CHAPTER I INTRODUCTION

CHAPTER I

Introduction 

Bilingualism

Bilingualism is the practice of alternatively speaking two or more languages (Weinreich, 1953). It exists in sizable numbers of individuals in communities throughout the world. However, in its natural habitat it extends beyond the simple acquisition of a second language. It involves the assimilation of a number of psychological, sociological, cognitive and cultural influences (Hamers & Nelde, 1980). This kind of bilingualism, encompassing both linguistic and cultural variables, finds its historical origins (among other factors) in ethnic development and cross-ethnic conflict allowing for the migration of peoples within and across national frontiers (Fishman, 1972; Lewis, 1980). 

The United States experiences this phenomenon in the process of coming into being Different 

peoples of diverse languages and cultures merged as streams of migrants interacted among themselves and established meaningful contact by means of a language of wider communication, English.

However, many have preserved their vernacular languages as a very useful link to their personal history. A developing biculturalism in fact accompanied their bilingualism. Many inmigrants held on, in part, to their vernacular in assumed acknowledgment of underlying heritage values. This prevailed under the influence of an assumed freedom of expression (Fishman, 1972). 

Immigration has been, in fact, the most influential factor promoting widespread bilingualism in the United States (Lewis, 1980). This is revealed in the United States population census figures of 1790. Out of 3.2 million inhabitants of the United States at the time (800,000 non whites), only 61% of the white population were of English origin (cited in Lewis, 1980). 

Non-English Vernaculars

Vernacular languages fell prey to the growing force of a socially cohesive and monolithic economic structure. Incoming groups of foreign stock were perceived as aien to newly formed societal values, thus discriminated against for being culturally different.

Compulsory public education was geared toward the rapid assimilation of incoming immigrants. Part of the “education endeavor” for children born of these families was the rapid removal of whatever idiosyncrasies that appeared dissonant to mainstream standards, including native languages (Sarason & Doris, 1979). This policy, in turn, prevented large numbers of ethnic children from receiving the benefits of public education. Busey [1969] (cited in Sarason & Doris [1979]), referring to mid-nineteenth century data, claimed that while nationwide 82 percent of the native white children attended school, only 46 percent of the children of immigrants attended. Furthermore, large numbers were not attending any school at all (Sarason and Doris, 1979). 

Beyond the school rooms, the agenda for national construction subsumed the concept of assimilation. The dominant culture (Anglo-Saxon) and the majority language (English) were closely linked to the American value system and seen as prerequisites to full citizenship. 

Pressure of political and economic origin and the passage of time influenced heavily the diminished language use of minority language communities (Fishman, 1980). However, large numbers of individuals held on to their linguistic and cultural traditions.

Several factors may have contributed to the strengthening of the vernaculars across the United States. As the nation grew and expanded, it was unable to absorb effectively language minority individuals into its economy. At this juncture, the democratic ideals which had bound together all citizens into a “melting pot” of progress fell into critical questioning. Ethnic consciousness deemed reawakened under the pressure of obligatory economic side streaming of language minorities (Fishman, 1971; Chaika, 1982). 

Hispanics, currently the most numerous of immigrant groups and the largest linguistic minority in the United States (Peñalosa, 1981) have encountered serious difficulties. Large numbers shared and still hold low socioeconomic levels matching other highly discriminated groups in the U.S.A., such as the blacks. Most have been here for more than a century, such as the Mexican-Americans and Puerto Ricans (colonized as a result of the Spanish-American War of 1898).

The Education of Language Minorities in the U.S.

Increasing concern over he role of education among language minority children grew as large numbers were and still are unable to graduate from the nation’s High Schools. School attrition has been higher  for language minority groups than for their monolingual counterparts. 

On the other hand, educators have known for years that language minorit children have difficulty succeeding in English monolingual schools. The U.S. Commission on Civil Rights declared in 1971 that no large-scale effort was ever undertaken to better the educational conditions of no-English speaking children of Mexican extraction (cited in U.S. Commission on Civil Rights, 1975). 

The Commission again conducted a study in 1972 which revealed that in that year the attrition rate for Puerto Ricans in the New York City Public Schools from 10th grade to graduation was 57 percent (cited in U.S. Commission on Civil Rights, 1975).

In 1968, the Bilingual Education Act, product of federal legislative action, provided funds to support bilingual programs. It came to be clearly recognized at federal levels that instruction in the native tongue of the child would help solve the grave problem of educational inequality for these children.

