26
$\begingroup$

I have noticed that quite a few questions tend to be answered in the comments.

This was apparently observed $12$ years ago to a similar extent. These questions (unsurprisingly) tend to be on the simpler side of Math SE, so that the answer is short enough that it fits in the comments.

My question is this: What is the intention of the Math SE in questions such as these? Are these "comment-answers" a satisfactory completion of the question? The original post gave reasons why this is harmful and they seem exactly the same as today.

It seems nothing has changed in these $12$ years and I am wondering if this was intended, or if there is still an underlying issue that needs to be addressed.

I will note that these comment-answers are most commonly posted by user with high reputation, that I see quite often posting high scoring answers to lots of other questions. These users are prominent members of the Math SE and if they are the ones comment-answering, I assume there is good reason they are doing it.

$\endgroup$
10
  • 4
    $\begingroup$ To answer your question, ending in a question-mark, is rarely, if ever appropriate to give a comment as an answer. It persists because there aren't enough users who flag, and not enough people reviewing new first answers, and late answer queues. And it persists because new users without sufficient rep regularly post comments in answer fields. When you see it, flag it. This is not appropriate now, just as it wasn't then. $\endgroup$
    – amWhy
    Commented Feb 3, 2023 at 20:02
  • $\begingroup$ Indeed, you are correct that there are select high rep users who post one sentence "answers". I would recommend flagging them for moderator attention, because despite The Enforcement of Quality Standards, and a flurry of flags sent to mods to handle them, for most of them, they persist to do so. $\endgroup$
    – amWhy
    Commented Feb 3, 2023 at 21:03
  • 60
    $\begingroup$ Rather ironic that this question was answered in comments... $\endgroup$
    – Clyde Kertzer
    Commented Feb 3, 2023 at 21:09
  • 4
    $\begingroup$ Maybe often people give aswers in the comments (correct answers) because they fear downvotes. Often someone says that the answer should be instead a comment. Once, I gave a (correct) answer, which received two upvotes by high reputation users and a downvote. One of these high reputation users said that she didn't understand the dowvote, but maybe my answer didn't respect the policy of SE, maybe should be a comment. My answer said that the theorem in the question couldn't be proved because it was false, and I gave a counterexample. I'm still wondering how this could be against the SE policy. $\endgroup$
    – BakerStreet
    Commented Feb 4, 2023 at 14:00
  • 3
    $\begingroup$ This is what happens also in other sites, of which I have experience, many answers are in the comments, and I wonder why. But, in general, one reason is that they fear that the answer can be considered non sufficiently elaborate, and have downvotes. $\endgroup$
    – BakerStreet
    Commented Feb 4, 2023 at 16:29
  • 23
    $\begingroup$ I think a simpler explanation is that a one-line comment takes less effort, whereas people would usually want their official answers to be more polished. So if they don't want to spend much time on this, they just give a brief answer in a comment. $\endgroup$
    – PatrickR
    Commented Feb 4, 2023 at 17:34
  • 3
    $\begingroup$ I also had a case that there was a question with many comments, which I did not feel like reading in detail. I gave an official answer, which was then followed by a bunch of discussion with the OP to make him understand. Then one of the persons who had originally commented chimed in and then posted her own official answer, saying she had already answered first in a comment (which turned out to be true) and implying that therefore she should get the credit. :-) (which the OP dutifully agreed to :-) $\endgroup$
    – PatrickR
    Commented Feb 4, 2023 at 17:39
  • 2
    $\begingroup$ Answers to simple questions have little chances to be upvoted and quite often not accepted by the questioner who maybe forget about the question. Answering in comment seems natural for such questions. An answerer wants a reaction to his answer, and if there is no reaction, which happens quite often and annoys, thinks the question does not deserve full standard answer which assumes votes and comments. $\endgroup$
    – kludg
    Commented Feb 5, 2023 at 7:48
  • 8
    $\begingroup$ I can't say anything about the policy (though see the comments of this question). I am among those who often "answer" in comments. Well, not quite answer: often I think a good hint is more useful to the OP and I write this hint as a comment, as it's not enough for a full answer. Sometimes it works, the OP has his Aha! moment, sometimes not. $\endgroup$
    – Jean-Claude Arbaut
    Commented Feb 5, 2023 at 13:26
  • 5
    $\begingroup$ @Arbaut I agree with you. I'm new here but I've a long experience as I'm a moderator of a site of mathematics. My impression: the problem are not partial answers in the comments. But sometimes if you give the answer, you receive a hail of criticism that there is already an answer in the comments. Answer in the comments often prevent the question to be answered. I notice here an enthusiasm of many people in closing questions and criticize all, often arbitrarily and hastily, following a 'policy' (often subjective) they have in mind rather than thinking about the well functioning of the site. $\endgroup$
    – BakerStreet
    Commented Feb 5, 2023 at 17:01

