Wikimedia Diversity Conference 2013/Documentation/roundup3 geographic diversity

Roundup-Session 3 - geographic diversity: language pluralism (best practices & new ideas) edit

A. What project has worked well and why? edit

1. Wikidata -has capacity to include many languages

2. Language Plurality Project (India) - Language plurality in India. How many exist, how many are spoken; recently deceased languages. Only 24 Wikipedias in Indic languages exist. Created a project to connect editors to work on some of these languages. Both online and in person.multilingual projects (e.g.India:780 languages), how to guarantee growth and mutual learning, e.g Start multilanguage projects to get people working on common themes, e.g. women scientists of India. Uploading sources on Meta and creating articles. Builds a sense of competition between communities. Project called Lilavathi's daughters: first female mathematician. Book was written by Indian academy of science on top 50 woman scientists.http://meta.wikimedia.org/wiki/Lilavati%27s_Daughters_Edit-a-thon

3. Swedish for Immigrants [SFI https://meta.wikimedia.org/wiki/SFI] project: Sweden was homogeneous in terms of language. More than 99% Swedish, less than 1% Suomi. Now a large population of immigrants - so hundreds of languages in Sweden: how to get them to work on their own language Wikipedias? Sweden has broadband internet everywhere. Kick-start small language communities out of Sweden, where the infrastructure is. Immigrants are being taught Swedish, so they can interact, but as part of the classes, they can use WIkipedia as a tool. Translate articles about Swedish culture to their own languages: also make it easy for immigrants to learn about the culture. 3 teachers, 10 signed up, but volunteer. Second group had 18 students. So far, Spanish, Thai, Arabic, Somali. Has a lot of work to be done. Some Hindi now.

4. Status quo of Africa: also diversity of languages. Also faces the problem of orality: how to gather this knowledge into Wikipedia. Jo'burgpedia: Outreach event, created one article aboutSatyagraha House in JHB attracted about 8-9 attendants, and had 6-7 language speakers. Goal was to translate the article about the house into local languages. Clear goal, so it was done successfully. Markup language was a stumbling block, as was understanding of the user-space; how to add images. Had experienced editors to guide volunteers. Joburgpedia has 12 sites so far, so the idea was to do this on each site: create English version; then arrange a reachout event where small-language editors would translate that into their languages.

Editing in Xhosa: Sinenjongo High School, Milnerton : attempt to re-invigorate Xhosa Wikipedia through the students. Get them to understand that WP exists in their own languages. They soon realised that Xhosa Wikipedia pages were bot-created, some had bad vandalism, and one specific editor managed to improve the article on Nelson Mandela in Xhosa: it said nothing about him, and only spoke about sports. She wrote some information in isiXhosa about Mandela. Not a large article but an encouraging start.

Translating help pages? No: the opportunity caught us unawares. Have seriously thought how to assist the children, e.g. with templates etc. At least now they know they can edit it.

5. South America: Children to indigenous languages (guaraní and quechua), 12-13 year-olds

Not a great resource yet, but encouraging results so far. Started with gramatically correct editing. Created articles in Spanish so far, and prefer that. Between 10 and 15 children out of 200 want to edit in the indigenous languages.

6. China Tibetan Wikipedia - Not so good: Tibetan teacher was curious to start Tibetan Wikipedia. Tried to edit, run the Tibetan Wikipedia, but he got overwhelmed and isn't working on it any more. Professor from Shanghai working on Classical Chinese-Zhuang typing system but no-one is using on this any more. No-one uses classical chinese drawings anymore, so not going well. Minority languages: some working better than others. Most not successful.

B. WHY did it work well? edit

1. Wikdata:

Conversation and search interface

2. Lilavathi's Daughters:

craeate common pool of resources, structure for projects, used social media, acknowledgement of editors, appreciation (thanks for new articles), Template created for user pages on each language project.

3. Swedish immigrants,

Pool of resources on meta that can be used by other chapters: e.g. mentor list. Institutionalising it with teachers. How to add to learning plans etc. Scaling well. Win-win situation for teachers and students and community.

4.1 Jo'burgpedia:

outreach keep it simple clear goal./expectation from volunteers Translation as an entry point.

4.2 Editing in Xhosa:

Let the kids do what they wanted to. Tell them they "can't break Wikipedia - go wild!"

Tell them that it exists: leave it to them.

Keep it fun. WP:IAR

5. Teach to translate into quechua and guaraní:

Tell the kids that there will be no test. Do it if you want, at your pace.

No pressure.

Enthusiasm of volunteers!!

Scalability: It's contagious. If 10 participate, the rest want to join, but if only 2-3 do it, they're seen as the weird ones.

group identity (in-group)

C. What new ideas result from the conference? edit

Create a label-a-thon: language-related project with Wikidata e.g. Lilavathi's daughters labeled in many languages.

From Siko's presentation: Invitation aspect: invite people specifically.

Show people who's involved.

Ask for help from WIkipedians /Wiki[p|m]edian expert/language expert with the projects

=== D. Who likes to be the contact person for people who want to get more involved or like to adapt the project in their home country? Everyone in this room: ===

  • David Richfield
  • Dumisani Ndubane
  • Addis Wang
  • Gerard Meijssen
  • John Andersson
  • Martin Rulsch
  • Andres Maggese
  • T Vishnu Vardham
  • Kwaku BBM
  • Awula Serwah
  • Cornelia Trefflich

E. What future projects should be started and who likes to take part? edit

Encourage contributions in the Twi language (Ghana) ( part of the Akan language family)

Wikidata-based Commons search

Wikispeech: Encourage any person with mobile connectivity to record any content (e.g. stories, lullabies, stories, etc.) Note: Might connect to Google's similar project