Writing Instruction for English Learners: Examining Instructional Practices in Fourth-Grade Classrooms

Abstract This qualitative study examined writing instruction in two linguistically diverse fourth-grade classrooms in order to determine the genres taught and the instructional practices favored by teachers. Researchers observed writing instruction and interviewed teachers in a culturally and linguistically diverse elementary school in Southern California to study how teachers worked with children to develop their capacity to negotiate different writing tasks. Findings revealed that students were engaged in content-area expository writing, and the writing assignments were influenced by assessment requirements. Both teachers evidenced explicit instruction of academic language, attention to genres, and scaffolding for writing. Observations revealed that teachers tended to focus their feedback on word and sentence level discourse during classroom instruction. For lengthier pieces, teachers used mentor texts and heavy scaffolding to ensure that every student in the class would be able to produce the writing required. Implications for professional development are discussed.


Classrooms
In this qualitative study, we investigate writing instruction in two culturally and linguistically diverse fourth-grade classrooms in one elementary school in Southern California.
Over the course of one semester, we observed the writing instruction provided to English Learners (ELs), students who had been reclassified as fluent in English, as well as monolingual English speakers, in upper-elementary mainstream settings. We were interested in examining writing instruction for upper-elementary ELs because research demonstrates that writing is a challenging language domain that requires explicit instruction and modeling from teachers (Gibbons, 2002;Reyes, 1992;Rose & Martin, 2012); yet writing instruction is often overlooked in the intermediate elementary grades (Gilbert & Graham, 2010). Despite recommendations by the National Commission on Writing (NCW, 2006) to double the amount of time spent on writing instruction, writing continues to be a neglected discipline with teachers dedicating an average of 15 minutes per day to writing (Gilbert & Graham, 2010).
Even though children in upper elementary school are expected to produce a wide variety of increasingly complex texts (Pytash & Morgan, 2014), students most often engage in writing to learn (Gilbert & Graham, 2010). Writing to learn is a term used to describe the use of writing as a thinking tool to enhance one's own understanding of a topic (Fry & Villagomez, 2012). While some research has demonstrated the benefits of writing to learn to enhance student learning in the content areas (Bangert-Drowns, Hurley, & Wilkinson, 2004), writing to learn tasks do not usually focus on apprenticing students into learning to negotiate different text genres. Gilbert and Graham (2010) noted that for students in grades four through six, writing to learn tasks included taking notes, composing short responses, or completing worksheets. Such writing activities are insufficient for developing expertise in creating the complex texts now required by the Common Core State Standards (2015). In the fourth grade, students are expected to write opinion pieces and informational texts that incorporate evidence and logical argumentation (Bunch, Walqui, & Pearson, 2014;Common Core State Standards Initiative, 2015). In this study, we investigate how fourth-grade students are taught to produce these lengthier and more complex texts.
Alongside concerns about time dedicated to writing and opportunities to write in different genres, the NCW (2006) suggests that more professional development is needed to help teachers cultivate the confidence and ability to create and deliver engaging writing curricula, particularly with culturally and linguistically diverse populations. The NCW has acknowledged that the needs of ELs are important to consider since this growing population brings unique language resources and life experiences with them as they learn to write and express their ideas in English.
When designing and implementing effective instruction for linguistically diverse youth, teacher preparation is essential (Ballantyne, Sanderman, & Levy, 2008;Garcia, Arias, Harris Murri, & Serna, 2010). However, only about 29.5% of mainstream teachers in the United States have had any relevant coursework or professional development, and only four states require preservice teachers to complete coursework focused on teaching linguistically diverse students (Ballantyne et al., 2008). Although California requires preservice teachers to complete coursework focused on teaching culturally and linguistically diverse students and obtain the Cross-Cultural Language and Academic Development (CLAD) credential, research demonstrates continued inequities in the education of ELs (Gándara, Maxwell-Jolly, & Driscoll, 2005; UC Linguistic Minority Research Institute (UC LMRI), 2008). Gándara and her colleagues have consistently documented that ELs lack access to appropriate education and qualified teachers.
They argue that this has resulted in underperformance on standardized measures, with EL students posting steady declines in their English Language Arts scores as they progressed through middle and high school (UC LMRI, 2008). Gándara et al. (2005) outline four essential attributes of teachers who are effective teachers of ELs: • Ability to communicate with students; • Ability to engage students' families; • Knowledge of language uses, forms, mechanics, and how to teach these; • A feeling of efficacy with regard to teaching English language learners (p. 3) In the present study, we focus on the third attribute-the knowledge of language uses, forms, and mechanics and how to teach them-and on two experienced teachers who were identified by the district as prepared to work with ELs. Our purpose is to better understand how these teachers instruct their students in the writing genres used in their fourth-grade classrooms. In this paper, we explore the kinds of writing tasks that students are assigned and learning to produce, and we examine the instructional approaches used. Our specific research questions include: 1. What kinds of writing are ELs asked to do in school? 2. What genres are students learning to write? 3. How do teachers instruct students to write these genres? a. What instructional scaffolds are used? b. What kinds of feedback do students receive during whole-class instruction?

