Jump to content

Wikipedia talk:Consensus

Page contents not supported in other languages.
From Wikipedia, the free encyclopedia

This is an old revision of this page, as edited by WhatamIdoing (talk | contribs) at 17:09, 29 October 2021 (→WhatamIdoing issues: Reply). The present address (URL) is a permanent link to this revision, which may differ significantly from the current revision.



Consensus redux

If a current consensus (by way of RFC) is sought on a topic and there was a prior consensus (also reached via RFC), of which is there is no notice of on the talk page, from a few years ago should the newer request for a consensus or RFC be closed for not stating it was an RFC to overrule a prior RFC? — Preceding unsigned comment added by DonkeyPunchResin (talk • contribs) 17:05, 29 October 2020 (UTC)[reply]

No. WhatamIdoing (talk) 01:54, 6 February 2021 (UTC)[reply]

"Wikipedia:EDITCONCENSUS" listed at Redirects for discussion

A discussion is taking place to address the redirect Wikipedia:EDITCONCENSUS. The discussion will occur at Wikipedia:Redirects for discussion/Log/2021 May 6#Wikipedia:EDITCONCENSUS until a consensus is reached, and readers of this page are welcome to contribute to the discussion. Dudhhr (talk) 05:18, 6 May 2021 (UTC)[reply]

"Consensus" is not a standalone policy

"Consensus" should never be cited as a standalone policy (often used as a substitute for "I don't like it"). It should always be based on other existing guidelines and policies that are cited. If no other policies are cited, then the "there is a consensus" claim should be ignored as just a local tyranny of the majority. OTOH, if that majority is saying "our consensus is based on this and this policy," then the word suddenly means something and discussion can be constructive. Otherwise not.

We already follow this principle in RfCs and AfDs, where a minority of editors can prevail because their arguments are based on policy. Arguments must be more than complaints. They should cite "guidelines and policies". This is from WP:AfD:

AfDs are a place for rational discussion of whether an article is able to meet Wikipedia's article guidelines and policies. Reasonable editors will often disagree, but valid arguments will be given more weight than unsupported statements. When an editor offers arguments or evidence that do not explain how the article meets/violates policy, they may only need a reminder to engage in constructive, on-topic discussion. But a pattern of groundless opinion, proof by assertion, and ignoring content guidelines may become disruptive. (BOLDING added)

One often sees a gang of fringe editors who cannot cite policies (because their fringe ideas -- which they dare not voice as their real motivation -- don't have a leg to stand on), so they cite "consensus" as their justification for deletion of properly sourced content they do not like. In such cases, "consensus" is a weak and pitiful argument. It's like seeing bogus scientific claims without any evidence, or Daddy forcing obedience "because I said so."

We need a sentence or two that encapsulates the principles above. -- Valjean (talk) 01:12, 12 May 2021 (UTC)[reply]

Is that not what is summarized in WP:CONLIMITED? – The Grid (talk) 01:31, 12 May 2021 (UTC)[reply]
I think that is mostly concerned with attempts to change policies. We need something that makes clear that if one invokes "consensus", one must also invoke specific guidelines and/or policies, as it is them, not the "consensus" policy, which has most weight. "Consensus" alone has no weight at all. -- Valjean (talk) 14:45, 12 May 2021 (UTC)[reply]
But in that circumstance you would not be reverting solely for "no consensus" because there are substantive content issues being discussed on the talk page. Certainly a distinction that we should make clear if the community decides to add text here that "no consensus" is not, on its own, a valid argument. Butwhatdoiknow (talk) 06:07, 14 May 2021 (UTC)[reply]
@Crossroads: Would including text taking into account the appropriate use of a "no consensus" objection while a discussion is taking place resolve your concern regarding Valjean's proposal? Butwhatdoiknow (talk) 05:08, 17 May 2021 (UTC)[reply]
I'm not really convinced a change is needed, but I don't feel that strongly, so I recommend proposing a specific wording here and seeing what if any comments on it come in. Crossroads -talk- 02:42, 19 May 2021 (UTC)[reply]

I sympathize but flatly disagree. I think that wp:consensus introduces a bit of hysteresis (higher bar for change, and a little extra weight for the status quo) into such decisions (which is a good thing). It is valid stand-alone within it's limited process scope and IMO that should not be in essence deprecated as proposed. It is our process rule for discussion and decisions. Those discussions and decisions are based on all of those other guidleines, policies and considerations, but the process rule is not dependent on them.

Of course, use of it is at the center of a massive amount of Wikilawyering. Which usually involves maneuvering to see who gets to claim "status quo" status for their preferred version/outcome thus requiring the persons with the opposite view to get a supermajority to get their preferred outcome in. For better and for worse, this is how the big fuzzy Wikipedia system works. North8000 (talk) 12:37, 19 May 2021 (UTC)[reply]

North8000, my concern is when fringe editors are ignoring policy-based arguments for keeping long-standing and properly-sourced content and just saying they have a "consensus" to remove it because they are in the majority in the discussion. They have no other policies to justify their move to remove content they don't like. When they do mention a policy, it's often "due weight", which is an equally vague and abused policy that should also be based on other policies. In such a case, "consensus" is not good enough. They must be able to cite other policies to justify removal or change. Their "consensus" must be a consensus that those policies are the justification for their actions. We see lots of attempts to whitewash Trump-critical content based solely on a "consensus" among such editors. Keep in mind we have plenty of editors who still believe The Big Lie and who try to remove what they see as fake news by the dishonest mainstream media, which are precisely the sources we consider RS. Long-time editors are doing this, so they know better than to reveal such thoughts. They just cite "consensus" and delete good content. -- Valjean (talk) 15:06, 19 May 2021 (UTC)[reply]
If, after discussion, no consensus emerges then the status quo prevails. If you don't think that is plainly stated then we can add language to make it explicit. The solution Valjean proposes deals with editors skipping over discussion and proceeding straight to "no consensus." I suggest that these are mutually exclusive fixes. Butwhatdoiknow (talk) 15:50, 19 May 2021 (UTC)[reply]
I'm with you in spirit 100% but not on the specific structure being discussed. Your hypothetical example was in essence a local consensus to make a bad edit. (I stuck to the "hypothetical case" portion of your post because I expect that the real case is more complex.) Wikipedia tries (via "not a vote" etc.) and often fails to make it more than "two wolves and one sheep discussing and reaching a 2/3 consensus on what to eat for dinner" but, repeating myself it often fails at that. The Wikipedia "Plan B" is to get a broad and widely advertised RFC which I think would help in the hypothetical case. You mentioned "due weight"; IMO that portion of WP:NPOV was flawed and unusable even 15 years ago and now with the changes that have happened in the media has become a recipe for doing the opposite of NPOV. But that's a different topic. North8000 (talk) 17:31, 19 May 2021 (UTC)[reply]

Proposed text re "no consensus" not standalone

At Crossroads' suggestion, I offer a first draft of proposed specific wording.

Lack of consensus, standing alone, only supports removal of a change when the content of the change is the subject of (a) a current substantive discussion or (b) a prior discussion that reached consensus. In such cases removing editors should link to the applicable discussion. In all other cases editors should instead cite policy, sources, or another substantive basis for the removal.

Any suggestions for improvement? Butwhatdoiknow (talk) 04:43, 30 May 2021 (UTC)[reply]

To clarify, I am proposing this text for wp:EDITCONSENSUS to discourage substance-free WP:JUSTDONTLIKEIT reversions). Butwhatdoiknow (talk) 20:29, 31 May 2021 (UTC)[reply]
I agree with Crossroads, this change doesn't seem needed. Nikkimaria (talk) 12:46, 30 May 2021 (UTC)[reply]
Well, I agree with Valjean that it does seem needed. Do you foresee the text making the policy worse? If not, any other reason to not add text that other editors believe is helpful? Butwhatdoiknow (talk) 15:17, 30 May 2021 (UTC)[reply]
First off, I think adding text that isn't needed to policy is always going to be problematic - if we want policies to actually be read, we want to avoid bloating them unnecessarily. Second, see point below. Third, this addition seems to contradict other parts of the policy, particularly NOCON. So yes, I foresee it making the policy worse. Nikkimaria (talk) 15:58, 30 May 2021 (UTC)[reply]
To your points: (1) Why do you think this text is not needed? Does it repeat something said elsewhere in the article? Do you think editors reverting solely on the basis of "no consensus" or "talk first" is not a significant enough problem? Something else? (2) See below. (3) wp:NOCON says what happens after the "D" in BRD fails to result in consensus. The proposed text would appear in wp:EDITCONSENSUS, which deals with the "B" and "R" in BRD. Would you please help me see the conflict that you see? Butwhatdoiknow (talk) 17:07, 31 May 2021 (UTC)[reply]
Your stated purpose in putting this forward is to discourage IDONTLIKEIT reverts; I see it as encouraging ILIKEIT edits, and privileging those edits to persist. I certainly don't see requesting discussion when there is disagreement to be a problem; not all potential cases are covered by existing policies or consensuses, but that doesn't make them inherently wrong and meriting change. Regarding point 3, your text is not limited in the way you suggest: it requires there to have been a prior consensus in order to justify a revert, but a prior discussion resulting in no consensus may place the onus elsewhere per NOCON. Nikkimaria (talk) 01:43, 1 June 2021 (UTC)[reply]
Ah, I think I see the problem. Would it resolve your concern if I revised the proposed text to make it clear that (a) the goal is that reverts be supported by a substantive reason ("policy, sources, and common sense") and (b) "lacks consensus" is not a substantive reason unless there was a prior consensus or an ongoing discussion? Butwhatdoiknow (talk) 23:27, 1 June 2021 (UTC)[reply]
I am not sure how this would be different from your existing proposal, which seems to have the same intent? Nikkimaria (talk) 02:28, 2 June 2021 (UTC)[reply]
Right same intent (stopping IDONTLIKEIT). I had thought you are okay with that goal but think the proposed text has unintended consequences (encouraging IDOLIKEIT). Are you not okay with my intent? Butwhatdoiknow (talk) 21:26, 2 June 2021 (UTC)[reply]
I don't see how you can effectively achieve the first without doing the second. What you've suggested as a revision seems to be the same as what you initially proposed. Nikkimaria (talk) 01:25, 3 June 2021 (UTC)[reply]
Clearly, I'm having trouble seeing how the proposed text would encourage ILIKEIT edits. Would you please elaborate? Butwhatdoiknow (talk) 15:35, 3 June 2021 (UTC)[reply]
Your proposal suggests that editors must have either a substantive rationale or an existing/in-process consensus in order to revert a change, but applies no such constraints on making the change in the first place. Nikkimaria (talk) 01:28, 4 June 2021 (UTC)[reply]
Thank you. How about something like this:
Explanations supporting a bold edit or revert should cite policy, sources, or another substantive basis for the change. Lack of consensus, standing alone, only supports an edit when the content of the change is the subject of (a) a current talk page discussion or (b) a prior discussion that reached consensus. In such cases editors should link to the applicable discussion.
Does that resolve your concern? Butwhatdoiknow (talk) 15:47, 4 June 2021 (UTC)[reply]
That's less problematic, but what do you mean by "explanations"? And what happens if there was a previous discussion that resulted in no consensus, as opposed to just no previous discussion? Nikkimaria (talk) 00:30, 5 June 2021 (UTC)[reply]
If you'll permit me, let's work on your first issue first. "Explanations" in this circumstance is synonymous with "statements," "rationales" (reasons). or "arguments" - what an editor would put in an edit summary or talk page post. Do any of those words, or something else entirely. sound better to you? Also, I'm thinking maybe "supporting" should be "in support of." So, for example, the first sentence could begin with "Statements in support of a bold edit or revert ..." Butwhatdoiknow (talk) 05:38, 5 June 2021 (UTC)[reply]
Or maybe ditch "explanations" altogether and go with "Editors should support bold edits and reverts with citations to ..."? Butwhatdoiknow (talk) 15:54, 5 June 2021 (UTC)[reply]
I'm not sure either of those options will be readily understood by those not participating in this discussion. I'm also wondering about how this applies in practice. For example, if I make an edit with the summary "ce" (ie. copyedit), does that count as a "statement"? Is that a "citation"? Nikkimaria (talk) 00:28, 6 June 2021 (UTC)[reply]
Okay, now we have three issues. With your indulgence, I'd like to keep our focus on the first one for now. How about this as the beginning of the first sentence: "Edit summaries and talk page posts in support of a bold edit or revert should include a citation to ..."? Butwhatdoiknow (talk) 01:44, 6 June 2021 (UTC)[reply]
Same question: If I make an edit with the edit summary "ce" (meaning "copyedit"), does that meet this proposed standard? Nikkimaria (talk) 02:17, 6 June 2021 (UTC)[reply]
How about this:
Base edit summaries in support of a bold edit or revert on policy, sources, common sense, or another substantive reason for the change.
"Base" and "common sense" come from wp:DISCUSSCONSENSUS and, I think (you may disagree), is broad enough - in conjunction with "another substantive basis" - to encompass copy editing. (Note: Since the proposed text would appear in wp:EDITCONSENSUS, I've taken out "and talk page posts.") Butwhatdoiknow (talk) 15:27, 6 June 2021 (UTC)[reply]
I would hope anyone making any edit would think it's supported by common sense - but common sense isn't necessarily so common ;-). I don't think this sentence is objectionable as a principle, although I still have concerns about how it may be applied in practice. Nikkimaria (talk) 23:31, 8 June 2021 (UTC)[reply]
Great! If it is applied inappropriately in practice then we can tweak it. On to the second part of the proposed text. Butwhatdoiknow (talk) 00:11, 9 June 2021 (UTC)[reply]
Substituting in the improved first sentence, we have:
Base edit summaries in support of a bold edit or revert on policy, sources, common sense, or another substantive reason for the change. Lack of consensus, standing alone, only supports a bold edit when the content of the change is the subject of (a) a current talk page discussion or (b) a prior discussion that reached consensus. In such cases editors should link to the applicable discussion.
Above, you asked "what happens if there was a previous discussion that resulted in no consensus, as opposed to just no previous discussion?" I'm okay with changing "... (a) a current talk page discussion or (b) a prior discussion that reached consensus" to "... a current or prior talk page discussion." Does that resolve your concern? Butwhatdoiknow (talk) 00:11, 9 June 2021 (UTC)[reply]
My interpretation of the modified second sentence is that lack of consensus in an ongoing discussion is appropriate justification for a bold edit. That's quite a bit broader than even BRD suggests. Was that intended? Nikkimaria (talk) 01:17, 9 June 2021 (UTC)[reply]
Dang those unintended consequences! What if we changed that sentence to "Lack of consensus, standing alone, only supports a bold revert when the content of the reverted change is the subject of a current or prior talk page discussion"? Butwhatdoiknow (talk) 05:59, 9 June 2021 (UTC)[reply]
Does "bold revert" mean something different from just "revert"? Nikkimaria (talk) 12:52, 9 June 2021 (UTC)[reply]
Nope. It's just left over from the prior text ("bold edit"). I'm happy to take it out. Any other concerns? Butwhatdoiknow (talk) 16:03, 9 June 2021 (UTC)[reply]
I'm worried that now we've come back to privileging ILIKEIT over IDONTLIKEIT. Unfortunately I'm not sure of what wording to suggest that might get around that. Nikkimaria (talk) 00:29, 10 June 2021 (UTC)[reply]
Maybe it would help if we had a concrete example to work from. Would you please provide a scenario in which an editor might make a non-revert "no consensus" edit during an ongoing discussion? Butwhatdoiknow (talk) 02:25, 10 June 2021 (UTC)[reply]
If the editors who are abusing this policy to try to win arguments ignore CONLIMITED and the fundamental definition which includes "Decision making and reaching consensus involve an effort to incorporate all editors' legitimate concerns, while respecting Wikipedia's policies and guidelines" then why wouldn't they just ignore this addition, too? It seems like it would just shift the debate to words like "current", "substantive", and applying consensus to something in the past. — Rhododendrites talk \\ 15:46, 30 May 2021 (UTC)[reply]
To your first point: The fact that abusive editors ignore policies should not prevent us from improving polices for the rest of the editing community.
To your second point: First, I welcome any proposed improvements to the text. Would removing "substantive" make it better? Second, this text speaks to the "R" of BRD, not the discussion that ensues thereafter, and would appear in wp:EDITCONSENSUS. Would it help if we changed "only supports removal of a change" to "only supports a revert of a change"? Or can you propose something better to make it clearer that the proposed text does not apply once a discussion has begun? Butwhatdoiknow (talk) 17:07, 31 May 2021 (UTC)[reply]
  • The problem with “no consensus” is that a lack of consensus to do X is too often interpreted as a consensus to do Y. Perhaps we should include a “Consensus is mixed” for situations where viewpoints are fairly evenly split. And perhaps: “Consensus is to NOT do X, but there is no consensus on what TO do instead.” Blueboar (talk) 17:25, 31 May 2021 (UTC)[reply]
I see the problem you are pointing out regarding wp:NOCON. That is certainly worth further discussion. Do you think adding my proposed text to wp:EDITCONSENSUS would impede developing your idea? Butwhatdoiknow (talk) May 31, 2021
Disagree with proposed change. This would just add more wikilawyering as a result of an attempt to reduce wikilawyering. For example if there was material that is just a somewhat bad idea to put into the article (that does not violate any policies), this could be used to wikilawyer to force the material in. The reason is: Most content decisions are the result of weighing multiple factors. Trying to instead prescribe the end result is usually problematic. North8000 (talk) 18:08, 6 June 2021 (UTC)[reply]
Please stand by. I'm (slowly) working on taking into account the concerns expressed above to generate an improved proposal. Butwhatdoiknow (talk) 20:37, 6 June 2021 (UTC)[reply]
I like it. Or something like it. It's super frustrating to me to have edits reverted, they say "reach concensus" first (no other reasoning), then you ask on the Talk page and don't get any feedback at all, then they still won't let you publish your edit. Maybe something like "if concensus is the only negative reasoning given, ask for review on the Talk page, if the other editor does not engage in discussion, see https://en.wikipedia.org/wiki/Wikipedia:Responding_to_a_failure_to_discuss" or something like that... Rogerdpack. (talk) 18:08, 8 June 2021 (UTC)[reply]
Yep, it's a problem alright. I've just added a See also section to wp:Responding to a failure to discuss with two other guides which seem to have independently been written to deal with the problem of the wp:I just don't like it editor. I'm (slowly) working on modifying my draft to take into account concerns expressed above. Please watch this space. Butwhatdoiknow (talk) 20:14, 8 June 2021 (UTC)[reply]
  • Ok… there are several scenarios we need to break out here:
FIRST: there is the situation when an issue HAS been discussed in the past and there IS an existing consensus about it. In this case a revert with “No consensus” or “Get consensus” really means: “Reverting because there is consensus against this.” It may be lazy wording, but it is valid. In this scenario, it is then up to the editor making the edit to go to the talk page and convince others that the existing consensus needs to be re-examined (ie “consensus can change” applies). That will be difficult (especially if the issue has been discussed multiple times). Responding editors should (at a minimum) state that the issue has been discussed before and has been decided. Ideally they would also point to the thread(s) where the consensus was reached. However, we do need to understand that there are some issues that have been discussed SO MANY TIMES that editors grow tired of having to have the same argument over and over again. This is one situation where it is easy to get BITEY. It helps to pin a note explaining the consensus to the top of the talk page (so it does not get archived)… and point to that note when you revert, rather than simply say “no consensus”.
SECOND is the scenario where the issue HASN’T been discussed, but the text has been in place for a long time (ie, there is a “silent consensus” for the current text). I agree that if you revert this with “no consensus” or “get consensus first”, you are indicating a willingness to discuss, and should do so. At a minimum, you should assert “silent consensus” on the talk page, and be ready to defend that assertion.
THIRD: there is the scenario where the issue has not been discussed, the text isn’t that old, and someone reverts with “get consensus”. Here the reverting editor is essentially saying “IDONTLIKEIT” and is making the other person jump through hoops in the hope that they will go away. This is a form of WP:OWNERSHIP, and should be discouraged. However, it is still (ultimately) up to the editor who wants the change to make the first move, and open a discussion. Call in outside opinions to establish consensus.
FINALLY: there is the scenario where the changing editor DOES open a discussion, and no one responds at all. Here is how I think the changing editor should REACT in this scenario: a) if, after a reasonable time (a day or two) no one responds to that thread, I would say it is OK to try the edit again (with an edit summary of “see talk - is there a new silent consensus?”. b) if the edit is again reverted, and STILL no one responds to the talk page thread, you can complain that the reverter is not acting in good faith, and call on outside help (dispute resolution, perhaps even ANI). Blueboar (talk) 21:18, 8 June 2021 (UTC)[reply]
If the edit summary for a revert says "no consensus" and nothing else, how does the reverted editor know which scenario applies? Butwhatdoiknow (talk) 23:25, 8 June 2021 (UTC)[reply]
Um… by going to the talk page and asking. Blueboar (talk) 16:45, 9 June 2021 (UTC)[reply]
In that case I'm not seeing the difference between an edit summary that says "no consensus" and one that says "guess!" Or maybe I'm misunderstanding what you are saying. Butwhatdoiknow (talk) 02:08, 10 June 2021 (UTC).[reply]
There is also the scenario where it has been discussed and no consensus was reached - see elsewhere in this discussion. Nikkimaria (talk) 23:31, 8 June 2021 (UTC)[reply]
True… but that scenario is already dealt with in the guideline… unless there are BLP or copyright issues, we default to the version of the article that existed prior to the edit in question. That does not mean we stop our attempts to come to a clearer consensus… discussions can continue… it just means that “for now” we keep things as they were UNTIL a clearer consensus is achieved. Blueboar (talk) 16:45, 9 June 2021 (UTC)[reply]
  • And there is also a scenario when some user identifies some problem with existing content (e.g., some statement "X") and explains how that content violates our policy (e.g. WP:V), but no consensus is achieved about that (i.e. some users think "X" is well sourced, which means it meets the policy, whereas others think otherwise). What should we do in that situation? Should it be interpreted as "The is no consensus that the statement "X" violates WP:V", or "There is no consensus that the statement "X" meets WP:V"? Currently, the policy explains only one situation: deletion discussion. However, this type discussions are a quite separate case, because if the decision is "delete", the content is permanently removed from WP. Meanwhile, in all other cases, the removed content remains preserved in the page history, which means the deletion rules in that case do not have to be that stringent.
If we apply the same criteria to the articles deletion and deletion of ordinary content, which is preserved in the page history, we seem to have a conflict with other policies. Thus, if WP:V says that all material in Wikipedia mainspace must be verifiable, but the discussion of some statement fails to come to a consensus that it is verifiable (although there is no agreement that it is not verifiable), it seems obvious that that content should be removed, because there is no agreement that it meets WP:V. However, taking into account that removal is supposed to be supported by consensus (and that consensus has not been achieved, because some users still believe the content is verifiable), we cannot remove it per WP:CON. I think, we need to clarify the policy to fix that.--Paul Siebert (talk) 01:25, 9 June 2021 (UTC)[reply]
Please do not confuse “Deletion” (which on WP applies to entire articles) with removal of content. We have different rules for retention/removal of article content than we do for Deletion/Retention of entire articles. Unlike a deleted article, removal of content is NOT necessarily permanent - for example, if content is removed because it is not properly sourced, it can be returned at a later date when a proper reliable source is found. If content is removed because consensus thinks it is not really relevant enough to mention, it can be returned if consensus changes and editors agree that it is in fact relevant (this might happen because a source is discovered that was not considered in previous discussions.) Blueboar (talk) 16:45, 9 June 2021 (UTC)[reply]
Blueboar, that is exactly what I mean: deletion and removal of content are two different things. However, I frequently face a situation when there is a disagreement about removal of some content (not deletion). In a situation when opinia are split about compliance of some content with our policy (e.g. WP:V or WP:NOR), users cite WP:NOCON and claim that that content should stay because there is no consensus to remove it. In connection to that, I propose to clarify the policy. In reality, it seems WP:CON does not explain what should we do in a situation if no consensus is achieved about compliance of some content with our policy: should the content be removed/changed, or it should stay?--Paul Siebert (talk) 17:16, 9 June 2021 (UTC)[reply]
This is why we have policy Noticeboard pages... so when local editors csn not reach consensus, and need outside opinions as to whether or how a policy applies to a specific bit of content, they can ask experienced editors to share an opinion. Another option is to file a formal RFC to determine how to interpret a policy or guideline in a given situation. Blueboar (talk) 17:29, 9 June 2021 (UTC)[reply]
The situation described by me is very common, and even some experienced admins seem not to understand the policy correctly. That means it probably makes sense to modify the policy instead of explaining the same things endlessly. We should explain that if there is a disagreement about compliance of some content with WP:V, that means there is no consensus about compliance of the content with WP:V, and, therefore, the content should be removed. Otherwise, we have a conflict between WP:V and WP:CON.
If we take WP:V as an example, compare these two statements.
It is easy to see that that interpretation leads to a direct conflict between WP:V and WP:CON.--Paul Siebert (talk) 17:40, 9 June 2021 (UTC)[reply]
No… at best it may lead to a potential conflict. The problem is that, in the scenario you describe, the violation of policy is itself the question that has yet to be determined. We have not (yet) determined whether there is a violation of policy or not. So, except for things like BLP or copyright, we default to “we will temporarily say there isn’t a policy violation, but this isn’t final.” Blueboar (talk) 18:54, 9 June 2021 (UTC)[reply]
I am not sure I understand you. If we have a dispute over some potential policy violation, and the opinia are split nearly evenly, the result of that discussion is "no consensus", and that decision is final (for example, that may be a summary of the discussion after a formal closure). However, the question is: ""no consensus" about WHAT?" That the content does NOT violate the policy, or that the content DOES violate it? --Paul Siebert (talk) 19:22, 9 June 2021 (UTC)[reply]
Neither… there is literally no consensus as to whether the content violates policy or not. Further discussion is needed. (And note that there is no such thing as a permanent consensus… because “Consensus can change”. A new discussion, especially one that presents arguments not considered in previous discussions, may result in a changed consensus.) Blueboar (talk) 17:38, 10 June 2021 (UTC)[reply]
I am still not satisfied with your answer. WP:NOCON specifically explains what should we do if the discussion's outcome (i.e. a final result) is "No consensus": Discussions sometimes result in no consensus to take or not take an action. What happens next depends on the context: .... Note, NOCON does not say "further discussion is needed", so this your interpretation is incorrect. Your reference to CCC does not work either, because it is equally applicable to any final outcome of any previous consensus building process. "No consensus" is as a legitimate final outcome as "Support" or "Reject", and NOCON explains what action should be taken if a discussion had not resulted in any consensus. Therefore, I am reiterating my question:
If no consensus is achieved about violation of some policy, should we interpret is as "There is no consensus that the policy has been violated", or "There is no consensus that the policy has NOT been violated"?
It seems that my question is by no means an idle curiosity: the current NOCON wording implies that in a case of disagreement, long lasting policy violations are supposed to be preserved. I am not sure that is in accordance with, e.g., this: This policy is non-negotiable, and the principles upon which it is based cannot be superseded by other policies or guidelines, nor by editor consensus.
--Paul Siebert (talk) 21:47, 10 June 2021 (UTC)[reply]
@Blueboar, are you taking an 'innocent until proven guilty' approach to verifiability?
(@Paul Siebert, I'm not satisfied with the 'status quo for articles' line in NOCON. It is a significant oversimplification of reality. But – speaking as the person who originally inflicted that section on this policy – the NOCON section is meant to be a copy of whatever the other policies and guidelines say, and not a set of new rules itself.) WhatamIdoing (talk) 22:35, 12 June 2021 (UTC)[reply]

