Why “younger” is not always “better” in foreign language learning

TProfessor Jean-Marc Dewaelehis post was contributed by Professor Jean-Marc Dewaele of Birkbeck’s Department of Applied Linguistics and Communications.

Politicians can be forgiven for not having much time to read the fine print when asking advisers to translate research findings into workable policies. Or does it work the other way round? Do politicians decide on policies first and subsequently ask advisers for appropriate research findings to back up the policy?  This seems the case when considering the wide consensus across the world about the benefits of early introduction of foreign languages (FLs) in pupils’ school curriculum. The expression “younger is better” in education sounds perfectly plausible, is simple and convincing, and must be a vote winner.

In the UK, FLs used to be introduced in secondary education. Estelle Morris, then Secretary of State for Education, changed this policy in 2002, scrapping compulsory modern FLs for 14- to 16-year-olds, and introducing them in primary schools. She claimed in 2006 that: “Starting at a much younger age is the best way of making sure we get more pupils taking exams and, more importantly, more of them enjoying and feeling confident about speaking a language other than their own”.

In other countries, FL teaching has even been introduced in nursery schools.  There seems to be a universal consensus among politicians that an early start in FLs will lead to a smoother, quasi-effortless learning process leading to high levels of proficiency in the FLs. Is this a myth?!

Spanish ClassCounter-intuitively, research suggests that adolescents and adults progress more quickly than children when learning FLs in a school context (so-called “instructed FL learning”). Many researchers have serious doubts about age of onset being the most important variable in successful FL learning. Indeed, research shows quite clearly that starting age is only one of many independent variables in very complex question.

A crucial distinction exists between so-called naturalistic and instructed FL learning.  Research on naturalistic learners, typically immigrants, shows that younger children are indeed more likely to become undistinguishable from native speakers of the FL compared to their parents and older siblings. However, the picture is not so clear in research on instructed FL learning, a crucial distinction that is commonly overlooked.

A large-scale project on Age of Onset of Acquisition (AoA) in formal foreign language teaching, the Barcelona Age Factor project has looked at effects of starting age and the comparisons were always of groups with same amount of instruction hours (200, 400, 700, and 800 hours of instruction). Earlier exposure (ages eight to nine) to English (as a third language) in a classroom did not result in better performance. Learners who started English at age 11 and those who had started at age 14 were found to progress more quickly than early learners but, after a similar number of hours of exposure, the differences between the groups were limited, with older starters still having a slight advantage. In another study with young adult learners who had 2500 hours of instruction, AoA was not found to have an effect but amount and type of exposure had a positive effect. In other words, input seems more important than AoA.

A Swiss study (Pfenninger, in press) found no advantages of an early start among Swiss learners of English even after five years of instruction. The writing skills of late starters caught up with those of the early starters within six months.  One possible explanation is that older learners have greater metalinguistic, metacognitive and strategic skills.

Munoz points out that from the observations that younger immigrants and immersion students in naturalistic settings seem to outperform older peers “an inferential leap is made in the assumption that learning age will have the same effect on students of a foreign language, when they are exposed to only one speaker of that language (the teacher, who is not usually a native speaker) in only one setting (the classroom) and only during very limited amounts of time”.

This does not mean that there are no age effects at all in learning and later use of the FLs.  Indeed, younger children seem to be more motivated in learning FLs. In my own research on language choice and self-perceived proficiency among more than 1500 adult bi- and multilinguals, I found that early starters in a FL felt more proficient in speaking, comprehending, reading and writing their FLs. They were also more likely to choose the FL for the expression of anger and feelings, for inner speech and mental calculation.  Interestingly, the effect of mode of instruction was even stronger than age of onset: participants who had acquired the FL naturalistically or in mixed mode (formal instruction combined with authentic use) outperformed participants who had learned the FL through classroom instruction only.

In their excellent overview of the literature on age and the teaching of FLs, Lambelet and Berthele point out that more research is needed on improving age-appropriate teaching techniques in order to boost motivation levels and metalinguistic awareness of FL learners of all ages. Moreover, extra thought needs to be given to the primary school teachers who are suddenly expected to teach a FL and who may lack in confidence and competence. In other words, those arguing for an early introduction of FLs at school need to take the nuanced research findings into account and avoid promising miracles.

At what age did you start learning a foreign language? How do you think this affected your fluency and confidence in the language? Please leave your comments below.