In 1970 the first expression of executive policy in the area of bilingual education was effected when the Department of Education and Welfare (HEW) issued its May 25 Memorandum requiring districts to provide assistance for language minority children. Failure to comply would be considered a violation of Title VI of the Civil Rights Act of 1974. 

In Lau v. Nichols, the Supreme Court affirmed the interpretation of Title VI’s scope declaring that children of Chinese ancestry had a clear educational disadvantage because they were being taught in a language incomprehensible to them, English.

This decision of 1974 in fact changed the whole concept of American public education: the nation’s judicial system intervened on behalf of equal educational opportunity. Public education, until then, meant bringing children up to conform to wholesome mainstream standards This also meant leaving out those who persevered in being different. It now appeared to shift its focus on accepting and tolerating cultural differences under the mandate of law. 

Bilingual education grew and expanded, partially due to a longstanding state of educational inequality for the linguistically different. The Bilingual Act of 1074 was enacted as a culmination of a series of accomplishments and advances on behalf of language minority individuals. It was supported and is now continued (reauthorization Bill of 1984) by the need to make the educational system work for the minority language child, an effort rarely observed in the annals of American public education.

Bilingualism and Bilingual Education

  Bilingual education has been and still is under heavy criticism in the United States. Most troubling is the general negative attitude of the majority population toward bilingualism of any kind. That maintenance of "old country" habits and behaviors (that is bilingualism and biculturalism) is frequently seen as suspicious or un-American  ( Fishman 1980). Other reasons are frequently stated. Expressions about the lack of effectiveness of bilingual education appear frequently in the popular media. Why allocate funds, mostly coming from the majority population, to implement programs which are not improving these children’s education? To the critics, then, it is futile to provide instruction to these children in the vernacular (Bowen, 1977). 

On the other hand, alternative reasons (other than linguistic) are presented to explain the minority child’s delayed progress in school. One is the low socioeconomic level of the parents which weighs on these children’s psychosocial development. Other reasons may fall under the umbrella of socioeconomic deprivation and poor school related program effectiveness.

However, no fully convincing body of experimental work has been put forth at this time to overcome contradictions and steady the course of bilingualism and bilingual education in the United States.

Statement of the Problem

School Bilingualism

Hispanic school children, as well as other language minority children in large urban cities is the United States, have long been influence by the controversial issue of bilingualism. That these children need to attain well rounded skills in English has been generally agreed upon by everyone truly concerned with the education of language minorities.

However, what has remained in the limelight of controversy is whether the language they bring with them to school, their vernacular, is to be kept in any way alive or allowed to disappear as rapidly as possible.

This longstanding issue seems to be centered on the academic usefulness of the vernacular. Opponents of bilingualism see it as a burden and a hindrance to acculturation and mainstreaming. On the other hand, those who support bilingualism maintain that is in fact and aid highly instrumental to the child’s speedy transition into English.

Research in this particular field has experienced considerable growth during the present decade. However, significant studies which specifically address the Spanish languge contribution (or lack of it) to English language development on urban bilingual children are not available in the current literature.

Cummins’  Hypothesis

Professor James Cummins, of the Ontario Institute for studies in Education 9Canada) and one of the most active supporters of bilingual education, has developed a comprehensive bilingual developmental interdependent hypothesis.

Cummins’  (1979) hypothesis states that “the level of L2 competence that a bilingual child attains is partially a function of the type of competence the child has developed in L1 at the time when intensive exposure to L2 begins” (p. 23). If a language minority child has failed to develop strong vocabulary concepts in the vernacular, it would be highly unlikely that this child would acquire competence levels of proficiency in the second language during his schooling years. Furthermore, his home language may eventually suffer further loss without adequate school support.

Empirical support for this hypothesis comes from correlational studies, studies of the optimal 

age question in L2 acquisition, and immersion program evaluations. This latter set of studies demonstrates the successful transfer of language skills across languages (Cummins, 1980).

Cummins also attempts to integrate linguistic and non linguistic factors in order to provide a better understanding of the problems currently confronting language minority children. He supports the idea that academic language deficit (deficit understood in terms of failure to compete with monolinguals in linguistic academic tasks) by itself does not fully explain a bilingual child's lack of adequate progress in school. 

Academic achievement, language attitudes, socioeconomic level, and teaching methods may also interact and explain the language minority child’s success or failure in school and in society. First language skills, however, seem crucial to competent school progress given the present conditions influencing language minority children (Cummins, 1979). 

