Wikipedia talk:Consensus/Archive 1

From Wikipedia, the free encyclopedia
Jump to navigation Jump to search

well i like the way this is starting. I'm quite fond of a good consensus! Erich 05:42, 11 Jul 2004 (UTC)

I like how this page thinks about the problems of consensus in big groups. We are seeing this on policy pages, where it is increasingly difficult to change things because of the sheer weight of numbers. This doesn't often affect article pages, thank goodness. Pcb21| Pete 11:22, 11 Jul 2004 (UTC)

Unavoidable conflict

Wikipedia is a consensus. Since anyone may edit, anyone has veto power. This may seemingly be "over ridden" by the collective will of the wikicommunity, but if a minority of one editor participates in the process at all, it once again resembles consensus.

Consensus, like most relationships, is easiest when few people are involved. There are two potential consequences of this:

  1. Wikipedians may attempt to discourage growth or new users in an attempt to facilitate consensus
  2. Consensus may become impossible to maintain as ever growing numbers of new people toss articles around

Thus wikipedia should institute practices which facilitate consensus. An additional benifit may be that initially hostile new users may realize there is a system set up not with the purpose of blocking them, but with facilitating their edits. Links:

Hyacinth 05:08, 11 Jul 2004 (UTC)

As has been stated numerous times, consensus is not unanimity. →Raul654 16:20, Jul 11, 2004 (UTC)
Then I would be wrong, wikipedia does not run on consensus but on unanimity.Hyacinth 00:30, 12 Jul 2004 (UTC)
You are wrong. Wikipedia does run on consensus, not unanimity, because unanimity is paralyzing (as you have observed). →Raul654 09:37, Jul 12, 2004 (UTC)
not wanting to call anyone wrong... but... it seems to me that Wikipedia runs on neither unanimity or consensus. Too many people must walk away from debates feeling frustrated. I suspect we never even here from most of them... they just walk away frustrated or not that interested. Whether they are a silent minority or a silent majority we don't know. The absense of rules, and mass action makes wikipedia a very political business. To achieve your goals you need to impress people, make allies, form alliances, and most of all persist. The most persistent with the most allies is the most successful.
either that or you avoid conflict. Erich 11:46, 12 Jul 2004 (UTC)
Conflict, in the sense of disagreements about article content, is absolutely necessary for the existance and improvement of wikipedia, and thus unavoidable, avoidance is not even desireable. Conflict, in the sense of fighting, while undesireable, is also often unavoidable: Wikipedia:Wikihate. Hyacinth 21:13, 12 Jul 2004 (UTC)
So, to correct: "Wikipedia, when it works, works on consensus."? Hyacinth 21:17, 12 Jul 2004 (UTC)

agree with the above. soo.. it seems we agree:

  1. debate is healthy
  2. consensus is the ideal method of forming articles
  3. avoiding conflict is undesirable
  4. Wikipedia has scope to improve in the way debate, consensus and conflict are manged

or do we? as for point 4... I guess that's why we're here! Erich 23:18, 12 Jul 2004 (UTC)

I have a point that belongs there too. Too many people think that their ideas are valid just because they believe them - the idea that you must use reasoning and logic to back up your position is foreign to them. On Wikipedia, if you want to be taken seriously, you have to back up your assertions with actual reasoning, not "this is my opinion and Wikipedia should be open to all ideas so then it should be included" (which is something you are apt to hear quite often). →Raul654 00:23, Jul 13, 2004 (UTC)

I'm not sure what you mean. I've never heard that argument, in fact, the opposite is true. People attempt to align their personal believes with a Scientific Objectivity that is, in the eye's of wikipedia policy, the same thing. Hyacinth 00:27, 13 Jul 2004 (UTC)

Just read anything written by Plautus Satire, or Drbalaji md, etc. And for the record, wikipedia (like scientific objectivity) rests on proof, not faith and opinion. Wikipedia's policy most definitely makes a distinction between the two. →Raul654 00:47, Jul 13, 2004 (UTC)

Actually, an opinion is a fact. Hyacinth 01:02, 13 Jul 2004 (UTC)

A statement of opinion ("My favorite color is red") is a fact. An opinion ("Chocolate cake is the best kind of cake") is not a fact. →Raul654 01:08, Jul 13, 2004 (UTC)
Every opinion express an indirect fact about its author vision. There is an implicit ("I think that") before each statement of opinion. anyways, such a fact is not relevant on wikipedia Izwalito

