何莫邪 譯注/陳國華 盧培培
【譯者言】語言與邏輯之間有著十分密切的關(guān)系,然而似乎各種語言里都有一些不符合邏輯的表達(dá)式。中文里大家習(xí)以為常的一個(gè)例子就是“在他沒……之前我就……”。這一期英文原文里的All that glitters is not gold 也是如此。Google Translate譯成“閃光的不是金子”,DeepL譯成“金玉其外敗絮其中”,顯然都不對,可見這兩個(gè)機(jī)器翻譯軟件里都沒有配備英漢對應(yīng)的成語庫。邏輯學(xué)家常以這句英文成語為例,說英語的否定句有時(shí)如何不合邏輯;文法學(xué)家也常以這句話為例,說明英文的否定范圍(scope of negation)可以如何違反常規(guī)。事實(shí)上,這一不合邏輯的否定方式不過是偶然因素導(dǎo)致的一個(gè)例外。
據(jù)《牛津英語成語詞典》(Oxford Dictionary of English Proverbs, 3rd ed. p. 316)和英文短語網(wǎng)站The Phrase Finder (https://www.phrases.org.uk/index.html),這句話的現(xiàn)存最早說法是一句拉丁文成語,出現(xiàn)于13世紀(jì)后期,有[1]和[2]兩種形式:
[1] Non omne quod nitet aurum est.
= Not all that glitters is gold.
并非所有金光閃閃之物皆為黃金。
[2] Non teneas aurum totum quod splendet ut aurum.
= Do not take everything for gold, which shines like gold.
切勿以為所有金光閃閃之物都是黃金。
以上兩種說法邏輯上都毫無問題。
大約1300年,法文里出現(xiàn)了下面的說法:
[3] Que tout n’est pas or c’on voit luire.
= That one sees shining is not all gold.
金光閃閃者,并非皆黃金。
法文的這一說法在邏輯上也沒有問題。
14世紀(jì)末,英文里有了與拉丁文成語類似說法。與[1]相似的說法出現(xiàn)在喬叟約于1374—1385年寫的長詩《傳言之屋》(House of Fame)里:
[4] Hyt is not al golde that glareth.
= It is not all gold that glareth.
= That which glares is not all gold.
金光閃閃者,并非皆黃金。
與[2]相似的說法出現(xiàn)在英文版《圣經(jīng)》的第1位譯者約翰·威克里夫(John Wycliffe)約于1380年寫的一篇布道詞里:
[5] Men shulden not holde al gold tat shynet as gold.
= Men should not take all that shines like gold for gold.
人們不應(yīng)將所有金光閃閃之物當(dāng)成黃金。
這兩種說法在邏輯上也都沒有毛病。
喬叟之后又有至少3位英國文人有過類似說法。一位是詩僧約翰·利德蓋特(John Lydgate),他有下面[6]和[7]兩種說法,[6]見于《佃戶與鳥》(Chorle and Byrde),[7]見于《論人間事物之可變》(On the Mutability of Human Affairs),二者都創(chuàng)作于約1430年:
[6] All is not gold that shewyth goldishe hewe.
= All that shows golden hue is not gold
顯得金黃之物都不是黃金。
[7] All is not golde that outward shewith bright.
外表閃亮者都不是黃金。
另一位是翻譯家阿瑟·戈?duì)柖。ˋrthur Golding),此人1577年翻譯了著名宗教改革家約翰·加爾文(Jean Calvin)用法文寫的《論〈以弗所書〉布道詞》(Sermons on The Epistle to the Ephesians),里面說:
[8] But al is not gold that glistereth.
閃閃發(fā)光之物都不是黃金。
第3位是埃德蒙·斯賓塞(Edmund Spenser),其1590年出版的長詩《仙后》(The Faerie Queene)第2卷中有這樣兩行:
[9] Yet gold all is not, that doth golden seem.
= Yet that doth seem golden is not all gold.
然而外顯金黃者并非皆黃金。
以上4種說法,除[8]以外,邏輯上也沒有問題。
之后就是莎士比亞約于1596年創(chuàng)作的《威尼斯商人》里那句流傳至今的名言。第2幕第7場里,輪到摩洛哥王子選匣子,決定他能否娶鮑霞為妃。他選中的是金匣子,打開一看,里面的紙條上寫著:
[10] All that glisters is not gold;
Often have you heard that told.
Many a man his life hath sold
But my outside to behold.