Cummins’ hypothesis may help predict the negative cognitive as well as affective consequences of exposure to school in a second language. This may occur when language minority populations are being taught in a developmentally recent second language at a time when their first language has not matured up to academic readiness levels.

Cummins lends support to the idea that the first language is part and parcel of the child’s well-rounded development (as language belongs to the mind-ambience context of a fully developing individual). Its dynamic nature demands completion and maturity (Anisfeld, 1984). When a child’s first system stops growing because negative values have been attached to it and further exposure has been arrested, a sense of frustration may be felt by the child, hindering second language development (Saville-Troike, 1975).

Second language learning not only involves the task of applying-verbal skills toa new language system, but also the mental and psychological effort of adjusting to a highly complex context of societal realities. This introduction to a new social system, very much linked to second language acquisition, is not engaged fully  by the child until he or she enters a classroom. The double burden of acculturation and language learning is no small endeavor. It becomes ever more difficult when experienced at school, a totally new experience for a language minority child. A child’s fully developing first language system seems to be a springboard from which he or she may move on to other tasks (academic and social), part of which is second language learning. 

In conjunction with his “linguistic developmental interdependence hypothesis” Cummins (1979) has also formulated his threshold hypothesis. It proposes that there must be minimal level of linguistic development in L1 before the child can attain competitive level of achievement in school . Academic language development is seen in the context of mainstream language proficiency. A language minority child may have developed good second language skills to function adequately among his or her bilingual peers. Yet, this may not be sufficient for true academic progress in comparison to the wider domain of native speakers of English, that is, the majority population.

Furthermore, when addressing the linguistic characteristics of language minority children, Cummins (1980bc) makes a distinction between communicative proficiency (accent, fluency), which he gives the acronym of BICS (basic interpersonal communicative skills) and cognitive-academic proficiency, which he refers to as CALP (cognitive/academic language proficiency). The former (BICS) is contextualized or context embedded language. The speakers can actively exchange meaning in the context of paralinguistic and situational transactions (Cummins, 1984). 

In contrast, CALP depends mostly on linguistic cues to express meaning in logical terms (Cummins, (1984). Language proficiency tests con be categorized as measures of CALP  and may serve as good descriptors of what cognitive academic language proficiency means.

The use of this terminology to distinguish these two aspects of language has been the subject of recent debate (Rivera, 1984). Criticisms of Cummins’ linguistic distinctions have basically addressed two issues. One is Cummins’s assertion that academic cognitive/linguistic  lags may explain the bilingual students’ delay in academic achievement. Troike (1984) and Wald (1984), two of his major critics, maintain that sociocultural factors as well as cultural attitudes, expectations and teacher behaviors are at the crux of the problem. Cognitive-linguistic phenomena may only be indicators of the level of acculturation attained by these students; CALP is only a symptom, not a cause of poor academic achievement.

The second issue questions Cummins’ understanding of BICS and CALP as separate and independent proficiencies. Canale (1984) describes it as a “dichotomous characterization of communication” (p. 31) claiming that language must always exist in some context in order for it to be meaningful. Cummins’ (1984) hast clarified this issue by positioning BICS and CALP as a single proficiency, each describing separate points in a continuum. The former, BICS, describes the most context embedded and least cognitively demandingproficiency and the latter, CALP, the most context reduced and cognitively demanding proficiency.

Previous Research

Abundant research evidence has accumulated during the last seven years in support of  Cummins’ (1979) developmental interdependence hypothesis. A large proportion of it has taken place either in foreign setting (Canada and Sweden) Czico, 1976; Swain et al. , 1976; Skutnabb-Kangas and Toukomaa, 1976). or in the western and norther part of the continent,with immigrant Mexican (Modiano, 1968) and Hispanic children (Walters, 1979; Sancho, 1980;  Hadis, 1984; Guerra, 1984) and indigenous populations (Leslie, 1977).

Research encompassing this specific area of Cummins’ hypothesis addressing Hispanic children and adults has followed along similar methodological lines as that previously mentioned. These studies have been limited in scope by basically considering correlations among linguistic variables. The longitudinal aspect of language development measured by observations of a single sample of bilingual children throughout the span of several years has not been the subject of previous research about Cummins’ linguistic interdependence hypothesis. 

The present research endeavor, in turn, will focus on the collection of correlated measures pertaining to students’s test scores of first and second languages throughout a period of three consecutive years. This will allow for observations and analysis of a wider spectrum of language variables within and across time parameters as well as within and across languages. An additional statistical procedure, not observed by previous research on Cummins’s hypothesis, will be implemented by the inclusion of the subjects’ intelligence quotients (IQ). These scores will be covaried with second language scores in order to determine whether our subjects’ intelligence scores contribute significantly to second language development.

