Abstract
The thesis investigates the adverbials in Squliq Atayal which is spoken in Chian-shih, Hsinchu County. Following Cinque (1999), we attempt to establish the relative order of Squliq adverbials. The first two chapters are introduction for a sketch of Squliq Atayal and literature review. Chapter 3 is discussion for adverbials in Squliq. Chapter 4 discusses temporal adverbials and topics related to subject-sensitivity. We divide the adverbials into two groups. One group is composed of predicate-like adverbials, which behave as a predicate in a sentence; the other group is made up of adjunct-like adverbials, which behave more like adverbs in English and Chinese and are assumed to have hierarchical order as claimed by Cinque (1999). The predicate-like adverbials are further divided into two parts, Type-I: stem + voice marker (e.g. m-k’ial/ pk’ial-un (violently)) and Type-Ⅱ: si + (p-) stem (e.g. si p’siang (quietly)). In addition, we use a four-layer structure, which is inspired by Rizzi (1997), to help demarcate the relative order for those adjunct-like adverbials. Furthermore, we find that temporal adverbials, like suxan ‘tomorrow’ and knwan ‘when’ in Squliq, form counterexamples in those syntactic operations, such as cleaving, topicalization, questioning, which are claimed to be subject-sensitive. Chang (1997) proposes voice-sensitive to explain the phenomena. However, his analysis cannot provide complete explanation for those adverbials base-generated under VoiceP. On the other hand, Pesetsky & Torrego (2001) claim that Nominative Case is an instance of an uninterpretable Tense feature (uT) on D. Following them, Tsai (2004) hypothesizes that temporal adverbials, just like nominative arguments, bear Tense features. Sentences with marked information structures (e.g., topic, focus, question, etc.) have two options to check off uT on C: One is to raise a nominative argument with uT further from [Spec, TP]. The other is to raise a temporal adverbial with iT. In this way, a temporal expression enjoys the freedom of either remaining in-situ, or raising to [Spec, CP]. Thus, Tsai’s (2004) hypothesis is a better account for it explains temporal adverbials both above and under VoiceP.