Most of the efforts above are in essence trying to prescribe the outcome of particular situations in this policy. While well intended and appreciated, IMHO trying to do that is fundamentally flawed. This policy describes a core mechanism that is used to navigate discussions and decisions in Wikipedia. And a place that it comes up in in contested changes (including reversions of other changes) of all types. This policy can be invoked where the only motivation is to see this process followed, or to try to help sway the outcome on a disputed item. Questions/assertions about the consensus process itself are usually nuanced (vs black-and-white) which is closely related the other nuanced consideration (who gets to claim status quo status). And then those are usually combined with other nuanced policy and guideline considerations in the final fuzzy Wikipedia decisionmaking process. There are trillions of possible sets of considerations and IMO trying to handle the concerns expressed above by trying to prescribe the outcome for particular ones of those is fundamentally flawed. So this policy often simultaneously involved at two levels. (A consensuses about the invocation of consensus :-)) The other thing to keep in mind is that besides creating and giving guidance on the consensus process, this policy is where we attempt create and define the "status quo" in Wikipedia.

Similarly, the trigger for this thread which is reversions done by somebody simply/only saying "no consensus" can mean anything from undoing an un-discussed bold edit (thus making a more thorough discussion the required next step to pursuing it further) to complex wikilawyering (usually under the mild common meaning of that term) to achieve the person's desired outcome with a whole range of possibilities in between. So the situation-of-concern is really a range of fundamentally different situations. So we need to get more specific on what the situation-of-concern is. My guess is that a common one is to in essence assert that the reverted-to version is the status quo and to require the opposing side to meet the more laborious and higher-bar process of achieving consensus to change it. Even this more specific case is actually two assertions in one, and can be normal and legit or quite the opposite. IMO if we want to evolve this policy to give more guidance for the situations of concern, trying to prescribe the outcome of discussions where it is invoked is a fundamentally flawed way to do it. Sincerely, North8000 (talk) 13:08, 10 June 2021 (UTC)[reply]

You say "...the trigger for this thread which is reversions done by somebody simply/only saying 'no consensus' can mean anything from ..." I say "exactly." That is why I am proposing text to require more information when editors use that phrase. Butwhatdoiknow (talk) 15:14, 10 June 2021 (UTC)[reply]
I don’t think we need policy text to require more information… if someone reverts with “no consensus”, and you want more information, just ASK for more information. Blueboar (talk) 11:50, 11 June 2021 (UTC)[reply]
Just checking, do we agree that "no consensus" has so many meanings that it is, in effect, no information whatsoever? Butwhatdoiknow (talk) 16:53, 11 June 2021 (UTC)[reply]
That edit summary provides some information. Specifically, the editor claiming 'no consensus' is claiming that other, unnamed editors agree with the person who says there is no consensus. WhatamIdoing (talk) 22:36, 12 June 2021 (UTC)[reply]
A revert is a rejection of another editor's proposal. It's a matter of civility. The reverting editor has a reason for the revert. It takes a few seconds to give that reason in an edit summary. The proposing editor may accept the reason and the matter is closed. Why choose to instead provide no meaningful rationale and force the proposing editor to take more than a few seconds to open a talk page discussion? Compare WP:SUMMARYNO. Butwhatdoiknow (talk) 16:53, 11 June 2021 (UTC)[reply]
I would agree that “no consensus” conveys no helpful information. I also agree that it is great when editors explain their edits in edit summaries. I do not agree that there is an obligation or requirement to do so. If someone is not sure why an edit was reverted… they can ask for clarification. Blueboar (talk) 18:34, 11 June 2021 (UTC)[reply]
It seems you and I have reached an impasse. One last request: please read the middle two paragraphs of WP:REVEXP with an open mind. Butwhatdoiknow (talk) 20:24, 11 June 2021 (UTC)[reply]
  • The problem is that there absolutely are situations where editors will revert something with a demand that consensus be produced for it (even in situations where there is no existing dispute), then refuse to engage further or to specify their objection. This is a form of stonewalling - it makes it extremely difficult to proceed, especially on low-traffic pages, because it's hard to answer an unvoiced objection. If someone specifies a problem I can try to answer it, then go to the appropriate noticeboard, or start an RFC, or whatever; but if they just say "rv, get consensus" for something that wasn't previously in dispute, there's no easy way to move forward. --Aquillion (talk) 22:55, 11 June 2021 (UTC)[reply]
  • I think the key point is that a specific objection should have to be raised in order to break WP:IMPLICITCONSENSUS; editors should be discouraged from reverting with no explanation at all (or to demand that someone produce a consensus for an edit that was previously, otherwise, uncontroversial - since it previously had implicit consensus, after all.) However, I would not go beyond that and say that the objections need to be policy-based, because while that looks logical at first glance it starts to get a bit convoluted as to what counts as policy - after all, "it reads better the old way" ought to be sufficient; it's fine to revert based on that, at least as the starting point for a discussion. The only thing we ought to be discouraging is circular arguments along the line of "I am reverting this because it lacks consensus; and it lacks consensus because I am reverting it." That is not the starting point of a discussion, that's basically trying to WP:OWN the article. --Aquillion (talk) 22:55, 11 June 2021 (UTC)[reply]
And and the solution to that is to ESTABLISH a consensus. Open a talk page discussion, post a neutral request on a relevant noticeboard asking other editors to swing by the article, examine the situation and give their third party opinions, file a formal RFC if need be. We already have procedures to deal with this… there is no need for more. Blueboar (talk) 11:46, 12 June 2021 (UTC)[reply]
Open a talk-page discussion saying what? Go to which noticeboard? Remember, the premise here is that we're discussing an editor who reverts with no explanation, and who refuses to explain themselves (or asserts that they are not required to explain themselves) when asked on talk. Without an explanation for why they object to the edit, it's difficult to take discussions elsewhere or to attract outside opinions. Just to be completely clear so we're on the same page: Do you believe it is acceptable for an editor to repeatedly, as their standard way of dealing with disputes, revert edits with an edit summary of "rv, get consensus" or some equivalent, in situations where there is no other reason to believe the reverted edits are contentious, without ever engaging beyond that or providing any explanation for their objections? I do not - I think that that is a straightforward conduct issue (as a form of WP:OWN and WP:STONEWALL) and that an editor who persists in it should be told to stop and eventually sanctioned if they don't improve. Like all possible conduct issues, occasional slip-ups aren't a serious problem (no editor is perfect and I'm not suggesting that policy should strictly require an explanation for every edit in all cases), but an editor who consistently behaves that way, or who asserts that they have no need to provide any explanation whatsoever for their reverts, is a problem who needs to be dealt with, and WP:CONSENSUS should (gently) warn editors away from that behavior. Yes, there are ways to work around a disruptive editor like the one I described, provided you can get more people involved - you can go through talk page discussions, messageboards, and even RFCs - but that is true for most conduct issues; it doesn't absolve the editor in question. As I said above, "I am reverting this because it lacks consensus; and it lacks consensus because I am reverting it" isn't a practical or acceptable way to approach consensus-building, and ought to at least be discouraged. This doesn't require formal procedure, just some wording that makes it clear that an editor who asserts that something lacks consensus ought to point to or explain the objection, in situations where it is not plainly obvious. --Aquillion (talk) 16:43, 12 June 2021 (UTC)[reply]
If they leave no edit summary at all, I would ASK why they reverted… open the discussion with: “I made this edit (link)… it was reverted… please explain why?”
If they leave a summary saying “get consensus” or “rv - no consensus” (or similar) then I would start the discussion with a link, then note that the other editor reverted, and end with “… so I am starting this discussion to see what the consensus is”. I would follow that up with an explanation as to why I thing my edit is appropriate.
What happens after that depends on whether other editors engage (or not) and if so, what they have to say. If they continue to be unresponsive… or display an OWNERSHIP mentality… well THAT is all covered under our dispute resolution guideline.
The point is… BE PATIENT. If you don’t know why you have been reverted, FIND OUT. If you DO know why, then explain your side of the argument as best you can, and see how or if others respond. If the “other guy” is continues to be recalcitrant, and unwilling to engage constructively, then follow DVR and seek help. This is standard procedure. Blueboar (talk) 19:13, 12 June 2021 (UTC)[reply]
Just to be completely certain: Do you accept that if a user regularly, across dozens of different articles and in interactions with dozens of different people, reverts with no explanation, or with an explanation of just "rv, get consensus" (in a situation where there is no other reason to believe the edit is controversial), that this is a conduct issue, in and of itself? Is an editor who behaves in that way behaving appropriately? I am not asking for your advice on how to deal with such editors; I am asking what advice you would give to such editors - to an editor who asserts that reverting without explanation is appropriate because WP:ONUS / WP:CONSENSUS means they don't have to justify reversions, and declares that they intend to continue doing so with most disputes they encounter in the future. Do you agree with them that they are doing nothing wrong? --Aquillion (talk) 21:19, 12 June 2021 (UTC)[reply]
It depends. I would want to look at the specific editor’s edit history to see if there is a pattern to their reverts. Blueboar (talk) 15:14, 13 June 2021 (UTC)[reply]
IMO the very specialized situation that you described is problematic and possibly (but not necessarily) a conduct issue. North8000 (talk) 21:25, 12 June 2021 (UTC)[reply]
I echo the points in the last several posts by Blueboar. Expanding on that, a simple revert saying "obtain consensus before doing this" is simply an assertion or claim that consensus must be obtained before making that change. Such covers trillions of possible scenarios and it is problematic to try to make a blanket statement that prescribes what happens next for all trillion of those. What happens next is determined by numerous other particulars. North8000 (talk) 21:17, 12 June 2021 (UTC)[reply]
In the vast majority of cases, editors are not required or expected to obtain consensus before making an edit, and insisting that they do so is highly inappropriate. –dlthewave ☎ 12:45, 13 June 2021 (UTC)[reply]
Yes, this is sort of the gist of what I think needs to be conveyed. To me, "rv, not an improvement" is fine (provided you genuinely think something is not an improvement); but "rv, get consensus" is something that should, at least, be highly discouraged, unless the fact that an edit is controversial is manifestly obvious. Part of the problem that I have with the way some people interpret WP:ONUS in particular is that they believe it allows for and even encourages that - ie. they feel it is appropriate to revert with no more than "rv, get consensus"; and then, when you ask them what their objection is, they'll say "I don't have to explain, you have an WP:ONUS to get consensus for every change you want to make." That isn't true! Since edits are presumed to enjoy consensus, you only have an onus to demonstrate consensus in the presence of an objection; which means that "rv, get consensus" is a circular objection. Now, there are situations where that's not true because the objection is self-evident and well-established - obviously if an edit is manifestly controversial because eg. they've tried to make a similar edit before, there's no need to explain why again - but if you're reverting otherwise-uncontroversial edits with no objection at all beyond a desire to require that someone establish consensus before every single thing they do, that's inappropriate. --Aquillion (talk) 22:38, 15 June 2021 (UTC)[reply]

