Negotiating Co-Teaching Identities in Multilingual High School Classrooms

Abstract School districts in the U.S. are increasingly calling on content area and English as a Second Language (ESL) teachers to work together to plan and deliver instruction in classrooms with linguistically diverse students. Such programming presumes, however, that collaborative teaching dynamics are unproblematic. The aim of this article is to examine ESL and content area co-teaching dyads at an urban high school in the U.S. southeast. Data were drawn from a year-long qualitative study of these classrooms and were analyzed using sociocultural perspectives on learning (Lave & Wenger, 1991; Lave, 1998). Findings highlight structural factors that inhibit the development of positive co-teaching relationships, including top-down decision-making, inadequate training for co-teaching, and lack of time for co-planning. Positive relationships were formed on the basis of shared personal and pedagogical visions, flexibility and adaptability. In addition to recommendations for school-level changes, implications of this study center on the need to prepare teacher candidates for collaborative teaching through cross-disciplinary coursework that includes opportunities to practice and reflect on co-teaching.


Introduction
English learners (ELs) constitute the most rapidly growing segment of the student population in American schools, and developing the means to improve their academic outcomes is one of the most pressing issues in current educational policy and practice. In response, many school districts in the U.S. have adopted collaborative teaching policies as a means of ensuring that ELs receive the linguistic support they often need in academic content area classes, and as a means of reducing budgetary and physical constraints in over-crowded schools (Gottlieb, 2006;Honigsfeld & Dove, 2010;McClure & Cahnmann-Taylor, 2010). Some of the benefits of collaborative teaching include a smaller student-teacher ratio, greater opportunities for smallgroup instruction, and access to comprehensible linguistically and academically sophisticated content (Davison, 2006;York-Barr, Ghere, & Sommerness, 2007). However, school administrators often assume that the development of a co-teaching relationship is uncomplicated and inherently advantageous (Arkoudis, 2006). In reality, co-teaching is a complex and dynamic process involving multiple agendas and personalities, and teachers are often unprepared or unwilling participants in co-teaching situations (Davison, 2006;Murawski & Lochner, 2011).
A growing body of research has begun to examine interpersonal dynamics within coteaching relationships in multilingual classrooms. Notably missing from the research, however, is an examination of co-teaching relationships in classrooms at the secondary level, and classroomlevel data on collaborative teaching in particular (Hornberger, 2006). Through classroom observations and interviews with five co-teachers (three content-area and two ESL teachers) at an urban high school in the U.S. southeast, this article captures the perceptions of co-teachers as they negotiated administrative and curricular demands, pedagogical goals, physical space, and diverse teaching styles and philosophies. The study was guided by the following research questions: How do high school ESL and content area teachers in co-taught classrooms engage with each other and their culturally and linguistically diverse students? How do teachers view themselves as coteachers within a multilingual classroom setting? And more specifically, how do teachers negotiate pedagogical goals and approaches in co-taught classes? This research evidence suggests that while collaborative teaching across a variety of content areas can be beneficial, structural factors can inhibit the development of positive co-teaching identities and outcomes.

