The project page associated with this talk page is an official policy on Wikipedia. Policies have wide acceptance among editors and are considered a standard for all users to follow. Please review policy editing recommendations before making any substantive change to this page. Always remember to keep cool when editing, and don't panic.
This is the talk page for discussing improvements to the Consensus page.
This page seems to require talk page discussion before changes to policy pages. The policy policy however allows the bold editing of policy pages. I personally feel that bold editing should be allowed and the wording here modified, although I agree we should admonish caution when making changes. I think an outright ban on editing policy pages(eliminating the possibility of bold, revert, discuss and totally contrary to WP:BOLD) is a bad direction to take the policy of policy, after all, that is how all these pages were created in the first place. Simultaneously (somewhat ironically) I am tepid enough to talk about this discrepancy first, rather then just editing the page. Thoughts? Crazynast04:16, 12 October 2011 (UTC)[reply]
I do not see any conflict—they both say the obvious, namely that normal procedure would be to discuss a proposed change to a policy rather than jumping in and substantively changing it. However, this is not a bureaucracy and there may be exceptions, so "no prior discussion" is not reason to revert a good change. Johnuniq (talk) 06:20, 12 October 2011 (UTC)[reply]
I don't read this page or any page as prohibiting the bold editing of policy. If you can improve a policy page, by correcting or improving a description of policy in practice, then do it. What is discouraged is the bold editing of policy as an attempt to "change the rules" as a means to changing practice. Policy pages, freely edited, should lag current practice and describe recent practice. New ideas, and new rules should be discussed first. --SmokeyJoe (talk) 06:59, 12 October 2011 (UTC)[reply]
Should vs. May
The older but still valid method is to boldly edit the page. Bold editors of policy and guideline pages are strongly encouraged to follow WP:1RR or WP:0RR standards.[1]
and Substantive changes should be proposed on the talk page first[2] seem to directly conflict. If multiple methods are allowed we should say that or at least not imply that there is only one course of action for change. If the community view (or established practice) is that the only time we don't start at the talk page is when we're ignoring all rules we should reflect that consistently across the policy pages as well. The simple change I attempted was to change should to may in the second passage, undone with the comment "should" is still correct, with rare exceptions per usualWP:BURO and WP:IAR. It seems to me that saying should is inherently bureaucratic as a requirement, where leaving it as a strongly encouraged option would be sufficent. Thoughts? Crazynast08:28, 12 October 2011 (UTC)[reply]
Are you disagreeing with the statement that substantive changes to policy pages should be proposed on the talk page first? How would replacing "should" with "may" (diff) help? Johnuniq (talk) 10:52, 12 October 2011 (UTC)[reply]
"Should" works better than "may". There is a slice of Lie-to-children here. If you need to read this page to know what to do, then you should propose first. If you don't need to read this policy, then it matters little what the precise words are. Other constructs may improve upon "should". --SmokeyJoe (talk) 10:56, 12 October 2011 (UTC)[reply]
I see your point. Perhaps the solution is a re-write from "Substantive changes should be proposed on the talk page first...:" to something like "Accordingly, the best practice is to propose substantive changes on the talk page first ..." Butwhatdoiknow (talk) 10:59, 12 October 2011 (UTC)[reply]
There are probably three kinds of changes being mixed up here. One is merely cosmetic changes to policy pages (spelling corrections, wording improvements, even quite substantial reorganization) - these shouldn't require prior discussion any more than changes to WP articles would. The second is changes that change the substantial meaning of the policy page (i.e. change the rules, if policy pages are seen as rules) to bring it into line with what is actually accepted best practice. The third is changes that aim to change what is accepted as best practice. It's the third that obviously ought to require prior discussion (you can't say that the community accepts something that it doesn't yet do unless it explicitly says it wishes to start doing it). But the first two types of changes ought to take place just like changes to any other page. Unfortunately in practice they often don't, since certain people are very trigger-happy with the revert button on policy pages, having got i into their heads that these pages contain some kind of semi-sacred text (or at least, that they always describe best practices eloquently, accurately and authoritatively, which is unfortunately not the case). --Kotniski (talk) 11:32, 12 October 2011 (UTC)[reply]
Consensus is not necessarily unanimity. Ideally, it arrives with an absence of objections, but if this proves impossible, a majority decision must be taken. More than a simple majority is generally required for major changes.
This edit introduced a novel policy; it contradicts WP:NOTDEM, and does not seem ever to have been discussed at this page. Its effect, if taken literally, is to replace consensus by majority vote any time a closer chooses to invoke it.
I trust that it is intended more narrowly: that sometimes a decision must be taken, and if it is absolutely necessary to act one way or the other, the majority prevails.
We are not a government; there is very little we must do, that cannot wait for discussion to include as large a proportion of editors as possible. On the assumption that this was what was meant, I shall edit to:
Consensus, on Wikipedia, is not necessarily unanimity. Ideally, it arrives with an absence of objections, but often we must settle for as wide an agreement as can be reached. Silence on the disputed issue is sometimes an option, but in those cases where a decision is absolutely necessary, and compromise is impossible, it may be necessary to follow the majority. More than a simple majority is generally required for major changes.
Please note that even this redefines "consensus." Off Wikipedia, it does mean "Agreement in opinion; the collective unanimous opinion of a number of persons," as the OED puts in (italics mine). SeptentrionalisPMAnderson19:04, 29 October 2011 (UTC)[reply]
Not sure what you're saying here, but I think the proposed change is an improvement. The way I read the new paragraph, it does not imply that unanimity is a prerequisite to consensus. Crazynast20:10, 29 October 2011 (UTC)[reply]
No, sorry, I wasn't objecting to the new paragraph, just pointing out that the original change to this page (way back whenever) that introduced talk of a majority was hardly a novelty - decisions have been called "consensus" without being unanimous for as long as I've been on Wikipedia, and probably ever since the total number of editors rose above about five. --Kotniski (talk) 20:37, 29 October 2011 (UTC)[reply]
Let me distinguish:
For us, decisions don't have to be unanimous to be consensus. I agree, and that is in the new text.
But for us, decisions made solely by majority vote are and ought to be vanishingly rare. They would have to be both situations where we couldn't simply say "no consensus" and leave the matter alone and ones where there was no hope of attracting wider approval by amending the proposal ("not A or B, but A'. ")
The novelty here consisted in suggesting that majority votes were somehow normal.
Consensus, on Wikipedia, is not necessarily unanimity. Ideally, it arrives with an absence of objections, but often we must settle for as wide an agreement as can be reached. When there is no widespread agreement, consensus-building involves adapting the proposal to bring in dissentients without losing those who accept the proposal.
When no widespread agreement is possible, we call that no consensus. Often this results in no change or in silence on the disputed issue. If there is a case where a decision is absolutely necessary, and compromise is impossible, it may be necessary to follow a simple majority. More than a simple majority is generally required for major changes.
The change to the policy was made without consensus, and is also contrary to the spirit of the policy. I think it should be removed completely. By "completely" I mean that we should not replace it with a watered down or modified version. Wikipedia does not work by majority votes. To suggest "we work by consensus, but when we can't get it we settle for a majority vote" is nonsense, because it amounts to no more than a roundabout way of saying "we work by majority vote", which is just not true. JamesBWatson (talk) 21:26, 31 October 2011 (UTC)[reply]
If there is a case where a decision is absolutely necessary, and compromise is impossible, it may be necessary to follow a simple majority. More than a simple majority is generally required for major changes.
That consensus does not require unanimity has been in this policy for years. If "consensus is not unanimity", then consensus is (sometimes) a (weird kind of) majority (including, importantly, that you "count" the !votes of the whole community when you're determining the majority, so that any one editor easily outvotes any number of vandals).
The current discussion is prompted by the RFC at WT:V. To avoid any time-wasting allegations about "gaming the system", I strongly suggest making no changes until that RFC is firmly, totally, definitely over with, including plenty of time for whoever feels aggrieved by the outcome to go off and complain in whatever (usually multiple) forums they choose. If this line can be improved, then it can be improved a month from now.
There are times when "no consensus" is an inappropriate, or even impossible, outcome. We cannot simultaneously delete and not-delete images; we cannot simultaneously elect and not-elect people to the Arbitration Committee, and we cannot simultaneously block and not-block users. Rather than pretending that a true consensus is always possible, we actually pre-define the meaning of "no consensus" in some situations (e.g., "no consensus" at AFD is treated exactly like a full-consensus keep). This policy has to cover all of those situations, not just the text-changing disputes that are clearly on the OP's mind. WhatamIdoing (talk) 22:02, 31 October 2011 (UTC)[reply]
I was prompted to read the section by a quote at the WT:V discussion, but that is not why I object to it; I did not realize, despite citing this policy for years, that Kotniski had added a reference to majoritarianism, and I would think it unwise if nobody had cited it. I would object to a restoration of the text at the head of the page; this entire discussion casts doubt on it ever being consensus - or indeed what Kotniski actually meant to say.
As for What's examples: those are the sorts of thing that belong in guidelines and other policies; specifying them here makes them too hard to reconsider. Do any of them fall outside the range of no consensus = no change (deletions, moves, blocks) or no consensus = silence? SeptentrionalisPMAnderson22:13, 31 October 2011 (UTC)[reply]
Certainly there are exceptions. If there is no consensus on whether something is a copyright violation, we normally delete it—a change. If there is no consensus behind a block, then the account gets unblocked—a change. RFAs normally pass with 70%, even though that's technically "super-majority" rather than "consensus". More importantly, "no change" can be meaningless. Imagine that I create an article. The next day, you add a sourced paragraph to it. Some people say your paragraph is off-topic; others believe it's relevant. The discussion dies out with no consensus. What does "no change" mean in this situation? No change, so your newly added paragraph is reverted? Or no change, so your newly added paragraph is kept? Either of these choices could legitimately be construed as "no change". WhatamIdoing (talk) 01:28, 1 November 2011 (UTC)[reply]
Guys, this is really getting away from the spirit of consensus. If you want to make a meaningful revision, I think you want to lean this way:
Consensus is a different concept than a numerical count. It is not unanimity and not majority rule, but rather a process of negotiation and balancing intended to create an outcome that most people will consent to, even if they don't agree with it 100%. Consensus in the natural outcome of a healthy collaborative process, where editors have worked together to craft an article they can all live with.
In some cases consensus is not achievable, usually because editors working on the article are unable to edit collaboratively due to ideological differences or personal frictions. In such cases editors may seek outside comments to break the stalemate, using tools such as RfC's. But these are not intended to be votes or efforts at establishing some 'majority opinion'; they should be considered extensions of the consensus process that bring in new editors to overcome the failure of collaboration on the page. Often, in fact, an RfC will produce a result unrelated to its original request, merely because new editors add fresh perspectives to the dispute.
Whatever we do we must not remove or hide the key information that on Wikipedia (in contrast to many cases in the real world) the word "consensus" is often used to describe a situation where there is far from unanimous agreement. One gets the feeling that this page, like so many other policy and guideline pages, is not being written with the intention of conveying to people clearly how we actually do things, but in a weaselly way in order to produce (or eliminate) statements that can be used as "arguments" elsewhere.--Kotniski (talk) 11:03, 1 November 2011 (UTC)[reply]
There is likely to be a difference between Wikiepedia's practices and its rules. That is common for institutions and is not necessarily a bad thing because it allows the solution of difficult cases. Definitions or rules are reductive and, as Kotniski notes, are often written with an eye to how they might be misappropriated; this, too, is normal. This page tries to express the practice of consensus on Wikipedia without writing in an excess of procedure and that rightly leads to editors using "consensus" and its related policies on the hard cases. I don't think we want to discourage that. --Ring Cinema (talk) 14:12, 1 November 2011 (UTC)[reply]
One more thing. There is a sensible rule of thumb for when a reasonable consensus is reached: when a majority of the minority accepts a proposal, that is very good evidence that the minority views are recognized and incorporated. --Ring Cinema (talk) 14:19, 1 November 2011 (UTC)[reply]
The issue is what to do under a very specific circumstance: (1) consensus cannot be reached and (2) a decision must be taken. For example: an image has been reported as a suspected copyright violation. The discussion is strongly divided. We cannot both "keep" and "delete" the image. What do we do? Pretend that we have a consensus anyway? Keep the discussion open, possibly for years (a process that puts the WMF in legal jeopardy if it actually is a copyvio), in the hope that a true consensus will magically appear someday?
We all want consensus. Actually, we all want perfect unanimity, because it's simple and easy to interpret. But it is not always possible to achieve this. And sometimes (not most of the time), we actually have to make a decision despite the obvious absence of consensus.
Three days ago, this policy talked about what happens in that (rare) circumstance. Now it doesn't, and we're back to pretending that consensus will always appear, if only you talk long enough (and block irritating people often enough).
Think about the possible copyvio: There's no consensus, with strong arguments and good editors on both sides, and we must ultimately make a decision one way or the other. What would you do? Endlessly extend the discussion? Always keep the disputed image? Always delete the image? Go with the majority? Stick your head in the sand and pretend that choosing to take no action is not itself a decision? WhatamIdoing (talk) 15:36, 1 November 2011 (UTC)[reply]
New paragraphs are covered in WP:BURDEN: the burden of establishing consensus rests on those who would add material, so it defaults to no addition, which is no change from the extablished text.
Copyvio (and BLP) fall under the same principle: adding material requires consensus. They also have a touch of IAR, since we are acting under legal constraints.
Nevertheless, I don't think that we should list such cases here: discussing things on two different pages leads to divergence. Is there a list of places where the meaning of "no consensus" is pre-decided? That would be a useful addition, as another subsection. 15:44, 1 November 2011 (UTC)
I don't find it useful to think of an IfD as a case where a decision must be taken; if that were the case, once an image was kept, that would settle it indefinitely. But there are, in practice, three resolutions: delete, keep and stop asking, and no consensus. If there is no consensus, there can easily be, and often are, repeated IfDs until consensus does form. SeptentrionalisPMAnderson15:50, 1 November 2011 (UTC)[reply]
Once positions are ossified, there should be an implicit burden on the minority to craft a proposal that their own like-minded editors endorse, as that mitigates the possibilities of stonewalling. If a minority's proposal is accepted by a majority of the majority, that should be good enough, as a practical matter. Similarly, a majority proposal accepted by a majority of the minority should end the discussion as a practical matter. That recognizes that unanimity is not always possible, that sometimes a decision must be taken, that decisions are revisable, that minority stonewalling is bad and that, in the absence of anything else, majorities are not irrelevant. --Ring Cinema (talk) 16:30, 1 November 2011 (UTC)[reply]
Thinking about that paragraph, does a brand-new stub really count as "established text"? If you've had something for a few years, then it's easy to claim a "prior consensus" (a claim, BTW, that violates the CCC provisions of this policy), but it's silly to make such a claim for two paragraphs whose age differs in only a matter of minutes or hours.
About the image in my example, the fact is that we don't know whether it's a copyvio: we can't make up our minds. Deciding to choose the default is, itself, a decision.
Certainly we should say something about how the "consensus" principle is applied in different areas. It's quite different with deletion decisions, for example, than with day-to-day article editing decisions, and different again for major policy changes (or even minor policy wording changes that people get emotional about, like this "verifiability not truth" nonsense).--Kotniski (talk) 08:42, 2 November 2011 (UTC)[reply]
Consensus seems to offer the best method to establish and ensure neutrality and verifiability
Why do we need to say this in the lead? What is it even supposed to mean? What other methods have been considered? Is this the sole or main raison d'etre of the consensus principle?--Kotniski (talk) 08:44, 2 November 2011 (UTC)[reply]
But why specifically "to ensure neutrality and verifiability"? Surely in so far as it works for those two things, it works just as well for all other things we find desirable? (And again, what other methods are we comparing it with? Who has done the comparison? On what basis do they conclude that consensus "seems best"?)--Kotniski (talk) 13:49, 2 November 2011 (UTC)[reply]
Neutrality, because a truly neutral text really needs agreement from rational readers from all the contending sides; finding them is the problem. Verifiability needs the same readers, as a check on what we need to verify.
As for the rest, this is a program; those who want to see what an encyclopedia genuinely produced by majority vote would look like, with all the sockpuppetry and meatpuppetry that implies, are welcome to take a fork. SeptentrionalisPMAnderson19:35, 2 November 2011 (UTC)[reply]
Well, majority vote is not the only alternative to what we call "consensus" (assuming there really is meaningfully a "method" corresponding to our use of that word). And my objection to saying "verifiability" and "neutrality" was not that consensus fails to produce these goods, but that consensus is equally successful (or arguably unsuccessful) at producing our other goods (completeness, clarity, usability, etc.), so there seems to be no particular reason to single out these two.--Kotniski (talk) 09:06, 3 November 2011 (UTC)[reply]
Completeness and clarity are less valuable to us; we are more willing to tolerate an incomplete or turgid article, on the grounds that it will get fixed before WP:DEADLINE, than OR or bias. Also, they are more likely to be attained by individual work, not requiring consensus. But if you want to say something like including neutrality and verifiability, suggest language. My reason for not saying something myself is that I don't see how to qulaify without verbosity SeptentrionalisPMAnderson13:20, 3 November 2011 (UTC)[reply]
I'm not convinced that it's the best method to establishing verifiability. For example, recruiting, training, and retaining subject-matter experts would often be more effective.
I more or less agree with the claim that it produces NPOV compliance. Additionally, it normally produces what we might call good editorial judgment: if everyone agrees that these two articles ought to be merged, then that probably is the best choice. WhatamIdoing (talk) 17:16, 3 November 2011 (UTC)[reply]
Consensus is a process in which editors seek to reach a mutually acceptable decision through discussion. Sometimes the term is used to refer to an ongoing process (a consensus discussion) and sometimes to the result of a past discussion (a current consensus). In both senses of the term, consensus implies that editors have raised any concerns they think are pertinent and weighed them in the light of the needs of the article and the policies and principles of the encyclopedia. Consensus is the natural process of healthy collaboration, and all editors are expected to make good-faith efforts to engage and respect it.
Consensus is not a numerical process; it is not the same as unanimity or majoritarianism or any other system of 'counting votes'. When things that look like votes happen on wikipedia (as in RfCs, AfD discussions, or other administrative actions), they are actually intended to solicit a broader range of discussion, not merely to tally up agreement. Such discussions are usually closed when they reach a saturation point in which most of the participants share a loose agreement about which policies and principles are pertinent and how they should be applied to the debate. The aim of consensus is to reach understanding, not agreement - it is expected that people can disagree with the outcome but accept it because they understand the principles in play.
Consensus is not always achievable for the same reasons that collaboration sometimes fails - strong ideological convictions, interpersonal frictions, or simple good-faith disagreement can frustrate the natural discussion process. Sometimes these discussions are left as no consensus, which ends the discussion without any action being taken, and sometimes they proceed on to dispute resolution mechanisms in order to seek a solution in a different way.
Ronz reverted [3] with an unclear edit summary. Are there any actual problems with this revision (which I think sums up the concept nicely)? --Ludwigs216:14, 3 November 2011 (UTC)[reply]
"Actual problems" This is called "poisoning the well." It's deemed disruptive in consensus-building.
The question is, is this an overall improvement? I think not.
Better to keep the "What consensus is" introduction concise.
Looking at it briefly, Ludwigs' text seems to be an improvement - it actually attempts to explain things reasonably fully, which is what I would have thought a section titled "What consensus is" ought to be doing. It's not an introduction (that's what the lead is for), this is the heart of the policy, where tricky concepts (as this one undoubtedly is) need to be explained at whatever length is necessary. --Kotniski (talk) 17:48, 3 November 2011 (UTC)[reply]
Ronz: I just wanted to know what your 'actual' objections were. The current 'what consensus is' section is largely meaningless and even ungrammatical at points, so that objection can't really be correct. you still haven't said anything specific about the issues you see, and I can't really credit this as a meaningful objection until you do. But let's wait a day and let others chime in.
"I can't really credit this as a meaningful objection until you do" Being dismissive of others is disruptive to consensus-building. --Ronz (talk) 18:42, 3 November 2011 (UTC)[reply]
I like this text in general. If there is a grammatical error or an unclarity, I do not see it.
I do disagree, however, with "The aim of consensus is to reach understanding, not agreement". No; if someone understands my arguments, and still thinks them evil and mistaken, he is not part of the consensus for them. The aim of consensus is acceptance: that everyone will tolerate the consensus position, without active efforts to overturn it. SeptentrionalisPMAnderson20:38, 3 November 2011 (UTC)[reply]
Good distinction. It's acceptance as opposed to agreement, and "understanding" is a very poor word to use because of its multiple meanings. --Ronz (talk) 22:26, 3 November 2011 (UTC)[reply]
Interesting. The idea behind that, really, was that the more someone understands your perspective, the less likely they are to think your perspective is stupid/evil. They may still think it's mistaken, but mistaken beliefs are grounds for conversation and compromise, rather than combat. In the philosophical presentation of consensus-decision making sources talk about creating the correct form of discussion, because setting up the right form is essential to productive consensus processes; but that form always involves trying to understand what your opposite is saying in his/her terms.
the theory behind consensus decision making is a little high-toned for use on project, but it would be nice to incorporate some of the principles. --Ludwigs223:01, 3 November 2011 (UTC)[reply]
Tout comprendre c'est tout pardonner is an apocryphal piece of late Enlightenment optimism; even Mme. de Staël, to whom it is attributed, never went quite that far. To see it after the twentieth century is astounding. SeptentrionalisPMAnderson00:24, 4 November 2011 (UTC)[reply]
Ludwigs, some good stuff in what you have proposed that I think we shall want to incorporate. However, I'm not so sure that "Consensus is a process", more that, "Consensus is the level of agreement needed for a proposal to be accepted within WP's decision-making process". Cheers, Uniplex (talk) 06:56, 4 November 2011 (UTC)[reply]
Yes, we should really decide how much abuse of the word "consensus" we are prepared to accept in the policy. Really it's a term that Wikipedia has borrowed from the real world and become accustomed to using to describe any facet of its decision-making process, which is a complex and internally inconsistent thing for which there is almost certainly no real word.--Kotniski (talk) 08:23, 4 November 2011 (UTC)[reply]
@PMAnderson: I hate to put it quite this way, but if you want a consensus system, then you have to accept an optimistic view on human nature. if you don't want to accept an optimistic view on human nature, then we should scrap consensus as an ideal and move to a more centralized authoritarian system. I'm trying to work within the principles of the project here, which insist on cood faith, openness, and a degree of mutual trust.
really, the problem here is that the project is caught between two mindsets
there is a strong ideal on-project about freedom, liberty, and personal rights (pure excessive enlightenment idealism)
there is a strong sense of paranoia about the behavior of others (primal clannism, in-group/out-group fears and aggressions)
When you mix those two there are only two possible outcomes (short of authoritarianism): rise above it to something collegial and collaborative, or a descent into Hobbes' world. You can't have both, and if you don't choose consciously you'll always end up with the latter.
In other words, I agree with you, but I think you're wrong. Think about that a bit.
@Uniplex & Kotniski: I'm just reflecting and clarifying a frequent cause of confusion on project. I thinking of the following kind of dispute:
Editor A: "You're not following consensus" (a process claim)
Editor B: "You're not obeying consensus" (an established-rule claim)
The term gets used in both ways, and I don't think people always realize that difference in usage, so I think it's a good idea to clarify it. --Ludwigs215:48, 4 November 2011 (UTC)[reply]
Authoritarianism is certainly a conceivable outcome; it is one way the project could fail. (The result here would be mass abandonment by editors, since emigration is so easy.) The Unbearable Lightness of Being is partly about a combination of Enlightenment idealism and paranoia in the real world, and the effects of its -er- change of state, where emigration was not possible. We will at least avoid defenestration.
Consensus requires more optimism than Hobbes; that's not a high threshold. But it does not require that the Barlanders and the Foolanders come to a reign of sweetness and light by understanding each others' positions; most such pairs understand each others' positions perfectly well now, but abominate them. It merely requires that, most of the time, there will be a sufficient number of uninvolved editors to decide between them, and balance on a reasonable course, while maintaining elementary civility - the advantage of a large editor pool, not saturated with any one or two factions. SeptentrionalisPMAnderson16:43, 4 November 2011 (UTC)[reply]
So in fact we can't say that the (realistic) aim of consensus is for everyone to do anything - those whose views are based on the "wrong" motives may need to be eliminated from the approval process, along with those who persist in disrupting the process in any of various ways. In fact our decision-making process is quite a long way from being based purely on what "consensus" would be generally understood to mean - often decisions are made on the basis of a respected but previously uninvolved adjudicator weighing the arguments and deciding which are stronger (partly on the basis of how many people have been convinced by them), while tending to give a (somewhat arbitrary) amount of extra weight to the position that we should leave things as they are. Making editors happy, though desirable, is (and should be) only a subsidiary goal.--Kotniski (talk) 18:04, 4 November 2011 (UTC)[reply]
It might be a purpose (particularly in the case of the lamer disputes), but the central purpose, surely, is to reach decisions which are the right ones (in terms of Wikipedia's mission and principles).--Kotniski (talk) 17:03, 5 November 2011 (UTC)[reply]
This discussion is so central to the Wikipedia problem…
PMAnderson: I understand your perspective, but I still disagree with it - take that both as a statement and as an example of what I'm talking about. In academic circles (where most everyone is a high-functioning intellectual) you see this all the time: two people who are absolutely dedicated to defeating each other's position, but who go about it in a thoroughly reasonable, thoroughly civil way, with great respect for each other. On wikipedia it's entirely different: people do not easily differentiate between disagreement and disapproval, or agreement and approval: they get confused and think - as you seem to suggest - that understanding=agreement=approval, so one cannot understand an opponent without getting all namby-pamby, lovey-dovey, group-huggish. So instead, they get hostile where they should just respectfully disagree. trust me, academics never get lovey-dovey, but they are very good at working through disagreements.
