會長程祥傚教技在開幕式上致開冪司 。 左起．鄧景濱丶 1王惠迪、巧家傑丶程祥微、冼為鏗、陳頌罄, . • --.;;,a~•••nu, I j,, ^ 海島市市政廳王席馬家傑致詞 新華社，奧門分社室教文體部顧問冼為鏗致可與會代表合照Hong Kong, Ma - da C 澳門理工學院副院長李向玉致詞
程祥徽教授ng,Macau e 01 ao d量一ia I 張國憲硏究員尹德剛教授張卓夫副理事長汪惠迪顧問 施關淦副主編---·--··-~· Jo PELA Associ 黃坤堯博士 熊懷芘女士沔関華講師 龍裕琛理事邵朝陽秘書長 鄧景濱理事長
敎育的應用"(1994: X X X) 。讓我總結一下。在我看來，形式主義與功能主義最基本的分歧是關於句法的獨立性(Syntactic autonomy) 的爭論。功能主義者拒絕接受這個概念，而形式主義者認為句法學是可以獨立於語意學及語用學的。 Newmeyer(1991)引述 Tomlin (1990: 7)的話可表示典型的功能主義的信念江Her(l997) 也引述〕句法學不是獨立於語意學及語用學的……針句法獨立的宣言是基於某些嵗察，這些嵗察包括一些語法形式的運用是和某些語義或語用的語篇功能的出玭有密切的關係，甚至有決定性的關聯。總的來說，功能主義者不會滿足於祇考慮語言的句法特徵而忽視其別的（包括社會）功能。2.3 匯合雖然以上這兩種研究方法互相敵對，它們也有些共同的特點。形式語法學者和功能語法學者都關心語言解釋。喬姆斯基的語法尋求｀｀形式的＂（內在的）解釋；而功能語法用語言的功能來解釋語言結構（外在解釋） (external explanation)。他們都以語言結構為研究對象並且以此為出發點。喬氏的語法在語言本身作抽象的概括，但功能語法學者多在不同的語言裡發現結構規律。這導致形式語法學者嘲笑功能（類型）語法學者祇做純描寫性的工作（收集語言資料）。相反地，功能語法學者譴責形式語法學者忽視基本的語言事實。無論如何，我認為這祇是一件事情的兩方面－－一側重的分別吧了。正如里奇 (Leech1983)說的，如果要完全理解語言，忽視其中一方面都是不智的。我認為這兩個研究方向假以時日（儘管一生的努力）是可以得到平衡的。霫多式和功能的爭論，我不在此多說。有興趣的學者，請參Newmeyer 即將出版的《語言形式和功能》 ("Language Form and Function") 。3. 漢語語言學町句法學）現代漢語語法研究可分為幾個階段(Huang and 匕 1997) 。階段始於馬氏文通(1898), 這是一部依印歐語法寫成的第一部現代漢16
語語法。第二階段是馬氏文通之後到六十年代。在這期間有關語法範疇的討論非常熱烈。其中也有一些重要的著作出現（例如黎錦熙(1954))。趙元任的中國語文法(1968)的出版可算是把漢語語法研究推向另一高峰。在之後十年間受這部語法影響的學者不計其數。正如曹逢甫說(1991 因ii)' 他的確｀｀曾為它裡頭許許多多有趣且詳盡精闢的分析而感到神迷＂。雖然趙元任的文法為後來的漢語研究打好基礎，些語言學家認為漢語語法可以從形式或功能方面着手研究（因為趙元任的語法是結構主義語言學的代表作）。1963 和 1965 的兩篇文章是最早以生成語法來分析漢語的兩篇重要文章。雖然王士元把興趣移到音韻學和語言變異，他的學生繼續研究漢語生成語法。在轉換生成語法最盛行的時期（在喬姆斯基發表 Aspect 之後），整個語言學界都在激烈討論深層結構（參看Harris 1994)。其中兩大陣容就是以 Lakoff 為首的生成語義學和以喬姆斯基為首的解釋語義學。與此同時，語言學家如湯姆遜(Thompson1973)和陸孝棟(Lu 1977)也辯論漢語的結果補語複合詞是否由詞彙規則(lexical rules)生成的，不過他們都沒有積極參予這場深層結構辯論。期間，其他語言學家亦發表了對漢語語法的研究（例如，Hashimoto 1971、湯廷池(1977),梅廣 (1972)和鄧守信(1975)等。在八十年代之前，語言學家很重視硏究語法的規則系統。但姆斯基提出他的管束論時(1981), 語言學家開始集中研究人類語言裡的普遍原則與參數。在漢語方面，黃正德(1982)的漢語的邏輯關係和雷拉去珝肆侖 (Logical relations in Chinese and the Theory of grammar) （中譯本由黑龍江大學內部發行）採用管束論寫漢語的邏輯關係，文不單對漢語語法研究有重大影響，就是普通語言理論（管束論）都受它影響。黃說明了漢語和英語的詞序是被參數化的 (parameterised)（參數祇容許兩個可能）。漢語的疑問詞 (wh-word)要保持在原位的(remain in situ), 就是說，無論在陳述句或疑問句裡，疑問詞都是保持在同一位置上。不過，在英語裡，疑問詞在疑問句裡就一定要移到句首才合語法。試比較：What do you want to eat? Ni yau chi shenme? 你要吃甚麼？英語的疑問詞在疑問句裡被移到句首位置，但反觀漢語的疑問句詞｀｀甚麼＂，它的位置跟在陳述句裡的位置是一樣的（此謂 wh-words
remain in situ) 。 亦因為甚麼＂是｀＇吃＂ 的補語 ( Complement), 而補語在漢語裡總是出現在動詞後面的。此外， 漢語的 wh-移動被認為是在邏輯形式上進行的。在這個原則參數語法裡， 我們可把兩種表面上不同的語言用普遍原則把它們連繫起來。這樣，我們發覺語言不是毫無限制地相異的。存在語言裡的差異（例如漢語和英語 ）其實可說是在軫把普遍語法的參數調整。從這觀點看來， 兒童語言習得祇不過是要固定這個參數，與外在世界的語言配合。其中一個有名的參數是 Pro-drop 參數，又名空主語假設(null sub」 ect hypothesis)。拿英語和漢語為例， 在英語裡，一個合語法的句子一定有（明顯的）主語。 例如 he speaks (他說話 ）， 但speaks 是不合語 法的。 （ 喬姆斯基 (1986) 提 出延伸投射原則(Extended Projection Principle ) 要求每個句子都要有主語 ） 。 然而， 葡語或漢語的主語就不一定在句子裡出現。例如 £ala (他／她說話）和｀｀說＂ 。 允許句子沒有主語的語言叫 pro-drop 語言。 漢語一般被視爲pro-drop 語言 ， 但有些研究者 （如 Cook 1988)認卮漢語不是一種完全pro-drop 的語言像葡語和意大利語， 而是介乎完全 pro-drop 語言和完全非 pro-drop 語言（如英語）之間的語言。 理由是漢語沒有顯示所有pro-drop 語言的特性。 兒童習得語言時就調定這個參數。自黃(1982)後， 許多語言學家都採取管束論或別的生成語法理論來研究漢語。 一個好例子就是李艷惠 (A Y~H Li) 的 Order and Constituency in Mandarin Chinese(l990 ) 。 它把好些黃(1982)沒有處理好的問題重新研究，例如漢語的抽象格(Abstract Case) 。 其他學者也用不同的理論來研究漢語。例如黃居仁(1987)就是用詞彙功能語法對漢語的｀＇的＂字結構進行研究。黃居仁－直致力構造漢語的詞彙功能語法。 像黃居仁一樣，何萬順(1990)亦提出一個修改的漢語詞彙功能語法。最近， 他把詞彙功能語法併入互動語法(Interaction Grammar) 裡(Her 1997) 。在功能分析方面 ， 曹逢甫(1976,1991)對漢語語法作語篇分析。當曹逢甫提及主題(topic )分析時 ， Li and Thompson (1976)也利用歷史證據支持其對漢語的主題分析。 八十年代， Li and Thompson (1981) 又著成漢語語法(Mandarin Chinese: a functional reference grammar) 。這部語法用功能主義的觀點來描述漢語。 儘管有些例子的合法度令人懷疑， 但它已成為一部標準的參考語法。至於其他較少影響力的語法也被應用於漢語的描述中。李英哲(1971)和湯廷池(1972)就是用格語18
法(Case Grammar)來分析漢語，不過， 現在已不合時宜了 。 劉香梅(1986)卻採用範疇語法(Categorial Grammar ) 作理論背景來描寫漢語的複合詞及短語。在功能類型研究方面，匕和 Thompson 仍然是非常活躍。 LaPolla (1990)在角色指稱語法之下以功能觀點研究漢語的語法關係，而且還引用歷時 (diachronic) 及共時 (synchronic) 的諠據。LaPolla 認启漢語的主語和賓語還未被語法化(gramma ticalised), 故主語和賓語這兩個概念未必適用於漢語。孫軔奮(1996)討論了漢語的語法化，尤其是把字句的語法化。儘管現時有各式各樣的研究課題，這方面的漢語類型研究還需要漢語語言學者的努力。至於其他的功能語法，系統功能語法在中圉受到重視。 自從八十年代以來，系統功能語法就發揮其影響力，因篦有些中國學者直接或間接受敎於韓禧德。韓禧德本人亦驚訝於系統功能語法存中圍迅速發展起來；他在澳門大學主辦的園際語篇分析研討會上對我說他也料不到他的語法理論在中國這麼有影響力。系統功能語法在中國的發展可見軫其中一個熱心的系統功能語法學者胡壯麟的著作。其中包括《系統功能語法概論》 (1989) 、《語言系統與功能》(1990) 、《中國功能語法的進展》 (1997)等。除了上述以上的漢語論文著作，很多學報都致力於發表漢語語法的研究。例如中國語言學學報 (Journal of Chinese Linguistics) 、Language 、 Linguistic Inquiry 、 Linguistics 及 Natural Language and Linguistic Theory 等。這些學報不單祇發表關於漢語的文章，而且很多文章都討論理論問題。譬如說，最近，討論論元結構 (argumentstructure)的文章有增無減。 Li(1990)討論漢語結果補語的論元結構。匕把 Grimshaw (1990) 的論元結構理論併入管束論來分析結果補語。這是一篇非常有影響力的文章，而且有很大的反響。 Huang and Lin (1992)採納了 Douty 的原型角色(Proto-role)理論，提出一個詞彙功能語法的分析來探討結果補語的論元結構。此外， Huang and Cheng (1994)也加入辯論駁斥了匕和 Huang and Lin 的理論。對漢語結果補語的論元結構的研究並沒有終此，我的一篇論文(Fong 1998)將對漢語結果補語作結構語法的分析。我亦期待更多語言學者以不同的理論觀點來研究漢語的結果補語。漢語語法中還有其他方面等着語言工作者去探討的。在美國、歐洲卜中圉、台灣及香港等地，以不同的理論來研究漢語的論文著作不計其數。遺憾地，我們不能逐一把它們介紹或討論。地區19
linguistic theory, s血ply because western linguistic theory is designed for western languages, especially English. So, it is not applicable to Chinese. There are, however, linguists who believe that the study of linguistic theory can better the study of Chinese. This paper will demonstrate that the former claim is a misconception while the latter is a more advisable 唧roach to Chinese linguistics. As a matter of fact, many Chinese scholars who had their education in America have been contributing to linguistic theory in general and Chinese linguistics in particular (eg for functionalism, see Tai and Hsueh 1989, Biq et al 1996; for generative grammar, see Huang and Li 1996). As Huang and 匕 (1996) point out, the study of Chinese has already made some significant contributions to general linguistic theory over the past decade. My atte唧t here is not to examine all these theories but to acknowledge that Chinese has been well studied within most of these theoretical frameworks, and some kind of unification may be achieved in the future. In fact, the recent 'Interaction Grammar'(eg Hsieh 1991; Her 1995) has been conceived to meet such a challenge. Although there are various linguistic theories available, it seems to me that linguists in Macau haven't