The Great Wall of “English Only”: Teacher Perceptions of Classroom Ecology in Arizona’s Post-Proposition 203 Era

Abstract In 2000, Arizona voters enacted Proposition 203, which limited the use of children’s native language in public schools and instead, mandated an accelerated one-year English immersion model for children with limited English proficiency. This study examines the results of a 20-question survey and follow-up interviews with 117 Arizona K-12 teachers of English Language Learners (ELLs) to determine their views on the effects of the “English–only” model on teaching and learning. The results revealed that restrictive language policies have upset the balance of classroom ecology, resulting in teacher frustration and concern. Teachers’ recommendations to improve ELL education are also discussed.


Introduction
In 2000, Arizona voters approved the most restrictive English-only education law in the country, which effectively banned textbooks, materials, or teaching in languages other than English. The "English for the Children" initiative, as it was promoted, virtually eliminated bilingual education programs that combined instruction in Spanish, American Indian languages, and English to accelerate the acquisition of academic English. These latter programs had been hailed for many years as an effective way to promote biliterate and bicultural children ready to function in a 21 st century multicultural, multilingual world, and to assist Arizona American Indian tribes in revitalizing and maintaining their languages.
To help schools and teachers comply with the new law, the state and its then newlyelected Superintendent of Public Instruction, developed an English-only immersion program to have English language learners (ELLs) join mainstream classes by their second year.
Accordingly, most teachers of the state's 155,789 ELLs (Payan & Nettles, 2008) were required to participate in workshops and training in English immersion teaching methods, as well as to obtain teacher certification endorsements in Structured English Immersion (SEI). Schools were required to track students who tested out of the one-year English immersion programs and to institute tutoring programs for students that made insufficient progress in English after one year.
Few schools have successfully implemented various language models mandated by the state since passage of Proposition 203. Most schools lacked funds to adequately support programs, and the state was unable to offer technical guidance or resources to help teachers with day-to-day requirements of the law (Arizona Republic, Feb 28, 2006). Adding to this confusion, during the last five years, the office of the Superintendent of Public Instruction changed the ELL exit criteria and the testing instrument (AZELLA) more than four times using different measures and evaluation criteria. Vigorous advocates of the new policies, however, continued to assert that these reforms were showing progress (Horne, 2004;Judson & Garcia-Dugan, 2004). On the other hand, many experts in language acquisition (Combs, Evans, Fletcher, Parra, & Jiménez, 2005;Crawford, 2003;Krashen, 2004;McSwan, 2004) questioned the effectiveness of these programs and provided data suggesting that there are detrimental effects of forcing rapid immersion programs on children at the expense of linguistically accessible content and early learning in their home languages. In spite of claiming large academic gains, the Arizona Department of Education mandated implementation of the first phase of a revised SEI model for ELLs during the 2008-2009 academic year. This "revised" model now requires four hours of segregated, daily, direct English language instruction (Clark, 2009). Thus, the debate has become increasingly convoluted over the processes of effectively educating ELLs as well as determining how much to spend on programs to better serve them. This convolution has been exacerbated by the imposition of "No Child Left Behind" requirements. It appears that the big losers from implementing these politically-oriented mandates continue to be the underserved linguistic minority students and their teachers.
The language and intent of Proposition 203 was seen by many scholars in the language acquisition field as guided by hysteria and unsound research targeted against the language, culture, and identity of a growing population perceived by some as a threat to the English language as a fundamental symbol of American social identity. Moreover, by choosing 'English-only' language policies as an expression of this symbol, advocates subscribed to a popular myth that treats bilingualism as a divisive force that threatens national unity (Johnson, 2005;Stritikus & García, 2005). Not only is this myth largely unfounded, but it ignores the fact that language plurality is normal for most countries in the world (Crawford, 2008).
The increasingly visible and growing number of Latino families in the U.S. has prompted a re-examination of the contours of our American identity as occurred during prior periods of U.S. history, particularly during times of economic crisis and international conflict (Suarez-Orozco & Suarez-Orozco, 2001). The most recent iteration, as manifested in the "English Only" movement, has instigated the enactment of voter initiatives in several states in the form of severe accountability and language mandates, which largely affect culturally-diverse populations educated in already under-resourced schools.