Research on Writing Instruction for English Learners
Research documenting writing instruction for English learners in the upper elementary years is an area of growing interest, particularly as the diversity of the school-aged population increases. Studies examining EL writing instruction present conflicting views regarding the value of process approaches for ELs (Edelsky, 1986;Reyes, 1992;Peregoy & Boyle, 2013;Rose & Martin, 2012). Current research points to the importance of explicit instruction in writing genres and language forms and functions (Schleppegrell, 2004), along with the provision of time and space to explore topics and themes that are meaningful, authentic, and culturally relevant (Schecter & Cummins, 2003).

Process Writing Approaches and English Learners
Edelsky's (1986) early research exploring young bilingual children's writing development demonstrated the importance of providing rich contexts for writing and many opportunities to write. Edelsky asserted that many features of children's writing could not be traced to direct instruction, and proposed that the children learn to write through an inquirybased process of hypothesis testing. However, some research has highlighted limitations for using process approaches with ELs, particularly when teachers fail to provide sufficient explicit instruction related to the organization of texts, or the development of an argument or point of view (Fang, Schleppegrell, & Cox, 2006;Reyes, 1992;Rose & Martin, 2012). This research calls into question the notion that students' writing will develop and improve naturally as long as students have many opportunities to write and are exposed to good writing models (Reyes, 1991;Rose & Martin, 2012). For example, Reyes' (1992) study on the use of dialogue journals and literature logs to develop bilingual students' writing skills demonstrated that students did not attend to the teacher's suggestions related to correct form, grammar, and usage unless these corrections were made explicitly. Similarly, McCarthey and García (2005) found that in the classroom where dialogue journals were implemented, students lacked a clear understanding of the purpose of their teacher's comments and questions, and so generally ignored them. Rose and Martin (2012) contend that when teachers employ process approaches without clear modeling, scaffolding, and instruction, the end result is to limit student writing production and growth. Rose and Martin (2012) point out that process writing instruction was developed for and oriented towards the middle class and "was not designed to provide access to education for marginalized groups" (p. 3). They argue that students must be appropriately prepared to engage with a wide range of topics, themes, and genres. In order to address these concerns, researchers have been working with teachers to apply systemic functional linguistics (SFL) and genre-based approaches to teaching writing in the elementary years.

