Volume 30, No 1
Ting Pang-Hsin 丁邦新
“Special words” is a term used by Professor Chao Yuen Ren to denote words that are pronounced in a special way, either deviating from the main phonological development or being different from the recordings of the Qieyun, Guangyun or Jiyun. To some extent, we may consider these words having irregular pronunciations.
If any “special words” in any dialects were pronounced irregularly at the time when dialects branched out from the mainstream of Old Chinese or Middle Chinese, these words would be shared by all the sub-dialects of a particular dialect group. In other words, the “special words” in a dialects group would become a common property of that group, and would not be pronounced the same way in other dialect groups. These “special words” of different dialects, however, could be treated as one of the criteria for dialect classification, similar to any phonological characteristic feature. “Special words” in all Chinese dialect groups will be discussed in this paper.
This paper examines the syntactic structures of Chinese serial verb constructions with two verb phrases denoting two separate events. It is suggested in Li and Thompson (1981) that this kind of serial verb constructions can have the consecutive, alternating, purpose or circumstance interpretation. Although it has often been argued since Li and Thompson (1973) that constructions with the consecutive, alternating or circumstance interpretation have a coordinate structure and constructions having the purpose interpretation have a subordinate structure, it is argued in this paper that constructions with the circumstance interpretation should, like those with the purpose interpretation, have a subordinate structure. This analysis is grounded on the Coordinate Structure Constraint argued in Ross (1967) and the fact that constructions with the circumstance interpretation can have topical ization without its meaning being affected. Though constructions with a purpose interpretation are, following Li and Thompson (1973), argued to have a subordinate structure, it is suggested that the serial verb phrase should be treated as a complement of the Head verb. The Generalized Phrase Structure Grammar (GPSG) approach to syntax is adopted as the framework of the analysis, and it is argued that the structures proposed are suitable for computational implementation.
这篇论文研究中文连动句的句法结构。Li & Thompson (1981)曽指出，中文连动句中不同动词短语代表不同活动的句子可分为连续、交替、情况及目的四种，而他们认为首三种分类的连动句有并列结构，而第四类句子则有从属结构、本文作者则认为具情况意义的连动句也应有从属结构，因这类句子与有目的意义的句子皆容许宾语出现在主语之前 (topicalization)。本文采用 GPSG (Generalized Phrase Structure Grammar) 句法分析法作为句法分析的结构，而所议的句法结构较易于计算机运作。
This paper presents an analysis based on new acoustic data from tones and tone sandhi in Wenzhou dialect. The data provide evidence for the independent existence of a Depresssor as well as the Tonal Register and Tonal melodic component in the tonology. For example, a convex [343 or 342] pitch is shown to result from a depressor effect on a tone with high falling pitch. Comparisons are drawn with acoustics of other Wu dialects and Zulu. It is suggested that ‘murmur’ in many Wu dialects is actually Depression, but that Depression is not associated with non-vocalic segments, as it typically is in some African tone languages.
本文是根据在对温州方言的声调和连续变调观察中获得的新的声学数据所作的分析。这些数据证明压抑机制，跟调[音]域调势一様，是声调学中一个独立的成分。例如，本文显示一个高降的音调会由于压抑机制的作用而变为先升后降[343 或 342]。本文将温州方言声调的声学特征与若干吴方言和袓鲁语作了比较。本文建意许多吴方言中所谓的 murmur（”浊气流”）实际上是压抑机制的结果，但这种压抑机制并不象在一些非洲声调语言中常见的那様与非元音性成分有关。
Chun-fat Lau 刘镇发
This article deals with two questions: (I ) The criteria for Chinese dialects classification and (2) The rationale for the Hakka Dialects or even the Hakka, Yue and Gan dialects as independent dialect groups.
