This study compares the use of Japanese hearsay evidentials:(-rashii 'it seems,' soo 'I hear') by native speakers of Japanese (NJs in L1 Japanese) and English speaking advanced learners of Japanese (NE learners in L1 English and L2 Japanese), when they report second-hand information. This study also examines whether NJs and NE learners differentiate the use of Japanese hearsay evidentials to whom they report hearsay information.
Ishida (1999) used Kamio's (1997) territory of information model to investigate the use of Japanese evidentials. He found that L2 learners whose L1 was English used fewer hearsay evidentials when conveying hearsay information in English and L2 Japanese than native speakers of Japanese in Japanese.
Ishida's study involved a sentence-level oral questionnaire. However, the data in the present study consists of 12 interviews with four informants. The informants were assigned a topic about a fire disaster in California beforehand. Then they were asked to report to addressees: casual friends and teachers, who did not know much about the disaster.
Findings show that the NE learners used fewer overt hearsay evidentials in both English and Japanese when compared to the NJs. The NE learners' low frequency of use could be attributed to their low use in English. This study also found that the NJs used hearsay evidentials more frequently when the addressee was a superior than when the addressee was a friend. When the addressee was a superior, the NJs may create more psychological distance from the second-hand information, due to a politeness factor. However, the NE learners did not show this sensitivity to the addressee variable. The use of hearsay evidentials shows the speaker's lack of precision towards the information, and may be functioned as a hedge. Such a sociopragmatic factor may affect the use of hearsay evidentials especially when the addressee is a superior