First draft. Comments welcome.

The Semantics of Abstract Nouns in English:Is Anything Lost in the Translation Into Chinese?

David MoserBeijing Foreign Studies University

International Symposium on Translation国际翻译学术研讨会

Nov. 1-3, 1997

1. IntroductionThe Chinese language has often been characterized by Western scholars (and by a good

many Chinese scholars, as well) as a language which does not accommodate the expression of abstractions as readily as do the Indo-European languages of the west. Though a good deal of early speculative nonsense in this area has been dispelled by a more careful scholarship, the issue remains a controversial and open one. The question first arose when sinologists began to probe the reasons for the relative indifference of Chinese philosophy to the abstractions that were central to the Greek philosophical tradition.1 One of the most common speculations has been that the perceived Chinese indifference to abstractions might be because the Chinese language itself was an unsuitable vehicle for abstract ideas. Such comments as these by famous sinologist Arthur Wright (1953) are typical:

[T]he Chinese [language] was relatively poor in resources for expressing abstractions and general classes or qualities. Such a notion as ‘Truth’ tended to devolve into ‘something that is true’. ‘Man’ tended to be understood as ‘the people’ — general but not abstract. ‘Hope’ was difficult to abstract from a series of expectations directed toward specific objects. (p. 287.)

Wright is speaking of the classical language here, but the characterization of Chinese as a canonical “non-abstract language”2 is usually applied to the modern language, as well. Such opinions are often based on the perception of some translators (particularly the translators of philosophical texts) that abstract terms and formulations in Western languages often seem to have no ready equivalent in Chinese, or are lost entirely, or must be left to context. The issue is complex, and while I am not in agreement with the more simplistic claims about the supposed lack of abstraction in Chinese, it is worth noting that there are significant differences in the way Chinese handles linguistic marking of abstractions, and these have interesting implications for translation between Western languages and Chinese.

1 For background information on this topic, see Graham (1978, 1989), Hansen (1983), Wright (1953), and others. The controversial work of Alfred Bloom (1981) is also significant in this respect, and although I do not concur with many of his conclusions, some of his work serves as a jumping-off point for this research.

2 The notion that some languages are “more abstract” than others is certainly not a new one, nor did originate in response to comparisons of Chinese with Western languages. French is sometimes considered more conducive to abstract word formation than German (see for example, Ulman 1952); the Latinate vocabulary of English is sometimes characterized as more abstract than the Anglo-Saxon. Ellis (1993) also points out a correspondence between aspects of German and the greater reliance on abstraction in German philosophy. (p. 89-91)

1

Through a general linguistic survey, and by using evidence from some psycholinguistic experiments of my own design, I will attempt to show that: (a) A lack of overt markers for abstract nouns and abstract propositions does not entail any relative lack of abstract thought in Chinese speakers, nor does it imply any difficulty in the expression and handling of abstract propositions; the semantics of abstract nouns and abstract formulations is often evoked contextually and syntactically in Chinese; and (b) the relative lack of explicit markers for abstract entities in Chinese does not necessarily entail difficulties in translation, when one takes into account the fact that Chinese tends to accomplish the expression of abstract and concrete through covert3 rather than overt means.

There is a kind of standard assumption that the ideal translation is one which causes “the same thing” to occur in the head of the speaker of the target language. It would seem like a good idea to first determine whether or not “the same thing” occurs in the daily psycholinguistic world of the target language speakers, whether the cognitive process involved is handled in the same way, or appears with the same frequency. It is this type of question that will direct the inquiry of this paper.

2. The translation of abstract nouns from English into Chinese.The semantics of abstract nouns has occupied Western linguists for centuries. There is a

general consensus that the philosophical motivation for much of Plato’s theories was based on abstract nouns in the language, and there have been many attempts to explain the referent of various kinds of abstract nouns. A full treatment of this complex issue is beyond the scope of this paper, but I will bring up a few specific issues related to translation between Chinese and English. Before we deal with this question, we must first give a brief overview of the differences in the way Chinese and English mark abstract nouns.

Prior to the modern period, the only device available to signal an abstract noun in Chinese was syntactic. A word could change word classes rather flexibly according to context alone, requiring no morphological change whatsoever in the word itself.4 The Neo-Mohist Canon (the 《墨辩》) there is the following passage:

小圆之圆与大圆之圆同.“The circularity of a small circle and the circularity of a large circle are the same.”

From the English standpoint, the second yuan 圆 of 小圆之圆 must denote a quality or feature of the circle—something like “circularity”—and not the circle itself. We rightly assume that this interpretation must be essentially equivalent to what the Mohist had in mind as well, or the sentence is meaningless and unmotivated. But perhaps it goes without saying that this interpretation is not a property of the second instance of the character yuan 圆 . The shift of 3 Whorf’s (1956) notion of covert categories and overt categories (also called cryptotypes and phenotypes). Overt categories are meanings encoded in the language by explicit markers. Examples in English are pluralization (marked with either -s, -es, or a vowel change), and past tense, which is marked in various ways according to verb type. Covert categories, by contrast, are more elusive meanings not coded in any particular word, but rather true of a covert class of words, whose common semantic features become evident only in certain linguistic contexts.4 The tendency of the classical language in particular is an extreme economy of means, requiring more interpretive work on the part

of the reader. The great linguist Wang Li (1947) characterized Western languages as fa zhi 法制 “rule-governed” languages, and

Chinese as ren zhi 人制 “human-governed”, saying “The ancient [Chinese writers] believed in not letting the text interfere with the

interpretation of the meaning; the style of Western writers is to not give the reader the opportunity for interpretation at all.” “古人主张不以辞害意,西人的行文却是希望不给读者以辞意的机会.”(p. 283.)

2

meaning is a result of contextual and syntactic factors only—the word itself has no explanatory function at all in Chinese (as it would here in English or Ancient Greek). For convenience sake we say “The first instance of the character means ‘circle’, the second instance means ‘circularity’.” From the standpoint of the ancient Chinese, however, there would be less of a compelling reason to analyze these as different words. That they refer to a different type of conceptual entity does not necessarily entail a conscious awareness of any word change.

This syntactic process by which words undergo class change in Chinese is analogous to what is called conversion in English, or zero derivation (see, for example, Quirk et al 1972). Conversion is a derivational transformation by which a word changes grammatical class without changing form:

“I commute every day. It’s a long commute.”

In the first case the word is a verb, in the second a noun. Other examples are “Use a hammer [noun] to hammer [verb] a nail.” “I’ll fax you. When you receive the fax...” and so on.5 Both suffixation and conversion processes exist in English to carry out class changes. We may say “They are certain to acquit him” or “His acquittal is certain”, and this change involves a the addition of a suffix, marking the class change explicitly. But in the sentences “They are certain to release him,” vs. “His release is certain”, the change is accomplished through conversion.6 The differences in the process look like this:

verb deverbal nounsuffixation: acquit —> acquittalconversion: release —> release

There are a number of ways in which transformations from adjectives to abstract nouns are marked in English, depending on such things as whether the adjective is derived from the noun or vice versa, and a host of other factors. There is a large class of adjectives that are transformed into abstract nouns by the adding of a distinctive suffix (e.g. kindness), as well as a smaller class of adjectives whose nominal form is the same (e.g. evil), and a comparison of these two will be more useful in providing a case analogous to that of Chinese:

adjective abstract nounconversion: He is evil. —> His evil knows no bounds.suffixation: He is kind. —> His kindness knows no bounds.

With the word evil here, note that our folk linguistic theory does not tell us that evil is a different word in the two sentences. Even though the class change is (unconsciously) perceived, the word is

5 This is a quite productive aspect of English, and is not merely a matter of certain classes of words or words with dual class membership. Almost any noun, verb, or adjective, given the right semantic context, can undergo on-the-spot conversion. An advertising blurb for a science fiction movie proclaims: “It out-E.T.’s E.T.!” A secretary says to a government welfare official, referring to a group of clients in a waiting room, “You’ve got an angry, a polite, and a scared out there. Which do you want to see first?” Examples are endless. But note that when the word changes class, it becomes subject to the morphological features of the new class, i.e. it can take plurals if changed to a noun, be put into past tense if made into a verb, etc.6 This example due to Quirk et al (1972), p. 1558.