Kotniski: The system we have now is only a consensus system at the margins, where no one is looking. in the center we have a dysfunctional system where the person/group which most effectively casts him/her/itself as a victim dominates the discussion. that's why this policy needs strengthening and clarification. --Ludwigs220:02, 4 November 2011 (UTC)[reply]
True, but setting up guidelines for civil, collaborative discourse thing a less subjective and less game-able. right now I'm getting massacred elsewhere because editors have managed to warp the concept of consensus discussion so far out of true that any effort to establish a discussion looks uncivil - it's a bizarre manifestation. Being able to point to what a proper discussion should look like would help a lot. --Ludwigs223:23, 4 November 2011 (UTC)[reply]
So to return to Ludwig's original proposed wording at the top of this thread, does anyone have any problem with it except for the sentence "The aim of consensus is to reach understanding, not agreement - it is expected that people can disagree with the outcome but accept it because they understand the principles in play."? If
that sentence were left out (or altered somehow), would we accept the text?--Kotniski (talk) 13:06, 5 November 2011 (UTC)[reply]
It still requires substantial tweaking:
Such discussions are usually closed when they reach a saturation point in which most of the participants share a loose agreement about which policies and principles are pertinent and how they should be applied to the debate is not how I would describe closing; I'm not sure we need this here, but if so, it would be closer to say that discussions are closed when the discussion stops; to avoid editors repeating themselves until everybody else is bored and then declaring victory, we may need to say is substantially over, and few substantive contributions seem likely. SeptentrionalisPMAnderson15:28, 5 November 2011 (UTC)[reply]
Consensus is not always achievable for the same reasons that collaboration sometimes fails - strong ideological convictions, interpersonal frictions, or simple good-faith disagreement can frustrate the natural discussion process. is tendentious; frustrate and natural are emotion-laden and beg the question.
The following sentence is simply wrong; we should encourage, and we must describe, the preferred reaction to no consensus; coming back with an amended proposal which addresses the concerns of both sides. We should also mention that discussions quite often emerge with a result which was neither the original state nor the proposal; otherwise we will be quoted as prohibiting that useful practice. SeptentrionalisPMAnderson15:28, 5 November 2011 (UTC)[reply]
Oh, there was also an objection to "Consensus is a process..."' - could that be fixed by saying something like "Consensus-based decision-making is a process..." or "Consensus is the basis for a process..." ?--Kotniski (talk) 13:08, 5 November 2011 (UTC)[reply]
As something of a newcomer to this page I know little of its history and motivation, however, having thought a little more since my comment above, I have to say that I'd expect a policy page to more prescriptive than descriptive, and that the proposed text seems to be more the latter (though so is the existing text to a certain extent). If we're trying to state "What consensus is", I think we should aim for an as clear and concise definition of such as we can manage. Perhaps (most of) the proposed text would be better as an essay? Uniplex (talk) 17:22, 5 November 2011 (UTC)[reply]
We ought to have less differentiation between prescription and description than most places; see WP:Policy: Functioning policies do describe what we actually do; if there is real consensus for something, it is what most people will actually do when they think about it. SeptentrionalisPMAnderson17:29, 5 November 2011 (UTC)[reply]
Ludwigs, someone once told me that the way to eat an elephant is one mouthful at a time. I'm guessing it might be good advice here; proposing one change (or at least, fewer changes) at time might make for slow progress, but hey, it's progress! Uniplex (talk) 09:02, 6 November 2011 (UTC)[reply]
Uniplex - lol - you apparently have not noticed what a big mouth I have. Strange, because everyone else seems to have.
seriously though, that approach is entirely appropriate for fleshing out a developed concept, but it doesn't work for developing a concept in the first place. The concept of consensus on wikipedia is confused, disjointed, and confounded largely because people have been trying to approach it one bit at a time rather than create an overriding concept; it's a vague hodgepodge rather than a clear principle. 'Consensus' is almost impossible to use effectively on project no one has a clear idea what it means, so everyone interprets it is self-serving ways.
so lets try a revision, incorporating what I read above:
Consensus is a process in which editors seek to reach a mutually acceptable decision through discussion. Sometimes the term is used to refer to an ongoing process (a consensus discussion) and sometimes to the result of a past discussion (a current consensus). In both senses of the term, consensus implies that editors have raised any concerns they think are pertinent and weighed them in the light of the needs of the article and the policies and principles of the encyclopedia. Consensus is the natural process of healthy collaboration, and all editors are expected to make good-faith efforts to engage and respect it.
Consensus is not a numerical process; it is not the same as unanimity or majoritarianism or any other system of 'counting votes'. When things that look like votes happen on wikipedia (as in RfCs, AfD discussions, or other administrative actions), they are primarily intended to solicit broad input in order to ensure that the outcome has broad acceptance across the community. Mutual acceptance is the goal, not complete agreement, and editors should work together to create such outcomes.
Consensus is not always immediately achievable; many factors can derail the discussion process and prevent acceptance from one side or another. When this happens - sometimes referred to as a 'no consensus' result - editors should reevaluate and revise their position to account for the objections raised by others, and then try again. Consensus is an ongoing negotiation, and it may take several attempts to construct something that is mutually acceptable. In the most difficult cases, Wikipedia has several dispute resolution mechanisms which can be used to help create a proper atmosphere for consensus discussion, or to find outside participants to move the discussion past some sticking point.
I think I'm coming to agree with the point that was made before, that we shouldn't abuse language too much, and that we should avoid using "consensus" to mean a process - consensus should mean the (desired) result, not the way of getting there.--Kotniski (talk) 19:16, 6 November 2011 (UTC)[reply]
if you're going to make that distinction, make it the other way - the process is what's important here. not the result. If we restrict consensus solely to the result then what we are going to see (as we already see so much of) is people asserting that a 'consensus' has been established and then aggressively destroying any conversation which might change that state. That is a worst-case scenario in which editors starts using the term 'consensus' to assert more-or-less absolute totalitarian rule.
Don't get me wrong, if you want totalitarianism then we can discuss that - there are advantages to totalitarian systems for a project like this - but I object to saying that we rely on consensus and then acting in a totalitarian fashion. that's just confusing for everyone, and disrupts the project on a fundamental level. --Ludwigs220:19, 6 November 2011 (UTC)[reply]
I didn't mean that we shouldn't talk about the process - just that it would be linguistically more satisfactory not to call the process "consensus", but to call it something like "the consensus-based decision-making process". That way "consensus" has only one meaning (at least, only one spectrum of meanings), and corresponds somewhat better (I think) with real-world usage.--Kotniski (talk) 12:31, 7 November 2011 (UTC)[reply]
Well, if you want to make that distinction clearer then we should talk about 'consensus decisions' and 'consensus discussions' as two separate things. the problem is the people on wikipedia fail to make the proper distinction - when you here someone say 'You're violating consensus!' it's almost never clear whether they mean "you're trying to overturn what we've already agreed on" or "you're not discussing things properly", and that distinction needs to be clear for the conversation to progress. Do you see what I'm getting at? I don't want to restrict the word 'consensus' to one sense or the other, because that will work against the way the word is commonly used, but I do want to clarify the potential for confusion between the two senses. --Ludwigs220:58, 7 November 2011 (UTC)[reply]
Lede
How about we simply remove, "and seems[weasel words] to offer the best method to achieve the goals of the project, including neutrality and verifiability, within its collaborative editing model" and be done with the weasel word? It's redundant with "all legitimate editorial concerns" and while I can see how someone might want to emphasize WP:NPOV at a later time, I don't understand the need for any emphasis of WP:V. And until we have one or more definitions, restore "There is no single definition of what the term "consensus" means." Something like:
Consensus refers to the primary way in which editorial decisions are made on Wikipedia. There is no single definition of what the term "consensus" means – it does not mean that decisions must be unanimous (which, although an ideal result, is not always achievable), but implies that the decision-making process involves an active effort to incorporate all legitimate editorial concerns instead of relying on numerical voting.
Editors usually reach consensus as a natural and inherent product of editing; generally someone makes a change or addition to a page, then everyone who reads it has an opportunity to leave the page as it is or change it. When editors cannot reach agreement by editing, the process of finding a consensus is continued by discussion on the relevant talk pages.
Yes, the weasel words should go. In practice, "usually reach consensus" is perhaps a little too much, "often reach consensus" perhaps. Uniplex (talk) 12:43, 5 November 2011 (UTC)[reply]
Yes, as I suggested before, I think this whole sentence about neutrality, verifiability etc. could happily be dispensed with. Obviously the community uses this method (inasmuch as it can be called a "method") because the community thinks it's the best method. I'd also be happy to dispense with the "no single definition" bit, even though I seem to recall putting it there myself sometime in the past.--Kotniski (talk) 13:02, 5 November 2011 (UTC)[reply]
Existing:
"Consensus" on Wikipedia does not mean that decisions must be unanimous (which, although an ideal result, is not always achievable), but implies that the decision-making process involves an active effort to incorporate all legitimate editorial concerns instead of relying on numerical voting.
Suggested:
"Since the ideal of unanimity is rarely attainable, Wikipedia's decision-making process involves an active effort to find a consensus which incorporates all legitimate editorial concerns without relying on numerical voting."
Other wording might be used to similar intent. My immediate concern was just the sentence structure. LeadSongDogcome howl!15:44, 5 November 2011 (UTC)[reply]
That implication isn't really the one we mean, though - if you say "since the ideal of unanimity is rarely attainable", the next clause would have to be something that means "we don't require unanimity".--Kotniski (talk) 16:56, 5 November 2011 (UTC)[reply]
Which would imply leaving unanimity out of the first sentence and making something like this the second sentence: "Consensus" on Wikipedia does not mean that decisions must be unanimous; the ideal of unanimity is rarely attainable. First define, then qualify. SeptentrionalisPMAnderson17:32, 5 November 2011 (UTC)[reply]
I don't think it's accurate to say that unanimity is rarely attainable. To the contrary, the overwhelming majority of edits are accepted more or less as is. The exceptions stand out because they are contentious, but normally it's not a problem to figure out what to do. Secondly, it's a serious error of judgment to think that we can define consensus as practiced here on Wikipedia without reference to the goals and mission of the site. Definitions and rules are reductive. The practice of consensus on Wikipedia is what this article should try to point to, and a definition will not accomplish that goal. --Ring Cinema (talk) 00:58, 6 November 2011 (UTC)[reply]
The problem with having no definition is that we are redefining the word. Non-Wikipedians have every reason to be baffled by any of our "consensus" discussions; to anyone consulting a competent dictionary, consensus is unanimity. We must say (and where better than here?) that we mean something different. SeptentrionalisPMAnderson03:01, 6 November 2011 (UTC)[reply]
I agree with that sentiment. However, there is not a widespread confusion about the practice here. More than likely that is partly attributable to what it has said on this page heretofore. Although there is a lot of self-satisfaction in taking out the "weasel words", I see it differently. As far as I can tell, it is exactly correct that consensus "seems" to be the best method to achieve Wikipedia aims. Sometimes "seems" is the fact of the matter. As it has read, it says that we do it this way because we think this is what works, even though we can't be sure this is why it works. The practice has to be connected to the goals to make it clear that if the practice doesn't serve the goal, it is not correct. So this is an important lynchpin to my way of thinking. Over-specification of the concept makes it brittle and that is not what will work. Instead, we need a flexible concept that allows sensible people to use their sensible judgement to solve the problem of putting together articles. Again, we want to point in the right direction, not give a step by step, inch by inch diagram of how we think someone else should go. If we set our egos aside, I don't see why we want to alter what is working unless it is incremental. --Ring Cinema (talk) 03:39, 6 November 2011 (UTC)[reply]
Ronz's proposal does not state that unanimity "is rarely attainable"; it has "is not always achievable". Also, it has "all legitimate editorial concerns" which of course includes WP policies. An observation that something seems to work is not particularly helpful: it just leaves the reader wondering why the statement is there. If we want to be more explicit, I suggest to incorporate the sentiment of this sentence from the article body: "Consensus is ultimately determined by the quality of the arguments given for and against an issue, as viewed through the lens of Wikipedia policy, not by a simple counted majority." Uniplex (talk) 08:43, 6 November 2011 (UTC)[reply]
Right, but someone above said unanimity is rarely attainable and that is not a good justification for Ronz's proposal. To be clear, I do not disagree with the ideas in Ronz's proposal and the definitions it offers "seem" to state things well. However, Uniplex, I don't think that the status quo says simply that things seem to work; it says that Wikipedia goals are served by consensus to the best of our knowledge. That is a good attempt to be extremely accurate. It states as clearly as possible why consensus remains the policy without going beyond the limit of understanding. In this article, it is important to tie the policy on consensus to the larger means and ends of Wikipedia. There is plenty of time later in the article to cash out the practice (and the practice -- not the meaning of a word -- is what matters, since 'consensus' is just the verbal hook on which we hang this). --Ring Cinema (talk) 13:49, 6 November 2011 (UTC)[reply]
Ok, my "rarely" was overly pessimistic. How about:
"Since the ideal of unanimity is not always attainable, Wikipedia's decision-making process involves an active effort to find a consensus which incorporates all legitimate editorial concerns without relying on numerical voting."LeadSongDogcome howl!14:44, 6 November 2011 (UTC)[reply]
It still seems to be a nonsequitur. The two clauses are not related in the way that the word "since" would imply: the first clause is not the reason for the second.--Kotniski (talk) 18:40, 6 November 2011 (UTC)[reply]
well, if we want to revise the lead as well, I'd suggest the following (mostly because it follows better compositional style):
Decision-making on Wikipedia uses a consensus model in which editors work collaboratively to seek mutually acceptable results. Often this is a natural, inherent and transparent by-product of collaborative editing; changes are made to an article, other editors revise or amend those changes, and eventually the article stabilizes on a new 'consensus' version. When it does not work transparently, editors turn to the article's talk page to discuss article changes overtly.
Consensus is not a 'vote counting' model. Instead, editors suggest revisions, raise policy considerations, debate the relative importance of particular points, and in general try to convince other editors that particular changes are in the best interest of an encyclopedic presentation of the article topic. The ultimate goal is to create a good article, but editors should keep in mind that there will be differences of opinion about what constitutes a good article. Consensus discussion is as much about finding common ground about what makes for a good presentation of the topic as it is about setting specific disputes.
Yeah, but perhaps a bit unfriendly to the average reader, who just wants an explanation in everyday language, not a philosophical treatise right at the start. We need to be clear about the basic points that we all know, but a reader coming from the real world might not be aware of (chiefly that around here, "consensus" doesn't mean that nothing can be done without unanimity).--Kotniski (talk) 19:11, 6 November 2011 (UTC)[reply]
well, point taken, but allow me to make the (somewhat snarky) point that it is far easier to 'dumb-down' a solid philosophical concept than to 'smarten-up' a vague hodgepodge of disconnected sentiments. Saying that it's 'a bit unfriendly' is not as helpful as it could be: either it's a bit unfriendly but right, or a bit unfriendly but wrong. if it's the former then we can work to make it more accessible. --Ludwigs220:24, 6 November 2011 (UTC)[reply]
Let's try that with verbs; I do not intend any change in meaning:
Wikipedians make decisions by what we call consensus: editors work collaboratively to reach a mutually acceptable text for the encyclopedia. Often this is a natural, inherent and transparent by-product of collaborative editing; changes are made to an article, other editors revise or amend those changes, and eventually the article stabilizes on a new 'consensus' version. When it does not work transparently, editors turn to the article's talk page to discuss article changes overtly. Much the same process is used for Wikipedia's internal pages.
Consensus does not count votes. Instead, editors suggest revisions, raise policy considerations, debate the relative importance of particular points, and in general try to convince other editors that particular changes are in the best interest of an encyclopedic presentation of the article topic. The ultimate goal is to create a good article, but editors should keep in mind that there will be differences of opinion about what constitutes a good article. Consensus discussion is as much about finding common ground about what makes for a good presentation of the topic as it is about settling specific disputes.
This is some progress but it is getting away from plain English. It has been implied here that 'consensus' should be so well laid out in this article that a neophyte can read it and jump right in with perfect knowledge of Wikipedia practices on consensus. That is not a reasonable or attainable goal. The practice of consensus will only be understood by experience and no definition will express that. Please, get used to that idea; consensus cannot be defined adequately, it can only be shown. Also, to the extent this article doesn't connect the practice of consensus to the larger aims of the encyclopedia it will be a failure. Is it a non sequitur to connect consensus to other Wikipedia values? Well, they must be connected because consensus is the means to the attainment of those ends! Consensus is not the chosen method on account of metaphysical or ideological reasons. No, it is a practical method and its implementation has practical results. We need to place consensus within the galaxy of Wikipedia values to show how to use it. That is not trivial; it is essential. --Ring Cinema (talk) 08:24, 7 November 2011 (UTC)[reply]
So what do you propose doing about the sentence in the lede that currently serves that purpose, but has been marked (for understandable reasons) with a "weasel" tag?--Kotniski (talk) 12:37, 7 November 2011 (UTC)[reply]
Ring: I dispute your claim that consensus cannot be defined adequately. In fact it's easy enough to do it (as we have started to do here), it's just not easy to put it in terms that a layman will immediately find accessible. However, it would be much better to have a correct-but-not-immediately-accessible definition than to have immediately accessible handwaving. we should give them a correct definition to ponder, and then let them understand that definition through experience; we should not leave them hanging with a vague "you'll figure it out in time" thing. Proper instruction requires properly framing material so that people have something to work with. --Ludwigs215:37, 7 November 2011 (UTC)[reply]
Kotniski: I find the weasel tag inaccurate. Sometimes "seems" is the right word. It's in the language for a reason. Knee-jerk labeling is cheap, not smart. I touched on this above, but it says "seems" because, among other reasons, there is not a counterfactual comparison or perspective of perfect knowledge from which to judge. It seems that consensus does something for Wikipedia that, say, majoritarianism or authoritarianism does not (and it goes without saying that verifiability and reliability are more important values than consensus).
Ludwig: I am not saying that we shouldn't do our best to define the term. That has always been in the article and I am on board with the bit above that tries to improve on that. But it's not handwaving to say that the practice is what matters and to be aware that definitions are reductive, especially when the matter at hand is complex. Again, the idea that it is handwaving to allow the editors to work out the practice as needed is a bit too hierarchical for our purposes. If there was no definition it would be preferable to an over-determination that does not speak to the ends that consensus seems to serve. --Ring Cinema (talk) 16:52, 7 November 2011 (UTC)[reply]
While the term "seems" is sometimes the right word, it has no place here. Seems means to "give the impression". We should not be talking about impressions here.
I suggest we take Ronz' original proposal, and incorporate suggestions from LeadSongDog (parenthetic clause on unanimity is awkward), PMA (define first, then qualify), Ring Cinema (saying unanimity is rarely attainable is inaccurate), Uniplex (ultimately determined by evaluating arguments through lens of policy), Kotniski and others...
Consensus is the primary way in which editorial decisions are made on Wikipedia. Consensus is a decision-making process in which an active effort is made to incorporate all legitimate editorial concerns instead of relying on numerical voting. Unanimity is an ideal result, but not always achievable.
Editors usually reach consensus as a natural and inherent product of editing; generally someone makes a change or addition to a page, then everyone who reads it has an opportunity to leave the page as it is or change it. When editors cannot reach agreement by editing, the process of finding a consensus is continued by discussion on the relevant talk pages, and ultimately determined by evaluating arguments through the lens of Wikipedia policy and guidelines because they reflect the consensus of the broad community.
I think it's an improvement, but there remain some issues. Is consensus really a process, an objective, or simply a transient state of a dynamic article (or paragraph, or sentence)? Do we not really mean "that wording to which the editors currently consent"? The "process" is more applicable to consensus building than the consensus itself. We should also make rather more explicit that the dictionary "majority rule" definition is not what we mean. LeadSongDogcome howl!19:00, 7 November 2011 (UTC)[reply]
Good points! Is this explicit enough?
Consensus, or broad agreement, is the primary determining factor in making editorial decisions on Wikipedia. Consensus building is a decision-making process in which an active effort is made to incorporate all legitimate editorial concerns instead of relying on numerical voting. Unanimity is an ideal result, but not always achievable.
Editors usually reach consensus as a natural and inherent aspect of editing; generally someone makes a change or addition to a page, then everyone who reads it has an opportunity to leave the page as it is or change it. When editors cannot reach agreement by editing, the process of finding a consensus is continued by discussion on the relevant talk pages. To separate legitimate editing concerns from personal preferences, determining consensus is not a matter of counting all opinions and positions on a given issue, but is evaluating arguments through the lens of applicable Wikipedia policy and guidelines because they reflect the consensus of the broad community.
I don't think so, and, in fact, suggest de-emphasizing if not eliminating other references to "compromise" on the page. I mean, what exactly in the consensus building process is supposed to be compromised? What standards are to be lowered? What principles are to be ignored? What policies dismissed? --Born2cycle (talk) 20:54, 7 November 2011 (UTC)[reply]
You're right: the first port of call should be to evaluate arguments through the lens etc. Maybe we can lose the final "because" clause? Seems to make the sentence a bit long and I'm not sure that following Ps & Gs needs much justification. Uniplex (talk) 21:25, 7 November 2011 (UTC)[reply]
Yeah, it is clumsy and needs improvement, but let's not toss the baby. I think it's important to convey that it's about determining the consensus of the community, not about determining the consensus of the handful or whatever of editors who happen to be paying attention. Is this better, or worse?
Consensus, or broad agreement, is the primary determining factor in making editorial decisions on Wikipedia. Consensus determination is a decision-making process in which an active effort is made to incorporate all legitimate editorial concerns instead of relying on numerical voting. Unanimity is an ideal result, but not always achievable.
Editors usually comply with consensus as a natural and inherent aspect of editing; generally someone makes a change or addition to a page, then everyone who reads it has an opportunity to leave the page as it is or change it. When editors cannot reach agreement by editing, the process of determining consensus is continued by discussion on the relevant talk pages. To separate legitimate editing concerns from personal preferences, determining consensus is not a matter of counting all opinions and positions on a given issue, but is evaluating arguments through the lens of applicable Wikipedia policy and guidelines.
Since policy and guidelines reflect broad consensus at Wikipedia, following them is how issues are resolved to be consistent with broad consensus. Productive discussions are less about what individual editors think should be done in a given situation, and more about what agreed-upon applicable policies, guidelines and conventions indicate should be done.
No, that is wrong, Born2cycle. Seems doesn't mean to "give the impression" -- at least not here. Rather, in this context it means what I said: it means that we don't know for sure if the practice is responsible for the result but we think so. And really 'seems' is only the lesser part. It is a naive mistake to divorce the practice of consensus from the ends that it serves. By stating the ends and their relationship to the means, we offer the more enduring guide for the editors. All your definitions are not worth much if they don't lead to the a better practice. If the purpose of this revision is to change that practice, what is the reason for that change? If the purpose is not to change the practice, why change the text? I'm sensing a lot of unfortunate ego and it's not pretty. --Ring Cinema (talk) 04:49, 8 November 2011 (UTC)[reply]
Wrong? What is wrong? How do you distinguish "give the impression" and "we don't know for sure if the practice is responsible for the result but we think so"? I know they don't mean exactly the same thing, but it's pretty damn close. For example, if I say that I'm under the impression that the U.S. went to war in Iraq because of oil, and you say don't know for sure if the U.S. went to war in Iraq because of oil, but you think that's why, aren't we expressing pretty much the same opinion?
But your point about needing to explain why we use consensus is well taken, though I think we can and should do it without using "seems" or the synonymous terms or phrases just discussed.
The purpose of these revisions, as others have noted, is to better explain what consensus at Wikipedia means, because it seems to be a source of confusion to newcomers. --Born2cycle (talk) 08:46, 8 November 2011 (UTC)[reply]
Some copy-editing and attempt to incorporate above concerns:
The term consensus refers to the process of making editorial decisions on Wikipedia, and describes the broad agreement, and reflection of Wikipedia's policies, that this entails.
Consensus usually forms naturally, as an inherent aspect of editing: when someone makes a change or addition to a page, everyone who reads it has an opportunity to leave the page as it is or change it further, towards the goals of the encyclopedia.
Occasionally, when editors do not reach agreement by editing, the process of determining consensus is continued by discussion on the relevant talk pages. In order to keep aside personal opinions on a given issue, this does not rely on numerical voting. Instead, editors build consensus by making an active effort to incorporate all legitimate editorial concerns, evaluating given arguments through the lens of applicable Wikipedia policy and guidelines. Unanimity is an ideal result, but is not always achievable.
Since policy and guidelines have broad consensus at Wikipedia, an editorial decision which is based on them, and made by a small group of editors, inherently achieves broad consensus: productive discussions are less about what individual editors think should be done in a given situation, and more about what agreed-upon applicable policies, guidelines and conventions indicate should be done.
I must say I don't much like the increased emphasis on policies and guidelines. The stupidest results seem to come out of discussions where people argue that we have to do things some way "because WP:XYZ says so". Policies and guidelines may or may not reflect community consensus, may or may not reflect Wikipedia's practice, and may or may not be sensibly applicable in a given situation. The "lens" through which we should evaluate arguments is that of Wikipedia's principles and practices, not its pseudo-laws.--Kotniski (talk) 08:44, 8 November 2011 (UTC)[reply]
Well certainly the polices and guidelines do not reflect broad consensus perfectly, but reflect broad consensus is what they're supposed to do, and largely do do. Otherwise, what would be the point in having them?
I mean, just look at what we're trying to do here. We're trying to convey as clearly as we can what CONSENSUS actually means in Wikipedia, in terms of how most people in the community actually see it.
The same principle applies to the creation and revising of any policy and guideline. It's not a perfect system (if nothing else because it's trying to reflect something, consensus about this and that and everything else, that is itself constantly evolving) but it is a self-correcting system (we're in the process of "correcting" right now, hopefully). When we're trying to apply policy/guidelines in a given situation, and realize they say something that does not reflect broad consensus, we can propose revising it accordingly.
So when we're trying to decide what policies and guidelines should say, we should be giving consideration to what broad consensus is about the particular relevant questions and issues. But when we're trying to determine what consensus is about a given editorial questions, we should be considering what broad consensus is about that particular issue as reflected in the appropriate/applicable policies and guidelines. That's what gives us common ground, and the ability to achieve agreement. Otherwise we all come to the table working with a different set of rules - you drive on the right, I'll drive on the left... you go on green, I go on red... no problem! Yeah, right...
