Keywords

Introduction

A common characteristic of all three North American countries is that their populations consist of settlers, the descendants of Europeans who settled in North America more than four centuries ago, and Indigenous communities, who lost most of their ancestral lands in the settlement/invasion by Europeans. The physical eradication of many Indigenous communities during the initial conquest and later territorial expansion by European settlers was compounded by what the Truth and Reconciliation Commission of Canada (2015) called “cultural genocide,” which Indigenous children experienced in residential schools that operated in Canada and the United States for more than 150 years. These schools were explicitly designed to eradicate Indigenous languages and destroy children’s Indigenous identities. In Canada, children were shamed and physically beaten for speaking their languages, and many experienced sexual abuse and torture at the hands of the religious orders which operated the schools (Truth and Reconciliation Commission of Canada 2015).

The legacy of residential schools and the racism that gave rise to them is large-scale underachievement among Indigenous students in Canada and the United States. Although most Indigenous students speak English (or a dialectal variety of English) as their home language (L1), many do not acquire sufficiently strong levels of English academic skills to pursue college and university qualification. Although sustained education decolonization projects have been undertaken in some North American Indigenous contexts (see, e.g., López-Gopar [2016] for Mexico, McCarty [2008] for the United States, and Walton and O’Leary [2015] for Nunavut in Canada), structural challenges such as the shortage of formally qualified Indigenous teachers have constrained the impact and scalability of these projects.

In Canada and the United States, most school-age learners of English are from immigrant backgrounds. From the beginnings of European settlement, both countries have sought and attracted large numbers of immigrants seeking new opportunities and a better life. For example, the province of Quebec in Canada was predominantly settled by French speakers, and French is the only official language in Quebec, although English is one of the two official languages at the federal level across Canada. Thus, English is taught as a second language (L2) to French-speaking students in Quebec. There are also pockets of minority francophone communities across other Canadian provinces, and these students have the right to attend French-medium schools, where English is also taught as a second language. Within Quebec, with some very limited exceptions, immigrant students are required to attend the French school system, and thus there is minimal teaching of English as a second language in Quebec English-medium schools.

Over the past 20 years, annual immigration to Canada has been around 250,000, with an increase to more than 300,000 since 2015. This has resulted in large numbers of students who come to school from homes where languages other than English or French are spoken. In Ontario, about 20% of the school population has grown up speaking a language other than English or French, and in large cities such as Toronto and Vancouver (in British Columbia), more than 50% of the school population comes from multilingual homes.

In the United States, almost 5 million students, representing about 10% of the school population, are identified as “English language learners” (ELLs), and this number is considerably larger in major urban centers across the country (National Center for Educational Statistics 2018). The largest group (3.8 million) is comprised of Spanish speakers, but many other languages are also represented (e.g., Chinese varieties, Arabic, Vietnamese). According to Sanchez (2017), California has 29% of all ELLs nationwide followed by Texas (18%), Florida (5%), and New York (4%). A large majority of ELLs (also termed “emergent bilinguals” in this paper) are born in the United States and are US citizens (85% pre-K through grade 5; 62% grade 6 through 12).

English language teaching (ELT) in Mexico differs from ELT in Canada and the United States insofar as Spanish is the language of instruction in almost all schools, except for some English-Spanish bilingual programs mostly in private schools and some bilingual programs involving Indigenous languages. Thus, English is taught as an additional language to students whose L1 is predominantly Spanish or, in some cases, an Indigenous language.

Nature, Trajectories, and Outcomes of ELT

Nature of ELT

In order to understand students’ English language learning trajectories and outcomes, it is necessary to distinguish between social and academic language or what Cummins (1981a) has labelled conversational fluency and academic language proficiency. Conversational fluency reflects our ability to carry on a conversation in familiar face-to-face situations where meaning is supported by facial expressions, gestures, eye contact, intonation, and the immediate environment. This dimension of language proficiency is developed by the vast majority of native speakers of any language by the time they enter school at age 5 or 6. Phonology and fluency, in particular, reach a plateau with minimal further development after age 5 or 6. Conversational language use involves high-frequency words and expressions as well as relatively common grammatical constructions.

Academic language proficiency, by contrast, represents an individual’s access to and command of the specialized vocabulary and functions of language that characterize formal schooling. It reflects the extent to which a student can comprehend and use the oral and written language that appears in the subject matter of academic disciplines and in discussions about these disciplines. It involves knowledge of less frequent vocabulary and more complex grammatical constructions, which are seldom used in face-to-face conversation. For example, the passive voice is a common feature of academic language, as is nominalization, where an abstract noun is created from a verb or adjective (e.g., acceleration). Unlike conversational fluency, students’ proficiency in academic language continues to develop through the school years and beyond both among native speakers and learners of English.

