Bilingual Language Control in Speaking and Translating

In the growing European Union internationalization goes hand in hand with an
increasing number of language backgrounds among its citizens. Being able to
speak several languages and switch between them is of increasing importance for
many citizens for both professional and private reasons. Although the ability
to speak to languages is often taken for granted, it is hotly debated how
bilinguals effectively keep their two languages functionally separate. Numerous
studies have shown that bilinguals cannot prevent interference from the other
languages (e.g., Christoffels et al, 2003; De Groot et al., 2000; Hermans et
al, 1998; Rodriguez-Fornells et al., 2005). Nevertheless, bilinguals have the
striking ability to confine their speech to just one language but are also
remarkably apt at changing language at will. Considering that languages are
represented in anatomical overlapping regions (e.g., Vingerhoets et al., 2003)
the key question is how language is controlled in our brain.
Sometimes the ability to voluntarily speak a specific language is lost
(Paradis, 2004). Bilingual aphasic patients sometimes mix languages
pathologically or switch between them without other linguistic impairment
(Fabbro et al, 2000). For example, in an exceptional case of alternate
antagonism, comprehension was normal but the patient was only able to speak
Arabic one day and French the next. On days she could not speak French, for
example, she was still able to translate from Arabic into French (Paradis et
al, 1982). This not only indicates that language systems may be preserved even
when language control is impaired, but also that language production and
translation may not be equivalent.
The Inhibitory Control (IC) model accounts for different recovery patterns in
aphasic bilinguals by assuming that their ability to control language is
damaged (Green, 1986; 1998). In this model a language-nonspecific control
device regulates the activation of various competing task schemas that specify
how a certain task (e.g., picture naming in English) is performed. It is
assumesd that processing in one language comes about by inhibition of the
other. Language switching experiments provide compelling evidence for the IC
model: It takes longer to switch from the second language (L2) to the dominant
language (L1) than vice versa (Meuter & Allport, 1999). This is explained by
assuming that L1 takes longer to recover from inhibition than L2. Note that
language-switching creates a situation in which both languages are frequently
used (mixed-language context), which is an aspect of the bilingual processing
that has been largely ignored. Yet, we found that a mixed-language context
profoundly slows down L1 without affecting L2 (Christoffels et al, 2007). The
question of language control becomes more pertinent when considering verbal
translation, an ability that emerges from acquiring another language. Here,
both languages are implicated at the same time (De Groot & Christoffels, 2006;
2007). Indeed, according to the IC-model even single word translation is a
high-conflict task. Even more complicated is simultaneous interpreting, where
the interpreter must continuously listen to and comprehend the speaker*s speech
and produce its translation at the same time. This is especially puzzling
because comprehension and production activate largely overlapping temporal
brain areas (Christoffels et al, in press). How is interpreting at all possible
if the input language is inhibited when both languages are involved and
activated simultaneously? In current models of bilingual processing translation
and interpreting are poorly accommodated (Christoffels & De Groot, 2005). Some
important issues are as yet unanswered, such as which brain areas support
language control, what are the neural correlates of translation processes and
what is the causal relation between control areas and language selection? We
believe that merging insights from psycholinguistics with a neurocognitive
approach can help to provide a framework to address these issues now that we
are able to use imaging methodology is to language research. Note that overt
language production has not been applied often in fMRI due to motion artifacts,
but recently used it successfully in single word production (Christoffels et
al., in press; 2006).

The aim of the current study is to identify the neural substrates of bilingual
language control in single language production and translation from one to
another language. This aim is important because insights in language control
may help to understand and treat the problems faced by bilingual aphasics and
improve language education generally and the training of professional
translators and interpreters, specifically. Further, these insights are
imperative to develop a neurological plausible model of language processing
that can explain bilingual speech production and translation.

The primary objective of this study is to gain understanding of the
neurocognitive basis of language control. To this end, we will acquire fMRI
data and behavioral responses of healthy bilingual adults.
Secondary Objective(s): There is still a dearth of information on the topic of
second language processing and translation. The design is rich in the sense
that many different aspects are worthwhile addressing, some of which still
exploratory. The secondary objectives are: (1) to study single language
production in the native tongue and in the second language, (2) to study
translation processes, (3) to investigate how single language production and
translation depend on similar or different brain structures by directly
comparing the two tasks, (4) to study language switching, (5) to study effects
of language context, (6) to distinguish between sustained and transient aspects
of language control, (7)to study the relation between proficiency in the second
language switching-related activity in the brain.

An experimental design is used. Participants will perform a picture naming or
translating task under different conditions. Brain activation during task
performance is measured using standard event-related functional MRI (see
protocol for details).

There are no known risks associated with participating in an fMRI study. This
is a noninvasive technique involving no catheterizations or introduction of
exogenous tracers. Numerous children and adults have undergone magnetic
resonance studies without apparent harmful consequences. Some people become
claustrophobic while inside the magnet and in these cases the study will be
terminated immediately at the subject's request. The only absolute
contraindications to MRI studies are the presence of intracranial or
intraocular metal, or a pacemaker. Relative contraindications include pregnancy
and claustrophobia. Subjects who may be pregnant, who may have metallic foreign
bodies in the eyes or head, or who have cardiac pacemakers will be excluded
because of potential contraindications of MRI in such subjects. Although there
is no direct benefit to the participants from this proposed research, there are
greater benefits to society from the potential knowledge gained from this
study. Insights in the neural basis of language control is instrumental for
understanding and treating language deficits, for training of professional
translators, and for the design of language education of the general public.