recognised their existence and their application in Chinese. The purpose of this paper is to stimulate Macau linguists to conduct more research on Chinese within current linguistic packages. It is hoped that information from this study may be useful in other fields that assumes certain familiarity with current linguistic thinking and methods, such as (second/ foreign) language teaching. In what follows, I shall make some comments on the formal and functional approaches to linguistics. The paper ends with a note on one of the potential applications of linguistic theory—language teaching 2 APPROACHES TO LANGUAGE STUDY The linguistic theories I mentioned in the introduction can be further categorised into two broad orientations—functionalism and formalism. Theses two orientations have been characterised as opposed to each other simply because partisans of these two paradigms have quite 24
different ideas about how to do linguistics, and what counts as genuine linguistic pursuit. It will take up a whole book to write on this topic. What I'11 do here is just to briefly introduce these two approaches to linguistics and their relevance to the study of Chinese. I hasten to add that the umbrella terms of formalism and functionalism consist of various theories and approaches. 2. 1 FORMAL (GENERATIVE) APPROACHES As we all know, Chomsky's grammar has always been a generative grammar in the sense that it is explicit and formal. This can be illustrated by looking at the inadequacy of the traditional notion of sentence, which states that a sentence has a complete idea. Chomsky defined a sentence as a string consisting of a noun phrase and a verb phrase. That's a precise and accurate definition of sentence. A generative grammar should achieve three levels of adequacy, ie observation adequacy, descriptive adequacy and explanatory adequacy. Explanatory adequacy is the highest level, which attempts to explain how and why Children acquire their native language so 唧idly. The ultimate goal of Chomsky's Universal Grammar (UG) is to achieve this explanatory adequacy. Since 區 Standard Theory, Chomsky has been searching for Universal Grammar because he believes that the central aspect or core grammar of language is born with us一the innateness hypothesis. Chomsky's innateness hypothesis focuses on the difference between human beings and animals, but not between human beings themselves. It's natural that Chomsky postulates an idealised society where humans share the same grammatical c叩petence within the same linguistic community. Linguistic theory is primarily concerned with an ideal speaker-listener, in a completely homogeneous speech community who knows its language perfectly and is unaffected by such grammatically irrelevant conditions as memory limitations, distractions, shifts of attention and interest, and errors (random or characteristic) in applying 25
his knowledge of language in actual performance (Chomsky 1965: 3) Although this statement is often criticised by functionalists and sociolinguists, Chomsky has no interest in the relations between language and society. Chomsky postulates the ideal speaker-hearer and the homogeneous speech community as artificial constructs (Brown 1997) . The 這ealisation is necessary for any scientific enquiry. Chomsky has no interest in the actual use of language, either, though he of course will not deny that variations in language exist. We thus make a fundamental distinction between competence (the speaker-hearer's knowledge of the language) and performance, the actual use of language in concrete situations. (Chomsky 1965: 4) What Chomsky and his associates study is competence but not performance, while the functionalist study language performance but not competence . In other words, the formalist studies what Chomsky calls I-language (internalised language) and the functionalist stud民SE-language (externalised language) (Chomsky 1986). Thus, 正s the psychological aspect of language that the formalist is interested, while the functionalist is more interested in the societal aspect of language. Attempting to understand competence, ie the knowledge of language, Chomsky hypothesises that human beings are endowed with an innate language faculty to acquire a first language. UG then is the initial state of a language LO. Humans are said to share the same UG whose principles are universal but as the child makes contact with the world, UG will set the parameters for the language spoken in the outside world. So, a Chinese child and an English child will have the same UG but different parameters . We will look at the wh-movement parameter of Chinese and English when we deal with Chinese syntax in the next section. This 唧roach to UG is thus called the Principles-and-Parameters Theory (Chomsky 1981, 1986), which has been dominant in linguistics since the 1980s. This approach has recently 26
been incorporated into Chomsky's Minimalist Programme, though it has some different assumptions and has not been fully developed yet. Chomsky's model of Universal Grammar has changed over the last forty years from a system of specific rules such as the passivisation rule in English to a system of general universal principles that can explain phenomena of different languages. Chomsky's UG has the advantage over others that it can tie together a whole set of seemingly different linguistic phenomena across languages by a few general principles and parameters. ® In the 1990s, Chomsky has downplayed the importance of formalisation and attempted to make the theory as economical as possible reducing the principles to minimum, so that it would reduce the burden of children's language acquisition process (Chomsky 1993, .1995). In the following, I will only comment on the ma」or differences between Chomsky's theory and other rival generative theories like RG, LFG, GPSG, HPSG, CG. Like GB or PPT, other models are interested in seeking Universal Grammar. But PPT is a multistratal theory that it postulates more than one level interrelated by transformational rules, while other theories are monostratal in that they don't postulate a deep structure but generate the surface structure directly. No transformational rules are needed in these theories, and derivation doesn't play a role in the theories. PPT takes a modular approach and constructions are'epiphenomenal' (Newmeyer forthcoming) . Othe「唧roaches are not entirely modular like GPSG and not modular at all like CG. As influenced by the autonomy thesis, PPT doesn't recognise any connection between form and meaning except the abstract level of logical form. All of the theories above are claimed to be generative in nature. Nongenerative theories are the topic of next section to which we now turn. 2. 