15

Recently, the question Was Eric Flint's 1632 the ONLY prominent SF&F work/universe where some (approved) fan fiction is officially considered canon? was asked.

The question asked for a single example, in order to avoid being a list question (list questions are considered a poor fit for this platform, and are considered not constructive by definition).

I feel that if the answer is "yes, Eric Flint's 1632 is the only SF&F work/universe where some fan fiction is officially considered canon", then the question is too localized, as it reflects a single point in time (all it takes is one other franchise approving fan fiction as canon and now the answer is incorrect).

If the answer is "no, there are others", and then we explicitly limit the answer to a single example in order to bypass the list question issue, then the answer isn't terribly interesting, and, by extension, neither is the question ("here's the only example of a franchise that includes fan fiction as canon" might be somewhat interesting, as it represents a unique situation, but "here's two examples of franchises that include fan fiction as canon" does not seem interesting).

I perceive this as a limitation of questions that look for a yes/no answer, where the answer can be proven by a single example. However, that doesn't mean that there could not be other questions in this format that might be of more general interest.

Do we, as a community, feel strongly one way or another about these types of questions? If some should be allowed, what are the criteria of identifying good questions from bad? Or are questions asking for a yes/no answer based on a single example not a problem?

1

4 Answers 4

10

These are some great answers posted, and it seems (so far) that the consensus is that these types of questions do have a place here.

However, I don't feel that any of the other answers sufficiently address how we fix the potential issue of these being treated as list requests.

Perhaps we're over complicating things.

I propose that any question asking "does x exist?" be worded to make no mention of examples. Any good answer will, by definition, include examples. A great answer will contain a fairly comprehensive collection of examples, and explain the context.

Mentioning examples (whether it is "I'm looking for examples" or "I only need one example") seems to be the flag many people identify as indicating a list question.

If the question is worded to simply focus on the problem, we may be able to avoid the issue altogether.

So, using some of the examples I've seen presented here:

  • "Do some vampires suck the heat out of the human body while drinking their victim's blood?" rather than "Are there any works that have vampires sucking the heat out of the human body while drinking their victim's blood?"
  • "Can a person be both a werewolf and a vampire at the same time?" rather than "Are there known werewolf containing works which show someone who is both a werewolf AND a vampire at the same time?"
  • "Can sorcery be defined as distinct from magic?" rather than "What are some examples of works were sorcery is defined as distinct from magic?" (the latter would be pretty clearly a list question, anyway)
  • "Can super villains succeed without super strength or super intelligence?" instead of "Are there super villains that are neither uber smart or uber strong?" (this one is trickier, as it feels more like a list question than the others, largely because "super intelligence" would need to be clearly defined, as would "super villain", and even possibly "succeed", although success could be part of the definition of super villain.
  • "Can an SF franchise involving space travel succeed without green slave girls and FTL?" is an appropriate wording (for a hypothetical, admitted absurd example), although "succeed" would need to be defined.

On major caveat, which is alluded to in the examples above: A pitfall I am seeing in a number of the examples, including the original question that prompted this debate, is that subjective criteria need to be clearly defined within the scope of the question, or omitted altogether. Words like "prominent", "successful", "succeed", or even "officially considered" need clearly defined criteria.

I will attempt to edit the original question, and see if it helps. If this solution seems agreeable, I will also vote to reopen.