The "Great Linguistic Wall"
Implementation of "English-only" policies replicates, albeit in a more symbolic form, the barriers or "walls" being erected along the Mexico-U.S border. These barriers -both in a symbolic and real context -serve to not only "wall-out" the migration of unwanted, undocumented persons, but they also exert extreme pressure against the language and culture of Latino migrants and citizens who are already here. Accordingly, this infrastructure denies access to and/or restricts government services, including education. As documented by federal research, 76% of all ELLs in the U.S. are of Spanish language background (Payan & Nettles, 2008) and most were born in the U.S. Thus, the quality of education they receive as citizens, especially in forms that marginalize and "wall-out" their home culture and languages, will have a great impact on the economic future of the U.S.
Efforts to construct socially fortified boundaries based on language are reminiscent of the infamous Berlin Wall. For 28 years, this wall effectively curtailed the drain of labor and economic output associated with the daily defection of large numbers of professionals and skilled workers from East to West Berlin, which had political and economic consequences for the Communist regime. A symbolic fortification can be as effective as a physical barrier.
Segregating speakers of languages other than English has the same effect as a physical barrier by "walling-off" their social, linguistic, and cultural capital that possesses the most currency in the American social mainstream. This is evidenced by the fact that ethnically diverse students in this country, especially Latinos, are not doing any better educationally than they were a decade ago.
Their academic achievement has remained static and little progress has been shown across national academic measures (Espinoza-Herold, 2003). For example, less than one-quarter of Arizona students passed the 2007 National Assessment of Education Progress test in reading, math, and science (Arizona Community Foundation, 2008). This report also indicated that less than one-half of the state's graduates were eligible for admission to college. Arizona student demographic analysis indicates that approximately 40% of all students are categorized into ethnically diverse sub-groups. This may serve as an indication that the state's current model of language instruction is not meeting the needs of this growing student population. Arizona ELLs consistently fall far behind their peers on assessments of academic achievement. Large disparities are found for both mathematics and reading on tests developed by both the state and the U.S. Department of Education. Those gaps range from 24 to 32 percentage points, depending on the subject and whether state or national tests are considered (Editorial Projects in Education, 2009).
In 2000, Ron Unz, (the wealthy software developer who sponsored Proposition 227 in California and 203 in Arizona) coined the term "English for the Children" and successfully obtained the attention of the average voter who chose to eliminate bilingual programs in public schools. However, the underlying notion behind the slogan emphasizes divisions in society, largely associated with the languages used by non-white Americans and undocumented immigrants. The ideology of "English for the Children" and similar movements are being actively promoted in other states, while ignoring or discounting effective and proven methods used to educate thousands of U.S. born language minority children. Overall, Proposition 227 has failed to produce substantial change in programmatic efforts for ELLs as there have been no marked differential achievement gains on standardized exams.
In a retrospective analysis of California's education initiatives, as well as those that have been enacted in other states, Bali (2008) notes that criticisms of declining educational quality have been levied on schools, mostly in urban areas, that have received large numbers of immigrants and minorities. Public policies that target minority group access to state public services are more likely to occur in areas with large minority populations, especially recent immigrants. Such policies can be interpreted within the framework of Key's (1949)  The citizen initiatives that target eliminating bilingual education are "subtractive" to linguistic diversity and deny enriching and proven approaches associated with cultural and linguistic resources of children and their families (Stritikus & García, 2005). In fact, replacing bilingual education with "English only" approaches may have a reverse affect on the academic achievement of language minority learners. On the 2007 National Assessment of Educational Progress, fourth-grade ELLs scored 36 points below non-ELLs in reading and 25 points below non-ELLs in math. The gap among eighth-graders was even larger -42 points in reading and 37 points in math (Goldenberg, 2008).
Subtractive approaches designed to move diverse children away from the influence of home, language, and culture toward a presumed reduction in these performance gaps have not proven effective. A meta-analysis of 17 research studies concluded that teaching ELLs to read in their first language and then in their second language, or in their first and second languages simultaneously, boosts their reading achievement in the second language compared with teaching them to read in their second language only (Goldenberg, 2008).
Another study concluded that students placed in dual language programs had a deeper understanding of English concepts than those who learned in mainstream programs (Burt & Dulay, 1979). In Texas, a study of bilingual teachers demonstrated that the "daily and long-term evidence experienced by bilingual teachers provides support for bilingual education as an effective means for educating language minority children" (Bustos-Flores, 2001, p. 264).
