2

I wrote a question and edited it fairly quickly (within two hours) to make it more focused by explicitly asking for textual evidence.

Two answers were posted before the edit and two after, but none of them really present textual evidence. Some of them are decent answers but they don't answer the current question, or even the original question (IMO).

Would it be appropriate to flag them as not-an-answer? If not, what can I do to encourage the answerers to actually answer the question?

  • As it turns out a mod seemingly did agree and has deleted the two answers I flagged. I have commented on the other two requesting textual evidence to back up their speculation. – curiousdannii Jun 15 '14 at 11:47
  • Did you consider that when the answer to your question is supported only by negative evidence there is no textual evidence? For example, in your question, if it is true that "no one knows anything more about Val", is it substantial proof that there exists no passage that indicates the opposite? If they did know more, there could be a passage to prove this, if not, there may simply be a lack of such passages. Your demand for textual proof may be a false pretense, as such, since they are addressing the actual question, failure to provide a quotation does not make them non-answers. – Gorchestopher H Jun 21 '14 at 16:00
  • @GorchestopherH the answers don't say that, they make positive assertions without evidence. If they had done that I would have been happy. – curiousdannii Jun 21 '14 at 23:33
  • They are making negative assertions with circumstantial evidence. You question asks if a party A has more information than party B which causes them to act differently. Both answers assert that there appears to be no evidence that party A is more informed, and argue that their difference in behavior demonstrates ignorance rather than superior information. How can this be proven other than a lack of evidence to the contrary? – Gorchestopher H Jun 22 '14 at 21:19
  • I just want them to say there's no evidence! – curiousdannii Jun 22 '14 at 23:08
8

TL;DR: We don't delete answers just because we don't like them, or because we don't think they're good enough. We only delete answers because they're not answers.

Extremely related: Is deleting really wrong answers ok according to the rules?

As I see it, there are four basic categories of answers (YMMV):

  • Great Answer - an answer that clearly addresses at least most of the issues brought up in the question - with evidence and/or references to back it all up. These are the answers that get my up-votes, regardless of whether or not they are necessarily "correct".
  • Marginal-Good Answer - an answer that clearly address at least some of the issues brought up. These can sometimes be speculation based. Evidence or references may or may not be presented, and may only support the concepts, but not offer definitive proof. These are the answers which may or may not receive an upvote from me.
  • Marginal-Bad Answer - an answer that only addresses some of the issues brought up in the question - usually with little evidence or few references to back it all up. Speculative answers with no evidence or proof would fall into this category. These answers may or may not get a downvote from me for not being useful.
  • Not An Answer - a comment or response to the question or another answer - with no actual attempt made to provide an answer to the question. These are the ones I'll flag for deletion.

To me, the answers on your question fall into the "Marginal" categories of answers. They're not great answers, but they make an attempt using logical speculation. As such, they may not be "useful" or "helpful" to you, and they may not be correct at all - but they ARE answers.

On this site, we generally encourage users to vote up answers that are good and helpful, vote down answers that are not not good or not helpful, and only delete answers that aren't actually answers.

  • I like your four categories of answer types -- succinct and distinctive. Makes sense to me. We don't delete answers just because they don't match what we might be envisioning as the perfect response. ETA: And editing and drastically changing the question after it already has answers is the height of unfairness, IMO. Bad form, even if that wasn't the intention of the OP when editing. – Slytherincess Jun 16 '14 at 14:00
  • If the question is "what textual evidence is there for X?" on what grounds can you say that an answer which doesn't discuss any textual evidence is an actual attempt? – curiousdannii Jun 17 '14 at 12:17
  • @curiousdannii 1) You didn't ask for "textual evidence" until two of the answers had been posted. Though, to your point, of of the remaining answers was posted after your edit. The other was possibly in the process of being written when you edited. 2) Sometimes an answerer doesn't have quotes/references handy. The lack of sources doesn't make their answer wrong, just unreferenced. 3) Sometimes there aren't any "textual evidence". In such cases, derived logical speculation is as good as it's going to get. – phantom42 Jun 17 '14 at 12:35
  • @curiousdannii Like I said in my answer above, whether or not the two existing answers are "good" or "helpful" is debatable and up to each user - but they are answers in the eyes of this community. – phantom42 Jun 17 '14 at 12:36
  • @phantom42 well that's what I'm trying to figure out. This community's standard for 'answer' is different that what I expected. I can't force it to change but I have been seeking clarification. When a question is asking for textual evidence then speculation doesn't answer it. Answers should either provide textual evidence or state there isn't any. If someone isn't in the position to look for references then why should they feel they are in the position to write an answer? These are sites for experts after all! If we wanted speculation we'd go to reddit or some other site. – curiousdannii Jun 17 '14 at 12:42
  • Sorry, "Though, to your point, ONE of" – phantom42 Jun 17 '14 at 12:45
  • We've given you clarification. You've got five answers here echoing the same basic points. If you don't agree, that's fine, but the community consensus is clear. – phantom42 Jun 17 '14 at 12:47
  • I don't write answers unless I have some sort of reference I can use to justify my answer, but I don't - and can't dictate how other users work. If they choose to write an answer that can't be backed up, that's their prerogative. You and I get to vote on whether their answer is helpful, or not - or even whether or not their answer even is an answer. If you don't think it's an answer, then go ahead and flag away. But the community disagrees with you. – phantom42 Jun 17 '14 at 12:50
  • I'm definitely getting closer to understanding the consensus. As I wrote to Richard above, I'd like to understand what "address the question" means exactly. – curiousdannii Jun 17 '14 at 12:51
  • Is it wise to bring "helpful" into the debate? I'm sure there are lots of potentially very helpful or interesting non-answers. That doesn't excuse not answering the question. – curiousdannii Jun 17 '14 at 12:52
  • You're welcome to try. But what a vote means to each person, and what "helpful" or "not helpful" means varies from one person to the next, and we've generally avoided ever trying to tell a user how or why they should upvote/downvote. – phantom42 Jun 17 '14 at 12:56
  • The problem is that many questions are unanswerable and launched by false pretense. Works of fiction are created from a discrete number of perspectives. Exactly never are all possibilities directly addressed in text. Some questions ask for things that simply do not exist and as such are addressed with a hypothesis. – Gorchestopher H Jun 22 '14 at 21:32
10

