Educating Students Who Do Not Speak the Societal Language: The Social Construction of Language-Learner Categories

In establishing the category long-term English learner, a research report titled Reparable Harm was the most influential. Positioned as a “wake-up call” for educators and policy makers in California about the number of students who still remain classified as English learners after many years of study in California schools, the report called for explicit action by identifying “promising practices” and providing suggestions for district and system level reform (Olsen iii). The label, in particular, provided a “generative metaphor” (Schön) that resonated with legislators and other state stakeholders as a conceptualization of an existing social problem in need of a solution. It was well received in context a in which the public discourse reflected a growing concern about immigration policy, an unease about the racial and ethnic composition of the state (e.g., Hanson), and an actual set of educational challenges.

This specific example of the construction of categories and labels matters because it is a clear example of how coexisting discourses and language ideologies provide a set of cultural rules, conditions, practices, and power relations (Chilton; Goodwin and Duranti; Lindstrom; Van Dijk, “Contextual Knowledge Management” and “Discourse”) that lead to the uncritical acceptance and reification of those categories. More important for this essay, it calls attention to the politics of language in educational contexts in this particularly vulnerable period of time in which economic, political, educational, and theoretical shifts intersect with mass migratory flows. We focus here briefly on two shifts.

The first involves the global education reform movement, termed GERM by Pasi Sahlberg, a Finnish scholar. This movement involves international test comparisons like the Program for International Student Assessment (PISA) and Trends in International Mathematics and Science Study and emphasizes the knowledge economy, competition-based education, standardized testing, and the evaluation of teacher effectiveness. According to Sahlberg, GERM is threatening even those countries, like Finland, where what matters most is good schools for all children. PISA, a single, two-hour examination that evaluates education systems worldwide by testing the knowledge of fifteen-year-old students, is a key part of this global effort at standardized educational reform. Each PISA report and rankings provokes either delight or debates and concerns about the quality of education in participating countries. Not surprisingly, in this competition-based context, immigrant status and language background of students and their effects on standardized scores are being carefully analyzed (e.g., Strong Performers; Christensen and Stanat; Thomson et al.; Knighton, Brochu, and Gluszynski). Among the Organization for Economic Cooperation and Development (OECD) countries, for example, the United States has the sixth largest proportion of students with an immigrant background. But the share of students with an immigrant background explains just four percent of the performance variation between countries. Despite having large proportions of immigrant students, some countries, like Canada, perform above the OECD average. Immigrants, however, and their language proficiency are a source of increasing concern in many nations.

The second shift involves language itself. The terrain has changed rapidly in applied linguistics. Currently, there are many theoretical debates in the field of second language acquisition (SLA). What have been termed “the social turn” (Block) and “the multilingual turn” (May) in that field, for example, have raised the following fundamental questions:

• What needs to be acquired in SLA?
• Should an implicit linguistic system be acquired or a set of structures and forms, or both the system and the set? Or is only the ability to use the second language (L2) effectively important?
• How are second languages acquired?
• Is SLA an individual cognitive process through which individuals move in similar ways? Is it a process of getting the bits and pieces and conforming more and more to a uniform target language (as spoken by idealized native speakers)? Or is SLA “a mediated, social semiotic activity” (Kramsch 97) that results from experience and use?
• What is the end state of SLA?
• Is native-like mastery or complete acquisition of the target language possible? Does SLA result in two full language systems kept separate in use? Does it lead to the development of plurilingualism (Beacco)—that is, of the ability of people to use more than one language in social communication, whatever their command of those languages might be—or of linguistic repertoires that grow and change to meet communicative needs without reaching, as Diane Larsen-Freeman suggests (“Second Language Acquisition”), a native-like endpoint or ultimate attainment?

In the field of SLA and applied linguistics, there is increasing agreement on the following points: SLA is a highly variable and individual process. It is not linear. The highest attainment for most L2 learners does not result in monolingual-like language, even when an L2 is acquired by very young children (Ortega). Teaching may not cause learning (Larsen-Freeman, “Standards”).

Immigrant linguistic-minority students, across the world, must acquire a majority-societal language, whether in a monolingual, bilingual, or multilingual program. In an era of global educational reform, there are increasingly serious consequences, from this pressure, to official language-learner categories. The categories draw from and contribute to the public perception of immigrant students, help or hinder their educational success, and shape the policies that regulate their educational trajectories. Seemingly neutral and commonsense descriptions of student characteristics can have a great effect on the academic lives of youngsters who are sorted in ways that limit their access to opportunities and resources (see Callahan; Xiong and Zhou).

