The tag has seen a lot of misuse recently, since apparently 28% of our questions are science-based. Since we have killed off its erstwhile companion tag , I believe it now makes sense to start enforcing a separation between questions that ask about whether a particular concept breaks suspension of disbelief and questions that really are about show-me-the-numbers hard science for those more obsessive-compulsive meticulous among us.

It is my understanding that this is in line with the initial purpose of the tags. My suggestion is to clarify the tags and start enforcing this distinction better.

What does the community think?

EDIT: The consensus seems to be building around having a tag (see the rationale in the answer + comments below).



Oh, dear, let's NOT have another editing run for . I'm ~32 edits closer to the 'copy-editor' badge after the events of last night/whatever-time-it-was-for-y'all, but we really screwed up the "active" tab, which led me to bump a few recent questions up via minor edits to restore some sanity.

Anyway, I suspect that this question is the result of a discussion we had in chat while relaxing after the genocidal editing, right? I'll rehash my position: should contain hard science. I'll quote Vulcronos' entire answer to an older question here:

I totally agree [that our default position should be based on science and logic]. Unless the question says there is magic and gives a decent description of what is possible, I assume we are sticking to real world physics. Science fiction and steampunk stories have enough variety already, adding magical answers with no knowledge of if magic exists or how is just unproductive and isn't likely to be useful to the OP.

So if I ask a question about, say stone-eating insects (been there, done that), I'll assume that everyone knows that I want a sensible answer (or answers). I did add , though (which could have different connotations). No magic, please!

So I totally agree that should call for hard numbers and strong evidence, backed up by sources, preferably papers (for good, free pre-prints, see arXiv). If I were to add the tag to my insects question, that would let people know that anything remotely speculative is a no-no. I want evidence, observations, and scientific theories that are backed up.

I actually like the idea more from my traditional standpoint: that of an answerer. It's evident that I use math quite a lot, possibly more than any other user on Worldbuilding. I have a tendency to go a bit overboard from the perspective of others, bringing in calculus and the like. The reason I do that is because so much of science uses mathematics. And not just algebra. Differential equations (e.g. $\frac{d^2x}{dt^2}-\frac{dx}{dt}=2e^x$) can be very important. Unfortunately, many people that want scientific answers use the - which is fine, but it does give me the impression that they want hard science, which typically involves a lot of math, papers, citations, etc.

So if we agree to use the tag for only answers with hard science, I'll know when to add in equations and hard science and when to stop myself from delving in too deep. And it'll make it better for those who dislike it when I write answers like that. This proposal would be awesome.


This would have been a comment on Monica's answer, but it was going to run way too long, so I figured I'd add it here.

sort of smells like a meta tag, but if we end up using it as meaning "hard science", then it could be a very useful meta tag. Going back to my question on insects, if I wanted an answer based only in science, I would use the tag. As it is, I recognize that the idea isn't too plausible (logistically, too), and so I didn't expect that any of the answers would have much to back them up. That's a case where science is expected, as it is in most answers. But if I asked for hard science, most answers would be partly or entirely invalidated for the sole reason that there's pretty much nothing we know on the subject of stone-destroying insects.

That said, if you were to use the tag on a question you mentioned, perhaps wouldn't be needed. As it is, you used and , which added all the specificity needed.

So now I'm on the fence. could be a meta tag, or it could be useful as a tag explaining just how much science needs to back an answer up.

Second Addendum

Okay, I'm now 100% behind (see the rationale in the comments below, as well as the reasons in the other answers for sort of getting rid of ). We've reached a point where almost all answers are backed up with science, so I think it's no longer needed.

I've gone ahead and written a test question for the tag:

How can a Type II civilization influence accretion rates from a debris disk to a passing star?

