogy) principles, I emphasize the important role of explicit instruction, as provided through SFL genre-based literacy intervention lessons, in allowing students to encounter different types of genre text over a 15-week course. I aimed to answer the following questions: (RQ1) How will L2 writers change from novice to experienced writers through the TLC?; (RQ2) How do learning tasks influence understanding of the target genre essay through the TLC?; and (RQ3) How does the EFL students' understanding of generic structure, lexicogrammatical features, and ideational meaning, experiential meaning, and text meaning exposition of the target genre essay change through the implementation of writing lessons using SFL GBA L2 writing?

Method
The research participants (N = 27; F = 20; M = 7) comprised first-year students in their second semester of the 2018 academic year, who attended two different classes taught by the author of this research (Table 1). Each class took six or seven different PEP English lessons (90 minutes per lesson) per week, such as reading, writing, speaking, and listening practice, in preparation for future study abroad programs. All students took the SFL-GBA writing lessons at a private university in Japan from April to August 2018. They were assigned to different classes according to their Test of English as a Foreign Language (TOEFL iBT) scores. Before delivering the informed consent materials, the study rationale and meaning of the research were explained to all participants and repeated each time the participants submitted relevant materials to the researcher. All participants understood the research purpose and that their identities would remain anonymous. Total 27 Note: A = The number of experienced students of SFL-GBA language learning lessons, B = Never previously learned the target genre, C = English Proficiency Group, D = Mean score of TOEFL iBT

Classroom Interventions
In the SFL model of reading and writing tasks, it is possible to design a curriculum and teaching plan with genres and to provide explicit guidance for each dimension of the tasks (Rose, 2015). A set of coherent learning stages derived from Feez and Joyce's (1998) TLC was applied to the course. These learning stages comprised: 1) building a context; 2) modelling and deconstructing an existing text; 3) joint construction of a text; 4) independent construction of a text; and 5) linking related texts. Three cycles of the TLC were introduced during the 15 weeks of the semester. Each TLC cycle took three or five weeks to complete.
The original five stages in the TLC were modified and extra scaffolding tasks were introduced for the EFL learning classroom contexts in Japan (Table 2). Table 2 provides an overview of the teaching methodology through the TLC. First, students listened to the explanation of generic structure and lexicogrammatical features of the target genre essay and deconstructed a model essay with their teacher. They then deconstructed and analyzed the model as the target genre text.
Second, students found authentic texts similar to the target genre essays in terms of structure and lexicogrammatical features. They deconstructed these texts to understand how to write the target essay. The essay topic was then introduced and students listened to the teacher's explanation and students had sufficient time to individually research the relevant information via the Internet before they began to write the target genre essay. They then engaged in a timed writing during class. The researcher wrote feedback on the target genre texts using a rubric-evaluation table of her own devising, which was based on Deng, Yang, and Varaprasad (2014), Melissourgou and Frantzi (2017), Ingebrand (2016), Whitfield (2000), and Pessoa, Mitchell, and Miller (2018). Before analyzing peers' essays, students had the opportunity to revise their essays. The students deconstructed and analyzed their peers' essays to understand the uses of lexicogrammatical features and the generic structure of the exposition (analytical) genre essay. For instance, students identified target vocabulary items such as auxiliary verbs (i.e., as modalities to express their opinions). In Task 2, they summarized the results of their peer genre analysis. All of these SFL GBA lessons were conducted by the researcher of this study as the teacher. Table 2. Example of teaching and learning cycle in the ESL writing content for this study (weeks 6 to 10)

) Building the context
To understand the purpose of the text and when the target text is used (A) To have peer and group discussions about why they are writing this target genre essay, its purpose, and reader-writer relationships (B) To connect their future academic experience (study abroad) and the target genre essay Week 6 Week 7 2) Modelling and deconstructing an existing text To understand the structure and lexicogrammatical features of the target genre text (A) To find an authentic text as similar to the target genre text as possible, through online sources; (B) to analyze its generic structure and vocabulary; (C) to write a reflection sheet (process/procedures were based on the Nagao, 2018); (D) to analyze the model text that the teacher prepared using the SFL framework rubric Week 8 3

) Joint construction
Before students write the target genre essay: To find similar texts to the target genre text Note. This teaching and learning cycle has been modified from Nagao (2018).

