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. Before editing this page, please make sure that your revision reflects consensus. When in doubt, discuss on the talk page. Always remember to keep cool when editing. Changes to this page do not immediately change policy anyway, so don't panic.

This talk page is automatically archived by MiszaBot II. Any sections older than 30 days are automatically archived. An archive index is available here.

Archives

Wikipedia talk:Consensus/Archive 1 Wikipedia talk:Consensus/Archive 2 Wikipedia talk:Consensus/Archive 3 Wikipedia talk:Consensus/Archive 4 Wikipedia talk:Consensus/Archive 5 Wikipedia talk:Consensus/Archive 6

"Consensus is a partnership between interested parties working positively for a common goal." -- Jimmy Wales

suggested rewrite for lead

I changed the lead to read as follows (got reverted, which is cool, so let's take it up here). explanations and reasons given below:

Consensus on wikipedia means that all parties concerned with a particular article agree to the result of an edit or series of edits, or to the content of the article as a whole. It is an idealistic concept which is often implicit and imperfect, but is central to the proper editing of articles, particularly with respect to achieving neutral point of view. Typically, consensus is reached as a natural and inherent product of the wiki-editing process: one editor makes a change or addition to a page, and then other editors will revise the edit, offer objections or make requests based on other wikipedia policies, engage in talk page discussion to iron out problems, or accept the change silently. More pervasive and more contested edits should generally seek broader consensus through RfC and other consensus-generating processes; This ensure there is adequate exposure to the community. In particular, policy and process pages need a higher standard of participation and consensus than other pages.

Consensus can be a confusing term, so care should be taken with its use. Editors often use it in a loose way to talk about the current accepted version of an article, but in general if any editor objects to an edit or to article content - regardless of how unreasonable that objection might seem - consensus does not exist. Consensus among a limited group of editors, at a single place and time, cannot be confused with consensus in its ideal sense, and cannot over-ride community consensus on a wider scale. It may be necessary on occasion to administratively override the objections of given editors for any number of reasons, usually involving disputes over the inclusion or exclusion of questionable content in the encyclopedia. When this happens, it should not be seen as achieving or restoring consensus, but rather as an unfortunate administrative act that was necessary to cope with a failure of consensus.

reasons:

1. this version actually gives a definition of consensus in its ideal form, which seems useful for an article on consensus
2. it also focusses on consensus as something that has to be worked on continuously, rather than something that can be achieved (which will help resolve a lot of paradoxical disputes where editors try to assert some 'consensus' over the objections of other editors)
3. it distinguishes (at the end) between a consensus on an article, and a version of the article that was imposed administratively (which is not a consensus, though it is occasionally necessary).

basically, I rewrote this to get across the idea that consensus is a continuous process which involves finding agreement with other editors, and that the preferred mode on wikipedia is to go out of our way to achieve consensus. I think some of these revisions need to be made (particularly something along the lines of point 1). suggestions or comments? --Ludwigs2 23:43, 19 August 2008 (UTC)

Ludwig, you've been active for about 3 months at WP. Welcome and thanks for joining the team! But maybe you should hang-out for a while and see how the processes work before trying to modify the core policies of the project. --Kevin Murray (talk) 00:30, 20 August 2008 (UTC)

Don't mind Kevin, he feels that newbies who get involved in policy discussions deserve to be bitten. I agree that it would be good to have a definition of consensus, instead of having the page dive straight into suggestions of how to achieve consensus. The problem is that on Wikipedia, in practice, consensus has extraordinary variability; it's neither unanimity nor supermajority nor even necessarily agreement. Your language of "all parties agree" is, in this regard, incorrect. One stubborn person cannot, in essence, hold a page hostage until his or her views are represented -- if Wikipedia did function this way, it would not function at all.--Father Goose (talk) 01:25, 20 August 2008 (UTC)

FG, that's a bit harsh. Newbies are welcome to join discussions, but should not be trying to boldly rewrite policy pages. --Kevin Murray (talk) 04:43, 20 August 2008 (UTC)

Sure they should try. Then they can find out how things really work! What people really need to understand is that consensus on a wiki is different from the common language meaning, or the idealized meaning used in other formal consensus decision making processes. It is less determined by our ideals than by the reality of what happens when you put a wiki on the web and invite all comers to participate. -- ☑ SamuelWantman 05:13, 20 August 2008 (UTC)

Kevin - I understand the sentiment: there's a value in any community to having a history. it establishes trust and familiarity, among other things, and so I can't really fault you for your suggestion. however, I don't see that as a reason for me not to try revisions that I think are useful. at worst I'll learn something (which happens all the time); at best, I'll make some sensible suggestions. I happen to know a lot about this particular topic, and I'm not an over-enthusiastic kid or some ham-handed goon (which I know are recurrent problems on Wikipedia). In short, I acknowledge your concerns, but I'd ask you to look at what I'm suggesting on its own terms, rather than prejudging it because you think I'm too new to say anything smart.

Father G, Sam - 'consensus' only has one meaning, and that is an idealism that is almost never achieved in the real world (well, outside of small groups: you can get a perfect consensus in any group that's small and intimate enough for face-to-face communication). most collective decision processes are designed to give an effective approximation to consensus while offsetting the various problems that are inherent to large groups (factionalization, miscommunication, troublesome individuals, etc...). consensus on a wiki is no different; wikis just create their own unique set of issues and constraints. what I had thought I was emphasizing in this edit (and maybe I missed the mark) was that people in a consensus system have to keep their eye on the ideal of consensus, even when they can't quite achieve it. if they don't, then the system breaks down into factional disagreements (i.e. competition between the beliefs of groups that have little or no interest in coming to an agreement with each other), and consensus becomes nothing more than a buzzword that each group uses to explain why the other groups are all wrong. see what I mean, and/or how can I put that better? --Ludwigs2 07:43, 20 August 2008 (UTC)

Well, you have to recognize that Wikipedia uses something more akin to rough consensus than the unanimity that you are suggesting (which, mind you, is only one possible definition of consensus. Further, Wikipedia sometimes uses words in different ways than they are used in the "real" world (cf. Wikipedia:Notability).

