一 一 －一 ·一 — -——一·一 ·— 一· —· 一·一一·一 ·— ＇一·-—·n 其慟第總期1 第年9 。。2 3" 3· r· e. b" m~ U· N" . · . . . · . . . . . . · . . · · . . . －一 ·r —-—-- -·- — 一— 一 ·一一一－sophisticated cultures must have needed to express themselves (1964: 104-118, 274-322). Although the'war'seems to be over by now, the controversy remains unresolved. In contemporary linguistic literature, one can easily take the monosyllabic approach for granted,c21 while others may freely take the disyllabic property as givenY1 It seems that no one would deny that modern Chinese is a monosyllabic language in the sense that every syllable has a meaning, while disyllabicity is also a property in its morphology. So the problem we face is: if Chinese is a monosyllabic language, how could it contain more than 80% disyllabic words? On the other hand, if disyllabicity has become a characteristic of the language, how could it still sustain a monosyllabic property? As Tang (1989:569) has put it: "We can certainly assert that the tendency towards disyllabicity of the Chinese lexicon will be developed further rapidly. Of course, we must realize that although the number of monosyllabic words is not high, the frequency of their occurrences is much higher than that of polysyllabic ones. As a result, the question of whether Chinese is a monosyllabic language or a polysyllabic language is still difficult to 1. Introduction There has been a'war'in the field oflinguistics over the syllabic property of Chinese language/s: whether Mandarin Chinese is a monosyllabic language or a polysyllabic language. The arguments in the literature for the last thirty years have resulted in two paradoxical generalizations: 1. (i) Mandarin Chinese is a monosyllabic language (Ii) Mandarin Chinese is a poly-(di-)syllabic language. The monosyllabic approach (see, Karlgren 1929, 1949; Li, F. K, 1951; Chao & Yang, 1962; Chao Y.R, 1968; Chou, F.G. 1982, and others) insists that since almost every syllable of Chinese has a meaning, Chinese should be characterized as a monosyllabic language. Thus, Chao writes: "The so called'monosyllabic myth' is in fact one of the truest myths in Chinese mythology (1968:139)." The polysyllabic 唧roach (see Kennedy, 19 51; Defrancis, 1950; Jing, 1969; Li & Thompson, 1981 and many others) observes that more than 80% of words in a running text or an ordinary Chinese dictionary consist of two or more syllables卫 If the majority of words are polysyllabic, there is no reason to consider Chinese a monosyllabic language. As Kennedy has reasoned: we have no record of Chinese ever having even as many as four thousand distinct syllables, a number which if representing only words of one syllable would have been quite inadequate to represent the scores of thousands of expressions that the Chinese with their highly decide." (My translation) In this paper, I will propose a two-rule system of Chinese morphology in section 2, based on the theory of prosodic morphology developed by McCarthy and Prince (1993) and Feng (1994, 2000). I shall demonstrate that the morphological system of Chinese can be constructed by two independent rules: a [1]According to Chou's calculation (1982:10), there are 1,478 (13%) monosyllabic forms, 6,816 (60%) disyllabic forms, and 3,072 (27%) polysyllabic forms in Readings in Sayab/e Chinese (eds. by Chao, Y.R.1968). [2] See, for example, Duanmu 1993, Footnote 7. [3] See, for example, Tang 1989:3,4,568; Li, 1989:114) •
．．一，- . - - .' - - . - . —-泅,, 语 -t4)-i•J -~-~~~~ .~?.~-~~~~- ?~ -~i-~~ 凹~-t!?.~ · -· · ···.. . ..- --·-— · 一·一·一 一一－—----- --- -——一·一 ·-Morphosyllabic Rule (MR) and a Foot Formation Rule (FFR).141 Furthermore, I will argue, in section 3, that these two rules apply at two different levels of the grammar: MR applies at the morphemic level, and FFR at the phrasal (or syntactic) level. Given the two levels for the application of these two rules, it follows that the input of FFR must take the output of the MR. Under this system, the monosyllabic property is determined by the rule of morphosyllabicity (see, 2.1) and the characteristic of disyllabicity is derived from the rule of Foot Formation (see, 2.2). Section four investigates the interactio~ between MR and FFR, in which the [2-to-1] and [l-to-2] morphological process are discussed. Given the two-rule system hypothesis, a number of theoretical consequences will be summarized in sections five and SlX. 2. Two Rules in Chinese Morphology 2.1. Morphosyllabic Axiom It has been widely recognized that morphemes in Chinese are overwhelmingly monosyllabic. However, there are neither generalizations nor linguistic representations to consider the monosyllabicity as a morphological constraint of the language. Traditionally, linguists only gave general statements such as "speaking of morphemes, Chinese is basically monosyllabic language" (Chou, 1984), but never considered it as a rule in Chinese. Some crucial reasons, I think, are these: First, we lack a theory of how it is possible for a language to [4] Therefore in Chinese, syllabic writing is de facto morphemic writing, and thus to call it morphosyllabic is correct, but not fundamentally different from calling it logographic or morphemic. (William G. Boltz 1989. Reviews (11) Sino-Platonic Papers, 14) observe a monosyllabic rule. Second, there are disyllabic monomorphemes in everyday speech, even though they are very rare. For example (a"." before a syllable indicates it is neutralized): 2. hudie'butterfly' Yuan.yang'mandarin duck' tanglang'mantis' bo.li glass ' pu. tao grape , However, as we will see below, these disyllabic morphemes cannot disprove the general fact that morphemes in Chinese are overwhelmingly monosyllabic. For reasons that will be given below, I would like to suggest that the property of syllabic monomorphemic formation should be considered a general constraint in Chinese morphology. This constraint will be stated using Defrancis'term (1986: 187) as "morphosyllabicity" and formulated in (3): 3. Morphosyllabic Rule (MR) ("M" stands for morpheme and II o II for syllable): 『。In Chinese, a syllable must correspond to a morpheme. According to MR, the M node dominates a syllable directly hence a syllable is directly correspondent to a morpheme. The mapping between an 11M11 and a II a II will immediately result in a type of output of one syllable with one morpheme.
--.. .. . . . . . . . . . . . Number 33 I 2009年第 1期總第暈期·· ······ ·· ·· ·········.. ··········· ·· ··· ············· ······· ··· ·························· ·············· 一 ·- --r- -一·--- --·.一 一 －－ 一—一 一一—一—一—一－－一 －一一一一 一 一一·一｀——The MR can also be naturally derived from the theory of Prosodic Morphology developed by McCarthy and Prince (1993). More specifically, it is derived from the theory of alignment. Recent studies of the syntax/prosody interface have resulted in an edge based theory (Chen 1987; Selkirk 1986; among others), which proposes that the domains of sentence phonology are specified by rules like "The right/left edge of some grammatical constituent coincides with the corresponding edge of some phonological constituent" (cf. Selkirk 1986). This rule has been defined in terms of edge alignment (ALIGN), with the following general schema by McCarthy and Prince (1993): 4. General Schema for ALIGN In ALIGN (GCat, GEDGE, PCat, PEdge), the GEdge of any GCat must coincide with PEdge of some PCat, where, Gcat= Grammatical Category, among which are the morphological categories Mcat = Root, Stem, Morphological Word, Prefix, Suffix, etc. PCat = Prosodic Category=µ, a , Foot, PrWd, PhPhrase, etc. MEdge, PEdge = Left, Right McCarthy and Prince's edge alignment schema extends the Chen/Selkirk theory in two ways: the grammatical and prosodic categories subject to alignment are the word?-internal morphological constituents, root, suffix, etc., and the word internal prosodic constituents, syllable, foot, etc.; and alignment of different edges may also be required. It has been observed that many languages require free-standing (nonclitic) words to be of a minimal prosodic size (typically disyllabic or bimoraic). Subminimal items are either barred from the lexicon entirely or brought up to code through various augmentation processes. As pointed out by Kenstowicz (1993), in Australian language Yidiny, all roots conform to a CVCV(CV) template, and therefore are minimally disyllabic. If there is a general correlation between certain prosodic categories (mora, syllable) and certain morphological categories (root, morpheme, words) in human languages, then it is reasonable that the MR in Chinese might simply be the result of the general principle of Edge Alignment. The alignment principle is parameterized in terms of a coterminous between a prosodic category syllable and a morphological category morpheme. In other words, the MR is a subcase of a more general principle of Edge Alignment between prosody and morphology. Given this, the Morphosyllabic Rule can be interpreted in terms of Edge? Alignmentl5J between syllables and morphemes ('M' strands for root morphemes and'o'for syllables): 5. ALIGN: [M] = [a] This constraint relates the prosodic category syllable to the morphological category morpheme, demanding that they begin and end together. This constraint has some interesting linguistic consequences in Chinese grammar (Feng, 1994/2000). It is important to note here the implication that any operation which [5] In the on going discussions, we use the symbol ALIGN to indicate an alignment of both edges ALIGN L for left-edge alignment, and ALIGN R for right-edge alignment. Furthermore, ALIGN: [X]=[Y] represents a situation where X and Y not only begin and end together, but also are interchangeable: If it is X, it must be Y. However, ALIGN:[ ]X = [ ]Y stands for a situation where X and Y begin and end together, but Y may not necessarily be X (although X must be Y). •
