1. Forum
  2. >
  3. Topic: Duolingo
  4. >
  5. Statistics on all incubator l…


Statistics on all incubator languages

Hi. Like so many Duolingo fans, I love Duolingo, and hope to see it grow. Like a great many of those users, I bug my friends who are native speakers of other languages and very good English speakers (or bilinguals) to sign up at the incubator.

(If I had a nickel for every time someone has asked me what I am getting paid to plug Duolingo...)

Well, a sobering thought occurs to me in that the people I recruit now may be waiting a long time or may never be considered. That's because of a fairly recent post where Luis mentioned there were some 20K applications!

So it would be great to know what languages have teams assembled, which are still looking, and how many people in each queue. That way Duolingo promoters can (1) stop filling the queue for languages that have teams in place and (2) spare our friends and acquaintances from bad experiences waiting endlessly to hear from the incubator.

So how about a list of all the languages for which applicants have submitted? Showing the name of the language, whether the team is in place, number of team members in place, number of vacancies, number of applicants, current position in the queue, current status of the language. There may be other information that would be helpful, but this page would be so appreciated by anyone like me.

eg: Name, Team in place, # on team, # of vacancies, # applicants, position in applicant queue, Status Russian to English, Yes, 4, 0, bajillion, 121, Beta

I made up some of the numbers, of course. Would be great if the table could be sorted by different columns, too.

Thanks for considering these suggestions.

January 6, 2014



Unfortunately, that's not how the incubator process goes, so I don't think you'll be able to collect much information here. There are no "teams in place" that you can't see, no "vacancies", no "queues". What you see on the incubator page is what exists right now. As incubator courses are successfully completed they add more, slowly enough to comfortably manage them. And it's more or less dependent on finding ~2 people qualified and motivated enough to work on a language pair that DL is interested in.

People who care about this should always apply and let their presence known to Duolingo--it would only help. At the same time creating the courses is not the only way to help; there is modding, contributing, beta-testing, being a discussion-monster, etc. And there are also people who are super qualified but are only able to help in other ways, as well as people still sending us applications to make the Turkish course even though it's been out on beta for more than a week now :-)


I appreciate the reply, but in particular the persons I am working on have no prior experience with Duolingo. Instead they are often language teachers in their own right. Good command of English.

My "hook" is that I tantalize them with how much their students would benefit from combining limited access to their live instruction with 24/7 access to Duolingo via web and app.

Since their first hands on Duolingo experience is the incubator, it really does matter quite a bit to me whether I am wasting their time or not.

As for "teams" not existing, you should read posts from moderators of incubator languages explaining why no additional moderators/contributors are needed pre-beta release for their languages.

I have read those posts, and if I had a link handy would gladly share it with you.


I believe that one of the key considerations in picking moderators is that they are already on duolingo and have a good usage history. I think that this is because they want people who already understand how duolingo works and have shown enough interest and enthusiasm in the way it works to use it for a decent amount of time.

So it you are encouraging people to sign up for the incubator who aren't already on duolingo then it might be a wasted effort, both for you and for them.


Well it would be nice to know if that's the case. Perhaps a Duolingo team member could clarify? Or you've seen that from one of them in a different post?

But in any case, the criterion you've just suggested is absent from the incubator application form, unless I have forgotten it. So I have proceeded accordingly since October.


I imagine that between two similar applications (similar in quality in both languages), Duolingo will prefer someone that is on Duo for some time and is active here on Duo, compare to someone who created an account on Duo, made an application and never came back on Duo and also compare to someone who just created an account.
Only because there is less risk for someone (among people applying) that work everyday on Duo and/or help regularly by answering other users' questions in forums etc. to quit Duolingo some days after being made moderator of a new class compared to someone who never really used the website.

Well, that's my guess.


no reason at all to assume a specialist would take more time than anyone else, much less do anything less well than someone with no experience as a teacher.


not faulting the logic :) but experienced teachers also bring something to the table. and key is aptitude at languages not aptitude at duolingo.


[Answer to tariqnisarahmed's beginning with "not faulting the logic :)"]

Yes but Duo's model is crowd-sourcing, so it seems to me to imply that they prefer a course well done quickly then made better during the beta phase by the crowd to an (almost) perfect course made by specialists but taking a more time to do (because less motivation/less time to do it) and with high risk (due to the fact that Duo don't know if this users that aren't active on Duo will stay at least some days) that of no progress.

And experienced teacher will still bring something to the table during beta test if (s)he is motivated to take the beta-course, and to report (with detailed comments) the issues (s)he finds.


[Answer to tariqnisarahmed's beginning with "no reason at all to assume a specialist"]

It was still under the comparison: fluent+very active vs. specialist+not active (at all), (within the crowd-sourcing model of Duolingo).

Of course, Duo would choose any perfect bilingual person who happens to be a professor and is highly active on Duo to do the course, well it would be strange not to do so. But that's a really rare species :) :) :)

EDIT: And I'm not sure that the teaching experience (of the concerned language) adds so much (compared to a fluent/bilingual person), since the task in incubator is to translate the Duo database of target language's sentences into the "from" language (then translate it back to all different possible meaning into the target language).