I notice that "as viewed through the lens of Wikipedia policy" is long-standing in the current body and we've qualified it in the proposed lede with "applicable", so maybe it's not so far off the mark? If we change "policies" in the 1st para to "principles" (we have "goals" in the 2nd para and then "applicible Ps & Gs" in the third), would that be about right? Uniplex (talk) 10:00, 8 November 2011 (UTC)[reply]
I would keep any mention of policies and guidelines out of the lead altogether. This is supposed to be about consensus, and consensus (if it exists) takes precedence over pretty much any written rule. The rather complex interaction between consensus and policies/guidelines can be explained at appropriate length in a later section of the page.--Kotniski (talk) 10:06, 8 November 2011 (UTC)[reply]
Wow. The consensus that matters most in WP decision-making is the consensus of the community at large. This is what consensus is at its very essence. The consensus reached by two, three, six or even 56 editors in a given discussion pales in terms of importance in comparison to the broad consensus that is supposed to be, and is largely, conveyed in policy and guidelines. To leave this fundamental concept out of the lede is to completely misconstrue what consensus is. It's like leaving out wheels in the lede of Bicycle. --Born2cycle (talk) 10:20, 8 November 2011 (UTC)[reply]
Well, I don't think policy and guidelines are fundamental to consensus. To the extent that they do reflect the genuine broad consensus of the community, and to the extent that the general rules they contain can be expected to make sense in a given specific situation, they certainly can be helpful in deciding matters where no clear local consensus exists. But both of those extents are limited, and anyway, this is really a part of the topic of what to do if consensus can't be clearly determined, not the fundamental essence of consensus (part of which is that decisions are made not by following set rules, but by deciding collaboratively what is the right thing to do).--Kotniski (talk) 10:32, 8 November 2011 (UTC)[reply]
Okay, fair enough. But what is fundamental is that in a given situation the consensus of those participating based on who-knows-what is not what should be determined; rather, what should be determined is what the consensus of the community at large is with respect to that situation. For example, if the 12 people who can't agree on whether to name an article A or B decide to name it A and B, that's a local consensus, but it's trumped by the consensus of the community at large that says the title has to be A or B (or maybe C), but notA and B. Or just because the only 2 guys working on a BLP reach a consensus on whether to include some unsourced controversial statement does not mean it's okay to include it... We have a higher standard, and that is the broad consensus of the community. That's why the discussions in determining consensus should not be about determining a consensus that merely appeases those who happen to be involved, but it should be about what broad community consensus is with respect to that particular issue. That's the fundamental part in theory. In practice, it means taking policy and guidelines into consideration, because that's how we collaboratively agree in specificity what broad community consensus is. So policy and guidelines are fundamental to consensus determination, in this indirect sense. --Born2cycle (talk) 10:47, 8 November 2011 (UTC)[reply]
As to the latest revision, we're not addressing Ring Cinema's concern about not explaining why we use consensus in WP, which used to be explained with the "seems" phrase. I find the clause about policy in the first sentence to be awkward - I don't think we need it. And "consensus" is not a process - but a reflect of broad agreement. Also, I think the first clause in the last paragraph, "Since policy and guidelines have broad consensus at Wikipedia, ...", really needs to say reflects rather than "have" (and if we were to keep "have", it would be "have broad consensus support". I attempt to address these concerns here:
Consensus is the primary determining factor in making editorial decisions on Wikipedia. It is the broad agreement that is achieved in the decision-making process. Consensus decision-making is the best known way to achieve the goals of the project, including neutrality and verifiability, within the Wikipedia collaborative editing model.
Consensus usually forms naturally, as an inherent aspect of editing: when someone makes a change or addition to a page, everyone who reads it has an opportunity to leave the page as it is or change it further, towards the goals of the encyclopedia.
Occasionally, when editors do not reach agreement by editing, the process of determining consensus is continued by discussion on the relevant talk pages. In order to keep aside personal opinions on a given issue, this does not rely on numerical voting. Instead, editors build consensus by making an active effort to incorporate all legitimate editorial concerns, evaluating given arguments through the lens of applicable Wikipedia policy and guidelines. Unanimity is an ideal result, but is not always achievable.
Since policy and guidelines reflect broad consensus at Wikipedia, an editorial decision which is based on them, even though made by a small group of editors, inherently achieves broad consensus: so, productive discussions are less about what individual editors think should be done in a given situation, and more about what agreed-upon applicable policies, guidelines and conventions indicate should be done.
I still disagree strongly with the last part in particular - my experience is that discussions become productive when people stop mindlessly saying "per WP:XYZ" and start giving genuine arguments as to why the encyclopedia is improved by doing one thing rather than another.--Kotniski (talk) 10:48, 8 November 2011 (UTC)[reply]
Depends on the situation; may be simply common sense about what is going to inform users better or help them more, may be a (hopefully shared) awareness of what Wikipedia's principles and practices are. Policies and guidelines are supposed to describe the latter, but do so with limited success, and so their use as sources of wisdom should be limited accordingly.--Kotniski (talk) 11:21, 8 November 2011 (UTC)[reply]
I think the situation Kotniski is describing should be covered by IAR (also policy), but also that the need to IAR should be relatively rare. If so, maybe we can put the lid on the lede for now, and move on to the body to address these and other details? Uniplex (talk) 11:04, 8 November 2011 (UTC)[reply]
Lede (random break 1)
I agree IAR is more than sufficient to cover those situations where community consensus does not apply, or is not reflected well in policy, etc. I do have a suggestion for a change to the first part, and some slight rewording. Well, the introduction of the term, community consensus. Here it is.
Consensus in general, or community consensus, is what the community broadly agrees about, but "consensus" can also refer to agreement among those participating in a given discussion. Consensus decision-making is the best known way to achieve the goals of the project, including neutrality and verifiability, within the Wikipedia collaborative editing model.
Consensus usually forms naturally, as an inherent aspect of editing: when someone makes a change or addition to a page, everyone who reads it has an opportunity to leave the page as it is or change it further, towards the goals of the encyclopedia.
Occasionally, when editors do not reach agreement by editing, the process of determining consensus is continued by discussion on the relevant talk pages. In order to keep aside personal opinions on a given issue, this does not rely on numerical voting. Instead, editors build consensus by making an active effort to incorporate all legitimate editorial concerns, evaluating given arguments through the lens of applicable Wikipedia policy and guidelines. Unanimity is an ideal result, but is not always achievable.
Since policy and guidelines reflect community consensus at Wikipedia, an editorial decision which is based on them, even though made by a small group of editors, inherently complies with community consensus: so, productive discussions are less about what individual editors think should be done in a given situation, and more about what applicable policies, guidelines and conventions indicate should be done.
These successive attempts to write the lead seem to me to be gradually making things worse; they make the whole matter even less readily comprehensible than it is at the moment, and introduce extremely dubious new concepts and simply untrue statements, such as "policy and guidelines reflect community consensus" - they might be supposed to, but we all know they often don't; and even when they do, they don't necessarily reflect it in regard to the matter at hand (which will often involve aspects the community never considered in its deliberations, if there were any, on the wording of the general rule).--Kotniski (talk) 11:21, 8 November 2011 (UTC)[reply]
The concept of local versus broad consensus is already discussed in the body, so I wouldn't describe it as new as such; as for the term we adopt for this, I've no particular preference. I tend to think that policy and guidelines do reflect community consensus, at least to the best of our ability. Granted, they may be incomplete, or they may be "wrong", but in these cases, the best course of action is often to address this directly (at guideline talk) rather than at article talk. Again, I'd suggest that this could be covered in the body. Uniplex (talk) 11:47, 8 November 2011 (UTC)[reply]
Kotniski's points above are resonating with me. Editorial discussions that turn on policy points don't make for good articles. And to me it is a bit of a conceit that policy reflects a broader consensus or participation of WP editors. Who is here but us chickens? Just a few of us, it seems, self-appointed experts all. I also would offer for consideration the view that the sort of "defining" evident here easily goes astray because of the easy case/hard case problem, i.e. it is easy to say what consensus is when there's agreement, but what is it when editors dig in their heels? It seems useless, yet it remains the method by which editorial decisions are made -- supposedly. To me, given that there is no crisis of consensus in the land, we should be making incremental changes only and observing their effects before more changes are considered. The best argument for a rewrite is that supposedly a newcomer won't understand consensus as practiced here, but I would say that is perfectly all right. The practice is in the editors, not in the abstractions of a definition which will with 100% certainty fall short of explicating the practice. --Ring Cinema (talk) 15:46, 8 November 2011 (UTC)[reply]
Once again we return to the descriptive/prescriptive divide. In practice, we start out just doing something, then someone describes it, and if that description is well regarded it is used until it fossilizes into policy over time. We're losing the point with abstract discussion though. We all know that the vast majority of constructive editorial decisions are tiny in scope, often just one or two words changed with little if any discussion. If I may suggest a more descriptive lede:
As we incrementally change article text, wording which survives does so by the consent of all editors, expressing our consensus.
I too understand and agree with Kotniski's point. Policy does not always reflect community consensus. But let's not throw out the baby with the bathwater. There is a big difference between local and community consensus, and we might as well make the distinction in the lede, so we can talk in terms of using local consensus to determine what community consensus is and how it applies in a given specific situation. The community consensus reflected in applicable policy is often instrumental in that process.
Consensus in general, or community consensus, is what the community broadly agrees about, but "consensus" can also refer to agreement among those participating in a given discussion, often regarding how community consensus applies in a specific situation. Consensus decision-making is the best known way to achieve the goals of the project, including neutrality and verifiability, within the Wikipedia collaborative editing model.
Consensus usually forms naturally, as an inherent aspect of editing: when someone makes a change or addition to a page, everyone who reads it has an opportunity to leave the page as it is or change it further, towards the goals of the encyclopedia. Wording which survives incremental changes does so by the consent of all involved editors, and ultimately represents the consensus perspective.
Occasionally, when editors do not reach agreement by editing, the process of determining consensus is continued by discussion on the relevant talk pages. In order to keep aside personal opinions on a given issue, this does not rely on numerical voting. Instead, editors build consensus by making an active effort to incorporate all legitimate editorial concerns, evaluating given arguments through the lens of applicable Wikipedia policy and guidelines. Unanimity is an ideal result, but is not always achievable.
Since policy and guidelines generally reflect community consensus at Wikipedia, an editorial decision which is based on them, even though made by a small group of editors, usually inherently complies with community consensus: so, discussions consistent with community consensus are usually less about what individual editors think should be done in a given situation, and more about what applicable policies, guidelines and conventions indicate should be done.
Sorry, but no it does not. Elegance matters, particularly in a lede. If we ask newbie editors to wade through something that complex, the few who do will not understand it anyhow. The application of consensus to policies and guidelines is far from being the main application, in fact it should probably not make the immediate following paragraphs. Most editors are not "policy wonks" and should not be burdened with lengthy discussions of the philosophy behind the making of policy. They simply want to know how to go about editing articles. LeadSongDogcome howl!17:46, 8 November 2011 (UTC)[reply]
I agree editors should not be burdened with lengthy discussions of the philosophy behind the making of policy. This does not do that; it does not discuss the making of policy at all. It simply explains that community consensus can often be found in policy because policy usually/generally reflects community consensus. In order to removal all mention of the making of policy in this wording, and how consensus is applied to polices and guidelines, not one byte would have to change. --Born2cycle (talk) 18:06, 8 November 2011 (UTC)[reply]
(ec) I feel the same way as LeadSongDog. Some of these points might be usefully made somewhere on the page. But this is not how the lead should be written. --Kotniski (talk) 18:08, 8 November 2011 (UTC)[reply]
Well, can you be more specific? I honestly don't understand LSD's objection at all since the wording says nothing about making policy. --Born2cycle (talk) 18:14, 8 November 2011 (UTC)[reply]
In my case, the objection (apart from the ones I've made before) is simply that this is too complex and unfathomable for the lead to this page, which ought to be something that introduces the topic in a reasonably transparent way for those who might be totally unfamiliar with it. --Kotniski (talk) 18:25, 8 November 2011 (UTC)[reply]
I thought I addressed your earlier objections. Which do you believe I did not address?
Well, it's now four short paragraphs. If that's too complex and unfathomable, I don't know what to say. There are a number of key points regarding consensus, and it merely touches on each of them. Isn't that what a lead is supposed to do? --Born2cycle (talk) 18:37, 8 November 2011 (UTC)[reply]
I just arbitrarily chose to look at WP:AGF... check it out. Its lead is also very similar to this - four short paragraphs. Is that too complex and unfathomable too? --Born2cycle (talk) 18:39, 8 November 2011 (UTC)[reply]
Does anyone remember what started this long discussion? It was a concern about how to understand consensus in the context of revising a core sourcing policy. "Follow the policy" is not a useful answer to "What shall we say in this policy?"
I like pointing out the role of policies and guidelines, but it's a very much article-focused approach, and consensus is not an exclusively article-focused issue. That paragraph should probably be qualified suitably, with words like "in resolving disputes over article content", so we're clear that "follow the policy" doesn't actually apply to 100% of consensus-related disputes. WhatamIdoing (talk) 19:05, 8 November 2011 (UTC)[reply]
I pasted the above text by Born2cycle in this readability calculator and found that it requires an estimated six years of post-graduate education to understand (grade 22). Yes, that is too complex and unfathomable. The wording I suggested came to grade 14, but I suppose it could be further simplified:
As we move from one change to another, text survives by the consent of all editors, expressing our consensus.
Well, I'm sure it can be explained better, but that might mean making it longer. But for now let's focus on whether we have local consensus here on whether the proposed wording accurately describes what consensus is. --Born2cycle (talk) 19:46, 8 November 2011 (UTC)[reply]
WhatAmIDoing, hmm, AGF (as an example) is a consensus-supported policy that applies in all discussions whether about article content or not. So, when we comply with WP:AGF we are also doing what community consensus wants us to do. So, maybe I'm missing something, but I think following (applicable) policy does apply to all disputes. No?
And isn't it useful, in explaining what consensus is, to point out that simply following policies and guidelines is usually being consistent with consensus? --Born2cycle (talk) 19:46, 8 November 2011 (UTC)[reply]
"Follow the policy" tends to result in proposals being mindlessly opposed on the (specious) grounds the rules say that we have to follow the (existing) policy. We have no policies that tell us what WP:V's lead should say; we have many advice pages that tell us what the lead to Barack Obama should say. WhatamIdoing (talk) 07:06, 9 November 2011 (UTC)[reply]
This seems to be hard work. Perhaps, for now, we should shift focus back to the beginning of this discussion, i.e. what to do about the weaselly lead sentence. Maybe:
In Wikipedia, consensus is the broad agreement, in accordance with our core principles, that we seek in making editorial decisions.
Wait, though I like this wording, the latest proposed lede above scores no worse in terms of complexity than the current lede, so this complexity is no reason to abandon this effort and start from scratch again. The weasel wording in the current lede was removed from the latest proposed lede many versions back. --Born2cycle (talk) 20:30, 8 November 2011 (UTC)[reply]
True, but there were concerns that the proposed lead sentences were not as accessible or engaging as perhaps they should be, especially to the new editor. And I wonder if we might faster progress by taking smaller steps (including where possible, the above efforts). Uniplex (talk) 20:52, 8 November 2011 (UTC)[reply]
Yes, elegance is challenging. It is worth the effort. This lede "sentence" is difficult for no good reason:
Consensus in general, or community consensus, is what the community broadly agrees about, but "consensus" can also refer to agreement among those participating in a given discussion, often regarding how community consensus applies in a specific situation.
In fact it isn't even a real sentence. It is comma spliced to death. The initial "or" can be taken two ways: "these are synonyms" or "these are alternatives". The following "is" (rather than "are") implies that they are intended as synonyms. In that case we should simply lose either "Consensus in general, or" or else "or community consensus," from the lede sentence. The part following the word "but" is also simply wrong. It implies that sometimes only the participants consent. In truth the silence of eligible non-participants also deems consent until they end that silence. LeadSongDogcome howl!21:24, 8 November 2011 (UTC)[reply]
Lede rewrite - making it more accessible
Uniplex, we got here pretty much one small step at a time. I believe we have addressed all concerns with respect to what it says, now we're working on making it more accessible. To that end, I incorporate LDS's latest suggestions.
Consensus is either broad community agreement, or agreement among those paying attention to a given discussion. Consensus decision-making is the best known way to achieve the goals of the project, including neutrality and verifiability, within the Wikipedia collaborative editing model.
Consensus usually forms naturally, as an inherent aspect of editing: when someone makes a change or addition to a page, everyone who reads it has an opportunity to leave the page as it is or change it further, towards the goals of the encyclopedia. Wording which survives incremental changes does so by the consent of all involved editors, and ultimately represents the consensus perspective.
Occasionally, when editors do not reach agreement by editing, the process of determining consensus is continued by discussion on the relevant talk pages. In order to keep aside personal opinions on a given issue, this does not rely on numerical voting. Instead, editors build consensus by making an active effort to incorporate all legitimate editorial concerns, evaluating given arguments through the lens of applicable Wikipedia policy and guidelines. Unanimity is an ideal result, but is not always achievable.
Since policy and guidelines generally reflect community consensus at Wikipedia, an editorial decision which is based on them, even though made by a small group of editors, usually inherently complies with community consensus: so, discussions consistent with community consensus are usually less about what individual editors think should be done in a given situation, and more about what applicable policies, guidelines and conventions indicate should be done.
That first para gets down to a 19, but it can get to 12 by leaving out externalia.
Consensus is either broad community agreement, or agreement among those paying attention to a given discussion. Coming to a consensus is the best known way to achieve the goals of the project.
Consensus forms naturally as an inherent aspect of editing. Whenever someone edits a page it is subject to further change by others. Language that survives intact represents the consensus perspective.
That gives us:
Consensus is either broad community agreement, or agreement among those paying attention to a given discussion. Coming to a consensus is the best known way to achieve the goals of the project.
Consensus forms naturally as an inherent aspect of editing. Whenever someone edits a page it is subject to further change by others. Language that survives intact represents the consensus perspective.
Occasionally, when editors do not reach agreement by editing, the process of determining consensus is continued by discussion on the relevant talk pages. In order to keep aside personal opinions on a given issue, this does not rely on numerical voting. Instead, editors build consensus by making an active effort to incorporate all legitimate editorial concerns, evaluating given arguments through the lens of applicable Wikipedia policy and guidelines. Unanimity is an ideal result, but is not always achievable.
Since policy and guidelines generally reflect community consensus at Wikipedia, an editorial decision which is based on them, even though made by a small group of editors, usually inherently complies with community consensus: so, discussions consistent with community consensus are usually less about what individual editors think should be done in a given situation, and more about what applicable policies, guidelines and conventions indicate should be done.
Got para 4 from a Grade 37 (!!!) down to a Grade 16. para 3 is already a Grade 17; not bad.
Consensus is either broad community agreement, or agreement among those paying attention to a given discussion. Coming to a consensus is the best known way to achieve the goals of the project.
Consensus forms naturally as an inherent aspect of editing. Whenever someone edits a page it is subject to further change by others. Language that survives intact represents the consensus perspective.
Occasionally, when editors do not reach agreement by editing, the process of determining consensus is continued by discussion on the relevant talk pages. In order to keep aside personal opinions on a given issue, this does not rely on numerical voting. Instead, editors build consensus by making an active effort to incorporate all legitimate editorial concerns, evaluating given arguments through the lens of applicable Wikipedia policy and guidelines. Unanimity is an ideal result, but is not always achievable.
Policy is usually an expression of community consensus. So, editing decisions that are based on policy usually comply with community consensus. Therefore, for decisions that achieve community consensus, discussions are usually less about what individual editors think should be done, and more about what policy indicates should be done.
Occasionally, when editors do not reach agreement by editing, discussion on the relevant talk pages continues the process toward consensus. To avoid personal opinions on a given issue, we do not rely on majority voting. Instead, we build consensus by working to incorporate all legitimate concerns, seen through the lens of applicable policies and guidelines. Unanimity is ideal, but is not always achievable.
Policy expresses community consensus. Editing decisions that are based on policy usually gain community consensus. Discussions that form consensus are usually less about editor opinions than about what policy indicates should be done.
NOTE: Ease varies from approximately 90 (comic books) to 10 (legalese) and lower (I've seen negative scores).
Para 1: Grade 12/Ease 40
Para 2: Grade 13/Ease 19
Para 3: Grade 14/Ease 20
Para 4: Grade 12/Ease 30
Consensus is either broad community agreement, or agreement among those paying attention to a given discussion. Coming to a consensus is the best known way to achieve the goals of the project.
Consensus forms naturally as an inherent aspect of editing. Whenever someone edits a page it is subject to further change by others. Language that survives intact represents the consensus perspective.
Occasionally, when editors do not reach agreement by editing, discussion on the relevant talk pages continues the process toward consensus. To avoid personal opinions on a given issue, we do not rely on majority voting. Instead, we build consensus by working to incorporate all legitimate concerns, seen through the lens of applicable policies and guidelines. Unanimity is ideal, but is not always achievable.
Policy often agrees with community consensus. So, decisions that comply with policy tend to agree with consensus as well. Discussions that form consensus are usually less about editor opinions than about what policy indicates should be done.
Any remaining suggestions or objections? Does anyone have an issue with replacing the current lede with this?
I'm sorry but this is not an improvement. It seems very much the product of a committee. No change is necessary and perhaps this effort should be scrapped.
I think it is a mistake to remove the bit about verifiability and reliability, since it is by paying attention to the dissenters that we catch the mistakes of the crowd.
The last paragraph doesn't seem to say anything at all.
Perhaps this is a better try:
On all Wikipedia pages, editorial decisions are made by consensus. This method has been chosen because it is believed to be the best way to achieve the goals of the project. In practice, consensus takes the form of broad community agreement, or agreement among those paying attention to the alteration of an article.
Consensus usually forms naturally in the process of editing. When a page is edited, those edits may be changed by another editor. The portions that remain without alteration become the consensus.
On those occasions when editors do not agree on a particular editorial decision, discussion on the article's talk pages continues the process toward consensus. Consensus is built in discussion by working to incorporate all legitimate concerns. Ideally, editors accommodate each other to reach an acceptable draft. Unanimity is ideal, but not always achievable.
The first sentence should ideally be "consensus in a nut-shell": readers should be able to read the first sentence and ask: "Has this told me all I need to know?", and if we've done a good job, the answer will sometimes be "yes" and the reader need read no further. I think Ring's 1st sentence is more likely to fit this bill; one thing missing though is an indication that WP consensus is different from that in the world at large (i.e. broad, not unanimous agreement). Also, I'd prefer if we can (as I tried above) to mention the project goals in the 1st sentence to further increase its potency as a "nut-shell". I'm not keen on the "believed to be the best way" bit as it seems to state the obvious: all policies are believed to be the best way; if they weren't, we'd change them accordingly. Was consensus "chosen" as such, or is it more a consequence of the pillar that "anyone can edit"? Uniplex (talk) 07:21, 9 November 2011 (UTC)[reply]
Honestly I still prefer the lead we've got at the moment (assuming the problem of the "weasel word" can be sorted out somehow). All these alternative suggestions seem to be introducing quite controversial philosophical viewpoints that need at least to be addressed at length, not presented upfront as if they were fundamentals of the subject. --Kotniski (talk) 09:05, 9 November 2011 (UTC)[reply]
I would say, though (and this applies to both the current lede and the suggested ones) that it's not only editorial decisions that are made by consensus - it's any decision that we're capable of making.--Kotniski (talk) 09:06, 9 November 2011 (UTC)[reply]
Okay, here's an attempt at improvement of just the lead sentence, and incorporating your point.
Shown first bare, then with embedded justification:
⋈ In Wikipedia, consensus is the broad agreement, based on the project's goals, that we seek in making any decision relating to the content or the running of the encyclopedia.
In Wikipedia [wide view of scope, since 'consensus' may mean something else outside WP], consensus is the broad agreement [basic definition per current practice], based on the project's goals [qualify 'agreement' per the project ideals], that we seek [an active effort by all concerned] in making any decision relating to the content or the running of the encyclopedia [local view of scope: we seek consensus for all decisions in WP; occasionally ArbCom over-rides, but we seek it none the less].
IMHO, this is more complete and concise that the current lead sentence—I can't see why we should emphasise any of the 5 pillars in particular: they're all pillars. Uniplex (talk) 10:46, 9 November 2011 (UTC)[reply]
Guys, this is getting worse and worse. The versions presented in this section are unclear, poorly written, and are landing fairly far away from the theoretical concept of consensus. What in heaven's name is going on? Nothing in this section is usable - I'm happy to give extended details on that if you like - so let's close it and start a new one. --Ludwigs212:39, 9 November 2011 (UTC)[reply]
Perhaps you could give some details of your thoughts on the sentence marked ⋈ above? If it's the syntax, please suggest a better cast. If "broad agreement" is too far from the theoretical concept of consensus, which should change, the term "consensus", or the interpretation of our decision-making as being based on "broad agreement"? Uniplex (talk) 12:56, 9 November 2011 (UTC)[reply]
I don't think we have any big problems with the current version. Given the importance of the subject, it would be a bad idea to alter the lede like this. Perhaps there are some incremental changes but I doubt it. --Ring Cinema (talk) 13:57, 9 November 2011 (UTC)[reply]
The present version is badly broken. Policy should be written in language that supports "anyone can edit", not just "anyone with post-doctoral reading skills". LeadSongDogcome howl!14:16, 9 November 2011 (UTC)[reply]
@Uniplex: the ⋈ sentence is not well-written, and its meaning is ambiguous. the core of the statement is that consensus is "the broad agreement that we seek in making any decision", which is difficult to parse and problematic. When you twist it around so that the result is first in this way you end up with people playing mid-east politics (the standard trope among mid-east dictators is that they have a consensus to rule and that anyone who disagrees is a rebel who must be suppressed to protect that consensus; we see too much of that on wikipedia pages already). Further, the way you toss the 5P in there chops up the sentence, making it harder to parse, and reinforces a 'rules first, discussion second' mindset that antithetical to any theoretical conception of consensus. I mean, I understand what you're getting at, but a naive reader is going to take entirely the wrong message from that line.