- 這一一 －一＝－~沭,, 语 ,t4).;,J I ~-~~~?-~?.~-~~~~-?~ 旦~~~-i~_t_i~? ..... ... .. .. .... .. ....... .. .. ... .................. .. ...... ......... . －－ － －＿＿＿＿＿＿＿一一 一 一一一一 一 一 —一·—- -- -breaks the desired relation between the morphological and prosodic constituency of a form, will be a violation of the constraint, since ALIGN requires sharply defined morpheme edges. For ALIGN to be satisfied, the morpheme final consonant or vowel must occupy the final position in the corresponding syllable, and the morpheme initial C or V must occupy initial position in that syllable. Consequently, a "morpheme mid syllable/consonant" will de align a morpheme (see McCarthy & Prince 1993 :38). This explains why there is no re syllabification in Chinese, as shown in (6) (where "?" marks the morpheme boundary): 6. cv I eve - *(CVC) vq Another important implication of the constraint [ALIGN: M= o ] is that the minimal or primitive constituents for morphological operations in the language are monosyllabic morphemes or words. From a prosodic viewpoint, any combination of two or more morphemes will interfere with the organization of syllables . From a morphological point of view, any association of two or more syllables will affect the morphological structure. Because of this, the organization of the grammar beyond individual syllables (in the prosodic system) or individual morphemes (in morphological system) will involve principles or j constraints, both prosodic and morphological · Note further that the [ALIGN: M= o ] also entails that, the following structure given in (7) must be ill-formed: ,. A o o In (7), the M node dominates two syllables, hence disyllabic morphemes will be generated. Obviously, if MR is a general rule in Chinese morphology, structures like the one given in (7) must be ill-formed. In other words, as a result of MR, there would be no syllable that is not a morpheme and no morphemes that contain two or more syllables. In next section, I will provide evidence to support the poss加lity of taking MR to be a general rule (or constraint) in Chinese morphology. 2.1.1. Morphemization of polysyllabic words As shown in (2), seemingly counterexamples to the MR can easily be found in Chinese. However, as Sproat & Sh巾 (1993) points out, within some disyllabic morphemes such as hudie (butterfly), in which the two syllables represent only one morpheme, one of the two syllables can be used as an independent morpheme in combination with other morphemes . For example: 8, hudie'butterfly'die-yong ' butterfly stroke' mayi ·ants'xiong-yi male ants' luosi ' snail' chanchu'toad' luo-wen chan-su ' whorl' 、 toad cake' tanglang'manti s'tang如 dang-che'antis arm stop car'(an antis stops a car with its arm) Yuan .yang'mandarin duck'yuan-lU'affectionate company' Bo.Ii ·glass' bo-qian 、 class fiber' hu.li ' fox' hu-shou 'fox smell' zhi.zhu'spider' zhu-wang cobweb' pang .xie'crab' xie-qing greenish-grey (color)' luo .tuo'camel' tuo-mao ' camel hair' Note that if die-yang is a compound formed by two —_ _II_ `
一－－—·一·一 ——----——一· —-—-- --- - - - -—-- —— - -—· 一————Number 33 2009年第 1期總第遣期................. ..... .......... .. .... ..... ... ......... .......... .... .... .... .. ... ... ..... ... ... ..... ．－－一 一·一- - — 一 一 —· —· —· 一 一 4一 一 —- — ·一 一一一一 一· 一 一 ·一·一一morphemes, die must be considered as an independent morpheme. If die is a morpheme, then the first part of h11-d比， i.e. hu, must also be analyzed as a morpheme. That is, given the following 山agram,if B in [B C] of (9b) and in [B D] of (9c) is a morpheme of compound words [B C] and [B D], then "a" in [a B] must also be considered as a morpheme. 9. a. a B b C. CD BB This treatment is parallel to a morphological analysis of cranberry, huckleberry and boysenbeny If benJ'is a morpheme, the other part of the words, i.e. cran in cranberry, huckle in huckleberry, and boysen in boysenberry must also be analyzed as morphemes although no such forms cran, huckle and boysen exist in the English lexicon. That is, the morphological theory must allow some morphemes that are not meaningful in isolation. These type of morphemes acquire meaning by virtue of their connection with other morphemes to form words. Likewise, if die in die-yang is a morpheme, there is no reason not to consider hu in hu-die to be a morpheme, even though hu may not be independently listed in the lexicon analogous to cran, huckle and boysen. If hu is a morpheme according the above analysis, then the correlation of a syllable with a morpheme satisfies the MR given in (3). In other words, as long as hu and die are two morphemes, they meet the morphosyllabic constraint: a syllable is directly dominated by a morpheme. If this is the case, the so-called Butterfly-cases would not be'true' exceptions to the hypothesis that there is actually a rule in Chinese morphology that a syllable must coincide with a morpheme. Su (1989) and Sproat & Shih (1993) have provided rich documentation of a strong tendency in Mandarin compound formation to pick one morpheme in a polysyllabic form. This tendency can be seen as a process of morphemization of polysyllabic forms, i.e. to pick up one syllable from a polysyllabic word and make it a morpheme by combining it with other morphemes/words that already exist in the language. Note that the morphemization process happens not only in those where one part of the polysyllabic form was used in classical Chinese (Sproad & Shih, 1993: 194), but also in loan words borrowed from other languages. For example, 10. Fute
泅,, 语t4)-r•」 , ~-~~~? .~ 翌~~~~-可旦:~~~_i紅位-~ ·-···········•·"'''''''"''''''' • "········ · · · ····· ······.. ·· ·· ··· · － 一 一一一一一 一一 一一 一 一 一 －— ·— 一 —一·一 ·一一—--- - --------11 a . Binglang 'betel nut' b . ningmeng 'lemon' c. pangguang 'balder' d . hu.lu ' bottle gourd' Furthermore, there are syllabic epenthesis forms in the language: 12 a. xue.me 'look for' b. zha.me'blink' The two syllables in the above examples are inseparable, and none of them has been used to form a part of another compound . Therefore, these are exceptions to the analysis given above, meaning that the MR may not be the only rule in Mandarin Chinese. Actually, as we will see below, the MR given in (3) must be violated since the Foot Formation Rule (FFR) is also a constraint in the language. We will discuss the interaction between the MR and FFR below. For now, we may safely say that MR applies only to root-morphemes 161, hence it is not sensitive to non-root morphemes. Since the disyllabic forms in (llll and the second syllable in [6] Root-morphemes are morphemes that can be used to form a syntactic compound . Thus, xie (crab), in zi-xie (purple crab) is a root-morpheme, because it serves as a head of the syntactic compound, but zi in yi zi (chair) is not a root-morpheme, because it is a suffix (functional morpheme), and the compound is derivative, rather than syntactic Epenthesis syllables like the ones in (10) are naturally excluded from root-morphemes . [7] They can also be accounted for in terms of loan words that are exceptional to native rules. [8] The distinction between root morphemes and non-root morphemes (excluding functional epenthesis syllables) can also be seen from the fact that re-syllabification in Chinese does not exist between root morphemes demanded by the Edge-Alignment given in (3), but sometimes happens to functional morphemes such as a in the following Tian-a Heaven-particle Tian-na Goodness! (12Y81 are not root-morphemes, the MR would not see them within this system. 2.1.2. Anti-Disyllabicity Although disyllabicity has been believed to be a strong tendency in modern Chinese, anti-disyllabicity can also be observed in the language. That is, disyllabic forms of some sort tend to reduce to a monosyllabic form again. Chou (1982) pointed out that the reduction of the nominal suffix er must be considered as a process of anti-disyllabicity (1982 :9). The reduction of er is further elaborated in Xu (1990). He argued that Chinese exhibits a tendency for disyllabic forms to be reduced to monosyllabic forms, by observing that, hai-er (child-son) is a disyllabic word formed by hai (child) plus a monomorphemic nominal suffix -er which was developed from the middle Chinese word er meaning ' son, small.'However, the second syllable er in almost all nouns of Mandarin Chinese has been reduced to only a /r/ feature fused on the proceeding syllable yielding a monosyllabic word. This type of process can be foi·mulated as follows: 13. Monosyllabic Word ---Disyllabic Word ---+ Monosyllabic Word hai hai-er hair He then concludes that the operation of two syllable words becoming one syllable words, i.e . [2-to-l], is a very active morphological process in many Mandarin dialects. The operation here is thus generalized as follows: first, the last morpheme in a two syllable compound loses its meaning with its syllabic status; second, the phonological feature of the reduced syllable is joined to the first one according to general •