Phase 1: Survey to Understand the Influence of the TLC on Students in EFL Classrooms
To explore RQ1, how learning tasks influence students' understanding of the target genre essay through the teaching learning cycle, students answered the survey once they had completed the TLC. The survey questions were divided into two parts. Part 1 evaluated students' awareness of their engagement in tasks during each stage of the TLC. They assessed each statement on a 5-point Likert scale (5 = "strongly agree," 1 = "strongly disagree"). Survey questions were qualitatively analyzed to evaluate how useful the students felt tasks were in each stage of the TLC (Table 3). In part two, the multiple-choice questions were designed to assess students' difficulties when participating in the tasks.

Pre-Instruction and Post-Instruction Essays
In Phase 2, to obtain in-depth qualitative insight into students' generic structure and lexicogrammatical features awareness, the exposition (analytical) genre essays were collected before and after the classroom intervention. In the first data collection (before the classroom intervention, Week 5), participants were assigned a timed writing task (pre-essay writing task) for 50 minutes during class time. All participants wrote an analytical exposition essay without explicit instruction on how to construct their texts. However, they were briefly informed on how to write the target essay generic structure. The topic for the pre-analytical exposition genre essay was whether to support or oppose the self-responsibilities of journalists in conflict zones. Explicit explanation of the relevant background information on the topic was introduced by the teacher before the students started to write the pre-essay.
Regarding the post-essay timed writing task in week 9, the topic of the post-analytical exposition genre essay was to support or oppose early EFL education in elementary schools in Japan. When they had completed the classroom interventions of stages 1, 2, and 3, all participants completed the timed writing task (50-60 minutes). The post-essay data were collected during stage 4 ( Table 3).
The target genre in this study was analytical exposition, introduced during the second TLC (Weeks 6 to 10). The purpose and generic structure of the target genre were to introduce a perspective related to the topic and then provide supporting evidence. The generic structure of the analytical exposition genre is: Statement of position ^ Arguments ^ Reinforcement of statement of position (Adiantika, 2015;Emilia, Habibi, & Bangga, 2018;Martin & Rose, 2008). Writers provided examples and statements that supported their arguments; in other words, they used condensed language and specific information from other texts (Hasanah, 2017;Schleppegrell, 2010). According to the author's observation in the classroom, many Japanese university students struggle to employ textual resources (e.g., cohesive links) in the analytical exposition genre.

Phase 2-1: Scoring the Target Genre Essays Using the SFL Framework Rubric
To explore RQ2, participants' essays were rated (scored) using the SFL framework rubric (Pessoa et al., 2018). The first part of Phase 2-1 comprised comparison between the higher and lower English proficiency groups' understanding of three metafunctions of the pre-and post-analytical exposition genre essay (Pessoa et al., 2018). Additionally, the second part of Phase 2-1 comprised selecting data from three students from the lower and higher rating scores groups on the pre-essay task and analyzing their pre-and post-essays to understand the change in their lexicogrammatical features based on the three metafunctional categories as rubric (Pessoa et al., 2018).
To track changes in students' understanding of lexicogrammatical features of the analytical exposition genre essay, one rater assessed the students' pre-and post-essays using the assessment rubric deductively ( Table 4). The rater for phase 2 was a researcher who was familiar with the theoretical aspects of SFL. The inter-rater reliability text was α = 0.80. The criteria of the rating rubric (see Table 4) were adopted from Pessoa et al. (2018, p. 87), and also referred from Martin and Rose (2008), Rose and Martin (2012), Butt et al. (2012), Knapp and Watkins (2005), Emilia et al. (2018), and Promwinai (2010). To summarize the rubric, regarding ideational meaning, students were required to provide accurate and relevant knowledge on the topic from online resources or subject reading materials. They were also expected to use specialized or technical vocabulary related to the topic to show both perspectives. Regarding experiential meaning, students were required to present their opinions using appropriate lexicogrammatical features rather than merely stating the facts of the events. This involved using external voices to provide relevant evidence for the central claims of the essay by summarizing the factual contents from the reading materials. Finally, regarding the textual meaning of the analytical exposition genre essay, students were required to use specific vocabulary items and fixed phrases related to coherency and cohesion. These criteria were explained during the lesson and students were given exposure to analyzing the target genre text using these criteria by applying genre analysis to their peers' essays and writing up the analysis sheet. The essay is grounded in accurate and relevant knowledge from the source text(s). (For post essay: a set of journal articles was provided to students; three articles in Japanese, three articles in English at stage 3: joint construction of the teaching and learning cycle).

