Wikipedia talk:Consensus/Archive 18

Archive 15

Archive 16

Archive 17

Archive 18

Archive 19

Archive 20

→

Archive 24

Not all "local" consensuses are equal

I note that we have recently had a lot of debates about whether various WikiProject specific guidelines conflict with more generalized guidelines (especially the MOS). I also note that in almost all of these debates, sooner or later someone points to WP:LOCALCONSENSUS to make the argument that the generalized guideline should trump the WikiProject guideline. The assumption being that the generalized guideline reflects a broader consensus than the "local" WikiProject's consensus.

I don't want to re-argue those specific debate... my question is whether WP:LOCALCONSENSUS really applies to every WikiProject guideline? The problem, of course, is that not all WikiProjects are equal. Some are small - with only a few active editors involved... and for these I completely agree that LOCAL applies. Others, however, are very large - with hundreds of active editors... and I am not so sure that it is appropriate to call a consensus reached by one of these larger WikiProjects "LOCAL". Indeed, it is possible that a consensus at a WikiProject's guideline may actually reflect a broader consensus than one reached at a "generalized" guideline.

I don't have a firm answer to this dilemma ... but I do think we need some discussion about it. The recent debates tell me that we need to reach a clearer consensus on what does and does not qualify as a LOCAL consensus. Blueboar (talk) 12:42, 6 April 2014 (UTC)

Guidelines are advisory, so it seems editors are free to ignore them. Even policies might be ignored if a local consensus develops for that. In fact, that is the mechanism or process by which policy changes. When something mistaken is corrected, it begins with one editor mentioning it. --Ring Cinema (talk) 18:13, 6 April 2014 (UTC)

Nope -- certain polices can not be overridden by "local consensus" including WP:NPOV, WP:RS and WP:BLP, for example. Policies can be changed through consensus on the policy pages and we can not presume to rewrite policies on any other page or article. If one disagrees with a policy, the mechanism is to propose changes at the policy, and see if one can persuade others of the gains to be made by such changes. Collect (talk) 18:22, 6 April 2014 (UTC)

Have a look at User:Andrewa/consensus is consensus for some other thoughts on this. Comments welcome there or here. Andrewa (talk) 19:11, 6 April 2014 (UTC)

The problem lies with equating policies and guidelines. It would be accurate if it said " ... participants in a WikiProject cannot decide that some generally accepted policy ... does not apply to articles within its scope. Wikipedia has a higher standard of participation and consensus for changes to policies ..." But the inclusion of "and guidelines" makes it inaccurate.

Guidelines are advisory and often ignored, some more than others, and regularly changed without much discussion. The GA criteria, for example, don't include compliance with most of the MoS. That's an example of a group of editors deciding that a certain guideline need not apply to certain articles, and everyone is fine with that.

This page used to say "In the case of policy pages a higher standard of participation and consensus is expected." This was changed in 2008 to "[i]n the case of policy and process pages," then later someone added a link to Wikipedia:Policies and guidelines, and later still someone changed the sentence to "policies and guidelines." SlimVirgin ^(talk) 19:14, 6 April 2014 (UTC)

Thanks, SlimVirgin, for making that distinction. Perhaps the page has moved backward over time. And I have to point out that, although there is an understandable belief that certain policies cannot be trumped by local consensus, I think we are obligated by the facts to accept that editors who agree to ignore policies are free to do so because who is there to stop them? It's just one of those things that isn't even a problem. Presumably, they would have good reasons for doing so, even if we can't exactly conjure them in the abstract. We don't quite express that disputes are settled by policy but policy is silent when there is not a dispute. --Ring Cinema (talk) 20:35, 6 April 2014 (UTC)

In that WP:IAR is also a policy, you could even argue that local consensus (which I'd prefer to just call consensus) trumps so-called community consensus. Consensus is consensus, see also User:Andrewa/creed#consensus.

The problem mainly arises when consensus is claimed to have been wrongly assessed, as in the current discussion at Wikipedia:Move review/Log/2014 March#Black crowned crane. But there are always avenues of appeal, and these need to be respected and followed.

No consensus decision can be ignored simply because it's claimed to represent local consensus. But when the term is used, that's nearly always exactly what the person using the term is wanting to do. That's why we have problems, in my opinion, and why I'd like the term local consensus dropped from the policy. Andrewa (talk) 00:23, 7 April 2014 (UTC)

Iterating: Non-negotiable policies which are "ignored" by editors are still policy and can be enforced by any editor at all who actually understands why the policies exist. If we allow any group to make up their own rules contrary to core policy, then the project is Wertlos utterly. Cheers. Collect (talk) 03:13, 7 April 2014 (UTC)

Agree. And nor can we allow any group to impose their particular interpretation of a policy or even of a guideline. That's what is tricky about working by consensus, and not having an editorial board, but rather letting the contributors write and even authorise and approve and even audit and ensure compliance with the MOS themselves. And that's a key difference of our MOS to those of other encyclopedias.

It's a plus and a minus, but as you say, it's not negotiable. It's a core policy. Andrewa (talk) 03:41, 7 April 2014 (UTC)

I don't think that people should be pointing to the broad LOCALCONSENSUS policy about that. They should be pointing to the very direct and specific WP:WikiProject advice pages guideline, which tells WikiProjects (==groups of editors) that their advice is no better or worse than any other small group of editors' advice, and if they want a proper guideline, then they need to make a WP:PROPOSAL just like everyone else.

There is nothing magic about calling yourselves "a WikiProject". There's nothing magic (or even exclusive) about declaring some articles to be "within your scope". WikiProject participants are just like anyone else when it comes to writing guidelines.

Obviously, there's a substantial backlog, which involves good advice not being adopted by the whole community (when it should be) and questionable advice having been self-declared a "guideline" by self-appointed advice givers years ago (especially in the MOS subpages). But the "rules" are clear now: write whatever essays you want, but if you want the entire community to feel bound by it, then you have to get the entire community to adopt your advice—and when they do, that advice quits belonging to your WikiProject, and starting belonging to the whole community. (See the history of WP:MEDRS and WP:MEDMOS for two pages that were done correctly.) WhatamIdoing (talk) 03:51, 7 April 2014 (UTC)

It's not just Wikiprojects. WP:DRN has a core of about half a dozen volunteers, and by design we have no power over anyone other than the power of persuasion. What if we decided to give ourselves some power and to delete any comment that we felt was "not helpful"? Would our local consensus be allowed to overrule WP:TPOC? --Guy Macon (talk) 04:22, 7 April 2014 (UTC)

Agree. The situation here is that neither side has a clear consensus, although one side is claiming it, and things that should have been done to produce one, in hindsight, have been neglected. And that's understandable... we come here basically to work on articles, not policies and guidelines! They are in that sense a necessary evil.

Of course I'm so far on the side of WikiProject Birds so far. This is partly because there are positive arguments at several levels to back them up, including things that others (who don't seem to have my background in relatively recent linguistics) don't seem to have even noticed and which I can perhaps contribute, but I confess (disclose) that it's now also partly because of the nature of the campaign against them which has now turned on me too. Not everyone on the "other" side has participated in this of course, and none of it has quite passed the boundaries where a behavioral issue should be raised, but it's often been close to the line... false accusations, repeatedly aggressive language, long posts of meaningless or at least poorly phrased argument and unchecked links, assumption of bad faith, gloating over my supposed blunders, unwanted reindenting of my arguments making them in my opinion harder to read... the list goes on. And in amongst all of that I notice the accusation that WikiProject Birds has been shouting down their opposition, when it appears to me to be more the other way around. It may just be that I've only seen one side of this, but see my off-wiki essay for another possibility of what is going on there.

I don't want to raise it now as a behavioural issue through dispute resolution, it's borderline at best so far IMO and in any case I don't think that would help to bring consensus. Again, the "other side" have threatened this (and the first step of WP:DR of course is to discuss it with me on my talk page, which has been done [1], as I've also done on one matter I mention above [2]). And if "they" do choose that path we just have to follow it. But I think for now it's just a matter of keeping the behavioural aspects in mind. In a bitter and long-running dispute like this it certainly needs to be watched.

We need to build a real consensus one way or another. And I suspect it will not be easy. I could be wrong, I hope I am. Let's give it a go. Andrewa (talk) 14:37, 7 April 2014 (UTC)

Non-negotiable policies which are ignored are of course subject to enforcement, but that still ignores the reality. Absent a dispute, policy is silent. Editors can fill articles with the content of their choosing as long as they agree to do it, and that includes ignoring policies. And, again, that is a good thing. The only way, it seems, to change policy (the mirror of practice) is to change practice. By definition, a new practice that would require a change in policy will violate a policy when it is introduced. So be it. --Ring Cinema (talk) 14:57, 7 April 2014 (UTC)

I demur. A policy violation is a policy violation no matter how long it is unnoticed. A person going 120 mph on I-95 is still speeding and in violation of the law even if he outruns the police car. Saying it is a "good thing" to violate policy is, IMO, a "bad thing." Collect (talk) 15:21, 7 April 2014 (UTC)

Not convinced it's a good analogy... not quite sure why not, though. Andrewa (talk) 15:57, 7 April 2014 (UTC)

I think that's because traffic laws don't have Ignore all rules as policy, but we do. When one policy is ignored in order to improve the encyclopedia, IAR makes it compliant with policy; anyone who wants to "punish" that "policy breach" needs to argue how following the letter of that policy would be better than the suggested improvement. Diego (talk) 10:17, 14 April 2014 (UTC)

Deep but well put. See WP:creed#rules. But I'm not sure where it leads us... Andrewa (talk) 15:57, 7 April 2014 (UTC)

This is a road where the speed limit is set by the speed of the drivers. So the superficial, somewhat grasping point is that "a violation has occurred". The deeper reality is that practices may change and policy reflects practice. No, it is not consonant with an authoritarian model, but so what? I understand that some might prefer that this page decides style for others. Sometimes it doesn't. Nothing lost. --Ring Cinema (talk) 18:51, 7 April 2014 (UTC)

The core policies have been vetted by literally hundreds of "drivers" and you would allow a single driver to assert "the speed limit signs that hundreds of have editors agreed on formally as being 'non-negotiable' do not apply to me because I do not like them"? Sorry -- I do not buy your version of the speed laws and your desire to make Wikipedia "non-negotiable policies" into an "ignore if you don't want to follow them" category. Collect (talk) 19:41, 7 April 2014 (UTC)

Vet it up! Basically, editors want to follow policies and they do their best to do so. That's the reason that compliance is widespread. However, it cannot be contradicted by experience that, until someone objects, policies are moot. No one can opt out of that reality. If editors want to ignore policies, they're free until another editor wants to follow the policy. This isn't taking a point of view, making a judgement, assigning a value, suggesting a preference, making a wish or offering a theory. It's an accurate description. The good thing about it is that sometimes policies are bad and editors don't follow them for good reason. That's how things get solved. --Ring Cinema (talk) 20:06, 7 April 2014 (UTC)

One distinction is that if an editor were to ignore BLP repeatedly, he would end up being blocked or banned, but I can't imagine anyone being sanctioned for ignoring the MoS.

To return to Blueboar's point, lots of WikiProjects make their own style decisions, based on the specialist sources or the editors' own preferences. Some of these projects have larger numbers of editors engaged than the MoS does, so to call the former "local" consensus in order to undermine it seems wrong-headed. I agree that we don't want local groups deciding to ignore core policy, but when it comes to guidelines I can't see that anyone has the right to tell a WikiProject what to do. SlimVirgin ^(talk) 18:33, 8 April 2014 (UTC)

• The gist of this thread amounts to elevation of wikiprojects to independent authorities on anything they declare themselves authoritative about; it would erect 100+ "special" editorial boards. It would undermine WP self-governance by essentially merging Category:Wikipedia guidelines and Category:Wikipedia essays (or Category:WikiProject Essays).

  A far more productive course of action would be to short-circuit the entire us-vs-them WikiProject system. Move them to something verb-based (WikiWorking Football?, I dunno), so people stop territorially identifying with them, and take other measures to put an end to this factionalism and walled city-state mentality, which has been worsening not improving over time. WikiProject insularity is WP:Esperanza all over again, times 100, with far worse and farther reaching negative effects.

  No one cares about a wikiproject essay coming up with a style "rule" that doesn't grossly violate normal English language usage in ways that bring the project into general disrepute among mainstream observers and distract or anger average readers into mentally rebelling against what they're reading, for no net gain. Our readers (who of course may individually be specialists in some area) grossly outnumber any specialists insisting on some parituclar quirk, usually in conflict with some other specialty's demands, and in conflict with our needs, and in conflict with major style guides on English language writing.

  When something the more vocal members of a project insist on may trigger concerns like these, the burden is solidly on the their shoulders to convince the community that what they're doing is the option that is most helpful to the encyclopedia and the largest percentage of its readers (note this is not the same as "common in that field's specialist publications"). On style matters, this is best done at WT:MOS, because far too few people watchlist its subpages or the NC pages to gauge actual consensus, and it's quite trivial to quietly engage in WP:GAMING and poll-stacking those backwaters to erect WP:False consensus. All of this "MOS isn't really a guideline and doesn't really have consensus" business is untenable. MOS is one of the most-watchlisted projectpages on the whole system. It is our style guide, and it does reflect site-wide consensus. Consensus does not mean "unanimity" nor does it mean "exactly 70% or higher" or any other "gotcha" loophole anyone would want to exploit. If you remove the {{Guideline}} from MOS and take it to WP:MFD you'll be laughed at. If someone adds that same tag to some wikiproject essay that conflicts with MOS, and you remove that and take that page to MFD, you may well succeed. Big difference. [Note that some projects studiously keep their "style guide" essays as part of their main project page...] As with any page, someone can push a pet peeve in MOS for a little while (WP:BOLD is policy, so people can change stuff at any time), but it won't stick if there is no community buy-in over the long haul. It's frankly reckless to suggest undoing this centralization and having wikiprojects make up their own rules in conflict with MOS (or any other major site-side guideline).

  There's also a lot of serious failure going on here to understand that policies are not set "against" guidelines; it is not a "my rule is a policy so your stupid guideline rule can go @#$% itself" pissing match (cf. nonsensical attempts to get WP:AT to "override" WP:MOS on style matters people can't get consensus for at MOS). Policies and guidelines both represent the same level of community consensus; they simply differ in what they address and how. [This gives the lie to attempts here to say "policies are non-negotiable" in the same breath as "guidelines cannot require anything".] Policies address concerns of vital importance to the functioning and survival of the project. Guidelines set out norms that allow the community to operate smoothly.

  Misperceiving the relationship as a simple hierarchy is a fundamental error, like thinking that one's employer has authority over one's church (or vice versa), when they're entirely different kinds of "authority" in one's life. Nerds may like to think of it as something like the difference between HTML and CSS; they're both standards with equal buy-in, they work together, serious errors in either ruin your site, but their purposes and scope are very dissimilar, and HTML (policies) are required to have a site at all, while CSS rules (guidelines) are not, but the site would be unusable without them. Regardless of analogy, WP essays, including "wikiproject style guides " and "wikiproject naming conventions", are something else entirely. They're opinions (usually of one or a very small number of editors, but some like WP:AADD approach broader guideline status) that set out particular reasoned positions on something, and nothing more. The logic in them may or may not outweigh or even be applicable to other concerns addressed by and balanced in a guideline or policy.

  PS: Another obvious flaw with the reasoning that launched this thread is that wikiprojects are not hive minds, and project members often strenuously disagree with pundits in the same project advancing WP:LOCALCONSENSUS "rules" that conflict with site-wide guidelines, because they know it's going to lead to nothing but strife. "WikiProject Foo says...." and "WikiProject Foo wants..." are rank, obvious fallacies, but they underlie this entire suggestion to change how consensus operates with regard to guidelines and project. WP:CONLEVEL is just fine as it is, and correctly recognizes that projects are nothing but pages at which individual editors happen to be collaborating on something. Wikiprojects are inanimate. In an important sense they don't exist. A webpage cannot want, decide, think, feel, demand, expect or prefer anything at all, any more than a stone can. I say all this as the originator of several active projects, BTW.