Genre Pedagogy and Systemic Functional Linguistics
The genre pedagogy approach to teaching writing is rooted in systemic functional linguistics (Halliday, 1978), a theory of language that emphasizes the relationship between the socio-cultural context of communicative events and the grammatical choices that speakers and writers make. Functional linguists focus on analyzing grammar as a meaning-making tool (Martin & Gebhard, 2011). In any communicative exchange, individuals make linguistic choices related to the three interconnected metafunctions of language-the ideational, the interpersonal, and the textual. Our language choices are influenced by the topic, social setting, our goals, our status in relation to other people, and the mode of communication (oral, written, e-mail, etc.).
Taken together, these linguistic choices create different registers, while text genres or text types indicate the broader social purpose of a text (Rose & Martin, 2012). Genres are staged, goal oriented, and social (Martin, 2009). For example, the stages of an argument or exposition require the author to state the thesis, state and explain the arguments in relation to the thesis, and then conclude with a recommendation. Genre pedagogy encourages teachers and students to engage in the analysis of texts in order to identify and describe the different stages, purposes, and audiences associated with particular genres.
In their work in diverse public schools in Sydney Australia, David Rose and J. R. Martin (2012) and their colleagues developed the teaching/learning cycle for introducing genres to children and young adults. The teaching/learning cycle is an iterative process of deconstructing model texts, building background knowledge of the field or topic, joint construction of texts, and eventually independent production of specific genres. Rose and Martin (2012) explain that in the Australian context, as in the U.S., writing instruction focused on process approaches, and teachers avoided direct instruction of grammar, which was seen as the rote memorization of meaningless rules. The teaching/learning cycle was formulated to build students' and teachers' knowledge about language and allow them to explicitly discuss what makes particular texts more or less effective.
In the U.S., educational linguists have argued that it is difficult for teachers to instruct students to be effective writers of academic texts because teachers themselves "often have not developed an explicit understanding of how language works in the texts they routinely require students to read and write in school" (Gebhard, Chen, Graham, & Gunawan, 2013, p. 108). Some initiatives to increase K-12 teachers' knowledge about language have included the Access to Critical Content and English Language Acquisition (ACCELA) Alliance at the University of Massachusetts (Brisk & Zisselsberger, 2011;Gebhard, Willet, Jiménez Caicedo, & Piedra, 2011) and the California History Project (Achugar, Schleppegrell, & Oteíza, 2007;Schleppegrell, Achugar, & Oteíza, 2004;Schleppegrell, Greer, & Taylor, 2008). During both projects, participating teachers analyzed state-mandated curricula to determine the kinds of genres that students were required to learn. Through analyzing curricula and texts, teachers developed a deeper understanding of the relationship between the functions of a text and grammatical choices, as well as an understanding of the stages associated with different genres. Teachers were able to apply their new understandings of grammar and genre to the analysis of their students' writing, as well as lesson development (Brisk, Hodgson-Drysdale, & O'Connor, 2011;Gebhard et al., 2013). After providing instruction in textual analysis, teachers involved in the California History Project found that their students performed better on standardized tests and were better able to navigate and produce expository genres (Achugar et al., 2007;Schleppegrell et al., 2004).
These interventions demonstrate that given professional development and sufficient time, teachers are able to engage their students in textual analysis. This in turn can empower young writers and result in improvements in their writing capabilities with regards to lexical diversity, complexity, and cohesion (Brisk & Zisselsberger, 2011;Gebhard et al., 2011;Schleppegrell et al., 2004;Schleppegrell & Go, 2007).
Developing teachers' knowledge of academic language and the features of different genres can also improve teachers' feedback on student writing (Aguirre-Muñoz, Park, Amabisca, & Boscardin, 2008). Aguirre-Muñoz and her colleagues found improvements in teacher feedback on student writing after a one-week introduction to the theoretical framework and instructional applications of SFL. The authors noted that before the professional development (PD), teachers' feedback focused on paragraph structure and mechanics (Aguirre-Muñoz et al., 2008, p. 305).
Such feedback did not always lead to improvements in student writing because teachers focused on correcting rather than teaching. For example, teachers asked students to increase the clarity or cohesion of texts, but did not demonstrate how to do so. The PD sessions assisted teachers in analyzing texts to identify the rhetorical features they were attempting to teach. After the PD, subsequent observations of instruction revealed that teachers who implemented the SFL approach to a high degree found improvements in students' writing in regards to clarity, varied noun phrases and verb choices, cohesion, and theme development. Based on the results of these studies, it appears that SFL and genre pedagogy hold much promise for improving writing instruction for ELs in upper elementary school. Still, McCarthey and Ro (2011) caution that teachers may employ genre-based approaches without attending to the social meanings or purposes of text types. Approaching genre pedagogy as the rote learning of specific text types that can be realized by following a recipe or graphic organizer could undercut the broader goal of genre pedagogy, which is to provide all students with access to academic language and "democratiz(e) the outcomes of education systems" (Rose & Martin, 2012, p. 4).

Methods
Due to the social and interactive nature of writing and writing instruction, we employed qualitative research methods, primarily observations and interviews, to investigate the kinds of writing in which children and their teachers engaged. Data were collected during the 2011-2012 academic year. The names of the school and participants have been changed to preserve confidentiality.

Research Context
Monument Elementary, which was selected on the basis of access and demographics, is located in a large urban area, and served 581 students during 2011-2012. At the time of this study, approximately 93% of the students were Hispanic, compared with 78% citywide, and 283 students were classified as English learners (48.7%). The primary language of the ELs was Spanish (281; 99% of English learners). In addition, 518 of the families (89.2%) qualified for free or reduced-price lunch. Teachers in this school district had received on-going professional development to address the needs of English learners. Prior to the 2011-2012 school year, the emphasis of the district professional development efforts had been on improving teachers' instruction of academic English by focusing on academic sentence structures (Dutro & Moran, 2003) and academic vocabulary (Marzano, 2004), specifically through the use of sentence frames.

Participants
Fourth-grade classroom teachers and students were invited to participate in this study.
We elected to focus on fourth grade because expectations for students' writing increase substantially from third to fourth grade. In the fourth-grade, students are expected to be able to integrate evidence from outside sources into their writing and are expected to create more complex texts (CCSS, 2015). Teachers were selected based on the principal's recommendation.
The researchers met with the principal, explicated the purpose of the research, requested permission to conduct research, and asked the principal for recommendations regarding participants. The principal selected two fourth grade teachers that she felt were exemplary, Ms. Nara and Ms. June.
Ms. Nara. During our interview, Ms. Nara revealed that she was a language learner herself, immigrating to the United States from the Philippines when she was four years old. She is fluent in two languages, English and Tagalog; with Tagalog being her first language. She is also proficient and certified to teach in Spanish, as she possesses a supplemental authorization known as a Bilingual, Crosscultural, Language and Academic Development (BCLAD) certificate. In addition, Ms. Nara has a Master of Arts in Multicultural/Bilingual Education. She stated that the district had provided her with several day-long trainings to teach writing (Step Up to Writing) and English Leaners (Specially Designed Academic Instruction in English, SDAIE).
She has been teaching for a total of 22 years, and has been assigned to fourth grade for the last six years. During her 22 years, she also taught kindergarten and first grade in the primary language (Spanish) program.

