Is Language Study No Longer Necessary if Machine Translations are Perfect?
Recently, "Pocketalk," a translation device with the ability to accurately interpret dozens of languages in real-time, has been all the rage. The Japan-developed device is certainly revolutionary in its accuracy. Just by talking into it, the user can get a sentence interpreted in seconds into the target language. With its customized SIM card, it can be used anywhere in the world without the need to adjust the setting every time one enters a new country. Such a device really helps people to imagine a future where the language barrier is no longer a barrier, and technology will allow people speaking different languages to communicate naturally in real time.
Indeed, it is a project that is becoming more and more possible with the advent of Big Data and Artificial Intelligence. By studying human-made translations in the millions, machines can detect concrete patterns that make future translations more accurate and speedy. Eventually, it is hoped that accumulating data will render human intervention for accurate translation and interpretation completely unnecessary. Machine translations may become just as good, if not better than, human translations, with the additional benefit of greater speed, cheaper cost, and more widespread availability of professional-level output.
It begs the question of whether study a foreign language is still needed when the machines become that good with translating and interpreting. After all, if machines allow for seamless communication with people speaking different languages, then what is the point of investing in years of effort to learn a new language? Such logic is very much justified for those who think of language as a mere tool for communication, but if one thinks of language as a gateway for understanding culture, then machine translation cannot replace good-old learning by talking with native speakers of the language.
To put simply, while machine translation may become better and better at interpreting what is said in another language, it remains unable and not designed to interpret why the native speaker of another language something in a particular way. Language is not simply a collection of vocabulary and grammar rules, it reflects the very mentality of their speakers. Certain word usage and expressions reflect cultural values that are widespread in the society to which the language is native. While machines can do their best in putting those expressions in a different language literally, they are unable to denote the cultural significance behind them.
If anything, boiling languages down to the literal meaning of words and sentences risks miscommunication even if both sides are perfectly in sync in terms of understanding what are said. Communication is not done simply by the literal meanings of words; quite often, what is not said is more important to the main point of the conversation than what is said, and the nuances of the words not directly related to the literal meaning, such as sarcasm and euphemisms, are culturally unique. Machine translation does not reflect those unsaid words and hidden intended meanings.
Given the alternative ways people of different cultures communicate besides using direct, literal expressions, overreliance on machine translations may actually be rather dangerous. People might become so sure of the objective accuracy of machine translations that they refuse to consider the possibility of subjective idiosyncrasies that guide the conversation. As people become surer of the literal meanings, they find themselves scratching their heads over what they see as the not-so-meaningful situations when certain phrases are uttered to them.
Faced with the confusion, it is necessary for people to head back to learning languages for themselves, rather than just relying on machines. When people learn languages, especially in immersive environments where they constantly interact with native speakers and their culture, they become aware of cultural differences between speakers of different languages that go beyond the language itself. And when they learn the language, they come to understand not just the language, but the mentalities of the speakers that dictate how the language is used and not used.
Thought this way, those who argue that perfect machine translations rid the very need for humans to learn new languages may be saying so to avoid the uncomfortable reality of having to think in a new way that comes with learning a new language. Their erroneous belief that the communication method used is the same across all languages and cultures is bound to set back the prospect of meaningful and effective international communication, despite advances in machine translation. In a world where more and more communication has become global in scale, that is just plain dangerous.
Indeed, it is a project that is becoming more and more possible with the advent of Big Data and Artificial Intelligence. By studying human-made translations in the millions, machines can detect concrete patterns that make future translations more accurate and speedy. Eventually, it is hoped that accumulating data will render human intervention for accurate translation and interpretation completely unnecessary. Machine translations may become just as good, if not better than, human translations, with the additional benefit of greater speed, cheaper cost, and more widespread availability of professional-level output.
It begs the question of whether study a foreign language is still needed when the machines become that good with translating and interpreting. After all, if machines allow for seamless communication with people speaking different languages, then what is the point of investing in years of effort to learn a new language? Such logic is very much justified for those who think of language as a mere tool for communication, but if one thinks of language as a gateway for understanding culture, then machine translation cannot replace good-old learning by talking with native speakers of the language.
To put simply, while machine translation may become better and better at interpreting what is said in another language, it remains unable and not designed to interpret why the native speaker of another language something in a particular way. Language is not simply a collection of vocabulary and grammar rules, it reflects the very mentality of their speakers. Certain word usage and expressions reflect cultural values that are widespread in the society to which the language is native. While machines can do their best in putting those expressions in a different language literally, they are unable to denote the cultural significance behind them.
If anything, boiling languages down to the literal meaning of words and sentences risks miscommunication even if both sides are perfectly in sync in terms of understanding what are said. Communication is not done simply by the literal meanings of words; quite often, what is not said is more important to the main point of the conversation than what is said, and the nuances of the words not directly related to the literal meaning, such as sarcasm and euphemisms, are culturally unique. Machine translation does not reflect those unsaid words and hidden intended meanings.
Given the alternative ways people of different cultures communicate besides using direct, literal expressions, overreliance on machine translations may actually be rather dangerous. People might become so sure of the objective accuracy of machine translations that they refuse to consider the possibility of subjective idiosyncrasies that guide the conversation. As people become surer of the literal meanings, they find themselves scratching their heads over what they see as the not-so-meaningful situations when certain phrases are uttered to them.
Faced with the confusion, it is necessary for people to head back to learning languages for themselves, rather than just relying on machines. When people learn languages, especially in immersive environments where they constantly interact with native speakers and their culture, they become aware of cultural differences between speakers of different languages that go beyond the language itself. And when they learn the language, they come to understand not just the language, but the mentalities of the speakers that dictate how the language is used and not used.
Thought this way, those who argue that perfect machine translations rid the very need for humans to learn new languages may be saying so to avoid the uncomfortable reality of having to think in a new way that comes with learning a new language. Their erroneous belief that the communication method used is the same across all languages and cultures is bound to set back the prospect of meaningful and effective international communication, despite advances in machine translation. In a world where more and more communication has become global in scale, that is just plain dangerous.
Comments
Post a Comment