Correct, and to quote Wikipedia:Neutral point of view:

  • "The neutral point of view policy states that one should write articles without bias, representing all views fairly...The policy doesn't assume that it's possible to write an article from just a single unbiased, "objective" point of view."
I didn't say that a single POV should be represented, I said that not all of them should be - that's when you get conspiracy theorists and alike. That's also why we require evidence to back it claims. Furthermore, it leads to terrible prose and cross-fire like debates within articles. That was the issue dealt with when Plautus satire was banned. →Raul654 02:13, Jul 13, 2004 (UTC)

Consensus is not unanimity

OK, so what is it? According to Wikipedia:

The process of achieving consensus involves serious treatment of every group member's considered opinion, and a collective trust in each member's discretion in follow-up action. In the ideal case, those who wish to take up some action want to hear those who oppose it, because they count on the fact that the ensuing debate will improve the consensus. In theory, action without resolution of considered opposition will be rare, and done with attention to minimize damage to relationships.

Yes, that's not unanimity, but it's a lot closer to unanimity than what we see in many Wikipedia policies, which often come down to some sort of supermajority vote.

anthony (see warning) 13:26, 10 Aug 2004 (UTC)

A consensus is a general agreement of the members of a group, which make a decision possible without the need to vote. A consensus can be uninamity, but is not obligatorily uninamity. Consensus as a method of deciding, tries to put in light on the validity of each participant opinion, to take in count every valid opinion and refuses to validate a choice when someone has a valid disagrement to object.Izwalito

Consensus is not precisely defined in official Wikipedia policy, but it is widely accepted that 80% is a fair threshold; by this definition, the minimum number of people that can build a non-unanimous consensus is five (a 4 to 1 vote means that 80% agree). ~leifHELO 04:13, 5 Dec 2004 (UTC)

There's no consensus at all if the fifth doesn't feel satisfied with the conclusion. That's the point. As far as that's concerned, consensus is unanimity. It certainly isn't the majority steamrollering the minority -- at least, not in the real world.Dr Zen 11:44, 1 Jan 2005 (UTC)
"Accordingly U-3 and lesser degrees of unanimity are usually lumped in with statistical measures of agreement, such as "80%, mean plus one sigma, Two-Thirds, 50% plus one" levels of agreement. Such measures do not fit within the definition of consensus given at the beginning of this article." Consensus decision-making. Should be required reading for anyone who thinks a poll creates consensus.Dr Zen 11:47, 1 Jan 2005 (UTC)

It's insane to think that methods of community consensus will port to Wikipedia, where 'community' itself is ill-defined. Voting works here because the marginal difference that one vote makes is seldom the deciding factor, so there's not the same sort of motivation to try and game the system. 80% is not consensus by the definition given above, but the definition above leads to madness, and doom. grendel|khan 05:26, 2005 Jan 3 (UTC)