2 FUNCTIONAL APPROACHES® Like formalism, functionalism also has different varieties. We have already mentioned some, eg SFG, and FG, but there's also 27
functional-typological linguistics (eg Comrie 1981, Givon 1979, 1984, Hopper and Thompson 1981, Croft 1992)®. They, however, share some assumplions. In general, functionalists view language as a social phenomenon rather than a psychological one, and they place the emphasis on the communicative functions of language. Functionalists feel that formalists fail to account for some 血portant phenomena of language when the language is formalised. Functionalism is closely linked to typological stud民s, since the typological linguist regards it as'the hypothesis that linguistic structure should be explained primarily in terms of linguistic function' Croft (1990: 2) . This view of functionalism is similar to the one shared by Hallidayans. Halliday (1979) , for example, views 'language as a social semiotic', and that the fundamental function of language is social communication. Like functional-typological linguists, Hallidayans also attempt to explain linguistic structure in terms of linguistic functions. Construction grammarians have closer link with Hallidayans in that they don't treat semantics and pragmatics as separated from syntax. Hallidayans don't even make a distinction between semantics and pragmatics. 今Somewhat different from the Hallidayan 唧roach is that of cognitive linguistics (Lakoff 1987, Langacker 1987, 1991), which had its root in generative semantics in the 1970s. Newmeyer (forthcoming) characterises these theories including RRG as e:i·ternal functionalism . Another kind of functionalism is integrative functionalism, whose representative is Emergent Grammar (Hopper 1987), which emphasises the discourse function of language inseparable from the speaker. I am hesitant to include Fillmore's (1968) Case Grammar in functionalism, but Fillmore no doubt was interested in meaning and language functions. Fillmore found that the prepositions in English have various functions/ meanings, and they express cases like agentive, dative, instrumental, goal, etc. This is to me a purely semantic-based grammar, and was developed as opposed to Chomsky' 28
s TG. As the theory phased out of the scene, Fillmore (eg 1988) and his associates at UC Berkeley has been developing Construction Grammar, which finds its root in Case Grammar. However, as I remarked earlier, CG is a generative grammar and it is understood to be a formal grammar, though it stresses the semantics and pragmatics of language. Incidentally, Hudson (1976) also attempted to present a generative systemic grammar. (I want to add that sometimes, people are confused with or don ' t make a distinction between a semantic account and a functional one.) The Chomskyan 唧roach is frequently contrasted with the Greenbergian approach to linguistics (cf Croft 1990). Greenberg conducted a large survey of many of the world's languages and produced several volumes of his findings, one of which was the 'implicational universals'of morphology and word order. The type of universals is implication as opposed to Chomsky's absolute universals The Greenbergian approach is basically inductive, starting from the known facts of language to generalisations. This is condemned by Chomskyans for being not scientific enough because scientists usually adopt a deductive method. So, Chomsky's linguistics is deductive in that hypotheses are constructed before examining their validity in the languages available. Given that generalisations can be made by considering a single language like English, Chomsky has only studied a single language, his most fam山ar language一English . Other languages have been studied by his associates though. If the generalisations work for a language, they should also work for other languages. Otherwise, they are not universal principles! In addition to the Greenbergian 唧roach, the Chomskyan 唧roach is also compared with the Hallidayan 唧roach to language analysis Although the Greenberg皿唧roach sprang from the same tradition of the Chomskyan approach, the Hallidayan 唧roach has its root in a different intellectual tradition一the London School of linguistics after Firth, whose 沮eas were greatly influenced by the anthropologist and sociologist Malinowsky. Halliday first developed the scale and 29
