Why There Will Always be a Need for Human Translators


2020-09-16 03:20 GALA


阅读模式 切换至中文

The practice of translation is essentially an interlingual form of communication. In order for that communication to be effective, every nuance and concept derived from the original text or speech must be analyzed and relayed in such a way that the meaning of the target text matches that of the source text as closely as possible. This is a skill that requires a high level of creativity, mixed with the capacity to think critically and in an integrative manner. This factor is perhaps what most strongly sets human beings apart from their mechanical counterparts (i.e. machines and robots). Indeed, machines do possess the capacity to learn and improve their productive output, but what they lack is part of what makes being human so unique: cognitive creativity. For this reason, as well as others which shall be outlined below, machines will never wholly replace human translators in the field of interlingual communication. Posing as a fundamental threat to the profession of translators, there are myriad reasons why machine translation just doesn’t make the cut. Perhaps it would be best to begin this discussion with an example of why automated machine translation will never outshine human translation services. Below is a short excerpt from a poem by Alexander Pushkin—the founding father of the modern Russian language—in the original Russian and followed by two translations. The first translation is the result of copying and pasting said poem into Google Translate with no further modifications and the second translation is the result of human translation performed by a professional certified translator. Original poem by Pushkin (1825): Александр Пушкин Если жизнь тебя обманет... Если жизнь тебя обманет, Не печалься, не сердись! В день уныния смирись: День веселья, верь, настанет. Translation performed by Google Translate: Alexander Pushkin If life deceives you… If life deceives you Do not be sad, do not be angry! On the day of gloom, humble yourself: Believe, the day of fun will come. Translation performed by Evgenia Sarkisyants: Alexander Pushkin Should ill-fortune cause you sadness... Should ill-fortune cause you sadness — Don't live dwelling on the pain! Don't be bitter, don't complain: It will come, your day of gladness. For starters, the machine translation failed to replicate both the rhyme scheme and the syllable pattern of the original poem, whereas the professional translator—EvgeniaSarkisyants—adhered to these elements and avoided word-for-word translation in order to give the English variant of this poem the same mood, elegance, and vocal character. Perhaps a machine could be designed purposely to imitate elements such as rhyme schemes when translating poetry but, given the intricate nature of such texts and the endless possibilities on the part of the translator, it is absurd to think that automated machine translation could compete with human translation services in terms of contextual creativity and meaning transfer. If more proof is needed as to why the creativity of human beings is indispensable to the art of translation, look up the several additional English translations of this exact poem and notice how each one differs from (and bears similarities to) the last. Stemming off this point, let us point out some of the areas of language comprehension that robots and machines fail to master. For one, they are constructed and programmed by humans and, therefore, do not have innate personalities and cultures with which to relate their worldview. They may have a certain form of memory, but a lack of life experience prevents them from mastering creativity. They are programmed for a purpose, unlike humans who must pursue aspects of life that lead them to find their own purpose. Because of this, machines do not have an understanding of cultural elements such as slang, shorthand speech, idiomatic expressions, and dialects. Furthermore, machines tend to work based on logic