What is meta? ×
Meta Stack Exchange is where users like you discuss bugs, features, and support issues that affect the software powering all 133 Stack Exchange communities.

It seems all the non-English language-community proposals on Area 51 are having a hard time getting off the ground (1, 2, 3, 4, 5, 6, 7, ...). Even iRosetta, which was the first site of this kind, has essentially been reduced to a two-man show and seems pretty dead in the water (go go, Peter and Donald! ;)).

Is the Stack Overflow model simply not very suited for language discussion communities? Or is there another problem, like missing localization or too little overlap between linguists and technical-minded people? Or is the community as such too small to reach a critical mass?

What are your opinions?

share|improve this question

5 Answers 5

Fragmentation doesn't help. Professional software developers saw Stack Overflow because Joel and Jeff talked about it and everybody was happy. There is not a population of Japanese language students who are waiting for a Japanese language Q&A site to be opened to contribute on it. They simply don't know it.

And they don't need it, either (or at least they won't think that they do if they see it). For learning a specific language, there are already excellent resources to trust: books, tutorials, etc. And I also guess the number of self learners is lower for languages than for software, because (1) it is much more boring to memorize information about a language on your own, (2) conversation practice is essential and this enforces the model of physical groups with a teacher.

Related to that, there is also a trust problem. For software questions, thanks to testing, if you tell me that it is ok to dereference a NULL pointer in C++, I will find out that you were wrong. I can initially trust your advice to build my solution, but I know the testers in our team will verify everything, and prior to that I can look at it myself and see whether what you said was true. But what happens with language? If someone says that "どうもありがとう!" is impolite, I cannot verify it by myself - I can only cross-check other sites. This makes the Q&A model less reliable in this case.

share|improve this answer
    
So you're essentially saying it's a problem of advertising? I'd agree with that. I do think though that there is a need, since teachers or natives aren't always at hand, but especially with language you usually want a question answered as soon as you come across it. And BTW, the Japanese page seems furthest along so far among the ones I mentioned... :) –  deceze Aug 18 '10 at 8:09
    
@deceze: I updated my post. Now it refers more clearly to several reasons: "know" (advertising, as you say), "need" and "trust". –  Daniel Daranas Aug 18 '10 at 8:28
2  
Interesting point about trust. But then again, there're many language communities on the web (e.g. the WordReference forums, Leo.org) which essentially deal with the same issue, successfully. And SO reputation is meant to address this point, as is the voting. A site that has a reasonable number of active users should be trustworthy, since it's unlikely that everybody is giving you false advise. –  deceze Aug 18 '10 at 8:33
1  
@Daniel re the trust issue, that is exactly where the voting system comes in. On a half-way decently frequented site, false information will be pointed out quickly, just as on Stack Overflow. That is one of the main points why I think the SO system would work great for language communities. (update ah, @deceze was quicker) –  Pëkka Aug 18 '10 at 8:33
1  
+1 for trust, good point. I've asked an English question and got 4 answers, but what are the credentials on them? I was hoping for some references, not "it looks weird" or "I wouldn't do it". I just can't accept an answer... –  Kobi Aug 18 '10 at 9:27
    
@Kobi You will probably always have this issue with language questions, since language is an evolving, arbitrary thing and an "official" language is mostly popular opinion. Hence it's hard to give an authoritative answer to a language question, except for citing more or less official sources (which was done in your case). That doesn't mean that there's no value in language Q&A sites due to trust issues. Look through the WordReference Forums, there are great and enlightening discussions to be found there. –  deceze Aug 18 '10 at 9:45

I think the potential and need is definitely there. The basic idea differs a bit from the sites you quote, but my English-German translation proposal was inspired by the Forums on Leo.org, a German-to-many online dictionary. Those forums are extremely well frequented.

The SO format would be perfect for language and especially translation questions (where nuances, style, subtle cultural differences play an important role.) - even though many questions would end up as polls.

As to why language related sites are not taking off, I think the main reason is that the topics don't overlap much with the programming community that is SO, so there is not much of a "native" audience driving them.

The fact that there are no localized versions of SO available doesn't help, either. How can I build a "xyz language" community when the interface, and commitment phase, are english only?

In my own case, there is also great reluctance to go and advertise my proposal. I'm an enthusiast, but have not many translation-related contacts I could make aware of this. And I'm not going to barge in in some forums and advertise what is effectively competition aimed at taking away loads of traffic from them: It would feel inappropriate.

share|improve this answer
    
I know what you mean, I was tempted to try something like this for iRosetta, but also felt it would be quite inappropriate. So it seems what Area 51 sites need is a better way to advertise. Any suggestions...? –  deceze Aug 18 '10 at 8:28
2  
+1, I'm pretty sure that overlap is the problem. How many SO/SF/SU are 'bilingual' French/English? and of those how many read the blog or are on meta? and of those how many have made it to Area51? and of those how many have bothered to click through to page 9 (currently) to see this? area51.stackexchange.com/proposals/1979/…. I've got some top-notch users on there, but critical mass is a long way off... –  Benjol Aug 19 '10 at 5:10

Is it worth trying to combine all (or most) of these sites and create a generic translation Q&A with specific tags for different languages?

