Volume 29, No 1
Dai Qingxia 戴庆厦
The paper is a preliminary exploration of the historical development of the causative category on the basis of a comparison of more than 20 Tibeto-Burman languages. The methodology adopted is one of the comprehensive comparison between phonetic forms and semantic content and of the analysis of the language system (structural system and kinship language system). We conclude that the causative category is one of the oldest and most general grammatical categories of Tibeto- Burman languages and its historical development is constrained by prosodic, semantic and functional factors.
This article offers a detailed analysis of the so-called ‘plural’ suffix -men in Mandarin Chinese. -Men occurs with both personal pronouns and nouns. However, while it is regularly attached to personal pronouns, it appears only sporadically – though not randomly – with nouns. Its occurrence after nouns is highly constrained (humanness, incompatibility with simultaneous counting, definiteness, non-genericity, modal overtones) and is manifestly not governed by considerations of number alone, unlike the plural in Indo-European languages. Clarifying the conditions of the occurrence of -men with nouns was the major unsolved problem, since this optional use cannot be captured in terms of syntactic or any other mechanical rules. It is discourse-motivated and depends on pragmatic factors. The dichotomy between obligatory and optional use prompted many scholars to distinguish two values for -men, a plural after pronouns and a collective after nouns. I argue that -men is in all cases a collective. What actually triggers the occurrence of -men is the conjunction of number and person. This is the case par excellence of the ‘plural’ of personal pronouns, which does not correspond to an addition or multiplication of elements, but rather to what Benveniste (1966) calls ‘amplification’ of persons. I show that in narrative contexts the group referred to by N-men is constructed relative to a given character as opposed to the speaker/narrator. The character is not only the locator for the group but also the spatial point from which things are viewed. This shifting of perspective (internal versus external point of view) is not structurally conditioned, which explains why nouns in narration are not obligatorily suffixed by -men. In summary, the suffix -men always marks a subjective grouping, either relative to the speaker/narrator or to some character whose point of view the narrator provisionally adopts. This amounts to unifying nominal and pronominal -men. The collective at issue should be characterized as personal, in the sense we talk about ‘personal’ pronouns.
“严" 及 “凡”开合分韵后例外字的分析
Guey-lin Chen 陈贵麟
The purpose of this paper is to solve an unsettled question on the few unusual rhyme words in the proto-form of Yan 严 and Fan 凡 rhyme sets. The few words can not be taken as a proof of violating the principle of the complimentary distribution because they come from the different sources. In the view point of the structural analysis and reconstruction rule, I study the Qie-yun 切韵 rhyme books and reach my conclusion. The rhyme sets of Yan 严 and Fan 凡 would be complimentary perfectly on labials if the Fan 凡 rhyme set were not mixed with dialectal words and elusive words from Archaic to Middle Chinese. It would be better to say that the few words on dentals and gutturals are very sporadic and should be treated as exceptional cases. These exceptional cases, though very complicated, may be summed up that they result from the two-syllable words, the abnormal dialectal words from Qin rhyme-base 侵部, and the new rhyme words misplaced in the rhyme book. They should be removed in the reconstructed process of Yan 严 and Fan 凡 rhyme sets. In some versions of Qie-yun 切韵, like Wang Ren-xu 王仁昫 rhyme book, its Shang-sheng 上声 words of Yan 严 are diffused from another Yan 琰 and its Qu-sheng 去声 words of Yan 严 are diffused from another Fan 梵. Those words drifting into the accidental gap of Yan 严 rhyme set will not interfere with the reconstruction rule of Yan 严 and Fan 凡 rhyme sets.
This study aims to demonstrate the existence of pre-Qieyun phenomena in the Old Sino- Korean Corpus. According to the toponyms in the Samkwuk saki (1145), the alveolar stop (Duan) initials of Old Chinese remained unchanged in Old Sino-Korean regardless of the degree to which they changed to palatalized or retroflexed stops in the Middle Chinese period. The palatalized alveolar stop (Zhang) initials of Old Chinese were also realized as unpalatalized alveolar stops in Old Sino-Korean, while they changed to affricates in Middle Chinese. Based upon these findings, this article argues that the Chinese characters were transmitted to ancient Korea at latest by the fourth to fifth centuries A.D.
Picus Sizhi Ding 丁思志
Contrary to previous analyses that generally regard ba in Mandarin to have been grammaticalized into a semantically empty preposition, the present study argues that the nature of the evolution of the verb ba is a semantic one. Taking the trend of semantic shift from concrete domains to abstract domains into consideration, it is proposed that the meaning of ba has been grammaticalized from ‘to hold (something)’ to ‘to bring about a resultative state’ in marking a resultative construction. This path of grammaticalization is independent of a well-attested path of grammaticalization responsible for changing ba into an instrumental preposition in Old Mandarin.