I agree with you that "consensus" is claimed on Wikipedia about as often as "God is on our side" is claimed off-wiki. The only way to combat such bullshittery is to get more people involved, who hopefully agree with you. (Attracting more participants to an issue that is of importance to you, however, is often difficult.)

You're also right that as broad a consensus as possible should be sought in preference to majority rule. This is what's wrong with Kevin's response above, in wanting to automatically exclude you from discussions such as these, when your ideas are potentially sound if not yet fully informed (and none of us can claim to be fully informed). Such an approach takes patience and even benevolence on the part of the people involved, however. Those are rare qualities, and in finite supply even for those who do have them. Then again, Wikipedia is host to more fair-minded people than I've met in any other sphere, so, I am generally encouraged. It's worth it to urge such benevolence on Wikipedia's editors, but unrealistic to insist on it. With that in mind, what changes would you continue to suggest for this policy's wording?--Father Goose (talk) 09:37, 20 August 2008 (UTC)

• Part of the problem is that words develop into euphemisms at WP and then these become the titles for our policies and guidelines under the newly evolved definition. While Ludwigs may be more accurate in describing consensus in the real world, the definition here has evolved after much hard work and "consensus." There is a fine line between improvements and constantly reinventing the wheel in such a way as to further destabilize an inherently unstable project. Our goal should not be a utopia of personal growth, but producing the best encyclopedic source for our readers. --Kevin Murray (talk) 14:02, 20 August 2008 (UTC)

An improved Wikipedia:Consensus would further that goal.--Father Goose (talk) 17:27, 20 August 2008 (UTC)

(re-dent) Father G.: I'm not disagreeing that wikipedia usually uses 'rough' consensus rather that 'true' consensus. that's normal and expected: a true consensus would be impossible here, like most places in the world, for practical reasons. all I'm trying to suggest is that this policy should remind editors that even as we use 'rough' consensus procedures, we should all keep in mind that what we're aiming for is 'true' consensus (even though that's impossible). otherwise it's too easy to get confused about what consensus is or means.

pardon the silly example, but this is kind of like like the fifth commandment: "thou shalt not kill" taken literally is impossible (we'd starve to death without killing, unless we got mana from heaven or learned to photosynthesize). But it's still a darned good orienting principle that we should aspire to.

with that in mind, I think the policy needs to give

• some description of the ideal form of consensus
• an acknowledgement that the ideal form is not practical, but should still be aspired to
• (eventually) a description of how consensus processes actually work in wikipedia (to the extent that's possible), and how they relate to or approximate the ideal form.

the goal would be to give a reader an understanding of the high-minded goal, and some explanations of how to steer actual discussion towards that goal as much as possible. I thought I'd given a decent stab at the first too points, but judging by reactions, apparently not. can you tell me where I've missed?

Kevin: I don't think I'm trying to reinvent the wheel; I'm not actually sure this particular wheel has been invented yet. the page is not well-developed. however, I can certainly see not wanting to disturb what is here too much. I trust, though, that we can keep that from happening. --Ludwigs2 17:54, 20 August 2008 (UTC)

Sincerely, welcome to WP. Nothing is well-developed here, because everything is a compromise. "Consensu" seems to work OK at the articles, but it is a joke at the process pages (process pages are policy and guidelines). Have you heard that a camel is a horse designed by committee? Well WP has more humps than a fancy house during Fleet Week. Part of my sensitivity on this issue is that we just spent about three months rewritting this monster. Sometimes stability is more important than perfection. Cheers! --Kevin Murray (talk) 18:35, 20 August 2008 (UTC)

You missed the second point, by emphasizing the "ideal" almost to the exclusion of the practical (and actually practiced). But describing what is "actually practiced" in matters of consensus is deceptively difficult; "what usually happens" is not necessarily "what should happen", and even "what should happen" is not always what should happen. And the more specific the guidance on any given policy page, the more likely it is to be wikilawyered. As a result, sometimes you end up with pages like Wikipedia:Consensus, which are a bit gutted, but also don't screw the wiki up too badly by saying the wrong thing.

I invite you to keep trying to improve it, though I don't intend to apply a whole lot of attention to it myself; it's too big a thing for me to wade into right now.--Father Goose (talk) 03:44, 21 August 2008 (UTC)

Kevin: Haven't we seen how this works? Now you're the teacher, and Ludwigs is the student, I'd say. Ironically perhaps, the teacher often learns more than the student. I know it's hard to let people edit your hard work of many months (or even years in my case), but if I can survive it, do you think you can too? :-) --Kim Bruning (talk) 17:23, 22 August 2008 (UTC) I wouldn't say consensus is a joke at process pages. I thought my demonstration to you was really effective; to wit, you wouldn't be here to say such things if they were really true O:-)

ok - taking Father G.'s comments into consideration, how about a revision like this?

Idealistically, consensus on Wikipedia means that everyone concerned with a particular article agrees to the content of that article. While it is pragmatically impossible to get explicit and overt agreement from everyone, consensus is expected as the natural and inherent product of the wiki-editing process: one editor makes a change or addition to a page, and then other editors will revise the edit, offer objections or make requests based on other wikipedia policies, engage in talk page discussion to iron out problems, until (eventually) all accept a version of the article (often silently). Consensus can be a confusing term, so care should be taken with its use. For instance, editors often use consensus in a loose way to talk about the current accepted version of an article, but this kind of status quo should not be confused with consensus in its ideal sense. Consensus can change as new editors and new information enter into the editing process, and a consensus shared among a limited group of editors, at a single place and time, cannot over-ride community consensus on a wider scale.

