Brigitte Höhenrieder
CULTURAL LINGUISTICS - A CHALLENGE FOR CHINESE LANGUAGE RESEARCH IN THE 21ST CENTURY ?
In the preface of the Encyclopedia of Chinese Cultural Linguistics published in 1993, we can read the following statements about cultural linguistics by Dai Zhaoming:
"Cultural linguistics is a new school in linguistics that sprang up in China after the mid-eighties and has integrated Chinese history and culture into the study of the languages of China. Within very few years, it has already started to form a fairly independent new discipline with its own objects, assignments and methods of research, and it has exerted a definite influence on the linguistic circles in China and abroad and even on fields of other humanities." (p.1). "Being a discipline in which language and culture intersect and permeate and the objective is to explain the cultural intension of language, cultural linguistics can make up for the shortcomings of modern linguistics brought about because of an absorption in static description of language structures, and is of a value and an effect that is irreplaceable by other linguistic disciplines including sociolinguistics and ethnolinguistics which are similar in nature" (p.10)
These lines confront us with some interesting claims about the importance of cultural linguistics: 1. Cultural linguistics is a new and independent discipline. 2. It has a certain impact on linguistics and on other humanities. 3. It can do better than established modern linguistics. 4. It does have a nature similar to sociolinguistics and ethnolinguistics, but nevertheless is decisively different from them.
Advocates of cultural linguistics in China generally propagate to supplement or even to replace established modern linguistics by cultural linguistics, at least in the research field of Chinese languages. Those should not be analysed on the premise of an autonomy of language as done in modern linguistics, but by including the specific cultural context of China. "Established modern linguistics" in this respect refers to all kinds of universalistic and structuralist, descriptive and formalistic linguistic approaches that have come to China from the West during the decades before the Cultural Revolution. Cultural linguistics developed in China during the revival of academic life after the Cultural Revolution.
Against this background my talk will in the beginning give a short introduction to cultural linguistics, its development and main lines. Afterwards it will present the positions within linguistics that cultural linguistics ascribe to cultural linguistics: as related to sociolinguistics and ethnolinguistics/anthropological linguistics, but nevertheless different and independent. These views of main representatives of cultural linguistics then have to be critically examined in order to find out, first, whether we can accept these so-called differences as real ones in a Western perspective, too, and second, whether cultural linguistics could be something new in Western linguistics as well - or at least in Chinese language research in sinology. In the end I intend to formulate my own thesis about the importance and consequences cultural linguistics could and should have in present Chinese language research.
THE MORPHOSYNTAX OF FOCUS AGREEMENT IN MANDARIN:
LUMPING TOGETHER CAI, JIU, DOU, YE (AND ZAI)
This paper proposes and defends two major claims; both concern the use of parametric (focus-related) cai, jiu, dou, ye and zai as in (1).
(1) a. Zhiyou XINGQITIAN ta cai gongzuo.
only Sunday (s)he CAI work
'(S)he only works on SUNDAYS.'
b. Zai ZHER women jiu neng wanr.
at here we JIU can play
'We can have fun HERE.'
c. Lian XINGQITIAN ta dou gongzuo.
even Sunday (s)he DOU work
'(S)he even works on SUNDAYS.'
d. Jiushi GUOWANG lai wo ye bu qu.
even.if king come I YE not go
'Even if the KING comes I won't go.'
e. MINGTIAN zai gongzuo ba!
tomorrow ZAI work PRT
'Let's wait until TOMORROW and work (only) THEN!'
The first claim is that the use of the particles listed above is an agreement phenomenon: the verbal complex of the clauses agrees with preceding foci of specific types. The specific kind of focus may either be marked overtly by function words (zhiyou, lian, jiushi in (1)) or
contextually as in (1b) and (1e).
The second claim concerns the specific categories that have been syntacticized in this domain. I will show that each of the particles is the syntactic reflex of a specific focus quantificational type. Possible focus quantificational types in this domain are restricted by the interplay of existential/universal quantification and negation with respect to sets of alternatives: cai-sentences must involve negated existential quantification over sets of alternatives (There is no other day apart from Sundays on which (s)he works; cf. (1a)); jiu-sentences involve negated universal quantification over sets of alternatives (Not all other places are places at which we can have fun; cf. (1b)); dou-sentences involve universal quantification over sets of alternatives (All other days are also days on which (s)he works, cf. (1c)); ye-sentences involve existential quantification over sets of alternatives (At least one occasion apart from the occasion when the king comes is such that I will not go on this occasion, cf. (1d)). This core system of four quantificational types leaves open the possibility of having more specific focus quantificational types which can still be analyzed as being based on the basic quantificational types. Zai as in (1e) is a case in point. I claim that zai in its parametric (focus-related) use is not just restricted to contexts in which the focus alternative temporally precedes the reference time of the asserted sentence, I also claim that the number of (excluded) alternatives considered is necessarily restricted to 1. This restriction will help us to understand why, varying with the context, zai-sentences are sometimes felt to involve only-semantics, and sometimes not.