Ms. June.
Ms. June shared that English is her first language, although she is also fluent in Chinese. During the interview, she explained that she went to Chinese school on Saturdays as a child. For her undergraduate degree, she majored in English. She also possesses a supplemental BCLAD authorization. She has participated in the same professional development opportunities as Ms. Nara with regard to writing and English learners. She has been teaching for 17 yearskindergarten for one year, second grade for 14 years, and fourth grade for the past two years.

Students.
All of the students in both classrooms, regardless of language background, were asked to participate. Of the 63 students, 33 (52%) chose to participate, with participation consisting of consenting to observation of classroom instruction and collection of writing samples. Student achievement data were not collected, but the ELs' language proficiency data were provided to the researchers. Six of the 33 participating students (18%) were classified as EL and their English language development levels ranged from intermediate to advanced. An additional 16 of the 33 participating students (48%) had been recently reclassified as Fluent English Proficient. The study also included 10 monolingual English-speaking students (30%).

Data Collection and Analysis Procedures
Observation of classroom instruction was the primary method of data collection for this study. Two of the authors conducted observations at the research site. To begin, pre-observation interviews were conducted in order to discuss the teachers' approaches to writing instruction and to ask each teacher when the researcher would most likely observe writing instruction. The in-person interviews were audio-recorded, and transcribed by the researchers later. The interview protocol consisted of 25 items focused on the teachers' philosophy, training, and instructional approaches related to writing (see the Appendix). Next, researchers conducted non-participant observations of each classroom over the course of an entire school day in order to gain insight into the ways in which writing and writing instruction were employed. Subsequent observations were scheduled based on when writing activities or writing instruction were most likely to occur, including in content areas. Teachers were not asked to prepare special lessons. Over the course of the spring 2012 semester, researchers observed between 15-20 hours in each classroom taking detailed field notes during each observation. Samples of student work were collected during each observation session in order to examine students' responses to instruction and to support field notes. Researchers engaged in informal conversations with the teachers before and after observation sessions, both in person and via e-mail, to clarify the purpose of writing activities, discuss student progress, and confirm or refine researchers' observations and interpretations.
Field notes and interview transcripts were coded by all three authors using an open coding procedure (Miles & Huberman, 1994); we used a qualitative data analysis software (Atlas.ti) to assist with this analysis. Codes were developed based on our review of the literature, National Commission on Writing (2006) recommendations, as well as themes that emerged from the data. For example, our review of the literature pointed to the importance of engaging in explicit discussions about language choices with students, and so we looked for evidence of these kinds of exchanges. Assessment of writing, standardized testing, and test prep are examples of codes that emerged from the data. During our interviews and observations, it was clear that assessment was a key issue for our participants-standardized testing was cited as both interrupting regular teaching and also guiding instructional choices. Therefore we developed codes related to standardized assessments. The first author coded a subset of the data to create an initial coding list, which was disseminated to the remainder of the team. Each researcher used this coding scheme to do a first pass through the data, and the code list was modified and refined on subsequent readings of the data. Finally, the first author revisited the entire data set using our refined coding schema to be sure that all data were coded consistently. Through the coding process we were able to document patterns across classrooms, and compare interview and observation data. While coding assists with the reduction and organization of data, this process may also lead to a certain amount of decontextualization of the data (Gagnon, 2010). Therefore, narrative summaries were constructed to synthesize the themes that emerged, and to summarize what we could conclude about each of our research questions (Gagnon, 2010). Gagnon (2010) explains that the purpose of constructing narrative summaries of cases is "to organize the evidence that supports the patterns identified in the analysis… and most importantly to elaborate on these patterns by returning them to their specific context" (p. 80). The researchers read each narrative summary to check for accuracy, look for non-conforming cases, and suggest modifications. Data analysis revealed several important themes including the prevalence of writing in the content areas, explicit instruction of writing genres, and clear connections between the teachers' beliefs about teaching writing with ELs and the instructional strategies observed.

Findings: Teaching Writing to ELs
We have organized our findings to correspond with our research questions and to highlight the themes uncovered in our data analysis. We begin with a discussion of the writing tasks teachers and students engaged in, the genres taught, and the instructional strategies used. In particular we explore the kinds of scaffolding teachers used to support students' writing development and understanding of genres taught.

