This study aims to investigate the translation procedures adopted for the English translation of Chinese empty words. To begin with, the methods which translators adopt to tackle empty words are identified by examining examples in a specially-constructed parallel corpus, which includes Chinese literary texts and their English translations. Eventually, eight translation procedures (1) Match; 2) Paraphrase; 3) Shared Match; 4) Implicitation; 5) Amplification; 6) Grammatical Conveyance; 7) Borrowing; 8) Omission) and one non-procedure (Mismatch) are identified. It is noted that Grammatical Conveyance is a procedure which could be deemed as a newly-identified method. As a further step, the proportion of these procedures/non-procedure is investigated to identify the most-/least-adopted ones (Match and Amplification respectively) and to discuss category-specific ones (Shared Match and Borrowing). In addition to identifying the procedures for translating Chinese empty words into English, this research also makes contributions in the following two aspects. Firstly, this study, to my knowledge, is the first research which examines all Chinese empty words at a time to identify the ways translators tackle them. Secondly, it is known from the results that difficult-to-tackle empty words are found in the categories of Adverb and Particle. In other words, not all empty words are difficult to translate as former scholars have described.

Chinese does not allow causative alternating unaccusative verbs or object-Experiencer psych verbs. It employs analytical causative (headed by shi 'make'), resultative (headed by de) and compound causative (with an activity predicate V1 and a result predicate V2 forming a V1-V2 compound) constructions to represent causativity. The present study explores syntactic and semantic properties of these verbs and constructions within the generative framework. It is proposed that Chinese unaccusative verbs and psych verbs involve a single VP and that Chinese resultative and compound causative constructions involve a functional category AspP. An empirical study is conducted to look into the mental representations of these Chinese verbs and constructions in English-, Japanese- and Korean-speaking learners' L2 Chinese grammars. The aim of the research is to find out whether Ll transfer persists in L2 Chinese and whether L2 learners can acquire the functional AspP involved in Chinese resultative and compound causative constructions. The experiment consists of a cloze test, a production test, an acceptability judgment test and a comprehension test. It involves 55 English speakers, 56 Japanese speakers, 73 Korean speakers and 28 native Chinese as controls. Both developmental patterns and variations between different language groups are examined to see whether L2 Chinese learners can acquire native-like mental representations of the structures examined and whether L2 groups from different Lls show any variation in their L2 Chinese mental representations. The results suggest that L2 Chinese learners are more likely to make causative errors with Chinese unaccusative verbs than with psych verbs and that compound causative constructions are more difficult to acquire than resultative constructions in L2 acquisition of Chinese. It is concluded that Ll transfer does not happen everywhere and that functional categories unavailable in the learners' Lls can be properly represented in L2 Chinese grammars.

In this thesis, I discuss the grammatical constructionalization involving BA in some depth. I aim to present how a constructional approach can be used to rethink diachronic change in Chinese Expressions involving BA. In previous research of relevance (e.g. Sun 1996; Xing 2006), BA is considered to be a single item that has undergone grammaticalization. From a constructional perspective (e.g. Goldberg 1995, 2006; Traugott 2003a; Trousdale 2008a, 2010; Traugott and Trousdale forthcoming), this view requires some revision. I attempt to provide a constructional account in which an Expression involving BA is understood as a “construction” at the micro-level. As constructions change over time, the form-meaning pairings involving BA are associated with more schematic higher-level constructions. I show the constructionalization process in which an Expression involving BA is reanalyzed and incorporated into different constructions with increased schematicity at higher-levels. Therefore, I assume that a network of correlated constructions is implicated. The diachronic research is conducted mainly based on data collected from the BA corpus, which is a subcorpus I compiled based on the Peking (CCL PKU) Corpus. Using the constructional framework, I also analyze Expressions involving BA in Contemporary Chinese. The aim of the synchronic research is to explore how an Expression involving BA is used currently, and to associate the synchrony with diachrony in some respects.