Tariq: Like I mentioned above, I am a moderator of an Incubator course, and so is jrikhal. It may be a little more illuminating to read what we have written under that light.

And thank you very much for spreading the good word. I'm sure Duolingo will seek help from some of the people you have galvanized. If you tell me about the language pairs for which you were able to find skilled bilinguals for, I may then maybe be able to tell you more. Other than that, there are really no statistics or numbers to reveal. Sorry! :-)


Wow, thanks, Baris. i reread your first comment, and still escapes me where you revealed your superhero/moderator status. :)

But THANKS is too short a word for your work to bring Turkish to Duolingo. I spread the word to a Turkish coworker, so that he could let more recent Turkish immigrants know. And I know thousands of Harmony school kids who'll love it when English to Turkish rolls out.

I also had no idea or had forgotten that j's a moderator. So some egg on face perhaps for that. The language pairs I am most heavily recruiting for are Arabic to English, Bosnian to English, and Urdu to English. I have encouraged others, too, but those are the big three for me.

* LOL, just noticed the bright green M. Okay, that was staring me in the face... *


I suppose I didn't really pronounce it too loudly myself. I was saying that people were sending "us" applications in the last sentence. And jrikhal was too humble to boast about it :-)

Your community organizing efforts are absolutely commendable. Immigrants in need and school children are definitely people DL wants to cater to. It will be closer to March when Eng-to-Tr will be ready. Do you work at one of the Harmony schools? It would be amazing if you could facilitate it so that the entire Harmony system can benefit from it. Definitely keep in touch with me regarding that. I am a part of the movement that recently founded the first ever charter-type school in Turkey as well and I am slowly but surely making it part of the curriculum myself.

I'm fairly certain Arabic will be in the next batch. And there is a good chance it will be Egyptian Arabic. In general Duolingo prefers a "universal dialect" as opposed to many different ones of the same language. Do let us know if you think another dialect of Arabic could be more suitable as the universal one. Bosnian sounds sweet, as I am of slight Bosnian descent, with my last name 'Obdan' being a Bosnian word itself :-) And Urdu is also very important that--millions of its speakers aside--it would be serving a severely underserved region and people in desperate need. I do have some acquaintances who are philanthropists in the Pakistani charter school community that I was going to contact about this. I guess you can be of more immediate help.

Hope this more or less clarifies the near future of the incubator process a bit for you. Cheers!


[In reply to your comment that Urdu is very important.] Yes, it is a "severely underserved region and [the] people [are] in desperate need." But do they have internet access? I think that Duolingo wouldn't really be helping if the targeted people can't access the material. Otherwise, I completely agree with you. I'm just curious about general internet availability in that region.


[to S7D] (You were on the 5th layer of that thread and there was no 'reply' option anymore.) Technology and internet penetration is an issue, but this is exactly the right place to start. You create the material, and then find one high-caste family to donate computers to 15 villages and the ball gets rolling. It's all about giving people the tools.


You can reach me via my duolingo id at gmail. Some of my wife's friends work at Harmony. Several cousins and family friends have kids there. And I would be happy to help in any way.

When you consider that literate Arabs of all regions understand and write in fushaa, what is commonly called classical Arabic or MSA (modern standard arabic), then I would cast my preference for that dialect-free tongue. Fushaa is also free of the pronunciation issues, especially the hard G that authentic Egyptian speech substitutes for the much softer and more French-like J sound. Beyond that, whereas Fushaa is a stepping off point from which students can delve into any dialect or classical texts, Egyptian is not so robust.

Last but by no means least: countless classical texts from Arab scientists, philosophers, ethicists, historians, and more have been scanned by Google and other preservationists worldwide. Some of those texts are in poetry, even science texts. But the prose is a trove of material that -- were Duolingo to translate it -- would be a world heritage cultural goldmine. Fushaa is the door to that trove, not any dialect.


RE Urdu Interestingly mobile access is better than hardline access in much of the country, even cities, partly because hardlines for everything: water, electricity, phone, and internet are all often held hostage by graft. While mobile access is limited by electricity, for charging, but requires less power and less stable power than fridges, tvs, computers, etc


Okay. The info on Fushaa is very useful. I have your contact info now and hopefully I'll have reason to keep in touch. It may take time tho, so let's all be patient :-) Cheers!


Will do. I can also put you in touch with educators who taught me my appreciation for fushaa over 'ammiyyah (dialects).


I'm sure one day I'll get around to learning more from the Sami languages. I'll make sure I hit you up for tips and guidance then. Thank you very much for your genuine offer. Who knows? Maybe you'll be the one teaching me Fushaa through Duolingo ^_^ Cheers!


We'll see! One more idea to pass on: since vocabulary/idiom and pronunciation are the most easily noticed differences between Arabic and its dialects, perhaps 'Ammiyyah could be supported on a Fushaa tree via "bonus skills," the way that there have been bonus skills in other languages.