tɕhiIʔ tsɨ vE (“吃仔饭”)is a common expression in Shanghai and Suzhou meaning “have eaten (the meal)”; tɕhiIʔ (吃) is a verb meaning “to eat”, vE (饭）is its object meaning “a meal”, and is the marker for the perfective aspect. When we try to date this expression, that is, to find out the date of its earliest antecedent, we get two different results. The structure “V+perfective marker+O” first arose in the 11th century. But the etymology of tsɨ (仔)is 著, and in Shanghai and Suzhou, 著 becomes tsɨ (仔) only in the phonological stratum inherited from the Jiangdong dialect of the 5th and 6th centuries. This paper attempts to account for the 500 year difference given by two dating criteria, grammatical and phonological. The paper argues that the perfective marker 著 tsɨ (仔) was previously a locative marker, attested in 5th and 6th century texts and preserved in Min dialects. There are also transitional dialects such as Qingtian 青田 in southern Zhejiang and Xiuning 休寧 in southeastern Anhui in which the forms cognate to are used both as the perfective marker and the locative marker. The reason why tɕhiIʔ tsɨ vE (“吃仔饭”) gives two dates is tsɨ (仔) because (or more precisely, its etymon) changed its grammatical function while retaining its phonological characteristics.
本文讨论为什么用语法，音韵两种标准给吴语“吃仔饭”断代，会得到两种不同的结果，其间相差五百年有馀。第一节说明“吃仔饭”这种[动词+完成貌词尾+宾语]的结构，至早产生于十世纪以后。第二节说明， 属于鱼虞有别那个层次的“着(仔) ” 字，是承继南朝的江东方言。第三节解释为什么两种断代标准所得的结果不同。第四节是馀论和结论。
Volume 29, No 2
Tze-wan Kwan 关子尹
Acclaimed as the father of general linguistics, Wilhelm von Humboldt’s scholarship on the Chinese language was also a remarkable one. For Humboldt, language universals are nothing but some general principles pertaining to the function of language as a major intellectual activity. These principles, however, have to be exemplified by individual languages with all their linguistic particularities. Humboldt considered languages to be like organic entities capable of consistent development over ages. Starting from specific linguistic idiosyncrasies, individual languages tend to follow their own consistent and self-adjusted line of development. In the Chinese language, initial linguistic features such as “phonic poverty”, “phonic isolation” (monosyllability) etc. prevented the Chinese from developing a grammatical system on the basis of inflection, as is the case in most Indo-European languages. In regard to grammar, the Chinese language seems to have found a way of its own. Instead of relying on externally precipitated sound forms, Chinese grammar has developed in such a way that the intellectual power tends to take direct control over the sentence. For this to function smoothly, a number of linguistic devices (such as tonality, the use of particles and idioms etc.) were also developed as compensation measures to offset the relative weakness of the Chinese sound system. For the same reason, in the development of the Chinese script, the graphical elements were allowed to share with the phonic elements the burden of meaning discrimination. With the measure of “analogy of script”, as highlighted by Humboldt, the hidden creativity of the Chinese mind is liberated to its greatest extent. Besides being an interpretation of Humboldt’s theory, this paper is also an attempt for reconstruction. In so doing, various works of modern scholarship on the Chinese language/script were cited, of which the works of two Chinese linguists, Hu Pu’an and Sun Yongchang were given special attention.
There is a traditional belief that a bare verb in the ba construction is never allowed. While often observed, the rule does not always hold true, as I propose to show in this paper. In poems, for example, bare verbs are commonly used to form ba constructions. In everyday speech, disyllabic bare verbs in the ba construction are also allowed, whereas a monosyllabic counterpart is strictly banned in the same environment. Given this, it is argued that the Bare Verb Effect in ba constructions is constrained neither by syntax nor by semantics, but primarily by prosody. The argument made here strongly supports the proposal made in Feng (1995) that the interactions between syntax and prosody are bi-directional: Syntax governs prosody and prosody also constrains syntax. The prosodic constraints developed in this paper can also be viewed as well-formed conditions on all ba sentences which cannot surface until all of the relevant types of structural constraints (syntactic, semantic and prosodic) are satisfied. The prosodic constraints could, therefore, be one type of interface condition under the assumption that all conditions are interface conditions and that a linguistic expression is the optimal realization of the interface conditions.
