The Art of Translation
Man Versus Machine
Translation is the art of rendering one language into another for the purpose of conveying an equivalent meaning in the second language. Depending on the nature of the subject matter, it may often be best to translate as literally as possible. On the other hand, it is occasionally more appropriate to deviate from the original modes of expression and create a new work that is to some degree inspired by the original, rather than constrained by it.
Since translation is a dynamic intellectual process, it demands mental focus and a great expenditure of energy on the part of the translator. Translators must also possess unique skills and qualifications that are acquired only after years of intensive study. Thus, the translator is a rare breed of expert whose learning must extend from the broadest brush strokes of a given subject, down to the minutest of its details. Far from simply glossing isolated words and phrases, the translator must interpret meanings and may need to conduct research into specialized subject matter. In today's world, even more is expected from the translator than ever before. Translators of today must absorb new and unfamiliar information readily, adapt to new realities in a world that evolves rapidly, and master the latest technologies in order to keep their skills relevant.
The Cambridge Translation Company understands what is required to produce a quality translation. That is why we place so much emphasis on our translators. All of our translators are language experts who have passed a rigorous battery of tests and have demonstrated a mastery over the languages and the subject matters in which they work. Many of them possess advanced degrees from some of the world's most elite universities. When you rely on the Cambridge Translation Company for your translation needs, you have access to some of the finest translators in the world.
Human Translation
(Reading + Research + Reformulation + Writing + Revision + Results)
To this day, the only way to guarantee an accurate and reliable translation is to find professionals who know how to read a text critically and conduct any necessary research before they reformulate meanings and convey the equivalent sense in the target language. To achieve this goal, the translator must be a skilled writer with an excellent command of grammar and style and an ability to make difficult decisions about how best to reflect the style and diction of the source document, without compromising clarity or precision in the translation. Professional translators are team-players who work with others to revise their translations until optimal results have been acheived for the client.
Despite remarkable advances in computing technology over the years, human beings still have a clear competitive advantage over machines when it comes to performing any creative intellectual task. As a result, professional translation is still an absolute necessity whenever there is a need for an accurate rendering of a text into another language. This is because computers, to put it simply, are not creative beings. They can do nothing other than what they have been programmed to do by their human masters. A computer cannot reason, cannot research, cannot discern, cannot distinguish between irony or some other rhetorical figure or flourish. To a computer, all languages are dead. The computer does not create meaning anew, as we humans do effortlessly on a daily basis.
Would you trust a computer to give you legal advice and interpret the law on your behalf? Would you trust a computer to perform surgery or diagnose conditions and prescribe you the appropriate drugs? Neither should you trust a computer to translate complicated text. Perhaps it is better to think of the machine translator as an over-the-counter solution to an everyday problem, which loses its usefulness as soon as the problem grows in size and complexity. While many web users are aware of the truly remarkable accuracy of certain online translation modules when it comes to finding the right word or phrase in another language, few are aware of the terrible inaccuracy of such tools when the textual input involves a moderate degree of complexity, or when the user selects a language pairing that is less common than Spanish into English.
The reason for the machines' apparent skill at the art of translation lies rather in their computational aptitude than in any other advantage they have over their human counterparts. The most accurate computer modules of today employ a method known as statistical machine translation. This method makes the most of the machines' computational speed by scanning millions of documents within mere fractions of a second, in order to provide the human user with a selection of what the module determines to be the statistically most likely candidates for the meanings of a given word or phrase. The computer makes its determination on the basis of the frequency with which a word or phrase in a given language is translated by another word or phrase in another language.
While the computer's method has some obvious advantages, it also involves some significant drawbacks. As already mentioned, the accuracy of the module is severely limited if a user wishes to select relatively uncommon language pairs. In addition, the modules do not yet control for the nature of the texts that the user uploads into them. Words and phrases that have a particular meaning in one set of contexts might have totally different meanings in another set of contexts. As a result, the module may often introduce significant errors into the translation. Only a highly trained and skilled professional will be able to know the difference. The intelligibility of the computer module's output will also drop off dramatically as the level of complexity and technicality increases, thus rendering much of this output useless for practical purposes. At last, there are many legal questions that come into play when a human user wishes to rely upon a module that is designed to process vast amounts of data that may actually be protected by copyright. The work products of countless individuals are first scanned by the module, but seldom with any permission being granted for the use of the material. Selections are then copied and mapped to "strings" found in other texts, which may likewise be copyrighted. It has yet to be examined whether such methods go beyond mere meta-analysis of data, or whether they in fact constitute a wholesale infringement of copyrighted materials.
