According to the Unaccusative Hypothesis (Perlmutter 1978), intransitive verbs fall in two distinct classes: the <strong>unaccusatives</strong> (whose subjects originate as direct objects) and the <strong>unergatives</strong> (whose subjects originate as subjects). Although there are studies of split intransitivity in Modern Japanese and European languages, very few exist for earlier stages of Japanese. To fill in part of this gap, this thesis presents a comprehensive investigation of split intransitivity in Old Japanese (largely, 8th century Japanese). The descriptive and analytic work of this research is based on the newly developed ‘Oxford Corpus of Old Japanese’ (OCOJ). It consists of original and romanized Old Japanese texts, with a wide range of information including the original orthography, part-of-speech, morphology and syntactic constituency in the form of XML tags following TEI conventions. It is part of a larger collaborative research project: ‘Verb semantics and argument realization in pre-modern Japanese: A comprehensive study of the basic syntax of pre-modern Japanese’, in which my DPhil project is situated. As part of my DPhil project, I took part in the analysis and tagging of the OCOJ, in addition to contributing to translation. My original contribution to knowledge is a comprehensive investigation and in-depth analysis of the lexical-semantic aspects of split intransitivity in relation to its morpho-syntactic expressions in Old Japanese. This includes: exploring to what extent intransitive verbs could be classified as unaccusative and unergative, what factors are involved in the classification, how they interact, what are the possible ways of representation, and the theoretical implications it brings to linguistic theory in general. Syntactically, I looked into manifestations specific to Old Japanese (e.g. perfective auxiliary selection), and also examined to what extent diagnostics – which show split intransitivity in English, Italian and Modern Japanese (e.g. N+V compounding and resultative construction) – could be applied to Old Japanese. Semantically, I investigated various semantic factors and proposed basic and complex models of the interaction between intentionality and affectedness in Old Japanese. I also proposed a ‘complex format for representing simple event structures’ which enhances the understanding of semantic aspects of split intransitivity. As such, the results of my research not only contribute to a detailed understanding of Old Japanese verbs, but also have implications for linguistic theory in general.

abstract: This thesis offers a contrastive analysis of the causative alternation phenomenon in English and Standard Arabic variety. This phenomenon has received a lot of attention in the literature on argument structure. It has traditionally been presented in terms of the causativization of inchoative verbs/unaccusatives. It is argued here that this analysis conflicts with the way the causative alternation is molded in Arabic. Causative alternation in Arabic is not only limited to inchoative verbs, but it incorporates unergative verbs as well, which play a vital role in this alternation. The implication of this observation is that the different syntactic behaviors between English and Arabic may reflect people’s different perception of events and lead to different syntactic computation. Therefore, this thesis highlights the role of this subset of intransitives/anti-causatives in the Arabic causative alternation and answers one of the highly considered questions on the causative alternation; that is, which version of the alternation is the lexical base, and which one is derived? This thesis also reveals that there is some significant difference between English and Arabic in terms of the alternatability of unaccusative verbs. Therefore, this study shows that most of the Arabic unaccusative verbs, except denominal verbs, have a causative alternant. This thesis also addresses the vital role of the Arabic verbal template in clarifying this phenomenon. In sum, this thesis provides an overview of the semantic, syntactic, and morphological properties of Arabic verbs undergoing the causative alternation. Besides employing the researcher’s native-speaker intuition, the English/Arabic Lexicon Dictionary and Arabicorpora are consulted to support the validity of the argument. / Dissertation/Thesis / Masters Thesis English 2015

本論文是以Kellerman (1978) 的U型學習理論為基礎做擴大研究。本篇論文旨在探討英文程度不同的中文為母語之人士，在學習非賓格不及物動詞時是否也會出現U型曲線，除此之外，非賓格不及物動詞與非賓格及物動詞，被動態，主詞生命度之間的互動也將做討論。 在這個研究中，我們採用語法判斷(Grammaticality judgment)來測試受測者對於非賓格不及物動詞的理解和中英轉譯(Chinese to English translation)來測試受測者對於主詞生命度和主被動態之間的影響。此研究共有123位受測者，他們根據學習英文的長短被分為四個組別，分別是低、低中、中、與中高程度。 此研究可歸納為以下結論。(1) 不同英文程度的中文為母語之人士，在學習英文非賓格不及物動詞時也會出現U型曲線。此即意味著U型曲線不僅僅出現在母語為荷蘭語學習非賓格及物動詞Break的學習上，更可擴大到母語為中文學習非賓格不及物動詞上。(2) 中文為母語之人士無法正確使用非賓格及物動詞，並且會把非賓格及物動詞當作非賓格不及物動詞。(3) 在學習非賓格不及物動詞中，主詞生命度確實會影響學習者使用主被動態之不同。當主詞有生命時，句子傾向使用主動態，當主詞是無生命時，句子傾向使用被動態。 / The study is based on Kellerman’s (1978) U-shaped leaning on break to do further study. The study aims to examine if the learning of non-alternating unaccsatives for L2 Chinese learners of English with different proficiency presents a U-shaped curve. In addition, the interactions among alternating unaccusatives, non-alternating unaccusatives, passives and animacy effect are discussed as well. In the study, we use grammaticality judgment task to test participants’ understanding of non-alternating unaccusatives, and adopt Chinese to English translation task to test animacy effect in non-alternating unaccusatives. 123 participants involve the experiment of the study. Among these participants, they are classified as four groups, low, low-intermediate, intermediate, and high-intermediate, according to how long they studied English. The results of the study are summarized as follows. (1) There is a U-shaped curve in learning of non-alternating unaccusatives for L2 Chinese speakers learning English. It suggested that U-shaped learning is not only in alternating unaccusatives break in L1 Dutch but also in non-alternating unaccusatives in L1 Chinese. (2) For L2 learners, they are unable to use alternating unaccusatives correctly and tend to view alternating unaccusatives as non-alternating ones. (3) Animacy effect does influence the choices of voice forms. The study showed that participants tend to use active voice while the subject is animate and prefer to use passive voice while the subject is inanimate.