Bilingual children in the crossfire

Language, power, and pedagogy: Bilingual children in the crossfire

Petrovic, John E

In the past few years, bilingual education has faced arguably the most strident attacks and setbacks in its long history in the United States. These attacks have come in a most insidious form: the ballot initiative. Armed with a dearth of data but plenty of duplicity and sound bites, people like Ron Unz, Linda Chavez, and Christine Rossell have funded or supported initiatives to severely curtail bilingual education in public schools for all children and to end it for those who need it most.1

The arguments against bilingual education have become as repetitive as the indisputable evidence that debunks them. There is the claim that “my ancestors came here and had no bilingual education”; but the records have shown that, historically, millions of immigrants did have bilingual education (Kloss, 1998). There is the claim that the United States is going to fall apart, to “balkanize”; but there is little chance or evidence of that (Petrovic, 1997). There is the claim that the high Hispanic drop-out rate demonstrates what a bad idea bilingual education is when in fact there is a demonstrated positive correlation between bilingual education and staying in school (Curiel, Rosenthal, and Richek, 1986). Readers of this journal will be well aware of the many other arguments against bilingual education.

In the midst of this bilingual turmoil, Jim Cummins’ Language, power, and pedagogy: Bilingual children in the crossfire provides an articulate, reasoned, and thoroughly supported argument as to the importance of using children’s first languages in schooling. Much of the content will not be new to those familiar with Cummins’ more than two decades of invaluable research. Here he pulls it all together in a passionate defense of bilingual education. The book is divided into the three major sections examining language and educational theory on both macro and micro levels in various contexts in the United States and around the world.

In the first section of the book, Cummins presents language theory as grounded in context and in dialog with research, policy, and practice. Emphasizing that his theoretical constructs emerge from educational practice, he admits that other interpretations may be possible and consistent with the empirical data, and that these data emerge from highly politicized environments. He offers strong empirical evidence to support a unified theoretical perspective focusing on the power of positive, interpersonal educator-student relationships to challenge coercive power structures, to help students negotiate positive identity, and to promote success in the school context. Drawing from cognitive psychology, applied linguistics, and educational research, he contends, “bilingualism is associated with enhanced linguistic, cognitive, and academic development when both languages are allowed to develop,” (p. 4) and he calls for teacher education programs and school systems to implement programs responsive to these findings.

In the second section of the book, Cummins develops the idea that conceptualization and assessment of language proficiency derive from power relations in society and that they have far-reaching effects on student opportunities in that society. Students need to acquire specialized academic literacy and to do well on assessment measures. Cummins reviews debate over the nature of first or second language proficiency and arguments about appropriate assessment of proficiency.

In regard to proficiency, Cummins revisits the distinction he has defended between conversational and academic language proficiencies. Readers of the Bilingual Research Journal will recognize these as the two constructs that he popularized: Basic Interpersonal Communicative Skills (BICS) and Cognitive Academic Language Proficiency (CALP). Cummins explains that the original purpose of the constructs was “to warn against premature exit of ELL students (in the United States) from bilingual to mainstream English-only programs on the basis of attainment of surface level fluency in English” (p. 58). Here he provides an interesting review of theoretical constructs that he has borrowed (those of Lev Vygotsky, Jerome Bruner, and Catherine Snow, among others) to develop his BICS/CALP distinction, including its later elaboration into the four quadrants model (along the intersecting continua of context embedded/ reduced and cognitively demanding/undemanding).

Cummins also takes up recent criticism of his BICS/CALP distinction by asserting that it is neither a “deficit theory” nor “autonomous” from its sociopolitical situation. In organizing conversational and academic aspects of proficiency in the BICS/CALP schema of cognitive demands in relation to contextual supports, he consistently emphasizes the complexity of language learning. He argues for an expanded conceptualization of proficiency that must include more time for ELL students to achieve on the level of their monolingual classmates. Consistent with broader notions of proficiency, he notes that more attention should be given to adult and child language assessment. Cummins suggests, “language development is characterized by increasing differentiation according to particular contexts and tasks,” (p. 55) and he notes that certain proficiencies or registers are merely different from, not better than, others. He argues for greater access to registers by means of developing literacy and language proficiency in schools.