Learning Trajectories

Newcomer immigrant students who arrive in the early years of schooling typically pick up L2 conversational fluency quite rapidly when there is exposure to the language in school and in the wider environment (e.g., on television). One to 2 years of exposure to and learning the school language are usually sufficient for young learners to acquire a comfortable degree of fluency in that language. Students who arrive at older ages (e.g., in their teenage years) may take longer to acquire L2 fluency and may retain traces of their L1 accent in the new language. By contrast, newcomer students typically require at least 5 years, on average, to catch up academically (Collier 1987; Cummins 1981b); this is because of the complexity of academic language (e.g., many more low-frequency words) and the fact that students are catching up to a moving target – native-speaking students who continue to make gains in vocabulary, reading, and writing skills every year.

Outcomes of ELT

United States. Sanchez (2017) summarizes the academic outcomes for ELLs in the United States as follows:

Many ELLs remain stuck in academically segregated programs where they fall behind in basic subjects. Only 63 percent of ELLs graduate from high school, compared with the overall national rate of 82 percent. In New York State, for example, the overall high school graduation rate is about 78 percent. But for ELLs, it’s 37 percent, according to the National Center for Education Statistics. Of those who do graduate, only 1.4 percent take college entrance exams. (p. 8)

Collier and Thomas (2007) similarly note that students taught English-as-a-second-language (ESL) as a subject at the secondary level or placed in ESL pullout programs at the elementary level frequently fail to catch up academically: “Our research findings across numerous school districts in the USA indicate that the average achievement levels of high school graduates who were initially placed in ESL pullout programs is the 11th percentile” (p. 344).

The challenges facing immigrant-background students in US schools are also reflected in the findings of the Organisation for Economic Cooperation and Development’s (OECD) Programme for International Student Achievement (PISA). PISA data regarding the academic performance of 15-year-old students from about 70 countries around the world have been reported since the year 2000, and supplementary analyses in some years have identified the performance of first- and second-generation immigrant-background students (e.g., Christensen and Segeritz 2008; Stanat and Christensen 2006). In the 2003 assessment of reading skills, first-generation immigrant-background students (born outside the United States) performed 50 points below the mean, while second-generation students (born in the United States) performed 22 points below the mean. The PISA mean for all countries is 500 points with a standard deviation of 100 points.

The PISA findings also highlight the role of socioeconomic status (SES) in determining educational outcomes in the United States. The United States is similar to many European countries insofar as the educational levels of immigrants and asylum seekers are significantly less, on average, compared to those of the “mainstream” population. Furthermore, the impact of socioeconomic variables on academic achievement is considerably greater than is the case for countries such as Canada where immigrant-background students perform relatively well in comparison to the non-immigrant background student population. Despite spending more per pupil, on average, than most other OECD countries, there are significant disparities among states in funding allocations to school districts serving students of different SES backgrounds due to the fact that funding predominantly relies on local taxes rather than on centralized federal allocations (e.g., Boykin and Noguera 2011). Consequently, many immigrant-background students who come from socially disadvantaged backgrounds attend schools that are under-resourced in comparison to schools attended by more affluent non-immigrant background students. The under-resourced nature of these schools includes not only per-pupil funding but also the experience and qualifications of teachers and school leaders.

In short, the OECD (e.g., OECD 2010a) has consistently emphasized that equity is a strong predictor of excellence. Indeed, countries that demonstrate greater equity across social groups also tend to perform more strongly on the PISA tests than those characterized by socioeconomic disparities (see also Darling-Hammond 2010). Thus, the underachievement of immigrant-background students in the United States can be attributed at least in part to the socioeconomic disparities that characterize its schools and society. Furthermore, the fact that immigration remains a volatile and divisive political issue in the United States reflects a social and educational climate that is less conducive to promoting both integration and equity in education.

Research focused specifically on the impact of ELT programs in the United States presents a complex picture. Callahan et al. (2010) analyzed nationally representative data to assess the effects of student placement in English as a second language (ESL) programs at the high school level on academic achievement and placement in college preparatory courses. They reported a strong negative relationship between ESL placement and both academic achievement and placement in college preparation courses, even when controlling for prior achievement and multiple background variables. Callahan et al. (2010) suggest that “disparities in language minority student achievement may be due in part to schools’ placement of students into ESL and policies regarding ESL students’ access to academic content” (p. 24). They specifically point out that “students placed in ESL coursework exit high school with significantly less academic content, even when accounting for English proficiency, prior achievement, generational status, ethnicity, parental education, years in U.S. schools, and school level factors” (2010, p. 26).

Callahan et al. (2010) note that this seemingly counterintuitive finding is consistent with the descriptions of some high school ESL programs that have emerged from ethnographic research, which refer to the “ESL ghetto”; specifically, students identified as “ESL” often experience reduced access to grade-level academic content because their level of English is not deemed sufficient to master this content. The authors note that their findings do not address pedagogical approaches within ESL classes. The negative effect of ESL placement is largely due to the fact that language-focused ESL coursework takes up space in the student’s schedule that might preclude students’ access to more academically rigorous and engaging coursework. They argue for the need to expand academically challenging content-based language support services at the high school level.

