Thank you all for your helpful suggestions.

I realized that translating from English instead of the other way around is certainly a better way to approach this.

Using ContentTranslation seems the best option, even though it currently does not contain support for auto translating the en/cs language pair.

It is important to me to be able to see somebody else's unfinished translation and as far as I know, ContentTranslation does not support that yet. Is this feature planned?

Making subtitles or translating subtitles has the advantage that the student can work on small chunks each time. A subtitle translated, a minute of video transcribed. It does not feel so daunting a task as translating a whole article may be. It seems to me a good idea to add this activity to the course. It gives students an opportunity to work with video content, which is something the course currently lacks.

I am not sure how much work it is going to be to oversee the students, though. I worry it would be too much for me. I would probably have to pick one of either subtitles or articles and focus only on that at first.

I will think it through some more, talk to the course coordinator (I am just a lowly teaching assistant, you know ;) and to the people at Czech Wikipedia who got in touch with me regarding my first e-mail to the mailinglist and we will figure something out. Thanks for your help.

Finally, I apologize for this thread's meaningless subject line.

That is my tl;dr response. Before I put it together, I responded inline to all your messages:

On Thu, May 21, 2015 at 6:10 PM, Leigh Thelmadatter <osamadre@hotmail.com> wrote:

I very much do recommend you contact the Czech education people, who are a great bunch and can give you invaluable hands-on support.

I got a private response from one of them. There apparently is somebody working at the University's rector's office I can talk to about this.

In my experience, I have found having students write new text in their non-native language to be extremely challenging, and you have to be sure that students are up for it. Translation gives the basic structure (a +) but it also has problems with L1 interference in L2. (and vice versa but particularly problematic for L1--> L2)

The articles which cs Wiki has and en Wiki does not tend to be about Czech and Slovak geography, municipalities and people. The language does not tend to be very convoluted for topics like that. I see your point that translating into a foreign language is rather difficult. There is about 1200 students taking the course each semester, though, and finding a couple who would be both willing to participate and up to it should be possible.

Translating into one's native language is probably a better proposition. It is also much easier on me, the person who has to check everything that students hand in. Every year, there is about 50 people in the course whose native language is not Czech or Slovak (med students from the Near East and Erasmus exchange students, mostly). I would have to find somebody else to check their work if they decide to participate in this activity.

1) Have students review articles in English on Czech topics for inaccuracies and/or out-of-date information and/or missing details or citations. The Visual Editor tool has made article improvement a bit easier, especially the addition of references.

From my experience, giving students too much freedom does not work. If I just told them to go looking for low-hanging-fruit edits, they would probably ignore that activity. The only people who may participate are existing WIkipedia editors who would use it to get points for something they are already familiar with. Furthermore, there is the question how many points should they get for their work and what the work actually constitutes. Should they get points for telling me that they did not find anything, that a page is all-right?

2) Wikimedia Commons (commons.wikimedia.org) has videos in English that need subtitles. One teacher at my school Karen Mazanec, had students create English subtitles for English video as intensive listening practice. https://outreach.wikimedia.org/wiki/Education/Newsletter/April_2015/New_to_Wikipedia:_A_personal_perspective

This sounds interesting. I toyed with a few ideas about subtitling videos myself. My biggest hurdle here is teaching students to use any software they would need. I planned to use Amara.org as a subtitle editor, mainly because on-line services tend to focus more on user friendliness and streamlining than desktop apps. Amara can subtitle any video, as long as it is possible to directly link to the video file. I guess it should be possible (technically and legally) to use Amara for Wikimedia content, but I have to check.

Interesting to to get your mail today as I had a meeting where they are talking more about modualizing (not a word, I know) courses. If you could send me a link about your course at Masaryk, I would appreciate it greatly.

Is it supposed to mean the same thing as modularizing? ;) I'll write you an e-mail about how is it working for us.

On Thu, May 21, 2015 at 6:27 PM, Kavya Manohar <sakhi.kavya@gmail.com> wrote:

Hi,

If you are planning to introduce wiki editing to your students by translating articles, I suggest you to try out the content translation tool.
See https://www.mediawiki.org/wiki/Content_translation.

It provides an easy interface for translations. The users need not bother about the wiki text formatting, rather can concentrate on the content and the language.

I know the tool. I was one of the people who campaigned for having it enabled on cs.Wikipedia ;)

On that occasion I made a comparison table for all available translation tools for WIkipedia I know of (Duolingo, Google Translate Toolkit, ContentTranslation) and CT is a clear winner. Duolingo has problematic ToS and drops formatting and references, Google Translate for WIkipedia is not being actively worked on and drops some formatting and references. I guess as the developers you know all about that ;)

What I like about Duolingo is that the translation unit is a sentence, not a paragraph. It is much easier for a language learner to work with individual sentences first and focus on whole paragraphs only later while revising. Having to translate a whole paragraph at once does feel overwhelming at times. What Duolingo users tend to do is to nibble on easy sentences and gradually move to translating more difficult ones. This is a workflow that ContentTranslation does not support. (I am not saying that it should support it. There are strong reasons for preferring paragraphs, too. I remember reading something about that on CT's page.)

On Thu, May 21, 2015 at 6:33 PM, Amir E. Aharoni <amir.aharoni@mail.huji.ac.il> wrote:

Indeed :)
When do you plan to do it? I'd ve happy to give you tech support with this.
(Disclaimer: I'm in the team that develops ContentTranslation.)

October 2015 at the soonest. November is more likely.