38
$\begingroup$

Suppose you come across a question that has been closed for lack of context, has high quality answers, and on its way to being deleted. Can you save the post by editing it to include more context?

Here is the deciding question:

Can the question be salvaged without changing the author's intent?

So, what are some things you could do?

  • You can fix or add MathJax.

  • You can clarify the author's English or phrasing.

  • You can add in relevant definitions, theorems, or simple background and motivation.

  • You can add a source for the question.

  • You can add context which was provided by the original author in the comments.

Now let's look at a few examples of actions that would change the author's intent.

  • Do not add an attempt. Adding an attempt modifies the author's intent by adding presumed information about their background, prior knowledge, mistakes or lack thereof in earlier parts of the question, and their understanding of mathematics in general.

  • Do not expand the question with so much context that it changes fundamentally. Doing so modifies the author's intent by presuming the author's level of sophistication and mathematical maturity.

Some might be thinking, "But what if the latter types of edits produce a great question?"

Well, yes. It is undeniable that one can write a great question by removing enough of a bad question and adding enough good. And math.se needs more great questions! However, we don't want to get into a ship of Theseus situation here. With few exceptions, a question should primarily remain the product of its author, while being helped and improved by editors, not commandeered by editors.

So, how can one save a question that can't be salvaged without changing the author's intent?

If the question is unsalvageable, you can rewrite it entirely by posting a new question. Here is the procedure:

  1. Ask a new question that does include proper context. The new question must be well above what would normally be the bar for acceptability. It must be well-written, including background, motivation, source if applicable (especially for ), etc. In particular, the context in the new question should not be an "attempt." If you're unable to formulate the question with context that is not an attempt, then the question is not valuable enough to be saved.

  2. Leave a comment on the question stating your intent to rewrite their question.

  3. Wait for 7 days. This gives the OP an opportunity to object to the rewrite.

  4. Flag for the answers to be merged into the new question.


Q: Which questions can be saved via a rewrite?

First, let me stress two points:

  • Most questions do not need to be saved.

  • Most questions that can be saved are salvageable through edits.

It is crucial to note that the rewrite process uses moderator resources, which are a limited commodity. Therefore, to qualify for a rewrite merge:

  • The original question must have at least one high quality answer.

  • The new version must have meaningful, distinct substance from what is contained in the original, and provide notable value to future users.

Furthermore, at least one of the following must be true:

  • The original is in direct risk of deletion, or has been subject to a delete/undelete war in the last 90 days.

  • The original predates the new context standards (i.e., was asked prior to 2015).

Q: What do I say in the comment warning the user that the question will be rewritten and merged?

Link this post and the context section of How to ask a good question.

I have rewritten your question to include context pursuant to the rewrite guidelines. I have flagged this question to be merged with the new question in seven days. If you would prefer to edit context into your question yourself, please raise a flag before that time so that the merge may be delayed or cancelled.

Source for copy-paste:

I have rewritten your question to include
[context](https://math.meta.stackexchange.com/a/9960) pursuant to the 
[rewrite guidelines](https://math.meta.stackexchange.com/q/32402).
I have flagged this question to be 
[merged](https://meta.stackexchange.com/q/158066/) with the
[new question](https://math.stackexchange.com/q/$FOO) in seven days. If you would
prefer to edit context into your question yourself, **please raise a flag**
before that time so that the merge may be delayed or cancelled.

(Make sure to replace $FOO with the appropriate link.)

Q: Should rewritten questions be Community Wiki?

No. Community wiki isn't a tool for reputation denial.

Q: My question has been rewritten and I don't like the new version. What can I do?

Edit proper context into your question prior to the merge date. If you need more time, raise a flag on your question (the original) requesting the merge be postponed. If the question has already been merged, this cannot be undone.

Q: What sort of context can be edited into a question?

The new question ought to contain motivation for the question and canonical background such as relevant definitions and theorems. In the case of textbook problems, it is a plus to add a citation to the source; in the case of contest math questions, this is a requirement.

Regardless of whether you are rewriting the question or editing context into the original question, do not simply add an "attempt". In the case of editing context into someone else's question, adding an attempt changes the intent of the author, so it is disallowed. Posting a new question that includes an attempt will make it so we do not consider it a "rewrite" and thus it will not be eligible for merging, and will also be open to being closed as a duplicate.

Q: When editing a question, should I signal that the parts that I add come from me and not the OP?

No. We have revision history for that. In particular, don't use forum-like syntax such as "EDIT:" or "ADDED BY {your name}:" Each version of a post should read as though it was posted by a single author in its current form.

Q: When is it appropriate to rewrite a question vs. edit context into the original question?

Qualitatively, edit context into a question if it is salvageable, and rewrite only if the question absolutely cannot be salvaged.

To elaborate on this, there are many "solve my problem" questions which no amount of editing can save. However, a new question might put the particular problem into a broader context, so that the original question becomes an abstract duplicate of the new question. If this can be accomplished while preserving the integrity of the original answers, merging them in improves future value of the question to other users, which can then serve as a duplicate target for similar questions.