The findings of Callahan and colleagues (2010) may not be generalizable to the classification of immigrant-background students as ELL in the early grades. Shin (2017) investigated the issue of whether an initial designation of students as ELL influences their later academic achievement. She reported that among students near the cutoff for designation as ELL or non-ELL, the classification had significant positive effects on ELLs’ academic achievement in the elementary grades and, to a lesser extent, in the later grades.

Canada. A synthesis of research findings from Montreal, Toronto, and Vancouver demonstrated that, in general, immigrant-background students tend to perform relatively well in Canadian schools (McAndrew et al. 2009). This study reported that the academic performance of students whose home language differed from that of the school exceeded what would be predicted based on various risk factors such as low SES:

In some sites, the results of the target group are even slightly higher than that of the comparison group [native-speakers of the school language] with regard to graduation rates, performance in various subjects, and most of all, participation in selective or university-bound courses. (2009, p. 16)

This general pattern is also apparent in the OECD’s PISA findings. The OECD (2010a) summarizes the performance of Canadian immigrant-background students in reading abilities as follows:

PISA results suggest that within three years of arrival in Canada, immigrants score an average of 500 on the PISA exam, which is remarkably strong by international standards. For comparison’s sake, in the 2006 PISA assessment of reading, Canadian first-generation immigrants scored an average of 520 points, as opposed to less than 490 in the United States and less than 430 in France. Canada is also one of very few countries where there is no gap between its immigrant and native students on the PISA. (By contrast in the United States the gap in reading is 22 points, and in France and Germany it is around 60 points). Second generation Canadians perform significantly better than first generation Canadians, suggesting that the pattern is of progress by all students over time. Finally, Canada is one of the few countries where there is no difference in performance between students who do not speak the language of instruction at home and those who do. (pp. 70–71)

The OECD (2010a) attributes the relative success of immigrant students as a group to the fact that their average socioeconomic status is equivalent to that of native-born students and they attend schools that are of equal quality to those attended by other Canadian students. The report also points to the fact that immigrants are welcomed as part of Canada’s commitment to multiculturalism which “provides a distinct philosophy that seeks to both respect the importance of native cultures while also incorporating immigrants into a distinctively Canadian identity” (p. 71).

However, this apparent success masks considerable variation in students’ academic outcomes. Studies in Alberta (Derwing et al. 1999; Watt and Roessingh 1994, 2001) revealed that large proportions of ELL students failed to graduate with a high school diploma (60% in the Derwing et al. (1999) study and 74% in the Watt and Roessingh (1994) study). More recent studies from British Columbia also show a high “disappearance” or non-completion rate among ELL high school students (Gunderson et al. 2012; Toohey and Derwing 2008). Immigrant students from higher socioeconomic backgrounds tended to perform considerably better than those from refugee and/or low socioeconomic backgrounds. In some contexts, the extremely strong performance of some groups of socially advantaged students masks the relatively weaker performance of students from less advantaged groups.

Some of the largely positive results for Australia and Canada can be attributed to selective immigration that favors immigrants with strong educational qualifications. In both countries, the educational attainments of adult immigrants are as high, on average, as those of the general population. In Canada, about 60% of immigrants fall into the “economic” category, selected for their potential to contribute to the Canadian economy, with the remainder distributed between refugee and family reunification categories. In addition, both Canada and Australia have encouraged immigration during the past 40 years and have a coherent infrastructure designed to integrate immigrants into the society (e.g., free adult language classes, language support services for students in schools, rapid qualification for full citizenship, etc.). Both countries have explicitly endorsed multicultural philosophies and policies at the national level aimed at promoting respect across communities and expediting the integration of newcomers into the broader society.

The impact of SES on school achievement differs significantly between Canada and the United States. Although there are significant SES disparities among the student population in Canadian schools (albeit not nearly as large as in the United States), the impact of these disparities on academic achievement is among the lowest in OECD countries (OECD 2010a), whereas the association between SES and achievement in the United States is about average for OECD countries. Canada also ranks among the strongest performers on PISA with respect to the proportion of “resilient” students, those who perform well despite coming from disadvantaged backgrounds (OECD 2015).

Volante et al. (2017) note that there are significant regional disparities across Canadian provinces with respect to performance on PISA. In some provinces, immigrant-background students tend to perform better on the PISA tests than the non-immigrant population, but in others (e.g., Quebec), immigrant-background groups underachieve significantly. Volante and colleagues suggest that these variations reflect a complex array of intersecting factors including the countries of origin of immigrants in different provinces, integration policies in different provinces and their effects on the ethnic and socioeconomic composition of schools, and the fact that non-francophone immigrant students in Quebec attend French-medium schools and learn French in a North American context dominated by English in the broader social sphere. By contrast, in anglophone provinces, immigrant students are learning a school language that is widely reinforced by social media, cultural influences (e.g., movies), and its international status.

