Volume 33 No 1
Yeshes Vodgsal Atshogs 意西微萨 阿错
This dissertation analyzes two language facts. At the very beginning, based on the newly-discovered language —Daohua, this dissertation intends to analyze the deep-contacting relationship between Tibetan and Chinese from the mixing complicated levels of phonetic system, lexical system, and syntax. A typical feature of the complicated levels is the heterogeneous-conformity mechanism of lexical and syntax system. This kind of situation makes us think of that, in the historical relationship between Tibetan and Chinese, there is a kinship relation represented by the basic vocabulary, as well as there is a great typological difference. Also, it let us bethink the famous viewpoint proposed by Mantaro J. Hashimoto (1985): Tibetan is related with the Sino-Tibetan family mostly in vocabulary, but in syntax it is more close to Altaic. Based on this, then the dissertation has farther analyzed the language fact related to the viewpoint proposed by Mantaro J. Hashimoto. When we compare the fact of deep-contacting mechanism of Daohua with the correlative fact of the famous judgment proposed by Mantaro J. Hashimoto, we can find there is an astonishing similarity between the two facts above. 本文主要包括两个方面的语言事实分析。首先是以最新发现的一种特殊语言一 “倒话”为例，分析藏、汉语言在深度接触中，从音韵、词汇到语法的复杂混合层次.这种复杂层次中的一个典型表现是词汇系统与句法系统异源的整合机制.这种情形让人想起藏、汉语言历史关系中，一方面是基本词汇的语音对应所体现的语言亲缘关系，一方面是语法结构类型的巨大差异； 想起桥本万太郎先生 (1985) 曾说道 “藏语在词汇方面和汉藏语系大有关联，然而在句法方面更像阿尔泰语” 的著名论述. 于是，又进ー步分析了与桥本先生这ー论述相关的语言事实. 当把现实语言深度接触复杂机制，与藏、汉语言特殊历史关系加以比较吋，则可以看到两类事实之间惊人的相似性.
In this study, I investigate how Mandarin Chinese speakers manage the information flow of nominal referents across argument roles in different types of discourse, in trying to understand the relationship between Chinese discourse and the shape of its grammar. One conversation and two types of oral narratives are taken as distinct types of speaking for purposes of examining the potential text-type difference against information status. The unique and recurrent patterning of information status across conversation and narrative discourse suggests a pragmatic motivation for word order. The preverbal roles, be it A, S or O, have a great propensity for given information; the postverbal roles, either S or O, maintain a much higher incidence of new information. This lack of accusative or ergative alignment of argument roles has a parallel in grammar where the major syntactic processes treat A, S and O in the same way, suggesting an iconicity between discourse and grammar in Mandarin Chinese.
Information flow Chinese discourse Text-type difference Information status Iconicit
试论所谓的标志动态的词素 "既 " , "已" 的句法和语义
Barbara Meisterernst 梅思德
This paper deals with the question whether the so-called aspectual markers ji 既 and yi 已 in Classical and Han period Chinese are genuine markers of the perfective aspect. The discussion is not based on the historical evolution of both markers but on their syntactic distribution in the sentence and on their semantic interplay with the verb they modify. We hypothesize that syntactically ji 既 and yi 已 are genuine adverbs serving to modify the aspectual value of the whole VP and not aspect markers serving to indicate the aspect of the verb comparable to the aspectual suffix -le 了 in Modern Mandarin. A comparison of unmarked and marked VPs clearly demonstrates that in Han period Chinese a verb phrase does not have to be modified by one of the two morphemes to express perfectivity or related notions. Depending on the situation type (Aktionsart) of the verb or the verb phrase, different aspectual values can be expressed without any obvious marking. Adverbs generally serve to emphasize particular nuances of the VP and the adverbs ji 既 and yi 已 serve to highlight nuances such as the factual occurrence of an event or a state, frequently a state that results from a preceding event. They are rather comparable to aspectual adverbs such as yijing 已经 ‘already’ in Modern Mandarin or ‘already’ in English.