3

not perceived as being polysemous, as would words like bank (as in “river bank”, “to rob a bank” etc.) and ball (as in “a child’s ball”, “a masquerade ball” etc.)

Of course, modern Chinese now has both overt and covert classes of abstract nouns. This is because a number of characters have come to be employed as suffixes in the twentieth century in response to the explosion of foreign scientific and philosophical works translated from Western languages (many of these being borrowed from Japanese). Such suffixes include -xing 性 as in

kenengxing 可能性, “possibility”, -hua 化 as in xiandaihua 现代化, “modernization”, -du 度 as in

sudu 速 度 , “speed”, and so on. Though these suffixes are far from fully productive, they are applied to a significant number of lexical items, and have become a standard part of the language, especially common in technical and scientific literature.

Interestingly, as these suffixes become more productive, one can begin to see all sorts of new nominalized forms in Chinese that are not the result of translation, and in fact have no idiomatic English translation at all. For example, the suffix hua 化 (roughly equivalent to -tion) is

now used quite freely with a wide variety of words: laonianhua 老年化 (translated awkwardly as

“age-ification”), lühua 绿化 (“greenification/greening”), nianqinghua 年轻化 (“youth-ification”),

and so on. The People’s Liberation Army now talks of the geminghua 革命化 of the military, a term which can only be awkwardly translated with the tongue-twister “revolution-ization” (since the word “revolutionizing” has a different, non-political sense). And finally, I once saw, in a scholarly article on the poetry of Tang poet Wang Wei, the mention of a certain poetic quality called bukepangguanxing 不可旁观性, which I can only render as “un-observe-ability”. Someone looking at only modern Chinese and English might conclude from such examples that it is English, rather than Chinese, that has trouble expressing abstractions. It is perhaps ironic that the Chinese language borrowed the suffixation technique for signaling abstract nouns from Western languages, and now uses the device to coin words that Western languages have no ready equivalents for.

There is also a class of abstract nouns that constitutes a covert category in modern Chinese (see Zhao Yuanren, 1968). Nouns such as bing 病 “ sickness”, 恩 “ kindness”, huo 祸 “disaster”, daoli 道理 “reason”, keneng 可能 “possibility” etc. can only take classifiers like

zhong 种 , lei 类 , pai 派 , etc., and partitive measures like xie 些 and dian(r) 点 ( 儿 ), but no

individual measures (including ge 个). Thus one must say yi zhong changshi 一种常识 “a kind of

common sense”, but not *yi ge changshi 一个常识 “a common sense”, yi zhong manzu 一种满足 “a kind of satisfaction”, but not *yi ge manzu 一个满足 “a satisfaction”. I will just point out here that the syntactic behavior of these nouns bears a resemblance to that of mass nouns such as shui 水 “ water”, which also can take some of the same classifiers (zhong 种 , lei 类 , etc.) in

addition to partitive measures, as well as container measures such as yi bei 一杯 “one cup” etc. It should be noted that not all linguists agree with Zhao Yuanren’s categorization of abstract

nouns in Chinese. Wang Li (1984), does not recognize Zhao’s covert category, but does consider a limited class of philosophical terms to qualify as abstract nouns:

“What we call nouns do not have quite the same scope as English nouns. Generally speaking, our nouns, aside from philosophical terms, refer only to specific things, and one might even say they refer only to perceptible entities. It can be said that there are no words in the Chinese dictionary at least, that correspond either to abstract nouns in English derived from adjectives (such as kindness, wisdom,

4

humility, youth, etc.) or abstract nouns from derived from verbs (such as invitation, movement, choice, assistance, arrival, discovery, etc.). We cannot, from the form of the Chinese word alone, distinguish any abstract feature; abstract nouns are identical in form to adjectives and verbs. As stated above, I do not advocate classifying words according to their function, thus we cannot offer a sentence like “Wo xihuan ta de congming” 我 喜 欢 他 的 聪 明 [“I like his cleverness”] as evidence to prove that

congming 聪明 is an abstract noun, nor from the sentence “Ta fei le chang shijian de

xuanze 他费了长时间的选择 conclude that xuanze 选择 is an abstract noun. If we

make our distinctions conceptually, in Chinese congming 聪 明 is definitely an

adjective, because it expresses an aspect of character; xuanze 选择 is likewise clearly a verb because it expresses an action....

Words such as zhengfu 政府 “government”, yihui 议会 “convention”, tuanti

团体 “organization”, zhengzhi 政治 “politics”, jingji 经济 “economy”, refer to things that cannot be perceived by the senses, yet it should be clear that they are no less concrete because of this, and their status as nouns remains undeniable.

Finally, there remains a class of philosophical terms such as dao 道 “way”, de

德 “virtue”, pin 品 “quality”, xing 性 “essence”, etc.; only terms such as these can be considered true abstract nouns. If it can be said that Chinese has abstract nouns, it is only this small class of items.” (pp. 21-22, translation mine)

Wang Li’s analysis basically amounts to the observation that abstract nouns do not constitute an overt category in Chinese, which is obviously the case with ancient Chinese, as well. Nevertheless, given a clear syntactic context, Chinese speakers have no more difficulty understanding a sentence like Shibai shi chenggong zhi mu 失 败 是 成 功 之 母 ¸ “Failure is the mother of success” than English speakers do in understanding “Money is the root of all evil”, where the word evil has no morphological marking to indicate its change of status from an adjective to an abstract noun.

There is a small class of words in Mandarin derived from verbs that seem to exclusively denote abstract entities. Among these are words like siwang 死亡 “death”, and yuedu 阅读 “[the

process of] reading”, chouhen 仇恨 “hatred”, and xinyang 信仰 “belief”, all unambiguously abstract nouns.

他死了—> 他的死 (or 他的死亡) He died his death

These words form a quite small class in Chinese, and most seem to be formed by compounding two synonyms or near synonyms: yue 阅 “to look at” + du 读 “read” —> 阅读 “reading”, si 死 “die” + wang 亡 “disappear, die” —> 死亡 “death”, etc.

Since, as mentioned above, modern Mandarin now has a set of suffixes corresponding to those that mark abstact nouns in other languages (-ness, -ity, etc.), this creates the possibility for some small class of words to hypostasize almost exactly like their English counterparts:

这件事很重要 —> 它的重要性This matter is very important its importance

5

The suffixation process (zhongyao 重要 “ important” + xing 性 “-ity, -ness”) makes the word clearly an abstract noun in any context.

However, the fact that such devices exist does not imply that the full-scale morphology of Western languages has been imported into Chinese. These devices are still far from being as fully productive, common, and idiomatic as similar principles of word formation in most Western languages. In general, one does not see in Chinese the kind of routine hypostatization7 of concepts that is so common in a language like English. One can observe this by comparing say, translations of People’s Daily articles in the English-language China Daily, or English language quotes translated in other Chinese newspapers. To take just two examples of English sentences translated into Chinese newspapers:

Ex. 1: “Without an understanding of the larger trends that are restructuring our society, the enactment of economic reform will be based on out-of-date assumptions.”

Translation in a Chinese newspaper:如果不了解正在使我们社会进行结构改革的大趋势 ,我们只能根据过时的假设来进行经济改革.Backtranslation: “If [we] do not understand the larger trends that are restructuring our society, we can only enact economic reform based on out-of-date assumptions.”

Ex. 2. “A foretaste of the seriousness of incivility is suggested by what has been happening in Houston.”

Translation:休斯顿所发生的情况预示:如果不讲文明,将会产生何种严重的后果.Backtranslation: “The situation happening in Houston foretells: If [one] is not civil, [it] will produce serious consequences.”8

Though more literal translations rendering the hypostasized meanings are possible in Chinese, they are not in keeping with the customary modes of expression of the language; translators who translate between English and Chinese are aware of this fact, and numerous textbooks and treatises on translation provide stock solutions to the problems presented by the constant reification of verbal and adjectival concepts in English.