一. -,--- —------——一——·——一一一·— 一—一一＇一一一一—一—一 一一— . 期暈第總期1 第年9 。。2 3~ 3· r· e b. m~ u. N~ · . . . · . · · . . . · . . phonological constraint/s, yielding what Kratochvil (1968) called a fusion syllable. Given the observations provided by Chou (1982), Kratochvil (1968), Xu (1990), and Wang (1994), it seems that even if functional morphemes are beyond its scopes, MR still wants every root-morpheme to be formed by itself independently. Therefore, the suffix has a tendency to lose its syllabic status. If it is so, the er-reduction not only provides examples of anti-disyllabicity, but also argues for the dynamic power of MR. 2.2. Foot Formation Rule Following Feng (1994, 1998), I would like to suggest further that there is another rule in Mandarin morphology: 14. Foot Formation Rule ^ 。。A foot must is formed by at least two syllables As indicated in previous studies (see, among others, Chen 1979, Sh巾 1986, and especially Feng 2000:Ch2), Chinese employees disyllabic foot structure. For example, 15. A: Jintianji hao? Today what date? 'What date is today?' B a. *Wu. 'Five.' b. Wuhao f 1ve-number'Five.' [9] Neutralized suffixes are also motivated in this regard. c. Chu wu Begmning five'Five.' d. Shi wu ten five'Fifteen.' 16. a. Emei, Taihang, Hua-shan, Tai-shan, Jinggang-shan dou shi Zhongguo de ming shan. Emei, Taihang, Hua mountain, Tai Mountain , Jinggang mountain are all China's famous mountains. Emei, Taihang, Hua, Tai and Jinggang are all famous mountains. b. Emei, Taihang, *Hua, *Tai, Jinggang dou shi Zhongguo de ming shan. Emei, Taihang, Hua, Tai, Jinggang are all China's famous mountains. Emei, Taihang, Hua, Tai and Jinggang are all famous mountains. The examples in (15) show that a monosyllable word cannot be used independently within the context cited above. The examples in (16) show that monosyllabic words cannot form an independent foot co-occurring with other feet. These examples suggest that FFR must be considered a prosodic constraint in Chinese grammar. Of course, exceptions to the FFR may be found in cases like the following: 17. Fan, wo yidianr dou bu xiang chi. Food I little all not want eat. As for food, I don't want to eat it at all. A monomorphemic word is used as an independent foot as seen above. However, it must be supported by a pause after it (indicated by'#'), that is: 18. Fan# wo yidianr dou bu xiang chi Food# I little all not want eat. •
._--齿,, 语t4)i•J -~~?~? -~~ 卫~~~~-~~ -~i-~~~.i~_t!~.~ ...... .. . .. . . .. ...... . .... As for food, I don't want to eat it at all It is well-known (see Chao, 1968:67) that after the topic (and also the subject), there is an overt grammatical pause between the topic and the comment in Chinese, and this is especially true when a monosyllabic form acts independently as a topic or a subject. Hence monosyllabic forms are structurally limited only to appear in the topic and subject positions in Chinese. In this situation, if we take the pause to be some kind of prosodic epenthetic device for the single syllable, this type of foot would be structurally represented as follows (a syllable with a prosodic epenthetic pause): 19. / Fan [pause], wo yidianr dou bu xiang chi. Food, I don't want to eat 1t at all. If this is so, the generalization that a monosyllabic word cannot form an independent foot can also hold in situations where a monosyllabic foot occurs. 3. Levels for Operation of MR and FFR Given the two rules (MR and FFR) outlined above, we are facing a paradoxical situation in Chinese morphology. By MR, monosyllabic words are licensed and hence legitimate in the language. If this is so, why would disyllabic words develop at all? By FFR, any instances of phonological words must be disyllabic, hence all monosyllabic words must be ruled out as a violation of the prosodic minimality. If this is correct, why are there monosyllabic words co-occurring with disyllabic forms and why is there a tendency of anti-disyllabicity as seen above? Obviously these two rules are mutually exclusive. That is , if there is a rule that requires prosodic morphological units to be formed by only one syllable, a disyllabic rule could not be allowed in the prosodic morphology. On the other hand, if a disyllabic rule is operative in the prosodic morphology, the monosyllabic rule would have been eliminated in that system because these two rules are incompatible. As a result, the operation of one rule will be to the expense of the other. However, these two rules co-exist in Chinese and their output occurs side-by-side. As a result, a theory of Chinese morphology must be developed in such a way that these two rules operate freely in the morphological system. In this paper I shall propose that the MR and FFR are actually located at two different levels of grammar: the MR applies at a level where morphemes or monosyllabic words are constructed, while FFR applies at the post-morpheme level where morphemes (bound or free) are put together, as seen from the following analysis of(20a-b) ('M'strands for morphemes and'Wd' for words). 20. a. s B. XP/S ·一．XP XP XP/Wd XP/Wd 丨 I 丨 丨Wd Wd M M I I I I M M 。 。\ \ aa aa (20a) represents a structural analysis where syllables are organized into morphemes (M), |、
-—· 一· -一 · — - · 一·一-'· - - . Number 33 .. ..... ..... .. .. .... ......... .. ... .. ..... .... ........ .... .. ..... .... .. ..... ... ..... ...... .... ... .. ... ` 七－－－ －－－ － 一一 一－－— 一 — - -— -- —一·一. -morphemes are combined into words (Wd), words are Note that if the language has only a FFR without grouped into phrases (XP), and finally, phrases are the MR, then free-standing words (non-clitic) in Chinese structured into a sentence (or a larger phrase). This is a would all have been constructed as disyllabic forms. possible hierarchical structure English. Note that there is no constramt on one-to-one correlation between for languages 21. an individual syllable and like an individual morpheme in English. However, since MR is a rule in Chinese, a structure like (20a), which is possible in other languages, must be re-structured as (20b) in Chinese. That is, one syllable must correlate directly to a morphological unit (either a morpheme or a monosyllabic word). Since a morpheme (bound/free) plus a morpheme (bound/free) in Chinese must be formed according to the syntax, (20b) is a natural result of morpheme (or syllable) combination. Given this, it is clear that only after the structure organized by the insertion of lexical items produced by MR, can the application of FFR take place, as illustrated in (21). A. XP ^ XP/Wd XP\Wd ^^ M MM M MR¢I I I I 。。。。XP ^ FFR今 f f XP/Wd XP\Wd ^^ M MM M I I 丨 la a a a B. Since FFR is syllable-based and since MR demands each syllable to be a morpheme, a grouping of syllables will inevitably result m a grouping of morphemes (bound/free). Since the FFR cannot apply without a grouping of syllables, and since syllables (morphemes) cannot be grouped without syntax, the application of FFR on syllables will inevitably involve an operation of syntax. 2009年第 1期總第這期This is because monosyllabic words violate prosodic minimality. Subminimal items are either barred from the lexicon entirely or are brought into it through various augmentation processes. However, dynamic rule in Chinese, and most crucially, it functions different level from the FFR. Therefore, an application of one rule cannot stop the operation of another. That is, MR cannot go beyond the morphemic level to prevent the operation of FFR, and FFR cannot apply to levels lower than syllables to interfere with MR. at a MR 1S also a Since each of them has its own domain of application, MR freely Prosodic Morphemic produces phrasal level, while Words MR morphosy llabic FFR forms uno bstructedl y can be represented as in Table 1. Feng 2000)) below generates (PrWd, see morpheme level. As a result, Chinese morphology is determined by not only FFR but also MR. This scheme above Table 1. Two-Rule System of Chinese Morphology ··· ··--·- ···--· ···---- ··· ·------ ·-··· 一 . ...... . .. . .... . . . ... . . . ..... .. . . . .... . ..... . ..LEVELS RULES PROCESS LEXICON Phrasal FFR PrWd Disyllabic Forms ^ MM I I 0 0 ... .... ............................ 一.. . "'........ _ ............................ .. ......... . Monosyllabic Forms Given this analysis, we see that these two levels make it possible for the co existence of monosyllabicity with disyllabicity in Chinese morphology. 4. Interaction Between MR and FFR As we have seen before, either MR or FFR has a •
