Currently a lot of questions are being closed as "primarily"-opinion bad. We've started to come up on a consensus on what counts as "primarily"-opinion based but we need to decide on exactly what subjective questions we want to allow and what we want our site scope to be. Judging by previous meta questions, answers, and comments, it doesn't seem like we want to completely kill of all subjective questions.

Especially since this is a natural language site, learning is a process which is different for everyone, so more often than not questions will probably be subjective.

Since we don't want all answers to need to cite a research paper, what subjective questions do we want? Tips? How important is X? Pros and Cons of X? What are we going to consider "on-topic" subjective questions?

  • 1
    Maybe we could also address how to convert some subjective questions into more quantifiable ones? (eg. How similar is X to Y? vs What is the linguistic distance between X and Y?, perhaps?) Commented Apr 6, 2016 at 17:33

1 Answer 1


First read Good Subjective, Bad Subjective.

Beyond that discussion, the reason you typically want to avoid certain types of questions (in this type of Q&A format, at least) is that some questions simply cannot be answered authoritatively and definitively in the space of a post — keep those two words in mind; they make a good acid test to judge the suitability of a question.

Some subjects simple need a lot of discussion or debate to cover the issue adequately. Stack Exchange is not a good fit for the type of round-table discussion needed to answer some questions, so we tend to forgo those types of questions entirely (by design, actually).

What is the best way to learn French? — BAD SUBJECTIVE

But be careful about issuing a knee-jerk response every time you see one of these verboten words like "best", "most", or "easiest". These tend to elicit a handful of close votes without actually looking if the underlying question can be answered at least somewhat authoritatively.

What is the best age to start teaching a second language?

In the study of language acquisition, I suspect this question is at least somewhat answerable in the context of this site (apologies if it isn't — it's just an example). Maybe the answer includes a bit of "it depends…", but if those conditions make the answer all that more interesting, you can still answer it.

But don't guess at an answer; if you don't have enough information to help that user specifically, don't just throw out a version of something that might pass as an answer maybe (and please don't half-answer/guess in comments; that's even worse). But looking back at these trigger words, you can usually look beyond the unfortunate wording to see if it translates into something actually answerable in this format:

At what age is the adolescent brain most capable of taking on a second language without interfering with their primary language development?

It's the same question as above, better asked. Such verbosity probably belongs in the body of the post, but if you can improve the tone of the question using the correct vernacular, improve it. We are building a wiki-based, collaborative compilation of knowledge. Titles are especially important. They should express what is being asked without having to open the question. If you can improve a title and opening, even a little, you should edit to make them more clear. Always.

The trouble signs of a question that should usually be closed as too broad or too subjective usually has some of these earmarks:

  • It would need a book to answer
    Some questions are simply too soon for a site like this. Folks often want to initiate a broad discussion or a solicit a list of tips and tricks they can pick through later. But a good question (in the context of this site) is usually about a very specific problem you might encounter in your day-to-day study; questions that can be answered reasonably well in the space of a post.
  • The topmost answer(s) will not stand alone
    If you can envision every answer being followed with a "yeah, but…" followed by a long comment exchange, it's probably going to be a debate or on-going discussion. Yes, comments are meant to suggest improvement to a post, but that post should essential start and end the discussion. This is what I mean by a definitive answer. Debates, discussions, and hot-button controversy are better left to the threaded discussion forums.
  • Potentially dozens of answers, all valid
    The hallmark of a good Q&A site is the best-constructed, "correct" answer rises the top. If the top answer (or the top few) don't complete the issue being asked about, that is neither definitive nor authoritative. Questions soliciting suggestions, recommendations, anecdotes, or polling of the community aren't generally a good fit for this site.

    There are rare, rare exceptions to this. Some communities have made a few well-regarded "resource" questions actually work on their site. We call these the "big list" questions (e.g. must-read books/blogs/web sites/resources). But communities should generally agree that these precious few questions are so helpful and so beloved, they are worth the on-going maintenance of meticulously curating and culling this ever-growing list on an on-going basis. If you allow more than the tiniest handful of these question on your site, they won't work. Such questions are too easy to ask, and letting folks solicit a growing number of these large lists will mean most of them will almost certainly fall into disrepair.

This list is not exhaustive. Feel free to carry on and explore this issue from here. But I would caution not to get too eager in coming up with a bunch of rules around this subject. You don't want a situation where folks are rubber-stamping every other post with a "problem" because we tried to anticipate everything that can every go wrong before there's an actual problem affecting the performance of the site.


You must log in to answer this question.

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