Natural Language Processing’s Benefits in Learning Language and Its Limitation
Language barrier has proven to be a huge obstacle in communication since the dawn of mankind. The advent of machine translator however, helps to combat this aspect and machine translator are becoming increasingly proficient at their job. Ultimately, we are reaching that state where translator would become so efficient that there will no longer be a language barrier. But there is a catch to it.
In this writing, I will be discussing the following points:
- What language is and the importance of learning new languages.
- What natural language and its importance.
- Natural language limitation.
- My take on natural language processing’s effect.
What is a language and why is it important that we learn it?
Oxford Living Dictionaries defines language as “The method of human communication, either spoken or written”. If we were to go a level deeper, language is considered to be anthroposemiotics- which translate to the ability to cooperatively share, exchange ideas and information through vocal interaction and body language
Dr. Uri Hasson — Director of Cognitive Neuroscience Program, provides quite an interesting explanation to human reaction when it comes to vocal interaction. As humans communicate, our brains begin to receive the soundwave from other and form our own thought as it processes the sound and imagines the scenario. Since the brain uses imagination to create our thought, the brain produces neural waves based on how the sound is sent to the receiver. An example is how when Dr. Hasson experiments “The Moth” story by a professional storyteller Jim O’Grady with MRI scanner to see 5 different listeners’ brain waves. As the 5 listeners rest before the story, their brain waves are in 5 completely separated states as expected. But when the experiment starts, the 5 brain waves are slowly going up and down accordingly to the story’s rhythm until the waves almost synchronize perfectly. Dr. Hasson called this phenomenon as neural entrainment, which is the brain’s synchronization to the soundwave from the environment to recreate visual, auditory or tactile for us human’s imagination. Correspondingly, how is this applicable to learning new languages?
When it comes to learning foreign languages, our brains have to rewire itself in order to understand new concepts, which results in shaping our thought, experience into a new way of thinking.
Not only learning language helps you develop new insight from what you’ve already known, but also build your cognitive skill. As mentioned earlier regarding brainwave and human’s reaction, the cerebral cortex is responsible for these actions that also known as the gray matter, which play a role in controlling attention span, forming thoughts, learning languages and many more. As we’re learning new language, the cerebral cortex optimize performances in order to maximize proficiency in obtaining new knowledge.
For example, let’s take the shades of green in Vietnamese as reference:
xanh lá cây, xanh lá chuối and xanh ngọc
The words share one common word which is “xanh”, meaning green. Analyzing the word “xanh” furthermore, by adding more adjective next to it we get:
lá cây → leaf,
lá chuối → banana leaf,
ngọc → emerald.
Hence, learning foreign languages isn’t only just about memorization but also learn how to learn more efficiently.
Speaking of efficiency in learning, natural language processing is one of the most developing technology in solving our language barrier to a much greater extent.
What is natural language and why is it important?
Natural language processing is a branch of computer science field consists of information engineering, and artificial intelligence that deals with computer’s interaction with human languages; particularly, on how programming and analyzing natural language.
Despite how complex natural language processing sounds, it is something you use daily without knowing it.
Take Google as an example, whenever you type a line in the search bar, it always show the result before you could finish your sentence. This is known as predictive typing that autocomplete fills the sentence based on how common the sentence is asked for. There’s also autocorrect that allows the user to find the answer based on natural language data. Sequentially, how do these 2 natural language processings help us in learning language compare to the traditional way of learning language? Let’s assume you want to learn how to say hello in French to meet your French relative. The first step is to type “how to say hello in French” that shows a page of Google Translate along with the line “Hello (English) → Bonjour (French)”. But you want to how to pronounce “Bonjour” accurately without using the text-based pronunciation of /bɑn.ˈʒʊɹ/. Google would highlight the speaker-shape symbol to let you hear the pronunciation of Bonjour. However, learning new words tend to be easy to forget the next day. Then you try to write an email to your French relative with an opening “Bonjor” since you forgot the spelling of Bonjour. But luckily, natural language processing would highlight the word and suggest with the correct version effortlessly without flipping through each page in the dictionary until you find Bonjour. Moreover, natural learning processing also encourages students to learn language easier and more engaging.
Little Pim is a self-learning language software for toddlers that currently becoming more affordable and common for all families. The object of this natural language processing is to develop young boys and girls early access to learning language; the toddlers are encouraged, engaged with cartoon characters and fun puzzle as a way to maintain their learning interest. Thus, natural language processing can be really useful in learning new languages; however, users have to keep in mind that it’s a tool to support learning. not a crutch to take advantage of.
Thus, one question remains, If natural language processing in the future allows for real-time, perfect translation, should human then ignore studying languages?
The answer is: It depends.
Natural language limitation
For a normal person, the needs to communicate in other languages is more beneficial than the understanding of the language itself. By saving time and allocating space in our brain we can use it to learn something else that can not be easily obtained through AI.
However, as it was mentioned above on how Google Translate aids us in learning language, there is another example you should be aware regarding A.I translation.
Let use the case of translating “A Fairy Song” by Shakespeare to another language? If you weren’t a native english speaker, the translation would present significant amounts of mistranslation and misinterpretation. Let’s kick it up a notch, what if Google Translate was the real-time, perfect translator; would the poem be translated ideally as the author intended? The problem still lies at hand if natural language processing can translate to keep its hidden meaning, rhythm or wordplay that only our human mind appreciates. Unlike machine translators, human learn languages not only through word by word like machine do, but also the culture and history of the language. Hence, even if there’s a universe where machines have the capacity to learn languages with humanlike minds, though while we can achieve the result of communicating our thoughts and ideas, we skip entirely the process of learning how to learn and the culture/history behind the language itself.
Natural language processing’s effect.
Learning a new language is something that we can’t fully master or expect A.I to do. But we should utilize technology to extend our reach in learning the language more proficiently. Because of our differences when it comes to language, we could admire the diversity of our worldview and beauty in each native language. In short, language natural processing has helped us a long way in learning new languages, but we should be mindful on how, when to use it rather using technology to limit ourselves from improving our intellectual growth.