Wikipedia has various processes to help build consensus and resolve disputes that block consensus. Third Opinions Requests for Comment can attract uninvolved editors to give comments and help resolve differences in an article; Wikiquette alerts can help resolve personal problems or arguments that sometimes occur between editors; if necessary administrator assistance can be requested through various processes (see Resolving_disputes). Except in the most extreme cases, though, these processes serve only as public, neutral grounds to help build a consensus that failed in article editing; it is always best to try to build consensus first through discussion.

that second paragraph is what I meant by "actually practiced" - just a list of the various processes wikipedia has available, and how they are designed to support consensus rather than override it. does this version work better? --Ludwigs2 21:55, 22 August 2008 (UTC)

well, I haven't heard back in a few days, so I'll try editing this in and see what response it gets. --Ludwigs2 21:27, 30 August 2008 (UTC)

This material could be more appropriate for Silence and consensus perhaps. /NewbyG (talk) 21:53, 30 August 2008 (UTC)

the whole thing, or just particular bits? I guess I'm not seeing what you mean. aside from the inclusion of a definition of consensus (and the second paragraph about consensus-oriented processes), most of this edit is just a rewrite of what was there before. I think it's an improvement, of course... what is it that you don't like? --Ludwigs2 20:39, 31 August 2008 (UTC)

<sigh...> ok, I'm not getting any feedback here, so (maybe tomorrow) I'll try editing in a proper definition of consensus again, and see if that flies. I'll keep it minimal - maybe that will help. --Ludwigs2 00:57, 11 September 2008 (UTC)

Consider this to be yet another objection to your efforts here. If anything, this policy wants to explicitly disclaim the idea that consensus requires unanimity. We frequently reach a consensus among all but one editor, but six or eight against one, especially when the proposals of the one are widely perceived as breaking major policies like V or NPOV, is still widely considered to be a consensus. WhatamIdoing (talk) 22:03, 12 September 2008 (UTC)

again, the inability to distinguish between unanimity as an ideal (which is highly desirable) and unanimity as a criterion (which is impossible) doesn't do a lot of credit to the argument you're making. it's akin to saying that since 'true love' is an impossible ideal, we should treat all our personal relationships like financial transactions. not a pretty picture... --Ludwigs2 00:30, 13 September 2008 (UTC)

definition of consensus

Newbyguesses - I'm curious why you removed that first couple of sentences that I edited in - namely: - "Idealistically, consensus on Wikipedia means that everyone concerned with a particular article agrees to the content of that article. While it is pragmatically impossible to get explicit and overt agreement from everyone, consensus is expected as the natural and inherent product of the wiki-editing process...". without arguing about the specific wording (which maybe could be better) I put that there because I feel that this policy needs some effective definition of what we mean by 'consensus'. without that, we leave a huge, gaping hole where any editor can pretty much assume that consensus means whatever it is he wants it to mean, and the policy becomes next to useless as a guiding tool. If you don't like what I wrote, that's cool, but please offer a suggestion as to how it can be revised so that we have an effective definition of consensus. --Ludwigs2 08:47, 1 September 2008 (UTC)

This is because consensus can not ever be defined by "overt agreement from everyone", that number being equal to the potential number of people with internet access, and therefore increasing (changing) from second to second. Consensus can never be defined by participants participants in the process. Consensus can only be defined by the basis on which it is achieved, known variously as evidence, facts, sources or some tangible, agreed upon constant--mrg3105 (comms) ♠♥♦♣ 14:11, 1 September 2008 (UTC)

This text WikiEN-l mailing list July 2005 used to be in WP:PRACTICAL on the project page. Now it's in the See also section. Cheers, /NewbyG (talk) 23:37, 1 September 2008 (UTC)

mrg3105: that is a misunderstanding of the nature of consensus. "consensus can not ever be achieved by "overt agreement from everyone", for the very reasons you stated, but there is no other way to define it. Consensus is an Ideal type, and should always be treated as such, otherwise you're not really talking about consensus at all. I think it's essential that we discuss this ideal type before we get into the details of how it is 'typically reached'. without that kind of definition, this policy is effectively neutered. --Ludwigs2 00:20, 2 September 2008 (UTC)

“Consensus” on wikipedia, as opposed to as in any dictionary publication, is a goal or an ideal (as per Ludwigs). It defies robust definition. We know it when we see it, but it looks different every time. If we could say with precision what consensus is, we could save the hassle of debate and negotiation and just go there. The next best thing is to describe the process that leads to consensus. There’s a history of difficulty here. Perhaps a picture would help? Do_you_want_me_to_draw_a_picture_or_something. --SmokeyJoe (talk) 02:05, 2 September 2008 (UTC)

Thanks SmokeyJoe, but that picture is wrong. In disagreement with Ludwigs2's contention that consensus defies robust definition, because it seems to me that the Wikipedia approach is more akin to the scientific method where its about testing theories (wikispeaky - "original research") against available data.

What is required to define consensus, is to rename the "Start" in the picture to "Author". The first "edit" is an authored article, with no preceding edits. What the author initially offers to Wikipedia is a theory. If the offering is completely lacking in cited references it is, despite the often obvious nature of the contents, original research until that data is substantiated from third party verifiable sources. How this is done, and what degree of agreement editors, the group that subsequently tries to turn theory into fact for, and even impart understanding of the subject to the reader, is consensus.

Another helpful definition is that of an "edit". Often used, in fact it does not mean change, but "a bringing forth" or "producing", and involves value adding. Opinion does not add value, data does. When editing is performed, that is not the grammatical or style editing, but content editing, effort is made to attain a better and clearer knowledge and understanding of the subject, and not just adding of data. Sometimes clarity can be achieved by removing data from the content.

Simply stating opinion or belief as to the truth or falsity of the theory, regardless of how many agree, is called theorising, not editing. Its right up there with the medieval belief in flat Earth.

Providing sources is called evidence. Verifiable sources are acceptable evidence.

Adding citations (preferably with page numbers) to the theory is called evidence-based substantiation of a theory.

