In the last ten years or so, a new term has appeared in the field of bilingualism, most notably in the area of bilingual education – translanguaging. One of its most visible proponents is Professor Ofelia García of the Graduate Center of the City University of New York. In a 2015 publication with Ricardo Otheguy and Wallis Reid, she defines translanguaging as, “… the deployment of a speaker’s full linguistic repertoire without regard for watchful adherence to the socially and politically defined boundaries of named (and usually national and state) languages”. Ofelia García has very kindly accepted to answer a few of our questions and we wish to thank her wholeheartedly.
Interview conducted by François Grosjean
You are a strong defender of allowing bilingual students in school to use their entire language repertoire through translanguaging. You write, “… forbidding bilinguals to translanguage, or assessing it negatively, produces an inaccurate measure of their language proficiency.” Can you expand on this a bit?
If we are truly interested in knowing what bilingual students know and what they can do with language, we must separate their ability to use certain forms of one language or another from their ability to use language. For example, in schools, students are asked to find the main idea of a text, to support an argument with text-based evidence, to infer, to make a convincing oral presentation, to work out a math problem. Especially emergent bilingual students may not be able to show that they can do these things if only allowed to use the language legitimized in school. Only by drawing from their entire language repertoire will bilingual students be able to demonstrate what they know, and especially what they can do with language. Being able to perform with language-specific features legitimized in schools is not the same as having general language ability or being knowledgeable of content.
Translanguaging in the classroom functions well when the children and the teacher share the same minority language. But in many other classrooms, students come from many different language backgrounds. How can translanguaging take place then?
Translanguaging pedagogy requires a different type of teacher, a co-learner. Classrooms are increasingly multilingual in the world. It is impossible for teachers to know all the languages of students. But it is possible for teachers to build a classroom ecology where there are books and signage in multiple languages; where collaborative groupings are constructed according to home language so that students can deeply discuss a text in the dominant school language with all their language resources; where students are allowed to write and speak with whatever resources they have and not wait until they have the “legitimate” ones to develop a voice; where all students language practices are included so as to work against the linguistic hierarchies that exist in schools; where families with different language practices are included. Any teacher, including a monolingual one, can take up translanguaging to enable their bilingual students to make deeper meaning and legitimize their home language practices.
Given that bilingual students, as they grow older, will increasingly find themselves interacting with monolinguals, especially in their future work place, should schools also encourage them to “keep to one language” and measure their proficiency at doing so?
Translanguaging leverages the fluid language practices of all bilingual students and communities to learn deeply, while also equipping students to recognize when to use what features for what purposes. Of course schools must develop bilingual students’ critical metalinguistic awareness, and their ability to suppress some language features from their repertoire at appropriate times. And schools must give students affordances to perform with only some of their features at times. And teachers must assess what students can do using their entire repertoire of language features, as well as selected ones. Yet, teachers should not evaluate bilingual children’s performances using only certain of their language features as being valid performances, and should not compare those performances to those of monolingual children in that language.
Turning to the more linguistic and cognitive aspects of translanguaging, how does the notion that a bilingual is not two monolinguals in one person – something I stressed as early as 1985 – tie in to translanguaging?
Your notion is essential to the concept of translanguaging. Translanguaging is based on François Grosjean’s idea that bilinguals are not two monolinguals in one (see here). That is why their performances in one or another language cannot be compared to that of monolinguals, since they would then be expected to draw only from less than half of their entire repertoire. This puts bilinguals in an unfair position.
In what way is translanguaging different from interacting with other bilinguals, changing base language freely, translating whenever needed, and intermingling one’s languages in the form of codeswitching and borrowing?
There is an epistemological difference between the theoretical position on language contact that has led to the constructs of borrowing, code-switching, calques, language interference, etc. and the concept of translanguaging. Language contact studies start with named languages as categories, and then look across these named categories. Linguists often refer to the behavior of bilinguals when they go across these named language categories as code-switching. It is an external view of language. But translanguaging takes the internal perspective of speakers whose own mental grammar has been developed in social interaction with others. For these bilingual speakers, their language features are simply their own. Translanguaging is more than going across languages; it is going beyond named languages and taking the internal view of the speaker’s language use.
The linguistic phenomena mentioned in the previous question have been studied by researchers for more than sixty years. What are the benefits of replacing them with “translanguaging” when the behavior is clearly the same?
The behavior may look to be the same from the external social perspective, from a perspective that doesn’t question why named languages and language hierarchies exist or the relationship between language and power. But seen from the internal perspective of the bilingual speaker, translanguaging behavior is clearly different. Translanguaging legitimizes the fluid language practices with which bilinguals operate. It posits that bilinguals have a much more complex and expanded repertoire than monolinguals. Bilingual speakers then appropriate all their linguistic features, no matter their social standing, instead of categorizing them as belonging to one national group or another to which they may not belong.
In your 2015 paper with Ricardo Otheguy and Wallis Reid, you state that we all have our own idiolect – a collection of individual lexical and structural features – and that there are large areas of overlap among idiolects so that we can communicate with one another. People call this sharing a language (e.g. English), something that is psychologically real for them, and yet you state that language is only a social and political construct, not a lexical or structural one. Can you explain?
Linguists can only describe linguistic features, but they cannot say as linguists which features constitute one language or another. The naming of a language is always a social, political and economic decision, not a linguistic one. Thus, it is not true that separate named languages have linguistic reality. However, it is true that languages are social constructs that have had very important real and material consequences in the lives of people, some bad, and some good. This social construction of language also has an important identity function for groups and individuals that cannot be denied. But named languages have often been constructed by a process of standardization that leaves out the practices of minoritized populations.
Bilinguals, according to you, only have one idiolect made up of more lexical and structural features than monolinguals and a more complex socio-cultural marking of which features to use and when. What evidence do you have, cognitive or neurolinguistic, that bilinguals do not have separate idiolects, one for each of their languages?
I am a sociolinguist who has specialized in the education of bilinguals. My evidence comes from the classrooms and from listening to bilingual children. From the bilingual eleven year old who once told me: “Spanish runs through my heart, but English rules my veins,” concretely expressing that for him there is one language system that cannot be separated because otherwise it would lead to his death. From the five year old Spanish-speaker who was learning English by repeating after the teacher, “That tree is grander.” From the teacher who tells me that her young first graders are always talking to themselves in Spanish, even when working through English. The work of neurolinguists is also beginning to show that when bilingual speakers perform linguistically, all the features of their repertoire are activated and available, even though speakers selectively suppress some features depending on the communicative situation in which they find themselves.
García, Ofelia (2009). Bilingual education in the 21st century: A global perspective. Malden, MA and Oxford: Blackwell/Wiley.
Otheguy, Ricardo, García, Ofelia & Reid, Wallis (2015). Clarifying translanguaging and deconstructing named languages: A perspective from linguistics. Applied Linguistics Review, 6(3), 281-307.