Well, would there be anything wrong with that? I always liked doom, myself. ᓛᖁ♀ 06:36, 3 Jan 2005 (UTC)
I wonder whether Grendel would like to have a go at making that POV sound less like something ey pulled from eis arse? I think consensus can be "ported" here. Yes, the coming and going of members makes the community rather fluid, and yes, that type of "consensus" is much easier to create than it would be in a smaller group (although one reminds oneself that it is precisely the decision-making method in diplomacy, where treaty wordings are most usually decided by consensus -- although, of course, one wouldn't forget that treaty language is circumscribed). But look at Blankfaze's proposal to expand CSD as an example. There is consensus that "something needs to be done". Ey could have proposed changes, looked at any opposition, incorporated others' views and in time built a policy that all involved felt was satisfactory (even if it wasn't precisely what any of them wanted). Instead, ey has gone for a series of votes, which, if any were to pass, say, by 80-20, giving the percentage Grendel suggests represents "consensus", there will be 20 editors who were interested enough to vote and do not want the policy (plus however many of the 80 who voted in favour who might have rather had a somewhat different policy but plumped for what was on offer). How is that "consensus" under any definition, let alone the one that leads this article?Dr Zen 10:07, 3 Jan 2005 (UTC)
Unanimity: "everyone being of one mind". Consensus: "general agreement or accord". It's a judgement call whether 80/20 represents "general agreement". Dbenbenn 00:12, 11 Jan 2005 (UTC)
That's not very "general". I'd describe that as "most agree" rather than "generally agreed". When we talk about a consensus in science, we mean things like quantum theory or evolution by natural selection. These are very widely agreed, not just by four-fifths of the community.Dr Zen 01:17, 11 Jan 2005 (UTC)
To reiterate some of the above points: In consensus everyone has veto power. Despite this it is not unanimity because opponents may wish to abstain from using their veto power, and often do. Thus 80/20 is not consensus, 100/0 is. Hyacinth 00:19, 18 Jan 2005 (UTC)
I'd rather you address the issues I raise and not my "arse", Dr Zen. NPOV states that We should not attempt to represent a dispute as if a view held by only a small minority of people deserved as much attention as a majority view. Look, we already have instances in which consent doesn't work, because the process is inherently adversarial---debates on deletion and featuring, for instance. I did not mean to suggest, also, that 80/20 is consensus. It's not. I take issue with the idea of... well, what do you do when one person demands that movies from 1980 be in Category:1980_films and another resolutely demands that they be in Category:1980_movies? There isn't a useful gray area in every debate. What do we do---do we even need a rigorous model? Is there one that fits, and is useful? What is the place of a consensus process on Wikipedia? grendel|khan 05:44, 2005 Jan 19 (UTC)
In my opinion it is very rare that decisions are bipolar. I think that we tend to build simplifying models of the world that yield such bipoles but there are usuallu other options and it is the surfacing of those options that is the strength of consensus. Consider cross-mapping synonomous categories for the movies/films, for example. --Theo (Talk) 10:22, 1 Mar 2005 (UTC)

I just came across this discussion. My $.02: Consensus does not require unanimity, but it does require that there be no objections or vetoes. 80/20 is not consensus. 100/0 is unanimity. 90/0/10 (yes/no/abstain) is consenus without unanimity. Consensus is an "ideal". Some wise soul once said something to the effect of "Ideals are like the stars, you can use them to guide you, but you never actually reach the stars." I think we have to keep in mind that the goal of Wikipedia is producing a free encyclopedia--like the wiki software, consensus is merely a means for achieving that goal. This point is reated to the notion found at Wikipedia:What_Wikipedia is not#Wikipedia is not an experiment in anarchy. Ideally, all parties should be able to reach agreement through consensus on every issue. Unfortunately that is not the reality. Attempting to reach consensus should always be the first avenue for resolving issues, but some issues are intractable (or in cases, some individuals are simply unwilling to compromise). Voting should be a last resort for resolving such intractable issues, because the policking that often accompanies voting can produce ill will between factions. But I'd like to think of voting and the roughly 80% threshold as something like exercising cloture in the U.S. Senate. Rather than allow a determined minority to hold proceedings hostage, a super-majority can determine to go ahead despite the objections of the minority. However, it should be an exceptional procedure to be used only after it is clear that consensus cannot be acheived and there is a substantial majority with a coherent position. Ideally (there's that concept again), the objections of the minority should somehow be noted rather than entirely discarded. olderwiser 03:36, Jan 19, 2005 (UTC)

I think you mean consensus is not always possible, that constant consensus is an ideal. Hyacinth 03:03, 20 Jan 2005 (UTC)
Good point. But actually, given the radically open and fluid constituency of Wikipedia, I honestly think true consensus IS impossible for MOST community-wide issues and for SOME deeply-felt content issues. True consensus, when it works, works through trust, which is built over time. With a radically open constituency, where a significant portion is transient--it is very difficult to establish the sort of trust required to reach consensus. For issues affecting the entire community, achieving true consensus is most often impossible and supermajority voting (after a period of discussion and debate to attempt to reach the most acceptable phrasing) may be the best we can do to demonstrate approval of new proposals. For content issues, I think consensus is at least possible in most cases and voting on versions should be a last resort. olderwiser 13:55, Jan 20, 2005 (UTC)

Consensus is a far cry from anarchy. I think that if Wikipedia, like anything, is to use consensus it must require it always. The only way movies from 1980 would be categorized in this scenario is through compromise, abstenion, or being blocked or banned for refusing to participate in consensus. In this case consensus isn't impossible.

