Many bilinguals have a story or two of being surprised, if not shocked, when a person suddenly uses a language they did not expect, even though they know it. Some report not understanding what is said at first. There are even threads on the web which refer to this topic. Thus, an American in Japan writes that he gets many shocked looks from Japanese people when he speaks Japanese in front of them for the first time. A Serbian notes that she is very surprised when foreigners in her country speak Serbian. And a teacher in India reports that her eyes nearly popped out when she heard a French student utter something in Hindi.

Language choice by bilinguals - basically which language you use with whom and for what - is governed by a number of factors several of which pertain to the participants in the exchange (see here). For example, when you do not know the speaker, visual cues can prime a particular language as has been shown in a recent study.

Shu Zhang and her colleagues at the Columbia Business School in New York wanted to know if facial characteristics could lead to language expectation. They asked a number of Chinese students who had been in the United States for about a year to take part in a computer-mediated communication task. The participants viewed the photo of either a Chinese or a Caucasian face of a certain Michael Lee and listened to him on a recording talk with a standard American accent about campus-life topics. The very same recording was played irrespective of the photo. The students then spoke about these same topics in English and were recorded.

When the researchers assessed their verbal fluency and measured their speech rate when speaking to Michael Lee, they found fascinating results. The students' fluency ratings were lower, as was their objective speech rate, in the Chinese face condition than the Caucasian face condition. Basically, they had expected the Chinese face to speak Chinese and hence their English language output was disrupted when interacting with Michael Lee in English.

Of course, only static pictures were used in this study, and they were clearly associated with visual cues representing different cultural groups. But what about situations where you associate a specific language to a given speaker you have actually seen speaking that language? And what impact does this have, not so much on your production of speech, as in the previous study, but on its perception? If, as some researchers still believe, both languages are always active when bilinguals process speech, then it shouldn't have an impact on your perception (nor should bilinguals be shocked by an unexpected language). If, on the other hand, one language is activated and the other deactivated, or even inhibited, when faced with a particular interlocutor who is known for speaking a particular language, then this should show up in the results obtained.

Monika Molnar and her colleagues at the Basque Center on Cognition, Brain and Language in Spain examined this very issue. They first asked proficient Basque-Spanish bilinguals, who had acquired both languages before the age of three, to familiarize themselves with three types of interlocutors via video recordings: Basque speakers, Spanish speakers, and bilingual Basque-Spanish speakers. The interlocutors introduced themselves and talked about various topics such as their family background, hobbies, work, etc. The Basque and Spanish speakers only spoke their respective language whereas the bilingual speakers used both Basque and Spanish and code-switched from time to time.

In the second part of the study, the bilinguals participants saw each interlocutor on video produce a word or a made-up word and they simply had to decide whether the item was a real word or not by pressing one of two keys. Seventy five percent of the time, the Basque and Spanish interlocutors presented words in the language they had spoken during the familiarization phase (the experimenters called this the congruent trials since they corresponded to the language they had used before) and the remainder of the time, they produced words in the other language (the incongruent trials). Thus, the monolingual speakers surprised the participants at various points by uttering words in the wrong language. Until this happened, the participants did not know that the interlocutors also spoke the other language. As for the bilingual interlocutors, they presented words in the one or the other language equally.

The results showed the importance of interlocutor identity when bilinguals process speech. The participants responded faster to words produced in the language associated with the monolingual interlocutors during the familiarization phase (e.g. Spanish words produced by Spanish interlocutors) than to words produced by them in the other language (e.g. Basque words produced by these same Spanish interlocutors). As for the words produced by the bilingual interlocutors, in the one or the other language, they were responded to the slowest.

With these two studies, researchers have been able to bring into the laboratory a phenomenon that bilinguals experience from time to time in "real life" - being confronted with a person using the wrong language. They have also confirmed that a language can be deactivated in certain situations. Future studies of this type, in which sociolinguistic factors are taken into account, will be fascinating to follow as they will certainly extend our knowledge of how bilinguals process their languages - individually, or together in a bilingual mode (see here).   

Note: An earlier post on this topic examines non accommodation in bilingual interaction, i.e. situations where speakers share at least two languages, but for various reasons use different languages with one another in the exchange (see here).

For a full list of "Life as a bilingual" blog posts by content area, see here.

Photo from the author's private collection.


Zhang, Shu, Morris, Michael W., Cheng, Chi-Ying & Yap, Andy J. (2013). Heritage-culture images disrupt immigrants’ second-language processing through triggering first-language interference. Proceedings of the National Academy of Sciences, 110(28), 11272-11277.

Molnar, Monika, Ibáñez-Molina, Antonio & Carreiras, Manuel (2015). Interlocutor identity affects language activation in bilinguals. Journal of Memory and Language, 81, 91-104.

François Grosjean's website.

You are reading

Life as a Bilingual

What You Didn't Know About François Grosjean

Fifty years in the field of bilingualism

Bilingual Infants Learning New Words

How do bilingual infants manage to learn similar sounding words?

Do Musicians Make Better Language Learners?

Second language learning and musical ability