7

There is a variety of people who come here with different intentions. There are people who are interested in particular conlangs (e.g. the languages of Tolkien), and there is no debate that their questions are perfectly appropriate here. But there are of course also people who want to make constructed languages themselves¹, and their questions will be much more linguistic in nature. I would go as far as to claim that for anyone who wants to make a naturalistic conlang² will predominantly have questions that might just as well feature on linguistics.SE. The obvious question then becomes whether these questions are on-topic here. Some example questions that fall into this grey zone:

  1. What is the difference between tense and aspect?
  2. What are common origins of accusative case markers?
  3. Consider an isolated, close-knit community; which characteristics is their language likely to have?
  4. What is the difference between an Isolating and an Analytic language?

Of these, 1. and 4. are questions on linguistic terminology, while the other two ask about tendencies in natural languages. All of these questions would presumably be welcome on linguistics.SE. At the same time, these are questions that an aspiring conlanger might have and (moreso in the case of 2. and 3. than for the terminology questions) will need answers to to properly construct their language.

Are these questions on topic here?


¹ such as myself
² which is likely a majority among conlangers

  • 2
    While the featured tag may have been a bit too much, I would like to sincerely disagree on this being a “relatively obscure scoping issue”. This issue is (in my opinion) a major deciding factor on whether the conlanging community as a whole ends up considering this site useful or not. I don’t see this as minor. – Adarain Apr 5 '18 at 21:36
9

Yes, I believe it is imperative that we allow these questions here.

The reasoning is quite simple: Yes, there is another place here on StackExchange you could go ask that question, but other places to ask questions have always existed. You could just as well go to reddit, or a conlanging forum, or the linguistics department of your favourite university, and ask there. This should not be a sufficient reason to ban these questions here.

On the other hand, I believe this kind of question brings immense value to the community: without them, a large section of active language constructors would be completely excluded from this site. I myself would likely never ask any questions anymore on this site, and I know quite a few people who feel the same. Conlanging depends on linguistics, and as such, excluding it would be (as I wrote elsewhere already) like excluding questions on music theory on a composition forum. As such I believe that any linguistics question that is relevant to conlanging should have a place here.

However, ...

I do believe that questions of this kind could be framed in a better way at times. Often, the person asking such a question has a very specific situation in mind that relates the question more closely to conlanging. I believe that where possible, we should encourage bringing the questions into context (by asking for context in comments, for example). To make an example, consider this (also grey-zone-y) question asked on this site by myself:

How do tones disappear from a language?

In this instance, I did not actually have any context to add. I am not currently working on any tonal languages that I might de-tonalize and was in fact just curious about it. However, imagine this was the case, then perhaps a better question might be phrased more along these lines:

In my constructed language, I have a tonal system with three tones (high, low, falling). Many words are mono- or bisyllabic, and there is a large amount of minimal pairs between words only distinguished by tones. I am currently deriving daughter languages from this language and would like for some of them to lose tone. However, I fear that in doing so, too much ambiguity will be created. What can I do to combat this?

Now, this question essentially asks the same thing, but the context makes it clearly conlanging-related, and on top also allows for more directed answers. While the former lands in the controversial grey zone discussed in this thread, the latter would be without any doubt on topic and a good question.

That said, I don’t believe that inventing context should be done either, which is why I’m not going to edit that old question of mine to contain the invented context above.

0

Added to Adarain's answer, I believe some questions can be edited to be more conlang-like. Here are the examples in the question:

  1. What is the difference between tense and aspect?
  2. What are common origins of accusative case markers?
  3. Consider an isolated, close-knit community; which characteristics is their language likely to have?
  4. What is the difference between an Isolating and an Analytic language?

Here are my revisions:

  1. (Don't know)
  2. What can my accusative case markers develop from? / What is a naturalistic development of accusative case markers?
  3. Which characteristics is my conlang likely to have, in an isolated, close-knit con-community?
  4. (Don't know)
  • 1
    While I agree with Adarain's suggestion that posts be edited to tie more closely into a specific conlanging problem when possible, I do not agree that any question that cannot be re-worded in such a manner is automatically more suitable for Linguistics SE. Linguistics SE assumes a different level of basic knowledge of linguistics and different goals than Conlangs SE does. Answering questions #1 and #4 would be approached differently in a community where the focus is on linguistic knowledge for its own sake rather than for the sake of constructing or describing one's own conlang. – Sparksbet Apr 8 '18 at 19:52

You must log in to answer this question.

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