Language apps and chatbots make it possible for us to learn languages from our phones. Thanks to development in machine learning, these automated translators are becoming more accurate and efficient.
So, with all of these new technological developments, should we even bother to learn a new language?
Wouldn’t it be much easier and efficient if the whole world chose to speak Mandarin, Spanish, or English?
In my opinion, no. Along with countless other problems, monolingualism would cause all of the interesting, useful, eye-opening ideas that exist in all of the world’s languages to evaporate.
At its core, losing a language would mean losing cultural diversity and human heritage.
However, recent years have seen a major decrease in college students learning foreign languages. Promises such as ‘enhancing career prospects’, just don’t do it anymore when we can just use Google Translate.
Besides, why would anyone pay college tuition to learn Spanish when they can do it for free on Duolingo?
“Everytime a Language Dies, it’s the Equivalent of
a Bomb Being Dropped on the Louvre.”
Every one of the world’s 6,909 languages have a unique way of seeing what has been woven into it over thousands of generations. Whenever a language dies, an immense amount of cultural heritage disappears with it.
Renowned MIT linguist Ken Hale has said that every time a language dies, it’s the cultural equivalent of a bomb being dropped on the Louvre.
The Significance of Words
Within every word is a ledger of history that can be traced back to reveal parts of history and aspects of humanity.
Take the Irish word ‘craic’ for example. Anyone who has been to Ireland will have seen this painted onto touristy pubs and splashed across postcards.
For those who aren’t familiar with the word, it refers to something that’s fun or entertaining. For example, you might do a jig on the bar or start a snowball fight for the ‘craic’.
You may even here Irish people asking “What’s the craic?” instead of “How are you?”.
We are all aware that languages borrow freely from one another. Looking at the word ‘craic’ would lead us to assume that it must have come from the Irish language and survived even after being banned under British rule in the 18th century. However, this is not the case.
What makes this word so interesting is that it was originally an English word meaning ‘to talk loudly or boldly’. The word died out in England but lived on in Ireland and transformed to become something many see as quintessentially Irish.
What is really mind-boggling is that lately it has been readopted in England and is creeping back into everyday use. There is even a campaign to change the spelling to ‘crack’ to fit better into the English language.'Craic', the iconic Irish slang, is originally an English word #craicorcrack #nocrackClick To Tweet
These transliterations continue to occur in modern times. For example, the word “emoji” is a Japanese compound of “e”, which means ‘picture’ and “moji”, which means ‘character letter’.
It has been successfully integrated into the English language for two reasons. Firstly, it matches the familiar prefix ‘e’ that we find in words like ‘e-mail’or ‘e-reader’.
Secondly, the word emoji appeared at the perfect time. Some of the MSN generation may be familiar with the word ‘emoticon’ — a combination of ‘emotion’ and ‘icon’.
These icons were no longer smiley faces or sad faces but can be anything from a toadstool to a bento box. They outgrew this word just as ’emoji’ swooped in.
Words like ‘craic’ and ‘emoji’ have so much history packed into them. Words also express cultural diversity and often reveal an insight into their speakers.
Many of these words cannot be translated, they simply are not a named concept in cultures outside their own.
The word ‘hanyauku’ from Rukwangali, Namibia, describes walking on your toes on warm sand. The German word ‘schilderwald’ refers to a street with so many road signs that you become lost. ‘Tsundoko’ is a Japanese word that describes the constant act of buying books without ever reading them.
These unique words are among so many other eye-opening examples. They demonstrate the complexity of the human mind and all the different perspectives of the world that are possible.
Languages are not just tools to describe the world, but also determine how we see the world
In an NPR podcast called ‘Hidden Brain’, Lera Boroditsky, a cognitive science professor at the University of California, described her experience with an Aboriginal community in Northern Australia.
She found that learning Kuuk Thaayorre, the language this community spoke, showed that the mind can work in different ways.
In this language, the concepts of right and left don’t exist. Instead, everything is placed in cardinal directions like north, south, east and west.
For example, if you were to do the hokey pokey in Kuuk Thaayorre, you might put your southwest leg in, and take your northwest leg out.
Interestingly, speakers of this language have been shown to orient much better than it was believed that humans could.
Boroditsky explains how we always knew that certain species of animals had abilities to orient which we believed were better than humans due to having magnets in their beaks or their scales.
As it turns out, humans do not have any biological excuse for why we can’t do it. In fact, if our language and culture require us to keep track of this information, humans can stay oriented surprisingly well.
Cultural Ideas Can Be Reflected in Language
Language can be shaped by the way humans experience the world, but can they play a role in the way we think about the world?