In another review, Glenn Martinez (2001) criticizes what he considers a “considerable amount of time and energy” spent on defending the BICS/ CALP distinction. He claims the criticisms against Cummins’ theoretical model “have been in some cases unfair and in other cases misguided.” The fact that Cummins does spend an entire chapter responding to critics indicates that they cannot be as easily dismissed as Martinez would have us believe. Carole Edelsky, Terry Wiley, and Jeff MacSwan are all scholars to be reckoned with and their critiques have forced Cummins to rethink or at least clarify his positions and terminology. Indeed, it would seem that the influence of MacSwan’s (2000) critique is evident throughout the text.

MacSwan’s basic quarrel is with the widely discredited notion of semilingualism that, he argues, is perpetuated in Cummins’ theories. In Language, power, and pedagogy, Cummins uses terms like “proficient,” “register,” and “function” as if they were interchangeable. At one point, he uses all three of them noting “bilingual students in some contexts have failed to gain access in either LI or L2 to the kinds of language registers (or proficiencies or functions) that are required to participate effectively in schooling” (p. 100, emphasis added). He also makes these things synonymous with CALF. It seems that Cummins wants to make little or no distinction between these terms. The problem is that they are not synonymous and, in fact, refer to different phenomena. Here Cummins gives an essentially correct take on MacSwan’s position: “As far as I can judge, for MacSwan, the threshold hypothesis crossed the boundary into a deficit position because it talked in terms of levels of language proficiency rather than in terms of access to academic registers” (p. 106). This is because MacSwan, a linguist, recognizes that children are fully proficient in their first languages by the time they get to school and possess a grammar that is essentially indistinguishable from that of an adult member of their community. They may, however, still lack the registers of language or language repertoire required for the peculiar purposes of schooling.

Cummins acknowledges several times that children are proficient in their first languages by the time they come to school. This in itself indicates the evolving nature of Cummins’ ideas, and the debt he owes his critics, since in the 1970s he was drawing from research supporting the imprecise idea that “some bilingual children who have been exposed to both languages in an unsystematic way prior to school, come to school with less than native-like command of the vocabulary and syntactic structures of both LI and L2” (Cummins, 1979, p. 238). Nevertheless, he continues throughout the book to equate fluency or proficiency with CALP, unwilling to concede too much to this very basic linguistic fact. On two different occasions he offers the comparison of a five year old and Toni Morrison: “The argument that language development is largely complete by age five essentially claims that Nobel Literature laureate Toni Morrison has no more language proficiency than a five-year-old child” (p. 108). The problem for Cummins here is that this statement is linguistically true. To the extent that this can be read as blaming the five year old for not having the literary repertoire that schooling has yet to give her, one could indeed interpret it as a deficit model. We could take Cummins’ comparison a step further and compare a fifty-year-old monolingual Mexican peasant woman and Toni Morrison. Could we possibly claim that this peasant woman is not proficient in her language? And could we possibly make such a claim because she did not hold the same linguistic repertoire as Toni Morrison? To do so would be elitist at best. Despite his explicit denial, Cummins’ unfortunate example adds to the view that he holds CALP in a superior, hierarchical position to other language registers or that he views literacy as a stage of linguistic development (which it is not).

One might also argue that Cummins’ position on testing for CALP supports the interpretation that he views CALP as a stage of linguistic development. He defends discrete point tests as a way to identify a student’s level of CALP. He also claims that these tests could be used to measure LI CALP. Since the construct came about as a way to explain ELL students’ struggles upon L2 mainstreaming and to avoid premature mainstreaming, assessment of L2 CALP is defensible. The only purpose for measuring L1 CALP, however, would be to make the argument that bilingual students are somehow not fully proficient in their L1, that they are “semilingual” (as per Cummins, 1979b, the definition of semilingual still seems to be his position given his continued defense of the threshold hypothesis). Note that native English speakers would never have to take such a test in their L I and, therefore, could never be judged to be semilingual. With this, Cummins may inadvertently be adding fuel to the critical fire by promoting a deficit model.