$\endgroup$
41
  • 5
    $\begingroup$ What is the mod's opinion on this edit? It does not quite fall into the first category (Mathjax ....), but looks more like adding a failed attempt. In general I think working out a failed attempt, or just some particular case makes a valuable context. If the purpose of this post is to serve as a guideline in editing closed post, I hope the points can be more clear. $\endgroup$ Aug 19, 2020 at 5:07
  • 5
    $\begingroup$ Thanks for posting this. I had some thoughts churning around recent meta posts impinging on the etiquette/wisdom of all this. $\endgroup$
    – hardmath
    Aug 19, 2020 at 5:17
  • 8
    $\begingroup$ Related older discussions: Editing someone else's question to add context and What kind of context is good to add to a contest math question? $\endgroup$
    – Martin R
    Aug 19, 2020 at 6:41
  • 7
    $\begingroup$ I just wonder if this should be split into a question and an answer. $\endgroup$
    – Martin R
    Aug 19, 2020 at 8:05
  • 8
    $\begingroup$ I think this is a well thought out procedure. I will adjust my practices on the rare occasions when I am prone to do something like this. $\endgroup$ Aug 19, 2020 at 8:49
  • 4
    $\begingroup$ @user400188 as explained in the post they do have the option to say they need more time. Before merging we can also check if the account was online or seems active. In a case of an active account that just was not online this one week for whatever reason we might decide to give an extra delay. To the best of my knowledge it is not technically possible to duplicate the answers. If anyone is worried their questions are merged while they are away or they will need to fix it at an inconvenient time, the solution is simple, fix the questions already before somebody has the idea to do so for you. $\endgroup$
    – quid Mod
    Aug 19, 2020 at 9:43
  • 9
    $\begingroup$ @ArcticChar the intent is the latter, i.e., there should be no attempt. That said, the line can be somewhat blurry. If the new posts explains that the problem is interesting because contrary to similar problems some standard approach fails then this is not an attempt although the information could also be presented via an attempt, but it is not quite the same. $\endgroup$
    – quid Mod
    Aug 19, 2020 at 10:11
  • 6
    $\begingroup$ @ArcticChar the advantage of having a new post is that the person that wrote the post owns it. Eg, if down the road somebody has a question and comments it'll notify the user competent to answer. (I know that it is in principle possible to get notified about comments on other posts and to notify editors the point stands though.) For attempts, generally, it is not clear what the point of an attempt is unless it is the actual attempt of the person that asks the question. Even then we hardly ever want attempts; we accept them because arguably for some users it is the only thing they can do. $\endgroup$
    – quid Mod
    Aug 19, 2020 at 12:50
  • 8
    $\begingroup$ @quid As some users who merely wants to preserve the content, I don't really see owning the post has any advantage. TBH I don't want to be pinged by post that I merely want it reopened. Also, in most situations the post concerned are done and dusted: the high quality answers are all there already. (And I still don't see what is wrong to directly edit the original post, if the poster's gone) $\endgroup$ Aug 19, 2020 at 14:54
  • 9
    $\begingroup$ Although this seems well thought out, it does seems like a rather complicated and time-intensive system to preserve content. As I read it, for content to be preserved someone must put in a lot of effort, and then their work has to be "approved" by a moderator. Editing the original post requires much less time and effort on everyone's part. $\endgroup$
    – user1729
    Aug 19, 2020 at 15:06
  • 6
    $\begingroup$ @user1729 This is why we strongly suggest editing instead whenever possible (in particular in the first Q in the Q&A section). Rewrites should be the exception, not the rule. $\endgroup$
    – Alexander Gruber Mod
    Aug 19, 2020 at 15:10
  • 7
    $\begingroup$ @user1729 "it does seems like a rather complicated and time-intensive system to preserve content" thus it is hoped you will chose the content you try to preserve especially carefully. Regarding the much less effort on everyone's part. Unfortunately that is not true as a lot of effort goes into pursuing and resolving conflicts related to this. $\endgroup$
    – quid Mod
    Aug 19, 2020 at 15:12
  • 12
    $\begingroup$ +1, this was long overdue, especially with lot of ongoing debate on closure / deletions. I hope this gets finalized soon with minor tweaks (if needed based on comments here). Thanks a lot to the mod team!! $\endgroup$
    – Paramanand Singh Mod
    Aug 19, 2020 at 16:21
  • 7
    $\begingroup$ Perhaps this "question"/statement should be designated a frequently asked question? To aid in users' referencing it as needed? $\endgroup$
    – amWhy
    Aug 19, 2020 at 16:31
  • 4
    $\begingroup$ @AlexanderGruber Suppose you have a question which is self-contained and well-posed in terms of definitions and the question asked, and gets answers accordingly. Should another user be able to rewrite all the terminology to use what they decide are "standard terms," even to the extent it requires existing answers to be modified to make everything line up? $\endgroup$
    – rschwieb
    Aug 20, 2020 at 14:09

0

You must log in to answer this question.

Browse other questions tagged .