15
$\begingroup$

I'm new to SE but practically every day I come across constructive, thought-out questions that are very quickly marked as duplicates -- while a simple comparison with the alleged duplicate shows very clearly that they are asking a quite different thing.

Often, the two questions are about the same topic but clearly focus on different aspects of it. Other times, the older one (or one of its answers) is an answer to the newer one, which could not be known in advance by the person who asks the new one.

Since users are enabled to mark a question as duplicate, do we also have a way to "defend" it when we think that such a mark is wrong?

$\endgroup$

2 Answers 2

Reset to default
16
$\begingroup$

There are a few ways:

  • If you have >3k rep, you can directly vote to reopen. This will put the question in the reopen review queue.

    Once in the reopen review queue, users with >3k rep can vote to reopen. If five users vote to reopen (either through the queue or directly) the question gets reopened. If the question receives enough Leave Closed votes (which I think is three votes) then it leaves the queue; people can still vote to reopen directly (in which case it will possibly go back to the queue) but it won't be shown to reviewers.

    Once the review is completed, its results are publicly available (example) and they are accessible from the question's timeline (example).

  • If a closed question is not on the reopen queue, editing it will bring it into the queue (exact criteria here). You can edit directly if you have >2k rep; if not, then your edit will be peer-reviewed (i.e. it needs two approvals on the suggested edits review queue).

    This is the most important part of the reopening process. If a question is close to a previous post but it is not quite answered by it, then it is important to make sure the new question references the previous ones and explains why it's different. Ideally this should be done by the original poster, but if you see a clear opportunity for an edit like that which does not conflict with the OP's intent, then go for it.

    There are important reasons why we close duplicates, most centrally because duplicate Q&As dilute and fragment the knowledge base we're building, and diminish the usefulness of this site as a reference source. If new questions cover new ground they should direct answers onto that new ground, because what we really want to avoid is duplicate answers in separate threads.

  • Finally, if you think a question is not a duplicate, leave a comment explaining why! Sometimes people are hasty (because reviewers are human), but a well-argued comment explaining why a question differs from the proposed duplicate can and does make a difference when reviewing.

For closure, it is possible for users with under 3k rep to flag as off-topic; this won't cast a closevote but it will place the question in the closure queue if it's not already on there. For reopen votes, this is not possible; it was asked on the Mother Meta and is currently status-declined - see the answer there for the reasons why.

Finally, let me comment on your second paragraph for a bit:

  • Often, the two questions are about the same topic but clearly focus on different aspects of it.

    This is a legitimate reason for reopening, but it often requires an edit on the new question to sharpen the differences and direct the answers onto the new ground.

  • Other times, the older one (or one of its answers) is an answer to the newer one, which could not be known in advance by the person who asks the new one.

    This is not a reason for reopening. This could indeed be known to the person asking the new question - we expect posters to search for previous questions before asking. Very often, the proposed duplicate will actually have been shown to the OP as they were typing the question:

    If people ask a question that's already been asked and answered, then no biggie: the question gets flagged as a duplicate, the OP gets directed to their answer, and the site gets a new waymarker towards that answer that can help future askers with the same question find the answer. We don't hold asking-duplicates against people (unless it's consistently repeated), but that doesn't mean the duplicate question should remain open.