In the eyes of Chinese dialectologists, the vast numbers of Chinese dialects in China, which are mutually unintelligible among themselves but having regular phonetic correspondence, are “Chinese dialects”. But what are the relationships among them? Borrowing the concept of biological classification, dialectologists tried to classify them, but they lacked consistency. For a long period in the past, we were misled by superficial humanistic phenomena and used artificial boundaries (like provincial or prefecture boundaries) plus the feeling of the speakers to separate dialects into groups. In the Southern provinces like Yue, Xiang, Gan, Min every province were assigned a provincial dialect plus a “HakIca dialect group” as a result of Hakka identity. After fixing the number of “dialect groups”, they used the authoritative dialect of the group to represent the group, without considering the historical linkages between the dialects and the migration history of the speakers. During classification, they mixed up the early and late historical events and seemed to have no strict scientific principles. Apart from Min and Wu, every dialect group consists of dialects that are put together regionally rather than historically. In this article, apart from a preliminary definition for “early” and “late” events, I also supplemented a principle of Zhuo-Shang/Zhuo Qu sound change pattern to distinguish dialect groups. In this way, Chinese dialects can be classified into Northern, Wu, Min and Gan-Yue dialect groups, so that the present irrationally separated three groups of dialects — Hakica, Yue and Gan — are included in the last group. In the past centuries, because of the conflicts between the Jiaying- and Cantonese-speaking speakers, both groups built up hatred and discrimination against each other. The flaying dialect speakers were under pressure and invented the label “Hakka Identity” for self-esteem. Unfortunately, their story was accepted by dialectologists at once as a criterion for dialect classification. Therefore, this article also studies the relationship between “Hakka Identity” and the “Hakka Dialects”, so as to point out how nonsense it was to use non-linguistic factors for the definition of dialects. Finally, I also discuss the definition, limit and the position of the “Hakka dialects” in the field of Chinese dialectology.
在汉语语言学家的眼中，中国大地上现存无类互不能通，但互相之间有严密语音对应规律的方音，都是”汉语方言”。但是，汉语方音之间的关系是怎様的呢？方言学家便借用了 生物学的概念，将方言分类。可是分类的方法和标准没有一定的规则。在过去的一段时间里，我们都被一些表面的人文现象误导，将人为的界限（例如省、州界）甚至说话人本人 的认同或感情取向，用来作为方言界限。南方各省首先产生了”一省一语”的粤、湘、赣、闽等方言，以及基于”客家认同”的客家方言。订立了”方言”的数量之后，更采取权 威方音作方言的代表，作为该方言的特征，没有考虑方音间的历史联系，和方言人口历史的来龙去脉。分类的时候，将早期和晚期历史条件混合使用，显得没有严谨的科学原则。 除了闽语、吴语之外，每个方言区包括的方音，主要都地域关系而不是历史关系。 我在这篇文章中，除了对”早期”和”晚期”的条件初步做出定义外，还补充了一条浊上／浊去演变类型的原则。这様，汉语方言可以主要分为北方话区、吴语区、闽语区和赣粤 语区。而目前不合理地分为客、粤、赣的三种方言，大部份也涵盖在赣粤语区内。 在过去的数百年间，由于嘉应话和粤语羣体间的矛盾，造成两个羣体间的憎恨，仇视和敌对。操嘉应话者因为受到排挤、压迫，编造了一个抬高自己身份认同的标签－－”客家” ，结果也被方言学家不假思索地，顺手捡来作为方言分类的理据。因此，本文也探讨了”客家认同”和”客家话”的关系，指乢利用非语言学因素来介定方言的错谬。同时，也顺 带说明了”客家话”的定义、范围和其在汉语方言分类上应有的位置。
This paper examines the syntactic behavior of a certain set of verbs in the language of oracle-bone inscriptions (LOB) that can be associated with four nouns or valents. It proposes that these verbs constitute a specific class_of verbs with the semantic features [+request] and [+rituality]. There are six such verbs attested, all of which allow the use of a fourth valent characterized as “instrument,” the role borne by a sacrificial victim. This is a kind of object which is distinct from ordinary direct and indirect objects and is here referred to as “Oins.” In the LOB an Oins is indicated not by any particular marker, but by the syntactic position it occupies in sentences bearing certain semantic constraints imposed by the core verb. I have attributed this syntactic feature to a specific cultural factor, namely, a belief in the idea of reciprocity implicit in the ritual of sacrifice or gift-giving. In this sense, the sacrificers or gift-givers are rewarded with what they want in life, be it material or spiritual. This “reward-cum-sacrifice” phenome-non evident in the semantically driven syntax of the ritual verbs represents an important aspect of Chinese civilization in general and Shang culture in particular.