Robert Iljic
L'EXPRESSION DE LA PLURALITE EN CHINOIS
En chinois contemporain, le suffixe -men, dit " de pluriel ", figure aussi bien derrière les pronoms personnels que derrière des noms. Toutefois, alors qu'il est adjoint régulièrement aux pronoms, il n'apparait qu'occasionnellement - bien que de fa,con non aléatoire - avec des noms. En fait, la plupart du temps les noms ne sont pas marqués, le contexte foumissant l'information nécessaire.
-Men est soumis à des contraintes importantes: (1) il fait en principe toujours référence à des etres humains; (2) il est incompabble avec un comptage simultané (* san ge háizimen); (3) il est associé au défini (il suppose une deuxième mention d'entités posées auparavant); (4) il véhicule parfois l'empathie, c'est-à-dire a des conmotabons modales.
Visiblement, son emploi avec des noms n'est pas régi exclusivement par des considérabons de nombre, contrairement aux désinences de pluriel des langues indo-européennes. La thèse que nous défendons ici est que -men n'est pas un vrai plariel mais, dans tous les cas, un collectif; et qu'il relève non seulement de la catégorie grammaticale du nombre mais aussi de celle de la personne.
La personne grammaticale se ramène à une localisation dans l'espace centré sur une origine subjective, c'est fondamentalement une topologie. De manière générale, le prétendu " plariel " des pronoms personnels est en réalité un colloctif: il ne consiste pas en une addition ou mulhplication d'éléments mais plutBt en ce que Benveniste (1966 :236) appelle une amplification des personnes. Les pronoms personnels définissent des positions par rapport au locuteur, respecOvement par identification ( I -ère personne), différenciation (2-ème personne) et rupture (3-ème personne) avec le locuteur. Bien que -men présuppose le nombre, n > I n'est que la condition nécessaire mais non suffisante de son emploi. Ce qui déclenche l'apparition de -men est en fait la conjonction du nombre et de la personne. C'est le cas par excellence du " plunel " des pronoms personnels. C'est pourquoi -men est obligatoire avec les pronoms personnels dès qu'on a affaire à plus d'une occurrence.
A la différence du pronom, le nom ne fait pas en soi référence à une origine subjective. Ainsi, en narration, qui est par définition en rupture avec la situation d'énonciation, le suffixe est optionnel. Y recourir c'est regrouper des enhtés/individus par rapport à un personnage narratif, loquel est non seulement le localisateur du groupe mais aussi le point spatial d'où les choses sont envisagées. Il y a à proprement parler un changement de perspechve. Le narrateur prend le point de vue d'un protagoniste, construisant de la sorte un point de vue interne au récit. En revanche, en allocudon N-men équivaut à un vous qualifié, et dans ce cas le recours à -men est systématique.
En conclusion, le suffixe -men marque dans tous les cas un regroupoment subjeclif, soit par rapport au locuteur/narrateur, soit par rapport à un protagoniste dont il adopte provisoirement le point de vue. C'est pourquoi on doit le qualifier de collochfpersonnel, au sens où l'on parle de pronoms personnels. Ce qui revient à unifier men nominal et -men pronominal.
MACRO AND MICROSTRUCTURES OF CLASSICAL CHINESE TEXTS
The fruitful research on the configuration of classical Chinese texts by V.S.Spirin has proved the existence of certain structures of sentences or sentence patterns, including 2, 3, 4 or 5 phrases, as well as two-dimensional structures of two by two or three by three types (the first, as it was shown by V.S.Spirin, was originally called qi, the second -jing). The description and classification of such units, undertaken in the submitted paper may contribute to the reconstruction of microstructures in a classical Chinese text. The paper argues that these structures were made explicit not only by the numbering, but also by definitions of "lines" and/or "columns" in the text itself.
It was also previously shown that certain texts can be divided into such structures without leaving any "residue" (cf. the division of the Xi ci zhuan by V.S.Spirin and the submitted analysis of Dao de jing composition). So one may come to the conclusion that the above mentioned texts as a whole have a certain microstructures, though of a comparatively simple kind. At the same time one may expect to find the most intricately organized texts among the pieces belonging to the genre fu (paral1el prose). Also an elaborate organization is to be expected of the canonical texts dealing with music.
The submitted paper argues that the texts of Wen fu, Wen xin diao long and the first three divisions of Yue ji certainly have rather complicated and elaborate macrostructure's formed by microstructures and sentences fol1owing each other accordingly to certain rules. The structure of Wen fu as a whole is rather similar to the structure of chapters 1, 33 and 34 of Wen xin diao long. At the same time each division of the Wen fu has similar structure imposed by the structure of initial divisions of the text, as well as the structure of chapters 33-34 of Wen xin diao long is determined by the structure or the first chapter of this text.