2
The thesis uses specialized/technical vocabulary to characterize an overarching claim.

3
The supporting claims are relevant and clearly create an analytical framework for the essay.

4
Readers can clearly understand when, where, what, and who because relevant lexicogrammatical features are used in this essay (particularly in the first paragraph).

Interpersonal meaning 5
The essay uses expanding resources (attribute) to bring in the source text. (i.e., learners are allowed to cite information from other materials that were recommended by the teacher.) 6 The essay uses contracting resources (endorsement) to show how the cited material supports the claims.

7
The essay uses lexicogrammatical features related to the writer's opinions to show their support for or opposition to the topic (e.g., modalities: auxiliary verbs and -ly adverbs).

8
The writer avoids using lexicogrammatical features related with personal pronouns, especially "I," and instead uses other words to replace subjective words.
Textual meaning 9 The writer uses reporting verbs, e.g., Author's name (2005), such as claimed that, argued, supported, explained, criticized that.  Vol. 13, No. 6;2020 scale was: 1 -Poor execution or no use or almost no use of linguistic resources, 3 -Limited use of linguistic resources, 5 -Only fair or problematic use of linguistic resources, 8 -Fairly good use with minor problems or inconsistencies using linguistic resources, and 10 -Excellent use of linguistic resources.

Phase 1: Students' Reflections on the TLC in EFL Writing
The results of descriptive statistics for the 5-point Likert scale demonstrate that most students found stage 3 to be the most beneficial: Joint construction of the text, helpful for understanding (or writing) the target genre essay (M = 4.26, SD = 0.81). Participating in stage 3 enhanced their understanding of SFL ideational meaning such as the participants and circumstances surrounding events on the exposition genre essay topic. Furthermore, many students found editing and revising the essay, individually and with the teacher's guidance, useful for understanding how to write the target essay. This above task was conducted between stage 4: Independent construction of the text and stage 5: Joint text, in Week 10 (M = 4.22, SD = 0.64). Additionally, students' awareness of Stage 2 of the TLC was useful for writing the analytical exposition essay was also higher (M = 4.07, SD = 0.83). Finally, students' understanding of these three metafunctional features was also higher than other tasks: e.g., interacting with peers (exchanging information with peers) to discuss the essay topic inside/outside of the classroom was helpful for writing the analytical exposition essay (TLC stage 1. building the context and stage 3. joint construction of the text); analyzing and referring to the model text of the analytical exposition essay was useful for writing the analytical exposition essay; 2. modelling and deconstructing the text). However, fewer students found that their prior writing experience of analytical exposition genre essays (e.g., at high school) was extremely useful for writing the target essays (M = 2.96, SD = 1.37). The results of former writing experience of the target genre differed between higher and lower English proficiency groups. For instance, most higher proficiency students reported having no experience of writing training in exposition genre analysis (M = 2.77), while six students reported having 70% (rather yes) of the writing analytical exposition essay experience, with a mean value of 3.14 among the lower proficiency groups (Appendix A). This result implied that lower proficiency students might have had previous writing experience of this target genre essay. However, according to Watanabe (2016), high school students are exposed to limited varieties of genre for their entrance examinations, with typical genres they practice comprising personal recount and personal opinion essays. In other words, the novice L2 writers in this study were probably not equipped with the skills required to evaluate information and perspectives, select and interpret evidence to support their claims, or to control a range of linguistic resources (Pessoa et al., 2018, p. 81).
Additionally, most higher proficiency students chose "It was challenging for me to identify the zig-zag writing pattern technique in a peer's analytical exposition essay" (M = 4.38). This may be because identifying the zig-zag writing pattern was less common in their L2 writing history. Successful L2 writers are expected to employ a zig-zag pattern of theme progression in their essays (Emilia et al., 2018).
The second part of the survey was multiple choice and asked students what tasks were challenging during the TLC. The findings demonstrated that (a) reading the journal articles (n = 5), (b) revising the summary from the journal articles (n = 5), and (c) identifying the zig-zag writing pattern technique in a peer's analytical exposition essay (n = 5) were challenging for higher proficiency students, while summary writing of the contents from journal articles was the most difficult task for lower proficiency students (Appendix B).
Many lower proficiency students chose "It was challenging for me to read the journal articles that the teacher provided (six different journal articles)" (M = 4.29). This may be because these students had less experience reading academic journals and had limited experience evaluating information and perspectives and selecting and interpreting evidence to support their claims.

