Chinese-style topics as indexicality

'Topic' is one of the most studied and the least understood subjects in Chinese linguistics. One major problem is the so-called 'Chinese-style topics/dangling topics'. Shi (2000) was the first to establish a typology of Chinese-style topics. Later studies were primarily concerned with the validity of his typology (Huang & Ting 2006; Pan & Hu 2002, 2008) and with how Chinese-style topics, if they exist, are semantically licensed (Hu & Pan 2009). More problematic and less discussed is the question as to how Chinese-style topics are syntactically derived. Based on previous studies and new tests, I argue that Chinese-style topics do exist, although not only in Chinese and not all Shi's six types are Chinese-style topics. I only identify Shi (2000)'s types 3 and 4 as Chinese-style topics, contrary to the conclusion of all previous studies. Furthermore, I argue that the Chinese-style topics which I identify share properties which non-Chinese-style topics do not have, namely Chinese-style topics necessarily or preferably stand before other topics and do not show Weak Crossover and Relativized Minimality effects. To explain these properties, I adopt Giorgi (2010)'s Indexicality Hypothesis and propose that Chinese-style topics, which have the interpretable [iDeictic] feature, sit at the specifier of the C-SpeakerP at the leftmost layer of the CP. This approach can shed new light on the famous dichotomy, that of topic-prominent languages vs. subject-prominent languages (Li & Thompson 1976).