Types of Writing Activities and Genres Taught
Writing to learn. When teachers engage their students in writing to learn, they are teaching students how to use writing to deepen their understanding of a topic (Bangert-Drowns et al., 2004). In both classrooms, most writing activities observed related to writing to learnwriting definitions, summarizing, and composing responses to literature. With regard to genres of writing, significant instructional time was dedicated to teaching response to literature and summary. Both teachers provided a substantial amount of modeling and scaffolding to help students recognize and produce these genres. During one observation, students were working with the response to literature genre after reading a short story, "First Choice." Students were given the following prompt: "What was the author's message in the story First Choice?" Students were then given explicit instruction in crafting a thesis statement, using textual evidence to support their main ideas, writing a conclusion, and revising a first draft. Based on our observations, both teachers clearly and repeatedly modeled the stages associated with the response to literature genre and guided their students through this genre multiple times. Students appeared to be familiar with the purpose and structure of this genre. Similar lessons were observed where students were explicitly taught to write summaries of texts.
Report writing. In addition to composing summaries and responses to fictional pieces, students in both fourth-grade classrooms composed research reports on the California missions.
Reports are factual texts used to classify and describe information and inform the reader about a topic in a succinct manner Schleppegrell, 2004). Students must become experts on the topic and then describe the information to a less informed reader in a clear, but interesting manner. In this case, the students' reports also exhibited features of accounts (Schleppegrell, 2004, p. 85), because they were required to narrate the life and times of specific missions, and interpret and analyze the information presented. Throughout the report writing process, we observed that both teachers employed clear and consistent scaffolding. The teachers guided the students through the research and writing about the Mission of San Juan Capistrano prior to students conducting research more independently and composing their own reports.
One way that Ms. Nara assisted students with such content area writing was by providing exposure to mentor texts. Mentor texts are high quality, authentic texts that are published for a wide audience that are used by teachers to model the kinds of writing valued in different disciplines (Gallagher, 2014;Pytash & Morgan, 2014). By studying mentor texts, teachers and students can notice and discuss the features of different genres, and the linguistic moves used by the author to realize his or her communicative goals. During our interview, Ms. Nara explained that she read sample texts aloud to familiarize students with relevant academic vocabulary and discourse features: I will find material, like for photosynthesis, I would read for them the way the paragraphs are supposed to sound if you were writing a paragraph about photosynthesis. Not that they were writing it at that moment, but sample writings, or sample pieces, so they can hear how language… science language and how it's used.
At several points during our interview, Ms. Nara discussed the importance of using mentor texts to help students analyze language choices, a key aspect of the teaching/learning cycle.
In addition to providing students with mentor texts to examine and discuss, Ms. Nara also utilized joint construction of a text. In these cases, Ms. Nara began by providing most of the writing herself with the students generating a limited amount of original content. The following is an excerpt from the extended writing frame that she created for her students: A mission is a religious settlement. Name of country built the missions in California.

Name of country purpose in building the missions was purpose in building the mission.
Spanish priests came to work in the missions in order to the priest's purpose. How many missions were built missions were built in California. Indians' way of life changed because how the Indians' way of life changed.
Students were instructed to use the teacher-produced example to guide their writing by filling in the underlined sections with information from their reading and research. The teacher provided similar frames for all six paragraphs in this jointly constructed report on the San Juan Capistrano mission.
Ms. Nara strongly believed that teachers needed to spend more time modeling good writing. For both teachers, modeling writing included examining mentor texts, thinking aloud and sharing one's thought processes during writing, as well as and providing students with examples and scaffolds, such as the writing frames used with the mission reports. In our interview, Ms. Nara explained that it is important to show students what you expect: I do a lot of the writing myself. I would actually write the paragraphs. I tell the teachers when they say, "Well, you know, I don't have any kids who can write a 4 1 ," "Have you modeled the 4? 'Cause if you model it, I bet you, hands down, there'll be three or four kids who can do what you model." But the thing is, there's a certain amount of lazinessthey (the teachers) don't want to pick up the pen, they don't want to have to think like that. So, I write pieces at home. The things I give my kids, I already have it written out as a paragraph, but then I'm collaborating with them. I'm actually trying to draw the sentences out from them, by giving them clues, "How could I say this, or how could I say that?" (Interview, 03/15/12) Ms. Nara is clearly describing her own version of the teaching-learning cycle (Rose & Martin, 2012), as she provided students with mentor texts, models of what she expected, and engaged in joint construction of texts prior to expecting students to construct successful reports on their own. Both Ms. June and Ms. Nara taught students how to write by using teacher-created exemplars, while gradually releasing their use over the course of the unit so that by the end, students could develop their reports with greater independence.