The use of abstract terms allows a language like English certain ambiguities that must be dealt with in the Chinese translation. For example, note in the second example above that the abstract noun “seriousness” in the English sentence is translated more specifically in Chinese as yanzhong de houguo 严重的后果, “serious consequences”, even though the translator could have

opted for the more awkward abstract noun yanzhongxing 严重性 “seriousness”. The result of this general resistance to hypostatization results in entire domains of routinely reified concepts in English being less idiomatically expressible in Chinese. One such domain is that of English terms describing attitudes, attributes, personality traits, etc., which usually must be expressed more concretely in Chinese.

7 The term “hypostasis” refers to the linguistic and conceptual treatment of an intangible abstract process as a noun-like entity—the reification of an abstraction.8 This example from Lian Shuneng 连淑能, Yinghan duibi yanjiu《英汉对比研究》 “Contrastive English-Chinese Studies”, 高等教育出版社, 1993.

6

For example, a sentence from a Time magazine article, “He was said to be impressed by Deng’s flexibility” is rendered in a Chinese translation as 据说他对邓的灵活态度印象很深 “It

was said he was impressed by Deng’s flexible attitude [taidu 态 度 ].” The word linghuo 灵 活“flexible” simply does not lend itself as readily to nominalization, and the term linghuoxing 灵活性 “flexibility” does not convey the same idea as the English word, and seems odd when applied to a human being. In many similar semantic contexts, such as sentences of the type “She showed a lot of Y”, where Y is a quality such as vivacity, enthusiasm, reluctance, tolerance, wilfulness, stubborness, likability, sincerity, coldness, warmth, cowardice, fear, indifference, etc. the term seems to resist wholesale hypostatization, and requires various forms of circumlocution, usually involving the addition of an additional element such as gan 感 “feeling” (e.g. kongju gan 恐惧感 “ fear[ful] feeling”, i.e. “fear”) , taidu 态 度 “ attitude” (e.g. renzhen de taidu 认 真 的 态 度 “sincere attitude”), or xin 心 “heart” (kuanrong xin 宽容的心 “tolerant heart” i.e. “tolerance”). Overall, the Chinese language seems less accommodating to what might be called “disembodied” abstractions. A sentence such as “There is still some honesty in the world” is more likely to be translated as 世上还是有一些诚实的人 “There are still some honest people in the world.”

Another implication of this state of affairs is that emotions are less likely to be treated as reified abstract entities in the Chinese psycholinguistic world. Lakoff & Johnson (1980), Lakoff (1987) and Kövecses (1989) have shown that patterns of idioms in all languages reveal relatively stable metaphor systems for various aspects of human existence (for instance, arguments are often conceptualized as wars, the process of communication is conceptualized according to conduit metaphors, etc.) Emotions in particular have a rich set of underlying metaphoric bases; that is, a set of phrases in English dealing with anger such as “He really exploded at me”, “I could feel the anger welling up in me”, “He blew up in my face when I told him the news”, and so on, reveal an underlying metaphoric conceptualization for anger that Lakoff & Johnson characterize as ANGER IS A PRESSURE INSIDE THE BODY. Other metaphoric categories for anger include ANGER IS AN OPPONENT (“I’m struggling with my anger”, “She fought back her anger”, “Anger took control over him”, etc.), ANGER IS INSANITY (“I just touched him and he went crazy!”, “One more complaint and I’ll go berserk”, “She went into an insane rage”, etc.) and many others. There are metaphor systems involved in English expressions concerning fear, worry, joy, relief, and for virtually all the human emotions labeled in the language.

An ongoing question has been to what extent these metaphor systems represent human cognitive universals, and to what extent they are culturally or linguistically determined. King (1989), applying Lakoff & Johnson’s framework to a survey of modern Mandarin idioms found metaphor systems similar to those in English in many semantic domains. Interestingly, however, he noted less of a tendency for Chinese to reify emotions, or to treat them as noun-like entities. Thus metaphoric idioms such as “Her frustration got the better of her”, “I detected a lot of nervousness”, “The child’s excitement spilled over into the next day”, “Her calmness in the face of disaster was admirable” “He made no secret of his dislike for her”, “There was panic in her voice”, and so on, often have no similar counterparts in Chinese, such meanings being most often formulated verbally. (Though it is easy to construct literal translations of such sentences, they tend to sound unnatural to Chinese speakers. Readers familiar with Chinese might try to formulate translations of these sentences to prove the point.) With regards to “fear” and “anger”, King has this to say:

7

[T]he most startling difference is in the observation that the metaphorical ways of talking about fear seem to be either unavailable to the speaker of Chinese or at best dispreferred... In English the function of [fear] metaphors is to highlight the most important causes of fear, such as death and physical pain. However, these causes all view fear as some kind of entity or thing which has a negative relationship between the emotion and the self. In Chinese these kinds of metaphors are generally dispreferred, although they may exist in the contemporary literature as a result of Western influence. In fact, except for ANGER, we have found very little evidence for the kinds of conceptual metaphors that appear to structure the domains of emotional experience in English. Even with ANGER we find that it is either difficult or impossible to entitize the emotion. For example, whereas in English we can find ANGER IS AN OPPONENT; ANGER IS A WILD ANIMAL; ANGER IS A BURDEN, we are once again prohibited from constructing such metaphors by constraints in Chinese.” (pp. 203-204)

King’s observations are interesting, though I disagree on his last point. It is unclear what specific “constraints” in Chinese he is referring to. There is nothing in the grammar or syntax of Chinese to prevent the formulation of such metaphorical utterances; rather there are deeply engrained habitual cognitive modes of expression which merely make such formulations less likely or less natural sounding in Chinese. Translators attempting to translate such metaphorical phrases into Chinese are dealing with a problem of semantic tendencies in the two languages, not grammatical or syntactic ones.

The question of semantics brings us back to the issue of the semantics of abstract nouns in English (and other Indo-European languages, of course). A full treatment of this problem is well beyond the scope of this paper, but I wish to bring up an important consideration for translating from a language like English, which employs morphological marking strategies, into a language like Chinese, which is relatively lacking in morphology. Very often the presence of an abstraction-marking suffix does not necessarily entail any significant semantic shift. Have we said anything different when we say “She shows a lot of spunk” vs. “She shows a lot of spunkiness”? And so with tact vs. tactfulness, joy vs. joyfulness, etc. To choose another example, if we ask a group of English speakers to characterize the semantic difference between “white” and “whiteness”, those few who don’t consider the question too silly to answer will very likely tell us that “white” refers to a color, whereas “whiteness” must refer to some quality of whiteness, in some more abstract sense. Yet if pressed for a more complete account of what this mysterious quality might be, such English speakers will very likely return in circular fashion to the color white itself. If faced with the task of translating, in a philosophical context,9 a term like “whiteness” into Chinese, should the translator at least consider applying some marker now available in the language to produce some lexical curiosity like baisexingzhi 白色性质 “white-quality”?

Jesperson (1924) makes some observations relevent to this question. He notes that, since terms like white and whiteness are both abstract, the problem is determining the semantic difference between the two:

“This usage, according to which concrete’ stands chiefly for what is found in the exterior world as something palpable, space-filling, perceptible to the senses, and abstract’ refers to something only found in the mind, evidently agrees with

9 I might mention in passing that even rigorous philosophical thinkers such as Plato and Aristotle often used leukotệs “whiteness” and to leukon “the white” in the same context with (apparently) no significant distinction involved.

8

popular language, but it does not assist us in understanding what is peculiar to such words as whiteness in contradistinction to other substantives.” (p. 134)

“[It is correct] that white and whiteness are equally abstract (in the sense separated from individual things’), but not...that the two are absolutely identical in meaning. The difference may be slight, but it is nevertheless a real one, else why should all nations have separate words for the two ideas? Observe that we use different verbs in the two cases: being white = having whiteness; the minister is (becomes) wise, he possesses (acquires) wisdom.” (p. 135-36)

Aside from the fact that his point about “all nations” having separate words for the two notions is wrong (and question begging), the more important question concerns the observation that whiteness is treated as a quality taking the verb to have. Since one can make the case that the concept “whiteness” merely reduces to “the quality of being white”, it is not clear what this difference tells us, other than that white can be an adjective while whiteness is a substantive. The real question is the semantic difference between white and whiteness as nouns.