In the third section of the book, Cummins points out that despite increasing numbers of bilingual education programs around the world and increasing amounts of research data, bilingual education remains controversial in the United States because neither policy-makers nor the general public understand its theoretical bases. This is partly because even discourse about bilingual education reflects power relations in a society. In these constricted frameworks, discussion may be limited to surface issues such as comparing transitional bilingual education to structured English immersion while ignoring more important dimensions such as whether a program is “enrichment-oriented” or “remedial-oriented”

Cummins categorizes current transitional bilingual education programs in the United States under Ruiz’s (1984, 1994) “language-as-problem” paradigm that associates societal problems with linguistic diversity. Ruiz contends that neither the “language-as-problem” nor the “language-as-right” orientation will resolve problems because divisiveness inherent in these orientations make them inadequate for long-term language planning. The “language as resource” paradigm offers the most benefits to all in a society.

The language as a resource paradigm cannot take hold, however, until we carefully re-examine the role of theory in guiding research, informing policy, and implementing practice. Presently, Cummins astutely observes, research is directly tied to policy making instead of informing hypotheses with broad explanatory and predictive power. Under his research-theory-policy paradigm, answers to questions fundamental to policy making become strikingly simplistic. For example, to the question of whether or not increased instructional exposure to English results in greater English achievement (the intuitively appealing time-on-task argument used so effectively by bilingual opponents), the answer is simply “no.”

Given the empirically supported theories behind bilingual education, we must reemphasize positive, collaborative micro-interactions between educators and students that enhance “both language generation (learning) and identity formation” (Cummins, 2000, p. 172). These are the building blocks of the “transformative pedagogy” needed to empower ELL students to engage critically with content matter in school, to extrapolate lessons to the larger society, and to question (and perhaps transform) coercive power relations in both contexts.

Cummins’ “transformative pedagogy” seems to differ in no substantive way from the critical theory and pedagogy defended by others (McClaren, 1998; Shor, 1980, 1996; Wink, 2000). The new label is superfluous and forces Cummins to review critical pedagogy in its broadest sense. The result is a reduced focus on language issues and their relationship to critical pedagogy, leaving a number of questions to be dealt with here more thoroughly. For example, on the one hand, Cummins denies that academic language is inherently superior to conversational language. However, he defines academic proficiency as characterized by such things as “low frequency vocabulary,” “complex grammatical structures,” and greater knowledge of “language itself’ (Cummins, 2000, pp. 35-36). But as MacSwan (2000) correctly argues, school or academic language is a distinctly middle and upper middle class language register. Thus, one could argue that Cummins’ position is contradictory-serving the perpetuation of linguistic elitism.

But, in Cummins’ defense, it is all too easy for middle class, English speaking, Anglo radicals-among whom we include ourselves-to take this tack. While it is essentially correct, it holds little purpose or value for language minority students or students who speak stigmatized varieties of English. It is here that we must heed the arguments of Lisa Delpit, whom Cummins also invokes to escape his seeming conundrum. It is here that transformative pedagogy is important to the extent that it emphasizes both acquisition of English and promotes critical analysis of the power arrangements that “standard English” symbolizes.

Nevertheless, transformative pedagogy certainly does not go far enough-at least as Cummins presents it-to challenge the linguistic status quo that the construct of CALP helps to perpetuate. His position is that variation in command of academic registers exists both within and between groups. Differences between groups are the result of coercive relations of power operating in both school and society. His answer is for all students, regardless of language background, to accept and adopt “academic language proficiency” as it currently exists. Nowhere in his transformative pedagogy is academic language proficiency challenged. In this way, CALP becomes a tool of prescriptivism. As far as we can tell, Cummins’ pedagogy is transformative only in that it demands respect for other languages and language varieties for the overly utilitarian purpose of acquiring CALP. His transformative pedagogy pays little attention to what CALP consists of or who decides that. In other words, Cummins’ present formulation of transformative pedagogy does not controvert the implication that a Shakespearean soliloquy is more CALP– laden, and therefore more valuable, than a Snoop Dog rap. This may not point to an inherent weakness of transformative pedagogy but to the need for a far more radical expression of it, its purposes, and its expectations.