Phase 2-1 Comparison of Rating Scores between Lower and Higher English Proficiency Groups
To explore the changes in EFL students understanding of the metafunctions of lexicogrammatical features, the raters' scores on the pre-and post-essays were examined using the rubric assessment criteria. The quantitative analysis involved reliability statistics for these ideational, interpersonal, and textual metafunctions and demonstrated Cronbach's alphas of 0.95, 0.80, and 0.79, respectively, for the lower proficiency group and 0.86, 0.88, and 0.82, respectively, for the higher proficiency group.
The results of comparison of mean scores (average scores) on the pre-and post-analytical exposition essays demonstrated that, overall, the mean scores of post-essay results improved in both groups (Appendix C). In detail, the results of the mean values of the pre-essay scores (see Appendix C and Figure 1) demonstrated that both groups scored higher on ideational meanings than other metafunctions.
The results of comparisons between post-scores and pre-scores illustrate that the greatest improvement among the lower proficiency group was in interpersonal meaning, which is related with impersonality, personality, subjectivity, and objectivity of the target text (post score = 20.79, pre-score = 6.14). The result of comparison mean scores for the lower proficiency group is that larger improvements occurred in interpersonal meaning (post score = 25.0, pre-score = 13.3) and textual meaning: cohesion and coherence of the target text (post-score = 19.3, pre-score = 7.8) (See Figure 1).

Figure 1. The results of comparison of mean scores (average scores) on the pre-and post-analytical exposition essays
Note. The lower English proficiency group is shown on the left and the higher English proficiency group is shown on the right.
Ideational meaning is related with students understanding of background information regarding the essay topic. Regarding ideational meaning among the lower proficiency group, the scores improved dramatically for the features: 1. grounding in accurate and relevant knowledge from the source texts; 2. using specialized/technical vocabulary to characterize an overarching claim; and 3. it seemed difficult to support claims that are relevant and clearly create an analytical framework for the essay (Appendix C) while clearly writing specific features (when, where, what, and who) in the first paragraph. For the higher proficiency group, the mean value score of the above features on their pre-essay was slightly higher than for other features, while the other three features of the ideational meaning scores were similar. The post-essay mean value score showed improvement in 1, 2, and 3, but 4 improved only slightly.
Interpersonal meaning refers to expressing ways of instituting relationships with others. At the pre-essay stage, the mean score of interpersonal meaning among the lower proficiency group was higher for item 8: Replace subjective vocabulary such as personal pronouns, especially "I," with other words than for the other three features. Additionally, they demonstrated their difficulties in using lexicogrammatical features related to the writer's opinions to demonstrate their support for or opposition towards the topic (e.g., modalities: auxiliary verbs and -ly adverbs in the body) at the post-essay mean score. The higher proficiency group made dramatic improvements in the mean score on 5. The usage of expanding resources (attribute) to bring in the source text and 6. The usage of contracting resources (endorsement) to demonstrate how the cited material supports the claims in interpersonal meaning. They also acknowledged how challenging it was to improve the usage of modalities such as auxiliary verbs and -ly adverbs.
Textual meaning refers to cohesive and coherent text production. In both proficiency groups, few lexicogrammatical features were found in their essays that were related with 9. The uses of reporting verbs, such as author name (published year) claimed/argued/suggested that and 10. Nominalizations (e.g. 10. nominalization: M = 0.36 for lower; M = 0.75 for higher English proficiency group). On the other hand, most students managed to use 11. linking or signpost words and 12. sub-claims at the beginning of the paragraph in the pre-essay. In the post-essay, the score of textual meaning, the score for 9. to use reporting verbs, increased sharply in both groups (i.e., difference post -pre: 4.43 for lower and difference of 3.92 for higher English proficiency group).