在汉语的句法研究中，一般认为把字句中的动词不能挂单。本文提出：这一规则尽管一般如此，但并不绝对。譬如，在诗歌韵文中，把字句的动词，无论双音节还是单音节，挂单均屡见不鲜；在口语里，双音节的挂单动词同样允许。口语里只有单音节动词决不挂单，即使其意义与双音节动词完全相同。据此，本文提出，动词挂单的把字句，既不能从句法上来解释，也很难从语义上来说明。 这里真正起作用的是“韵律”。 韵律是制约把字句动词的根本所在。这一分析，为“句法控制韵律，韵律也制约句法”（Feng, 1995）的理论，提供了有力的证据。本文认为：韵律规则可以看作一种不同类型的交接条件 （Interface Condition）。如果语法规则可以视为交接条件， 如果语言表达都是交接条件相互作用的最佳条件，那么，把字句的表达，同样必须在其韵律条件与其它交换条件的相互满足中，才得以实现。
Jian Kang 康健
In the context of aspectual studies, there has been little scholarship on the perfective aspect value of the directional verb compounds (henceforth DVC). The aspectual nature of the directional verb complements (shang, xia, hui, guo, kai, qi, and lai, qu) was almost overlooked until 1992, when the Chinese grammarian Fang treats them as the perfective aspect particles. However, the exact nature of the aspectual meaning of the directional verb complements is frequently unclear. This article provides a detailed study of the perfective aspectual meaning of the DVC. The study is made in terms of aspectual structure developed by Tenny. In this framework, directional verb compounding is viewed as an operation over aspectual structure. It is an aspectual operation whereby the PATH, TERMINUS or the MEASURE aspectual role of directional complement(s) is added to the empty aspectual role grid of the simplex verb (mainly an Activity) they combine with. As a result, the verb compound acquires the MEASURE or the PATH, TERMINUS aspectual role and is converted to a complex Accomplishment. As such, the DVC describes a telic situation. In a past context, or in combination with the perfective le, they denote attainment of a goal, thus contributing to the perfective meaning of a sentence. Directional verb compounding represents a highly productive process that relies on the property of delimitedness. They are clearly instances of grammaticalization of the telic notion in Chinese. They contribute to the perfective reading of a sentence.
汉语中对体态(Aspect)的研究,极少涉及到趋向复合动词有助于表达完成体 (Perfective Aspect)的语义功能。趋向动词 (上，下，进，出, 回，过，开，起，来，去)的表达体的语义几乎完全被忽视。直至一九九二年，中国学者房玉清在其着作“实用汉语语法”中，阐述了趋向补语表达体的语义及作用，并明确地称这些动词为完成体的动态助词。然而语法界对趋向动词表达体态的功能仍缺乏明确的认识。鉴于此，本文就趋向动词表达体的语义作了详尽的论述。本文作者认为趋向动词不是表达完成体的动态助词，而实质上是终结动词(Accomplishment)的情状标记。趋向动词在与活动动词(Activity)结合时，给活动动词加上了终结点(endpoint) 或目标 (goal) 的概念，从而使只表示持续的，匀质的，不引向结果的活动动词转变成为表示非匀质的引向结果的终结动词。用Tenny (1994) 研究发展的体结构(aspectual structure)的理论来分析，趋向动词与活动动词结合的过程可看作是体结构中的运作过程。运作的结果，趋向动词将其表达途径，终端，或度量(PATH, TERMINUS, MEASURE) 的语义特征转加到了活动动词的空白的体作用格中(aspectual role grid). 由此，不具备终点特征的活动动词转变成终结动词。因此趋向复合动词描述的是一个有终结点的事件 (a telic situation)。 在表示过去的语境中，或与完成体的动态助词“了”共现时，趋向动词表示事件的完成或其结果的实现。在此意义上，作者认为趋向动词表达了完成体的体语义。趋向动词与活动动词的结合在汉语中是十分丰富与活跃的。因此作者进一步认为趋向复合动词与结果复合动词一起，是汉语中完成体体系中的一部分。
The Temple Names (廟號) of Shang Kings (ca 1,700 to 1,100 BC) found in the Oracle Bone Writings are remarkably similar to the King List compiled by Ssu-ma Chi’en (ca 145-90 BC), the Grand Historian of Han Dynasty, in the Book of History. The Temple Name of Shang King usually contains only two characters and the last character always corresponds to one of the ten heaven stems (天干). Although several theories have been advanced to offer explanations of the use of heaven stem in the Temple Name, the significance of the first character in the Temple Name has not been discussed in the literature. Here, we argue that the first character 盤 (Pan) in the Temple Name of the 19th Shang King, Pan Keng (盤庚 or 般庚), was given posthumously and that it was given in honor of his lifetime achievement in the massive relocation of the Shang Capital, from Yen (奄) to Yin (殷) and his use of the newly developed hang-tu (夯土) technology in a gigantic city-palace construction work. We further argue that the word 般 was derived directly from the hang-tu (夯土) technology. My argument was based on the following considerations: (i) the structural analysis of the oracle bone character that corresponds to 盤 or 般 ; (ii) the Shang history on Pan Keng as recorded in the Book of Documents, the Bamboo Annals, and the Book of History; (iii) the archeological excavation of the palace and temples remains at Anyang city, the last Shang Capital that Pan Keng established about 3,400 years ago; (iv) the possibility that the phonetic value of Pan was derived from the sound generated during hang-tu production; and (v) the word pan (槃) was used in the Book of Odes to indicate construction work.
This paper presents a comparative study on the A-NOT-A Questions in Mandarin and Taiwanese Southern Min. A dialectal investigation is conducted in the city of Hsinchu to grasp the patterns of A-Not-A questions in Taiwanese Southern Min. The survey turns up a range of divergent grammatical patterns of these questions as well as other related linguistic features that are closely correlated with the age factors as manifested in the discrepancy between the older and younger generation. The linguistic behavior of the younger people as native speakers of Taiwanese Southern Min shows signs of nascent change under the burgeoning influence of Mandarin.