所有的發(fā)光物并非黃金;
這句話您肯定熟知于心。
許多人葬送了寶貴性命,
僅因?yàn)槲彝獗砹料寡劬Α?/p>
莎士比亞的[10]明顯直接抄自戈?duì)柖〉腫8],只不過語序略有調(diào)整,格律也由前輕后重的抑揚(yáng)格變成了前重后輕的揚(yáng)抑格。這兩句話按照字面意思分析都經(jīng)不起推敲:既然“閃閃發(fā)光之物都不是黃金”或“所有的發(fā)光物并非黃金”,那么黃金是一種發(fā)光物,照理也不是黃金。不過莎劇讀者或觀眾從不這樣推敲,因?yàn)榇蠹抑?,詩人享有一種詩學(xué)特許(poetic license),即“詩人假設(shè)自己享有一種權(quán)利,允許他改變或顛倒標(biāo)準(zhǔn)句法,或背離常見修辭或發(fā)音,以使其所作之詩符合詩歌的格律或聲調(diào)要求?!?明白了這一點(diǎn),人們會(huì)把這行詩解讀為“并非所有發(fā)光物都是黃金”(Not all that glisters is gold)。莎士比亞的這一詩行抑揚(yáng)頓挫,末尾突出“并非黃金”,比符合邏輯的其他說法更受歡迎。結(jié)果是詩學(xué)碾壓邏輯。
Many exotic2 logical features have been claimed for3 the Chinese language throughout the ages. For example, it has been claimed that negation somehow does not have its full logical force in Chinese. In fact it turns out that such illogicalities as cumulative negation4 and such phrases as “all that glitters is not gold” are in fact quite alien to Chinese grammar, and if we are to compare logicality with respect to negation, then it turns out that Chinese is very considerably more logically transparent and rigid than ancient Greek (c, 1). It has been claimed that Chinese is very strange in lacking proper word classes, but Shakespearean English turns out to be very close indeed in its treatment of word classes to Classical Chinese (compare “but me no buts5”), and in any case there is a sound basis for distinguishing the functional properties of ancient Chinese words, not to speak of modern ones. They often even have morphological word class characteristics (c, 4). The stylistic differences in explicitness in Greek versus implicitness in ancient Chinese are real enough, but they are only matters of degree, and it must be pointed out that Greek can be elliptic in many places where Classical Chinese cannot (c, 5). I find that there are indeed many semantic/logical configurations that are perfectly possible but cannot be represented in Classical Chinese. But it turns out that these do not generally seem to be essential for the articulation of scientific thought. The syntax of Classical Chinese turns out to be rich enough to express the thought of Plato, though it is poor enough to necessitate considerable syntactic and logical revamping (c, 6).
The history of logical concepts in China has, in recent years, received much more attention than the grammatical and linguistic problems introduced6 so far, and this7 area remains controversial. A. C. Graham8 remarked that “it is perhaps inevitable that in the no-man’s-land on the common borders of linguistics, philosophy, and sinology, among those from whom one looks for stimulating new approaches9 (Rosemont10, Hansen11, Hall12) most generalisations about the Chinese language start from totally obsolete assumptions”. Graham goes on to give an example: “To speak of Chinese sentences as ‘strings of names’ is to revert to the grammatical knowledge of the ancient Chinese themselves (and not quite catching up with. Names and Objects).13” Graham argues: “But a claim that, for example, there are no sentences in English, offered without even an alternative account of the grammatical differences between the so-called ‘sentence’ and the nominalised clause, would be meaningless14; why is the claim supposed to be meaningful if the language is Chinese?”
古往今來,有許多帶有異國特質(zhì)的邏輯特征被說成是中文特有的。例如,有人聲稱,不知為何否定式在中文里沒有充分的邏輯效力。實(shí)際情況卻是,像多重否定式這種不合邏輯的現(xiàn)象和All that glitters is not gold(直譯是“所有發(fā)光物并非黃金”)之類的語句在中文文法里相當(dāng)異類;如果我們比較否定方面的邏輯性,那么事實(shí)證明,中文在邏輯上要比古希臘文透明和嚴(yán)謹(jǐn)?shù)枚啵ㄒ奵, 1)。有人聲稱,中文非常奇怪,缺乏正兒八經(jīng)的詞類,但事實(shí)證明,莎士比亞英語在對詞類的處理上與古典中文非常接近(比較but me no buts),而且不管怎樣,區(qū)分古代中文詞功能屬性的做法是有堅(jiān)實(shí)基礎(chǔ)的,更不用說現(xiàn)代中文詞了。中文詞常常甚至有形態(tài)學(xué)上的詞類特征(c, 4)。希臘文的顯性和古代中文的隱性在文體上確有差異,但這只是個(gè)程度問題,而且必須指出,希臘文在很多地方可以省略,而古典中文卻不可以(c, 5)。我發(fā)現(xiàn),確實(shí)有很多完全可能的語義/邏輯配置在古典中文中無法表示出來。但事實(shí)證明,一般來說,這些配置對科學(xué)思想的清晰表達(dá)似乎并非至關(guān)重要。經(jīng)證明,古典中文的句法既十分豐富,足以表達(dá)柏拉圖的思想;又比較貧乏,須經(jīng)過相當(dāng)可觀的句法和邏輯改造,才堪當(dāng)此任(c,6)。
近年來中國邏輯概念史受到的關(guān)注比迄今為止中文文法和語言問題受到的關(guān)注多得多,而且邏輯概念史領(lǐng)域仍存在爭議。葛瑞漢曾說:“在語言學(xué)、哲學(xué)和漢學(xué)共同邊界上的無人區(qū),當(dāng)你從羅思文、陳漢生、郝大維等人那里尋找具有啟發(fā)性的新思路時(shí),得到的有關(guān)中文的一般性論斷大都基于完全過時(shí)的假定”,這也是在所難免的。葛瑞漢接著舉了一個(gè)例子:“把中文句子說成 ‘一串串名稱’是倒退到古代中國人自己的文法認(rèn)知上(而且尚未完全跟上)?!备鹑饾h說:“舉例來說,有一種說法認(rèn)為,英文沒有哪個(gè)句子不可以拿來說明(甚至沒有其他辦法說明)所謂‘句子’及其名詞化之間有何文法差異。這種說法毫無意義。如果中文同樣是這種語言,這一說法又有什么意義呢?”