8
$\begingroup$

While browsing new questions asked on this website, it is common to find some of them as duplicates of older ones, and it is then generally productive to indicate it as such, and redirect the OP to the post where their problem was already discussed and solved.

However, I sometimes witnessed a tendancy to mark as duplicates questions about a special case/simpler case of a more general problem. For instance, recently someone asked about Gauss lemma in arithmetic in the context of integers, and a user here suggested that it is a duplicate of an older question discussing Gauss lemma in the context of polynomials.

When I first saw the above situation, I felt like it is not quite right, and I've started wondering whether questions about special/simple cases of a more general problem be marked (closed) as duplicate.

If the OP asks the question formulated in this very special case, it is most likely to me that they are not aware of the generalization. Now, there could be two specific reasons for this. Either the OP just didn't know about it, and it would be enriching for them that somebody points out their problem as a special of a broader statement ; either the OP does not have the mathematical level yet necessary to understand the more general statement. In the first case, such a question may indeed be marked (closed) as a duplicate, whereas in the second, I feel that it would be wrong.

Going back again to the example I gave above, because the OP explicitely asked about Gauss lemma using the formalism of integers, it seemed to me that they probably didn't know about abstract algebra neither about polynomials. Redirecting them to a more general version of Gauss lemma then seems conterproductive to me.

I would like to know what your thoughts are on this matter, and what would be the guideline when such cases occur.

Thank you for reading my post.

EDIT: Eventually, as I found back the question I am referring to, it seems that the user who suggested it was a duplicate deleted their comment. Even though my example may not be relevant anymore, I guess that my interrogation still is valid and maybe could refer to something you also already witnessed.

$\endgroup$
8
  • $\begingroup$ As I think you suspect, there can be distinct issues raised by special cases of general results that admit "simpler" proofs. The impulse of some to identify previous Questions that give a more general result is understandable, and in some instances such a link can be satisfying to the OP. If I understand the focus to be on an elementary treatment of the special case, I often put the link in a Comment that describes it as Related or Closely Related, rather than as a Possible Duplicate. $\endgroup$ – hardmath Aug 4 '18 at 1:25
  • 3
    $\begingroup$ Related (at least to some extent): How close a match we need to close a recurring question as a duplicate? $\endgroup$ – Martin Sleziak Aug 4 '18 at 6:35
  • 3
    $\begingroup$ Re: it seems that the user who suggested it was a duplicate deleted their comment. With the first vote to close as a duplicate, the Stack Exchange software automatically generates comment of the form: "Possible duplicate of ..." When the post is closed as a duplicate, the comment is deleted (again by the SE software). So if it was this type of comment, it might have been created and deleted automatically, without the user typing the comment and deleting it later. $\endgroup$ – Martin Sleziak Aug 5 '18 at 13:49
  • $\begingroup$ ah, this is a good one. Ok, say a there's a popular question about a lemma. We cannot close a popular question about the corollary to that lemma stating it's a duplicate. Also, the same notion applies to polynomial equations. Can we close all solve this second degree polynomial equation questions by citing go see the quadratic equation? I think not. $\endgroup$ – Nick Aug 5 '18 at 18:43
  • 1
    $\begingroup$ @Nick Frankly, if a question would be better answered by pasting it into Wolfram|Alpha, it shouldn't be on this site at all in my opinion. $\endgroup$ – Derek Elkins left SE Aug 6 '18 at 19:02
  • $\begingroup$ @DerekElkins That is equivalent to saying thoughts shouldn't be written down because there are already other brains who can think them. $\endgroup$ – Nick Aug 6 '18 at 19:55
  • 2
    $\begingroup$ @Nick You say Can we close all solve this second degree polynomial equation questions by citing go see the quadratic equation? Frankly, I think that the site would be very well served with by canonical "How do you solve $ax^2+bx+c=0$?" question, with an answer that develops the quadratic formula and another couple of answers that work through a few examples using smaller hammers (such as factoring in a case where things work out nicely). I do think that all questions about solving quadratic polynomials should be closed as duplicates of each other. $\endgroup$ – Xander Henderson Mod Aug 16 '18 at 11:54
  • 1
    $\begingroup$ I don't see any point is having tens or hundreds or thousands of questions that are all asking basically the same thing, only with slightly different numbers. $\endgroup$ – Xander Henderson Mod Aug 16 '18 at 11:54
6
$\begingroup$

I think a rule of thumb is that if you can in the span of $1-3$ comments mention how the question is a special case of a more general question (without relying on too much jargon) then it is worth flagging as a duplicate, and if the user is feeling generous, maybe even leaving said comments.

Otherwise, even if SE is intended to be a repository of knowledge, if either hobbyists or undergraduates have no way to access that knowledge because everything is done in maximal generality, we are probably failing in that goal.

I think usually for copy/paste questions that are overly specific, the rule of thumb can be applied to mantain quality of the site and also possibly direct OP to resources that cover some of the terminology for searching SE or google.

$\endgroup$
2
  • 2
    $\begingroup$ I will often choose to close as duplicate a PSQ that would otherwise be closed as off-topic for lack of context. This seems to me consistent with using duplicates to make the best content more easily discoverable. $\endgroup$ – hardmath Aug 16 '18 at 16:53
  • $\begingroup$ Well that makes sense. If a question will be closed and there is a similar one, then there obvious utility in linking it’s $\endgroup$ – Andres Mejia Aug 16 '18 at 17:37

You must log in to answer this question.

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