Yes, I wasn't overly enamored with the structure (but I think the subject matter is right for the lead sentence), hence my invitation to re-cast. Uniplex (talk) 17:57, 9 November 2011 (UTC)[reply]
@LeadSong: it's not clear what you're referring to when you say 'The present version…'. which version are you objecting to, specifically? --Ludwigs214:42, 9 November 2011 (UTC)[reply]
Wikipedia uses a process called 'consensus decision-making in every area of the project. As an ideal, this is a simple concept: editors work together to seek out a result - a consensus - which everyone can accept as reasonable and valid. However, consensus decision-making in practice is more complex. It must take into account the needs of writing a good article, satisfy the principles and policies of the project, cope with participants who do not understand the process, and in other other ways adapt to less-than-ideal conditions. The golden rule is to write for the encyclopedia and assume that everyone else is doing the same; that approach creates the best atmosphere for resolving problems and reaching some mutually acceptable goal.
Consensus decision-making does not have to involve discussion or debate. Collaboration on articles is a form of 'invisible' consensus decision-making: changes are made to an article, other editors revise or amend those changes, and eventually the article stabilizes on a new revision that everyone accepts - a consensus is reached. When this type of 'invisible consensus' doesn't happen, then editors use the talk page to communicate with other editors and hammer out revisions that all participants can accept. If necessary, more participants can be invited into the discussion though processes like Third Opinion or Request For Comment, or problems can be referred to administrative pages, but those too operate under the principle of consensus.
The most important - and difficult - aspect of consensus decision-making is that it is not authoritarian. It is not majority rule or voting, does not require unanimity, and is always open to revision and improvement. Even policies are not rules in the generally accepted sense, but are instead consensus-based principles that have broad acceptance across the Wikipedia community. What is important is finding common ground about what makes for a good presentation of the given topic.
I think this covers the major points highlighted above, and separates out the 'ideal' of consensus from the 'practice' of consensus (which is where a lot of the above confusion rests). --Ludwigs214:28, 9 November 2011 (UTC)[reply]
In other words, you like the first twolines - lol. ok.
Look, here are the points I think it is vital to outline in the lead. maybe we can start from an outline like this and build it up:
consensus is a decision-making process that aims for mutual acceptance (the more general that mutuality, the better)
consensus and collaboration are essentially the same thing - consensus is really just a more abstract, verbal manifestation of practical collaboration
consensus is an ongoing process that is never fixed or settled: one can't 'claim' a consensus, one can only 'reach' one, and it may need to be reached over and over again
editors out to debate from the perspective of the encyclopedia, because that is the best arena in which to find mutual acceptance.
If these points aren't covered then we are not helping people to understand what they are supposed to do with this policy. so how do you want to write it? --Ludwigs215:18, 9 November 2011 (UTC)[reply]
I think we need to roll back a little further still. The first thing we need to do is agree what is wrong with the current lead; without this in place, we have nothing concrete against which to judge new proposals. You've mentioned "the major points highlighted above"; in reality, I think a number of editors have raised a number of points at various times, but we haven't agreed as a group which of these points are valid and should be addressed. For example, is the current weaselly statement, superfluous, useful and clear enough to most, useful but not as clear and concise as it should be, or a shining example of written expression? We've had various views. Should we be writing conversational-like prose or legalese-like prose? We've seen proposals of each. Is the problem that the lead doesn't effectively summarise the body, or are we trying to update the lead to reflect changes in practice or ideals (that we shall then flesh out in the body)? We can't even answer questions such as "How should we approach changing the lead: small incremental changes or a redraft?" until we've answered the more fundamental questions. Ludwigs, your points 1-4 may well help here, if we agree that a) the lead doesn't already cover them, b) they are right (w.r.t. WP principles), and c) they're not not over-detail for the lead. Uniplex (talk) 15:32, 9 November 2011 (UTC)[reply]
The first problem I'd suggest for the lead (which is kinda where we started) is that it ought to start by establishing where it sits in the hierarchy of WP procedural docs: 5P delegates an aspect of WP's operation to be handled in this policy; the first thing this policy ought to do is acknowledge that that is its purpose and context. It currently seems to try to do this but in a vague way, referring to two other policies (linked) and one fundamental principle (not-linked), instead of just referring to the principles statement. The second problem is that it refers to just editorial decisions, whereas it's use as arguably wider than that. Uniplex (talk) 15:56, 9 November 2011 (UTC)[reply]
Thanks for the draft, Luwigs. I like it. Your bullet points are also well taken. I have a concern about the divide between the lede and the rest of the article. Would it be preferable to confine the lede to a more concise summary of the major points with the understanding that whatever is missing there is filled in below? Maybe you are striking the right balance but it is worth considering. And, Uniplex, too, is correct: what is wrong with what we've got? Its imperfections have functioned well so what is a rewrite supposed to accomplish? But Ludwigs's draft is a good start. --Ring Cinema (talk) 16:07, 9 November 2011 (UTC)[reply]
Specific comments on Ludwigs' points 1-4:
I think B2C's view (as I understand it) is well worth incorporating here: that consensus should probably always be viewed as being community consensus i.e. the entire community's view is always present in any decision by virtue of it being made in reference to (as applicable) guideline, policy, or principle (and that these are updated as necessary in line with any change in community view).
Maybe, I'm not sure how useful it is to state this though.
I think there are a couple of points here, both correct: that WP:CCC and that consensus cannot be claimed or declared (it can of course, on occasion, be judged). First point at least, is already in the body; whether it needs to be explicit in the lead or can be inferred, remains to be seen.
@Ring: well, most of the current body text is from a revision I made months ago, so the principles should be roughly the same as what I'm looking for here. there will probably have to be some juggling done, though. Mostly (right now) I want to get an 85% draft - one that we all mostly agree is clear and on-point, but probably wordy. then we can filter it down by moving things into the body as needed.
@Uniplex: your first point is one I agree with, but it's difficult to get at. In my own head I think of the ideal of consensus as something that anyone would accept: it's a kind of Kantian notion that we try to write content that any person, anywhere, would judge as an acceptable-if-imperfect presentation of the topic. Then a debate between a small number of editors becomes a microcosm of that effort to craft universal acceptance. I'm just not sure how to say that without sounding mildly psychotic.
With regard to Ludwig's points listed above, I concur with Uniplex above. The following point needs to be made:
Policy and guidelines usually reflects general consensus, and, so, decisions that comply with policy usually agree with consensus.
I think it's important to say that here in the lede because it's a point that is often lost or overlooked in many discussions. It explains how consensus and policy are related and thus why following policy is important. We don't follow policy because it's the rule - we follow policy because the community has agreed to do what policy says to do. I don't care exactly how it's worded, but it needs to be in there, as it's crucial to understand what, in practice, one actually does to find out what consensus says to do in a given situation (you do it by looking at policy and guidelines). --Born2cycle (talk) 17:21, 9 November 2011 (UTC)[reply]
Well, hmm, it seems I keep having to point this out - policy and guidelines often don't answer such questions, for the same reasons I keep giving - policy as written might not properly reflect actual consensus or practice, there might to good reasons to deviate from normal practice in whatever case is under consideration, and of course policy often doesn't provide an answer to the question anyway.--Kotniski (talk) 17:55, 9 November 2011 (UTC)[reply]
Yes, but no. For 99% of changes, the policies and guidelines do answer such questions for article content: May I spam my website all over Wikipedia? WP:EL says no. May I say that Magic Johnson has AIDS? WP:BLP says yes, but only if you support the claim with an inline citation to a reliable source. May I add a little well-sourced information about a minority opinion on this subject? WP:NPOV says yes, within limits.
It's only in a very small number of situations that the policies and guidelines don't provide a useful answer (either saying nothing, or saying something potentially misleading), and even then, the policies and guidelines are themselves the jumping-off point for discussions. We don't really need to write this policy to emphasize the <1% of most challenging disputes. WhatamIdoing (talk) 18:08, 9 November 2011 (UTC)[reply]
Once again I made the error of presuming that it goes without saying that policy is imperfect; apparently not. Of course there are questions for which policy and guidelines do not provide an answer (often because there is no consensus for how those particular questions should be answered), and of course there are situations to apply IAR - but that's no reason to not look at policy and guidelines for guidance. And it's certainly not a reason to avoid pointing out that policy and guidelines usually reflect consensus. I mean, there is general consensus about how decisions should be made on this project, and there is (imperfect and incomplete) manifestation of that consensus in writing, which is stated in our policy and guidelines. --Born2cycle (talk) 18:08, 9 November 2011 (UTC)[reply]
(ec) There's a logical chain: guidelines (and essays), policies, principles; if you can't find applicable guidance in one, move to the next—I'm not sure there's any other course of action we can take. If guidelines are lacking in an area and this is causing you problems with articles, add to the guideline (with consensus of course). Uniplex (talk) 18:15, 9 November 2011 (UTC)[reply]
The other course of action, obviously enough, is to THINK FOR OURSELVES. Guidelines and so on will sometimes answer questions for us; most times they won't, and nor should we expect them to. --Kotniski (talk) 18:21, 9 November 2011 (UTC)[reply]
(ec, reply to WAID) I'm not sure what universe of disputes you're considering. Most of the questions that policies will answer definitively are not serious disputes (usually everyone except one newcomer knows the answer, and they just point him to the relevant policy by way of explanation). But once there is a genuine dispute between experienced editors, then it's almost certain either that it's a matter that no policy or guideline will settle, or that one side or the other is putting forward valid arguments to deviate from those rules.--Kotniski (talk) 18:17, 9 November 2011 (UTC)[reply]
I'm considering the universe of actions that are bound by this policy, not just those that you would label as a "dispute". WP:Consensus applies to every single action, not just content disputes. If you fix a typo, your action is subject to this policy. That's why we have the section that begins "Consensus is a normal and usually implicit and invisible process across Wikipedia..." This policy applies to those "invisible" forms of consensus every bit as much as they apply to knock-down, drag-out fights with people screaming about the consensus clearly being on their side. WhatamIdoing (talk) 18:29, 9 November 2011 (UTC)[reply]
This is all becoming very abstract. If I fix a typo, I'm just fixing a typo, I'm not concerned whether Wikipedia has a policy that says "fix typos". If I change the wording of a sentence, I'm just being bold and doing what I think best - highly unlikely that any policy will settle the matter of which wording is superior, and even if there is an obscure guideline somewhere that appears to address it, it's probably just something written by one person off the top of their heads, and not worth a great deal.--Kotniski (talk) 18:36, 9 November 2011 (UTC)[reply]
It's not abstract at all. It's a practical result of good policy that we have few disputes in those areas where policy is clear. I suggest the evolution of policy on BLPs is a good example of that. Over time, we worked out a lot about what should and should not be in BLPs, and this is reflected at WP:BLP. Anyone who doesn't know is quickly reminded and can easily refer to it.
The fact that those areas where we do have a lot of disputes is also where consensus (and, thus, policy) is not so clear, really makes this point. It's also an indication that there's an area where general consensus should be worked out so policy can be tightened up, so the incidence of disputes in that area can be reduced. Policy and guidelines are intimately connected to consensus in this way. --Born2cycle (talk) 18:57, 9 November 2011 (UTC)[reply]
It doesn't seem the least bit abstract to me. If your bold change sticks, then there is true consensus for your change. This policy has said that for years. Consensus does not have to involve a single word of discussion. Everything that is accepted, even if that acceptance is silent, counts as consensus.
Oh, and the policy that tells you to fix typos is WP:Editing policy, especially the WP:IMPERFECT and WP:HANDLE subsections. The Manual of Style and WP:Basic copyediting can help you identify which minor changes and re-wordings are likely to be appropriate. While you are a good enough writer that you don't need to consult them normally, none of these are "just something written by one person off the top of their heads". WhatamIdoing (talk) 18:50, 9 November 2011 (UTC)[reply]
So if five years ago someone added "Your nose smells" to some article, and no-one noticed it until now, then that addition has "true consensus"? And OK, the MoS is quite often cited and is a useful resource, but the other guidelines you mention are unlikely ever to answer an actual question. People do what they do mostly guided simply by common sense, and long may they continue to do so.--Kotniski (talk) 18:59, 9 November 2011 (UTC)[reply]
Come on. Of course we're assuming we're talking about changes that others observe and don't object, not about changes that get in undetected.
And of course consensus and policy are often in sync with common sense, so simply following common sense often works, and will continue to work. But that just means common sense often approximates consensus and policy, and, so, following one's perception of common sense is usually a good indication of what to do, not that common sense is the base principle at work. --Born2cycle (talk) 19:21, 9 November 2011 (UTC)[reply]
I think this should have been put the other way round - consensus and policy are often in sync with common sense, so following consensus and policy often works, but that doesn't imply that they are the base principle. The base principle is intelligent people working (alone at times, together at times) to produce what they intuitively know will be a good resource for humankind. Policy is mostly quite a good description of certain aspects of what they do; consensus reflects the fact that most of them more or less share the same vision of what this resource should be like.--Kotniski (talk) 19:57, 10 November 2011 (UTC)[reply]
Claiming consensus
I also don't get this "can't claim a consensus" thing - of course you can claim that there's (enough of) a consensus to imply a decision do something - so what is it y'all mean by that?--Kotniski (talk) 17:58, 9 November 2011 (UTC)[reply]
(ec) Strictly speaking no, if you reach consensus, it's obvious to all involved that this has happened. In my experience, if this doesn't occur, more often than not, it's due to there being no clear guideline (so fix this as above). Uniplex (talk) 18:15, 9 November 2011 (UTC)[reply]
I still don't know what you mean. Many times there are valid differences of opinion about whether a consensus has been reached or not. --Kotniski (talk) 18:19, 9 November 2011 (UTC)[reply]
Kotniski: last comment before I'm off for the day. the problem with 'claiming a consensus' is that you only need to do it when someone disagrees with you. if you claim that we have a consensus and I agree with you, what's the point of making the claim? if you claim we have a consensus and I disagree with you, then in a fundamental sense your claim is wrong. At best you get into a situation where some editors say "we've all developed a consensus, and you're not part of it; sorry", which is sometimes necessary but shouldn't be done without hearing what the person has to say (i.e., trying to reach a consensus, and failing). in short, it's bad to use 'we have a consensus' as a tool for shutting down or blocking discussion, not unless discussion has already happened and reached a deadlock where it needs to end. consensus requires hearing people out, even if you don't want to. --Ludwigs218:33, 9 November 2011 (UTC)[reply]
Most disputes involve such small numbers of people that you can easily count heads and see if everyone accepts the outcome. If everyone accepts that ____ is achievable (i.e., won't tempt others to the sin of edit warring) and no worse than what we've got, then you have a consensus for it. WhatamIdoing (talk) 18:41, 9 November 2011 (UTC)[reply]
(ec, reply to Ludwigs) I agree with the last part of what you say, but certainly not the first - there, you're claiming that if there's a disagreement about whether there's consensus or not, then there isn't consensus. Which would only make sense if we had a reqiurement for unanimity or some other obejctive test for consensus, which we don't and won't. Since it's subjective, there will and should be disagreements about it, and hence there can and should be "claims" on both sides.--Kotniski (talk) 18:43, 9 November 2011 (UTC)[reply]
And Ludwigs is right: if I am screaming that "X" absolutely must be in the article, and four editors are pounding on the talk page saying "not-X is the only acceptable outcome!", then we don't actually have consensus.
If those four say that we have a consensus, and I say we don't, then I'm signalling that I'm not prepared to accept the fact that I've been outvoted (yes, outvoted) in this instance. I know that in this situation, I'll lose—as described in the section on acceptance immediately below—but the fact remains that we don't have consensus about what ought to go in the article. At best, we might have a consensus about what will go in the article. (Everyone here can count well enough to realize that no single editor is able to out-edit-war four determined opponents.) When I accept that I've lost, then we can claim a consensus, but when I don't (yet) accept my loss, we do not have a consensus. WhatamIdoing (talk) 18:58, 9 November 2011 (UTC)[reply]
What you say would be true if "consensus" on Wikipedia meant what it means in the outside world - but it doesn't. Here, it doesn't (for practical reasons) require unanimity. (And if you say it does - well I disagree with you, so sorry, your view doesn't have consensus...)--Kotniski (talk) 19:02, 9 November 2011 (UTC)[reply]
Kotniski is spot on about this. The only way "having to claim a consensus means there is no consensus" is true is by presuming the 100% agreement of all involved parties definition of consensus, which we don't use at WP.
For example, I claim there is a consensus that all articles about U.S. cities that are not on the AP list are to be predisambiguated with ", State", even if their names are unquestionably unique. I happen to strongly disagree with that rule, but the claim that consensus supports that rule can never-the-less be made. --Born2cycle (talk) 19:13, 9 November 2011 (UTC)[reply]
I agree that consensus is not unanimity on WP, and would add that in my experience effective consensus decision-making in real life does not require unanimity either (WP article on Consensus decision-making discusses this at length).
I may not be quite clear on what is meant by "claiming consensus." It seems to me that closure on a discussion is often beneficial. This may require someone acting as recorder, checking whether all legitimate concerns have been addressed and summarizing the discussion. Sometimes one or more participants may not be clear as to what a legitimate concern actually is or that unanimity is not required (some participants may be new to Wikipedia or habitually tendentious). I agree that consensus is usually reached, but find that there are also circumstances where closure needs to be facilitated. It seems to me that may mean claiming consensus. Sunray (talk) 19:25, 9 November 2011 (UTC)[reply]
Sunray, in my view reasonable editors will always reach a reasonable consensus. This is because reasonable editors are more interested in being accurate than in being right, and they will find a way to balance differing viewpoints against each other to reach something that everyone can accept. The problem comes from unreasonable editors - editors who for one reason or another are unable to accept some particular content point or perspective. When there are unreasonable editors, we expand the discussion (through RfC's, policy and policy noticeboards, mediation and the like) in order to reach a consensus on whether the particular point they are being unreasonable about is a valid concern for the encyclopedia. So yes, closing a discussion can be beneficial and can be a valid part of consensus, but it has to be a reasoned closure with an opening for further discussion: efforts to lock a particular 'consensus' so that it can never be discussed and never be changed violates the spirit of consensus.
My own view is that we have to treat consensus as an educative process: any consensus debate must be concerned with raising the understanding of the various participants so that they can see all sides of the problem. Not everyone will be able (or willing) to do that, obviously, but it's only by making the effort that we can reach effective consensus. using the concept of consensus shut down all discussion is anti-educative - we're not telling people what they need to know to understand the consensus, we're just telling them to shut up and go away - and that's destructive. --Ludwigs215:32, 11 November 2011 (UTC)[reply]
Yes, I certainly agree that "claiming consensus" should not be used to shut down discussion. I also think that consent is an important component of a consensus decision. That breaks down when one person is being unreasonable (and most of have been unreasonable at one time or another). Good policy should cover all eventualities. So how do we stay educative and still get closure on discussions? Sunray (talk) 17:53, 11 November 2011 (UTC)[reply]
I mean, there comes a point where the discussion ought to be closed - we can't waste our time discussing certain things eternally, and the point of these discussions is to make a decision about what's best for Wikipedia, not to soothe editors' souls. Sometimes there are going to be decisions that quite significant numbers of reasonable people disagree with - that's inevitable - and there are also going to be decisions where quite significant numbers of reasonable people disagree with the closer's assessment - and (given that we will never have an objective definition of "consensus") that's inevitable too. Neither of these things are bad. What I think is worse, in fact, is the debates that go on and on without receiving the attention of a legitimate closer - that means lots of editors' time wasted, and the actual "decision" more likely to be made through edit-warring than through proper evaluation of the arguments (and it's often the unreasonable editors who are also the more likely to edit-war).
And, of course, a debate that has been closed can be reopened after a reasonable time has passed - or even earlier, e.g. if someone's got a significant argument that was overlooked in the previous debate.--Kotniski (talk) 18:06, 11 November 2011 (UTC)[reply]
I agree with these points, but I don't think we can cover all that in this policy without being confusing. Maybe the compromise is this: describe what consensus is, give some kind of metric for when consensus fails, but don't try to talk about remedies: make a link to dispute resolution processes, which can each have their own particular rules about decision-making. For example, something like:
If editors spend more than a week debating a point with no progress, or if talk page discussion begins to get heated, they should turn to one of the available dispute resolution procedures and treat the decisions made there as a new consensus.
Again, it depends what kind of discussion it is. If it's something like an RM or AfD, then the procedure is already in motion, and you might as well just sit back and wait for someone to close it (although this policy ought also to give advice to closers on how to assess consensus). If it's a more generic discussion, the most general advice we can give in these situations (and I think we already do) is to bring more people into the discussion (in a way that avoids charges of canvassing). --Kotniski (talk) 18:39, 12 November 2011 (UTC)[reply]
We operate with two kinds of 'acceptance' here. One is that we work out a compromise that leaves everyone more or less satisfied, e.g., we tweak language so it is a little softer or stronger than any individual really wants, but which finds the sweet spot where "just barely strong enough for me" overlaps with "just barely soft enough for you".
The other is that one or more participants accept the fact that they've lost. Knowing WP:How to lose is an important skill. Sometimes continued discussion is just a waste of time, because there is no overlap in the participants' positions. If I am absolutely determined that the only acceptable title is "Ganges", and you are equally determined that the most common name is "Ganga", then there is no half-way position. For technical reasons, we must have exactly one article title: we cannot choose two titles, and we cannot choose zero titles. At some point, one side or the other has to admit defeat (at least temporarily). For better or worse, the decision in this uncommon class of disputes is usually made more or less as a majority vote.
The article title problem makes a somewhat extreme example, but this happens all the time: I have a great new idea that gets shot down by a dozen editors. I may still believe in my idea, but I accept that "consensus", as embodied by a WP:SNOW storm, opposes it. I may want to re-write an article to reflect what I (an individual, and therefore individually biased, person) believe is the current state of knowledge, but if I encounter strong opposition, I may well accept that further efforts are a waste of everyone's time (for now), because the article's current balance is supported by a "consensus". In all cases, I'm accepting that I've "lost", not accepting a compromise that leaves everyone (barely) satisfied. WhatamIdoing (talk) 18:25, 9 November 2011 (UTC)[reply]
This is exactly the kind of thing that needs to be explained in the body of the policy, but not in the lede. Perhaps a subsection on "acceptance"? With WP:ACCEPT as a redirect to it? Looking forward to be able to say, "Looks like I need to accept that ..." --Born2cycle (talk) 18:35, 9 November 2011 (UTC)[reply]
I concur, for the reasons given above: major/unsolvable disputes may be high-profile but they are very, very few compared to, for example, a revert with an edit comment of "per such-and-such guideline" and then no more is said. Uniplex (talk) —Preceding undated comment added 18:39, 9 November 2011 (UTC).[reply]
Born2cycle's suggestion of a separate section seems like a good idea. Perhaps it should go just above the new (and still incomplete) section on no consensus. WhatamIdoing (talk) 19:02, 9 November 2011 (UTC)[reply]
Basically what I've said above: You can have consensus ("agreement") because you all accept a particular result as being good (or good enough), or you can have consensus because you accept that you've lost. In this latter case, you are not accepting the particular result as something that is even remotely appropriate from your perspective, but only accepting that no desirable outcome is actually achievable. Both of these situations count as "consensus" on Wikipedia. WhatamIdoing (talk) 02:29, 10 November 2011 (UTC)[reply]
OK, so we have consensus when everyone agrees what to do, and we have consensus when everyone agrees that there is sufficiently wide agreement as to what to do. Those are the easy cases. The hard cases are when there is disagreement both about what to do and about whether there is sufficiently wide agreement as to what to do. What do you want to say about those cases? Nothing that would imply, I hope, that there can be no assertion of consensus in such cases, since that would effectively bring us back to a situation where unanimity is required in order to do anything.--Kotniski (talk) 08:38, 10 November 2011 (UTC)[reply]
It seems to me that legitimate concerns are very important in cases of protracted disagreement. If one or more participants in a discussion are raising what they believe to be a "legitimate concern" (i.e., something that clearly addresses, or goes against, a WP policy), it needs to be addressed before there is closure to a discussion. A problem arises when there is only one individual making such a claim. In such cases, it is perhaps best to secure agreement from at least one other individual that there is, indeed, a legitimate concern that needs to be addressed. Such cases would often require an outside opinion (e.g., from an appropriate noticeboard). Sunray (talk) 18:47, 10 November 2011 (UTC)[reply]
If I'm reading this correctly, at least part of the concern is about "claiming consensus" when certain legitimate concerns have not been addressed. Frankly, I've seen improper claims of "no consensus" too - by which I mean a position whose legitimate concerns have not only been addressed, but dispensed with, but is still considered in deciding there is "no consensus" because some significant number of people continue to hold that position despite there being no legitimate concerns favoring that position.
In both types of situations, the encyclopedia would be improved if the arguments on both sides were given more consideration and objectively evaluated and weighed against each other - to determine consensus - rather than paying attention to the number of individuals who are supporting one position or another.
