It may be that MO is not in need of publicity. The AMS Notices article may have sufficed.
But some here might be interested in what CS Theory, which is definitely in need of publicity, is doing: gathering "best questions."
http://meta.cstheory.stackexchange.com/questions/457/nominations-for-publicity-poster-at-focs

@Willie: June/July 2010:
http://math.stanford.edu/~vakil/files/mathoverflownotices.pdf

In fact, I tossed around the idea of doing an community wiki AMS-Notices-style article for SIGACT news/EATCS for the cstheory site. Now I'm encouraged to explore this further.

@grp: What do you mean by "internationalize"? MO has contributors from many countries.

Oops. I was thinking about multilingual websites, but not saying so. In this case, internationalize is in the "web-site" sense, which is to convey certain aspects of the site in several languages. At the very minimum, it would mean having the FAQ and how to ask pages in several languages. It might mean having mathoverflow.xyz domains, where xyz is a string standing in for .de, .in, .cn, and other international domains. There are other possibilities, but my question asks about requests to extend the site in any fashion related to (geographical forms of) language and culture.

Gerhard "Ask Me About System Design" Paseman, 2010.10.18

grp, I believe the answer to your first question is, "basically none". There was some discussion about language policy when Laurent Fargues wrote an answer in French, and I think the consensus was that we should demand questions in English and allow answers in other languages.

It is far from clear what anyone would gain from having alternative top-level domains. Do people still type full URLs nowadays?

@grp: Ah, okay now I see what you mean. As a "crowdsourcing" site I suspect MO would only diminish itself if it fragmented into language specializations.

Dear Andrew,

Very nicely put!

Best wishes,

Matt

English is the de facto language of mathematics for good reason---because mathematics is an international endeavour.

... and we English-speaking mathematicians can't be expected to manage other languages?

Personally, I don't see why people shouldn't ask questions in other languages. Of course, by doing so they will knowingly be limiting the pool of possible answerers, but surely that's their choice?

Of course it is their choice, but keep in mind that the question is not just for them---there are many others who could potentially profit from their question. By asking or answering a question in their own personal language they limit its usefulness to those who speak their language, thus shutting out the rest of the community of mathematicians. I definitely hope that people are free to use their own language if they want, but I wouldn't particularly want to encourage it. In my view, mathematics wins when our community has a shared language, so that everyone involved can participate and communicate together, but that doesn't mean I would want to force other people into conformity.

Ryan makes a good point too.

@grp,

Unfortunately, there are no such plans. For one, the lack of control over our software would prevent us from even getting started. If at some point we switch to some entirely new software, and this becomes possible, I'd be happy to help make it happen if others did the heavy lifting of translation work.

In the meantime, I don't particularly mind if people ask questions in other languages, as long as they are aware that this may reduce their readership. As we're in absolutely no danger of having "too much" foreign language use, it behooves us to be welcoming.

Dear Ben,

Your recollection agrees with mine. Also, one might note that already there are several comment exchanges in Russian (and this is the only language in which I'm aware of there being multiple such exchanges).

Best wishes,

Matt

Dear Noah,

Of course there are also native speakers of other languages (say Chinese) studying in France who learn French but not English (or at least for whom English is yet another language to have to learn, having already learnt French).

Best wishes,

Matt

Those Chinese speakers in France will have to learn English as well.

I think that was the point Matt was trying to make. For a Chinese student studying mathematics in France, it is possible for him to get by without learning English at all. If he further does not intend to publish in journals other than those in France and those in China, he does not have to learn English. While it may be a good idea for said student to indeed learn English, your (apparent) stance seems to remove the decision from the individual and prescribe that everyone must learn English, which is a stance I am rather uncomfortable with.

In other words, you, Shevek, may choose not to interact with anyone who doesn't speak English if you so please, but if I see a Chinese student whose English is practically non-existent, I can choose to interact with him in Chinese; ditto if I see a Vietnamese student who speak French, and whose English skills are unfortunately poorer than my limited French skills.

I know you wrote earlier that you don't want to force other people to conform. But some of your follow-up posts have the connotation, if not the denotation, that is quite the opposite.

---

I agree in general that we should encourage the use of English as a common language on MO. But I think that's the degree it should be: an encouragement. Users should be made to understand that if they do post questions/answers in other languages, it may not reach as wide an audience. But they should not be barred from asking/answering in foreign languages.

I should have read the label on this can of worms before opening. While discussion of what implicitly is
"the official language of MathOverflow" and whether "one language is better than many" have import, I
remind people of the title of this thread.

