The adaptability of laboratory phonemic perception training protocols to common second language instruction situations

Abstract

For the past two decades, second language pedagogy has de-emphasized phoneme-level pronunciation training. During the same period, laboratory experiments have demonstrated a variety of benefits from training non-natives in the perception of difficult target language contrasts. Experiments have shown that learners’ perceptual performance improves (Jamieson & Morosan, 1986; Flege, 1989 & 1995), that the improvements generalize to new talkers and words (Lively, Logan, & Pisoni, 1993), that perceptual training triggers production improvements (i.e., without specific pronunciation training, Bradlow et al., 1997), and that both perceptual improvements (Flege, 1995; Lively et al., 1994) and production improvements (Bradlow et al., 1999) continue over several months. Protocols for these training experiments are generally intensive, and may include sessions of 40 min, three sessions per day, or sessions for 15 consecutive business days. Such onerous protocols are acceptable for basic research with compensated subjects. They are unacceptable for normal L2 instruction, with paying learners and with limited time for pronunciation training. The present study analyzes the training protocols from selected research articles, and documents the types and amounts of training. It then suggests acceptable alternative regimens for achieving comparable training in typical learning situations, including 16-week college semesters and shorter language institute terms.