F ~--泅,, 语 t4).;,J -~-~~ 的-~翌~~竺？［巒g団研~-~ ........ ... ........... ... ..................... ... ........ .. ....... . .separate domain of their 唧lication, and because of this, it becomes possible for disyllabic forms to coexist with monosyllabic forms. However, although neither of these two rules would prevent the application of the other, each of them affects the output of the other. This is because the output of MR must appear in phrases where FFR dominates. As a result, the output of MR must be satisfied by FFR at the post-morpheme level. On the other hand, MR is also a rule that dominates at the morphemic level, hence disyllabic morphemes will violate the requirements of MR. The domination relationship is thus formulated as below (where ">>" means "dominates"): 22. Syntactic Level Morphemic Level FFR>>MR MR >> FFR This situation can best be interpreted in terms of the Optimality Theory (OT). That is , when FFR dominates MR, the output of MR must be considered as violation of FFR. When MR dominates FFR, the output of FFR may or may not be considered as a violation of MR, depending on whether the disyllabic compounds produced by FFR are represented by one or two lexical morphemes. If the disyllabic forms are represented by a combination of two lexical morphemes, the MR is satisfied, and if it is represented by only one lexical , morpheme, the result would be a violation of MR. Thus, the intervention of one rule on another will inevitably result in the following situation: monosyllabic words always violate the FFR at the post-morphemic level, thus they must be combined with other/s to 唧ear at the phrasal (syntactic compund) level. On the other hand, MR also intervenes to prevent epenthesis of extra syllables from occurring everywhere, i.e. in a two syllable unit with only one lexical meaning. Given this, it is not surprising that the dominating relationship of FFR >> MR 祠1 give rise to enormous disyllabic forms (or combinations) as exemplified in (23a c) and also causes some prosodically constrained syntactic consequences as shown in (23d g) 23. A. to look for 尋 *zim (Middle Chinese) xue.me (Mandarin) B. to blink 眨 *tFEp(Middle Chinese) zha .me (Mandarin) C. Mama, wo yao shui 惲ao)Mom, I want sleep Mom, I want to sleep. D. *zhongzhi shu plant trees E. *ba lian ca ba face wipe to wipe face F *" . .」 mxmggaicarry out change zhongzhi shu mu plant trees woods ba Iian ca-ca ba face wipe wipe to wipe face , , JlllXlllg ga1 ge carry out change change G. *Ma, woe le, wo yao chi le, wo yao chi fan Mom I hungry Asp. I want eat hungry Asp., I want eat-rice . 'Mom, I am hungry, I want to eat.' am hungry, I want to eat.' Ma, woe Mom, I 'Mom, I The examples given in (23a b) show that some monosyllabic words in middle Chinese have developed into disyllabic forms by adding an epenthetic syllable in Mandarin Chinese. Example (23c) shows that, under the pressure ofFFR, some intransitive verbs have developed as intransitive VO compounds. Examples (23d g) show that monosyllabic words (ob」 ects or verbs) must be ruled • -
一一-------"•一一 一·一 ·一· 一 —一·一 ·一一 .-一Number 33 2009年第 1期總第遶期.. , ........ ............. ........... ...... ..................................... ........... .... .... ... . out prosodically because of their incapab山ty of forming a foot and hence realizing the Nuclear Stress (NS) at the end of the sentence (see Feng 1994, 2000). Contrary to the requirement for disyllabicity demanded by FFR, the dominating relationship of MR >> FFR functions to motivate and thus protect monosyllabic forms as illustrated in (24). 24. [2 to 1] Process A. Hai+er hair child son children B. jin + ri jir This day today C. Shen-me sha What what Morpholization D. Hudie Butterfly E. fute Volt d頂 yong ' butterfly swim' butterfly stroke fu volt F. sai en si ke xue'branch (of academic) study' science science As seen before, the reduced syllables in (24a c) may be initiated under purely phonological conditions, but it is important to realize that the poss洳lity of all [2 to I] morphological processes is possible only if they are protected or licensed by the MR, because the results of 、uch a process would otherwise be impossible under the requirement of FFR. Given this, it would be better to I consider the Chinese morphology as a system governed by the cooperation and intervention of the two rules. l The themy p
-泅,, 语t~i'」 , ~-~ 笠?.~?.~.~~~~- ?~-~ 四嚥t_i?.~ . '··· ···· ···· ···· · · ··· ···· · ·· ·.. .... . .. . ···················......... 5. Redundancies in the Chinese Lexicon Given the two-rule system of Chinese morphology, a natural conclusion about the lexicon of Chinese would be that the morpMlogical operations of MR and FFR will result in a certain degree of redundancy of wordhood in the language. That is, one meaning may possibly be represented by two types of forms: a monosyllabic form produced by MR and a disyllabic form resulting from the FFR. This, as shown in (26), is exactly what happens in Modern Chinese. 26. A. xiang wo xiang ma.ma xiangnian Wo.men xiangnian zuguo Miss I miss (my) mother Miss We miss (our) country J .. laO jiao hua Jiaoguan jiaoguan huasao Water water flowers Water water plants fan wo ai chi fan Mifan wo chi mi-fan Rice I like eat rice 「ice I eat rice Suan wo qu mat suan Dasuan wo mai da-suan Garlic I go buy garlic Garlic I buy garlic B. na-li (that-place, there) zhe-li (this-place, here) jin-ri (this day, today) nar (there) zher (here) jir (today) ming-ri (next day, tomorrow) mir (tomorrow) Both types of words co-exist in the language. The monomorphemic words meet the MR and disyllabic forms satisfy the FFR. However, when monosyllabic morphemes/words are used alone in natural speech, they violate the requirement of the FFR, thus a monosyllabic form must co-occur or be grouped with others to satisfy the FFR at the phrasal level. When a combination happens to match perfectly (i.e., a minimal foot with a minimal phrase), the prosodic morphological system will immediately identify it as a PrW d according to the FFR given in (25). Furthermore, under the prosodic pressure and pragmatic usages, the two elements in a PrWd will be'frozen'or'idiomatized'and finally become what is called a compound word in Chinese (Feng 1997). 6. Summary In this paper, I have proposed a two-rule system for Chinese morphology. Although these two rules have been observed in previous works and stated in different terms, misinterpretation of these two rules has caused great confusion regarding the syllabic property of the Chinese language (Defrancis, 1986: 177 188). Given the analysis in this paper, it is clear that the confusion is mainly caused by the lack of a proper theory to capture the rule system, and to explain how it operates in such a way that these two rules cooperate with each other as well as intervene in the outputs produced by each other. The theory presented here leads naturally to some desirable theoretical as well as empirical consequences. J;lirst, a natural plausible conclusion for the mo1;10syllabic myth in Chinese is that there is only a system of rules which interact to derive the lexicon and whet~er Chinese is monosyllabic or polysyllabic is merely a result of rule-interactions. Second, the notion
'. .'. . '. Number 33 ; 2009年第 1期總第遑期.................... ............................... ..... .......... ................... ..... ... .. .. .... -- . . of PrW d as seen in Table 1 and (25) may be crucial in resolving the traditional problems of distinguishing phrases from words in Chinese morphology (Feng 2000). Given the notion of PrW d, any instance of combinations of two (monosyllabic) morphemes will fall into the category of PrW d, hence examples such as bai zhi ('white paper') and kan po ('see through') are prosodic words. In the present theory, PrW ds are outputs of the morphological system in the sense that a compound must first be a PrW d, even if a PrW d may not necessarily be a compound. Therefore, although kan po is a PrWd, it may be used in other environments as a phrase such as kan de po'be able to see through,'where a functional marker can be inserted in between the two morphemes. According to this analysis, it is natural to have the following categories: 27. Phrase: bai.de zhi'papers that are white' kan.de po'be able to see through' PrWd: bai zhi'white paper; white paper' kan po'see through; understand thoroughly' Compound: bai shu'white potato, sweet potato' *bai.de shu'sweet potato' gai shan'change better, improve' *gai.de shan'be able to change better' Given the prosodic category of PrW d in Chinese morphosyntax, the peculiar behavior of cases like bai zhi and kan po is accounted for systematically. Third, although the rule system of Chinese morphology produces a certain degree of redundancy in the lexicon as seen in (26), other systems of the language will be expected to operate on these forms so that they are not pure redundancies. The fact that monosyllabic forms and disyllabic forms have served different stylistic and prosodic purposes in the language, confirms the hypothesis made here. Li J. (1989), for example, observed that disyllabic forms tend to be used in formal occasions while monosyllabic forms are generally very casual. Furthermore, as seen in (23d g) above, strong prosodic positions tend to attract disyllabic forms and exclude monosyllabic forms, indicating that the monosyllabic and disyllabic forms have played different roles in Chinese prosodic system. Finally, if the two rule system of Chinese morphology is correct, further research may provide evidence from language acquisition as to whether or not these are different steps of acquiring these two rules. Actually, Tang's study (1989:43-92) has already suggested that children exhibit a tendency to first acquire MR with an unproductive FFR. If this is so, it is not surprising that monosyllabic forms exhibit a high frequency of occurrence, given that the stylistic usage of FFR (i.e. forming a foot by using classical words or morphemes) would be a more advanced adult grammar. Obviously, if the theory given here is correct, it may also shed some light on language acquisition regarding the diglossic grammar of children and adults. *This paper was originally composed in 1994 in order to initiate the field currently known as prosodic syntax. Today, this new field is flourishing in the Chinese-speaking world and the question of the 'monosyllabic myth'discussed here has become more important than ever. I have therefore revised the old paper with great gratitude to the editor, Professor .