Reviewing five major studies conducted in the 1990s, Valencia (2002) concluded that bilingual education was more effective in providing the kind of learning environment that children from diverse language backgrounds needed "in order to gradually narrow and close the academic gap between themselves and the average native English-speaking child" (p. 177). In addition, Krashen & McField (2005) noted: "study after study has reported that children in bilingual programs typically outperform their counterparts in all-English programs on tests of academic achievement in English or, at worst, they do just as well" (p. 34). More recently, López and Tashakkori (2006), concluded that bilingual approaches in education: (1) provide all students access to a quality education despite their language differences, (2) enable students to have broader career opportunities as our economies increasingly bridge international borders, and (3) do not slow the development of the student's English and helps them relate other subject areas to each other. Overall, an abundant body of research demonstrates that linguistically additive approaches accelerate the acquisition of academic English in language minority children and do not mitigate the perceived threats to a significant American social symbol.

Classroom Ecology and Teacher Voice
Classroom ecology refers to the actualities of classroom life as experienced in the interactions among students and teachers (Fuller & Clark, 1994;Tharp, 1989). Classroom ecology is developed through the interconnected processes which impinge upon behavior in the teaching environment, such as routines, practices, perceptions, and interactions in classrooms that serve cultural minority children (Ladson-Billings, 1994). The enactment of voter initiatives involves a political process, which translates into school and classroom practices or ecologies. This classroom environment has varying effects on families, communities, schools, students and especially on teacher functioning.
As the primary mediators for educating children, teachers seem to have little voice in the debate about the needs of the children they serve and the effects it has on classroom ecology, especially where language minorities are concerned. Lemburger (1992) stated, "much of the literature on bilingual education focuses mostly on its legal, political, and methodological aspects; teachers' voices are greatly missing from the literature" (p.1). Concurrently, in a survey of California teachers, Gandara, Maxwell-Jolly & Driscoll (2005) noted: Teachers are both on the front line and responsible for the bottom line when it comes to providing these students with the skills and knowledge they will need to survive and thrive in U.S. society. Yet seldom are teachers invited to share their experiences and their concerns with those who shape education policy. It is critical to ascertain the perspectives of teachers who have so central a role and such a large stake in these issues if instruction for EL students is to significantly improve (p. 2).
Among other groups of California educators, García and Curry-Rodriguez (2000) found the policy mandated by California's Proposition 227 resulted in confusion in schools throughout the state over issues of quality instruction for ELLs, and that teachers reported receiving little guidance. Further, teachers felt that top-down reforms limited their pedagogical classroom decisions.
In another study on teacher reactions to English-only, researchers noted that state mandates combined with high-stakes testing and increased accountability, as imposed by "No Child Left Behind", created "a tense and in many instances, demoralizing environment for teachers". They further noted that "…the key to understanding the challenges educators face in implementing school reform lies in centralizing teachers' experiences inside the classroom, and considering the intersection of these multiple policies with pedagogy" (Alamillo & Viramontes, 2000, p. 12). Most research studies concerning effective practices and programs for ELLs concur that it is not the actual "language" utilized in instruction that is the most important factor, but rather it is the quality, rigor, high expectations, and teacher preparation that impact academic achievement of ELLs (Hamilton, 2006). Also, research is lacking on social and cultural needs specific to ELLs given that most of the focus has been related to oral and written proficiency (Yoon, 2008).
Studies that focus on teachers and their implementation of "English only" programs found some common denominators. In a California teacher study, Gandara, et al. (2005) found the need for: (1) more and better materials for ELLs, (2) more time to teach students and to collaborate with colleagues, and (3) more paraprofessional assistance. They also noted that teachers in small and rural districts face challenges similar to urban districts but often have fewer resources than larger districts.
In response to the Arizona "English for the Children" initiative, Wright and Choi (2006) observed that "lack of teacher voices in this debate is troubling" (p. 4) because teachers have been given the charge to implement "English-Only" policies at the classroom level. Teachers know their students best. yet they are not given credit for their knowledge and experience in the debates over language policy.
In their study of third-grade ELL teachers, Wright and Choi (2006) noted near unanimity by teachers in explaining that children must learn English, but not at the expense of their home language. They also noted that teachers did not oppose the notion that language minority children should be fully bilingual in both English and their home language. After extensive teacher interviews and examining test data, the authors concluded that there was no evidence that Proposition 203 and its mandates for the English-only immersion "model" have improved the education of ELLs. Additionally, they concluded that teachers hold personal views that contrast with the ideology of Proposition 203 and its supporters. Based on the viewpoints shared by teachers in the study, Wright and Choi recommended: Administrators should allow certified, endorsed, experienced teachers to make professional curricular and instructional decisions within their own classrooms and schools based on their students' current levels of English and academic proficiency. The state and school district administrators need to find ways to increase teacher morale and create incentives for teachers in ELL impacted schools to remain to prevent high teacherturnover rates. The state should establish a system to allow the input of experienced ELL teachers into the educational policy-making process for policies which affect ELL students (p. 49).
