An optical imaging study of semantic and syntactic processing by bilinguals Lilian C. Scherer a,b,c,*,1 , Francine Giroux a,d , Fre ´de ´ric Lesage a,e,f,g , Noureddine Senhadji a , Habib Benali a,d,f,g , Ana I. Ansaldo a,d a CRIUGM, Centre de recherche, Institut universitaire de ge ´riatrie de Montre ´ al, 4565, chemn Queen-Mary, Montre ´l, Que., Canada H3W 1W b UFSC, Universidade Federal de Santa Catarina/SC, Brazil c UNISC, Universidade de Santa Cruz do Sul/RS, Mestrado em Letras, Brazil d Faculte ´ de Me ´decine, Universite ´ de Montre ´ al, Montreal, Canada e Ge ´nie E ´ lectrique, E ´ cole Polytechnique de Montre ´al, Canada f ISERM, Paris, France g CRM, Montreal, Canada Accepted 6 July 2006 Introduction The combination of neuroimaging and psycholinguistic approaches may contribute to bridging psycholinguistic theory and experimental neu- roscience. With regard to the processing of syntax and semantics, most PET and fMRI studies are accomplished with monolinguals, and report controversial findings. Thus, although researchers agree upon the fact that both semantic and syntactic aspects are crucial for determining sentence meaning, the underlying computations remain controversial. Hence, some researchers claim for the existence of modular systems, specialized in one single language component, and predominantly left lateralized (Fodor, 1983), whereas others state that syntactic and semantic processing depend upon a widespread and cooperative network of brain areas in both hemi- spheres (McClelland, 1987). Two perspectives have thus emerged: one pos- its that syntax is processed prior to semantics (Friederici, Meyer, & von Cramon, 2000), while the other assumes that both semantic and syntactic processing occur simultaneously (Hagoort, 2003; Just, Carpenter, Keller, Eddy, & Thulborn, 1996). Further, it is considered that semantic and syn- tactic processing could be sustained by a common network including Bro- ca’s and Wernicke’s areas, with activations in Wernicke’s area preceding those in Broca’s area during semantic processing, and activations in Bro- ca’s area preceding those in Wernicke’s area with syntactic processing. However, the temporal chronometry of sentence comprehension remains to be determined. Further, given that monolingualism is increasingly becoming the exception rather than the norm in the world, it appears nec- essary to accommodate current models of language processing to multilin- gual language use. Among neuroimaging techniques currently available, OI (optical imaging) has a potential to elucidate the issues of temporal chronometry in language processing, given that its temporal resolution is superior to that of PET or fMRI (Weng, Ding & Volkow, 1996; Kennan, Kim, Maki, Koizumi, & Constable, 2002). Its efficiency to examine the neural substrate of syntax with unilingual populations (Noguchi, Tatsuya, & Kuniyoshi, 2002) has already been demonstrated. However, there is still a need for more overlapping measurements with OI, particularly with bilinguals. The present study used OI to identify the neural substrate of semantic and syntax processing with proficient bilinguals. Methods Participants Ten right-handed healthy young adults (age range 20–35), Brazilian Portuguese native speakers, highly proficient in L2 (French), acquired after the age of 7. Material One hundred and sixty sentences (80 per language) were developed, so as to accomplish semantic or syntactic judgment, with 20 correct and 20 incorrect sentences per condition. The syntactic condition tapped on ver- bal agreement (subject and verb agreement), whereas in the semantic con- dition the verb attributed a correct or incorrect sense to the sentence. In order to allow for comparisons of data across languages and across con- ditions, the number of words and propositions was statistically controlled, and the sentence pattern was maintained across languages and conditions. Procedures Behavioral and optical imaging data were recorded simultaneously. Sentences were presented centrally on a computer screen, in blocs alternat- doi:10.1016/j.bandl.2006.06.107 * Corresponding author. E-mail address: lilian.cristine.scherer@criugm.rtss.qc.ca (L.C. Scherer). 1 Sponsored by CAPES/Brazil. www.elsevier.com/locate/b&l Brain and Language 99 (2006) 8–219