The organization of each unit or these structures implicates a certain rhythm of succession of sentences and phrase pattern. The macrostructure of texts can be divided into those of 12-units kind, as well as three units (sonata form-ABA) and five units (rondo form-ABACA or ABABA) kinds. The first kind has as model the set of 12 ly, the last two the sets of sancai and wu xing (wu yin), thus corresponding exactly to the san wu concept as it was reconstructed in publications by V.S. Spirin, A.I. Kobzev and me. This rhythm may be made explicit by repetition of certain key words such as yue or li in the Yue ji. The text may be also divided into two symmetrical parts having the same or similar structure.
RECENT LEXICOGRAPHICAL TRENDS IN TAIWAN
The past decade saw a considerable increase in the number of reference works for the Southern Min variety of Taiwan (hereafter: Taiwanese). This paper focuses on lexicographical aspects of three recent dictionaries, viz. Wu Shouli's Comparative Mandarin - Taiwanese Dictionary for Practical Use (2000), Qiu Wenxi and Chen Xianguo's Practical Comparative Mandarin - Taiwanese Dictionary (1996), and the first two volumes of the Lexicon of Southern Min, issued by the ROC Ministry of Education (1998, 1999). It compares the macro-structures of these dictionaries as well as the accessibility, relevance, and quality of individual entries.
Special emphasis is laid on the intricacies of spelling modern Taiwanese expressions in the traditional Chinese character script. The compilers of the three reference works reviewed here share the claim that only the use of Chinese characters can do justice to the genetic affiliation of Taiwanese with Mandarin. However, the question of orthodoxy in the representation of Taiwanese expressions by means of characters continues to be a bone of contention. Etymological correctness requires that entries be represented by established characters for Mandarin cognates. On the other hand, the prominence of loan characters and the existence of conventional dialectal characters in popular writing traditions often outweigh etymological considerations.
LES "ESPACES BLANCS" ET LA TRADUCTION POETIQUE
Tout texte se compose d’une partie explicite et d’une partie plus ou moins importante de sous-entendu. Dans les textes littéraires, et surtout dans les textes poétiques, le sous-entendu acquiert une importance incontournable. On pourrait identifier les "espaces blancs" ou "vides" comme la voie de respiration et donc de vie d’une œuvre d’art. Au cours de la traduction se pose toutefois le problème de séparer les vides fonctionnels au texte littéraire de départ et qui lui sont propres des vides qui surgissent au cours de la transposition en raison de l’ignorance du monde culturel de la langue d’origine dans le monde culturel de la langue d’arrivée. Au cours du travail, les traducteurs devraient traiter les uns et les autres de façon différente, les premiers nécessitant d’être conservés, les deuxièmes au contraire d’être comblés.
En m’appuyant sur des exemples tirés de la traduction en chinois de poèmes italiens de Montale, je montrerai la nette tendance des traducteurs à remplir les "blancs fonctionnels" du texte original : à plusieurs endroits, ils expriment le non-dit et explicitent des relations syntaxiques supposées. Je mettrai en évidence la perte d’intensité des métaphores transformées, en chinois, en simples comparaisons. Les traductions chinoises de Montale donnent en générale une impression d’étouffement qui dérive de l’introduction de constructions explicatives, telles que génitifs, similitudes, etc.. Ces dernières ne concourent pas véritablement à la compréhension du texte. Il en résulte un affaiblissement de la fonction esthétique du langage en faveur d’une pseudo - fonction communicative.
En revanche, l’écart culturel suscite des vides injustifiées dans la version du traducteur. Ceux-ci ont en outre l’inconvénient de rendre le texte peu compréhensible pour le lecteur chinois. Face à ce problème culturel, les milieux traductologiques proposent une alternative, soit expliciter à l’intérieur du texte les indices culturels, soit avoir recours aux notes en bas de page. Bien qu’en poésie aucune des deux solutions ne soit tout à fait satisfaisante, des notes sont tout de même préférables à la désinvolture du vide. Cependant, dans le corpus des traductions chinoises de Montale, les notes sont très rares. Il est évident, dans plusieurs cas, que les traducteurs chinois eux-mêmes n’ont pas su lire convenablement plusieurs références à la culture occidentale, ce qui a donné lieu toujours à des obscurités, parfois à des contresens. Par exemple, dans le poème Anguilla, Qian Hongjia comprend et traduit sirena ("sirène") dans le sens de qiti sheng, "son de la sirène", d’où la grave distorsion du texte original, où l’on fait référence en réalité au personnage mythique mi femme mi poisson. Il faut cependant signaler que, dans une autre traduction, Liu Ruting interprète justement sirena comme meiren yu.
Li Renzhi (Dennis)
DOUBLE AND TRIPLE NEGATIONS
OF MODAL AUXILIARY VERBS IN MANDARIN CHINESE
This paper deals with the complexity in form and in meaning when modality occurs in a context with two or three negative markers. It describes the various syntactic structures of the sentences containing modality with two or three negators, examines the intricate relation between double and triple negations, and analyzes the semantic variation arising with the use of different negative words. It explores to what extent that the double and triple negations of Chinese modality is consistent with modal logic and in what ways they do not coincide. It explains why some modals (e.g. ‘dei’) allows no double or triple negation while some others do (e.g. ‘yao’ and ‘gai’) but semantically do not obey the logical equivalence suggested by Aristotle. It also explains how an negative marker (e.g. ‘fei’) evolves into a positive, emphatic necessity modal marker. It presents some views which are different from those documented in traditional Chinese grammar.