本文想讨论“既”，“已”在汉代汉语里是否为标志动态词素的问题 。这研究不建立在“既” ， “已”的历史发展上。而以它们在句法和语义中的功用作为研究基础。作者假定 “既” ， “已” 在语句构成上是真正的有修饰VP的功能的副词，不是标志动态的助词，不可以和现代汉语指示完成的动态词尾 “-了” 比较。 作者想说明在汉代汉语中动词的动态无必要有标志。根据动词的语义在汉代汉语中没有标志的动词可以表达或动 作完成 ，或动作进行，或事情的状态 。因为修饰VP的副词一般有强调VP意义的功能， 所以作者认为副词 “既”，“已” 的主要功能首先是着重指示动作的完成或成果或结果的状态 (resultant state)；其次是着重指示动作或事实状态的发生。我们认为“既”，“已”可以和现代汉语中的副词“已经”，英语的副词 “already”相比 。 不同于动态词尾“-了”。
Classical Chinese 古代汉语 Han peiriod 汉代 Perfective aspect 完成式 Situation type 动态 Adverb 副词
Hui-chuan Hsu 许慧娟
As a departure from Cheung’s (1986) rule-based study on syllable contraction in Cantonese, the present analysis employs the notion of constraint interaction in Optimality Theory (Prince and Smolensky 1993) to solve the critical issue concerning the prediction of the output nucleus. Distinct from the sonority-oriented dialects such as Taiwanese Southern Min, Hakka, and Taiwan Mandarin, Cantonese is characterized by the determination of output nucleus by vowel length. In Optimality Theory, language variation appeals to different ranking of a set of universal constraints. This paper also illustrates that as a product of fast speech, syllable contraction may not follow language-specific phonotactic constraints.
Syllable contraction 音节缩减 Optimality theory 优选理论 Phonotactic constraint语音配置限制
Chaofen Sun 孙朝奋
This paper is an effort to identify the etymology that makes possible for the Early Modern Chinese verb zhuo, written as 著 or 着, to function as a causative verb in a serial-verb (or periphrastic) construction. Although it is no longer a productive causative construction in modern standard speech, it is still widely used as such in Early Modern Chinese and continues to be used in some formal writing in modern time. The meaning of indirect causation is found to be derived from, or related to, zhuo’s original, verbal meaning “to place” in Old Chinese. Both metonymization and metaphorization, as are defined within the theoretical framework proposed by Traugott and Dasher (2002: 4), are involved in the relevant historical changes. The former gives rise to the verbal meaning of “to use” in Middle Chinese, which, in turn, leads to a number of metaphorical processes that ultimately produce the Early Modern Chinese causative meaning. The initial intuition for such a hypothesis comes from a somewhat formal Modern Chinese verbal compound shi-yong 使用 “to use”, which is comprised of two morphemes shi “to cause” and yong “to use.” The Chinese story reported here then adds to those observed by Heine and Kuteva (2002: 328) still another lexical source, i.e., “to use,” from which meaning of causation may arise.
本文旨在调查“着”字在早期官话联动式中作使动动词有关的历史。“着”字这样的非直接使动用法，虽然在现代汉语口语中几乎已经绝迹，在现代汉语书面语中仍然经常可见。调查发现，该使动义与其古汉语中的放置义有着刀切莲藕斯不断的关系。在历史演变中，Traugott 和 Dasher（2000: 4）理论框架里的两个演变机制，语义引伸（metonymization）和隐喻（metaphorization）都起到了很大的作用。首先，“着”字从较基本的动词性放置义引伸出中古的使用义，然后又由于中古的“使”与“用”语义相通，“着”字又通过“使”而取得了使动义。“使”和“用”在中古的内在连系还反映在现代汉语“使用”这个复合动词上。本文的发现更为Heine and Kuteva (2002: 328)所列出的使动词可能源头又增加了一个新的可能。
Zhuo 著 Shiyong 使用 Causative 使动 Semantic change 语义变化 History of Chinese 汉语历史
Volume 33 No 2
Ivan Chow 周允文
An acoustic study was conducted to investigate the prosodic devices used by Mandarin speakers to resolve temporary syntactic ambiguity in speech production and perception. It was set in the fields of phonology and experimental phonetics – the study of the manner in which the relationship between phonological units at various levels affects the production of the combined string in the spoken form; and the physical and mathematical measurement of the spoken language in experimental conditions. Two separate groups of native Mandarin speakers were asked to read and listen to specially designed syntactically ambiguous sentences with two possible interpretations. Sequences of homophonous syllables were embedded in test sentences within which the interpretations were derived by placing prosodic boundaries at different syntactically motivated locations. According to two different syntactic structures imposed on the sequences, speakers were asked to disambiguate the sentences by conveying and recognizing the prosodic boundaries. Experimental results showed that, in addition to pauses and pre-boundary lengthening, pitch reset (a melodic juncture marker) was also an effective device for the resolution of temporary syntactic ambiguity in both speech production and perception. However, the quantitative identity of the prosodic foot (disyllabicity or trisyllabicity) was not a significant factor in the parsing process.