Can we all agree that it's okay to "claim consensus" even though some are not agreeing, when their disagreement has no basis in legitimate concerns? --Born2cycle (talk) 19:29, 10 November 2011 (UTC)[reply]
Yes; but also, sometimes we have to make a decision even though there are legitimate concerns on both sides, and there is no way of fully addressing all those concerns. Sometimes, as in everyday life, we just have to weigh up the advantages and disadvantages of all of the (imperfect) options, and choose one of them, warts and all.--Kotniski (talk) 19:48, 10 November 2011 (UTC)[reply]
Maybe. I think such situations are far rarer than many presume them to be. People in the real world as well as at WP are often too quick to assume there is no consensus solution, and all too willing to compromise when there is no need. --Born2cycle (talk) 20:30, 10 November 2011 (UTC)[reply]
Yes, those situations are quite rare. However, they do exist, and this policy used to address them, by saying that if all else fails, we'll have to take a more-or-less majority "vote", at least as a temporary measure. That this applies to (in my experience) fewer than 1% of talk page discussions and noticeboard actions doesn't mean that it never happens. WhatamIdoing (talk) 20:35, 10 November 2011 (UTC)[reply]
I have mentioned this elsewhere, but I think it is worth keeping in mind: In cases where a straw vote is taken and the number of editors is more than a few, a proposal from the majority that is acceptable to a majority of the minority (the straw poll minority) ought reasonably to have some standing. At the same time, this is a fallback position. --Ring Cinema (talk) 04:50, 11 November 2011 (UTC)[reply]
I don't consider these situations to be particularly rare - particularly among the cases that are likely to cause this policy to be consulted (i.e. those where it's not clear whether there's enough of a consensus to do something). In most cases of dispute there are pluses and minuses to each available course of action, and it's quite common for different people to attach different weights to these pluses and minuses; it's also quite common that no solution has an obvious supermajority supporting it, so there could well be dispute about whether sufficient consensus has been reached or not (thankfully we don't often debate that issue, we tend to accept the decision of the closer in the case of the kind of discussion that gets formally closed, and that of the most dedicated edit-warrior(s) in other cases).--Kotniski (talk) 11:20, 11 November 2011 (UTC)[reply]
I have a good example. [[4]] has been a long conversation that turns on the relative merits of navigation boxes and an infobox field with more or less the same information. --Ring Cinema (talk) 14:45, 11 November 2011 (UTC)[reply]
I've seen a similar issue debated related regarding infobox album. In that case, IIRC, they found distinct uses for each and so all were happy. It seems that we have two categories of issue: 'style' or 'accessibility' issues, and 'content' issues. For the latter, numerical voting should be disallowed, otherwise we risk to lose NPOV. For the former, numerical voting is not quite as worrying (V & NPOV issues are much less likely); however, the best way to judge is probably via feedback from readers: maybe try one style on one group of pages and the other on another, then see which gets better ratings (we have "Rate this page", although presentation isn't currently included). Uniplex (talk) 16:55, 11 November 2011 (UTC)[reply]
Dramatic change are in store for the WP:Article feedback tool, so that particular method won't work, even if we could validate it as effective (which I doubt).
Article content issues, especially where WP:FRINGE applies, are naturally worrisome. However, some of the worst fights in en.wiki's history have been about completely non-content matters (anybody else remember the date de-linking ArbCom case?) that were ultimately so unimportant that they could have been settled by a straight-up majority vote—or by flipping a coin and taking the outcome—without a single word of content being harmed either way. Both "2011" and "2011" are equally compliant with all major content and sourcing policies, and yet we had people screaming for months (or was it more than a year?) about the "lack of consensus". WhatamIdoing (talk) 17:25, 11 November 2011 (UTC)[reply]
Virtually every (contested) move discussion is an example of this sort of situation. The article titles "policy" (WP:AT) even lists a set of criteria that may have to be weighed up against each other in deciding on a title. Different people will weigh these in different ways in a given case - and when the numbers of people on each side (as there usually are only a finite number of sides - you can't very often find a compromise title) are fairly equal, different closers might take different views about which way to decide it. None of this implies that anyone is being unreasonable or that anything has gone wrong with the process - indeed, it would be very worrying if there were no such cases.--Kotniski (talk) 18:19, 11 November 2011 (UTC)[reply]
Yes, article titles are a good example of the problem:
You must pick exactly one title. It is not (due to the software) possible to have a page without a title. It is also not possible to have two or more actual titles for a single page.
Thoughtful, reasonable, experienced editors sometimes disagree about what the best name is.
Their reasons are often very much policy-based and very similar in strength: Do we pick the simplest name or the most precise name? Do we pick the most common name for all English-speaking sources or the most common name for sources specifically from the local area?
These discussions often do not, and cannot, end up as a compromise that (at least barely) satisfies all parties (the ideal for an NPOV dispute in the text). There is no way to set the article title so that it is partly Leukemia and partly Leukaemia and partly Leukæmia. These discussions are basically decided as a sort of majority vote (taking into account the strength of arguments, not just mere numbers), with the minority's "consent" being only their acceptance of the fact that they had their say and still "lost". WhatamIdoing (talk) 19:29, 11 November 2011 (UTC)[reply]
But it can wind up with a situation which (almost) everyone is willing to accept, in this sense: There is no move request active on Leukemia; as far as the talk page will show, there never has been. (This may well be because the dialectical difference is explained, in bold, in the first line; that 'is a compromise, as much of one as we can easily manage.) SeptentrionalisPMAnderson19:56, 11 November 2011 (UTC)[reply]
Compromise is over-rated. It encourages "solutions" that are worse than either of the original positions. I suggest Fixed-winged aircraft is probably the best known example of that (never mind all the rationalizations contrived to justify it), but more recently is the (thankfully brief) local consensus reached by electronic game console enthusiasts who couldn't agree on Sega Genesis or Mega Drive and so "compromised" on Sega Genesis and Mega Drive. In that case reason ultimately prevailed (but it wasn't easy) and it's now at Sega Genesis, which is good precisely because it's one of the two, and not a compromise.
The adoration for "compromise" is way overblown, IMHO. Compromise titles should be the rare exception, and ideally eliminated, by definition (a compromise is not the ideal by definition). --Born2cycle (talk) 20:44, 11 November 2011 (UTC)[reply]
I agree that compromise (in general) should take a back seat to evaluation of the arguments, so suggest to change the diagram in the Process section accordingly. Uniplex (talk) 20:56, 11 November 2011 (UTC)[reply]
Yes, although there are many cases too (not so much with article titles, but with matters of content) where a "compromise" is a good, and perhaps the best, solution. But the emphasis should be on finding a solution that best satisfies all the legitimate concerns, rather than necessarily one that best satisfies all the editors (though in most cases the latter criterion will be a good enough reflection of the former). --Kotniski (talk) 09:34, 12 November 2011 (UTC)[reply]
Septentrionalis, it's true that there hasn't been an active dispute over its name, probably because it's neither a controversial article nor a subject that attracts less-mature editors (unlike, e.g., certain video game disputes).
But imagine that it was a brand-new stub (or, if it's easier for you to imagine, pretend it's still one of the many subtypes of leuk(a)emia that are still redlinks), and there was a dispute about the spelling for the article title: Two editors have contributed equally to the article. They differ in their spelling. Three other editors offer an opinion: two favor the British spelling and one favors the American spelling. All five reject the ligature ("æ") spelling.
Wouldn't it depend on what arguments had been advanced? If it was just "I'm British so I prefer the British spelling" and so on, then I guess the rules at WP:ENGVAR would be applied, rather than considering the (essentially empty) content of the discussion. --Kotniski (talk) 09:31, 14 November 2011 (UTC)[reply]
ENGVAR is useless:
An article on a topic that has strong ties to a particular English-speaking nation should use the English of that nation—it doesn't have any ties to any particular nation.
When an English variety's consistent usage has been established in an article—no "consistent usage" has been established.
the variety used in the first non-stub revision is considered the default—it's still a stub, and the contributions of the disagreeing editors are equal.
There really aren't any possible policy-based arguments for this issue, because none of the choices are more verifiable, neutral, etc. than the others. And we must make a decision: we cannot have multiple titles, and we refuse to tolerate edit warring over this. WhatamIdoing (talk) 22:13, 14 November 2011 (UTC)[reply]
I guess it's just left as it is, then (or at whatever title it was originally created at). This seems more of a matter for the MoS (where ENGVAR appears) than for here.--Kotniski (talk) 08:15, 15 November 2011 (UTC)[reply]
It's not properly a MoS issue: it's an AT issue, and AT specifically says that the decision must be made according to this policy (second sentence, first section). So what do you think the consensus is in that situation? Do you think that any consensus exists? WhatamIdoing (talk) 01:46, 16 November 2011 (UTC)[reply]
Well, no. Although if there were an argument about it, I suspect a consensus could be reached about how to apply the principles of ENGVAR and the national varieties section of AT to resolve the situation.--Kotniski (talk) 10:40, 16 November 2011 (UTC)[reply]
I think it's highly unlikely for a debate to commence about the English variety of an article's body and its title while it is still a stub. At any rate, even if such a debate begins, one is better off creating content than huffing and puffing pointlessly on the talk page. The first one who makes enough of a contribution to make the article not a stub establishes "the variety used in the first non-stub revision", and that becomes the default for the title as well as the article, per ENGVAR. As long as the article remains a stub the point of ENGVAR being useless is technically accurate but ultimately moot. The only potential issue then is perhaps about which revision made enough of a contribution for the article to no longer be a stub. Again, debate about that too is unlikely in practice. For example, consider the initial revision of Yoghurt (which was at Yogurt - speaking of examples where ENGVAR is ignored)... clearly not a stub at a creation: [5]. --born2cYcle17:51, 16 November 2011 (UTC)[reply]
And what would you do in between now and when that happened? Permit the two original contributors to continue their move-warring, because some day, someone else will finally add another paragraph and then you can finally apply ENGVAR? WhatamIdoing (talk) 17:57, 16 November 2011 (UTC)[reply]
I (oh, you weren't asking me, but still) would put it back to the title it was at when it was created (or whichever of the alternative reasonable titles was used first) (or whichever it was most recently stable at). If the wording of AT and ENGVAR isn't tight enough to cover this situation, then that's a matter for those pages, not this one.--Kotniski (talk) 18:02, 16 November 2011 (UTC)[reply]
So you would make up your own rule about how to handle it. I believe, by the way, that your proposed rule was deliberately rejected by ENGVAR.
The reason why this example matters for the discussion at this page is that WP:AT sets consensus as the most important criterion for selecting an article title. This policy used to address the problem of "no consensus" when a decision must be made (at least temporarily, until some sort of consensus appears): it said that as a last, desperate resort, you might rarely need to choose a majoritarian approach instead of pretending that consensus was always going to appear. I think it still needs to address this issue directly.
There are very few realistic options about what to do, as a general rule, when no consensus is currently available: You have suggested a first-come, first-served rule. This is normally labeled the "prior consensus version" rule. We can do that in a very few cases, including article titles. However, it makes a poor general rule, because it applies to so few cases. For example, is the "prior consensus version" the one just before I added a sentence in September, or the one just before you reverted it in November?
Alternatively, we can refuse all contested actions. We can accept all contested actions. We can apply a (more or less) majority vote. As far as I can tell, that's about it for our options.
IMO following the majority (or plurality, and except where other policies specify the outcome for a no-consensus discussion result), although basically an undesirable option, is probably the least-evil choice. WhatamIdoing (talk) 02:29, 19 November 2011 (UTC)[reply]
Process diagram
About the diagram in the process section, though - it seems to me to be a gross over-simplification, and would belong better in some sort of essay than in a policy.--Kotniski (talk) 09:42, 12 November 2011 (UTC)[reply]
The diagram was the focus of long discussion some time ago. The view that I put, which seems to have prevailed, was that the diagram should be simple and immediately understandable for newcomers, that "policy" pages primarily exist to serve as an introduction for the newcomers. What policy pages should avoid is the becoming of forums for high-level oblique philosophical discussions between highly experienced Wikipedians. A number of diagrams were produced. Kim came close to collecting them all together, in an essay, I think. A more detailed version of the diagram that was there for some time previously can be found, half shaded, at WP:BRD. Kotniski, "gross" would seem to imply that you think the simplification is too great, to the point of taking it into error. What exactly is wrong with the diagram, or what is missing that is essential? --SmokeyJoe (talk) 12:34, 12 November 2011 (UTC)[reply]
Well, all right, if it were labelled as a "simplified diagram of the process" or something, I guess I wouldn't object to it. But when we make things simple for newcomers, we do them no favours if we make them think that it really is that simple. (If there's an error, then I guess it's the reference to "compromise", which - as noted above - is not necessarily the right solution.)--Kotniski (talk) 14:37, 12 November 2011 (UTC)[reply]
Here are some observations on the current diagram:
You don't have to edited a page to disagree with a change made by someone else, but the diagram seems to suggest otherwise.
The time to wait to see if the article is edited further is unspecified (might not seem significant but I've seen arguments based on the length of time between changes)
Evaluation of the arguments for and against the change (e.g. checking more sources) should be given greater prominence—compromise may not be necessary.
Yes, those are all good points - particularly the first one, as we seem to be encouraging people to edit-war by telling them that their disagreement doesn't count if they don't physically revert.--Kotniski (talk) 15:34, 12 November 2011 (UTC)[reply]
Uniplex: that's not bad at a glance, but there are a couple of obvious revisions that need to be made
you need to move the 'make an edit' box up, so that it's clear that that is the effective starting point of the diagram - otherwise there's no clera entry point, visually speaking
say 'make an edit or propose a revision', to cover both cases
'outcome clear' and 'change needed' are ambiguous - they need clearer referents
It is probably not accurate to believe that the policy is examined primarily by newcomers. Another major use occurs when disputes arise and editors look for a way to resolve it. This double use accounts for easy case/hard case problem. For newcomers we want to show how the practice works in most cases (the easy ones), but when there is disagreement the same page is supposed to provide a way out of hard cases. --Ring Cinema (talk) 16:50, 13 November 2011 (UTC)[reply]
Here's an updated diagram with some additional notes to hopefully address and clarify some of the points made. Ludwigs, one thing that the diagram is trying to show is that the process needn't start with an edit: it can also start with someone reading a page (and agreeing with what they see, or taking to talk if they don't). Ring, I think the hard cases are covered by spending a lot of time in box #1: the two eventual options are to give up, or to somehow find a clear way to proceed (which might for involve, for example, compromise or ArbCom) and allow box #3 to be done. Uniplex (talk) 18:11, 13 November 2011 (UTC)[reply]
Again, the more that is specified, the more that seems to need to be specified. The diagram works better when the categories are vague. Let the editors figure out how it works, because they are smarter than us and have real problems to solve. 2. I'm not sure which box is supposed to be number one, but in any case the presumption that editors are reasonably open to changing their views is not what I observe at all. This is an easy case solution applied to a hard case reality. Elsewhere in this discussion I read the view that compromise is overrated and editors should just figure out who has the better argument; in fact, that is not how it goes on the hard cases. The only way to break a deadlock of determined editors without recourse to authoritarian or democratic processes is through compromise. At least, I haven't heard or seen another way. --Ring Cinema (talk) 18:39, 13 November 2011 (UTC)[reply]
The proposal is not that compromise should be disallowed or even discouraged, but that it should not be given undue weight (in the diagram) over other debating techniques (such as persuasion, and objective discussion, which are listed ahead of compromise in the policy text). Uniplex (talk) 20:38, 13 November 2011 (UTC)[reply]
RE Uniplex’s three points
Here are some observations on the current diagram:
You don't have to edited a page to disagree with a change made by someone else, but the diagram seems to suggest otherwise.
The diagram is not about rights to disagree. It is about productive methods for moving forward. It is not productive to merely disagree. To be productive, you need to have a solution. This is where the starting edit comes in. If you disagree, then it is likely that only you know what you disagree with, and only you have read the archives specifically reading into the point of disagreement. You are then ideally suited to make the edit that fixes what you consider to be the problem.
If you can’t make you edit? Is it because you don’t have a solution? Then this is idle hand wringing.
The great advantage of making the edit is that it provides a focus for subsequent discussion, if there is disagreement.
The time to wait to see if the article is edited further is unspecified (might not seem significant but I've seen arguments based on the length of time between changes)
Good point. Why don’t you suggest some lengths of time. I suggest you do it in the caption, not in the figure. For me, an edit is ratified after other regular edits further edit the page. I think “time to wait” is better measured in “edits by others” than in units of time.
Evaluation of the arguments for and against the change (e.g. checking more sources) should be given greater prominence—compromise may not be necessary.
I think this sounds like a call for a new guideline on logical evaluation of arguments. Ring Cinema has a good point about rejecting weaker arguments in favour of stronger arguments. Where differing opinions are mutually exclusive, it does not make sense to “compromise”. However, it can be helpful to refocus on what is important and to sidestep the point of disagreement. --SmokeyJoe (talk) 00:22, 14 November 2011 (UTC)[reply]
Uniplex, I must object to you making discussion here that references an offsite hosted image. One problem is that for me it is today inaccessible. Another problem is that the use of offsite images does not sit well with the copyright licensing for this discussion. Please upload your images onto Wikipedia, with free licensing, before referencing them. --SmokeyJoe (talk) 00:22, 14 November 2011 (UTC)[reply]
If it's faster and easier to upload them elsewhere, then I don't mind discussing them while hosted offsite. The final image would need to be uploaded to be useful to us, but the drafts seem unimportant to me. WhatamIdoing (talk) 00:36, 14 November 2011 (UTC)[reply]
Having images integral to the discussion offsite is crossing the line with respect to conducting offsite discussions. I therefore object to this discussion progressing with the image(s) not being available on and from this site.
Personally, I object, because for whatever reason, the site hosting the image(s) is blacklisted to me. I should not have to deal with issues of access to other sites in order to participate here. --SmokeyJoe (talk) 01:11, 14 November 2011 (UTC)[reply]
User:Uniplex's 1st image referred by him above, reasonably assumed that he licenses it the same as he does his text that discusses it.User:Uniplex's modified image referred by him above, reasonably assumed that he licenses it the same as he does his text that discusses it.
OK, I got access to Uniplex’s images (easily uploaded and placed left and right). I must strongly object that these diagrams are a model for consensus building. His flowcharts introduce endless cycles with don’t involve edits to the page. Endless cycles of discussion and evaluation before an edit is allowed leads to discussions loosing focus, with tangential lines and debate scoring taking a life of its own, and newcomers to the debate being unable to decipher the point of the discussion. The point of any discussion should be the improvement of the page. The asserted improvements are evidenced in the actual edits to the page. With further discussion, especially where reverting occurs, the so called improvements are modified, and without modified edits being made to the page, it becomes increasingly unclear what edit is subject to the current debate. If in any cycle, an edit to the page is not even attempted, then the process is failing.
Contrary to some of Uniplex’s legend, “Every change to a page should be the results of following the blue/green path”, the ideal situation of wiki editing involves purely direct editing to the page. Discussion is only needed if there is disagreement, or the edits are too fast/complex for the edit summaries to suffice. Indeed, a huge number of decent mainspace pages don’t even have activity on their discussion pages. Uniplex’s model places committee work ahead of actual progress, and this is directly contrary to the wiki-method.
Edit warring and talk page filibuster are two extremes of failure to work productively. One is an extreme lack of discussion. The other is an extreme lack of editing. If there is any conflict, both discussion (to allow for a verbose description of perspectives) and direct edits (to focus the discussion) is needed in every cycle. --SmokeyJoe (talk) 02:24, 14 November 2011 (UTC)[reply]
SmokeyJoe, One of us is misunderstanding the process depicted in the diagram. Per the diagram on the right, here's how I see a normal edit progressing:
1. Editor "evaluate[s] existing text per sources, policy and guidelines; consider[s] further change"
Is it clear [to our editor] how to proceed?
Yes → [Based on that evaluation], was a change decided upon [by our editor]?
3. Yes → [Our editor] make[s] the change
In other words, (1) and the two questions loosely represent the mental processes all editors should go through when editing. Step (2), which is about involving others, is only entered when it is unclear how to proceed. Ideally the editor realizes when this is, but sometimes it takes BRD to find out...
I think this is a good description of the flow that actually occurs, though of course in reality many edits are not always preceded with as much evaluation of existing text against sources and policy as might be ideal. --Born2cycle04:01, 14 November 2011 (UTC)[reply]
Born2cycle, yes, we made different interpretations. Your interpretation is to me a stretch of interpretation in matching the diagram to good practice. However, even your interpretation as stated above places a high hurdle before the editor who has a first edit to make. It is clearly contrary to WP:IAR, and contrary to the message “you can edit this page right now”.
Regarding File:Consensus Flowchart.svg (the current diagram), I take it as granted that at the “Edit” step, the editor will have thought first. However, little more than that. It should not be necessary, even recommended, that an editor evaluate policy, guidelines or past discussions. (Evaluate sources, yes, no argument there).
RE: “Involving others, is only entered when it is unclear how to proceed.” By making the edit, the editor defines the focus of discussion that is too follow. If the editor is unsure and needs to ask a question, fine, do so, but asking questions first is not wiki-editing. It is good, to be encouraged, but is not wiki-editing and should not be mandated.
My interpretation of the suggested diagram is that it calls for pre-agreement before editing. This amounts to requiring edit-request notices to be placed and for positive response(s?) before editing occurs. This would kill the vitality of the project. --SmokeyJoe (talk) 05:47, 14 November 2011 (UTC)[reply]
(ec) Thanks, SmokeyJoe, for uploading the diagram (I didn't do it myself as I could find a suitable sandbox for images but I guess it makes no odds). The diagram tries to capture the overall process of making an edit to Wikipedia and to cover the vast majority of cases. If it only helps level our understanding here at talk, then it's been a useful exercise; if it can be rolled out further, all well and good, but of course, it must be 100% clear in what it is saying before this could happen. The trickiest part of the process to explain or advise upon is the details of box#1 as this amounts to "how do you make a decision?", and "how do you win a debate?"—whole books have been written on those subjects, so I don't think we should try do this in the diagram; we can use the text for this aspect. Box#1 is most often a mental process: someone reads the page or just a recently changed portion and compares the presented information with their own knowledge (for which sources may well be available); whether they realise it or not, they often evaluate against guideline and policy (perhaps thinking "that sentence is too long to read easily", or "why should I believe that: it's not cited", or "that's true, but it's only half the story"). If they don't think any of these things then the blue/red path applies and the consensus status has been reinforced a little. I'll see if I can update the text on the diagram to make the mental aspect clearer. Uniplex (talk) 07:01, 14 November 2011 (UTC)[reply]
Well, I'm quite confident that Uniplex did not intend to convey that pre-agreement is required before editing. But since you interpreted it that way, it's likely others will too. That needs to be rectified. How about just changing the text in box #1 to say:
1. Evaluate existing text as needed per sources, policy and guidelines to determine if change is warranted
Okay, the proposed diagram does not reflect WP practices. The first thing that happens when an editor sees something in an article that should be changed is simple: the editor makes the edit. That's not what the proposed diagram says. Is this an attempt to smuggle in a change in the policy? As things stand today, there is not a crisis in consensus. We don't need to change anything. The best way to have a well-functioning editorial process is to let experienced editors do their editing in a context of stable, predictable policies. But a change in the policy is not called for and this diagram would definitely undermine the current practices of Wikipedia. --Ring Cinema (talk) 06:59, 14 November 2011 (UTC)[reply]
Born2cycle, I’m not seeing anything close to an improvement over the current diagram. It is clear on the two essential points: (1) If you disagree with a previous edit, then you must discuss; (2) The point of the discussion is the next edit.
If there is any problem with the current diagram, the only one I can see is the point made by Ring Cinema. Sometimes, the discussion leads to the conclusion that someone’s argument for change is rejected. In this case, the word “compromise” doesn’t fit, and there is no subsequent edit. However, the rejection of an edit, and discussion that answers a challenge, is not, per se, consensus building. It is mere maintenance of the existing consensus. --SmokeyJoe (talk) 07:01, 14 November 2011 (UTC)[reply]
W.r.t. your (1), it doesn't have to be a previous edit (per se) that you disagree with: you may just read a page and find something that you think is wrong. Uniplex (talk) 08:13, 14 November 2011 (UTC)[reply]
Very confusing. The arrow points from 2 to 1??? What does that mean? Start at 1, obviously proceed to 2, then, no wait, follow the arrow, but then what makes it 2? Okay leave out 2 but it is second so it must be important. So then maybe start with 3. Yes, that is clear. Read the notes outside the boxes, ignore the labels, follow the path, try to get to 1.
What is the problem with consensus that this confusing diagram addresses? We already have the policy operating with all kinds of good results. If the diagram is different from the written policy, which should be followed? If there is no difference, why is it there? It's not clear on any point that I can see and it doesn't have the practices correct. Believe it or not, most editors know how to manage without our help. Let's not get in their way. --Ring Cinema (talk) 07:18, 14 November 2011 (UTC)[reply]
You find the diagram confusing? You start at the top at Provisional Consensus and then follow the blue arrow to box #1 where you evaluate the text, and from there to the bottom decision diamond. Making that decision will take you either to box #2 (black/no arrow) to consult with others or to the decision diamond in the center (blue/yes arrow). From the center diamond you either go to make the edit (box #3) or, because you've decided there is nothing to change, you go back to the top and off to evaluating more existing text. --Born2cycle07:23, 14 November 2011 (UTC)[reply]
(edit conflict)The problem with the current diagram that the proposed diagram addresses is that the current diagram implies that edits are made without consideration for anything. The edit-with-consensus process does not start with making an edit; it starts with evaluating the existing text. I mean, the current diagram applies to vandalism edits that aren't noticed the same as it applies to consensus edits... there is no distinction.
So I don't understand the objection to the proposed diagram, especially with my suggested "as needed" change to the wording in Box #1. It addresses the problem with the current diagram I just described, and accurately describes how edits are normally done (as I outlined above). --Born2cycle07:23, 14 November 2011 (UTC)[reply]
Ring, I added the numbers primarily to facilitate discussion; they should probably be removed in a final draft. Further simplifications are possible: the two notes on the left, though useful, are not essential, so could go to the text. The note asscociated with box 2 could be incorporated into the text of box 2. Uniplex (talk) 08:05, 14 November 2011 (UTC)[reply]
Break (process diagram)
Just my impression on coming back into the discussion - I'm not seeing anything useful in the (latest version of the) diagram that's been proposed. I can't imagine how anyone would understand anything better on looking at this diagram than they would by reading text describing the processes. I would rather do without a diagram altogether - to be honest, the activity we're talking about doesn't follow any set process or algorithm, and it's misleading to imply that it does.--Kotniski (talk) 09:40, 14 November 2011 (UTC)[reply]
I'd be quite happy to lose the current diagram: it purports to be general but it's scope seems very limited; it's purpose is unclear (we've had suggestions that is both for and not-for newcomers). Uniplex (talk) 10:11, 14 November 2011 (UTC)[reply]
The diagram proved so successful that the problem it addressed has largely gone away. The problem was that many otherwise intelligent editors couldn't seem to rationalise direct wiki editing with a consensus building process. There was in some places a lot of Bold editing turning into edit warring, and in other places endless discussion that didn't move anywhere. Many diagrams were suggested. Some had more detail than others, but they coalesced around one core feature – the editing cycle during conflict includes both discussion and editing. The purpose of the diagram is to provide a very simple picture to communicate a proven method for consensus building through wiki-editing. The diagram is applicable for all editors, whether newcomer or old timer.
