Journal of Chinese Linguistics vol.19 (1991) 中国语言学报 19 卷 (1991)
Volume 19, No 1
Chinese Pidgin English is a contact language widely used along the coast of China in the 18th and 19th century. It was formed in a very restricted contact situation to fulfill limited communication need between English-speaking traders and their Cantonese-speaking servants. Many linguistic features of this pidgin can be considered as contribution from Cantonese. A few features can be analyzed as simplification from both languages. No exclusive English structural feature is found in the data of pre-20th-century Chinese Pidgin English.
Although both Chinese and European languages exhibit long distance reflexives, the properties of long distance reflexives appear to be different in Chinese from those in European languages. In this paper, we focus on two differences. First, in Chinese, long distance reflexives exhibit “blocking effects” while there is no blocking effect in European languages. Second, in European languages but not in Chinese reflexives can only be bound across certain clause types. We suggest that long distance reflexives involved successive cyclic head movement in all languages and that what distinguishes the distribution of reflexives in Chinese from that in European languages is the effect of rich or impoverished morphology.
长距离反身词在中文和欧洲语言里均属常见，但显示的特点却不同。本文重点讨论两者的不同之处。第一，中文在运用长距离反身词上有所谓的“阻碍现象”，而欧洲语言没有。第二，中文里的长距离反身词，不像欧洲语言，不会受到某些特定句式的影响。我们对上述差异所提出的解释是：所有语言里的长距离反身词都必须经过连续循环的“主要语提升” (head movement)，而中文与欧洲语言之间长距离反身词的差异则取决于该语言词语结构变化的多寡。
Most likely southern poets of the Yuan wrote songs using the ru-tone of their southern dialects. If their usage differs consistently from how northern poets used these tones, it could be another indication that the ru-tone had actually disappeared from the northern spoken dialects. In the following study I have analyzed all 2ru-tones in the 53 songs written to the Heiqi nu song form. The results are only tentative. There are appended a list of all 2ru-tones in the songs and a statistical register of the tones in each line.
This paper reexamines the grammatical category status of coverbs in Mandarin Chinese. It considers the properties of coverbs used in earlier works to identify them as verbs, prepositions, or both, and demonstrates that none of these properties serves to distinguish the class of coverbs from the class of verbs in Mandarin, or to identify coverbs as prepositions. It illustrates problems with Li’s 1985 assumptions about thematic role properties central to her Case structure analysis of Mandarin and uses thematic properties to distinguish ba from the class of coverbs. Finally, it accounts for the major properties of coverbs in terms of semantic features, thematic role assignment, and phrase structure configuration.
本文对汉语中“同动词 (coverb)”的语法分类问题进行了重新探讨。在以前的论着中，把“同动词”划分为动词，介词或动词兼介词。本文分析了这些划分的特征，认为这些特征既不能把“同动词”与动词加以区别，又不能证实“同动词”为介词。1985年李博士 (Audrey Li) 提出了关于主题作用特征的设想。此设想是她进行汉语“格”结构分析的核心。本文对她的这个设想提出了一些问题。并运用主题特征把“把”字从“同动词”的词类中区分了出来。最后，本文从语义特征，主题作用及短语结构上阐述了“同动词”的一些主要特征。
New Publication 新书
Volume 19, No 2
Chung-yu Chen 陈重瑜
Thirty-six instances of confusion between the nasal endings -n and -ng in dictionaries and rime books have been cited. The data here show that among all the finals, the pairs -in/-ing and -en/-eng have been most unstable. There are indications that final -in will be the first to lose its identity. Data of current actual speech in Singapore and Taipei reveal a predominant trend of -in merging into -ing and -eng merging into -en. A total of 104 attested examples of confusion or interflow between the retroflexed and the dental obstruents have been found in rime books, dictionaries and, in a few cases, reported current speech. There are 66 instances of de-retroflexion; 46 of them have lost retroflexion completely; 20 are in a two-reading stage. 15 examples of acquisition of retroflexion have been found; 13 can probably be attributed to influence of graphs sharing the same phonetic component. There were 23 characters that started out with both types of initials; 20 of them in Middle Chinese and 3 recent formations in the 1932 system. Of which, 11 retain only the dental readings, 5 retain only the retroflexed readings, 4 preserve both types of readings and 3 have become obsolete. The data here indicate that the confusion between the two types of initials can be traced back to the time of MC.
