9

I believe that questions asking for a list of languages that possess a certain feature have more reason to be allowed at Linguistics.SE than similar questions at other SE sites.

Because that's an imporant part of typological research, and that's where a researcher could get valuable help from the community by asking other linguists to list the languages of the type he is interested in. (Different linguists have worked with different languages, so collecting their answers is valuable for getting an idea of the possible languages of the interesting type.)

As an examle, two questions I've just posted:

They ask to list the languages with the corresponding features.

Of course, I could phrase them "Is there a language which has ...", but that would be not honest, because the more such languages will be listed the more value for the corresponding typological and theoretical research.

(BTW, should such questions be marked with a dedicated tag?)

7

I may be undermining the original vision that the original founders had for this site, but as a linguist, I would heartily welcome questions that ask for languages or lists of languages with a certain feature. So much of theoretical linguistic research involves formulating hypotheses or building models based on data from a single language or a cluster of languages and then turning to analogous data in other languages to see if the original hypothesis is supported cross-linguistically or to lend credibility to the model.

I have yet to post a question on this site, but this is precisely the type of question that I am most likely to bring to a forum like Linguistics SE, because it is the most difficult kind of research to do systematically using other more traditional methods (i.e. it's hard to know where to look for something when your question is precisely of the form "Where would I find X"!). This site provides a unique way to harness the power of an engaged community with a common interest but an incredibly diverse palette of linguistic knowledge.

It seems to me that one of the most powerful aspects of SE system is the flexibility built into it, which allows a unique culture to develop organically for each specific SE site in a way that makes sense for that field.

Imagine a situation in which I'm trolling recent questions to see if there are any that I might be able to answer, and I come across one of the form "Are there any languages with feature X" or "I'm looking for a bunch of languages with feature X", and someone has provided a perfectly acceptable answer with, say, a list of examples from three different languages. If I happen to know two more languages with the relevant feature, I'm not going to pass over the question because my list would be shorter and therefore somehow inferior. And I'm also not going to repeat the original three languages in my own answer alongside the additional two in an attempt to create a more comprehensive super-list of five languages! I'm simply going to add my answer with the two languages that I have, trusting that this is a "the more the merrier" type of situation. This was precisely what I did in this situation, where the two people who had responded before me had provided perfectly acceptable answers but I happened to think of a couple of additional examples that I thought were worth sharing.

As for how to handle upvoting, one sensible way to approach upvoting on answers to "list" questions is simply to upvote every answer that fulfills the criteria sought out in the question. It wouldn't be a matter of picking the best answer but rather acknowledging every accurate answer, thereby rewarding each person who responded with helpful information for making a valuable contribution.

Incidentally, I believe this may currently be disallowed, but I think it would be great if whoever asked the original question could accept more than one answer as "the correct answer" (this may actually be a topic worthy of its own question here in Meta!).

  • I'd be glad if Linguistics.SE would develop in the direction you draw in your reply. Implementing a technical facility or a policy to allow to collectively build lists of languages (or other observed data) would benefit Linguistics.SE;moreover, it'd be also useful to collect also negative (as well as positive) examples for a feature being looked for: knowing that a language doesn't meet the criteria is different from not knowing for sure the relevant aspect of the language.We could abuse comments for posting languages that we have looked at and discovered that they don't conform to the request – imz -- Ivan Zakharyaschev Oct 26 '11 at 14:40
1

Lists don't fit well at SE (as you've noted in links to other questions an meta questions). How do you vote for a better list? also is the expectation that a list is comprehensive? That is desired but how could it be expected?

And if you request one item per answer, how do you judge a single item as better than another (it's supposed to be a list).? A case can be made for the latter, but to mimic a list a person would need to submit more than one item, which is just psychologically weird, whatever the rational utility.

But is linguistics and typology in particular somehow different? I'd guess the powers that be would say no.

However I think there is a place here for lists like you want. See what kinds of answers you get. You know you can answer your own question (to at least start off with examples you already know).

1

This Meta question is old but the subject isn't. Due to recent issues, I'm going to write something about this.

First of all, no open-ended lists.

A question that asks for a reasonably scoped set of answers is OK. You don't need to ask a question with one answer, but asking for too much is not OK, either. Some issues about this are:

  • How do we handle them? They need constant attention, fixing new users' answers, etc. This is not part of Moderator duties, so this is the first problem.
  • Turn them to CW? Community Wikis are not the answer to any question that doesn't fit the site. Community Wikis are for those questions that don't fit the site but represent a special case. If we do this for any open-ended list, then we're not dealing with special cases anymore.

So, if you ask for a feature shared by languages which are just a few, then it could be fine. But if you expect this to turn into a huge amount of answers, then there is a problem.

The StackExchange sites are not forums, unlike they're sometimes not appropriately called. They are Q&A sites, so they expect to bring a reasonably narrow set of information for a certain question.

If you have more than one issue, why not simply posting multiple questions that address each issue? If your question is addressing a huge issue, why can't you word it to make it more narrow? I understand that Linguistics is not exactly like StackOverflow, which is very practical, while we handle theoretical issues. But we cannot let everything go by itself.

I can ask for some advice from other mods, and see what ideas I can get, but don't expect miracles. :) I'm not the one who makes the rules, but the rules are clear.

You must log in to answer this question.

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