-3
$\begingroup$

A question was submitted by a new contributor. The question covered an advanced topic in matrix calculus and received a very good answer which provided detailed coverage of those topics. However the question was nevertheless closed. There was an option to reopen the question by suggesting improvements; I submitted an edit which was reviewed and accepted. The edited question also received positive marks. But then the question was closed and the edits were removed. The reason given was that my edits '[did] not clearly align with the original intentions of the author'; anyone with basic understanding of matrix calculus would realise this was not the case. The moderator responsible for removing my edits also suggested that I had 'hijacked' the original question. This is a ridiculous suggestion since I have no privileges that allow me to reedit and submit unilaterally. Furthermore all my edits were entirely within the original scope of the question. The original author had tried standard differentiation using the indexed components of the matrix but did not go to enough detail to show why that approach did not work. My edits provided this missing detail. Another reason given for removing my edits was that they invalidated the submitted answer. But anyone with basic knowledge of the subject would know that my comments reinforced the validity of the submitted answer. I give very little importance on whether I get good marks or not; but I am not going to allow arbitrary interventions that disrupt the process of learning on this site to go unnoticed.


Key issue: Edits to a question are always placed in a queue. They are only posted when approved by moderators. What is the point of this procedure if after the edits are approved then the edited question is closed again?

$\endgroup$
8
  • 10
    $\begingroup$ I understand your frustration, but I don't think that we should undertake major rewrites of questions, however strongly we feel that the rewrite better captures the intent. There are alternatives. Post your own question along with your solution and attach relevant two way links. Edits should be limited to things like formatting and grammar. $\endgroup$
    – lulu
    Commented Aug 28 at 23:33
  • 2
    $\begingroup$ I have, on occasion, used the comments to propose clarifying language which the OP (without prompting) has cut and pasted into their post. As another alternative. And I have posted solutions which began with a statement of my interpretation of the problem. As yet another. $\endgroup$
    – lulu
    Commented Aug 28 at 23:34
  • 6
    $\begingroup$ "Edits to a question are always placed in a queue. They are only posted when approved by moderators." is false. Moderators are not necessary and almost never involved. $\endgroup$
    – Nij
    Commented Aug 29 at 10:01
  • 2
    $\begingroup$ @GEdgar An edit made by a user with under 2k XP requires review. As Ted Black currently has 1137 XP, their suggested edits undergo review. $\endgroup$
    – Xander Henderson Mod
    Commented Aug 29 at 19:13
  • $\begingroup$ Sorry for a meta of a meta question - the main page shows "Thanh Nguyen modified the post yesterday" - I can't see in the page history what modification was done by this user. Where do I find this information? $\endgroup$
    – Srini
    Commented Aug 31 at 21:01
  • 3
    $\begingroup$ @Srini The said user posted an answer, but deleted it soon after it is posted. Hence the action is invisible to users with < 10k reputation points. $\endgroup$ Commented Sep 2 at 15:15
  • 1
    $\begingroup$ This is shown in the timeline of this question. $\endgroup$ Commented Sep 2 at 15:16
  • $\begingroup$ Ted: There is a meta Math tag specific-question that might be appropriate for your Question here. Of course you might be aware of that tag and have intentionally not applied it, so I'll leave it up to your discretion. $\endgroup$
    – hardmath
    Commented Sep 3 at 17:46

2 Answers 2

16
$\begingroup$

I'm sure you had the best of intentions, but let's review the editorial comment you left when updating the original post:

Defined problem more clearly; added an alternative approach that is difficult to resolve. Gave a reference that has examples of matrix calculus applied to similar problems.

Such changes are inconsistent with the Community's policy of preserving the original poster's intent. "Light" editing to fix obvious typos (misspellings and the like) are permitted and encouraged, and if you wanted to Answer a Question using the "alternative approach", explaining the gap between the OP's original thinking and your preferred outlook, that would avoid any issue of the changing the OP's intent.

That would be a better way to give "a reference that has examples of matrix calculus applied to similar problems," and might well be the help that the OP needed to create valuable site content.

$\endgroup$
5
  • 2
    $\begingroup$ The comments I posted on the edit were not properly phrased. I was definitely not trying to 'add an alternative approach' (if I had an alternative approach I would have posted an answer). I should have written that I was trying to elaborate the approach tried by the original author, namely differentiation w.r.t. to the components of the matrix. The reference given did not change the focus of the original question, namely the application of matrix calculus. I really cannot understand why my edits have created such an adverse reaction, I wonder if there is something else going on. $\endgroup$
    – Ted Black
    Commented Aug 28 at 23:42
  • 7
    $\begingroup$ @TedBlack In the past I advocated a similar approach, using somewhat radical edits on occasion, when I wanted to salvage a question I found interesting. Later the community decided that it was too much. This policy on context rewrites was then formulated by the moderator team. Please study that. I think it is a worthy compromise. Not optimal (probably from any angle), but that's what compromises are! $\endgroup$ Commented Aug 29 at 13:06
  • $\begingroup$ @TedBlack I completely agree with the negative responses that your posting has generated. Beyond helping to embed a diagram, or inserting paragraph/line breaks to help readability, I feel that a user should confine their reaction to referring the OP to a protocol article, such as this one or some other article. Certainly, the reviewer can also leave a comment that indicates specific aspects of the posted question that the reviewer feels need to be improved. ...see next comment $\endgroup$ Commented Aug 31 at 20:39
  • $\begingroup$ @TedBlack However, I think that it should then be left to the OP to either invest the time/trouble to improve the question or not. $\endgroup$ Commented Aug 31 at 20:39
  • $\begingroup$ @JyrkiLahtonen thanks for your constructive comment $\endgroup$
    – Ted Black
    Commented Sep 1 at 4:13
6
$\begingroup$