Instructional Strategies to Support Writing
Scaffolding writing with sentence frames. Both Ms. Nara and Ms. June employed a variety of scaffolding techniques to instruct students in the linguistic features of the genres they were learning. These teachers specifically cited sentence frames as a form of scaffolding on which they had received professional development and frequently used. The teachers used this strategy to support student writing and, in some cases, to highlight language functions. During one observation in which students were composing responses to literature, Ms. Nara began by discussing the differences between summarizing and responding with her students. One of her students explained that a response to literature is different than a summary because in a response to literature "they write down their understanding of the story." Then, Ms.
Nara asked students to write about something they could relate to in the story by using a sentence frame: "In the story ______________ by __________, I can relate to ______________ because ______________." The sentence frame helped students remember that they were trying to connect with and interpret the text using specific examples from the story and their own lives to develop their ideas. In Ms. Nara and Ms. June's classrooms, the sentence frames were used to help students think about the goal of the writing task, the function of conjunctions and other language forms, as well as to provide temporary support until students could begin to use these forms independently.
When working in the content areas, the fourth-grade teachers used sentence frames to guide students in writing definitions and summarizing their learning. For example, during a science lesson, Ms. Nara worked with her class to discuss an experiment on electrical charges.
Students were given two sentence frames, which included key vocabulary: When you rub two objects together, ____________________________________.
(Key vocabulary: like/same, unlike/opposite) Ms. Nara used these sentence frames to emphasize and teach a key science concept (the behavior of electrically charged molecules), and the linguistic function of the term while. In modeling how to use these sentence frames, Ms. Nara compared the use of while to the use of but: Ms. Nara: It's like using the word but, remember we discussed this when we were talking about the covered wagons. So, if while refers to opposites, what are the two opposites we have on the board? Students: attract and repel! This lesson provides a useful example of how focusing on language forms and functions can enhance students' understanding of content-area concepts and allow students to express their learning in increasingly sophisticated ways. Ms. Nara's English learners were able to participate in this activity at their varying language levels. The examples below show the writing produced by students classified as Intermediate and Fluent, respectively. The underlined portions indicate student writing (words were crossed-out by student and students' misspelling was maintained); the teacher provided the themes and the conjunctions in bold: When you rub two objects together, I learned that one attracts and another repels.
Oppisites Unlike charges attract, while same like charges repel.

(Intermediate Level Student)
When you rub two objects together, they could attract or repel depending on the numbers of ⌃ electric charges.
Opposite charges come together while like atoms can separate.
When atoms are opposites they come together while like charges separate or come apart.