When the group agrees that the evidence is correct in the way and degree of substantiation it offers, it is called proof.

When all possible implications of a theory have been addressed by offering proof, the theory becomes fact.

Consensus is therefore a process motivated by quality improvement by a group, which, through aggregation and summary, transitions unsubstantiated, or partly substantiated theory or opinion presented in the article into fact through use of perspective(s) and contextualised evidence derived from relevant and verifiable expert sources on the subject in the effort to verify all statements made in the article that may be questioned if they lack proof of being true or false.--mrg3105 (comms) ♠♥♦♣ 05:00, 2 September 2008 (UTC)

Which picture is wrong? The current one? There have been many pictures. I dispute that is it required (by who?) to define "consensus". Considering your offered definition: Do you think it likely that if you found a newcomer (the intended reader of this page), and told them this definition, that it would lead to improved contibutions and ultimately a superior product? On the other hand, could that defination do damage, when a newcomer reads that wikipedia works by consensus, follows the link here, reads esoteric verbiage, and gives up? --SmokeyJoe (talk) 09:59, 2 September 2008 (UTC)

mrg3105: actually, I don't disagree with what you said here (though the way you said it made my head spin - lol). the process of achieving consensus is very much akin to the scientific method. but you neglect to consider that the goal of the scientific method is create theories that are universally accepted, not to create theories that are true. no amount of evidence in the world will prove that a theory is true, but sufficient evidence will convince everyone that they should use it as though it were true. if scientists don't pay attention to that last point (that their goal is to create consensus among everyone that the theory works) then more often than not they end up in the fringe theory category, preaching their 'science' to a bunch of true believers. likewise, if wikipedians don't keep focussed on the point that they are trying to create consensus among everyone that an article correctly portrays a topic, they will (more often than not) end up trying to push what they think everyone ought to believe is true about the topic, and that will just disintegrate into POV wars.

SmokeyJoe: I don't think it's difficult at all to say what consensus is. consensus is a state where everyone agrees that the article correctly portrays the topic. it's impossible to reach, of course, but it's easy to state. all I'm arguing for here is that we state it outright, and then get down to discussing the processes that move us in that direction. --Ludwigs2 22:40, 2 September 2008 (UTC)

I make no apologies for the rather complex definition of a rather complex concept. Although the concept of consensus has existed for some years, we seem to still be trying to formulate a definition, so five lines is an improvement on nothing :)

Given that we live in the world of miniaturisation, the definition is only likely to become smaller, but I did try to include all necessary ingredients.

I disagree that the goal of the scientific method is to create theories that are universally accepted. By virtue of being a theory they are never universally accepted, even if they are universally believed to be true. By definition, a theory that is true, is - a fact!

Evidential sufficiency is only applied to render a theory conditionally true, that is it is true within constraints of certain assumptions, conditions or conceptual environment. It still remains untrue if universally applied. Hence applied science, that seeks to identify all possible such conditions through experimentation. Consensus there, is also not required because it is also evidence-base, i.e. I believe it when I see it.

Category:Fringe theories is sadly lacking :), but surely would qualify for not applying the original research policy to its articles. However, a good many scientific facts have evolved from fringe theories, and have been turned into applications that seem to work just fine with no one laughing. "What, you want to light a fire below the decks of a ship?" - Napoleon responding to the first suggestion of steam-powered vessels.

"consensus is a state where everyone agrees that the article correctly portrays the topic" has the problem of not specifying why, how and within what criteria, so maybe between your one-liner and my five we can arrive at something less scary and more comprehensive?--mrg3105 (comms) ♠♥♦♣ 01:15, 3 September 2008 (UTC)

lol - oh please! theories should never be elevated to the throne of fact (to do that is to violate the very thing that makes science science, and render the whole enterprise futile). and my one-liner is not intended to specify how, why or what - it's only purpose is to set the lofty, unreachable state of true consensus in its proper place as the proper goal, so that when we go on to specify how, why, and what, we know precisely what they are aiming at. this is good ol' Wittgenstein, my man: you cannot make any measurements at all until you first point out what it is you're measuring against. --Ludwigs2 22:20, 3 September 2008 (UTC)

Theories are elevated to the status of fact all the time.

If you don't intend something, don;t say it

Its hard to measure Wikipedia consensus because there is only one other example that I know of where a group of complete strangers with no means of evaluating their capacity for making true statements are called upon to provide a decision on a question. The one example is an abject failure at consensus, born of a society that has been in an economic recession since the 1980s, and is quite literally dying despite being one of the most technologically advanced in the World, so comparison is not easy...with Japan--mrg3105 (comms) ♠♥♦♣ 12:06, 4 September 2008 (UTC)

mrg3105 - you are simply being argumentative now. if you simply want to refuse to make the distinction between ideals and practices, there's nothing I can do about that (except roll my eyes), but studied ignorance is no replacement for an actual argument. point by point:

• "Theories are elevated to the status of fact all the time." true, and every time they are it's a mistake. that's not the way science works.
• "If you don't intend something, don't say it." read the entire sentence, because my intention is clearly spelled out in the latter half.
• "Its hard to measure Wikipedia consensus because..." of some left-field analogy to post WWII Japan? hunh?

If you have a real point to make, make it, otherwise please stop trying to oppose the proper development of this page. --Ludwigs2 01:48, 7 September 2008 (UTC)

Ludwigs, you seem to be trying to write the article Consensus instead of documenting the actual practice as it happens in writing Wikipedia. This project page exists solely for documenting and guiding the actual practice. In actual practice, unanimity is not required. The definition of Wikipedia-consensus, as opposed to some other kind of consensus, therefore does not require mentioning either "overt agreement" or "everyone". WhatamIdoing (talk) 22:08, 12 September 2008 (UTC)

WhatamIdoing: you seem to be misconstruing the nature of a policy. Policies are not there to 'document actual practice', but rather to 'offer guides to best practice'. 'best practice' is itself an idealistic term (since it is assumed that not all editors will innately know what best practice is), and on this particular topic (consensus) 'best practice' is meaningless without some discussion of the ideal towards which consensus reaches. there is plenty of space and opportunity in the policy to discuss your' real world application', but we need to determine what we are making 'real world applications' of before we do anything else.