4 Answers 4

Reset to default
17
$\begingroup$

TL;DR: please flag answers in the comments for moderator review.

The assumption of the answer which was written 12 years ago is that the question itself is welcome on the site, and that the commenter who posts an answer in the comments field should be nudged towards giving a proper answer. When a question is truly welcome on the site (i.e. it meets the quality standards of the site, as outlined in the meta question How to ask a good question), this remains the correct course of action:

  1. politely suggest that the commenter flesh out their comment into a real answer, and post that answer, or
  2. if the commenter chooses not to, flesh out the answer yourself and post it;
  3. once the answer has been posted, flag the comment for deletion as "no longer needed".

However, there are also cases in which a commenter will post an answer to a question in a comment when that question is not welcome on the site, e.g. the question lacks context, or is a duplicate. Such questions should not be answered—answering these kinds of questions encourages other users to post more questions of similar quality, which ends up diluting the usefulness of Math SE as a repository. Instead of posting answers to these kinds of questions (whether as answers or in the comments), users should be using the comments to encourage the asker to make edits which will improve the quality of the question. Once the question meets the standards of the site, the first paragraph (above) applies.

If you see someone answering questions in the comments, and you believe that the question itself requires improvement, please flag the comment for moderator review. You are also welcome to politely nudge the commenter away from posting answers in the comments, but you should feel no obligation to do so, and you should disengage if they attempt to argue the point.

