In my experience, "language-specific grammar and usage" seems to be the most popular close-vote reason on this site. And it certainly makes sense—we get a fair number of people asking about the usage of an English word, or whether a certain sentence is grammatical, which is off-topic for linguistics.

However, I've also seen this reason applied to everything from "what is the etymology of X?" to "what sound changes produced Y?" to "are Z and W related?". None of these seem to be grammar and usage questions, and the latter certainly isn't language-specific.

Now that we're out of beta, and can potentially have more reasons-to-close, might it be worth revisiting what exactly "language-specific grammar and usage" is supposed to mean? Right now it mostly seems to be used as "no other reason fits", which is misleading to the questioners.

  • I think the use of "language-specific grammar and usage" reason is getting practically out of control. Some questions that I subjectively considering interesting and about linguistics, and that are objectively not about "language-specific grammar and usage", are getting closed or nearly getting closed using this reason. It would strike me as odd that linguistically-minded people wouldn't realize that "language-specific grammar and usage" means that it has to be language-specific and either grammar or usage related (or under a strict reading, both, but that's not how English works). – LjL Nov 16 '19 at 1:34
  • So, what might it be? That it just gets used as a wildcard when one can't think of a reason (and that has actually been stated as the reason by someone, when I previously asked in a comment), but that's inexcusable, because there certainly is the option to state it's for a custom reason, and I trust nobody who cares about this site would pick the wrong reason just to avoid having to type the actual reason. Or... maybe there are people who just want to disruptively close questions they don't like, despite them being legitimately about linguistics, and so they can't state the real reason. – LjL Nov 16 '19 at 1:37

The primary question is, what should constitute "off topic"?. We also have to figure out how to communicate that viewpoint in the help center. The four specifically off-topic areas are:

  1. Advice or help on learning a foreign language

  2. Translation requests

  3. "Please make me a syntax tree" — Don't ask us to solve your homework problems for you, but it's OK to ask questions that would require drawing of syntax trees to gain a better understanding of the subject, so that you can learn to solve the homework problem on your own

  4. Identification of a language or a script based on specific samples and/or its translation — Questions about the scholarly aspects of translation and language identification are still welcome.

2 is a subcase of 4: 2 can be deleted. I would rewrite 4 as "Translation or identification of samples of a language or a script. Questions about the theory of translation and language identification are still welcome". 3 misses the boat. It's not about trees, it's about "do my homework", "solve my problem" questions. Put simply, "Don't ask us to solve your homework problems or answer your test questions for you". Even if the most common offender is syntax trees, that does not justify making it seem that (a) questions about syntactic trees i.e. constituency are OT and (b) asking for answers to phonology problems is on-topic. At the moment, I have little to say about the language-learning ban.

Questions about proper grammar, proper usage – the normative questions – already have an SE-wide close reason: "primarily opinion based". It seems to me that we do not need to do or say anything special, although seasoned users may need to get on board with the idea that primarily opinion-based questions should be closed (this reason is under-utilized). What remains of the "language specific" reason is that it refers to non-generalizable questions.

My proposal is that we need to dump the "language specific" way of thinking about the matter, and instead focus on the problem of non-generalizability. Saying "Don't ask non-generalizable questions" is too vague, but still, I propose this as a target statement. Is there a general class of non-generalizable questions that we would like to encourage or at least tolerate?

  • 2
    Even homework isn't the real issue, it's that for most syntax questions there are innumerable answers depending on the linguistic framework used. If a question specifies a framework it's often fine. The problem is most don't, and many askers don't even know what to call their framework. That's why we close the tree questions and direct them to talk to their tutors. – curiousdannii Aug 15 '19 at 8:07
  • 1
    This is true, but that also points to the problem that almost all syntax questions fail to suggest a theoretical framework. E.g. "What is the subject of a passive sentence", "Can I form a morphological condition like this". We have "too broad" as a remedy for questions that are framework dependent. – user6726 Aug 15 '19 at 13:44
  • 1
    Or "unclear" because we don't have enough info to give an answer. – curiousdannii Aug 15 '19 at 23:44
  • "Questions about proper grammar, proper usage – the normative questions – already have an SE-wide close reason: "primarily opinion based"." I'm not sure I'd agree with this: "is sentence X valid English?" is a question that can be answered objectively (or at least as objectively as "what is the etymology of Y?"). – Draconis Aug 16 '19 at 2:39
  • If so, is there a legitimate reason to preclude questions about "proper English"? – user6726 Aug 16 '19 at 4:33
  • 1
    @user6726 Mostly that they're not particularly interesting, linguistically. They can be answered by any native speaker, and there's already a separate SE site for that (ELU). – Draconis Sep 5 '19 at 2:36

You must log in to answer this question.

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