The "don't ask" page of the site says, a good question should insist on answers where opinion is backed up with facts and references.

But when a question - even when well written and focussed - explicitly asks for answers which give references like research papers or statistics, it will often get closed as off-topic (for being a third-party resource request). That looks somewhat contradictory to me.

Any suggestions how to resolve this contradiction? Don't get me wrong, I am strictly against allowing questions which are exclusively or mainly third party resource requests. But can we find a way to make questions less prone to closevotes (and downvotes) just because they contain a buzzword like "research paper/ link /example for ...."?

3 Answers 3


Unfortunately, the dont-ask page, where the "backed up with facts and references" line comes from, is not editable. However, I don't see a contradiction here.

Questions should not ask for resources, including articles, blog posts, books, or papers. Instead, the question should be focused on a problem that someone is facing. When writing answers, answerers should be encouraged to not only share the experiences but back up their experiences with references to things that others have said as well. A question that has a lot of answers that cannot be supported with references is likey to be a poor fit. An answer that expresses opinions without supporting facts is also likely to be a poor fit.

Automatically downvoting or closing questions because they contain buzzwords is highly inappropriate. I think it's hard to change behavior, so I would encourage anyone who sees a question that is casually asking for a reference to consider editing the question to remove those buzzwords and help focus on the problem at hand. This could help alleviate some of those downvotes and close votes.

  • 1
    "Automatically downvoting or closing questions because they contain buzzwords is highly inappropriate." - I cannot agree more. Unfortunately, I think we have way too many community members here who give me the impression they believe it is appropriate and have no interest in editing buzzwords out, though they have more than enough rep to do so. Just look at the previous meta question asked before this one. The result of the "buzzword closing" was that the asker was driven away from here and asked at a different SE site.
    – Doc Brown
    Commented Feb 17, 2021 at 20:07
  • 1
    @DocBrown Yeah, I'm not really sure how to solve that particular problem. There is a problem with explicitly asking for references, and that is the ease of which people can just throw references in without addressing the problem. But the solution is for editing to refocus or reframe the problem, rather than down voting and voting to close. Downvoting is a problem since too many downvotes and it's off the homepage for people to see, write good answers to, and hopefully edit to be an example of a good question.
    – Thomas Owens Mod
    Commented Feb 17, 2021 at 20:18
  • 1
    To be honest, I always feel a little guilty about removing buzzwords from questions. I always feel like I am fundamentally changing what they are asking. I always get the feeling that requests for papers, references, blog posts and such are really what the OP is after. But I also feel guilty about down-voting and closing these questions. It all depends on how much effort they put into the other parts of their question. Commented Apr 5, 2021 at 20:25

Your response to my prior answer prompted me to consider this problem from a different angle. I've posted this as a new answer as this is distinct from what I addressed before.

I've started to consider that question that ask for references fall into one of a few categories, and I'm struggling to come up with any category where, at the same time:

  • Explicitly asking for references is meaningfully adding something to the question/answers, more so than an (IMO default) implicit expectation that references would always generally improve an answer's quality.
  • The need for the references is considered on-topic and within the guidelines.

If you (or anyone else) can give me an example of a counterpoint where both bullet points are adhered to, I'm more than willing to reconsider the conclusion of this answer.
The conclusion I make here is made based on the absence of a valid use case.

That doesn't mean that references are bad form. They definitely increase the quality of an answer. However, I am considering that expecting references is bad form.

To put it differently, what is the purpose of asking for references? What does it do, that wouldn't already be the case if you hadn't asked for references?

  • Requiring publically available references?
    • SE should not be used as a human flesh search engine.
  • Requiring privately available references?
    • SE is not suited to "word of God" style questions.
  • Optionally asking for good answer quality?
    • That is already an implicit expectation anyway.
  • Indicating that non-reference answers shouldn't even be posted?
    • Answers without references are still answers. For disagreements on the quality of the answer, that is what the voting system exists for. Question writers do not wield authority on what is or isn't allowed to be posted (provided it actually attempts to answer the question at hand).
  • Suggesting to others that non-reference answers should be downvoted?
    • That is not a soapbox that question writers (or anyone) should have. Users should vote for their own reasons, not because someone tells them to.
  • Because question writers should be able to be the sole authority on what can('t) be considered an answer to their question?
    • If that were the case, then question writers would be given the right to freely delete answers to questions they wrote. Clearly, this is not the case, and the only way to address your dislike of an answer is either through voting or moderator flagging.