This investigation utilizes the work of Wright and Choi (2006) to extend the input of teacher perspectives beyond grade 3, and extends the scope of participant viewpoints to 117 K-12 Arizona educators working in 16 urban and rural school districts. In addition, it situates teachers' views on actual classroom practices and the effects on classroom ecology. This study examines teachers' perceptions almost a decade after the passage of Proposition 203 and subsequent implementation of the state-mandated SEI model to instruct English learners. It is a good measure of day-to-day challenges in pedagogy as well as their influence on ELL academic achievement and teacher satisfaction.

Purpose of the study
The survey and interview protocol in this study were designed to answer three research questions related to the views of K-12 teachers: (1) the impact of SEI on ELL academic English development and academic achievement, (2) the impact of SEI on classroom ecology, and (3) program improvements that are needed to increase the effectiveness of instruction for ELLs. The teachers' views provide a refreshing barometer to assess the effects of language programs in the classroom. Collectively, these teachers provide the best assessment to date of how the state-mandated English acquisition model was implemented, and the impact it had on the education of ELLs. minutes. The survey instrument was driven by informal conversations with in-service K-12 educators who suggested the need to develop this study, and by the experience of the author as a former K-12 teacher in Arizona.

Methodology
The survey was administered at selected large and medium-sized school districts in urban and rural Arizona that had large numbers of ELLs. Approximately 80% of the participants worked at inner city schools in the Phoenix area, while a minority of participants was employed at rural school districts and Native-American reservations. All teachers had ELLs in their classrooms. Letters were sent to school administrators and participating teachers in spring 2007 describing the study and assuring anonymity for schools and participants. The letters also indicated that participation was voluntary. Undergraduate and graduate student volunteers assisted in implementing surveys and follow-up interviews later that spring. In many cases, undergraduate students implemented surveys in school districts where they were completing their practical internship requirements. In addition, graduate students who were employees of participating school districts identified colleagues whom they knew were teaching ELLs.
Of the 188 surveys sent to teachers, 117 or 62.2 % were returned to the investigator. Data were analyzed using a mixed-method approach (see Greene & Caracelli, 1997a;Greene & Caracelli, 1997b;Miles & Huberman, 1994) which provides for the objectivity of quantitative approaches combined with the ability of subjects to qualitatively detail their experiences through interviews (Heron, 1981). This type of analysis also serves as confirmatory analysis, where quantitative measures could reveal additional aspects of the qualitative analysis as well as the converse. For quantitative data, responses for each question were entered into SPSS (a statistical analysis software program) and were compared to a random distribution using a Chi-square or Kruskal-Wallis test (SPSS, 1990). For qualitative data, open ended responses to questions and any comments or reflections made by teachers were assembled and analyzed for patterns and common themes (Aronson, 1994). The qualitative data provided invaluable insight that allowed us to explore at greater depth the quantitative data obtained through the selected response survey items.

Survey limitations
Although the study was diverse in terms of rural versus urban populations of ELLs, it relied on schools in which university student interns were placed. As such, the survey was not based on a randomized sample of schools, but rather a cross-section of high-need schools that required more student interns. On the other hand, we were only able to analyze 117 out of 188 surveys or 62% of the participants. The remaining 71 surveys, or 38% of responses, were excluded from analysis because 17 participants did not return surveys and 54 participants did not complete all survey questions. Overall, performing a review of biases in this sample compared to the full population is beyond the scope of this manuscript.

Participants
Most respondents (89%) were female, which is consistent with gender representation in the national teaching force, particularly in elementary grades. Most participants (63 or 68%) taught grades 1-4, 31 teachers (26%) taught grades 5-8, and only 7 teachers (6%) taught high school ELLs. The majority of participants (81%) were of European-American ancestry, which is consistent with the demographics of the national teaching force. Many teachers (46 or 39%) had from 1 to 5 ELLs in their classes. On the other hand, 26 teachers (22%) had 6 to 10 ELLs, and 34 teachers (29%) had 11 to 20 ELLs. Only 11 teachers (9%) had more than 20 ELLs. The majority of participants were experienced teachers: 88 of 117 respondents (75%) had 4 or more years of teaching experience, and 64 teachers (55%) had at least a Master's degree. In addition, 47 teachers (40%) had earned a full endorsement to work with ELLs (ESL, Bilingual or the new SEI endorsement). The rest of the teachers were in the process of obtaining the newly mandated SEI endorsement. In terms of the student population, a large majority of ELLs in the study (84%) were of Mexican-descent. The remaining students were from the middle-East, Africa, and Asia.