  • $\begingroup$ Agreed. Science based = show me the math. Reality check = is this believable/plausible/possible/likely (even if it sorta violates science)? $\endgroup$ – James Mar 18 '15 at 19:56
  • $\begingroup$ It sounds like you're saying that, to you, science-based means "prove it more stringently". If I ask a question about, say, how multiple moons interact with tides (been there), I expect answers to be based on astronomy even without that tag. Answers that prove what they say will tend to be better answers. $\endgroup$ – Monica Cellio Mar 18 '15 at 20:02
  • $\begingroup$ @MonicaCellio Yes, science-based answers tend to be better answers. The way I read the suggestion underpinning this particular answer is that if the question is tagged [science-based] then anything that can't be backed up by scientific references is off limits in answers. And I believe that was the original intent of the tag. [reality-check] is more of can this idea work, given my imaginary world?. $\endgroup$ – user Mar 18 '15 at 20:09
  • $\begingroup$ @MichaelKjörling thanks. So [science-based] is really more like [stringent-science-required]? $\endgroup$ – Monica Cellio Mar 18 '15 at 20:16
  • 1
    $\begingroup$ @MonicaCellio Well, consider the tag wiki excerpt for [science-based]: For questions that require answers based in hard science, not magic or pseudo-science. The way I read it, that means hard science answers, which means that (while it doesn't necessarily have to be backed up in the answer itself; few will dispute the claim "during daytime on Earth, the sky is blue") any claim made must be possible to back up with science as we know it. $\endgroup$ – user Mar 18 '15 at 20:18
  • $\begingroup$ Wow, I just got back to check this out. @MonicaCellio - Michael said basically said what my response would be. $\endgroup$ – HDE 226868 Mar 18 '15 at 20:21
  • 2
    $\begingroup$ Does science include sociology? Anthropology? I just think we're asking for trouble with this tag, and the requirements for answers should be stated in the question. $\endgroup$ – Monica Cellio Mar 18 '15 at 20:36
  • $\begingroup$ @MonicaCellio Then perhaps I'll modify my stance for now. I'm not leaning one way or the other on whether it should be kept, but this I how I feel if it is kept. $\endgroup$ – HDE 226868 Mar 18 '15 at 20:38
  • 2
    $\begingroup$ @MonicaCellio That's basically what my old post (which HDE linked to an answer to) was arguing, and I think we had a fair amount of agreement there given the voting on the answers. $\endgroup$ – user Mar 18 '15 at 20:38
  • 1
    $\begingroup$ I agree with Michael's older question (and the top-voted answer), both of which have a lot of community support. Re the addendum, would [hard-science] better convey what you're after? $\endgroup$ – Monica Cellio Mar 18 '15 at 20:43
  • $\begingroup$ @MonicaCellio I think that would be a much better solution to the whole problem. $\endgroup$ – HDE 226868 Mar 18 '15 at 20:44
  • $\begingroup$ I think you and I are the biggest math answerers on the site (though I have no way of confirming that). My position is that science based should be hard verifiable science - that means no subjective sciences like philosophy. We could have something like [social-sciences] for psychology/philosophy/sociology, if necessary. $\endgroup$ – ArtOfCode Mar 19 '15 at 9:34
  • 4
    $\begingroup$ I like the edit to "hard-science". Our default answers seem to be soft-science or science-based extrapolations. hard-science sounds stricter than science-based and therefore is more likely to be used correctly. $\endgroup$ – Mourdos Mar 19 '15 at 13:58

If we're going to have the tag, yes. But I really don't like either this tag or reality-check. Here's why:

All answers on this site should be consistent with the baseline in the question. Answers on a question about unicorns don't get a free pass just because they're magical beasts; answers need to jive with what we know (or what the asker has stated) about unicorns. And all questions should specify -- right there in the question -- what they're looking for to the extent that they can. A question about science-y things should get science-y answers because that's the question that was asked, not because of the presence or absence of tags. A question asking "is X realistic given factors Y and Z" should get an analysis of X with respect to Y and Z regardless of the tags. I'm having trouble seeing the added value and, demonstrably, the tags get used a lot and in ways that people don't seem to understand.