A step back: edit summaries in support of bold edits

A lot of interesting discussion above. Perhaps we should go back to basics. In Wikipedia:Consensus#Through discussion we already say that editors should:

... try to work out the dispute through [talk page] discussion, using reasons based in policy, sources, and common sense ...

I propose we add similar text to Wikipedia:Consensus#Through editing and offer a first draft:

Edit summaries supporting bold edits or reverts should refer to policy, sources, common sense, or another substantive reason for the change.

What changes would improve this draft? Butwhatdoiknow (talk) 03:05, 13 June 2021 (UTC)[reply]

  • Leaving detailed edit summaries is nice, but is NOT required. In fact, we are not required to leave ANY summary at all. I oppose any attempt to mandate edit summaries. Blueboar (talk) 09:14, 13 June 2021 (UTC)[reply]
How about "Edit summaries supporting bold edits or reverts, if provided, should refer to policy, sources, common sense, or another substantive reason for the change"? Butwhatdoiknow (talk) 15:25, 13 June 2021 (UTC)[reply]
If I were a wikilawyer, I would argue that as long as my summary mentions the word “consensus”, it IS a reference to policy (ie this guideline)… or I would simply stop giving any summary at all. Blueboar (talk) 20:22, 13 June 2021 (UTC)[reply]
Reading between the lines, it appears that "if provided" resolves your first concern regarding mandatory edit summaries.
Regarding your second concern, I think you are reading the proposed one-sentence text correctly: it does not prohibit empty edit summaries and it does not prohibit "no consensus" edit summaries. Would you have it read differently? Butwhatdoiknow (talk) 22:24, 13 June 2021 (UTC)[reply]
I would have it not mention edit summaries … at all. Blueboar (talk) 01:51, 14 June 2021 (UTC)[reply]
Something like "Bold edits and reverts should be supported by policy, sources, common sense, or another substantive reason for the change"? Butwhatdoiknow (talk) 06:19, 14 June 2021 (UTC)[reply]
  • You skipped the most common reason. That the person making the edit believes that it will improve the article. And nearly any edit can be called a bold edit. At best such would a be gigantic change for the entire English Wikipedia. North8000 (talk) 02:20, 14 June 2021 (UTC)[reply]
Ah, but is it an improvement? An edit summary saying something more than "I think this is a good idea" goes a long way toward showing other editors that it is. Butwhatdoiknow (talk) 06:15, 14 June 2021 (UTC)[reply]
Agreed. So does that lead to saying that we should create a rule to mandate it? North8000 (talk) 23:10, 15 June 2021 (UTC)[reply]
Yes. But I realize that's not going to happen. So my hope is that we can encourage substantive edit summaries. Hence, my proposal below. Butwhatdoiknow (talk) 23:19, 15 June 2021 (UTC)[reply]
In what way does the proposed text change "the entire English Wikipedia"? Can you cite some policy or guidance that it alters? Butwhatdoiknow (talk) 06:15, 14 June 2021 (UTC)[reply]
I'll give you the precise logical answer even though it might unintentionally seem short and flippant. The policy that it changes is WP:consensus. My previous post was a description of the impacts of such. North8000 (talk) 23:08, 15 June 2021 (UTC)[reply]
Which previous post is that? Butwhatdoiknow (talk) 23:19, 15 June 2021 (UTC)[reply]

Blueboar's proposal

What about: “When making bold edits or reverts, be prepared to explain and discuss them on the talk page if another editor raises questions or concerns.” I think this would cover all eventualities. Blueboar (talk) 11:40, 14 June 2021 (UTC)[reply]

Nice. What do you think about adding a sentence after that saying that supporting an edit with a meaningful edit summary may avoid the need for a talk page discussion? Butwhatdoiknow (talk) 14:44, 14 June 2021 (UTC)[reply]
Meh… I would rather not mention edit summaries at all. The problem is that edit summaries are (by design) brief … too brief to be really meaningful. If there is a need for editors to communicate meaningfully, the talk page is a better venue for doing that. Blueboar (talk) 20:02, 14 June 2021 (UTC)[reply]
Here's a sample revert edit summary: "violates wp:NPOV." Would you characterize that as not really meaningful? How about one that says "unsourced, see wp:SOURCE"? Butwhatdoiknow (talk) 22:31, 14 June 2021 (UTC)[reply]
I don't see which is your point. Are you saying that they are short yet valid thus seeking to refute Blueboar's post, or are you saying that those are problem edits because they are too short? North8000 (talk) 00:23, 15 June 2021 (UTC) '[reply]
I think these short edit summaries are meaningful. But perhaps Blueboar doesn't. In that case I would like to understand why. Butwhatdoiknow (talk) 01:35, 15 June 2021 (UTC)[reply]
No, I would not call them meaningful. Sure, they give me some clue as to why the revert took place, and they point me towards policies I should review, but they don’t give me any details as to how those policies relate to the revert or to the material that was reverted. I would still have to go to the talk page and ask some questions if I had concerns.
The same is true for an edit summary of “get consensus” or “no consensus”… sure, those tell me that the policy I should review is WP:Consensus, but for details I would need more info from the reverter. For that, I would have to go to the talk page.
I am not opposed to edit summaries (they do have a use). But they are not good for giving details or explanations. Blueboar (talk) 11:56, 15 June 2021 (UTC)[reply]
"I would still have to go to the talk page ... if I had concerns." Right. But will you allow that in some cases a reverted editor who reads "unsourced, see wp:SOURCE" might find that explanation sufficient and not have any further questions or concerns? Butwhatdoiknow (talk) 16:48, 15 June 2021 (UTC)[reply]
Sure, they might (or might not). Just as someone reading “No Consensus” or “Get consensus” might feel that that explanation is sufficient (I would). In fact, someone seeing an unexplained revert with no edit summary (at all) sufficient.
I was under the impression that this entire thread is focused situations when one editor doesn’t find an brief edit summary sufficient. Blueboar (talk) 14:50, 15 June 2021 (UTC)[reply]
My view is that the entire thread - all the way back to "Consensus" is not a standalone policy - is about whether "no consensus" is a meaningful edit summary for a revert. But the current discussion is focused on whether we should add a sentence to your proposed text that says supporting an edit with a meaningful edit summary may avoid the need for a talk page discussion. If we shouldn't then the "no consensus" question is probably moot. Butwhatdoiknow (talk) 21:50, 15 June 2021 (UTC)[reply]
  • . Just as someone reading “No Consensus” or “Get consensus” might feel that that explanation is sufficient (I would). Just to be completely clear again: Do you accept WP:EDITCONSENSUS as policy, which establishes that all edits are presumed to enjoy consensus? That being the case, what exactly do you feel a "get consensus" edit summary actually conveys that a blank edit summary does not? I am honestly completely baffled that an experienced editor would believe a "rv, get consensus" edit summary is appropriate for reverting otherwise uncontroversial edits - to me it is no different from a blank edit summary, and routinely reverting with that rationale would be a conduct issue indistinguishable from routinely reverting with no edit summary at all. Policy plainly establishes that all edits enjoy consensus until someone objects; that being the case, a revert whose rationale demands consensus for an edit that nobody has raised any other objection to is reverting with no rationale at all. --Aquillion (talk) 22:51, 15 June 2021 (UTC)[reply]
Thank you. Will you also allow that a substantive edit summary is more likely to provide an explanation sufficient to avoid the need for a talk page discussion than a blank edit summary? Butwhatdoiknow (talk) 21:50, 15 June 2021 (UTC)[reply]
  • The same is true for an edit summary of “get consensus” or “no consensus”… sure, those tell me that the policy I should review is WP:Consensus, but for details I would need more info from the reverter. No, absolutely not. "rv, WP:UNDUE" (or any other non-circular objection) is completely fine as an edit summary and I would have no issue with it, since that makes it clear what the fundamental objection is. Ideally I would obviously want them to be able to give more details on talk, especially if the rationale is unclear, but even if they don't, that much is still more than enough to start discussions, ask on an appropriate messageboard, and even go for a proper and reasonable RFC ("does this edit put WP:UNDUE weight on [crux of the edit]?") An edit of "no consensus", on the other hand - in situations where there is no reason to think the edit being made is particularly controversial, eg. no pre-existing disputes related to it - is completely inappropriate because WP:EDITCONSENSUS establishes that edits are presumed to enjoy consensus. Reverting an edit solely because you want the person who made it to establish consensus is therefore a circular objection - "I am objecting to this edit because it lacks consensus; and it lacks consensus because I object to it." It provides no information whatsoever - a blank edit summary would, to me, actually be preferable, because at least it makes it unambiguous that the edit is being reverted without a rationale. Reverting with "get consensus" is using legalese to obscure that fact that in a way that could confuse or mislead inexperienced editors, and can waste time by falsely implying that there's an existing objection or consensus that the reverted edit goes against. Again, the issue isn't "sufficient explanation" vs. "insufficient explanation"; the issue is whether they provide any explanation at all. "rv, get consensus" is not an explanation for a revert and provides no information beyond what would be contained in a blank edit summary. --Aquillion (talk) 22:51, 15 June 2021 (UTC)[reply]
I disagree. I find “rv, get consensus” to be perfectly acceptable as an edit summary for a revert. It isn’t hugely meaningful, but it is acceptable. At minimum, it is a request for a discussion on the talk page. What is wrong with that? Blueboar (talk) 00:45, 16 June 2021 (UTC)[reply]
The problem with a "no consensus" edit summary. Butwhatdoiknow (talk) 06:16, 16 June 2021 (UTC)[reply]
Blueboar's idea looks good to me. It's not a big change but reinforces and puts a point on correct practice and putting it in this key place would reduce the issue that started this thread. North8000 (talk) 21:35, 14 June 2021 (UTC)[reply]

More discussion of the "no consensus" edit summary

  • This completely misses the core issue, which is that all edits are presumed to enjoy consensus until someone objects and that lack of consensus is therefore never a valid rationale to revert an otherwise unobjectionable edit. --Aquillion (talk) 22:51, 15 June 2021 (UTC)[reply]
That doesn't seem right....could you expand on that? I think that the common meaning of consensus in Wikipedia is that some type of discussion has occurred and concluded, with this policy providing the framework when needed. North8000 (talk) 23:14, 15 June 2021 (UTC)[reply]
Right, Blueboar's text does not speak to whether "no consensus" is a meaningful edit summary. That said, do you have any other objection to the text? Butwhatdoiknow (talk) 23:26, 15 June 2021 (UTC)[reply]

Ok… let’s break this down further, to see where we can agree and where we don’t. Here is how I see it:

  1. An edit summary of “Not consensus” is a clear assertion that a consensus exists, and the edit violates that existing consensus. Solution: Go to the talk page for details and discussion.
  2. An edit summary of “Get consensus” is a clear assertion that a consensus does NOT yet exist, but the reverting editor would like one to be formed. It is a request to halt bold editing and shift to discussion mode. Solution: Go to the talk page and discuss.
  3. An edit summary of “No consensus” is between these two, and thus not as clear. Another editor seeing this summary is not sure whether the reverter is saying “this goes against existing consensus” or “there is literally not a consensus and I would like to form one”. Solution: Go to go to the talk page and ASK.

In other words… in all three situations, the solution to any objections, questions or confusion is the same: talk page discussion. And because all three are clarified/resolved by talk page discussion, I don’t really care what the edit summary says, or whether it is “meaningful” or not… We end up in the same place! Blueboar (talk) 10:15, 16 June 2021 (UTC)[reply]

One of the things that struck me as I was catching up on this conversation today is the Wikipedia:Product, process, policy difference. The example of "rv, WP:UNDUE" indicates a perceived problem with the substance of the edit. "Get consensus" does not indicate any problem with the content; it indicates a perceived problem with the process the editor used to make the edit – that the process is the most important thing, to the point that no content-related objections were even worth mentioning.
I don't feel like we should treat perceived procedural problems as being as big a deal as perceived content problems.
@Blueboar, you have left out several options, including the one in which I revert you because I disagree, I know that consensus is on your side (or at least not against your edit), and I'm just trying to prevent consensus from being implemented. "I can't win on the merits, but I can drag this out for weeks" is not exactly unusual among our more determined wikilawyers. WhatamIdoing (talk) 22:38, 16 June 2021 (UTC)[reply]
Thing is, griping about whether an edit summary is meaningful or not is also a procedural complaint. Especially when we remember that we are allowed to not leave any edit summary at all. Blueboar (talk) 11:38, 17 June 2021 (UTC)[reply]
An interesting point, which has led me to suggest at Wikipedia talk:Editing policy#Edit summaries for reverts that we start requiring descriptive edit summaries for reverts. Levivich 13:47, 17 June 2021 (UTC)[reply]
@Blueboar, I don't think I agree. "I won't tell you what's wrong with your content; I'm only going to tell you what's wrong with your process" is not equivalent to "Please tell me what's wrong with my content". WhatamIdoing (talk) 17:31, 18 June 2021 (UTC)[reply]
  • WP:CCC already says: Editors who revert a change proposed by an edit should generally avoid terse explanations (such as "against consensus") which provide little guidance to the proposing editor (or, if you do use such terse explanations, it is helpful to also include a link to the discussion where the consensus was formed). Furthermore, WP:BRD already says that the editor whose edit was reverted is expected to discuss the proposed change instead to starting an edit war. I agree that BRD is not a policy (and therefore is not mandatory), so it probably makes sense to make it a part of the policy.
However, the most important point here is as follows. Frequently, users speak on behalf of some consensus without bothering to make sure such a consensus had ever been achieved. In my opinion, it is disruptive. Although CCC already explains that, I would propose to add a separate clause that stipulates that the user who refers to some consensus (for example, by reverting with the edit summary such as "against consensus", "no consensus", "please, obtain a consensus first") MUST provide an exhausting information about the previous discussion when the explicit consensus was achieved on that matter upon request. A refusal to do so automatically nullifies that type arguments. In other words, if a user B reverts a user A with an edit summary (or a talk page post) "I revert you because you are acting against consensus", a user A can legitimately argue: "Please, explain when and how that consensus had been achieved?" If a user B is unable to give a satisfactory answer, the text modified by a user A cannot be considered a consensus version, and a de novo consensus building process may start. We also must explain that "no consensus" etc refer only to the consensus that was achieved through a discussion, not to the implicit consensus achieved through editing. Similarly, a long standing version should not be automatically considered as a consensus version.--Paul Siebert (talk) 23:16, 16 June 2021 (UTC)[reply]
I fully agree with @Paul Siebert that Frequently, users speak on behalf of some consensus without bothering to make sure such a consensus had ever been achieved.
However, I think the proposed solution might need more work. For example, editors should not be expected to provide information about previous discussions for blatant policy violations; pointing to the policy should be sufficient. (Of course, those editors shouldn't be claiming 'no consensus', either; they should just say 'rv massive BLP violation' or whatever the problem is.) WhatamIdoing (talk) 17:35, 18 June 2021 (UTC)[reply]
Both are better than no explanation at all - and, as your examples demonstrate above, "no consensus" is a non-explanation.
To your comment: All reverts do not require full blown "better" explanations. Actually, make that most reverts do not require full blown "better" explanations. If you can avoid the bother of opening up a talk page discussion with a simple and informative edit summary then why not do it? Butwhatdoiknow (talk) 19:16, 18 June 2021 (UTC)[reply]
Perhaps it’s because I don’t find opening or contributing to a talk page discussion to be a bother… while I do find remembering to leave edit summaries to be a bother. I rarely bother to leave edit summaries, and don’t pay much attention to the edit summaries left by others. On the other hand, I use talk pages a lot. Blueboar (talk) 20:44, 18 June 2021 (UTC)[reply]
If the community adopts this proposal and you open talk page discussions instead of writing edit summaries then I can almost guarantee you that no one will say you have violated the spirit or intent of the proposal. Butwhatdoiknow (talk) 01:04, 19 June 2021 (UTC)[reply]
I agree with Blueboar. Plus trying to write something that is prescriptive about edit summaries for particular situations IMO is bound to be messy and sort of WP:Creep. Plus inevitably whatever is in the edit summary is inevitably going to be a weak version of what is in talk and in rough situations you'll see people preying upon that. North8000 (talk) 20:56, 18 June 2021 (UTC)[reply]
"Explain edits reverts" doesn't seem particularly messy to me. What am I missing? Butwhatdoiknow (talk) 01:10, 19 June 2021 (UTC)[reply]
Does “Explain edits” include explaining the initial bold edit, or does it only apply to a revert? Blueboar (talk) 12:54, 19 June 2021 (UTC)[reply]
I have edited to clarify. Butwhatdoiknow (talk) 16:56, 19 June 2021 (UTC)[reply]

