Language immersion counter-productive, says Finnish prof

Being a naive progressive, my opposition to the illegal policy of compulsory Catalan language immersion has always been based on appeals to the old liberal idea that we should be free to use whichever language we choose, with the odd reference to “rights” enshrined in international conventions and publicised via absurd junkets like UNESCO’s International Mother Language Day and the struggle waged by Catalan speakers during the dictatorship.

A polemical piece by Tove Skutnabb-Kangas entitled The right to mother tongue medium education – the hot potato in human rights instruments suggests that I (as well as parties like the PP and Ciutadans, and the sane wing of the PSC) might be better off making a pragmatic case–that language immersion is likely to inflict severe long-term economic damage:

Minority children have to become minimally bilingual through their formal education. Bilingual education of all kinds is a very specialized and sensitive area of both research and policymaking. However, detailed knowledge of it is a prerequisite for being able to make sound recommendations. An important complicating issue is that some of the scientifically sound and practically proven principles of how to enable children to become high-level multilinguals with the support of the educational system are in fact counter-intuitive: they go against common sense.

If indigenous or minority children who speak their mother tongue at home, are to become bilingual, and learn the dominant/majority language well, one might, with a common sense approach, imagine that the principles of early start with and maximum exposure to the dominant language would be good ideas, like they are for learning many other things – practice makes perfect.

In fact, sound research shows the opposite: the longer indigenous and minority children in a low-status position have their own language as the main medium of teaching, the better they also become in the dominant language, provided, of course, that they have good teaching in it, preferably given by bilingual teachers, just as the Hague Recommendations on the Educational Rights of National Minorities and the UNESCO Education Position Paper Education in a multilingual world (2003) recommend.

I shall give two examples of recent very large-scale longitudinal and methodologically extremely careful studies from the United States, Ramirez et al. (1991) and Thomas & Collier (1997, 2002; see also other references to them in the bibliography).

The Ramirez et al.’s 1991 study, with 2,352 students, compared three groups of Spanish-speaking minority students. The first group were taught through the medium of English only (but even these students had bilingual teachers and many were taught Spanish as a subject, something that is very unusual in submersion programmes). The second group, early-exit students, had one or two years of Spanish-medium education and were then transferred to English-medium. The third group, late-exit students, had 4-6 years of Spanish-medium education before being transferred to English-medium.

Now the common sense approach would suggest that the ones who started English-medium teaching early and had most exposure to English, the English-only students, would have the best results in English, and in mathematics and in educational achievement in general, and that the late-exit students who started late with English-medium education and consequently had least exposure to English, would do worst in English etc.

In fact the results were exactly the opposite. The late-exit students got the best results, and they were the only ones who had a chance to achieve native levels of English later on, whereas the other two groups were, after an initial boost, falling more and more behind, and were judged as probably never being able to catch up to native English-speaking peers in English or in general school achievement.

The Thomas & Collier study (see bibliography, under both names), the largest longitudinal study in the world on the education of minority students, with altogether over 210,000 students, including in-depth studies in both urban and rural settings in the USA, included full MTM programmes in a minority language, dual-medium or two-way bilingual programmes, where both a minority and majority language (mainly Spanish and English) were used as media of instruction, transitional bilingual education programmes, ESL (English as a second language) programmes, and so-called mainstream (i.e. English-only submersion) programmes. Across all the models, those students who reached the highest levels of both bilingualism and school achievement were the ones where the children’s mother tongue was the main medium of education for the most extended period of time. This length of education in the L1 (language 1, first language), was the strongest predictor of both the children’s competence and gains in L2, English, and of their school achievement. Thomas & Collier state (2002: 7):

the strongest predictor of L2 student achievement is the amount of formal L1 schooling. The more L1 grade-level schooling, the higher L2 achievement.

The length of MTM education was in both Ramirez’ and Thomas & Collier’s studies more important than any other factor (and many were included) in predicting the educational success of bilingual students. It was also much more important than socio-economic status, something extremely vital when reflecting on the socio-economic status discussions and choices in relation to the Roma in most recommendations. The worst results, including high percentages of push-outs, were with students in regular submersion programmes where the students’ mother tongues (L1s) were either not supported at all or where they only had some mother-tongue-asa-subject instruction.

This research helps explain the disgraceful rates of school failure for Spanish mother-tongue speakers in Catalonia–at 42.62%, more than twice those for Catalan speakers–and I think provides a key to understanding why–and this is personal experience–they are frequently incapable of expressing themselves adequately in either Catalan or Spanish.

Useful comparisons can be made between some aspects of the Catalan language regime and pre-democracy policy in South Africa. I think Nkonko M Kamwangamalu hits the nail on the head when in Language policy and mother-tongue education in South Africa: The case for a market-oriented approach, writing of the use of apartheid mother-tongue education as enshrined in the 1953 Bantu Education Act as a means of excluding blacks from public goods and services, he says that

the main problem with mother-tongue education is not whether it is good or bad but rather whether it can empower those to whom it is targeted.

