Naturalness of interjections produced by German learners of Japanese

This study investigates the naturalness of utterances containing interjections produced by German learners of Japanese. Ten participants with beginner-level proficiency and limited experience living in Japan took part in a pronunciation experiment. They were presented with conversational scenarios in both Japanese and German and asked to produce the sentence “<interjection>, Manami-san” with appropriate emotional expression. Two native Japanese speakers performed the same task for comparison. Acoustic analysis revealed a moderate correlation between the duration of the interjection and the following phrase (r = 0.59, p < 0.01), and a strong correlation between the maximum pitch of the interjection and that of the following phrase (r = 0.85, p < 0.01). Subsequently, an impression rating task was conducted in which 10 native Japanese speakers evaluated 150 recorded utterances on a 5-point scale (1 = very unnatural, 5 = very natural). Utterances matching native pitch patterns in both the interjection and the following phrase received the highest average score (M = 4.37). When only the interjection matched, the average score was 3.60; when only the following phrase matched, it was 3.84. These findings suggest that the naturalness of learners’ speech depends not only on interjection prosody but also on the prosodic features of the subsequent phrase.