1. Forum
  2. >
  3. Topic: Duolingo
  4. >
  5. Good News, it's the weekly "W…


Good News, it's the weekly "When do we get new languages" question again!

Well I have definitely noticed this is a very unpopular question, and really all us hardcore duolingo users know the answer anyway.... But -

It would nice to have some insight on how it is progressing. Will the user created courses be managed or curated by Duolingo staff or will it be a Wild West approach sort of like Memrise? Will it be like a wiki, or will it be up to the course author alone or their select friends? Are the myriad of different writing systems the world has to offer giving you as much of a headache as they gave me as I was learning them (still working on Chinese, thats a tough nut...) ? Have you decided how teach non latin scripts or will you rely on (gasp) transliteration?

Pleeeeeeeeeeeeeeeeeeeeeeeeeeeeeeease? O.O

July 30, 2013



Hi all! As you guessed language requests come in quite a lot. We also get tons of community members asking to help us in these efforts, so we've decided to create tools that will able you, the community member, help us add more languages. We love including you all in the process. We've just gotten started. I can guarantee you that we'll be updating you all along the way as we find answers to all these questions :)


In the meantime, my polite suggestion is that you put up the most basic of FAQ pages that could have just the one question and answer, this most frequently asked 'More languages?' one. The rest of the FAQ could be filled in later. But having that one Q & A should stop at least some of these redundant forum posts.

This particular post wasn't downvoted because it was asked in a more knowledgeable way but I feel bad when I see the less informed new duolingo users posting the question and getting heavily downvoted, while also understanding why people downvote it.


Indeed, it is sort of like the automatic "RTFM" response. Except its a bit unfair when there isn't one and new users get downvoted into the -50's without knowing why everyone is being so mean :)

To be honest the total silence is just as good a sign anyway, it means you lot are prioritising actually working on the thing rather than waffling on about it on some blog :). I just can't wait to see duolingo be applied to every language in the universe...


It's great to hear this from you guys, I cannot wait. I really want to learn a couple of languages that are not on here yet and I am really looking forward to seeing this introduced.


Really liking duolingo after I returned to Latin based languages. I wanted to suggest a way in which duo can teach Japanese Kanji (should Japanese ever become a language offered). That being through teaching via Kanji Radicals (aka Kanji Damage). No commercially written book teaches this way (it's a growing online movement), it's been the best way a non-native can learn Kanji as fast as I'm learning vocab on duo. For instance, the word for Rain: 雨 becomes a radical in almost every weather oriented word like cloud 雲 snow 雪 or lightning 電. Apologies if this isn't the thread for it, but, this seemed the place to put it!


Sounds like James Heisig's Remembering the Kanji; it's not really an online movement, I'd say.


A little kind of different than his method; though there -are- similarities. A lot of the failures I've seen every 'published' book -- even Genki -- is that they separate the words by the Grade Level. I more refer to ( http://kanjidamage.com/ ) that method. What works, is that it creates relationships without relying on the lie of 'Pictograms.'


Hesig's definitely the market leader here, and it is a book, though there are supporting websites. Hesig differs from KanjiDamage in that he teaches the meaning first, readings later (in another book). KanjiDamage includes some reading information. You might like to look at WaniKani, which rolls up the most common reading along with the kanji, and then uses learnt Kanji to introduce vocabulary.

However, the underlying point of this is creating an effective beginner Japanese course is a whole lot harder than the current English to/from European languages Duolingo offers.


I sure hope it doesn't take the Memrise approach... I tried using it, but quickly quit it. I much prefer Duolingo.


My main issue with Memrise is the lack of grammar instruction. It's pretty much all vocabulary all the time. In addition, they only quiz you on the vocabulary one way: they give you the word in your native language, and you have to translate it. Once my Spanish grammar is solid, I'll likely go back to Memrise to expand my vocabulary.