` —.1_I 齿 f1 语t4}i•」~-~~~? -~~-~-~~-~~-~~-~ 嵒_i~_t!~-~ -- -- - · ·-Zhou Jian, whose enthusiasm is responsible for finally making it available to the English-speaking community. References Chao, Y.R. 1968. A Grammar of Spoken Chinese. University of California Press. Berkeley. California. Chao, Y.R. & L.S. Yang. 1962. Concise Dictiona,y 司SpokenChinese. Cambridge, Mass. Chen, Matthew Y. 1979. Metrical Structure: Evidence from Chinese Poetry. Linguistics Inquiry I 0,371 420. _ .1987. The Syntax of Xiamen Tone Sandhi. Phonology Yearbook, 4:109-150. Cheng, Xiangqing. 1982. Xianqin Shuangyinci Yanjiu [A Study of Disyllabic Words in Pre Qin] in: Cheng,Xiangqing (eds.) X山11qin Hanyu Yanjiu [Studies of Pre Qin Chinese]. Shandong Jiaoyu Chubanshe [Shandong Educational Press]. _. 1985 Lungheng Fuyinci Yanjiu [A Study of Polysyllabic Words on Lunheng]. in: Cheng,Xiangqing (eds.) Liang Han Hanyu Yanjiu [Studies 司 the Han Chinese]. 1985. Shandong Jiaoyu Chubanshe (Shandong Educational Press). Chou, Fa-kao. 1962. A 祏storical Grammar of Ancient Chinese. Part II: Morphology. The Institute of History and Philology, Academia Sinica. Special Publications No.39. Taipei. —1982. Zhongguoyu Danyinjiexing zhi Zai 」iantao [Re-examination of Monosyllabicity in Chinese]. Zhongguo Yuwen Ywv·iu [Chinese Study], No.5. 1982:5-12. DeFrancis. 1950. Nationalism and Language Reform in China Princeton. Reprinted New York, 1972. _.1986. The Chinese Language: Fact and Fantasy University of Hawaii Press. Duanmu, San. 1990. A Formal Study 司Syllable, Tone, Stress and Domain 切 Chinese Languages. Doctoral Dissertation. MIT. _. 1993 和me Length, Stress, and Association Domains. Journal of East Asian Linguistics. 2.1: I 44. Feng, Shengli. 1994. Prosodical/y Constrained Syntax in Chinese. Doctoral Dissertation. UPENN 1997. Prosodic Structure and Compound Words in Classical Chinese. In: Packard. J. (ed.) Word Formation in Chinese. Trends in Linguistics: 197-260. Mouton de Gruyter. .1998. Lun Hanyu de Ziran Yinbu [On Natural Foot in Modern Chinese]. Zhongguo Yi1wen [ChinesePhilology} I :40-47 2000. Hanyu Yim/U J1ifaxue [Prosodic Syntax in Chinese] Higher Educational Press. Beijing Guo, Shaoyu.1985. [1938] Zhongguo Yuci zhi Tanxmg - -Zuoyong [The Elastic Property of Chinese Word Length] in Zhaoyushi Yuyan Wenzi Lwlji [Collection of Linguistic and Plzilological Works}. Shanghai Guji Chubanshe [Shanghai Classics Press], and also Ya1lji11 Xuebao24, 1938. Hayes, Bruce. 1989. Compensatory Lengthening in Moraic Phonology. Linguistic Jnq11i1y 20:253 306. . I 981. A Metrical Tlzeo,y c:,f Stress Rules. Doctoral Dissertation. MIT. IULC. 1980. Jing, Yunjing. 1969. Disyllabicity of Modern Mandarin. Zhoimgguo Wenhua Jikan (Chinese Culture). Vol.10.No.4:88 104. Karlgren, Bernhard. 1940. Grammate Serica: script and phonetic in Chinese and Sino Japanese. Bulletin of the Museum of Far Eastern Antiquities 12: I 47 I. _. 1929. Sound and Symbol in Chinese. London. _.1949. The Chinese Language. New York. Kennedy. George. A. 1951. The Monosyllabic Myth. Journal of the American Oriental Society. Vol.71. No.3: 161 166. —.1955. The Butterfly Case . We11ti 8, March 1955 _ .1964. Selected Works of George A. Kennedy. Edited by Tien-yi Ii. New Haven: Far Eastern Publications. Kenstowicz, Michael. 1993. Evidence for metrical constituency. The view from building 20 --- Essays in linguistics in honor of Sylvain Bromberger, ed. By K. Hale and S. J. Keyser Cambridge, Mass.: MIT Press. Kratochvil, Paul. 1968. The Chinese La11guage Today. London. Li, C.N, & Thompson, S.A. 1981. Mandarin Chinese: a Functional Reference Gramma1: Berkeley, CA: University of California Press. Li, Fang Kuei. 1951. Zang-Han Xi Yuyan Yan」 iu Fa. Guoxue 」ikan [Journal of Sinology] 7: 165-175 .1973. Language and Dialects of China. Journal of Chinese Linguistics I: 1 13. 1980. Shang Gu Yin Yanjiu [Study of Old Chinese Phonology] Shangwu Chubanshe .Bei」 ing: Commercial Press. Li, Jiashu. 1989. Danyinjieci he Shuangyinjieci zai Kouyu he Shumianyu shang de B國oxian [Monosyllabic and Disyllabic Words in Everyday oral and Writing Languages]. in Li Jiashu Yuwen Yanjiu heYuwe11 Jiaoxue [Chinese Study and Chinese Teaching). Shangwu Yinshu Guan. Hong Kong. 1989: 1 15. 區erman. M & Sproat. 1992 . The Stress and Structure of Modified Noun Phrases in English. In: Ivan A, Sag and Anna Szabolcsi (eds.), Lexical Matters. Stanford University. Lu, Bingfu. & San, Duanmu. 1991. A Case study of the Rhythm Between Rhythm and Syntax in Chinese. Paper presented on the Third American Conference on Chinese Linguistics. Lu, Zhiwei. 1947. Gu yin Lue Shuo [Summary Discussion of •
