Although computer scientists have toiled for decades to produce machine translation comparable to that rendered by humans, they have yet to succeed. In critical moments when human translation or interpretation is simply not an option due to logistical constraints (e.g. rescue efforts during the recent earthquake in Haiti), machine translation can be a literal lifesaver; however, in cases where style, originality, or real-world context count for something, call upon a human translator to deliver the best results.
Previously, computer scientists attempted to “teach” the computer the linguistic rules of two languages in the hopes that the computer would piece together something intelligible in the target language. These days, the newest machine translation technology available through Google Translate takes a different approach. Using powerful search techniques and Google’s vast library of books, Google Translate turns in a reasonable performance based on matches found among thousands of documents produced for organizations such as the United Nations and the European Union by human translators.
While machine translation does serve a purpose now and again, the current technology’s formulaic approach depends upon the works that skilled human translators have already created. In the case of truly original works with zero precedent to be found online, computers fail to render the nuanced translations created by humans every time.
Read more about the machine translation debate in this article by The New York Times.
Read our articles:
Google Translate and the Struggle for Accurate Machine Translations
Google Strikes Deal to Translate European Patents
Machine Translation vs. Human Translation: Pay Less, Get Less
When Never to Use Google Translate