Background
Co-teaching is a practice that is well established in the field of special education (Murawski & Lochner, 2011;Villa, Thousand, & Nevin, 2013), and more recently in the field of bilingual and English language education. Research has begun to examine ESL/content-area teaching practices and outcomes (Honigsfeld & Dove, 2012) from a variety of perspectives, including social justice and inclusive models of instruction, teacher leadership (Theoharis, 2009), and collaboratively designed standards-based curriculum and instruction (Short, Cloud, Morris, & Motta, 2012). Nonetheless, research on collaborative teaching between ESL and content area teachers has largely focused on co-teaching in primary school contexts (McClure & Cahnmann-Taylor, 2010), and in school contexts outside of the U.S. (Creese, 2006;Gardener, 2006). Despite the growing popularity of this model, the benefits of co-teaching arrangements for exceptional children and ELs are not at all clear in the literature to date.
Relationships can differ within each co-teaching model, each impacting the way in which instruction is carried out. Particularly useful to this article's analysis is Davison's (2006) framework of co-teaching relationships. According to Davison, collaborative relationships between an ESL teacher and content area teacher can take several forms, ranging from pseudocompliance and passive resistance to a more productive and creative co-construction. Pseudocompliance or passive resistance is the first level of collaboration. This exists when there is an implicit or explicit rejection of collaboration and preference for status quo with little or no real investment of time or understanding of why co-teaching needs to happen. Compliance occurs when there is a general expression of collegiality with well-meaning teachers who make minimal efforts to collaborate in planning and teaching. Convergence occurs when teachers embrace opportunities for professional growth and much more effort is made to engage in dialogue. The most productive model of collaboration exists when teachers can reach creative co-construction. Davison (2006) explains, Teachers' roles become much more interchangeable, yet more distinct, [a] high degree of trust of other is evident, responsibilities and areas of expertise continually [are] negotiated, informing documents [are] seen as actively co-constructed and teacher-developed, conflicts in roles [are] seen as inevitable, accepted, even embraced, as a continuing condition which will lead to greater understanding (p. 468).
Among the documented challenges to co-teaching in multilingual settings are differences in ideology with regard to language use in the classroom, pedagogical approaches, and contentarea background. For instance, although ESL teaching necessitates knowledge of linguistics and second language acquisition theory, the field is often viewed as "strategy-driven" and nonacademic, and bilingual and ESL teachers are often marginalized within school contexts (Harper, de Jong & Platt, 2008). With this background in place, this article interrogates how co-teaching relationships and identities unfold in different ways in three different high school classrooms.

Theoretical Frame
The review of literature suggests several competing themes regarding the impact of collaborative teaching on teachers themselves, particularly when co-teacher relationships are taken into account. As our research examines co-teacher identities in classroom practices, in theorizing our inquiry, we drew on sociocultural learning theory, and particularly on the notion of communities of practice (Lave & Wenger, 1991), which are "a set of relations among persons, activity, and the world, over time and in relation with other tangential and overlapping communities of practice" (p. 98). As a construct, communities of practice rests on the theory that learning occurs through social interaction, and that learning shapes and is shaped by our identities in different contexts. Power dynamics, whether characterized by conflict and subordination, or by consensus and egalitarianism, are embedded within communities of practice and determine the potential for learning. Wenger (1998) argues that the experience of learning and one's view of oneself as a learner within a community of practice is a function of one's imagination that connects particular activities to future possibilities: "Participation here refers not just to local events of engagement in certain activities with certain people, but to a more encompassing process of being active participants in the practices of social communities and constructing identities in relation to these communities" (p. 4). Newcomers who join communities initially learn tasks that may be considered less important than those performed by key members of the community. Hierarchy is evident in any setting where individuals inhabit different roles, and, in relation to this study, academic content area teachers may view the work of ESL teachers as auxiliary, rather than viewing ESL teachers as experts of their teaching domain (Arkoudis, 2006).
Legitimate peripheral participation is that which is considered tangential, yet essential to the workings of a community project. In order for legitimate peripheral participation to occur, new members of the community must be granted on-going and continuous access to "old-timers, and other members of the community; and to information, resources, and opportunities for participation" (Wenger, 1998, p. 101). Control and selection, while present in all communities of practice, give rise to potential for manipulation of access, which can prevent legitimate participation.

Research Context
As a region with a historically low immigration rate, the U.S. southeast has seen a rapid increase in its immigrant population over the past twenty years, and the changes precipitated by this rapid demographic transformation are particularly visible in the region's public schools. The population of ELs enrolled in public schools in the state where this research took place nearly doubled between 2002 and 2007 (from 60,149 to 112,532 students). This demographic shift took place during an era of new education reforms, such as the No Child Left Behind Act of 2001 and Race to the Top. These reforms have prioritized high-stakes testing and teacher evaluations as means of reducing the achievement gap between white and historically marginalized students, including immigrant language learners. The focal school district began implementing collaborative teaching programs district-wide in 2007 as part of a "Strategic Staffing and School Turnaround Initiative," which was intended to improve academic outcomes in its low-performing schools.
This study took place at Shady Oaks High School 1 over the course of one academic year. Located in a high-wealth neighborhood, the school had historically enrolled middle, upper middle, and upper class students living in the neighborhoods surrounding the school. District zoning policies had recently diversified the school's student demographics, leading to a 30% increase in its enrollment of low-wealth immigrant and racial minority students. The majority of the school's ELs were immigrants from Mexico and Central America, and refugees from Southeast Asia and Africa. There were four ESL teachers at the school. Students designated by the district as Limited English Proficient (LEP) were enrolled in at least one 90-minute ESL class per day in which they received instruction focusing on English language and literacy development. Besides ESL classes, ELs were enrolled in content-area courses required for graduation, as well as various elective courses, such as music and art. Roughly half of these content area classes were co-taught by content-area specialists and an ESL teacher.