I don't really see an issue with the existing wp:Consensus text. I think one of the bigger issues occurs when one side acts as if they have consensus when the other side doesn't agree. In such cases edit wars can occur and often are "won" by the side with more numbers. If anything I think it should be emphasized that if editors don't agree that a consensus has been reached then we assume one hasn't. The next steps would be either follow the rules of NOCON and/or get an uninvolved editor to help decide if consensus has been established (typically a RfC). I agree that No Consensus by itself isn't a valid reason to revert but in most cases there will be an obvious talk page discussion at which to point. It is always best to make sure the edit summaries include more than just "no consensus" but sometimes context is clear in the talk history and edit history. TLDR/ I don't think a prescriptive change is needed, I could be open to a suggestion that says "if an edit is reverted based on no-consensus, its best practice to include an edit summary that points to the discussion that reached the no-consensus and it is rarely appropriate to revert a new edit (ie not restoring recently disputed content) as "no consensus" with no additional explanation. Springee (talk) 03:12, 19 June 2021 (UTC)[reply]

@Blueboar: re "What about a revert with an edit summary saying “Please discuss on talk page”" Actually, that is a good question. I know at least one article, World War II where the users achieved a local consensus that all important changes should be discussed on the talk page first. In my opinion, that approach is fruitful in such high profile articles, and the reverts with such an edit summary are quite ok. Maybe, we should add that into the policy specifically for high profile or good/featured articles. However, I am not sure that approach can be universal, although I do not oppose to it either.
re "What I still do not understand is the focus on edit summaries." I see absolutely no problem if the edit summary says just "see talk page for further details". To me, it does not matter where and how the revert is explained: it may be either a talk page or an edit summary. The point is that if some change is reverted because it ostensibly violates consensus, that revert must be explained with the reference to some consensus that had been explicitly achieved in the past. We cannot and we should not describe any edit that goes against consensus that was achieved through editing and the edit violating consensus. The reason is simple: consensus through editing is a series of subsequent additions/deletions, which last until it is settled to some more or less stable version. However, the process of consensus building through editing has no clear final point, so it is impossible to say when and an which point it had been achieved. Therefore, any new edit may be equally considered as a continuation of the editing process. Therefore, any reference to previous consensus should refer to some discussion that explicitly lead to some consensus.
@WhatamIdoing: Obviously, policy violations are a totally different group of cases. I am talking specifically about the cases when the current text was supported, or is claimed to be supported, by some previously achieved consensus, but neither the old or a new versions violate any policy.
In general, the main situation we need to address is as follows. Sometimes, a group of users claim that the current version is supported by consensus, but they repeatedly refuse to provide further evidences. In that case, a user who repeatedly requests for such evidences may be accused of stonewalling and reported at ANI. From my experience, that sometimes leads to severe actions against that user, because the community frequently does not analyse evidences, and when several users claim that the consensus had been achieved in the past, the community is prone to believe to them, and repeated requests for evidences is sometimes interpreted as a continuation of stonewalling behaviour.
In connection to that, the policy must make it mandatory that any reference to previously achieved consensus must be supported by evidences that should be provided upon a request.--Paul Siebert (talk) 20:00, 19 June 2021 (UTC)[reply]
  • “Please discuss on talk page,” standing alone, is not okay because it is not an explanation. It gives the reverted editor (and the rest of the community) no meaningful information regarding why the edit was reverted. (And, I would add, it violates the intent and spirit - if not the letter - of wp:EDITCONSENSUS.) That said, if (as in the example above) there is a local consensus to pre-discuss "important changes" (whatever those are) then "Please discuss on talk page" with a link to the discussion establishing the local consensus would be okay. Butwhatdoiknow (talk) 20:23, 19 June 2021 (UTC)[reply]
An edit summary of “please discuss on the talk page” isn’t intended to be an “explanation”… it is a request, indicating a desire to follow the “consensus through discussion” approach to achieving consensus. Blueboar (talk) 21:18, 19 June 2021 (UTC)[reply]
Problem no. 1: WP:DISCUSSCONSENSUS is designed to resolve disputes. “please discuss on the talk page” gives no indication of the existence of or reason for any dispute.
Problem no. 2: Whatever the intent, by accompanying the "request" with a revert the reverting editor is making a demand, not indicating a desire. And let's say the reverted editor gives in to that demand and starts the talk page discussion the reverting editor couldn't be bothered to start? What would the reverted editor say? I suppose it would be "I'm here. What are we discussing?" Why not answer that question in the revert edit summary? (Or, if the explanation is lengthy, by starting a talk page discussion.) Butwhatdoiknow (talk) 22:21, 19 June 2021 (UTC)[reply]
Re 1: the fact that an edit has been reverted indicates that it is disputed. Even if the person who wants to make the edit doesn't know why, they should be able to provide a justification for the edit, or else why would they have made it? Re 2: isn't this the whole point of BRD - if you make a bold edit and it's reverted, you should start discussion? The reverting editor could of course, but saying that they must is in opposition to that suggested procedure. Nikkimaria (talk) 00:13, 20 June 2021 (UTC)[reply]
Yes, there IS a difference. When a user "A" reverts a user "B" that means that the latter disagrees with the former, and they are in equal positions: each of them are supposed to justify the ideas they are advocating: the user "A" has to provide additional evidences and/or argument to support the new version, whereas the user "B" is supposed to prove that the current version is better.
In contrast, when the user "B" says the edit was reverted because there is "no consensus", it implies that the user "A" is acting against some consensus, so the editor's "A" activity may be potentially disruptive, so a prolongation of that dispute may be interpreted as stonewalling, which may inflict sanctions on the user "A". Therefore, the "no consensus" edit summary (and variations thereof) can be seen as implicit accusations of potential misbehaviour, and, to put "A" and "B" in equal positions, our policy is supposed to put additional burden of evidence on those who refers to some "consensus" during editorial disputes.--Paul Siebert (talk) 00:33, 20 June 2021 (UTC)[reply]
When a user "A" reverts a user "B" that means that the latter disagrees with the former – if only that were always true, and we didn't have editors who reverted bold edits on (bad) principle, including in cases when they personally agreed with the edit.
@Blueboar, imagine that you've made a reasonable change. I revert you. My edit summary says "Please discuss on the talk page". Do you feel confident that you can guess what I want you to discuss? WhatamIdoing (talk) 01:52, 20 June 2021 (UTC)[reply]
Generally, yes… I know you want to discuss the bold edit I made. I may not know why you want to discuss it, but I would respect your request nevertheless. I would open the discussion by asking why you reverted. Blueboar (talk) 10:48, 20 June 2021 (UTC)[reply]
Assuming good faith, I am reading your answer to say "no, I would not feel confident that I can guess what you want to discuss about the edit I reverted." That brings us back to my unanswered question: Why not extend the reverted editor the courtesy of an edit summary that explains the reason for the revert? Butwhatdoiknow (talk) 15:34, 20 June 2021 (UTC)[reply]
Because I can explain the reason for the revert betterand more completely on the talk page. And that is even MORE courteous. Blueboar (talk) 21:35, 20 June 2021 (UTC)[reply]
You know the issue isn't edit summaries vs. talk page discussions. But I'll play along and rephrase the question: Why not extend the reverted editor the courtesy of a contemporaneous explanation (in an edit summary or a talk page discussion) that gives the reason for the revert? Butwhatdoiknow (talk) 05:30, 21 June 2021 (UTC)[reply]
It's the 'contemporaneous' aspect that bothers me.
Bob reverts Alice, and tells Alice to start a discussion. Alice has no idea why Bob reverted it, but she starts a discussion. It says "Hey @Bob, I'm starting the discussion. Why did you revert this edit?"
Then – ideally – Bob posts exactly what he should have posted in the first place.
Why are we trying to force Alice to post a useless, zero-content comment before Bob posts the explanation? What value is upheld, or what benefit is provided by making Alice start the discussion, instead of having Bob start the discussion? WhatamIdoing (talk) 15:57, 21 June 2021 (UTC)[reply]
I think you and I may be on the same page. Here's what I am trying to ask: Is there any good reason a reverting editor might have for not giving an explanation for the revert contemporaneously with the revert itself? Isn't that what you are asking as well? Butwhatdoiknow (talk) 16:42, 21 June 2021 (UTC)[reply]
Yes, but I have the impression that @Blueboar did not agree with us. WhatamIdoing (talk) 04:53, 22 June 2021 (UTC)[reply]
WhatamIdoing, this section is discussing only a very specific category of reverts, namely, the reverts of the edits that go, or that are perceived to go, against consensus. Therefore, I am not sure your comment is relevant.
With regard to "Please discuss on the talk page", usually, I use that type edit summaries in two cases: first, when I am reverting the edits in such high profile articles as World War II, and the edit is too bold and/or obviously non-professional, and, second, if some new user made some edit that affects the topic that is being actively discussed on the article's talk page. --Paul Siebert (talk) 01:57, 20 June 2021 (UTC)[reply]
Re 1: Is the edit disputed? All we know from “please discuss on the talk page” is that the reverting editor wants to talk. Re 2: If it's reverted with an explanation and you agree with the explanation then you should not start a discussion on the talk page. Again I ask why not explain a revert in the edit summary? Butwhatdoiknow (talk) 00:32, 20 June 2021 (UTC)[reply]
Re "why not explain a revert in the edit summary?". It is always better to explain, although in most cases it is not mandatory. IMO, there is one case when that should be made mandatory: when the reverting user refers to some consensus (see my response to Nikkimaria).--Paul Siebert (talk) 01:49, 20 June 2021 (UTC)[reply]
If it is always better to explain a revert then why not say that editors should always do it? Butwhatdoiknow (talk) 02:18, 20 June 2021 (UTC)[reply]
Because in many cases that is not absolutely necessary. If we introduce artificial requirements, the rule will not work. We should make some requirement mandatory only when it is really needed, otherwise we discredit our policy.--Paul Siebert (talk) 05:33, 20 June 2021 (UTC)[reply]
Why is it "really needed" when a reverting editor refers to some consensus? What makes that case different from the "not absolutely necessary" cases? Butwhatdoiknow (talk) 06:18, 20 June 2021 (UTC)[reply]
  • Consider the following situation… a section of an article was worked on by several editors a few years ago… It took time, but eventually it stabilized into a consensus version. However, this was achieved without any discussion. Instead, everyone boldly tweaked each other’s edits and had their edits boldly tweaked (including occasionally being reverted and trying a different approach). Back and forth … until everyone was satisfied. Textbook “consensus through editing”.
Now, a new editor comes by… and breaks the consensus everyone worked so hard on (perhaps unknowingly re-adding something that was considered and rejected over the course of the back and forth). So, one of those original editors (knowing the editorial history of what was previously considered and rejected) reverts back to the consensus version. Since there is no discussion to point TO… what should the reverting editor put as an edit summary? Blueboar (talk) 21:35, 20 June 2021 (UTC)[reply]
(Note - Personally, I would just put “see talk” and give the details there… however, BWDIK wants the edit summary itself to have an “explanation” and so this isn’t acceptable… so my question stands.) Blueboar (talk) 22:32, 20 June 2021 (UTC)[reply]
Blueboar, you know that is not what I want. Do I read your comment above to say that editors should give contemporaneous explanations ("the details) of reverts in either an edit summary or a talk page? If so then we are in complete agreement. Butwhatdoiknow (talk) 05:47, 21 June 2021 (UTC)[reply]
I understand what you mean, but the opposite situation is even more likely: some new user sees an obvious blatant error, or NPOV violation etc, but the user who was active in this article before and who, for some reason, wants that error to be preserved, reverts the attempt to fix it under a pretext that the reverted text is ostensibly a consensus version.
If we assume that all long lasting content is by default the content people had been working hardly with, then we must openly say that any long lasting version should be considered a consensus version. I personally believe that would be a grave mistake.
A more concrete answer to that question is as follows. If the user who reverts a bold edit can demonstrate that the stable version was worked on by several editors a few years ago, and that it had stabilized into a consensus version (no serious dormant disputes, no strong conflicts etc), then it is easy to demonstrate, and these evidences must be provided. If the user who reverts the bold edit was a part of that implicit consensus building process (or at least is familiar with the details of that process), they can easily provide needed evidences. Thus, a good example of edit summaries can be This had been discusses in the past and rejected 3 years ago. See the article history and talk page discussion. If you believe you have fresh arguments, please, present them on the talk page first.
The burden of evidence must be on the user who refers to a previous consensus. Actually, the policy always says about that, although it sounds more like a recommendation. I propose to make this requirement mandatory.
With regard to: "Personally, I would just put “see talk”", absolutely agree. That is what I myself am doing frequently. It does not matter where and how the information is provided, but it must be provided.--Paul Siebert (talk) 22:24, 20 June 2021 (UTC)[reply]

Let's look at this structurally. "No consensus" like a zillion other commonly used edit summaries and things that Wikipedians write is not logically specific which means that it can cover many different statement and assertions (including assertions of what situation exists) in the context of many different situations that actually exist. Let's narrow the infinite possibilities down to just 20 situations that actually exist and 20 different assertions that the person could mean for each which brings it down to about 400 situations covered by a (potential) "no consensus" edit summary. And so we're having discussions above about a handful of those 400 to formulate a prescriptive policy (yes, this is a policy page) to prescribe what must be done for those situations. So following this concept, we're going to have to first list the 400 different situations covered by a (prospective)"no consensus" edit summary and then write a policy on what should happen for each of those 400. When we're done with that then we can start with #2 of those zillion.

Or we can write a policy that requires explanatory and logically explicit edit summaries. Which inevitably would be violated about 100,000 times per day so we'll need to expand WP:ANI capacity to handle those. The brief edit summary that I'll use for this edit ("+comment") will become one of those violations.