I think these are meta tags, but more than that I think they're unhelpful meta tags. I would prefer that we be willing to spend a few more words in the question to say what we actually want ("science" is pretty broad too, after all), and I'd like to see answerers and askers work together to nail down the level of desired detail when that's not clear. Incremental answers are great for this -- here's an answer with first-order citations and key quotes/equations/etc, and the asker (or anybody else) can ask more detail about such-and-such point and then you show more math, and so on.

  • $\begingroup$ As noted in comments on HDE's answer, I think hard-science would address the problem. $\endgroup$ – Monica Cellio Mar 19 '15 at 23:42

I've been battling this in my mind. Basically, I believe I feel much the same as I did back then: if a question doesn't specifically state anything at all, we should assume that answers need to be consistent with the known sciences as we know them. That leads to at least two conclusions relevant to this particular question:

  • The tag isn't really needed. In the absence of some specific indication to the contrary, we should assume the semantics of that tag as defined in the tag wiki: that such questions require answers based in hard science, not magic or pseudo-science. The tag could in principle be deleted and the site should not suffer from that. The voting on the old meta question indicates fairly strong community agreement with basically this position, and it agrees well with the original mission statement of the site from the Area51 proposal. Assume our real world, unless the asker specifically states otherwise.

  • If we are going to keep the tag, then it really should mean what the tag wiki states, and not something else or half-hearted. The kinda-sorta half-hearted variant is what (a whole similar question in itself, but unrelated to this) is meant to cover.

Whether or not we keep the tag itself, a good litmus test for whether an answer fits a question that matches the above might be: if I take this one random claim that is made in an answer, and post it on one of the science-based sites with a request basically amounting to "please explain this statement to me", then would it get shot down as not based in actual science? That means sites like for example Space Exploration, Biology, Earth Science, Astronomy, Physics, History, ... well, you get the idea. If it would get shot down there for that reason, then it doesn't belong in the answer and it follows that anything based on it also doesn't belong. An answer doesn't necessarily need to provide scientific references for every claim made within the answer itself, but every claim made in an answer to a question seeking answers based in science should match known science and be possible to explain in the context of the relevant sciences as we know them.

  • $\begingroup$ Lots of actual science questions get shot down as 'speculative' on those sites. $\endgroup$ – Serban Tanasa Mar 18 '15 at 22:02
  • $\begingroup$ @SerbanTanasa You're missing my point. I'll edit later, but try this: If I'm writing a science-based fiction story about a colony on Titan, and for whatever reason want some aspect of the story to depend on the color of the sky, then I might ask here how to explain that. If someone writes in an answer "property X of Titan's atmosphere causes the sky to have color Y", I should be able to take that to an appropriate site, perhaps Physics, and ask "why would property X cause the sky to appear color Y when viewed from Titan's surface?" and the answer won't be "science doesn't allow property X". $\endgroup$ – user Mar 18 '15 at 23:10

I introduced science-based to solve the problem of people asking scientific questions with magical answers. The culture of the site is now well enough established that I do not think it is needed any more, I've certainly not seen many magic answers to science questions recently. In other words the tag has succeeded. The question is whether the presence of the tag is still needed or whether those answers would start sneaking back in if we removed the tag.

There are other reasons that might encourage keeping it as well. For example some people really dislike magic-based questions. Being able to filter on only science-based questions is a useful feature for those people.

So I'm on the fence here, I think it was useful. I'm not sure if it's still useful. It's certainly popular which suggests that people like it.

  • $\begingroup$ I don't know how well the filtering works; there are lots of questions that have other science tags but not science-based. Do we have a lot of questions that invite magic but aren't tagged magic? Filtering works in both directions -- favorite and ignore. $\endgroup$ – Monica Cellio Mar 19 '15 at 12:53
  • $\begingroup$ I guess we do this with a text search for the word magic where tag != magic and text !contains "no magic". Something like that? $\endgroup$ – Mourdos Mar 19 '15 at 13:56

You must log in to answer this question.

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