Volume 2000 (2000) Issue 117 Pages 71-100
The aim of this paper is to explore the verbalization pattern of subject, especially that of personal pronoun subject, in Japanese discourse. Wallace Chafe (1972, 1974) remarks that the concepts of speaker and hearer are ‘given’ and are pronounced with weaker stress and lower pitch. However, the Japanese equivalents of ‘I’ and ‘you’ are not always omitted.
In this paper, we shall argue that there are four types of subject in the light of its semantical function, and that this semantical function controls the verbalization vs. non-verbalization of the subject. The four types of subject are: 1) Topic of absolute contrast (Tac) / Nominative of complete exclusion (Nce), 2) Topic of relative contrast (Trc) /Nominative of partial exclusion (Npe), 3) Topic of logical uniqueness (Tlu) / Nominative of logical uniqueness (Nlu), 4) Topic of absolute uniqueness (Tau) / Nominative of absolute uniqueness (Nau).
In the case Tac/Nce, the subject has to be verbalized. On the other hand, for the case Tau/Nau, the subject is normally not verbalized.
Trc/Npe are divided into two types according to their predictability. When the subject is predictable through the factors below, it is omittable. When it is not, it must be verbalized. The factors which make a subject predictable are ‘Speaker-Hearer Predicates (SHP)’, ‘Shared Information’ and ‘Repetition or Substitution’. SHP indicate certain types of predicates which demand the first person or the second person subject. 'Shared Information' is the nformation shared between a speaker and a hearer. The third factor 'Repetition or Substitution' means repetition or substitution of something which has been uttered
previously in the discourse. If any one of these factors are present in the predicate or in the context, the hearer can predict the subject, and it can hence be omitted. But in the case that none of these factors are present in the predicate or in the context, the subject must be verbalized.
Tlu/Nlu are uniquely identifiable by shared knowledge between speaker and hearer or from context, and these subjects are omittable.