Finally, for pronunciation materials writers, teachers, and teacher trainers, we call attention to the following points. First, theJapanese flap, a segmental, appears to be linked to Japanese speakers' accent in English. If the goal is to reduce the Japanese accent in English, then a focus on reducing the transfer of this flap and improving accuracy in pronouncing /i/ and /1/ would be a sensible endeavor, especially if, as was seen in this study, flap substitution does not diminish easily or
quickly over time. Second, because flap substitutions may vary widely across speakers, and because individuals' EFL speech develops in different ways, pronunciation pedagogy may need to become more individualized. If individualized instruction is impossible, a partial solution may reside in self-access labs and increasingly effective computer software applications that allow students to work independently on individual problems (see, e.g., Lambacher, 1999, for the teaching of /i/ and /1/ to Japanese). Third, although it is unrealistic to expect most students to improve quickly, teachers should recognize that at least some students manage to reduce their L1-L2 transfer and improve their segmental production and
global foreign accent over time.