ELL academic English proficiency and achievement
Teachers overwhelmingly agreed in their views related to ELLs lack of success learning academic English ten years after the dismantling of bilingual education in Arizona. When asked if ELLs were learning to speak English faster under the new SEI model, 23 respondents (19.7%) responded "yes" while 91 teachers (77%) responded "no". In question 7, when asked if ELLs were learning to read and write academic English faster with their current language model, 16 teachers (13.7%) responded "yes" and 96 teachers (82.1%) responded "no". For both questions, there were a significantly higher number of "no" responses than expected compared to if teachers had answered "yes" or "no" with equal frequency (speaking: Chi-square test, Χ 2 = 40.6 (df = 1), N = 114, p < 0.001; reading and writing: Χ 2 = 57.1 (df = 1), N = 112, p < 0.001).
Several teachers (3.7%) did not have an opinion or did not respond. Their neutrality may stem from confusion with regard to the state-mandated SEI language program for which many teachers had not received adequate training.
In post-survey interviews, the majority of teachers expressed their frustration with the SEI model and the rate of ELL's English language development and academic achievement.
Following are some typical responses in regard to students learning to read and write academic English: "I am a frustrated teacher. Believe me! Proposition 203 has not been beneficial in teaching ELLs….The necessary scaffolding to make connections between the students' native language and English has been removed." Another teacher commented that the SEI model may have helped ELLs acquire conversational English faster but not the language that these students needed to succeed academically: "Maybe they are learning conversational English faster than they used to, but it is at the expense of content area and knowledge! This is a net loss for the students!" Two experienced teachers noted: "I think full immersion in English without support in bilingual is hindering students. They can't associate English with what they know, and they are unable to relate the new material to what they know in the full immersion program. Arizona's version of SEI needs a bilingual component at least at the beginning and intermediate levels." "Students are speaking English faster, but not academic English. It is difficult for them to talk about new concepts, when they have a limited language foundation." On the other hand, a minority group of teachers (19.7%), indicated satisfaction with the SEI model in the following way: "Yes, they are learning English faster because they are fully exposed to English so that they pick up the language faster. I do not see much of a difference-motivation, not legislation is the key." "The students are working more closely together with teachers and other students on their language acquisition skills. They pick up on the language by being immersed in it all the time they are at school." The post-survey interviews corroborated the quantitative results of the surveys in almost equal proportions. The majority of teachers felt that the SEI model was not producing the results in academic English and achievement for ELLs anticipated by Proposition 203 several years after its implementation. For some teachers, ELLs may be acquiring conversational English relatively fast but at the expense of content knowledge and literacy.

Classroom challenges resulting from the SEI model
Teachers were prompted to rank seven classroom-based challenges that they were facing as a result of implementing the new SEI language model (1=highest challenge, 7=lowest). These seven challenges included: (1) lack of materials and teaching strategies, (2) one-on-one teaching, (3) teaching language and content simultaneously, (4) no bilingual education program, (5) different proficiency levels, (6) students lack motivation or have bad behavior, and (7) no communication with students or parents. The survey also provided an "other" category so that the respondents could enter an option not provided in the survey. Based on the total number of one through three rankings for each challenge, the top three challenges were: (1) ELLs need oneon-one teaching/tutoring (92 of 112, 82.1%), (2) ELLs have different proficiency levels which makes it hard to effectively help them (79 of 111, 71.2%), and (3) teaching English language and content simultaneously (52 of 112, 46.4%). For the need of one-on-one teaching/tutoring, 40% of the respondents listed this as the most important challenge, 22% indicated this challenge was their second highest priority, and 20% listed it as their third priority. This robust finding indicates that teachers believe that ELLs need instruction in a more individualized format given the multiple proficiency levels that they possess in reading, writing, and speaking.
Responses were also analyzed using a Kruskal-Wallis test (non-parametric analysis-ofvariance) to determine if teacher experience (independent variable) affected the ranking of responses (dependent variable). The four levels of teaching experience were 0-3, 4-5, 6-10, and greater than 10 years of experience. The results show that teacher experience did not affect the ranking for any of the seven responses (P > 0.25 for all seven challenges). Note that even when there were 117 respondents, the total number analyzed was lower (N=111 or 112). Four respondents listed other challenges, and two respondents did not rank all seven options or were missing at least one response. Lack of administrative support, absence of professional development, and excessive testing were listed by the four teachers in the "other" category.