Participants
Data were comprised of interviews and observations of five teachers (two ESL teachers and three content-area specialists) at Shady Oaks High School, all of which were conducted by the lead investigator of the research team. Though more co-taught classes were observed, we purposefully selected the case studies herein for a variety of reasons. First, as the focus of our research is on co-teacher identities and relationships, observational data illustrates how individuals negotiate space, and embody subject positions differently. Second, taken together, the teacher interviews illustrate the range of responses to the co-teaching environment and practice. Finally, focusing on three cases allowed us to go into greater depth than would otherwise be possible with a larger sample size.
The teachers had between seven and twenty-five years of teaching experience, and each had little to no training or experience co-teaching. The co-teaching dyads were assigned by school administrators, and none of teachers within each dyads had taught together prior to the year in which this research took place. The principal researcher first observed a co-taught 10 th grade biology class taught by Mrs. Thomas, a biology teacher in her 40s, who had been teaching at Shady Oaks High School for eight years, and Ms. Elway, an ESL teacher in her 60s who had taught ESL for twenty-five years. Until this year, Ms. Elway had taught earth science and biology classes to 9 th and 10 th grade native speakers of English, many of whom she considered high achieving based on their performance in the class. One week prior to the beginning of classes, she was asked by a school administrator to co-teach a biology class with Ms. Elway. Mrs. Thomas had no training in working with ELs and had not attended any of the optional district-level workshops on teaching ELs (e.g., using Sheltered Observational Instruction Protocol approaches) or on co-teaching techniques. Ms. Elway had not taken coursework in biology since her undergraduate training, which she completed in the late 1970s.
The second pair (Mr. Johnson and Ms. Elway) taught algebra to 10 th graders. Mr. Johnson, who was in his mid 30s, began his career as an elementary school teacher, specializing in math. He had been teaching at Shady Oaks High School for almost five years at the time of this research. Like Mrs. Thomas, Mr. Johnson also had no formal training in teaching ELs. On the other hand, Ms. Elway had not taken mathematics coursework since earning her undergraduate degree. Although both teachers had co-taught with other teachers at the school, neither of them had attended co-teaching training sessions prior to or during their partnerships. The third pair (Mr. Wilson and Mrs. Stevens) taught 11 th grade civics. Mr. Wilson, who was in his late 40s, had been a social studies teacher at the school for 19 years. He had taught one social studies class with another ESL teacher. He had attended workshops on teaching ELs in contentarea classes, but openly expressed that these professional development opportunities presented him with no new insights on how to work with this population. Mrs. Stevens, who was in her 50s, had taught ESL for 10 years, and this was her first co-teaching experience. However, she held an undergraduate degree in history and felt, in her estimation, somewhat confident with the content of the civics class.