Or else we can use the current system for those which is that if the edit/revert is contested a discussion ensues generally starting with talking about what the current situation is and an expansion on what the editor is asserting. Sincerely, North8000 (talk) 18:14, 21 June 2021 (UTC)[reply]

I already proposed to narrow it down to just one case: if some user writes "no consensus" (be it an edit summary or talk page post), that user must explain what consensus is meant, and when it was achieved. Without that, the policy implicitly puts the burden of proof that no consensus exists on those who proposes a change. That contradicts to the policy's spirit. The policy says:
A consensus decision takes into account all of the proper concerns raised.
When a user X writes: "No consensus", they imply that the edit made by a user Y is close to the previously proposed version that had already been rejected, and that the arguments Y presents had already been addressed in the past. That means the user X has that information, otherwise the statement made by X would be an obvious lie. That means it would be easy for X to share this information with Y, by doing that, X shows respect to the opponent.
Furthermore, if Y starts browsing the article's history in attempt to find when and where the consensus had been achieved, that may take a lot of time, and, importantly, the search may lead to somewhat different result, which by no means facilitates the consensus building process. Meanwhile, by explaining what consensus they refer to, the user X puts this information into a common knowledge domain, and that, per Aumann's agreement theorem, greatly facilitates a consensus building process.
In general, this theorem must be made a core of this policy, because it significantly narrows the range of issues where people can agree to disagree.
If the requirement to explain the revert is too strict, we can do the following:
"If some user reverts the edit with such an edit summary (or a talk page post) as "No consensus", "against consensus" etc, that user must provide an exhaustive information about the consensus they refer to upon request. If that user is unable to do so, the reference to the previous consensus is deemed non-legitimate."
--Paul Siebert (talk) 18:59, 21 June 2021 (UTC)[reply]
I totally agree in spirit… a revert should be accompanied by an explanation. But calling an unexplained (or poorly explained) revert “non-legitimate” is just an invitation to Wikilawyering. Encourage, don’t mandate. Blueboar (talk) 22:29, 21 June 2021 (UTC)[reply]
I don't think that we need to require exhaustive information. Just some information should usually be enough. "See talk page" (if there's a relevant discussion there) gives the bold editor a starting point. Even "previously rejected" gives you a hint about what the problem is. An unexplained "No consensus" could mean anything from "I don't like it" to "this is good edit, but I want you to jump through some hoops first" to "I can't believe that you missed the huge fight about this last month". WhatamIdoing (talk) 05:02, 22 June 2021 (UTC)[reply]
Blueboar and WhatamIdoing, actually, I didn't mean explanations are mandatory. I meant that if you revert someone with the edit summary "no consensus" etc, you must be ready to explain that your action when you are asked about that. If you fail to explain when and how that consensus had been achieved, the reference to consensus is not deemed legitimate, and you cannot say that the user whom you reverted was acting against consensus. That includes possible complaint at ANI, AE, etc. Moreover, if you fail to provide evidences and continue to maintain the proposed change violate previous consensus, that your activity may be considered disruptive. Obviously, the consensus achieved through editing is not covered by that rule, because in that case it is hard to discriminate between a consensus version and just a last stable version.
Specifically, that is needed to prevent some users from speaking on behalf of the community in a situation when they hardly have a right to do so.--Paul Siebert (talk) 01:14, 24 June 2021 (UTC)[reply]
I can certainly agree with the principle that an editor should explain their edits/reverts when asked to do so. I see such a request as being the bridge between “consensus through editing” and “consensus through discussion”.
In “consensus through editing” We don’t actually need to explain our edits… we simply edit (and reverting is just another type of edit - part of the back and forth of the editing process). However, as soon as someone effectively says: “hold on, I want to discuss this” (whatever the language used to convey this request) we should shift to “consensus through discussion” - and explain.
It does not matter who initiates this transition. It is usually the editor who gets reverted (by asking “why”)… but the reverting can certainly initiate the transition (by anticipating the request and preemptively posting an explanation). Blueboar (talk) 12:23, 24 June 2021 (UTC)[reply]
Yes, and I think we need to add that to the policy. Currently, there may be a conflict between WP:ADMINSHOP and WP:CCC. The latter just recommends that terse explanations should be avoided, whereas the former says: (Editors who refuse to allow any consensus except the one they insist on, and who filibuster indefinitely to attain that goal, risk damaging the consensus process. As a result, a user who is repeatedly asking for evidences that some text reflects previously achieved consensus may be accused of filibustering or stonewalling. Indeed, very frequently, users respond to such requests just by saying: "Please, stop it. You are acting against consensus.", and the current version of the policy perfectly allows them to do so. We must fix that.
The following change should be made to WP:CCC. Instead of:
"Editors who revert a change proposed by an edit should generally avoid terse explanations (such as "against consensus") which provide little guidance to the proposing editor (or, if you do use such terse explanations, it is helpful to also include a link to the discussion where the consensus was formed)."
I propose:
"Editors who revert a change proposed by an edit should generally avoid terse explanations (such as "against consensus") which provide little guidance to the proposing editor. If you do use such terse explanations, you may be asked to provide a link to the discussion where the consensus was formed. In general, in any discussion, an editor who refers to some previously achieved consensus may be asked to provide additional information on that matter. If the editor fails to do that, their reference to the consensus that was ostensibly achieved in the past cannot be considered valid, and their arguments may be disregarded"
--Paul Siebert (talk) 16:06, 24 June 2021 (UTC)[reply]
  • I just realized another reason for my hesitancy regarding all this: There seems to be an assumption that a summary that just says “no consensus” means the reverter thinks a previous consensus exists.
I think that assumption is flawed. Sure, It could mean that, but it could ALSO mean that the reverting editor dislikes the edit (or is simply unsure of it) and is requesting discussion in order to establish consensus on the edit. It is an attempt to say that there is “no consensus” (yet)… so let’s discuss and FORM one. Sure, there are probably better ways to say that… but we should assume good faith, and a blunt “no consensus” can be interpreted as a good faith attempt to request a discussion. Blueboar (talk) 01:30, 25 June 2021 (UTC)[reply]
Do I correctly deduce from this that you also take the position that "I reverted this edit because I don't like it" is a good faith edit summary? Butwhatdoiknow (talk) 02:28, 25 June 2021 (UTC)[reply]
Yes, it may be a good faith edit summary, if a reverter is ready to elaborate on that and join a talk page discussion Actually, that is what WP:BRD says.--Paul Siebert (talk) 03:48, 25 June 2021 (UTC)[reply]
  • I couldn't find that language at wp:BRD. Is it from somewhere else?
What I did find at BRD is "briefly explain why you reverted." The problem, of course, is that wp:JUSTDONTLIKEIT is not a meaningful explanation. Or are you saying that an explanation doesn't have to be meaningful as long as the editor is prepared to be meaningful in a talk page discussion? In other words, is it good faith for an editor who is willing to engage on a talk page to say "yellow octopus" in the edit summary of every revert they make? Butwhatdoiknow (talk) 05:25, 25 June 2021 (UTC)[reply]
"I reverted this edit because I don't like it" is a good-faith edit summary (when reasonably accurate). The person doing this isn't trying to harm the encyclopedia; the editor just thinks the edit in question did not improve the page.
Saying "no consensus" when the problem is "I (personally and individually) don't agree" is potentially deceptive. We usually interpret "no consensus" as meaning "many editors disagree" – not just me. WhatamIdoing (talk) 02:13, 28 June 2021 (UTC)[reply]
Perhaps I should have asked: Is "I just don't like it" a good faith reason to revert? Is wp:JUSTDONTLIKEIT wrong? Butwhatdoiknow (talk) 15:24, 28 June 2021 (UTC)[reply]
I think it can be. It isn’t a Policy based reason, but not every change/revert needs to be policy based. Of course, the reverting editor should be prepared to follow up and explain WHY he “just doesn’t like it” (if asked). Blueboar (talk) 17:23, 28 June 2021 (UTC)[reply]
Right, edits need not be policy-based (e.g., "copy edit" is fine). The problem with "I don't like it" is that it is not substantive - it adds nothing to the fact of revert itself. The reverting editor is giving the reverted editor no "why" information. Butwhatdoiknow (talk) 19:37, 28 June 2021 (UTC)[reply]
"Copy edit" doesn't give you "why" information either - if I add a comma with the edit summary "copy edit", that provides a reasonable description of what I did, but not why I think there should be a comma there in the first place. Nikkimaria (talk) 22:31, 28 June 2021 (UTC)[reply]
"Copy edit" gives you a "what" and implies a substantive "why" (to improve the text (as opposed to the content)). "I like it" and "I don't like it" give you neither. Butwhatdoiknow (talk) 16:40, 29 June 2021 (UTC)[reply]
Don’t care… (once again) if you want to know WHY someone “just does not like it”… just ASK them! It isn’t onerous. Blueboar (talk) 16:55, 29 June 2021 (UTC)[reply]
The fact that you don't care is not a reason "based in policy, sources, and common sense." Assuming, without conceding, that forcing other editors to open a discussion to obtain the reason for an edit is not onerous, then it isn't onerous to force the editor who makes the change to explain it from the get-go. I will (once again*) ask the question that you have yet to answer: Why not save the step and have editor who makes the change provide a contemporaneous explanation?
* See above at: 22:21, 19 June 2021 / 02:18, 20 June 2021 / 05:30, 21 June 2021 / 16:42, 21 June 2021 Butwhatdoiknow (talk) 20:41, 29 June 2021 (UTC)[reply]
Because I can not control what some other editor puts (or does not put) in their edit summaries. What I CAN control is how I respond to whatever they put (or don’t put). If I need clarification, it’s up to ME to ask for clarification. Blueboar (talk) 22:40, 29 June 2021 (UTC)[reply]
But we're not talking about what you can control, we're talking about what our policies should say. Our policies are full of aspirational requirements and prohibitions that seek to "control" editor behavior (with varying levels of success). Why shouldn't requiring contemporaneous explanations for edits be one of those rules? Butwhatdoiknow (talk) 23:40, 29 June 2021 (UTC)[reply]
Is "I have a good reason for this edit. Start a talk page discussion and ask me if you want to know what it is." a good faith edit summary? Butwhatdoiknow (talk) 19:37, 28 June 2021 (UTC)[reply]
Sure, if the person making the edit is willing to do just that if someone asks. It may not be to everyone's preference, but not all edits are easily explained within the confines of an edit-summary - even something like "per WP:V", while technically correct, may oversimplify. Nikkimaria (talk) 22:22, 28 June 2021 (UTC)[reply]
  • "not all edits are easily explained within the confines of an edit-summary" Can't argue with that. In such cases the appropriate practice* would be to explain it on the talk page contemporaneously with the edit (with an edit summary of "see talk").
* I say "appropriate practice" because a substance-free edit "explanation" impedes consensus-building and creates extra work for any editor who wants to know the edit what and why, Further, in the case of reverts, it is discourteous to the reverted editor. Butwhatdoiknow (talk) 16:40, 29 June 2021 (UTC)[reply]
  • In the case of an edit that will obviously be controversial, an immediate talk-page post is reasonable. But in other cases, an edit summary that simply states what and not why is entirely acceptable, because no one is likely to question it. This is why a lot of our anti-vandalism tools use just a "Reverted..." summary by default - I don't really need to explain why I think the article's text should not be replaced with "poop", and I don't particularly worry about being discourteous to the editor who thought it should. Nikkimaria (talk) 20:45, 29 June 2021 (UTC)[reply]
Fair enough: bad faith editors don't deserve an explanation. But what about good faith editors? Let's not throw the baby out with the bathwater. Make vandalism an exception to the general rule. Does that resolve your concern? Butwhatdoiknow (talk) 22:21, 29 June 2021 (UTC)[reply]
@Butwhatdoiknow, this is kind of a tangent, but you keep asking whether something is good faith, and it's bugging me.
"Good faith" means that the person is trying to help. That's it.
If your neighbor gets up at sunrise and plays his trumpet so loudly that the entire neighborhood woke up, that's "good faith" unless and until you discover that your neighbor was actually trying to hurt you. Even if the neighbor was trying to wake everyone up on purpose (maybe because he'd been reading that people who wake up earlier live longer, and he wants everyone to live longer), that's still good faith.
For that edit summary to be "bad faith", you'd have to have a reason to believe that the editor was actively trying to harm Wikipedia (or another editor) with that edit summary. It's possible to cause harm in an edit summary, but those tend to take the form of vandalism, WP:OUTING, lying, or other really obvious problems. Merely reverting something that I don't think improves a page (BTW, that falls under "common sense" in the policy) or claiming that there's no consensus (maybe there isn't!) is not evidence of trying to harm Wikipedia. It may be a suboptimal edit summary. It may be undesirable. It may be inefficient. But it's not an active attempt to deceive or hurt people.
Bottom line: Please stop asking whether those edit summaries are made in good faith. They are. WhatamIdoing (talk) 01:53, 30 June 2021 (UTC)[reply]
Between active good faith and active bad faith lies a gray area of laziness, thoughtlessness, and discourtesy that - whatever the intent - harms Wikipedia by making editing more frustrating and making the community less collegial. Would it bug you less if I referred to such behavior as lacking in civility? Butwhatdoiknow (talk) 05:23, 30 June 2021 (UTC)[reply]
It would certainly bug me to have edits labelled as lazy, thoughtless, or discourteous, simply because you didn't like the associated edit summary. Nikkimaria (talk) 13:06, 30 June 2021 (UTC)[reply]
Me too! My understanding is that the topic of this discussion is substance-free edit explanations, not the edits they fail to meaningfully explain. Butwhatdoiknow (talk) 15:01, 30 June 2021 (UTC)[reply]
Blueboar, let me remind you that the section that we are talking about is WP:CCC, i.e. this section specifically discusses the cases when the reverter thinks the previous consensus does exist. I agree with what you say in general, I think we should not restrict users in the methods of interaction and collaboration. However, in this concrete section we must focus on one specific case: what should we do when some users believe they are defending a consensus version, whereas others express a doubt that that version really reflects consensus? I proposed the solution, and I am not pretending it is a universal rule: that rule regulates a relatively narrow type of cases. If you have an idea on how to explain that in a more clear way, please, propose your version. --Paul Siebert (talk) 03:23, 25 June 2021 (UTC)[reply]
But I think the current language is just fine as it is. Blueboar (talk) 13:51, 25 June 2021 (UTC)[reply]

This discussion covers trillions of possibilities.... a common, incomplete edit summary which inevitably gets somehow clarified in the next few exchanges. IMO trying to get prescriptive here about what should happen is not a good idea. North8000 (talk) 14:30, 25 June 2021 (UTC)[reply]

Yup… We can give advice, outlining things that we find helpful when trying to reach a consensus (which we currently do) … but we can not (and should not) proscribe how to reach a consensus (nor can we proscribe how not to do so). If we find that a particular comment or approach isn’t helpful, we shift gears and try something else.
Fact is, reaching a consensus can be messy and is often frustrating. There will be false starts and miscommunications. Sometimes it will be hard to assume good faith… and yet we try to do so. And occasionally, no matter how hard we try, we find that consensus just isn’t possible. Blueboar (talk) 16:58, 25 June 2021 (UTC)[reply]
Actually, we should, can, and do proscribe how to reach consensus. wp:EDITCONSENSUS calls for "Substantive, informative explanations." wp:DISCUSSCONSENSUS calls for "reasons based in policy, sources, and common sense." Granted, once each side has stated their case the path forward may "cover trillions of possibilities." But we're talking about whether each side must state their case (that is, must say more than "I disagree"), not what happens after that. Butwhatdoiknow (talk) 19:02, 25 June 2021 (UTC)[reply]
It says “should”, not “must”… it’s (good) advice, not a proscription. Blueboar (talk) 20:23, 25 June 2021 (UTC)[reply]
"Must" is a synonym of "should." And the sentence that precedes the "should" sentence says "Explain the purpose of the edit" without equivocation. (It recently said "Use clear edit summaries that explain the purpose of the edit," also without equivocation.) - Butwhatdoiknow (talk) 23:36, 26 June 2021 (UTC)[reply]
Must is not a synonym of should. Most policies use those words in a manner that is consistent with RFC 2119. One of the recommendations in Wikipedia:Policies and guidelines is to use those words precisely (specifically, to not use should to describe requirements: you must not violate copyrights). WhatamIdoing (talk) 02:17, 28 June 2021 (UTC)[reply]
(assuming you meant "prescribe" and not "proscribe") By "prescriptive" I meant saying very specifically what should happen. "Prescribe" can have broader uses including what you refer to. North8000 (talk) 20:34, 25 June 2021 (UTC)[reply]
Do you mean the policy is not mandatory? To the best of my understanding, we have a separate category pages, called "guidelines" that contain non-mandatory advice. In contrast, the policy says something that we must observe without reservations. We cannot engage in original research (not "should not"), we cannot post non-verifiable information (not "should avoid" doing that), etc. Why WP:CON is an exception? The WP:CCC is written in a style that is more appropriate for guidelines, not a policy, and I disagree with your approach that de facto mixes the latter with the former.--Paul Siebert (talk) 20:45, 25 June 2021 (UTC)[reply]
Did you intend to address that to me? Because I don't see how it relates to what I said / intended to say. North8000 (talk) 20:54, 25 June 2021 (UTC)[reply]
No, I didn't. As you can see, our posts were separated by less than 10 minutes, I would simply have no time to respond.--Paul Siebert (talk) 00:07, 27 June 2021 (UTC)[reply]
Paul, the line between policy and guideline is not so clear cut. Many of our guidelines contain firm “rules” (worded as “do”, “don’t” and “must”), and many of our policies contain advice (worded as “should”). Blueboar (talk) 21:22, 25 June 2021 (UTC)[reply]
Yes, and I see it as a problem that we should fix.
In particular, the consensus policy is needed to address one important issue: to emphasize the difference between consensus and vote. That is the key aspect, because, by its nature, Wikipedia is being written by amateurs, for amateurs, and mostly in an amateurish way, although some (few) users may be professional experts or the amateurs who are digging deeper into the subject. If a consensus building process is not focused on a quality of arguments, amateurs (who are much more numerous) will always dominate over professionals, which negatively impacts WP quality. To avoid that scenario, the policy's preamble says that consensus is not the result of a vote. However, the same preamble says that consensus on Wikipedia does not mean unanimity, which means that even when some user disagrees, we still can talk about consensus. The problem is that the user (or users) who disagrees with the majority may be the one whose opinion is based on the most deep and thoughtful analysis of sources. I recommend you to read this article (I hope it is freely available online), where the discussion on the Vietnam war talk page is being analyzed by a scholar who is systematically studying Wikipedia. He notes that Wikipedians efficiently " changes the criteria we use to judge expertise, albeit, I would argue, without replacing them with much that could be construed as progressive". In other words, in a discussion that involves several superficial users and one or few users who are digging deeper into the subject the former quickly start to dominate, because the arguments of the latter fall into deaf ears, and by repeating them, the later is engaged in what is called "stonewalling".
By making WP:CON deliberately vague (by imposing a guideline style on it), we greatly facilitate that outcome. We need more strict procedure that would not allow some users to refer to "consensus" in a situation when no consensus was achieved, or the consensus building process was just an implicit !vote.--Paul Siebert (talk) 00:07, 27 June 2021 (UTC)[reply]
Well, now you are getting into something more fundamental than reverts and edit summaries. I would agree that the way in which we reach consensus has evolved over the last few years. consensus discussions are increasingly being structured as !votes… battles between pre-set options… whereas we used to structure them as more open-ended discussions, where people attempted to find compromise. Perhaps Wikipedia was less “litigious” in those days. Blueboar (talk) 17:47, 28 June 2021 (UTC)[reply]
Yes, I see the problem as a fundamental, and the problem is in an extreme inhomogeneity of the Wikipedia community and Wikipedia content. Many (if not majority) articles, such as anime, rock bands, computer games, small towns, local schools, etc, are devoted to very simple topics, and they are based on relatively low quality sources. Other articles talk about science, history, etc., and much more serious sources are available on those subjects, and, importantly, these topics require more expertise and more specific approaches towards sources selection. However, the policy is the same for all those topics (BLP, MedRES are rare exceptions). Wikipedia policy is inherently more friendly to amateurs than to experts, and it is more focused on creation of a comfortable and collaborating atmosphere for users, whereas the debates between professionals may be rather hot, and they may require many rounds of arguments and counter-arguments. To non-expert, behaviour of an expert looks close to stonewalling, and that may lead to elimination of the expert from the discussion, or even to their topic bans (we all know the examples). Therefore, we need a tool that would hepl an expert in that situation. We need the rule that would allow a user who brings strong arguments and high quality sources to prevail in a dispute even if that user is outnumbered by non-expert users who believe they reflect consensus. And I described it above.
To avoid misunderstanding, let me re-iterate: we need a multilevel concept of consensus that would allow a user (or a small group of users) who brings higher quality sources, stronger arguments, and who are digging deeper, to elevate a consensus-building process to a more professional level and to protect them from the users who may say: "Look we already have a consensus, and we are comfortable with it. We do not want to return to that issue anymore, because the issue is already resolved, and if you will continue your stonewalling, you will be reported at ANI". I think it is extremely important, because it may significantly increase quality of high level Wikipedia article, and prevent true experts from leaving the project.--Paul Siebert (talk) 19:44, 30 June 2021 (UTC)[reply]

Policy disagreement

Zero0000 and I disagree on the application of this policy.