Modernising, republican elements in Catalan nationalism echo the public voice of Italian fascism in the 20s and 30s, interpreting their mission as being one of preventing marginalisation, of providing equal opportunities for all, irrespective of their origins. In fact, I think it is fair to say that language immersion here is a manufactured disaster which serves to exclude, not empower.

Whilst it is understandable that its success is celebrated by ethnic and class warriors who deride Spanish as a language of and for maids and servants, the poor and the vulgar, illiterates and those of little class, it’s difficult for me to understand why it continues to enjoy the support of many who regard themselves as being on the freedom-loving left.

Similar posts


  1. Same story everywhere. Left wing protection of the rights (language and otherwise) of the presumed disposessed becomes a permit, within the fiefdoms of the latter, to operate as the most draconian authoritarians of the right – ethnicity-based dictatorships finding their apologia in post-war socialist liberalism. Can anyone tell me how Québec language policy empowers anybody “liberated” by it in an English-speaking continent? No matter where it takes place, it is purely an exclusionary instrument, not entirely distinguishable from the yellow arm band.

    And now the political profiteering from the issue in Madrid…

  2. The whole article is a piece of crap. I have trilingual children. My kids learnt English through immersion, their English is equal or better than their American peers. Their academic results are above average and they speak, in addition to English, an Asian language and a Latin language.
    The academic failure rate that you mention, has nothing to do with immersion. It has to do with the economic and cultural backgorund of the families that you allude to and the crappy education system in Catalonia and Spain (crappy with and without immersion).
    The only reason why you are against an extensive use of Catalan in Catalonia, is because it is inconvenient for you and those foreigners that would like to live in Catalonia without this annoying thing called Catalans.

    And by the way, which English speaking continent is Charles talking about? In most of the continent, they speak like in Cazorla.

    Ian Llorens
    Massachusetts, USA

  3. Ian,

    Point well taken about the quality of English spoken (or castellano in Cazorla, for that matter). No argument here.

    But I’m not sure what your point is. Is it that regional governments should support unilingualism in a bilingual social context? I don’t think so, judging by your waxing poetic on the prowess of your own children.

    Regardless, the end result of linguistic exclusionism and protectonism is really no different from that of the two centuries that les Québecois spent having all their relations with the outside world mediated by the clergy and certain elites. That’s not Cataluña, by the way, despite the parallelism that many there see between La Belle Province and the Condado. It’s Euskadi.


  4. Ian, if you think the article is a piece of crap then you should really address the points it makes.
    * I’m really glad your children are doing well, but your personal experience doesn’t refute the authors’ statistics-based claims.
    * Defenders of immersion in Catalonia often blame the dreadful results for Spanish speakers wholly or principally on the latter’s general poverty and alleged lack of culture, but Skutnabb-Kangas refutes that in the last para, writing that “The length of MTM education was in both Ramirez’ and Thomas & Collier’s studies more important than any other factor (and many were included) in predicting the educational success of bilingual students. It was also much more important than socio-economic status”. The message seems clear to me: anyone who says linguistic immersion is compatible with action to ameliorate social inequality–in Catalonia that’s everyone on the “left” apart from Ciutadans, and I imagine you don’t regard yourself as being on the “left”–is either a fraudster or a fool.

  5. Trevor,
    I am one of the millions of Catalan children, whose mother tongue is Catalan (even though my mother was not a Catalan born, she decided for some strange reason to speak to me in Catalan since I was born) and that from 1939 to the 80s underwent immersion in Castilian only schools (I was not even allowed to use Catalan with my schoolmates, that’s why I still talk in Castilian with those childhood friends that attended school with me, you know how difficult is to change language with someone you know).

    My personal experience for 15 years in this kind of schools, where I met hundreds of Catalan speaking children in this Castilian immersion environment, is that they (we) did not perform below average, much to the contrary.

    I challenge those psychologist that you quote to make a study of those millions of native Catalan speaking people aged 70 to 25, and correlate their academic results and average income with those children who were raised in Spanish and attended the same Castilian only schools (all schools were Castilian only during those 40 plus years).

    Until this study is published, based on 100s of thousands of cases readily available with statistically significant samples, I continue to proclaim that those studies are a piece of manure and after they are published it will be clear that what they say is a fallacy.

    They can also conduct similar studies with Chinese children (raised in mandarin or other Chinese dialects by their parents) and educated in US schools in English only.

  6. A couple of caveats:
    1) Good quality education is more important than the language used
    2) I am not for pure immersion. I wholeheartedly support trilingual education in Catalonia (Catalan, Spanish and English)
    3) The Catalan education system sucks
    4) For those kids joining the local education late (in second grade or later), a special phasing-in program should be put in place
    5) Speaking the language or languages of the local community without foreign accent is definitely a plus. I come across very often, outplacement services reports, which quote the heavily accented speech of the clients as a reason to fail to find proper employment in USA

  7. It would be great if someone did do the kind of study you suggest, but until different research comes along you’re going to have to accept that the only intellectually respectable position is that linguistic immersion is generally A Bad Thing.