THE TURKIC PHONOLOGICAL SYSTEM AND THE YÙNJÌNG RHYME TABLES
The nomadic Turkic and the sedentary Chinese people shared a long history of contacts along the northern Chinese frontiers. Since the fourth century, the Altaics established several empires in northern China. The Tabghach empire (Northern Wei), in particular, adopted a sinification policy which nurtured a large number of Turkic bilinguals. These bilinguals developed contact-induced sound changes. The effects of such sound changes are reflected in the Qieyùn system.
For a medieval Turkic bilingual, Orhan Turkic of the eighth century was the primary language and the Chinese secondary language. Features of the primary and secondary languages are examined for their markedness or un-naturalness under contact condition. A major interference is attributed to the Turkic opposition of front/back (I.e., bi-polar: front and back poles) vowels and sound harmony. Sound interference resulting from Turkic bi-polar sound harmony was investigated in relation to the four-grade division of the Yùnjìng rhyme tables. The latter is an analytical tabulation of Qieyùn.Attempts to adhere to Turkic bi-polar sound harmony were registered in the four-grade division of Yùnjìng. Grade 1 sounds are associated with low#back vowels. Grade 1 was unmarked but modulated only for the acute consonants due to Turkic consonantalism, yet naturalized at the time of Qieyùn.
Grade-3 and grade-4 sounds were associated with the high#front vowels. Grade-4 sounds were unmarked and grade-3 marked to a Turkic bilingual. Sound changes were motivated by the vowels, but manifested in the dispersal or labialization of medieval Chinese initial consonants. The Turkic modulation of grade 3 sounds varied for different articulatory gestures: the dental, the gutteral, and the labial etc. For the dentals, it created palatals (grade 3) and retroflex sounds (grade 2). For the coronals, it caused principled centralization of tongue position. For the labials, velars and laryngeals, it caused principled rounding of the lips.
THE DEMARCATION OF WORDS IN MODERN CHINESE
The problem of demarcation of words did not pose for the Chinese until the twentieth century, when various systems of phonemic transcription were developed. With the use of Chinese characters, viz. the use of a morphemic script, the word boundaries are not visible in writing. Every character has the same distance from the others. But as soon as a phonemic script is employed, the units of speech have to be separated by gaps in writing in order to make a text readable. These units of speech are usually called "words ", however dubious this term may be.
The difficulty in Chinese lies in the fact that the Chinese speak by means of words, but on account of their long tradition of writing by means of characters they have never developed a sense of "wordness", to say nothing of coining a term for it. The term ci "word" did not appear until in modern times.
In China, a great discussion about the concept of the "word" set in in the fifties, when pinyin was being worked out. Since then, a pinyin orthography has been elaborated. But in many ways, the rules of orthography don’t seem convincing, mainly because the concept of the word underlying these rules is not as precise as it might be.
In this paper, some criteria for delimitating words and phrases will be examined. Such criteria are offered in Isaenko’s Opyt kitajsko-russkogo foneticeskogo slovarja (1957), Paul Kratochvìl’s The Chinese Language Today (1968) and in San Duanmu’s essay "Wordhood in Chinese" (1998). The attention is focussed on modifier-noun constructions, verb-object constructions and the combinations of verbs with the so-called "resultative complements", viz. resultative verbs.
EARLY SYSTEMS OF GUANHUA CHINESE ROMANIZATION
Since their first contacts with Chinese-speaking people in the Philippines or in China, missionaries of various order began to study the Chinese language, thus preparing the first tools to apprehend the language and facilitate social and religious intercourse with those populations. First in the Philippines and eventually in China western missionaries prepared various dictionaries of the Chinese language, all left in manuscript. In the Philippines, Augustinian, Dominicans and Jesuit missionaries chiefly edited materials on various Minnan Chinese languages, the Chinese dialects spoken by local Chinese immigrants. Eventually in China first the Jesuits and eventually the Dominicans missionaries began to compile the first bilingual dictionaries of Mandarin or Guanghua Chinese.
In order to compile these list of words or dictionaries the first problem they had to solve was how to develop systematic systems of Chinese romanization able to represent and distinguish the various peculiarities of the phonological system of late Ming Guanhua Chinese, such as aspiration, initial sonorant consonants, final consonants, tones, etc.
The paper will present a short history of these early romanization systems from Matteo Ricci's first systems, developed around the year 1600, until Basilio Brollo's system worked out at the end of eighteenth century. By comparing these systems and the solutions therein adopted, it will be possible to acquire new data about early Ming - late Qing Chinese guanhua phonology. These data will be also useful to determine genetics relations among these romanisation systems, determining also which and why some systems or solution adopted were most lasting and were also able to influence much later systems, such as those created by Protestant missionaries during the nineteenth century, thus establishing a bridge between the activity of nineteenth century Protestant missionaries and their Catholic forerunners.
