Non-experts closing questions

Question

I have recently noticed what, to me, is a pretty irking trend in the tag I contribute to most often (algebraic geometry). It seems to me that there is a contingent of 'serial closers' who insist on closing questions despite those questions having (at least from my perspective) non-trivial academic merit. What bothers me the most is these 'serial closers' seemingly have zero knowledge about the threads they condemn—or, more accurately, essentially have never contributed to answering questions in the relevant tag.

Is it possible to restrict people's moderation abilities largely to certain pre-approved tags where, of course, pre-approval should be related to one's contribution to those tags? This is perhaps not tenable given the size of this website/nicheness of some tags, but I feel like something should be done.

EDIT: When I wrote the above(/the previous edits) I was frustrated, and consequently was fairly pompous and overly accusatory. I apologize for that. My main concern is for students who can't get their legitimate questions answered because of non-optimal procedures for closing questions. I hope that the tone of my question doesn't distract the reader from this.

Answer 1

The first obvious issue (which I feel is what may have triggered this post, and has been identified by other users in the comments) is that experts can guess the context from a question, and appreciate why it is relevant, even if there is zero detail given in the original post. They may also know most of the background and guess what the motivation behind asking the question is. As such, some questions seem perfectly reasonable for them, that non-experts will find not up to standard. From this point of view, the frustration you expressed is completely understandable. However, we definitely cannot (and will not) ask people not to help us moderate certain posts of the site simply because they are not experts on a particular topic.

On the other hand, one can draw a parallel between this situation and writing a paper for a non-specialized or general journal: if your audience is wide, then an extra dose of context and background may be necessary for your article to be appropriate for the journal. You may find yourself defining certain terms that may be very common in your subfield of expertise, and reminding the reader of previous work that is absolutely well-known in your area, but that is the trade-off we need to make to reach our desired audience.

In some way, posting a question on the site follows this idea: the background and expertise of our user base is very, very varied, and the volume of questions we get every day is constantly increasing. As such, many users that may not be experts in an area will still be helping our efforts to tackle low-quality posts. The issue arises: do low-quality questions on relatively advanced topics (for example, schemes) deserve the same treatment as low-quality questions on pre-calculus and algebra?

I believe that we can meet half-ways: any post in the site indubitably benefits from context, background, and a show of effort from the OP, and there seems to be no strong argument to change our standards for advanced questions. At the same time, maybe for these, we can be slightly more relaxed as to how far we want to take this. For example, someone may be trying to compute the first or second page of a spectral sequence, and requiring them to show their full attempt may be counterproductive.

Answer 2

No question in mathematics is unmotivated or intrinsically interesting. Despite the fact that I, myself, often use the phrase "this is a natural question," no question in mathematics is actually natural. A question becomes interesting or natural once one has enough of a background to appreciate where a question comes from.

For example, I think that the following is an extremely natural question to ask:

Let $X \subseteq \mathbb{R}^m$ be a metric space, and $f: \mathbb{R}^m \to \mathbb{R}^n$ an $\alpha$-Hölder map. Is it true that $$ \operatorname{dim}_{\text{As}}(f(X)) \le \frac{\operatorname{dim}_{\text{As}}(X) - 1}{\alpha},$$ where $\operatorname{dim}_{\text{As}}(X)$ denotes the Assouad dimension of $X$?^[1]

Anyone who can answer the question likely already knows what all of these terms mean, and why the question is interesting, and so on. It is likely an "intrinsically interesting" question to experts. That doesn't make it a good question on Math SE.

Off the top of my head, this question would be much improved by

1. Providing a reference for basic definitions—in particular, the Assouad dimension is, maybe, not the most familiar object. A citation to an appropriate paper in which the dimension is defined would be incredibly helpful. An actual definition would be even better, but is maybe not required.