I am really shocked by the amount of resistance I am getting towards adding 'ideally, consensus involves the agreement of all concerned parties'. what is it about that phrase that you object to? it would be nice if you'd spell it out, point by point, so I can disabuse you of whatever worries you have here. --Ludwigs2 23:32, 12 September 2008 (UTC)

Actually Ludwigs2, policy is the attempt to prevent repetition of past errors based on history of previous practice.

Policy can never approach the ideal state because of the factor of change which is endemic to the functioning of Wikipedia.

In order to formulate policy on Consensus in Wikipedia, one has to firstly define

• 1 Its purpose (as applied)
  • The ideal purpose
• 2 Its process (as applied)
  • The ideal process
• 3 Policy constraints
• 4 Foreseeable change factors

As I and many others see it on daily basis, and the example of the "Soviet invasion of Manchuria" is but one, is that consensus is misused as a form of vote. Therefore its purpose as applied now is to force change despite article quality policies. Therefore it its process is not about what consensus establishes, but how a group of co-believers incorporates their opinions into Wikipedia articles. There are many such examples, and in fact they dominate Wikipedia articles and consensus formation daily. Just watch it happen if I start nominating unsourced articles for deletion, or deleting any block of text or section in articles which are unreferenced. Most editors don't even have a clue about how to reference, which is not suggesting an entire book as a substantiation of contributed text.

Now quite frankly I don't know you from Joe Blogs, and can only form an opinion based on your communication, as you do on mine. What you don't seem to realise is that anything that is not referenced, i.e. anchored in some recognised and authoritative published source, anything you add to the Consensus policy is likely to be completely revised sooner or later, with the emphasis on sooner. This is at once a constraint and a factor in writing Wikipedia policy. Most Wikipedia policy is not worth the code it takes to display it because it is not authoritative. It is worthless as policy because it changes. Policy is meant to prevent dynamic reflections of policy writers, and help those following it, the executors - that's ME and the vast bulk of other editors who should not have to look over their shoulders as we author and edit articles for an appearance of a group of administrators who can impose their will regardless of any policy because they can. When this happens, policy is subsumed by anarchy...--mrg3105 (comms) ♠♥♦♣ 02:20, 13 September 2008 (UTC)

Ludwigs, your idea about how Wikipedia policies happen is entirely wrong, and is directly contradicted by many long-standing pages. This is not a top-down bureaucracy. Policies document actual practice. See, for example, what Wikipedia:Policy#Sources_of_Wikipedia_policy says:

Policy change comes from three sources:

1. Documenting actual good practices and seeking consensus that the documentation truly reflects them.
2. Proposing a change in practice and seeking consensus for implementation of that change.
3. Declarations from Jimmy Wales, the Board, or the Developers, particularly for copyright, legal issues, or server load.

In practice the first option is the most effective...

See also Wikipedia:PPP#Policy: It is often preferable for policies and guidelines to spring up organically through codifying existing practice, rather than to be imposed from the top... and Frequently, we simply write down what already happens. Anything that describes the usual outcome of a common process, is a good guideline for the future...

I've noticed over the last few weeks that several different people on several policy and guidelines pages have politely suggested to you that you get a bit more editing experience before trying to fiddle with Wikipedia's policy pages. Perhaps you will consider their suggestions. WhatamIdoing (talk) 05:51, 13 September 2008 (UTC)

• I like the definition consensus is defined as a lack of active opposition. I think this is wiktionary's definition. --Surturz (talk) 17:06, 14 September 2008 (UTC)

Do you dig: "the logic may outweigh the logic"?

From Wikipedia:Consensus#Participating in community discussions: On 17:57, 7 September 2008, I changed seemingly confusing: "Minority opinions typically reflect genuine concerns, and the logic may outweigh the logic of the majority." to more explicit: "Minority opinions typically reflect genuine concerns, and their strict logic, if present, outweighs a point of view of the majority.", because such seemed to be intention of the sentence. On 20:40, 8 September 2008, User:Jfdwolff reversed it. Do you agree with the reversal, please? 162.84.184.38 (talk) 04:15, 9 September 2008 (UTC)

Or, in other words, is it true, that "the logic may outweigh the logic", if nothing CANNOT outweigh itself, please? 162.84.184.38 (talk) 19:41, 9 September 2008 (UTC)

Maybe we can we lose "outweigh" altogether? How about Minority opinions typically reflect genuine concerns. Convincing arguments typically sway community consensus. Or something like that. No need to emphasize a dichotomy with "majority" nor a metaphor with "outweigh". /NewbyG (talk) 01:02, 11 September 2008 (UTC)

Flaw in the flowchart

The flowchart shows a pathway: "Was the article edited further" → No → "New consensus". Actually, very often, the fact that no new edits were made indicates only that people are tired of arguing and/or unwilling to edit-war, but not that consensus has been reached. There are plenty of pages here that have reached a kind of stable equilibrium that nevertheless has no consensus. —Angr 21:51, 12 September 2008 (UTC)

This is not so much a flaw in the flowchart as reflection of a real weakness of wiki editing. Can you offer a suggestion for a solution? The best I've seen is to point people to Wikipedia:Silence and consensus. If you don't oppose it, you haven't oppsed it. --SmokeyJoe (talk) 06:21, 13 September 2008 (UTC)

That's a slightly different issue, though. I don't mean you've never opposed it, I mean you have opposed it over and over again, but the people who disagree with you have also opposed you over and over again, and eventually you just give up trying. Once you give up trying, the page reaches stability, but it would be a huge mistake to treat the stable version of the page as having even temporary consensus. —Angr 09:13, 13 September 2008 (UTC)

I think you're misreading the chart. If you're talking about a change someone else has made then you should look at the branch "Was the article edited further?" → Yes. You're supposed to read the chart in respect to your own changes, not to third parties' changes. --Gutza ^{T T+} 21:49, 13 September 2008 (UTC)

P.S. In that case, you'd probably end up at "Seek a compromise", after deciding you don't agree with the edit. The fact that you voluntarily choose not to seek a compromise (for the zillionth time) is IMO outside the scope of this page. (Not that I don't agree with you, I've been there as well.) --Gutza ^{T T+} 21:52, 13 September 2008 (UTC)

Angr, is this a situation where you and others have debated something to death without reaching a conclusion? In that case, the flowchart is correct in that the process can go on forever and that eventually reaching a consensus is not guaranteed.