Methods
This study is comprised of three individual case studies of the co-teacher dyads described above. Merriam (1998) defines case study as "intensive descriptions and analyses of a single unit or bounded system, such as an individual, program, event, group, intervention, or community" (p. 19). The six cases reported herein are considered instrumental case studies that attempt to generate a general understanding of co-teacher relationships (Stake, 1995).
Research captured "naturally occurring, ordinary events in natural settings" (Miles & Huberman, 1994, p. 10), the groundedness of data in local contexts over sustained time periods, richness of data, and an emphasis on people's lived experiences in connection to their social worlds. Four 90-minute observations were conducted by the lead researcher in each class for a total of 18 hours of observational data recorded in field notes. Observations centered on social interaction within teaching practices, including verbal discourse between the teachers and with students, as well as their negotiation of space and materials. Three thirty-minute semi-structured interviews following the same interview protocol were conducted with each of the teachers by the lead author to gain a first-person sense of their self-identities, their attitudes toward co-teaching, as well as their perceptions of learning outcomes, classroom phenomena, social interactions, and pedagogical approaches. A total of nine hours of interview data were obtained, which we later transcribed. Transcripts were completed by the lead author and co-authors, and then were crosschecked by all.
Data analysis occurred in several stages: 1) organizing data, 2) generating categories, 3) coding the data, 4) testing the emergent understandings, 5) searching for alternative explanations, 6) writing the report. Following methods outlined by Patton (2002) and Heath and Street (2008), upon transcription of interviews, working together, we read through the data and generated codes from interview responses, background literature, and the conceptual framework. Coding was generally organized around the interview questions posed, as well as the research questions that guided this study. Using a constant comparative, or recursive perspective, we juxtaposed data from observations and interviews with our underlying assumptions or hunches, as well as theories and concepts from the literature to create a dialogue between existing explanations and ongoing data collection and analysis (Heath & Street, 2008). We then separated the data analysis into two stages: within-case and cross-case stages (Stake, 1995). The within-case analysis focused on three descriptive, comprehensive cases, looking at contextual variables that factored into each. In the cross-case analysis, we looked for themes generated across all cases, documenting individual differences that arose between them. Each case was comprised of eight categories and subcategories. The cross-case analysis focused on themes that emerged in all cases, as well as themes that were not common across the cases. Once analysis was complete, the lead author of this study conducted member checks in order to validate, or in some cases, dispel initial interpretations of data, and to verify the accuracy of findings (Creswell, 2007).
We stress that our findings represent our analyses of three co-taught classrooms at one high school in a district that was working to respond to the academic and linguistic needs of its growing EL population. These cases are not intended to represent all co-taught classes at the high school level. Findings can, however, challenge educators and scholars to consider conditions under which well-intentioned policies are enacted.

Findings
The three cases presented reveal significant overarching themes as well as unique, casespecific characteristics, each of which is detailed below.

Biology
Biology is a required for graduation in the state, though biology was not a tested subject within the state's academic accountability program. There were 33 students in Mrs. Thomas's (biology teacher) and Ms. Elway's (ESL teacher) co-taught 10 th grade biology class. Twenty-four of the students were classified as ELs who had been in the U.S. for varying lengths of time. Eighteen of these students were Hispanic, four were Vietnamese, and two were Congolese. The nine non-EL students were retaking the course after having failed it the previous semester. The course was taught in a biology lab, which was comprised of several fixed tables, and students sat on stools facing the front of the room. Mrs. Thomas created a seating plan that placed ELs together by language group, and native speakers of English on the opposite side of the room. Ms. Elway was not consulted when the seating plan was devised. During one observation of the biology class period, there existed physical and emotional distance between Ms. Elway and Mrs. Thomas. Field notes from one observation depict a recurring scene in this classroom: Mrs. Thomas began the class by having students work on questions on a worksheet related to DNA. Students pulled out a packet of worksheets Mrs. Thomas had given them the previous week. The worksheets contained content students were expected to know for a state-mandated end of course exam. [Ms. Elway was given the same packet to prepare for the course.] Ms. Elway walked around the room monitoring and offering assistance with vocabulary to ELs who asked for it. The classroom was noisy, with students getting out of their seats. Ms. Elway attempted to keep non-ELs focused, but the students ignored her. After a few minutes, Mrs. Thomas orally went over responses to the questions. There were few visuals around the room to assist students with comprehension, and content was not contextualized. A male EL in the right corner behind me seemed to know a lot of the answers to her questions to the group, which he uttered under his breath. Mrs. Thomas called on the same students to respond, none of whom were ELs. Students were then given five minutes to review their workbooks before they were to take a pop quiz on DNA. There was lots of chatter and the class wasn't focused during those five minutes. One student was at the pencil sharpener the whole time. Another student was sketching an anime character in their notebook. Another student put his head down on the table. Mrs. Johnson never established direct eye contact with Ms. Elway, and did not speak with her, and Ms. Elway, in turn, rolled her eyes at Mrs. Johnson and tried to keep herself busy by helping students stay on task.
As the content specialist in biology, Mrs. Thomas assumed all responsibility over lesson plans and the development of curricular materials. She gave all students the same unmodified materials, which Ms. Elway generally received the day before class. Ms. Elway's offers to meet to plan together were consistently turned down, and she felt she had no input either in planning, instruction or classroom management. In an interview she shared, "I have tried to set up meetings to plan, but [Mrs. Thomas] never can. Instead, I get the materials at the same time students do, though I feel like I should be one step ahead. But I'm basically just like them, except that I can understand the language and the general idea of what they're covering. It's demeaning to me that I come to class knowing about as much as the students do." Evident in this relationship was unwillingness on Mrs. Thomas's part to give Ms. Elway open access to course content.
Ms. Elway felt an investment in seeing students succeed, whereas she viewed Mrs. Thomas as not being overly concerned with their needs, as shown in her decision not to scaffold their learning through such means as modified materials, or visuals, and by not calling on English learners to respond to group questions. Mrs. Thomas also generally ignored classroom management issues. In contrast, Ms. Elway often circulated around the classroom to, in her own words, "put out fires." Ms. Elway admitted to deferring to Mrs. Thomas in spite of her desire to assert more control in disruptive situations: "I'd rather not get into an argument with Mrs. Thomas in front of the class over student behavior, so I do my best to work on the edges to keep kids on task." Observations across several class periods revealed that the co-teaching relationship was defined by avoidance, submission, and a lack of trust. Mrs. Thomas shared that she was reluctant to co-teach with a non-science instructor because of the extra planning it involved, and because she preferred to teach on her own. In a separate interview, Mrs. Thomas admitted that she felt unprepared to meet the needs of ELs, and that she "looks forward to next semester, when I will teach AP biology, and other higher level classes, and when I'll have my classroom to myself…unless [administration] decides something else. I admit I'm not as good at working with ESL students." In contrast, Ms. Elway had generally positive experiences in other co-teaching situations and carried certain expectations regarding co-teaching relationships into this biology classroom. Instead, Ms. Elway related that she felt, in her words, like "a servant or a secretary" in Mrs. Thomas's room. These findings suggest that in the absence of connection, and willingness to collaborate, co-teaching can be greatly limited in its student and teacher engagement potential.