$\endgroup$
12
  • 5
    $\begingroup$ Amen. A related problem is (in my opinion) that of some high-rep users posting comments as answers. One-liner banal answers to banal questions. Then again, it is not always clear whether an answer as a comment is intended to be a hint posted for the purpose of making the asker see the light or an attempt to get around the rules on bad questions- When/if I find the time I may open another thread to discuss the differences. $\endgroup$
    – Jyrki Lahtonen
    Commented Feb 5, 2023 at 5:31
  • 2
    $\begingroup$ You seem to propose that no matter how low-quality a question may be, it must be answered with great respect. Please respect answerers too, they may be fed up of bad questions. $\endgroup$
    – kludg
    Commented Feb 5, 2023 at 9:57
  • 3
    $\begingroup$ @JyrkiLahtonen Well, sometimes it's a way of not being called a rep whore. Though it's still possible to post the answer as community wiki to achieve the same effect. $\endgroup$
    – Jean-Claude Arbaut
    Commented Feb 5, 2023 at 13:28
  • 1
    $\begingroup$ @Jean-ClaudeArbaut Quite. It's not clear to me what is the best way of dealing with the situation when you know that this piece of mathematical knowledge has already been distributed on the site. Many of the alternatives each have problems: A) you don't want to post dupe answers, most certainly not duplicate yourself (lest the dirty word apply), B) you don't want to (vote to) close the question as a dupe to the one you answered anno dazumal (ditto), C) you don't want watch another veteran rake in more points for banalities,... $\endgroup$
    – Jyrki Lahtonen
    Commented Feb 5, 2023 at 14:29
  • $\begingroup$ (cont'd) D) you don't want to deny a promising new user a chance to show their skills, E) but you still want to explain them the rules, F) if you use comments as a "hint" you may either not give the asker enough, or make those promising noobs unwilling to post an answer (lest they feel like they are copying from comments), or G) both. Also, when judging others there are obvious beam and mote issues because we cannot read their motives (the stats may point at a certain direction, but...) $\endgroup$
    – Jyrki Lahtonen
    Commented Feb 5, 2023 at 14:35
  • 4
    $\begingroup$ @kludg I am not sure if your comment is directed at me, but I don't think that I have suggested that every question must "be answered with great respect". What I have said is that people should be treated with respect, and that low-quality answers should not be answered, either in the answer box or in the comments. $\endgroup$
    – Xander Henderson Mod
    Commented Feb 5, 2023 at 14:38
  • 1
    $\begingroup$ @XanderHenderson I don't think that answering in comment is a big problem and requires moderator attention. If a question is good I would answer it in the standard form which allows better communication with the questioner and other users. But if the question is pure and it is likely that the questioner will show no reaction I think comment-answering is normal. $\endgroup$
    – kludg
    Commented Feb 5, 2023 at 15:04
  • 7
    $\begingroup$ @kludg The official policy on SE (that is, across the entire network) is that comments are ephemeral, and are meant to be used only for suggesting improvements to a question or answer. Posting answers in the comments is officially against the rules, and answers-in-comments should be flagged for moderator attention. $\endgroup$
    – Xander Henderson Mod
    Commented Feb 5, 2023 at 15:06
  • 9
    $\begingroup$ And I very much disagree with you regarding the nature of the problem. Math SE is already perceived as a homework mill by many outsiders. One of the reasons for this is that low-quality, no effort questions are often answered here (either in the comments, or as actual answers). If a good question is answered in the comments, then the comment should be moved to an actual answer, and then flagged as "no longer necessary". Answers to bad questions in the comments should simply be deleted (again, "no longer necessary" is the appropriate flag). $\endgroup$
    – Xander Henderson Mod
    Commented Feb 5, 2023 at 15:09
  • 1
    $\begingroup$ @XanderHenderson Homework questions is a different thing that deserves moderator attention. It is a big problem for online education. It is annoying to see how questions from paid courses are answered here. I did not mean such questions. $\endgroup$
    – kludg
    Commented Feb 5, 2023 at 15:29
  • 7
    $\begingroup$ @kludg Answers don't belong in comments. If the question is good, the answer should be moved to an answer, and the comment deleted as "no longer necessary". If the question is not good, the question itself should either be improved or closed, and the answer-in-the-comments should be deleted. In either case, comments-as answers should be flagged for moderator review. As an aside, note that Math SE has no "homework" policy. The context standards are a compromise between those who welcome homework questions, and those who do not. $\endgroup$
    – Xander Henderson Mod
    Commented Feb 5, 2023 at 16:54
  • 1
    $\begingroup$ If you flag someone's comment does that punish the commenter in some way? Like subtracting reputation and/or 3 flags and you're permabanned? How would this work? $\endgroup$
    – user1112591
    Commented Feb 9, 2023 at 19:42
16
$\begingroup$

Anyone who sees such a thing may copy the answer into an Answer box (with credit, we hope) and thus reap the reputation points awarded to it. Or the downvotes, if appropriate.