·---一~一 一－ 一 — - . ---,, 一 — 一一 一· —· —· —. . 一· —- . — 一 · — Number 33 2009年第 1期總第遁期········ ··· ······ ···· ··· ··· ···· ·· ·· ······· ··· ·· ·· ······ ·· ···················· ············· ·· ···· ·· ·· ·· 一一 一 一一 一 —·一 一 一 一一 一 一一·— － 一 一·— 一一 · 一 — Old Chinese Pronunciation]. Yanjing Xuebao [Yanjing Joumal] Monograph no. 20. Taipei: Xuesheng Shujiu (Student Book co.) McCarthy, John. I 979. On Stress and Syllabification. Linguistics /11q11i1y. 10: 443 466. McCarthy, John & Alan, Prince. 1991. Prosodic Minimality ---Lecture presented at University of Illinois Conference the Organization of Phonology. . & .1993. Prosodic Morphology I. ms. Naspor M. and I. Vogel.. 1986. Prosodic Phonology. Foris Publications. 1986. Norman, Jerry. 1988. Chinese. Cambridge University Press. Prince, Alan. I 980. A Metrical Theory for Estonian Quantity. Linguistic Inquily. 11, 511 526. Selkirk, Elisabeth. 1980a. Prosodic Domains in Phonology: Sanskrit Revisited. In: Mark Aronoff and Mary Louise Kean (eds.) Juncture Anma Libri, Saratoga, CA. 107 129. . l 980b. The Role of Prosodic Categories in English Word Stress. Linguistic Inquily. 11, 563 605. .1981. Epenthesis and Degenerate Syllables in Cairene Arabic. in: Hagit Borer and Youssef Aoun (eds.) The Theoretical Issues in the Grammar of Semitic Languages. MIT Working Papers in Linguistics, vol.3 . 1986. On Derived Domains in Sentence Phonology. Phonology yearbook 3:371-405. Shih, Chi Jin. 1986. The Prosodic Domain of Tone Sandhi in Chinese. Doctoral Dissertation. University of California San Diego. Sproat, R. & C.L. Sh山 1993. Why Mandarin Morphology is Not Stratum Ordered. Yearbook of Morphology. Clair Academic Publishers. Su, Xiyu. 1989. Shishuo Xiandai Hanyu Fuyin Danchunc1 Zhongde Yin」 ie Fuha Xiang Yusu de Zhuanhua [Discussion on Morphemization of Syllabic Symbols in Modern Chinese Polysyllabic Forms] 和wen Xuexi [Philology Study], No.2 Tang, Tingchi. 1989. Hanyu Cifa J叩 [Studies in Chinese Mo1phology and Syntax]. (2 Volumes) Taipei: Student Book Company. Wang, Li. 1980. Hanyu Shigao [A historical Grammar of Chinese]. Zhonghua Shu」 u Press Wang, Hongjun. 1994. Two Phonetic Word Formations in Common Use in Chinese. 粒yan Yanjiu [Language Study] I :65 78. Xu, Tongqiang. 1990. 」ieguo de Bu Pinghengxing he Yuyan Yanbian de Yuanyin [Structural Asymmetricity and the Cause of Language Change]. Zhongguo 汕wen [Chinese Philology}. No. I. Zhang Hongming. 1992. Topics in Chinese Phrasal Phonology. Ph.D. Dissertation. University of California, San Diego. Zhu, Dexi. 1956. Xiandai Hanyu Xingrongci de Yanjiu [A study on Ad」 ectives in Modern Chinese]. Yuyan Yanjiu [Language Study] I :83 112. -•
--—-—一—一－一·— 一 一一一~一·————· 一一· — · — 一一一一 —· — · ——· — · —一· — ·-Number 33 _ 2009年第 1期總第禮期.......... . .. . . . . . ...... ........................... . ................ . . . ........ .. . . . .. . ... . ...... .... ~- －一－－－－－一一—·一 －－— 一一－一·——一——一 一 —·一 一 ·一一一一·一 ·— 一一一- -馬禮遜多方面的業績和貢獻中， 《華英字典》最受囑目，影響最深遠，貢獻最大。三部中，第一部稱為《字典》，是世界上最早編寫出版的漢英字典，它的漢字詞條參考《康熙字典》排序，馬檀遜給予英語釋義，這是馬禮遜學養和智慧創造胜的心血工程。第二部稱為《五車韻庫》，馬禮遜在前言中說明是以陳藎謨、胡含一師生兩代的心血結晶為藍本，詞條按漢字音序排列，馬禮遜再參考相當有關韻書和資料，加上英文釋義而成。 1註,,1 第三部稱為《 An English and Chinese Dictionary》，詞條以英文字母為序排列，除以漢語為英語單詞釋義外，還針對用者需要，加輯實用詞語，如同義詞、反義詞、近義詞、常用成語、方言俗諺，並從名著《論語》《孟子》《幾何原本》《草本綱目》《紅樓夢》《三國演義》等選譯章句，作句型範例。在馬禮遜多方面的業績和貢獻中， 《華英字典》最受矚目，影響最深遠，貢獻最大。一、世界上第＿部實用性顯著的英漢字典馬禮遜編的《 An English and Chinese Dictionar y》，不管從形式到內容，都具有鮮明的針對性，因而具有較高的實用價值。1、規模可觀，體例猖創
.... 泅,, 语 t4)-i•J~~~箜-~~y~~~七？［竺~-i述?.~ ... ... .. .. . .. . . .. . .. .. . .. . .. . .... . ...... ... .. .. . .................. ..一一一 . 一 — . . -- - - -- - - - ----- --,--.., 和假借六書；還舉了多個片語和例句，以示其｀｀品性＂、 ＂名聲＂的詞義。馬禮遜對六書的介紹，顯然不準確，但這一詞條足以顯示編者的意圖。如此編排是《字典》的獨特性。對外漢語教學的一大難點是漢字的聲調，馬禮遜也十分注意，在 "ACCENT條＂中介紹漢字｀｀四聲" , "平上去人＂。同時，馬禮遜已理解到漢語單字配詞的張力，充分運用熟字構新詞的途徑，擴大讀者的辭彙量，十分重視羅列實用片語。例如P.388的詞條 "SHED tears" 先釋以三個漢語同義詞： ＂流淚、下淚、灑淚"'再補充一個．，＇墮淚＂，對照英語 "let fall tears" ; 最後還利用熟字以擴充新詞"Shed tears of blood 泣血。儘管這解釋流於字面，但編者苦心可見一斑。®著意介紹中華文化四百多年前，羅馬天主教特別是耶穌會數以百計的神職人員，通過澳門已把過千華文典籍運回西方，其中部分巳譯成拉丁文或其他歐洲文字。西方對中國學術思想有一定的接觸而且比較景仰。馬禮遜在編撰過程中也十分重視把中華文化引進字典，引進英語世界， 《論語》、 《孟子》的章句不少被引用，例如P.353 "其為人也孝悌而好犯上者鮮矣。 P.384 "知恥近乎勇"。 "PROMULGE條" "promulge or diffuse virlue and cause families to be well regulated; states to be orderly governed, and the world to enjoy tranquility推之德而使家齊國（台平天下＂。 《字典》不但介紹了孔子、孟子、程子、朱子，連董仲舒也被稱為｀｀董子＂，並在P.283 "moral writing勸世文＂後引入他的名句： " opposed this principle and said正其誼不謀其利，明其道不計其功＂。對清王朝（合用的典章制度、連我們中國人編的字典也不一定列人的官場用語、文人謙辭，馬禮遜重視羅列，例如 "To purchase promotion捐了前程在身上"; "AGE條＂中羅列｀｀貴庚、高壽、弱冠＂等詞； "YOU條＂，除釋｀＇你、你們、爾、爾等、爾輩、汝＂外，還羅列｀＇尊駕、先生、仁兄、足下＂四詞。此外， 《四庫全書》《數理精蘊》《天下地輿圖》《本草綱目》｀＇意筆、工筆＂、 ＇壽山石＂等典籍和名詞也被選輯，說明他對中華文化的重視。 《字典》中還選列了不少民間諺語。例如： ｀｀養兒防老，積穀防饑＂ ＇＇人心不足，得隴望蜀＂ ｀｀刻薄成家，理無久享＂ ｀｀若要富，險中做＂ ｀｀寧為雞口，勿為牛後" "取快一時，沉淪－世＂ ｀｀人棄我取，人取我棄＂……都是作者體會到語文與文化的依存關係， 《字典》重視把語言與文化相互聯繫的例子。®地方色彩豐富鮮明。馬禮遜在中國生活25年，只局限於廣1州、澳門兩處沿海地區；工作限於東印度公司的翻譯員和本身傳教事工，其生活體驗頗受局限。他學習中國語言文字，閻接經驗比直接體驗多得多。反應在《字典》中知識面不平衡，這是它的弱勢，但對海事活動、商貿活動、與中國官府來往等，由於直接深人，了解全面，體會深刻，將這些寫人辭書，形成《字典》的另一優勢。在人、物、事、工中，凡馬禮遜所熟悉的，都特別豐富完備，為其他字典所不及。例如與海事有關的 "ship" , 他能列出 14個名詞和多個語句（共佔37行，）； "boat" , 佔23行中，首以｀｀船、艇、舟＂三個單音詞釋義，接著列舉了｀｀尖底船、平底船、西瓜扁船、大艇、花艇、巡船、三板、渡船、駁渡＂九種船名。 "ABROAD" 條， （佔32行）同義、近義詞語俱備，包括中國沿海特有的｀＇外洋、洋貨＂等詞； "BOATMAN" -詞在釋義中除｀＇艇人、船夫、船家＂三個名詞之外，還有一個同義詞是｀｀蛋家＂且加以註釋： "an egg house man" "So called probably from the 唧earance of the boat, which with its covers, to kelp off the sun or rain has an oval appearance like an egg." (見P.46) 這是廣東沿海地區對｀｀蛋家＂的傳說，如此解釋未必有事實根據，也與中國權威字典的解釋有所不同，但可作為當年民間一種說法，算-
~ - . . . '- .'. . . 日其慟第總期1 第年9 。。2 . . 3 3.r. e. b· m~ u. N" . . . · . . . . . . . . . · . . · · . . . . . . 是馬禮遜的個人學習心得。馬禮遜在中國的日子，25年工作都沒大的變動，為甚麼與犯罪訴訟有關的內容《字典》也特別詳實？原來與當年具體的歷史條件有關。 當年凡涉及英國人犯法事宜，兩廣總督都要東印度公司承擔處理，案件往往牽涉人命、處理不善還會遭到官方停止貿易、以及對英商不供應食水食物的制裁。 馬禮遜作為東印度公司翻譯員，對這方面的工作如寫狀子，譯口供都必須格外謹慎。因此對有關用語，他必須熟悉， "LAW條＂共佔72行，收｀｀法律、法例、律文、條例……大清律例＂ 。 還有許多實用語句 。 "MAGISTRATE條＂中，有｀｀官府、憲＂兩釋義外，又列了｀｀稟官府、稟大憲、打官司＂三個片語。 "CUSTOM條＂收錄了｀＇稅＂ ｀｀餉稅" "稅館＂｀｀關口＂ ｀｀關部＂ ｀｀關鈔的官＂ ＂關口的爺們＂ ｀｀徵收稅的官＂系列有關詞語。此外，如"ACCUSATION條" "ACQUIT條＂等不少詞條都有這類內容。®學術態度理性開明著重有關宗教用語，對其他宗教敏感也是《字典》的鮮明特色。最詳盡的詞條 "HEAVEN" , 共佔444行，自然很多內容涉及宗教 。 其次是 130行的"GOD", 單詞中已列出： ｀＇神鬼、神祇、上帝、河神、山神、天、仙、神仙、菩薩、上帝＂，還有許多語句 。 （見P.189-190) "CREED條＂佔54行，全部內容是基督教的六大信條。 "ATONE,to,for crimes贖罪條＂，列出的例句也著意宣傳教義： "Jesus has atoned for our sins, 耶穌代贖我們的罪＂ （見P.32) 。"MORAL writin矗"就是介紹基督教的勸世文的（見P.283) 。又如 "RELIGION條＂共佔54行，較全面列出各大宗教： ｀｀三教、儒教、釋教、道教、天主教、西洋教、＂，其中還有一首介紹佛教的律詩，又引一句對所有宗教作出肯定的句子： "A spirit of filial and fraternal duty, of faithfulness and truth, is what all religious delight in, 孝悌忠信之風而各教之所樂＂。可見馬禮遜對自己的信仰雖十分虔誠，但對於其他宗教也能尊重，態度理性開明，值得肯定。較多："MAHOMOD穆罕默德＂ （見 P.264) ; "MUSSLMAN回回人（見P.288) 更難得，他以讚美異教先知的話作例句： "Pure and true religion of the most noble and holy man Mahomed清真教穆罕默德至貴聖人＂。 （見P.345 "PURE條＂）《字典》對不同宗教的人、物、事、工往往還用模擬方法，例如"ABBESS條＂解釋為｀｀女修道長＂，又解釋為｀｀師太（見 P.2) 。有關各大宗教的內容除了宗教內容外，引伸佛家辭語也很多，如＂六根＂ ｀｀六塵＂ ＂圓覺＂和＂圓悟＂，可惜都只看表字面意義，未能觸到本質。 也有可能這些佛家語在編者心目中，已經融人中華文化中，屬他所關註的人文哲理，精神靈性的辭彙範疇而已。基督教是｀｀神本＂的思想體系，作為傳道人，馬禮遜對人道的關懷還是值得肯定的。 《字典》 "MAN條＂共佔80行，即可見一斑。®《字典》有意無意間成了清末粵方言載體對《字典》中的標音，本人未作具體研究，學者譚樹林認為｀｀它更接近廣東方言"。 1計SJ 明、清時期，外國人聘請中國教師學習中菌語文是犯法的，馬禮遜沒有多少選擇的餘地，所以他的中文教師有不同的口音，有不同的教育背景，他日常接觸的中國人幾乎都是操粵語的。在這種清況下，馬禮遜的確處於口語和書面語｀｀若即若離＂的狀態。 《字典》，粵口語非常多 。 有的他註明是粵語，例如： "MOOR條＂"Moor and Perses in Canton are called 白頭人＂當年粵語區把摩爾人稱為白頭人，因為他們用白頭巾裹頭 。 ＂事仔＂、 ＂事頭＂ 註明是粵方言。馬禮遜專門介紹的粗語穢言也註明是廣）寸1慣用的。 （見P.298) 但是，更多的粵語口語是沒有註明是粵方言的，與書面語混在一起：如 "Adam's apple喉欖（喉結）＂、"Ankle腳眼（腳踝）＂、 "Begin to do起頭做（開始做）＂、 "BEGUN曾經起手（已開始） " "A little _, - •
,.. ——· 一·一·一—— 一一·一 ·—— 一一·一一一 ——·— 一 一 — 一· -一 ·— 一 一—·— -- - - -— · 泅月语t#!)i'」 MACAO JOURNAL of Linguistics • .... .. ... .. .. .. . ..... .. . .. ... ... .... .......... ... .. , ... ... .. ................. .. .. .... ... ... .. .... .. ...... ... .. . — －－－一—-,- --.. 1nore要些添（多要一些）＂、 "Still more越發多（更 簡單音譯為 "Urh-yay" 、 "Laou tae-tae" 。類似的多）＂、 "Mad發癲（瘋狂）＂、 "Catamite契弟＂ 錯譯實在不少。歐訓以小時作時間計算單位，一天（鑾童）、 "continue to wait 等下（繼續等候）＂、"To befriend in business particularly幫趁（向商販購買）＂、 "fare price of passage in a vehicle by water水腳、盤費' (交通費）……賓在太多，不能盡舉。這本外國人在19世纪初編的《字典》哈我們纪錄了不少清末南粵地方語言。3 、 《字典》有三方面較顯著的錯誤從以上五大特色中，我們既看到《字典》的許多優點，也發現一些明顯的不足。叩同音混淆漢字中同音字太多，容易混淆，外國人學漢字更難－一分兩。例如 "DOUBT條＂中馬禮遜為讀者列出＇＇滿腹狐疑＂、 ＇｀狐疑不信＂之後再列一個四字成語： ｀｀狐思亂想＂。這是不對的，應為｀＇胡思亂想＂（胡亂、思想兩詞構成並列式四字成語）；又如"CONCUBINE條＂中的｀｀妾氏＂應為｀＇妾侍＂；"SERVANT條＂中的｀｀事仔＂，應為｀｀侍仔＂ （廣東口語，侍候、服侍別人的人，書面語稱｀｀侍應生＂）都是同音字混淆所影響。 1,r,1＠釋義不確由於民族文化差異，釋義以偏概全，難以準確。這類錯誤任何外國人都難以避免，但作為字典卻不容有失的。如 "COUSIN條" " son of an uncle堂兄弟"、 " daughter of an uncle堂姐妹＂。英語單詞"UNCLE" 實在包括了漢語的｀｀伯父"、 "叔父＂、｀｀舅父"、 "姑丈＂、 ＂姨丈＂五義。因此 "son of an uncle" 除譯成堂兄弟外，還有可能是表兄弟。 《字典》這的釋義就過於簡單了。因文化差異，漢語中的人倫閼係及其稱謂自然是外國人學中文的一大難點，馬禮遜也不例外。例如在 "CONCUBINE條＂中，他選譯《紅樓夢》的文句，只把｀｀二爺＂、 ＇｀老太太＂24個小時；中國則把一天分為12個時辰，每個時辰是2小時。儘管明末開始，歐洲生產的時鐘（我們也稱之為時辰鐘）已輸人不少，特別成為清皇室的寵物。馬禮遜對中國的計時單位也認識不足。