$\endgroup$
11
  • $\begingroup$ Thanks for the time to give this practical answer. About the second part, namely "other times, the older one (or one of its answers) is an answer to the newer one, which could not be known in advance by the person who asks the new one": Although this is the minority of cases, it definitely happens, even after googling; put plainly, one doesn't know what they don't know. As about the proposed answers while typing, they prove to be useful a small percentage of the time (and how could they be otherwise). $\endgroup$
    – Helen
    Apr 25, 2017 at 16:52
  • $\begingroup$ @Helen The point is that we don't take it against people when they ask duplicates (unless there is a clear lack of due diligence or repetitive behaviour), but that doesn't mean that they're not duplicates. It is OK to not know what you don't know - but once the duplicate is pointed out, we evaluate the questions on their own merit. $\endgroup$
    – Emilio Pisanty
    Apr 25, 2017 at 16:59
  • 3
    $\begingroup$ I have a personal counterexample - it's not PhysicsSE but it's handy. I asked why a certain software behaves differently than before under a specific linux edition; and this was marked as duplicate of a question about which exact code this software invokes. The answer to my question has to do with the fact that this software now calls another code, which obviously didn't/couldn't know, therefore the linked question and its answers form the answer to my question. Bottom line? There is a difference between asking the same thing and asking something whose answer has already been written down. $\endgroup$
    – Helen
    Apr 25, 2017 at 17:14
  • 2
    $\begingroup$ @Helen I fail to see the distinction. Whether the OP could have known that their question is a dupe or not is irrelevant to whether it is a dupe or not and to whether it should be closed as a dupe or not. That third one is entirely dependent on whether it's a dupe or not and entirely independent of whether the OP could have known beforehand. If they couldn't have known - again, we don't hold grudges. If you want to discuss this further, it's best to take it to chat. $\endgroup$
    – Emilio Pisanty
    Apr 25, 2017 at 17:42
  • 1
    $\begingroup$ @Helen Duplicate questions themselves aren't really the problem (although we do expect people to avoid asking obvious dupes). As Emilio says "what we really want to avoid is duplicate answers in separate threads". So the precise content of the new question is only of secondary concern, the main thing is if the answers in the old question (aka the dupe target) adequately answer the new question. In some cases, it may be helpful to post a comment to the new question that explains how the old answer applies to the new question. $\endgroup$
    – PM 2Ring
    Apr 28, 2017 at 6:43
  • 2
    $\begingroup$ FWIW, I'm only a newbie here in Physics, but I'm quite aware of the general Stack Exchange policy re dupes. Sometimes on Stack Overflow the answers on the best dupe target may contain the info required to answer the new question but it may be hard for a beginner to see how to apply that info to their particular problem. In such cases, before dupe-closing we write a short answer that summarizes and synthesizes the necessary info and add a snippet of code that shows how to apply that info to the OP's specific case. This works ok over on SO, it may be less applicable to Physics.SE. $\endgroup$
    – PM 2Ring
    Apr 28, 2017 at 6:59
  • $\begingroup$ I think @PM2Ring 's last comment illustrates very well the distinction between the "answer existing somewhere (under a different question)" and "asking the same question". $\endgroup$
    – Helen
    May 4, 2017 at 16:31
  • $\begingroup$ @Helen We can keep discussing for discussion's sake if you want, but in both cases the question needs to be closed as a duplicate, so I don't know if there's anything new to be said here. $\endgroup$
    – Emilio Pisanty
    May 4, 2017 at 16:36
  • $\begingroup$ Yes, if the examples with the software questions cannot illustrate the point, then we obviously interpret the same facts differently. $\endgroup$
    – Helen
    May 4, 2017 at 16:43
  • $\begingroup$ @Helen I simply don't understand what it is you're trying to say. Yes, there are cases where OP could not have known their question was a duplicate. So what? Are you claiming that those questions should remain open? Are you proposing any change in policy for those questions? Or is it just discussion for the sake of discussion? Have you got any concrete examples on this site? Or is this all just academic? $\endgroup$
    – Emilio Pisanty
    May 4, 2017 at 16:55
  • $\begingroup$ It's sure not intended to be academic, but the length of the discussion has blurred the point. I think I can summarize what I try to say in another way: "If someone who has done his research cannot see that the answer for A also forms the answer for B, then asking B will also be useful for future users." Apart from that, I'll try to defend such individual cases more eloquently wherever / if they occur. $\endgroup$
    – Helen
    May 4, 2017 at 17:25
4
$\begingroup$

I think the most important thing to do after getting a comment which suggests that your question is a Possible duplicate of some other question, is to explain prominently in your question why you think it is not a duplicate. It is important to do this as soon as possible, before votes accumulate, and certainly before the question is put on hold. Reopening is very much less likely.

Posting your reasons in a comment is not as effective as posting them in your question. When I review a question for reopening I rarely consider more than the edits made to the question after it was closed. (This is inconsistent of me, because I post a comment when I nominate a question for reopening.) I review comments before casting a closing vote.

Posting a comment pinging the user who voted against your question as a duplicate is more effective than posting a general comment. Again it is important to act quickly before votes start accumulating. Much as we like to think we vote independently, we are more likely to cast closing votes when we see that others have already done so.

I also have an inbuilt bias against reopening questions; I think most reviewers would admit to having the same bias. To have any kind of a chance I think you must make a convincing argument. It is not enough to say that the answers in the other question were not helpful or that you could not understand them. You must explain in what way they fail to satisfy you.

Neither (in my opinion) is it enough to say that the two questions focus on different aspects. The issue is whether or not one of the answers to the other question answers the aspect you are focussing on. It often happens that an answer is more general than the question asked, so it might cover your case also.

You might pre-empt votes to close your question as a duplicate by citing similar questions and explaining what is different about your question. If you cite possible duplicates, reviewers are less likely to search for them. However, this can also be a liability, inviting your question to be marked as a duplicate if you do not make a good argument for it being different or the other answers failing to answer your question.

As a last resort, if your question is not reopened and you are convinced that you have a good case, you can post a question in Physics Meta asking why your question has been closed. The discussion might persuade voters that your question ought to be reopened.

$\endgroup$
1
  • $\begingroup$ I think this gives real insight. For the record, I was triggered to ask by seeing others' questions marked -wrongly, in my opinion- as duplicates, and I wanted to find out what a third person can do to stop such cases, so it's a little different than what the OP can do, but still. $\endgroup$
    – Helen
    May 4, 2017 at 16:28

You must log in to answer this question.

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