THE NEGATIVE WEI IN SHIJI
This paper deals with the negative adverb wei in the Shiji, a text dating from the Former Han (206 B.C. - 8 A.D.) and its function as a marker of the aspect of the verb. According to Pulleyblank, it corresponds to the adverb ji which is assumed to function as an adverb indicating perfective aspect. Etymologically, wei seems to be a fusion of the negative root *m- in wu and the adverb ji meaning "not already" > "not yet" or "never". The adverb ji is found in combination with the final yi which indicates that a change of state takes place, while wei is found in combination with the final ye which indicates an unchanged, continuing state. Sentences with wei and ji seem to form the same dichotomy as sentences with meiyou and verbs followed by the suffix -le in Modern Mandarin do.
Differently from the Indo-European Languages, in which the categories tense, aspect and actionality are mainly marked by the inflectional system of the verb, in the Chinese language, particularly of the Classical and Han Period, these categories are almost exclusively marked by adverbs and a few final particles. The language of this period does not show any suffixes indicating different aspects of the verb as Modern Mandarin does. Possible morphological indications of aspect are obscured by the writing system, therefore one is confined to syntactical evidence to build up a coherent system of markers of tense and aspect.
This study, which is part of a larger study of temporality in Shiji, will show the constraints in the employment of the negative wei. Wei belongs to the group of adverbs proper that indicate tense, aspect and actionality of the verb. All these adverbs are confined to preverbal position - in contrast to those adverbs and adverbial phrases which can occur either in sentence-initial or in preverbal position - and are not seperable from the verb by any noun or noun phrase except for a prepositional phrase. In the case of the negative another exception is found, which is still operating during the Han-time: pronouns referring to the object can be inversed between negation and verb.
This paper will give some evidence for the interpretation of wei as one of those adverbs indicating tense, aspect or actionality of the verb. For this purpose, I will focus on the exact position of wei in the sentence, particularly in combination with prepositional phrases and additional adverbs. The possible combinations with additional adverbs, for example with another modal or with a manner adverb, and their relative position, will make obvious the precise position in the adverbial hierarchy occupied by adverbial wei. The semantics and actionality of the verbs negated by wei will be analyzed to find out if their are any semantic changes caused by the negation of the verb by wei. Besides this, the combination of wei with final ye will be analyzed.
WORD, SYLLABLE, OR MORPHEME IN PUTONGHUA.
This presentation is a langue discussion about the status of the word and the morpheme in the description of present-day Chinese, its main thrust being that reading Chinese is just a very complicated way of reading a syllabic notation. The crux of the argumentation is the status of the morpheme in words with internal constituent structure; in the domain of roots and compounds, verbal Chinese features an abundance of ‘bound morphemes’, which only appear in the syntactic and pragmatic context of an utterance; they are, therefore, not language proper, but verbal material that exists as a linguistic abstraction only.
In a European Saussure perspective the problem area is to what extent ‘bound morphemes’ in the domain of roots and compounds have a well definable component of meaning and to what extent they are productive. The stance on ‘morpheme’ in contemporary linguistics seems to be that the tag can only be used about an element which appears in another environment as an autonomous word in its own right, and that an element only deserves the right to be designated ‘morph’ if it instantiates a specific morpheme with a discernible, independently attested component of meaning.
in the synchronic mode a bound element is not a morph at all. Just a string of phonetic material.
My primary aim is to establish the word, rather than the Bloomfieldian morpheme, as the base unit in the description of present-day Chinese, my secondary to demonstrate that characters represent strings of phonetic material which a phonological analysis will divide into syllables rather than Bloomfieldian or Post-Bloomfieldian ‘morphemes’ as currently believed. In my analysis the prototypical word has two elements: dongxi has dong and xi; xiaoshuo has xiao and shuo, huoche has huo and che, etc. The central question addressed is whether these elements themselves have discernible components of meaning ¾ whether they are morphemes, or not.
My ultimate target is the assumption that the string of phonetic material, cut out by a Chinese word, morpheme, or character is co-extensive. I will argue that a character in isolation merely represents a specific stretch of phonetic material, whereas its meaning is indeterminate until the stretch has been embedded in a carrier sentence. My thesis about the status of ‘the word’ in present-day Chinese is informed by several theoretical sources, including Ferdinand de Saussure, Emile Benveniste, Robert H. Robins, Roy Harris, Stephen R. Anderson, Peter H. Matthews, and Jaromir Vochala. It is illustrated by investigating the domain of roots and compounds as it was originally set out in pioneering treatises by Chao Yuen Ren, for instance ‘A Grammar of Spoken Chinese’, or was subsequently reproduced in textbook presentations by Kratochvil, Henne, Li & Thompson, Ramsey, Norman, DeFrancis and Yip Po-ching.