Alternatively, is this a case where some editors are routinely defeated in debate, but they (or others?) return again and again to revisit the same question. In this case, either the page does not properly document a common perspective, or you are dealing with a troll, and I would suggested using the method of slow reversion. --SmokeyJoe (talk) 11:30, 14 September 2008 (UTC)

I'm thinking mostly of the former, though the latter could arise too even without anyone trolling. The point is that sooner or later, the page doesn't get edited anymore (or not with respect to the contentious material), but the mere fact that the page is stable with respect to that issue does not actually mean that consensus has been reached. The chart implies that lack of further editing implies that consensus has been reached, but in fact that is not always the case. Sometimes the lack of further editing merely implies that side A has given up trying. No compromise, no consensus, just an exhausted retreat from the battlefield, leaving side B in charge of the page. Side B's version of the page stays untouched for months and months, but nevertheless it doesn't have consensus. —Angr 12:07, 14 September 2008 (UTC)

I think I know what you mean. Something that makes it worse is the auto-archiving of talk pages. Perhaps some "disputed" tag would accurately record the situation? --SmokeyJoe (talk) 13:08, 14 September 2008 (UTC)

Yes, though sooner or later those get removed too when no discussion of the dispute is still happening, because it's all been discussed to death already. —Angr 13:16, 14 September 2008 (UTC)

Odd. It's always possible to reach *some* consensus, if only to represent all sources fairly and let the reader decide (this works equally well for articles as for policy). There will always be an NPOV position that can be taken (so far I have never seen the NPOV concept actually outright fail. If you show me an example to the contrary, I'd be very interested indeed!) . People often get hung up on false dichotomies or on their own arguments. If this happens to you, you can try several things, like try a different perspective on for size, or get a little help with the whole consensus shebang! :-) Remember, you're not here to fight, wikipedia is not a zero sum game, always try to seek a way for everyone to win, or surely all will lose. --Kim Bruning (talk) 22:34, 15 September 2008 (UTC)

Well, not everything is an issue of NPOV, and sometimes dichotomies aren't false. One example I'm thinking of is the whole "ß vs. ss" issue in articles like Wilhelmstraße, Franz Josef Strauß, and several dozen others. There are only two ways to spell it, Wilhelmstraße and Wilhelmstrasse, with no possibility of compromise between them. (Well, technically, writing it in all caps, WILHELMSTRASSE, would be a compromise, but that would not be an acceptable article name either.) So the article gets written with one spelling, and then there's an RM to move it to the other spelling, and sometimes the RM succeeds because during that particular week enough people from one side show up to make the "vote" 2 to 1 or more, and sometimes it fails because the "vote" is less than 2 to 1, and sometimes the "vote" is less than 2 to 1 but the article gets moved anyway despite the absence of consensus. And from time to time someone points out that this decision shouldn't be made on a case-by-case basis, as that results in inconsistency, but rather there should be a decision made at the policy/guideline level that will apply to all articles. But since the two sides are about evenly split, no consensus can or will ever be reached. Nevertheless, at some point, people give up arguing about it and the articles settle down and are stable at one spelling or the other, giving the illusion of consensus where in fact none exists. —Angr 07:14, 16 September 2008 (UTC)

In the case of requested moves, no one tried to find consensus in the first place. RM is simply broken. Note that RM has no policy behind it, and half the time doesn't even follow any discernible polling guideline. If you spot a requested move happening on a page, I've found that it's possible to simply remove it (at least early on when someone has just posted it) , and people won't protest too loudly. You can then proceed to actually find consensus. :-)

Removing an rm later is a bit iffy, since it may contain many people's opinions. You can still remove the poll header and footer, and start a discussion on the basis of opinions. Finally, you can always simply follow the normal structured discussion methods for polls, if by some miracle the poll is actually properly structured for once. O:-)

Even splits are pretty normal. This is where you start to actually negotiate with people. In the case of page names it's all just ego-wankery though, because we have a perfectly serviceable redirect system. Don't bother with those at all, and/or kill them on sight. (Of course having multiple 'equally-good' names analogous to hard links would kill that entire class of debate all at once, but no-one has gotten around to writing that ... yet ).

So the example you came up with is a system that actually doesn't do consensus so well (or at all), and you can just wipe it if it gets in your way (because -as a rule of thumb- consensus typically wins out over other systems). The reason this system still exists is because some people enjoy feeling important. But now you know what to do: Either just let those people have their fun and waste their time; or if you need stuff done right now and your time is precious to you, kill the process. People will accept that, if you explain what you are doing clearly and carefully. Do you have more examples? --Kim Bruning (talk) 09:07, 16 September 2008 (UTC)

An example from policy: the notorious WP:NFCC#8. Much drama over whether it should say "Non-free content is used only if its presence would significantly increase readers' understanding of the topic, and its omission would be detrimental to that understanding" or simply "Non-free content is used only if its presence would significantly increase readers' understanding of the topic". Lots and lots and lots of discussion with no consensus. Finally, this edit with a laughable edit summary. Since then, the page has stayed stable with respect to the contentious clause, giving the illusion of consensus where none exists. But in my first example, I wasn't asking what to do in such cases; I was merely pointing out that page stability is not actually an indicator of consensus. —Angr 09:23, 16 September 2008 (UTC)