本文考察了甲骨文中一组四价动词的句法行为。这一特定类别的动词现在发现共有六个，它们都具有[ +要求]和[ +礼仪]的语义特征。这些动词都选择工具宾语作为其第四个论元 （其他三个论元是主语、直接宾语和间接宾语），而这个工具宾语是由祭牲类名词充当的。在甲骨文中，这种”工具宾语”不是由特定语法标志来标示，而是由其在句子中所处的 句法位置来显示的，这种句法位置受特定的语义约束。本文把甲骨文中的这种独特的句法特征归因于一个明确的文化因素，即在祭祀中暗含的互惠观念：祭祀者会得到报答，无论 他们所求的是物质的或精神的。”祭祀－受益”的现象代表着中国文明（尤其是商代文化）的特点之一。它在礼仪动词的句法中明显可见，而这种句法是由语义驱动而形成的。
Volume 30, No 2
Tsu-lin Mei 梅袓麟
This paper tries to answer two questions: Do we need a Chinese historical phonology with special Chinese characteristics? What sort of historical phonology with special Chinese characteristics should we develop? The answer: In the work Karlgren, Fang-kuei Li, Yakhontov, we already have an Old Chinese phonology with special Chinese characteristics. (I )All these scholars use phonetic compounds as one of the sounces for Old Chinese reconstruction, and the use of Chinese characters is a distinctive feature of Chinese culture. (2)All modern scholars of Old Chinese phonology incorporate in their work two great achievements of Qing philologists: first, the establishment of the rhyme categories of the Sizijkig, and second, Duan Yucai’s famous diction “words in the same phonetic series belong to the same Shying categories.” Section three argues that the concept “alternation by a single sound”(yi sheng zhi zhuan) propagated by Wang Niansun in Guangya shuzheng led to the pseudo-scientic method used in etymological studies such as Wang Li’s Tongyuan zidian and Xing Gongwan’s Han-Thai yu bijiao shouce. Section four shows that Old Chinese had prefix *s- and cognates could be formed by adding *s- to the root.
本文第一节提出两个问题：我们需要不需要一种”有中国特色的汉语历史音韵学”？我们需要哪一种”有中国特色的汉语历史音韵学”？第二节说，现在主流的汉语历史音韵学，如李方桂《上古音研究》、雅拱托夫《汉语史论集》、潘悟云《汉语历史音韵学》，里面已经包含着”有中国特色的”成份：（１）把方块字当作基本数据之一，（２）青代朴学家的古韵分部，（３）段玉裁的”同声必同部”。此外还用了西方传入的描写语言学和历史比较方法。结论是说我们现在已经有了”有中国特色的汉语历史音韵学”。 第三节说，王念孙《广雅疏证》中所用的”一声之转”的观念，在二十世纪产生章黄学派的训诂学，而且影响王力《同源字典》、邢公畹《汉台语比较手册》这两本书里研究同源词的方法。”一声之转”虽然是”有中国特色”的方法，但不能算作科学的历史语言学的一部分。 第四节举例说明，用汉藏比较的观点怎样去研究上古汉语的同源词。
Chung-Yu Chen 陈重瑜
This paper attempts to explore the nature of sound change, whether it is phonetically gradual or abrupt, as well as the catalytic effects of word frequency in the prelude leading to the final actuation of the change. Various types of tonal changes in Modern Peking have shown strong indications of having been frequency-induced. Three types of on-going changes are presented for illustration. Accumulation of occurrences (frequency) is accumulation of effects; sufficient effects eventually trigger a change of state. While the accumulation of effects is gradual, the triggering of change is abrupt. Hence, sound changes are phonetically both gradual and abrupt, a gradual prelude leading to an abrupt change-of-state. In the process of lexical diffusion, frequency may have been a criterion for the selection of participating members in various types of tonal changes between Middle Chinese and Modern Peking. However, in retrospect, only the abrupt change-of-state is visible, the gradual prelude built on frequency is perhaps invisible from a historical perspective. In that sense, gradualness is compressed into abruptness. This is perhaps the middle ground between the Neogrammarians’ view of sound changes being phonetically gradual and the Diffusionists’ view of their being phonetically abrupt.
This article examines various properties of the focus operator dak in Cantonese. The focus operator dak can occur in two positions in Cantonese: a postverbal position and a preverbal position. In this article, it is argued that these two focus dak’s are not the same: postverbal dak is a verbal suffix whereas preverbal dak is a verb. It is suggested that the preverbal dak construction can be treated on a par with the cleft construction. In addition, the relation between dak and the focus sentence final particle zoo, the relation between dak and negation, and the conditions of occurrence of several postverbal dak’s in Cantonese will be discussed.
Yunji Wu 伍云姬
The aim of this paper is to use data from 86 localities in Hunan to reconstruct the old layer of the personal and demonstrative pronoun system of the Hunan dialects. According to the data provided in the Modern Hunan dialects, the following system is hypothesised to have existed in Hunan at an early stage, although no single locality in modern times has maintained the entire system.