As far as I understand it, the burden to achieve consensus is on the person who is removing referenced information which is not obviously wrong and doesn't have BLP issues. Zero0000 believes that per WP:ONUS the burden is on the editor who wants to keep the information even if it's been there for a long time. Would be happy to hear third opinion. Alaexis¿question? 07:01, 8 June 2021 (UTC)[reply]

I agree that "the burden to achieve consensus is on the person who is removing referenced information which is not obviously wrong and doesn't have BLP issues". --Bduke (talk) 07:35, 8 June 2021 (UTC)[reply]
I agree that the burden to achieve consensus is on the person who is removing longstanding content which has WP:IMPLICITCONSENSUS. I believe the key word in WP:ONUS is "achieve": The onus to achieve consensus for inclusion is on those seeking to include disputed content. Content with implicit consensus has already achieved consensus. This has been debated exhaustively in the archives here and at WT:V. Kolya Butternut (talk) 08:12, 8 June 2021 (UTC)[reply]
(Note the relevance of this to the discussion in the previous section.). The difference between WP:ONUS and WP:NOCONSENSUS is that WP:ONUS specifies who has the responsibility for obtaining a consensus, while WP:NOCONSENSUS specifies what "commonly" happens if an attempt to achieve consensus has failed. The very purpose of ONUS is to distinguish between inclusion and exclusion. What NOCONSENSUS most definitely does not mean is "it has to stay in the article until I agree to remove it". The case that Alaexis hopes to get support on by asking an "in principle" question shows exactly that in my opinion. Note that the argument for removal includes an RS challenge that Alaexis has not even responded to. Zerotalk 08:34, 8 June 2021 (UTC)[reply]
Please don't misrepresent my position. I'm not saying that the information in question needs to stay until I agree to remove it. I'm saying that contentious changes need to be discussed and this discussion has barely started. Alaexis¿question? 09:09, 8 June 2021 (UTC)[reply]
I'm generally in favor of long term material staying but that's not an iron clad way around ONUS iirc there was a hoohah over some medical stuff sitting around for ages and that had to go. If there is a valid objection to the material, accuracy, sourcing, the usual, it needs to come out.Selfstudier (talk) 09:27, 8 June 2021 (UTC)[reply]
You are right, however none of this is applicable here: it's not a medical or BLP question, the source is not prima facie unreliable and it's quoted accurately. Alaexis¿question? 09:41, 8 June 2021 (UTC)[reply]
It would help if you gave us a link to your discussion. Is it a discussion between two people who haven't found a middle ground? Perhaps you should seek a third opinion regarding the substances of your dispute rather than regarding what to do when there is no consensus. See wp:3O. Butwhatdoiknow (talk) 15:39, 8 June 2021 (UTC)[reply]
Thanks, I believe we would be able to reach an agreement. If not we would seek third opinion. Alaexis¿question? 05:38, 9 June 2021 (UTC)[reply]
(Butwhatdoiknow: I linked to it above.) There is already a third opinion and Alaexis has reverted again without bothering to respond to my RS challenge. Sorry folks, this doesn't belong on this page. Zerotalk 08:56, 9 June 2021 (UTC)[reply]
@Butwhatdoiknow, it seems to be a question about who counts as an 'immigrant' when the border changes. It reminds me of this joke:

An old woman is applying for housing in Russia.  She fills in the form that the bureaucrat gives her.

Where were you born?  St Petersburg

Where did you go to school?  Petrograd

Where did you get married?  Leningrad

Where do you want to live?  St Petersburg

The joke, of course, is that these are all the same city.  The only thing that changed is the legal name of the city. One approach is to treat everyone who lives in the same place as 'immigrating' to a new country whenever the national border shifts. Another approach is to treat everyone as staying put and not immigrating, even though they are obviously affected by the change. The solution for the article is probably to spell it out in more detail – the equivalent of "during the Handover of Hong Kong, the 6.5 million residents of Hong Kong stopped being counted as residents of UK territories and started being counted as residents of China." WhatamIdoing (talk) 22:56, 12 June 2021 (UTC)[reply]
More pointfully for this page: If two editors can, in good faith, claim that Policy Says™ the opposite things, then we need to fix these policies. WhatamIdoing (talk) 22:57, 12 June 2021 (UTC)[reply]
I had attempted to fix this with an RfC which failed. My feeling is that part of the problem is that we are using a content policy as a conduct policy. Kolya Butternut (talk) 00:44, 13 June 2021 (UTC)[reply]
@Kolya Butternut, you might be right about that (using a content policy as a conduct policy). Wikipedia's policies are rather "organic" (unstructured?) in nature, and we do sometimes use a hammer when a screwdriver would be more logical. WhatamIdoing (talk) 01:47, 13 June 2021 (UTC)[reply]

Everything in Wikipedia articles needs to be referenced or reference able. So "referenced" is the norm, not a reason to give some material special status. North8000 (talk) 02:14, 14 June 2021 (UTC)[reply]

Edit summary vs explanation in policy change

Butwhatdoiknow changed the text of the policy under "Through editing", from "edit summaries" to "explanations". Although I agree somewhat with the change, I think it warrants clarifying that edit summary preferably should not be omitted when making an edit and adding an entry in the talk page explaining the edit. I tried to include such language but the aforementioned editor undid my edit, with the rationale that it was instruction creep. As the policy previously stated, "edit summaries are especially important when reverting another editor's good faith work." If we just replace "edit summaries" there with "explanations" without further clarification, some editors will likely read this policy and just leave an explanation in the talk page without an edit summary, when actually doing that is cumbersome for other editors because instead of just going through the history of the page and reading edit summaries, they would have to be clicking on the talk page and looking for the relevant thread about the edit change. Therefore, I think if the policy is to be changed to include "explanations" instead of "edit summaries", a clarification should be made that most of the time is preferably to write an edit summary and if needed add a talk page thread as well about the edit. Thinker78 (talk) 15:11, 23 June 2021 (UTC)[reply]

  • Sorry, I didn't read your text as saying what you say above. With regard to what you say above, I'm all for adding text that a "see talk" edit summary should be added when that is where the editor leaves an explanation. Going beyond that places a burden on the "talk page explanation" editor to explain twice.
You suggest this is a reasonable thing to do because folks checking edit histories will otherwise have to hunt for and find the associated talk page post. However, as a practical matter the overwhelming talk page history views of our hypothetical edit will occur within days of the edit itself. In those cases finding the talk page post should be a snap. Later, because the rare curious viewer would know the date of the edit, finding the talk page post should not present the high hurdle you posit for finding the relevant post. Butwhatdoiknow (talk) 23:18, 23 June 2021 (UTC)[reply]
  • My take on this is that edit summaries are a convenience, not an end unto themselves. They are nice, and are (usually) helpful, but they are not necessary. What IS necessary is dialog between editors whenever disagreement or confusion crops up.
It does not matter where or how that dialog takes place (it can be in edit summaries, it can be on the article talkpage, it can be on a user’s talkpage , it can be at a relevant noticeboard, etc.) It does not matter who initiates the dialog (if the other editor doesn’t, you can). It is the dialogue that is important, not the specific method or venue. Blueboar (talk) 14:29, 24 June 2021 (UTC)[reply]
  • I don't think edit summaries are in general the right place for dialogues, other than in a limited fashion. One of the reasons is because other editors are not privy of getting involved in that dialog, unless they make edits themselves. That's the purpose of the talk page. I would say also if there is a certain dialog developing in edit summaries in a page history it might indicate edit warring. Thinker78 (talk) 16:54, 24 June 2021 (UTC)[reply]
Depends… let’s say editor A adds some new bit of information to an article, but editor B thinks it is worded poorly. B might rewrite the addition with an edit summary such as “How about this?”… and that might be followed by another tweak by editor C with “Or this”. Now let’s say the initial editor (A) thinks these rewrites omit something important, he might revert with “needs to say ‘XYZ’”. So B tries a slightly different rewrite (this time saying ‘XYZ’), with “Is this what you mean?”… Those edits and summaries are a dialogue (of sorts). And this is perfectly acceptable.
I have even seen “dialog” take place with just edits (ie without anyone leaving any summary whatsoever)… especially in the early stages of building an article, where everyone is trying out different ways to word the same basic information.
That said… yeah… I would agree that the more complex a disagreement is, the less we should rely on an edit summary to convey our concerns. Talk pages are the best venue to discuss more complex disagreements or concerns. Blueboar (talk) 17:56, 24 June 2021 (UTC)[reply]
  • There are additional recent changes which are also additional policy changes. Some that pretty much say that any unexplained edit is a violation of policy. I was going to revert but stopped just short of doing so.North8000 (talk) 14:00, 24 June 2021 (UTC)[reply]

Explanation of July 7 edit

On June 28 I moved this sentence:

If an edit is reverted and further edits seem likely to meet the same fate, create a new section on the associated talk page to discuss the issue.

from the 4th to the 3rd paragraph of wp:EDITCONSENSUS. On June 30 I edited the sentence to read:

If you can't [think of a compromise edit that addresses the other editor's concerns], or if you do and your second edit is reverted, create a new section on the associated talk page to discuss the dispute.

Another editor noticed the original sentence had been removed from the 4th paragraph (but, I suspect, did not realize that it had been moved to the 3rd paragrah and edited) and restored it to the 4th paragraph. Now we say the same thing twice: once in the 3rd paragraph and once in the 4th paragraph.
I'm removing the sentence in the 4th paragraph and copying it over the sentence in the 3rd paragraph. If there is no revert of that edit then, after a week or so, I'll try restoring the edited version of that sentence. Butwhatdoiknow (talk) 03:27, 7 July 2021 (UTC)[reply]

Thanks for explaining. I think your edited version is fine. --Matt Smith (talk) 03:44, 7 July 2021 (UTC)[reply]

Not so obvious

The second paragraph of wp:EDITCONSENSUS begins:

All edits should be explained (unless the reason for them is obvious) ...

I suggest that the reason for an unexplained edit is never obvious to an editor scanning a Watchlist or a Revision history. Accordingly, I propose removing "(unless the reason for them is obvious)" from this text. Butwhatdoiknow (talk) 01:45, 3 August 2021 (UTC)[reply]

Disagree. For example, I am replying to this edit with the edit summary of "re"; someone scanning a watchlist or revision history will understand that I'm replying to something without me needing to explain why. Nikkimaria (talk) 02:31, 3 August 2021 (UTC)[reply]
I would say that your "re" is an explanation. As you say, it gives someone scanning the history a clue regarding what the edit is about. My concern is the blank edit summary, which contains no information whatsoever. (That said, I can see having a special rule for talk pages, where almost all of the edits are replies.) Butwhatdoiknow (talk) 05:45, 3 August 2021 (UTC)[reply]
It's an explanation of what the edit is, but not the reason (why the edit was made). Similarly in mainspace "create", "ce", and many more potential summaries say what but not why, but the proposed change would disallow them. If your intention is rather to disallow blank summaries, whether entirely or by namespace, this isn't the optimal approach to achieve that (plus that needs a sitewide RfC). Nikkimaria (talk) 12:12, 3 August 2021 (UTC)[reply]
What if we change the first sentence:
All edits should be explained (unless the reason for them is obvious)—either by clear edit summaries indicating the reason why the change was made, or by discussion on the associated talk page.
to
All edits should be described and, unless implicit in the description, explained—either by an edit summary indicating the reason for the change, or by discussion on the associated talk page.
That makes it clear that, for example, "ce" is okay because it implies a reason: to improve the text. Butwhatdoiknow (talk) 16:00, 3 August 2021 (UTC)[reply]
I don't think there is a significant different in how people will be interpreting "obvious" vs "implicit" - and as I said, if your intention is to mandate edit summaries (and actually this new proposal goes beyond that), this isn't the way to go about it. Nikkimaria (talk) 00:37, 4 August 2021 (UTC)[reply]
This edit now makes it clear that edit summaries are not mandated, Incorporating that change into the proposed text, using "obvious" instead of "implicit," and doing a bit of copy editing, we have:
The Wikipedia community strongly encourages editors to describe and (unless obvious from the description) explain all edits—either in the edit summary or in a post on the associated talk page.
Do you have any objection to that text? Butwhatdoiknow (talk) 04:58, 16 August 2021 (UTC)[reply]
What benefit do you see in that change? Nikkimaria (talk) 12:17, 16 August 2021 (UTC)[reply]
I'll be happy to answer your question after you have answered mine: Do you see any harm in the proposed change? Butwhatdoiknow (talk) 15:48, 16 August 2021 (UTC)[reply]
It makes the wording more complex - I think this is generally the wrong way to go for policy pages unless necessary, and I'm not seeing a "necessary" here. Nikkimaria (talk) 00:00, 17 August 2021 (UTC)[reply]
Thank you. Yes, instructions should be worded as simply as possible. On the other hand, important points should not be sacrificed at the alter of simplicity. I suggest a two step analysis: (1) Is the point important enough to make? (2) If it is, what wording that will make the point as simply as possible?
Step (1) brings us to the answer your question: the benefit would be the elimination of the "unless the reason ... is obvious" exception from the current text (The Wikipedia community strongly encourages meaningful explanations for all edits (unless the reason for them is obvious) ...). This exception is problematic because (a) noting is obvious from blank edit summary, (b) all edits are obvious to the editor making them, and (c) lazy editors use the exception as an excuse to not leave an edit summary. Butwhatdoiknow (talk) 05:49, 17 August 2021 (UTC)[reply]
So, putting aside the precise wording for now (that's step (2)), do you agree that we should eliminate the "unless the reason ... is obvious" exception? Butwhatdoiknow (talk) 05:49, 17 August 2021 (UTC)[reply]
This brings us back to what I said above: if you're trying to mandate edit summaries, this isn't the way to do that. Nikkimaria (talk) 12:54, 17 August 2021 (UTC)[reply]
As I said above, and as Blueboar notes below, this edit now makes it clear that edit summaries are not mandatory, What I am trying to do is eliminate the "unless the reason ... is obvious" exception because it is problematic (as I explain above) and confusing (as Blueboar says below).
Again I ask: putting aside the precise wording for now, do you agree that we should eliminate the "unless the reason ... is obvious" exception? Butwhatdoiknow (talk) 16:05, 17 August 2021 (UTC)[reply]
No. I can easily think of edits that are genuinely obvious without any explanation - for example, a page creation - and if we're not intending to require edit summaries then the concern of "excuses" doesn't apply. Nikkimaria (talk) 00:28, 18 August 2021 (UTC)[reply]
Before I respond I want to make sure I understand your first point. Would you please give me a couple more examples of edits that are obvious? Butwhatdoiknow (talk) 01:37, 18 August 2021 (UTC)[reply]
Most things in your own userspace would also fall into that category, whether that's drafting (apart from some thing that do require summaries, like copying from another page), editing your userpage, etc. Nikkimaria (talk) 02:35, 18 August 2021 (UTC)[reply]
I can see an argument that explanations for such edits are less useful because third parties are unlikely to search those pages' Revision histories. But I'm having trouble seeing how the substance of such edits would be "obvious." What am I missing? Butwhatdoiknow (talk) 15:37, 18 August 2021 (UTC)[reply]
I think by your reference to the "substance" of the edit, I've spotted a difference of interpretation. In context I read the current phrasing as not requiring providing a "why" for particular edits; am I correct in understanding that you're using it as not requiring a "what does this edit do?" summary? Nikkimaria (talk) 03:59, 19 August 2021 (UTC)[reply]
Let me try answering your question with this categorization: (1) Blank edit summaries have no substance. An editor would have to check the diff to know what the original editor did. In short, the substance of the edit is not obvious. (2) Edit summaries such as "revert," "reply," and "added content" have some substance. An editor looking at a Revision history would know the general "what" of the original edit but the specifics or the "why." (3) Edit summaries with what and why information such as "rvv" or "added content re Cicero didn't say it" have more substance and, in my opinion, are the gold standard. Butwhatdoiknow (talk) 16:18, 19 August 2021 (UTC)[reply]
I would disagree on all points. (1) is not universally true - if I create a page it is clear even with no edit summary what I am doing for someone glancing at the edit history. (2) is not universally true - even leaving aside the case of an inaccurate summary, sometimes it is sufficient to say only what and not why. Giving this edit a summary of "re" is not any less useful than "re because I have something to say", and arguably copying my entire comment into the summary (which is a more comprehensive "what" but not a "why") would be more useful than the "why" summary for someone glancing through the edit history. And so (3) is not universally true - explaining why can be helpful where there is likely to be good-faith disagreement about a particular edit and where that explanation can reasonably be included in an edit summary... but that is not every case. Nikkimaria (talk) 03:00, 20 August 2021 (UTC)[reply]
Yes, I see why a page creation "edit" needs no explanation. But is that the only exception? I'm looking for a couple more examples of edits that are obvious. (And by "obvious" I mean someone looking at a blank edit summary in a Revision history would have any information regarding what the original editor did. Please let me know if you are using a different definition.) Butwhatdoiknow (talk) 05:22, 20 August 2021 (UTC) @Nikkimaria: please reply. Butwhatdoiknow (talk) 05:30, 25 August 2021 (UTC)[reply]
Another example would be a page move - eg. moving a draft to mainspace. Nikkimaria (talk) 13:18, 25 August 2021 (UTC)[reply]
I've never moved a page but, looking at the graphic shown at Help:How_to_move_a_page, it appears the process involves a pre-written edit summary that the moving editor can modify. Since a move preserves the edit history, it is not clear to me how the move would be "obvious" if the edit summary were blank. Please elaborate. Butwhatdoiknow (talk) 16:34, 25 August 2021 (UTC)[reply]
A user can optionally add an edit summary for a move, but whether they do or do not, there is a system message that makes clear what the move was. This, for example, is a page move made without an edit summary: you can clearly understand that the user moved that page from draft to mainspace, without them having to use an edit summary to explain. Nikkimaria (talk) 20:36, 25 August 2021 (UTC)[reply]
The diff you link to has an edit summary ("EnPassant moved page Draft:Lions Lighthouse to Lions Lighthouse"). I'm still looking for a second example of a blank edit summary that conveys any information at all other than an edit took place. Or are you suggesting we change "(unless the reason for them is obvious)" to "(unless the system automatically generates an edit summary or the reason for them is obvious)"? Butwhatdoiknow (talk) 00:55, 26 August 2021 (UTC)[reply]
That's not an edit summary, it's a separate system message. There is, as I noted, a separate edit summary field, which in this case the editor did not fill in. And I'm not suggesting we change anything; I don't feel the present wording needs changing. Nikkimaria (talk) 01:43, 26 August 2021 (UTC)[reply]
According to Help:Edit summary, "An edit summary is a brief explanation of an edit to a Wikipedia page." It does not add "generated by a human editor." Does it say somewhere else that an explanatory message generated by the system and appearing in the Revision history edit summary parentheses is not an edit summary? Butwhatdoiknow (talk) 03:28, 26 August 2021 (UTC)[reply]
This page details what are considered automatic edit summaries; that is not among them. That being said, I think continuing to argue about this is not productive and not really the point of this discussion. Nikkimaria (talk) 12:50, 26 August 2021 (UTC)[reply]

