This study investigates the nature of L2 sentence processing mechanisms by examining whether or to what extent the sensitivity to prosody would affect the resolution of relative clause attachment ambiguity. The study examines how native speakers of a high attachment language would resolve the attachment ambiguity of their second language which has the opposite preference. Twelve Korean-speaking L2 learners of English and twelve native English speakers participated in a listening task to examine the effect of different types of prosodic information, such as prosodic boundary and focal prominence, along with a silent-reading task. The results showed that L2 learners' high attachment preferences in English were consistent in both tasks, but the focal prominence factor reinforced this preference in the listening task. The findings suggest that L2 learners' parsing routines generally followed their L1-specific high attachment, and this default preference was reinforced by convergent prosodic prominence and intonation contour. Emphasizing that L2 sensitivity to prosody reflects a specific modular organization, the study argues that an L1-induced default projection in processing of relative clause attachment can be enhanced or reduced by specific accentual patterns in L2 sentence processing, without the need to claim shallow processing of L2 learners.