Suffixes such as -ness and -ity can perform a wide range of semantic functions, some rather subtle. Could it be that in some cases speakers are encouraged by the morphological difference to manufacture a kind of “dummy node” for an illusory meaning?10 Note that it is not sufficient to merely point out that native speakers intuitively feel there is a difference in meaning; such intuitions might be misleading. There is even much speculation that many of the dead-ends explored by Plato were the result of mistakenly ascribing some kind of reality to non-existent objects which could be grammatically expressed with the tools of the Greek language (see Reding, 1986). That the Chinese philosophers never fell into this trap may be due to the fact that Chinese does not encourage such wanton hypostatization.

One is forced to speculate about similar elusive semantic distinctions when encountering seemingly parallel notions in the Greek and Chinese texts. Plato, for example, was concerned with the abstract commonality underlying all specific instances of number, and his discussions invoked the notions of “unity” and “duality”:

“...and therefore you accept no other cause of the existence of two (tou duo) than participation in twoness (tôs duados)... and whatever is to be one (en) must participate in oneness (monados)...” (Plato, Phaedo 101c)

It is significant that the very distinction so crucial to him was already marked in the language (or, perhaps more importantly, that there were standard devices for marking such distinctions as the need arose. For example, “unity” is also expressed as to en, “the One”). When we encounter passages that at least appear to invoke similar notions in, say Zhuangzi or the Neo-Mohist Canon, it is natural to ask how the Chinese themselves must have perceived such statements:

10 I do not believe this is an outlandish idea. Many years ago a Chinese friend asked me if my near-sightedness was slight enough

to allow me to occasionally go out without wearing my glasses. Pointing to my glasses I replied “Bu hui, wo li bu kai tamen” 不会，我离不开它们。 This remark amused my friend, who politely informed me that the word yanjingr 眼镜 “glasses” is singular in Chinese. This led me to wonder whether my mistake was a semantic one, or merely a formal grammatical one, like the speaker of a romance language mistakenly applying a gendered pronoun to an inanimate object when speaking English. Do English speakers really conceptualize glasses as plural objects, exactly like “gloves” or “shoes”? How about a word like “pants”? “Scissors”? “Outskirts”? “Wages”? “Pliers”? The fact that speakers will sometimes lazily say “Hand me a pliers” is perhaps significant. It may well be that the semantics of these words are not binary values, but rather “fuzzy” combinations of singular and plural. At any rate, this constitutes some evidence that the grammatical behavior of a word does not necessarily entail the full concomitant semantic content the category normally carries.

9

「一二」不相盈.“ ‘One’ and ‘two’ do not fill each other.”11

Did the Mohist here have anything like the Greek monados (“oneness” as opposed to “one”) and duados (“twoness”) in mind? Certainly any abstract reference to, say, the number 6, must refer to something like the abstract principle underlying any concrete instance of six pebbles, six horses, or six whatever. Is this entity any different than what would be denoted by the special word “sixness”? The suffix seems to indicate that the entity is to be interpreted as a quality of some kind, but what quality could that be, if not simply the abstract principle that all sixes have in common?12

Clearly there was a felt need for suffixation devices to translate abstract nouns into Chinese, or the common practice of using such new suffixes as -xing 性 and -hua 化 would not have arisen. But Chinese has been and presumably still is capable of accomplishing the same communicative task through purely syntactic means (just as zhe zhong keneng 这 种 可 能 and zhe zhong

kenengxing 这 种 可 能 性 are both equally available to the speaker). It remains a task of the translator to determine whether or not there is any true semantic distinction between the two forms in the target and source language, and to translate accordingly. This task will obviously be harder in the context of philosophical texts, in which highly abstruse concepts and distinctions are constantly dealt with.

The Chinese language shows a reluctance toward “disembodied” or stranded abstractions. This is perhaps seen in the certain aspects of the historical tendency toward bisyllabic compounds in the language. Word formation of verbs for common actions such as eating and talking are predominantly VO (verb-object) structure: chifan 吃饭 “eat-rice” [eat], shuohua 说话 “speak-

talk” [talk], dushu 读 书 “ read-book” [read], zoulu 走 路 “ walk-road” [walk], kaiche 开 车 “drive-vehicle” [drive], zuofan 做饭 “make-rice” [cook], xiezi 写字 “write-character” [write],

huahuar 画 画 儿 “ paint-painting” [paint], maren 骂 人 “ curse-person” [curse], sharen 杀 人 “kill-person” [kill], etc. We see other principles of word formation in the class of verbs, such as

verb-verb combinations, of course, such as shibai 失败 “ lose-fail” [fail], siwang 死亡 “die-

disappear” [die], xuexi 学习 “study-familiarize” [study], etc., but the most common form became one in which the action of the verb has an object. Most of these compounds have a certain amount of synctactic mobility in the modern language, functioning as abstract nouns in nominal position (Chifan shi bu neng couhe de 吃 饭 是 不 能 凑 合 的 “ Eating is something you can’t treat casually.”). But though in many cases the literal sense of the compound is effectively a dead

11 The point of this particular passage is that, although certain pairs of notions are inseparable or logically linked as a unit, they are still considered as two distinct things. Specifically, the Mohist characterizes width and breadth as being analogous to “hard and white” (坚白), which is to say, examples of two things that mutually pervade each other; hard and white were considered two, but one in the stone, just as width and breadth are two, but one in the object. “One” and “two”, on the other hand, were not considered to be mutually “pervading”.12 Interestingly, I have encountered certain Chinese translations of passages from Aristotle and Plato that do not attempt to render such distinctions marked in the original Greek. For example, there is the following passage from Aristotle: “The elements of number, according to them [the Pythagoreans] are the Even and the Odd...Unity [to en, “the One”, “one-ness”] consists of both (since it is both odd and even); number is derived from Unity.” (Metaphysics, I, v, 5-6). Feng Youlan (1967) translates the last sentence ¡u¤ C (p. 549) [Backtranslation]: “They take ‘one’ as coming from both (since ‘one’ is both odd and even); from ‘one’ comes all number(s).” The question remains: Does yi 一 here convey the same sense as “Unity” [to en]?

10

metaphor, the compounds always remain decomposable to some extent, and do not hypostasize quite as readily as do English verbs.13

3. The translation of the abstract generic from English into ChineseThe definite article “the” in English often serves to signal what is sometimes called the

“abstract generic”, that is, a singular term that denotes an entire class or type of items, as in “The tiger is in danger of extinction.” Most treatments of English grammar maintain that generic use of the definite article, indefinite article, and plural (or zero article) function equivalently to denote the entire class:14

(a) A panda eats bamboo leaves.(b) The panda eats bamboo leaves.(c) Pandas eat bamboo leaves.

Despite the apparent functional equivalence in this and other cases, apparently there are semantic differences between these usages. (And all three would be translated by the same sentence in Chinese, Xiongmao chi zhuzi ye 熊 猫 吃 竹 子 叶 . Often the definite generic is analyzed as referring to the class as a whole, whereas the plural denotes the members of the class.15 The indefinite article, in contrast to the definite generic, supposedly picks out indifferently any member of the class. From the point of view of abstraction, however, none of these characterizations quite captures all the senses of the definite generic. Other contexts bring out other distinctions:

The alphabet was invented by the Phonecians.*An alphabet was invented by the Phonecians.?Alphabets were invented by the Phonecians.

The indefinite article here, “an alphabet,” loses the generic sense, and the plural “alphabets”, while acceptable to some ears, seems more to evoke the dazzling variety of various world alphabets rather than the austere abstraction “the alphabet”. And here we see a special function of the definite

13 One could at least make the case, for example, that a noun like sharen 杀人 denotes a narrower concept than the English killing,

which can be extended to animals and insects; dushu 读书 refers to books, and probably is more resistant than the word reading is to metaphorical extensions like “reading your mind”. Native Chinese speakers speaking English as a second language often make sentences such as “Do you like to read books?”, “I’m going to sing songs”, “I can drive a car”, where native English speakers more often simply say “read”, “sing”, and “drive” in such contexts.