鼻韵尾 -n 与 -ng的混淆在词典与韵书里已有迹可循：本文举出的36个例证显示 -in/-ing 与 -en/-eng 两组最不稳定；而目前台北与新加坡两地的实际语音里的变化则是以 -in 变向 -ing, -eng变向 -en进展得最快。卷舌音与不卷舌音的混淆在中古音里已不乏其例，而自中古音以降，中原音韵，1932系统，1963-85系统各阶段之间，读音亦多有变更。文中的104个字例只有笔者在短期内搜寻所得；实际上的字数当远多于此。此处104个例证可说是北京一地的变化。在北京以外，台湾的国语，新加坡的华语，及中国大陆许多地区的普通话里，卷舌正迅速地消失。这种现象可能促使卷舌音在北京一地的变化步伐加速。
Anne Yue-hashimoto 余霭芹
The present paper explores the problem of linguistic stratification in syntax by presenting the case of V-not-V question forms in Southern Min as seen in several colloquial Southern Min documents from different historical periods–The Litchi (Mirror) Tale, The Golden Flower Girl, Su Liu-niang and The Schoolmate and Zither Book–as well as in the contemporary Yilan dialect of Taiwan and the Shantou dialect of Guangdong. Our conclusion is that there are at least three different strata of this syntactic form in Southern Min: a premodern colloquial stratum with the form 可/豈 VP-NEG from which may have sprung the modern colloquial ADV-VP form, a standard Southern Min stratum with the form VP-NEG which span several centuries and still survive to this day and a non-native stratum in the form of VP-NEG-VP as a result of contact with Northern dialects. In this new light of syntactic stratification, we can understand why more than one syntactic form exists in Southern Min as equivalents to the V-not-V question form and still support Professor Zhu Dexi’s general claim that the two major types of V-not-V question forms are mutually exclusive in distribution among the dialects, although we may think that perhaps there are three, instead of two such major types. What is of paramount interest is that similar kinds of interrogative structures not only exist in other Sino-Tibetan languages and neighboring non-Sino-Tibetan languages, but also seem to indicate the same type of complementary distribution. It will be of immense typological significance to investigate further into the problems of the origination and diffusion of these question types. Then, the problem of syntactic stratification of not only Southern Min dialects or the Chinese dialects but the Sino-Tibetan languages will then come to light.
Linguists seem to have a different opinion on the syntactic properties of the PP’s in Classical Chinese. While many assume that Classical Chinese had predominantly postverbal prepositions, many others believe otherwise. This study finds that the latter view is perhaps correct and that the PP’s in Classical Chinese could be either postverbal or preverbal. On text-count level, PP’s of the two types are about equally divided. Furthermore, this paper shows that the adposition yi 以 can be used as a preverbal preposition, a postverbal preposition, or a preverbal postposition in Classical Chinese. Evidence will be given to contend that the variation between the preverbal and postverbal uses of yi seems to be motivated by discourse factors. The nominals following the preverbal yi seem to be much closer to their antecedents than those after the postverbal yi. In addition, the high degree of variability of yi may follow from an earlier grammatical change, i.e. changing from postpositional to prepositional. Thus, it is hypothesized that the postpositional use of yi perhaps was simply a vestige of earlier Sino-Tibetan syntax.
Passive forms in contemporary Chinese require an obligatory post-verbal constituent, which was not required in ancient Chinese. This change has three possible explanations. The first one is the rapid development of the V-R structure. In ancient Chinese, what was needed to express a factitive meaning was the V-O structure, while today the V-R structure is the dominant one and R cannot be omitted. The second explanation is related to the aspect of passive verbs. Passive forms generally express the accomplished aspect, in which case the verb is marked by an aspectual particle. When a non-accomplished aspect verb is in the passive form, a post-verbal constituent is required in order to add the accomplished meaning. Consequently, either in the former or in the latter, the post-verbal constituent becomes indispensable. The last reason is that the R holds the key meaning in the V-R structure. Semantically speaking, V plays the role of manner circumstantial complement and R is the focus of the whole passive sentence. Thus, R is essential and cannot be deleted.
At least three distinct approaches to the analysis of Chinese regulated verse have appeared in the literature of the past half century: linear, metrical tree, and metrical grid. Wang Li (1957), Downer and Graham (1963), T’sou (1968), and Jakobson (1970) have all offered linear accounts of this verse, and all of these accounts have been argued to be inadequate in Chen 1979. Chen gives the second approach: he applies generative theories of meter that were developed in more recent years and comes up with a tree analysis that has received wide acclaim (as in Graham 1980, Yip 1980a & b, Xue 1989). While some linguists have offered explicit criticisms of varying parts of Chen’s analysis (such as Liu 1980, Schlepp 1980b), others have offered evidence that goes counter to certain details of Chen’s analysis without citing Chen’s work (Boyce 1980, Cheung 1980, Ripley 1980). But only two works that I know of have offered entire alternative analyses to Chen’s. One is the linear analysis in Lorentz 1980; the other is the metrical grid analysis in Napoli 1989. In this paper I argue that no single approach, whether linear, arboreal, or grid, is adequate to account for all aspects of Chinese regulated verse.
上半世纪的研究论着，对律诗的分析至少有三种不同的探讨方法线性的分析，诗韵律的树状图和诗韵律的格子状分布。王力(1957), Downer和Graham（1963）, T’sou (1968)，以及Jakobsen（1970）都对律诗提出线性的描述，1979年Chen指出所有的这些描述都缺少足够的立足点。Chen提出第二个探讨方法：他用的是诗节奏有繁殖能力的理论；这是较近几年发展出来的树状图分析法。这种分析法受到广泛的赞扬（如Graham 1980，Yip 1980a 和 b，Xue 1989）。一些语言学家对Chen论文中不同部分提出明晰的批评（如Liu1980，Schlepp 1980b）；也有一些语言学家对Chen某些分析的细节提出相反的证据，但是没有指明是针对Chen的文章（如Boyce 1980，Cheung 1980，Ripley 1980），据我所知，目前只有两篇文章针对Chen的理论，提出对律诗完全不同的另外一套分析；一篇是Lorentz 1980年提出的线性分析，一篇Napoli 1989年的诗韵律格子状分析。在这篇文章里，我论证如果只采用一种分析法，不管是线性的也好，树状图的也好，格子状分布的也好，都不能涵盖中国律诗的各个方面，得过来看，树状图的方法最适合分析韵脚的形态，线性分析法最适合声调形态的研究。