Today’s unease with the diagram I find difficult to understand. What, in theory or practice, is wrong with it? There was a problem with “compromise”, which Kotniski has just recently largely attended to, but no suggested problem of structure. Uniplex’s diagram suffers a very severe flaw – it includes cycles that don’t involve any editing. Experience, experience from before the rise of popularity of WP:BRD, was that endless discussion disconnected from actual editing would lead to tangential meandering, off-topic arguments, impenetrable talk page threads for newcomers, and exhaustion.
I think the question of what's wrong with it has largely already been answered, by Uniplex near the start of this thread - for one thing, it implies there must be an edit and a revert (or further edit) before any discussion can start, and that the only initial way of expressing your non-acceptance of an edit is to fight it.--Kotniski (talk) 12:52, 14 November 2011 (UTC)[reply]
The implication is right. There is little point in starting a discussion in the absence of an edit and a revert (or further edit). Maybe a point is for editor education, but it is a discussion not directed at improving the page. Some people seem to like discussion for the sake of discussion, but it is not so productive. --SmokeyJoe (talk) 13:53, 14 November 2011 (UTC)[reply]
I've seen many people start discussions questioning existing text or proposing an addition, without any recent edit having been made—these are all part of the consensus-forming process. Uniplex (talk) 14:12, 14 November 2011 (UTC)[reply]
Yes, I can't agree that Joe's interpretation is either current practice or good practice. Often discussion is started before making a change (if you know it's going to be controversial, or you have doubts about it yourself, or...); and often discussion of others' edits is started without reverting or editing further. I think such behaviour would even be regarded as more admirable in many cases, as it avoids edit-warring and maintains a more respectful atmosphere.--Kotniski (talk) 14:37, 14 November 2011 (UTC)[reply]
Perhaps you two are only thinking about high traffic pages? I've just gone through several random articles. Many of them don't even have talk page content! Of course if something is controversial, then discuss. But how do you know it is controversial. If you think you can improve it, and there is no history of related edits or discussion, you should make the edit. --SmokeyJoe (talk) 01:33, 15 November 2011 (UTC)[reply]
Yes, this is the point - sometimes one approach is right, sometimes the other, but the diagram would have people believe that there's only one route to follow.--Kotniski (talk) 08:23, 15 November 2011 (UTC)[reply]
Oh, and another problem I'm seeing, looking at it again, is that the "implementation" of a "compromise" is treated the same as a bold edit, not as the establishment of a new consensus. If we are to assume the compromise really was accepted by (almost) everyone in thorough discussion, then this is wrong - it would imply that any one editor who doesn't accept the consensus and wants to be a pain about it can effectively keep the discussion going (or returning to the prior status quo) indefinitely.--Kotniski (talk) 13:01, 14 November 2011 (UTC)[reply]
Again, I don’t see the problem. Implementation of the compromise may yield unforseen issues. An implementation of compromise should not be quickly assumed to be a new consensus. The change may attract new participants. It is quite right that this may go on for some several, even many cycles. If there is a disagreement, it is not normal for a consensus to appear after a single cycle. Eventually, if the process is followed, if most of the edits are not reverts, the process finds a consensus. --SmokeyJoe (talk) 13:53, 14 November 2011 (UTC)[reply]
Depends what you mean by a "cycle". It seems you're considering a process where consensus is reached chiefly by editing, and it may be that this is the process that the author of the diagram mainly had in mind. However an important part of the process in practice is the bit where editing alone fails, and agreement is reached through discussion. In this case, I think we have to say that the agreement reached (provided certain conditions are met, such as sufficient publicity and correct closure) does represent a new consensus, and indeed an even stronger one than the implicit one we say exists when something hasn't been disputed for some amount of time. If explicit consensus decisions are not to be respected as consensus decisions, then we go back to the situation where nothing can be done without unanimity except by resorting to edit-warring.--Kotniski (talk) 14:44, 14 November 2011 (UTC)[reply]
Yes, the author of the diagram (who was not me), is a strong advocate of “consensus is reached chiefly by editing”. I was a convert.
RE: What “represent a new consensus”. Do you not accept the definition: “Consensus is achieved when there are no further changes made”? You seem to be creating an artificial definition with too many caveats for it to be workable. Your notion of a “consensus decision” I submit is elusive in practice.
RE: “where nothing can be done without unanimity”. This goes to the question of how the group decides to reject the weaker argument. It’s a good question. --SmokeyJoe (talk) 01:42, 15 November 2011 (UTC)[reply]
No, I don't accept the definition that "consensus is achieved when there are no further changes made", whatever that would mean in practice (presumably that consensus is achieved when Wikipedia is switched permanently into read-only mode). Consensus on Wikipedia means that, when valid arguments are considered, most people agree (or something like that). It does not mean that everyone agrees, or that everyone who's prepared to edit-war over the matter agrees. --Kotniski (talk) 08:23, 15 November 2011 (UTC)[reply]
Maybe a longer version? Would you accept the definition that "consensus is achieved when there are no further changes made (directly relating to the point in question) after the same editors make other accepted edits nearby, and there are no unanswered objections on the talk page. " (obviously page protection, or editors departing exacerbated is not evidence of consensus). The problem I have with your definition lies in the pragmatics. Who decides when valid arguments have been considered? Who decides who and how many can be excluded from the “most people agree” camp? --SmokeyJoe (talk) 22:12, 15 November 2011 (UTC)[reply]
An uninvolved admin (or other uninvolved editor of good standing). Obviously that then raises further questions, but thankfully we don't have disputes over such matters too often. But I don't think we can have a definition that makes consensus (ultimately) dependent on "no changes being made" - that would effectively entitle any single belligerent-minded editor to prevent a matter being settled in a way he/she doesn't like. --Kotniski (talk) 10:46, 16 November 2011 (UTC)[reply]
My objections are that the proposal diagram doesn't reflect current practice, it can't be edited, and it doesn't self-destruct when an editor realizes that it misrepresents the contents of the article. Again, if the diagram says something different from the article, I assume that the diagram should be ignored. Where does it say that? And if that is the case, what is the purpose of the change? --Ring Cinema (talk) 13:09, 14 November 2011 (UTC)[reply]
Ring Cinema, I don’t get what you are saying. What is “it”? Are you saying that the current diagram is in conflict with the policy text? If so, can you explain more precisely? --SmokeyJoe (talk) 13:55, 14 November 2011 (UTC)[reply]
I think Ring's point is that whilst policy pages should be edited only with care, it should still be possible for anyone to edit them, including diagrams. This may be a fundamental problem with WP, that it doesn't provide a diagram editor. Uniplex (talk) 14:17, 14 November 2011 (UTC)[reply]
I have to apologize for a confusion. I misread or misspoke "current diagram" in some places above as the current proposal. So, to be clear, the current diagram (as found now on the page) is not a problem with me. It states less than the policy so it's fine. The proposal diagram on this page is a problem. It's not an accurate statement of the practice. Again, my apologies. --Ring Cinema (talk) 14:26, 14 November 2011 (UTC)[reply]
Ring, you said "if the diagram says something different from the article, I assume that the diagram should be ignored"—the current diagram does say different to the policy. Cherry-picking is not fine: it's misleading; were similar to occur in an article it would be called POV. Uniplex (talk) 14:44, 14 November 2011 (UTC)[reply]
My point is that if the diagram is different from the policy, a mistake was made in the creation of the diagram. For that reason, the current diagram -- simplified and therefore clearly not a full statement of the policy -- is useful but the muddled proposal on this page is bad (i.e. confused, unclear, contradictory). There's no question of cherry-picking with the current diagram because it's a stripped-down representation that can't be mistaken for the full policy and probably does help a newcomer. --Ring Cinema (talk) 15:48, 14 November 2011 (UTC)[reply]
Most newcomers have their edits/arguments rejected because they don't "get" V. When they start to get V, they then fall foul of RS, and their next hurdle is often DUE. Suggesting to compromise doesn't help them along this road of understanding at all. I'm not proposing that the diagram that I drafted should necessarily replace the existing one (in fact, it seems that there are valid arguments for having no diagram); at this stage it's primarily a vehicle for discussion and invites helpful suggestion for improvements to get it to a state where it might be useful for further dissemination. Uniplex (talk) 16:32, 14 November 2011 (UTC)[reply]
(edit conflict) I have to say, this section itself is a perfect example of typical work through consensus, and it is not represented by the current diagram on the policy page, but is represented by the proposed diagram. It is not represented in the current diagram because the current diagram starts with someone editing the article (or, in this case, the policy page). Yet no one has done that here. A proposal has been made, it was not "clear how to proceed", so it went to "seek more opinions". Now we're all in box 1 together talking about whether change is warranted. This is very typical for how edits are often done.
The other way edits are often done is also represented in the proposed program - where the box 1 deliberation is done within the mind of a sole editor, who decides it is clear how to proceed, and he goes ahead and edits. Now, some comments about what was said above:
Kotniski, first you said, " I'm not seeing anything useful in the (latest version of the) diagram that's been proposed.", then you said, "Often discussion is started before making a change, ...". Well, what's useful in the proposed diagram is that it clearly depicts that discussion is often started without making a change", while the current diagram on the policy page does not. Do you not see that, or am I missing something?
Then a bit later you said, "another problem I'm seeing, looking at it again, is that the "implementation" of a "compromise" is treated the same as a bold edit,", so now you are seeing not only one problem, but two? Is that with the proposed diagram, or the current diagram? Or maybe you're saying the problems you see in the current diagram are not resolved by the proposed diagram? Well, this particular problem is addressed by the proposed diagram, I think, by intentionally conflating both kinds of consensus (bold and compromise) into what he calls "provisional consensus". But is that not accurate? I mean a bold edit is often likely to be accepted by consensus, and "compromise" consensus edits often aren't. Do we really prefer compromise consensus to bold consensus? I mean, sure, when you know a particular change is against consensus, that's to be discouraged, but that's going beyond bold editing, isn't it? So, like Ring, I don't see a problem with conflating compromise and bold consensus. The "provisional consensus" notion is actually quite brilliant, if you ask me.
Ring, you say "My point is that if the diagram is different from the policy, a mistake was made in the creation of the diagram." How is the proposed diagram different from the policy? I mean, I think everyone understands that a diagram can only be a high level overview, and that you have to look at the text to get the details of what exactly each box means, but do you see anything depicted in the proposed diagram that is incorrect or misrepresentative of policy? I don't, but I do see misrepresentation in the current diagram on the policy page (it implies everything starts with edits, which ignores situations like this very one).
SmokeyJoe, you said, "There is little point in starting a discussion in the absence of an edit and a revert (or further edit)". I just noted above how this entire section is an example of exactly that - starting a discussion in the absence of an edit and a revert. I find this to be very familiar and typical. Do you not? --Born2cycle16:43, 14 November 2011 (UTC)[reply]
Yes, I find this familiar and typical, but not in a good way. There is an implied edit here – the removal of the figure. This conversation would likely be clearer had someone actually removed the figure. At the moment, this conversation confuses the removal of the figure entirely and the discussion of new versions. This confusion hinders progress. --SmokeyJoe (talk) 01:54, 15 November 2011 (UTC)[reply]
How can something be so clear and obvious to one (me) and so confusing and muddled to another (you)? Specific issues with the current diagram have been identified by multiple people. There is a proposed version which is evolving in the obvious hopes to replace the current one. Some people have suggested removing the original one and not replacing, but no one wants that enough to actually have done that, apparently. So the discussion is mostly about replacing the current diagram with the proposed diagram (or a revision of it). --Born2cycle02:00, 15 November 2011 (UTC)[reply]
How can something be so clear and obvious to one and so confusing and muddled to another? Easy. You can have a lengthy meandering discussion that is decoupled from editing. This is my point about what is good about the current diagram, directly contrary to others’ positions. This thread “Process diagram” begins with a post suggesting removal of the diagram. If this thread is no longer advocating removal of the diagram, then how is it a reasonable discussion, especially to someone newly arriving? “Specific issues with the current diagram have been identified by multiple people”. They mixed through the discussion, and not the focus of it, and I disagree that their issues read as specific. --SmokeyJoe (talk) 21:59, 15 November 2011 (UTC)[reply]
Most everything starts with edits. It's a distortion to imply that garden variety editing involves anything other than edits as a starting point. Then there's discussion in search of a compromise. So for a newcomer -- who doesn't contribute to policy pages anyway -- the current diagram is a useful streamlined summary. It's not complete but it doesn't pretend it is. --Ring Cinema (talk) 17:16, 14 November 2011 (UTC)[reply]
Sorry, maybe it's lack of coffee, but I just cannot understand your objection. One of the most common acts I see newbies do is post on talk pages making suggestions for the article, obviously because they're not sure if it's appropriate to add or whatever. And the difference between an experienced editor and such a newbie is that the experienced editor can do a better job in box 1 in his own mind, so he can answer the first question affirmatively, and proceed with the edit. But this is also true for newbies when it comes to relatively minor changes about which they are confident. All of this is covered in the proposed diagram, and yet you say it's a "streamlined summary" and "not complete". Of course it's an overview, but that does not mean it's incomplete. What, exactly, do you think is missing? And, even if something is missing, how is the proposed diagram still not better than the current diagram which is blatantly incorrect by implying everything does start with an edit, when it never does (unless it is vandalism or poorly conceived edit, it always starts with evaluation of the text, and consideration of other factors like UNDUE, as depicted in the proposed diagram). --Born2cycle17:25, 14 November 2011 (UTC)[reply]
To clarify my objections in a few words: my objection to the present diagram is that it's simplified to such a degree as to make it wrong about certain things; and my objection to the proposed diagram is that it looks so complicated that it's not going to help anyone understand (or want to understand) what we're trying to get across. Until someone can come up with a magic solution that is both reasonably simple-looking and reasonably accurate, I think we do best to dispense with any diagram and just use ordinary text like we do for almost everything else in our policies.--Kotniski (talk) 18:00, 14 November 2011 (UTC)[reply]
The diagram is appropriately simple. There is nothing in it that is wrong. There is no case to remove the simple diagram, it works well and faultlessly. I suggest a more detailed diagram can be added after the simple one. Please review the several in the archives. --SmokeyJoe (talk) 02:01, 15 November 2011 (UTC)[reply]
With respect, I think the reason you don't see anything wrong with the diagram is that you yourself are wrong (or out of sync with the community's views) about certain things. You seem to think that consensus is purely an editing process (when people stop editing, we have consensus), which is commendable up to a point (I agree that discussion is generally overrated, and constructive co-editing underrated, as a means of producing better content), but it can't be only that - discussion and explicit consensus-forming/decision-making inevitably play an important part too, and the diagram doesn't properly reflect that. In fact I don't think any diagram can, as the two processes are often concurrent and interact in too many ways for us to be able to predict.--Kotniski (talk) 08:31, 15 November 2011 (UTC)[reply]
I don’t think I am “wrong” or “out-of-sync”, I think the problem is that you want more from the diagram than it gives. I certainly don’t pretend that the diagram alone covers everything. It is necessarily a simplification of the text that it supports. I personally support including additional diagrams with more detail. We attempted that years ago, and I think the effort is worth continuing. However, the current diagram is not wrong in what it does contain. It represents a very good method for finding consensus through coupled editing and discussion. It is obviously more complicated that WhatamIdoing’s very simplest and most common case below, which occurs when there is no opposition. And it does not cover situations of entrenched oppositions, where calls for further opinion are needed, or dispute resolution, etc.
I am sure that the proposed diagram is faulty because it doesn’t tightly couple the discussion to the focus of the discussion.
I do think that consensus is best achieved through editing. Discussion provides an abstraction of what people want. Editing demonstrates what people want.
When people stop editing, we have consensus, yes, except of course unless editors have been bullied or bored out of the process.
RE: “explicit consensus-forming/decision-making”. What is this? I don’t see it in the proposed diagram. It would be great it you could formulate a recipe or diagram for doing it, and I will be impressed if you can. If you can, it would have obvious application in the guidelines on how to close an RFC.
Can you clarify for me: Do you want to remove the current diagram? Do you want to work on a more comprehensive diagram? Do you think that the current diagram is worse than no diagram? --SmokeyJoe (talk) 21:48, 15 November 2011 (UTC)[reply]
I would happily remove the current diagram (though I could live with it as long as it's explained that it's a simplification, and preferably in what ways it's a simplification). I don't want to work on a more comprehensive diagram, since I don't believe this is an issue that can be conveniently represented diagrammatically (though someone may prove me wrong). And yes, I believe that the current diagram is worse than no diagram (again, because it misleads people into thinking there's some simple algorithm at work here).--Kotniski (talk) 10:52, 16 November 2011 (UTC)[reply]
Oh, and about “explicit consensus-forming/decision-making” - I mean the taking of decisions as a result of talk-page discussion (either through the discussion being formally closed, or through the result being clear to everyone). Of course I don't have an algorithm or a "recipe" for doing this, but it's an inevitable and important part of the process. (Personally I wouldn't say that this kind of decision-making is necessarily based on "consensus", but the word is so much part of wiki-religion that people have come to use it as jargon to describe any kind of decision that originates from editors rather than The Powers.)--Kotniski (talk) 11:01, 16 November 2011 (UTC)[reply]
The current diagram has six shapes connected by lines: so does the proposed diagram; however the proposed diagram describes about 99.9% of cases whereas the existing diagram covers a much smaller number. The notes could move to the text (if they're not already there) to yield a) a simpler looking diagram and b) text that can easily be revised. As long as WP keeps its collaborative editing model, the geometry/topology of the proposed diagram is unlikely to ever change. Certainly, we shouldn't show a diagram that's misleading, but if the diagram is clearly presented (and we're still waiting to hear what exactly is not clear) then those readers for whom a picture paints a thousand words can focus on the diagram; those who prefer text can focus on that. Uniplex (talk) 00:25, 15 November 2011 (UTC)[reply]
I think we could all agree on the very simplest and most common case, which looks something like this:
No, you're thinking about the editing process. The consensus process begins with someone's desire to change something. Until a change is wanted, then there's no point behind trying to decide whether everyone accepts the change. WhatamIdoing (talk) 02:54, 15 November 2011 (UTC)[reply]
The current diagram has the virtue of being so simple that it will not be mistaken for the actual policy. It is clearly a representation of something more complicated with the details left out. Since that is clear, that will be seen as its purpose. It is true that different editors go for consensus in different ways, and we are better off letting them do that. Much as I admire the intensity of this group's respect for WP, we must humbly accept that we are self-appointed experts who likely lack the magic wand of consensus. I think it is possible to take issue with the current diagram because it smuggles discussion in under "compromise"; it's a reasonable summary. --Ring Cinema (talk) 16:10, 16 November 2011 (UTC)[reply]
RC, no one, including you, has identified how the proposed diagram is an inaccurate or misleading summary of what policy is. I say it is accurate.
We agree the current diagram is over-simplified. You say "so simple it will not be mistaken for the actual policy"; I say, "so simple it does not reflect policy at all". --born2cYcle16:55, 16 November 2011 (UTC)[reply]
WAID, I'm definitely not thinking about the "editing process". AFAIK, the "editing process" is: "click on edit; edit; save".
I wouldn't word how the consensus process begins as "with someone's desire to change something", but I essentially agree. An editor decides an improvement could, or might, be made, with a change. If the decision is it could be made and it's clear what to do, he does it. If it's unclear it would be an improvement, or it's unclear exactly how to proceed, he consults with others. All of this is depicted in the proposed diagram.
Look, I had nothing to do with coming up with this proposed diagram. I just came along and recognized how amazingly accurate, and yet simple, it is. It's amazing in that it has as many shapes as the current diagram, and yet is comprehensive. I agree the presentation/layout could be clearer, but the flow depicted is spot on as far as I can tell. So far everyone who has objected to it cannot identify something wrong or inaccurate with it. It applies to every situation I've ever been or, or can imagine anyone being in, from the most trivial copy edit task to a months-long debate involving dozens of editors, countless proposed revisions, and multiple straw polls, etc. The linchpin of the diagram is box #1, which ingeniously represents the mental processes of individuals as well as points and reasoning shared in discussion which are key to maintaining and finding consensus. I think Uniplex deserves a barnstar for this stroke of genius. --born2cYcle16:55, 16 November 2011 (UTC)[reply]
The proposed diagram leaves out the most common editing process. According to the proposal, an editor who is clear about how to proceed makes an edit and arrives at consensus immediately. Another editor reverts that edit and we're not on the diagram any more. So according to the proposal the two editors are in different places -- one says it's clear how to proceed and the other is probably over in box 2. This is what I meant above when I said it lacks a point of view. The current diagram is from the perspective of the person reading the diagram; they don't have to read anyone else into it to figure out how it applies to them right now. So in that way it's a summary that works for newcomers: it answers the question, "What do I do now?" --Ring Cinema (talk) 18:24, 16 November 2011 (UTC)[reply]
Huh, you say they are in different places but don't identify where each is. No. That's not my interpretation at all. You're looking at it at different times. The first one is at "clear how to proceed" before he edits. After he edits they're both at provisional consensus. Then the second editor goes through loop, and reverts. The the fist editor either accepts the revert, or they both end up in #2/#1... What I see in the proposed diagram:
an editor who after evaluating text per box #1 is clear about how to proceed and so makes an edit and arrives at provisional consensus immediately.
a second editor who after evaluated text including 1st editor's change per box #1 is also clear about how to proceed and makes another edit (a revert, hopefully with a thoughtful/helpful edit summary), and also arrives at provisional consensus immediately.
now the first editor either accepts the revert, or goes through the process again, but this time via box #2 to bring in others, at least Editor #2, together with him into box #1 which which ingeniously represents the mental processes of individuals as well as points and reasoning shared in discussion which are key to maintaining and finding consensus.
They're both in the diagram and at the same place... right back where they started with the same text... at the top in provisional consensus. The difference is that presumably this time in box #1 the first editor will take into account the 2nd editor's input, or, if necessary, will "seek more opinions" per box #2 before proceeding.
The proposed diagram also answers "what do I do now?". The difference with the current diagram is that the proposed diagram gives a correct answer. --born2cYcle 18:50, 16 November 2011 (UTC) Revise --born2cYcle19:00, 16 November 2011 (UTC)[reply]
Your description doesn't seem to match the diagram. The first editor goes straight to consensus (blue blue blue green green). Every consensus is provisional, no? In the current diagram, he waits to see what happens to his edit ("Wait"). So the proposal seems wrong and over-specifies the policy. Secondly, the second editor is in another part of the diagram (2, I think) saying things like, "Hey I have a different opinion." But the proposal says that the consensus is already done. Of course, I know what you mean and how you want me to read it because I know the practices of WP. But that's not the point, right? Tendentious or creative misreadings will happen if they are allowed. --Ring Cinema (talk) 20:36, 16 November 2011 (UTC)[reply]
How long should he "Wait"—a week, a month, a year? In reality, he need not wait at all; he can do anything he likes. "the proposal says that the consensus is already done": no, the proposal says that consensus is provisional. Uniplex (talk) 21:15, 16 November 2011 (UTC)[reply]
Also, the box "Is it clear how to proceed?" seems superfluous. I see an identity between knowing how to proceed and agreeing to a change, unless there will be no change, in which case the diagram seems to strand us in the middle forever. --Ring Cinema (talk) 20:39, 16 November 2011 (UTC)[reply]
I am baffled by what you're apparently seeing and saying. I with you up to where you ask if every consensus is provisional. Yes. Then you say that in the current diagram it shows the editor waiting after making his edit (true, that's what it depicts), and that that somehow makes the proposal diagram wrong and an over-specification of policy. I don't see how that second part follows the first part at all. The wait in the proposed diagram is implied, and, accurately, is not necessary. The editor, can, for example, decide to make another edit, or even revert himself. He will not necessarily wait, which the current diagram incorrectly states he will do.
Then you say the second editor is in another part of the diagram ("2, I think"). It's unclear if you're talking about the second editor before or after he reverts, but in either case, he never takes the black path through #2. He also goes blue-blue-blue-green-[edit (revert)]-green.
The proposal never says "the consensus is done" - it says consensus is always provisional (another brilliant aspect of it).
I don't see how the diagram strands us in the middle. To use our own situation as an example, you and I are circling from #1, through the diamond you say is superflous, answering no each time, and reentering #1 via #2 (at the moment only you and I are actively participating, but we're certainly seeking input from others too). Are we stranded here? Perhaps, but that's reality. Until we get to a point where one of us feels it's clear to proceed, there we will stay. Eventually, we will either agree on how to proceed with a change (from #1, blue-blue-green-green), or we will realize we can't agree on a change or will agree no change is needed (blue-blue-red). In either case we will finally answer "yes" to what you say is a superflous question, and then either yes or no at the diamond in the middle.
If you're now trying to say the diagram could be misinterpreted - and giving example of how it could be - okay, we can work with that, by trying to make it more clear. Would you agree to at least try to work towards making this diagram convey more clearly what we want it to say? Because that's very different from an effort to try to show it's unworkable. --born2cYcle21:17, 16 November 2011 (UTC)[reply]
Perhaps I could ask you to re-read my last post with an eye to the fact that I sometimes refer to the proposal and sometimes to the current diagram. Sorry if I do that in a confusing way. That said, I think my objection stands up just fine. If an editor starts at 1 (why is that on the side?) and follows blue-blue-blue-green-green, he can think of an edit change, decide it should be done, do it, and have a new consensus all on his own without coming across any other editors. At that point, according to the diagram, his work is the new consensus. That is what the diagram says. I am not making that up. And that does not seem to be the actual process. Since it is not the way the practice goes here, I think that constitutes a divergence of the diagram from the practice, and that divergence is what I object to. (I'm not sufficiently interested in this to repeat myself on the other objections, although I think they're definite problems.) --Ring Cinema (talk) 01:19, 17 November 2011 (UTC)[reply]
I've gone back and reread your last two posts with particular care to whether you're referring to the proposed or current diagram - my same observations and questions stand as posted. I was not confused about which diagram you referred to when. I am confused by statements like this:
In the current diagram, he waits to see what happens to his edit ("Wait"). So the proposal seems wrong and over-specifies the policy.