With respect to differences according to countries of origin, Canadian children of immigrants from East Asia (e.g., China) and South Asia (e.g., India) demonstrate higher educational attainment than those from Southeast Asia (e.g., Philippines), the Caribbean, Central and South America, and Southern Europe. The PISA findings from Canada and other countries also show that the concentration of socioeconomic disadvantage in schools plays a more significant role in school achievement than students’ individual SES alone. In other words, when students from low-SES backgrounds attend schools with a socioeconomically disadvantaged intake, they tend to perform significantly worse than when they attend schools with a socioeconomically advantaged intake.

Mexico. As stated earlier, Mexico differs from the United States and Canada in terms of the role of English as a medium of instruction. In most schools, with the exception of elite bilingual schools (English and Spanish) and very few Indigenous-strand schools (Indigenous languages and Spanish), the main language of instruction is Spanish (Sayer and López-Gopar 2015). According to the PISA findings, Mexico is significantly below average compared to other OECD countries in terms of general educational results. Nevertheless, Márquez Jiménez (2017) argues that the PISA results have been used by the media and education officials to produce alarming discourses (e.g., Mexico shows no progress in math achievement during the last 10 years). These discourses result from reductionist interpretations of the successes or failures of the Mexican educational system as well as myopic perspectives masking the complexities of the national sociopolitical and economic situation underlying the Mexican educational system.

In Mexico, ELT as well as education in general is constrained by social inequality. López-Gopar and Sughrua (2014) state:

Among the countries belonging to the Organisation for Economic Co-operation and Development (OECD), Mexico is the country with the widest gap between rich and poor people (González Amador 2013). Sixty million people, 50% of the country’s population, live in poverty, and 51.5 million experience food shortage (Enciso 2013). Despite these statistics, Mexico is home to Carlos Slim, the richest man in the world. In addition, 43% of the country’s wealth is controlled by 0.02% percent of the population. Only 20% of the population is considered neither nonpoor nor nonvulnerable (Olivares Alonso 2013), and 80% struggle financially on a daily basis. (p. 106)

López-Gopar and Sughrua (2014) go on to argue that economic inequalities are even more severe among the Indigenous population, especially in the southern states Chiapas and Oaxaca. For instance, in Oaxaca, more than 75% of the Indigenous groups live in poverty and suffer poor nutrition (Enciso 2013). Currently, “the [Mexican] national educational system is still showing wide inequalities in the distribution of educational opportunities … which mainly affect the lowest socioeconomic groups” (Márquez Jiménez 2017, pp. 4–5; our translation). Hence, the education of Indigenous groups has been reported to be in a critical state with alarmingly low alphabetic Spanish literacy rates and high dropout rates (López-Gopar 2016).

Despite social inequalities and nationalistic views that attempt to resist the “invasion” of the English language in Mexico (Hamel 1994), ELT is part of the Mexican education system both in private and public schools. In the private sector, English is present in elite bilingual schools. Sayer and López-Gopar (2015) argue that “nowadays the default meaning of a ‘bilingual school’ for most Mexicans is Spanish–English” (p. 576; quotations in original). Even though the term “ bilingual” indexes prestige and hence is used as a marketing strategy in elite bilingual schools, the term “bilingual school” actually covers many types of private institutions, ranging from language centers offering 3- or 5 h-per week English courses to elite English-medium schools (Sayer and López-Gopar 2015). Due to this wide range of ELT programs in bilingual schools, it seems almost impossible to determine their degree of success in developing bilingual students. Lethaby (2003), who has years of experience working with elite bilingual schools in Mexico, identified major problems in these schools, such as unclear and unrealistic linguistic goals, shortage of ELT teachers who can teach both language and content, and lack of appropriate materials for elite bilingual schools in Mexico. Most importantly, Lethaby (2003), along with Sayer (2015) and Pennycook (2016), argues that ELT in elite bilingual schools is fueled by a neoliberal discourse, which falsely equates English with economic success. Lethaby (2003) raises the issue of the detrimental effect ELT may have on Mexico’s Indigenous languages and even Spanish. Finally, elite bilingual schools are especially “elitist” in economically poor Mexican states such as Oaxaca where only 5% of the population can afford to attend these schools (López-Gopar and Sughrua 2014), leaving 95% of the population’s English instruction in the hands of public schools.

