I've just seen an answer closed as a duplicate.

The text in the blue box containing the information about the reason it was closed says "This question has been asked before and already has an answer."... but the thing it's marked as a duplicate of doesn't have an answer.

Is that still legitimate? If it is, shouldn't that reason be changed so it doesn't say the duplicate has an answer that it doesn't have? If the text can't be changed, wouldn't it suggest the intent is that both should remain open until one of them actually gets an answer?

[I'm not trying to suggest that there's something egregious here - and certainly I don't think there's anything but the right motives present - but if I find this quite confusing, it would not be surprising if the OP does. Some additional clarity on this issue would be good, I think.]

  • $\begingroup$ This is strange, because one can't flag a question as duplicate if the original does not have at least one upvoted answer. $\endgroup$ Apr 3, 2014 at 11:40
  • 1
    $\begingroup$ @Andre A moderator can close a question that hasn't been flagged though. $\endgroup$
    – Glen_b
    Apr 3, 2014 at 21:15

2 Answers 2


This is a gray area, but I think it would be best if we changed the text to "This question has been asked before". Consider the case of someone who posts an identical question several times in a row (maybe once per day or so) or because their question didn't get an answer. This seems to happen somewhat regularly.

However, I suspect the moderators don't have the ability to change that text. In light of that fact, we could just be somewhat more cautious about closing threads as duplicates of unanswered questions (although I would still do it in many cases).


If the message "This question has been asked before and already has an answer..." was not carelessly written, then it just covers cases where the question is a duplicate and the OP can find the answer elsewhere in the forum. Nevertheless what @gung refers to is also an existing matter that should be handled: serial posting of the same unanswered question, or just two identical questions, irrespective of whether the first has an answer.

Pehaps this should be turned into a "feature request" -to close the question as duplicate while having the choice to select one of two messages -the current version of the message or its truncated version that gung proposes.

Or, perhaps it would be cleverer to close the older one, since it appears that newer questions have more chances of being answered... which would require another message:"This unanswered question has just become a duplicate -of a newer one"!

  • $\begingroup$ +1 for "close the older one". I often think that's the better solution, but it doesn't seem to be used by default. $\endgroup$ Apr 6, 2014 at 17:06
  • 1
    $\begingroup$ @gung As a matter of policy I believe this should be decided by each SE separately, through a users-vote or a decision by the moderators. The problem is that the existing set up of closing questions should be enriched in order to close the older question and label it with the correct message, and this I think requires a feature request. $\endgroup$ Apr 6, 2014 at 19:02
  • 3
    $\begingroup$ Usually the older thread has accumulated comments and/or replies, indicating closure would be inappropriate. For orphan posts, though, which have gotten no attention at all, it does make a lot of sense to close them in favor of the newer one. $\endgroup$
    – whuber Mod
    Apr 7, 2014 at 20:42
  • $\begingroup$ @whuber This is indeed a needed distinction for older posts, thanks for pointing it out. About that "feature request", should somebody open a new meta-question for it, or it is up to moderators to take it to the SE engineers? $\endgroup$ Apr 7, 2014 at 20:55
  • $\begingroup$ Mods rarely contact the SE team about suggestions posted on our meta site unless they are truly urgent. Presumably, somebody on the SE team will review these posts eventually. For more attention to this issue--which concerns an SE-wide behavior--consider posting your suggestions on meta.SO. $\endgroup$
    – whuber Mod
    Apr 7, 2014 at 20:59

You must log in to answer this question.

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