How does the fact that the current diagram shows a wait after the edit support the conclusion that the proposal diagram is wrong and over-specifies the policy? Does the policy specify a wait? No (I even searched for the word "wait" in the text at WP:CONSENSUS... not there). It's the current diagram that over-specifies the policy by indicating that wait.
Now in this latest post you write:
If an editor starts at 1 (why is that on the side?) and follows blue-blue-blue-green-green, he can think of an edit change, decide it should be done, do it, and have a new consensus all on his own without coming across any other editors. At that point, according to the diagram, his work is the new consensus. That is what the diagram says. I am not making that up. And that does not seem to be the actual process. Since it is not the way the practice goes here, ...
Above you were told by both me and Uniplex that the numbers in the boxes are just labels. You always start at the top, at "Provisional consensus". Starting at the top the editor can follow blue-blue-blue-green-green. But from box #1, it's just blue-blue-green-green (the first blue got him to box #1 from provisional consensus). But that nit aside, yes, tracing that path is the normal path for a unilateral edit (if you will), and the end result is a provisional consensus for that updated text, which remains until that text is changed again, which could be a few seconds later, or not for years. You're not making that up. That is what the diagram depicts, and that is the actual process. Of course. What about this seems like it's not the actual process to you? --born2cYcle05:51, 17 November 2011 (UTC)[reply]
In general, more than one editor is involved in consensus-making. It seems more accurate and less likely to cause problems if it's recognized that consensus depends on some kind of acceptance, especially since it's implied that consensus is always provisional. There is a difference between a change that's worked over thoroughly and one that's just put up, even though both are revisable. --Ring Cinema (talk) 10:18, 17 November 2011 (UTC)[reply]
Of course there is a difference. But every change ever made, whether it lasted forever (so far), or just a few minutes, was 5 seconds old 5 seconds after the editor who made it clicked save. And if it was never changed again then the consensus support it had at 5 seconds was no different than the consensus support it had at 3 years. Time does not determine consensus for a change, but it helps confirm whether there is consensus, but even then it's always provisional. There is no bright line, and I fail to see how waiting is a factor in the consensus process we're trying to depict in the diagram. If you find something in an article that is problematic, it doesn't matter whether it was there 5 minutes or 5 years - you will, presumably, challenge it just the same.
Anyway, the current diagram implies that there is some unspecified "wait" period after an edit is made that determines whether there is consensus. But this is plainly wrong. First, even if the edit is reverted during the unspecified waiting period, that doesn't mean there was no consensus (in the extreme, it might have been a vandal that reverted, more realistically, it might be the one person who is against consensus that reverted). Second, even if the edit is not reverted, or not changed further, during the unspecified waiting period, that does not mean there is consensus. It could mean no one noticed. WP is riddled with unnoticed edits made by vandals that fall into this category. --born2cYcle18:25, 17 November 2011 (UTC)[reply]
I think that waiting to see how others react to an edit is part of what good editors do. It's certainly a better description of the process than the proposal diagram, which allows the erroneous interpretation that a new edit is consensus upon its adoption by a single editor. And your counterexamples seem to require bad faith edits to undermine the utility of the current diagram. This page is about consensus. I have other problems with the proposal as well, primarily based on its over-specifications. For example, the box labelled 1 seems to limit the valid reasons for an edit and I know from experience that if good edits have to be for specified reasons according to policy, then edits made for other reasons will be reverted for illicit reasoning. Sometimes that is good and sometimes it is bad, but this is not the place to take that up. --Ring Cinema (talk) 19:12, 18 November 2011 (UTC)[reply]
"than the proposal diagram, which allows the erroneous interpretation that a new edit is consensus upon its adoption by a single editor". No. The proposed diagram correctly presents that a new edit is provisional consensus regardless of whether it has been adopted by just one, or dozens, of editors, and that it achieves and retains this status as long as it is not changed again, or challenged, which could be five seconds, or five years or more.
You say that good editors wait, and that waiting is part of what they do. Wait for what? Wait before they do what?
Some editors edit and move on, and don't even know if their change is reverted or challenged. Others may watch the page to see what happens, but in the mean time they do others things too. I don't get the "wait" thing at all. To wait means to delay some other action. What are they delaying? I don't see how "wait" accurately describes what editors, including good editors, actually do, at all. Can you explain? --born2cYcle19:58, 18 November 2011 (UTC)[reply]
This is a sort of passing comment: If Ring is confused by the diagram, then it's likely to leave other people confused, too.
Also, the diagram gives a seriously misleading definition of "bold". Cleaning up someone else's typos is still bold editing, even though it's perfectly clear that typo correction constitutes an improvement. I really cannot support its current form. WhatamIdoing (talk) 02:34, 19 November 2011 (UTC)[reply]
The word 'bold' means 'courageous and daring'—experienced editors especially would not consider fixing typos to be bold. That said, my recommendation above was to minimize the text in the diagram per se and provide necessary notes as editable text. Arguably, mentioning bold editing is not necessary; it all depends how much you want to cover in the diagram and its notes. Uniplex (talk) 12:35, 19 November 2011 (UTC)[reply]
That's only in the real world. BOLD is wikijargon and has a substantially different meaning on-wiki, just like "Notable" means "qualifies for its own stand-alone article, as proven by the fact that multiple reliable sources have already taken notice of it" on-wiki, rather than what the dictionary says, which is closer to "worthy of someone taking notice of it (whether or not anyone actually has)". WhatamIdoing (talk) 21:12, 21 November 2011 (UTC)[reply]
Both good points. I think we're passed misunderstandings with RC and are actually at a disagreement about how consensus works. In particular, whether "waiting" is a fundamental aspect of it. Do you have a view on that question (please review the last few comments above)? Also, once we get that ironed out, I'd like to work on tweaking the diagram to make it easier to understand. And of course make wording changes to address problems like the one you just noted about bold editing. Any suggestions on what wording to use instead there? --born2cYcle06:00, 19 November 2011 (UTC)[reply]
I think some find the two chained yes/no questions hard to follow as they consider these to occur (in some cases at least) simultaneously. So maybe these should be collapsed into a single question: "Conclusion?", with three possible answers "Unclear", "No change warranted", "Change decided upon" or some such. Uniplex (talk) 12:50, 19 November 2011 (UTC)[reply]
I am not confused about what tie diagram says. For the record, I have no problem locating some parts of WP page editing in the proposed diagram. However, tendentious readings will be made under pressure and in that context detailed definitions become destructive and distracting. Wikipedia's strength is in its editors not its policies. --Ring Cinema (talk) 15:21, 19 November 2011 (UTC)[reply]
Good Lord, I've lost track of how many different different objections for the proposed diagram you've raised and, apparently, abandoned. If this is just a game of Whac-A-Mole for you because you're just against the diagram for some unstated reason, and so the stated reasons are not really important to you, we cannot have a productive discussion about this. --born2cYcle21:44, 19 November 2011 (UTC)[reply]
My objections are perfectly coherent; none of them are abandoned; all of them still stand. However, long posts are not read here. Careful consideration yields good decisions. I suspect the proposed diagram would not be a good idea and I am happy to share my objections in a eupeptic form. There is no problem before us and stable policies are a lot more important than adding a dubious diagram to a policy already functioning well. --Ring Cinema (talk) 13:56, 22 November 2011 (UTC)[reply]
"There is no problem before us"...!!! Have you attended any RfCs lately? They're mostly freakin' disasters, inordinately inefficient, and the just the tip of the iceberg of the petty squabbles going on underneath. "a policy already functioning well"... Why do you think we've arrived here at this policy page? Because we've observed it functioning so abysmally. Also for consideration, why do you think, as has been reported, that "good editors are leaving WP in droves"? Jimbo himself complains that the project looks backwards instead of looking forwards. I'm astounded by how far removed from reality your statements appear to be. Uniplex (talk) 07:52, 23 November 2011 (UTC)[reply]
I get that you're generally opposed to incorporating the proposed diagram, but your specific objections are a mystery, at least to me, and I've been trying to pay very close attention. Note that I at least recognized you were not confused about what the proposed diagram says. If you're just going to repeat vague general objections without engaging in a genuine effort to explain what your specific objections are so that they can be addressed, or so that others realize that they cannot be addressed, you're just being disruptive. You had two somewhat specific objections to the latest revision below. I addressed one, and Uniplex addressed the other. If those objections "still stand", please explain, in detail, how and why they were not adequately addressed by us. Anything else? --born2cYcle16:50, 22 November 2011 (UTC)[reply]
Break 2 (process diagram)
Consensus is the broad, community agreement that we strive for in making decisions on Wikipedia.
[Assessing ‘true’ consensus (observing that the entire community gives simultaneous agreement) is neither practical nor particularly useful since, because the community, its views, and the world at large on which those views are based, are in a constant state of constant flux, any determination of consensus is potentially out-of-date the moment after it has been made.]
Consensus in Wikipedia is deemed to be provisional, that is, it may be considered to represent ‘true’ consensus, providing that it is not subsequently challenged and replaced with a new, provisional consensus.
The decision-making process in practice is depicted in the diagram and starts from A, the current state of a page. As above, consensus is provisional.
Changes (adding, removing, or modifying content) are most often the result of a single editor judging that change is needed and editing the page. This is shown by the path A→B→C→D→A.
To determine if and what change is warranted, an evaluation (B) per sources and Wikipedia's policies and guidelines is made; this, even if by a single editor, best ensures that the result should constitute community consensus.
At D, all that is needed beyond editing the page (citing sources as appropriate), is to give an appropriate edit summary, including the reason for the change (e.g. "... per given source", or "... per guideline X").
Others who view the page may decide to change it further (ABCDA again) or give silent agreement (ABCA).
If it is unclear whether to make (or perhaps revert) a change, the issue is opened for discussion (E) at an article-, policy-, or other talk-pageas appropriate, and evaluation (now a collaboration by a group of editors) resumes at B.
Note that in "Seek more opinions", the opinions sought are not on article topics, rather on interpretation of policy, or on the reliability or significance of a source.
Though it may be revisited as necessary during the evaluation, the most important thing to first establish is the set of criteria by which the potential change will be judged.
This involves identifying which policies and guidelines apply particularly to the matter in hand, and determining which types of reliable source will be used to provide the evidence for the criteria to be assessed.
If multiple source types are considered pertinent, "weighting" them in some way is sometimes appropriate. For example, depending on the subject matter, mainstream press may be considered more or less authoritative than books on the subject. WP:RS and WP:NPOV discuss this in detail.
Once the criteria have been established, it should be a relatively straight-forward task to consult sources to gather the evidence required to make the decision. If the decision seems difficult then options include: gathering more evidence, reestablishing the criteria, or it may become apparent that the the discussion would be better moved to a more appropriate forum (e.g. from article-talk to guideline-talk).
Note that to maintain clarity in a group evaluation (whose members may "dip in and out" as their time allows), it can help on the talk page to keep established information (such as the criteria and the gathered evidence) separate from the journal of related discussion.
Pared down diagram with editable notes. Seeing as the proposed diagram takes a generalized view, and the existing diagram details a smaller set of specific circumstances, if they are clearly designated as such, there's no reason why we can't have both. Uniplex (talk) 16:40, 20 November 2011 (UTC)[reply]
It has the same problem as the previous proposal: a single editor goes all the way to consensus without any other editorial input? Plus, is there really an identity between consensus and provisional consensus? If so, why "provisional"? Seems to me experienced editors are aware that a consensus goes through stages of acceptance. And if not, it's an extra word. I'm not sure what problem is being solved by adding a diagram that can't be edited. --Ring Cinema (talk) 23:07, 20 November 2011 (UTC)[reply]
Yes, a single editor goes all the way to consensus without any other editorial input, just like in the current diagram...
[previous consensus] → [make an edit] → <was the article edited further?> (no) → [new consensus]
And, yes, there is a distinction between consensus and provisional consensus. There really is no such thing as consensus in Wikipedia - it's really always provisional, and the proposed diagram emphasizes this, which is good. --born2cYcle01:32, 21 November 2011 (UTC)[reply]
Ring, that consensus can be considered to be "provisional" is explained in the first note: "since pages are always subject to further change". The ABCA cycle occurs every time someone reads a page or a diff and lets it stand as is. Prior to such, the consensus for the page is provisional (subject to change), and after such, it is still provisional (subject to change)—this is the very nature of Wikipedia. You appear to have an objection to diagrams "that can't be edited"; one can only wonder as to what relevance this might have as: AFAIK, it applies equally to all diagrams in WP; if it doesn't, a helpful response would be to suggest a more suitable format to use—either way, the diagram is so simple that modification through recreation is an easy option. Uniplex (talk) 11:02, 21 November 2011 (UTC)[reply]
On the "wait" issue - it seems like there is an unstated general agreement that the period between an edit and any other edit to that page is significant to the consensus process. That is, if you make an edit it's especially provisional until someone else makes an edit; that if someone else makes an edit, without reverting your change, that somehow counts as overt acceptance of your change.... now there are two people who apparently support the change, which makes reverting it a bigger deal. Maybe this is what RC is trying to get at, and sees reflected somewhat in the current diagram but not in the proposed diagram? --born2cYcle16:38, 21 November 2011 (UTC)[reply]
The infinite loop running through BCE seems altogether too accurate a description of seriously dysfunctional disputes to be something that we want to recommend as a best practice. WhatamIdoing (talk) 21:16, 21 November 2011 (UTC)[reply]
But it doesn't suggest an infinite loop, since one way or another, some kind of conclusion is eventually reached, especially as "seek more opinions" escalates as per the dispute resolution process. Like it or not, this is the recommended and tried-and-true practice. --born2cYcle23:16, 21 November 2011 (UTC)[reply]
That's basically wishful thinking. Seeking more opinions on a content dispute does not always result in reaching a conclusion about what to do with the article. In fact, #1 under ==Pitfalls== is "too many cooks", i.e., you have sought more opinions and it has made the dispute worse, not better. WhatamIdoing (talk) 19:15, 22 November 2011 (UTC)[reply]
♣ The role of the diagram is not to encourage any particular path; that's a job for the text. The diagram merely reflects certain aspects of the text, specifically, the most common paths taken in the consensus process. Perhaps the title of the diagram could make its role clearer in this respect. Perhaps note E could express caution at progressing up the chain; perhaps all the notes could be worked into the main text (or are already there). The specific benefit that the diagram brings, is that in reference to it, the current text can be made shorter and still convey the same meaning. This is true for two reasons, paths may be referred to in the text very succinctly (e.g. ABCA) and some of the paths that are currently treated as being different in the current text, are in fact the same path, or the same path with only a modicum of textual qualification. Consider the diagram as a road map: a road map doesn't tell you the best route, but it is used in conjunction with a route-finder app that does. Uniplex (talk) 11:50, 23 November 2011 (UTC)[reply]
I don't accept the "up is down" logic that provisional consensus is distinct from consensus because all consensus is provisional (see B2Cycle above). 2) I don't accept the "two are one" logic that Uniplex uses to imply that consensus is the same kind of provisional both before and after subsequent edits. The true statement that all WP articles can be changed does not imply that consensus is constant after a new edit. Rather, there are several criteria one can reasonably employ to assess the "stickiness" of an edit, i.e. time, discussion, subsequent edits. --Ring Cinema (talk) 05:34, 22 November 2011 (UTC)[reply]
It has been neither implied nor stated that "consensus is constant after a new edit". Neither has it been implied nor stated that the diagram offers a means to assess the "stickiness" of an edit. It has been stated (many times now) that consensus is provisional after any edit: i.e. that the edit represents consensus providing that is it not subsequently challenged. This is the reality we find ourselves in at WP; one can never know if a challenge will subsequently occur. For example, this edit, was eventually reverted as practical-joke/mistaken after being present in the article in one form or another for 19 months. Did that edit represent the consensus view during that time? No, it just hadn't been properly assessed. Did that edit have provisional consensus during that time? Yes, it was taken at face value, copy-edited, and worked in with other text, but it was always subject to the eventual assessment that would see it reverted. Does it have "constant" consensus now? No, the current text (without the edit) could be challenged again today. Will WP articles ever have "constant" consensus? Unlikely, but that's not the point, such things are best discussed in the text: the diagram's role is to inform you or to remind you of most of what you need to know, for most of the time. Uniplex (talk) 07:47, 22 November 2011 (UTC)[reply]
Okay, let's try an example and see whether it helps. Here is an actual, good-faith change made yesterday by an editor. This person presumably:
A. Read the article (which had "provisional consensus" according to this diagram).
B. Considered a possible change.
C. Concluded that a change was needed.
D. Made the change.
and—according to this diagram—this change, which replaces a direct quotation from a reputable source that is firmly against this thoroughly discredited alternative cancer treatment with a made-up recommendation to use it—instantly acquired the status of "provisional consensus", even though we could trivially predict that nobody was going to agree with this person's change, and in fact it was reverted six and a quarter minutes later.
Now are you really prepared to declare that person's edit to be just as valid and just as consensus-driven as any other change made to the page? Did that change actually have any level of support that could be described as "provisional consensus" with a straight face? WhatamIdoing (talk) 19:23, 22 November 2011 (UTC)[reply]
I say again:
It seems like there is an unstated general agreement that the period between an edit and any other edit to that page is significant to the consensus process. That is, if you make an edit it's especially provisional until someone else makes an edit; that if someone else makes an edit, without reverting your change, that somehow counts as overt acceptance of your change.... now there are two people who apparently support the change, which makes reverting it a bigger deal.
How to address this in the diagram? What if we added:
[E. Wait for another edit to the page] → <is change reverted> → (yes) → [F. change has provisional consensus]/ (no) → [F' change is not consensus] → [A. provisional consensus]
In words, once the change is made whether it has provisional consensus is determined by the next edit to the page. If it's reverted it obviously does not; if it's not reverted, then it apparently does. Either way, you're back to Box A. --born2cYcle20:06, 22 November 2011 (UTC)[reply]
WAID, I can't declare that an edit is not valid until I've reviewed it, and nor can anyone else. Furthermore, there is no formal review process, no guarantee of when or if an edit will be reviewed. In the mean time, the edit has provisional consensus and readers may be viewing it. This is the very nature of WP's editing model: we trust our editors, but no more than to declare their editorial decisions as provisional. The probability of an edit having full, community consensus may go up depending on the breadth of the discussion that approved it, and the length of time it remains unchanged—we never formally measure this probabilty though, because our process doesn't require us to. (Note to self: probability of community consensus may also go down with time, as information may go out-of-date and the likelihood of a challenge goes up.) In these terms, it all sounds a bit crazy—how could it possibly produce a useful encyclopedia? But it does. Uniplex (talk) 21:16, 22 November 2011 (UTC)[reply]
Jumping back: Uniplex, we agree that consensus is mutable by nature. However, I think we part company in our willingness to leave the impression that consensus becomes sticky. The diagram that you defend says that all consensus is the same kind of provisional, and I don't think a project page on consensus should give that impression. Consensus is different before and after other editors have been on the scene. Yes, there is a sense in which both are provisional, but it is so equivocal as to be misleading to use the same expression for a fresh edit done this morning and a compromise solution arrived at after two weeks of discussion. I find it a misrepresentation of consensus rather than a useful simplification. This seems to be the position you are defending, although you might point to something in the diagram that makes this distinction clear. Thank you. --Ring Cinema (talk) 15:18, 23 November 2011 (UTC)[reply]
I suspect that both you and WAID are misinterpreting the purpose of the diagram, which is to supplement the text. The diagram shows the general case; how "sticky" consensus is depends on specific circumstances; the diagram makes no comment on it. The reply ♣ I made to WAID above may help. Also, what "bad thing" are you afraid of happening if we use the diagram? The rules are not being changed. Bad edits will still be reverted. Edit warring will still be disallowed. Uniplex (talk) 15:53, 23 November 2011 (UTC)[reply]
While it may in your mind be a supplement, it will be read tendentiously. Again, there seems to be an easy case/hard case problem. If I was going to design a diagram to supplement, I might think about box labels that match the article's section headings. Or a box for each process with arrows for segues. Discussion seems to get short shrift in this proposal even though it is the place where things get worked out. --Ring Cinema (talk) 16:29, 23 November 2011 (UTC)[reply]
Couldn't writing "See text for details" next to the diagram perform a mind transfer to the readers? Discussion of discussion is where we are trying to get to, if only we could get the basic ground rules in place first. Uniplex (talk) 16:49, 23 November 2011 (UTC)[reply]
Should we be able to? Sure, if every single editor were a cooperative, mature person who wanted nothing more than the ideal contents for the project. In practice, it doesn't work that way. WhatamIdoing (talk) 18:54, 23 November 2011 (UTC)[reply]
Sorry WAID, you've lost me there. Let's roll back: RC complains that the fact that the diagram supplements the text is in my mind and not in the readers' minds, is a problem. I suggest that we can transfer it from my mind, to a note—"See text for details"—and from there to the readers minds. RC also expressed an interest in the discussion area of the process; I suggested that we ought to try to do one thing at a time. What are you suggesting? Uniplex (talk) 19:07, 23 November 2011 (UTC)[reply]
In translation: "Rejoice! You are neither cynical nor a born bureaucrat!"
It does not matter how many disclaimers you add. If you put up a diagram that is wrong, or at least one that could be seriously wrong in some messy dispute, someone is going to pound on the table and demand that the diagram is the one True™ way. You can no more stop them from doing this than you can stop them from taking half-sentences out of context. WhatamIdoing (talk) 01:59, 24 November 2011 (UTC)[reply]
This comment is in reference to Uniplex's statement, "I can't declare that an edit is not valid until I've reviewed it":
I understand what you're saying, but the fact is that a change like this does not have consensus from the very instant that the change is thought about, even if nobody ever sees it. It is not possible to have consensus for a change that seriously and unquestionably violates major content policies.
"Consensus" is not "the agreement of the couple people who happen to notice the change". Consensus is based on the whole community's views. It is not possible for a POV pusher or a spammer to override the major policies even if nobody ever sees their bad changes. That's what we mean at WP:LOCALCONSENSUS when we say "Consensus among a limited group of editors, at one place and time, cannot override community consensus on a wider scale."
This edit was a serious violation of consensus from the very beginning. Any individual might not be able to act on that violation (by reverting it) until he has seen it, but there is zero consensus for the change even before any specific human realizes that we have a problem, because there is a strong, unshakable, community-wide consensus not to dramatically misquote sources. WhatamIdoing (talk) 19:02, 23 November 2011 (UTC)[reply]
WAID, I don't disagree with you, but I'm not sure what bearing this has on the diagram. Perhaps I can ask you the same question that RC steadfastly refuses to answer: What "bad thing" are you afraid of happening if we use the diagram? (Since no rules are being changed: bad edits will still be reverted; edit warring will still be disallowed, etc. etc.) If someone would answer the question then we might be able to do something about it. Uniplex (talk) 19:14, 23 November 2011 (UTC)[reply]
No, your accusation is inaccurate, Uniplex. I have repeatedly raised the ways in which the proposed diagrams can be read that are antithetical to current practice or good practice. Now, perhaps you want me to then say, "And the editors will then be allowed to insert falsehoods into the articles." However, a clear causal link from a bad diagram to a poor WP article might never be made. But that is not the point, right? I think that I have less faith in our ability to make a rule and know how it will be followed. (You are incorrect that adding the proposed diagram changes no rule; you are talking about your intention.) That is why I like to mention that the best path to WP excellence comes from experienced editors working within a stable set of practices that are widely understood. Changes to this page are not stable. Changes to this page that can be read in bad ways are not stable. So, what bad thing have I steadfastly refused to mention? I have been quite clear: the proposed diagram could make the work of good editors difficult and I've said it implicitly or explicitly in every post. --Ring Cinema (talk) 21:01, 23 November 2011 (UTC)[reply]
I assure you, Ring Cinema, you have been anything but clear. I cannot think of any discussion in which I have ever been involved on Wikipedia in which someone's comments were less clear or more nebulous than yours have been. After spending hours over a number of days reading and thinking about everything you've written, I have only the slightest and most general of inklings of what you're trying to say. By the way, saying you've been clear and that you've said "it" does not make it true.
If you really wanted to be clear, then you would do this:
State your objections as clearly as you can in a draft post,
Review everything in that draft that you've already said above and how others have responded to it, and
incorporate clarifications in your statement that addresses what others have said.
Uniplex, what matters is that your diagram says that this zero-consensus-from-its-conception edit has "provisional consensus". It doesn't have provisional consensus. It has no consensus whatsoever. It is exactly the kind of edit that is prohibited by all of the content policies. But you have incorrectly and misleadingly labeled that zero-consensus edit as having some kind of consensus, solely on the grounds that it happens.
The first "bad thing" that's going to happen is that some people will be honestly confused (especially people who don't read the text, and that's a lot of them) and thus believe that this zero-consensus edit has some sort of consensus. The second "bad thing" that's going to happen is that wikilawyers and POV pushers will find it very convenient for insisting that their anti-consensus changes have some sort of consensus. Neither of these are beneficial to the overall project. WhatamIdoing (talk) 05:39, 24 November 2011 (UTC)[reply]
WAID, firstly, referring to the diagram as "your diagram" belies and attacks the fact that it is the result of a collaboration: it has thus far, been updated to incorporate the comments of (off the top of my head) five other editors. Secondly, your comments belie the fact that virtually every WP policy, guideline, and paragraph therein, cannot be read in isolation. Thirdly, the diagram does not imply that deliberately bad changes have any sort of consensus: a deliberately bad change omits B: Evaluate existing text as needed per sources, policy, and guidelines, to determine if change is warranted. Uniplex (talk) 07:42, 24 November 2011 (UTC)[reply]
RC, though you still haven't answered the question in a way that could allow the diagram to improved, we now understand your objection: the diagram represents a change, and change, in your opinion, is bad. Here are some comments from Jimbo for you to ponder: "This enshrines old bad practices and privileges the past over the future ... That radically conservative attitude conflicts with WP:BOLD and WP:IAR". Uniplex (talk) 07:42, 24 November 2011 (UTC)[reply]
Is that what he's trying to say? It's change and change is bad? Really? Wow. In any case, the proposed diagram seems like a change to how the policy is presented, but not a change to the policy. --Born2cycle (talk) 08:56, 24 November 2011 (UTC)[reply]
This is particularly alarming: "the best path to WP excellence comes from experienced editors working within a stable set of practices that are widely understood". Okay, so I'm not Jimbo, but I'm certain he'd say something more akin to: "WP excellence will come from simple, efficient practices that are easily understood by new editors". His comments just above (BOLD, IAR) are in reference to policy improvement. Uniplex (talk) 12:38, 24 November 2011 (UTC)[reply]
Okay, so there we have it. You mistakenly believe that new editors, many of whom don't know what they're doing, will somehow make improvements, when it is pretty obvious that the backbone of WP is the experienced editors who do the bulk of the work. To undermine the experienced editors by destabilizing important policy is a destructive idea. So, yeah, wow, I have an idea about how to maintain a well-functioning institution. You don't.