  Finally, the fact that not all LOCALCONSENSUSes are equal is irrelevant. None of them are equal to or better than site-wide guidelines and policies. Any matter of LOCALCONSENSUS that should be a part of policy or real guidelines is something that an application of patience and reason (not system-gaming and tendentiousness) will soon enough see accepted. But most of you are using "local consensus" wrong. {{em|Actually go read WP:LOCALCONSENSUS. A local consensus in WP terms is not valid; it's something you're supposed to avoid, not seek to establish and crow that you have! "Our project has a local consensus to..." means "Me and some editor friends have decided the rules do not apply to us with regard to...". It makes you look like you believe this is a WP:BATTLEFIELD, and seem that you are WP:NOTHERE to build an encyclopedia but to engage in weird forms of WP:ADVOCACY. — SMcCandlish ☺ ☏ ¢ ≽^ʌⱷ҅_ᴥⱷ^ʌ≼  22:23, 8 April 2014 (UTC)

These somewhat prolix comments don't seem to square with reality or common sense. For one thing, trying to claim any consensus is site-wide is founded in a fallacy and seems to be a bid for authoritarianism. 2) It seems to say that a local consensus isn't valid; that is clearly at odds with the facts, since no consensus is made by editors who are absent. In fact, editors taking exception to policies and not following them (for good faith reasons, to state the obvious) is an essential part of the process of eliminating bad policy. Every dissent begins with a lone dissenter. The wish to impose one's views on other editors is anathema to WP. --Ring Cinema (talk) 01:19, 9 April 2014 (UTC)

Did we perhaps make a mistake when we used the term "Local". That focuses people on location of the consensus, rather than the size of the consensus. I think we all agree that a consensus reached by a small group of editors should bow to a consensus reached by a larger group of editors... but I would say that is true regardless of what page the two consensuses formed on. I am tempted to propose that LOCAL CONSENSUS should really be renamed SMALL CONSENSUS. Blueboar (talk) 21:50, 9 April 2014 (UTC)

Perhaps we should have levels named by the type of article or consensus they apply to: Thus "Community-wide policy and guideline consensus" (level 1) (having broadest likely participation), "Project wide or policy noticeboard level consensus" (level 2) (participation generally those interested in a broad topic or a specific policy, or in discussions regarding a single noticeboard or Wikipedia discussion board), and "single article consensus" (level 3) (smallest usual participation limited primarily to those interested in that single topic), with each properly ceding precedence to the higher level. Collect (talk) 22:29, 9 April 2014 (UTC)

The problem is that you are still assuming that guideline pages always reflect a wider consensus than Project (or even article) consensuses. That may be true in most cases... but it simply isn't true in every case. I have been involved in RFCs on Project pages that brought in hundreds of participants... and discussions on guideline pages that only involved about ten or so participants. Suppose these two pages end up "conflicting"... which would you say reflects the consensus of the wider community? Which has the higher "level" of consensus? In such a situation, I would say it was the consensus at the project page. Blueboar (talk) 22:17, 10 April 2014 (UTC)

Actually I make no such assumption. What I do assume is that the guideline and policy pages are "superior" to individual article pages in that the guideline or policy affects many pages, the article page only affects one. That which has the greatest effect on the project, should also have the greatest weight when people are seeking not to abide by the consensus at the page which applies to many pages. Else we could have a cheerful anarchy where a guideline is set forth, and not a single article abides by it <g>. Therefore, that which is of higher project importance (that is, affecting the greatest number of articles) should be given greater weight. And, as a rule, the main pages are indeed followed by more editors than are most individual articles or projects. Collect (talk) 22:48, 10 April 2014 (UTC)

What about things like the reference desk guidelines? On the one hand, just because they apply only to the local reference desks that doesn't mean that the wider community did not give consensus. The other side of the coins is that perhaps the only editors interested in creating a reference desk guideline might tens to be those who answer questions on the reference desks. Are the instructions ate the top of the WP:DRN page really different than some future "DRN guidelines"? --Guy Macon (talk) 03:33, 10 April 2014 (UTC)

Added such to "level 2" as clearly being less in importance that specific policies or guidelines, but of greater importance than single article discussions. Collect (talk) 16:48, 10 April 2014 (UTC)

I don't think the relevant distinction is large group/small group. This strikes me as an incorrect insistence on some kind of hierarchy, since editors working in small groups on an article might make better decisions than a large number of editors. It makes just as much sense to think the larger group -- abstractly writing necessarily more general principles -- will fail to appreciate the requirements of a specific article. However, I think there is some value in distinguishing matters that are entirely conventional from those that are not. "Go on green" is a convention that could be decided the other way as easily, but it's not something to dispute. Conventional matters don't suffer from top down decision making and that might become part of this project in my view. --Ring Cinema (talk) 14:36, 10 April 2014 (UTC)

You know... this is reminding me a lot of the debates we used to have over whether WP:GNG "trumped" project specific SNGs (after years of debate... the consensus there settled on "no... both should be given equal weight"). Blueboar (talk) 22:17, 10 April 2014 (UTC)

No wonder --- WP:GNG specifically states: A topic is also presumed notable if it meets the criteria outlined in a subject-specific guideline listed in the box on the right. The conflict appears to have been quite resolved in that case. Collect (talk) 22:52, 10 April 2014 (UTC)

It emphatically doesn't mean random essays put up by wikiprojects (or whoever). The WP:PROPOSAL process is how most "subject-specific guidelines listed in the box" were elevated to guidelines. Even when they weren't it was after a long consensus discussion at the parent page. This is true of subject-specific notability guidelines, s.-s. style guidelines, and s.-s. naming conventions guidelines. One of the problems with much of the discussion above is the fallacious reasoning that because a localized consensus of editors interested in some partcular things (the articles withi the scope of a wikiproject, or how to handle something in one particular article) that they should just tell the wider consensus (be it at MOS or where ever) to go to hell. This is not how WP works. If you think you've found a hole in a major guideline or policy, you go talk about it at that guideline/policy's talk page, not just your little backwater talk page, to gain consensus to change the guidance in question to account for a variance, or a consensus that the variance is unusual (WP:IAR) or obvious WP:COMMONSENSE) enough that no variance is need, or that you're wrong (WP:DEADHORSE) and pushing some pet peeve. The incessant problem of wikiprojects, and various groups of pundits on controversial articles, thinking that WP:LOCALCONSENSUS doesn't apply to them somehow is a constant and worsening source of pointless conflict.  — SMcCandlish ☺ ☏ ¢ ≽^ʌⱷ҅_ᴥⱷ^ʌ≼  21:12, 12 April 2014 (UTC)

The WP:PROPOSAL process is how most "subject-specific guidelines listed in the box" were elevated to guidelines.

As a point of fact, pretty much all of the subject-specific guidelines tagged before 2007 did not have formal proposals—and that's most of them. Before 2007, the actual process was a few people wrote something, and eventually someone slapped a label on it. The MOS pages in particular have a long history of problems, although the massive effort to reconcile the MOS a few years ago dealt with the worst problems. WhatamIdoing (talk) 18:38, 14 April 2014 (UTC)

It's completely up to the editors on a page to agree how to follow policies and as long as they are satisfied no amount of authoritarian wishing can change that. It is first of all just a fact of the matter, and secondly the corrective on the mistakes of the self-appointed experts crew who frequently get things wrong and are slow to fix them. Policy reflects practice, so when practice changes, a few eggs might get broken. No harm done. --Ring Cinema (talk) 22:45, 12 April 2014 (UTC)

No. No consensus on any page can choose to negate core policies such as WP:BLP, WP:RS andWP:NPOV. Period. And saying that all policies can be ignored is a sure way to make Wikipedia a total anarchy. Cheers. Collect (talk) 23:25, 12 April 2014 (UTC)

Yep. There's not a question of negating anything. Period. It's already the case that any policy can be ignored by the editors on a page and nothing will change that, even those who want to pretend they are telling other editors what to do. With respect, your view is based on the idea that other editors just have to do it the way the policy tells them to. In fact, compliance is voluntary and most editors are perfectly willing to follow the most common conventions. What prevents anarchy is that the policies are good and reasonable. If that perception was undermined, as occasionally happens, policies would be ignored, as occasionally happens. --Ring Cinema (talk) 01:19, 13 April 2014 (UTC)

The wikipedia system is "fuzzy " in all respects. How policies and guidelines are written, how they interact with each other, how they guide or influence the outcome of things. And so I think that trying to derive a broad specific answer out all of that is like trying to herd cats. But a accepted core part of that fuzziness is that certain metrics determine how strongly a policy or guideline determines the outcome of a discussion. These include:

• It's place in the Wikipedia hierarchy, with WWF by-laws policies at the top of the list, followed by core policies, followed by policies, then prominent guidelines, then non-prominent guidelines, with project guidelines near the bottom of the list. Higher in this hierarchy means more people gave this more serious and careful consideration.
• To what degree the policy etc explicitly and clearly and categorically weighs in on the item at hand. (vs. creative claims of such)

The bottom line is that when a core policy clearly and categorically weighs in on the topic at hand, that trumps pretty much everything else. Things get fuzzier as the "metrics" get weaker. North8000 (talk) 13:53, 13 April 2014 (UTC)

Generally agree, but I would note that if a project guideline is in fact broadly followed across a whole class of articles, that also evidences a wide editing consensus. Alanscottwalker (talk) 14:02, 13 April 2014 (UTC)

(added later)That could mean merely that it is usually a good thing to do, not wide agreement that it should be binding. North8000 (talk) 15:20, 13 April 2014 (UTC)

Except that all guidelines are the usually "good thing to do" and in theory actually embody the good thing done. Alanscottwalker (talk) 15:30, 13 April 2014 (UTC)

As I suggested above, they would still be subservient to site wide policies, but superior to individual articles -- I think that is your point? Collect (talk) 15:00, 13 April 2014 (UTC)

No, my point is it potentially raises an actual conflict - as two different sets of editors interested in this over here do not agree how the guidelines properly apply over there, which are of interest to another set. As for policy, it is often written in the general, whereas guideline are more specific, then some guidelines are more specific than others -- and articles are written in the specific, not the general. Alanscottwalker (talk) 15:09, 13 April 2014 (UTC)

There is no amount of insistence that changes the fact of the matter that when editors agree to ignore a policy they can do it. Unless there's a mechanism I don't know of whereby a policy prevents editors from making changes or additions to pages. Sure, if there's a conflict about it the policy will normally be followed but that is again a simple case of consensual compliance by the editors who become involved. But if the editors involved ignore a policy, it will actually be ignored and the consensus will be to ignore the policy. Presumably, it is done for good faith reasons; no words can change that reality. --Ring Cinema (talk) 17:12, 13 April 2014 (UTC)

Well, right until a "disruption" is declared by the powers that be (yet, another consensus/process) and pages protected and people are exiled or self-exiled (this may be a long or short process or series of processes) -- the hope is that none of the parties will let it reach that stage -- but those things will likely occur sooner rather than later in what are seen as core affected areas. Alanscottwalker (talk) 18:20, 13 April 2014 (UTC)

What powers that be? The authoritarian impulse is not even a good metaphor, but it seems to have been used literally. In fact, good faith edits that ignore policies generally aren't even a remotely good reason to discipline an editor. The normative aspect of policy compliance is extremely widespread in its acceptance for reasons that are self-evident, but there is no cop on the beat and when the policy is wrong editors are doing everyone a favor by ignoring it, since that is the corrective on inferior policy-making. Policy reflects practice, so by definition a change in policy might postdate a change in practice. No problem! --Ring Cinema (talk) 22:31, 13 April 2014 (UTC)

What powers that be? A host of other editors, administrators, arbcom. Good faith editors that repeatedly ignore those (and policy) will often be disciplined ("good reason" or not). Alanscottwalker (talk) 00:55, 14 April 2014 (UTC)

Sure, Ring Cinema (talk · contribs), when a local consensus of editors agrees to ignore a policy they can do it; but there can and often should be consequences. But such cases should be aberrations and still should fall into one of the three following categories.

1. A change that is later reversed for compliance with policy.
2. Part of a pattern that eventually leads to a corresponding policy adjustment.
3. A truly unique case unforeseen by policy where IAR applies and a no corresponding change to policy is practical.

I think it's important for editors who decide to go against policy to understand which of these three cases they believe they are in, and to state this clearly in the corresponding discussion. --B2C 18:12, 14 April 2014 (UTC)

Although it is widely misunderstood, wp:iar is also policy, and rightly so. IMO it's main use and applicability to to prevent mis-use of policies (e.g. by wikilawyering) contrary to their intended purpose rather than to override their intended purpose. North8000 (talk) 22:52, 13 April 2014 (UTC)

You can always consider the one in a billion exceptional case where some supposedly absolute rule fails (i.e. comes into conflict with improving Wikipedia as judged by a consensus of editors), so only WP:IAR will always apply. And since Wikipedia has a huge number of articles, a one in a billion exception may well occur from time to time. Count Iblis (talk) 00:42, 14 April 2014 (UTC)

No, that's a misstatement of the facts. There are no absolute rules in this context so it's unclear what you're referring to. Decisions are made by consensus, in general. While authoritarians wish to impose their rules on others, it's out of place on Wikipedia, which doesn't operate that way. Poor policy decisions are not uncommon and editors are rightly the ones to decide how to apply policy on the articles they edit; that is what makes Wikipedia great. Bad policies are ignored, yet policy compliance is high. No problem! --Ring Cinema (talk) 01:50, 15 April 2014 (UTC)

Process

Perhaps add something along the lines of: "Where two guideline writing projects conflict, and the matter cannot be settled, a sitewide RfC, constructed with the aide of Mediation or Dispute Resolution (by project members and others interested) should occur on its own dedicated page or at VPP." Alanscottwalker (talk) 13:46, 13 April 2014 (UTC)

It might be simpler to say, "if two advice pages conflict, see WP:POLCON". We don't need to duplicate the existing policy on resolving conflicts between policy and guideline pages. WhatamIdoing (talk) 18:40, 14 April 2014 (UTC)

Suggested change

Given my concerns (above) about the fact that, sometimes, a consensus reached at a large WikiProject may actually represent a wider consensus than one reached at a relatively obscure guideline page... I am beginning to think it best to simply omit the issue of "WikiProject consensus" vs. "Guideline consensus". So... I would like to discuss the following suggested change:

• Consensus among a limited group of editors, at one place and time, cannot override community consensus on a wider scale. For instance, unless they can convince the broader community that such action is right, participants in a WikiProject cannot decide that some generally accepted policy or guideline does not apply to articles within its scope. For example, a few editors working on a particular article cannot unilaterally decide that a generally accepted policy or guideline does not apply to that article. They need to seek a wider consensus, to determine whether the wider community agrees that an exception should be made.

My suggestion is far from perfect... and I am not proposing it as a finalized product. I'm thinking conceptually at this stage, and not focused on specific language. Please discuss. Blueboar (talk) 12:42, 17 April 2014 (UTC)

Demur. And your wording would suggest that any Wikiproject could decide that any policy simply "does not apply to this project" if "more people opine at the Wikiproject than opined at the policy or guideline page" which is the path to anarchy. If the members of your posited large Wikiproject want to make the change, they should do so at the policy or guideline level and see if the broader community (This does not mean "more editors make for a stronger position, by the way, only that people with a broader range of interests are able and likely to participate) concurs with their elision of such a policy or guideline. If the broader group does not concur, then the elision would be improper. Suppose "Wikiproject:Evil People" decided to say "WP:BLP does not apply to articles under this project" - your wording would allow them to do so. And because of that possible reasoning, I demur on any removal of the primacy given to policies and guidelines over Wikiproject discussions. Collect (talk) 14:32, 17 April 2014 (UTC)

My suggested wording actually says nothing about WikiProjects (one way or the other). That was intentional.

The reality is that a large WikiProject would never reach a consensus to ignore a core policy (like WP:BLP or WP:V) in the first place. It just won't happen. So there is no need to create a "rule" about the possibility.

Now, there is the possibility of disagreements between large WikiProjects (with perhaps hundreds of active editors) and relatively obscure guidelines (with perhaps as few as 10 active editors)... In those situations, I disagree that the WikiProject needs to edit the guideline, and somehow make their consensus "official". The there is no need for a guideline to list every single exception to its guidance... the idea that there might be unwritten exceptions to guidelines is a given. That is why we wrote WP:Ignore all rules and promoted it to POLICY level - WP:IAR explicitly allows a WikiProject to say: "Guideline X may well work for every other topic area in Wikipedia... and because it does work in all those other articles, there is no need to change it... however, guideline X makes no sense in articles relating to topic Y... so in Y articles we are going to simply ignore guideline X". Blueboar (talk)

I would still prefer that certain policies be considered as non-negotiable, and that no project or article may violate such policies. I find IAR to be a strange animal of exceedingly limited relevance to the editing of articles, and appears primarily to be applied to non-editing actions rather than allowing elision of policy. Collect (talk) 13:13, 18 April 2014 (UTC)

But all policies are negotiable, were negotiated, will be negotiated, per the wishes of the editors. The idea that Wikipedia editors shouldn't be trusted isn't consistent with anything that happens here. Compliance with policy is widespread because the policies are good. When the policies are bad, editors do what's best. And the brute fact is that editors are completely free to ignore any policy as long as there is a consensus for it. Or, to put it another way, policy interpretation is subject to the constraints of consensus. That is not something to fear, it's the strength of Wikipedia. Policy is silent absent a dispute. Misapplied authority is a bigger problem then misapplied consensus. --Ring Cinema (talk) 15:26, 19 April 2014 (UTC)

Again, the reality is that a large WikiProject would never reach a consensus to ignore one of our "non-negotiable" core policies (like WP:BLP or WP:V). It just won't happen. Why? Because the vast majority of the editors participating the large WikiProjects support the core policies. If someone at a large WikiProject suggested ignoring BLP, the other editors in that project would quickly squash the idea. Where the potential for conflict between "the rules" and consensus might occur is at the guideline level. Guidelines don't enjoy the same universal support that the core policies do, and thus it is more likely that a consensus will form saying that an exception should be made. That's OK. Blueboar (talk) 16:56, 19 April 2014 (UTC)

Um-- nope. Some policies are specifically stated as not negotiable, and I have more faith in them than in any prayer that "no group would ever do anything wrong". "It just won't happen" is a classic example of "they couldn't hit an elephant at this distance." Collect (talk) 06:01, 20 April 2014 (UTC)

Indeed. It's also the same reasoning as "He couldn't be a serial killer, he's too quiet and nice", and "our government couldn't possibly turn into a police state, this is a democracy", and "I don't need to lock my door in Toronto; Canada is safe", and "it's perfectly safe for a toddler to play with my 130-pound mastiff, he's never bitten anyone", and "Why not point this gun at my head and play with it? It not like someone would be stupid enough to leave bullets in it."  — SMcCandlish ☺ ☏ ¢ ≽^ʌⱷ҅_ᴥⱷ^ʌ≼  07:49, 20 April 2014 (UTC)