9
  • 1
    I really like your proposal. MOST of the extra wording you see as needing to be cut is an attempt to defend the questions from being closed "as list", and not intrinsic to the question's content/intent. Feb 3, 2012 at 16:07
  • 2
    One minor quibble - you - or at least your wording - seem to be conflating "is a list question by design" problem with "is a poor question that CAN generate a list due to vague wording of definitions problem ("super intelligence"). The latter is, indeed, a VERY legitimate concern as you noted in the last paragraphs, and should be addressed, but I strongly feel that it is 100% separate from listness or yes/no questions. Feb 3, 2012 at 16:09
  • I agree that ambiguous definitions are a very different problem, and aren't just restricted to list or yes/no questions. However, it seems like you may be conflating list and yes/no questions? If so, then I feel just as strongly that they are separate topics. If you want support for list questions, then this is the appropriate discussion. I am much less sympathetic to true list questions ("what are some...?") than to the "does x exist?" questions we're addressing here.
    – Beofett
    Feb 3, 2012 at 16:52
  • 1
    Um... I personally distinguish them, but conflated them here because people close yes/no based on being a list. Feb 3, 2012 at 16:56
  • Then I guess I'm missing the point of your previous comment... the "is a list question by design" problem is stand-alone, and nothing in my answer is intended to support questions asking for a list of answers. I mention the pitfall of ambiguous definitions only because they have come up repeatedly between the examples in this thread and your initial question. I didn't mean to imply that this was a problem specific or unique to yes/no or list questions, merely that it is a generic problem that seems to come up slightly more frequently in these types of questions.
    – Beofett
    Feb 3, 2012 at 17:43
  • Sorry - I mean there's 2 classes of problems: the questions with bad/imprecise wording (which isn't directly list related), and the questions which are sorta list related - the latter separated into real list questions and yes/no questions. My 1632 one was accused of BOTH classes of problems, and so is "super villains" one. Better? Feb 3, 2012 at 18:05
  • Yeah, that makes more sense. Thanks for clearing that up.
    – Beofett
    Feb 3, 2012 at 18:25
  • 1
    @DVK, Beofett: In practice, people do not pay to the fine details of wording that you are obsessing over. They just answer with their favorite example. It doesn't matter that the question asks “does some SF have X” rather than “are there examples of SF that has X”. Mind, I don't see the difference either; both are calling for an overview of the SF that has X, and both in fact get examples of SF that has X.
    – user56
    Feb 3, 2012 at 21:25
  • FWIW, I've put the version of how I would edit the question here: dl.dropbox.com/u/2095704/1632-alternate.txt. I go further than your edits, but I think it's a similar goal.
    – Tony Meyer
    Feb 4, 2012 at 8:15
9

Personally, I'd find questions like this useful in a number of cases. I know this is an absurd example, but I'm doing that so it stays away from any real-world questions. Suppose someone is researching a topic, like successful SF franchises, shows, or movies that involve space travel and every example they think of includes specific elements, like FTL and green slave girls. The more he researches, the more he find this is true in every example. So it becomes a question, "Can an SF franchise involving space travel succeed without green slave girls and FTL?"

One exception proves the rule, two or three would support the point that it's not a requirement.

As I see it, the problem occurs with all the "me too" answers where each person wants to cite one example. So I think the issue is not so much in allowing such questions, but in what can be done to prevent endless lists of answers with short (or single item) lists.

9

Questions that explore the state of science fiction and fantasy, the history of the genre, and how and why things turned out the way they did are fascinating questions that aren't just trivia about a specific universe, and are (for at least some) the sort of question that we want to be encouraging.

I don't buy the "too localised" argument at all. The majority of the questions on the site (also the majority that the StackExchange CHAOS employees are pushing) deal with universes that are still being developed. That means that the correct answer can change at any time, with a new book, new episode, new issue, new sequel.

(Ours isn't the only site where this is true. Even on Stack Overflow languages are continually evolving, so, e.g., a new release of Python may change the accuracy of an existing answer).

"Too localised" is, e.g., asking about WorldCon 2011 rather than WorldCon in general. The correct answer may always change (and editing is nearly always possible); the possibility of the answer being less correct at an undefined point in the future does not make a question too localised in time.

I don't buy the "isn't terribly interesting" argument, either. There's no guarantee that any question gets terribly interesting answers, and I can certainly imagine interesting ones to questions of this type. An answer that I would find interesting wouldn't just randomly pick one "no" example, it would pick the most interesting (most well known, earliest, most significant in the history of the field, etc) and explain how this came to be.

IMO the problem with the specific example here is that because there's such rabid fear of any question that may use a list in the answer @DVK went overboard in adding text to the question to demonstrate that it wasn't a list. I think it would have read better if it skipped all of that and just asked about the history/state of an author accepting fanfic into their universe. (OTOH, those with the rabid fear would then probably have closed it before it even made it to the chat auto-post).

I don't want to edit the question since it's still under discussion, and it seems wrong to ask an alternative version for the same reason. So I've written up an alternative and put it here. ISTM that it would answer the original yes/no question, but it's integrated into a wider question (c.f. Beofett's answer).

5
  • 4
    indeed. That question could have been significantly improved had I not been forced to practice "defensive medicine" against "this smells like a possible list" knee jerk closings - and your last ellipses are 100% correct as that is indeed what had happened to some of my questions in the past when they did NOT get posted with such annoying fencing. Feb 3, 2012 at 6:11
  • Tony - how do you feel about my proposal (currently voiced as comments on Gilles's answer to the linked meta post) to change the rules for closing list questions from "looks like a list question" to "generated over N answers"? Feb 3, 2012 at 6:12
  • 2
    @DVK did not show that his post wasn't a list request. “I'm only asking if X exists” is still a list request: it's asking for items, not for an answer.
    – user56
    Feb 3, 2012 at 21:27
  • @Gilles -I think that you are trying to read into intentions . I can agree with your attempt to judge that "wording blah blah" MAY (likely|unlikely) produce a list of answers", but if, for the purposes of the asker, a single example is enough, then it's NOT a list request. It's something that's poorly worded to potentially act like one. Whether it's worded SO poorly as to be beyond redemption is, IMHO in need of a case by case decision, not a blanket rule. I'm very willing to be convince of your rightness by data, my main source of disagreement with you is statistics of my own questions Feb 3, 2012 at 21:35
  • @Tony - let's try... scifi.stackexchange.com/questions/37223/… Jun 22, 2013 at 4:46
7