14 For example, Quirk et al (1972) say “The distinctions between definite and indefinite, and between singular and plural , are important for specific reference. They tend to be less crucial for generic reference, because generic reference is used to denote the class or species generally. Consequently, the distinctions of number which apply to this or that member, or group of members, of the class are neutralized, being largely irrelevant to the generic concept.” (p. 265)15 Jesperson (1927), for example, says “[P]lurals denote all members of the species, but they do not denote the species itself as the-sg’ [the + singular noun] does.” This analysis seems to work better for contexts like “The panda is in danger of extinction” than in the above case, where the issue is suitableness of individual pandas as pets, not the species as a whole. It should be noted that there is considerable disagreement as to whether the bare plural and the definite generic actually express any semantic difference. Rundle (1979) has this to say: “The predicates attached to the cuckoo’ may sometimes require the plural cuckoos’, sometimes the singular a cuckoo’, but, figurative uses aside, one or the other is surely always in place. Which seems to confirm the original suggestion: generality can be introduced in a number of different ways involving verbs, determiners, adverbs, and other parts of speech. Generic the’ is one such device, but it is in no way creative’. It gives us a deceptive variant on the other forms, but it does not mark ascent to a new range of entities.” (p. 206.) Here Rundle is mainly speaking about the definite generic used to denote species, as distinct from some other uses dealt with in this section.

11

generic: it often refers, not to the members of the class or the class as a whole, but rather to a conceptual amalgam of essential and defining features of the class. “The alphabet” denotes that particular composite of only those essential features shared by all alphabets—a “Platonic” alphabet.

For this reason, the definite generic is frequently used when speaking of inventions, which are typically envisioned as a set of abstract design features, independent of any particular past, present, or future physical instantiation (“The wheel was invented by cavemen,” rather than “Wheels were invented by cavemen.”) Of course this analysis is true for any object conceptualized as an abstract structure, as in “The mammalian eye evolved independently from the insect eye,” where abstract evolutionary design features are being emphasized.

In speech and writing, speakers of languages with the singular-plural distinction constantly alternate between singular and plural generic references, often as a result of various contextual pressures or subconscious imagery, but sometimes for no apparent semantic reason at all. The following passage, a verbatim excerpt from a call-in talk radio show, is typical in this regard:

“The American President is unique among world leaders, in that he returns to civilian life in a very short time, at most eight years. A president has got to do whatever he’s going to do in his short term of office, and so by and large, presidents don’t get to do all they set out to do by the time they’re done.”

The speaker begins with the singular generic reference, commonly used in contexts such as this where the subject is envisioned as a timeless abstraction. The next reference shifts to the indefinite article, perhaps since the context has shifted to the more specific behavior of American presidents. The final shift to plural may be simply an arbitrary choice, or may entail some subtle influence of imagery in mind of the speaker. The translator rendering this passage into Chinese might very well choose to ignore the three separate forms of expression, and simply opt for Meiguo zongtong 美国总统 in each case.

A full treatment of the semantics of the definite generic is beyond the scope of this study. But it seems clear that the use of the singular is often semantically significant, whether denoting the class as a whole, positing a set of essential features, or merely summoning up the image of a singular instance as a representative of the category (what Quirk et al, 1972 refer to as “the class as represented by its typical specimen.”) While plural and singular generic uses overlap and blend into one another depending on semantic context, as a rule the singular is most likely to be used in cases where there the referent seems to be some core of defining characteristics: “the internal combustion engine”, “the Victorian novel”, “the American family”, and so on.

It is difficult to compare the frequency and patterns of usage of the definite generic with modern languages, but examining the texts one gets the impression the device is at least as pervasive as it is in English. The definite generic seems so natural to speakers of the Indo-European language group that it receives little mention in most comparative grammar books or translation treatises. It is one of the aspects of language that scholars doing comparative linguistics between European languages take for granted, until examining languages of the Far East or other “exotic” languages.

Chinese has no definite article, and thus it is technically impossible for the language to directly signal grammatically the kind of abstract entity that the English device evokes. Before we ask whether or not there are other methods for signaling such abstract entities, we must ask: Is this singular, abstract generic lacking from the Chinese psycholinguistic world?

12

Clearly, explicitly asking native Chinese speakers such a question would be fruitless. Introspection is a notoriously unreliable tool, and subjects are prone to confusion and confabulation when asked to consciously reflect on linguistic processes that usually take place unconsciously and spontaneously.

3.1. Psycholinguistic tests for the abstract generic in ChineseIn order to circumvent such direct introspection, I devised several test stimuli which would

test for the presence of a singular, abstract generic in the Chinese psycholinguistic environment. The tests were quite preliminary and exploratory, and designed merely to test the waters with regards to singular-plural imagery in Chinese. Chinese subjects16 were given sentences with various nouns in subject or object position, and in varying contexts designed to evoke the generic concept to greater or lesser extent. They were then asked to replace the nouns with either singular or plural pronouns. For example,

熊猫正濒临灭亡 —> ——?—— 正濒临灭亡“[The/a] panda/s is/are on the verge of extinction.”

where the blank is to be filled in with either ta 它 “it” or tamen 它们 “they” (i.e., either ta zheng

binlin miewang 它正濒临灭亡 or tamen zheng binlin miewang 它们正濒临灭亡). Though the answers do not, of course, always result in natural-sounding Chinese sentences, they presumably are able to provide clues to the underlying mental imagery with regard to the noun in question. (It was stressed at the time of the test that the resulting sentences would not always be natural-sounding.) The methodological assumptions of the test were as follows:

(1) Given the lack of singular-plural marking, one might assume that the semantics of generic reference in Mandarin would always be neutral, never singular or plural. If so, then the the results of the tests should be close to chance. This is because, presumably, subjects would either (a) consider the question nonsense and choose randomly, or (b) attempt to answer on the basis of other idiosyncratic factors. It is also possible that subjects might attempt to answer the question by determining from context whether the actual referent of the noun in question is plural and simply replace all nouns interpretable from context as plural with the plural pronoun tamen, “they.”

(2) If the semantics of generic reference in modern Mandarin is always plural, then presumably subjects would straightforwardly replace all the nouns in question with the plural pronoun tamen (i.e., the equivalent of the above mentioned comment of the professor’s wife: “Either you are talking about a single kangaroo or about all kangaroos. What else is there?”). This state of affairs would in principle be indistinguishable from the case in (1), though it could be assumed to be more consistently applied in this case.

(3) If, on the other hand, the semantics of generic reference is context dependent—that is, neutral in some contexts, plural or singular in others—then we should expect to see contexts where a significant number subjects choose one possibility over another. Given the fact that the choice of whether or not to use the singular generic in English is extremely variable (as shown in examples above), we can expect wide statistical variation, with some contexts eliciting a strong singular or plural response, and others eliciting only statistical tendencies in one direction or another. In

16 Subjects for the various tests were Beijing Normal University undergraduates, and a group of Chinese students and visiting scholars in an English class in the U.S. See Appendix 1 for full details.

13

addition, we should not be surprised if the results do not completely tally with English language preferences in using the singular generic.

Note that, if the above assumptions are valid, any significant tendency for subjects to replace nouns denoting groups or pluralities with the singular pronoun ta would be evidence for the psycholinguistic presence of the singular generic in Chinese.

The results of the test indeed seem to reveal evidence for the generic singular in Chinese. For example, for the sentence

在美国汽车很贵.“In America cars/ the/a car are/is very expensive.”

66% of subjects replaced qiche 汽车 “car”, with the plurual pronoun tamen 它们. But in another sentence, the concept of the car as an invention is highlighted:

汽车是福特发明的.“The automobile/automobiles was/were invented by Ford.”

In this context, fully 93% of Chinese subjects replaced the word qiche 汽 车 “ car” with the

singular pronoun ta 它 , suggesting that subjects considered the referent to be a singular abstract

notion of car. In the sentence Zhinanzhen shi cong Zhongguo chuan dao xifang de 指南针是从中国传到西方的 (“The compass/compasses came to the West from China.”), the role of the compass as an invention is stressed. Again, most subjects (86%) opted for a singular pronoun. By contrast, in sentences where inanimate objects were are clearly quantified as plural, most subjects replace the noun with the plural pronoun. (For the sentence Zhe xie shouyinji tai gui le 这些收音机太贵了“These radios are too expensive,” 94% of subjects used tamen, “they”.)