Further reading

  • Dewaele, J. M. (2009). Age effects on self-perceived communicative competence and language choice among adult multilinguals. Eurosla Yearbook, 9, 245–268.
  • Enever, J. (2011). ELLiE. Early Language Learning in Europe. London: British Council.
  • Lambelet, A. & Berthele, R. (2014). Âge et apprentissage des langues à l’école. Revue de literature. Fribourg: Research Centre on Multilingualism.
  • Pfenninger, S. (in press).The literacy factor in the optimal age debate: a 5-year longitudinal study. International Journal of Bilingual Education and Bilingualism.
  • Muñoz, C. (2011). Input and long-term effects of starting age in foreign language learning. International Review of Applied Linguistics in Language Teaching, 49, 113–133.

Other posts by Professor Dewaele:

Other blogs about linguistics:

. Read all 9 comments . Category: Social Sciences History and Philosophy . Tags: , , , , ,

Pardon my foreign accent!

This post was contributed by Professor Jean-Marc Dewaele from Birkbeck’s Department of Applied Linguistics and Communications.

In an increasingly multilingual world, more and more people are exposed to foreign accents, and if we use foreign languages ourselves, the chances are that we have our very own foreign accent.  Does a foreign accent matter?  Yes, says Moyer (2013) because it is so salient: “it is the means by which we make ourselves understood, and the yardstick by which others judge us, whether we like it or not”. A foreign accent can have unexpected social consequences.  Indeed, those with a strong foreign accent are often judged as less competent, less educated, less intelligent, and less trustworthy (Fuertes et al, 2011).  In other words, a foreign accent can be a heavy burden. Veronica Glab (2014), a Canadian-born Pole currently living in Madrid, describes how her “Polish and American accents came up like bile” when speaking Spanish, and how it made her feel like an outsider, causing a sense of failure.

It is really hard to get rid of a foreign accent when the foreign language learning started later in life as Glab can testify.  Only a small minority of late learners manage to speak a foreign language with a native-like accent.  They are typically people with good auditory working memory, i.e. people who can keep sounds in their short term memory for a longer time.  Other psycho-cognitive predictors are phonetic coding ability, an ability to sing and a high level of empathy (Hu et al, 2012).

Personality has also been mentioned by multilinguals as having an effect on their attitudes towards foreign accents. Glab wonders whether her foreign accent might be linked to her timidness and self-consciousness.

Having a slight French-Dutch accent in English, I was never too bothered by it as English speakers usually like French accents, which they tend to find “cute”.  My accent became a liability only once, in 2003, at the time of the second Gulf war.  Having been stopped for speeding in Independence, California, by a patriotic police officer, who told me the helicopter had followed me for half an hour through the desert, and caught me going over the 60 miles per hour limit .  Did I have an explanation for that?  My answer, but probably even more my French accent, did not please the officer.  Didn’t I realise, he asked, that the French were not welcome in the US? (It was the time of the “Liberty fries” – because the Americans weren’t allowed to use the F-word anymore).  I hastened to point out that I was Belgian, and that Belgium had not voted against the US in the UN.  Little did it matter.  My infraction, compounded by my French accent cost me $ 150.

What this little episode shows is that attitudes toward foreign accents vary according to the accent in question, and to the general socio-historical context.  It also shows that different people have different attitudes.  So what exactly affects attitudes towards foreign accents? We explored this question more systematically in a study that just came out (Dewaele & McCloskey, 2014). We collected data on attitudes towards the foreign accents of other people and the own foreign accent via an online questionnaire.  A total of 2035 multilinguals from around the world participated. We found that extraverted multilinguals, who were emotionally stable and tolerant of ambiguity were significantly less bothered by the foreign accents of others. Only more neurotic multilinguals were bothered by their own foreign accent. Unexpectedly, participants who knew more languages to a higher level were more negative about the foreign accent of others and their own. However, participants who grew up in an ethnically diverse environment, who had lived abroad and who were working in an ethnically diverse environment were significantly more positive about foreign accents. Women had a more negative attitude towards their own foreign accent – but not that of others.

It thus seems that how much we are bothered by foreign accents falls partly outside our conscious control as it depends on personality, language learning history, current linguistic practices and sociobiographical background.

Other posts by Professor Dewaele:


Dewaele, J.-M. & McCloskey, J. (2014). Attitudes towards Foreign accents among adult multilingual language users. Journal of Multilingual and Multicultural Development

Fuertes, J.N., Gottdiener, W.H., Martin, H., Gilbert, T.C., & Giles, H. (2011). A meta-analysis of the effects of speakers’ accents on interpersonal evaluations. European Journal of Social Psychology, 42, 120-133.

Hu, X., Ackermann, H., Martin, J.A., Erb, M., Winkler, S., & Reiterer, S. (2012). Language aptitude for pronunciation in advanced second language (L2) learners: Behavioural predictors and neural substrates. Brain and Language doi:10.1016/j.bandl.2012.11.006.

Moyer, A. (2013). Foreign Accent: The Phenomenon of Non-native Speech. Cambridge: Cambridge University Press.

