A systematic literature review of automatic speech recognition in L2 learning
Danial Mehdipour-Kolour, Walcir Cardoso · 2023
Despite the increasing popularity of using speech as a text input method, there is surprisingly little research on the use of automatic speech recognition (ASR) to aid in writing. This lack of attention is particularly notable given the recent advancements in ASR technology. This chapter systematically reviews the empirical literature on the potential and limitations of ASR in second/foreign language learning in general, with a secondary emphasis on second language writing. Including studies published between 1990 and 2021, our literature search identified 79 journal articles that explicitly employed the technology for second language pedagogy ( n = 64) and as a writing tool for native speakers ( n = 15). The studies were thematically analyzed using two coding methods (inductive and deductive). Our results indicate that automatic speech recognition has been mostly used as a pronunciation tool ( n = 33), not for developing second language writing. Findings also reveal some benefits (e.g., ASR promotes a process-oriented approach to writing) and limitations (e.g., ASR increases the complexity of text formatting). Our discussion highlights the importance of these findings and suggests directions for future research on the implementation of ASR in second language pedagogy, particularly for writing development.