The "not an answer" flag system is used for identifying answers that are basically pointless.

Since you can't expect every moderator to be an expert on every canon universe, the general rule-of-thumb should by that any reasonable person should be able to recognise that the answer provides no useful content.


Per Anna Lear's answer to this question on another stack

A bad or a wrong answer is still technically an answer. Moderators aren't here to judge the correctness of answers. That's what the voting system is for, so the right way to handle those is to downvote, edit, or leave a comment.

The "not an answer" flag is for posts that are either completely unrelated to the question, a "me too" kind of post, a follow-up question asked in an answer, etc.

  • So in my situation the answers do not answer the question but instead answer a different related question. Is it appropriate to flag them? – curiousdannii Jun 15 '14 at 11:46
  • @curiousdannii - If the answer sheds no light on the question, then you should flag it. – Valorum Jun 15 '14 at 12:20
  • 3
    @curiousdannii - FYI: The two current answers might lack referencing but even the most cursory glance suggests that they do address the question. As such, neither would be flag-worthy – Valorum Jun 15 '14 at 12:26
  • can you point me to a definition of "address the question"? I understand that wrong answers don't get deleted. A wrong answer would be one that says "on page 123 it says blah" when it doesn't. "I think character x thinks blah" isn't a wrong answer to a textual evidence question, it's a non-answer. – curiousdannii Jun 17 '14 at 12:48
  • @curiousdannii - you're plain wrong. If the answer relates directly to the question (e.g Q. "what is 2+2", A. "The answer is 5") then it's still an answer. The 'rightness' of it doesn't matter. – Valorum Jun 17 '14 at 15:46
  • I agree that's a wrong answer, but I don't think it's what's happened with the answers on my question. Anyways I'd like to drop this all, though if "address the question" is defined anywhere I'd like to read it. Thanks you for your comments. – curiousdannii Jun 17 '14 at 23:58
  • 1
    @curiousdannii If we applied your logic to questions, we'd close yours as it demands text that does not exist. – Gorchestopher H Jun 22 '14 at 21:40
4

No, it wouldn't be appropriate to flag them as not-an-answer.

Not an answer flags are only for, well, things that aren't answers - e.g. requests for clarification that should've been comments, spam, insults (like, "that's a stupid question, everyone knows...")

You should never use not an answer flags for answers that are wrong. It doesn't matter how wrong they are: totally off-base or (as in this case) just missing supporting quotes. The moderators who look at your flags can't be (and aren't expected to be) experts on every work that might be asked about here, they're role is to handle exceptional cases that ordinary users can't.

Instead, downvote them or leave comments (as you've done).