$\endgroup$
18
  • 1
    $\begingroup$ Not everyone will misbehave if it is allowed. $\endgroup$
    – kludg
    Commented Feb 5, 2023 at 12:02
  • 7
    $\begingroup$ It may be appropriate to mark such answers Community Wiki. $\endgroup$
    – Peter Phipps
    Commented Feb 5, 2023 at 12:57
  • 2
    $\begingroup$ @kludg - What does misbehaving have to do with anything? $\endgroup$
    – Accelerator
    Commented Feb 5, 2023 at 13:11
  • 2
    $\begingroup$ @Accelerator I think stealing answers from comments of other users is misbehaving, but I don't want to discuss it more. $\endgroup$
    – kludg
    Commented Feb 5, 2023 at 15:14
  • 8
    $\begingroup$ When I find myself in this situation, if the question and answer are both good, I typically prompt the commenter to post their comment as an answer. If they don't, after a day or two I'll come back and do it myself in the form of a community wiki answer with appropriate acknowledgement. $\endgroup$
    – KReiser
    Commented Feb 5, 2023 at 16:04
  • 15
    $\begingroup$ @kludg If someone posts an answer in the comments, they are breaking the basic structure of the SE models, which is that some users post questions, and others post answers. Posting an answer as an answer makes that answer more visible to search engines, including the on-board search engine. If someone answers in the comments, and someone else uses that comment as the basis for an actual answer, that's fine. There is no "theft" involved---the person who posted an answer in the comments isn't following the rules, and if they wanted "credit" for an answer, they could have posted an answer. $\endgroup$
    – Xander Henderson Mod
    Commented Feb 5, 2023 at 20:02
  • 10
    $\begingroup$ Again, it is fair to ping that user before posting an answer (e.g. to see if they want to post an answer), and there is nothing wrong with giving them credit in your own answer, but, at the end of the day, answers should be posted as answers, not comments. $\endgroup$
    – Xander Henderson Mod
    Commented Feb 5, 2023 at 20:03
  • 1
    $\begingroup$ It seems rather stupid to me if someone will extract complete answer from such nudging comments and post it as answer. $\endgroup$
    – kludg
    Commented Feb 6, 2023 at 6:57
  • 11
    $\begingroup$ Please don't (mis)use Community Wiki for reputation denial. See, e.g., meta.stackexchange.com/q/227290/160917. Also, there is no need to deny reputation to someone who follows the practice that Stack Exchange wants to incentivize and writes answers in the answer box. $\endgroup$
    – D.W.
    Commented Feb 6, 2023 at 8:34
  • 10
    $\begingroup$ @kludg This is a question and answer site, not a tutoring site. Posting answers and "hints" in the comments runs counter to goals of the site. These are bad practices, and should be avoided. Period. If you don't want someone to "steal" your answer, then post an answer, and not a comment. $\endgroup$
    – Xander Henderson Mod
    Commented Feb 6, 2023 at 14:21
  • 3
    $\begingroup$ @XanderHenderson Well, thanks for sincerity. I rarely answer now, neither in comments nor in answers, and, after being warned, will even less, so I am not going to be a big policy violator. Period. $\endgroup$
    – kludg
    Commented Feb 6, 2023 at 15:19
  • 4
    $\begingroup$ Copying someone else's writing without proper credit is plagiarism. It is also a copyright violation because content contributed by users to stackexchange is covered by the CC-BY-SA license, which requires attribution. $\endgroup$
    – Nick Alger
    Commented Feb 6, 2023 at 22:47
  • 12
    $\begingroup$ @NickAlger That would matter if someone were to simply copy-paste an answer. However, it becomes less relevant if someone takes the comment and fleshes it out into a full answer. It would be polite to indicate that a full answer was inspired by a comment, but it is not always necessary---there are, generally speaking, only a few ways to answer any given question in mathematics. $\endgroup$
    – Xander Henderson Mod
    Commented Feb 7, 2023 at 12:20
  • 1
    $\begingroup$ @D.W. That is not a misuse of Community Wiki. That's precisely how CW posts have been used, (descriptively, historically) and CW posts invite other users to add addition information, to make a collaborative answer, prescriptively. $\endgroup$
    – amWhy
    Commented Feb 16, 2023 at 19:59
  • 1
    $\begingroup$ @NickAlger Note that copying a problem from a book, without proper credit, is also plagiarism. I think that happens here much more often than copying a comment. $\endgroup$
    – GEdgar
    Commented Feb 17, 2023 at 1:41
8
$\begingroup$

The accepted answer mentions "Math.SE as a repository" and I think this divide is exactly crucial to understanding this behavior. Math.SE, like all SE sites (in theory), simultaneously wants to be a repository of useful Q/A, and a community.

When a user posts a question that is neither bad nor good -- a near-duplicate, something not fully thought out with a simple issue, or etc. -- a typical reaction is to close the question with a vague justification. This is good for the repository, since the question is unlikely to be useful to other visitors. On the other hand, it's bad for the community, since this new user isn't likely to learn what they did wrong from these pro forma closures, they're likely to learn that this place is hostile, and quit. That happens a lot! It's not an edge case!