SOME ISSUES IN CANTONESE POSSESSIVE CONSTRUCTIONS
Among the Cantonese classifier constructions (Matthews and Yip 1994) we shall consider in this presentation only possessive constructions and their use of classifiers. Only one (Pacioni 1998) of the scanty works describing Cantonese classifiers (Cheung 1989, Killingley 1983) is devoted to analysing possessive constructions and their use of the classifier. The aim of this talk is to further investigate some of the semantic and syntactic issues concerning these possessive constructions in Cantonese by exploring the differences among the four possessive constructions:
|
(1) |
ngo5 |
(ge3) |
tuhngsih |
(2) |
ngo5 |
ge3 |
tuhngsih |
|
I |
(GE) |
colleague(s) |
I |
GE |
colleague(s) |
||
|
‘my colleague’ |
‘my colleague(s)’ |
||||||
|
(3) |
ngo5 |
go3 |
tuhngsih |
(4) |
ngo5 |
di1 |
tuhngsih |
|
I |
CL |
colleague |
I |
DI |
colleague(s) |
||
|
‘my colleague’ |
‘my colleagues’ |
||||||
In formal Cantonese there is another possible possessive construction using DIK, the Cantonese reading for the Mandarin DE, as in:
|
(5) |
ngo5 |
dik1 |
jat1 |
sang1 |
|
I |
DIK |
one |
life |
|
|
‘my whole life’ |
(Matthews and Yip 1994:107). |
|||
We will also contrast these constructions in Cantonese with what is usually intended for possessive constructions containing classifiers in other languages in order to underlay that the descriptions in the literature are quite different from what is found in Cantonese. Recent findings from studies on Cantonese and on Child Language Acquisition of some semantic feature will also be considered.
REFERENCES
Cheung, Samuel H. N. 1989. The study of the use of Cantonese measure phrases (in Chinese). In Proceedings of the Second International Symposium on Chinese Studies, 753-774. Taipei: Academia Sinica.
Killingley, Siew. Y. 1983. Cantonese Classifiers. Syntax and Semantics. Newcastle upon Tyne: Grevatt & Grevatt.
Matthews, Stephen and Virginia Yip. 1994. Cantonese: A Comprehensive Grammar. Routledge: London - New York.
Pacioni, Patrizia. 1998. Possessive Constructions, Classifiers and Specificity in Cantonese. in (ed) S. Matthews Studies in Cantonese Linguistics. Hong Kong: Linguistic Society of Hong Kong. pp. 63-80.
Marie-Claude Paris
STAGE-LEVEL AND INDIVIDUAL-LEVEL PREDICATION IN CHINESE (AGAIN)
In this paper I will show how a good number of unrelated syntactic constructions in Chinese can be accounted for by the semantic distinction between stage-level and individual-level predication. In particular, I will concentrate on the distribution of constituents which can precede and follow durational complements.
First, I will try to explain the difference in grammaticality between (1) and (2), which are built on exactly the same syntactic pattern.
(1) ta jiao shu shi nian le
he teach book ten year F.P.
he has been teaching for ten years
(2) ??t�a� kan shu shi tian le
he read book ten day F.P.
Second, I will concentrate on the distribution of the final particle le. Why is the pair (3)/(4) acceptable, whereas in (5)/(6), (6) is ungrammatical?
(3) ta ku-le shi fen zhong Ø
he cry-Sfx. ten minute Ø
he cried for ten minutes
(4) ta ku-le shi fen zhong le
he cry-Sfx. ten minute F.P.
he has been crying for ten minutes
(5) ta gai-hao-le na dong fangzi hen duo nian le
he build-finish-Sfx. this Cl. house already many year F.P.
it's been many years since he built that house
(6) *ta gai-hao-le na dong fangzi hen duo nian Ø
he build-finish-Sfx. that Cl. house already many year Ø
Third, I will account for the relationship between the verb and its object as well as the presence or absence of the verbal suffix -le. Why are (7)-(8) well formed, but not (9)?
(7) ta gai-Ø Ø fangzi (yijing) hen duo nian le
he build-Ø Ø house (already) very many year F.P.
he has been building houses for many years (already)
he’s been an architect for many years
(8) ta gai-(hao)-le na dong fangzi hen duo nian le
= he build-finish-Sfx. that Cl. house already many year F.P.
(6) it's been many years since he built that house
(9) *ta gai-le Ø fangzi (yijing) hen duo nian le
he build-Ø Ø house (already) very many year F.P.
THE FINE STRUCTURE OF SHI...DE CLEFTS
Studies of the shi..de focus construction over the past two decades have clarified many aspects of its structure (Paris 1979, Huang 1982, 1988, Chiu 1993). One unanswered question, however, is the precise nature of the complement selected by shi in the various subtypes of the construction. We argue that the status of the complement differs in the adjunct focus pattern (1) and the subject focus pattern in (2). We observe that the former disallows Object Fronting (1a), universal dou (1b) or negation (1c) after the focused adjunct:
(1a) *Tamen shi zai 1997 nian Gugong qu de (cf. Tamen shi zai 1997 nian qu Gugong de.)
