Talk:2019/Grants/COVID-19 translation

Willingness to collaborate
I proposed this project in a certain way centered where I am at the School of Data Science at the University of Virginia. However, I am willing to join anyone else who wants to develop or share COVID-19 content. I am especially interested to make this a multi-team effort. Teams which especially appeal to me are Wikimedia Medicine, the Translation Task Force, SWASTHA, WikiProject Humanitarian Wikidata, Wikimedia New York City, and Wikimedia Sweden, because all of these organizations do health translation in Wikipedia.

If other organizations would join, I am ready to defer to any common plan and participate according to that.

Having this money come to the University of Virginia is not urgent. If anyone else can manage money to do a COVID-19 project then I can support that. I wish that the sponsorship could encourage collaboration, and will support whatever plan seems best to make that happen.  Blue Rasberry  (talk)  00:20, 7 April 2020 (UTC)

Criticisms
I have received some private criticism of this proposal. I will respond to that here. If anyone wants to discuss further, then speak up!  Blue Rasberry  (talk)  00:34, 7 April 2020 (UTC)

Will you pay translators to produce content to publish in Wikipedia? Paid editing is taboo in Wikipedia.
This project is too complicated to plan far in advance. If paying translators makes sense then I support it.

In some cases, especially for underserved languages in the developing world, there is not a pool of volunteer Wikipedia labor. There is no way to use money to quickly recruit volunteers to do translation, or if there is, the cost of recruiting a volunteer is higher than the cost of paying someone outright for translation. Whatever process this project uses will be transparent in the usual Wikipedia way.

One possible process is this:


 * 1) start with English language content
 * 2) use translation consultant to produce first draft
 * 3) seek review of that draft from existing Wikipedia community, who cannot do the translation but do find it easier to review an existing draft
 * 4) if recommended, finally seek review from native speaker who also is a health care professional and who can review technical terms and phrases

 Blue Rasberry  (talk)  00:34, 7 April 2020 (UTC)

Why not use medical professionals for everything?
Ideally all of Wikipedia should be written by experts. Wikipedia does not have the goal to encourage everyone to edit, but instead, letting everyone edit is just a consequence of not having resources to get professional experts to produce perfect content in every language. It would be ideal to have experts produce first drafts of everything then let volunteers critique or revise those first expert drafts.

Unfortunately history has demonstrated that we cannot recruit volunteer experts to write all of Wikipedia. Also it is too expensive to pay them.

This is a fairly inexpensive project by United States standards, and this budget is insufficient to pay health care experts to produce all the desired content. Instead, this project plans a mix of some strategic expert review, and some Wikipedia volunteer review, and some staff administration to document what we did to be ready for future disaster response on Wikipedia. This is an attempt to have the response that is cheapest and best.  Blue Rasberry  (talk)  00:34, 7 April 2020 (UTC)

Seeking community review
Anyone with any comments, criticism, or support can post here. Anyone is also welcome to say "I do not feel this model for translation will be effective", and further either give any or no explanation. Thanks.  Blue Rasberry  (talk)  00:36, 7 April 2020 (UTC)
 * Support Happy to be involved in drafting the initial content. Doc James (talk) 20:01, 8 April 2020 (UTC)

Responding to COVID-19 in June 2020
COVID-19 shut down the world in March 2020 and the Wikimedia community responded as described in "Wikipedia on COVID-19: what we publish and why it matters".

Now in June 2020 people are still developing COVID-19 content, but the environment has changed, as the initial volunteer drive to develop content has slowed and the development of Wikipedia content has become complex as various language projects diverge more from each other. There are 131 language versions of the "COVID-19 pandemic" Wikipedia article, which seems like a lot because this is the world's most largest collection of language versions of this topic. However, there are still language communities online which are seeking information and who either do not have basic COVID-19 articles available in their language, or where the articles exist but only barely and without key messages.

This project seeks to translate information into those underserved languages, which will have the dual effect of providing information on this pandemic which has no end in sight, and also by setting a precedent and model for getting information to these language communities in the case of the next crisis or disaster.  Blue Rasberry  (talk)  21:38, 19 June 2020 (UTC)

Administration
This project was originally proposed in March 2020 to include some administration labor. On talking the situation over with the funders, they explained that they are able to support labor costs against invoices. This works for our school's researchers to reduce our administrative costs because we typically manage larger projects with more overhead, and setting up this project with outside help to send payment to translation contractors simplifies the process we have in place.

Something else that has changed is that since the advent of COVID, our department has begun to develop other crisis management processes in Wikipedia. While this project proposed here will develop COVID-19 content in languages which we otherwise would not address, we can share some of the team structure, quality control processes, and design from the other projects to get this one off to a good start with some different organizers.

I took admin costs out of the original budget which leaves us to only fund labor for production in this proposal. Even without the admin costs here, I think we can get more effective outcomes because now several months later, we know more about what a COVID-19 response should be in general and in Wikipedia and as part of an effort to address other similar disasters.  Blue Rasberry  (talk)  21:48, 19 June 2020 (UTC)

Languages chosen
When I drafted this in March I was unsure what languages to target. I wanted languages from a small language Wikipedia community because there are fewer precedents for working with these, but they are still important.

I wanted languages which students study at the University of Virginia. This school has 25,000 students and a proportionately developed language program, so there were lots of options. After asking around, I got the most interest and some leads on research organizers for Tibetan and Nepali languages.

I think these are great choices for a small language Wikipedia pilot and that this fits with what our school is ready to do.  Blue Rasberry  (talk)  22:17, 19 June 2020 (UTC)