In their open ended responses and in subsequent interviews, these educators revealed their frustration for not being able to provide individual assistance to their ELLs, and secondly, for the lack of bilingual assistants and tutors in their classrooms. In positioning themselves relative to ELLs, teachers identified themselves with a traditional model of teaching content in a single language to a class and small groups within classes. They perceived the addition of ELLs as challenging their professional competence to adequately meet the language learning needs of ELLs. Further, teachers positioned themselves as having an unfair burden imposed upon them by Proposition 203, which created impossible new teaching challenges in classrooms already overburdened with accountability standards and administrator imposed responsibilities. One teacher noted: "The number of expectations that school administrators have for us is overwhelming. If I could only teach-and not be part of various school committees that waste our time and the number of endless projects that do not resolve anything-I could devote more time to my classroom and children. Our classrooms are too large in Arizona to meet individual needs of students and especially, to have time to help ELLs." Teachers often associated ELLs as a form of "special needs", similar to those of special education students. Associated with these needs, teachers indicated that 45% their ELLs needed one-on-one teaching help, 16% indicated that different proficiency levels of students confounded the teaching role, and 16% noted that ELLs need bilingual education programs. One respondent noted: "Teaching language and content simultaneously is impossible. ELLs need one-on-one help. My class is too large to give them the academic support they need to ensure that they are comprehending the material. I try to help as much I can, but I need to advance with the curriculum I should be teaching. Students are not motivated to learn anymore and they display bad behavior. I spend a lot of time on classroom management issues. When children do not comprehend English, they are easily distracted and bored. I am so frustrated with the lack of support that I have been thinking of leaving the teaching profession." Most teachers indicated that they were frustrated at not being able to meet the individual needs of ELLs given their varied English proficiency levels as well as content knowledge. Almost a quarter of teachers interviewed (24%) indicated that schools should return to the bilingual education model used for ELLs before Proposition 203. The remaining teachers seemed willing to adapt to the presence of ELLs in the classroom if additional personnel like teaching assistants worked in the classroom. Several teachers listed other challenges such as lack of communication with parents, students' lack of motivation, insufficient training, and professional development.
In particular, many teachers indicated that the prohibition to speak a language other than English had presented new challenges. They commented that many parents who used to attend school functions had stopped doing so for fear of their inability to speak in English with teachers. One teacher commented: "This law has infused fear and mistrust among some of the parents of my students who are not coming to school anymore. I wonder if this new SEI model had taken research into consideration-we should never outlaw someone else's language!" During post-survey interviews, teachers also noted that some challenges can be mitigated by a better knowledge of cultural background.
"Research indicates that teachers should educate themselves about students' culture and language because so much of their learning comes from their cultural background. I have enrolled in an intermediate Spanish class at a local community college because I want to be able to communicate with my ELL parents." "It is very helpful to understand a student's cultural background. It gives you insights to the students' interests and experiences and it gives you something to draw on when teaching or it helps you get the children hooked."

Improving the current model
The 117 participating teachers were asked their opinion on potential improvements and urgent changes needed to better serve the educational needs of ELLs in Arizona. The six potential changes included: (1) incorporate multicultural education, (2) return to bilingual education, (3) self contained ELL program, (4) eliminate testing for the first years, (5) have a pull out program, and (6) hire assistants/tutors. Based on the total number of one through three rankings for each improvement, the top three improvements were: (1) hiring assistants/tutors (88 of 113, 77.9%), (2) eliminating tests for the first few years (69 of 113, 61.1%), and (3) incorporating multicultural education (57 of 113, 50.4%). It is important to note that the response "returning to bilingual education" was a narrow margin behind the response "incorporating multicultural education" (49 of 114, 43.0%). However, if we only look at the number of teachers that ranked these two improvements as their first priority, then "returning to bilingual education" was more important (N=27) than incorporating multicultural education (N=15).
Providing a pull out program was considered to be of moderate importance (52 of 113, 46.0%), but only 6 teachers identified this as their first priority. Providing a self contained ELL program was considered to be of low importance (27 of 113, 23.9%). The majority of teachers (74%) also completed the "other" category which provided them an opportunity to list improvements that were not in the survey. Many teachers were creative in their selection of possible solutions and changes to the SEI model. Several listed computer-assisted instruction programs, having two certified teachers in classrooms with 10 or more ELLs, smaller class sizes, school district's autonomy to create their own language models, or instituting merit pay for teachers who were willing to learn the students' native language.