3pl be at 1997 year imperial.palace go de 'It is in 1997 that they went to the imp.p.'
(1b) Tamen shi zuotian (*dou) lai de
3pl be yesterday dou come de 'It is yesterday that they came.'
(1c) Tamen shi zuotian (*mei you/*bu ken) lai de.
3pl be yesterday neg / neg want come de
The subject focus pattern also disallows Object Fronting (2a) and dou (2b) but allows negation under pragmatically appropriate circumstances (Paris 1979), as we see in (2c):
(2a) *Shi tamen Gugong qu guo de.
(2b) *Shi tamen zuotian dou lai de.
(2c) Shi tamen zuotian bu ken lai de
be 3pl yesterday neg want come de 'It is them that didn't want to come yesterday.'
The data in (1) indicate that the complement projection with a focused adjunct cannot contain a landing site for Object Fronting, nor host dou or negation. These facts suggest that this projection is no larger than vP. The data in (2) show that the projection in the subject focus pattern, on the other hand, must be large enough to contain negation but not dou or the landing site for Object Fronting. Paul (2000) shows that the landing site of Object Fronting is higher than the position of negation. Taken together, these facts indicate that there must be a projection containing a position for subjects above negation but below the projection(s) hosting dou or fronted objects. We label this projection FP. The resulting structures are:
(3) [Tamen shi [vP zai 1997 nian [vP ttamen qu [VP Gugong ]]] de ] (Adjunct Focus)
(4) [Shi [FP tamen zuotian bu ken [vP ttamen lai ] ] de ] (Subject Focus)
It remains to be explained why the structure (4) is unavailable for nonsubject focus (Paris1979):
(5) *[shi [FP zuotian tamen [vP ttamen lai ] ] de ]
We propose that the focused element in the shi..de construction checks a focus feature on shi. This captures the widely held intuition that the shi ..de focus pattern is a type of cleft construction: in clefts there is a grammatical relationship not only between the copula and its clausal complement, but also between the copula and the focused element. In (3), the adjunct checks only a focus feature; the case-related features of shi are checked by the raised subject tamen. In (4), however, the non-raised subject must check both sets of features. We suggest that this is blocked by NPs or PPs interposed between the two (5).
ORIGIN AND EVOLUTION
OF QUESTION WORDS IN ARCHAIC CHINESE:
A COGNITIVE APPROACH
With the exception of the pronoun shei "who", which appeared for the first time in the Shi jing, all question words in Archaic Chinese are different from those of Contemporary Chinese, both in form and in use.
This paper will examine how the various forms of these question words arose as well as their evolution from the Early Archaic period (11th-6th c. BC) through the Late Archaic one (5th -2nd c. BC) to the Han dynasty Pre-medieval period (1st c. BC - 2nd c. AD). All the occurrences of question words meaning "who", "what/which", "how", "why", "when" and "where" will be presented in charts, as attested in the following documents: Shang shu and Shi jing for the Early Archaic period; Zuo zhuan, Lunyu, Mengzi, Mozi, Zhuangzi, Xunzi, Han Feizi for Late Archaic; Shi ji, Hanshu and Lun heng for Pre-medieval.
The four main claims made in this paper are:
Reasons, which might explain these evolutionary phenomena and characterizations of the differences between certain synonyms, will be proposed.
Some conclusions will also be drawn concerning the cognitive contours of the conceptual categories that correspond to the interrogative words in Archaic Chinese: PERSON > OBJECT > PROCESS / QUALITY > SPACE > TIME. The study of the transfers of meanings from one domain to another will also suggest that reduction in semantic complexity is far from being crucial.
LES CONNECTEURS DU CHINOIS CONTEMPORAIN
Les connecteurs, guanlianci en chinois, comme leur nom l’indique, couvre la classe des mots-outil qui remplissent la fonction de connexion entre deux phrases ou deux constituants (nous mettrons l’accent sur la jonction entre deux phrases dans ce présent travail). Ils désignent non seulement la catégorie traditionnellement dénommée conjonctions, lianci en chinois, mais aussi les adverbes ayant la fonction de connexion, guanlianfuci en chinois, que nous dénommons les adverbes conjonctifs.
Dans les recherches antérieures, notamment en langue chinoise, les connecteurs sont regroupés en deux grandes catégories : ceux qui relèvent du processus de coordination et ceux qui relèvent du processus de subordination. Les critères établis sont essentiellement, voire exclusivement logiques et sémantiques. Ainsi, nous trouvons des classes généralement assez confuses et hétérogènes entre les unes et des autres selon les auteurs. Il y a plusieurs questions qui préoccupent notre esprit:
dans la deuxième phrase ?
(4) y a-t-il des différences de comportement syntaxique lorsqu’un connecteur s’emploie isolément et lorsqu’il se trouve dans une structure particulière (en connecteur double et en construction corrélative) ?
