6

This question is inspired by an answer to What is biggest problem facing Genealogy & Family History Stack Exchange? where @lejonet says:

Focus on English language and thereby no user base for curating questions from an international multilingual community. This is a huge problem, even for regular users, because e.g. Americans won't find help with their European, non-British ancestors.

The Stack Exchange network certainly has a Non-English Question Policy but it states its scope at the outset:

here’s our official policy towards non-English questions on Stack Overflow, Server Fault, and Super User.

It appears to have been based on We need to help non-English-speakers somehow and there seems to have been an attempt to widen its scope in Do posts have to be in English on Stack Exchange?

However, I do not read anything in the above links that says Genealogy and Family History has to be "English only".

I would like to propose that here we try a policy where a question can be asked and answered in multiple languages but with an aim that one of them is always English. How I think it could work is as follows:

  1. Question is asked in any language, preferably a well known one, and preferably (but not necessarily) with at least some form of English translation. The title should preferably be written in English. Tags should preferably be in English.
  2. When Question is seen by anyone that recognises the language that it is written in, they check for whether that language has been included as a tag. I suggest this because with my high school French I could recognise and crudely translate a question but it would be far better to tag it so someone with far superior French-English translation skills than mine could look for it and do much better.
  3. When anyone translates a question into English (and theoretically they might translate a question from English too - to try and attract a wider audience) they use the Horizontal Rule button below the original text and write "Translation (via Google Translate)", or "Translation (by @user1234 with high school French)", etc, and then the translated text.
  4. Part of the translation should be to provide an English title to the question
  5. Part of the translation should be to provide English tags to the question
  6. Part of the translation should be to double-check that the question has been tagged for the language it is originally written in (unless that is English)
  7. For answers we likewise be supportive of everybody's translation efforts by using a Horizontal Rule to separate original from translated text.

The above policy could I think have these advantages:

  • engage more genealogists world-wide than we do now because Google searches in their own language will find some Q&As that they can more easily read
  • have a big benefit to those of us keen to understand records in languages other than our own
  • keep close enough to any Stack Exchange policy that is bent towards English, but open our doors to non-English speakers
  • make G&FH SE more attractive to non-English speakers without making it hard for the majority of our current users which I suspect is English only speaking.
  • taking advantage of our wiki style editing capabilities to curate those Q&As which become multilingual will help separate the value we offer from that of regular Q&A, message/bulletin board, and discussion forum sites.

Playing Devil's Advocate, what happens if a question is asked in a language with very few speakers and that no one here (or Google Translate) has the faintest idea how to translate? My answer would be that we could leave it open for a while, but that the option to close it as "Unclear what you are asking" remains valid to do, and if it is stumbled upon by someone who does have the skills to translate it then it should be possible to simply re-open it.

  • How will users of other languages 1) find this site and 2) know that they can ask questions in another language if they don't speak English to begin with? – user47 Sep 29 '14 at 17:42
  • @JustinY These are both good points. I'm thinking that if it became policy then a "soft launch" (i.e. non-English Q&As starting to be found by search engines) rather than mass advertising may be the way to go while we iron out any issues. There would appear to be a need for some additions to our help with some lessons perhaps learned from french.stackexchange.com/help/on-topic. Doing this for multiple rather than just a few non-English languages would become difficult. – PolyGeo Sep 29 '14 at 21:52
  • If you do this, you need to think carefully about how you will identify obscenity and hate speech in languages you don't speak and can't run through Google translate. – user104 Oct 1 '14 at 8:33
  • @ColeValleyGirl That is certainly a concern and I think the best I can suggest is a reliance on flags by users who do speak those languages, and if something goes unflagged and untranslated for a few days closure/deletion may be the only sensible option depending on how "of concern" it looks superficially. I have not yet investigated how wide a range of character sets SE can handle but initially I am thinking this may be an expansion into European rather than all possible languages. – PolyGeo Oct 1 '14 at 11:04
  • I would certainly seek advice from other Stackexchange sites who've attempted something similar. – user104 Oct 1 '14 at 12:28
4

This effort is admirable, but it would hurt more than help.