Algebra 1
Algebra 1 was a required course for graduation, and a course in which students were subject to state-mandated tests. Mr. Johnson (math teacher) and Ms. Elway's (ESL teacher) 10 th grade algebra class was comprised of thirty students, twenty of whom were English learners. The English learners were from Vietnam, Somalia, Pakistan, Mexico, Guatemala and Honduras. The remaining ten students were repeating the course having not passed the previous year. Field notes from one class observation suggest a radically different relationship from that of Ms. Elway and Mrs. Thomas: Both Mr. Johnson and Ms. Elway stood at the doorway greeting students as they entered the classroom. Students returned their greetings and took their seats, which were arranged in pods of three to four desks with ELs and non-ELs sitting together in "a team," as Mr. Johnson referred to them. Students worked individually on a warm-up exercise while Mr. Johnson and Ms. Elway checked in with one another and walked around the room checking students' work. Mr. Johnson then went over the warm up on a Promethian board by eliciting and demonstrating algorithms and how to arrive at the correct response. Imagery, large, simplified text, uncomplicated slides, and manipulatives were used. Students were mostly attentive and took notes. Ms. Elway circulated the classroom throughout the period, helping students decipher the meaning of symbols and word problems, and ELs and non-ELs responded to her help and classroom management styles as much as they did toward Mr. Johnson. At one point, Mr. Johnson asked Ms. Elway to give an example of a problem he asked students to solve….
The two teachers floated in and out of shared spaces and drew on and legitimized each other's areas of expertise. The material used was the same level as other algebra classes at the school, although it was taught at a slower, incremental pace. Mr. Johnson shared his vision in an interview: All students, ELs and non-ELs alike, need patterns. If they have access to tools to succeed, they will do the work. This means that I need to explain and repeat things slowly and over and over again until they get it. And when they do it's like a light goes off, and I give that student a leadership role in the class so that they can reach other students who don't get it.
Mr. Johnson was aided by having had many of these students in class the semester before; he believed he had sufficient understanding of their academic needs. ELs were often reluctant to speak in his class, he shared, and he believed that support from an ESL teacher was needed in his classes to develop English learners' language skills, better gauge their learning, and help maintain a positive classroom environment, which both teachers felt was sometimes challenging given the diversity of student needs. Mr. Johnson and Ms. Elway met regularly to discuss plans, adapt materials and coordinate classroom roles. Their relationship was amicable and open, and they gave regular feedback to one another on teaching. Ms. Elway shared, "Many teachers see working with ELs as a service; Mr. J. sees it as a privilege. And he welcomes having me in the classroom, which makes my job a lot easier and more meaningful than when I feel like I'm a bother to the other teacher." Ms. Elway believed that co-teaching with Mr. Johnson elevated her status as an ESL teacher within the school, in part because of Mr. Johnson's popularity among ELs and non-ELs alike.
Overall, because of their shared teaching philosophies and approaches, Mr. Johnson and Ms. Elway maintained a collegial relationship, which influenced students' attitudes towards each other and their teachers. They displayed a professional disposition toward their work, which was evident in their planning and balanced instruction. They also both seemed to value each other's areas of expertise. Mr. Johnson recognized the importance of Ms. Elway's focus on content-area language, and Ms. Elway, in turn, respected Mr. Johnson for his content knowledge, his teaching, and his efforts to connect with students. This case illustrates that successes in co-teaching have much to do with the degree to which both teachers feel valued and needed for their knowledge, effort and approach to teaching collaboratively, and where teachers position themselves and each other positively.