Whether this discussion is "productive"

We are discussing whether to retain the "(unless the reason for them is obvious)" exception to the "explain edits" text. I have posited that no edits are obvious. You have responded: "I can easily think of edits that are genuinely obvious without any explanation" (00:28, 18 August 2021 (emphasis added)). So, yes, let's step away from "edit summary" and focus on "edit explanation." - Butwhatdoiknow (talk) 17:33, 26 August 2021 (UTC)[reply]
I have asked for examples of obvious edits and, with the exception of page creation, your proposed examples have not involved edits without explanations. Instead, they have involved edits with system generated explanations. Would you please provide another example of an edit that is obvious, i.e., self-explained when it appears in a Revision history with no explanatory text between the parentheses? - Butwhatdoiknow (talk) 17:33, 26 August 2021 (UTC) @Nikkimaria: please reply.- Butwhatdoiknow (talk) 15:35, 1 September 2021 (UTC)[reply]

@Nikkimaria: Unless you advise otherwise, I will read your silence as agreement that (a) edits are rarely self-explained ("obvious") to editors scanning page revision histories and (b) "(unless the reason for them is obvious)" is wp:CREEP we can safely remove from wp:EDITCONSENSUS. Butwhatdoiknow (talk) 15:12, 15 September 2021 (UTC)[reply]
Not agreeing - as I said, I don't think continuing this line of discussion is productive. Nikkimaria (talk) 01:14, 16 September 2021 (UTC)[reply]
@Nikkimaria: What you said (12:50, 26 August 2021 (UTC)) was "continuing to argue about [what constitutes an "edit summary"] is not productive and not really the point of this discussion." I agreed (17:33, 26 August 2021, first post) and returned to the point of the discussion (17:33, 26 August 2021, second post): your assertion that the text should remain because there are many "edits that are genuinely obvious without any explanation" (00:28, 18 August 2021 (emphasis added)).[reply]
Resolving that assertion is productive because it helps you and me reach a consensus regarding whether the disputed text is CREEP. Accordingly, I renew my request that you "provide another example of an edit that is obvious, i.e., self-explained when it appears in a Revision history with no explanatory text between the parentheses." Butwhatdoiknow (talk) 06:38, 16 September 2021 (UTC)[reply]
At this point I don't think we've even reached a consensus on what we're discussing. See the exchange of 19-20 August. Nikkimaria (talk) 00:15, 17 September 2021 (UTC)[reply]
In the August 19-20 period I offered (16:18, 19 August) three statements. The first one was "Blank edit summaries have no substance. An editor would have to check the diff to know what the original editor did. In short, the substance of the edit is not obvious." You replied (03:00, 20 August) to this point by saying "not universally true - if I create a page it is clear even with no edit summary what I am doing for someone glancing at the edit history." I, in turn, replied (05:22, 20 August) "Yes, I see why a page creation 'edit' needs no explanation. But is that the only exception? I'm looking for a couple more examples of edits that are obvious."
I think we are discussing whether there are any other examples of edits with no explanation that convey meaningful information to an editor reviewing a revision history. What do you think we are discussing? Butwhatdoiknow (talk) 07:34, 17 September 2021 (UTC)[reply]
At the moment we're down a rabbit hole about what "explanation" means, and I'm not sure why we're focusing on that; the text that you propose removing regards "reason", not "substance". Nikkimaria (talk) 02:49, 18 September 2021 (UTC)[reply]
Note: The words I propose to remove appear in context as follows (emphasis added): "All edits should be explained (unless the reason for them is obvious)." Butwhatdoiknow (talk) 05:38, 18 September 2021 (UTC)[reply]
What do you propose we focus on to help us decide whether to keep or remove the parenthetical? Butwhatdoiknow (talk) 05:38, 18 September 2021 (UTC)[reply]
See Jc37's post below. Nikkimaria (talk) 13:11, 18 September 2021 (UTC)[reply]

Other discussion

  • With the change to “encourages”, I think it clear that we are NOT mandating edit summaries. And since we are not mandating them, I think stating an “exception” might be confusing. After all, it’s not like we discourage summaries when the reason for the edit is obvious. Blueboar (talk) 13:41, 17 August 2021 (UTC)[reply]

Huh?

Ok, I just read through all of the discussion above, but it seems to me - and maybe I missed it in someone's comments - but it seems to me that the parenthetical is being taken out of context of the whole sentence and section.

The meaning (to me at least) is clear: If you are being WP:BOLD, then the explanation should appear in the edit summary, if as a result of a discussion, then explain there. Remember, we're talking about how this refers to wp:con here, not just any old edit summary.

And with that in mind, the text is saying that if the average editor looks at an edit and the reason is obvious, then explanation isn't required, but if it isn't obvious then an explanation (either through edit summary of discussion) is then required.

I don't understand why the parenthetical should need to be removed, since it merely documents current practice, as far as I know.

So what am I missing here? - jc37 11:59, 18 September 2021 (UTC)[reply]

What you are missing is the question: What happens when there is disagreement as to whether the reason is “obvious”?
After all, what is obvious to one editor may not be obvious to another.
My answer to that question is to “Assume Good Faith”. If someone makes an edit without leaving a summary, we have to assume that this editor was acting in good faith, and thought that the edit was “obvious” enough to not need a summary (even if the reason wasn’t obvious to YOU). If we start with an assumption of good faith, it then becomes the responsibility of those who don’t find the reason obvious to ask for clarification. Blueboar (talk) 13:14, 18 September 2021 (UTC)[reply]
Oh ok, ty. And sounds like what we do here. (or in other words, would seem to be the current process.) - jc37 13:54, 18 September 2021 (UTC)[reply]
Well, we are talking about writing, so I think another underlying assumption is that for the vast (super)majority of edits the "why" is obvious (that seems to be the actuality for, 'it stays, until someone reverts or challenges'), or the why can readily be intuited by anyone with a modicum of intelligence, even when one disagrees. Wikipedia does not suspend the role of written communication: 'make sense'. In the minority edits, if others can't tell why, then the edit will likely be discarded, which means the proponent of the edit is the party that 'losses' if others are confused by the edit, and thus the disadvantage already falls on the proponent, and also means we do not need any bureaucratic/mandatory rule for summaries. In short, 'you want your edit to stick', it's already functionally all on you, in the edit and in summary: be advised, don't confuse others, make sense. -- Alanscottwalker (talk) 14:30, 18 September 2021 (UTC)[reply]
We should also encourage information in all edit summaries because they "help other editors by (a) saving the time to open up the edit to find out what it's all about, ... and (c) providing information about the edit on diff pages and lists of changes (such as page histories and watchlists)" (from Help:Edit summary). — Preceding unsigned comment added by Butwhatdoiknow (talk • contribs) 15:46, 18 September 2021 (UTC)[reply]
@Jc37, I think what you're missing is that there seems to be a difference of opinion in what the "patrolling" editor needs to do to figure out what's going on. Imagine that I've reverted poop vandalism. An obvious kind of edit, right? @Butwhatdoiknow wants to be able to figure out that I was reverting poop vandalism without looking at the diff – while only looking at the edit summary. Therefore, I need to write something like "rvv" in the edit summary, because the long-standing recommendation of "Undid revision whatever by User:Vandal" doesn't explain what happened. Everyone else assumes that editors will look the diff, and therefore editors only need an explanation if the situation is not obvious to someone who has already looked at the diff. WhatamIdoing (talk) 21:29, 19 September 2021 (UTC)[reply]
To actually patrol, you have to look at the article. To edit intelligently at all, you have to look at the article. -- Alanscottwalker (talk) 22:26, 19 September 2021 (UTC)[reply]
Fair enough. But does that mean edit summaries are not useful?) And, if they are useful, shouldn't we encourage them? See generally, Perfect is the enemy of good. Butwhatdoiknow (talk) 05:53, 20 September 2021 (UTC)[reply]
The standard practice is to care about an "explanation" (the why) only if an experienced editor would be confused by the diff. For example, if you are removing a perfectly good sentence from an article, then "Rm duplicate" explains why you are doing so. "Copyedit" would only explain what you are doing. WhatamIdoing (talk) 15:58, 20 September 2021 (UTC)[reply]
Not sure what your point is here. Are you saying that blank edit summaries are okay if an experienced editor would not be confused when looking at a diff? Or, put another way, that there is no problem with a Revision history full of blank edit summaries? Butwhatdoiknow (talk) 21:24, 20 September 2021 (UTC)[reply]
No. I'm saying that there's a difference between "explaining" why you made an edit and "describing" what you did, and that an actual "explanation" is only necessary sometimes. The default undo summary – which is strongly recommended for vandalism – is not "an explanation" of why you undid an edit. It is a factual "description" of what you did. WhatamIdoing (talk) 01:15, 22 September 2021 (UTC)[reply]

Fresh take on "unless the reason for them is obvious"

The discussion above reveals that one of the problems with the current text is the meaning of "explained." Does it mean "describe" or "give reasons"? To resolve that issue I propose changing

All edits should be explained[under discussion] (unless the reason for them is obvious[under discussion])—either by clear edit summaries indicating the reason why the change was made, or by discussion on the associated talk page.

to

All edits should be[under discussion] described. Unless the reason for an edit is obvious, it should also be explained either briefly in the edit summary or by discussion on the associated talk page.

Your thoughts? Butwhatdoiknow (talk) 07:36, 22 September 2021 (UTC)[reply]

"Explained" is correct and I don't see any room for confusion. People who just want to see what changed (i.e. a description) can look at the diff; what is at issue here is the reason for the edit (i.e. an explanation). "I changed 1920 to 1921" is useless, but "1921 is what the source has" is useful. Zerotalk 09:07, 22 September 2021 (UTC)[reply]
(1) While "explained" is perfectly clear to you, the discussion above shows that it isn't perfectly clear to everyone. (2) Edit summaries are not only viewed on watchlists, they are also viewed in revision histories. While editors can check each and every diff, it is very time consuming. (3) Some edit descriptions are self-explanatory ("copy edit," for example) and don't need a separate statement giving the reason.
Or are you saying we should eliminate "(unless the reason for them is obvious)" altogether? Butwhatdoiknow (talk) 20:41, 22 September 2021 (UTC)[reply]
@Zero0000: I look forward to your reply. Butwhatdoiknow (talk) 19:23, 27 September 2021 (UTC)[reply]
@Butwhatdoiknow, do you have WP:NAVPOPS installed? Because of NAVPOPS, I don't think that it is very time-consuming to check each and every diff. WhatamIdoing (talk) 16:44, 7 October 2021 (UTC)[reply]
Never heard of it before. And, I'm guessing, I am not in the minority of editors in that regard. Butwhatdoiknow (talk) 18:29, 7 October 2021 (UTC)[reply]
It is one of the most popular gadgets on wiki. Please go to Special:Preferences#mw-prefsection-gadgets and turn it on for a few days. It's the sixth item in the first section on that page, so it's one of the easier ones to find. WhatamIdoing (talk) 01:10, 8 October 2021 (UTC)[reply]
Once you've got it on, the key point is: go to your watchlist or a page history, and hover over the "prev" button. Assuming you have an average-to-good internet connection, you'll be able to see most diffs without clicking on anything. WhatamIdoing (talk) 01:11, 8 October 2021 (UTC)[reply]
Well, I've tried it and I find it more irritating than helpful. I wonder what "most popular" means in terms of the percentage of Wikipedia editors who use it. And, anyway, why use a work=around when simply leaving a meaningful edit summary solves the problem from the get-go? Butwhatdoiknow (talk) 06:38, 10 October 2021 (UTC)[reply]
Because what you're calling a work-around is automatic and self-explaining, whereas leaving an edit summary, redundant to what is apparent from the content of the edit itself, is adding an unnecessary burden to each contribution. And in what way are popups irritating? EEng 11:59, 10 October 2021 (UTC)[reply]
(1) Navpop is automatic only if you turn it on and hover over edits one-by-one. (2) You and I disagree regarding whether the better practice is (a) editors leaving descriptions as they edit or (b) all subsequent editors having to inspect each and every edit to learn what each one is all about. (3) Navpop is irritating in at least three ways: (a) once you start it the darn popups appear on every page you're looking at (not just Revision history pages), (b) the popups cover up text so you have to move your mouse to read the text underneath, and (c) if the popups are at the bottom of the page you have to scroll down to read the popup. I'll add that, since the popups are only small boxes, they don't "self-explain" large edits. Butwhatdoiknow (talk) 16:38, 10 October 2021 (UTC)[reply]
So turn on navpops and hover over edits (the ones you think need reviewing) one by one; that's what I do and it works great. I haven't had your "darn popups appear on every page" problem; the bottom-of-page problem is rare; for large edits I open the diff. I think your scenario of all subsequent editors having to inspect each and every edit to learn what each one is all about is an exaggeration: on my watchlist, I probably feel like learning the substantive content of the diff for maybe 1 in 10 pages that appear (and for those, I typically use the x diffs since last visit link to see all changes wrapped into one); on page revision histories I usually use the Compare selected revisions button for similar efficiency. I think cluttering up watchlists with comments like "I removed a comma" makes the watchlist harder to use. I'm certainly never going to bother entering such stuff, and you're never going to get consensus for requiring it. EEng 05:47, 11 October 2021 (UTC)[reply]
I am intrigued by your comment that edit summaries with descriptions of changes "makes the watchlist harder to use." Would you please elaborate? Butwhatdoiknow (talk) 15:46, 11 October 2021 (UTC)[reply]
Especially when the edit summary is longer than a few words, the the line wraps. A watchlist in which every entry is on a single line is easier to scan than a watchlist in which some are one, some are two, and a few are three lines. WhatamIdoing (talk) 18:32, 11 October 2021 (UTC)[reply]
Well, the example given was "I removed a comma." I'm still at a loss regarding how Eng finds that makes a watchlist "harder to use." (And, in any case, what are you doing looking at edit summaries? I thought you were a Navpop user.) Butwhatdoiknow (talk) 02:31, 12 October 2021 (UTC)[reply]
Because it clutters the watchlist with trivia. If you're just removing a comma, ce or punct will do as an edit summary; I don't need to read I ... removed ... a ... comma. When you do something that needs explanation meta to the edit itself, then use the e.s.: "After consulting Smith, Jones, Birch, and Calton, it's clear that the figure in Calton is a typo -- S, J, and B all agree on the correct figure, and S in particular points out that Calton corrected the figure in his British edition." That's what edit summaries are for. EEng 03:17, 12 October 2021 (UTC)[reply]
I'm with you 100% - brevity is best. To my mind, "ce" is a perfectly valid description. And, when an editor gives that description, the reason for the edit is obvious (the editor believes it improves readability) and the edit needs no "meta" explanation. An editor who sees that summary and is curious can open or "navpop" it to see the change.
The problem with the current text is that says that a blank edit summary is okay if - after opening the edit - the "meta" explanation is clear. How can we improve
All edits should be[under discussion] described. Unless the reason for an edit is obvious, it should also be explained either briefly in the edit summary or by discussion on the associated talk page.
to say that you should give a description (even if it is only "ce" or "punct") but don't need to give a meta explanation if the description makes the meta explanation obvious? Butwhatdoiknow (talk) 06:36, 12 October 2021 (UTC)[reply]
Honestly I've lost track of what we're arguing about, but I do think every edit should have a least a faint indication of what it "is": ce or punct (as above) or + or cut detail or oops. It's when someone says the edit has to be "explained" or "described" that I get nervous. EEng 10:58, 12 October 2021 (UTC)[reply]
Yep, that's the problem we're working on: how to characterize the minimum requirement for an edit summary. "All edits should be ..." what? Currently it says "explained," which you and I agree doesn't hit the mark. You don't like "described." Maybe "summarized"? "Identified"? Or perhaps we just need to add "briefly" in front of one of these words. Butwhatdoiknow (talk) 17:37, 12 October 2021 (UTC)[reply]
I'm at a loss as to how to characterize it. Obviously if there's something about the edit that isn't apparent from the diff itself -- something other editors need to know to understand why you've done what you've done -- then you say that (as in my Smith, Jones, Birch, Calton example above). If that's not the case, then when I'm about to save I think of someone who, for some reason, is untangling a complicated list of diffs on the rev history page -- what can I do to help them sort out who did what? "OK, let's see again ... right, the edit by JoeEditor that says 'drop detail', that's where he cut that irrelevant stuff, and then the edit EEng marked as 'typo' is where he fixed that spelling error, then right after that UnionJack made that edit he marked 'BrEng' because actually EEng had mistakenly changed the British spelling to American spelling ..." So for most edits the edit summary really doesn't have to be much at all, at least given the way I see their function. But again, I don't know to express that. It's possible that all we need to say is that most edits should have something in the edit summary field, and beyond that we let water find its own level.
BTW, I'm talking about pages other than discussion pages here; on discussion pages, my e.s. is almost always simply "+" because, well, I posted something. EEng 13:50, 13 October 2021 (UTC)[reply]
I prefer edit summaries that say something (i.e., something relevant). But I still wonder: Do we need to say that on this page? Maybe it's enough to have that information in the WP:Editing policy and similar pages. WhatamIdoing (talk) 14:23, 13 October 2021
As I've said (below): Feel free to propose that in a separate discussion. Butwhatdoiknow (talk) 19:31, 13 October 2021 (UTC)[reply]
EEng: How about a "best of the worst"? Which term is least objectionable to you: "briefly explained," "briefly described," "briefly summarized," or "briefly identified" (or something else)? Butwhatdoiknow (talk) 20:24, 13 October 2021 (UTC)[reply]
I'm sorry, I just don't know. I only wandered in here looking for that bathroom. But here's another thought: the context here is how edit summaries are involved in consensus-building that takes place directly in the context of editing. It's not about the humdrum everyday edits that aren't involved in any sort of issue. So, to the extent this guideline makes recommendations about edit summaries, maybe instead of All edits should be explained (unless the reason for them is obvious it should say If you think your edit might be questioned, it should be carefully explained by} ... no, that won't work. I give up. EEng 23:09, 13 October 2021 (UTC)[reply]
Hate to see you go. Just one more thing: Is it fair to say that you don't see much of a difference between "briefly explain" and "briefly describe"? And, if so, that you have no objection changing:
All edits should be explained[under discussion] (unless the reason for them is obvious[under discussion])—either by clear edit summaries indicating the reason why the change was made, or by discussion on the associated talk page.
to
All edits should be[under discussion] described. Unless the reason for an edit is obvious, it should also be explained either briefly in the edit summary or by discussion on the associated talk page. Butwhatdoiknow (talk) 06:00, 14 October 2021 (UTC)[reply]
I'm sorry to let you down, but I just can't say. EEng 14:02, 14 October 2021 (UTC)[reply]
This Navpop discussion seems to go to whether we should encourage "explanations" for all edits. That is the subject of the Wikipedia_talk:Consensus#Something_better_than_"all_..._should"? section, below.
The question up for discussion in this section is whether - until and unless we do away with the current sentence altogether - to (a) clarify that "explanation" includes both "describe" and "give reasons" and (b) clarify when a description is sufficient and when reasons should also be given. I welcome your thoughts on that issue. Butwhatdoiknow (talk) 16:38, 10 October 2021 (UTC)[reply]
It doesn't feel to me like anyone else shares your concerns/confusion over this sentence. Unless I've missed something, maybe we don't need to change anything. It can be CREEPY to change a policy when only one editor needs a more substantial explanation. WhatamIdoing (talk) 15:30, 11 October 2021 (UTC)[reply]
How about we just remove the sentence completely, since how and when to use an edit summary or to pre-emptively post an explanation on the talk page isn't really a core point for the policy about consensus? WhatamIdoing (talk) 16:45, 7 October 2021 (UTC)[reply]
Feel free to propose that in a separate discussion. (That said, I suggest the sentence is appropriate because it speaks to a procedure that enhances consensus building.) Butwhatdoiknow (talk) 18:29, 7 October 2021 (UTC)[reply]