category grammar in the 1960s, and nowadays, his grammar is referred to as systemic-functional grammar or systemic grammar for short. Although SFG has been influential in Britain and Australia (after Halliday moved to Australia in the 1970s, many linguists have been trained under him or his associates there), it hasn't received much attention from linguists in America (eg Fries is one of few close collaborators with Halliday living in America). Halliday's theory regards language as a system (hence, systemic) and as having the primary function of communication in society. Contrary to Chomsky' s grammar, SFG uses an extravagant rather than economical framework since language can be used to communicate many meanings. SFG is devised with assumptions that human beings need to express ideas and social relationships with others (Graddol et al 1994). Thus, language must have ideational and interpersonal functions. Moreover, since human beings communicate in pieces of speech and writing, these two functions conspire to make a textual meaning. These three types of meaning constitute the grammar of human language and also the backbones of SFG. Another 血portant aspect of Halliday's grammar is the concept of transitivity, which, different from the traditional notion, expresses the argument structure of sentences, ie the relations between participants and processes. In terms of 唧lication, Halliday lists in 犀 Functional Grammar (1994) more than twenty ways that a theory of language can be 唧lied. For him, the test of a linguistic theory is to see if'it facilitates the task in hand'. Halliday's own theory in fact'has been used for a variety of purposes: analysis of texts, spoken and written; stylistics; computational linguistics; developmental linguistics, and study of socialisation; study of functional variation of language; and for a number of educational 唧lications'(1994: xxx). Let me make one final remark about functionalism. The most fundamental difference between formalism and functionalism in my view is .the tenant of syntactic autonomy. Functionalists reject the 30
conception of autonomous syntax, while formalists believe that we can do syntax without considering semantics and pragmatics. Newmeyer (1991) cites Tomlin (1990: 7) to characterise the functionalist position as follows(also quoted in Her 1997): Syntax is not autonomous from semantics and pragmatics… The rejection of autonomy derives from the observation that the use of particular grammatical forms is strongly linked, even deterministically linked, to the presence of particular semantic or pragmatic functions during discourse. To sum up, funcitonalists will not be satisfied with the sole consideration of syntactic features of language regardless of the other (social) functions of language. 2. 3 SOME CONVERGENCE Although the two approaches are opposed to each other, there are some basic features that they share. Both of these approaches are concerned with explanation in linguistics, though the Chomskyan 唧roach looks for formal (internal) explanation in conjunction with the autonomy hypothesis, while the functional approach explains the structures in terms of function (external explanation). They both examine the structure of language and treat it as the starting point. The Chomskyan 唧roach makes a lot of abstract generalisations within languages while the functionalist abstracts linguistic patters across languages. This leads to the reason why formalists often have contempt for functionalists'purely'descriptive'work (collection of language data), while functionalists accuse formalists of overlooking the basic linguistic facts. However, I think this is simply the two sides of the same coin一a difference in emphasis. As Leech (1983) remarks, it'll be foolish to ignore either one of the aspects of language if we want to have a full underslanding of language. I believe that there would be a place for these two approaches to linguistics to converge although it may take a life time's work. The form-function issue has been one of intense debates in linguistics, and I refer the 31
interested reader to Newmeyer's forthcoming book Language Form and Function for this issue. 3 CHINESE LINGUISTICS®(SYNTAX) Several stages of the study of modern Chinese grammar can be identified (Huang and 匕 1997). The first stage began with the publication of Ma S加 Wen Tung (1898) which was based upon European grammar. The second stage began after Ma's until the 1960s. Within this period, Chinese grammarians debated about the issue of grammatical categories and produced some important work like L」 (1954) . The next stage saw the appearance of Chao's (1968) A Grammar 吁 Spoken Chinese, which follows the structuralist tradition of linguistics particularly influential in America before the , Cognitive Revolution'in the fifties. Since then, linguists working on Chinese has been influenced by Chao's monumental work. As Tsao put it (1991: vii), he'was greatly impressed by its many interesting, and often detailed, observations and analyses about Chinese grammatical structure'. Though Chao's grammar had laid the foundation for subsequent research on Chinese syntax, some linguists found that Chinese grammar could be better handled from a formal or functional model. W S-Y Wang's Some Syntactic Rules in Mandarin (1963) and Two aspect markers of Mandarin (1965) were the seminal papers in early Chinese generative linguistics. Though later Wang moved on to language change in phonology, his students continued the tradition of analysing Chinese within generative grammar. In the heyday of transformational-generative grammar, especially after Chomsky's Aspect was published, debates about deep structure surrounded almost the whole linguistic community (cf Harris 1994). This led to the debates between interpretivist semantics led by Chomsky and generative semantics led by Lakoff. Linguists like Thompson (1973) and Lu (1977) also argued about whether Chinese resultative verb compounds were generated by lexical rules, though they didn't actively participate in the deep-structure debate. At the same time, 32