ELT instruction begins for most Mexicans at the middle school or junior high school level within the public educational system. Without having a clearly spelled out language policy (Ramírez-Romero and Sayer 2016; Reyes Cruz et al. 2011), ELT instruction began in Mexican middle or junior high schools in 1954 (Ramírez Romero et al. 2014). Moving from a grammatical and structuralist approach to the communicative approach in the mid-1990s (Terborg et al. 2007), this ELT instruction has yielded poor results (Davies 2007). In fact, Ramírez Romero et al. (2014) affirm that “even after six years of English [three years of secondary schools and three years of high school], most high school students (ages 15 to 18) have minimal communicative abilities in English” (p. 1022). López-Gopar and Sughrua (2014) explain that these poor results are caused by “limited hours of instruction, inadequately prepared teachers, and incongruous curricula” (p. 107). Consequently, following the younger-the-better language ideology, since the early 1990s, the Mexican government has developed five different, yet overlapping, ELT programs aiming to start at kindergarten age: (1) ELT state programs (from 1990 to 2009), (2) English Enciclomedia (from 2001 to 2006), (3) the National English Program in Basic Education (PNIEB) (starting in 2009), (4) the Program for Strengthening the Quality of Basic Education (PFCEB) (starting in 2013), and (5) the National English Program (PRONI) (starting in 2015). (See Ramírez Romero and Sayer 2016 for a detailed explanation of these programs.)

Focusing on the last three programs, which have exactly the same curricular structure, Ramírez Romero and Sayer (2016) have identified several main accomplishments: increase in number of students in ELT classes, solid design of new curriculum, status of ELT as part of the elementary curriculum, free books widely available, diversity of teacher training opportunities, and more positive attitudes toward ELT. Conversely, these authors have also pointed out remaining challenges such as uneven coverage of the program leaving out students in marginalized communities, the new curriculum not being used by many teachers, severe problems with the distribution of textbooks, insufficient number of qualified teachers, poor working conditions for ELT teachers, and the need of a more multilingual and intercultural approach in ELT in Mexico. Finally, Mexican ELT policies and planning have ignored the needs and challenges of thousands of retornados (migrants who return to their or their parents’ home country Mexico after having lived outside of the country for a period of years). These retornados arrive in Mexico at different ages with English often as their dominant language and at different academic language proficiency levels (Kleyn 2017).

The Emergence of Content-Based Approaches to ELT

The seeds for the emergence of pedagogical approaches that attempt to integrate the teaching of academic content and second languages were sown in the evaluations of bilingual and L2 immersion programs that were initiated in North America in the 1960s and 1970s. These programs (e.g., Lambert and Tucker 1972) demonstrated that it was not only possible but highly effective to develop L2 skills while focusing instructionally on academic content (e.g., science) taught through that language. Krashen’s (1981) concept of comprehensible input was highly influential in explaining how L2 acquisition required L2 input that learners could understand. For example, Krashen’s work was adopted by the California State Department of Education’s (1981) theoretical framework for educating ELLs, and, in subsequent years, various “sheltered instruction” or scaffolding strategies were elaborated by educators and researchers to facilitate content-based instruction for ELLs in mainstream classes.

These instructional strategies typically include the use of visuals such as images and graphic organizers, concrete demonstrations, paraphrasing, and other strategies for developing conceptual knowledge while reducing the linguistic demands associated with that knowledge. In addition to scaffolding strategies designed to make instruction comprehensible to L2 learners, content-based instructional approaches also emphasize the importance for teachers to articulate language objectives together with content objectives in their lesson plans.

Following the publication of its 1981 theoretical framework, California became the first US state to institute formal certification requirements for teaching ELLs. The Crosscultural, Language, and Academic Development (CLAD) certification authorized teachers to provide withdrawal or pullout ESL instruction as well as Specially Designed Academic Instruction Delivered in English (SDAIE) within mainstream or “sheltered” content-based classes for ELLs. A bilingual variant (BCLAD) intended for teachers in bilingual programs was also offered.

Several content-based instructional models for teaching ELLs were elaborated in the United States during the 1980s and 1990s. Among these were the Cognitive Academic Language Learning Approach (CALLA) (Chamot and O’Malley 1986), which incorporated three major components: high priority academic content, academic language development with a focus on literacy, and explicit learning strategy instruction. The Sheltered Instruction Observation Protocol (SIOP) model was developed in the late 1990s (Echevarria et al. 2004) and continues to be widely used across the United States.

In Canada, the theoretical work of Bernard Mohan (1986) was highly influential in enabling teachers to think about the language demands of topics and tasks in content areas. Mohan’s Knowledge Framework proposed six major knowledge structures that underlie the organization of concepts and meaning across the curriculum (description, sequence, choice, classification, principles, evaluation). Cummins and Early (2015) described the Knowledge Framework as follows:

Each of these knowledge structures has distinct linguistic features that set it apart from the others. For example, Description requires the use of adjectives, usually the present tense, and verbs such as “to be” or “to have,” while Sequence would more likely use past tense and action verbs, as in a narrative.… Mohan also argued that each of these distinct knowledge structures can be represented graphically by key visuals. Key visuals, or graphic organizers, as they are commonly termed, are visual displays of information that both organize and simplify content.… They also make visible the cognitive structures that underlie the content. Because these key visuals express the cognitive structures in a way that reduces the linguistic demands of linear text, they are highly effective in helping ELLs to understand content. (p. 38)

Despite the pioneering work of Mohan and his collaborators (e.g., Early and Hooper 2001), Canadian teacher education programs have been slow to incorporate insights about scaffolding and effective teaching of ELLs into mainstream initial teacher education. For example, teacher education programs serving the urban centers of Toronto and Vancouver implemented mandatory courses on supporting English language learners only around 2015, about 30 years after similar requirements were established in California.