. Read all 5 comments . Category: Social Sciences History and Philosophy . Tags: , , , , ,

Multilingualism in psychotherapy

Professor Jean-Marc DewaeleThis post was contributed by Professor Jean-Marc Dewaele, who presented on this topic at a conference earlier this month to celebrate the 50th anniversary of the Department of Applied Linguistics and Communication. He originally contributed this blog post in 2013, when it won that year’s Equality and Diversity Research Award  from the British Association for Counselling and Psychotherapy for the paper Costa & Dewaele (2012).

Emotions play a crucial part in our daily lives. We share them, jokingly or seriously, in face-to-face interactions, in texts or emails, and this is a crucial social activity, which is crucial for our interpersonal relationships and our individual well-being.

It is much more difficult to communicate emotions in a foreign language (LX), because of gaps in the linguistic, pragmatic and sociocultural knowledge needed to express the full range of emotions.  LX users (and I’m one myself in English) can feel frustrated at not being able to project an accurate image of self. Interestingly, a majority of multilinguals report feeling different depending on the language they are using (Dewaele & Nakano, 2013). It typically takes a couple of months before LX users can be relatively confident that their communicative intentions in expressing emotions will be correctly decoded and that their capacity to infer the emotions expressed by their interlocutors is sound. The difficulty lies in the fact that as first language (L1) users we can express our own emotions precisely, and recognise other people’s emotions unerringly.  This sense of security is lost when having to communicate emotions in the LX (Dewaele, 2010).

I have explained in an earlier blog that emotion words can have a very different resonance in the different languages of a multilingual: swearwords typically don’t sound as offensive in an LX and expressions of love don’t sound as strong. A study on Turkish L1-English L2 bilinguals showed that emotional phrases presented in an L1 elicited higher skin conductance responses than the translation of these phrases in a L2 (Caldwell-Harris & Ayçiçeği-Dinn, 2009), in other words hearing Turkish taboo words made participants sweat more than the equivalent words in English.  It thus seems that the L1 is the language of the heart, while the LX often fulfills an intellectual function and is relatively emotion-free, creating a feeling of detachment or disembodied cognition (Pavlenko, 2013).  Research has shown that immigrants’ memories that were experienced in the L1 are generally richer in terms of emotional significance when recalled in the L1. When these L1 memories are recalled in an LX, some emotional intensity is lost.

This might not always be a bad thing, especially if the multilingual is talking about traumatic events like torture or rape.  It is crucial that psychotherapists are aware of this phenomenon.  Indeed, there are important psychotherapeutic implications of being multilingual both for the patient and for the therapist. Beverley Costa, director of the charity Mothertongue, was struck by a quote from a Greek-English-French participant in my 2010 book:

I think when I talk about emotional topics I tend to code-switch to English a lot. I remember when I was seeing a psychologist in Greece for a while I kept codeswitching from Greek to English. We never really talked about this…To my mind it may have been some distancing strategy. (p. 204).

Beverley contacted me to carry out a study on differences between monolingual and multilingual therapists.  The paper, which was published in 2012, showed that psychotherapists agreed that learning a language made them better attuned to other languages and to multilinguals. They also believed that through working across languages they had learned to think carefully about how they used language, to check understanding and to simplify their language. Although no therapist had tried out inviting other languages in to the therapy they were interested and saw the potential of trying this.  The judges from the British Association for Counselling and Psychotherapy who awarded us the Equality and Diversity Research Award (2013) particularly liked our recommendations for research, practice, training and supervision: “Firstly, it would be useful and interesting for further research to be conducted on language switching in therapy – how it is initiated and what it signifies. The second recommendation relates to practice. This research highlights the need for therapists to pay attention to the way in which the inherent self-disclosure is managed by the therapist who speaks multiple languages (…). It is also suggested that therapists consider if, when and how to initiate inviting languages they may not understand into the therapeutic space and the therapeutic implications of such an initiative. Finally, it is suggested that training of psychotherapists needs to include a component on the psychological and therapeutic functions of multi/bilingualism and underlying implications for therapy. Training and supervision for psychotherapists could also include practice for therapists to make formulations in different languages. With increasing numbers of multilingual people now accessing therapeutic services and becoming therapists, it seems timely for the curricula of psychotherapy courses and therapeutic practice for all therapists – mono and multilingual – to be revised in order to take into account the changing profile and language needs of users and providers” (Costa & Dewaele, 2012: 35).

The study and the award will be presented at the BACP Research Conference in Birmingham on 10-11 May 2013.

Other posts by Professor Dewaele:


Caldwell-Harris, C.L. & Ayçiçeği-Dinn, A. (2009) Emotion and lying in a non-native language. International Journal of Psychophysiology 71, 193-204.

