The neural correlates of grammatical gender decisions in Spanish Arturo E. Hernandez, CA Sonja A. Kotz, 1 Juliane Hofmann, 1 Vivian V. Valentin, 2 Mirella Dapretto 3 and Susan Y. Bookheimer 3 Department of Psychology, University of Houston, 126 Heyne Bldg., Houston, TX 77204 -5022, USA; 1 Max Planck Institute for Cognitive Neuroscience, Germany; 2 University of California, Santa Barbara; 3 University of California, Los Angeles, USA CA Corresponding Author: aehernandez@uh.edu Received 8 December 2003; accepted 8 January 2004 DOI: 10.1097/01.wnr.0000119452.05850.8c In the current study, nine participants were asked to make gender decisions for a set of Spanish nouns while being scanned with functional MRI (fMRI). Words were chosen in which a direct mapping between ending and gender (‘‘transparent’’ items such as carro fem or casa masc ) is present and those in which there is not a direct relationship (‘‘opaque’’ items such as fuente fem or arroz masc ). Direct comparisons between opaque and transparent words revealed increased activity in left BA44/45, and BA44/6 as well as bilateral activation near BA 47/insula and the anterior cingulate gyrus. These results reveal activity in areas previously found to be devoted to articulation of the determiner and to morphological processing. Taken together they support the notion that gender decisions for opaque items requires deeper and more e¡ortful processing during the retrieval of lexical and syntactic informa- tion. NeuroReport 15:863^866  c 2004 Lippincott Williams & Wilkins. Key words: Cognitive neuroscience; fMRI; Language; Spanish; Syntax INTRODUCTION Grammatical gender is a pervasive phenomenon in many of the world’s languages. However, this gender is not conceptual in nature. For example, the French word for ocean (mer) is feminine, while the Italian (mare) and the Spanish (mar) can be either masculine or feminine in Spanish. This clearly shows that gender markings are not completely systematic across languages and do not rely solely on conceptual information. Why does grammatical gender exist? What is it for? Some have suggested that gender helps listeners to keep track of the referents in a sentence, much like uniform numbers indicate the identity of players on a football field [1]. For example, a Spanish speaker will say something like ‘pasameLA’, pass it fem to me. This allows one to cut the number of potential referents in half. Furthermore, gender marked nouns have to agree with determiners and adjectives. Hence, gender plays a role both at the lexical as well as syntactic level. Recent discussion in the neuroimaging literature has considered the nature of syntactic processing across a variety of tasks [2,3]. These studies revealed a mixture of results potentially due to a variety of tasks such as sentential tasks [4,5], phrasal [6] and single word tasks [7]. However, all these results have shown increased activity in a set of areas which include BA 44 and 45 for syntactic tasks. Recent work by Miceli and colleagues has also found a similar locus for gender decisions of single nouns compared to phoneme and semantic decision tasks [8]. However, Miceli and colleagues did not explore, opacity, a crucial distinction in Italian. In Spanish, as in Italian, gender can be marked both in a transparent manner (-o for masculine, -a for feminine) and in an opaque manner (in which items can end in e,l,n,r,s,t and z which tend to be more irregular). Previous work has found that processing of opaque items results in longer reaction times [9]. However, to date no published study has investigated the neural correlates associated with the processing of opaque items. What brain areas will reveal increased activity for opaque items relative to transparent items? Recent work in computational linguistics suggests that the gender of a single word can be computed by word endings [10]. Whereas transparent items involve a one-to-one correspon- dence between word and ending, opaque items require more complex lexical processing in order to match a particular ending with a gender. Studies using fMRI in German have confirmed the finding that overt generation of the determiner for a noun results in increased activity in BA 44 near BA 6 [11]. Activity in this area has been associated with phonological retrieval [12,13]. If gender decisions for opaque items require more complex phonological retrieval, increased activity in superior BA 44 should be observed. Second, one would also predict that gender decisions should lead to increased activity in BA 44/45 in an area proximal to that observed by Miceli and colleagues. It remains to be seen if this area shows increased activity for opaque items relative to transparent items. BRAIN IMAGING NEUROREPORT 0959-4965  c Lippincott Williams & Wilkins Vol 15 No 5 9 April 2004 863 Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.