英文摘要 |
The traditional way of categorizing modal adverbs in Taiwan Southern Min is by means of semantic properties, such as the speaker's alethic, epistemic, boulomaic, deontic, or evaluative attitude towards the sentence proposition or the speaker's conjecture of the sentence possibility, probability, or necessity. This categorization is often diverse and captures little generalization. This research proposes a syntactic analysis based on the distribution and licensing assignment of modal adverbs. There are three types of modal adverbs: speaker-oriented, epistemic, and deontic adverbs, and they always appear in the order of the speaker-oriented type precedes the epistemic type and the epistemic type precedes the deontic type in a sentence.According to Travis(1988), adverbs are a defective construction due to their lack of maximal projection and therefore cannot be licensed via theta role assignment or predication. They must be licensed by the head to which they are adjoined. Based on this, we propose that in modal adverbs, the deontic modal adverb is licensed by the head of PreP(Preo), the epistemic modal adverb licensed by the head of IP(Io), and the speaker-oriented modal adverb by the head of CP(Co). With this licensing principle and the modal adverb categorization, not only the co-occurrence restriction between the modal adverbs and their particular heads but also the linear order of these three types of modal adverbs in a sentence can be accounted for in a systematic way. |