Japanese on Duolingo! Yay! … Or Not.

I recently read an article from Kuma Sensei, a Japanese learning blog, commenting on the recent addition of Japanese to Duolingo. I have used Duolingo in the past, both commending and criticizing it. When I saw that Japanese was added to Duolingo, I had to bite my tongue so that I wouldn’t start screaming about other languages that should be added. Before I jump into this article’s main point, I’d suggest reading the article first: https://kumasensei.net/learn-japanese-duolingo-review/.  Kuma Sensei offers a qualified and in-depth evaluation of Duolingo’s Japanese course, which, to my knowledge, is currently available only on iOS and eventually Android. Given that Duolingo is a primarily web-based application, this is a bit odd. Kuma Sensei’s overall evaluation seems to be summed up with one quote:

“Duolingo may just be what the doctor ordered for people who absolutely loathe using textbooks and want to just sit down and start learning Japanese for free.”

This is a totally fair observation, since in my experience, most language learners do not seem particularly keen on academically-oriented study programs. That said, Duolingo’s Japanese doesn’t escape Kuma Sensei unscathed. There’s a remarkable lack of grammatical explanation, which seems to be the case for most Duolingo courses.

Even for Italian and Spanish, arguably fairly simple languages in terms of grammar, the explanations of when to use certain verbal forms leaves much to be desired. And again, maybe that’s Duolingo’s appeal. But context-based translations and nuance, which are key skills to acquire as a language learner (no matter who you are) are completely lost on our beloved owl. However, Japanese’s more complex features, such as the mandatory mixed use of hiragana, katakana, and kanji are not at all explained, which I label as a serious deficiency of the course. Although, to quote Kuma Sensei: “You’re lucky you’re still in beta phase, punk.” It’s unfortunately apt that in Kannada (and most of India’s languages), being compared to an owl is to be considered unintelligent.

Which brings me to my point. I’ve been pushing for Kannada to be added to Duolingo for almost four years now, and I’ve yet to actually receive any kind of communication from Duolingo to discuss the potential project. My growing frustrations with Duolingo’s apparent disinclination to support minority languages, compounded with the flaws of the Japanese course are eating away at my faith in its ability to support language learning. I’m well aware that Duolingo is not a great tool for those aiming to become even conversational in a given language, but ostensibly, that is what Duolingo purports to do.

I want to like Duolingo, really, I do. The game-like aspects make it a really powerful starting tool for language learners, but unfortunately no more than that. There’s a lot of further work to be done on your own, which is kind of unavoidable. Duolingo has a lot of potential for bringing up minority languages, which it already has shown it can do, given the availability of Welsh, Irish, Vietnamese, and Turkish courses. Granted, these languages are rendered in Latin script anyway, so that may make things easier. But knowing that the Japanese course is so flawed, it might not be that these other courses are any better.

I’d be really glad to hear anyone’s thoughts on this, and please don’t forget to share this on your social media!

しりとり (Shiritori) and Word Games

Today, while hanging out with a few of my Japanese friends, I learned about a game called しりとり (shiritori), which is a type of word game where people say words, take the final kana (or syllable) and use that to find another word that begins with it. It was pretty difficult for me, since I have a fairly limited knowledge of Japanese words. So, that means if I say umi, the person after me has to say word that begins with mi. Obviously, you have to know the kana spelling of a word in order to play this game properly. The catch is that you cannot play words that end in the kana ん (n), since no words in Japanese end with this kana. On top of that, you can only play common nouns, so no names of places or people. If you are in a position where you have no choice but to play a word that ends in ん, then you lose. A similar game called “word chain” exists in English, though this version has way fewer way to ways to lose, since very few letters in English are like ん for the purposes of the game.

Now, what this made me think about is the fact that the idea of “spelling” is an almost unique thing to English, since nearly all letters have more than one possible pronunciation that overlaps with other letters. In Spanish and Italian, for example, spelling is fundamentally unimportant, since every letter has a one pronunciation and one only, and all words are spelled exactly the way they sound. French could conceivably have spelling-based games, since more letters are ambiguous the way English is. Even if the letter or symbol of a language has multiple pronunciations depending on the position of it in a word, spelling is insignificant so long as there no overlaps with other letters. For example, the letter “f” and the combination “ph” make the same sound, but are used to spell things in different ways. “Ph” is used in almost exclusively words of Greek origin, like “philosophy” or “philanthropy”, and “f” for everything else. But for the unlearned player of word chain, these words have ambiguous spellings.

Another thing that this pointed out to me is that in many languages, this game can end very quickly. For example, in Italian, nearly every word ends in a vowel, and that significantly shrinks the bank of words you can use for the game. Spanish has a similar problem, since relatively few words end in consonants other than and s. In many (if not all0 Indian languages, this game is not feasible, at least if it’s played like shiritori. Using the final syllable is very difficult, since even though Indian languages use abugidas, where each letter is almost always syllable unto itself. The problems come up when you have a syllable that has more than one consonant in it. For example, if I were to use the Kannada word ಮಿತ್ರ (mitra), the next word has to begin with ತ್ರ (tra), of which there are very few. It’s even worse if you play a word that ends in the sound ಋ (ṛ), since there are very, very few words that actually start with this letter. It’s just that the writing system is not suited for such games. For what might be obvious reasons, Chinese languages cannot play this game, since hanzi don’t work that way. Using radicals to determine the next word requires too much knowledge on the part of the player. Also, pinyin finals can’t always start a word, and tones restrict syllables even more.