• Oppose proposed change or anything like it: We already have too many wikiprojects who think no rules they don't like apply to them. A change like that would be immediately and broadly misinterpreted as a huge "victory" for wikiproject "sovereignty", followed by a mad rush to ignore every inconvenient rule (especially MOS rules, but plenty others too) with impunity, specifically because wikiprojects were no longer mentioned explicitly. Further up this page, there's discussion of expanding this section to include more specific examples, e.g. reference desks, because they too often wrongly conclude they can ignore broader, site-wide consensus with impunity. WP:BLP and WP:V are red herrings here. It's a real problem that projects are ignoring MOS generally, MOS:ICONS, MOS:CAPS, etc. more specifically, as well as the naming conventions guidelines, and others. It doesn't really matter if Blueboar thinks that projects shouldn't have to gain consensus at WT:MOS to change the style guidelines. That's just how it works. It's how it needs to work to avoid complete chaos. If their reasoning makes sense to other editors and readers, it will prevail; if it doesn't, it won't. If the discussion only happens on their own project talk page, their reasoning will always prevail, not matter how faulty it is, because they're simply asking themselves for permission. Projects talking to themselves is not a consensus. PS: Just because a project lists 100s of editors doesn't mean it has 100s of active ones, much less that they all care enough about style or other guideline matters to comment on it. Just because the MOS pages are pretty stable and are not being actively edited all day every day doesn't mean they don't have a remarkable number of participants. MOS is one of the most-watchlisted Wikipedia-namespace pages on the system. The underlying logic of this "my wikiproject is bigger than your guideline crew" attitude is seriously faulty to begin with (aside from the us-vs.-them posturing). The average Westerner believes in astrology, but that doesn't make it any less pseudoscience; numbers are no substitution for reason. Wikiprojects overvalue reasoning that suits their prejudices, often totally discounting, even mocking, the concerns of editors trying hard to achieve balance in a encyclopedia with the broadest possible audience. So, no thanks. Keep that mention of wikiprojects precisely where it is in this policy!  — SMcCandlish ☺ ☏ ¢ ≽^ʌⱷ҅_ᴥⱷ^ʌ≼  07:49, 20 April 2014 (UTC)
• Oppose as a recipe for chaos. We have enough anarchy here already. This would open the door for even more walled gardens in style, among other problems. No, no, a million times no. --John (talk) 08:08, 20 April 2014 (UTC)
• Strong oppose - This is for the birds. Collect got it right. --Stfg (talk) 09:32, 20 April 2014 (UTC)

Except that not just policies but site-wide guidelines were included for a reason. As far as I can tell, the main reason some want to change this section is to undermine the MOS having any ability to meddle in the city-state affairs of sovereign wikiprojects. <rolling eyes>  — SMcCandlish ☺ ☏ ¢ ≽^ʌⱷ҅_ᴥⱷ^ʌ≼  11:40, 20 April 2014 (UTC)

I agree, but why the "except"? Collect said: "I demur on any removal of the primacy given to policies and guidelines over Wikiproject discussions". (My italics) My position is identical to that. --Stfg (talk) 14:50, 20 April 2014 (UTC)

Durrr! I missed that; had been reading a different comment of Collect's that just mentioned policies. :-)  — SMcCandlish ☺ ☏ ¢ ≽^ʌⱷ҅_ᴥⱷ^ʌ≼  21:31, 20 April 2014 (UTC)

Amazingly enough, I do not always use the full set of words which are implicit in my stated position <g>. Collect (talk) 21:40, 20 April 2014 (UTC)

Right! It's just in this particular discussion, a larger than usual number of participants see guidelines as only glorified essays, and see "policy" in the mass-noun sense as only including "polcies" not "policies, guidelines and other ways things work consensus here", so it's important disabuse them of that notion quickly before the discussion goes off the rails. >;-)  — SMcCandlish ☺ ☏ ¢ ≽^ʌⱷ҅_ᴥⱷ^ʌ≼  07:59, 22 April 2014 (UTC)

• Oppose. If the local consensus is the wider consensus, then no change is needed, since the local consensus will have no trouble convincing itself. -- JHunterJ (talk) 11:15, 20 April 2014 (UTC)

  The problem is that there are often a few "the guideline is always right" diehards who refuse to accept that there is a consensus to ignore the guideline. They keep pointing to the guideline and saying "Your consensus "violates the rules" ...the guideline says X, you must do X"... such editors refuse to accept the larger consensus that says "Yes, we know... and for other articles X is excellent advice... but not in these articles... these articles are an exception to X". Blueboar (talk) 12:53, 20 April 2014 (UTC)

  The solution to that is to seek a revision of the site-wide guideline on its talk page, not to sidestep it. --Stfg (talk) 14:50, 20 April 2014 (UTC)

  Exactly. If the local consensus really is wide enough not to be local, then it is wide enough to update the site-wide guideline to reflect the new consensus (which is no longer local). -- JHunterJ (talk) 15:16, 20 April 2014 (UTC)

I disagree... there is often no need to "update" what is usually excellent guideline advice to account for an exception that may well affect only a limited number of articles. That is simply instruction creep. By explicitly mentioning what might well be a very limited exception, you highlight it. Doing this can cause more argument, confusion and misunderstanding than not mentioning it. Blueboar (talk) 15:50, 20 April 2014 (UTC)

If it's not the broader consensus, then it's a local consensus, and to be avoided. This is usually the point at which WP:IAR gets brought up, and ignoring the rules is fine when there is consensus that the encyclopedia is improved by that ignoring of the rules. And there's still no need to change this, and good reasons not to change it: the encyclopedia will not be improved by additional local consensuses for very limited exceptions that don't improve the encyclopedia. -- JHunterJ (talk) 16:02, 20 April 2014 (UTC)

Yes... I am talking about situations where there is a solid and wide consensuses to invoke WP:IAR... I am talking about situations where the consensus is that the encyclopedia will be improved by ignoring a particular rule. The way the CONLIMITED is currently written, it implies that such a solid wide-ranging consensuses can not form on WikiProject pages... but they can. And when they do, they don't necessarily require editing the policy or guideline page to account for them. Blueboar (talk) 16:58, 20 April 2014 (UTC)

Such solid and wide consensus should be reflected in the guidelines. The guidelines are specifically there to describe the solid and wide consensuses of the encyclopedia. Such solid, wide-ranging consensus cannot be held solely by a Wikipedia Project; when they are solid and wide, the guidelines need to be updated to reflect them, since they are then the current practice of Wikipedia that the guidelines are there to describe. If there is no consensus to put them in the guidelines, the consensus is not yet solid or wide enough (and so is still a local consensus). -- JHunterJ (talk) 17:21, 20 April 2014 (UTC)

Yep. Blueboar, you seem to be contradicting yourself. If it's so trivial that it'll only affect a limited number of articles (your emphasis), and taht adding mention of it to the guideline would be just instruction creep, it can't possibly represent something that is a solid and wide consensus to invoke WP:IAR (your emphasis again). It can't be a molehill and a mountain simultaneously. BTW, are we talking about bird common name capitalization again, perchance? If so, your perception of a solid and wide consensus for it is a house of cards.  — SMcCandlish ☺ ☏ ¢ ≽^ʌⱷ҅_ᴥⱷ^ʌ≼  21:53, 20 April 2014 (UTC)

• Oppose. Reading "simple omit the issue of "WikiProject consensus" vs. "Guideline consensus".", I was headed towards "of course, yes, how did that get in there." Defining "WikiProject consensus" vs. "Guideline consensus" is a bad idea. "Consensus" is a nebulous concept. We couldn't source the topic at "Consensus", and had to move it to Consensus decision-making. Consensus is a target that is built or developed, and it defies definition. The best we can do is in hindsight. A consensus is shown to have been a consensus if it is used or followed, and not objected to, by a significant number of people, and a significant period of time.

However, the sentence suggests to be cut refers to the situation of differeing prevaling opinions ("convince" implies an opinion being advocated) at a WikiProject page and at a policy or guideline page. This is quite reasonable. I would propose instead to change

"They need to seek a wider consensus, to determine whether the wider community agrees that an exception should be made."

to

"They need to seek a wider consensus participation, to determine whether the wider community agrees that an exception should be made policy should be modified."

NB. This implies, importantly, that conflict between documented policy (on a policy of guideline page) and WikiProject advice should be resolved at the policy/guideline page. --SmokeyJoe (talk) 12:45, 20 April 2014 (UTC)

Smokey, I can agree with your consensus participation suggestion... but not with an exception should be made policy should be modified suggestion. That simply invites unnecessary instruction creep. Modifying a policy or guideline to explicitly state every exception simply clutters up the guideline (distracting editors from what may well be excellent advice in every other situation).

Let's suppose that a WikiProject has a concern that some provision in a guideline does not make sense for the articles in their topic area... however, they understand why the guideline includes the provision, and agree that it is good advice for other topic areas. They file a wiki-wide RFC on the project talk page to explain their concerns, and after due discussion (and perhaps some compromising) the wider community reaches a consensus that agrees with their view. The guideline is still good advice. There is no need to modify the guideline so that it explicitly says "Do X, except in articles relating to Y". Indeed, explicitly stating an exception might actually cause more confusion that clarity, resulting in more misunderstanding of what is generally excellent advice. Blueboar (talk) 14:57, 20 April 2014 (UTC)

I generally concur with SmokeyJoe, but would go with "policy or guidelines should be modified" or something like that (it's not all about policies, per se). Blueboar, the kind of situation you're describing is almost always WP:Specialist style fallacy-based, insular exceptionalism, and should not be allowed. In the rare cases it does make sense, it most definitely does need a change to the guideline in question. The very fact that you're talking about wikiprojects coming to their own ivory-tower conclusion they need to do something different from all other topics isn't just a red flag, it's a parade of huge red banners.  — SMcCandlish ☺ ☏ ¢ ≽^ʌⱷ҅_ᴥⱷ^ʌ≼  21:53, 20 April 2014 (UTC)

Yes, "policy or guidelines should be modified" was intended. It gets tedious to keep writing "policy or guidelines", and it is usually taken that a generic small p policy includes guidelines. --SmokeyJoe (talk) 07:43, 21 April 2014 (UTC)

I don't have time to read everything that's been added here in the last couple of days. I just want to add this fact: The reason that we have this statement about "participants in a WikiProject [deciding] that some generally accepted policy or guideline" doesn't apply is because the community kept having to deal with drama over this misconception. It was not a random example. It was added to address the specific and moderately widespread problem of people deciding that whereas they and their buddies over at the WikiProject hated infoboxes, or thought obscure original publications in a franchise were more important than the one that all the sources were talking about, or whatever their pet idea was, that they could impose their decision on articles that they'd had nothing to do with writing, because it's "within our scope". WP:Advice pages also addresses this explicitly, and has for several years, but we get people claiming that they get to boss around other editors because they declared themselves to be "a WikiProject". WhatamIdoing (talk) 01:07, 24 April 2014 (UTC)

• Oppose the proposal. A best case scenario is that it does nothing. Worst case is that it causes chaos, as others have noted. --BDD (talk) 19:04, 28 April 2014 (UTC)

Further CONLEVEL discussion

Moved from Wikipedia talk:Manual of Style § WT:CONSENSUS discussion to remove mention of wikiprojects from WP:LOCALCONSENSUS

SMcCandlish, I don't think the proposal would have the effect you fear. The idea behind the proposal is that, when two groups of editors reach a conflicting consensus, the locations of the two consensus discussions matters less than the relative size of the two consensuses.

I think everyone would agree that a consensus reached by hundreds of editors should out weigh a consensus reached by a limited number of editors ... because hundreds of editors better reflects the consensus of the "wider community" than one made by a limited number of editors. The problem is that some of our more obscure guidelines are the product of a very limited number of editors, while some of our larger WikiProjects have hundreds of active editors. In such a situation, we have to ask: which better reflects the views of the "wider community"? Blueboar (talk) 12:39, 20 April 2014 (UTC)

Blueboar, this talk of "consensus" is an erroneous use of the word. Hundreds of editors do not make a consensus if there are other editors elsewhere working with a different view, and the two groups have not engaged directly. Consensus buiding requires involvement of all stakeholders. If there are two groups in different places, they need to be made to work together in the same place. I think the obvious solution is to discourage pseudo-policy documentation being hosted on WikiProject pages. Instead, the WikiProject members need to engage other interested editors at the policy page. --SmokeyJoe (talk) 12:57, 20 April 2014 (UTC)

OK, but the same needs to apply in the other direction. Hundreds of editors (at a guideline) do not make a consensus if there are other editors elsewhere (at a WikiProject, or even on a specific article) working with a different view. Now... if we accept that, how does it mesh with what WP:CONLOCAL says? The entire rational behind CONLOCAL seems to indicate the opposite... that some consensuses (those at a policy/guideline) in some way trump other consensuses (those at WikiProjects).

I completely agree that when two groups reach conflicting consensus, they need to come together, discuss the issue and reach a compromise... that's how Consensus is supposed to work. The problems arise when the two groups refuse to listen to what each other is saying, and start engaging in pissing contests over which page trumps the other (for example Guideline vs WikiProject page).

Another issue is the question of where the new (compromise) consensus must be reached... does it have to be reached at the relevant Policy/Guideline page, or can that discussion take place on the WikiProject's page? (my opinion is that it does not matter)... however, if it takes place on the WikiProject page, does CONLOCAL apply to that new consensus? I hope we could agree that it does not really matter which page the discussion takes place at. Unfortuately, the current wording of CONLOCAL can be interpreted as saying that the (new) consensus will continue to be discounted, simply because it took place "at a WikiProject" and not at a policy/guideline page. There will be diehards who refuse to accept that a new "wider" consensus has been reached... simply because of where the discussion took place. These diehards will continue to argue that the (new) consensus is "LOCAL" even though it isn't. Yet another unintended flaw with the current language of CONLOCAL. Blueboar (talk) 13:48, 20 April 2014 (UTC)

I don't agree with the language on levels of consensus at WP:Consensus. Instead, I would say a group in isolation may think it has a Wikipedia consensus, but it is mistaken. It may be operating under a false consensus. The term "local consensus" is not helpful, it implies independent separate parts of the community.

Where should separate groups come together? In theory, it needn't matter. In practice it does. They should come together at the policy/guideline talk page, or elsewhere if agreed and advised at that talk page, elsewhere such as a subpage of the talk page, or a WP:RfC subpage. WikiProjects are special interest areas, and are not appropriate places to document community practice. --SmokeyJoe (talk) 14:25, 20 April 2014 (UTC)

[End of material copied from WT:MOS].

@Blueboar: I've given clear reasons why the proposed change would have the effect I'm concerned about; your assertion that it wouldn't isn't persuasive. I think your summary of the idea behind the proposal is correct, and that it is precisely why it would happen. Your might-makes-right, power-in-numbers argument on this is not how consensus works here. The arguments presented by the stakeholders (who are often not divided into "sides") are what is weighed, not a head-count, otherwise it's simply a vote, a popularity contest. That would be totally untenable here. The majority of Westerners, if not people in general, believe in both astrology and the gambler's fallacy, but that doesn't somehow make them not pseudo-scientific superstitions, as just one example of the flaw in your reasoning

It's both intentional and crucial that CONLEVEL does not flow both ways. There is no shortage of "consensuses" of "hundreds of editors" to push all kind of WP:POV crap in articles, and it's easy to manufacture more of them simply by canvassing Internet forums to upset the balance in particular articles or even whole topic areas, but much more difficult to change WP policies in this way, because peole from all over the system, all walks of wikilife, watchlist our policy and guideline pages and are resistant to questionable changes being made to them.

SmokeyJoe (talk · contribs) is correct that a big knot of editors in one wikiplace does not constitute a consensus if they're not communicating with another cluster of editors with a contrary view. And I certainly agree with his statement that the obvious solution is to discourage pseudo-policy documentation being hosted on WikiProject pages. We even have a guideline about this at WP:PROJPAGE, though clearly very few editors are familiar with it. But that analysis doesn't go quite far enough here, as I'll elaborate in a bit. I also agree that the term "local consensus" isn't helpful; it really is about insularity, about groups in effective isolation making up their own rules (and it's why, for example, MOS supersedes it's style sub-guidelines; few of them are well-watchlisted, and they frequently get POV-forked on various points.)  — SMcCandlish ☺ ☏ ¢ ≽^ʌⱷ҅_ᴥⱷ^ʌ≼  21:27, 20 April 2014 (UTC)

PS: The rather obvious WP:CONLOCAL shortcut we were using somehow didn't actually exist yet, so I created it.  :-)  — SMcCandlish ☺ ☏ ¢ ≽^ʌⱷ҅_ᴥⱷ^ʌ≼  07:16, 21 April 2014 (UTC)

Oh no! There are already too many shortcuts. And these buzzword style shouty shortcuts have led, in my opinion, to a dumbing down of debates where they are used. And I thought we were agreeing that the term was negatively productive? --SmokeyJoe (talk) 07:46, 21 April 2014 (UTC)

We can MfD it later. Or we can refactor this to use CONLEVEL and speedy the shortcut now. Heh.  — SMcCandlish ☺ ☏ ¢ ≽^ʌⱷ҅_ᴥⱷ^ʌ≼  08:03, 22 April 2014 (UTC)

re: ...a big knot of editors in one wikiplace does not constitute a consensus if they're not communicating with another cluster of editors with a contrary view.... I actually agree with that. My concern is that the current language does not make it clear enough that this goes in two directions. A big knot of editors (at a guideline page or a WikiProject) does not constitute a consensus if they are not communicating with another cluster of editors (at a WikiProject or a guideline page) with a contrary view. My point is that guideline pages do not automatically "trump" WikiProject pages ... when they disagree, a wider - joint consensus is needed. Blueboar (talk) 13:23, 23 April 2014 (UTC)

But I think they do. The point isn't the size of the knot of editors, but the scope of the policies and guidelines. This page, the MOS, etc, are project-wide. Subject-specific wikiprojects aren't. --Stfg (talk) 14:40, 23 April 2014 (UTC)

But that causes problems... some of the more obscure guidelines may have a wiki-wide scope... but don't actually have wiki-wide consensus. Blueboar (talk) 23:39, 23 April 2014 (UTC)

Example please of any such guideline where any proposal to emend it gets shot down so that a project would need to assert it is inapplicable? If you can name one, I will gladly assist in rewording it to avoid being ignored by the affronted wikiprojects. Cheers. Collect (talk) 00:44, 24 April 2014 (UTC)

WikiProject Composers tried it. It took weeks of drama and a couple of RFCs to convince them that their project page did not actually trump WP:INFOBOXUSE, and that they therefore could not ban the use of (dis)infoboxes on every article that they said was within their scope. The community will not tolerate changes to the guideline at INFOBOXUSE that say "The use of infoboxes is neither required nor prohibited for any article, except if the article is about a musical composer, in which case it's banned" (which is what it would take). WhatamIdoing (talk) 01:15, 24 April 2014 (UTC)

@Blueboar: I think you are confusing consensus with majority. Actually, every consensus anywhere on Wikipedia was hammered out by a tiny minority of editors. The point is, they took the trouble to hammer out the agreement, maybe needing to settle on compromises en route. For anyone to say, well, me and my friends weren't there, so it's not our consensus and we'll do whatever we like, is just the road to factionalism. The way to move forward is to go for further discussion in the main scope, not to create walled gardens. --Stfg (talk) 10:23, 24 April 2014 (UTC)

Thanks for stating the reality. Consensus is made by small numbers of editors. This discussion highlights that a small number of self appointed experts would like to believe they can dictate to the rest of the editors. That's not how it works. The fortunate reality is that policies are followed because they are good policies, and it does no good to pretend that anything is non-negotiable or in some way supersedes or trumps the rest. That's just an authoritarian fantasy and doesn't exist here, even if someone says it does. --Ring Cinema (talk) 22:26, 24 April 2014 (UTC)

Exactly my concern, Ring. When guideline pages clash with WikiProjects, what we often end up with are two competing groups of "self appointed experts" ... the "experts" at the guideline vs. the "experts" at the WikiProject ... with both groups trying to dictate to the other. Neither group is actually seeking consensus. That's the problem I am trying to address. Blueboar (talk) 14:34, 25 April 2014 (UTC)

Stfg, as usual you put that very well. Blueboar, "problems" caused by WP having one MOS not 200 different ones run by wikiprojects and PoV gangs are problems we can certainly live with, and it's much more managemable than the reverse. MOS has emerged as a centralized resource because it was needed by the community, not because some people are jerks and don't like the way you want to hyphenate, or whatever it is you're unhappy with MOS about. WP:AT was elevated to policy for similar reasons in part, WP:N evolved out of scattered topical notability guidelines and someone's userspace notability essay for similar reasons, and WP:CENT exists for similar reasons, and so on. Some people don't like centralization simply because it's centralization and some they're anarchic by nature (note Ring Cinema's continual refrain about "authoritarians"; that word being applied to me is totally absurd if you know anything about my professional history!). Certain things just functionally work much better when centralized. A style guide is such a thing, pretty much by definition.