other linguists also produced original works on Chinese syntax, for example, Hashimoto (1971), Tang (1977) Mei (1972), and Teng (1975). Up to the 80s when Chomsky launched his theory of government and binding, considerable attention was paid to the rule systems of grammar. Since then, linguists have concentrated upon finding the universal principles and parameters of human language. In respect to Chinese, 」 Huang (1982) Logical Relations in Chinese and the Theory in Grammar (the Chinese translation was produced by Hailungjiang University) was a significant work both on Chinese and on general linguistic theory. Adopting a GB 唧roach, Huang shows that the Chinese and English word orders are parameterised (parameters only allow two possibilities). In Chinese wh-words are said to remain in situ, meaning it stays in the same position both in declarative and interrogative sentences. In English, however, wh-words must be moved to the beginning of the sentence in interrogative sentences. Consider: What do you want to eat? Ni yau chi shemne? you want eat what Now the wh-word is moved to the front of the sentence in English but in Chinese it remains in situ, because Shenme is a complement of the verb eat, and complements in Chinese are always after verbs. It is also suggested that Chinese covertly undergoes wh-movement in Logical Form—the interface between syntax and meaning. In this principle-and-parameter approach, we can capture the similarity of two seemingly different languages, and so universal principles can 唧ly to languages. In this way, we show that languages don't vary without limit. The differences that exist in languages (in this case Chinese and English) lie in the setting of parameters in UG. The acquisition of language by Children is seen as s血ply a matter of fixing the parameters settings for the language spoken around them. One well-33
known parameter that explains the distinction between languages like English and Chinese is the so-called pro-drop parameter. This pro-drop parameter is also ref erred to as null sub」 ect hypothesis. In English, a well-formed sentence must have a (overt) sub」 ect, for instance, he speaks but *speak (s) (Actually, Chomsky (1986) postulates the extended pro_」ection principle that every clause must have a sub」 ect (in all languages)) . However, in Portuguese and Chinese, a well-formed sentence may not contain a (overt) subject at all, for example, fala (s/he speaks) and shuo (speak) as contrasted with the ungrammatical'speall in English. Languages that allow sub」 ectless sentences are called pro-drop languages. Chinese is generally considered to be one of them, but some researchers (eg Cook 1988) show that Chinese is not an entirely pro-drop language like Portuguese ,and Italian, but is somewhere in the middle between the completely pro-drop languages and completely non-pro-drop languages like English. Children are claimed to acquire their native language including syntactic structure by setting the pro-drop parameter. After 」 Huang's contribution to Chinese and linguistic theory, numerous linguists have worked within GB and other generative frameworks. One good example was A Y-H Li's Order and Constituency in Mandarin Chinese (1990) which takes up the matter that Huang didn't deal with-the issue of abstract Case in Chinese. Others also produces excellent works on Chinese in different frameworks, for example, C-R Huang (1987), which adopts the LFG in the analysis of the Chinese DE, and since then, C-R Huang has been developing a Chinese LFG . Following Huang, Her (1990) also developed his modified version of Chinese LFG and verb subcategorisation, and more recently, he integrated LFG into the more general interaction grammar (1997). Along the functional line, Tsao (1976, 1991) produced a discoursal anyalsis of Chinese sentences. While this touched upon the notion of topic, Li and Thompson (1976) argued for a topic analysis of Chinese 34
from historical evidence. In the early 80s Li and Tho唧son (1981) further produced a functional grammar of Mandarin Chinese, which has become a standard reference of functional Chinese linguistics. Other less influential frameworks have also been applied to Chinese 匕 (1971) and Tang (1972) presented a Case Grammar of Chinese, though it has been out of the fashion now. Liu (1986) adopts the Categorial Grammar approach to analyse Chinese. For functional-typological studies of Chinese, Li and Thompson are still among the most active scholars in this field. LaPolla (1990) is a study of the grammatical relations in Chinese within a broad functional framework such as RRG drawing on both synchronic and diachronic evidence of Chinese. It's further argued that'sub」 ect'and'ob」 ect'has not been grammaticalised in Chinese. Sun (1996) discusses the grammaticalisation in the history of Chinese, and gives a detailed account of the grammaticalisation of the BA construction. However, this area of Chinese linguistics still remains the field that needs further research. In terms of other types of functional linguistics, SFG has stood out in China. SFG has been influential in China since the 80s, when many now famous Chinese linguists had studied under Halliday or his associates. Halliday (pc) remarked to me in the International Conference on Discourse Analysis 1997 held at Macau University that he hadn't realised that SFG linguists would be so active in China. The recent development of SFG in China can be witnessed by some publications by one of the devoted proponents Zhanglin Hu, for example, A Survey of Systemic-Functional Grammar (1989), Language System and Function (1990), Advances in Functional Linguistics in China (1997). Apart from the book-length studies made by Chinese linguists, many journals have devoted to the study of Chinese, for example, Journal of Chinese Linguistics, Language, Linguistic Inquiry, Linguistics, and Natural Language and Linguistic Theory, etc. Not only many Chinese linguists publish their articles in these journals, but many of 35