Civics
Like biology and algebra 1, civics was a required course for graduation. Mr. Wilson (civics teacher) and Mrs. Steven's (ESL teacher) 11 th grade civics class was comprised of 23 ELs and eight native speakers of English. The English learners were from Sierra Leone, Somalia, China, Vietnam, Russia, Mexico, El Salvador and Guatemala. The classroom was arranged with desks in five rows, all facing forward. The room had an LCD projector, and there were posters and state and world maps hung around the room. At the start of class, Mr. Wilson routinely walked around the room greeting students by their names and joking with them in a serious, but affable manner. The atmosphere was welcoming and inclusive. He shared, "Teachers here generally view ESL students as second class -that they bring the school's reputation down. I really work against this in the way I set up my classroom. I want all kids to feel welcome." Mrs. Stevens checked in with the ESL students but typically stayed at the back of the room while Mr. Wilson greeted them.
In one class, Mr. Wilson reminded students of an upcoming exam, as well as how they should prepare for it. He told students their exam would differ from that of prior civics tests in that it would focus on vocabulary words as opposed to essays that emphasized students' understanding of political processes. In a post-observation meeting, Mr. Wilson related that he slowed down instruction and offered more scaffolding to this class, though he expected students to be exposed to the same subject matter as the non-ESL Civics classes. Field notes illustrate this practice: This lesson focused on the ratification of the U.S. Constitution. Mr. Wilson passed out a worksheet containing guided notes because a foldable activity meant to increase students' understanding of target vocabulary that students had previously done was not as successful as he had hoped. Mr. Wilson gave directions on how to take notes on the guided notes sheet, and provided a model of a partially completed worksheet. He asked students to give synonyms used several metaphors as he explained key terms and events. Mrs. Stevens stood at the back of the class during the entire lesson.
In a post-observation interview, Mr. Wilson shared: The reality is that we are teaching in English and that's really hard for them. Language is not a vehicle for them for teaching the content, like it is for American kids. I find myself using a lot of images. Like if I have to explain what a raw material is, I show them a picture of a cotton field, and then a machine, and a shirt, just to show them the progression between a raw material and a manufactured good. This is something you take for granted with an English speaker. You can't go with a slug approach; you have to go with a buckshot approach and reach them as many ways as possible, whether it's a one day approach or a three day approach. Something's going to catch. The mastery thing is something I've worked on. This is why it's important to work with an ESL teacher. If we combine our ideas, something's going to work In contrast to this statement, Mr. Wilson admitted that there was a power struggle between himself and Mrs. Stevens, and that the previous ESL teacher he worked with was more of a "team player," who shared resources and teaching ideas. He felt as though Mrs. Stevens was too passive and that she should be more proactive in helping students. On the other hand, he appreciated her help getting materials to students who had been absent from classes or who needed tutoring outside of his class. From her perspective, Mrs. Stevens believed that she was not given room to be proactive in the classroom, and while well-intentioned and a "very good teacher," Mr. Wilson was reluctant to give up control of the classroom. She added, "He just gets going and doesn't stop long enough for me to contribute. It's frustrating, because I'm not sure why I'm even there when I could be more effective with these students." Mrs. Stevens and Mr. Wilson did not plan together, nor did they share materials. These two teachers, nonetheless, displayed a collegial relationship in front of students. Mrs. Stevens concluded, "We need to have teamwork, because we cannot morally and professionally fail a student because they don't speak the language. But I don't feel as though I'm effective when I can't help a student with the language and content." In this case, Mr. Wilson and Mrs. Stevens appeared to work at cross-purposes. Mr. Wilson felt confident in his abilities to teach the class on his own, while Mrs. Stevens believed her expertise was necessary, but underutilized. The tension between Mr. Wilson and Mrs. Stevens largely stemmed from incongruent attitudes toward co-teaching as a practice and mandated policy within the district.