This paper investigates the suo construction in Classical Chinese and attempts to provide a proper syntactic analysis of it by comparing it with its modern Chinese counterpart. I extend and modify Ting’s (2003a) analysis of the modern suo construction to account for the Classical suo construction. Like its counterpart in modern Chinese, the Classical suo is a clitic in overt syntax, raising from N0 to I0 in overt syntax. This explains its fixed position with respect to other elements in the clause, as well as the fact that it may stand for grammatical object, location, but not grammatical subject. The reason why it may stand for manner and reason as well as grammatical object of a preposition, in contrast to the modern suo, is due to different categorial status of coverbs in modern and Classical Chinese; namely, that modern Chinese prepositions are in fact verbs in Classical Chinese. Unlike the modern suo, which is a variable, the Classical suo undergoes further movement from I0 to C0 at LF to fulfill its operator status. I argue that this explains why suo is optional in modern Chinese, but obligatory in Classical Chinese. This analysis echoes the conventional wisdom, which may be traced back to Ma (1898) that the classical suo is a relative pronoun. But crucially I argue that suo is such a pronoun, not in the overt syntax, but at LF.
Ren Zhang 张韧
This paper focuses on the argument structure properties of Possessum-object constructions in Chinese, as illustrated in Xiao3wang2 si3le fu4qin ‘Xiaowang dieLE father’ (Xiaowang had his father die on him). Three issues are discussed from the perspective of conceptual structure and semantic unification within the framework of Conceptual Semantics (e.g. Jackendoff 1990, 2000), drawing on Nunberg’s (1995) work on predicate transfer. First, the systematic relation between such constructions and their alternants such as Xiao3wang2 de fu4qin si3le (Xiaowang’s father died) is argued to be better captured by conceptual inference rules, rather than syntactic derivations. Second, the possessive relation between Xiao3wang2 and fu4qin ‘father’ is encoded as a conceptual binding relation within conceptual structure representations. Third, the predicate expression in such examples must designate a Property-based State or an inchoative Event, in order to be compatible with the conceptual structure of the construction as a whole. This need is facilitated by the particles le and guo, which serve to adjust the lexical conceptual structures of verbs. In addition, the NP within the predicate of the construction must show some functional need (Pustejovksy 2001), which accounts for directionality restrictions on predicate transfer (*zhe4wei4 fu4qin si3le Xiao3wang2/hai2zi ‘this father dieLE Xiaowang/child’ (This father had Xiaowang/the child die on him.)).
本文从概念语义学 (Jackendoff 1990, 2000) 的角度发展 Nunberg (1995) 的述谓延伸理论，在此基础上研究属格作宾结构 （如：小王死了父亲）的论元结构。作者提出这一结构与其变体（如：小王的父亲死了）的关系应由概念推理规则处理，而非句法转换操作。结构所含的所属关系是一种概念约束关系。结构的谓词须表达一状态或状态的变化，并含一特征 (Property)，以便与结构的总体意义一致。这一要求通过‘了’和‘过’对动词概念结构的调整而得到满足。述谓中的名词短语含某种功能需要，不具指称独立性，这解释了述谓延伸的方向性限制，如不能说，*这位父亲死了小王/孩子。
Meichun Liu 刘美君
This paper argues for the inclusion of constructional inferences in semantic representation of verbs. By examining the transitive pattern of the Mandarin verb GǍN (趕) ‘rush’, it is found that verbal semantics can only be adequately represented if constructionally coerced information is taken into consideration. The construction [GǍN + NP] renders specific interpretations that cannot be directly derived from the lexical meaning of either the verb or the object NP. The construction itself carries salient information for the appropriate interpretation. The ‘ellipsed’ eventive information is analyzed from the perspective of Construction Grammar (Goldberg 1995). Moreover, a compositional framework based on Qualia Structure (Pustejovsky 1995) is utilized to resolve the potential ambiguity arising from the construction-triggered meaning. By combining the two complementary approaches, this paper attempts to account for and represent contextualized information as well as varied meaning facets associated with a partially-filled construction.
Shin Yong Robson 容新
This paper discusses the temporal relations that the particle le reflects in Mandarin sentences. While the verb suffix le1 and the sentence final particle le2 respectively represent the different phases of the perfective-inchoative aspectual opposition, these aspectual views predict different relations between E(vent)-time and R(eference)-time of a situation. Le1 marks the perfective aspect of an actually occurred situation. E-time of this situation is either anterior to or included in its R-time. Le2 indicates the inception of a newly developed state; R-time of the described situation is always included in its E-time.