8

Recently I've noticed multiple questions about etymology getting close votes and, in at least one case, actually getting closed. The reason given was that language specific usage and grammar questions are off-topic

Whilst these questions are language-specific (being about the etymology of a single word in one language), they are neither usage nor grammar questions. As such the reason given for closing these questions does not seem to apply as written

This is not a good situation, and so there ought to be some sort of clarification. Either we should explicitly include etymology as part of the reason to close (this seems like a bad option, as many etymology questions have received several upvotes and contained interesting information), or explicitly excluding etymology from the description of this reason to close (this could possibly cause confusion as people may skim, see etymology, and assume that means it's included)

If etymology is on-topic, what can be done about wrongfully closed questions? Votes to reopen haven't generally succeeded, and it is often the same people voting to close etymology questions, is there a way to "officially" remind people of what the off-topic rules actually are and/or remove their ability to vote to close if they continue to wrongfully vote to close?

cf what should be done about "language-specific grammar and usage"

4

I've felt for some time that something along the lines of this close reason ELU has might be helpful for keeping unwanted etymology questions at bay:

Please include the research you’ve done, or consider if your question suits our English Language Learners site better. Questions that can be answered using commonly-available references are off-topic.

That is, we require etymology questions to accompany the research the poster has done and make ones commonly-available resources like Wiktionary and EtymOnline can easily answer subject to immediate close.

If it's still true that we can have no more than three canned site-specific close reasons, the second and third can be combined to make room (i.e. "Questions seeking help with identification of, translation of, or making syntax trees of specific samples of text are off-topic").

3

I've been someone who's been in support of pretty tough lines in regard to etymology questions. I still stand by the quote below: linguistics is fundamentally about language systems not just language, so the etymology of arbitrary words doesn't really qualify. But I think it would be helpful if we could build up a list of gold standard etymology questions so that we can see what kind of things we don't want to lose. I started a list, but it needs more people to add to it.

Linguistics is the study of language systems. Etymology questions belong here when they're asking about systematic issues. This could involve sound system changes. It could be about languages which have borrowed wholesale and systematically from another language. And questions on grammatical morphology are almost always on topic. But non systematic arbitrary borrowings or meaning changes don't really belong here. (from this answer)

I suggested this in a comment, and think it could be a good principle to adopt:

Maybe we could say something like "etymology questions must situate themselves within a historical context of language change or contact." This context would include both positive content as well as negative, where the OP has actually tried but been unable to substantiate any evidence for cognates.

Most low-quality etymology questions show no effort of prior research or give any context for why the question is even remotely interesting. What we want are interesting questions that require experts (and not just someone else to search an etymological dictionary for you.) So if we started requiring all etymology questions to provide contextualisation, then that would make a big difference.

If that's a principle the community would support, then we can start thinking about how to change the close reason descriptions.

(PS: I've been interpreting the current first close reason as implicitly including single-language, single-word meaning questions. I assume other people do too - no one wants questions like "what does X mean?" here, even though that's not grammar or usage. It's only a tiny step from single-word meaning questions to the single-word etymology questions that get closed.)

5
  • 1
    I would understand "what does X mean" as a usage question, the sense of a word is essentially determined by its usage. I do think your suggestion of requiring contextualisation is good though – Tristan Dec 15 '20 at 13:50
  • @Tristan I had thought of "usage" questions as more directly asking how to use some part of a language. Either way, it's not particularly relevant to this discussion on etymology questions.. Hopefully we can get some consensus on etymology questions as they've been one of the big disputes on this site the last few years. – curiousdannii Dec 15 '20 at 14:12
  • yeah, the current situation just seems a particularly unclear one that doesn't really do anyone any favours – Tristan Dec 15 '20 at 14:18
  • 1
    For what it's worth, I would think "what does X mean?" is a question about identification and translation of specific samples of text, which we've also said are specifically off-topic (but as a separate reason from language-specific grammar and usage). – Draconis Dec 15 '20 at 21:59
  • 1
    I agree with this answer. If we change the close reasons, it may be good to add a link to a community wiki answer somewhere with etymology resources with which users can do basic research. If after that there is something they don't understand it would fall in scope I think, as the question is then about how to apply a certain methodology. – Keelan Dec 18 '20 at 6:52
3

I agree with the position that linguistics is fundamentally about language systems not just language facts; but I also maintain (vigorously) that theories of the system are empirical hypotheses supported by facts of language. We tolerate very many questions about communicative behavior and text-processing, a fact that I find hard to square with a supposed allegiance to the scientific study of language systems, and I conclude that Linguistics SE is not just about the theory of language system, it is in fact more broadly "language and linguistics" (as judged by clientele).