I'm interested in hearing a concrete example of why explicitly asking for references adds something that is both considered good form and doesn't already exist without having to ask for it.

In absence of such a concrete reason, the downvotes that such questions receive are arguably not wrong. Not because of predetermined community standards enshrined in the site scope, but because the explicit request for references at best is superfluous, or at worst shifts the purpose of the question to an off-topic nature or infringes on the definition of what an acceptable answer is (which question writers do not have the discretion to override - if they did, they'd be given rights to outright delete answers to their own questions).

  • Thanks for sharing your insights. So I guess instead of allowing to add certain request for references to questions, we should consequently forbid them? This would actually mean a change of the don't ask page, which says "Constructive subjective questions insist that opinion be backed up with facts and references". Well, that is definitely another approach to resolve the conflict I observed, not the one I had in mind, but surely an idea worth to think about it.
    – Doc Brown
    Commented Aug 11, 2021 at 10:25
  • @DocBrown: Whether you explicitly forbid them or not (and then leaving it up to the community to regulate whether they dislike or ignore the requests) is an interesting question, one that I don't have a definitive opinion to. I do agree that the quoted text should be rephrased, but you could simply rephrase it as describing an implicit answerer guideline (i.e. "your answer quality is increased when you back up any non-objectively-provable claims you make") instead of what a question writer can demand (i.e. not "you are allowed to insist that answers provide references").
    – Flater
    Commented Aug 11, 2021 at 10:29
  • The whole don't ask page clearly addresses askers, not answerers. And it does not only allow questions to ask for answers with references, but encourages to write this into a question.
    – Doc Brown
    Commented Aug 11, 2021 at 10:32
  • @DocBrown I suspect your comment was written before I changed mine? I was indeed wrong in my initial response as I had misread the context of the quoted text. I think my revised response actually agrees with yours, or am I misunderstanding?
    – Flater
    Commented Aug 11, 2021 at 10:33
  • ?? Your rephrasing idea in your former comment - in the version I see - is surely a good suggestion, still the resulting sentence would not fit to the don't-ask page then any more, it would then belong to the "how-to-answer" page.
    – Doc Brown
    Commented Aug 11, 2021 at 10:42
  • ... and I agree to you that explicitly asking for 3rd party references does not contribute to a question's quality, so downvoting could be justified. But I see actually more the close-votes to be an issue - don't forget there is an predefined close-reason, where its literal wording may encourage some kind of buzzword closing.
    – Doc Brown
    Commented Aug 11, 2021 at 10:46

It is a contradiction, and its used as a truism by incredulous individuals wishing to exercise arbitrary control to their narrow preference on what questions should be asked. Take this question here:

Why is/was Bjarne opposed to adding strong/opaque typedefs to the core language of c++?

  1. It was closed for being "opinion" based.
  2. When I asked how that made sense, the reasoning given was that he would actually have closed it as being a resource request:

One option is it turns into a hunt to find a particular blog post, paper, or book where the person clarifies their stance.

Another option is that the answer hasn't been given anywhere and people guess or share their opinions.

A third, but extremely rare option, is that the one person (or in some cases, one of a small group of people) shows up here to answer the question. We expect that questions are answerable by the broader community and not require a specific individual.

The point of the matter is that every question asked here is guilty of being off topic for something. When one reason fails to live up to scrutiny, the reasoning can just shift to something equally truistic and broadly appliable. In this case, "Third party resource request" which technically applies to every question ever asked.

Any suggestions how to resolve this contradiction?

Its tough but I would say:

  • Change the prohibition from third-party resource requests to third-party resource reccomendations.

Suppose this were a law stack exchange:

  • Bad: A monkey stole my wallet. Who is a good lawyer for suing monkeys?
  • Good: Is there any case law that applies to a monkey stealing my wallet at the zoo?

And here:

  • Bad: Where can I find interviews of Bjarne talking about c++ typedefs?
  • Good: Why is Bjarne opposed to adding strong/opaque typedefs?
  • 1
    Well, the site's on-topic page already uses the words "Product or service recommendations". The predefined close reason uses the words "Questions asking us to find or recommend tools, libraries...". Me (and others) often use the term "3rd party resource requests" as a short form of "questions asking us to find or recommend 3rd party resource".
    – Doc Brown
    Commented Oct 1, 2021 at 12:48

You must log in to answer this question.

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