I think that such questions should be on-topic, but DO need to be posed with more care to avoid some pitfalls.

This answer will contain 3 parts: The pluses of the questions, the proposed negatives of them with my counter-arguments, and the possible downsides of them with proposed mitigation approaches.


PART A. why are "yes/no" (prove/disprove) questions useful?

They have several possible distinct, though related, uses:

  1. They are a great way to obtain detailed confirmation for a broader research done for a wider question.

    For example, take a hypothetical "What are all the classifications of vampire feeding" question? I'm writing a mega-great answer with 20-point taxonomy with examples, and feel that there ought to be a class for "vampires who suck heat out of human body via drinking blood". But I can't recall any vampire doing so. So, I can either exclude a possible great class from taxonomy, or ask on SFF.SE - as a separate question - if there indeed is a work where vampires suck blood to obtain victim's thermal energy.

  2. They are the best way to back up uncited/unsubstantiated claims made about SFF world. Basically, a SFF.SE version of Skeptics questions. Such as:

    • The claims made in a specific post as part of a logic chain (or the entire answer).

      E.g. an answer to a hypothetical question about werewolves states that a person can never be both a werewolf and a vampire. I have doubts about validity of the answer due to reliance on such an unproven fact, but don't remember any counterexamples. So in my head there is a VERY VERY specific, objective, and constructive SFF question: "Are there known werewolf containing works which show someone who is both a werewolf AND a vampire at the same time". When someone has an objective answerable SFF question, it would be nice if there was a place where one could ask it and have the experts answer it...

    • The claims made in general on SFF.SE.

      E.g. Gilles made a comment on Meta yesterday that there are works where "sorcery" is cleanly different from "magic". Far be it from me not to trust Gilles, BUT he didn't include an example, and I wanted to know at least one such work (why? because I would love to read it! Same reason people seem to like story ID questions). This is a special case because the claim already presupposes that the work like that exists.

    • The claims made elsewhere.

      E.g. some article/book (cracked.com?) states that all the supervillains can be classified into uber smart or uber strong. That classification rings false to me, but I'm not a comic expert. To resolve this, I would like to ask "if there are supervillains that are neither uber smart or uber strong"?

  3. As Tango have eloquently pointed out in his answer - and so did Tony - a major use would be a research into some characteristics of SFF as a genre.

    I would have expected this site to be HEAVILY predisposed to favor such goals!


PART B. Addressing common criticisms of yes/no questions?

  • Criticism: A "No" answer is too localized as someone may write a work which changes this "no" to a "yes" later (source: Beofett's question above).

    Answer: Leaving aside Tony's wonderful answer addressing this, I have only one thing to say: while that is indeed a remote possibility, using logic like this we should make entire Star Wars off-topic. Lucas can release the next version changing ANY of the canon facts at any time.

  • Criticism: A "yes" answer is too content free to be useful/interesting (source: Beofett's question above).

    Answer: Again, a great answer was in Tony's post. I will merely quote statistics: the question that was closed already had 3 up-votes - meaning at least 4 people found it interesting (out of 92 views due to being on an obscure topic with obscure tags - not a bad ratio by this site's standards IMHO, or ANY site's). I apologize, but an opinion of one, or two, or 3 people that something is "not interesting" IMHO is 100% invalidated when at least 4 people find something interesting.

  • Criticism: Such questions may degenerate into a list of "yes" answers, with no possibility of objectively picking a correct one.

    Answer: First, if they actually do for a specific question, I would be the first one to agree that it was a bad question and will vote to close it myself. Writing such a question in a way that does NOT encourage (and discourages) multiple list type answers should be the task of the asker.

    However, I strongly feel that if the asker DID succeed in wording the question carefully enough that it didn't degenerate into a list of answers, there is no basis for closing.

    I will address specific approaches/strategies for preventing lists in the third part. Going forward, if the community decides that such yes/no questions as a concept are on-topic, we may want/need to develop "best practices" for asking them in a constructive list-avoidance manner. Given my past experience, I can 100% assure you that it IS possible to do.