Studies have shown that when grammatical gender is applied to all nouns, non-bilingual speakers saw the world in a very gendered way.
For example, if you are taught that a fork is categorized grammatically as feminine, as in French, you are likely to think forks are more feminine.
Whereas a bilingual person who knew that fork was feminine in French but masculine in Spanish would be more likely to determine that the word is only gendered due to a formal property of the language.
Studies have shown that the categorization that language provides becomes psychologically real in the minds of its users.
On one side lie languages such as Spanish and Hindi where nouns are gendered. In the middle, we find languages like English where nouns are not gendered but pronouns are.
Finally, on the far side of the spectrum, we have languages like Persian and Finnish where often gender doesn’t even come into the conversation unless it is directly referred to. How does this translate to the way we think about people in our daily lives?
Numerous studies have revealed that young kids who were learning to speak in Hebrew, a language that uses gender cases heavily, figured out whether they were a boy or a girl almost a year earlier than kids learning Finnish, a language with little gender marking.
It makes you think; could this language filter the way sexism affects the world without us even realizing?
The World Health Organisation has compared gender equality and gender parity norms across the world. Taking pay, representation in government, and education into consideration, they have found that equality between the sexes correlates with the level of gender use of the language spoken by the population of that country.
Putting Google Translate to the Test
We have established that culturally, socially, and psychologically, languages matter a great deal. They are at the very core of our human expression and identity.
However, can technology enhance our learning of a new language?
Despite every language teacher I’ve ever come across cautioning against its use, Google Translate, since it’s 2006 debut, it has served over 500 million monthly users in need of 140 billion words a day.Since its 2006 debut, @GoogleTranslate has served over 500 million monthly users in need of 140 billion words a day.Click To Tweet
In 2011, Google Brain, the AI department at Google, implemented the principles of artificial “neural networks”. These neural networks learn by trial and error while inputting data into Google Translate.
The online translator is now able to learn not only all grammatical rules and vocabulary but also classifications and predictions based on its abilities to recognize patterns in data.
This is a major improvement — translations on the platform have become far more natural sounding and precision has improved drastically.
At first glance, Google Translate is a serious threat to linguists and translators alike. Once we automate a task, we downgrade the relevant skill that was involved before.
However, the current technology involved in machine translators like Google Translate still cannot match a human interpreter.
Let’s not forget that language calls for the ability to adapt. To operate in a language, especially one that is not your mother tongue, you need to be able to improvise when information is lacking or absent entirely.
To speak and learn a language, you have to be able to handle the messiness of thinking on the spot and handling unique grammar rules that operate in inexplicable patterns that seem to never make sense.
Technology has not learned to fully handle such ambiguity just yet.
For example, if you ask Google translate to tell you how to say ‘I’m Irish’ in the Irish language, it will give you the answer “Tá mé Gaeilge”. This translation is pretty far from correct. To retranslate it back to English, you would be left with ‘I am the Irish language’; not exactly a common phrase.
Automated translators also cannot handle the nuances or artistry of language that human interpreters can. The most exact word for word translation is not always the best option.
Take for example the French phrase “Se taper le cul par terre“. In English, this means “to laugh hysterically”. However, if you were to ask Google Translate, it would tell you it meant “Butt hitting off the floor”.
A misunderstanding that I’m sure has led to endless disasters
What can Language Apps Provide?
In my search to try to find out whether the days of translators are numbered, with 25 million active monthly users, ‘Duolingo’ seemed like a good language app to start with.
According to their website, Duolingo is a: “free science-based language education platform that has organically become the most popular way to learn languages online’.
Valued at $700 million, offering the service for free seems like the least they could do.
The ‘science’ aspect of this statement is that the language app aims to create a language programme that is tailored to each learner’s specific needs. This is done by using adaptive algorithms which speed up or slow down according to how fast you retain information.
I have to admit the app is well-designed and engaging, yet it’s definitely not without errors. Recycling and repetition, however, are at its core.
Duolingo’s main tactic is audiolingual drilling. If you have a good memory, you are far more likely to succeed in this language app. But, I suppose that’s the basis for the education system as we know it, so you can’t hold that against them.
The language app takes you through translation, spelling, aural, and oral trials. Each lesson is broken down into a bite-size chunk.
It’s a useful way to fit a little bit of learning into your everyday life. I really enjoyed doing something productive on my phone for a change.
With this language app, You can conjugate in the coffee queue or recite food vocab while you’re making dinner.
However, I did find myself passing through the French prepositions section without really having a clue how to use them. When I made a mistake, it didn’t tell me why but instead just prompted me when a word was misplaced or wrong.