(Redesignated Fluent English Proficient (RFEP) Student)
The final sentence was produced independently (although key vocabulary related to the unit was available on the board). Ms. Nara frequently encouraged students to compose additional sentences without the support of a sentence frame. Hence the sentence frames provided support so that students at different language levels could successfully engage with the task. It is interesting to note that the intermediate child provided a subjective response to the first sentence frame, inserting what he learned and writing in the first person, while the more advanced student did not. This may reflect developmental differences, since novice writers are more likely to use personal experience to support claims, and may also highlight an opportunity to teach novice writers about different ways to express knowledge claims and create a more authoritative stance . Nonetheless, the sentence frame aided the intermediate child in expressing his learning accurately using scientific language. This provides an illustration of the ways that Ms. Nara worked to teach content area concepts while simultaneously supporting language development.
Verbal scaffolding of student writing. In addition to extensive use of sentence frames, another way these two teachers assisted students was through verbal scaffolding. Ms. Nara and Ms. June encouraged students to talk through their ideas and frequently directed students to write down ideas that they had just finished expressing orally. The following excerpts from our interviews with Ms. June and Ms. Nara highlight this orientation towards teaching writing: I tell the kids, "What you are able to say, you can write, so we have to work on what you say and how you say it." A lot of the kids they don't make (that) connection. They have the good, the interpersonal communication skills and that is very different from their CALPS, which is Cognitive Academic Language, and so, I believe that all kids need help.
(Nara, 03/15/12) Have them talk it through and then have them talk to each other; not talk, but tell each other what they are going to write. Listening to their partner, repeating it, "Okay, what did your partner say? Okay, now, write it." So having them orally produce it a lot.
(June, 03/15/12) These statements reflect the teachers' beliefs regarding the connections between oral language and writing, and highlight the notion that teachers of ELs should provide students with opportunities to practice academic language orally, as a method of scaffolding writing. Verbal scaffolding was used in a variety of ways in both classrooms: to brainstorm ideas and jumpstart the writing process, to reflect on the meanings of specific words and to expand students' lexical diversity in writing, and to encourage reflection on connections between language choices and text purposes.
The following snapshot from Ms. June's classroom shows how she used dialogue to stimulate the writing process. Here Ms. June was guiding her students to write a summary and response to a story called "Going West: St.: disaster journey? T: that wouldn't really be considered a disaster… Okay, now that you have written your heading, write a title, chose one or write your own.
St.: It might be a long title, but can we do? My Diary Going West?
T: Okay, love it! Here Ms. June provided students with general evaluations regarding the accuracy of their word choices, but this conversation might have been more useful had the students and teacher discussed why particular words were more or less suitable.
Both teachers often provided general feedback during these kinds of whole class discussions (e.g., "that would be a great title"), and frequently focused on grammar (e.g., "if we say pioneers, do we need to say 'they' right after?"). However, we also observed that the teachers encouraged students to consider their readers' engagement, and gave students examples of ways to use dialogue or figurative language to create a more interesting or effective text. In order to help her students brainstorm more specific language for their "Going West" diary entries, Ms. June used sensory prompts, asking students to think about what they would hear, see, taste, touch or smell on their journey. After several students shared ideas for each category, Ms. June prompted them to expand on their answers. When one student stated that he would see dirt, she asked, "Where would you see that? What would that feel like as you kicked dirt up as you walked?" Through these brainstorming activities, Ms. June used verbal scaffolding to encourage students to use the academic vocabulary they were learning during the unit, as well as draw on relevant prior knowledge.
In Ms. Nara's classroom, verbal scaffolding came in the form of direct feedback on writing in the moment, as well as conversations about the links between grammatical choices and the purposes or functions of different genres. In the following excerpt, Ms. Nara's students were developing sentences to show what they had learned about the pioneers. The teacher began by providing students with sentence frames and then students shared what they wrote: St. 1 (reading aloud): The pioneers traveled by boat or covered wagon.
St. 2: Is mine a compound sentence?
T: Read the sentence.
St. 2: Most pioneers went on horse drawn wagons, but some got sick.
T: Yes, that is a compound sentence. Remember what happens if we use the word "but"?
You need to have something that relates to or explains the beginning part of your sentence. (Teacher helps student change her sentence to, "Most pioneers went on horse drawn wagons, but some had to walk.") This passage provides an example of the kinds of conversations that Ms. Nara had with her students about using specific language choices to express ideas more accurately. During our interview, Ms. Nara explained that she discussed the functional purposes of different conjunctions and connectives with her students: We're working on modeling complex and compound sentences for them so they can learn to structure their sentences differently and it makes a greater impact on their writing.
When we were modeling for the state standardized testing and they were having to respond to literature… I have a boy who's (in the gifted program) and his writing, everything is "Then, then, then, then, then, then." Transitions! Many of them are stuck in first grade with first, next, and last. That's all they have in terms of writing transitions and I told them, "That doesn't always lend itself to your writing though. What if you're writing about something that's not ordinal? Then you don't have first, next, and last." Such conversations and modeling of the connections between form and function may help Ms. Nara's students become more conscious of their language choices and how those choices affect the readers' reception and comprehension of their writing. These kinds of discussions might be enhanced if Ms. Nara pointed out similarities and differences between conjunctions used to connect ideas within a sentence and connective devices used to connect ideas in a paragraph or longer piece of written discourse.

Discussion
This study documents writing instruction in two fourth-grade classrooms in California to gain insight into the types of writing activities and genres of writing that English learners (ELs) are requested to produce, as well as the kinds of instructional approaches used by mainstream classroom teachers.

Text Types and the Common Core State Standards
With regards to the kinds of writing students were assigned, we observed that student writing tasks focused primarily on writing to learn-tasks such as composing reading responses or summaries-and report writing. Our findings are consistent with Gilbert and Graham's (2010) survey study in that teachers spent little time on narrative writing. This trend may reflect the Common Core Standard's increased emphasis on informational text, analyzing text, and writing for different audiences and purposes. With the implementation of the Common Core English Language Arts Standards (2015), students are expected to produce more textual analysis (of both literary and informational text) and less personal narrative (Shanahan, 2012).
High-stakes writing assessments clearly had profound influences on the kinds of writing that children engaged in and the genres they were learning to produce. In California, the fourthgrade writing assessments required students to produce a piece of writing on demand, and Ms.
June and Ms. Nara were observed providing explicit instruction in the genres that students were likely to encounter on the writing test. The performance-based tasks which are a requirement of the Common Core assessments require creativity and critical thinking, which means that students will need to generate written answers on their own. Thus students need increased opportunities to practice independent writing production.