Regardless though, the comparison to Memrise here is that the Memrise staff doesn't decide what courses are good and which ones aren't. I had absolutely no problem with that, there were plenty of really good courses that were easy to find. Did you have a problem with that aspect of it?


You shouldn't only use one tool to learn. I personally find that DuoLingo is a little lacking in vocabulary, but Memrise is lacking actual "instruction" (it doesn't really explain grammar or anything, just introduces you to new words). I use both of them together and they really do compliment each other nicely.

DuoLingo for actual "lessons", Memrise for learning new words.


I think one of my problems was the lack of grammar, also. Having to translate from English is not a problem for me, since I would rather practice that more. I find translating from German easier, so practicing the harder thing is good.

A second problem was the memes, having to choose something to help me remember it, which didn't help me actually remember the word I was supposed to remember. To me, if felt like I was learning something wrong. How things were "said" in the memes was not how the words were actually supposed to be pronounced.

I probably shouldn't really say much against Memrise since I didn't try it very long. Maybe after more practice with it I would be okay with it. Plus, I know a lot of other people have success there, so it really can't be that bad. Just not my style, perhaps.


Memrise is really a general purpose thing though, it is only as good as the author of the particular course makes it. I like it for memorising chinese glyphs, and it's ok at that since the context is not amazingly important, just how reliable your recall is. I haven't tried it for anything else because I am much more of an audio learner, I prefer to load something on my music player and study that while I am doing something else, then it is a zero cost effort for me. Memrise and Duolingo are just nice for those times when I have nothing to do and I actually want to drill, so the finesse of the system is not really my priority :)


I agree. Even though Memrise had just about all languages there was no specific aproach to learning whatever language. Users just created a lesson and then put random words without grammar. I found it quite difficult when I was trying to learn Russian.


As I see it above though, it really is just a drill, and if you use it in that capacity, its quite handy. Just like you would use a drill to put up a picture, but probably wouldn't to saw a table in half...

Like it or not, drilling is crude, but still effective at forcing information to stick in your head. I go back to glyphs I haven't seen for months and think to myself "How the hell do I remember what that one means?"

Just speaking up for memrise, since I don't actually see memrise and duolingo as equivalent tools (or even in competition)


Yes, they definitely have to be used for different things, and I don't see them as competing much either. I do a lot of practices here, and study on my own while offline. Yesterday I was trying Memrise some more, like I said I should somewhere above. It is good at drilling vocabulary, as you said, chilvence. Just because I don't care for it doesn't mean other people should not at least try it.

Though I am someone who doesn't really trust user-created content (look too much around they internet and you may see why, I hope there will be some sort of quality control), being able to make your own can really help a person remember things, also. The memes are more of a distraction for me, and I have more success when I don't pay them much attention. For me, I focus on the shapes of words, and remember those. Sort of like how we read English words once we are good at reading.

This discussion has made me really think about what works and doesn't work for me.


It's funny you say that, more often than not it is the irritation of dumb meme's that spurs me into thinking of my own, and that's like pouring concrete into your brain.


Heh, yeah. Similarly, I prefer to write things down in a notebook. That helps me remember. Early on when learning something, I remember it by remembering where I wrote it in the notebook.


As long as this question has been raised, I'd like to ask a similar one. I was just wondering which languages might be slated for inclusion first? I'm guessing most asian languages are on hold, due to the different alphabets, but I admit that something like Esperanto would be interesting to see.

Whatever happens, I'm happy enough. I never thought I'd learn Italian, and here I am, over 600 words in! Even if this site stays as is, it's the best thing I've ever encountered on the internet!


The different alphabets question is the one I would most like them to answer. I am obsessive at learning alphabets, and I genuinely think that they are alot easier to learn than most people think - the only writing system that will give you a real challenge is Chinese/Japanese glyphs, and only because there's rather a lot of them.

Fundamentally the only thing stopping anyone is the keyboard issue, everyone has the latent capacity to be literate with another set of symbols, but it is not really trivial to add a new keyboard layout, and your keyboard certainly doesn't come with the right markings anyway unless you bought it on holiday. Google seems to have a nice transliteration input system on their translate page, and it is a lesser problem on phones because of virtual keyboards, but still, its a tough one and I am not creative enough to think of a solution.

