Abstract
The grammaticalization of verbs of ‘say’ into complementizers, subordinating conjunctions and other grammatical functions has not been well-investigated in the study of Taiwan Mandarin (TM). Most previous studies on the grammaticalization of shuo in Taiwan Mandarin mainly focus on the discussion of universal mechanisms, but seldom touch the issues about language contact and social factors. In addition, the historical track on which verbs of ‘say’ changed has not been well-studied. This dissertation aims to investigate the grammaticalization of shuo in Taiwan Mandarin. Shuo originally functioned as a verb; afterwards it also functioned as a complementizer; finally, it has come to have reportative function and can also be a pragmatic marker expressing such meanings as counter-expectation. Based on Heine and Kuteva’s (2006) integrative model of grammaticalization, we combine both language internal and external factors to account for the grammaticalization process of shuo in Taiwan Mandarin. On synchronic aspect, we compare the degree of grammaticalization of shuo in Taiwan Mandarin, Peking Mandarin (PM) and Taiwanese Southern Min (TSM), and get the rank：TSM＞TM＞PK, which indicates that shuo in TM should be accelerated by the external language contact with kóng in TSM. We also retrace the origination and replacement of saying verbs in Chinese history, and find a track of the development of verbs of ‘say’: yue □ dao□ shuo. Each of the verb of ‘say’ plays the leading role in different historical periods. Once the new verb of ‘say’ come into being, it follows the paths of the previous verb of ‘say’ and further develops its new usages. The historical retracing provides us the clue of some sequences for the internal grammaticalization process. In addition, our data-driven investigation employs instruments such as questionnaire, TV talks corpus and spontaneous spoken Mandarin Corpus. The purpose of the questionnaire mainly shows the language varies according to the social factors: age and mother tongue and confirms the validity of Givón’s (1980) binding hierarchy. In addition, the results of all the data we collected coincidentally show that the frequencies and degree of acceptance of shuo as pragmatic markers are much lower than those of complementizers. As a complementizer, the frequencies and degree of acceptance of shuo co-occurring with connectors is much more than those of shuo with speech act verbs or cognitive verbs. As a pragmatic marker, utterance-initial shuo has higher frequency and degree of acceptance than utterance-final shuo. The grammaticalization processes of shuo as a complementizer and as a pragmatic marker originate from the same source but diverge at a later stage. The route for shuo to evolve into a complementizer only follows some parts of the hierarchy posted by Givón and it seems to have reached the ultimate stage. On the other hand, the route for shuo to grammaticalize to a pragmatic marker is the combination of language universal and language contact, especially on shuo as an utterance-final pragmatic marker. Unlike shuo as a complementizer, pragmatic marker shuo has not reached a mature stage and the usages are not as productive as shuo as a complementizer in Taiwan Mandarin.