That particular sentence is a redundant partial restatement of NFCC#1 O:-). There doesn't appear to be a current discussion. As the discussion has moved to archive, you could place the sentence back in, and see what happens as per BRD (thus reviving the discussion), but you may feel that you already have a list of people to discuss with. Find the strongest opponent and figure out their opinion in detail (if you don't know it yet). If you think it conflicts with the foundation issues, explain your reasoning to them. Find out whether they object to the foundation issues (which states simply that all content must be GFDL or CC-BY... any use of non-free-images already contradicts that :-P but exceptions prove the rule and all that.), or if they merely object to your reasoning. O:-) --Kim Bruning (talk) 01:45, 18 September 2008 (UTC)

You're still missing the point. I'm not asking what do to resolve this, I'm just pointing out that the stability of the page over the last few months creates a misleading illusion of consensus. As for the phrasing of the NFC criterion itself, I'm past caring. At some level, the contested phrase is just a restatement of the first part of the sentence anyway: the original phrasing amounted to "If A then B, and if ¬B then ¬A", and now it just says "If A then B". —Angr 05:57, 18 September 2008 (UTC)

Well, you can try to change it any time, right? But will you succeed? --Kim Bruning (talk) 15:24, 23 September 2008 (UTC)

RfC: definition of consensus

A user has requested comment on Wikipedia policy or guidelines for this section. Within one hour this page will be placed on the {{RFCpolicy list}}. When discussion has ended, remove this tag and it will be removed from the list.

An apparently irresolvable debate over whether to include a definitional statement such as: "Idealistically, consensus means that all concerned parties agree to the substance of an article or edit, though in practice this is rarely feasible. Debate seems to revolve around the following points:

• Pro: Policy should reflect 'best practices', and no discussion of the 'best practices' in consensus is meaningful without some recognition of 'true' or 'ideal' consensus
• Con: Policy should reflect 'actual practices', and since ideal consensus is never achieved or achievable, it should not be mentioned or encouraged.

I am on the Pro side, and may have misconstrued the Con side - please edit the above to give a better (single-line) synopsis, if necessary. --Ludwigs2 00:16, 13 September 2008 (UTC)

• Con. In reliable sources that talk about human decision-making, "consensus" is usually opposed to "unanimity", in the sense that "unanimity" requires agreement by all concerned, whereas "consensus" is less than that. Here's a sample quote and citation:

Whereas previous treaty revisions had a unanimity requirement, the convention, as we demonstrate, relied on "qualified consensus" as its decision-making rule, something less than unanimity but more than a simple majority. —König T, Slapin JB (2006). "From unanimity to consensus: an analysis of the negotiations at the EU's constitutional convention". World Politics 58 (3): 413–45. 

With this in mind, it is better to use the word "unanimity" to describe the situation where everyone agrees, in contrast to "consensus". Eubulides (talk) 01:58, 13 September 2008 (UTC)

• Oppose trying to pass off unanimity as the only form of consensus. Wikipedia policies explicitly attempt to document the real world. See WP:PPP. WhatamIdoing (talk) 05:53, 13 September 2008 (UTC)
  • please don't misrepresent the issue - I'm trying to be fair about this, so should you.
• In my humble opinion a request for comment can not be used for a policy since all participating editors are involved parties! Having said that I suggest that reflecting in the policy what happens is the first step to acknowledging the practice, whether good or bad. Comparing this practice to best practice outside of Wikipedia (conducting comparative analysis) would be the second step in creating a policy definition of the ideal practice. Adopting this would require consensus...which can not be achieved because of the constant change in consensus group membership. The only relatively stable policy (even if not an ideal one) would therefore be one that is closely integrated with other policies, as was my suggestion in relationship to the article quality policies (sorry to harp on about it) because interdependencies tend to create strengths in policy creation by using multiple points of view and therefore a more critical thinking approach--mrg3105 (comms) ♠♥♦♣ 15:00, 13 September 2008 (UTC)
• Pro. this addition creates a proper frame for understanding what consensus means. without such a frame, it is far too easy (and far too frequent, in my experience) for editors to mistake a local consensus (between a handful of editors about a particular version of an article) for a real consensus (something which is generally acceptable to readers and editors alike). this leads to edit-warring as editors try to defend a local consensus against outsiders, and general confusion as one groups of editors tells another that they can't disagree because it violates consensus (when in fact disagreement means there's no real consensus to violate). this addition would help to lead editors back towards finding consensus, and away from aggressively defending consensus. --Ludwigs2 21:37, 13 September 2008 (UTC)

• Sorry Ludwigs, but your question is phrased terribly.
  • I wouldn’t take serious anything that begins with “idealistically”.
  • The use of “all” renders the definition certain to fail, either outright, or through subsequent weird redefinitions of words like “party”, “agree” or “substance”.
  • Why should it be policy that a certain “practice is rarely feasible”? If this is a statement of fact or opinion, it belongs in an essay.

