Introduction

Theoretical models of language processing abound, but, unlike memory where animal models are informative, only experiments involving human volunteers seem adequate to probe the complexities of human language. In order to determine which of these models is most appropriate, cognitive scientists are increasingly turning to brain imaging to provide answers to old questions as well as to provide a platform to ask new ones.

The advent of functional magnetic resonance imaging (fMRI) has provided the opportunity for scientists from a wide variety of disciplines to investigate multiple facets of human cognition in healthy volunteers and clinical populations. Imaging with fMRI is attractive compared to previous methods like PET (positron emission tomography) and X-Ray scintigraphy because investigations can be repeated without exposing the study subject to ionizing radiation (Raichle, 1998). fMRI reveals parts of the brain where blood flow increases or decreases in response to performance of a cognitive task (Jezzard, Matthews & Smith, 2001). This review covers what has been learned regarding the topography of brain activation when bilinguals process different languages. The effects of cognitive demands of different languages, levels of processing, proficiency, and language background are considered.

Studies on bilinguals in clinical populations

Despite their heterogeneity and varied rigor, studies based on clinical populations are important because they provide information about brain regions that are critical to particular cognitive functions (as opposed to being merely participatory).