Note that some nouns in certain English contexts conventionally take the definite article (“play the piano”), some conventionally take the plural (“She likes to ride horses”, “He’s afraid of snakes”), while others with mass-like behavior take no article (“watch television” as opposed to “listen to the radio”). Again, we should not expect whatever these Chinese subjects’ mental imagery might be to always correspond to idiosyncratic English usage; even between closely-related Indo-European languages there are significant differences in patterns of generic usages. One constantly encounters generic usages in translation that violate our conventions:

北京人已经使用天然火. 他们用火烧烤事物...Beijing ren yijing shiyong tianran huo. Tamen yong huo shaokao shiwu…“Peking Men already made use naturally occurring fire; they used fire to cook food...”

The customary English way of expressing such a sentence would use the abstract singular generic, “Peking Man...he...”. The fact that in English we routinely envision individual exemplars in such contexts (“Peking Man”, “Java Man”, “prehistoric man”, etc.) is no doubt largely conventional.

Note that singular pronoun responses in the test give no evidence as to the English distinction between the definite generic (“The car was invented by Henry Ford.”) and the indefinite

14

generic (“The average American drives a car to work”, i.e. “any random member of the set of all cars”). Since Chinese lacks the definite article, it is questionable whether or not such a psycholinguistic distinction would play a role in the language. It should be noted, however, that indefinite generic is quite common in Chinese for some contexts, usually when the speaker or writer wishes to introduce a hypothetical member of a class into the discourse:

当一位作家拿起笔来写文章的时候, 他脑子里必须有构思.Dang yi wei zuojia na qi bi lai xie wenzhang de shihou, ta naozi li bixu you gousi.“When a writer picks up his pen to write an article, he must have a plan in mind.”

However this usage is almost always restricted to cases when the person or object is being considered as random member of the class for some rhetorical purpose, and does not function to represent the class as a whole. A thought such as “Writers have a hard life” is expressed (ambiguously) without quantification: Zuojia de shenghuo hen jianku. 作家的 生 活 很艰苦 “[The/a] writer/s [+ genitive particle] life/lives is/are hard.” A general statement like “A good

child obeys its mother” is usually expressed (ambiguously) as Hao haizi ting mama de hua. 好孩子听妈妈的话. To preface such general statements with yiwei 一位 or yige 一个 sounds jolting and

unnatural to Chinese ears (though the pluralizing suffix men 们 can often be added to the noun with no problem).

These issues are almost intractable and intrinsically blurry. Certain of these forms are obviously functionally equivalent in most contexts, while others some have subtly different semantics which are revealed through patterns of use. Still others are essentially fossilized conventional usages (akin to dead metaphors) and have little psycholinguistic significance. With generic reference, as with so many linguistic phenomena, context is everything. Judgments are subjective, and a great number of usages considered ill-formed or ungrammatical on some account turn out to be perfectly natural in some unexpected context. This is not so surprising if one holds a model of fluid, subtly nuanced and blurry semantic meanings in the mind being expressed by the much coarser-grained and more discrete system of natural language.

These tests and collected examples are meant to be merely preliminary and exploratory probes into the reality of something like the definite generic in Chinese. Some results are no doubt a result of methodological problems, or subjects’ misunderstandings about the nature of the task. A thorough study would have to more carefully isolate factors such as the position of the word in the sentence, as well as distinctions such as class vs. kind, genus vs. species, animate vs. inanimate, count noun vs. non-count noun, etc. It was not my intent to carry out an exhaustive cross-linguistic survey of specific and generic reference in Chinese and English, but merely to demonstrate that singular and plural imagery play a psycholinguistic role in generic and abstract reference in Chinese, despite the fact that the singular-plural distinction is not marked as it is in English. Whatever the final implications of this sketchy data, I believe they suggest at least these tentative conclusions:

(1) Despite the standard ambiguity with regard to number in Chinese, Chinese speakers seem to have singular and plural schemas for unmarked nouns (though they are no doubt evoked less consistently and obligatorily than they are in English).

15

(2) These schemas seem to exhibit context-dependence.(3) Some patterns seem somewhat parallel to forms such as the definite or indefinite generic

in English (and Indo-European in general), which suggests that these patterns have some cognitive universality.

One does see some evidence in usage patterns of singular and plural pronouns for something akin to the definite generic in modern Chinese. Despite the fact that use of the pronoun is much less frequent in Chinese than in English, and thus there is often no evidence in the sentence of any underlying singular or plural imagery associated with a noun, one does encounter a number of clear examples in written and spoken language. The first example is from Lin Yutang:

普通男人是美国民主主义的基石, 因为代表最多数的是他, 而不是美国绅士.17

Putong nanren shi Meiguo minzhuzhuyi de jishi, yinwei daibiao zui duo shu de shi ta, er bu shi Meiguo shenshi.“The ordinary man is the cornerstone of American democratism, because he [他] and not the American gentry represents the majority.”

我人为还是意大利的妈妈最伟大, 她一切是为了孩子...18

Wo renwei haishi Yidali de mama zui weida, ta yiqie wei le haizi…“I think the Italian mother is the greatest, she [她] does everything for the child...”

It is difficult to say how much such usages are an “indigenous” part of the Chinese psycholinguistic world, since they have been subtly effected by exposure to Western languages and Western works in translation. Lin Yutang is a perfect example of a Chinese writer steeped in Western languages and modes of expression, and it could be that his choice of pronouns was affected by his exposure to English. At any rate, such usages now seem to acceptable in the language, though many Chinese people characterize them as somewhat Ou hua 欧化 “Europeanized”. It is also possible that the distinction was always present psycholinguistically in the language, but has become more evident with increased use of the pronoun.

The definite generic is used routinely to distinguish between a class conceptualized as a unity vs. as a collection of individuals. It is quite common to see this distinction applied in writings about the Chinese Communist Party, for example:

高速度增长的经济迫切需要大量人才. 中国共产党已明白此理, 所以他们正在培养一批学者…Gao sudu zengzhang de jingji poqie xuyao da liang rencai. Zhongguo gongchandang yi mingbai ci li, suoyi tamen zheng zai peiyang yi pi xuezhe…“The rapid economic growth urgently requires a large talent pool. The Chinese Communist Party already realizes this, so they [ 他 们 ] are now training a group of scholars...”

17 Lin Yutang sanwen 《林语堂散文》 (Lin Yutang Prose) Vol. 1, Hebei renmin chubanshe, 1994, p. 194.18 Beijing radio broadcast.

16

共产党它自己正在反省,而且整个社会的变革, 共产党它自己正在支持.Gongchandang ta ziji zheng zai fanxing, erqie zheng ge shehui de biange, gongchandang ta ziji zheng zai zhichi.“The Communist Party itself [它自己] is in the midst of self-reflection, and

in addition, the Party itself [它自己] is supporting the reform of the whole society.”