One thing I do know though, is that anyone wanting to learn Arabic, Hindi, Russian etc is not going to be satisfied just learning a 'romanised' version of it that is going to be useless in the real world, and I hope Duolingo doesn't sacrifice that fact in the name of being 'accessible'. It is much more valuable to show people just how easy it is to learn a new writing system, then they have something they can be proud of. Give a man a fish they say...


The Korean alphabet is really easy to learn. I for one would vote for Duolingo Korean.


Well, most alphabets are easy to learn really. What's special about the Korean alphabet is how nicely thought through it is and how well it gels stylistically with Chinese glyphs. I'm really surprised it hasn't spread throughout the region.


I'd personally take this (http://www.apronus.com/internet/ruskey.htm) as an inspiration. It shows you an onscreen keyboard, takes your key inputs and converts them to the respective Cyrillic inputs.

Arabic, Hindi and Cyrillic all fit onto a keyboard, so the same approach can be used for all of them. Chinese actually has two different ways of being input, one being via strokes mapped to the keyboard, and the other being the pinjin latinized version which is used by many chinese. For example 北京 will be written as Běijīng, with the software suggesting different glyps based on what you've typed.


That's pretty useful. Unfortunately It's made me realise another thing, since I chose a custom Cyrillic layout closely matching QWERTY, I am useless at the standard layout. But I would probably be useless at dvorak as well. So that's another trade off, teach the more unfamiliar, but long term usable layout, or use a more familiar layout to take the pain away at the beginning?

I think I have also posted here before how I don't imagine it will be easy to make a Chinese course without pinyin. There is nothing fundamentally difficult about learning the glyphs, but input by stroke order is probably enough to make beginners brains explode. You fundamentally have to learn that stuff from a calligraphy lesson with an expert teacher and on real bona fide paper. And I say that even though I am quite proud to have worked a lot of it out myself.


The stroke order thing isn't that hard, after all primary school children learn it in china just fine. Still, I agree that teaching chinese by stroke order might be complicated. On the other hand, there should be a way to teach other alphabets as along with the language.


It's easy to say that yeah, but the average primary school kid in China already knows a lot more about Chinese than I do, even in 1st grade :)

What I was really getting at, is for a skill that requires so much attention to detail and dedication to master it is a lot better to learn with some confident and experienced instruction, especially since it is so far removed from the western idea of writing. I don't deny that anyone can do it, I just mean that the Duo team will have to work really damn hard to make is seem as easy as the languages they already have.


I doubt Duolingo would construct romanised courses for Russian, Japanese etc. because (as I have been lead to understand) their business model centres on providing free training for their for-profit crowd-sourced translation business, and real websites are printed in the native scripts. All things considered, this isn't necessarily a major obstacle.

I have been typing in Japanese (as often as the need arises) for over two years. I use a Linux-based application called "Anthy" which converts Romaji to Kana and Kanji (with an arrow key menu to scroll through and select what I intend to write, when there are multiple homophonic possibilities). Pretty much all modern OS's can be configured with similar tools, and instructions for enabling them can be found here: http://www.localizingjapan.com/blog/

I know that Pinyin to Hanzi inputs can be configured just as easily, and I would be very surprised to hear that aren't similar tools for Hangeul. Duolingo could integrate tools like this into the website itself, but I imagine that it would eat up a lot of bandwidth and cause (more) lag for the users; far better to guide users through setting these tools up on their own machines, since they will want to use them for other applications anyway. In any case, the first few lessons on the tree of any exotic-script-language would be devoted to learning to use the input mechanism.

As for variant keyboard layouts, conversion stickers can be ordered from many suppliers for about $1-3 dollars each. You can add and switch between layouts through your particular OS's input options. If I were to start learning a language like Russian or Greek (with its own alphabet and standard keyboard layout), I would probably buy a USB keyboard to convert for the purpose. How Duo will deal with Cyrillic, I am far less certain.