Phase 2-2: Individual Changes; Comparing Raters' Scores
In this section, to understand how individual students' scores on ideational, interpersonal, and textual meaning changed from the pre-to post-essays, each of the score elements was analyzed. The lowest rating score at the pre-essay, mid-rating score at the pre-essay, and highest score at the post-essay from both proficiency groups are summarized in Figure 2 and Appendix D. Students with lower scores on the pre-essay had limited understanding of the three metafunctions at the beginning and, in particular, their understanding of the textual meaning of the essay did not improve significantly in the post-essay. For instance, according to the analysis sheet of student No. 8 in the lower proficiency group, which was collected between stages 4 and 5 (Table 2), "in terms of writing nominalization in my and my classmate's essay, no nominalization was found. Personally, I have some difficulties with writing sentences using the participial construction as nominalization in the essay. I usually write simple structure sentences (e.g., subject, verbs, and objectives, such as 'The Japanese government will introduce the English curriculum at elementary schools') and I realized that I need to practice nominalization to be able to use this structure in my essay" (quote from student No 8's genre analysis sheet, personal communication, 11/28/2018/). The change in both groups' metafunctions between the pre-and post-essays were relatively similar. In the post-essay, the scores of the three metafunctions were dispersed; thus, the score for ideational meaning was highest, followed by interpersonal meaning and textual meaning on the post-essay tasks. These results may indicate that it is easier to improve understanding of ideational meaning while understanding of textual meaning can be challenging.
The common feature of students who scored higher in their pre-essay was that the textual meaning score was lower than the other two metafunctions in both the pre-and post-essays. The genre analysis sheet of student No. 9 in the higher proficiency group was collected between stages 4 and 5 (Table 2), as this learner focused on using reporting verbs and nominalization to explore the lack of use of textual meaning in his or her essay. "Nominalization: I found two sentences were nominalized in my essay while I could not find any in student No. 3's essay. I assume that student No. 3 could start with -ing sentences instead of writing sentences that started with using non-human nouns often. In addition, for the textual meaning in this essay, I used the reporting verbs 'stated' and 'reported' and I found the reporting verbs 'investigated' and 'advocated' in student No. 3's essay. To show supportive or opposing evidence on the topic, the vocabulary item 'advocated' is appropriate. I have never used this word" (extract from the genre analysis sheet of student No. 3 in the high proficiency group, personal communication, 11/20/2018/).

