This paper demonstrates differential, systematic, cross-modal responses to language by contrasting the regularities of infant movement behavior in contexts in which infants are presented with language stimuli, with those exhibited in the context of music. Using a detailed coding system, it shows that infants recognize the social underpinnings of language and respond to language stimuli with vocal, gaze, head, and torso, (but not arm or manual) movements that differ from those exhibited to music stimuli. It proposes that measures of these sorts of bodily gestures can not only provide a reliable supplement to looking time measures for gauging infant language abilities, but also uncover novel, richly textured nuances to our understanding of infant language acquisition.