Costa, B. & Dewaele, J.-M. (2012) Psychotherapy across languages: beliefs, attitudes and practices of monolingual and multilingual therapists with their multilingual patients. Language and Psychoanalysis 1, 18-40.

Dewaele, J.-M. (2010) Emotions in Multiple Languages. Basingstoke: Palgrave-Macmillan. doi:10.1057/9780230289505.

Dewaele, J.-M. & Nakano, S.  (2013) Multilinguals’ perceptions of feeling different when switching languages. Journal of Multilingual and Multicultural Development 34 (2), 107-120. DOI: 10.1080/01434632.2012.712133

Pavlenko, A. (2012) Affective processing in bilingual speakers: Disembodied cognition? International Journal of Psychology 47, 405-428.

. Read all 3 comments . Category: Social Sciences History and Philosophy . Tags: , , , , ,

“Most obscene title of a peer-reviewed scientific article” – an amusing award for a serious academic paper

This post was contributed by Professor Jean-Marc Dewaele, from Birkbeck’s Department of Applied Linguistics and Communication.

This post contains strong language.

As an applied linguist and a multilingual I have always been interested in the communication of emotion in a person’s multiple languages.  It seems that telling jokes in a foreign language, declaring love or promising something in a foreign language does not quite have the same resonance as it typically has in a native language (see also my taster lecture – contains strong language!).

One particularly interesting area is how multilinguals swear.  Indeed, swearwords in a foreign language don’t sound quite as bad as the ones in the native language, and students spending some time abroad are eager to pick up some of these words that their teachers did not want to teach them.  However, these enthusiasts abroad quickly realize that the people around them do not necessarily approve of the liberal use of swearwords.  What sounds like “funny” words to the foreign-language user can in fact be deeply upsetting words to the native language-users.

I remember how Livia, my trilingual daughter (English, Dutch, French as first languages), aged 7, playing with a Belgian bilingual boy (Dutch, French as first languages), who, when he heard she also had English as a first language, exclaimed that he knew English too, after which he uttered Fuck you!, which made my daughter jump and switch to English: But you can’t say this!  The boy looked surprised at her emotional reaction.  He had clearly no idea that this funny expression would upset his friend.

What matters when swearing is to know how to do it “appropriately”, in other words, know the context in which certain swearwords and expressions may be tolerated or appreciated.  For foreign language-users it may take years, and even then they typically avoid them because the reactions they elicit may differ from native users using the same words in the same situation.  I think that it is because swearing is an indication of “in-group” membership.  However, if you have a foreign accent you clearly don’t belong to the “in-group”, and you’re expected not to use these words, and not make fun of the head of state or queen/king.

In 2010, I published a paper:  ‘Christ fucking shit merde!’ Language preferences for swearing among maximally proficient multilinguals. Sociolinguistic Studies, 4 (3), 595-614. (doi : 10.1558/sols.v4i3.595). I investigated language preferences for swearing among multilinguals using an on-line questionnaire. They consisted of 386 adult multilinguals who had declared that they were maximally proficient in their first and second languages and used both languages constantly.

I discovered that despite similar levels of self-perceived proficiency and frequency of use in the first language and second language, the first language was used significantly more for swearing and first language swearwords were perceived to have a stronger emotional resonance. An analysis of additional interview data confirmed the findings of the quantitative analysis, also highlighting cultural issues in swearing.

The working title of the paper was Language preferences for swearing among maximally proficient multilinguals until I heard an Anglo-Canadian author, Nancy Huston, who has lived in Paris for many years, being interviewed on France Inter about her swearing behaviour.  She explained that when she needs to express a strong emotion, like sudden anxiety, or when dropping a hammer on her foot, she swears in English. The journalist then asked her Vous dites quoi? ‘What do you say’?  Nancy answers: Je dis Christ fucking shit merde! ‘I say Christ fucking shit merde!’  (“merde” meaning ‘shit’, is a high-frequency French swearword).  She’s surprised at the presence of the French swearword and adds: Ah je peux ajouter merde! ‘Ah, I can add merde!’.

I thought this quote would be the perfect illustration of my paper, namely that while multilinguals generally prefer swearing in their first language, some second language swearwords may creep into their core emotional vocabulary as a result of years of affective socialization in the culture of the second language.

I had to argue with the guest editor and the general editor Sociolinguistic Studies to keep the swearwords uncensored in the title.  I explained that it would make no sense to censor them, as the code-switching would become invisible, and that it was exactly the phenomenon I was interested in. They agreed in the end.

A few days ago, to my amused surprise, I won the award for “for most obscene title of a peer-reviewed scientific article”.  Merde alors!

Other blog posts by Professor Dewaele:

. Read all 30 comments . Category: Social Sciences History and Philosophy . Tags: , , , , , ,