In the first quote, the Party is envisioned as the members in it, in the second as a singular whole, “chunked” on a somewhat higher, more abstract level. This usage would of course correspond to pronoun referents in English for nouns like “government”, “company”, “organization”, which can take singular or plural.19

Suppose we want to ask a very simple question (one that would have seemed strange to the Mohists, but seems reasonable to us), namely “For any given instance of the word junzi 君子 “gentleman/men, superior man/men” in a sentence, is the underlying mental imagery that of the whole class of junzi, a single exemplar of the class, or is the word semantically entirely neutral as to number?” (I hope I’ve made the case by now that this is not an entirely silly question.) Granted that the question is already complicated enough for a language with plural markings, are there any clues we might base our answer on? The fact that in English we almost invariantly translate the word in the singular tells us nothing except that we have some tendency to use singular reference in talking of exemplars and moral models (“the Sage”, “the Superior Man”), and does not entail that the Chinese a similar austere abstraction. Occam’s Razor might dictate that there is no imagery at all, or if context dictates, it must be plural. But we have seen from the example of modern Chinese, which is ambiguous in the same way classical Chinese is, that speakers do sometimes have singular mental schemas for plural objects. Do modern speakers of Mandarin tend to conceptualize a singular junzi, as we do? If one peruses any of the many baihua translations of the classical texts, one does occasionally see passages in which the translator evidently had such a singular generic junzi in mind. For example, in one such book, a sentence from the Analects, 子曰: “君子易事而难说也.” (Analects, Zilu 子路 13.25) is translated into modern Chinese as 空子说: “君子容易共事而难以让他高兴.” (“Confucus said: “[As for] junzi, [he’s] easy to serve, but it’s hard to please him

[他].”)20

It’s difficult to say how prevalent such interpretations are. I gave one group of subjects the following question in the same format as the test for the singular generic:

君子之交淡如水. ---> ——— 之交淡如水.Junzi zhi jiao dan ru shui. _____ dan ru shui.“[The/a] junzi/s’ social interactions are clear as water.”

Out of 23 subjects, 16 (70%) replaced the word junzi with tamen 他们 “ they”, which perhaps should not be particularly surprising, and suggests that at least some speakers have some awareness

19 There is some variation in this regard between varieties of English. In American English, companies tend to be conceptualized as singular entities (“IBM wants to be your personal computer company”) whereas in Great Britain they are treated as plural (resulting in sentences jarring to American ears: “IBM want to be your personal computer company”).20 Baihua si shu 《白话四书》Huang Piaomin 黄朴民 et al, (eds.) Sanqin Publishing Co. 三秦出版社, 1990.

17

of the possibility of singular abstract reference for such contexts. But it is also significant that this is in contrast to the habitual English-language custom of conceptualizing such generic classes as being represented by a singular, representative member.

There are some special cases in the Confucian texts where clearly the referent is to a singular junzi, but these represent rather special cases:

有人于此, 其待我以横逆, 则君子必自反也: “我必不仁也, 必无礼也.21

“Here is a man, who treats me in a perverse and unreasonable manner. The superior man in such a case will turn round upon himself— I must have been wanting in benevolence; I must have been wanting in propriety.’”

Here we have the junzi in a hypothetical context in which an individual is involved, a denotation somewhat different from an abstract, ideal junzi. In the absence of such a context, the number may usually have been neutral.

It seems likely that these same sort of singular generic schemas may have been only very sporadically and contextually evoked in Chinese, and occur in the mind of Chinese speakers in the process of discourse, unmarked and unnoticed.

4. The “ad hoc” classThere is a semantic device very much related to definite and indefinite generics, and that is

to treat a proper name or some well-defined situation generically as a common noun to posit an ad hoc class of individuals that share common traits with the exemplar. The noun in question becomes subject to quantification, pluralization, etc. and the change can be signaled by articles, the plural or any number of means. For example:

Hawkings is brilliant, but he’s certainly not an Einstein . China can’t take another Tiananmen.People opposed to our involvement in Bosnia want no more Vietnams.

This abstraction is in some sense the mirror image of the definite generic: whereas in the singular definite generic certain characteristics of a class are abstracted out and conceptually embodied in one individual, in the ad hoc class certain features of one individual person or situation are abstracted out and used to create a hypothetical class. Thomas Reid characterizes the device well:

“Sometimes the name of an individual is given to a general conception, and thereby the individual in a manner is generalized; as when the Jew Shylock, in Shakespeare says—”A Daniel come to judgment; yea, a Daniel!” In this speech, “a Daniel” is an attribute, or a universal. The character of Daniel, as a man of singular wisdom, is abstracted from his person, and considered as capable of being attributed to other persons.”22

This linguistic device has a certain relationship to nominalization. Though the usage usually involves proper names, the process involved is actually a very general cognitive one, involving the conceptual highlighting of certain aspects of any relatively unique concept or state of

21 Mencius Lilou Book IV, Pt. II, Ch. 28.22 Thomas Reid “Of Abstraction”, in Van Iten (1970), p. 111.

18

affairs. We isolate a certain state of affairs, “The stock market crashed in 1929”, which we nominalize and treat as a thing: “the 1929 stock market crash.” From then on we are free to treat the object we’ve created, divorced from its original context, as a class of situations: “The recent Wall Street plummet was not a 1929 stock market crash,” or, “We don’t want any more 1929 stock market crashes, or simply “any more 1929s.”

The ad hoc class does seem to be a part of the modern and ancient Chinese psycholinguistic world. I jotted down this sentence from a Beijing radio announcer:

我们国家的运动员中有几个郎平?Women guojia de yundongyuan zhong you jige Lang Ping?“Among our country’s athletes, how many Lang Pings are there?”

And the usage is also evident in the written language:

这不足怪, 因为大陆并非只有一个陈希同…23

Zhe bu zuguai, yinwei dalu bingfei zhi you yige Chen Xitong…“There’s nothing strange about this, because there’s not just one Chen Xitong in mainland China.”

...这个国家的今天绝对不是这股风貌, 或者, 它早已成为另一个印度了.24

…Zhe ge guojia de jintian jue bu shi zhe gu fengmao, huozhe, ta zao jiu cheng wei ling yi ge Yindu le.“...then this country today wouldn’t look this way; perhaps it would have already become another India.”

My own impression is that such usages are fairly common. Many Chinese report that it is much more idiomatic to omit any quantifiers, so that “There won’t be any more June Fourths [i.e., “any more Tiananmen Square Incidents”] in China” is more likely to be expressed as Zhongguo bu hui zai fasheng Liu Si “中国不会再发生六四.”, and not Zhongguo bu hui zai fasheng yige Liu Si “中国不会再发生一个六四.”25 Also, since Chinese lacks plural forms (except for limited cases where

the pluralizing particle men 们 applies), sentences like “Shakespeares are rarer than Napoleons” must be translated with awkward and unidiomatic verbosity: Shashibiya shi de renwu bi Napolun de renwu shao. 沙士比亚式的人物比拿破仑式的人物少 “Shakespeare-like people are rarer than Napoleon-like people.” This raises the question of whether, for ergonomic reasons, such utterances are might be somewhat rarer in Chinese.

23 Shijie ribao 世界日报, Oct. 29 1995, p. A524 Wang Shan王山 Di san zhi yanjing kan Zhongguo 《第三只眼睛看中国》·(“Looking at China with the Third Eye”), Shanxi People’s Publishing House, p. 40.25 There are some uses of yige 一 个 which might seem related to the ad hoc class, but are probably rhetorical or dialectical

idiosyncrasies. The Chinese song The East is Red begins with the line “东方红,太阳升,中国出了一个毛泽东...” (“The east is

red/The sun rises/China has produced a Mao Zedong...”) Also this example from the novel Wei Cheng《围城》: “你就象木头人似的, 一句话不说,全忘了旁边还有个我...” (“It’s like you were made of wood. You didn’t say anything, you completely forgot there was a me at your side...”)

19

It is not difficult to find what appear to be examples of this phenomenon in the classical texts:

便之居于王所, 以在于王所者, 长幼卑尊, 皆薛剧州也...26

“Suppose that all in attendance on the King, old and young, high and low, were Xue Juzhous...”

However, given the extreme terseness of classical Chinese is such that often the entire semantic relation must be inferred from context, such that a perfectly reasonable alternate translation would be “Suppose all in attendance... were [people like] Xue Juzhou...” But is there any difference between these two formulations? Clearly there must a strong functional equivalence, but does the difference in formulation have any semantic implications? Specifically, is one usage more abstract than the other in some sense? The question brings to mind, of course, the distinction between the poetic devices of simile (“Love is like a rose”) and metaphor (“Love is a rose”); is there any real difference between these? These issues touch upon the subtlest linguistic distinctions that are the meat-and-potatoes of translators and poets.

Numerous other instances seem to invoke an ad hoc class to greater or lesser extent:

一薛剧州, 独如宋王何?27

“What can one Xue Juzhou alone do to the Song king?”

自有生以来, 未有孔子也. (Mencius, p. 315 )“Since there were living men until now, there never was another Confucius.”