Cummins’ latest contribution should be seen as a must read for those interested in the areas of bilingual and language minority education, language policy and politics, and second language acquisition. At points, however, it does require substantial background reading or at least review for critical appraisal. Here we look forward to continuing debate of the specific criticisms. Given our own limited additions to that debate, we should make clear our belief that this volume is exceedingly important for its many direct links to practice and policy and for its unabashed commitment to bilingualism and bilingual education. Cummins opens the book in chapter one in a refreshing way, approaching it in several journal-like entries of trips, conversations, and conferences that foreshadow nearly every important topic covered, some in great depth, later in the text. As pointed out above, the concluding chapter, while convincingly argued, left us feeling that Cummins somehow stopped short of where his transformative pedagogy can take us. Perhaps that is a strength of the book that we need to discover. But the greatest contributions of Language, power, and pedagogy are a thorough review of the literature in a number of areas and Cummins’ clear, concise, and definitive answers to policy questions. The long anticipated National Research Council report by Diane August and Kenji Hakuta was disappointingly and unnecessarily gray on many such questions. Here Cummins sets the record quite straight.


1To be precise, these initiatives have not brought about the legal end of bilingual education for those who need it most, but certainly they have brought the practical end. One of the provisions of the initiative is that English language learners ten years or younger may not participate in bilingual education. This age range includes a huge number, perhaps the majority, of English language learners. The fact that parents still have the right to claim a waiver if they document the “psychological need” their child has is insulting at best and draconian at worst.


Cummins, J. (1979). Linguistic interdependence and the educational development of bilingual children. Review of educational research, 49 (2),222-251.

Cummins, J. (1979b). Cognitive/academic language proficiency, linguistic interdependence, the optimum age question and some other matters. Working papers on bilingualism, 19, pp. 121-29.

Curiel, H., Rosenthal, J., and Richek, H. (1986). Impacts ofbilingual education on secondary school grades, attendance, retentions and drop-outs. Hispanic Journal of Behavioral Sciences, 8 (4), 357-367.

Kloss, H. (1998). The American bilingual tradition (2nd ed.). Washington, DC: ERIC Clearinghouse on Languages and Linguistics.

Krashen, S. (1999). Condemned without a trial: Bogus arguments against bilingual education. Portsmouth, NH: Heinemann.

MacSwan, J. (2000). The threshold hypothesis, semilingualism, and other contributions to a deficit view of linguistic minorities. Hispanic Journal of Behavioral Sciences, 22 (1), 3-45.

Martinez, G. (2001). Review of Language, power and pedagogy: Bilingual children in the crossfire. Retrieved October 23, 2001, from http://

McClaren, P. (1998). Life in schools: An introduction to critical pedagogy in the foundations of education (3rd ed.). New York: Longman.

Petrovic, J. E. (1997). Balkanization, bilingualism, and comparisons of language situations at home and abroad. Bilingual Research Journal, 21 (2 & 3),233-254.

Ruiz, R. (1984). Orientations in language planning. NABEjournal, 8(2), 15-34.

Ruiz, R. (1994). Language policy and planning in the United States. Annual Review ofApplied Linguistics 14, 111-125.

Shor, I. (1980). Critical teaching and everyday life. Boston: South End Press.

Shor, I. (1996). When students have power: Negotiating authority in a critical pedagogy. Chicago: University of Chicago Press.

Wink, J. (2000). Critical pedagogy: Notes from the real world (2nd ed.). New York: Longman.

John E. Petrovic and Susan Olmstead

University of Alabama

John E. Petrovic is an assistant professor in Foundations of Education at the University of Alabama, School of Education, Department of Educational Leadership, Policy and Technology Studies, 206 Wilson Hall, Tuscaloosa, AL 35487-0302.

Copyright National Association for Bilingual Education Summer 2001

Provided by ProQuest Information and Learning Company. All rights Reserved