The other obvious action is to just answer the question. Format it nice, whatever, it's good for your rep, you get internet points, and the asker is happy. This is (sort of) good for the community, because it leaves the asker with a good experience and they're likely to come back. But it's bad for the repository because this question isn't great, it's not likely to have benefits to future searchers, etc., and the question might even get closed later (which is confusing to most non-power-users).

So when somebody understands the trade-off, and doesn't need to farm rep by answering easy questions, they'll often give a quick sketch of an answer in a comment. A "real" answer isn't needed because the question is unlikely to get a lot of attention anyway, and a comment is faster. Additionally, the small text also communicates the smallness of the answer, and functions as a soft admonishment to the asker, like they've used a heavy tool for a light task.

You might not like it. I'm sure people will continue to complain about this, and delete comments, and so on. But nobody has yet resolved the repository vs community issue -- and SE desperately needs to be both -- and this middle action has proven itself to be a useful middle ground between the two "correct" actions.

Now of course this is all assuming the question is neither bad nor good, which is usually what I see in this situation. If you see a question with a comment-only answer that's actually good, please, make a real answer (feel free to cite the comment if you want).

$\endgroup$
4
  • 4
    $\begingroup$ "Math.SE, like all SE sites (in theory), simultaneously wants to be a repository of useful Q/A, and a community" - This might be an accurate statement of the perceptions of many users, but I don't think it matches the mission statement that SE was designed for, at least as I understood it. The SE model is that SE sites are an archive of knowledge, in the form of questions and answers, and explicitly are not a social network or helpdesk or community (any more than Wikipedia is). I think part of the tension is that many users want it to be something it was not designed to be. $\endgroup$
    – D.W.
    Commented Feb 13, 2023 at 0:49
  • 3
    $\begingroup$ See, e.g., meta.stackexchange.com/q/65261/160917, meta.stackexchange.com/q/297483/160917, meta.stackexchange.com/q/342779/160917, meta.stackoverflow.com/q/367016/781723, meta.stackoverflow.com/q/388384/781723, meta.stackexchange.com/q/19665/160917. This seems to me to be a deep and core tension within the Stack Exchange model, and it generates a lot of passion. $\endgroup$
    – D.W.
    Commented Feb 13, 2023 at 0:55
  • 5
    $\begingroup$ I would certainly add that on Meta -- and among power users in general, but in particular on meta -- the primary view that SE should be primarily a repository, with new questions only grudgingly allowed insofar as they are the only way to grow the body of knowledge. This makes sense given the selection bias of being on Meta at all, in addition to the normal conformity effects of a small-ish and self-selected community. But I don't think this is the consensus off meta. $\endgroup$
    – Richard Rast
    Commented Feb 13, 2023 at 19:50
  • 5
    $\begingroup$ For myself I think it's obvious that in order to achieve either goal, the site needs to pursue both goals (for some definition of community), and in order to function as a community at all, the site needs to allow people to be people, and follow social norms such as politeness, courtesy, kindness, etc. These things are in tension with the goal of having a clean repository but the tension is inevitable, and must be navigated carefully, not "solved." But that's my opinion; my answer was only intended to explain why people do it, not prescribe what they should do. $\endgroup$
    – Richard Rast
    Commented Feb 13, 2023 at 19:52
1
$\begingroup$

Please forgive that this answer goes somewhat off a tangent. I have said something to the following effect on other posts on Meta: I do think this site Math Stackexchange is sometimes confused on what it is supposed to be, and I think that this topic is related to that.

Is the primary goal of this site about A) producing a library of high-quality questions and answers? Or is it as much about B) helping struggling students learn math? It can be that both are goals of this site I suppose, but keep in mind that these two goals often diverge. In particular, it seems to me that there is a difference between:

  • A quality question--a question that is original, challenging and interesting, posed by someone who clearly put a lot of thought into it, versus

  • A deserving question [according to many users anyway]--a question that is not "quality" according to the above definition in that it is not interesting or original, but one that is posed by a student who is clearly struggling to learn the concepts. An example of a question that is considered deserving according to the above definition, but not quality, would be the question of whether say $\{(x,y) \in \mathbb{R}^2; x+y=5\}$, is a vector space, asked by a student clearly struggling with linear algebra. Now, such a question is really a duplicate of sorts, questions as to what constitutes a vector space have been asked time and time again, but the student wants this particular example worked out for them.

