文/吉迪恩·路易斯-克勞斯 譯/鄧志輝

翻譯是藝術(shù)，還是數(shù)學(xué)題？

文/吉迪恩·路易斯-克勞斯 譯/鄧志輝

Universal translation1科幻作品往往默認(rèn)存在某種universal translator（國(guó)內(nèi)常譯為“通用翻譯器”或“萬能翻譯器”）幫助人類與外星人順暢交流。universal translation乃由universal translator轉(zhuǎn)化而來。此處譯文適當(dāng)進(jìn)行了顯明化處理。has long been motivated by a utopian ambition,a dream that harks back to Genesis, of a common tongue that perfectly maps thought to world2人類思維反映客觀現(xiàn)實(shí)，語言則是傳達(dá)思想的工具。若存在人類共同語，則能克服不同語言所具有的民族性，使全人類對(duì)于客觀世界的認(rèn)識(shí)（即思維）借助該共同語言得到毫無分歧的呈現(xiàn)。.

[2] Translation is possible, but we are bedeviled3bedevil使痛苦；虐待。by conflict. This fallen state of affairs is often attributed to the translators, who must not be doing a properly faithful job. The most succinct4succinct簡(jiǎn)潔的；簡(jiǎn)明的。expression of this suspicion55 suspicion懷疑。is “traduttore, traditore,” a common Italian saying that’s really an argument masked as a proverb. It means, literally, “translator,traitor,” but even though that is semantically on target, it doesn’t match the syllabic66 syllabic音節(jié)的；分音節(jié)的。harmoniousness of the origi-nal, and thus proves the impossibility it asserts.

[3] For now, efforts in the discipline of machine translation are mostly concerned with the dutiful assembly of “cargo trucks” to ferry information across linguistic borders. The hope is that machines might efficiently and cheaply perform the labor of rendering sentences whose informational content is paramount7paramount至為重要的。: “This metal is hot,”“My mother is in that collapsed house,”“Stay away from that snake.” Beyond its use in Google Translate, machine translation has been most successfully and widely implemented in the propagation8propagation傳播。of continentspanning weather reports or the reproduction in 27 languages of user manuals for appliances. As one researcher told me,“We’re great if you’re Estonian and your toaster is broken.”

[4] Warren Weaver9（1894—1978）美國(guó)數(shù)學(xué)家，被譽(yù)為機(jī)譯鼻祖，早于1947年就提出機(jī)譯設(shè)想，1949年發(fā)表一份以“翻譯”為題的備忘錄，正式提出并詳細(xì)闡述機(jī)器翻譯問題。, a founder of the discipline, conceded: “No reasonable person thinks that a machine translation can ever achieve elegance and style.Pushkin10（1799—1837），俄國(guó)著名詩人。此處用以指代注重語言使用之elegance and style的詩人群體。作者的意思是，機(jī)器翻譯無法實(shí)現(xiàn)優(yōu)雅和風(fēng)格，所以詩人們不必?fù)?dān)憂會(huì)被機(jī)器取代。從翻譯角度來看，直譯為“普希金們”略嫌隱晦，“以普希金為代表的詩人們”則過于繁瑣，所以只簡(jiǎn)單譯為“詩人們”。need not shudder.” The whole enterprise introduces itself in such tones of lab-coat11lab coat實(shí)驗(yàn)室的工作服，這里引申為“簡(jiǎn)樸實(shí)用的，不加任何修飾的”。modesty.

[5] In 1960, one of the earliest researchers in the fi eld, the philosopher and mathematician Yehoshua Bar-Hillel12（1915—1975），以色列哲學(xué)家、數(shù)學(xué)家、語言學(xué)家，尤以在機(jī)器翻譯和形式語言學(xué)中的成就聞名于世。,wrote that no machine translation would ever pass muster13pass muster及格；合乎要求。without human “postediting.” He called attention to sentences like “The pen is in the box” and “The box is in the pen14pen還有“圍欄，關(guān)押”等義。.” For a translation machine to be successful in such a situation of semantic ambiguity15ambiguity歧義；一語多義。, it would need at hand not only a dictionary but also a“universal encyclopedia.” The brightest future for machine translation, he suggested, would rely on coordinated efforts between plodding16plodding老牛拖破車似的；做事慎重而呆板的。machines and well-trained humans. The scienti fi c community largely came to accept this view: Machine translation required the help of trained linguists, who would derive increasingly abstract grammatical rules to distill natural languages down to the sets of formal symbols that machines could manipulate.