while simultaneously disregarding the context of the situation. Without considering the context, a translation could turn out to have a completely different meaning than the original text intended, especially in the case of word-for-word translations which are often how machines process text across languages. There is a popular practice now among translation agencies where, in order to maximize profit margin, they will perform a machine translation—for example, by copying and pasting an entire document into Google Translate—and then assign said “translated” document to a freelancer as a proofreading assignment. In essence, this allows the translation agency to offer a much lower rate per word in order to get the same result: a human translation service. If machine translation had the actual potential that many of its advocates claim it does, the subsequent proofreading service would not be necessary. In fact, the machine translations are usually so horrific that the “proofreader”—who is usually a translator by trade—essentially has to retranslate the original document from scratch while still being paid a fraction of what they should be earning with regard to their workload. Lastly, machine translation lacks a major part of what makes writing so unique: emotion. Machines cannot and never will manage to replicate tone as established by human emotion, which is involuntarily incorporated into every written text. This level of emotion cannot be attained any other way and also explains why each translator will produce a slightly different translation of the same source text. Take the preceding poem by Pushkin, for example. There exist several variations in English (as well as in other world languages) of this poem and each one is vastly different in terms of style and tone, but still manages to carry the message exactly as Pushkin intended when he composed it in 1825. Human translation services will never be wholly replaced by machines, though incorporating machine translation into the translation process doesn’t always hurt. It helps to expand your options and be able to logically pull material from different sources. Nevertheless, the level of creativity possessed by humans is unique and irreplaceable in the field of art, under which the craft of translation may be rightfully included.
翻译实践本质上是一种语言的交际方式。为了使交流有效,对源自原文或讲话的细微差别和概念都必须加以分析和传达,以使目标文本的含义尽可能与源文本的含义相吻合。这不仅是一项需要高水平创造力的技能,还需要综合思考和批判性思考的能力。这个因素也许是人类区别于机械同类(即机器和机器人)的最有力因素。的确,机器确实具有学习和提高生产力的能力,但机器所缺乏的正是人类如此独特的一部分:认知创造力。由于这个原因以及下文将概述的其他原因,机器永远不会在语际交流领域完全取代人类翻译。 作为翻译行业的巨大威胁,各种原因导致机器翻译无法进入这个行业。也许我们最好以一个示例说明为什么自动机器翻译永远不会胜过人工翻译服务。 下面是现代俄语之父亚历山大·普希金一首诗的节选,有俄文原文和两篇译文。第一次翻译是在没有进一步修改的情况下,将所述诗歌复制粘贴到谷歌翻译,第二次翻译是由专业认证译员进行人工翻译的结果。 普希金(1825)的原诗: 亚历山大·普希金 假如生活欺骗了你... 假如生活欺骗了你 不要伤心,不要心急! 忧郁的日子里需要平静 相信吧,快乐的的日子将会来临。 谷歌翻译的结果: 亚历山大·普希金 如果生活欺骗了你… 如果生活欺骗了你 不要伤心,不要生气! 忧郁的日子,平静自己: 相信,快乐的一天会到来的。 由Evgenia Sarkisyants翻译: 亚历山大·普希金 如果不幸给你带来悲伤 如果不幸给你带来悲伤- 不要活在痛苦中! 不要耿耿于怀,不要抱怨: 你快乐的日子终将到来。 首先,机器翻译无法复制原诗的押韵和音节模式,而专业译者Evgenia Sarkisyants坚持运用这些要素,并避免逐字翻译,以赋予这首诗的英语变体具有相同的情绪、优雅和声调。 也许可以将机器设计为在翻译诗歌时模仿诸如押韵方案之类的元素,但是鉴于此类文本的复杂性以及翻译者的无限可能性,我们认为自动机器翻译可以在语境创造性和意义传递方面与人类翻译服务竞争是荒谬的。如果还需要更多的证据来证明为什么人类的创造力对翻译艺术是不可或缺的,请查阅这首诗的其他几本英文译本,并注意每首诗与上一首诗有何不同(和相似之处)。 撇开这一点,我们可以指出一些机器人和机器在语言理解方面无法掌握的一些领域。首先,机器人由人类构建和编程,因此,没有与之关联的先天个性和文化。机器人可能有某种形式的记忆,但缺乏生活经验使他们无法掌握创造力。机器人是为了目的而设计,不像人类必须追求生活的各个方面才能找到自己的目的。因此,机器对俚语、速记、惯用表达和方言等文化元素并不了解。此外,机器倾向于基于逻辑工作,同时忽略情境的上下文。如果不考虑上下文,翻译可能会产生与原文本完全不同的意思,尤其是在逐字翻译的情况下,通常是机器跨语言处理文本的方式。 现在翻译公司中有一种流行的做法,为了利润最大化,他们会进行机器翻译--例如,通过复制并粘贴整个文档到谷歌翻译--然后将所述的“翻译”文档作为校对任务分配给自由职业者。从本质上讲,这使得翻译公司能为每个单词提供更低价,以获得相同的结果:人工翻译服务。如果机器翻译真的有许多拥护者所宣称的潜力,那么随后的校对服务就完全没必要。事实上,机器翻译非常可怕,以至于“校对员”--通常是贸易翻译)必须从头开始重新翻译原始文档,而他们所获得的报酬相对于工作量来说,只是校对员应得报酬的一小部分。 最后,机器翻译缺少使得写作如此独特的主要部分:情感。机器不可能也永远不会复制人类情感所建立的声调,而人类情感会不由自主地融入每一篇书面文本中。无法通过其他任何方式达到这种情感水平,这也解释了为什么每个译者会对相同的源文本产生略有不同的译文。以普希金的前一首诗为例。这首诗在英语(以及其他世界语言)中有几种变体,每一种在风格和音调方面大不相同,但仍然设法传达出普希金在1825年创作这首诗时的意图。 人工翻译服务永远不会完全被机器取代,尽管将机器翻译纳入翻译过程并不总是有害的。机器翻译有助于扩大选择,并能够从逻辑上从不同来源提取材料。 然而,人类所具有的创造力水平在艺术领域独一无二且不可替代,在这种创造性水平下,合理使用翻译技巧是理所应当的。