Moderating the Question

The original version of the question was four lines long (two, really, if the displayed equation were inlined). It was two sentences. The proposed edit was significantly more than four lines—it introduced an entirely new discussion about "index notation" which was not present in the original question (it is hinted at by the statement "I can obtain a solution in terms of the elements of the matrices," but it is not obvious that this means "index notation", nor even that the asker is familiar with "index notation"), there is an entirely invented computation, and there is an added reference which may or may not actually reflect the original asker's level of knowledge.

This edit goes far beyond what the community generally considers to be acceptable.

In moderating that question, I noted that the question was closed by the community at the time the edit was made. Then the inappropriate edit was made, and the question was subsequently reopened. I reverted the inappropriate edit and closed the question for the reason it was originally closed, thereby returning the question to the state it was in before the edit was made. I have no dog in this fight vis-à-vis whether or not the question is opened or closed, but I do want the community have the opportunity to act on a version of the question which clearly reflects the original author's intent.

I will also note that another edit was made to the question which incorporated a comment left by the original author. Such a comment does clearly reflect what the author was thinking—so, even though that edit more than doubles the length of the question, it is appropriate.

It also appears that the question is likely to be reopened (it currently has 4 reopen votes).

Some History and Context

The Math SE community has, after years of arguments, come to an uneasy consensus that editors should take a very light touch with respect to changing questions. It is fine to correct obvious typos, but it is not okay to invent context for a question. This compromise comes from a long-standing argument on the site between "deltionists" and "anti-deletionists" (or the "pro-homework" and "anti-homework" factions, or "exclusionists" and "inclusionists", or "archivists" and "tutors", or whatever else, depending on how, exactly, you want to draw the lines).

Math SE does permit users to post homework problems, but users asking those questions must present context for their homework. The argument is that this context informs answerers about the asker's background. What is their general level of mathematical knowledge? What do they know about the specific topic? What relevant tools are they familiar with? And so on. But these data are related to the original asker, and not any other random editor who might come along. Questions are personalized.

(As an aside, I'm not a huge fan of this compromise—I would rather just declare that homework questions are off-topic, drop the context requirements entirely, and focus on making sure that the questions on Math SE are, first and foremost, good questions. But that is not the way the world works, and we get compromises which no one likes. That's what a compromise is, after all.)

Because there are diamonds in the rough on the site, and there are good-mathematics-questions which are not good-SE-questions, we have developed a procedure for salvaging good questions which don't meet the site's standards. In a nutshell, first use the comments to suggest changes which might improve the question. If those changes are never made, feel free to post a new question, then flag the old question for moderator attention.

$\endgroup$
4
  • 1
    $\begingroup$ We keep hitting the same wall. The edit is deemed 'inappropriate' and yet not only it went through peer review (I have 1,137 reputation so I cannot edit unilaterally) but also received upmarks. I have read the policy document regarding edits of closed questions. I agree with the policy of removing edits that change the scope and substance of the original question. My argument is not that my edit should have been accepted no matter what; my argument is that we cannot have an edit accepted after peer review and then removed by another peer review. There should be an objective set of standards. $\endgroup$
    – Ted Black
    Commented Aug 29 at 19:46
  • 11
    $\begingroup$ "...we cannot have an edit accepted after peer review and then removed by another peer review.": we certainly can, it happens all the time. It's called 'retraction'. $\endgroup$
    – postmortes
    Commented Aug 29 at 20:18
  • 10
    $\begingroup$ The edit was not "accepted after peer review and then removed by another peer review". The edit was (incorrectly, per site policy) accepted per peer review, and removed by a moderator. This was not a peer review action, but the action of an elected moderator. But do note that it has happened before that an edit gets through the peer review process, and is later reverted by another peer (that's what rollbacks are for, after all). Whether or not an edit is appropriate is often a judgement call, and peers may disagree. $\endgroup$
    – Xander Henderson Mod
    Commented Aug 29 at 20:56
  • 3
    $\begingroup$ This edit changed the question by a large amount. How it was accepted by various peers eludes me. The point being "my argument is that we cannot have an edit accepted after peer review and then removed by another peer review" doesn't work because the first "peer review" was shambolic per site guidelines (see the Guidelines on Context Rewrites linked by Jyrki below), so it had to be reversed. No doubt some reviewers need to be told off here. The second peer review reversing the first one won't be rolled back because it did meet site guidelines (or so I think!) $\endgroup$ Commented Aug 30 at 6:45

You must log in to answer this question.

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