Some of the languages that I think are suitable for this game (using either the Japanese or English version of the rules) include Greek, Russian, Korean, possibly Vietnamese, maybe Irish, and Catalan. Correct me if you think I’m wrong. One of the keys to this game is that there has to be a letter or symbol that little to no words can start with.

I hope you enjoyed this post, and I highly suggest playing it for practice in the languages mentioned. Please remember to share this wherever you think people will be interested!

Progress on Kannada Duolingo

I recently completed a preliminary version of the Kannada Duolingo course by creating a course on Memrise. If you don’t know, Duolingo is a language learning site and application available on computers, iOS, and Android. Duolingo provides free language instruction to anyone who has a computer or a smartphone. This is a revolutionary service, since it is extremely accessible and democratic. The courses rely contributors from around the world to improve and revise the courses, using a cloud-based system. Naturally, they do screen contributors for genuine knowledge of the language in question and commitment. Now, there are already a few courses on Duolingo that teach minority or at least non-mainstream languages, including Welsh, Irish, Romanian, and Polish. Duolingo’s service is an important tool for ethnic groups around the world to preserve their languages and inform the world. However, because Duolingo is not working on languages that require learners to learn a new script to read the language, languages like Chinese, Japanese, and Hindi are significantly delayed. Duolingo also relies on its employees’ internal knowledge of the languages. I am pushing for Duolingo to create a course for Kannada, so as to preserve the language of the Kannada-speaking community, particularly in expatriate communities.

Memrise is another useful service for language learners, and has a simple but very accessible way for people to write their own courses. The Kannada learning course that I wrote and recorded is up and running already! You can check it out here: http://www.memrise.com/course/990976/learn-kannada-3/

I’m considering putting up the full text as a textbook at some point, but I’m having some people review at the moment. Please keep an eye out!

Making Multinationals work for Multilingualism

Making Multinationals work for Multilingualism.

This is an awesome prospect for any threatened or minority language. We should be making companies place such languages as their priorities if they operate in places where the language is spoken by the majority.

Why Minority Languages Matter

A lot of people will question learning minority languages such as Catalán, Navajo, or Irish. Many believe it is a waste of time, and that language death is inevitable. However, for the languages already mentioned, as well as several others, it is well within that community and other people’s capacity to help revitalize usage. Tom Scott makes a valid point about how if we let minority languages die, there are certain aspects of the human experience and capabilities of the brain that we let die with them. You can watch his video here: Fantastic Features We Don’t Have In the English Language.

Language is intimately linked to the way we live our lives. It is theorized that language evolved out of a method for human mothers to communicate with their children, and as human society became increasingly complex, involving multiple individuals in the process of raising children, it eventually became a medium for communicating with one another. Another hypothesis is that language is a vocal manifestation of one’s ideas. Ideas are apparent to oneself, in one’s mind, but not necessarily in comprehensible language. The idea is that humans needed a way to communicate their ideas and feelings regarding things, and that is why language evolved. Personally, the theory regarding mothers is a lot more plausible. There’s a reason, “motherese,” exists. However, these two hypotheses do point out crucial facts about the development of language. Languages have features based on the particular needs of a people in a certain place.

For example, the aboriginal language in Australia does not have words for left, right, up, or down, but rather assigns cardinal directions. As a result, most of the speakers of this language have an intuitive sense of direction..Some have proposed that due to the lack of landmarks for people to judge physical position in the Australian wilderness, language there had to have less arbitrary ways of describing direction. In a place like the Americas, the landscape is varied enough for people to judge direction based of off the various shapes of the land, and therefore, the language there can assign arbitrary directions, or at least directions revolve around a given point. The ability to distinguish direction in absolute terms is very useful, and demonstrates the capacity of the human brain to evaluate its surroundings as such. If this language dies out, we miss out on a generation of people who have this ability, and completely exclude it from the development of other people in the world.

Now, let’s look at a non-physical example. In Catalán, the construction no… pas is a nuanced one. It negates a predicate, and also indicates that this negation is contrary to a notion held by listener. This is a very useful feature, and is built into only a few words. It is for this reason that some non-native speakers of English can be very verbose, because they’re trying to express an equivalent sentiment of what might be a very short sentence in their native language. Implications and nuance are very important in some languages, especially in minority languages, where they can be unique to those languages. By letting such a language die, you allow a possibly more effective and expressive mode of communication die as well.

Perhaps the most grave loss in the process of language death is the loss of a culture and people. Language, as stated before, contains a great deal of history and knowledge behind the way people communicate. John McWhorter argues that language death and the loss of a culture are not necessarily linked. I refute this point, because of the reasons listed above. Skills and modes of expression that are exclusive to a particular language are part of a culture. A people lose a great deal of themselves in not being able to speak their language. There are things they will not be able to understand or express. Sure, they can maintain their traditions, but the meaning and history of those traditions is lost outside of the native language. By working to revitalize minority languages, even only within their indigenous areas, we maintain another part of the human experience. If it happened with Hebrew due to the work of Eliezer Ben-Yahuda, it can happen for any language at any time!

Languages are different for a reason. The subtle nuances and implications of certain words and phrases can often be lost in translation. There’s a reason that people who read manga in English will miss much of the symbolism, hidden meanings, jokes, puns, or wordplays that the original Japanese text might have. This is why I believe that translation can never do real justice to having a proper conversation in the language being translated. In a world with infinitely varied settings and circumstances, knowing other languages that express certain sentiments more accurately is paramount.

It’s been some time since I’ve written a full article. I haven’t really been doing much lately except writing language guides and subtitling Khan Academy videos (which you should do, if you know a language that you think people would benefit from having subtitles in).  I’d appreciate any comments on this, so feel free to leave some!