Analysis of the effects on pronunciation of training by using song or native speech

Abstract

This research is an investigation of whether Japanese speakers’ English pronunciation improves more after training on sung or spoken speech. The stimulus was a 14-word sentence taken from one English song's lyrics, and it had some words that are difficult to pronounce for most Japanese English learners. Thirty Japanese learners of English were recorded before training. Then, half of them trained by listening to the English song and singing it, and the other half trained by listening to a native speaker speaking the lyrics. Each group was allowed to train individually for 10 minutes, and then were recorded again. Then, 15 native or near-native English speakers at an American university and 100 native English speakers from Amazon Mechanical Turk evaluated those randomly-presented recordings. Listeners gave points for various phrases’ pronunciation, the overall accent and the overall intonation. As a result, we found out that training by using the music condition resulted in generally worse results than the regular speech training. In addition, perhaps surprisingly, intonation of the whole sentence had an additional significant negative effect following music training. These results seem to show that training by using regular speech is more effective for English learners than training by using songs.