Computer assisted foreign language teaching and learning by using speech technologies

Early works aiming at technical support of teaching and learning pronunciation of foreign language can be found in the 1980s, where technologies for speech analysis, synthesis, and recognition were examined. In the 1990s, with the help of statistical frameworks, speech recognition and synthesis technologies were greatly improved and CALL (Computer Assisted Language Learning) studies benefited much from these improvements. Since then, various CALL systems can be found in the market. Many of them had functions of scoring pronunciation and/or detecting pronunciation errors using ASR technologies. These days, not a small number of teachers use synthesized speech as model utterances in reading and shadowing practices. Using these advanced technologies, many examples of developing CALL applications are found in technical papers but fundamental studies are also still being done. Similarly to studies of speech recognition and synthesis, researchers’ attention is paid mainly to machine learning. In this paper, these CALL technologies are reviewed. Then, personal opinions on CALL are given and our recent research and development of CALL is introduced.