Speaking of things that haven't been articulated, there is a gaping hole where the reason to change the policy should be. Uniplex? B2C? Any ideas about what you're trying to fix? --Ring Cinema (talk) 15:35, 24 November 2011 (UTC)[reply]
I suppose I should mention, for the record, that B2C and Uniplex have misstated my objections rather crudely. I don't have one objection, and they land on different levels of detail. It is not hard to understand (e.g.) that the proposed diagram can be read as different from current practice -- even if the advocates claim they don't intend that -- but that reading is very simple. Since we don't need this diagram, we should do with it what we do with all the proposed edits we don't need. --Ring Cinema (talk) 15:41, 24 November 2011 (UTC)[reply]
Ring, I'm sure that you're already aware that Delete – No need is an WP:IDONTLIKEIT argument, one to be avoided. As for the reasons for looking for improvement, several have already been given; you did not respond. Here are some again:
"There is no problem before us"...!!! Have you attended any RfCs lately? They're mostly freakin' disasters, inordinately inefficient, and the just the tip of the iceberg of the petty squabbles going on underneath. "a policy already functioning well"... Why do you think we've arrived here at this policy page? Because we've observed it functioning so abysmally. Also for consideration, why do you think, as has been reported, that "good editors are leaving WP in droves"? Jimbo himself complains that the project looks backwards instead of looking forwards. I'm astounded by how far removed from reality your statements appear to be. Uniplex (talk) 07:52, 23 November 2011 (UTC)[reply]
What do you think happens to experienced editors? They leave. What will new editors become (if they don't run off screaming)? Experienced editors. Why do you ignore Jimbo's views? Do you even know who he is...? Uniplex (talk) 17:12, 24 November 2011 (UTC)[reply]
I see the genetic fallacy in that reference to Jimbo, Uniplex. I'm sure Jimbo can speak for himself. I also don't know what you mean when you say that experienced editors leave, since the definition of an experienced editor is one who stays. I would take issue with the idea that the proposed diagram would make RfC's more efficient or that the policy on consensus is the cause of any problem. Depending on consensus is intrinsically difficult so it's unsurprising that there are sometimes problems; it is removed from reality to believe that there are rules for consensus that would obviate those problems. Even less do I accept that a flawed diagram that misstates the practice of page editing could improve the editing of pages. But I am open to good ideas about how to incrementally improve WP practices. You may recall that I mentioned the majority of the minority as a group that might be undervalued in our thinking. --Ring Cinema (talk) 15:48, 25 November 2011 (UTC)[reply]
Ring, I guarantee that, one way or another, every experienced editor will leave WP; think about it. Sorry, but "the flawed diagram" is IDONTLIKEIT—there is nothing that can be done for such claims. The diagram does not embody any change to WP practice; it merely reflects (a substantial portion of) the current text. Uniplex (talk) 20:14, 26 November 2011 (UTC)[reply]
If that is really your argument, Uniplex, it falls spectacularly flat on its face. I feel quite certain that we do not have a problem with mortality attrition among WP editors, and if, as you claim, the proposed diagram was put forward to solve the problem of editors lost to death, you painted yourself into a corner. This makes it clear that the diagram isn't proposed for a useful stated purpose. As for your assertion that the diagram would not change WP practice, this is pure speculation on your part and, based on my experience, certainly false. --Ring Cinema (talk) 15:47, 27 November 2011 (UTC)[reply]
Break 3 (process diagram)
Let's remind ourselves of how this discussion started: problems with the current diagram. Despite its best intentions, the current diagram supports the following conclusions:
any edit (including bad ones) has consensus if no other edit occurs on the same page (after 24hrs, say)
any edit has consensus if other edits subsequently occur on the same page
NPOV editors have to compromise with POV editors (so that a partial amount of POV material receives consensus)
POV edit 1 modified by POV edit 2 makes for a strong, "sticky" consensus
These are serious problems for something that plays such a central role in WP's operation. It's no wonder that when day-to-day editing includes elements of the above, we end up with train-wreck RfCs. With this in the mix, editors just aren't sufficiently cognizant of WP's principles or best practice. If day-to-day editing didn't suffer from the above, most problems would be solved early on and few would even get to RfC.
The proposed diagram attempts specifically to address the problems in the current diagram, and thus be a good, base reference for what consensus is. It doesn't preclude having other diagrams which "unroll" particular paths worthy of specific comment. Problems with the proposed diagram raised and not addressed thus far:
possibly open to misinterpretation
The writer's challenge is to present ideas clearly and concisely—are we not up to that challenge? If not, why are we even here? If we use the diagram and a problem occurs, we'll fix it. If we can't fix it, we'll revert and think again, based on the valuable knowledge that we gained in trying—to make WP better. Uniplex (talk) 09:00, 25 November 2011 (UTC)[reply]
Do you have any evidence that your conclusions reflect reality? I think your conclusions are not realistic, certainly not under the scutiny that any bold edit attracts, and when push comes to shove, editors refer to the text, not to a simple diagram. However, I would like to encourage you make a more complete diagram, covering more of the details. This should be in addition to the simple diagram, which should come first. Experience is that "detailed and correct" inevitably is so complicated that it is not helpful as an introduction. --SmokeyJoe (talk) 09:13, 25 November 2011 (UTC)[reply]
Thanks for your thoughts Joe. I suspect that all of the "conclusions" I gave, do occur in WP to a certain extent (and I think WAID alluded to some of them, somewhere above); whether or not they occur because of the diagram, is hard to say—the existing diagram could be used to bolster such claims though. The proposed diagram covers a large number of scenarios and doesn't seem to have the same problems. In fact, just about everything that has been thrown at it so far, it has coped with. It doesn't include a "give up" option: I've hesitated to do so since a) it's fairly obvious that any process can be abandoned at any point, and b) it's a bit defeatist! (We have, of course, a text section on "no consensus"). Which are some of the most important details that you think should be included in the more complete diagram? I have an idea for describing best practice for consensus discussions, but I was thinking that this may be best in text... Uniplex (talk) 22:06, 25 November 2011 (UTC)[reply]
I don't see where any of the objections have been knocked down. The proposed diagram still has the same problems that were mentioned from the beginning. --Ring Cinema (talk) 01:23, 26 November 2011 (UTC)[reply]
As equally does the present diagram. Do people not see by now that consensus-forming is not something that follows a diagrammatic process, and we would do better simply to abandon all attempts to illustrate this policy with diagrams?--Kotniski (talk) 11:42, 26 November 2011 (UTC)[reply]
How do you distinguish processes that are "diagrammatic" from those which are not? I suggest that any process is diagrammatic, and that any process that is not diagrammatic is not a process. Are you saying consensus-forming is not a process? --Born2cycle (talk) 20:08, 26 November 2011 (UTC)[reply]
Well, yes, I suppose - at least, not a single process with defined steps and paths that can realistically be represented on a diagram.--Kotniski (talk) 11:10, 27 November 2011 (UTC)[reply]
"Consensus is the broad, community agreement that we seek in making decisions on Wikipedia"? This seems inaccurate and subject to many interpretations, some of which could be false. If the whole community doesn't participate in a decision, is it still consensus? According to this sentence, maybe not. Uniplex tried to include this on the main page already and I reverted it. Now, it's almost like he might want to smuggle it in where it can't be edited. I'm sure he's not thinking that way but here it is. --Ring Cinema (talk) 04:08, 28 November 2011 (UTC)[reply]
We strive for community consensus, but we only ever achieve an estimation of it, made frequently by no more than handful of editors. This is explained a bit more in the notes with the diagram above. @SmokeyJoe, per your suggestion, I've also added a first draft of an expansion of box B. Uniplex (talk) 13:25, 28 November 2011 (UTC)[reply]
Stability means not significantly challenged as well as not moved
This edit by Ring Cinema (talk·contribs) removed wording that clarified the meaning of stability with respect to article titles. The original words, with the removed part highlighted:
In article title discussions, no consensus has two defaults: If an article title has been stable (meaning without significant challenge as well as not actually moving) for a long time, then the long-standing article title is kept. If it has never been stable, or unstable for a long time, then it is moved to the title used by the first major contributor after the article ceased to be a stub.
Why was this removed? The edit summary says, "over-defining -- let the editors do the editing". Why leave such a general decision (what something means) ambiguous for the editors to debate on a case-by-case basis?
Consider two articles A and B. Say each was originally at some other title, but both have been at their respective current titles for over five years, despite repeated attempts by many people to move these articles, with RM discussions always resulting in "no consensus". Are these current titles "stable", or not? The answer to that question of course determines whether each stays at its current title or be moved back to the original title. If we don't clarify whether the repeated challenges means they are or are not stable, then anyone can argue either way, based on whatever title he or she prefers for some completely unrelated, irrelevant and unstated reason. In other words, it's meat for WP:JDLI rationalization. Why tempt like that?
Now, in theory, we could of course decide to define "stable" in this context either way. But in practice, I think we if we honestly look at the spirit of this idea to go back to the original title when the article is "unstable", we must admit we're talking about there being disagreement about the current title. Just because all the attempts to move it via RM failed because the current title is always favored in "no consensus" decisions does not mean it's stable. After all, the only reason it hasn't actually been moved repeatedly is probably due to the technical reason that you have to be an admin to move a title over an existing redirect that has history. I suggest it's inappropriate and unhelpful to define "stability" in a way that depends on a technicality like this.
Because it's out of scope. If you want to have that be the rule for article titles, then you need to convince the folks at WT:AT to agree with you. (Don't be surprised if it's rejected because by some on the grounds that it would be an outright gift the persistent complainers about the title chosen for the Ganga/Ganges River.) WhatamIdoing (talk) 02:24, 10 November 2011 (UTC)[reply]
The editors can handle this. They are smarter than us. My objection is that, while now there's a call to define "stable", next we'll have to define 'significant' or 'challenge', which begets definitions of the terms of those definitions. Definitions are not a solution. Experienced editors with an understanding of a stable set of practices are the key here. --Ring Cinema (talk) 03:48, 10 November 2011 (UTC)[reply]
I agree with WAID and would go further to suggest that this entire bullet point is out of scope. The policy should cover the principles of consensus; how to apply it in specific circumstances (or even commenting on such) ought to be in guidelines. Uniplex (talk) 05:37, 10 November 2011 (UTC)[reply]
Ring Cinema, subjective matters like quantifying "significant" are one thing, but it makes no sense to require editors to reinvent-the-wheel (in this case, decide whether "stable" refers specifically to no physical moves of a title, or also means no significant opposition to the title) on a case-by-case basis. This is a simple binary question. I get that we want to avoid defining too much in general, but this seems like exactly the kind of thing we do want to define so that it is interpreted the same way in all cases.
It makes no sense to decide that A is unstable in one discussion, because there has been significant opposition to it since the article was moved to A, even though it has never moved since, and also decide that Bis stable in another discussion, despite there being significant opposition since it was moved to B. And, yet, this is what we're allowing, arguably even encouraging, by refusing to answer this question one way or another. --Born2cycle (talk) 05:43, 10 November 2011 (UTC)[reply]
I think in practice, we default to the currently stable title regardless of how much it's been challenged. But I disagree that the entire bullet point is "out of scope" - this is a common and important example of how Wikipedia deals with consensus-related problems, and people coming to this page will want to know such things, not just be given meaningless airy-fairy ideals that have no clear practical application.--Kotniski (talk) 08:30, 10 November 2011 (UTC)[reply]
Policies and guidelines have distinct roles: examples of how policy might be applied in a given situation might be given to illustrate a policy point. However, this is a potentially dangerous thing to do since how policy actually is usually applied in a given situation is described in the appropriate guideline (which might be referenced from the policy page). Uniplex (talk) 09:02, 10 November 2011 (UTC)[reply]
Dangerous to clear and unambiguous expression, i.e. there is the potential that the reader will confuse might be applied and is usually applied as above. Policies and guidelines lie respectively in the realms of the general and the specific; whilst there might be a grey area where these realms meet, general editing of pages is clearly general, whilst precisely what to do about particular types of conflict over article titles (a specific mechanism, perhaps subject to change), is clearly specific. Examples in policy docs are best avoided, lest they be confused with guidelines; guidelines should be placed where they belong, in guideline docs. For example, no policy change would or should be needed to invent a new type of redirect—an "alt-title" redirect—which if used, would display it's name as the article title instead of, as currently, "Redirected from ..."; but as stands, this policy would likely need to be updated to reflect a minor change of mechanism. Uniplex (talk) 09:51, 10 November 2011 (UTC)[reply]
Examples frequently are helpful to assist in making the connection of theory to practice. It is true, though, that they have to be labeled as an example. The case here reads not as an example, but as a detail on the policy. Perhaps it was ill-advised to discuss the issue of no consensus in article title disputes, but at least what it says currently is accurate. --Ring Cinema (talk) 13:48, 10 November 2011 (UTC)[reply]
Uniplex, are you fully aware that all three of these examples are taken directly from actual policies? We're not making up new rules here. We're only repeating the bits of other policies (with links directly to those policies) that are relevant to consensus disputes. That's why I told Born2cycle that if he wants to change the actual policy on article titles, then he needs to go change the actual policy on article titles, not merely the summary on this page: these are mere pointers to the long-standing rules, not new rules. WhatamIdoing (talk) 20:40, 10 November 2011 (UTC)[reply]
Until a few days ago, the second para discussed reverting generally; after an edit reduced it's scope to that of bold edits, I regeneralised it. Given that this has just been reverted, I've restored the previous stable version and suggest we discuss whether or not the para should be general or specific to bold edits. Uniplex (talk) 13:53, 14 November 2011 (UTC)[reply]
Blatant violations of UNDUE, especially in BLPs, the list goes on and on... Picking up on a theme I just wrote about above, this is why I like the concept of "provisional consensus" so much, and that consensus stemming from both truly bold edits (when the editor genuinely believes there's a good chance the edit is consistent with consensus) and from so-called compromise consensus is pretty much the same thing - we don't prefer one to the other. --Born2cycle16:51, 14 November 2011 (UTC)[reply]
I don't see why any of these things are not "bold". The wording that Uniplex objects to is "Although other editors are perfectly entitled to make changes without prior discussion (to "be bold", in Wikipedia parlance), sometimes you might believe that a recent change is not an improvement. In this case you may decide to revert (undo) the change. When doing this, however, it helps to state (in your edit summary or on the talk page) the actual reason for your disagreement,..." I don't really see (a) how this restricts the principle only to bold edits, any more than the restored version of the paragraph does; or (b) why this sumary definition of "bold" would exclude all the bad types of edit mentioned above. If it's done without prior discussion, then as far as this paragraph is concerned, it's bold. (It might also be stupid, but it's still bold.) --Kotniski (talk) 17:54, 14 November 2011 (UTC)[reply]
When you reference WP:BOLD, what is said on that page is implied. And what is said at WP:BOLD is in reference to encyclopedia improvements. Implicit in a WP:BOLD edit is the presumption that the bold editor is making what he believes to be an improvement to the encyclopedia - that certainly does not apply to vandalism (which may be "bold", but they are not WP:BOLD). --Born2cycle18:08, 14 November 2011 (UTC)[reply]
OK, but when you revert vandalism, there's no particular encouragement to explain why you're reverting it (it will be obvious to everyone), so it's right that the paragraph should be narrowed in scope to exclude such reverts.--Kotniski (talk) 20:24, 14 November 2011 (UTC)[reply]
Well, "not okay", because vandalism isn't one of the many examples given above of non-bold edits. Naïve edits, uncited trivia, poorly written prose, and NPOV violations are very often BOLD edits. The opposite of a BOLD edit is a timid request on a talk page. The extreme of a BOLD edit is a reckless one. Edits do not stop being BOLD merely because someone else (or everyone else) disagrees that the change constitutes an improvement. Making a direct attempt at an improvement (even if made from an ill-informed POV) is enough to count as a BOLD edit. WhatamIdoing (talk) 22:26, 14 November 2011 (UTC)[reply]
Even discussed edits may be reverted—by someone outside the discussion. There are many edits that are reverted and the story ends. A bold editor OTOH, doesn't run away when someone reverts his change: he follows it up with a repeat of the same edit but with an additional cite, or a better explanation in the edit summary, or he discusses it BRD-style. An argument that if an edit was reverted, then it must have been bold, doesn't hold water---how does one know that it wasn't the revert which was bold? If we want to refer to un-discussed edits, just call them un-discussed edits. If we want to discuss reversion (which is I believe the intention of the para in question), discuss the mechanism and its implication to the process; don't conflate it with boldness—the reason for a revert is the same as for any other change: because the editor believes that it improves the article to do so. Uniplex (talk) 00:57, 15 November 2011 (UTC)[reply]
BOLD editors most certainly do "run away" when someone reverts their changes; in fact, I have repeatedly and strongly recommended that strategy to people who weed link farms. You can usually remove spammy links from ten or more articles in the amount of time that it takes to explain why you removed the links that you did. Ten improved pages usually contributes more to the ultimate goal of a well-written encyclopedia than one discussion about the links in a single article. WP:Bold, revert, discuss is an optional process. "Running away" is often the optimal strategy, and it, too, is a form of consensus: I'm accepting and agreeing not that your reversion is an improvement, but that my time is better spent in tasks other than discussing the changes with you.
High-volume editors do a lot of BOLD-and-ignore editing. Outside of the project pages, you could probably revert 90% of my edits without me even noticing, much less contesting, your reversion. I've edited more than 22,000 unique pages so far: it is impossible for any human to keep track of the changes in that many pages. WhatamIdoing (talk) 02:50, 15 November 2011 (UTC)[reply]
This discussion seems to have wandered off, so let's go back to the diff: [6], and look at the changes one by one. The first sentence, then: the old (restored) version has It is not necessary for editors to seek written permission in advance to make such changes, and the absence of a prior discussion does not prove that the change is not supported by consensus. Well, the first part has already been said in the paragraph above, and the second part has too many negatives to be fathomable (but is probably just trying to say the same thing again). Is it not better to replace this as proposed, with (something like) Although other editors are perfectly entitled to make changes without prior discussion (to "be bold", in Wikipedia parlance), ... This makes clear that we're not making a new statement, and gives a useful link and jargon explanation to boot. Is anyone saying this isn't better?--Kotniski (talk) 08:41, 15 November 2011 (UTC)[reply]
It is not necessary for editors to seek written permission in advance to make such changes, and the absence of a prior discussion does not prove that the change is not supported by consensus. However, sometimes you might believe that a recent change is not an improvement. When reverting an edit you disagree with, it helps to state the actual disagreement rather than citing "no consensus" or "not discussed". This provides greater transparency for all concerned, and likewise acts as a guide so that consensus can be determined through continued editing.
Although other editors are perfectly entitled to make changes without prior discussion (to "be bold", in Wikipedia parlance), sometimes you might believe that a recent change is not an improvement. In this case you may decide to revert (undo) the change. When doing this, however, it helps to state (in your edit summary or on the talk page) the actual reason for your disagreement, rather than just citing "no consensus" or "not discussed". This is not only polite, but also helps indicate what issues need to be addressed in subsequent efforts to reach a consensus on the matter.
I agree that the paragraph could use re-writing. I prefer the end of the proposed version, but disagree with beginning any paragraph with "Although". Both versions suffer the same fault of opening the paragraph in the negative. A paragraph is best opened be stating what it is stating. I suggest:
Sometimes you might believe that a recent bold change is not an improvement. In this case you may decide to revert (undo) the change. When doing this, however, it helps to state (in your edit summary or on the talk page) the actual reason for your disagreement, rather than just citing "no consensus" or "not discussed". This is not only polite, but also helps indicate what issues need to be addressed in subsequent efforts to reach a consensus on the matter.
Per its section title, the paragraph is about "Reaching consensus through editing". Why discard information pertinent to the subject? I.e. that undiscussed changes may be supported by consensus. Uniplex (talk) 23:00, 15 November 2011 (UTC)[reply]
Quite right. Would the following:
Sometimes you might believe that a recent bold change is not an improvement. In this case you may decide to revert (undo) the change. When doing this, however, please state (in your edit summary or on the talk page) the actual reason for your disagreement. Note that the absence of a prior discussion does not prove that the change is not supported by consensus. Reverting merely due to "no consensus" or "not discussed" is not polite. Citing the actual reason for your disagreement helps indicate what issues need to be addressed in subsequent efforts to reach a consensus on the matter.
"Not polite" is not a good choice of words here. "No consensus for this change" might actually be the politest way the editor can think of to say "Only a raving lunatic would believe that garbage".
The stronger reason to avoid that edit summary is that if you have a substantive objection, then "no consensus" doesn't communicate useful information about your legitimate concern. (If you don't have a substantive objection, then the edit summary is highly informative... but most people don't want others to know that they are thoughtless, bureaucratic obstructionists, so they, too, should avoid that edit summary.)
Where good faith is not to be assumed, such as for obvious vandalism or lunacy, I don’t think any of this policy page is intended to be applicable.
I personally am not attached to “not polite” and support the proposal to edit this paragraph being implemented, with all editors encouraged to improve the wording. To me, the paragraph says “give a substantive reason for reverts”, with the background reason “unexplained reverts are counter-productive to consensus building”. --SmokeyJoe (talk) 04:42, 16 November 2011 (UTC)[reply]
What's the motivation for rewriting the paragraph? Granted, the first and last sentences are a little wordy, so just a trim a few words (without changing meaning, such changes would be considered "minor" and not require prior discussion). Uniplex (talk) 06:17, 16 November 2011 (UTC)[reply]
So what do you think is (or should be) the substantial meaning of this paragraph (which is only part of the section). Or in other words, what points (about "Reaching consensus through editing" - the section topic) remain unaddressed by the first paragraph of the section, and thus need to be dealt with in a further paragraph(s). To me it seems we've already made the point about "bold is OK", and the points to make now are (1) "revert after bold is also OK"; (2) "explain your reverts". Is there anything else you want to say?--Kotniski (talk) 11:08, 16 November 2011 (UTC)[reply]
The second and third sentences are clearly stated and their meaning is self evident (and not covered in para 1). The last sentence gives justification for, and consequence of, the instruction given immediately before. The first sentence largely provides context for the second (which invites comparison with the first through "However")—of course, part of the first sentence has been copied to the first para so there is some overall redundancy. Uniplex (talk) 20:07, 16 November 2011 (UTC)[reply]
So do you agree that a shortish half-sentence beginning "Although..." would do a better job of seting the context than a rambling, negative-filled sentence (repeating something we've already said above) followed by "However..."?--Kotniski (talk) 21:13, 16 November 2011 (UTC)[reply]
As above, I'm quite happy to see the wordiness reduced; I've no particular preference between words such as "although" or "however"; minor copyediting changes don't need to be discussed. Uniplex (talk) 21:24, 16 November 2011 (UTC)[reply]
So if we take the proposed version, replace "When doing this, however.." with the original "When reverting an edit you disagree with..", and remove the reference to "politeness" (though I actually think that's quite a significant factor), would you (and others) be happy with it?--Kotniski (talk) 07:41, 17 November 2011 (UTC)[reply]
This explanation is complicated. Why don't you show us what you mean? I don't mind "polite". Make your edit and leave it to WhatamIdoing to further improve as required. --SmokeyJoe (talk) 09:51, 17 November 2011 (UTC)[reply]
Where does consensus begin and the rest of the editorial process leave off? Perhaps upon examination we enter consensus and its necessities always from a finite number of directions. Limited cases, limited reasons, narrow requirements. Where does consensus end and the rest of the process begin? Perhaps we return only with a new consensus or the failure to achieve consensus. Since this article is about consensus, we should discipline ourselves to stay on that subject only. --Ring Cinema (talk) 16:15, 23 November 2011 (UTC)[reply]
Goal
There's a line I like at User_talk:Jimbo_Wales/Archive_88#NOTCENSORED_and_illustrations that says "the proper goal of community consensus should be (generally) to reflect what is in reliable sources." This is specifically about article content (rather than advice pages or behavioral issues, and therefore would have to be labeled as such), but perhaps something like that would be a helpful addition to the lead. WhatamIdoing (talk) 18:41, 23 November 2011 (UTC)[reply]
Yes, it's great for content. However, the lead should be a precis of the body. Whether we ever get as far as reviewing the rest of the body, to update it to reflect current best practice, and further highlight/clarify what we all observe as horribly-bad practice, remains to be seen. If we ever do, per RC's section above, we should be careful to keep separate the essential characteristics of consensus from that of content policy. Uniplex (talk) 09:06, 24 November 2011 (UTC)[reply]
Jimbo's statement at the top of this talk page is also rather good. Someone asked recently if it is acceptable to begin a lead with a quote, to which the response was a resounding "no"; that was in relation to articles though. Uniplex (talk) 09:15, 24 November 2011 (UTC)[reply]
On first impression, I don't think it is significant and central to the policy of consensus. It is not really about consensus, but about the project goal. Perhaps add it to WP:Goal. --SmokeyJoe (talk) 09:16, 25 November 2011 (UTC)[reply]
I agree with Joe, it's a project goal: a goal of Wikipedia should be (generally) to reflect what is in reliable sources. Though I think WP:5P may the place for it, in the second pillar, which skirts round this sentiment. Uniplex (talk) 22:36, 26 November 2011 (UTC)[reply]