1. Forum
  2. >
  3. Topic: Duolingo
  4. >
  5. How much work to create a new…


How much work to create a new language course?

I would like to somehow initiate the making of a new duo language course (English to Northern Sami). I do not speak the language myself, and therefore I cannot apply to be a course contributor. However, I would like to spread the word and look for people who want to contribute, as there are quite a number of Sami speakers who are committed to the conservation of this indigenous language.

My question is, how much work will it be for the volunteers involved in creating a new language course? How many hours a week? How long does it take before the course is online? How much work after the course is up and running? How many translators are needed?

I haven't been able to find this information, so I hope some of you can enlighten me on how much work it is to make a new language course.


December 20, 2017


[deactivated user]

    I don't know, but A LOT


    I think that if you have a different alphabet, then it would take longer (for example, Hindi and Arabic.) I'm assuming that a course would take at least 6 months to 2 years with a good group of contributors, and a Latin-script language.


    Ahh, yes I guess so, I wish I had some numbers! It can't be impossible...?


    Having the statistics is not the problem, having the volunteers is.


    I want the statistics to let volunteers know what they are going to :)


    I don't know hard numbers, either, but there is this article in the unofficial Wiki that talks about all the work that has to be done: http://duolingo.wikia.com/wiki/Course_contributor_guide

    It's a cool article, with screen shots and links to a good discussion or two.

    If it's the first time a language has been taught FROM, then there's additional work to be done. For example, if you are teaching English for Sami speakers, and there are not yet any courses for Sami speakers, then a lot of work has to be done translating the basic Duo environment into Sami. All the menu options, the "Post" and "Cancel" buttons for discussion creation, the "Reply" and "Give Lingot" prompts under comments, the "Type what you hear" and "Select ALL correct answer" prompts, the online help, the descriptions for the app stores.... just tons and tons of work.


    Thanks! Very helpful article! Yes, so a Sami course for english speakers will not be as much work as the other way around.


    Looking at previous languages, it often takes up to 2 years, for 6 expert volunteers to make a course.

    However some must take a lot less (especially reverse trees), since most of the material of reverse trees is directly taken from the original tree.

    Different courses also have different numbers of words.

    It'd be useful for someone who actually works for Duo to respond to this query though


    However some must take a lot less (especially reverse trees), since most of the material of reverse trees is directly taken from the original tree.

    No, most of the material is not taken from the original tree. The structure of the tree teaching X from Y is based on the needs of Y speakers learning X. This pertains to what words are needed where, what grammar concepts must precede which other ones, etc. When you turn it in reverse, you will find that the speakers of X learning Y need to be presented with different linguistic features. So the structure of the reverse tree then matches those needs.

    The simple key to what I said is that every new word being introduced in a lesson typically needs to have three sentences defined for it, and those sentences (other than that new word) can only use words defined in the skills and lessons the user must have already gone through.

    As a consequence, reusing the material from the opposite direction is actually a big challenge that is best avoided. Sometimes it happens spontaneously, but most contributors are very unhappy when that happens.


    I wonder how many volunteers is ideal. Is it best with a smaller group of f.ex. 6 volunteers, or the more volunteers the better?


    The number of volunteers in the incubator is limited. Usually when they add a new course, only two people ("course moderators" who overview the whole process) are admitted in. When they've done a bit, they get "invites" they can use to bring more people in (and they can choose them from the other people who applied for the course.) The number of invites is limited, so the team will never be a big one. I'm not sure how many people exactly you can get in, but around 6 looks fine. If you have more people on the team, they will not be in the incubator, but maybe they can help with more general stuff or be ready to replace someone who needs to leave.


    Thank you! So two lucky volunteers get to be course moderators :)


    yeah but it would by a lot of work for some one


    I would imagine it would depend on the language, the number of volunteers, how well they get on with each other and whether they hit on insurmountable problems or not.

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