1.3 Literary, scientific, and technical translation
To better understand why certain types of texts are more suitable for the use of machine translation it is important to comprehend the differences between individual forms of translation.
“Literary translation is the translation of poetry, drama, and other literary works from one language to another” (Trujillo 4). Literary translation, moreover, has to convey the mood, feelings, and emotions expressed in the original text. The translators play here the role of artists, writers, and poets, they are more than mere linguists, they have to know the nuances of individual words perfectly, must have a sense for wording and creating a certain atmosphere. The translation is freer than the technical translation. The focus of literary translation is the message, legacy, and artistic value, the texts consist of complex sentences and are formed by an abundance of words.
Technical texts are often dry and tedious, they contain simpler sentences and focus on the meaning of words and the text as a whole.
“Technical translation is primarily distinguished from other forms of translation by terminology, although terminology usually only makes up about 5-10% of a text” (Newmark 151). Terminology is the crucial aspect of technical translation, the results of an inaccurate translation can often be an injury, a damaged device or appliance, or even worse: therefore the translation should convey the exact meaning of the source text. Also Mark Herman states in his article “Technical Translation Style: Clarity, Concision, Correctness” in the book Scientific and Technical Translation that: “Clarity, concision and correctness, the principal stylistic goals of technical writing, are simultaneously those of technical translation; an excellent technical translator is an excellent technical writer” (11). Technical translation involves the translation of texts from specialist disciplines in industry, commerce and sciences, it has its own format which often includes the use of the second person, present tense, and shorter simple sentences (instructions and manuals). Technical translators must have intimate knowledge of the vocabulary and even the subject matter of their field of specialization (Trujillo 4).
Newmark distinguishes the following varieties of technical styles: academic, professional, and popular (153).
Scientific translation and technical translation are quite similar. But the difference is that while technical translation should convey the translated information as clearly and precisely as possible, scientific translation contains discussions, explanations of theories, and used methods, and therefore is a combination of both technical and literary translation.
Simple sentences, approved terminology, a tendency to use certain parts of texts again and again, and the formal style make technical translation a perfect partner for machine translation.
1.4 Translation workflow
The change of the translation process was inevitable. The entire working process of a translator has changed from start to finish, and information technology plays an irreplaceable role already in work conditions negotiations, in accepting a translation job, and in its delivery. Contemporary technical translators receive original texts (in the source language) in a form of a text file or a set of files that can be subsequently processed in a text editor (however, this process is getting obsolete these days, too) or in a file format associated with a certain translation tool. They open the file in the corresponding program (text editor, translation tool, computer-aided tool). The advantage of technical translation is that it often is a part of a bigger project, in such a case the project contains original texts, appropriate term-bases with approved terminology and a translation memory (or translation memories). Translators then use the supplied term-bases and translation memories to help them translate a text, according to the expectations and needs of a customer. Translation tools provide translators also with validation tools, such as terminology checks, spelling and grammar checks, consistency checks, and often they can also verify that certain translation rules are followed throughout the whole file or an entire project. After all the checks have been done, the translator returns the translated file together with the output of required validation check back to the customer. As already mentioned, the workflow is done using information technologies.
Information and communication technologies have a significant influence on the translation itself and the workarounds used by translators. The amounts of texts requiring translation, predominantly in the form of files, are rising together with the development and rapid growth of the use of information technology by the general public.
1.4.1 Computer-aided translation
Computer-aided translation uses many features which can simplify the work of a translator significantly, it can be faster than mere manual translation and reduces costs per word. “In addition, the use of MAT (machine-aided translation) can result in improvements in quality, particularly in the use of consistent terminology within a text or for particular kind of client” (Trujillo 4). It is important to understand those tools and the differences between them.
The tools designed specifically for translators assisting translators in performing their daily routines are referred to as Computer-Aided Translation tools (CAT tools). They include translation memory systems, terminology management systems, electronic corpora and sometimes machine translation (Byrne 17).
A term-base is in fact a dictionary or wordbook which contains words (terms) approved by a customer or, in the scope of a bigger project, by a person responsible for the used terminology. Term-bases can be used everywhere, where the use of approved terminology is required. The pairs of source word (or a few words) and a target word are often accompanied by instructions regarding the use of these terms (when to use them, in which product, in which context), names of persons responsible, dates of edits and possible synonymy.
Term-bases can have many different forms, they can be well arranged in an Excel file (or another tabular worksheet or document) separately delivered to the translator, or they can be incorporated and attached to the text to be translated (a number of professional translation tools support this very convenient function – for example MemSource, WordFast, and SDL Trados Studio, to name just a few).
1.4.3 Translation memory
A translation memory consists of bilingual pairs of blocks of texts, individual sentences and sometimes even paragraphs. Translation memory simply reuses already existing translations. Trujillo states that:
One feature of certain types of text such as technical manuals and various kinds of report is the degree to which they resemble each other. The manual for one version of a word processor will be similar to the manual for the next version. Even the manual of one software application from one company will have much in common with that for other applications from the same company. They will include phrases such as “To save a file open the File menu” or some similar instruction. (59)
Translation memories make use of this property of technical texts, they save the source text together with the translated text (again the source text together with the text in the target language in the language pair), and whenever the new source text contains a similar sentence they offer the translator a suggestion (called a match). Depending on how high the level of resemblance between those two sentences is, translators can accept the suggestion, accept and/or edit the suggestion to suit better the translated text (to reflect the differences with the text being translated) or can reject the suggestion and translate the text themselves.
Translation memory tools allow translators to set the required ratio of such a resemblance, to narrow the search for an appropriate similar translation. The commonly used match value to use the translation memory efficiently is about 70 percent.
The above described features help to maintain the translations consistent within a file or a project, and further more across many files and many projects. Translation consistency is one of the quality factors often required by the customers.