It seems like there are constantly problems on the various Stack Exchange site I drop in on with people wanting to close questions, but not finding exactly the right reason to close them. Take for instance, 'If Christopher is a “carrier of Christ” then what is Jennifer carrying?' At least one reader of the question objected to it because it had an overly attractive title. Never in a million years will this reason for closing a question be added. The reader didn't like the question, but didn't have a convenient outlet to express the problem.

More personally, a question at the new Philosophy site was "closed as not a real question" with the description:

It's difficult to tell what is being asked here. This question is ambiguous, vague, incomplete, or rhetorical and cannot be reasonably answered in its current form.

At least one moderator suggested that it was really closed as a "general reference" question, but the site doesn't have that reason implemented yet. Maybe this is standard, accepted practice, but the close reason chosen could not be more misleading to the asker. The question simply can't be rephrased as a clear, concrete, complete, or true question that would be reopened. It was in fact reasonably answered (twice by people who voted to close!) in its current form. You might as well close it because it "smells too much like red fractals". At least that way nobody can claim that the close reason was meaningfully chosen.

Now I really hate the idea of closing questions, but I despise a system that forces people to behave dishonestly. Let's set aside the question of whether it's a good idea to close these questions or not (the first one hasn't even gotten any close votes at this time), and address the dishonesty inherent in itemized close reasons on Stack Exchange.

I've come to think of the moderation system on Stack Exchange as the Western posse system. At some point in these movies the sheriff of some frontier town will be convinced that he can't handle the demands of law and order by himself. To deal with the problem, he deputizes whichever of the citizens he trusts, and hands out guns and badges. On a good day, the posse rounds up the outlaws and dispenses rough justice. On a bad day, dead bodies are strewn along the main drag, salon, cemetery, etc. The posse becomes judge, jury and executioner.

You might think I dislike the system and you'd be partially right. Its prone to bad outcomes. But so are all the other moderation schemes I'm aware of. At least this scheme has the virtue of being extensible and somewhat democratic. If you aren't willing or able to establish a complete justice system, the posse system serves to keep law and order fairly well.

Filling out the reasons to close is a bit like handing a deputy their gun and listing the 7 (or whatever) reasons to fire it. You better be convinced that those reasons are comprehensive and bulletproof (pun!), because if not you'll end up with people not pulling the trigger when they should or else shooting an innocent man. I'm not convinced that the set of closure rules developed on Stack Overflow works as well on some other site with a different culture. And I'm certain that the rules developed on Stack Overflow are misapplied everywhere I look.

We've already given trusted users the power to vote to close questions and we know that people will pick the most likely reason to close a question rather than constraining themselves to the particular rules delineated by the system, so why not go whole-hog and let people write the specific reason they decided to vote to close? At least that way someone could vote to close a question for whatever reason it bothers them and everyone else can decide if they are right or wrong.

(I suspect I was once on the side of demanding rules for closing questions so that people wouldn't abuse that power. If so, my only excuse is that I'm older and therefore wiser now.)

Update: The problem isn't that there are rules of engagement, it's that they are:

a) potentially broken for some sites,

b) often ignored,

c) not regularly enforced,

d) and lead to dishonesty.

My guess is that the official answer will be: no. I'm fine with that, but not before the problems are at least put into collective consciousness.

