-2

I recently flagged this comment. The comment reads:

Probably overreading. Plato's ideal-transcendence can't be reduced to a quantifiable dimension.

I flagged it because it was an answer posted as a comment. Stack Exchange policy could not be clearer about these comments: they should be removed.

Comments are not recommended for any of the following:

  • Answering a question or providing an alternate solution to an existing answer; instead, post an actual answer (or edit to expand an existing one);

My flag was rejected.

Why is official Stack Exchange policy being ignored here? This seems like something that should be really simple to get right.

(Why does this rule exist? See here.)

2
  • 1
    If this was one comment I would agree. But there's a pretty clear pattern of people leaving answers as comments on this site.
    – user111
    Nov 14 '17 at 17:37
  • 4
    Is this question about a general pattern, or about that one comment? If the former, it's probably a duplicate of your previous question.
    – Rand al'Thor Mod
    Nov 14 '17 at 18:39
8

I'm going to ignore Stack precedent for a moment, and come back to it in a minute - it's not enough to say, "this is Stack policy," without investigating whether and why it might make sense to apply in this context. I'm not typically one to blindly follow policy, especially not when it asks me to remove information that doesn't exist elsewhere, so I am happy to go into more detail.


First, consider the difference between:

Probably overreading. Plato's ideal-transcendence can't be reduced to a quantifiable dimension.

And:

It may be helpful to start by thinking about whether Plato's ideal model can be reduced to quantifiable dimensions.

One comment is "clearly" an "answer" to the question. The other is "clearly" a signpost, or a helpful remark for future answers - a note about content that might be relevant or helpful. But they both contain exactly the same information. So let's ignore the two words that frame the comment like it's trying to be an answer, and focus instead on... well... what it says.

And when you look at it without that frame, even you've left a lot of signposts like that - comments that contain only directions or references for where one should look or start thinking. You're not the only one: they're very common here. And as far as I've seen, they can be very helpful in guiding answerers in navigating topics that are not necessarily simple or easy.


The comment you flagged was originally posted as an answer. I had converted it to a comment because it, well, didn't contain nearly enough information to be a proper answer. But removing it outright would have deleted something that at least thrusted towards an answer, in a way that isn't available on that answer page, and can't be easily searched or found elsewhere.

So the question I ask when I see it flagged is pretty straightforward: does this comment, right now, contain something valuable? And it clearly does: both for anyone passing through, but especially for someone trying to develop a full answer, it provides a signpost that gives a direction in which to start heading. That's valuable.

But it's especially valuable on a site where answers are often hard to build. It's especially valuable when these kinds of connections are not ones even any mythical expert is necessarily going to see. And up until someone posts an answer that develops the point, the comment contains no redundant information, and serves only to help the reader.

Once an answer is posted that contains and expands upon the information in this comment, or when the question is generally sufficiently answered, I'd say it becomes redundant and should/will(?) be removed.


On a policy level, this is not as clear cut as it seems. I would place this under "minor, but transient information." It's minor: someone's smacked a signpost into the ground, as if to say, "try that way, maybe?" And it's transient: it will only stick around as long as it remains non-redundant.

Even that equivocation aside, we as a site can't blindly follow Stack policies. We have to be judicious about what we do and why. We are both different and unique as a site in many ways, and much of the Stack framework (and existing policy) was not set up, and will not work for us. When considering whether that policy is helpful to apply, we need to look not just at what policy says, but why it says that, to see whether it makes sense here.

While on other sites, deleting anything answer-y in comments makes obvious sense, for questions like ours, some discourse around the question is often both inevitable and necessary to build good answers. Again, once that discourse becomes redundant, it should be cleaned up.

In basically all other cases I'm totally on board with Stack comment policy. But I think following it to the letter presents an absolutist view about useful comments that isn't functionally helpful in many cases - for us. (I'm also not convinced this is against Stack policy, anyway.)

1
  • 3
    It's times like these when I wish for a bounty function on meta sites. The distinction between following official SE rules and doing what's actually good for the site (in those possibly rare cases when there is a distinction) can't be emphasised enough. This main meta post by Shog9 addresses that same issue pretty well. And the rationale of not removing useful information is essentially what went into my answer here, but you've expressed it better and more convincingly.
    – Rand al'Thor Mod
    Nov 15 '17 at 0:54
1

I agree this is an attempt to answer the question. The person is asking if they're understanding correctly, and the comment is telling them they're overreading. That should be removed.

In chat, Mithrandir mentioned it wasn't an answer and was trying to lead on an answer — I disagree with that, it's definitely just an answer. Even if it isn't, it shouldn't stay: half-answers, not-quite-answers, leading-on-answers, and so on, don't have a place in comments.

Some of us here come from some sites where comments get super-discussiony and are virtually unregulated, but I'd prefer we keep things more clean & follow the rules on comments here. We're not huge and there's energy capacity for users & moderators to prune comments diligently.

Stack Exchange policy gives us some pretty good lines to draw (bolding theirs):

You should submit a comment if you want to:

  1. Request clarification from the author;
  2. Leave constructive criticism that guides the author in improving the post;
  3. Add relevant but minor or transient information to a post (e.g. a link to a related question, or an alert to the author that the question has been updated).

The first and second are black and white, and the third one has examples given.

Over on RPG Stack Exchange, we believe that answers-in-comments have a number of toxic effects, primarily in the way they bypass all of our quality control mechanisms we expect actual answers to go through:

Answering in comments does the following things.

  1. It bypasses question closes. They're closed for a reason.
  2. It provides an answer that can't be marked as an answer for future people's knowledge.
  3. It contributes to long comment debates as you can comment on an answer, but it's unclear what you're commenting on in a comment thread.
  4. It is "cheating" by locking your answer to the top. Answers with higher votes/accepted answers should go to the top to indicate their quality. Bypassing that by sticking your answer in a comment on the question is unacceptable.
  5. It bypasses all our quality control mechanisms: we can't downvote your "answer", edit it, or comment on it to request clarification or improvements. Answers also bump a question to the top so that people will scrutinize the answer; comments don't do this.

Since we collectively have the energy to moderate comments to avoid this, we do so, and I believe Lit.SE may have similar.

Anything at all that is a comment attempting to answer the question, or is trying to lead on answers or sort-of-answer, or voice opinion (comments aren't for discussion), or vaguely resembles those things, should be removed if anyone spots it and takes the trouble to flag it. It has poor effects, is a misuse of comments and a breakage in Stack processes, and its negative effects can and ought to be entirely avoided.

You must log in to answer this question.