在正文第一頁第一個詞條 "A" , "An hour" , 譯為｀｀一個時辰"'根本錯了；接著是 "A handful" 譯作｀＇－據＂也不準確，大多字典譯為｀｀一把＂或｀｀一撮＂比較準確。作為作者對首頁必認真審視，為甚麼第一頁兩詞都錯？原因最可能是因為理解未充分。®解說牽強。中國地大物博，歷史悠久，傳統文化博大精深，語言文字在幾千年的使用中凝聚了前人無窮的智慧，留下許多內（面深厚的典籍，積累了生動豐富的語言，名言、成語、典故、戲言、反話、隱語、歇後語……的確是窮一生都無法全面涉獯的。因此， 《字典》中許多釋義只按字面牽強解讀，有的還張冠李戴。譚樹林先生認為不準確甚至錯誤的一個例子：他把諺語｀＇跟起三姑學跳神，跟起好人學好人。＂英譯成："If you follow the three nuns, you will learn to dance to the gods; If you follow good men, you will learn to be a good nian." 漢語中｀｀三姑六婆＂有特定的語義。 ＂三姑＂是指尼姑、道姑、卦姑；六婆是指牙婆、媒婆、師婆、虔姿、藥婆、穩婆。譯成 "three nuns" 是明顯的錯誤。口101 《論語》中孔子說的｀｀民可使由之不可使知之＂， 《字典》解釋為 "The bulk of mankind may be make to act from (moral principle) but not to understand them. (見P.136) 孔子這句話，兩千多年來，中國專家也在爭論，單在斷句上，已有三種有理有據的不同分法。 l;lllll 要準確譯出這樣的傳統典籍，我們中國人也難做得到。馬禮遜雖有心，力實不逮。
一＿－－ ．..... ..... .... .............. ... .. ... ··········· ····· ········· ··· ·· ··· ········· ······ ········· ····· Number 33 2009年第 1期總第遑期二、瑕不掩瑜， 《字典》能享盛譽，實至名歸從具體的歷史條件分析，儘管《字典》存在的訛誤不少，但我們不能苛求於200年前的外國人，何況如此巨大的勞心工程的開拓？當年，馬禮遜是在秘密中孤軍作戰把《字典》完成的！由於明、清政府均不准華人教洋人中國文字， 不准洋人購買、印刷、運出中國書籍，馬禮遜的印刷工作間曾被官方沒收刻版，抓捕工人；他們收藏刻版於角落處，又被白蟻吃光，挫折頻生，一度心灰意冷也是人之常'I青。 1,ll: 121 能堅持13年在這種條件下完成任務，是十分難得的。何況《字典》本身也充分顯示出馬禮遜在靈修和學術兩方面的非凡成就。這些成就表現於下列幾個方面。L 內容雖不均衡，但不失為百科全害式的有分量的字典悰觀《 字典 》 479頁正文中，它側重語言和宗教，但舉凡政冶、經濟、社會、文化、自然各個領域都能收集基本的、常用的詞語。在17世纪以來逐步傳入中國的自然科學中，許多學科的基礎知識也見收集，特別是幾何學 、 天文學、 地理學、生物學、 醫學。 它確實為使用英語的讀者學習基礎漢語提供第一本實用的工具書。 一個開關而再閉關，只留澳門廣」什兩條通商小門的形勢， 多少西方學者遙對這文化大國，只能望洋興嘆。有歐洲植物學家希望馬禮遜能在物種王國中，代為收集堅果、家樹、野樹和灌木的品種。目 LJI 馬禮遜熱心助人也得到兩位檀物學家的幫助，因而在《字典》 P.174編出南中國每月開花的植物倩單，真是一個創舉！其中還為讀者編出中國樂譜與西洋樂譜對照圖表；各種數目字的寫法，十進的名稱；度量衡表；甚至影響中國民風深遠的《關聖帝君覺世真經》 全文及譯文。凡此種種都增加了 《字典》百科全書的性能。2、馬禮遜的中文造詣高，是 《字典 》 質素的主要保證馬禮遜刻苦、博學、認真，思维細密而活躍。 請看P.2 "ABDICATE條＂馬禮遜在釋義之後，引伸了｀｀To resign an office一語，以｀＇休致"、 "致仕＂兩個較文雅的同義詞以後， 還列出以下的詞作比較： ｀｀告休＂、 ｀｀乞休致"、 "告病＂、 ＂告老"、 "告終養'、 "讓位＂；又P.300 "OFFICE條＂，還補充了" to be compelled to resign勒休＂。又例如： P.40 "BEHEAD條＂他羅列了四個同義詞： ｀｀斬頭＂ ｀＇殺頭＂ ｀｀處決＂和｀＇梟首＂。目前我們常用的正文1500頁至4000頁的英漢字典，都只列前面兩詞以糰義。 《字典》中列了一些較生僻的詞，也能較準確釋義。如｀｀中焦，指腹腔上部的醫學用語"' " 系戀，牽掛依戀，與愛戀同義＂，以及上述的｀｀弱冠＂ ｀｀尊駕＂ ｀｀神祇' "休致＂ ｀｀致仕＂等等 。 理解和運用漢語虛詞是對外漢語教學的一大難點。從P.253 "Let or permit條＂更能看到嗚禮遜的功力。首先，他列出五個釋義的單音詞： ＂任＂ ｀｀許 ＂ ＂憑＂ ＂准＂｀｀由＂ 。其次，他在下列的例句中表示還可以用另外8個詞語來表達 ： 1. "容我做個東道主人Let me act as a master" ; 2. "讓我打扮你Let me dress you" ; 3. "給你聽 Let you hear" ; 4. "令其透氣let the air get ou t" ; 5. "俾各知所遵守Let every one know··· …" ; 6. " 務使點墨不留 Let not a spot of ink remain" ; 7 ｀｀聽其刷印let them paint off……" ; 8. "著照所請let it be done as is requested" 。第三，甚至不用任何與let相應的詞， 可直接表達： ｀｀放兵進來let the rebel troops mto the city" ; " 賞鑒let n1e see it" ; "落下篷來 letdown the sail" ; "放池中let it go into the pond" 。 從馬禮遜掌握漢語詞的豐富，可見其漢學造詣深厚。他希望《字典》 能給學華文漢語的歐洲人更多方便，他---•
. . . . .'. 丶齿,, 语t0i'] ~~~~? -~~-~ 的凡翌旦呣嚥卫?.~ ....... . ..... . .. ... ..... ... .... .. .... . . ... . . .... .. ...... .. ..... .... .，一 · — 一一一一·一—一一一一一一 一 一· — · — 一一一·— 一一一一·一 ·一一一一·一 一 一一·一 · -;- 丶 －是有心又有力的。3、按漢語構詞規律，編輯具科學思維在漢語豐富的詞語中，馬禮遜不但科學地掌握其構詞規律，而且還有意識地讓讀者能舉一反三，從規律中提高學習效益，編輯思维的科學性，對讀者的幫助更為重要。 P.282 "MONTH條＂他不但準確地介系召： "In each of the four seasons, the lst,2nd,and 3rd month are called孟仲季" ; " denote a decade or ten days" 是｀｀旬＂和｀｀澣＂ （作者註：該字通浣）；同時他又指出： " the 1st ,2nd,and 3rd decade are distinguished by 上中下 preceding either of these words" 換句話說，馬禮遜通過構詞法教了｀｀孟春＂｀｀仲夏＂ ｀｀孟秋＂ ｀｀季冬＂ ｀｀ 上旬 ＂ ＂中旬＂ ｀｀中澣＂ ＂下澣＂等時間詞，編者思維的科學性，有助讀者發展科學性思維。4 、 《字典 〉 兼顧文言和白話，適應特定年代的要求從《字典》中，我們還可以看到，例句不但引用文章、小說中的白話文，他自己也舉了不少白話文例句。馬禮遜已意識到當年正處於漢語文體正從文言走向白話的過渡階段，因而《字典》必須兼顧兩者，才能增強其實用性。綜觀全部正文，它的確已做到兩者兼顧。這是過去以耶穌會傳教士為主的西方漢學家所編寫的字典所未有的。例如在P. 1可見＇＇的、之＂同列： ｀｀縱欲的、放欲縱橫之人" ; P.2 "a character that 1s abbreviated 省寫的字，減筆之字＂ 。全文用文言表達者稍多：如P.14 "舟宜以載物"、 P.40"我既不以正道待人安望彼以良心報我"、 P.300 "酒席乃獻神者"、 P.394 "豈能描寫其萬一哉！＂……。白話表達的也不少：如P.32 "有船到了澳門" , P.282 "這是甚麼日子? " , P.300 "好妺妺你錯怪了" , P .3 16 "我看不見那件東西＂…… 。語±n 幺m丶三20世紀80年代末、 90年代初，我搜集澳門雙語精英資料，翻過馬禮遜的漢英—英漢字典。 2007年底決定聚焦於巨著的第三部： 《英漢字典》，從中體會馬禮遜的貢獻。前後十七年，不少資料已經淡忘，但對這位英嗣漢學家的評價國際國內都越來越鮮明，普遍得到讚揚和肯定。地球越來越小，人與人的距離在縮減；但與此同時，人與人的文化差異更霏要相互了解，矛盾誤會更需要諒解，衝突紛爭更需要和解。但是，缺乏溝通，即使兩敗俱傷的歷史教訓，也不會導向和平理性。 只有在溝通、理解的基礎上，才能摒棄爭奪 ， 互相尊重，導向了解—一諒解——和解的坦途，追求和平與雙贏的願景。在本文的寫作過程中，我的腦海經常浮現澳門歷史上這一群雙語精英的事蹟 ： 利瑪竇、 江沙維、 瑪姬士、 伯多祿和高美土。利瑪竇和馬禮遜同樣是向東方開拓傳教事業的闖將，同樣作出創業的犧牲；江沙维比馬禮遜長一歲，來澳遲了五年，同樣是長期刻苦鑽研中國語文，馬禮遜牧師則孤軍作戰，歷盡風險，翻譯聖經 ， 1815年至1822編出漢英一—英漢字典；江沙维神父卻率領82人的編撰大軍 ，從1831年開始持續11年，編完中文葡文字典、葡文中文字典，再編中文拉丁文字典、拉丁文中文字典。江沙維著的《漢字文法》，也是一部百科全書式的漢學專著， 享黌歐洲。 人們在他的墓誌銘上讚頌他是 ： ｀＇為葡國爭光、為澳門爭光的人＂。馬禮遜馳名西方，也憑百科全書式的中文英文字典，墓誌銘上也被刻上 ｀｀萬世不朽之人也" 。 1苴 14[ 兩人際遇不盡相同，但為人類提供溝通工具，促進東西文化交流、樹立永恆的功勳極為相似。這一群雙語精英都作出了超越時空的貢獻，超越初衷的成功，給澳門留下珍貴的歷史遺産。- I
