TTS, ASR, MT and the lot in EFL pronunciation training - a state of the art demo

Włodzimierz Sobkowiak
IN PWSZ Konin, IFA UAM

My Soczewka 2003 presentation is a simple demo of the state of the art in speech engineering used in the context of EFL pronunciation/phonetics training. I also offer some thoughts on the pedagogical issues involved, such as the viability of speech-enabled software in the didactic context, its consequences for phonetics teaching and learning, the main problems encountered and the pedagogical potential to be exploited. I briefly touch on Automatic Speech Recognition (ASR) and evaluation, as well as speech-to-speech Machine Translation (MT), but I concentrate on text-to-speech synthesis (TTS). Its contemporary quality is demonstrated, its feasibility as an EFL pronunciation model defended, the main benefits pointed out, and some possible applications and developments discussed. The perspective is thoroughly pedagogical and practical, rather than technological.