This essay defends the idea that language is a unique, genetically underwritten (informationally encapsulated) “module of the mind” and considers some consequences such a stance holds for the psycholinguistic study of nonnative language (L2) acquisition. As is well known, language as conceived within the tradition of generative grammar (e.g., Chomsky, 1965, 1975, 1980, 1986) is unlike other types of cognition and its basis (viz., Universal Grammar) is innately given; this is the position I support in this paper. Specifically, I begin by summarizing the main arguments for this position, after which I discuss (following Segal, 1996) four different conceptions of what a module is and then devote particular attention to the theory developed by Fodor (1983, 1985) on the architecture of mind (his “modularity thesis”). There follows a comparison between the distinct views held by Chomsky and by Fodor on the conception of “the language module” and on the structure of mind more generally. Arguing that the two views are not inherently incompatible, I offer a speculation on how to begin to reconcile them, which leaves me well positioned to advance three specific implications for the theoretical study of L2 acquisition within the framework of generative grammar, couched within a theory of the modular mind.