Subjectivity in Japanese: A Corpus-Linguistic Study

Abstract

This paper provides a corpus-linguistic study on subjectivity in Japanese, in an effort to arrive at how subjectivity, transitivity and grammaticalisation are related. 899 lexicons from nine grammatical categories (suffixes and prefixes, adjectives, particles, auxiliaries, nouns, adnominals, adverbs, and transitive/intransitive verb pairs) are examined. The findings reveal that Japanese is a subjective/objective-split language, and that subjectivity in affixes is facilitated by phonology: voiced/voiceless consonant alternation. The data also show that consonant-voiced prefixes and suffixes yield a subjective reading, while consonant-voiceless prefixes and suffixes render an objective meaning. Split subjectivity in adjectives is realised by morphology: しい-ending adjectives tend to be subjective, while い-ending adjectives are mostly objective. The differentiation of subjectivity in adjectives is further tied to the constraints on personal pronoun and verbalisation possibilities. Intriguingly, objective/subjective readings of しい-ending adjectives andい-ending adjectives are switchable. Furthermore, among transitive/intransitive verb pairs, intransitive verbs are likely to get grammaticalised, while transitive verbs are likely to be lexicalised and thus render a subjective reading. This is confirmed by change-of-state verbs and motion verbs. This paper therefore puts forward the hypothesis that the interrelationship of grammaticalisation and lexicalisation is orthogonal.