The issue is this. A very basic goal for all Stack Exchange sites is that we want to all but guarantee that when you ask a question, you get a good answer. This is a Q&A site, so questions and answers coming together is what makes these wheels turn.

On other websites, which fulfill other functions, it may work to allow users to post in all languages without concern for whether anyone responds. Not so here. Allowing anyone to post in languages other than English hugely decreases the chances that they'll receive an answer. Then they just wander away again, and the community here is left with questions that die on the vine.

You need a reliable feedback loop for a community to take root. That's why when the SE team made our first foray into Stack Overflow in another language, we carefully put in time and assembled infrastructure to ensure this basic premise would be satisfied.

Please, don't try and turn multi-language support on and off like a light switch.

An alternate idea for bolstering site activity

It looks like this discussion emerged because the community is trying to figure out how to increase site activity. In an effort to address that root question, I'm suggesting a different idea about how to further this goal; A Weekly Topic Challenge.

Other SE sites have sometimes leveraged these with rather positive results. Like the current idea of supporting other languages, it rallies the community and trains people in how to use Stack Exchange for Genealogy pursuits.

If this is appealing to you, I encourage you to ask a new meta question about it and get the ball rolling.

  • The problem with genealogy Q&A is that the majority of questions has a regional nature. When you look at the traffic on the sites that I visit, questions about methods or technology, which is what some of the original StackExchange sites are about, are a minority. And for the majority of questions, like Dutch researching French or German ancestors, the English community of this site is simply not good enough. That's not because people here are not smart, but because most lack the local knowledge that I need. – Enno Borgsteede Oct 2 '14 at 13:44
  • I understand that and sympathize, but the Q&A format isn't going anywhere. The fact remains that we're unable to support multi-language sites. That's why I offered an alternate suggestion which maximizes on the system we have available to us. – Ana Oct 2 '14 at 18:45
3

I think the proposal is worth a try.

G&FH SE has had issues several times with questions from individuals whose native language is not English. Their questions are phrased ambiguously, the authors seem to have problems understanding requests for clarification, and everyone gets frustrated. Requiring English only has definitely limited participation.

There are a number of genealogy forums that have no language restrictions.

What you are proposing would, when all the elements were complete, look something like this: Proposed layout for multi-language questions

Here are some points for consideration that could fine-tune your proposal:

  • review of tags for synonyms in other (major) languages (we don't want duplicate tags)
  • many country tags could do double-duty as potential language tags
  • some of our answers are long, often with many urls; at the answer level, the translation could be a précis instead of the full text
  • I agree fully with your third dot point - a translated précis rather than always a word for word translation seems very sensible. I'm thinking we should keep country tags separate from language tags in order to, for example, find the smaller but significant number of Canadian questions that might be expected in French rather than English. For the first dot point I was thinking we would create synonyms as we become aware of the same tag meaning being used in more than one language rather than reviewing them up front but maybe doing both would work too. – PolyGeo Sep 29 '14 at 4:11
  • BTW Your graphic is excellent and once we have had some discussion to see if we can gain consensus I think it could be the centrepiece of a policy because it conveys my general thinking much better than words. – PolyGeo Sep 29 '14 at 10:36
  • I like this idea other than I propose the English (even if only Google Translate be at the top) as per other points made. – CRSouser Jun 12 '15 at 18:58
1

Based on experience, I don't believe in translation. That's because translations are extra work, and I personally don't have time for that. I'm not a forum person anyway, because visiting sites takes way more time than mail lists or newsgroups, so when I use them, I'm always in sort of a hurry.

When questions are translated, there's also a risk that people reply in English only, which would mean that one needs a translator for the answer too, and when that translator happens to be me, I will very likely give up.

If you allow questions in other languages, I suggest that you accept that there are questions and answers that you can't read. It would be nice if they can be tagged by language, so that you can filter on that. There's a considerable amount of people that feel more at home when they can use their own language at will, and don't feel obliged to translate, or ashamed for writing bad English. I'm part of the Dutch community on Geni, and I see that it works.

