Abstract

Using a learner corpus, the TELC corpus, this study investigated L2 prepositional error patterns in terms of developmental trajectory and L1 influence in naturalistic data. The analyses were conducted across different score bands and L2 proficiency levels. The findings revealed a U-shaped pattern for L2 preposition error averages across both score bands and L2 proficiency scores. Frequencies of L2 prepositions and error correction patterns indicated that multiple sources contribute to the emergence of these errors, including but not limited to frequency. Further analyses of ungrammatical sentences with prepositional errors provided strong evidence of L1 influence across different essay score bands.