Research on Hispanics

Four linguistic interdependence studies involving Hispanic subjects are reported in the literature (Walters, 1979, Sancho, 1980, Hadis, 1984, [cited in Gingras, 1984]). All four provide support for Cummins’s hypothesis. However, these studies apply the interdependence hypothesis to address diverse aspects of the problem. 

The first study by Walters (1979) is limited to the pragmatic domain of language ability. It shows that a relationship exists between bilingual children’s two languages and that this relationship influences performance on certain tasks in the second language. However, it fails to address the cognitive-linguistic aspects of language development and its longitudinal dimension. Cummins’ hypothesis, on the other hand, underscores cognitive academic language proficiency (CALP) in the context of long term bilingual education (Cummins, 1980c).

The second monograph, by Sancho, compared the effects of two fundamentally different approaches to bilingual education, maintenance and transitional instruction in San Antonio, Texas. Analysis of the results indicated that the type of program had less influence on student achievement than did the degree of linguistic competence which the bilingual child initially brought to the school setting. A direct relationship was found between academic performance and initial proficiency in both languages. This supports the hypothesis that the development and maintenance of two languages in the classroom increases the ability of bilingual children to perform logical operations such as those required in mathematics.

However no observations where recorded on the developmental interdependence influences of both languages upon each other such that one, the native language, may be predictive of second-language development. Hadis (1984), on a similar study, fails to include treatment of the academic language performance of the Hispanic high school students who served as subjects. Furthermore, no data on elementary school children’s language performance was collected and analysed.

The fourth and most recent study of Cummins’  linguistic interdependence hypothesis by Guerra [1984] (cited in Gingras, 1984), analyzed the effect of three native language related variables and two English language learning related variables on six measures of error recognition and correction (error judgements) of the second language.

A sample of 182 Hispanic adults was randomly drawn from the total population of Hispanic students at Houston Community College. Guerra found that academic native language skills (Spanish) appeared to be predictive of the students’ ability to comprehend and recognize errors in the second language as well as t he ability to correct local errors in the second language when examples of syntactic errors were shown to these students.

Although this study provide evidence of Cummins’ interdependence hypothesis, it limits the sample to Hispanic adults and focuses solely on the subject’s ability to recognize and correct syntactic errors in both languages. Academic language proficiency for either children or adults includes other aspects of language, such as vocabulary knowledge and comprehension.

The above four research projects regarding Hispanics, in sharp contrast with our contemplated study, focused wither on linguistic proficiencies exhibited by adult Hispanics (Hadis and Guerra), on the one hand, or measured the communicative competence of children, in contrast to academic proficiency (Walters). Sancho’s study, although focusing on primary school children, observed the subjects’ initial proficiencies in the first and second language with respect to academic performance, failing to compare first and second language intercorrelations over time (or longitudinally).

The long term effects of first language proficiency on second language ability  were no subjected to analysis. Furthermore, the influence that intelligence may have on the subjects’ ability to learn the second language is not known from any of the studies reviewed here.

Research on Finnish Students

Skutnabb-Kangas and Toukomaa (1976) provided strong support for Cummins’s hypothesis, precisely by showing that Finnish immigrant children (mostly of indigent families) reached competence levels in Swedish, in direct correspondence with the levels of proficiency in their vernacular. In other words, children who know Finnish best also learned Swedish best. And those who did not attain progress in Finnish (the vernacular) only accomplished slow and poor progress in Swedish. Moreover, the authors observed vernacular loss by those children who followed the Swedish curriculum alone: their Finnish remained static, or stagnant. Ethnic values appeared to have been negatively structured into the curriculum, to these children’s disadvantage.

Skutnabb-Kangas’ (1979) research drew data from 700 Finnish immigrant children in two communities in Sweden. Picture vocabulary test scores were collected by her colleague Pertti Toukomaa (1979). Poor language productions on both languages were reported below the 30th percentile rank on picture vocabulary tests administered in Finnish and Swedish; however, Finnish  immigrant children exposed to a Swedish language medium of instruction exhibited a diminishing level of proficiency in Swedish across three consecutive years (first, second and third graders) on a cross-sectional study (independent groups). 