In summary, although content-based ELT programs are widely recognized in Canada and the United States as “best practice” by researchers and educators directly involved in teaching ELLs, current provision both within schools and initial teacher education programs is inconsistent and, in many cases, incoherent. This is partly due to the fact that different provinces (in Canada) and states (in the United States) pursue different policies with respect to curriculum and assessment. However, more fundamentally, the incoherence is due to the fact that despite progress in recent years, ELT in many contexts is still seen as marginal rather than mainstream within school policies and instructional practices. Particularly at the secondary level, there is no requirement or even expectation that school leaders should be familiar with the knowledge base regarding ELLs or with principles for effective teaching of these students. Classroom teachers, again particularly at the secondary level, still frequently view the development of ELLs’ academic language proficiency as the job of the ESL teacher rather than as a whole-school concern.

ELT Within Bilingual Programs

The volatile debate since the 1970s within the United States about the legitimacy and effectiveness of bilingual programs has largely been resolved with respect to the empirical data, although the ideological divisions regarding immigration and social equality that fueled this debate have intensified in recent years. Different types of bilingual programs for ELLs have been implemented in the United States. The following three categories capture the range among these programs:

  1. 1.

    Transitional bilingual programs are intended as a relatively short-term bridge (typically 2–4 years) to enable students to continue learning subject matter content in their L1 while they are catching up academically in English.

  2. 2.

    Developmental bilingual programs provide L1-medium instruction together with English-medium instruction throughout the elementary school years (kindergarten through grade 5 or 6) with the goal of developing strong literacy skills in both L1 and English. These programs are sometimes referred to as (one-way) dual language programs.

  3. 3.

    Dual language or dual immersion programs involve both students dominant in the minority language and dominant in English. The goal is to enable both groups of students to develop strong oral and literate abilities in both languages. These programs are sometimes referred to as two-way immersion or two-way bilingual programs.

Comparisons of these three bilingual program types with English-only options have consistently shown that ELLs in bilingual programs who are at risk of academic underachievement perform at least as well, and frequently significantly better, compared to similar students in English-only options (e.g., Collier and Thomas 2007). Dual immersion programs consistently yield the best long-term outcomes for Spanish-speaking ELLs. For example, Valentino and Reardon (2015) investigated differences in academic achievement trajectories from grade 2 through middle school among English Learners (ELs) in one monolingual program option (English immersion) and three bilingual options (transitional bilingual, developmental bilingual, and dual immersion) offered by the San Francisco Unified School District. They reported that the English language arts (ELA) test scores of ELs in all bilingual programs grew at least as fast, and sometimes faster, than those in English immersion. Bilingual program options worked particularly well for Latino students (as compared to Chinese-speaking students): “The by-ethnicity results suggest that Latino ELs perform the best in both ELA and math in the long term when they are enrolled in any of the bilingual programs, but especially have the most optimal long-term outcomes in DI [Dual Immersion]” (p. 630). By seventh grade, Latino students in the dual immersion and transitional bilingual programs showed much higher ELA performance than those who had received all of their instruction through English.

The fact that less instruction through the majority language entails no adverse effects for students’ academic development in that language has been attributed to the cross-linguistic interdependence of academic and conceptual knowledge and skills (Cummins 1981a). In other words, although the surface aspects of different languages (e.g., pronunciation, fluency, orthography, etc.) are clearly separate, there is an underlying conceptual and linguistic proficiency that is common across languages. This “common underlying proficiency” makes possible the transfer of knowledge and literacy-related skills across languages.

Despite the extensive evidence of cross-linguistic transfer, particularly between languages with many cognate relationships (e.g., Spanish and English), instructional practice in bilingual and dual language programs has tended to adopt a “two solitudes” model of bilingual proficiency where the two languages are kept rigidly separate from each other (Cummins 2007). These programs implicitly assume that cross-linguistic transfer will happen automatically without explicit teaching for transfer across languages. This instructional assumption is devoid of empirical support and operates to limit the potential effectiveness of bilingual programs (Escamilla et al. 2014).