Data were then analyzed to determine if teacher background influenced the ranking of responses in regard to how teachers would change the ELL program. Three measures of teacher background were included in the analysis: (1) level of education (Bachelor's degree, Master's degree), (2) grades taught (K, 1-2, 3-4, 5-6, 7-8, and 9-12), or (3) teacher experience (0-3 years, 4-5 years, 6-10 years, and 10+ years). We then analyzed the responses using a Kruskal-Wallis test (SPSS, 1990). The three measures of teacher background were used as the independent variable and responses to each of the potential changes were used as the dependent variable. The results show that none of the three measures of teacher background affected ranking for any of the six possible responses (P > 0.05 for all six improvements).
Corroborating the survey findings, several teachers who participated in interviews when answering "how would you change the current SEI model?" stated: "I would have more classroom aides available for teachers to help with the serious neglect of ELLs we have right now. We simply can't provide individual attention to students in so many English proficiency levels. Besides, we need to teach the content standards and prepare them for state testing. It is an impossible job! I would like to receive training and have the AZ Department of Education trainers show us how to do all this!" Also, in response to improvements to the current model, almost all of the teachers surveyed (95%) believe that teachers should be familiar with the home culture of their students and that this knowledge is essential as it aids communication (34%), and empowers students psychologically (20%). While knowledge of the home culture is essential, only 43 teachers (37%) had visited the homes of ELLs. In sum, this third theme regarding changes to the current SEI model produced a large amount of data and lively discussions during the interviews. The teachers in the study seemed anxious to see real changes and less classroom experimentation.
The demands and restrictions, unreal expectations, and the punitive actions against teachers and schools were evident in the survey responses and interviews. A very vocal and demoralized teacher told us: "I look forward to going home every day and I do not enjoy teaching anymore. The unrealistic demands of school board members and the superintendent of public instruction…people who have never taught and who will never appreciate the job public school teachers perform is taking its toll on me. The general public thinks that teachers can spend leisure summers, when in reality, summers are spent taking courses to recertify or SEI courses who do not help us with most classroom challenges we have to face. When our starting pay is $30K a year, I do not think I will be able to survive another year in this school." The goal of this study was to gain descriptive insights into the perceptions, practices, and routines that constitute the ecological classroom functioning involving ELLs and their teachers.
The testimonies and data collected in this study provide substantial evidence that Proposition 203 and its SEI model have affected in a substantial and negative way the ecology necessary for healthy and productive student-teacher interactions. The following section discusses the main findings and conclusions.

Discussion
The main finding of this study is that the overwhelming majority of teachers believe that Proposition 203 failed in its promise to increase the rapidity with which ELLs acquire academic English. Further, since bilingual education classes were discontinued and most ELLs were first mainstreamed into regular classes, (and five years later, segregated into self-contained environments) most teachers see one-on-one instruction as an urgent remedy. Most respondents felt that they do not have the resources or time to meet the needs of an academically diverse non-English speaking population. Compounding these latter problems, teachers also reported that they were challenged to teach English language and content simultaneously to students that vary in proficiency levels. Essentially, teachers felt that while they are under a mandate to teach ELLs, they lack the time and resources to adequately address the education needs of these children. As a result, ELLs are not learning to speak, read, and write academic English at a rate that would adequately support increased academic achievement.
The findings documented above covered three major themes and research questions in relation to the current state-mandated language SEI model and its impact on teaching, learning and academic achievement of ELLs. In this section, we draw three main conclusions based on the analysis of the findings: (1) the state-mandated SEI model has not accelerated the acquisition of academic English by ELLs, (2) the new model has not improved the general education of ELLs, and (3) Proposition 203 and its SEI model have created significant teaching challenges and disruptions to a healthy and productive classroom ecology.
