Until recently, much research has documented problems in the instruction of English language learners. There was a clear need for research documenting the need for improvement. When students are presented with conventional curriculum with no modifications, they tend to flounder, become overwhelmed, and mentally tune out or withdraw from active classroom participation (Gersten, 1999; Gersten & Woodward, 1994).

Over 20 years ago, the research of Moll, Estrada, Diaz, and Lopes (1980) poignantly delineated the pain and frustration that English language learners struggling to learn English sometimes feel when taught in all-English settings. Students may fail to understand what the teacher is talking about, and may become frustrated when they have an idea but cannot adequately express their thoughts in English. Moll et al. found that teachers tended to correct pronunciation errors (e.g., seyd for “said”) or interrupt passage reading with attempts to define simple English words (e.g., “surprise”, “guess”), thereby breaking the flow of the story. Moll et al. decried “the deliberate, slow pace of lessons with students in the low reading groups” (p. 305), and the lack of intellectual challenge and conceptual development provided to them.

This focus on the details of accurate English language production makes the students appear less competent than they really are. When Moll et al. (1980) followed the same students into a Spanish reading lesson, they observed that the students, although considered “low ability” by their teacher, were able to answer comprehension questions correctly on grade-level material, to develop and expand on ideas in the stories, and to process more complex text. Further, the students could read texts usually reserved for “high-ability” students.

Yates and Ortiz (1991) found that many teachers view language minority children as simply low-performing native English-speaking children. This tendency has led many to merely adopt a watered-down curriculum, including reading material well below the students’ ability to comprehend. This recurrent problem denies language minority children access to the type of instructional material they need in order to make adequate academic progress. This curriculum mismatch, in all likelihood, is one reason for the extremely low academic performance levels of many English language learners.

According to Fradd (1987), teachers who work with English language learners often tend to use “brief utterances such as ‘What is this?’ or ‘What color is that?’” (p. 146). Students learn to reply in like form, in one- or two-word utterances. Not surprisingly, little curriculum content or social expectation is communicated in this type of verbal exchange. In classroom observations of English language learners, Ramírez (1992) noted the same phenomenon regardless of teachers’ or districts’ philosophy of bilingual education.

For years, program evaluation research attempted to determine which model for educating English language learners produced the highest levels of student academic achievement (Baker & de Kanter, 1983; Danoff, Coles, McLaughlin, & Reynolds, 1977–1978; Gersten & Woodward, 1995; Ramírez, 1992; Willig, 1985). This research, as well as a series of research reviews and meta-analyses (Baker & de Kanter, 1981; Greene, 1998; Rossell & Baker, 1996; Willig, 1985) have often yielded differing conclusions. In a synthesis of almost 20 years of program evaluation research, Cziko (1992) concluded, “it may well be unlikely that this question [of which is the best approach for teaching English-language learners in the United States] will ever be satisfactorily answered regardless of the quantity and quality of additional evaluative research” (p. 15). The report by the National Academy of Sciences (August & Hakuta, 1997) concluded that “for numerous reasons, we see little value in conducting evaluations to determine which type of program is best” (p. 138).

Much of the early educational research on English language learners focused on determining the rate at which English language instruction should be introduced. These evaluation efforts were guided at times by theoretical issues, at other times by political issues involving bilingual education (August & Hakuta, 1998; Crawford, 1995).

The type of bilingual program model and the language of instruction employed, while important, have received far more attention in research and in public debate than the equally critical issue of how ideas and concepts are taught. Recently, however, the research focus has shifted away from searching for the “best” program model toward identifying useful and feasible instructional practices (August, 1999; August & Hakuta, 1998; Berman et al., 1992; Gersten & Baker, 2000; Tikunoff, 1985).

Differing Theories About and Approaches to Second Language Instruction

The goal of building competence in English without unduly frustrating students requires a complex balance between the utilization of the native language and the language to be acquired. In reality, many models of bilingual education exist (Ramírez, 1992; Rossell & Baker, 1996). For the purposes of this discussion, however, we will briefly describe two of the major approaches advocated for educating English language learners and the underlying rationales of each.