them also address current theoretical issues. Just take one example for our purposes. Recently, there has been an increasing interest among linguists in the study of argument structure—the interface between syntax and semantics. Li (1990) addresses the thematic relations between the Chinese resultative verb compounds and the arguments involved. This paper adopts a GB account with Grimshaw's (1990) prominent theory of argument structure, and encouraged many theoretical debates on this issue. Incorporating Dowty's Proto-role theory into their own work, C-R Huang and Lin (1992) give an LFG analysis of resultative verb compounds in Chinese. Huang and Cheng (1994) also 」 oined the debate by producing their GB analysis attacking C-R Huang's and Li's. The discussion of argument structure of Chinese resultative verb compounds has not ended. For example, I am preparing an analysis of resultative constructions within Construction Grammar (Fong forthcoming). I should expect more original works from various frameworks. This is only one aspect of grammar that has been studied intensely. There are many other aspects of grammar being explored, or waiting to be explored. In addition, there are other numerous unpublished manuscripts and dissertations on Chinese using most of these theoretical frameworks done in Taiwan, China, HK, Europe and the United States. Unfortunately, we will not be able to review them here. Regional and international conferences on Chinese linguistics have been being organised worldwide. Chinese linguistics has received unprecedented attention from scholars around the world. Considering the bulk of publications on the interaction between linguistic theory and Chinese linguistics, we can be assured that it is not impossible to work on Chinese within any theoretical packages. Of course, there are linguists who don't work within any established frameworks. This in fact characterises the current research of linguistics here in Macau. The advantage of not subscribing to a particular theoretical framework is that linguists have more freedom in choosing what they think is appropriate, and can be eclectic about 36
different approaches. Eclecticism seems to be endorsed by many Chinese linguists. Nevertheless, it will be i唧ossible to do any serious work in linguistics entirely atheoretica!Ly. 4 LINGUISTIC THEORY AND LANGUAGE TEACHING I've 」 ust surveyed (though briefly) the current linguistics and Chinese linguistics abroad, and hope that there will be a reconciliation between the formal and functional paradigms. The place for these paradigms to converge could be the area of (foreign) language teaching, which is often treated as an equivalent to applied linguistics . Chomsky has been denying the use of linguistic theory in language teaching and doesn't see any connection between the two. However, many applied linguists have attempted to incorporate linguistic theory in language teaching. This can be witnessed by the collection of papers on pedagogical grammar (Odlin 1994)閂 Language teaching is more closely connected with functional grammar than formal grammar for obvious reason. Foreign language learners, for example, would be more interested to know what linguistic patterns can achieve what functions. An interrogative sentence can be used to ask a question, make a request or even express disapproval, etc. The form一functionissue is no less relevant in foreign language teaching and learning . I know that many theoretical linguists are also interested in language teaching店pplied linguistics . In fact, the more we understand second/ foreign language learning/acquisition the easier we can teach the second/foreign language to our students . In the teaching of Chmese as a second/foreign language, I suggest that we could make some observations about linguistic facts of Chinese and present them to our students descriptively. As they advance, we could then make some abstract generalisations about Chinese, for instance, the null sub」 ecthypothesis or a topic analysis of Chinese. In this way, students will learn that in Chinese unlike English, the sub」 ect could be absent. We can also compare the student's mother tongue, say, English with Chinese in this regard. 37
5 CONCLUDING REMARKS To recapitulate, the purpose of this paper is to encourage more linguistic research on Chinese within these theoretical packages. To the best of my knowledge, there is little work if any in linguistics that has analysed Chinese within current western linguistic theories (as those discussed above) by linguists in Macau. But if we look at HK, Taiwan and China, we'11 find that linguists have already devoted themselves to all sorts of linguistic theories and have already been 唧lying various theories to the study of Chinese . There is no reason for us not to contribute to this exciting enterprise. I'll be 加ppy to see something similar in Macau, and more interested to see how Cantonese can be analysed in terms of western linguistic theory. To quote the famous linguist S-X Li.i(l980) (quoted in the preface to Tai and Hsueh by B Li.i 1992), We must know that the study of Chinese has become the world's interest; many scholars in their countries are studying the Chinese language, history and art. Their research findings are of great value to us. While we don't have to accept all of them, we must be aware of them. If we are still trying to solve the problems that have been solved by others, wouldn't it be absurd?® Selected References Biq Y et al 1996 (eds) Recent developments in functional approaches to Chinese. In Huang and 匕 1996Bresnan 」 1982 (ed) The mental representation of grammatical relations . MIT: Cambridge, MA Brown G 1997 Introduction. In Performance and Competence in second language acquisition. Brown G, K Malmk」ce r and 」覇lliams (eds) CUP: Cambridge Chao YR 1968 A grammar 吐 spollen Chine!fe. University of California: Berkeley Cheng LLS and Huang CTJ 1994 On the argument structure of resultative compounds. In In honor of William S-Y Wang. 38