While teachers in this study were often ambivalent about the merits of "English Only" versus bilingual education, they felt unprepared and under-resourced in terms of meeting the demands of a dramatic shift toward SEI as mandated by Proposition 203. Almost all teachers indicated their awareness of the social aspects of learning for ELLs as indicated by the 95% agreeing that teachers need to be familiar with home culture, while less than one-third had engaged home culture in their teaching. When teachers were asked to rank the top challenges brought about by the new state-mandated ELL program, over 80% stated that ELLs needed oneon-one assistance given their very diverse proficiency levels. In this regard, teachers resented the fact that before the elimination of bilingual programs, each classroom was given a teacher aide who took care of providing individual assistance to ELLs. The sense of frustration brought about by growing numbers of ELLs in the classrooms was a reflection of: (1) challenges to the competence of the teachers to meet the needs to ELLs without additional resources, (2) an added pressure for accountability on top of other responsibilities associated with classroom practice, and (3) a sense of disconnect and lack of knowledge of the home culture and language of the ELLs. In all, the teachers' expressions of frustration regarding ELLs reflect and underscore a growing crisis not only in English only states, but also in professional education that challenges the mainstream model of instruction and classroom management. Teachers are expected to do more with less, respond to increasing linguistic and cultural diversity without preparation and training, and to model and adapt their teaching to politically-defined populations and categories of students. On top if this there are greater accountability standards imposed by "No Child Left Behind," which requires all testing in English. The teachers' responses indicated that the bottom line is that quality of instruction is greatly decreased, as well as their sense of control over their own professional responsibilities. The attitudes and positioning of teachers in this study suggest a growing and acute dilemma in the teaching profession. If these problems are not resolved, teachers will continue to become disenchanted and leave the profession.

Conclusions
This study expands the research on third grade teachers by Wright and Choi (2006) to document the more robust finding that K-12 teachers, regardless of grades taught, years of experience, or educational background, overwhelmingly believed that Proposition 203 and its mandates for the English-only immersion "model" have not improved the education of ELLs. This study joins other investigations cited by Hamilton (2006) who noted: "a group of new studies is providing fresh evidence of what many researchers have been saying all along: English immersion has more political appeal than educational merit" (p.1). So, the dilemma faced by teachers in responding to the educational needs of ELLs is expressed not only in their concern for lack of resources, but also their ability to provide them appropriate and effective instruction. This dilemma is best captured in Hamilton's further observation (2006) that "there is evidence that it's not the language of instruction that counts, but the quality of education" (p.1). The teachers' voices in this study are in accord with ongoing research not only about ELLs, but also about those quality issues effectuated by non-educator citizen initiatives. Teachers in this study indicated that ELLs need substantial classroom support if they are going to succeed in American classrooms, as well as teachers who are free to exercise their professional knowledge and instincts. Public initiatives, which effectively ban bilingual education or any other research-based model, deny teachers the proper exercise of their responsibilities and certainly; quality is affected as a result. Initiatives such as Proposition 203 fly in the face of many recent research studies, which have sustained the efficacy of bilingual approaches in teaching language minority children -not only in student learning but in also effectively involving communities in the education of children. A sink or swim approach is both inadequate and inappropriate as it frustrates teachers and disserves ELLs. This approach also creates an unnecessary barrier or "wall" of demarcation between communities and schools, lessening the beneficial guiding effects of research and the professional judgment of teachers who must deal with the linguistic realities of children expecting to succeed in the educational system. In terms of classroom ecology, a healthy balance is almost impossible to achieve under such conditions. As classroom ecology is closely related to cultural and social characteristics of the community, the lack of balance becomes further exacerbated when connections to the community are disrupted, as indicated by the teachers in this study.
As the findings reveal, the Great Wall of "English only" created an ecological imbalance in schools and neighborhoods, and it also segregated entire language minority communities through the relegation of languages spoken in the U.S. to second class and 'undesirable' status.
As the pressure of immigration enforcement increases with the erection of walls between neighborhoods, so will the symbolic backlash represented in "English Only". While in time this will pass, as many other reactive social responses have, the damage to the integrity of the educational system, especially those that serve diverse communities, may become irreparable and irreversible. A sound, quality educational system must accommodate the judgment and intuition of well-prepared and competent teachers. Denial of competent exercise of responsibilities to teachers renders them frustrated, ineffective, and without proper control of the institutions, policies, and services that allow them to provide research-driven education to diverse children.
This study joins others in demonstrating that teachers are reacting negatively to interference with their experience and judgment wrought by citizen initiatives. As these initiatives take hold in classrooms, teachers are feeling both helpless and hopeless in being required to wage a battle that is not of their making, and without the support they feel they need to meet the learning needs of the children they serve. While walls are built to exclude, they can also confine. We have observed how teachers in this study can become "walled in" by laws and policies they did not favor while those they serve are "walled out" of effective teaching and research-based practices. Thus, a great divide is created that will surely result in undesirable long-term consequences to this nation in an increasingly internationalized economy and an environment of cross-cultural communication.