Abstract
This study examined the effects that communicative messages' relevance to speaker vs. hearer has on usage of sentence-final expressions. The expressions studied were indirect forms such as sodesu, yodesu, mitaidesu and the particle ne. In Experiments 1 and 2, subjects were asked to write what they would say in certain situations which were described to them. When the speaker alone supposedly knew what he/she was to communicate (Experiment 1), use of indirect forms increased when the message was relevant to the hearer but not to the speaker. The particle ne was rarely used in such a situation. When both the speaker and the hearer assumedly knew what was to be transmitted by the speaker (Experiment 2), both ne and indirect forms were used more often when the communicative content was relevant to the hearer than to the speaker. Experiment 3 demonstrated that direct forms were judged to be more natural when used for important than for unimportant topics. A model for differential usage of sentence-final expressions in the Japanese language was proposed.