Also (to reply to your later comment and Ring Cinema's, without forking my response to a different indentation level), your characterization of MOS disputes as just some self-appointed experts in conflict with some other self-appointed experts, so it's a toss-up, is missing something obvious: MOS's rotating cadre of grammarians and linguists and professional writers and English teachers and whatnot, with a strong background in language and communication, are experts (self-appointed or not) on style and writing, while topical experts are experts on facts in those topics, not on style and writing [except of course where they individually also coincidentally have expertise in that area]. It's a crucial difference. WP:SSF explains a lot of that difference's contours. Your position on this is directly equivalent to declaring that sci-fi fans trying to rewrite all of the physics articles to agree with what happens in "Start Trek" and "Star Wars" is just some conflict between camps of expert editors and thus a tie/draw. But there is no legitimate conflict in physics between physics experts and sci-fi experts, no matter how much sci-fi experts think they've absorbed all they need to know about physics from their sci-fi reading and writing and thus feel they, too are experts in that topic. They're just not [unless, again, they individually also have a professional or education background that coincidentally also make them real physics experts]. The idea that the average biologist is also a language usage expert is absurd.

Way more important than any of this is that MOS is a centralized guideline everyone can participate in and that a zillion people watchlist. It's hard to change on a whim, and it represenets a wide array of viewpoint. This is never true of any wikiproject talk page (much less WP:PROJPAGE somewhere proposing to be a "guideline". Even if MOS had no language experts on it at all, it would still presently the community-wide consensus on who to handle language issues of a genera style and grammatical and layout nature.  — SMcCandlish ☺ ☏ ¢ ≽^ʌⱷ҅_ᴥⱷ^ʌ≼  20:37, 25 April 2014 (UTC)

Expand description of Forum shopping

At present the policy on WP:FORUMSHOP says "Forum shopping, admin shopping, and spin-doctoring. Raising essentially the same issue on multiple noticeboards, or to multiple administrators, is unhelpful to finding and achieving consensus." (Talk page archives show different versions of this line.) I suggest expanding the description to say "Raising essentially the same issue on multiple noticeboards or talk pages,...." (Suggested addition in italics.) Thoughts? – S. Rich (talk) 23:12, 1 May 2014 (UTC)

WP:1AM

Please critique my essay about "one against many" consensus situations at WP:1AM. Comments should be posted on Talk:1AM. Thanks! --Guy Macon (talk) 09:06, 12 April 2014 (UTC)

I think it's very good but not sure how practical it is. I've certainly found myself in 1 against many situations, and I usually walk away. Much more common are a few against many situations. That's much more difficult. The history of the yogurt title comes to mind, and of course the US city name controversy, and much more current now is the RM discussion at Talk:Hillary Rodham Clinton.

To address a specific type of a few against many situations, I wrote this essay, intended primarily as advice for RM closers: User:Born2cycle/Yogurt Principle. --B2C 19:59, 14 April 2014 (UTC)

Damn... I was hoping that an essay with the short cut "WP:1AM" would be advice saying not to edit at 1:00 in the morning - after coming home from the pubs. Now that would be a useful essay. Blueboar (talk) 02:51, 2 May 2014 (UTC)

Bird article name (capitalisation)

There is a move discussion in progress on Talk:Crowned Crane about four articles related to birds species. The rationale is that there is no reason why bird names should be capitalised while Wikipedia recommends that all animal names should not be written with capitals. Please participate to the discussion.
Thank you! Mama meta modal (talk) 09:07, 9 March 2014 (UTC).

Request for comments

There is now also an ongoing request for comments on the same subject: Talk:Crowned Crane#Request for comments.

Do not hesitate to come and comment on this question. Mama meta modal (talk) 08:52, 16 March 2014 (UTC).

Consensus

The discussion was closed (and the pages moved) on 26 March 2014, see Talk:Crowned crane#Requested move for details.

Mama meta modal (talk) 20:49, 26 March 2014 (UTC).

Move review for species pages at Wikipedia:Move review/Log/2014 March. Cas Liber (talk · contribs) 00:32, 27 March 2014 (UTC)

See also Wikipedia talk:Manual of Style#Second proposal. Mama meta modal (talk) 06:25, 9 April 2014 (UTC)

New discussion

An important discussion started on Talk:Crowned crane and Wikipedia:Move review/Log/2014 March#Black crowned crane now moved to Wikipedia talk:Manual of Style#A new proposal regarding bird article names.

Mama meta modal (talk) 20:56, 9 April 2014 (UTC).

The consensus is now clear. The relevant pages will soon be checked and made consistent with Wikipedia talk:Manual of Style#Bird common name decapitalisation.

H. H. Wander Strata (talk) 23:07, 2 May 2014 (UTC).

New essay

Just started a new essay at Wikipedia:Confusing arguments mean nothing. Would appreciate any collaboration.--Paul McDonald (talk) 21:45, 26 May 2014 (UTC)

WT:EW

There's a discussion at Wikipedia talk:Edit warring#BRD cycle related to WP:EDITCONSENSUS and WP:BRD that may interest some editors who watch this page. WhatamIdoing (talk) 00:41, 2 June 2014 (UTC)

query

Should we differentiate between an "agreement between two editors" where no formal section seeking consensus is started from this process of actively seeking to accommodate the opinions of all the editors on an article or BLP?

I have run into such a case, and said that such an agreement is not the same as "establishing consensus" but one of my many fans has gone around posting that this is an absurd position for me to take, thus I present it here as a query.

Under what circumstances should an agreement between two editors to be regarded as any sort of consensus, especially where far more than two editors have opposed the consensus by rejecting it in edits on the article? Collect (talk) 06:18, 29 July 2014 (UTC)

• Two editors may presume a consensus subject to evidence otherwise, according to WP:Silence. One editor acting alone can also do this. I would hope that the editors have made some effort to gather evidence of the opinions of others recorded previously.

The two editors may not refuse to acknowledge a later objection by a third editor on the bases that they decided consensus previously. --SmokeyJoe (talk) 07:26, 29 July 2014 (UTC)

There were six other editors who objected to the edit in the case at hand. I think "agreement by silence" is not the proper interpretation. When the objections were given on the article talk page within days of the agreement, I think that claim would likely fail? There was no attempt to ascertain views of others AFAICT, and no section title on the talk page indicating that the agreement was going to be asserted to be stare decisis. I think that "consensus" should only be claimed where an attempt to obtain consensus has been openly sought - which would be a slight change? WP:SILENCE is an essay and of nugatory value IMO. It does however include Similarly, in the presence of a revert, there is neither silence nor consensus. which would seem dispositive even for the essay that as soon as a revert is made, consensus can not be asserted. Collect (talk) 11:54, 29 July 2014 (UTC)

We always have to remember that consensus can change... Two editors can indeed reach a consensus, but two people do not make a very strong consensus. The consensus they reach only lasts until someone comes by and questions it. At which point we need to re-open the discussion and either reaffirm the previous consensus, or establish a new consensus. Blueboar (talk) 12:21, 29 July 2014 (UTC)

My observation is that people who like to assert the existence of a consensus tend to be dogged, combative, and not partial to working towards a consensus. --SmokeyJoe (talk) 12:40, 29 July 2014 (UTC)

The appropriateness of "asserting" consensus really depends on how strong the consensus being "asserted" was. The stronger the consensus was (determined by how many people were involved in reaching it, and how many support it) the more we can say that the consensus is unlikely to change... and the more the consensus can be "asserted". For example: when we point to a policy provision to object to an edit, what we are doing is "asserting" consensus (that policy)... policy is, after all, nothing more than very strongly supported consensus that is extremely unlikely to change. Blueboar (talk) 13:19, 29 July 2014 (UTC)

I do a lot of dispute resolution work and have to look at questions like this fairly often. Let me note here that the agreement which asserts the two-person consensus was made in this edit on July 4, with a corresponding edit in the article that same day. The next edit to challenge it was, I think, made in the article text in this edit on July 18 and on the talk page in this edit on July 25. Unless there were some prior edits that I've missed that's 14 days between consensus assertion and objection-by-editing. In a "hot" article such as Donald Trump (i.e. one which gets a lot of editor attention), I have to say that I think that in ordinary circumstances it would be a very close call whether or not the subsequent objections were sufficient to prevent a consensus from being formed by the prior agreement and edit. However, this is a BLP article, it was a clearly controversial edit, and the quality of both sources and consensus (see the second bullet point of the WP:No consensus section of this policy) are of importance. On the whole, I either think that either there was no consensus formed due to the subsequent objection or that the quality of the consensus (and the interpretation of the source, though that's a different issue) should keep the disputed edit in play without further bickering over whether or not there was a consensus. Regards, TransporterMan (TALK) 14:56, 29 July 2014 (UTC) PS #1: Unusually, I disagree with Blueboar in this instance. A consensus is a consensus. However weak it may be, it cannot be set aside merely by objecting to it. A new consensus must be formed to set aside the old consensus. An objection to the prior consensus may be enough to form a new consensus, of course, if no one objects to the objection. PS #2: I strongly object to the idea that any sort of formality should have to be observed in order to form consensus. I could say more about that, but I don't think that's really necessary to resolution of the specific problem here. TM — 14:56, 29 July 2014 (UTC)

I don't think we actually disagree. I am certainly not saying that a previous consensus gets "set aside" merely because someone objected to it. I am saying that when someone objects, the previous consensus goes "on hold"... pending either a reconfirmation or the formation of a new consensus. Blueboar (talk) 02:15, 2 August 2014 (UTC)

Level? or Width?

Is consensus really something that is best described in terms of "Levels"? I am wondering if we should shift the language a bit, so we talk about consensus in terms of width and strength of consensus ... a narrow consensus (reached by a small group of editors) is not very strong and is easily challenged... a wider consensus (made by lots of editors) is much stronger and not as easily challenged. The reason why Policy/guideline pages are so difficult to challenge and change is that they represent a very very wide consensus (one that is Wikiwide). Blueboar (talk) 01:56, 2 August 2014 (UTC)

• Adjectives limiting the meaning of consensus (level, width, breadth, strength, long-standing, established, tested) tend to be used to attack/defend rules without speaking directly to the issue. They also allude to the notion that consensus is objectively measurable. They entirely misunderstand what consensus means, which concerns a negotiated outcome that takes into account all objections. I suggest that a consensus is best measured by how easily objections/challenges/questions are answered; and if answers are without substance, such as "it is the rule because this page says so and this page represents community consensus", then more work is needed. --SmokeyJoe (talk) 03:58, 15 September 2014 (UTC)

I have long had a problem with the concept of "local" consensus... I understand why we have this provision (one or two editors at a local page should not simply set aside wiki wide consensus)... and mostly I agree.

But the problem I have with the current language is that our phrasing pegs the "legitimacy" of a consensus on the page at which a consensus was reached... and that does not necessarily mean anything. For example, a wiki wide RfC (reflecting a very wide consensus of the community) might be held on a "local" page. ie the location of the consensus is less important than the number of people who participate and support it. I am groping for words to make this clearer. Blueboar (talk) 23:42, 15 September 2014 (UTC)

I completely agree. I think the problem is that consensus is not objectively measurable except through the test of time, and adjectives attempting to define precise degrees are doomed to fail. The process turns into a worship of the forms and loss of the principle. I think the answer is "no". --SmokeyJoe (talk) 06:38, 23 September 2014 (UTC)

Expand one sentence

The WP:ADMINSHOP section currently says It doesn't help develop consensus to try different forums in the hope of finding one where you get the answer you want. Until I saw that it was protected, I was going to expand this to something like If your position has been rejected in one forum, it doesn't help... Another user came to me asking for an administrative action, and I said no, but I also told him to go ask other admins (and said to tell them that I suggested that he ask them) if he thinks that I'm wrong. That's not rejection: that's an admission that I could easily be wrong and that other admins might know better. WP:ADMINSHOP is meant for people who ask the other parent surreptitiously, not people who ask the other parent because the first parent suggested it. With this in mind, would anyone object to my proposed expansion? Nyttend (talk) 05:48, 23 September 2014 (UTC)

• Maybe

However, there are a few "rules" creeping in under "WP:Consensus". I think Wikipedia:Consensus#Consensus-building_pitfalls_and_errors in general is becoming bloated with content about what consensus is not. WP:Consensus should not be a location to point to a rule. I think that WP:ADMINSHOP should be located primarily at Wikipedia:Administrators, and WP:Consensus should point to the information there. --SmokeyJoe (talk) 06:18, 23 September 2014 (UTC)

I was quite surprised to see that WP:ADMINSHOP came here, and I think I agree with you. Nyttend (talk) 12:38, 23 September 2014 (UTC)

Protected edit request on 23 September 2014

197.29.74.250 (talk) 14:53, 23 September 2014 (UTC)

Please don't use this template unless you're making a request. Nyttend (talk) 16:54, 23 September 2014 (UTC)

Returning to the previous consensus

Are there editors here who don't know that in the absence of a consensus the page returns to its last consensus? That is our policy. Rationalobserver, I am happy to discuss this matter with you and come to some kind of agreement. I let you make your case above, but when you denied that your proposal is dry, I thought it was a good time to return to the last consensus until we can hash out the basics. Sorry we are not in agreement. Do you have another proposal? I'm trying to think of a something we can both agree to. (BTW, BWDIKnow's objection was not about my content. His edit summary asked me to make my best proposal. I don't think that's good advice, but it's not an objection to my content, and it's erroneous to revert me for that reason.) --Ring Cinema (talk) 18:37, 17 September 2014 (UTC)

I don't know it, as a matter of principle. The "last consensus" may not have been. The proposal to change it is evidence that it is not. In the case of "no consensus", the consensus wording must be a tempered version, and failing that, removed entirely. --SmokeyJoe (talk) 06:33, 23 September 2014 (UTC)

A proposal to make a change is not evidence the current consensus is not the current consensus. That's false. --Ring Cinema (talk) 12:47, 26 September 2014 (UTC)

I think it is, but we may be applying very different weights to the word "evidence". A proposal for change is weak evidence, and us countered by a reasonable response. The point is that sometimes the status quo may be rejected, but there remains disagreement on what to replace it with. Sometimes, one of the options is better than nothing. --SmokeyJoe (talk) 13:38, 26 September 2014 (UTC)

A proposal is exactly one person's opinion, so it's pure speculation to assert it indicates there is a change in the consensus. If Editor A's text is changed by Editor B and reverted by Editor C, there's just one editor supporting B's draft, and that's B. So, yes, when there is a content dispute the page returns to the last consensus until there's a consensus for a change. As I stated above. --Ring Cinema (talk) 13:16, 1 October 2014 (UTC)

I didn't say "indicates". That would be way too strong, much as is your first sentence. A presumed consensus can be disproven without establishing a positive consensus for something different. --SmokeyJoe (talk) 13:30, 1 October 2014 (UTC)

Points well taken. --Ring Cinema (talk) 15:35, 1 October 2014 (UTC)

Same issue

Current text: "Consensus can be assumed if editors stop responding to talk page discussions, and editors who ignore talk page discussions yet continue to edit in or revert disputed material may be guilty of disruptive editing and incur sanctions."