Pyramid: Taipei Chomsky N 1968 Aspects of the theory of syntax. MIT: Cambridge Chomsky N 1981 Lectures on government and binding, Dordrecht: Foris Chomsky N 1986 Kno國edge of language , Praeger: NY Chomsky N 1995 The Minimalist Program , MIT: Cambridge Croft W 1990 Typology and universals. CUP: Cambridge Cromie B 1981 Language universals and linguistic typology Chicago: Chicago University Press D出 S 1981 Functional grammar. Dordrecht: Foris Fillmore C 1968 A case for case. In Universals in linguistic theory. Fillmore, C 1987 The mechanism of'construction grammar'. ELS Fong IWR (in preparation) Oiinese Resultative Constructions, Gazdar G et al 1985 Generalised phrase structure grammar. Harvard University Press: MA Givon T 1984 Syntax: a functional-typological introduction Benjamins: Amsterdam Goldberg AE 1995 Constructions. Chicago University Press: Chicago Haegeman L 1994 Introduction to government and binding theory. (2 ed) Blackwell: Oxford Halliday MAK 1979 Language as a social semiotics. OUP: Oxford Halliday MAK 1994 An introduction to functional grammar. (2ed) Edward Arnold: London Hashimoto A Y 1971 Mandarin syntactic structures. Unicorn 8, 1 -149 Harris R 1994 The linguistic wars. OUP: Oxford Her 0-S 1990 Grammatical functions and verb subcategorisation in Mandarin Chinese. Hawaii University diss (also published by Crane, Taipei 1991) Her 0-S 1997 Interaction and variation in the Chinese VO construction. Crane: Taipei 胡壯麟等 1989 系統功能語法概論，湖南敎育出版社胡壯麟 1990 語言系統與功能，北京大學出版社胡壯麟 1997 功能語言學在中國的進展，清華大學出版社39
Hudson R 1984 Word grammar. Blackwell: Oxford Huang C-R 1987 Mandarin Chinese NP DE. Cornell diss (also published by Academia Sinica, Taipei 1989) Huang C-R and Lin F-W 1992 Composite event structures and complex predicates. In Proceedings of the 3rd annual meetings of the formal linguistic society of America . Huang CTJ 1982 Logical relations in Chinese and the theory of grammar. MIT diss. Huang CTJ and AYH Li 1996 Recent generative studies in Chinese syntax. In Huang and 匕 1996Huang CTJ and A YH Li 1996 (eds) New hor國ons in Chinese Linguistics. Kluwer: Amsterdam Lakoff G 1987 Women, fire and dangerous things. Chicago UniversitY Press: Chicago Langacker R 1987 Foundations of cognitive grammar vol 1 Stanford University: California Langacker R 1991 Foundations of cognitive grammar 這 2.!3tanford University: California 黎錦熙 1954 新著團語文法匕 C and S Thompson 1981 1\llanclarin Chinese: a functional reference grammar. University of California: Berkeley Li YHA 1990 Word Order and Mandarin Chinese. Kluwer: Amsterdam Liu X 1988 A categorial grammar analysis of Chinese separable compounds and phrases. Student Book Co: Taipei 呂必松 1992 總庄。 在戴和薛((Tai and Hsueh) 1989 的中譯本Lu 」 H-T 1977 Resultative verb compounds vs directional verb compounds in Mandarin. Journal of Chinese Linguistics 5, 2, 276 - 313 馬健忠 1898 馬氏文通Martin C et al 1997 Working with functional grammar. Edward Arnold: London Mei K 1972 Stud認s in the transformational grammar of 1\llandarin standard Chinese . Harvard diss 40
Newmeyer F 1991 Functional explanation in linguistics and the origin of language. Language and Communication 11, 1 / 2, 3 - 29 Newmeyer F (forthcoming) Language form and function. MIT: Cambridge, MA Perlmutter D 1983 (eds) Studies in relational grammar 1. Chicago University Press: Chicago Radford A 1997a Syntactic theory and the structure of English CUP: Cambridge Radford A 1997b Syntax : a minimalist introduction. CUP: Cambridge Tai 」 and F Hseuh 1989 Functionalism and Chinese grammar. Chinese Language Teachers'Association. Tang TC 1972 A case grammar 吐 spoken Chinese. Hai-Guo Book Co: Taipei 湯廷池 1977 國語變形語法研究（第一集 ）， 學生書局Teng S-H 1975 A semantic study of transiti叨ty relations in Chinese UCP: Berkeley (also published by Student Book Co, Taipei) Tho唧son G 1997 Introducing functional grammar. Edward Arnold: London Tho唧son S 1973 Resultative verb compounds in Mandarin Chinese: a case for lexical rules. Language 49. 2: 361 - 379 Tsao F-F 1979 A functional study of topic in Chinese: the first step towards d為course analysis. Student Book Co: Taipei Tsao F-F 1990 Sentence and clause structure in Chinese: a functional perspective . Student Book Co: TaipeiVan Valin R 1993 Advances in role and reference grammar . John Benjamins: Amsterdam Van Valin R and R LaPolla 1997 Syntax . CUP: Cambridge Wang W S-Y 1963 Some syntactic rules for Mandarin. The Ohio State University Pro,」ect in Linguistic Analysis, 3 Wang W S-Y 1965 Two aspect markers in Mandarin. Language 41, 457 - 470 徐烈炯 1988 生成語法理論，上海外語敎育出版社41
Footnotes (DThis paper was based on a speech I gave in the conference of Linguistics Society of Macau. I thank the Society. I also thank Prof D-G Yin, who first suggested that I deliver a lecture on linguistics abroad, and suggested the Chinese title of this paper to me. Iok Feng Lam helped me with the Chinese translation of the abstract. I'm sure there will be some controversies about my views on the sub」 ect. I do hope, however, that this paper can stimulate more research on Chinese linguistics, both theoretical and descriptive. @Readers who are interested in Chomsky's model of grammar should consult the following works: Xu (1988) is written for Chinese readers, and it discusses the GB model. Radford (1997a, b) are complementary; one is a shorter version of the other. They are the most updated introductions to the Minimalist Program, albeit technical in some respects. Haegeman (1994) is a textbook introduction to GB. @Sells(l985) is a good introduction to three theories, ie GB, GPSG and LFG. Horrocks(l987) covers more ot less the same theories but in more detail. @For Hallidayan linguistics, interested readers should read Halliday's major work (1994). A series of books published by Deakin University for the course Specialised curriculum : language and learning adopts Halliday's framework Thompson (1997) and Martin et al (1997) are introductions to Halliday's Functional grammar. Chinese works include those by Hu (1989, 1990, 1997) as mentioned in the text. @For introductory work on functional-typological linguistics, see Cromie (1981), Croft (1990), Givon (1984), etc. ®In the International Conference on Discourse Analysis held at Macau University in October 1997, Halliday in response to my question remarked that he doesn't make a distinction between semantics and pragmatics. 叨Some people like my colleague Steve Schackne think that it is problematic to use the term'Chinese linguistics'if we understand linguistics as the scientific study of language. However, it would be clumsy to say'the scientific study of the Chinese language', but if we insist on the strict sense of the term, we cannot even speak of'applied linguistics'since it is already subsumed under 'linguistics'. I still use the obviously inaccurate term for lack of a better one. Besides, it is now widely used in the literature I'll only discuss the work on syntax here and hope that other aspects of Chinese linguistics like phonology, semantics, and discourse, will also be introduced in 42