[6] This paradigm17paradigm范例；典范。prevailed18prevail普遍存在；盛行。until 1988, year zero for modern machine translation, when a team of IBM’s speech-recognition researchers presented a new approach. What these computer scientists proposed was that Warren Weaver’s insight19指韋弗1949年在“翻譯”備忘錄中提出的觀點(diǎn)，認(rèn)為翻譯過程類似于密碼解讀過程，故可從這一角度來進(jìn)行機(jī)器翻譯研究。about cryptography20cryptography密碼學(xué)。was essentially correct but that the computers of the time weren’t nearly powerful enough to do the job.“Our approach,” they wrote, “eschews21eschew避開；戒絕。the use of an intermediate222 intermediate中間的。mechanism (language) that would encode the‘meaning’ of the source text.” All you had to do was load reams23ream〈非正式〉大量的文字（或?qū)懽鳎f parallel text24parallel text平行語料，指使用不同語言撰寫、相互間具有“翻譯關(guān)系”的文本。through a machine and compute the statistical likelihood of matches across languages. If you train a computer on enough material, it will come to understand that 99.9 percent of the time,“the butterfly” in an English text corresponds to “l(fā)e papillon” in a parallel French one. One researcher25指弗里德里克·賈里尼克（Frederek Jelinek，1932—2010），世界著名的語音識(shí)別和自然語言處理的專家，他在 IBM 實(shí)驗(yàn)室工作期間，提出了基于統(tǒng)計(jì)的語音識(shí)別的框架。本句所指原話有不同版本，其一是“Every time I fire a linguist, the performance of the speech recognizer goes up.”。quipped26quip講俏皮話。that his system performed incrementally better each time he fi red a linguist. Human collaborators, preoccupied with shades27shade差別；不同。of “meaning,” could henceforth be edited out entirely.

[7] This statistical strategy, which supports Google Translate and Skype Translator and any other contemporary system, has undergone nearly three decades of steady refinement28refinement（精細(xì)的）改進(jìn)，改善。. The problems of semantic ambiguity have been lessened by paying pretty much no attention whatsoever to semantics.The English word “bank,” to use one frequent example, can mean either “financial institution” or “side of a river,”but these are two distinct words in French. When should it be translated as“banque,” when as “rive”? A probabilistic299 probabilistic基于概率的；或然的。model will have the computer examine a few of the other words nearby.If your sentence elsewhere contains the words “money” or “robbery,” the proper translation is probably “banque.”(This doesn’t work in every instance,of course. A machine might still have a hard time with the relatively simple sentence “A Parisian has to have a lot of money to live on the Left Bank.”

[8] Many computational linguists continue to claim that, after all, they are interested only in “the gist300 gist要點(diǎn)；大意?！?and that their duty is to fi nd inexpensive and fast ways of trucking the gist across languages.But they have effectively arrogated31arrogate僭稱；霸占。to themselves the power to draw a bright line where “the gist” ends and “style” begins. Human translators think it’s not so simple. All texts have some purpose in mind, and what a good human translator does is pay attention to how the means serve the end, how the “style” exists in relationship to “the gist.” The oddity is that belief in the existence of an isolated“gist” often obscures the interests at the heart of translation.

[9] What mostly annoys human translators isn’t the arrogance of machines but their appropriation of the work of forgotten or anonymous humans. Machine translation necessarily supervenes on previous human effort; otherwise there wouldn’t be the parallel corpora32corpora語料庫，指為特定的應(yīng)用目標(biāo)而專門收集加工、具有一定結(jié)構(gòu)、可被計(jì)算機(jī)程序檢索的原始語料集合。that the machines need to do their work. I mentioned to an Israeli graduate student that I had been reading the Wikipedia page of Yehoshua Bar-Hillel and had found out that his granddaughter, Gili,is a minor celebrity in Israel as the translator of the “Harry Potter” books.He hadn’t heard of her and didn’t seem interested in the process by which a

無障礙型通用翻譯的靈感來源是一個(gè)令人聯(lián)想到《圣經(jīng)·創(chuàng)世記》的烏托邦式夢(mèng)想，即借助某種共同語言，架構(gòu)起人類思維與客觀世界間的完美橋梁。