It might sound crazy, but different languages works for Stack Overflow. =:-)

If you think it is a good idea, please let me know here and I'll start it (unless someone beats me to it).

(This will solve the first two of the three issues hightlighted in Daniel Daranas' excellent answer - and the trust issue will also be improved because of the many eyes (reviewers) principle which you will get with an increased critical mass of contributers.)

Update: I started a site (see link below) but then deleted it when Pekka pointed out that such a site already exists: Translation. May I recommend everyone interested in this topic turn their attention there?

Update 2: Today there are two new sites:

  1. German
  2. Portuguese

So we really need some clarity here. I'll flag this to get the moderators to intervene.

share|improve this answer
1  
I still think languages should be separate because they will be separate communities, but maybe a combined site is the way to go to reach critical mass. If there isn't one yet, why not start one! I'll follow. –  Pëkka Aug 18 '10 at 10:26
    
OK, here we go: area51.stackexchange.com/proposals/18281/ask-babelfish –  Wikis Aug 18 '10 at 10:43
    
If you want to support this, please contact the other existing language Q&A sites and communities and invite them to pool their resources here. –  Wikis Aug 18 '10 at 10:47
2  
I cannot agree with this: multiple language works for SO only because most of the users knows a great many language (even if they are skilled in only one or a few), therefore most of the questions can be of value/interest to most of the users. Multiple languages, on the other hand, are rare. Even bilinguals are rare! In addition, the history of computer language means that many of them share common features, while with languages even similar ones are difficult for non-speakers to decipher, thus making most questions useless for most users. –  Yi Jiang Aug 18 '10 at 12:53
    
@Yi Jiang - to some extent I agree - computer languages share common features (e.g. loops, variables) but then so do human languages (nouns, verbs). SOOooo - if you don't agree, what do you propose instead? –  Wikis Aug 18 '10 at 13:32
1  
I'm afraid that a site about all languages in the world may indeed be much too broad and not attract the attention it needs from the experts it needs. A real problem would be that you'd only be able to understand a fraction of the questions, if they all deal with completely different languages. That'd make moderation virtually impossible. Also, there's no place for language specific grammar or culture questions on a "Translations" site, which greatly diminishes the value of such a site. I'd say a site needs to focus on a particular language or group of closely related languages to work. –  deceze Aug 20 '10 at 8:14
1  
@deceze I agree about the problem of moderation. You would need a critical mass of people. Also agree with broadening to culture. The translation Q&A already exists, btw. But maybe, like Wikipedia, it is counterintuitive and we just need to try. At the moment, the independent langauges sites success level is variable at best, with most failing badly. –  Wikis Aug 20 '10 at 11:36

Of those, most of them are languishing, but one, the Japanese site, which I'm a member of, is actually nearing commitment. We need eight more followers and just a few more votes for on/off topic questions to get to commitment. Getting through commitment is a tougher thing unless the requirements in terms of reputation are lowered. The problem is that the commitment phase requires lots of people with reputation points on stackoverflow to join, whereas the people who I would like to see join the site may not even be programmers. In fact I would be wary of a language-learning site dominated by programmers.

share|improve this answer
1  
then the problem is to find around 300 people willing to commit to it. –  Stefano Borini Aug 18 '10 at 11:30
    
ouch! I proposed area51.stackexchange.com/proposals/14570/… and it adds a new follower every day or so. it's got a long way to go but is that languishing? –  hawbsl Aug 25 '10 at 22:05

I think it would be a better world and there would be less headaches if everyone spoke only one language! English would probably be the best candidate.

share|improve this answer
1  
Apparently it's been tried and wasn't a great success: en.wikipedia.org/wiki/Tower_of_Babel =:-) –  Wikis Aug 18 '10 at 10:40
1  
... or everyone believed the same religion (no religion), worked the same hours and shared their wealth, leaving all differences behind. Sounds wonderful, only it led to fascism whenever it was tried by more than 50 people... –  Kobi Aug 18 '10 at 10:44
1  
@Kobi : I think you don't really understand what fascism was about. –  Stefano Borini Aug 18 '10 at 11:35
    
@Kobi: also you are making connections that aren't really related and don't make much sense. –  Andreas Bonini Aug 18 '10 at 11:49
1  
You wouldn't be saying this had you learned a language sufficiently different from your mother tongue. There's much culture ingrained in every language that would simply vanish would we all speak the same. –  deceze Aug 18 '10 at 12:20
    
@deceze: I'm pretty sure Kop is Italian, actually. –  mmyers Aug 18 '10 at 20:36
    
I don't want to die of boredom. Flavor in the world is a good thing. –  Peter Ajtai Aug 18 '10 at 20:54
2  
@mmyers Italien and English are not that different. Try an Asian language. :) –  deceze Aug 18 '10 at 23:16
    
Akkor már miért nem rögtön magyar? –  Ákos Kovács Jun 30 '11 at 12:39

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .