It is one of the basic findings of decades of research in bilingual education that language submersion is not a productive way to educate minority students. ‘Language submersion’ refers to a situation where students are made to study exclusively through the medium of a language that they have not yet fully mastered; i.e. they are learning a new language AND curriculum content at the same time; usually in the presence of peers who are native speakers of the language of instruction and in the absence of any structured language learning support.
It is beyond doubt that this type of education produces poor results, both in regards to language outcomes and in regards to content learning. In Foundations of Bilingual Education and Bilingualism, Colin Baker (2006, p. 219) lists the negative consequences of submersion education:
Listening to a new language demands high concentration. It is tiring, with a constant pressure to think about the form of the language and less time to think about curriculum content. A child has to take in information from different curriculum areas and learn a language at the same time. Stress, lack of self-confidence, ‘opting-out’, disaffection and alienation may occur.
Indeed, in a famous court case, Lau vs. Nichols, the US Supreme Court ruled in 1974 that submersion education constituted a violation of civil rights:
[…] there is no equality of treatment merely by providing students with the same facilities, textbooks, teachers, and curriculum; for students who do not understand English are effectively foreclosed from any meaningful education.
Research that has demonstrated the negative consequences of submersion education has mostly been conducted at the primary and, to a lesser degree, the secondary level (see Baker 2006, Ch. 10 for an overview). The relative absence of research at the tertiary level is probably due to the fact that most attention has been devoted to students who do not speak the language of instruction at all. However, in higher education a certain proficiency level in the language of instruction is typically an admission requirement. Even so, it is reasonable to assume that students who meet the language-related admission requirements but do so at a relatively low level of proficiency will still be disadvantaged by the combined weight of having to improve their linguistic proficiency and having to learn complex academic content at the same time.
A recent study of the academic progress of pharmacy students in a four-year degree at the University of Otago in Dunedin, New Zealand, demonstrates exactly that (Green 2014). The study analysed the performance data of 297 students entering the program in three consecutive years and examined whether performance in the program could be predicted on the basis of student variables such as English language proficiency, ethnicity and residency status. 265 of these graduated in the end; 28 had to repeat a year; and 105 failed at least one paper at some point during their studies.
It is an admission requirement into the program where the study took place to sit an English diagnostic test. Those who fail may still enter the program but will be required to pass a remedial English paper in the first year. The number of students who were admitted despite failing the English diagnostic test was 48 and these were coded as having ‘weak English.’ The number of students who should be considered English language learners was probably higher but the study did not use further measures of English language proficiency. The data were coded for ethnicity, though (94 students were European/Maori; 186 Asian; and 17 ‘other’), and one might assume that the ‘non-local ethnicity’ students (Asian and ‘other’) included some more language learners even if their English might have been good enough to pass the diagnostic test.
The strongest predictor of success in the program (graduating within four years; not having to repeat a year; not failing a paper) was, unsurprisingly, academic performance on entry. The predictors of low performance (not graduating; having to repeat a year; failing one or more papers) were more complex, and included “having weak English, being of non-local ethnicities, being male, and having lower grades at entry” (Green 2014, p. 5).
In his discussion, the author (who is not a language but a health researcher) sums up the language problem as follows and, implicitly, provides a perfect description of language submersion in higher education:
Within our own university, the students reported on in this study that are identified as having lower English proficiency in the [admission] screening test are enrolled in a remedial programme that they are required to pass. All students are then re-screened in the second year of our programme, but none of those identified in the first year excel, and 77 % of them fail a subsequent screening test in second year, and are then directed to further remedial help and further rescreening. That the students who are initially identified in the first year continue to have academic difficulties, even at the end of the programme, in spite of having to seek remedial help, and being further re-tested suggests two possibilities. The first is that the remedial help is ineffective, but the second, and in my opinion more likely, possibility, is that students who start with weaker English will be improving their English skills over time during the course, but are unable to make up enough ground. (Green 2014, p. 8)
None of this will come as a surprise to anyone who knows anything about bilingual education. What is surprising is that – despite decades of research that show the detrimental effects of submersion education – submersion education in higher education has, paradoxically, come to be widely perceived as the IDEAL method of English language learning and higher education. This supposedly ideal method is, of course, not called ‘submersion education’ but ‘international education’ or ‘global education,’ and includes international students coming to study in English-dominant countries and the proliferation of programs with English as medium-of-instruction around the world.
As Green’s research demonstrates, the price for this misguided belief in the sink-or-swim method is, inter alia, paid in academic performance.
Baker, C. (2006). Foundations of Bilingual Education and Bilingualism. 4th ed. Bristol, Multilingual Matters.
Green, J. (2014). The effect of English proficiency and ethnicity on academic performance and progress Advances in Health Sciences Education DOI: 10.1007/s10459-014-9523-7