Parti des réflexions et des recherches sur ces questions, nous essayons ici de proposer une classification générale englobant les principaux types de connecteurs du chinois contemporain qui peuvent être définis sur la base de critères proprement syntaxiques. Cette classification se divise également en deux grandes classes, comme le font la plupart des linguistes jusqu’alors, mais nous obtenons des sous-classes très différentes des leurs, du fait que nous ne prenons pas uniquement en considération des critères logiques ou sémantiques mais également, et même en priorité, syntaxiques ou syntactico-sémantiques.
Ce travail a pour but principal d’établir des critères syntaxiques séparant les deux grands types de connecteurs chinois. Il se constitue en deux parties : dans la première nous traitons les connecteurs de coordination, dans la deuxième les connecteurs de la subordination.
TRANSPOSITION PHONÉTIQUE ET CHOIX DES MOTS :
LE CAS DES TOPONYMES DES PAYS
OCCIDENTAUX EN CHINOIS AU DÉBUT DU 19E SIÈCLE
Les toponymes des pays étrangers sont pour la plupart transcrits phonétiquement en chinois. La règle générale pour adopter un nom de lieu étranger consiste à trouver l’interprétation la plus proche des sons étrangers dans la langue d’accueil. Toutefois, il s’avère que très souvent, d’autres éléments qui interviennent dans cette "adoption", se superposent à la règle d’interprétation phonique et modifient les données et le processus de transcription. Dans le présent exposé, c’est à travers l’analyse des différentes versions des noms de lieux occidentaux, citées dans une dizaine d’ouvrages publiés dans la première moitié du 19e siècle et à travers la comparaison avec des toponymes de l’époque du 16e au 18e siècle introduits par des jésuites et ceux de la fin du 19e siècle, que nous essaierons de déterminer quels sont les principaux éléments qui conditionnent l’adoption d’une version des toponymes étrangers, tant au niveau phonologique qu’au niveau sémantique.
THE EVIDENCE FOR SINO-AUSTRONESIAN
Based on recent research, an up-to-date account of the phonological, lexical and morphological evidence for a genetic relationship between Chinese, Tibeto-Burman and Austronesian will be presented. This will include a statement of the sound correspondences, a list of shared basic vocabulary items, and a list of shared elements of morphology. An initial attempt will be made to characterize the way of life of the speakers of the parent language. The homeland of the proposed family, and the (pre-)historical circumstances surrounding its break-up will also be discussed.
CODE-MIXING BETWEEN CHINESE, JAPANESE AND
AUSTRONESIAN LANGUAGES: TOWARD A
DEFINITION OF GRAMMATICALITY CONSTRAINTS
This contribution is to deal with code-mixing between typologically different languages: Mandarin Chinese, "pidginized" Japanese and Austronesian languages, as spoken in Taiwan by Aboriginal speakers. The data are all drawn from a linguistic field study conducted in the Hualien district (East of Taiwan) in 1996-1997. The aim of the paper is to try and identify valid constraints on such mixing processes, and if syntactic are found insufficient, to complete the analysis with informational and pragmatic constraints.
There have becn a few theoretical studies dealing with syntactic constraints on codemixing. Let us cite David Sankoff and Shana Poplack (1981), who base their analysis on linear order of constituents. They formulate the "equivalence constraint", which predicts that alternations between two languages tend to occur in specific points of the sentence, namely points where juxtaposition of elements does not violate syntactic rules of both languages, especially rules concerning linear order of constituents. In a more recent paper (Poplack and Sankoff 1988), they formulate a constraint applying to languages with different typological word order for constituants such as S, V and 0. Switching is declared to be possible only at points where there is a corresponding position in both languages. This constraint differs from the "equivalence constraint", bocause it is stated in terms of absolute position of constituents, rather than linear order.
Sankoff and Poplack's research was based on code-mixing utterances involving English and Spanish. Research on other language pairs has shown violations of the principles evolved from Sankoff and Poplack's observations. However, these violations were often compensated in interaction by "saving" mechanisms, such as repetition or omission of constituents. However, once applied to our data, these principles can be shown to be violated more often than not, without there being any saving mechanism. How can this be explained?
The first remark to be made is that typological order differs greatly between the languages used in our data, be it order of S, V and 0, or that of smaller constituents or elements in the sentence. If it can be argued ttat Chinese languages are SVO, Japanese is rather SOV, and Austronesian languages of Taiwan VSO (or at least verb-initial). As for order of elements inside constituents themselves, many differences can be observed.
Thus, Sankoff and Poplack's constraints, based on linearity and typological order, are visibly too strong to account for language pairs such as encountered in our data: they exclude on principle grounds ALL code-mixing between these languages.
As we shall see, not only IS code-mixing possible, but it very often violates the linearity- or order-based principles. More interesting is the fact that many of these codemixing utterances DO violate the syntactic rules of at least one of the languages involved. These violations do not seem to render the utterances difficult to understand for competent speakers. We shall start by classifying the code-mixing utterances according to the syntactic slot where the mixing occurs (inter- or intra-sentential, core or peripheral). Violations of syntactic rules shall be analyzed. Finally, we shall try and state new constraints in order to explain the facts, drawing on syntactic as well as informational or pragmatic structures.