I would propose the following: "Editors who ignore talk page discussions and reinsert disputed material may be guilty of disruptive editing and incur sanctions. Returning content to the last consensus until discussion is concluded is a more orderly way to proceed." --Ring Cinema (talk) 06:41, 21 October 2014 (UTC)

No. Why? We are talking about articles where there never really was such a thing as "last consensus". By "last consensus" you just mean ... not sure what exactly. Last edit made before the revert? Last version two months ago? Year ago? The article in its pristine state as it was when it was first created out of Jimbo's womb? Believe me, I've seen the articles in their state of "last consensus" and they were an embarrassment. The whole point of the Wikipedia is that we can always improve articles and this "no, I managed to get my POV into the article for more than a month, therefore it must stay there forever!" attitude is just reactionary. And completely opposed to the Wikipedia ethos. Volunteer Marek  07:09, 21 October 2014 (UTC)

Wikipedia's Lamest Edit Wars here we come!

Alright, so now we've got no consensus as to what consensus is. And there's a bit of edit warring about it. If you take a step back it's sort of funny. If you take another step back, it's funny but there's obviously a deeper underlying issue here.

But nevermind that. Let's not get deep before we have to. My point is that that particular wording is not just wrong (I think it is, but I understand that others may think otherwise) but it's just a bad piece of writing. As in "it doesn't make sense". As in it's nonsense. As in it read like it was written by an eight year old. It confuses the reader, the editor who comes here for guidance. If *you* want it to reflect the fact that an edit which is lucky to survive, by hook or by crook, for a month or so automatically becomes "consensus", then write that down. Don't mince words. Just say it.

Like I said, the person who put that wording in there didn't actually mean for it to be permanent (that person got indef banned for other stuff, but nevermind). The relevant edit summary (you can go through the edit history yourself) pretty much says that this is just some imperfect wording that could be very much improved upon. But hey, it never got improved on, it just stayed there, saying silly things.

So remove it or reword it. Volunteer Marek  07:03, 21 October 2014 (UTC)

In the absence of a new consensus, the page is returned to the last consensus. Do that first. --Ring Cinema (talk) 07:06, 21 October 2014 (UTC)

Says who? What is "last consensus"? When some obnoxious clown has managed to wear out all the other editors in the discussion?  Volunteer Marek  07:13, 21 October 2014 (UTC)

Start a discussion, folks. Returned to last stable version, so let us work forward on this page instead of on the main page. Cheers. Collect (talk) 09:04, 21 October 2014 (UTC)

There are four different discussions, and an obvious lack of consensus to have ever included this content on a policy page. Edit-warring to include content for which there was never a consensus isn't the way to include the content. Collect, your edit summary is good advice; start a discussion for any changes, and there was never any discussion or consensus for this content in any form. This is a policy page that affects the way editing is done all across Wikipedia, it applies to every discussion; it needs to be done right, not slipped through and persistently reinserted when that addition is challenged. No consensus for the edit? Never was one? Not staying on the policy page. If it really does belong, then a consensus for it shouldn't be difficult, but there has never been a consensus for this material, and a rationale based on an interpretation of an essay is not cause to include content on a core policy page when no consensus for it's inclusion has ever existed. - Aoidh (talk) 09:16, 21 October 2014 (UTC)

Collect restored the policy to how it has been for four years. Why did you edit war by reverting? Johnuniq (talk) 09:34, 21 October 2014 (UTC)

The comment you're responding to answers your question, is that supposed to be rhetorical? - Aoidh (talk) 09:38, 21 October 2014 (UTC)

Tuning out of or avoiding harassment on ongoing talk page discussions is certainly not a !vote in favor of whatever stupidness may be occurring on them. This particular eddy the world of policy pages does not reflect actual practice on Wikipedia. It is too often gamed. Silence does not equal consent, it just means editors are tired of a pointless discussion. The longevity of any policy-wonky page on Wikipedia does not prove anything to the contrary about actual Wikipedia norms. - Wikidemon (talk) 09:44, 21 October 2014 (UTC)

If no one objects to a content change on a page, that actually is how 99% of edits are accepted. There's no place for editors to "agree" or "approve" or "consensify"; instead, they just move on. Perhaps you are thinking of a specific case. --Ring Cinema (talk) 12:40, 22 October 2014 (UTC)

Assuming consensus

I agree with Volunteer Marek: one cannot assume consensus because of silence. Editors can post to a talk page and, if they hear nothing back, can proceed with editing. That does not, however, mean that any sort of "consensus of one" exists. This would simply revert back to the BRD cycle. If a long argument has resulted in one party quitting in frustration we shouldn't construe that withdrawal as a silent assent. In Wikipedia editing, no means no. Chris Troutman (talk) 21:05, 20 October 2014 (UTC)

Thank you. There are actually several problems with that sentence. Here is the sentence:

"This obligation applies to all editors: consensus can be assumed if editors stop responding to talk page discussions, and editors who ignore talk page discussions yet continue to edit in or revert disputed material may be guilty of disruptive editing and incur sanctions."

The first problem is that it's a run on sentence which conflates two completely different things. Editors who ignore talk page discussion and continue to edit war are in a completely different category than editors who come and make a change after there has been no discussion on a talk page for a long time. Yet, putting both of these into a single sentence makes it seem like they're equally bad. At the very least, the sentence is badly written and confusing.

Second, the beginning of that sentence talks of an obligation. "An obligation is a course of action that someone is required to take, whether legal or moral. " [3]. But then it talks about something that "can be assumed". This is grammatically incorrect and logically absurd. And what is it exactly that an editor is obligated to do? Assume silent consensus? Why? The second part of the sentence does not make sense with the first part of the sentence. The way it is written is simply embarrassing.

And then there's the gist of the matter. Whoever wrote this probably has very little experience in actually editing articles. If you've been around Wikipedia then you know that all kinds of junk makes it into all kinds of articles for all kinds of reasons. If somebody vandalizes and article or inserts a hoax into it, then it stays there for several months, does that mean that this edit has "consensus"? Yes, I know that's a bit of an extreme example, but I think it illustrates the point. Wikipedia is suppose to be an ever evolving and ever improving encyclopedia. There's no tyranny of the status quo. Stuff that's already somehow made it into an article is not privileged.

Of course the way that the problem often manifests itself is that on some medium to low traffic articles, one or two obsessive editors can bully and shout down anyone who disagrees. Most people have better things to do with their time than spending it on pointless arguments with nutzoids. So they leave and talk page discussion goes silent, the tendetious editor does a little dance of joy and boom! Their edits have "consensus" no matter how bad they are. Then someone comes along awhile latter, notices that there's problems with the article but now they face an uphill battle because they are "against consensus". Ridiculous. Volunteer Marek  21:16, 20 October 2014 (UTC)

(edit conflict):I guess we disagree. The reason why I think silent consensus is valid, is because consensus is not permanent. Editors can come to an article at any time and form a new consensus. Articles are typically watched by multiple editors. Its reasonable to assume that if nobody objects to a bold edit or an edit proposal of a talk page, then there is tacit agreement. On the other hand, editors should not be allowed to stonewall by reverting a bold edit and then refusing to engage in seeking a new consensus on a talk page. I do however agree that filibustering is not helpful, so I don't object to some tweaking of the wording.- MrX 21:21, 20 October 2014 (UTC)

At the very least, can you at least agree that as is, the sentence is horribly written and doesn't really make sense?  Volunteer Marek  21:23, 20 October 2014 (UTC)

Of course! I don't have any objection to rewording as long as we don't loose the intended meaning.- MrX 21:28, 20 October 2014 (UTC)

Well, my intent is to remove it, because in addition to being badly written, it also is wrong headed. So I'm not the person to ask for a rewording. Volunteer Marek  21:34, 20 October 2014 (UTC)

How about In the absence of discussion on the talk page, editors can proceed until a reversion sparks discussion to determine consensus. Editors who ignore talk page discussions yet continue to edit in or revert disputed material may be guilty of disruptive editing and incur sanctions"? Chris Troutman (talk) 21:39, 20 October 2014 (UTC)

I agree that the current language is a problem. I've been part of many discussions in which some participants do seem to drop out from frustration; and certainly that cannot indicate consensus. Omnedon (talk) 21:51, 20 October 2014 (UTC)

I full heartedly believe that the wording is just fine the way it is. It's representing that editors cannot simply just revert from the sidelines and claim that there's no consensus, and makes clear that if you do just that you will incur sanctions. Besides, your reword changed the meaning from This obligation applies to all editors: consensus can be assumed if editors stop responding to talk page discussions, and editors who ignore talk page discussions yet continue to edit in or revert disputed material may be guilty of disruptive editing and incur sanctions. to In the absence of discussion on the talk page, editors can proceed until a reversion sparks discussion to determine consensus. It's moving the goalpost to saying that if there's any talk page discussion, not just if you stop responding. And it specifically mentions reverts, which I don't believe is a good way to frame this. Also, I don't believe you can just change guidelines or policies just like that, especially since this affects every single article on the site. Tutelary (talk) 21:52, 20 October 2014 (UTC)

To focus on just one aspect of this: you're OK with assuming consensus "if editors stop responding to talk page discussions"? I find that part of the current wording to be questionable and out of touch with reality. Omnedon (talk) 21:59, 20 October 2014 (UTC)

When people stop responding to talk page discussions, that means that they've decided to stop contesting or commenting on the behaviors of the article. And when somebody makes an edit that you agree with, you do the same. If you have a problem with a certain edit on the article, it's your job to contest that edit, and if you don't within a reasonable amount of time, then it can be assumed that yes, no one had a problem with it. I also feel that because I quoted this exact sentence verbatim on my talk page, Varek is trying to get it changed thinking that this is where all the SPAs come for and get their daily dose of knowledge of mediocre policy and guidelines. Tutelary (talk) 22:20, 20 October 2014 (UTC)

This discussion is in-keeping with WP:PGBOLD. An RfC will result should it look like we can come to agreement. As for your point, I don't think the winner of talk page battles should be the party willing to argue ad infinituum. The way the text currently reads, it intones exactly that. My point about reverts was that WP:BRD is the law of the land. Editors should continue to be bold until a reversion brings about a discussion. I don't think SPAs bother reading policy but I don't want this policy to hinder good-faith contributors. If this argument between you and Marek stems from some other conflict, then I was unaware. Chris Troutman (talk) 22:33, 20 October 2014 (UTC)

BRD is not a policy or guideline, and I would advocate against if it were put up for a vote on whether it becomes one. It's a really cited, really useful essay, but it does have exceptions. The policy regarding blatantly unconstructive and moving past still when consensus has already come to a head multiple times is WP:IDHT, not the consensus policy. Regardless, it doesn't mean that. It means that if you step away from the article for a good amount of time, and stop responding to talk page requests or discussion, that can be assumed as silent consensus, especially if you're editing other articles or doing other activities. (But I wouldn't want that in policy, but that's just one reason why.) A page on Wikipedia shouldn't be hindered because someone wanted to stop or cease discussion on the talk page. Also, if this were to be changed, people can come back months later and say 'See my talk page post? You still have no consensus for that' even though it was assumed since he stopped replying and stopped responding to the talk page that he was fine with it, when previously he'd been quite staunch at opposing it. Editors should continue to be bold until a reversion brings about a discussion. This is one example of good intentions, but would be bad played out. What about 0RR or 1RR where you can't necessarily revert something? What about administrative pages where you absolutely can't contest the changes? I believe it's well intentioned to remove or rephrase the bit for scope, but its wording does well now does itself do justice. We can't worry about what editors aren't doing, only what they are doing. The Varek thing you can see on my talk page. I quoted this specific sentence verbatim and here Varek comes to remove it from the policy. Tutelary (talk) 22:44, 20 October 2014 (UTC)

Actually, Tutelary, no. If an editor stops responding to talk page discussions, we don't necessarily know why. If s/he has clearly expressed an opinion, then stops expressing that same opinion, that doesn't mean that consensus magically comes into existence. Omnedon (talk) 22:36, 20 October 2014 (UTC)

I do like Chris troutman's wording much better.

" It's representing that editors cannot simply just revert from the sidelines" - uhhh... like this?  Volunteer Marek  21:58, 20 October 2014 (UTC)

• This statement: "consensus can be assumed if editors stop responding to talk page discussions" is quotable nonsense. While maybe not entirely backwards, it is very poorly written and does not belong in policy as written. --SmokeyJoe (talk) 23:46, 20 October 2014 (UTC)

May I respectfully ask how you consider it to be 'quotable nonsense' and to be 'poorly written'? Tutelary (talk) 23:52, 20 October 2014 (UTC)

It is quotable, as I demonstrate. In isolation, it appears to mean something. It is nonsense because it says that if you never give up bludgeoning, eventually you win. The text is appallingly unclear on how to balance WP:AGF and WP:STICK, and it is not the place to attempt it. WP:Silence, which is not just and essay, makes it clear that silence is the weakest form of consensus. My silence to date does not stop me from correcting anything. This sentence in particular unbalances the notions of WP:Silence.

The section "Consensus-building in talk pages" is suffering scope-creep. Paragraphs 3 & 4 especially. There is a perfectly good guideline at Wikipedia:Talk page guidelines, which is where paragraphs 3 & 4 belong. --SmokeyJoe (talk) 00:15, 21 October 2014 (UTC)

I really think you're looking too much into it, SmokeyJoe. Especially regarding what it says and what it doesn't say. It -does- say that consensus can be assumed if editors stop responding to talk page messages. There's a good reason for this to be a policy, because we don't want to have to email or off wiki message somebody to really see if they're still opposing something. When somebody staunchly opposes something, uses the talk page like a hawk, constantly uses a club to respond, and everybody's against him? Eventually stops responding, and the editors on the article edit the article according to consensus, even if it was just the dissenting editor stop replying. That's what's this supposed to do. Now, to the point of The text is appallingly unclear on how to balance WP:AGF and WP:STICK It doesn't need to. The guideline and the essay don't need to be covered in the policy of consensus. It's appropriately covered in their aforementioned articles. No need to insert some sort of specific clarification on here when it really isn't warranted. WP:Silence, which is not just and essay, makes it clear that silence is the weakest form of consensus. This is something I really want to comment about, because I reverted you for it. WP:Silence -is- an essay, to which no editor has any obligation to follow, under any circumstances. This makes it pretty much irrelevant for this discussion, since we're discussing the policy of Consensus. Why I reverted is because mentioning WP:SILENCE and trying to say 'it can just be used for WP:SILENCE' when you're removing something from a policy can be entirely destructive, because people are not obligated to follow essays, only policies, and you're effectively changing how editors are supposed to do things, since they look for guidelines and policies for guidance. Also, WP:SILENCE could be representing an extreme fringe group of people, and was not argued, discussed, or at all. I can create an essay saying that anybody who disagrees with me would be wrong. I'd be in essence wrong but I could do it, and I'd be emphasizing how low quality essays are in terms of editor obligation to follow them; none. My silence to date does not stop me from correcting anything. This sentence in particular unbalances the notions of WP:Silence. It doesn't need to, since WP:SILENCE has not had an RfC to determine content and does not have site wide implications if edited. It doesn't need to affect or employ an essay. Tutelary (talk) 00:26, 21 October 2014 (UTC)

The problem in the selection under discussion is the implication that during a content dispute -- presumably during the formation of a new consensus -- reverting would be considered a bad thing. Returning to the last consensus is the correct course of action while discussion is underway to form a new consensus, so why would that be disruptive? The disruptive editor would seem to be the one insisting on a change before there is a consensus for a change. --Ring Cinema (talk) 22:06, 20 October 2014 (UTC)

Ok. Putting aside the substantial question of whether "status quo=consensus" (which I think is a ridiculous idea, but anyway) can we at least agree that the current wording is just horribly written, and doesn't make sense? It talks about an "obligation" and then fails to specify what the obligation is, and even switches to a "you can if you want to" kind of wording. If you track down the history of how this happened, you'll find that even the person who actually put that wording in wasn't happy with it and it was meant as a temporary wording "place holder" kind of thing, meant to be rewritten. Then that person got busy with other stuff and eventually ended up getting indef banned (lol!), although for, AFAICT, unrelated issues. In a way, the fact that this awful wording persist (for so long) in this page is a perfect example of why "no discussion" does not equal "consensus". And how hard it is to change something - which is blatantly incorrect or faulty - once it has "baked itself" into the text. Volunteer Marek  04:51, 21 October 2014 (UTC)

We can put aside the non-question of "status quo=consensus" because there's no coherent alternative for our purposes. Yes, there is a problem with the selection but it can't be "fixed" by restating the mistake I mentioned. A new draft shouldn't imply that returning to the last consensus during a content dispute is disruptive. In fact, that's the only sensible thing to do. --Ring Cinema (talk) 05:16, 21 October 2014 (UTC)

Assuming you mean "cannot" rather than "can", the point is that as written the sentence is just badly written, is poor grammar, and doesn't make sense. Hence, whatever the intention behind it, it should be removed until the intended meaning can be clarified. IF the editors who support that particular interpretation of CONSENSUS want it back in, then they need to write that clause in a way which is not incomprehensible. So I'm removing it again for now. Just based purely on stylistic merits. Volunteer Marek  06:15, 21 October 2014 (UTC)

No, I mean "can"; there's no alternative. Perhaps this is not such a clear case as you imply, given that there are differences of opinion about it. Assuming the facts as you state them, it might be a good argument against inserting "place holders" rather than an exception to the sensible notion that a consensus is succeeded by a new consensus. --Ring Cinema (talk) 06:32, 21 October 2014 (UTC)

I'm not sure why or when this text was added, but it's absurd. That wording suggests that consensus is when you keep replying until you are the last comment, and that is what can decide consensus; it is not. The wording contradicts the spirit if not the actual wording of the rest of WP:CONSENSUS, and has no purpose in the policy page. Was there a previous consensus that determined that it belonged? If not, it needs to stay out until a consensus determined it belongs. - Aoidh (talk) 06:51, 21 October 2014 (UTC)

OK... here is my take on the issue: We are really discussing two different things... 1) "silence" vs 2) "lack of continued discussion".

