| Budget Amount *help |
¥3,770,000 (Direct Cost: ¥3,500,000、Indirect Cost: ¥270,000)
Fiscal Year 2007: ¥1,170,000 (Direct Cost: ¥900,000、Indirect Cost: ¥270,000)
Fiscal Year 2006: ¥2,600,000 (Direct Cost: ¥2,600,000)
|
|---|
| Research Abstract |
The main purpose of this research is to obtain quantitative functional load of segments for distinction and to find its usefulness in psychological experiments. The investigator has proposed a model of functional load in the past and obtained the dataset of "neighborhood" (i.e., a set of words sharing all but one phoneme with the target word) for Japanese. Preliminary experiments based on the dataset were carried out this year, The chief advancement in the term of the grant is, however, an elaboration of the model The index of the functional load in the previous research has been based solely on the number of minimal pairs and phonemic frequency. The new model includes a revised index which takes the leftward context of the opposition as the shrinking factor of the search space. When there is a long context before the opposition, the role of that opposition is small due to the fact that the search for the target word is already narrowed down by the context. This is based on the notion of "cohort' proposed in Marslen-Wilson (1989). Using a familiarity database for Japanese Lexicon (Amano & Kondo 1999), a statistical analysis of the proposed functional load was done. The results show that phonemes even within a natural class, such as stops, have a quite skewed distribution of functional load. The dogma in structural phonology says that phonemic system of a Language has a coherent and balanced system of phonemes. However, the skewed distribution of phonemes suggests that the phonemic system is quite unevenly utilized in reality.
|