  • PolyGeo's proposal does not require that the translations be available immediately. Indeed, a GoogleTranslate version could be added and then improved on by native speakers, over time, just like the all-English questions. The real beauty of the proposal will come when someone is asking for help across languages - looking for records in an ancestral country and language or tracing relatives that emigrated. Currently answering those questions that depends on those of us that are multi-lingual, and I don't think we cover enough languages. – bgwiehle Sep 29 '14 at 12:57
  • I think the big question is whether we want questions with no translation at all. Multi-lingual sites like Geni, GeneaNet, and My Heritage allow that, and that makes some feel more at home. Like Facebook and Google they don't demand anyone to use another language than their own, and that's how they attract visitors. Thing is that for them, audience comes first, not what the host thinks is beautiful. – Enno Borgsteede Sep 29 '14 at 20:13
  • 1
    For this to work I think there will need to be a translation to/from English involved because G&FH SE is moderated and at the moment the three moderators are English speakers. Adding some bi or multilingual moderators would be great but getting language coverage by a moderation team, that is unlikely to grow larger than three while we remain small, will be hard. I think the comment by @bgwiehle is exactly what I was thinking - only a crude (Google Translate copy/paste) and at times belated translation is all that is needed. The Q&A may be over by the time it is translated. – PolyGeo Sep 29 '14 at 22:02
  • 1
    Having just visited geni.com/blog/new-on-geni-multilingual-profiles-386125.html I'm impressed by the level of multilingual support being offered there, and that is certainly not something that I would advocate we try to match. I think our uniqueness comes from the ongoing curation available for our Q&As to keep getting better, and part of that getting better could be to have some of them readable and contributable by non-English speakers with no extra effort on their part being required. – PolyGeo Sep 29 '14 at 22:52
  • I think that part of the answer depends on the sort of Q&A you want covered here, and on the amount of extra visitors you want to attract. I'm convinced that you get the most when you don't demand anything, like Facebook does. There Dutch connect to Dutch, abandoning a local site that started earlier. If you want the same for genealogy, you need to think a bit different than for computing issues as they exist on the original StackExchange. – Enno Borgsteede Sep 30 '14 at 13:52
  • Reason behind this, is that in genealogy, both my language and location are way more important. When I ask about a software feature, like connecting to the FamilySearch tree, an answer from an American member will likely not fit very well, because that member doesn't know which of the candidates support my language, and which don't. Similarly, when I ask about research in Germany, and answer from a US member will likely be useless, because some resources available in the US are not available in Europe, because of legal restrictions. In other words, answers are not as universal as you think. – Enno Borgsteede Sep 30 '14 at 14:03
  • So the basic question that you need to ask is, whether StackExchange has an advantage over local sites, and if it has not, how you can create such an advantage. It does not exist for advice on software, I think, because software is localized. It may exist for research help, like me asking a fellow genealogist to look at a German LDS film, which I can't legally access in a FHC here in The Netherlands, but that is quite a rare case. In most cases, locals are the best help for locals, and that how Facebook and Geni work. – Enno Borgsteede Sep 30 '14 at 14:11
  • 1
    Enno, you make a lot of good points. For certain questions, people need local help. For other questions, the act of writing the question out can be beneficial, even if no one answers it. To ask a good question, the researcher has to organize any prior research. And it can be helpful at times to have a second person search for things, simply because they may have different skills in searching for things. It allows the person asking the question to get a fresh perspective on their question. – Jan Murphy Oct 2 '14 at 23:22
1

I like @bgwiehle's proposal other than I would keep English as core requirement of all posts and the English version always be at the top as well as the title, even if it is a generic Google Translate or another generic translation program doing the text translations. The original text can be located below.

My PRIMARY reasoning for this is I can only following my English speaking roots so far and then I go into a foreign language which I do not speak and if we have more diverse user base the odds of my non-English questions might get answered as well as I might be pointed to local sources in those regions I would have not otherwise found due to not speaking the language. I would assume the opposite would be true in non-English areas to English areas.

We also have a tag where we are asking for a translation, but this is more secondary or even tertiary.

To be clear though I the posting in multiple languages would only be required for original Non-English language posts and it would be denoted that it is an automated translation,

AND moderators would not be responsible for performing the translation.

You must log in to answer this question.

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