1. Silence is when a edit is made (or a proposal to change is made on the talk page)... and no one reverts or states an objection. In such a case, we can assume that (after a reasonable amount of time has passed), continued silence = consensus in favor of the edit/proposal.
2. Lack of continued discussion on the other hand, occurs when a discussion has started... everyone has said what they want to say... and the discussion dies out. A "lack of continued discussion", on its own, tells you nothing about consensus... Instead, you have to examine the discussion and determine why the discussion has ended. Indeed, a "Lack of continued discussion" often indicates that reaching a consensus is not possible. It usually occurs when people feel that they have argued enough and that further discussion is pointless... they may feel that they have clearly stated their views (for or against), and since these views will not be changed by further argument, there is no point in continuing to say the same thing over and over again. Blueboar (talk) 13:25, 21 October 2014 (UTC)

My two cents on the particular language in dispute: I think the problem is more in what people are reading into the language than what it actually says. If you read this language to mean that the last survivor of a talk page discussion "wins," you're reading a meaning into the language which it doesn't have or say. It says "consensus can be assumed if editors stop responding to talk page discussions" (emphasis added), but assumed for what purpose and to what end? To me, all this means is that if you're the last man standing that you can assume that you've at least got enough consensus that you can't be blocked or banned for disruptive editing if you then edit the article text in accordance with your talk page position. It doesn't mean that if when you make that edit one of the other participants in the talk page discussion (or some new editor) cannot revert you because you have consensus. Such a reversion simply means that your assumption — which was only an assumption — turned out to be wrong and your desired edit is still in dispute. That reading is supported by the end of the sentence which then goes on to say what can happen to editors who continue to insert or revert material which is still in dispute. The problem with this sentence is that the "and" in its middle should really be a "but" to make the meaning clearer: "This obligation applies to all editors: consensus can be assumed if editors stop responding to talk page discussions, but and editors who ignore talk page discussions yet continue to edit in or revert disputed material may be guilty of disruptive editing and incur sanctions." Regards, TransporterMan (TALK) 18:27, 21 October 2014 (UTC)

Re: "... if you're the last man standing that you can assume that you've at least got enough consensus that you can't be blocked or banned for disruptive editing if you then edit the article text in accordance with your talk page position." - not necessarily. The "last man standing" could have a serious case of "WP:IDIDN'THEARTHAT"... so convinced by his/her own arguments that he/she continues to push and push for a change when discussion shows that the change clearly does not have consensus (or, alternatively, repeatedly opposing a change that discussion shows clearly does have consensus). This is why you have to actually read the discussion and examine why there is a lack of continued discussion.

Being the last man standing really says nothing about whether your view has consensus (or not)... nor does it tell you whether boldly acting upon your view would be disruptive (or not). Being the "last man standing" is not some sort of "get out of jail free" card. What determines consensus (and the level of disruption) is the entire discussion... and the views stated by of each participants - as of when they stopped contributing to the discussion. In a heated debate, silence rarely indicate acceptance. It usually is an indication that everyone is simply fed up with repeating the same arguments over and over again... and have decided to no longer "feed the troll." Blueboar (talk) 15:20, 22 October 2014 (UTC)

I agree, but that's exactly one of the situations which is covered by the second part of the sentence (the part after the "and" which I think should be a "but," but which has the same meaning either way, just being clearer with the but). However, I probably should have said, more precisely, "you can't ordinarily be blocked or banned". (But if I may cavil for a moment, I can't imagine a case of IDHT which could be so bad that it results in the discussion dwindling away, but which would result in a block or ban after a single edit without warning and without some sort of prior churning — e.g. ANI or an admin's involvement — away from the article talk page upon which a charge of disruptive editing would be based rather than just being based only upon on the IDHT behavior at the article talk page discussion, but perhaps I'm just naive.) Best regards, TransporterMan (TALK) 15:54, 22 October 2014 (UTC)

Returning to the last consensus

Are there editors here who don't know that in the absence of a consensus the page returns to its last consensus? That is the policy. Presumably if you are going to edit here you know the section on No Consensus. If you haven't read it, this would be a good time. No consensus means no change, for obvious reasons. --Ring Cinema (talk) 06:48, 21 October 2014 (UTC)

Can you provide a link to this consensus, where was this discussed and a consensus formed that this content belongs? This is a policy page; the content doesn't belong until a consensus shows otherwise, being slipped in unnoticed at a prior point does not make a consensus. - Aoidh (talk) 06:53, 21 October 2014 (UTC)

You don't know what you're talking about. The paragraph has been part of the policy for a long time so it's hardly slipped in. Go ahead and follow the policy: return the page to the last consensus. --Ring Cinema (talk) 06:59, 21 October 2014 (UTC)

The whole point is that "has been part of the policy for a long time" is a pretty stupid argument. It's a logical fallacy in fact, but it's so stupid that it doesn't even get to have its own (Latin) name. Volunteer Marek  07:05, 21 October 2014 (UTC)

Sorry, you're badly mistaken. You don't know the policy you are attempting to edit. Let's see a demonstration by you of your ability to follow the policy. After that, it might be useful to hear what you have to say about the policy itself. First, return the page to the last consensus. --Ring Cinema (talk) 07:09, 21 October 2014 (UTC)

Yes, I do know the policy I'm attempting to edit. You, you seem to be using the Chewbacca defense. And that's being gracious. Volunteer Marek  07:12, 21 October 2014 (UTC)

Based on your behavior, you don't know about No Consensus. --Ring Cinema (talk) 14:36, 21 October 2014 (UTC)

You're getting pretty low on the pyramid there. Try addressing what is being said, instead of saying "you don't know what you're talking about" to those that disagree with you as if that means anything. - Aoidh (talk) 07:13, 21 October 2014 (UTC)

It seems I do, actually. Provide a link to the discussion, or I can only assume that there wasn't one, because I wasn't able to find one. Was it discussed or noticed in any way? When was it inserted? If you cannot answer these questions, that's not much of a claim for a consensus, especially on a policy page. Not noticing an edit is an assumed consensus, not an actual consensus that holds any weight when actually challenged. - Aoidh (talk) 07:13, 21 October 2014 (UTC)

The shoe is on the other foot. If you think this is the exceptional case where the text in the policy -- yes, this is a policy -- wasn't arrived at via consensus, document it. Until then, you're just speculating. I look forward to your documentation. --Ring Cinema (talk) 14:36, 21 October 2014 (UTC)

You're asking me to prove a negative? I cannot document something that was never written, if you're claiming there is a consensus, the onus is on you to prove that, it is not the job of others to disprove your unfounded claims. You say there is a consensus; prove it. If you're right, that shouldn't be difficult. - Aoidh (talk) 00:43, 22 October 2014 (UTC)

The text was placed in in this edit [4]. Even the person putting it in there wrote in their edit summary "not entirely satisfied with the result". There's also discussion in the talk archives about removing any links to WP:SILENCE since that's an essay not policy and shouldn't be in here. Volunteer Marek  19:10, 21 October 2014 (UTC)

Marek, are you aware that the Policy policy (not a typo) says (here: see the "Or be bold" sub-subsection) that while it's not best practice that it's at least an acceptable method of changing policies and guidelines to just edit and see if it sticks without any prior talk page discussion? If you jump up a couple of paragraphs, the second paragraph of the "Content changes" section has some language which suggests that such a change may not have "immediate effect" but that, in and of itself, implies that if it sticks long enough without being changed that it's part of policy. Regards, TransporterMan (TALK) 21:46, 21 October 2014 (UTC)

For the record, an edit that has gone unchanged for four years on a page that is frequently discussed is clearly the result of a well-accepted consensus. That an editor expressed the opinion that he could do better means nothing four years after. --Ring Cinema (talk) 04:49, 22 October 2014 (UTC)

You have a different definition of "clearly" than others then. Wikipedia consensus is that a "silent consensus" is assumed, and doesn't exist if objections are raised. That's as far removed from "well-accepted consensus" as one can get. The edit that the text was introduced in was massive, changing a huge chunk of the policy page; it's not surprising that it wouldn't be noticed amongst all the other changes. That isn't a consensus, not "clearly so" nor "well-accepted". An example of a "well-accepted consensus" is one that is discussed with a clear consensus outcome. That you are of the opinion that this makes a consensus is what means nothing. Per this policy page, Wikipedia has a higher standard of participation and consensus for changes to policies and guidelines than to other types of pages. and "well nobody saw or if they did they didn't say anything" does not even come close to meeting that standard. - Aoidh (talk) 05:39, 22 October 2014 (UTC)

I said nothing about a silent consensus on this passage. Please cite the other cases where content on a policy page goes unchanged for four years and in your view isn't the consensus. Thanks. --Ring Cinema (talk) 06:06, 22 October 2014 (UTC)

Then if you're talking about an established consensus and not a silent one, show a diff or a link to the discussion which established this consensus. If you cannot do this, there is no such consensus. If doesn't have to be "in my view", that's the policy as written. If you want to argue against policy you're welcome to do so, but policies have a higher standard for consensus and "oh nobody saw it" doesn't meet that. I'm assuming, since you seem quite adamant that there is a consensus, that you have a diff to back up your claim. Where is it? - Aoidh (talk) 07:46, 22 October 2014 (UTC)

So there is no evidence to support your view? Let us know if you find any cases where a policy unchanged for four years was not a consensus. Until then, please follow this policy scrupulously. Thanks. --Ring Cinema (talk) 12:32, 22 October 2014 (UTC)

...are you seriously asking if there's evidence that no discussion happened? I cannot prove a negative, but your continued failure to show any substance to your claims says all that needs to be said about the merit of your comments; there was never a consensus for this material. Don't ask others to follow this policy if you are unable to read what it says; this policy contradicts your claims, as does your refusal to address the very simple question of where this supposed consensus was given. - Aoidh (talk) 17:20, 22 October 2014 (UTC)

I agree with Ring Cinema, but I do want to say that this question — whether bold edits to policy without discussion are valid — is an issue which has been discussed for years. If you dig back through older versions of this policy you'll find that the question of consensus by silence in general and the separate issue of consensus for policy change by silence have fairly continually been formulated and re-formulated. For example (and only as an example) from March 6, 2009 until September 3, 2010 this policy said that while no-discussion edits could change policy they could do so only "if adequately exposed to the community." The latter edit changed that only-if restriction to what we now have in the CONLIMITED section which only says that policy changes "may be made without prior discussion, but they are subject to a high level of scrutiny. The community is more likely to accept edits to policy if they are made slowly and conservatively, with active efforts to seek out input and agreement from others." The problem is, frankly, a practical one. Policies here, like articles here, are formed through an accretion of fully-discussed changes, (which are in the minority), changes which are proposed on the talk page with no objections being raised, changes which are discussed on the talk page without specific agreement but which are edited in without objection, and changes which are simply made without discussion but which aren't reverted. If we go picking at them based upon how they were made then Wikipedia is going to become more of a lawyer's dream than it is already. Every attempt to change text or policy can be challenged by somebody digging back through the edit history and discussion history and proclaiming "Ahah, was no discussion!" The next step from there is,

• "There was discussion, but everyone didn't sign off on the result." or
• "There was discussion, and everyone signed off, but the discussion was lousy and missed the point." or
• "There was discussion and agreement but only two editors were involved (or three or four or pick your number) and that's not enough to form a real consensus." or
• "Okay, so it takes five editors to form a real consensus and there were six involved here, but only two of them 'really' took part in the discussion." or
• A million other variations on the theme of there was no discussion or the discussion wasn't good enough or was flawed.

Considering the complexity of the histories and discussions on some articles and policies, if we enable the challenge of article or policy text on the basis of whether or not it was adopted by consensus, someone might make a supplementary income hiring themselves out as a Wikipedia-history researcher. The real problem here is the failure to recognize that consensus is a means of resolving disputes and discussions, not a status or shield. Once the text of an article or policy has stabilized, as indicated by anyone's failure to revert or object to the last edit for a period of time long enough to indicate that the current discussion, if any, has died away, then consensus is needed to change the text if anyone objects to or reverts the next edit or proposed edit. That's what consensus is for, to decide how discussions are to be resolved, not to guard or insulate prior edits whether in articles or policy. (There are one or two very narrow exceptions to that concept but it would extend this already tldr comment to mention them here.) Regards, TransporterMan (TALK) 14:50, 22 October 2014 (UTC)

Protected

Since protection ended on the 18th, nothing has been done here but revert the same change 6 times. On the consensus page, the irony... I have reapplied full protection here for three months (up from the previous one month). I protected the page in the "old" version, I haven't looked at the change and have no interest one way or another. But discuss it here, get a clear consensus, and then make an edit request. No other option seems possible. Really... Fram (talk) 09:51, 21 October 2014 (UTC)

Wow, this is "textbook". Volunteer Marek  16:39, 21 October 2014 (UTC)

Fram, you are mistaken. Two different bold proposals were offered and both correctly reverted for lack of consensus. Some insisted their proposal should be policy before they got a consensus for it, which is a frightening ignorance of how Consensus works for someone who wants to contribute here. --Ring Cinema (talk) 12:36, 22 October 2014 (UTC)

With "here", I meant on the article, not on the talk page. My apologies for the confusion that understandabley caused. Fram (talk) 13:30, 24 October 2014 (UTC)

completely new proposal

Merely outlasting others does not mean one can assert "consensus." Consensus requires actual agreement on a compromise, not simply driving others off a page into silence. Unless one has achieved some sort of agreement on the article talk page, one should be wary of announcing a "win" on any issue, and acting in that manner may be viewed as "disruptive" to the collegial process.

Appears to me to be in the spirit of encouraging actual compromise about the content of any article. Collect (talk) 15:35, 22 October 2014 (UTC)

I oppose, not so much because of your first sentence (though I'd like for it to say "does not necessarily mean"), but the rest. Example: Editor X makes an edit, Editor Y reverts with the objection on the talk page that while the edit would seem to be otherwise appropriate, the sources are not reliable. Y asserts reasons on the talk page why they are reliable, and Y drops out of the discussion. From an objective point of view, the sources are clearly reliable and Y's objections are not supported by policy. The second part of your proposal would suggest that X doesn't have consensus for his edit because X and Y did not come to an agreement, actual or otherwise. That's clearly the wrong result. X does have consensus because his arguments were better than Y's. The fact is that consensus does not require actual agreement, even at the end of a discussion or dispute (heck, it doesn't even require a discussion at all), and while that's easiest to see in the two-editor situation, it also applies in the multi-editor one. Regards, TransporterMan (TALK) 16:26, 22 October 2014 (UTC)

Well... the outcome of your scenario really depends on whether the sources actually are reliable or not. X may be positive that they are, while Y may be just as positive that they are not. What is usually needed is a third party opinion to break the stalemate. Besides... even if X and Y reach a compromise ... a consensus of two is an extremely weak consensus. In fact, I would hesitate to call it a consensus at all. It certainly should not be held up as "the consensus" if editor Z comes along and questions it. Blueboar (talk) 15:19, 24 October 2014 (UTC)

Also oppose as the language is too contentious and tends to contradict the 'silent consensus' view discussed in the prior sections. I sort of agree with the last sentence, but I don't think it belongs in a policy document.- MrX 17:08, 22 October 2014 (UTC)

What about simply saying: Merely outlasting others does not mean one can assert "consensus." (see: WP:IDIDN'THEARTHAT). Blueboar (talk) 13:41, 24 October 2014 (UTC)

I was trying as best I could to accommodate the positions of as many editors given above as feasible. Collect (talk) 14:19, 24 October 2014 (UTC)

• This policy should not prescribe things better covered by WP:DR. A sustained and supported objection implies no consensus. Editing contrary to explicit discussion on the talk page is disruption and best described under that policy. Including everything that concensus is not is to bloat the document and to create overlap with other better attended policies. --SmokeyJoe (talk) 22:51, 24 October 2014 (UTC)
• I support User:Collect's suggestion, but even further, please read Collect's proposal on steroids below:

• I like the spirit of Collect's proposal but Blueboar's is probably more "vanilla" (not a bad thing) while still capturing the essence. Volunteer Marek  05:56, 26 October 2014 (UTC)

Collect's proposal on steroids

I just wasted about two weeks time, and got Jimbo involved because of a certain misunderstanding we had about the meaning of the word "consensus". In the world of professional group-facilitation, and as even defined in our own article on Consensus, it turns out that the word "consensus" has only some similarities with Jimbo's definition of, "partnership between interested parties working positively for a common goal." I am wondering if when Jimbo first specified back "centuries ago" that Wikipedia editing would be based on "consensus", if he yet knew the deep and rich history of this word. I'm a Quaker, and we have been using the "professional group-facilitation" understanding of the word for centuries. With Quakers, it's kind of like the Entmoot in Lord of the Rings. We sit around and debate, until finally we all get tired and even those who might still disagree, at least agree to be silent. The discussion process is usually led by a facilitator who is hopefully neutral and who is responsible to try to "synthesize unity" throughout the discussion. This person is also responsible to give all views a fair airing and to try to point out unreasonable behavior. Then what is left is described as "consensus", and acted upon. With Wikipedia as it is now, if Quakers were to try it, it would be like telling everyone in the Quaker Meeting, "go ahead and do whatever you want, then we can all come back and talk about it later, or hack apart whatever anyone person might think any other person did wrong." Obviously if that were the way the Quakers operated, nobody would have ever heard of them!

Yes, I support User:Collect's suggestion, but still even further, I am thinking that it might even be better if, should any editor strongly disagree with another editor's recent edit, then that editor in disagreement should have the right to revert the contested edit, and neither of the two or more editors in disagreement should be allowed to post an edit on that given point until they all reach a "professional-group-facilitation" style consensus on the article's talk page. Perhaps silence by an editor for a week, might be equated to a vote of no-contest, but if argument still persists after a week then an Rfc could be made, requesting a neutral facilitator. Perhaps that would not only act to calm the perpetual edit-wars, but would also result in more stable and well reasoned articles. Scott P. (talk) This comment first edited at 22:51, 25 October 2014 (UTC), last edited at 06:18, 26 October 2014 (UTC)

May I suggest you take a look at the proposal I made above. "Editors who ignore talk page discussions and reinsert disputed material may be guilty of disruptive editing and incur sanctions. Returning content to the last consensus until discussion is concluded is a more orderly way to proceed." I don't think it's controversial to advise discussion first and changing second when there is a content dispute. --Ring Cinema (talk) 03:21, 26 October 2014 (UTC)

Yes, I think Blue boar has it right. He definitely has the spirit of it right. But that question does assist. Wording should be added to the procedure clarifying especially that a specific verbal consensus should not be "stealth-edited" out at a later date, and that a an explicit verbal consensus is far stronger than the mere implicit consensus that forms automatically when nobody challenges and edit over time.