The above system is derived from the Archaic Chinese system. The changes and developments of the Hunan dialects that originated from Archaic Chinese are as follows:
i) The initial of the rd person pronoun shifted from dental to velar, possibly due to analogy; and
ii) The demonstrative pronouns “this” and “that” have phonological connection with the lperson ft wo in the Hunan dialects, while in Archaic Chinese the demonstrative pronouns had different sources.
In this paper, the relationship between personal pronouns, demonstrative pronouns, and personal and demonstrative pronouns, will be addressed to support the above hypothesis. Scholars have noticed that there are phonological relationships amongst pronouns. Zhou Fagan once pointed out that phonological-semantic relationships are often found among the 1st, 2nd and 3rd person pronouns and among the pronouns indicating inquiries, negation, etc (Zhou 1953:3). Lu Shuxiang claimed that the 3rd person pronoun and demonstrative pronoun have a very close relationship. In Chinese, it is only the lst person pronoun that has no relationship with demonstrative pronouns (Lu 1985:187). All of the above research uses written data from Archaic Chinese and discusses its relevance to modern Chinese dialects. The aim of this paper is to discuss the phonological relationships between personal and demonstrative pronouns, to reconstruct the earlier layer of the pronoun system in Hunan, and to describe its development by using modern data from 86 localities. 2 The data from Hunan demonstrates that there are phonological links between personal pronouns, demonstrative pronouns, and personal and demonstrative pronouns. In the older layer of the pronoun system in Hunan, the three personal pronouns and two demonstrative pronouns once all had velar initials. It is argued that in the Hunan dialects, the 1′ person pronoun has been related to demonstrative pronouns.
The following four issues will be discussed respectively:
i) The phonological links in personal pronouns in the Hunan dialects;
ii) The phonological links between demonstrative pronouns in the Hunan dialects;
iii) The phonological links between personal and demonstrative pronouns in the Hunan dialects; and
iv) The development of personal and demonstrative pronouns in the Hunan dialects.
This paper explores the relationship between the aspectual and thematic systems of verbs in Mandarin, focusing on the phenomenon of Aspectual Category Shift (ACS), the shifting of verbs between aspectual categories. Using a framework of aspectual features proposed in Jackendoff 1987, it demonstrates that the aspectual structure of verbs determines and constrains their ability to shift. Finally, it argues that aspectual features intersect with certain thematic features to determine the four aspectual categories of State, Activity, Achievement and Accomplishment.
本文旨在探讨汉语中动词的体类和体元系统二者间的关连，着重于体类变化以及动词在体类变化的现象。本文是依据Jacken doff 与1987年所发表的体特征一观点作为论述架构，主要论点在于证明动词的体特征会决定并限制其变化的能力。最后，本文亦针对体特征与某些特定的题元特征将共同决定「状态」，「活动」，「达成」，「完结」四种体类此论点作一论证。
The polysemous lexeme DO, literally meaning be-at or arrive, in Hakka, a language spoken mainly in Taiwan and China, carries multiple grammatical Functions ranging from a main verb, a coverb, a complementizer, to a verbal complement. In regard to these phenomena, the study has two claims. Structurally. DO is grammaticalized from a full-fledged verb, through the path of a coverb. and a complementizer, to a verbal complement. The development not only reflects the unidirectionality of a grammaticalization process but also exhibits a continuous grammaticalization chain by which a lexical item travels from a lexical-content word to a grammatical morpheme, and then from a less grammatical morpheme to an even more grammatical morpheme. (Hopper 1991: Hopper & Traugott 1991, 1993; Heine et al. 1991) Second, the meaning extensions of DO display the scale of a metaphorical abstraction: PERSON/OBJECT > SPACE > TIME > MANNER/RESULT (cf. Heine et al. 1991). The development also exemplifies the three semantic-pragmatic tendencies characterizing paths of semantic change in Traugott (1989). More specifically. DO denoting a person or an object existing or arriving at a certain place is metaphorically extended to specify a location at which some action takes place. or to mark a spatial or temporal goal. which some action has reached. Subsequently. the meaning is further developed to indicate the result or manner caused by some other action. Finally, the resultative meaning of DO may be embodied into a verb to denote achievement resulting from the action of the verb. Hence, the synchronic semantic connections between different lexical domains of DO are accounted for naturally.