--SmokeyJoe (talk) 10:58, 14 September 2008 (UTC)

• • • just stating common sense here. we all know what consensus means in the ideal, and we all know that it's not particularly practical; I'm just asking to make that explicit.--Ludwigs2 00:04, 15 September 2008 (UTC)
• Oppose By Ludwig's definition, if we have one person who is convinced that camel urine cures cancer, then we're stuck from removing it. This is so ridiculous, I'm not sure if I should laugh, cry, or think it's time to give up on Wikipedia. OrangeMarlin ^{Talk• Contributions} 15:06, 14 September 2008 (UTC)

  Laugh, cry, give up, then come back? ;-) You can convince the Camel Urine dude that your position is more solid, or you could include Mr. Camel Urine's sources if they exist (if there's sources, then you gotta represent them), of if there's no sources *and* Mr. C.U. is being entirely unreasonable (and only then), you can ignore Mr. C.U for being unreasonable, or finally, perhaps Mr. C.U. is actually right, you should never discount that option ;-) . It's a sequence of steps, that starts with assuming good faith. As long as there is good faith, there's always going to be a valid consensus everyone can live with. You only give up on consensus once you establish that good faith is not present, and all means of establishing a good faith rapport have failed. But often enough, people act in good faith, and who knows, you might learn something new about ... euh... camel urine. :-P --Kim Bruning (talk)

• A better definition of consensus (and I'm sorry but I forget who said it first, possibly Avi) is something like this: the version that causes the least amount of dissatisfaction among the smallest number of editors. Another definition is a version that everyone is willing to accept, even if they don't actually like it or agree to it. ☺ Coppertwig (talk) 17:57, 14 September 2008 (UTC)
  • I think you've phrased the first one wrong (shouldn't it be 'greatest number'?), but I'd be perfectly happy with the second, if that's preferable. this is just intended to encourage an expansive definition of consensus, rather than a constrictive one that can be abused. --Ludwigs2 00:04, 15 September 2008 (UTC)
• Comment: look guys, don't think that I'm suggesting this because I'm going to get something out of it. I know how to deal with people who misuse consensus as a way of pov-pushing, and I have no shame about making lives miserable when I see it happen. I'll be perfectly fine if this change isn't accepted; I might even enjoy myself more if I rid myself of the illusion that other editors take consensus seriously. however, I happen to think wikipedia will be a nicer place if there is gentler means of encouraging editors to use broader rather than narrower definitions of consensus. wikipedia is what we make of it - you should think about what you're making of it as you ponder this change. --Ludwigs2 00:12, 15 September 2008 (UTC)

Wikipedia is not supposed to be a "nicer place" as it is not an online social network site. It is meant to be an authoritative reference source. Ideally if all contributors to same article are on the "same page", litrally, there shouldn't need to be a need for consensus--mrg3105 (comms) ♠♥♦♣ 03:46, 15 September 2008 (UTC)

I am not surprised that the latest person to seek a "broader definition" is an editor that regularly finds his views excluded under the standard Wikipedia application of this concept. I'm sure that most experienced editors have seen previous attempts to redefine consensus as "something that includes my view"; perhaps it should be added to WP:PEREN. WhatamIdoing (talk) 05:39, 15 September 2008 (UTC)

I don't think that would particularly discourage such people; the people who continually try to redefine consensus probably genuinely and actually believe that there's a problem with it and it needs to be addressed. People on that thin line between OR/SYNTH and just "Way out in left field" aren't going to think it applies to them. Celarnor ^{Talk to me} 06:37, 15 September 2008 (UTC)

• Mu! (aka. Framing the debate): Ask the right question and you'll get the answer you want?. :-P Both pro and con are incorrect, so this is a false dilemma.

• (Perceived) consensus does exist on items for varying amounts of time: see: WP:SILENCE, so the con fork does not hold.
• The pro fork seems more correct, but it once again assumes that there exists such a thing as imperfect consensus, implying the existence of a dichotomy, so this fork is also incorrect.

Both forks are begging the question, by assuming that there is a dichotomy between ideal and non-ideal consensus. So this is a biased survey, since no matter which fork you pick, you always have to concede the assumed dichotomy. :-P

Actual consensus is simply the sum of preferences of all the participants. Whether it is possible to act on that consensus is another matter entirely

(if some people want to do A, and others want to do not-A, you're kind of stuck until someone changes their mind; this is where negotiation comes into play. You must convince others that your position is correct. For all the most important stuff on wikipedia -our content- we have plenty of time, and we can afford to do it this way, and can't afford to not do it this way .) .

--Kim Bruning (talk) 14:00, 15 September 2008 (UTC) Incidentally, one of the reasons why we demand consensus and not polling, is to eliminate this exact fallacy. (this is especially important for content, but other parts of the wiki benefit as well.)

• Neither Concensus is agreement, if this can't be found with civility and reasoning then dispute resolution. LeeVJ (talk) 23:00, 20 September 2008 (UTC)

• Pro a definition, as all the talk of how to go about achieving it is meaningless if there is no objective standard, particular as disputes arise not just over the content itself but over what the consensus is regarding content. CON the exemplar definition supplied as absolute agreement is no more consensus than a simple majority is consensus.LowKey (talk) 02:44, 25 September 2008 (UTC)

Definition of consensus on wikipedia is A: Local consensus: Absence of disagreement with the current state of a page. B:Global consensus: The average position of all wikipedians on a particular topic. C: Projected/predicted consensus: If we were to try to get everyone to agree to a particular compromise on issue x right now, that outcome is the projected or predicted consensus. --Kim Bruning (talk) 17:18, 25 September 2008 (UTC)

Objective measurements for each are:

• Local consensus: Act on it. If there are no edits, comments or reverts of your edits, you have local consensus, albeit temporarily (see WP:SILENCE, WP:CCC for limitations). This is an absolute measure.
• Global consensus: A: Hold a global survey (unreliable), B: Write a policy/guideline/essay and poll on that (VERY unreliable/ deprecated) C: interview a sample of people, and discuss the subject with them in depth. (reliability is usually good enough) D: Act on it locally, then see under Local Consensus. Similar limitations apply, note that it might take a while for enough people to become aware of your edit!
• Projected consensus: see methods under global and Local. Typically you use projected consensus when applying WP:IAR, or other situations where you're in a hurry. Some methods may need to be used post-hoc.

At all times, actual consensus is attained through action and negotiation --Kim Bruning (talk) 17:29, 25 September 2008 (UTC)

Consensus being replaced by silence at Wikipedia_talk:Reliable_sources_(medicine-related_articles)#Rfc_Promotion_of_MEDRS_to_guideline