Discussion
In this study, I modified a set of learning stages regarding the TLC to explicitly teach first-year EFL students how to write in the analytical exposition genre. I documented the writing development of twenty-seven students, focusing specifically on six students, particularly on experienced and novice writers. The quantitative analysis to explore (1) how learning tasks influence students' understanding of the target genre essay through the TLC found that most students in both the lower and higher proficiency groups lacked prior writing experience of analytical exposition genre essays. One of the challenges for students was to include description, explanation, and information reports in the target essay. Features of "subjective or objective" and "personal reflection or not," which influenced their lexicogrammatical features choice and meaning-making choice, influenced their exposition (analytical) genre essay writing. These features are related to the interpersonal meaning of the SFL rubric in Table 4. For example, in the pre-essay, 7 out of 14 in the lower proficiency group and 3 out of 11 in the higher proficiency group avoided using the personal pronoun "I" to express their opinions. Another possible technique is to use contracting resources (endorsement) to show how the cited material supports the claims and to use cited information with reporting verbs, e.g., AAAA (2019) claimed that / argued Multiple tasks at each stage of the TLC influenced students' understanding of how to write analytical exposition essays. The survey results demonstrated that the tasks related to "Modelling and deconstructing the text" (stage 2 of the TLC), particularly the task of highlighting specific vocabulary, were the most helpful for the lower proficiency group. The revisiting and editing task and exchanging information with peers in order to understand the topic content also enhanced their understanding of the generic structure, lexicogrammatical features, and three metafunctions. Students in the higher proficiency group found editing and revising their essay to be the most helpful task, followed by deconstructing the model essay, and peer interactions to exchange information in order to understand the topic content. Overall, Weber (2001) states that this model of teaching and learning writing can enhance awareness of particular areas of difficulty in writing the target genre essay. In this study, for both groups of students, some tasks related with "modelling and deconstructing the text" could impact their understanding of the target genre text.
The quantitative analysis found that students incorporated the targeted lexicogrammatical features in their writing in ways that enhanced their understanding of appropriate lexicogrammatical choices for the target genre text. To explore RQ2, how students understand generic structure, the lexicogrammatical features, ideational meaning, experiential meaning, and text meaning exposition of the analytical genre essay changed over the 15-week course in which participants wrote essays rated using the SFL framework rubric. Overall, participants' scores in terms of all metafunctions, ideational, interpersonal, and textual meaning, improved in the post-essays. However, the changes in ideational meaning, which is related to understanding of the background information of the essay topic, demonstrated only slight improvement. A notable result was that lower proficiency students understanding of interpersonal meaning and textual meaning were dramatically improved in the post-essay. Using the explicated analytical learning methods of SFL-GBA possibly enhanced their understanding of appropriate vocabulary choices (Yasuda, 2015). This was particularly the case for students who scored lower on the pre-essay. Through the SFL genre-based approach to writing lessons, their understanding of these two metafunctions, ideational and interpersonal meaning, achieved similar scores in the post-essay. The changes to metafunctions among individual cases demonstrated that textual meaning can be the key target feature in this study. Students with the highest rating score in the pre-essay demonstrated significant improvement in the textual meaning in the post-essay, while students with the lowest and mid-scores in their pre-essay had difficulties improving the textual meaning in the post-essay score.

Conclusion
It is important to acknowledge other factors that may have impacted students' writing development, such as students' previous learning experiences. The input tasks related to the target genre texts can improve the quality of feedback from instructors and peers, their input and output in terms of reading and writing in other lessons, and their clear understanding of the learning goals for writing class. Additionally, the contents of the SFL rubric need to be developed. The conclusion regarding the effectiveness of the teaching methodologies and introduction of relevant tasks may also be limited, as no control groups were available for comparison of the results. Further research could introduce more relevant reading and writing tasks (i.e. Reading tasks should be related to students' interests and curriculum; different types of tasks and editing tasks may be necessary during the Joint construction.) Despite these above limitations, the study findings discovered some positive support for the degree of explicit disciplinary writing instruction with an explicit focus on generic structure and lexicogrammatical features, particularly in narrowing the gap between different levels of English proficiency. Furthermore, this research highlights the value of introducing the writing teaching methodology at the tertiary level. Doing so exposed students to different genre texts to meet the needs of academic writing in higher education in Japan. Furthermore, a new assessment rubric using the SFL framework was developed in this research to enable SFL to be used in language learning classroom contexts. In the future, this SFL rubric will need to be tested in different reading and writing contexts. By connecting students' assessments related with their understanding of linguistics resources to their ability to write effectively, I hope that applying the SFL genre-based approach to writing in EFL classrooms will allow English students to see the structure of the texts and learn how to control their lexicogrammatical choices in order to express their meaning more fully when writing texts.