In an effort to make translation easier for the average person, computer scientists have created a translation process that involves automation through mathematics. Linguists protest the automation, saying that this translation technology has removed the very soul of language by completely disregarding the nuances, culture and feeling inherent in verbal and written communication.
Yet as translation technology becomes more and more accurate through the use of word pairings and algorithms, some have begun to wonder: is language an art form or a math equation? If machine translation can match the capabilities of a human, where does it end? There are already worries that we are over-reliant on this technology, with some skeptics even claiming that machines will (in time) overrun humanity as we know it. Others believe, however, that with the advancement of society will come an era of tech-enabled empowerment, one in which tech reestablishes itself as a re-humanizing force.
Depending on how we choose to view translation, it is easy to see the validity of both arguments.
Every language presents a unique structure, which can create confusion when that language is converted word for word by a machine. However, by continually refining the process, a researcher can “train” a computer to catch small nuances in language to achieve higher accuracy.
The process has a long way to go; some say that in order to create the perfect machine, you must first teach it to feel emotion. While it is unknown if this will ever be a reality, some researchers say that the potential for flawless translation through the use of mathematical processes may not be impossible.
Using Math to Solve Translation Problems
Within the realm of mobile translation technology, there are numerous gadgets that allow users a way to escape from becoming “lost in translation.” By pointing their phone’s camera at a menu or bathroom sign, avid travelers can now use apps such as WayGo and Word Lens to instantaneously translate words into another language. For those who don’t want to be bogged down by hand-held devices, Japanese telecom giant NTT Docomo is developing translation glasses for the 2020 Olympics in Tokyo, which will allow tourists to have instant translation right before their eyes.
The latest translation technology does not only grasp translation in textual forms but also aims at decoding verbal conversations. This feature is already available in apps such as Google Translate and Apple’s Heart to Heart, which translate spoken words in real time.
The most futuristic tool of all—often likened to the Universal Translator gadget in Star Trek—is Skype Translator. It is built on a machine learning platform that has the capability of software learning from training data examples. Combining speech recognition (SR) and machine translation (MT), Skype Translator aims to be a game changer for communications, revolutionizing the way we communicate across the globe.
Is Art Translatable?
Though automation has brought forth the advantage of quick and mobile translation, it is not wise to treat every translation as a math problem. For example, when you use these gadgets to read a book or translate a poem, more often than not, the meaning will be buried within a mountain of jumbled words and phrases. Sometimes the insertion of human emotion and knowledge into a translation is essential for it to make sense. If one chooses to think of translation as an art form, then translating must remain people-centered. This way, it does not risk becoming dehumanized and reduced to a jumble of numbers.
Similarly, within the medical field, it is best not to leave the translation process completely to machines, so as to eliminate any risk of incurring a life-or-death situation. Even for most critical product releases, machines are still not ready to handle the responsibility of making sure everything comes out correctly.
Some developers are already working towards finding a balance between human and machine translation by using the principles of social media. They combine the accuracy of the human touch with the speed of a machine by providing a platform for collaboration. Translators can contribute words and phrases into a database that others can quickly pull from in order to receive accurate terminology. Similarly, they can upload large files and collaboratively work on translation, while automatically storing previously translated material to be reused in future projects. Using basic translation material from Google Translate, translators can also build upon a translation and review it to make it sound more human.
Perhaps the future of translation does not involve a choice between art and mathematics, but instead involves a fusion of two important disciplines to create fast, accurate translations with all the emotion of the human spirit. Until then, you can trust your smartphone translator for getting directions to the airport, but leave that translation of Moby Dick up to the humans.