NOTE: I will try to address the ones that I saw posted, if I missed one please feel free to add a link or explanation as a comment and I'll address that too


PART C: Possible approaches to avoid pitfalls

As noted above, there is SOME (though in my experience, small) risk that, unless carefully worded, such a question CAN generate a poorly-suited-for-SE-format list of "yes" answers all of which are equally valid.

Possible approaches include:

  • Politely requesting people to refrain from posting additional examples if one is already posted in an answer.

    Please note that much-criticized "I am willing to take action on extra answers" addendum that I posted on 1632 question was NOT an example of such a polite request, but a signal to moderators/possible "list closers" that I'm willing to make the effort to prevent the question from turning yes/no into a list. In other words, should the community decide that yes/no is ontopic and NOT a list, said addendum would NOT be necessary at all.

  • Strongly encouraging people to add any additional examples to an EXISTING "yes" answer, thus making the content a lot more interesting while still being a good fit for Q/A format.

  • Flag it for moderator to "Protect" as soon as the asker accepts an answer (mods can set a special flag on a question that is designed just for that - it prevents posting new "me too" answers)

    This has a minor downside of extra work for mods, but the mods seem to be eager to put in extra work of closing yes/no questions anyway, so this should have net zero effect on workload at worst (and save the mods major discussions/griping in comments/on meta at best :)

  • As a community, treating any extra "yes" examples posted in contradiction to the question's specs the same way we would treat ANY duplicate answers (e.g. imagine any random question, where user A posted an answer, and user B came along, copy/pasted an answer with minor details changed, and added as his own). Those should be commented on, downvoted, possibly deleted while the extra example is edited into the first answer.

  • Discouraging posting "me too" yes answers with extra examples for rep gain by making the question "CW". I know that CW is not considered Good JuJu on StackOverflow, so this may not be the greatest approach, but it's an effective one - if there's no rep gain possibility, people are a LOT more likely to add any extra examples to an existing answer, AND such an answer will be more widely editable.

  • ...

9
  • I hope the downvote wasn't due to TL;DR :) Feb 3, 2012 at 7:20
  • 1
    Sorry, but TL;DR everything. I'll respond to a few point. A.1: ask for an overview, not for a list; the problem is getting the answerers not to post items. A.2: you could have asked in chat. A.3: I would like that, and I've (unsuccessfully) tried to promote it in the past (e.g. 1 2); please take past attempts into account in your proposals. Oh, and lists are not the answer.
    – user56
    Feb 3, 2012 at 21:31
  • 1
    B.1: if the only possible answer is yes, what's the point of the question? B.2: you see, an answer has to do more than literally answer the “is there an X”. Also, upvotes do not indicate a quality question. B.3: Past experience tells us that it's not really “may degenerate”, it's “has a 90% chance of degenerating, unless the question is ignored and receives no halfway worthwhile answer”.
    – user56
    Feb 3, 2012 at 21:34
  • 2
    C.1: As I've told you before, if “Yes, it exists because example A” is a valid answer, then so is “Yes, it exists because example B”. C.2: then what you want isn't a Q&A site, it's a wiki. C.3 see C.1. C.4 No, it's not a duplicate answer. C.5 No, community wiki is not a magic wand that hides a bad Q&A under the carpet (deletion is).
    – user56
    Feb 3, 2012 at 21:36
  • 1
    "A.2: you could have asked in chat" - Of course that is a possible approach, but that is making an assumption that I'm the ONLY one who'd be interested in such information. IMHO, that's not the case. YMMV Feb 3, 2012 at 21:44
  • @Gilles - "A.1: ask for an overview" - As a rule, I think that's a perfect guideline. I know I often fail to do so, but a lot of time NOT for the lack of intent, but just because I suck at communication. Feb 3, 2012 at 21:47
  • A3. - Would it help if drafts of such questions were FIRST aired out on meta/chat? I would be happy to do so, and I bet Tango as well. I'd rather have constructive feedback first and generate a great question with community input than post a poorly worded one and get hammered. May be I'm weird :) Feb 3, 2012 at 21:49
  • 2
    If you're going to post “demonstration” questions during an experiment phase, yes. In the long run, no, SE works by asking questions first and shooting later.
    – user56
    Feb 3, 2012 at 21:57
  • 1
    @Gilles - posting pre-announced-on-meta demonstration questions seems to be dangerous for one's health around here from past experience, even when they aren't MEANT to be demonstration :) Feb 4, 2012 at 20:48

You must log in to answer this question.

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