A Quantitative Analysis of the Utterances by Japanese Learners of English at Different Proficiency Levels and English Native Speakers in a Persuasion Role-Play: A Study Based on the Spoken Dialogue Module of the International Corpus Network of Asian Learners of English (ICNALE)

Using the data taken from the Spoken Dialogue Module of the International Corpus Network of Asian Learners of English (ICNALE), the current study analyzed the utterances by Japanese learners of English (JLEs) at four different proficiency levels and English native speakers (ENSs) in a persuasion role-play. Our corpus-based quantitative analyses have shown that (1) the quantity of speeches by JLEs is 10-20% smaller in terms of the number of turn takings and 10-35% smaller in terms of the number of tokens in comparison to that by ENSs, (2) JLEs use the second-person pronoun you much less than ENSs and use the words of contrast and exaggeration in a different way from ENSs, (3) JLEs overuse the non-lexical fillers and want, while they underuse the lexical filler well, the hedges such as like and the epistemic would and could, the second-person pronoun you, the amplifier really, and varied basic words often used in set-phrases, and (4) JLEs and ENSs are classified into four clusters (novice JLEs, lower-intermediate JLEs, upper-intermediate JLEs, and ENS), and these clusters are characterized by different sets of words and they are in accordance with four stages in JLEs’ L2 acquisition process.