  • Well that's different from other sites I'm more familiar with such as Christianity. Oh well... – curiousdannii Jun 15 '14 at 8:44
  • See, I don't buy that the not-an-answer flag should be so restricted, because spam and insults have their own flags. Requests and comments clearly qualify. But if you don't think they attempt to answer the question then you shouldn't feel shamed into not flagging. It's subjective. If you say not to flag these kinds of questions then the mods are effectively saying their opinions are the only correct opinions. – curiousdannii Jun 15 '14 at 8:54
  • 1
    No, that's exactly the opposite. You're asking the mods to delete answers you think are wrong. But SE policy is that users determine what's wrong or right (not mods, not the OP), and upvotes/downvotes are the way to express it. Wrong answers won't be deleted. – Avner Shahar-Kashtan Jun 15 '14 at 10:23
  • They don't answer the question wrongly because they don't address the question! I am an experienced flagger on other sites. This site's conventions just seem different (and worse, to me.) – curiousdannii Jun 15 '14 at 11:44
  • 4
    They certainly do seem to address the question, to me. They may not do so in a thorough enough manner for your liking, which is certainly a legitimate concern, and they not address 100% of the point you requested in the question (the textual evidence), but to say that they are not different than an answer saying "I didn't read the books but in the TV show she's hotttttt!!!1" (which is the sort of things I would flag as Not An Answer) seems a bit excessive. – Avner Shahar-Kashtan Jun 15 '14 at 14:09
4

I think you should alter your behavior in the future. Changing a question such that current answers no longer address the question is sort of abusing the edit feature on the site. Next time, post a new question that asks exactly what you're asking for. In that new question, reference and link your original question, and make it clear how the new question differs.

This way the questions and answers on the original don't cease to make sense, they're all addressing the same thing. For the original question, mark whatever answer you find acceptable as 'accepted', if one of them addresses the question you posted, and move on to the new question.

  • 1
    Yes it is unfortunate that I couldn't get the question right to start with. The intention was always the same, and the body was essentially the same, but the title suggested a different question. – curiousdannii Jun 16 '14 at 0:42
  • In the past I have edited some of my own questions so drastically that existing answers were rendered unrelated -- this was a terrible practice on my part and I'm glad you pointed it out. I bring this up in my comment to @phantom42 as well. – Slytherincess Jun 16 '14 at 14:11
1

It might probably help you to look at the text of the "not an answer" flag reason:

This was posted as an answer, but it does not attempt to answer the question. It should possibly be an edit, a comment, another question, or deleted altogether.

This is IMO inappropriate for these answers because they do attempt to answer the question. The only place they fall down is that they don't meet your criteria of textual evidence, but otherwise they are honest and in-good-faith attempts at providing an answer.

Of the four options given for dealing "not an answer", none of them are really appropriate either. The appropriate thing to do is for them to be edited to add the textual evidence you're looking for.

So that establishes that "not an answer" is not the appropriate way to deal with this. Adding a comment requesting such evidence is good, and you can always downvote the answer if you feel it's not helpful to you.

However it's worth noting that regarding textual evidence on this SE, and particularly with regard to a work that the author has not yet completed (cases also exist where an author has died before completing a work), the hard reality is that sometimes this evidence just does not exist. In cases like that you just have to accept that an answer without textual evidence is the best you're going to get.

  • Or the best would be for an answer to say there is no textual evidence... – curiousdannii Jun 17 '14 at 12:14
  • I guess I see "attempt to answer" as meaning attempt to answer the question on the questions terms, according to its specifics. Other people see it as meaning attempting to provide something on the same topic even if it doesn't strictly answer the question. – curiousdannii Jun 17 '14 at 12:15
  • 1
    @curiousdannii - You're always welcome to research your own answer or edit/improve existing answers with extra referencing. As long as you're augmenting the existing answer, that's perfectly acceptable – Valorum Jun 17 '14 at 15:51
  • 1
    @curiousdannii - I'm not sure that other people see it the way you're claiming. My experience on this site is that most users genuinely want to be helpful, even if the end result is not 100% perfect or to-the-letter. SE culture allows speculative answers without direct textual evidence provided an otherwise reasonable case can be made: blog.stackoverflow.com/2010/09/good-subjective-bad-subjective. I'm sorry that your expectations are different. – user8719 Jun 17 '14 at 16:00
  • 2
    Quite frankly, I would personally abandon any SE site that completely required me to answer every question exactly to specification by OP with threat of deletion if I didn't meet the OP's criteria well enough (doubly so if they're allowed to just randomly change the question after answers have been posted). *this assumes that my answer is an actual attempt at answering the question, and not just some random comment/complete garbage answer – phantom42 Jun 17 '14 at 18:59
  • @JimmyShelter thank you for that link. I guess I'm trying to navigate the difference between "Is X likely?" and "Is there evidence for X?" questions. Providing evidence for both is awesome of course, and if you can't I think we'd all agree that good solid speculation is fine for the first. But is it for the second? My thinking is that if someone said "There's no evidence but my theory is ..." that's a good answer, but if they don't say that there's no evidence it's not just a bad answer but a non-answer. – curiousdannii Jun 18 '14 at 0:07
  • @phantom42 I hear what you're saying, but no one seems to be able to objectively say what "actually attempt to address the question" means. It seems as subjective as the answers. – curiousdannii Jun 18 '14 at 0:08

You must log in to answer this question.

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