  • 1
    "At least one reader of the question objected to it because it had an overly attractive title." ... what the heck. There was recently an even worse case on SO. Jun 21, 2011 at 21:22
  • 5
    > You might as well close it because it "smells too much like red fractals"
    – jcolebrand
    Jun 21, 2011 at 21:23
  • @BoltClock: Long story. The reader may have a point actually since the title was potentially unrelated to the actual question. My point in bringing it up is that you can't really predict what sort of issues people might have with specific questions until they are asked. Jun 21, 2011 at 21:27
  • Hah, it ought to be the other way round: If a question is not ambiguous, vague, incomplete, or rhetorical then it surely doesn't belong on a philosophy site! Jun 21, 2011 at 21:33
  • @yoda: Ha, Ha. (Sigh.) I'd be curious to know if any other site has as high a closed question ratio as the philosophy site does. It's like Tombstone, Arizona over there. Jun 21, 2011 at 21:39
  • 2
    @yoda: In that respect, the Philosophy site is widely misunderstood; it is actually a site about logical reasoning, not pondering the meaning of life.
    – user102937
    Jun 21, 2011 at 21:57
  • @Robert: I think you will find that I respectfully disagree. ;-) (But there are bad questions over there and some of them are even closed.) Jun 21, 2011 at 22:06
  • @Robert: I have not checked that site out yet! My comment was just plain silly and a prime candidate for the "not constructive/off-topic" flag :) Jun 21, 2011 at 22:08
  • @Jon: Check the FAQ, and the Area51 proposal. They both fall squarely on the "Logical Reasoning" side. The site should have been called "Logical Reasoning," not Philosophy.
    – user102937
    Jun 21, 2011 at 22:23
  • @Robert: I think you will find that the proposal has been all but dismissed by a portion of the population. For instance, several of the "on-topic" questions from the definition stage are already closed as "off-topic". I'm not complaining, but I think the situation illustrates well some consequences of the posse system. Jun 21, 2011 at 22:27
  • Keep in mind that the Philosophy proposal is still in beta, and the on-topic definitions may still be in flux. Until that is settled, the site will remain somewhat unstable. I just noticed this meta question which illustrates the problem eloquently.
    – user102937
    Jun 21, 2011 at 22:34
  • @Robert: Preaching to the choir man, preaching to the choir. Jun 21, 2011 at 22:35
  • I'd really love to see the reason picked for voting to close this question. Jun 22, 2011 at 18:57
  • I can not tell what is being asked here or what the bug is. retagging to discussion
    – waffles
    Jan 24, 2012 at 5:43

2 Answers 2


A few thoughts:

  1. The closing system is modeled around those reasons that have been historically proven to cause online communities to derail. The only purpose for closing a question is to eliminate these reasons, thus keeping the site safe from the clutches of those who would derail it.

  2. The reasons a deputy might pull the trigger are not arbitrary. To the contrary, they are very specific, well thought-out, limited in number, and principle-based.

  3. If a close reason can be found that supports a substantial use case (e.g. General Reference) then it can potentially be added to the list of close reasons. But adding a close reason requires careful consideration about how it will affect the overall close dynamic.

  • 2
    @Jon: Regarding the "Attractive Title" question, why didn't someone with sufficient rep simply edit the title?
    – user102937
    Jun 21, 2011 at 21:52
  • I agree with these answers, but there's a key missing piece of item #2. In the real word, peace officers are given specific rules of engagement and they are drilled into them via training. Every case of a police officer firing their weapon is examined by some authority who has power to discipline the office in cases the rules are broken. In the real world, an officer of the law can be jailed for breaking the laws they are called to uphold. None of these conditions currently exist on Stack Exchange sites. Jun 21, 2011 at 21:56
  • 1
    @Jon: Moderators perform this function.
    – user102937
    Jun 21, 2011 at 21:58
  • @Robert: The day's young yet--it might happen. But the point has much less to do with the specific complaint as with the problem of someone feeling like something's wrong with a question and feeling constrained by the system to simply complain about it. (I just was wandering past when the problem flared up and it seemed like an interesting example.) Jun 21, 2011 at 22:00
  • Which function? The training seems lacking and I see no evidence of moderators reviewing close actions as a rule. I suppose users are disciplined for extremely bad behavior, but I really don't think anybody will run afoul of any rules in the cases I cited or most of the cased I didn't. Jun 21, 2011 at 22:04
  • 1
    Don't forget that users can also vote to reopen questions. FWIW, I don't think the PostModernist question should have been closed, but then again, the Philosophy site is more like MathOverflow than it is like Math.SE.
    – user102937
    Jun 21, 2011 at 22:05
  • Anyway, moderators review flags, not close actions (unless they are flagged). If you see a question that you believe shouldn't have been closed, flag it and if a moderator agrees, they will reopen it.
    – user102937
    Jun 21, 2011 at 22:09
  • @Robert: Don't get me started! Would that not be a reason for letting voters actually say what they mean by closing a question? If you really, honestly think a title attracts undo attention, wouldn't it be better to be able to say that? If it gets closed as "Not a Real Question" or whatever, the solution to editing a question for rehabilitating becomes that much harder. Jun 21, 2011 at 22:10
  • 1
    No, you can't make close reasons arbitrary. Using your deputy analogy, they could then pull the trigger for any reason that makes sense to them. Close reasons must be established under a highly-disciplined framework. The role of close voters is to decide whether or not to close questions within that framework, not to invent ad-hoc rules for closing.
    – user102937
    Jun 21, 2011 at 22:12
  • @Robert: Exactly. But users already apply ad hoc rules for closing questions. They just have a handy "here's my approved justification" checklist to click on. That's what's wrong with the postmodernism question in my opinion and why I brought it up! Jun 21, 2011 at 22:15
  • 4
    They're not supposed to. While the "Not a Real Question" close reason seems vague, it actually is quite specific; it's not a catchall for "I don't like the question," and using it this way to close questions is abuse.
    – user102937
    Jun 21, 2011 at 22:17


