Melanie Butler comments on the mixed messages emerging from Madrid’s bilingual programme.
For the last decade the largest language learning experiment in the world has been taking place in the state-run schools in Spain. The laboratory has been the regions which speak standard Castilian Spanish, rather than a local language. The intervention was the introduction of bilingual teaching of academic subjects, usually English and Spanish, in the school system from the first year of primary.
Educational authorities from California to Kazakhstan are looking to bilingual programmes to improve students’ language learning, but the scale of the Spanish experiment and its political support, particularly from the centre-right, is unprecedented in language learning. Bilingual education in Spain has been highly politicised for the last forty years. The centre-left nationalist parties in regions which speak other languages, such as Catalan and Basque, have been pushing bilingual education in two community languages since the country became a democracy in the 1970s. Not all regions have tracked the results, though the Basque Spanish system has produced some of the best research on bilingual education in the world. Much of what can be learned from the Basque region, though, is not replicable in the new programmes run in the Castilian Spanish regions. The Basques are learning two languages used in their community, while the children in the Castilian programmes learn a foreign language, predominately English. The Basques can rely on L1 subject teachers for both languages, while the new programmes use local non-native teachers. Are the new programmes working? There was no randomised control, and research studies have been thin on the ground, but now that the first cohort of students has been through the system the outcomes are beginning to emerge. Different regions have adopted different models and the results released so far are very varied.
In Leon, the English results at age sixteen have been weak: after ten years of being taught 30 per cent of their subjects in English, students only achieve an average level of CEFR A2. The regional education authorities have blamed the problem on the relatively low B2 level of English of the teachers, and promised furious parents that teachers will get a package of training and an army of native-speaker assistants. The importance of fluent teachers should come as no surprise. The bilingual schools start teaching English to children from age six, and we know from Early Language Learning in Europe, a longitudinal study in six European countries, that young children acquire very little language unless the teacher has a language level of C1 or above. This is even more likely in bilingual education, which relies on language acquisition rather than explicit learning: children pick up English through comprehensible input. If the teacher is not fluent, the children will acquire much less. They may also acquire less subject knowledge. One of the many studies recently released by educational authorities in Madrid (see November 2016 Gazette) found that primary children in bilingual schools where teachers had B2 English showed a statistically significant deficit in scientific knowledge compared to their peers in Spanish-medium schools. This was highest for students whose parents had low educational attainment. However, where teachers had C1 English, the effect was considerably diminished. In any case, the problems with science seem to be reversible. Another study suggests that the knowledge deficits disappear at secondary school. Indeed, test results show that bilingual schools’ academic results at age eighteen mirrored the regional average. What do the results in English show? They start off brilliant. At nine years old, 60 per cent of children attain an A2 level of English on tests of speaking and listening, with only around 6 per cent scoring below A1. Three years later on a test of all four skills, 40 per cent hit B1, with another 44 per cent passing at A2. Worryingly, 16 per cent remain at A1 or below. These results caused consternation among the members of Enseñanza Bilingüe, the Spanish association which support bilingual education. ‘Just 84 per cent?’ was their Twitter response when the results were announced by Madrid’s education council. ‘What happened to the rest of them?’ Well they certainly didn’t make it into the ‘bilingual section’ – the prestigious top stream in Madrid’s bilingual secondaries. Entry to these is based on a child’s level of English when they leave primary, which means that either a child has natural aptitude for languages or their parents have enough money to pay for additional private lessons.
The low achieving children join a lower stream, a ‘bilingual programme.’ Judging by the raw data, they learn very little. At sixteen years old, tests show that 14 per cent of students were still below A2, with an astonishing 4 per cent at A0. The left behind don’t appear to have learned anything at all since they were nine. Overall, however, the English level in Madrid’s bilingual schools is good: 57 per cent of sixteen year-olds scores at B1 or above, the kind of results seen among the best-performing countries in the EU, according to a pan-European test run by the EU in 2012. Even more astonishing, 7 per cent passed at C1, a level most native-speaker sixteen year-olds would struggle with. Seven per cent is a familiar number to linguistic statisticians. According to research from Swansea University, in a normal distribution of language ability around 7 per cent show a high linguistic aptitude, exactly the percentage of Madrid children scoring C1. These children are natural linguists, and their aptitude, which is not correlated to IQ, can be easily measured by a simple test. In most northern European countries these students would be passing B2 in English while learning a second foreign language. Seven per cent is also the number of children with language processing problems who will always struggle to learn a second language. In fact, they are typically children who talked late in their first language. The data shows that bilingual schools have fewer children known to have special needs than other schools. But what about special needs that nobody has tested for, like dyslexia? ‘Dyslexia is a failure to acquire reading skills that affects around 5 per cent of children despite adequate intelligence, education and social background,’ according to French neuroscientist Franck Ramus. Dyslexic children have difficulty associating graphemes and phonemes, especially in languages like English which have an opaque spelling system. Only 1 per cent of children will show dyslexia-related reading problems in Spanish, which is a transparent language, but the number who have reading problems will rise when they study English. Even born bilinguals can show dyslexia in one language but not the other; a case study of a bilingual Japanese–English speaker found that, although he had severe dyslexia in English, he had no problem reading Japanese. More problematic for a bilingual schools system is that, in many cases, dyslexics show a phonological deficit, a problem discriminating sounds, which means they will have particular difficulty learning a foreign language, specially English, even with specialist support.
How many of the children who are left behind in Madrid’s bilingual schools are actually undiagnosed dyslexics? And what about the 11 per cent of left-behind children who statistically are not likely to be dyslexics? What information do we have about them? We have no information at all. Although a vast amount of demographic data has been released, it is related only to the science results of children in bilingual schools, not their English results. Meanwhile the Madrid authorities have announced a raft of new measures for schools: increasing the number of bilingual schools; piloting English in pre-schools; even introducing a new English exam at age fourteen. Faced with accusations from the left wing that the programmes discriminate against the disadvantaged, president of the Madrid region Cristina Cifuentes responded by quoting the research on science results. Time to do some research on the English language results, Cristina. There is no doubt that the bilingual schools are doing a great job with 84 per cent of their students, but too many children are being left behind.
Pic courtesy: Daniel Friedman