In contrast to bilingual education models that implicitly adopt a “two solitudes” assumption, the Literacy Squared model developed for emergent bilingual students by Escamilla and colleagues (2014) explicitly sets out to develop literacy in both Spanish and English. Many Spanish-L1 students born in the United States enter kindergarten with varying degrees of proficiency in English and Spanish (e.g., as a result of attending an English-medium preschool program). Thus, a strong and equal instructional emphasis on both languages from the beginning of kindergarten makes more sense according to Escamilla and colleagues than focusing initially on one language to the exclusion of the other. They point out that the paired literacy instruction implemented within this approach “is unique in that it intentionally and purposefully connects Spanish and English literacy environments” (p. 2). Their ongoing research demonstrates not only strong correlations between Spanish and English literacy skills that increase as students go through the grade levels but also “shows the potential of Literacy Squared for developing biliteracy [and] also for accelerating literacy achievement in English for emerging bilingual students” (p 14).

In summary, multiple models of bilingual education that include ELT have demonstrated evidence of effectiveness for ELLs within the United States. In general, these programs show greater evidence of effectiveness with respect to English literacy development than is the case for English-only programs. Programs that provide sustained instruction through the two languages (e.g., dual language programs) and that teach explicitly for cross-linguistic transfer of academic language and literacy skills (e.g., Literacy Squared) are most congruent with validated theoretical constructs (e.g., the interdependence hypothesis) and with the empirical data.

Multilingual and Translanguaging Approaches to ELT

The term translanguaging was originally proposed in the Welsh context by Cen Williams (1996) to refer to the alternation of input and output mode in bilingual instruction. Thus, students may receive information through the medium of one language (e.g., Welsh) and then talk or write about this information through the medium of the other language (e.g., English) (Lewis et al. 2012). García (2009) extended this notion to refer both to the everyday interactional practices of bi-/multilinguals that draw on their full linguistic repertoire and to pedagogical approaches that acknowledge the integrated nature of students’ linguistic repertoire and the cognitive, academic, and personal affordances provided by students’ multilingualism.

Although some aspects of García’s conception of translanguaging have been critiqued, notably her assertion that discrete languages do not exist and thus there is no transfer across languages (e.g., Cummins 2017; MacSwan 2017), her analysis has stimulated a process of systematically documenting existing translanguaging instructional practices and also encouraging educators to explore the pedagogical possibilities opened up by this theoretical construct. With respect to existing pedagogical practice, many case studies (e.g., Cummins and Early 2011) demonstrating the effectiveness of multilingual instructional practices predated the emergence of the construct of translanguaging. For example, DeFazio’s case study of the International High School in LaGuardia Community College, New York City, documented how students “use both English and their native language for all phases of learning and assessment” (1997, p. 103). Chow and Cummins’ (2003) description of the “Dual Language Showcase” documented how multilingual elementary school students could create and digitally publish bilingual books in multiple languages (see http://schools.peelschools.org/1363/DualLanguage/Documents/index.htm). However, current attention to and controversies surrounding the construct of translanguaging have brought these formerly isolated case studies into broader focus and encouraged educators to explore the pedagogical possibilities opened up by students’ multilingual repertoires (e.g., García and Kleyn 2016).

In short, the translanguaging construct has acted as a catalyst for an intense dialogue between instructional practice and theory that has highlighted the legitimacy and benefits of drawing on students’ multilingual repertoires to advance both their learning of English and their overall academic development. The fact that teachers in highly diverse classrooms do not speak the vast majority of the languages of their students is no longer seen as an impediment to implementing instruction that connects with students’ lives and linguistic accomplishments.

Decolonization and Identity Negotiation in ELT

Within the field of applied linguistics, there is a large degree of consensus that the construct of “identity” is of central importance in understanding patterns of language learning and linguistic behavior generally (e.g., Norton 2013). Issues related to identity and societal power relations have also emerged as significant analytic constructs to account for patterns of academic success and failure among students from socially marginalized communities (e.g., Battiste 2013; López-Gopar 2016). Yet, within mainstream educational policy discourse, there is typically minimal focus on either societal power relations or the ways in which these power relations influence patterns of identity negotiation within schools. The focus of educational policy and classroom practice in most countries has been to increase the effectiveness with which national standardized curricula are transmitted to students. This “effectiveness paradigm” focuses on ensuring that students meet universal, one-size-fits-all standards, which are typically assessed by standardized or state-developed tests, all in the ultimate service of greater economic competitiveness.

Researchers have attributed the persistent achievement gaps between social groups to the fact that mainstream curricula and instruction typically ignore fundamental causal factors underlying the underachievement of groups that have been socially marginalized. Specifically, devaluation of identity associated with generations of racism, cultural genocide, and exclusionary colonial structures has operated within schools in the same way as in other societal institutions (Truth and Reconciliation Commission of Canada 2015). If devaluation of identity, fueled by coercive relations of power, operates as a fundamental causal factor within schools, then affirmation of identity must clearly be infused within the structure and operation of schooling in order to reverse this process (Cummins and Early 2011; Cummins et al. 2015; López-Gopar 2016).