We very frequently get questions about a linguistic aspect of a language, which then receives language-specific close votes. Often this includes questions of the kind "what is the analysis of this fact of language X", but generally the question is framed in an unsophisticated "language fans" fashion. One solution is to make small edits that retain the semantic core of the question so that it is more sophisticated in wording.

But ultimately, this is a consequence of the SE model and popular-vote models of action in general. Somebody somehow determined at some time that such-and-such would be a reason to close, somebody wrote the text(s) saying what's on-topic and what's off topic, so the solution to the injustice problem (the description does not match what you believe the rule should be) is to re-write the relevant official text (all of the text, not just one specific piece of text), and cause it to become official. I have whined more than once about the problem that "language-specific" and "usage" are about two completely different things. Questions are closed because people VTC: then people VTC, they see a specific set of choices. Therefore, primary blame should be placed on that set of choices. Make a concrete proposal. Secondary blame should be put on the rarely-read Help Center instructions to OP, if the wording does not clearly declare our sentiment.

2
  • Regarding "re-write the relevant official text": please propose what wording you would prefer. I'll change it after other people chime in. I wanted to make etymology off-topic, but many people were against the idea. Also please keep in mind that we can have a maximum of three close reasons. – prash Mod Jan 5 at 17:58
  • 1
    I'll give some thought to re-structuring the wording, and we'll see what I come up with in the next two weeks. – user6726 Jan 5 at 19:22
3

Unfortunately, as a long-time user of this site, I have seen tons of really bad questions on etymology, picking two superficially similar words of unrelated languages like English and Hebrew and asking for a connection, or questions that can be easily answered by looking up Wiktionary or another online dictionary. Most of those questions are not only closed but also have been deleted after some time. We once had a closing reason lacking basic research for such kind of question which is more fitting than the templates we have right now.

The main point is: Does the question and its answer adds something valuable to a question and answer site like this one? Will other users than the original poster come here via a search engine and find some valuable information? If yes, the question and its answers are worth keeping, otherwise not. Sometimes a bad question is justified by good answers.

Single word etymologies are problematic in this respect, because we don't want to build a comprehensive etymological dictionary disguised in Q&A format. An etymology needs to stand for some more general phenomenon to stay here.

2
  • these also seem like pretty good principles to me. As it stands though, closing such questions is not in accordance with the stated guidelines, creating a situation that is extremely unclear to askers (especially as most of the people asking such questions are new contributors, who don't have much experience of the unstated norms of the community). Do you know why the "lacking basic research" reason was removed? – Tristan Dec 17 '20 at 10:08
  • @Tristan I didn't engage in the discussion but when I remember right there is a limit to the number of templates for closing questions, therefore introducing a new one means also retracting an old one. – jk - Reinstate Monica Dec 17 '20 at 10:27
1

I'm of an unpopular opinion: I think we need more etymology questions to set a better standard.


  1. Etymologies aren't language-specific when their origin can be disputed. If we take linguistics to be the study of language systems, blocking off discourse here would prevent one angle of studying language systems. Understanding which conjecture on the derivation of a particular term requires a multi-faceted approach based on history, phonology and language-specific rules for multiple languages.
  2. Asking for "true" etymologies ought to be avoided, because this implicates other etymologies as phony or irrelevant. The term "ultimate etymology" should be encouraged in its place if the question seeks the first occurrence or earliest origins of a word.
  3. Seeing as this is a Q&A site, the way I see it asking a question here is a part of the research process and shewing people away for "not making an effort" or "not providing a justification" is an attitude that we need to get over.
  4. That being said, yes, there are too many "Are X and Y related?" questions. Millions of words across thousands of languages means an infinite combination of X-Y pairs, so there will always be a demand for this type of question. And I think the issue isn't rooted in the questions so much as the body. Here is an example of what a body ideally looks like:

I was reading about the etymology of the word Michigan. It said that it came from an Ojibwe word meaning "big water". And then, I saw that Mississippi was from another Ojibwe word meaning "father of water". I have come to believe these states are related. Is this analysis correct?

This is actually a very good body because it shows us the asker's point of view, they have read about this somewhere and are looking for a third opinion. I do realize that writing a good body is comparatively a harder job for non-native English speakers but we have flags/community edits!

In conclusion, instead of a new community-specific flag, I say we should instead use comments and needs improvement where readers are left in the blue and edits for clarity if there are minor grammatical errors.

2
  • I think we'd all agree we need more high quality etymology questions! But what we need to decide is what we do with the existing low quality ones. – curiousdannii Dec 27 '20 at 3:22
  • 1
    I believe I've already answered that question? Just flag for "needs improvement" and move on. If possible commenting the issue. Unless someone comes up with a community specific "Are X and Y related?" flag – madprogramer Dec 27 '20 at 17:10

You must log in to answer this question.

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