future work. ®This is a collection of papers devoted to language teaching from various theoretical perspectives including UG, LFG, etc. ®My own translation. 43
b. 身子非常軟。(8)a 我這樣的高幹子弟，被有的人稱為水箱。……打開門，程面通明；闕了，程面就黑暗，冷着。b. 冰箱程很冷。爲了表述的方便，本文把出現在 a 句中的形容詞稱之爲變化形容詞。情狀是一種意義類別，但在語法特徵上也有其形式上的表現：變化形容詞可以在"NP+ 已經十 十 了＂和"NP+ 沒+ "這兩種句法槽中出現，表示對某種狀態變化的肯定和否定，而狀態形容詞不具備此項語法特徵。l. 3 狀態與變化的關係恆定事件、動作事件和變化事件是漢語的三種典型事件，典型形容詞（性狀形容詞）善長於表述恆定事件，但有些非典型形容詞（變化形容詞）則可以表述變化事件。狀態與變化是兩回事，但不是渺不相關的兩回事，事實上是息息相通的，有着十分密切的聯繫。變化形容詞作為陳述性成分其所指有一個隨時間展開的漸變過程，它可以有起始點、終結點和續段。從系統論觀點出發，我們可以把變化過程以及變化所涉及的始末視為一個大的系統，這個系統由初始狀態、變化過程和完結狀態三個子系統構成。變化過程始於初始狀態，變化過程的結束便產生一定的結果，進入另一種狀態，所以從系統的角度說，過程祇是狀態的轉換。狀態和變化二者之間的關係可以用下圖來表示(T 是時間軸， I 代表過程開始的時點， F 代表過程結束的時點）：均質 I 異質 F 均質初始狀態 變化過程 完結狀態下面的例句可以很好的說明這種現象：(9)(她）病好了，一下子胖了，胖得像個禾場滾子，腰間一掄輪肉往下垂。前一個｀｀胖＂是動態的，表示由非胖狀態（初始狀態）的異質變化，｀｀了＂表示變化過程的完結；後一個｀｀胖＂是異質變化的結果，從而又轉入均質的靜態。可見，一種變化過程的終結，也就是一種新狀態的實現。變化過程的起始點處於始發狀態，變化過程的結束便進入完結狀態。由於變化形容詞表現一種隨時間展開的動態過程，所以與體密切相關。61
貳形容詞的體形式2.1 形容詞的體系統體反映句子所表達事件的構成，對事件的觀察可以有不同的角度：一是把事件作為一個不予分解的整體，從外部來表述事件的構成，以表現事件的未然(A 點）與經歷(E 點）；－是對事件進行分解，從內部來表述事件過程的各個階段，以表現事件的起始(B 點）、終結 (D 點）、持續和繼續(C 點）。圖示如下：A B C D E T . . ,- - -·. I 事件 j體作爲一種語法範疇，有其特有的體形式，現代漢語裡表示形容詞體意義的典型形態形式主要有三個：a 實現體｀｀了 ";b. 經歷體｀｀過"; c. 持續體｀｀着＂。非典型形態形式主要有兩個：a 起始體｀｀起來";b. 繼續體｀ 下去＂。本文着重考察體的典型形態形式。2.2 實現體：了2.2.1 形容詞翼｀了＂同現的句法格式｀｀了＂標記新狀態的實現，其作用是凸顯完結或終結狀態。由形容詞加｀｀了＂構成的線性序列主要有以下幾種形式：Sa: 名已形十了Sb: 名主汀戸了十數量S己（名主) +形+了十 名賓2.2 . 2 關於 Sa :名己形十了＄是形容詞實現體的基本格式，格式中的形容詞都具有[ +自變］語義特徵，它能和相關的詞一起構成一條連續變化的鏈，以表達自身動態變化實際到達的位置。就變化的方向而言，有些狀態的變化祇具有單向性：年輕一_老生—>熟新—>舊矮-高緊-------鬆快一>鈍在單向性變化鏈上，前項是客觀事物先天的固有的自然屬性，後項的情狀通常是在無外力［－創造］的狀況下事物發展的終結狀態，二者不具有可逆性。由此，前項屬於性狀形容詞，後項屬軫變化形容訶，祇有後項才能與｀｀了＂同現。例如：(10) 到妒今，大祜老了，沒人上門提晟了；拾來大了，辰得又高又大，堂受一條漢子，幹活·拿九分五的工了 。62
(31)初來時，穿一身扛本衣裳，見人就亻~1 眉眼，不肯說語。(32) (奶牛）針了尾巴拉下盆子大一堆糞來，那糞不成形，像甩下的一把稀泥。(33)眼角餘光程瞑見秘書長還站在那程，手一時收不回去，卻慢慢彎了指頭。由此，該格式中的一些詞語儘管從語法形式上看可以變換成 3 格式，但語義上並不一樣， S 是意志格式，而 S 則是非意志格式。比較：針了尾巴-尾巴針了 彎了指頭一_指頭彎了特別值得注意的是，通常不與｀｀了＂同現的［－自變］形容詞也可以進入該格式：(34)莊之蝶送走了柳月，就堅定了自己不再寫作的念頭。(35)但邱老康不敢多問，思想到遠在天涯的程順，恨賊小子一走那麼久，冷落了滾燙燙的匡子的心。由於 svo 句式形成了一個強勢語義場，該格式中的個別詞（如｀｀活躍、普及、突出、嚴蕭＂等）可以不借助｀｀了＂而仍然表示動性，這與 Sa和 3 迥然不同。從功能游移的角度看，這類用法十分接近於動詞。因此語法學者在操作上多處理為形動兼類詞。從我們收集的材料看，＄多為意志句。2.2.5 形容詞與｀｀了＂同現的三種格式大致可以反映變化形容詞的游移軌跡，因此，三種格式既是一種共時現象，又是一種歷時現象。在游移的過程中，［十自變］語義特徵是形容詞功能游移的主要誘因，隨着游移的歷程，其特徵不斷地弱化，由此[+自變］語義特徵在三種格式中由高至低的等級序列：Sa>Sb>Sc。伴隨着[+自變］語義特徵的弱化，動性和[+可控］語義特徵逐漸地增強，呈現出一個與[ +自］語義特徵相應的由低至高的等級序列： Sa
如：(22)我有買今晚妁戲票。(22a)我貫了今晚的戲票。(22b)我是買了今晚的戲票。(22c)我已經閌了今晚的戲票。例(22a)及(22c)屬於正常的、合乎規範的表達，以｀｀了"'"巳經＂之類來突出｀｀已然＂的語境。例(22) 的原句所要傳達訊息，除了｀｀買＂這一重要的涵意外，還着重｀｀有＂的肯定語氣。換言之，這句的意思是＇＇的確買下了＂或｀｀眞的買好了＂。如此一來，所帶的強調性便由｀｀有買＂這樣的表達法顯現無遺。當然，這一表述壓根兒是源於方言性的表達。如果以較規範的華語表達，祇能用｀＇是＂字句來發揮強調的作用。®借＂有＂起強調作用近乎英語的用"do"(過去式則｀油d")加於動詞之前，使之成爲｀｀羨餘成分＂，卻是顯而易見的｀｀強調式＂，如"do come"之類。釕郎良偉曾述及｀｀有＂字句的強調法；他言簡意賅地指出：強調動作或現象的發生的時態台灣官語不用動詞詞尾勺＂（如｀｀去看了他＂）而用助動詞｀｀有＂（七｀｀有去看他＂），這黠台灣官語的影菩大於善通語。®這類蘊含強調意義的｀｀有＂，稱之爲｀｀助動詞＂，並不為過。這有別於前文所論的副詞性的用法。不過，是否一定歸之為助動詞，在語法界仍非定論。祇要將一些帶｀｀有＂的強調句，與英語的相應表達相對照，當能深一層體會其修辭意味：(23)我們豈告訴他這宗案件，但他忘了。(23a)We 叫 tell him about this case, but he forgot. (24)我徑通卸他。(24a)I 輯 inform him. (25)我豈打電語絳他，但找到不他。(25a)I 血 telophone him, but I could not get him (26)我豈說過那林的語。(26a)l 血 say that. (27) 他徑拿我的東西。(27a)He 叫 take my things. 經此比較，不難發現漢語的｀｀有＂字居然微妙地發揮英語"did+V"的特殊修辭效果。 英語的"did"( 或"do")原來用於問句，但與動詞（不81
談補語陸世光從語言的比較來看，補語可以說是漢語的一個特點了，它的表達方式，表意內容都有其自身的特點和規律，有別於其他許多語言，對於初學的人尤其是把漢語當作第二語言學習的人是十分重要的，同時又是難於掌握的。補語，顧名思義是補充或補足的意思，語法書上很早就有補足語名稱，但那時所說的補足語與現在說的補語有很大的不同。 以前管補語叫作後附加語（與前附加語相對）或叫動詞形容詞後附加語，有的也叫後附的副詞性附加語，還有的叫動詞、形容詞補足語或末品補語等等。比較起來，現在稱作補語似乎又簡明又確切， 很能反映出所表示的內容，可以說是名副其實的了：｀｀補語是動詞或形容詞後面的補充說明成份＂（丁聲樹等）。早在五、六十年代就有不少著作對補語作過詳細的論述，其中有趙元任的（漢語口語語法》，丁聲樹等的《現代漢語語法講話》等，尤其是後者影響很大很廣，成爲敎學科研以及敎科書編寫的重要依據和參考材料。近些年來許多語法學家十分重視對漢語補語進一步探討，從各個角度，形式的、意義的、句式變化的、分類的等加以描述。李臨定先生的《動補格句式）是着重從形式上加以描寫和分類的，他把以動補格為謂語中心的句式分為五類：l. s ve 2. s VO ve 3. s vc 。4. s VOJ vc 02 5. S 把 01 ve 0 2 並對各類句式的構成，句式變換以及表示的語義等進行闡述。呂叔湘先生則從另一角度即語義方面對補語進行了精細的描寫。 他在153
雙聲與疊韻向倩"雙聲＂是指組合在一起的字音（音節）的聲母相同，＂堊韻＂是指組合在一起的字音（音節）的韻母相同。｀｀雙聲＂、｀｀聲韻＂是漢語的一種特殊的語音現象。在古代漢語裡，雙聲疊韻是一種重要的構詞手段。這是因為漢語是一種詞根弧立語，構詞大多以單音節詞根為組織成分。這樣，在運用復合法組合詞根構成詞時，就很容易出現聲母相同或韻母相同的博況，而人們之有意識地採取這種方式構造復音詞，便自然造成了大量的雙聲詞和疊韻詞。在現代漢語裡，也有很多新詞是用這種方法造出來的。因此，雙聲詞和疊韻詞在漢語中是屢見不鮮的。如：渺茫[n 丨幺叮1 忙］飄潑［女丨幺／女乞］紳聖［尸序尸乙｀］懸想[ T Ll 卑T 丨 jc."]洋溢［丨允汀｀］游弋[ I x'l'J 妥貼［六乂乞勺計廿］裊娜[}I 幺丐乂乞勺清擢[<.IL<.LJ勺祈求[ <. 戶丨沢］搖曳［丨幺＇丨廿｀］隱逸[ I 与汀｀］琵琶［女「女丫勺洞達［力乂乙汾丫I]以上是雙聲詞，以下是疊韻詞：讚嘆［？月｀六弓｀］蹣躅［女卑＊叨啓迪[ <幺力「］艱險［甘丨弓 T 丨牟］比翼［与鬥｀］情景[
演門語言學刊 第 6、 7 期 1998.5.主編成員程祥徽 邵景濱鄭煒明陳項聲張卓夫本期主編邵景潰攝 影郭濟修邵朝陽封面題字朱乃正封面設計 廖文暢出 版澳門語言學會會址 ：澳門大學中文學院 T203 室 電話 ： 3974201 電傳： 840353 贊 助澳門文化司暑植 宇溴門新藝電腦桂字排版公司印 刷澳門華輝印刷有限公司定價 30 元Revista da Ciencia Lin~uistica de Macau N? 6 e 7 Maio de 1998 Editor:Ching Cheung Fai 、 Tang Keng Pan 、 Cheng Wei Ming 、Chen Song Sheng 、 Cheong Cheok Fu Editor Executivo: Tang Keng Pan Fotografo: Kuok Chai Sao I, Sio Chio leong Caligrafia da Capa: Chu Nai Cheng Coordenac; 袖 grafica: Liu Man Cheong Publica函o:Associa啤o da Liencia Cinguistica de Macau Aderec;o: Sala T203, Institute de Chinesas Studies, Universidade de Macau Tel: 3974201 Fax: 840353 Patrocinio: lnstituto Cultural de Macau Composi亞o:san Ngai Tipografia Computadorizada Co lmpressao: Tipografia Welfare Co Prec;o: 30 Pa tacas
