Involuntary language switching in two bilingual patients during the Wada test and intraoperative electrocortical stimulation

Abstract

We present two bilingual patients without language disorders in whom involuntary language switching was induced. The first patient switched from Dutch to English during a left-sided amobarbital (Wada) test. Functional magnetic resonance imaging yielded a predominantly left-sided language distribution similar for both languages. The second patient switched from French to Chinese during intraoperative electrocortical stimulation of the left inferior frontal gyrus. We conclude that the observed language switching in both cases was not likely the result of a selective inhibition of one language, but the result of a temporary disruption of brain areas that are involved in language switching. These data complement the few lesion studies on (involuntary or unintentional) language switching, and add to the functional neuroimaging studies of switching, monitoring, and controlling the language in use.

Introduction

People are considered bilingual when they use two or more languages or dialects in their everyday lives (Grosjean, 1994). According to this definition, more than half of the world population is bilingual (French and Jacquet, 2004, Harris and Mc Ghee Nelson, 1992). Research on bilinguals has traditionally concentrated on finding differences in the cerebral representation between languages, initiated by the observation that in some bilingual aphasics there is a differential recovery of languages (Albert & Obler, 1978). Although various patterns of language recovery have been described in these patients, no general principle seems to be compatible with all cases. For instance, the language preferentially recovered is not necessarily the native language, whereas in some cases one of the languages is never recovered. In other cases, patients may use one language with the accent of the other, even though both languages were spoken with native-like fluency before the cerebral accident (Paradis, 1998). Overall, in clinical studies ‘no differences have been reported between unilingual and bilingual aphasics with respect to localization, frequency of occurrence or any other parameter’ (Paradis, 1998).

A second source of information regarding multilingual language representation comes from subjects without language disorders. When reviewing functional neuroimaging studies (fMRI, PET, MEG), however, the cortical organization in bilinguals remains controversial. A number of studies have shown differences in localization of first and second languages in healthy subjects (Kim et al., 1997, Simos et al., 2001), but these results are disputed by others (Chee et al., 1999, Mahendra et al., 2003). When part of the cortical surface of bilingual patients is tested with electrocortical stimulation during surgery, all (four) studies report that L1 and L2 are represented in both distinct as well as shared cortical sites (King and Schell, 1987, Ojemann and Whitaker, 1978, Roux et al., 2004, Walker et al., 2004).

As the representation of language in the brain of multilingual subjects remains controversial, even less is known of how subjects manage to switch between different languages (Fabbro, 2001). Language switching (also named code-switching or language mixing) allows multilingual people to use a given language at will and to adjust to the appropriate environment, while at the same time non-targeted languages are suppressed (Muysken, 2000, Poplack, 1980). Some authors have suggested that switching between languages involves aspects of central executive function and is processed by frontal areas, although specific neural systems are not known (Hernandez et al., 2001, Penfield and Roberts, 1959). Pathological language switching is very rarely reported. In brain-damaged bilingual patients language switching may be inappropriate (e.g., when the listener does not understand the language) and occur involuntary (Paradis, 1995). One patient with a large left frontal tumour displayed pathological language switching without any other linguistic impairments (Fabbro, Skrap, & Aglioti, 2000). Functional neuroimaging studies in normal bilinguals have shown involvement of left inferior frontal, supramarginal and dorsolateral prefrontal areas during language switching (Hernandez et al., 2001, Price et al., 1999). We know of only one recent study where switching was induced temporarily; these were two bilingual patients that experienced involuntary language switching after receiving repeated transcranial magnetic stimulation on the left dorsolateral prefrontal cortex as treatment for depression (Holtzheimer, Fawaz, Wilson, & Avery, 2005).

In this article, we describe two invasive methods of brain mapping (namely the Wada test and intraoperative electrostimulation) that transiently induced functional disruption in two bilingual patients. In both cases, following global (Wada) or loco-regional (stimulation) inhibition, there was a transitory switch from the first to the second language. We conclude that both methods temporarily induced a disturbance of the neural networks that are involved in language switching.