[2]翻譯誠然可為，但分歧依然令人苦不堪言。這種不盡人意的現(xiàn)狀常被歸咎于譯者——人們想當(dāng)然地認(rèn)為這必然都因他們未能忠實(shí)盡責(zé)所致。對(duì)此類不信任心態(tài)最言簡(jiǎn)意賅的表達(dá)，莫過于一句意大利諺語“traduttore,traditore”。這話貌似一句格言，其實(shí)只是一個(gè)觀點(diǎn)，其英文直譯是“translator, traitor”（“譯者，叛徒”）。不過英譯文雖然語義無誤，音節(jié)上卻無法再現(xiàn)意大利原文的對(duì)稱和諧之美，因此倒恰好佐證了原句所宣稱的翻譯不可為之論。

[3]就目前來看，機(jī)器翻譯領(lǐng)域主要是兢兢業(yè)業(yè)地以“貨車”組裝方式進(jìn)行語際間信息傳送，以期在翻譯那些信息成分至上的句子時(shí)，機(jī)器可以更廉價(jià)而高效，例如：“這塊金屬很熱”“我母親還在那棟倒塌的房子里”“離那條蛇遠(yuǎn)點(diǎn)”等。除了谷歌翻譯軟件以外，機(jī)器翻譯應(yīng)用最成功、服務(wù)范圍最廣的領(lǐng)域，當(dāng)屬洲內(nèi)天氣預(yù)報(bào)的傳播系統(tǒng)，或是家用電器使用說明書的27種語言翻譯系統(tǒng)。如一位研究者所說，“如果你是愛沙尼亞人，而且面包機(jī)壞了，這時(shí)你會(huì)發(fā)現(xiàn)我們的服務(wù)相當(dāng)不錯(cuò)”。

[4]機(jī)器翻譯的鼻祖沃倫·韋弗曾經(jīng)坦承：“但凡有點(diǎn)理智的人都清楚，機(jī)器翻譯永遠(yuǎn)無法實(shí)現(xiàn)語言的優(yōu)雅美感或風(fēng)格的藝術(shù)再現(xiàn)，因此詩人們不必恐慌。”——整個(gè)機(jī)譯行業(yè)都以這種樸實(shí)的語氣自謙。

[5] 1960年，該領(lǐng)域的先驅(qū)之一，同時(shí)也是哲學(xué)家和數(shù)學(xué)家的耶霍舒亞·巴爾-希勒爾發(fā)文宣稱：除非有人工譯者進(jìn)行后期編輯加工，否則機(jī)器翻譯的質(zhì)量絕對(duì)無法過關(guān)。他提醒人們注意一些歧義句，如：“pen（鋼筆）在盒子里”和“盒子在pen（籠子）里”，機(jī)器在處理此類語義歧義時(shí)，僅靠字典尚不足夠，還必須借助某種“萬能百科全書”才行。因此，在他看來，機(jī)器翻譯要實(shí)現(xiàn)最佳前景，必須依靠呆板的機(jī)器與訓(xùn)練有素的人工緊密合作，方有可為?？茖W(xué)界很大程度上逐漸接受了這種觀點(diǎn)：機(jī)器翻譯必須依靠專業(yè)語言學(xué)家的幫助，后者通過導(dǎo)出日益抽象的語法規(guī)則，將自然語言簡(jiǎn)化歸納為一套套正式符號(hào)，供機(jī)器識(shí)別使用。

[6]這一思維范式持續(xù)至1988年。在這個(gè)現(xiàn)代機(jī)器翻譯技術(shù)的元年，來自IBM公司的一個(gè)語言識(shí)別研究團(tuán)隊(duì)展示了一種全新方法。這些計(jì)算機(jī)科學(xué)家提出，沃倫·韋弗當(dāng)年從密碼學(xué)視角將翻譯視為“解碼”過程的看法本質(zhì)上沒錯(cuò)，但受當(dāng)時(shí)計(jì)算機(jī)技術(shù)的限制，從該思路出發(fā)無法實(shí)現(xiàn)機(jī)器翻譯。他們寫道：“我們的方法則避開了這一常規(guī)思路，不再依賴中介機(jī)制（語言）來對(duì)源文本的意義進(jìn)行編碼?！币龅闹皇峭ㄟ^機(jī)器載入大量平行語料，然后對(duì)語言間的對(duì)應(yīng)情況進(jìn)行統(tǒng)計(jì)分析即可。只要給計(jì)算機(jī)的訓(xùn)練語料庫夠大，它就會(huì)逐漸學(xué)習(xí)到,英語文本中的the butterfly在99.9%的情況下都與法語平行文本中的le papillon相對(duì)應(yīng)。有位研究者曾打趣說，每開除一名語言學(xué)家，他的系統(tǒng)運(yùn)行效率就會(huì)大幅上升。糾結(jié)盤桓于各種細(xì)枝末節(jié)“意義”間的人類伙伴似乎從此可以徹底退場(chǎng)了。