Cummins and colleagues (2015) have described a range of pedagogical strategies that challenge the devaluation of identities associated with coercive relations of power. They include connecting instruction to students’ lives, decolonizing curriculum and instruction, and engaging students’ multilingual repertoires. These instructional responses go beyond simply teaching the school language effectively through scaffolding and content-based teaching.

The construct of identity texts illustrates the fusion of identity affirmation and instructional scaffolding as key components of truly effective ELT. For example, scaffolding and identity affirmation are fused when newcomer students carry out creative writing tasks initially in their L1 and then work with various resources (e.g., teachers or other students who speak their L1) to translate this work into the school language. Cummins and Early (2011) described identity texts as follows:

Students invest their identities in the creation of these texts – which can be written, spoken, signed, visual, musical, dramatic, or combinations in multimodal form. The identity text then holds a mirror up to students in which their identities are reflected back in a positive light. When students share identity texts with multiple audiences (peers, teachers, parents, grandparents, sister classes, the media, etc.) they are likely to receive positive feedback and affirmation of self in interaction with these audiences. (p. 3)

Schools that engage in decolonization pedagogical strategies such as identity text work project a radically different image of the student in comparison to more typical schools that adopt a remedial orientation to students characterized as English language learners or disadvantaged. These latter terms implicitly define students by what they lack, and instruction often focuses on remediating presumed linguistic or academic deficits. By contrast, students’ identity texts reflect an image of themselves as intellectually and academically competent, and this transformed identity fuels further academic engagement.

Literacy Engagement as Fuel for English Academic Language Development

There is extensive empirical evidence that print access and literacy engagement represent a powerful determinant of the development of reading comprehension for both native speakers of a language and second language learners (e.g., Krashen 2004; OECD 2004, 2010b). Large-scale data from the OECD’s Programme for International Student Assessment (PISA) showed that “the level of a student’s reading engagement is a better predictor of literacy performance than his or her socioeconomic background, indicating that cultivating a student’s interest in reading can help overcome home disadvantages” (OECD 2004, p. 8). Subsequent PISA findings (OECD 2010b) confirmed these trends. Engagement in reading was assessed through measures of time spent reading various materials, enjoyment of reading, and use of various learning strategies. Across OECD countries, reading engagement was significantly related to reading performance, and approximately one-third of the association between reading performance and students’ socioeconomic background was mediated by reading engagement.

This finding assumes relevance in light of the fact that a large proportion of ELLs in the United States come from lower-income communities with significantly less access to print in their schools and homes than is the case for students from middle-income communities (e.g., Duke 2000; Neuman and Celano 2001). The fact that academic language (e.g., low-frequency grammatical constructions and vocabulary) is found predominantly in printed texts rather than in everyday conversation highlights the importance of ensuring that these students experience a literacy-rich school environment from the day they enter school. Unfortunately, the centrality of literacy engagement has not been incorporated into literacy and second language teaching policies in most North American educational contexts.

Conclusion

ELT in North American schools presents a mixed picture. In all three countries, considerable resources have been assigned to the teaching of English, but there are still significant gaps in the application of empirical research and emerging theory to ELT policies and instructional practices. For example, there is widespread agreement among ELT professionals and researchers about the pedagogical advantages of content-based ELT in comparison to teaching English as a second language in isolation from other academic content. Curricular packages such as CALLA and SIOP with credible evidence of effectiveness have been developed and extensively field-tested. However, across North American schools, English is taught as a separate subject far more frequently than as a vehicle for academic content. Integrated instructional practice where ESL teachers work together with mainstream content teachers is becoming more common at the elementary level but still very much the exception at the secondary level. There is little evidence, for example, that most science or mathematics or history teachers routinely generate language teaching objectives to accompany their content teaching objectives. There is also little evidence that schools have adopted policies to maximize ELLs’ literacy engagement despite extensive research supporting the effectiveness of this strategy.

Similarly, bilingual and dual language programs have consistently demonstrated more positive outcomes for ELLs than English-only programs but still represent only a small fraction of instructional practice in a handful of US states. Translanguaging approaches that attempt to mobilize students’ full multilingual repertoire draw on the same theoretical and empirical foundation as more formal bilingual programs (e.g., positive transfer between L1 and L2), but implementation of these approaches is still in its infancy.

The reluctance of many policy-makers to strongly promote bilingual and multilingual instructional approaches is rooted in unfounded sociopolitical concerns that maintenance of L1 might reduce emergent bilingual students’ motivation to learn English and integrate into the society. The fact that these concerns are totally without empirical support highlights the continuing influence of societal power relations on educational policy and practice. The analysis we have presented in this paper suggests that educational equity and reversal of achievement gaps between social groups will advance only when schools explicitly set out to challenge coercive power relations by implementing instruction that affirms the identities of marginalized group students.

Cross-References