I guess I sort of jumped in here without yet reading all of the most recent stuff. Sorry about that. Regarding your suggestion Ring, I agree with the spirit of it. I also see some of the other user's points. Here is my compromise suggestion:

"Consensus can be assumed if editors stop responding to talk page discussions for one week or more, or if no objections are made for one week or more. The preferred procedure for treating a disagreement with a recent edit is to return the concerned content to the state of its last consensus until a mutually acceptable agreement can be reached. If after one week, a consensus has not yet been achieved, an Rfc should be placed before making the proposed change. Editors who come via an Rfc should attempt to find harmony, rather than to take sides. If still no consensus can be found, then the last edit with consensus should remain. Any editors who might not comply with these procedures for consensus may be guilty of disruptive editing and incur sanctions."

I put the "one week" in there for clarity's sake. The treatment of silence here is roughly modeled on the Quaker consensus process. With the Quaker model, consensus can still be reached if a participant declares that they "Stand Aside", which is their registering of their personal objection, yet their willingness to let the process proceed via consent by silence. The Wikipedia equivalent of a Quaker "Blocker" would be the editor who is willing to at least post a single well reasoned and thoughtful objection every 6 days. I think that the Quaker model would work well in Wikipedia because as an encyclopedia, it is probably generally better to "err on the side of caution", and allow those who properly block change to have a slight advantage. Sure, this may be a little too adventurous for now, but vague speculations are still welcome. Scott P. (talk) This comment first edited at 06:09, 26 October 2014 (UTC), last edited at 13:36, 26 October 2014 (UTC)

The opening sentence is still is problematic... Again, there are three scenarios where discussion can die out:

1. A consensus has formed in favor of the proposed change (but one or two editors won't accept that fact)
2. A consensus has formed against the proposed change (but one or two editors won't accept that fact)
3. No consensus is likely to form (and everyone has become tired of arguing)

Yes, the fact that editors stop responding to talk page discussions is often an indication that a consensus has formed... but it does not indicate what that consensus actually is. And sometimes a lack of continued discussion is an indication that a consensus is not likely.

I think we need to limit the concept of "silent consensus" to situations where there is no discussion at all (Bold edits that are not reverted... proposals on the talk page that are not responded to at all). Once a discussion starts, however, we can not assume a silent consensus in favor of an edit should the discussion die out. Blueboar (talk) 14:03, 26 October 2014 (UTC)

The reason that I feel that some form of silent-consent after an objection would be necessary, would be because if not, then basically any "obstructionist" could easily over-rule any other editor by placing a single objection. It seems to me that such would give the "obstructionist" mentality a little bit too much power. What if an "obstructionist" simply posted the word "No", without any rationale? The Quaker "balance of power" lies slightly with the status quo, but not that much. Those who oppose must reasonably state their positions, and can not simply state their positions then leave the meeting. They must remain present and ready to answer any questions about their position until the end of the Meeting, just like everyone else. Scott P. (talk) 14:12, 26 October 2014 (UTC)

I understand your concern... but consider the situation where the "obstructionist" has reasonably stated his/her position, and the editor who desires the change simply refuses to listen to his/her concerns. The fact that the "obstructionist" subsequently walks away from the discussion in frustration does not mean there is a silent consensus in favor of the edit.

That said... I question whether "consensus" is really the right word to use when only two people are involved in a discussion. You might end up with "agreement", but I would hesitate to call that agreement a "consensus". Consensus requires multiple editors. Blueboar (talk) 15:11, 26 October 2014 (UTC)

In a Quaker Meeting, if the person proposing a course of action does not show at the meeting, the proposed course of action is shelved until they can attend a meeting to discuss it. Similarly, if the person proposing an edit doesn't show up for the "agreement/ consensus process" their edit is shelved and the pre-edit version retained until they can. That seems to make sense, no? Scott P. (talk) 15:43, 26 October 2014 (UTC)

Looks like a recipe for fringe and pov editors to be disruptive and keep their positions in an article. How will this avoid that? Dougweller (talk) 15:50, 26 October 2014 (UTC)

Scott... I'm not talking about the situation where those for (or against) a course of action don't show up at the meeting. My concern is the scenario where everyone has shown up for the meeting... and after discussing the issue for hours and hours, there is still disagreement. I am talking about the scenario where everyone has discussed their views fully (often multiple times)... and eventually people start to get up and leave the meeting in frustration. You can not assume a consensus because those people have walked out. Blueboar (talk) 16:20, 26 October 2014 (UTC)

(This comment out of time sequence, was sequentially below.) If a better length of time, such as two or three days for the wait period might be found to be better, that could certainly be substituted for the 7 day wait period I first proposed above. Scott P. (talk) 17:31, 26 October 2014 (UTC)

By the rules above, the older edit would remain only until either:

1. Agreement was reached for a change, either amongst the original discussion participants or amongst them plus an Rfc editor who is not supposed to take sides, or
2. The older edit was found to be in clear violation of some other WP that had been "missed earlier", which the newer edit fixed, or
3. The opposing editor(s) all disappeared for 7 days or more, especially the 7th day (which to me would seem quite unlikely if the opposition had good cause).

Then the newer edit would take its place.

Or conversely the older edit would be retained even after the attempted consensus if:

1. No agreement can be reached after an Rfc, or
2. The newer edit was found to be in clear violation of any WP, or
3. The proposing editor disappeared for 7 days or more, especially for the 7th day,

Then the older edit would remain in place after the process had ended.

Regarding the types of bad edits suggested by Doug just above, if implimented thoroughly, the bad edits would be generally blocked out at the outset by those who object. If bad edits are not objected to in a timely fashion, then it would be no different than it already is, more or less, they are presumed to have "consensus", no? Scott P. (talk) This comment first posted at 17:00, 26 October 2014 (UTC), last posted after a significant rewrite and list formatting at (done) 17:25, 26 October 2014 (UTC)

After more closely comparing the ramifications of both wordings, I can now see that, in order to better harmonize the two sections, the first sentence of the italicised wording above should probably read something like: Consensus can be assumed if the objecting editors stop responding for (the specified 'reasonable' time frame), (--added--) but the proposing editor remains, (/--added--) especially at (the last part of the time frame). (--added--) A failure to reach consensus can be assumed if no parties remain at the (end of the specified time frame) (/--added--) At least having some kind of a time-frame would seem to me to be better than the totally open ended system we now have. Scott P. (talk) 17:53, 26 October 2014 (UTC)

That makes it worse from my perspective... not better. Once a talk page discussion starts, and someone states an objection... you no longer have silence. People have spoken, and may well have firmly disagreed with the proposed edit. What you are proposing is that one person with a disruptive WP:IDIDN'THEARTHAT attitude can assume consensus simply because they outlasted everyone else. In fact the opposite is more common... it is quite likely that there is a clear a consensus opposed to the change, and people are simply tired of repeating themselves. Blueboar (talk) 23:01, 26 October 2014 (UTC)

I think I see your concern that the WP:IDIDN'THEARTHAT problem may still need more attention. I think that perhaps your concern may point to the possibility that 7 days is way too long to ask folks to endure, perhaps only one day, or even more specifically 24 hours. That way folks could move on to other concerns more quickly before losing interest. Also, another way that the Quaker consensus process has been adapted for modern decision making is the option to have certain vote threshholds, such as.... say.... 75%. Then the 4th person who may be an obstructionist might be more prone to try to compromise along the way, once it became clear that he was about to lose the vote.

Specifically what I see happening regarding the "silence problem", if the time-frame were shortened to 24 hours, would be this: If no "explicit consensus" has yet been reached, in order to make certain that their views are still duly registered, the interested parties would then normally still possess the small amount of self discipline needed to simply return to the discussion in one day's time, if only just to state that they are still supporting/ (objecting to) the proposed edit for the same reasons first stated. If they lack the self-discipline or concern to even do that much, in my opinion, such "hit and run" commenters should not be catered to, and it should generally be safe enough to count their vote out at that time. If a "hit and run" commenter happened to state a good relevant concern in their one-time comment, at least the others there would have had time to read it and to adopt the reasoning for themselfves if needed. So, to slow down a little, if you wanted, I could re-edit the whole thing about now to incorporate all of the changes so far discussed? Scott P. (talk) This comment first posted at 01:51, 27 October 2014 (UTC), last edited at 02:12, 27 October 2014 (UTC)

In the absence of a consensus for change, no change is made. That part of your proposal doesn't seem to have it exactly right. Also, I feel this is conflating two issues. The issue I take up and the issue Collect takes up don't have to be put together. The meaning of silence is one thing; returning to the last consensus is, I believe, something else. --Ring Cinema (talk) 18:09, 26 October 2014 (UTC)

I was thinking the same thing. Perhaps this tweak above will do it. Now to look at what you are saying about Collect's proposal. Scott P. (talk) 18:29, 26 October 2014 (UTC)

Regarding Collect's proposal and the way Collect's proposal does or may no work together with this proposal: Collect's concern seems to be that consensus could turn into a mere "contest of staying power" (which sadly is often now the case). I would think that by pre-defining a reasonable length of time, we could both give everybody a break from the greulingly long "staying-power-matches" that are now occurring, and it would force those who made comments during the same time-frame to perhaps be a little more "articulate" in their comments, and not do as our friend, Senator Cruz once did, merely reading to us all endlessly about Green Eggs & Ham! Thomas Jefferson once said, that to have rules to govern by, even if cumbersome, are better than to have no rules at all. As we now have it, there is no time frame, no clear and specific explanation of exactly at which instant a contested edit should stay or revert (except when the Admin tells you) and many other related procedures left to our imaginations. Scott P. (talk) This coment first edited at 18:43, 26 October 2014 (UTC), last edited at 19:12, 26 October 2014 (UTC)

So far the only change I'd support is Blueboar's. Dougweller (talk) 21:39, 26 October 2014 (UTC)

Will be back around 2300 UTC. Scott P. (talk) 04:35, 27 October 2014 (UTC)

I'm lost. What was Blueboar's proposal? So far, the only one I support is Ring Cinema's (the one which he repeats in this section at 03:21, 26 October 2014, above). I particularly oppose two concepts/proposals brought forth here: First, the idea that two people cannot form consensus, at least in the way we use the term consensus here; second, the highly regimented proposal by Scott P.

Let's say that there is a strong discussion in which 7 editors come to a clear consensus, fully agreed, but with an 8th editor opposed who does not concede.

• The article is edited in accordance with the consensus (Edit 1).
• The opponent does not revert immediately, but waits a few months during which the article remains stable, and then reverts (Edit 2) and no one opposes him or objects.
• Another couple of months pass and then one of the 7 happens to note the opponent's reversion and reverts it (Edit 3) saying that there was consensus for the version decided on by the 7.

What should happen? Answer: Unless a consensus can be obtained for Edit 3, the article should be returned to Edit 2 because that edit was made after the article was in a stable state and no one objected or reverted when Edit 2 was made. Why? Because consensus is a means of resolving discussion and disputes, not a status or shield, because consensus can change, and because the vast majority of edits which are made without objection or reversion are good ones and need to be able to be retained. If Edit 2 is, indeed, a bad one and a good decision was made back at the time of Edit 1 then there should be no real difficulty in forming a new consensus to revert it. (On the other hand, there is a possibility that the prior consensus, though strong, from an objective point of view was awful and wrong and the opponent was the sole voice of reason.) Consensus is not a status or shield, it's a means of resolving discussions and disputes, consensus doesn't create stare decisis because everything in Wikipedia is subject to being changed. The only question is whether the proposed change has consensus, not whether what went before had consensus. Regards, TransporterMan (TALK) 15:16, 27 October 2014 (UTC)

Blueboar's: ": Merely outlasting others does not mean one can assert "consensus."" Dougweller (talk) 19:01, 27 October 2014 (UTC)

Thanks, Doug. I'm okay with that, too. Best regards, TransporterMan (TALK) 19:26, 27 October 2014 (UTC)

I have a slightly different take on TransporterMan's scenario... since Edit 1 was the result of an actual consensus discussion... and the opponent was involved in that discussion... and thus the opponent knows full well that Edit 1 had consensus (and his/her preferred version did not)... the opponent was wrong to revert (Edit 2). By reverting to his/her preferred version the opponent is in fact acting disruptively. He/she is engaged in edit warring (all be it a very slow edit war).

If there is any assumption that should be made... The opponent should assume that the previous (Edit 1) consensus still holds ... pending a clear indication that new consensus has formed. Yes, consensus can change... the question that the opponent forgot to ask was whether it had changed. In this case, a silent consensus is not good enough. Blueboar (talk) 19:46, 27 October 2014 (UTC)

I don't find this case as the most difficult because, if the 8th editor is still alone, the consensus to remove the unsupported edit will still prevail. It's not so awful that she tried again some time later and, anyway, if the issue is really important, other editors will spot it. If it wasn't noticed, how important could it be? --Ring Cinema (talk) 20:32, 27 October 2014 (UTC)

Good question, and good answer. "What about later stealth edits attempting to subvert an earlier consensus process, perhaps even several months (or even perhaps years) earlier. The obvious answer is that an explicit dialogue based consensus is always stronger than a mere implicit time-test based consensus, and if any of the editors involved in the original explicit consensus, or anyone else who is aware of the original explicit consensus, were to catch the "stealth edit", then they would obviously have the right (and almost obligation) to revert the "stealth edit", and if it seems to be an intentional "stealth edit", probably to file an incident report.

Still, wording like this: Normally an edit that was made as the result of a consensus discussion should not have its meaning changed without first achieving a new consensus discussion specifically supporting the proposed change. "Stealth edits" made in an apparent attempt to override a consensus decision may risk xyz penalties. Scott P. (talk) 23:02, 27 October 2014 (UTC)

"Stealth edits" has no real meaning for the purpose of writing the policy. Good advice isn't policy. --Ring Cinema (talk) 01:36, 28 October 2014 (UTC)

Unfortunately I have to go out to something this evening. Will be back tomorrow night for longer. Scott P. (talk) 23:07, 27 October 2014 (UTC)

Scott, like it says in the Wikiderata: "And whether or not it is clear to you or what the essay says, there really is no deadline. Therefore be at peace with reverts, for longevity is more powerful than consensus and a damned sight better than BRD." Regards, TransporterMan (TALK) 13:38, 28 October 2014 (UTC)

Thanks TransporterMan, much appreciated. Scott P. (talk) 04:11, 29 October 2014 (UTC)

Proposal of 22 Oct.

There was some positive response to my proposal, so I'll offer it again.

Current text: "Consensus can be assumed if editors stop responding to talk page discussions, and editors who ignore talk page discussions yet continue to edit in or revert disputed material may be guilty of disruptive editing and incur sanctions."

I would propose the following: "Editors who ignore talk page discussions and reinsert disputed material may be guilty of disruptive editing and incur sanctions. Returning content to the last consensus until discussion is concluded is a more orderly way to proceed." --Ring Cinema (talk) 04:52, 22 October 2014 (UTC)

I'm sorry, but I think that the original wording is clearer. Your proposed wording seems to convey a different meaning than the original. - MrX 17:20, 22 October 2014 (UTC)

Actually, I believe this is a corollary to the No Consensus policy and it states it with admirable clarity. Take another look. --Ring Cinema (talk) 18:58, 22 October 2014 (UTC)

I should have been more clear. Your proposed wording, while not wrong, at least partly negates the original wording, which is also not wrong. Why not include both possibilities and wording that reconciles the apparent contradiction.

It's unnecessary to include "...and incur sanctions" since its covered in WP:DDE. I firmly disagree with the last sentence of your proposal, for the reasons I've stated above. Some editors stonewall the consensus seeking process by making a single comment in a discussion and then leaving as if they have no further obligation to respond to reasonable counterarguments. This has a potentially detrimental effect on content.- MrX 18:14, 24 October 2014 (UTC)

Ring Cinema's rewrite alleviates the problematic wording of the paragraph: I think "Consensus can be assumed if editors stop responding to talk page discussions" is open to gaming, especially when a policy hands an editor a mandate to assume something about the stance of other editors in a dispute. Editors often drop out of discussions because they have nothing further to add, not because they concede the point or believe they have consensus. By definition, there is no consensus until there is a consensus, so WP:NOCONSENSUS should apply until a discussion "concludes". Whether the discussion results in a new consensus pretty much depends on the nature of the discussion, and I don't think it is particularly helpful to assume something just from the fact a discussion has concluded; the conclusions need to be actively drawn from a discussion. Betty Logan (talk) 07:13, 24 October 2014 (UTC)

But the opposite is true, too: it cannot be necessarily concluded that there is no consensus from the fact that everyone fails to affirmatively state their agreement. When seven editors are saying X (for a good reason) and one is saying Y (for a bad reason) it is neither true that (a) Y has to expressly concede in order to form consensus nor that (b) all seven have to actually say that they agree with one another in order to form consensus. I'm fine with the change proposed by Ring Cinema, but we've got to remember that this pendulum swings both ways. Regards, TransporterMan (TALK) 13:22, 24 October 2014 (UTC)

I don't mind Ring's wording... but I do think it is incomplete. A lack of continued discussion can indicate one of three things:

1. A clear consensus has formed in favor of the proposed change (and most of the participants are tired of arguing with the few who oppose) - in which case, continued stubborn opposition to it is disruptive.
2. A clear consensus has formed against the proposed change (and most of the participants are tired of arguing with the few who support it) - in which case, continued pushing for the change is disruptive.
3. No consensus has formed (and most of the participants are simply tired of the debate) - in which case we default to the last firm consensus version, take a break for a while, and resume the discussion at some point in the future to see whether consensus has changed. Resuming the discussion after a reasonable break is not disruptive.