Now, as to how the goals of A) and B) as above diverge from each other: If we try to help the "lost" students--which generally means entertaining the not-quality questions by giving hints in the comments--we will only encourage more of such non-quality questions.In any event, even EoQS Enforcement of Quality Standards seems confused on the goals of this site: The OP in the EoQS Thread seems to say outright near the top that the goal is A) but then they say that the rule is instead to "Encourage New Users To Improve Their Posts" via comments--when it is damn near impossible to take a basic question from, say, Linear Algebra 101 that has been asked many times on here in varying forms and make it a "quality" question. In many instances the only way to "improve" such a question is to give a hint, which leads to the site having a bunch of lower-quality questions that are answered in the comments. [This is one of my beefs with EoQS that I expressed before on Meta, but I digress.]

Now to turn a bit back to the topic of this thread. From what I've observed: If someone feels the goal of this site is B) as well as A), then they will tend to often use the comments to give extended hints or answers in many instances. We see that in the comments here in this very thread. I personally think that instead, a question should either be answered [in the right place--the Answer box below] or closed. But, from previous posts here on Meta and from observing MSE itself, I am aware a lot of people disagree with me here, and so I don't think the answering in the comments will go away. A critical mass of posters here see this forum see the goal of this site to be B) as above as much as A), and maybe they don't mind this site having a bunch of questions seen as less-than-quality, that go officially unanswered but with a tutoring session in the string of associated comments below.

With that said, as it appears the lesser-quality questions are here to stay [not that I like that but it is what it is], I wonder if the following would work to make more people happy on what MSE is supposed to be: Make a tag/subsection, or whatever called Homework or Basic Help. A question posted there would stay up for maybe 1 month. The student would get his help, and it would not be worth the while for someone looking for easy rep to post there, as the question will soon be gone [and the rep that goes with it]. This may be viewed as a bonus for some here...

$\endgroup$
6
  • $\begingroup$ I think the problem is that if (B) is publicly announced as a "goal" then the majority will absolutely rush here knowing that (B) is going to be done. (A) will be a forgotten footnote, which it isn't yet thanks to some community efforts in that direction. I still think that if the site needs to promote the goal (A), then it needs to (1) Show that asking a better question gives you a better answer. This has to be implemented. (2) Reduce the rewards from (B)-type questions or allow an easier transition from (B) to (A)-type questions, which is also difficult to implement. $\endgroup$
    – Sarvesh Ravichandran Iyer
    Commented Feb 15, 2023 at 4:31
  • 1
    $\begingroup$ I think to some extent that MSE does that job properly : bad questions "should be" closed and deleted (and edited if possible) so that (B) type answerable questions reduce in number, allowing (A) type questions to receive attention. I believe that EoQS was an effort in this direction. On the part that says "edit the question and improve it", about a week into curation it became clear to me that we have a lot of questions that really can't be improved. We cannot make "good contributions" out of questions that unfortunately don't have anything to contribute: that's not necessarily a bad thing. $\endgroup$
    – Sarvesh Ravichandran Iyer
    Commented Feb 15, 2023 at 4:36
  • 1
    $\begingroup$ Version 13 of this answer. $\endgroup$
    – Gerry Myerson
    Commented Feb 16, 2023 at 21:04
  • 1
    $\begingroup$ @GerryMyerson probably. But as far as I see it, this is at the root of a lot of issues with MSE! $\endgroup$
    – Mike
    Commented Feb 16, 2023 at 21:07
  • 1
    $\begingroup$ What I meant was, you have now posted 13 versions of this answer. $\endgroup$
    – Gerry Myerson
    Commented Feb 17, 2023 at 1:48
  • 2
    $\begingroup$ @GerryMyerson that was infact exactly as I took it... $\endgroup$
    – Mike
    Commented Feb 17, 2023 at 2:35

You must log in to answer this question.

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