On a side note, I have been using Programmer Dvorak ( http://www.kaufmann.no/roland/dvorak/ ) as my keyboard layout for many years, and I simply can't bring myself to imagine how people type in French or German without these ("`^'éüçß) keys on their keyboard. I know Duo has the point-and-click "buttons", but still...


Well, at least we know they are working on them :)


They have an internal policy not to release dates. Although, I'm not sure if this includes what their current progress is. I believe the new languages will be added so each person can vote translations. For example, at the moment we have to submit new translations for sentences at the end of the tree since some aren't accurate. For this case, users would submit translations, and the crowd would vote them up, and trusted users will allow popular suggestions to be used. (Or something along these lines).


We release dates when we know what they are for sure. For example, when we launched the Android app we told the community months before. Same goes for releasing the iPad app :) We'll keep you looped in. Don't worry.


Luis told me you had a policy, but I'm not sure how restrictive it is so yeah haha, but thanks for the updates :)

EDIT: Let me know when the tools are available for testing because I would love to help start adding Dutch.


Are you going to do user based vocabulary expansion as well? I'd help adding words to the German/English course if it becomes and option.


Yea, I wish the vocabulary on some of the languages were broader than they currently are. :)


I guess my question is slightly different, but fairly similar in a way. How would someone who wanted to get involved in creating a new language course go about getting started? (Not me, I hasten to add, I'd be one of the end users, but before people throw a lot of effort into trying to create something that I think the duolingo structure could do better, an idea of whether it might be possible for them (and me, because I'd be willing to do the hosting and sysadmin) to do so here would be really useful).

In the meantime I'm torn between carrying on learning Italian or holding out for Corsican, which is the language I really want to learn (and which I'm somewhat more proficient in than Italian already ...)


I would hope that Duolingo would have an already layed out tree for that language and it would be similar to how the current five are. Then when you want to add words you can just make a few clicks, it gives you a set of words for you to translate and then it puts them into the database along with other people's inputs. An idea for verbs could be that a table/chart would pop up and you could fill in the conjugations for that verb in that verb tense.

This is an idea that I just thought of. Maybe Duolingo could use it :)


I really like your idea, I think this is the best way of adding a new language using an existing template. But for languages that use new letters (different to latin keyboard e.g. Russian, Arabic etc.) there will be added a new first branch called 'Alphabet'. Also the duolingo staff will have to work hard to monitor all the input to make sure it is of a good quality and is grammatically correct.

By the way, I will be happy to help with the development of Russian language :)


Yes. I think that it would need a new alphabet branch too that would include pronounciation of each letter. It would be even better, but harder, to add grammatical explanations to each letter saying when each letter sounds like what. For instance the German 'e' sounds like "eh" but when put in front of 'i' it becomes an 'i' sound and when it is put after the 'i' it sounds like "ee".

Yes the Duolingo staff would have to work extra hard to monitor the inputs of the general public. Maybe if there isn't any staff that know this language they might hire a specialist or native to make sure that its all correct.

On your note about Russian. That would be awesome! I am wanting to learn Russian :)


I think it would be quite simple if contributions are logged by user, and if a pattern of bad translations develops with one user it is very easy for an expert to prune/correct them. Other users can be market as trusted. New users could even be polled for which languages they are completely fluent in, which would be more of a trust thing, but at least a starting point.


Well I think the issue is more that someone who actually knows the language (someone who can tell when something is wrong and not by just using Google Translate) should monitor the inputs of the public. I think that the fluency poll would be useful since they could be trusted more than others with putting in information.


About your last point, I think you should just continue learning Italian. I'm still learning french, but I'm finding it a lot easier to pick up Spanish compared to how I found learning french originally.


español-italiano please!


more language from french please! =)

Learn a language in just 5 minutes a day. For free.