With the exception of conjunctions, it also failed to lay out the rules in much detail or to explain any of the countless exceptions to practically every grammar rule.
Anyone with a competitive streak will benefit from the gamification of the language app. The app (effective but annoying) reminds you to practice every day via notification and e-mail.
You not only compete with your own records but can connect with friends or even family to motivate you further. Also, as you move up the levels, your learning is reinforced with virtual rewards like bonus topics such as ‘flirting’.
Read more: How Useful is Gamification Marketing?
Not to brag or anything, but I managed to unlock this flirting level after just two weeks. After getting the go-ahead from my virtual owl tutor, I decided to put my new skills to the test.
My test subject was my boyfriend who is fluent in Italian. I approached him in the kitchen and said “Ci vieni spesso qui?”(“Come here often?”). This was greeted with a look of confusion, although I am sure my grammar was correct.
After a little explanation, he burst out laughing, not exactly the desired effect. Unfortunately, I couldn’t understand his response.
Later, he said no sane Italian would say something so cheesy and unnatural. Although he later admitted my accent wasn’t too bad, I don’t think I’ll be using that particular phrase anytime soon.
I found that although the language app was very user-friendly and entertaining, another problem was that I found myself buried in irrelevant vocabulary.
A key part of efficiently learning a new language is making it relate to your life and specific needs and putting your skills into practice through communication.
The Duolingo language app recently added chatbots to provide for this need. However, I found these were poorly executed and came nowhere close to having a real conversation.
One of the chatbot exercises left me in a virtual café with a waitress bot. It being after five o’clock (somewhere), I asked for ‘vino’. However, ordering something other than ‘caffè’ was incorrect.
Overall, I found my conversations with the chatbots to be more like guessing games than efficient communication tools.
On another note, admirably the founders of Duolingo have been forward-thinking in the possibilities the language app could offer.
They claim that the app offers free opportunities to learn languages to people who otherwise would have no other way.
It makes you consider how powerful technology could be as similar apps and ideas could be used to tackle global literacy problems.
Next up in the Language App Test, Tandem
Another language app, called Tandem, covers 150 different languages and a completely different approach.
Tandem works like a conversational exchange. The language app works primarily on the activity of finding a partner whose native language is the one you are trying to learn.
You can start off with communicating via chat or videos and determine who will be a suitable partner for you. I found that having similar interests and a sense of humor was far more valuable than having similar language proficiency.
Although, at first, I wanted to find someone whose English was as bad as my French, it worked out better to find someone who I actually enjoyed talking to.
The aim of Tandem is to converse with native speakers. Despite the fact that this may be intimidating for some beginners, having the added pressure of a real human being judging you can be very motivating.
Although Tandem doesn’t offer resources and games like Duolingo, it does give you the option to connect with professional tutors online. You do have to pay, but it’s a good idea.
Another positive aspect of Tandem was that the cultural aspect was present. I found out so much about life in France and was introduced to up to date and widely used slang. This is an exposure I simply couldn’t have got from an app like Duolingo.
After two weeks, I asked my partner for some feedback on my progress. She was very honest and agreed I had improved but, then again, our first conversation consisted of about two sentences from me.
One of the best aspects of the language app was that it forced me to use the French skills I had. It was easy to ignore emails or notifications from Duolingo, they weren’t from a human like my Tandem partner.
Language Robots: A New Frontier?
It seems that language apps are just the beginning. Soon, the possibility of having a robotic language tutor will be a reality.
This year, the European Commission has funded a project called L2TOR. Pronounced ‘el tutor’, these ‘social robots’ are part of a research project called Second Language Tutoring.
The project aims to design a child-friendly tutor robot that can be used to support teaching preschool children a second language. These robots were developed to counteract the numerous problems that apps like Duolingo and Tandem cause.
L2TOR teaches language by interacting with children in their social and referential world.
The L2TOR robot will be designed to interact naturally with children aged four years old in both their second language and the child’s native language.
In particular, the project is aimed at teaching Dutch and German to immigrant children speaking Turkish as a native language. This is only the beginning of what AI that is used to teach languages has to offer.
Language reflects human nature. Being inherently human, language portrays the way in which we conceptualize reality and how we relate to one another.
Learning languages opens our eyes to new ways of thinking and could lead us to discoveries which would otherwise never cross our minds.
As it stands, language apps like Duolingo and Tandem can help us to improve our language skills, and Google translator definitely has its perks.
But can machines truly understand the significance of the words we speak?
This is the next step for language apps and programs. A program may understand the word “Mother”, “Love”, or “Taco Tuesday”, but does it understand the significance? This is the next step in language app development, and it may be right around the corner.