A Theoretical Rationale for the Importance of Patterning in Language Acquisition and the Implications for Data-driven Learning

This article makes the case for the importance of exposure to language patterns in the context of second language acquisition. From the theoretical perspective of the Usage-Based (UB) model of acquisition, I argue that in the process of first and second language acquisition, the learner attends to frequently used form-meaning pairings that they experience. These then become entrenched as grammatical knowledge in the mind of the speaker (Ellis and Ferreira-Junior 2009). From a data-driven learning (DDL) perspective, I argue that the UB model provides a strong rationale for this approach because it offers a principled means of curating data from corpora to aid L2 learning. Work that brings the UB model and corpus linguistics (CL) together is emerging. For example, Römer (2019) explores patterns of verb-argument constructions (VACs) acquired by German beginner-level learners of English. Building on Römer’s work, this study focuses on certain patterns and explores these using data from the 55-million-word Cambridge Learner Corpus. It illustrates how patterns develop across proficiency levels in learner English and compares this with L1 patterns (using the British National Corpus, Burnard and Aston 1998). Overall, it is argued that, through a UB understanding of how language is acquired, DDL has the potential to bring an acceleration of language experience to the learner. However, the curation of language input needs to be sensitive to and differentiated for stages of learning. It is argued that UB theory coupled with CL analysis can aid this process