You have to examine the actual discussion to know which of the three applies. Blueboar (talk) 13:30, 24 October 2014 (UTC)

I was thinking that all the varieties you mention are not going to be covered here. The rest of the policy, I believe, covers the indeterminate part of consensus-forming that concerns you. --Ring Cinema (talk) 21:27, 24 October 2014 (UTC)

The problem with the current version is that it is poorly worded. I think the original intent behind that sentence was to state that a new consensus could only be arrived at through the conclusion of a discussion, but it reads as though once the discussion concludes a new consensus is formed, which we all know isn't necessarily true. Ring Cinema's wording corrects this to state that a new consensus requires the conclusion of the discussion i.e. a new consensus effectively concludes a discussion. Betty Logan (talk) 16:10, 24 October 2014 (UTC)

Yes, that's mostly true. Another scenario, which I have encountered many times, occurs when a change is proposed on the talk page (or boldly edited and a discussion is opened), and no one one responds. After a few days, its safe to assume that no one objects to the change, thus a 'silent consensus'. As long as we don't loose the spirit of that idea in the policy, and as long as we don't create a loophole for users to game the system by stonewalling progress, then I can live with just about any wording, including some neat combination of the all of the ideas in this thread.- MrX 18:28, 24 October 2014 (UTC

I don't disagree with the above comments, but when I attempted an amendment to my draft it fell into the trap of trying to do too much. This proposal is narrowly focused. It doesn't say everything about discussion and consensus, but it clarifies, I believe, a problem paragraph. The other question merits a different proposal. --Ring Cinema (talk) 21:46, 24 October 2014 (UTC)

There are probably many improvements that can be made, but they probably warrant a separate discussion anyway. This is just a basic correction, in matching the wording to the original intention behind the sentence. Betty Logan (talk) 22:13, 24 October 2014 (UTC)

• Support Ring Cinema's proposal as realistically reflecting actual good practice and having the right tone, describing good practice without prescribing method. --SmokeyJoe (talk) 08:56, 26 October 2014 (UTC)
• Support Since my proposal below may be a bit too "daring" for some, though I truly do hope that ultimately when people eventually weary of the incessant inbuilt edit-wars under the current structure, such a policy may eventually be accepted. At least this is a step in that direction. Scott P. (talk) 13:10, 26 October 2014 (UTC)
• Oppose because it's changing the basis of the 'consensus can be assumed' bit if no one responds or if talk page discussions cease. You're omitting that and I read the above and don't really see a good reason to do so. Betty Logan's response on that people can 'game' consensus if no one objects...well, when no objects, what do you call that? Silent consensus. Say I edit a highly trafficked page like a politican's and add a reliably sourced paragraph detailing family life. It doesn't get reverted and no one really objects, so no discussion is needed and is in effect a silent consensus (after a reasonable amount of time). Well, a month or so later, somebody reverts me as 'not really relevant to politician' even though pretty much no one had a problem with it before. It's not 'gaming' the system for the person who reverted me to get consensus since in effect, when I did my changes it was seen by editors, not remarked as problematic and I wasn't reverted nor discussion did ensue, why would I suddenly have to defend my changes a long time afterwards when it was deemed as beneficial or wasn't contested? I also fear that if the silent consensus bit is removed, people can contest all sorts of well deemed and well established things in the article just because 'there wasn't a consensus because only you added it and there was no discussion'. It seems overly bureaucratic. If you're wanting to reword it because it's 'badly written' then do so, but the original meaning should not be lost or displaced. Tutelary (talk) 13:28, 26 October 2014 (UTC)

It seems to me like the potential of an agreement may remain here. @Tutelary: if the existing idea: "Consensus can be assumed if editors stop responding to talk page discussions," were fully retained, would you then be Neutral? @Ring Cinema: would you be willing to fully retain that idea if that would get a Neutral here? Scott P. (talk) 13:46, 26 October 2014 (UTC)

Yes, that's pretty much my entire qualm. Rewording, fine as long as the original meaning is not lost. Ring Cinema's proposal is just that, rewording but then losing the original meaning. Tutelary (talk) 15:38, 26 October 2014 (UTC)

Except that would continue perpetuating the erroneous line of thought that editors stop reponding on the talk page because there is consensus. What happens more often than not is that if there is a conensus then editors stop replying on the talk page. That is what the policy should say; we should not draw assumptions simply because someone withdraws from a discussion, which is what the current wording implies. Betty Logan (talk) 19:24, 26 October 2014 (UTC)

Note: Ring has not posted on Wikipedia for 10 hours and is apparently out for a little while. Scott P. (talk) 14:27, 26 October 2014 (UTC)

• Support If there is a consensus then that inevitably marks the end of a discussion, and only then should the article's state be altered to reflect the new consensus. If discussion is ongoing then by definition a consensus has not been reached. This is the spirit of the policy and it should be worded to that effect. If editors simply stop responding on the talk page (which often happens when it becomes clear discussion is going nowhere) then it should not be assumed a consensus has been reached, as implied by the current wording. Betty Logan (talk) 19:24, 26 October 2014 (UTC)

I have no problem with Scott's fix per Tutelary's qualm. I was explicitly trying to separate the "meaning of silence" from "go back to the last consensus". May I suggest that I restore the meaning of silence part on this selection, but then it's removed if the discussion below bears fruit? I suggest that because I haven't seen anyone attempt to link the two concepts in some causal way. --Ring Cinema (talk) 00:12, 27 October 2014 (UTC)

I'm going to try to meet Tutelary's and BettyLogan's complaints. During an ongoing talk page discussion, editors should assume there is not a new consensus until it is concluded. Consensus can be assumed if editors stop responding to talk page discussions. Editors who ignore talk page discussions and reinsert disputed material may be guilty of disruptive editing. Returning content to the last consensus until discussion is concluded is a more orderly way to proceed. --Ring Cinema (talk) 04:52, 22 October 2014 (UTC)

"During an ongoing talk page discussion, editors should assume there is not a new consensus until it is concluded"

No. As a discussion approaches consensus, discussion tends to stray to minutiae or other matters. This sentence contradicts Wikipedia:Consensus#Reaching_consensus_through_editing, which is needed to keep focus on edits that may achieve consensus.

You would say this even if there is an active discussion? In that case, the consensus through editing would seem to be over. --Ring Cinema (talk) 02:16, 27 October 2014 (UTC)

"Consensus can be assumed if editors stop responding to talk page discussions."

No. This is an appallingly bad apparent definition of "consensus". Stated objections do not need to be repeated, and if unaddressed there is no consensus. A lack of response at best supported a weak presumption of consensus.

Tutelary had a problem with this, but perhaps he has a suggestion for a compromise. For example, would it work to start the sentence with "Sometimes..."? Personally, I would prefer it. --Ring Cinema (talk) 02:16, 27 October 2014 (UTC)

"Editors who ignore talk page discussions and reinsert disputed material may be guilty of disruptive editing."

Yes.

"Returning content to the last consensus until discussion is concluded is a more orderly way to proceed."

Yes. But note that this does not assert the the last consensus was any particular previous version. Note that any and every previous version may be broadly disputed. This policy must not be readable as declaring by formula something specific as "consensus". --SmokeyJoe (talk) 01:04, 27 October 2014 (UTC)

Perhaps you're right, but neither does the current policy specify these things. --Ring Cinema (talk) 02:16, 27 October 2014 (UTC)

Revised proposal

Tutelary doesn't agree with removing the first original sentence and SmokeyJoe doesn't like my new try in the first sentence. I can accept that there is no consensus for these changes, so let me propose this:

Consensus can be assumed if editors stop responding to talk page discussions. Editors who ignore talk page discussions and reinsert disputed material may be guilty of disruptive editing. Returning content to the last consensus until discussion is concluded is a more orderly way to proceed. --Ring Cinema (talk) 20:26, 27 October 2014 (UTC)

Discussion on this proposal continued below in #Revised proposal --Francis Schonken (talk) 15:29, 31 October 2014 (UTC)

Boiled down proposal from Collect's-proposal-on-steroids:

Here is what I think is a very boiled-down and simplified version of the proposal from the previous section:

An explicit-consensus in favor of a given edit is achieved, only if by the end of the 24th hour after the posting of the original objection, the following two conditions are met:

1. 75% or more of the editors involved have posted their clear approval of the edit, and,
2. At least one of the approving editors is shown to still be present in the discussion, minimally by posting, “still in support of the edit”, or something similar, during the 24th hour after the posting of the original objection.

One of the supporting editors should be present at the end of the 24 hour discussion window to both confirm that the edit is still supported, and, after the full 24 hours has ended, if consensus was acheived, to re-insert the now-explicitly-consensused edit.

Anything short of these requirements would result in a lack of consensus.

The preferred procedure for treating a disagreement with a recent edit is to first post an objection on the article's talk page. Such objections should ideally be full, well reasoned, and hopefully well focused on the edit and not the other editor. Next return the concerned article content back (revert it) to the state of its last consensus, until a mutually acceptable agreement can be reached. If after 24 hours, a consensus has not yet been achieved, an appeal to an Rfc can still be made to try to gain the required 75% threshold. Such an appeal would only extend the period for possible consensus by an additional 24 hours, with the same requirement that at least one of the supporting editors must be present during the 48th hour. Editors who enter the discussion via an Rfc should attempt to find harmony, rather than to take sides. If after 48 hours, still no consensus can be found, then the last edit with consensus should remain in place.

Normally an edit that was made as the result of an active consensus discussion (an explicit-consensus-edit) should not have its meaning changed without first first achieving a new consensus via a new consensus discussion using the above consensus-discussion-rules. Later edits which may attempt to override such an explicit-consensus-edit without first attempting to achieve a new consensus via a new consensus discussion should automatically be treated as edits with no consensus.

Regarding contested edits to material that is less than one month old, and for which earlier material there was no consensus discussion, i.e. for which only an implicit-consensus exists: For any such contested edits, such edits shall be considered as having achieved consensus if in their consensus discussions as above, only a 51% threshold is achieved. (Meaning, the requirements to revert or change such "younger edits" are not quite as demanding.)

Any editors who might not comply with these procedures for consensus may be liable for having acted in a disruptive manner, and may incur sanctions.

I know this rule would be a bit more "regimented" than what we now have, but due to its open-endedness, what we have often seems to me to create unnecessary difficulties, and in my experience, from time to time seems to incite disharmony and bad feelings amongst editors, if for no other reason, than simply as a result of its unpredictability and its "openness to loopholes". It is my belief that the proposed policy above incorporates at least something from all of the many excellent suggestions and comments above, and "boils it all down" to a simpler and more coherent form. Thanks to all. Comments? Scott P. (talk) 03:36, 29 October 2014 (UTC)

Instruction creep and likely to cause more problems than it causes. Seems to me that as USer:DeCausa USer:NeilN USer:Cullen USer:Francis Schonken USer:Morphh were discussing this with you at Jimbo's page they might want to comment also. Dougweller (talk) 09:40, 29 October 2014 (UTC)

Is there any specific thing at all that you see "wrong" with this proposal, other than the fact that it "creeps", so to speak. The concerns you listed in the previous section were ,1. regarding the possibility that these rules might make it easier for pov'ers and fringe editors to have their way, and 2. Regarding the possibility that editors in general might be able to ram edits through by their mere "staying power". In the rules above, I did my best to try to take your concerns into account. Are these two things still concerns of yours, and if so, why? If you might have any specific concern other than "instruction creep", your clarification and explanation of it would certainly be appreciated. If concerns 1 and 2 are still concerns of yours, then exactly why are they still concerns of yours? Thanks, Scott P. (talk) 11:38, 29 October 2014 (UTC) Scott P. (talk) This comment first edited at 11:38, 29 October 2014 (UTC), Last edited at 11:58, 29 October 2014 (UTC)

(e.c., @Dougweller:) Indeed, tx.

First consideration: Scotperry's proposal focusses on an algorithm to determine consensus in a near to mathematical approach: for this policy page this would draw attention away from its central message: discuss over disagreements until (at least) a modus vivendi is found. Use rational argument rather than weight of numbers to bring opposing views closer together. I.e. the "process of compromise" mentioned in the nutshell.

Second consideration: trying to win an argument "by weight of numbers" is usually rather divisive than something that brings opinions closer together. It is contradictory to the idea of ironing out differences in approach, but rather acerbates those differences, and is contrary to what experience has thaught to be the best mode of operation for Wikipedia.

Detailed procedures generally don't work too well together with "policy-level" guidance. WP:3RR is nonetheless such policy-level detailed procedure, designed to put an end to situations where editors disruptively fail in the consensus seeking process. That one works well in Wikipedia, as a "line not to cross", and nothing in that vein is needed here, in this policy-level guidance that explains what to do instead. --Francis Schonken (talk) 11:49, 29 October 2014 (UTC)

I also agree to Dougweller's "Wikipedia:Instruction creep" argument. --Francis Schonken (talk) 11:52, 29 October 2014 (UTC)

Absolutely not. This proposal would determine content by voting rather than the strength of arguments. SPA's and fringe theory pushers would salivate. --NeilN ^{talk to me} 12:01, 29 October 2014 (UTC)

Gone for the next 12 hrs. Scott P. (talk) 12:07, 29 October 2014 (UTC)

• Oppose Per Francis Schonken, Dougweller, and NeilN. On Wikipedia, consensus is determined by the strength of arguments, not the numerical supremacy of !votes. Rationalobserver (talk) 18:14, 29 October 2014 (UTC)

• Oppose: Though this would make our job over at dispute resolution much easier, I have to oppose since it focuses too much on quantity, not quality. TransporterMan (TALK) 20:36, 29 October 2014 (UTC)

• Oppose for quite a few reasons but the principles in WP:NOTAVOTE would cover many of them. I don't really see a problem that needs to be fixed in any case. I think WP:EDITCONSENSUS tempered with WP:BRD works well enough. DeCausa (talk) 21:51, 29 October 2014 (UTC)

What problem are we trying to resolve?

I think we may have lost sight of the underlying issues (I know I have). This often happens in lengthy discussions... as participants discuss proposals and counter-proposals, they develop different conceptions of what the problem actually is, and start to talk at cross purposes. And so... I think it might be beneficial to take a break from considering proposals and counter-proposals - to take a step back, and re-state what the problem actually is.
First we need to reach a consensus that it actually is "a problem" ... and then we can discuss how best to resolve that problem. Blueboar (talk) 13:09, 30 October 2014 (UTC)

I wholly agree and would add the suggestion that rather than just talking about this in the abstract that we might also consider the question of just exactly how and under what circumstances the problem would even come up in daily practice. That is, in what precise circumstances would the objection be made that "you can't do that, we had/have a consensus"? Best regards, TransporterMan (TALK) 18:48, 30 October 2014 (UTC)

Sorry to take up so much space here. Thanks to all who attempted to make constructive contributions to the idea that consensus, or lack thereof, might actually be something that is simple and quite clear to all, and that a clear well defined consensus should be required before edits are generally accepted. Who could ever be so foolish as to imagine that such a concept as a "consensus" might possibly ever be something simple and clear, a thing easily understood and agreed upon by all? Sorry for having been so bold as to have apparently let my imagination get the best of me, bating me with the hope that a consensus might actually be something intelligible, simple and clearcut. So I suppose the misty concept of whatever a consensus really is, and when one can actually act on whatever one is, must remain forever primarily in the realm of the mysterious administrator's mind. It would appear that the nay's now have it. Scott P. (talk) This comment first posted at 02:36, 31 October 2014 (UTC), last edited at 05:21, 31 October 2014 (UTC)

No one has yet answered Blueboar's question of what is the supposed problem trying to be fixed and TransporterMan's request of concrete examples of the "problem" actually arisings. DeCausa (talk) 07:27, 31 October 2014 (UTC)

The answer is here: #Assuming consensus, triggered by this bold edit. Two issues were raised:

1. A proposal to remove "consensus can be assumed if editors stop responding to talk page discussions, and..."
2. A proposal to reword the rest of the sentence: "...editors who ignore talk page discussions yet continue to edit in or revert disputed material may be guilty of disruptive editing and incur sanctions.", for clarity.

There are other ideas floating around, but it was this specific sentence that seems to be under discussion.- MrX 16:12, 31 October 2014 (UTC)

I find it to be of interest that while I have asked here whether simplicity, ease of understanding by typical editors, transparency of the consensus process, and clearcut results that do not typically require admin intervention, should be aims of this discussion, my voice appears to have fallen mostly on deaf ears, with apparently no mention of these specific aims by most others here, instead apparently preferring to debaqte fine points of language instead. No wonder the consensus process is such a relative mystery to so many. No wonder the editors on the page seem to have so much trouble even coming to a consensus about what a consensus should be. Interesting..... Scott P. (talk) 22:02, 31 October 2014 (UTC)

I agree that some more solid definitions would help - all-too-often, Wikipedia editors assume a simple majority of all the editors on an article is all they have to get accepting a position, rather than actually trying to accommodate the objections reasonably held by a minority.

We even have a current case where an RfC was closed finding that a specific source was reliable as a source for its own opinions (by a broad consensus), and then the "losing side" averring that now we have to decide whether that source, after a long discussion finding it to meet RS, now meets "weight considerations."

We have editors cheerfully using sources where it backs what they "know" to be the "truth" and contesting the exact same sources where it contradicts the "truth." We have editors following other editors around with the goal of inserting snarky comments about the tailed in as many locations as possible, and averring that they are the guardians of the "truth."

Rather than being a "Society of Friends", Wikipedia is often a society of editors bound by common interests trying to ensure those specific interests are dominant in the project. "Consensus" is not just a "mystery" in such cases, it is a miracle when it actually occurs. Collect (talk) 22:21, 31 October 2014 (UTC)