Cross-case Analysis and Discussion
The results of this study corroborate the findings presented by previous scholarship on coteaching in both special education and ESL fields (Davison, 2006;Scruggs, Mastropieri & McDuffie, 2007;Keefe & Moore, 2004) and add to the body of growing evidence of factors that either facilitate or impede positive teacher relationships in co-taught classrooms. We highlight cross-cutting themes below that have implications for how the teachers came to view themselves and their work.

Top-Down Partnerships
That school administrators assigned teachers to co-teaching dyads left the teachers in all but one case feeling resentful, powerless, and ineffective in the classroom. The teachers were placed in a situation of having to negotiate classroom space, roles, pedagogical practices and materials with co-teachers with whom they had little or no previous contact. In two classrooms, the teachers displayed relationships based on territoriality and incompatible pedagogical goals and methods. In the same two cases, teachers did not collaborate with one another on course planning, lesson plans were not shared, and the ESL teachers were given lower status, or, in the words of one teacher, "secretarial" roles within the classroom, as marginalized non-participants (Wenger, 1998). On the other hand, one of the cases presented a model of how co-teaching might look at Davison's (2006) accommodation or convergence stages. One of these classes centered on preparing for end-of-year testing, as opposed to broader learning engagement with the subject matter. The ESL students in the classroom, in turn, were less interested in the material, tended to give up when they failed to comprehend content, and were more apt to engage in off-task behavior.
To alleviate tensions around status and legitimacy, it is important that teachers be included in decision-making around co-teaching partnerships well before the beginning of the school year. To minimize differing epistemological assumptions (Arkoudis, 2006), ESL teachers who have knowledge in specific domains should pair themselves with teachers in that content area. Content area teachers need opportunities to dialogue with ESL teaching staff and to observe ESL classes before planning for co-teaching. Co-teaching must be understood from a relational perspective mediated by individual identities, as well as from an instructional one. Teachers must feel a sense of agency over staffing decisions, and feel confident with and knowledgeable of the expertise their colleagues bring to the co-teaching context.
The teachers who desired time for collaborative planning found that scheduling conflicts hindered meaningful and comprehensive collaborative relationships. Preparation times that did not align, and mandatory after school staff meetings meant that teachers were limited in terms of when they could meet. All of the teachers felt that competing demands also stood in the way of their ability to form collaborative relationships with their co-teachers. The ESL teachers felt they had to spend considerable time learning or brushing up on content knowledge, which left them with even less time to meet with the content-specialist teacher. Mrs. Stevens (biology), Ms. Elway (ESL), Mr. Wilson (civics) and Mr. Johnson (math) also suggested the need for longer-term collaborative relationships so that they could work together with the same students over the course of two or more years. They held that teachers and students would benefit from consistency in terms of teaching styles and language instruction. Similar to Davison's (2006) findings, teachers in this study who spent little time planning together were limited to a task orientation, as opposed to developing long-range objectives that focused on broader academic outcomes. Co-teachers need opportunities to get to know one another; as these cases illustrate, co-teaching relationships depend equally on shared or compatible knowledge and skills, and collegiality, and adaptability.