[7]這種統(tǒng)計(jì)法是如今谷歌翻譯器、Skype翻譯器及其他各類當(dāng)代機(jī)譯系統(tǒng)的技術(shù)基礎(chǔ)，且自其問世30年以來，一直處于穩(wěn)定改良中。語義歧義問題已有所減少，而解決方案居然是：徹底繞開語義。舉一個(gè)大家熟知的例子，英語中bank一詞同時(shí)有“金融機(jī)構(gòu)”與“河岸”之義，而在法語中這兩個(gè)意思分別對(duì)應(yīng)兩個(gè)完全不同的詞。究竟何時(shí)該將bank譯成法語的banque（銀行）、何時(shí)該譯成rive（河岸）呢？機(jī)譯概率模型會(huì)指引計(jì)算機(jī)查看附近的幾個(gè)單詞，如果句中其他地方出現(xiàn)“錢”或“搶劫”等類詞語，則可判斷恰當(dāng)譯法很可能是banque。（這當(dāng)然不適用于所有情形?！鞍屠枞说糜幸淮蠊P錢才能住在左岸”，這個(gè)句子本身雖然并不復(fù)雜，計(jì)算機(jī)在翻譯時(shí)恐怕卻得頗費(fèi)周章。）

[8]很多計(jì)算語言學(xué)家反復(fù)聲明，稱自己感興趣的僅限于信息“要義”，職責(zé)是尋找低廉而快捷的手段實(shí)現(xiàn)語際間的信息要義輸送。殊不知，他們?cè)诖诉^程中僭取了一個(gè)權(quán)利，即由他們來界線分明地判定何時(shí)“要義出”、何時(shí)“風(fēng)格入”。人類譯者則認(rèn)為事情并非如此簡(jiǎn)單。所有文本都自帶意圖，而一名優(yōu)秀人類譯者所做的，恰恰是關(guān)注手段如何為意圖服務(wù)、“風(fēng)格”如何與“要義”互存。悖論就在于：相信某種“要義”能獨(dú)立自存，這一看法反而遮蔽了翻譯本身的核心要義。

[9]最讓人類譯者生惱的還不是機(jī)器的這種倨傲，而是它們對(duì)無名或匿名人士勞動(dòng)成果的任意取用。機(jī)器翻譯無可避免要仰賴此前的人類勞動(dòng)成果，這也是為什么要建立平行語料庫的原因，否則機(jī)器無法工作。我曾與一位以色列研究生交談，說起我一直在讀維基百科上有關(guān)耶霍舒亞·巴爾-希勒爾的介紹，了解到他的孫女吉麗是《哈利·波特》系列小說的譯者，在以色列小有名氣。這名學(xué)生對(duì)她一無所知，談話過程中也沒表現(xiàn)出對(duì)出版商花錢引進(jìn)魔法類兒童讀物過程的興趣。但是，如果沒有吉麗·巴爾-希勒爾這樣的譯者一字一句精雕細(xì)琢、為一個(gè)用途非凡的平行語料庫做出4000余頁的語料貢獻(xiàn)，就不會(huì)出現(xiàn)支持希伯來語-英語互

〔〕

〔〕publisher paid to import books about magic for children. But we would have no such tools as Google Translate for the Hebrew-English language pair if Bar-Hillel had not hand-translated,with care, more than 4,000 pages of an extremely useful parallel corpus. In a sense, their machines aren’t actually translating; they’re just speeding along tracks set down by others. This is the original sin of machine translation: The field would be nowhere33 be nowhere 沒有取勝的機(jī)會(huì)；一無所得。 without the human translators they seek, however modestly, to supersede34 supersede取代，代替。. ■譯的谷歌翻譯應(yīng)用程序。在某種意義上，機(jī)器從未進(jìn)行真正的翻譯，而只是沿著他人鋪設(shè)好的軌道飛馳，這正是機(jī)器翻譯的原罪所在：若非借人類譯者之功，機(jī)器翻譯行業(yè)斷不能有任何建樹；然而它們盡管姿態(tài)極盡謙卑，卻一心圖謀要將人類譯者取而代之。 □

（譯者曾獲第五屆“《英語世界》杯”翻譯大賽一等獎(jiǎng)。譯者單位：中山大學(xué)外國(guó)語學(xué)院）

Is Translation an Art or a Math Problem?

ByGideon Lewis-Kraus