So, "Traveling to Cairo," or "Rocking the Casbah in Morocco," or similarly playful while useful exposures to 'Ammiyyah for persons interacting with people of a specific country and dialect.


The problem with only adding fushaa is that many Arabs don't use it in a daily basis, and will only use their dialect. So while fushaa is definitely great for formal things (news announcements, government statements, etc.) as well as art (most of the beautiful Arabic poetry is decipherable with fushaa), it won't be very helpful if you want to actually go out and talk to people. I would suggest having a fushaa tree, and then having an Egyptian/Moroccan/what have you tree only available after completing the fushaa tree. This gives you a solid "base" for all the dialects.


How do I get to try out a course that has been released in Beta but not graduated from the incubator yet?


Head on over to your "Settings" menu where you can select those languages in the "Learning Language" dropbox.


I agree with you; some details would be nice. But consider, today we have the first week after the christmas holidays and it is monday. I think they will come back to the office and roll out the next 10 courses or so, roll out the new web design and also start the first courses from English to Russian, Turkish and Dutch. So they have already a good workload. Just be patient and let them start the new workyear in a healthy manner ;-).


I have no idea how long it would take to compile this information, though I tried to think of information that may already have been tracked for Duolingo's internal project management needs.

But there's nothing impatient about this request. The process can only go as fast as it can. Imagine how many decades we waited for Luis and team to blow Rosetta Stone off the top of the hill with a refreshing Duolingo breeze?

Will be worth the wait. I'm just tryng to avoid recruiting people who are not needed, because I respect their time as much as that of the Duolingo staff.


As far as I know, teams in place are exactly the one that you can see in Phases 1, 2 or 3.

For the other courses, no team yet selected by Duo.


your guess is as good as mine. i'd rather have data.


It's not a guess.

Duo selects ~2 moderators of the next course they want to add in the incubator. When they have two users (that applied) that answer positively to their proposition to take care of the course, then Duo adds the course in Phase 1.

EDIT (after your edit [EDIT2: well you deleted it it seems..]): Arab->English has never been in Phase 1. So Duo never selected a team for the course.

EDIT3: If Duo didn't added a course doesn't necessarily means they didn't have good enough (in term of quality/motivation/... of appliance) candidates. It's also due to the fact that they take their time --- well at least in the creation of the first sets of course by users --- to keep quality, having time to resolves all the issue that come with each new courses and that moderators report to them etc...


thanks, jrikhal. if the duolingo team likes the idea, i hope they act on it. i'd also like more clarity from the moderators/duolingo about whether prior duolingo experience is a requirement or just a good thing to have.

otherwise, i don't think i'll invite or encourage more volunteers whose applications may be ignored.


[Answer to tariqnisarahmed's beginning with "scroll back up to the original post. re-read."]

Ah, of statistics... Sorry, it wasn't clear to me if it was the idea of your first post or an idea is this part of the conversation.

Well, about the statistics:
- Name: already available -> In the incubator, courses in Phase 1, 2 3.
- Team in place: already available -> In the incubator, courses in Phase 1, 2 3. Other (future) courses still don't have team. Either because Duo don't want (for the moment) to launch a too large amount of courses in order to maintain quality and/or because they don't found good appliance (but, and just here it's a guess, I guess it's more for the first reason in many cases)
- # on team: already available -> In the incubator, courses in Phase 1, 2 3.
- # of vacancies: There is no defined final number of contributors, so not really vacancies. If the ~2 chosen moderators 'feel they need'/want more help (or maybe if Duo find it isn't quick enough, but I didn't hear about such a case since I'm into incubator --- contributing to a course) they ask Duo and one or two extra contributors will be choose, then an e-mail immediately sent to them and if they accept to join the team, then they're added directly. So, there is no vacancy since almost as soon as extra contrib. are needed they're found, but on the other hand need of extra contributors can occurred more or less at any time.
- # applicants: could be given. But it's not because there are already 1000 appliances that a new one can't be better, so it would only discourage people to propose their help and so I don't really see it as a good point to release it.
- position in applicant queue: There are no classement in the list of appliances. The best ones are kept in case of a need of new contributors, the other ones receive an e-mail thanking them and saying that they have not been selected. So if you receive nothing it's that you're still in good position if extra help is needed. - Status: already available -> In the incubator.

(1) stop filling the queue for languages that have teams in place and (2) spare our friends and acquaintances from bad experiences waiting endlessly to hear from the incubator.

- (1): as said above, new appliance can be better than the 1000 before it.
- (2): as above, if no (negative) news, it's that you're appliance could be selected. But anyway, nobody should be stressed waiting an answer. You propose your help and either you're selected either not, nothing more...


scroll back up to the original post. re-read.


What I dislike is the low number of languages available in the incubator. I could certainly help with some basics of a few languages and more advanced people could build on that, but they aren't available in the incubator list. How do I get them there? We can't help build on something that isn't on that list. Some things incubate longer than others.

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