    Carmen Leal writes in today’s La Vanguardia:

    La inmersión consiste en un cambio de lengua hogar/escuela. Desde el primer día todos los conocimientos se transmiten en una lengua que desconoce el niño. Es una buena forma de aprendizaje de una segunda lengua, pero tiene unos riesgos para el desarrollo intelectual del niño si no cumple una serie de condiciones. En 1965 en Montreal, Lambert aplicó la inmersión a unos niños anglohablantes en francés. Aprendían francés y en francés. Fue un éxito, consiguieron mejor rendimiento intelectual y un nuevo idioma. Trasladado el experimento a Estados Unidos con los hispanos para que aprendiesen inglés, los resultados fueron catastróficos. Ni aprendieron inglés ni su razonamiento lógico mejoró. Lambert estableció tres requisitos para el éxito. Primero, alto nivel sociocultural de los padres; segundo, lengua materna del niño que tenga prestigio; tercero, tratamiento pedagógico específico y voluntario. Ninguna de ellas se cumplen en Catalunya. Ni el nivel sociocultural de todos los padres es alto, ni el castellano tiene prestigio en Catalunya, porque no se emplea para las funciones altas de la sociedad: Parlamento, Administración, docencia, etcétera, ni es voluntario porque no hay elección de lengua.

    You say you’d prefer a trilingual system, which may have merits. I’d give parents and/or children the right to spend their share of the educational tax cake on schools as they see fit, and if they want to learn (in) Mandarin and Aranese, then good luck to them, and may the market take the hindmost. In any case, immersion is economically damaging and morally embarrassing and has got to end.

  8. Trevor,
    Thanks for saving me the time to google counter-studies to prove that Ramirez’ and Thomas & Collier’s studies are wrong.

    Let’s compare some of Lambert’s statements to mine:
    Ian: “The academic failure rate that you mention, has nothing to do with immersion. It has to do with the economic and cultural backgorund of the families”
    Lambert: “requisitos para el éxito. Primero, alto nivel sociocultural de los padres”

    Ian: “and the crappy education system in Catalonia and Spain ”
    Lambert: “tercero, tratamiento pedagógico específico”

    I do not buy the second argument. I have asked my 2 year old who has learnt English in his 5 months at daycare whether his learning success is due to the prestige of English, and he was not able to elaborate. It is the same as Artur Mas trying to encourage children to learn foreign languages giving them tax deductions.

    The only point I agree with, although I do not know whether Lambert meant this or not, is that the will of the parents plays a role. If parents do not support it, success can be in danger. If parent do not see the positive side of having his childen fully integrated in the society, it is going to be more difficult.

    Trevor, your conclusion is absolutely out of whack.

    My conclusion is very different:
    If done properly, immersion works. Immersion not only provides a second language, but allows integration in the future. An accentless individual(accentless in both the mother and the acquired language) has a competitive advantages at all levels of society. But nothing comes for free, it may require some additional effort.

    It is clear that what you do not like is Catalan and Catalonia. Immersion is just an excuse. But with your own money, you can do what you want, I assume Gaelic and Frisian may be your languages of choice. But with the public money, we will do what our elected officials decide. This is democracy.

  9. Trevor

    I wouldn’t classify it as ‘bad’ but I’m not sure how well immersion works solely on its own merits. The thirty or thirty-five year history of the popular French immersion program in the province of Ontario has produced very few bilingual people. The point is probably that if the languages are not used at home, then a kid probably learns what he needs to. In Canada’s case, it probably provides a fertile breeding ground for future federal civil servants. And in any case, the mere possibility that it teaches kids that there is another world beyond the typically parochial one presented in most public education systems is plus in itself, regardless of its immediate success, or lack of.

  10. Sorry, I shouldn’t have used the word “generally”. What Skutnabb-Kangas and Lambert are telling us in their various ways is that immersion will have disastrous consequences for precisely the types of people to whom it is being applied in Catalonia. Lambert’s saying that in order for it to succeed here, we would need to find a way of miraculously improving the socio-cultural level of the Spanish-speaking underclass, ascribing prestige to Spanish, and to make teaching specific and voluntary, which is to say: abolish it.

    @Ian: When you say that “It is clear that what you do not like is Catalan and Catalonia. Immersion is just an excuse.” you remind me of Sharon at my primary school, who, when someone disagreed with her, used to stamp her feet and scream, “Why do you all hate me so much?” Can I call it Ian Syndrome?

  11. Oh yeah, English speaking children in Canada are rich, smart and have a prestigious language, Spanish speaking children in Catalonia are poor, stupid and have a infamous language. You sound pretty much like the “son of a bitch” Salvador Sostres.

    By the way, I do not hate anyone, and even less you. I simply disagree with you and try to make my point. I have the right to reply when someone criticizes my country, my language, my people, but still insists on living there and enjoying the many positive things we have. We are not perfect, I know, but who is? But if I would ever decide to hate someone, it would be someone more relevant than you.

    And you also remind me a daycare friend called Luisito, when you do not like something, you try to have the teacher ban it. And what you call stamp feet and scream, I call intelligent debate. But you know I was educated through immersion and what can you expect from such an individual?

Your email address will not be published. Required fields are marked *