Who would have imagined, only 20 years ago, that computers would come to play such an important part in our daily lives? Today, it would be impossible for many of us to imagine ourselves performing most of our daily tasks without the priceless help of our digital devices. We, translators, have also found a powerful ally in computers to perform our translation tasks.
Computer- asssisted translation
In the 20th century, professional translators were overwhelmed by a huge volume of printed documents to translate with the help of only a few dictionaries, experience and invaluable patience. However, the contribution of computers and assisted translation has made the translation task considerably easier. Computer-assisted translation (CAT) refers to software intended to simplify translation projects.
On the one hand, CAT tools divide the text into segments or phrases to be worked on by the translator. On the left hand side we find the original text, divided into segments, and on the right, the translated sentences are entered. By displaying the text in two columns, the translation task is made considerably easier and even more so in the case of proofreading completed translations.
As the translation units enter the system, we “feed” the program’s translation memory. This memory is of great use when having to translate the same term or phrase several times in the same project. It is especially useful for technical translations, where special attention should be paid to terminology. But it is also helpful when updating old translations, since it assists in maintaining consistency between different versions.
Translation memories, however, have certain limitations. Many terms, depending on the context, may have distinct meanings. For example, blade means cuchilla in Spanish, but in English, shoulder blade means omóplato, and wiper blade means limpiaparabrisas. In the end, it is the translator who must interpret the original text and decide which terms suit each specific situation.
On the other hand, there is automatic translation. Although it was seemingly invented quite recently, in fact, the first automatic translation was performed in the 1950s. Its original purpose was to generate text in language B from a source text in language A, without the need for the intervention of a human translator. To do so, it may base its translations on previously translated texts, performed by human translators, searching for the most frequently used translation using statistical methods. Or it may generate the translation using criteria combining linguistic and grammatical rules with common dictionary content.
These are still far from providing the quality of a professional translation; however, they are becoming increasingly sophisticated. Nonetheless, they can only aim to serve as an additional tool for the translator, who must ultimately interpret the meaning of the message as a whole and apply the most suitable translation.
Therefore, assisted and automatic translation can be important tools in the translator’s daily routine. If used correctly, they can help create high-quality translations in limited periods of time. But it takes more than a hammer to become a carpenter. Automatic translation engines can be useful in specific cases, when attempting to grasp the general meaning of an isolated sentence, but they cannot compete with a professional translator. If the translator is removed from the equation, the text will lack a natural flow, an effective message, and, ultimately, the correct transmission of the content. Once again, Google Translate is just not enough.
About Antonio Leal Fernández
Graduate degree in Translation and Interpretation from Universidade de Vigo (2013). Translator and proofreader in the German and English to Spanish combinations.