Native Language Emphasis.  For Latino students, the most commonly utilized model of bilingual education has a strong native language component (Cummins, 1989; Hakuta & Snow, 1986; Troike, 1981; Wong-Fillmore & Valadez, 1986). We use the term “native language emphasis” to describe this approach.

Troike (1981) cogently presented the conceptual framework for native language emphasis:

  1. People are more likely to learn anything, including English, if they understand what they are being taught.
  2. Students with limited English ability will not fall behind their English-speaking peers if they can keep up with subject matter content through their native language while they are mastering English. (p. 498)

Wong-Fillmore and Valadez (1986) applied Troike’s rationale for a native language emphasis to reading: “It is not possible to read in a language one does not know” (pp. 660–661).

In other words, until students obtain a reasonably good knowledge of English, particularly in conceptually complex areas such as reading/language arts and social studies, instruction should be in the native language. This approach ensures that students are not deprived of the experience of learning core concepts in the school curriculum during the years when their English language vocabulary is limited.

Theorists such as Cummins (1989) assert that once students succeed at comprehending complex academic material in their native language, they will transfer this knowledge to the same subjects taught in English. Although it may seem a belaboring process, it would seem more sensible to teach complex academic content to students in their native language first so that they can understand and discuss challenging material without the added demand of constantly translating or expressing ideas in a second language.

There remains great diversity in opinion and practice as to how rapidly students should be introduced to English language instruction and how long native language instruction should be maintained (August & Hakuta, 1998; Crawford, 1995; Gersten & Woodward, 1994; Ramírez, 1992).

Sheltered English/Structured Immersion: Merging English Language Instruction with Content Learning.
  Contemporary conceptualizations of education for English language learners acknowledge the participation of many monolingual teachers. Newer approaches, often called sheltered instruction (Echevarria, Vogt, & Short, 2000) or structured immersion (Gersten & Woodward, 1985; Genesee, 1984) emphasize the merger of English language instruction with content-area instruction.

The approach is currently used most frequently with Southeast Asian students in the elementary grades, and it is being used increasingly with all types of English language learners, including Latino students, at all grade levels. This is particularly true since the passage of Proposition 227 in California in 1998.

According to contemporary theorists, understanding of English can be obtained through well-designed content area instruction where English is used, but at a level that is constantly modulated—that is, adjusted and adapted so that it is comprehensible (Anderson & Roit, 1996; Chamot & O’Malley, 1989). Teachers attempt to control their classroom vocabulary, avoid use of synonyms and idioms, and use concrete objects, gestures, and visuals such as story maps to enhance student understanding of the essential concepts in academic material.

Teachers using the sheltered English approach do not shy away from teaching age-appropriate concepts such as “migration” to third graders or “peninsula” and “compromise” to sixth or seventh graders. Consciously making instruction highly interactive affords students many opportunities to verbalize their thoughts (even if the grammar or syntax is imperfect), so that they are able to grasp age-appropriate material.

In an articulate plea for the integration of reading with English language development, Anderson and Roit (1996) note: “Spoken language is fleeting and inconsistent over time. Text is stable and does not pass the learner by. It allows one to reread and reconsider that which is to be learned in its original form” (p. 2). Anderson and Roit demonstrate how the “potential reciprocity between learning to read and reading to learn has strong implications for developing oral language in English language learners, even as early as first grade” (p. 1).

Less than a decade ago, there were fierce controversies between proponents of structured immersion/sheltered English (Baker & de Kanter, 1983) and proponents of bilingual approaches that emphasize development in a native language (Wong-Fillmore & Valadez, 1986). In recent years, however, research and thinking has moved away from this dichotomy towards a search for coherent programs that employ an optimal mix of instructional methods. Researchers such as Barrera (1984), Saville-Troike (1982), Anderson and Roit (1996), and Gersten and Jiménez (1994) have stressed consistently that the key problem and issue is not the determination of the exact age or grade level at which to introduce English language instruction, but rather how to merge English language acquisition with academic learning in a fashion that is stimulating and not overly frustrating to students.