2. A similar result holds for the Hausdorff dimension. Why not mention that?

3. This question was posed to me by a colleague, so I don't actually know what motivated him to ask, but, given his research area, I have strong suspicions. Why not explain that? "This question is related to controlling the behaviour of differential operators near the boundaries of John domains."

4. I haven't really spent much time on this question, but if I were to post it here, I would almost certainly have put some effort into it. Explaining where the wheels fell of would be helpful. For example: "Paralleling the argument for the Hausdorff dimension in [reference], it seems that we should be able to make an estimate of the form [foo], but the inequalities end up going in the wrong direction at step [bar]."

I will also point out that this kind of question can serve to advertise the topic being studied. If you phrase a question so that only experts are going to understand it, then you are only reaching experts. On the other hand, if you phrase a question so that even a non-expert might understand where it comes from, you might convince the non-expert that the question or field of study is interesting, and put someone on a path towards expanding your area of expertise. Maybe you are one of those paranoid academics who is obsessed with priority and deeply fearful of getting scooped but, if not, isn't a Good Thing™ to attempt to be more inclusive an bring more people into the fold?

Even if a question is more advanced, context is required. Rather than chaffing against this requirement, why not think of it as an opportunity to spread the word about something you find interesting?

---

[1] As an aside, if anyone has an answer to that question, I would be mildly interested to see it—a similar question was posed to me by a colleague, and I had to reply with a shrug of the shoulders. One of these days, I'll sit down and see what's what, but haven't had time to think about it for a while.

Answer 3

Here is a thought. If you, an expert in a topic, see a problem-statement-question that is interesting, and if you know this question is not a duplicate, and if you judge that adding some context would increase the value of the question to MSE (in part by increasing the chance that someone would be willing to write a good answer before those pesky non-experts vote to close it), then edit the question lightly to include enough context to avoid it being closed.

I have seen this done in the comments sometimes, in fact I have done it myself in the comments, for example by adding a link to a definition which I know is important but perhaps not commonly understood.

Answer 4

This is really a double-edged sword.

I agree that it is frustrating to see good posts being closed, especially when the close-voters don't seem to have the background to understand the given contexts (even when explicitly given in the post).

However, it is equally frustrating to see low quality/no context posts not being closed. This could happen if the power of the close-voters are further restricted. From my experience (unfortunately I don't have statistics to support that), it is far more difficult to close a Riemannian geometry question then a post in e.g. Calculus or Linear algebra. My guess is that close-votes reviewer are in general less willing to vote on advanced topics (e.g. see this comment).

Also, the suggested proposal is quite absurd, since more knowledgeable users would be banned from some closing some elementary questions. e.g. I am then not allowed to vote to close questions in elementary-set-theory (which I never answered), or even algebra-precalculus (only 2 answers).

I fully agree with Asaf that there is no simple solution to this. As suggested (by Asaf), what one could do is to

• leave a comment below and describe how that is a good questions while the post is not yet closed, and

• vote to reopen when it's closed. Use the Reopen thread when necessary.

Answer 5

My experiences with having questions closed is that it can be arbitrary, with too much authority given to individuals whose rankings are sufficiently high to be able to cause a question to be closed.

Perhaps annotation flagging problems with questions could be extended such that if enough time goes on without raised issues being corrected, possibly combined with an assessment that the question has generated insufficient interest, the question is closed. Then it would be algorithmic, and less subject to what frankly feels like somebody's whim at times on this site.

One good test of a question's worthiness is if it generates any serious interest. If a question has been closed which in fact had generated interest, then the interests of some users are not being met.

In fact I reached this thread because I was looking at a geometry question I had bookmarked - only to find it had been closed: Prove that the 4 points are concyclic. While I agreed that the asker of this question could have provided more detail about their efforts, I also thought, for example in this particular case, that the person who asked the question had put considerable effort into the diagram provided.

Also, I have had several questions of my own closed that I know I put a great deal of effort into making as clear and precise as I was able, and that generated some interest prior to being closed.