See, this would put closing in the same place that down-voting has been for the life of the site: by default, there's no reason given; you must rely on the skill and good will of those voting to provide one.

I'm a huge fan of this when it comes to the normal voting system (the one used for ranking posts). However, there are two big differences when it comes to the moderation system:

  1. If your question gets closed, you won't get any answers.
  2. Close-voting isn't anonymous.

The reason #1 matters should be pretty obvious: since not getting answers defeats the whole purpose of asking a question, you should probably have some help understanding why it was necessary. Making explanations optional puts this in jeopardy. And by making the reason free-form text you are making it optional - as we've discussed so many times WRT down-voting, there's really no good way to force someone to enter a useful bit of text.

You might think #2 would work to mitigate the problems #1 would cause for free-form text... Surely peer-pressure would put things to right, eh? Weeellll... I'm not very confident of this. A lot of people are not terribly tactful when leaving criticism - in fact, it's more than a little difficult to leave constructive criticism without coming off as rude or condescending. Expecting five people to manage it on every closed question is a bit optimistic. So even in the best-case scenario where everyone does their level best to write up a good explanation for closing, you're still more likely than you are today to leave the author feeling insulted or belittled.

You better be convinced that those reasons are comprehensive and bulletproof (pun!), because if not you'll end up with people not pulling the trigger when they should or else shooting an innocent man.

Sure, that's a given. Even if this free-form thing was a reality, you can bet there'd be some massive Meta post with a list of reasons and explanations for folks to refer back to...

We've changed, added, removed, and refined the close reasons continually since they were introduced. Just last week, "Subjective and Argumentative" was replaced with "Not constructive" across the entire network, with an explanation crafted to better describe why a broad class of questions are simply not appropriate for these Q&A sites. Because folks will vote to close stuff they think doesn't belong, and simply pick whatever reason best matches the one in their head, we strive to choose reasons that are easily recognizable as good matches for valid reasons and helpful in describing those reasons to both closer and author.

If you want to add more detail, or think you can do better than the stock explanation, leaving a comment is always an option. If you think another user's chosen close reason is invalid, dispute it - in comments, by voting to re-open if the question has already been closed, or via a moderator flag if it's clear that intervention is necessary.

  • This is a fine answer. But I will point out that in many ways voting to close is an easier choice than downvoting. One reason seems to be that picking a reason from a list is easier than writing a critique. Further, its a far more powerful option to dealing with bad questions. I suppose I'll try reopening some of the questions that seem to have poorly picked close reasons and see if things improve. Jun 22, 2011 at 18:54
  • @Jon: that is by far the most effective response. And... state your reasoning - you might be surprised how far an effective argument can take you.
    – Shog9
    Jun 22, 2011 at 19:01

You must log in to answer this question.

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