Reserches in CCL related to the Japanese language first started in 1984 with a machine translation project from English into Japanese funded by the British government and then ICL(now Fujitsu-ICL). Prof.Jun-Ichi Tsujii joined the Department as head of CCL in the autumn of 1988 from the University of Kyoto. Since then CCL has implemented a number of research contracts from Japanese organisations such as ATR, EDR, Matsushita, Oki ,Sharp, etc. Curretly CCL has research collaboration with the 'Shin-puro' project of the National Language Research Institute. We closely work with Mr Hiroshi Nakano. Specifically we are investigating how and to what extent the quality of English written by Japanese can be improved. The nature of mistakes change from simple grammatical errors to more sophisticated irregularities relating to nuance and style according to how advanced Japanese writers are in English language proficiency. You can find many books on common grammatical mistakes made by Japanese, but there are very few books which analyse mistakes, as these become more 'advanced'. Therefore what we would like to find out is what makes the English written by Japanese authors still sound odd even when those authors have otherwise attained a high degree of competence in English.

We have selected some Japanese informants with the good command of English and asked them to tranlate short passages from Japanese books into English. Their tranlations are subsequently checked by an English native speaker. We will then analyse the tranlation carefully against the folowing criteria:

1)word choice
The wrong choice of word can be made at least in two ways: either wrong in meaning or wrong style. Two words can be synonymous at the lexical level, but only one of them may be suitable in a specific context.

2)tense and aspect
How accurately tense and aspect are represented in the translations.

3)stylistic coherence
Stylistic changes may happen at the sentence level as well as at the lexical level. We plan to examine stylistic irregularities at the sentence level.

4)unusual syntactic structures
A Japanese writer may construct English sentences influenced by the structure of the Japanese sentence.

5)flow of the text
Sequencing within texts is presented in a different way between Japanese and English. For example, an English text may require logical connections which are not shown in Japanese.

Having performed careful analysis of the tranlations against these criteria, we are actively consideing the possibility of narrowing our main focus down to tense and aspect. The Japanese language has different mechanisms for indicating tense and aspect from English, and yet both languages are cleary capable of describing the same event from the same temporal point. There is therfore a case for investigating the nature of linguistic clues in Japanese which influence Japanese autohrs' representation of tense and aspect in the producion of texts in English.

January 1996

ホームページ / インデックス / 前ページ / 次ページ