WhatamIdoing issues

@user:WhatamIdoing I acknowledge your suggestion that we remove the “All edits should be explained …” sentence altogether. That is a discussion worth having. But it is tangential to what the sentence should say until and unless we remove it. Butwhatdoiknow (talk) 01:56, 25 October 2021 (UTC)[reply]

To recap, I propose changing

All edits should be explained[under discussion] (unless the reason for them is obvious[under discussion])—either by clear edit summaries indicating the reason why the change was made, or by discussion on the associated talk page.

to

All edits should be[under discussion] briefly described. Unless the reason for an edit is obvious, it should also be explained in the edit summary or by discussion on the associated talk page.

First issue. @user:WhatamIdoing wonders (15:30, 11 October 2021) whether I am the only person who finds “explained” to be unclear. As evidence that I am not the only one, see (1) discussion (beginning at 02:31, 3 August 2021) regarding whether “re” and “ce” are explanations, (2) discussion (beginning at 03:17, 12 October 2021) regarding whether “ce” and “punct” sufficiently supports an edit, and (3) the distinction WhatamIdoing makes (at 01:15, 22 September 2021) that “rvv” is a description (not an explanation). Does that resolve the first issue? Butwhatdoiknow (talk) 01:56, 25 October 2021 (UTC)[reply]

No. WhatamIdoing (talk) 22:49, 26 October 2021 (UTC)[reply]
What more evidence would you want that I am not the only one who finds the requirement that edits be "explained" in the current text confusing? Butwhatdoiknow (talk) 23:37, 26 October 2021 (UTC)[reply]
Perhaps you could try to find any two other editors who are willing to explicitly and directly say that they can't figure out how to write an edit summary as a result of this sentence? WhatamIdoing (talk) 19:58, 28 October 2021 (UTC)[reply]
The text in question says "All edits should be explained (unless the reason for them is obvious)." Doesn't that speak to when - rather than how - to write an edit summary? Butwhatdoiknow (talk) 16:01, 29 October 2021 (UTC)[reply]
There is only one time to write an edit summary, which is when you make an edit.
Did you perhaps mean to ask "Doesn't that speak to whether - rather than how - to write an edit summary?" WhatamIdoing (talk) 17:09, 29 October 2021 (UTC)[reply]

Off topic discussion of usernames

For some reason, I didn't know Butwhatdoiknow and WhatamIdoing were different accounts. The similar account names are a strange coincidence. – The Grid (talk) 20:03, 28 October 2021 (UTC)[reply]

  • I can see how the brain would rearrange a few letters like that.
Back in the day, you could create accounts that differed only in their capitalization. Thus we have Mastcell and MastCell, for example. I see that there is also one named Thegrid (an editor at the German-language Wikipedia). WhatamIdoing (talk) 22:04, 28 October 2021 (UTC)[reply]

Whether to require explanations at all

I agree that best practice is to explain every edit as you make it, but it isn’t what most editors actually DO. Most edits are NOT explained until/unless they are challenged.
Policy is supposed to reflect actual practice, and we should not write “rules” that are not (and likely will not be) followed.
I have no problem with encouraging editors to leave explanations… we can note that it is “best practice” to do so… but we should not say “you must leave an explanation” when such a rule will simply be ignored. Blueboar (talk) 13:33, 3 August 2021 (UTC)[reply]

The fact that some editors violate the current rule does not mean that we should abandon it. That reasoning leads us to no rules whatsoever. Butwhatdoiknow (talk) 15:25, 3 August 2021 (UTC)[reply]
Except - the “rule” is: Wikipedia consensus usually occurs implicitly. An edit has presumed consensus unless it is disputed or reverted. Should another editor revise that edit then the new edit will have presumed consensus unless it meets with disagreement. My take is that there is no NEED to explain an edit that is presumed to have consensus - UNLESS it meets with disagreement. In other words, the need for explanations kick in when/IF there is disagreement. There is no requirement for an explanation when/IF there is no disagreement. Blueboar (talk) 17:04, 3 August 2021 (UTC)[reply]
If (a) there is no need to explain an edit that has presumed consensus and (b) all edits are initially presumed to have consensus then (c) there is no need to explain any edit. The problem with this sentence is, I think, that the need to explain an edit (that is, to tell the community what you have done) is not the same as an edit having presumed consensus. Butwhatdoiknow (talk) 18:40, 3 August 2021 (UTC)[reply]
Not quite… There is no need to explain any edit UNTIL (d) an edit is disputed. This is because once an edit is disputed, we reach the point where we can no longer presume consensus. Thus, once there IS a dispute, people DO need to give explanations. Blueboar (talk) 19:34, 3 August 2021 (UTC)[reply]
Explain can mean either "tell" (describe) or "support" (provide reasons). For the moment, let's accept your position that an editor need not support an edit until it is disputed. What about telling editors who scan Watchlists or Revision histories what you have done ("ce," for example)? Any reason to make that optional? Butwhatdoiknow (talk) 22:35, 3 August 2021 (UTC)[reply]
There is even less need for a “what” summary than there is for “why” summary. If you want to see what an edit was, you should look at the article edit history and actually see what was added or subtracted from the article. Don’t trust the edit summary… seriously, think how many idiots try to hide vandalism behind a faked edit summary. We catch it because we DON’T trust edit summaries. Blueboar (talk) 23:53, 3 August 2021 (UTC)[reply]
This seems to repeat the "don't make a rule because some folks will abuse it" reasoning that leads to no rules whatsoever. That said, most articles have just a few editors and they get to know each other. In such cases they do trust each other's work (including edit summaries). Further, edit summaries are not just for real-time monitoring, they are also used when folks look back at a history to see when a change was made. A list of empty edit summaries makes the editor "go fish" for the change. Butwhatdoiknow (talk) 20:33, 5 August 2021 (UTC)[reply]
All that said, I concede below that there is no consensus for language explicitly requiring edit summaries. Butwhatdoiknow (talk) 20:33, 5 August 2021 (UTC)[reply]
Any kind of wording in a policy page suggesting that editors must give edit summaries has been constantly rejected at the village pump, as far as I know. The closest to it is WP:UNRESPONSIVE. This change would be duplicating another page anyway, and this isn't the page where that should belong, but in any case it purports to set a requirement that consensus does not support. ProcrastinatingReader (talk) 20:39, 3 August 2021 (UTC)[reply]
WP:UNRESPONSIVE seem pretty definite to me. No qualifying words such as "best practice" or "ideally." Perhaps it would help if you provided links to a couple of the village pump discussions so we can see the thoughts expressed there. Butwhatdoiknow (talk) 22:24, 3 August 2021 (UTC)[reply]
This was the most recent. At a skim there's also [1][2] ProcrastinatingReader (talk) 22:41, 3 August 2021 (UTC)[reply]
Well, I'm not sure that I'd characterize those discussions as "rejecting" mandatory edit summaries. Rather, they seem to reflect a lack of consensus to either require them or not. Which brings me to the section I'm starting below. Butwhatdoiknow (talk) 20:33, 5 August 2021 (UTC)[reply]
  • Not seeing anything in those discussions that would remotely approach the necessary consensus to change the existing wording. If you think there's consensus to change it, I'd suggest an RFC. --Aquillion (talk) 06:23, 20 August 2021 (UTC)[reply]

Something better than "all ... should"?

Above, ProcrastinatingReader points to several discussions that reach no consensus regarding whether edit summaries should be mandatory. "Should" in the second paragraph of wp:EDITCONSENSUS seems to reflect this lack of agreement inasmuch as it has two meanings: ((a) "must" and (b) "ought to"). So one option is to retain the current wording ("All edits should be explained") as is.
Blueboar proposes changing the current text and has offered "Ideally, all edits should be explained" and "Best practice is to explain your edits" as alternatives. While I prefer the current wording, if we do decide to change the text I propose "The Wikipedia community strongly encourages meaningful explanations for all edits." Butwhatdoiknow (talk) 20:34, 5 August 2021 (UTC)[reply]

I could accept this. While there is not a consensus to mandate explanations, I think we do (and should) encourage clear communication (be it in an edit summary or via talk page threads). Blueboar (talk) 21:49, 5 August 2021 (UTC)[reply]
I oppose it if it includes talk pages. For example, it's quite common where an edit summary would be a bad, missing context, misleading or ineffective summary of the talk page post and I deliberately leave only a short non-explanatory edit summary as I will for this post. :-) Also, Wikipedians have limited volunteer time and trying to write an edit summary that avoids the above issues can take more time than the edit itself. North8000 (talk) 12:52, 19 August 2021 (UTC)[reply]
To clarify, that strong urging should be for edits in article space and for edits in places other than talk pages. North8000 (talk) 17:18, 19 August 2021 (UTC)[reply]
  • Strenuously disagree with weakening the language here. "Several discussions have reached no consensus" is the opposite of a a reason to change the wording of a longstanding policy; we need to reach a clear consensus on something else to replace it (ie. not just a discussion between a handful of editors, given that this is known to be a long-standing dispute that has attracted much more discussion), otherwise the old version ought to stand. Beyond that I'm utterly opposed to any changes that would weaken its wording - my comments above mostly touch on my rationale. Edit summaries are an essential part of collaborative editing, and while we don't police every single edit, an editor who never used edit summaries, ever, at all, or whose edit summaries were consistently grossly insufficient, can absolutely face sanctions for this - it is comparable to eg. WP:MINOR in that respect, where editors are given substantial leeway and occasional lapses or gray areas are fine, but the basic requirements do still apply and are not at all mere suggestions or "encouragements." --Aquillion (talk) 06:11, 20 August 2021 (UTC)[reply]
Can you give us some examples of editors who actually were sanctioned for not leaving edit summaries? I have never seen it done. Blueboar (talk) 12:04, 20 August 2021 (UTC)[reply]
@Blueboar, I saw a dispute along those lines a year or two ago. I think the conclusion was everyone agreeing that edit summaries are Good Things. It's more common to see complaints about misleading edit summaries, such as adding disputed content with an edit summary of "copyedit". WhatamIdoing (talk) 21:33, 19 September 2021 (UTC)[reply]
Unfortunately, nothing we are discussing will stop misleading edit summaries. And I agree that (except for misleading edit summaries) edit summaries are “good things”. I just disagree that we need to mandate “good things”. Blueboar (talk) 22:15, 19 September 2021 (UTC)[reply]
I agree with you. WhatamIdoing (talk) 16:01, 20 September 2021 (UTC)[reply]
Also agree.....not possible....vast majority of edit do not have a summary. Would be setting our selves up for mass conflict with litte benefits for our readers.Moxy- 13:35, 11 October 2021 (UTC)[reply]

Note

What is it like to ask for a consensus? I want to change a table. --IoATi (talk) 03:19, 3 September 2021 (UTC)[reply]

You have two initial pathways to choose from. The more common is to to follow Bold, revert, discuss. Just make the change you want to make. If it is reverted, don't revert back but instead open a new section on the article talk page requesting discussion for the change. Discuss and see if consensus can be reached. If necessary to come to consensus, try dispute resolution. The other possible path is to just start with the talk page discussion without making the change and waiting to see if it will be reverted. Regards, TransporterMan (TALK) 16:47, 3 September 2021 (UTC)[reply]

Volunteers

I've been thinking about Wikipedia:WikiProject Council/Guide/Task forces#Converting existing projects to task forces. I think we have some of the language wrong there, but there's a fundamental question that I don't think is addressed outside of Wikipedia:Wikipedia is a volunteer service: We can't form a consensus to force you to take any action. We can require you to stop ("stop it, or I'll block you"), and we can be coercive ("do it, or I'll block you"), but we can't actually make you do anything.

In the case of merging groups of people, we require a discussion with plenty of opportunity for people to decide whether they want to join a different group. (These discussions are usually boring, but they are sometimes emotionally fraught.) If someone doesn't want to join the other group, then we really can't make them.

In the case of a content decision, if I strongly disagree with the consensus – maybe everyone agrees to cite a source that I think is garbage, or to emphasize a viewpoint that I think is unimportant – then you can expect me to abide by the consensus (e.g., I won't revert you for IMO damaging the article by adding that garbage source), but you can't expect me to implement the consensus myself (e.g., I won't personally damage the article by adding that garbage source myself).

In the case of a behavioral decision, if an admin disagrees with a decision, then that admin is expected to not take that action. If the AFD closes as delete, but you strongly think it should be kept, then you let some other admin delete it. If RFA ends with a vote to promote, but you think the candidate will be a lousy admin, then you let some other bureaucrat handle it.

I don't know what you call this idea in the real world. First do no harm? The opposite of running to do evil? Freedom of conscience? Keeping your hands clean? Being mature enough not to jump off the cliff just because everyone else voted to do so? But I think that it might be nice if one of the policies pointed out that you can't vote to make someone else do something. Consensus means that the group agreed that I can do it myself, not that I can make you do it for me. WhatamIdoing (talk) 16:38, 20 September 2021 (UTC)[reply]

Retrieved from "https://en.wikipedia.org/w/index.php?title=Wikipedia_talk:Consensus&oldid=1052523636"