今以燕伐燕, 何为动之哉? (Mencius Gongsunchou Part II, 4.8 (p. 339)“But now with one Yan to smite another Yan—how should I have advised this?”

人皆可以为尧舜, 有诸? (Mencius, Gaozi pt. II, 12.2)“All men may be [regarded as] Yaos and Shuns, is it so?”

These examples again suggest universals of cognitive mechanisms dealing with abstraction, regardless of the specificity or clarity with which the language marks them. However, the picture that emerges from a close comparison of these phenomena in Chinese and English is that the kind of marked, abstract generic entity that played such a crucial role in the Western languages and philosophical systems was not an important or salient aspect of the Chinese psycholinguistic world.

5. ConclusionIt is in some ways puzzling that the Chinese language and Chinese philosophy should have

a reputation in the West as being relatively non-abstract in nature. Everywhere one looks in the Chinese tradition—from the Yi Jing to Tang poetry to the stuffiest passages of the Confucian classics—one finds subtle and deep abstract concepts evoked and extended in flexible ways. Yet

26 Mencius Book III, Part 2, Ch. 627 ibid.

20

there is a general tendency in the sinological literature to speak of Chinese philosophy and Chinese culture as a whole as relatively indifferent to abstraction. What is generally meant, I think, is that Chinese thinkers do not very often bring up for discussion the kinds of abstractions the Western philosophical tradition have deemed important; or if they do bring them up, they don’t seem to pursue them with the same dogged intensity that an Aristotle or a Plato did. Chinese philosophy is considered less abstract, then, because there are no fully-developed theories of abstraction. From this general observation comes the suspicion that perhaps Chinese people do not have the same proclivity for abstract reasoning that Westerners supposedly do, and since it seems unlikely that there is some “abstraction genotype” missing from the Chinese gene pool, one seemingly plausible explanation for this state of affairs has been that the Chinese language somehow inhibits abstract thinking.

If we think of abstraction as involving the high-level attention to and flexible manipulation of categories and qualities of the physical and mental world, then it goes without saying that there are no non-abstract languages; for as we have said, the creation and extension of abstractions is one of the very core functions of natural language. The essential role of language in cognition is to “abstract out” features of the world such as “four-legged”, “feathered”, “white”, “courageous”, “passive”, “human”, “beautiful”, “good”, etc. in order to make sense of the ever-changing stream of sensory input. (While languages don’t always pick out exactly the same abstractions, the set each language is equipped with is flexible and fine-grained enough to give speakers ways of talking about almost anything.) Once these abstractions have been codified, they can be used by the Greeks to develop a theory of essences and qualities, or by Chinese correlative cosmologists to catalogue the various objects in the world according to the principles of yin and yang, or by the Neo-Mohists in order to expound a theory of how names relate to things. By this standard, virtually all systems of thought are abstract because they all work by manipulating abstractions in various ways. Furthermore, despite the great variety of language systems, there is a great deal of commonality in the way human beings cognitively conceptualize and treat abstractions, and I hope I’ve been able—to underscore this fact.

Not all languages are equally transparent as to their various functions. Linguistic functions can be accomplished in a wide variety of ways—by word order, by specific morphological markings, by repeating certain elements, by changes in intonation or stress, etc.—and some of these are easier to notice and explain than others. That is to say, speakers of the language are able to master the devices the language employs to signal meanings, but they most often are unaware of how these devices work, or even of their existence. While this fact does not have great significance for routine language use, it becomes important when translators set to work rendering into one language the abstractions from another language.

A question remains whether or not there are aspects of language that can facilitate or hinder abstract thinking, and whether or not Chinese speakers routinely deal with abstractions to the same extent that speakers of other languages do. Students at the University of Michigan and at Beijing Normal University were given the following question, in their native language:

If 1 were greater than 2, and 2 were greater than 3, would 1 be greater than 3?

如果一比二大, 而二比三大的话, 一是不是比三大?

21

The results were rather striking. 27 out of 27 American subjects answered “yes” to the question (100%), whereas only 23 out of 34 Chinese subjects (about 68%) answered “yes” to the same question. (The methodology is based on an experiment by Bloom, 1981).

Subjects were asked to explain their answers. Virtually all the American subjects answered said things like “Given the premise, it follows logically” and “This is trivially obvious”. Chinese subjects responding “no” or “not necessarily” gave explanations like “The premise is obviously false, and reasoning from a false premise gives a false conclusion,” or “Though the structure of the argument is logical, the resulting answer is obviously absurd, and therefore the answer is no.” Many Chinese subjects refused to adopt the counterfactual premise, saying things like “1 could never be larger than 2, the question makes no sense.” The Chinese response is in some sense correct, of course. The problem is logically unanswerable. The question is why the Americans agreed so readily to the absurd premise, whereas Chinese speakers balked at it.

There is speculation that the result involves the subjunctive mood in English (“If 1 were greater than two...would 1 be greater” etc.), which signals the listener that the premise is not to be taken at face value (and thus is removed from considerations of practical reality), whereas the Chinese must glean the intent of the utterance from contextual and pragmatic cues, and is thus encouraged to think more deeply about the realistic implications of the premise. I have my doubts about this interpretation, but whether or not there is any truth to it, such questions about the different psycholinguistic worlds of Chinese and English have a direct and significant bearing on a multitude of practical translation questions, and these are at least worthy of further study.

22

Bibliography

Bloom, Alfred (1981), The linguistic shaping of thought: a study in the impact of language on thinking in China and the West. Hillsdale: Lawrence Erlbaum Associates.

Feng Youlan (1922) “Why China has no science—an interpretation of the history and consequences of Chinese philosophy”, International Journal of Ethics, 32:3.

Graham, Angus C. (1978) Later Mohist logic, ethics and science. Hong Kong: Chinese University Press.

Graham, A. C. (1989) Disputers of the tao. La Salle: Open Court Press.

Granet, Marcel (1934) La pensée chinoise. Paris: ditions Albin Michel.�

Hansen, Chad (1983) Language and logic in ancient China. Ann Arbor: University of Michigan Press.

Jesperson, Otto (1924) The philosophy of grammar. Chicago: University of Chicago Press.

King, Brian (1989) The conceptual structure of emotional experience in Chinese. Ph.D. dissertation, Ohio State University, Dissertation Abstracts International, no. DA-9011203.

Kövecses, Zoltàn (1989) Emotion concepts. New York: Springer-Verlag.

Lakoff, George (1987) Women, fire, and dangerous things: what categories reveal about the mind. Chicago: University of Chicago Press.

Lakoff, George & Mark Johnson (1980) Metaphors we live by. Chicago: University of Chicago Press.

Lin Yutang (1935) My country and my people. New York: John Day.

Moser, David (1996) Abstract Thinking and Thought in Ancient Chinese and Early Greek. Ph.D. Thesis, University of Michigan, USA.

Quirk, R., S. Greenbaum, G. Leech, & J. Svartvik (1972) A comprehensive grammar of English. London: Longman.

Reale, Giovanni (1987) A history of ancient philosophy. I. From the Origins to Socrates. John R. Catan (ed. and trans.), Albany: State University of New York Press.

Reding, Jean-Paul (1986) “Greek and Chinese categories: a reexamination of the problem of linguistic relativism.” Philosophy East and West, 36:4, 549-374.

23

Rundle, Bede (1979) Grammar in philosophy. Oxford: Clarendon Press.

Van Iten, Richard J. (ed.) (1970) The problem of universals. New York: Appleton-Century-Crofts.

Wang Li (王力) (1947) Zhongguo yufa lilun 《中国语法理论》(“Theory of Chinese grammar”).

Shangwu yinshu guan 商务印书馆 (Commercial Press).

Whorf, Benjamin Lee (1956) Language, thought and reality. John B. Carroll (ed.), Cambridge, Mass: MIT Press.

Wright, Arthur (1953) The Chinese language and foreign ideas. In Arthur Wright (ed.), Studies in Chinese Thought, 286-303.

Zhao Yuanren (1968) A grammar of spoken Chinese. Berkeley: University of California Press.

appendix 1

24