As a followup to Does stackexchange have any standard for sources? and Why do we need references?, and inspired by several answers I've seen that have been downvoted due to citing conspiracy-theory-quality sources only to have those downvoters accused of bias/suppression, I'd like to propose that we set some concrete minimum level of standard for acceptable sources, for when sources are provided.

As a starting point, I suggest that we prohibit using any sources from Media Bias Fact Check's "Conspiracy-Pseudoscience", "Questionable Sources", or "Satire" categories.

Answers based on these sources can get a "poor quality source" banner similar to the "unsourced" one we have, or the latter can be modified to be "unsourced or uses untrustworthy sources". Alternatively, they could potentially just be deleted as very low-quality if they rely solely on this type of source.

My hope is that by providing a specific list and a policy that can be pointed to, there will be less back-and-forth over this type of answer, and we will be more consistent about what gets jumped all over and what skates by.

  • 1
    I'd be happy already if the "unsourced"-label would appear more consistently and faster. If voters here +1 an 'answer that doesn't need sources' then the Q was way too primitive to begin with. // Your starting point is hopefully easily circumvented if a poster says "look what they write, look how they lie, here's why'? Feb 21, 2019 at 18:33
  • 1
    I think having some kind of standard is a good idea (so +1 on that), but I think all you're really doing is moving the issue from quibbling over the source itself to the list of acceptable sources. Now the argument could just be about how your list of sources is biased. I don't think it's as simple as pointing people to a rote list of what "we" consider acceptable and not acceptable.
    – user5155
    Feb 21, 2019 at 18:53
  • @JeffLambert That's a reasonable point. I'm hoping it will centralize the discussion into one place, and using MBFC's categories (or something equivalent from elsewhere) should hopefully avoid quibbling over individual sources. We can't eliminate all debate on this, because we'll continually get new users who haven't seen the list - the goal is minimizing it and consistency. And if someone objects to our choice of source list (MBFC or otherwise), then they're going to complain anyway so we're not any worse off.
    – Bobson
    Feb 21, 2019 at 19:05
  • 1
    About the angle: Are you really aiming for an authoritative list of un-canonical sources, those that then equal 'no-sources at-all' in treatment, or would you be happy with giving these lists or contents of these lists as 'bad examples' together with an explanation, more important IMO, of why they are bad sources and how to handle posts containing them as main references in favour of the argument the post makes? Feb 21, 2019 at 20:44
  • 2
    @LangLangC - That would be the link from the "poor quality source" banner I was thinking of. I'd love if we could have a "Here's what's bad, here's why, and here's how to fix it" FAQ for it we could point to, whether or not we actually create a banner.
    – Bobson
    Feb 22, 2019 at 18:35

1 Answer 1


To put it simply: No

The problem I have with "unbiased" lists of "other peoples' biases", or "what is fake news" is that there inevitably is some sort of bias that leaks in. A great example of this is Politifact, which has faced criticisms of bias, despite claiming to be neutral.

The other end is I've had some sources I've quoted not in the three categories you mentioned be slammed as belonging in them. If we have a list, there will be people who will no doubt attempt to silence otherwise fair sources that they simply disagree with.

To put it simply, trying to limit sources is a game of futility. What we can do is

  • Downvote - Nobody on Politics.SE is above downvoting things they disagree with
  • Respond - Post a comment or an answer refuting the questionable sources

You must log in to answer this question.

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