Genre Pedagogy
The influence of genre pedagogy was evident in that students were instructed in the stages of the different text types they were learning, especially response to literature. Both teachers understood the importance of linking language forms and choices to the purposes of texts. We clearly noted instances in which the teachers had explicit conversations with students related to "the functionality and appropriateness of different language choices" (Achugar et al., 2007, p. 11), which is an important aspect of developing students' academic literacy. For example, Ms. Nara stated that a student's choice of connectives should vary depending on the purpose of the text. Introducing teachers to the precepts of systemic functional linguistics (SFL) could provide them with tools and a common language that they could employ while teaching writing through genres. Engaging teachers in a more thorough-going analysis of the stages of each genre and how these stages are achieved might help teachers provide more specific and substantive feedback to their students (Aguirre-Muñoz et al., 2008). Rather than teaching students to memorize a list of key words associated with the text type of compare and contrast, teachers could engage students in a thorough-going analysis of texts in which comparison is used effectively. Similarly, making teachers and students more aware of the stages and rhetorical moves used in writing reports or accounts might also improve students' report writing, as was demonstrated in the California History Project (Schleppegrell et al., 2008).
The strategies that Ms. Nara and Ms. June implemented to draw students' attention to the functions of language included modeling the genre being taught, discussion of the features of specific genres, teacher feedback specific to language use in the genre, and the use of sentence frames to highlight language functions. All of these instructional practices have the potential to improve student writing with regards to clarity, cohesion, lexical variety, and lexical density (Aguirre-Muñoz et al., 2008;Crosson, Matsumura, Correnti, & Arlotta-Guerrero, 2012). Further professional development focused on using SFL to teach writing-deconstructing and reconstructing academic texts; discussing the language choices made by authors, and teaching children to talk about the texts they are reading and composing-could improve writing instruction for all children, but especially for English learners (Aguirre-Muñoz et al., 2008).

Teacher Preparation to Teach Writing with English Learners
This was a small qualitative study undertaken to illuminate the practices of two teachers from an urban district, who had been singled out as highly proficient at teaching writing to English learners. Linguistically diverse students need plenty of opportunities to learn and practice the academic language needed to perform successfully in school. In order for teachers to be able to support students' academic language development, they must "have extensive skills in teaching the mechanics of language and how it is used in different contexts and for different purposes" (Gándara et al., 2005, p. 3). The teachers in this study were well prepared to serve their culturally and linguistically diverse students-both had bilingual teaching credentials and many years of experience. Our findings revealed positive trends with regard to the explicit instruction of writing for ELs. However, some of the same concerns that were highlighted in the National Commission on Writing reports (2003,2006) remain, namely that writing instruction continues to be given short shrift with regards to the time allocation, and that teachers would benefit from increased support and professional development related to teaching writing.
Although both teachers in this study had specialized training in the teaching of ELs, they had received little professional development that was focused on writing. Both teachers reported feeling confident about their ability to teach writing to ELs, but they also noted the efforts they personally expended to create effective curricula. Both teachers stated that they often created their own materials, which they then shared with one another. While this may work fairly well for competent, trained, and experienced teachers, it certainly is not appropriate to expect novice teachers to be successful without the appropriate support.
Our interviews and observations revealed that the teachers utilized the materials and concepts introduced to them in their professional development workshops, which focused on using graphic organizers and sentence frames to support student acquisition of academic language. However, the kind of scaffolding we observed also resulted in student work that lacked variety in content, voice, or style. While students may have benefitted from the high level of scaffolding provided and were able to produce coherent reports, we wondered about the extent to which students were actually appropriating genres in ways that would enable them to draw on these language forms independently in new contexts. Continued observations of these EL students' writing development over time would be useful for determining how students used what they learned about report writing in subsequent assignments.

Scaffolding Writing with English Learners
This point leads us to reflect on the effective use of different kinds of scaffolds for English learners. The verbal scaffolding used by both teachers appeared to be useful and productive. During these conversations, students were observed actively considering the meanings and functions of different grammatical forms. However, no differentiation of the scaffolds, writing instruction, or the products expected from students was observed. Although Ms. Nara gradually released the scaffolds, we never observed the students writing without support. For example, in reading the mission reports that the students wrote independently, we noted minimal variation with regard to structure or tenor. This may indicate that students needed more time to practice the report genre. We speculate that because teachers were required to model, instruct, and practice multiple writing genres, they may have lacked sufficient time to allow for independent practice. Thus scaffolds may have been provided more on the basis of the teachers' goals and schedule rather than on student needs. Rose and Martin (2012) assert that when teaching students to recognize and use new genres of writing, robust teacher modeling is necessary. If students were subsequently provided with opportunities to write in different genres more independently, then perhaps the level of scaffolding we observed in the classrooms was appropriate. If students do benefit from extensive modeling and scaffolding in writing genres that are new to them, it would be useful to look at writing instruction longitudinally to see when and how teachers provide students with opportunities to practice genres independently. In future research studies, we would like to observe the writing development of specific EL students over the course of an academic year or longer. This would allow us to determine the influence of instruction on student writing, and would permit the assessment of the extent to which students appropriated and independently employed the language forms and academic vocabulary that they were exposed to during scaffolded writing assignments. *English learners are reclassified as Fluent English Proficient when they obtain an advanced score on the state's English language proficiency exam.