Co-Teaching Identities
In only one classroomalgebra 1did the relationship between the content area and ESL teacher epitomize learning in a community of practice (Wenger, 1998), as these teachers had a shared vision of curricular goals, and of their position within the classroom. The content-area teacher in this and one other classroom held a positive attitude toward ELs in spite of his lack of long-term experience with this population. Both the ESL and content area teachers in these classes attempted to foster inclusivity and a collegial community within these classrooms by engaging students in cooperative learning activities. ELs in these classes displayed general enthusiasm for the course content, and learning outcomes, in turn, were generally positive. These students were given legitimate peripheral participation in classroom discussions and other learning activities.
Two of the three content-area teachers expressed a desire for more classes where English learners are integrated with native-speaking peers. These teachers liked having English learners together in one class, as they felt they were able to get to know these students and target their needs and strengths more easily than in a classroom where they had to manage multiple student needs. Mr. Johnson (math teacher) noted a change in his attitude toward ELs as a result of his co-teaching experiences: "I used to see them as a challenge. Now I look forward to having them in my class. I wouldn't have felt this way without Ms. Elway and [another ESL teacher]." Both Mr. Johnson and Ms. Elway felt that had become better teachers and had learned from one another as a direct result of their collaborative teaching relationship. In sum, whose content area teachers who are better at co-teaching are simply better, more engaged, more motivated, more up-to-date teachers (e.g., Mr. Johnson), whereas those who struggle seem to be teachers who are simply not motivated to engage students in positive ways (e.g., Mrs. Thomas), irrespective of the co-teaching arrangement.

Implications
It is increasingly apparent in educational policy that teachers are responsible for working with both children and adults in the classroom. What remains underemphasized is that working with teachers in the classroom is a skill that may take time to develop. While we stress the benefits of co-teaching as an instructional approach, our research also points to some potential pitfalls. Before and during the school year, teachers should be given additional preparation time that aligns with their co-teachers so that they can develop more collaborative two-way relationships, whereby the ESL teacher becomes more of an active player in designing and disseminating course materials and methods, and can challenge the privileging of content area over language development. Coteachers should be included in sustained professional learning opportunities together, and be provided with guidance on how to successfully collaborate in instructional planning and in the delivery of pedagogical content (Honigsfeld & Dove, 2010). Moreover, in their capacities to adapt and collaborate, teachers should continue to question and to transform current policies that, in some cases, leave them feeling marginalized, and undervalued for their expertise, or territorial and mistrustful of their colleagues.
In addition, one way to address the concern raised over the legitimacy of ESL teachers' knowledge and expertise is to make this dichotomy less apparent in teacher preparation programs. Pre-service ESL teachers should receive more training in content areas, whereas content-area teacher candidates should receive more training in second language learning theory and ESL methodologies within their content areas. Teacher education coursework can be designed so that content-area and ESL pre-service teachers or teachers seeking advanced degrees have opportunities to work collaboratively on unit and lesson plans, designing materials, and selecting appropriate and engaging teaching approaches drawing on each others' areas of expertise. This necessitates that faculty within content-area and domains and bilingual and second language education work together in a collaborative manner to design coursework that offers students authentic opportunities to learn from their colleagues. It is imperative that teacher educators model the very skills and dispositions they wish to see in their students by working across disciplines to develop innovative coursework that highlights the benefits of collaboration, in terms of enhancing our knowledge and skills and adaptability (Darling-Hammond & Bransford, 2005). Fieldwork placements that give teacher candidates a chance to work across program boundaries can build capacity for collaborative work while still in a supervised environment that fosters critical reflection. Teachers' attitudes and personalities must be considered to make sure collaborating teachers can get along and embrace the idea of collaborating, and they should have opportunities for facilitated reflection in order to foster collaboration (Davison, 2006). Findings stress the importance of conceptualizing collaborative teacher relationships in terms of how individuals position themselves and are positioned by others in the classroom context (Hornberger, 2006).
On a final note, as ELs represent a growing population in U.S. K-12 schools, teaching professionals will likely continue to hone and develop new instructional strategies to facilitate their success in school. The present study shows that much can be gained by drawing on scholarship on co-teaching in other domains, such as special education. Future research could further examine where knowledge about co-teaching in each of these fields overlaps, and where differences exist. Research can also analyze co-teaching relationships over a longer period of time, and in various contexts, including during planning periods. Research could examine the efficacy of professional training for co-teaching. Additionally, although continued attention should be given to teacher's perspectives on co-teaching, student perspectives on the benefits and drawbacks of co-teaching could also yield important insights into the efficacy of co-teaching models. In which roles do students find co-teachers most helpful? How do students perceive of co-teachers in their respective roles? Do they attribute more power to one teacher over the other? Does co-teaching impact their learning positively or negatively? Continued attention to co-teacher relationships and practices is a critical step in ensuring that co-teaching is both effective and rewarding to the teaching and learning community within a multilingual classroom.