To further her hypothesis, Skutnabb-Kangas provided data (see table 1) on language proficiency scores (listening comprehension in Swedish) of Finnish children schooled in their native country (Finland). Strikingly enough, children schooled in Finland performed better in Swedish listening comprehension tests than Finnish children who never went to school in Finland. To advance her point, the author presented evidence of significant improvement for Finnish children who underwent schooling for more than three years in Finland. This author also demonstrated improvement in Swedish comprehension tests for those children undergoing bilingual education. 

TABLE 1

Listening comprehension (Swedish) and the Language of School Entry

	GROUP I	GROUP II
TEST MARK %	STARTED SCHOOL IN SWEDEN	STARTED SCHOOL IN FINLAND
		School years in Finland before emigration
	SWEDISH           FINNISH   CLASSES            CLASSES	1-2 YEARS      3 + YEARS

1-2 (-)       12	4	14             12
3             50	33	17            12
4-5 (+)      38	63	69          76
100%	100%	100%        100%
N   =  82	49	29            17

NOTE: From Language in the process of Cultural Assimilation and S
tructural incorporation of linguistic Minorities, (p.16), by Skutnabb-Kangas, T. 
(1979) Interamerica Research Associates, Inc


     The author, however, fails to indicate statically whether there are significant differences among the four groups of Finnish children (Table 1). A non parametric analysis of variance seemed to be the appropriate statistical operation. Two factors (independent variables) are represented on the frequency table described above: Finnish academic proficiency by virtue of language of school entry with four levels; and Swedish comprehension with three levels. 

Research of Amerindians

Leslie (1977), likewise, discovered a high correlation between childdren’s oral Cree competence and English reading skills in a study conducted with grade one and two students on the Hobbema Cree Indian Reserve in Alberta. Similar findings were reported by Modiano (1968) with mexican Indian children. Those who were first taught to read in the vernacular did better in English than those who were taught to read only in English. 

The common pattern revealed by these researchers is that first and second language appear to be correlated such that command of the native language at the onset of schooling may have significantly contributed to the acquisition of the second language throughout the education process.

Dissertation Rationale

This particular research work will attempt to discover, in fact, whether developmental interdependence and, consequently, additive bilingualism is observable among Hispanic children 
currently undergoing schooling in large urban centers on the eastern part of the mainland, specifically, in Newark, New Jersey. In contrast to research reviewed previously, Cummins’  hypothesis will be investigated by observing first and second language development in primary school bilingual children on a longitudinal basis; that is, the collection and analysis of language scores of a single group of children for three consecutive years.

Secondly, and sharply differing from previous studies, it will be designed to show statistically whether intelligence is to be held accountable in significant levels for the improved second language proficiencies of  these children in place of first language skills. Thirdly, it will perform a factorial analysis of first and second language proficiency across the span of three years, encompassing within language and across language interactions in concomitance with several presumed linguistic-related variables, thus presenting a wider picture of linguistic development as influenced by ethnolinguistic, socioeconomic and demographic characteristics.

And, fourthly, sampling criteria will provide for the observation of a wide repertoire of first and second language skills currently exhibited by children undergoing bilingual education from the onset of schooling (first grade) thus allowing for a comparative longitudinal analysis of children with lower and higher proficiencies.

Results of this study may most clearly indicate whether the development of first language skills by these children may be useful if second language skills are to be effectively fostered. Bilingual education aimed at developing first language skills will then be a most clearly convincing need.


In addition to better language performance, stronger cognitive-conceptual abilities may be developed by these children. Competence levels of English proficiency will consequently allow no delay in the general education of these children. Skutnabb-Kangas and Toukomaa (1976) observed that achievement in subject matter which required abstract modes of thought was appreciably improved when migrant children exhibited more sophisticated knowledge of their mother tongue. In addition, these children had increased probabilities of competing with native speakers of Swedish in the job market There, skillful use of cognitive and language abilities were requires. Thus, their chances of fully joining the social and economic structure of the receiving country were improved (Skutnabb-Kangas, 1979).

     Hispanics who have lived and grown up in large urban areas, such as Newark and New York City, are currently evidencing poor chances of structural incorporations into the economic system. Large numbers have not developed deep level academic, cognitive, and language skills to compete on an equal level with white Americans.

Drop-out levels continue to be as high as ever (Aspira of New York, 1983) for Hispanics of the New York City public school system. This may remain so until educators of the majority population realize that maintenance bilingual education is not to be seen solely as a legal issue, but as a pressing alternative in need of serious research. Only then will it become a truly effective instrument of educational and socioeconomic opportunity for disadvantaged language minority children.