Labels in education, moreover, are problematic. Raymond McDermott warns us that we must be concerned about labels that fundamentally work to keep people in their place and serve as display boards for all the contradictions of school systems and language teaching programs around the world. An established category in a school setting will produce a certain student body. Ours is an era of standardized tests not only in mathematics, reading, and science but also in language proficiency—in particular, the progress of children acquiring the societal language is assessed.

But this assessment is a complicated and difficult endeavor. As Glenn Fulcher and Fred Davidson contend, the practice of language testing “makes an assumption that knowledge, skills and abilities are stable and can be ‘measured’ or ‘assessed.’ It does it in full knowledge that there is error and uncertainty” and strives to make “the extent of the error and uncertainty transparent” (2). In recent years, there has been an increasing concern in the language-testing profession about the degree to which that uncertainty is made transparent to test users at all levels as well as to the general public. Elana Goldberg Shohamy has raised a number of important issues about ethics and fairness of language testing with reference to language policy. Attention has been given to the effect of high-stakes tests, to the uses of language tests for the management of language-related issues in many national settings (Spolsky), and to the special challenges of standards-based testing (Cumming; Hudson). Alister Cumming makes a strong statement about the conceptual foundations of language assessments:

A major dilemma for comprehensive assessments of oracy and literacy are the conceptual foundations on which to base such assessments. On the one hand, each language assessment asserts, at least implicitly, a certain conceptualization of language and of language acquisition by stipulating a normative sequence in which people are expected to gain language proficiency with respect to the content and methods of the test. On the other hand, there is no universally agreed upon theory of language or of language acquisition nor any systematic means of accounting for the great variation in which people need, use, and acquire oral and literate language abilities. (10)

This dilemma notwithstanding, educational systems develop their own sets of standards. These standards, developed as part of a policy-making consensus process, are generally based on the professional perspectives of educators or on the personal experiences and views of other members of standards-writing committees and not on empirical evidence or on SLA theories. Cumming points out that this approach involves a logical circularity, because what learners are expected to learn is defined by the standards, taught or studied in the curriculum, and then assessed “in reference to the standards, as a kind of achievement testing.” He cautions that the applications of such assessments “should not be misinterpreted as evaluations of proficiency or competency generally or by extension to contexts other than the curriculum standards or local educational conditions” (11).

According to Cumming, language proficiency assessments, as currently constructed, tell us very little about a student’s proficiency in a second language. They tell us only where a student scores with reference to the hypothesized sequence of development on which the assessment is based. Such scores are useful in that they allow educators to classify and categorize students and, in theory, to provide them with appropriate instructional support as the students acquire the societal language. Many would argue that in this imperfect world our educational systems are doing the best they can.

Given our growing concern today about classifications and categorizations such as so-called long-term English learners (Olsen), we should examine the politics of language that results in the labeling and categorization of immigrant students, whose number has greatly increased in an era of mass migration. We should consider what Ellen Bialystok, one of the most distinguished researchers on child bilingualism in the world, and Kathleen Peets have to say about categorizations:

Our ordinary conversational means for describing people’s language experience perpetuates a fiction so compelling that we accept the description as a meaningful category. We talk as though being bilingual, or being a language learner, or being literate in a language is an identifiable state with objective criteria and stable characteristics. Our faith in these descriptions as reliable and valid categories extends to education, where such categories are used to classify children and place them in various instructional programs, and to research, where experimental designs are built around the objective of uncovering the unique profile for members of the respective categorical groups. Practically, these approaches are useful and allow educational practice and research inquiry to proceed, producing outcomes that are largely positive. Theoretically, however, the categories are elusive, with individual variation within a category sometimes as great as that between two individuals in different categories. (134)

We are required by existing policy mandates to identify and classify immigrant students as second-language learners. The assumption is that accurate language categorizations can be created and students identified who fit into them. If Bialystok and Peets are correct, however, much harm can come to students if we expect consistent growth and development even though growth and development are known to be highly variable among learners, if we create language ghettos from which students cannot exit and, more important, in which they cannot develop their minds. In times of mass migration and movement of peoples, positive forms of social cohesion should be promoted, diverse groups should be integrated, and peace in the world should be secured. Developing the next generation of minds must be a fundamental goal for all educators.