There seems to be a lot of language sites not surviving: American Sign Language, Arabic, Aramaic, Bangla, Bulgarian, Catalan, Czech, Danish, Dutch, Greek, Finnish, Hebrew, Hindi, Hungarian, Indonesian, Irish, Jamaican, Lojban, Malayalam, Malaysian, Norwegian, Persian, Polish, Romanian, Samoan, Sanskrit, Swahili, Swedish, Tamil, Thai and Lao, Yiddish; Baltic Languages, Indian Languages, Nordic Languages, Semitic Languages, Slavic Languages, Turkic Languages; Translating and Languages; and probably more. Some, like Greek and Arabic, have been proposed multiple times.
What a waste of expertise...
Perhaps a solution is to create a Languages.SE (or Polyglot.SE) site. (This idea has gone through a few mental revisions---I'll tidy it up.)
1. Unique site feature: language label.
Each question has a unique, required language label. The community will have to work out how to do this (but we've figured out other complicated boundary issues, e.g. the religion sites). These should be chosen based on what would make a good stand-alone site topic.
(I originally suggested the 5 usual tags be distinct for each language, but maybe that's not necessary. The language tags tend to overlap heavily (translation, grammar, meaning) that it's probably unnecessary.)
I currently feel we should require diamond-moderator approval before creating a new "language label". This needs to be done with the utmost care.
It should be possible to "favorite" and "ignore" languages in some way.
2. Synchronized with existing language sites.
Each of the non-English language sites (Chinese, Esparanto, French, German, Italian, Japanese, Korean, Latin, Portuguese, Russian and Русский язык, Spanish, Ukrainian) are synchronized, in the sense that questions at these sites pop up on Languages.SE.
If you click on a question about e.g. German, you're redirected to the German site.
Why? (a) Many people have interests and expertise in multiple languages. (b) Languages overlap (off the top of my head, Chinese-Japanese-Korean and Urdu-Persian-Arabic).
3. Ability to ask about languages without an existing language-specific site.
At Languages.SE we can also ask local questions about Arabic, Malaysian, Hebrew, etc., for which there does not exist a unique site. We add the language label to categorize them.
4. Potential to launch successful sites.
When e.g. Vietnamese posts are sufficient to demonstrate it's viable as a standalone site, a separate Vietnamese.SE can be created from the questions with the Vietnamese label.
Splitting a site will split users reputation into Vietnamese.SE and Languages.SE, but leaves Languages.SE badges unaffected.
The questions now labelled Vietnamese are now redirect to the newly launched Vietnamese.SE, so there's no major seam.
The biggest problems I (currently) envisage...
Moderation and curation
Moderation and curation seems to be a major challenge with this proposal. I feel like there needs to be a hard rule here: local questions are strictly in English only. Why?
- It facilitates moderation, and allows users to cast educated votes.
- Expertise comes from polyglots, so I don't see this as a major inhibition.
- English is the default language of Stack Exchange.
- Sites "launched" from Languages.SE can subsequently admit questions in their language.
I can't think of a realistic way to allow non-English questions, while having them responsibly moderated and voted upon.
What's a language? What's a dialect? There are also questions [e.g. etymology] which straddle the boundaries between languages (but this is already a problem).
However, I feel like if this is done carefully in discussion with the community, it would be do-able. Proof-of-concept: Linguistics.SE has language-specific tags (Arabic, Turkish, etc.), and they're not constantly bickering about boundaries.
Perhaps a site running multiple scripts for multiple languages would result in an unresponsive site. Maybe additional scripts are best left for "launched" sites.
Many languages can already work in today's browsers:
- لكن لا بد أن أوضح لك أن كل هذه الأفكار
- קצרמרים החופשית לויקיפדים בדף מה אנציקלופדיה צעד
- สไลด์ชาร์จเซ็นทรัล แฮนด์ไฟลต์ปาสกาลเพาเวอร์ ไพลิน ดีลเลอร์นู้ดเซ่นไหว้ยะเยือกบัส
- chỉ đơn giản là một đoạn văn bản giả
- jest tekstem stosowanym jako przykładowy wypełniacz w przemyśle poligraficznym
- छपाई और अक्षर योजन उद्योग का एक साधारण डमी पाठ है
- საბეჭდი და ტიპოგრაფიული ინდუსტრიის უშინაარსო ტექსტია
I found some similar ideas Merging All the Language Proposals into one language website (2011) and What are the advantages of single language sites over collective ones? (2016). I also found Combinining all language related sites into one (2019) on meta.SE.
But I'm not proposing to eliminate or exclude single-language sites, but instead create a supersite: i.e., we have both.
What this fails to address is that many of these sites either require or at least allow questions in the site language (e.g. French or English on the French site), making a composite site particularly confusing for everyone involved.
The major problem with a multilingual site where most users won't speak all, or most, of the languages being used is one of moderation and communication.
Hence why it seems necessary to have the hard rule "local questions are in English".
- Often, languages are too far away from each other to be lumped sensibly
- Large languages will dominate over small ones too much
Is the first point even a problem? There are polyglots who are interested in languages in general. Moreover, Linguistics.SE exists and some of its questions requires expertise in multiple languages.
The second point is an issue at virtually every SE site (e.g. Minecraft dominates at Gaming.SE): some tags are used far more frequently than others. It's why we have indexing (i.e., tags).
If you make English the sole languages allowed on meta, then that'll alienate every non-anglophone on the site. Native English speakers will have a lopsided influence on the site.
English would need to be the sole language allowed
on meta for asking questions on the main site, and for posting meta questions. I do not see why this would "alienate every non-anglophone" more than other English-only sites (not having a site = 100% alienation). Realistically, "active users" have a lopsided influence, regardless of their native language.
I took some ideas from the My Filter page and combined them to give:
Question: Would a general Languages.SE supersite (synchronized with the current language sites) be feasible?
I'm still not 100% sure this is a great idea, but I'm yet to think up a "dead on arrival" problem. I currently feel that if people wanted to make it work, they could make it work.