I apologize for not taking earlier pains to be clear. I want to know if MathOverflow should be prepared for
international publicity. It seems to me that, even if it is decided that questions should be submitted
in English using only an approved LaTeX style file, it still is appropriate to put an international face
on the website to "welcome", as I think Andrew Stacey might put it, visitors who are still learning English.
Since the demand for such is low, it seems it is not time yet to move on the issue. I think it is not to
early to discuss this issue of enhancing MathOverflow for the purpose of being more inviting to the world.

(Yes, I know. Too many homework problems. OK, I'll come up with another proposal to limit those. I
still hope Anton et.al. add this to the list of potential "big plans" for MathOverflow. How about a sign
saying "Need Homework Help? Let MathOverflow Guide You. Click Here", and then dumping them on
some other site.)

Gerhard "Babel Towers: Apartments to Let" Paseman, 2010.10.20

Gerhard, sorry, I know the thread was for publicity concerns...

To repea myself on the one topic, about individual thoughtfulness rather than MO policy or software, Pete came up with a clever and gracious one-time approach that involved no MO features at all, but can be used from time to time by anybody who sees a need:

http://mathoverflow.net/questions/41097/how-to-ask-meta-question-in-math-overflow-closed
@JJC: You might try writing future questions in French as well as English. Some of us would probably understand you better, and this understanding can be communicated to the rest. – Pete L. Clark Oct 5

From what I can see, Jerome Jean-Charles held the initial 1980 French record (the age he gives is about right) for speed-solving Rubik's cube, see pages 12, 13, 14 in
http://www.jaapsch.net/puzzles/cubic1.htm

I like the bilingual approach, again from time to time, and under the assumption that an OP is willing to reveal original language and admit the possibility of language difficulties. There will often be someone, maybe not in a convenient time zone, who can either make a part translation into English, or, sometimes far more important, translate misunderstood comments by us into another language.

jbl, two days ago: "is rooted in low-level misunderstandings" http://tea.mathoverflow.net/discussion/714/

Dear Emerton,

Regarding posting in German, I offer the following (anecdotal) evidence that this would be significantly different than allowing French posting:

I am not fluent in French, and I have not taken a class in French since elementary school. However, I am able to read mathematical books and papers written in French with relative ease. Conversely, I have tried to read a number of papers in German. However, every time I have tried, it has resulted in abject failure (the worst kind of failure!)

I'm getting the vibe that some people here are unimpressed with my comments about the English language.

@Willie: Just to be clear: I _would_ like it if everyone learned (in addition to their native tongue) a single common language (so that we can all communicate well together, with fewer misunderstandings, etc. etc.); but wanting people or society to do something is completely different from wanting to _force_ people to do something: questions of choice, freedom, and self-determination trump (in my value system) the previous desire. I also do not see anything particularly distasteful in the former desire; to be honest it seems pretty clear to me that ideally there would be a common language that everyone could speak with an adequate level of facility.

Out of curiosity I've been attempting to hypothesize about some alternative viewpoints about this "shared language" issue to see if I should change my opinion, but I haven't had much success. I'm just thinking about what would be the "best" situation for mathematics. For example, this idea of publishing in non-English language journals. I mean, being able to publish in a journal in your native tongue definitely is great in that one can write with more fluidity and express your ideas with more precision and less difficulty, and so forth. This is all very good for people not fluent in English. But in the end isn't the purpose of writing a paper to communicate knowledge and ideas to other people? I fail to see how writing mathematics in more than one language is advantageous for our profession. (A paper is written once and read many times after all.) Of course people should have the choice to do what they want, I just can't see how making such choices is good for everyone. If anyone thinks I'm being blind or an idiot feel free to let me know (or not!).

In summary, Shevek thinks the world would be a better place if everyone had an adequate grasp of the English language and wrote up scientific work intended to be disseminated in English. I'm not sure what's controversial in thinking that the world would be a better place if A, B, or C about it were different (if banker's were less greedy, if people didn't litter, if everyone treated each other with compassion, if everyone could communicate academically via a single common language, ...).

+1 to Willie.

Shevek, I think D is considerably more controversial than A, B or C. (Perhaps there are economic theories in which A is controversial too, though?)

I'll leave it at that, for fear of further disrupting grp's useful suggestion. I've probably done too much damage already.

There are only three (3) viable options on the language issue:

1. continue to enforce the English-only policy, in which case you come off as insensitive

2. allow any language, in which case you look silly, like a polygot boarding house

3. make the site fully bilingual, Esperanto being the other language

Dear Maxime,

Thank you for this wonderful comment!

Best wishes,

Matthew