The relevant policies in the opposite scenario, no consensus, would be Wikipedia:Compromise and Wikipedia:Co-operation.

I ask: Would we rather people be blocked, banned, or otherwise, for refusing to participate in consensus than quit because their concerns where over riden, or be blocked or banned for attempting to have their concerns addressed? I'm not sure. Hyacinth 03:23, 20 Jan 2005 (UTC)

The reason I mentioned [[Wikipedia is not an experiment in anarchy is because Wikipedia is similarly not an experiment in consensus decision making. We shouldn't let ideological considerations of consensus distract from the practical goal of producing a free encyclopedia. I'm not sure I understand your last comments about blocking, banning, etc. Blocking and banning should be reserved for disruptive or abusive behaviors. If persons become abusive or disruptive they should be blocked, and if necessary banned. Building trust goes both ways, and if a minority position is not able to present sufficiently persuasive arguments in support of their position in a civil manner and resorts to abusive or disruptive behaviors, that is unnacceptable. olderwiser 13:55, Jan 20, 2005 (UTC)

Sorry if this isn't my place, but it occurs to me that subject consensus building could be interperted to directly contradict Wikipedia's priamry policy on NPOV. The neutral position isn't always the majority or even unanimous position. --Axon 13:21, 25 Feb 2005 (UTC)

The neutral position may not be the one held by the majority but it can be one that everyone can live with. If X says "x" and Y says "y", which are opposed, then the neutral position is to say 'X says "x" and Y says "y"'. In general, however, the problem for consensus is when Z, a lone voice, says "Z", a strongly held position that other participants consider to be irrelevant and that Z insists must be considered. An absurd example: In a discussion about the colour of the sea, some people call it blue, some call it green, some say it varies with the prevailing conditions and Z says that Brighton is by the sea. Consensus can not be reached until Z understands why Brighton is not relevant. A more subtle case is when Z insists that the sea sometimes has electric pink polka dots. --Theo (Talk) 11:04, 1 Mar 2005 (UTC)
How does one objectively differentiate the above situation from a situation where Z is the more neutral position but in minority? In the second case, consensus is actually non-neutral. --Axon 17:13, 2 Mar 2005 (UTC)
A thought-provoking question. Ultimately, however, is that not the premise of my first sentence; just recast it to replace "then the neutral position is to say" with "and Z says […], which is the neutral position". To reach consensus, the neutral position must acknowledge the views of the substantial vocal groups. --Theo (Talk) 17:42, 2 Mar 2005 (UTC)
Thank you, and thank you for your equally thought-provoking response. Apologies, but I'm not quite sure I understood your point. The situation I'm am contrasting is where X says "x" and Y says "y", which are opposed, then the neutral position is to say 'X says "x" and Y says "y"', but Z says 'z' which contradicts the 'neutral' position then how do we determine if Z is a crank or genuinely represents the neutral position but is merely in minority. Also, are we talking in terms of groups within Wikipedia? If so, then we shoudn't we discount the consensus within Wikipedia: it is the consensus outside of Wikipedia in the "real world" we should be reflecting? --Axon 18:16, 2 Mar 2005 (UTC)
I do not understand how "z" can contradict 'X says "x" and Y says "y"' without dissembling. In this case the neutral statement is objective and demonstrable. Assuming, however, that it is possible for Z to make an honest statement such as you describe, we need to assume that Wikipedia broadly reflects a world-view (palpable nonsense in some senses: we exclude the set of people with no Internet access, for example). Wikipedians in good faith will attempt to represent the NPOV as they see it but the size of the community should encourage accurate representation of the larger constituency. --Theo (Talk) 00:30, 3 Mar 2005 (UTC)
For example, "z" could state that X does not mean "x" and/or Y does not actually say "y". Chances are this could be dissembling but there could be cases where this is not always true.
I think the real question here is whether the consensus within Wikipedia reflects the consensus, and thus the NPOV, without. As you state, not only does Wikipedia merely represent the (very small, priviliged) minority of the world's population with Internet access, but the even smaller, self-selecting group of those who are motivated to edit Wikipedia. This is, by no means, representative.
OTOH, perhaps we could rely upon sufficient numbers within Wikipedia to make up a balanced consensus, but I think this would only hold for popular topics and would fail for obscure topics where the consensus could easily be ignorant of wider issues. --Axon 11:32, 4 Mar 2005 (UTC)
ULtimately, I believe that we must rely on Wikipedians to do the best that they can. NPOV is such an essential policy that it is hard to imagine anyone remaining unaware of it for long. There is nothing that we can do to make non-contributors contribute any more than we can give everyone Internet access: even Coca-Cola only aspires to reach everyone in the world. I like to think that enough editors are already doing a great job doing what they think is right, and that this tendency can only grow—I think that we are already at critical mass. --Theo (Talk) 14:26, 4 Mar 2005 (UTC)
The issue here is not trusting editors but whether we should rely upon consensus as a mechanism to determine the neutrality of articles. My point was that consensus can only work fairly if there is sufficient numbers of informed participants that are representative of wider opinion and that this cannot always be the case given that Wikipedia is not a random sampling of all opinion. This was not meant as a slur against Wikipedians, merely as an observation. Consensus may work in the general case of popular articles but, as I've stated above, may fall down in the case of obscure articles, not necesarily through any malicious fault of editors, but possibly through ignorance or lack of interest. --Axon 15:29, 4 Mar 2005 (UTC)
I agree with everything that you say. Be assured that I was not suggesting that you were slurring other editors (WP:AGF). My point was that we must do the best that we can with what we have and that our consensus is the best that we can get at any momemnt because we have no way of identifying who has the 'right' answer (and do not get me started on 'right').

I just found this page, and a proposal on how to proceed

Recently, at Wikipedia:Village pump (policy) I brought up the issue of a lack of policy on consensus, not knowing that work had already started at this page. Please take a look: Wikipedia:Consensus/Village Pump discussion

Several people responded positively, so I thought I'd start a wikipedia page about consensus. I typed in "Wikipedia:Consensus" and ended up here. I'm glad to see that work has already begun and glad to see familiar names attached to the comments on this page.

I have 28 years of experience with structured consensus decision making starting back in 1977 with the Clamshell Alliance (just discovered another article I have to write), and continuing with non-profits and community groups that run on consensus decision making. I have taken some facilitation training workshops over the years. I would call myself very experienced, but not quite an expert.

I've read through the project page and this discussion, and here's what I'm thinking:

  1. We need to concisely define what wikipedia's version of consensus is.
  2. We have to explain how consensus works.
  3. We have to create tools for implementing the process when needed.
  4. Once approved, the process has to be incorporated into all the other wikipedia policy pages in a way that it promotes the consensus decision making process.
  5. Clearly, when this topic comes up, people bring their own understanding of what consensus is, and that varies widely. While many people do not have experience with structured consensus decision making, almost everyone thinks they know what consensus is.
  6. While Wikipedia may not have been designed as an experiment in consensus decision making, it cannot avoid being one. I think that is a good thing.
  7. The process of consensus decision making will be different in different situations. We should think about how the process is to be applied, from the smallest editing conflict to the largest policy decisions.
  8. Consensus functions by "quiet consent". This means that not saying anything implies acceptance. This is the norm at wikipedia. Whenever anyone makes an editing change and it is not reverted, that is consensus in its simplest manifestation.
  9. Acceptance does not mean that you agree. Consensus can be reached without everyone agreeing. People that abstain are realizing that their concerns are not important enough to block a decision. It is often valuable for everyone else to know why people are abstaining.
  10. Consensus is about finding common ground. It is the duty of the majority to listen to the concerns of the minority and try to adjust their solution to address those concerns. There are ways to incorporate multiple views into solutions.
  11. Consensus does not trump the truth. It is the duty of those who are certain of the facts to convince everyone else of their mistake or be proven wrong.
  12. Blocking consensus is a rare event. If you block consensus you should be prepared to prove the validity of your position, and to show that the proposed consensus is bad for the group. People who abuse the process by blocking consensus regularly without cause lose their credibility.
  13. There are known techniques for helping groups reaching consensus.

I've been thinking about creating a framework for decisions that would get plugged in whenever necessary. This would be the work that a facilitator would normally do, but in our case, it can be done collectively. Something like:

  • Solicit opinions
  • Brainstorm possible solutions
  • Create a rough proposal
  • Hear people's concerns
  • Modify proposal to address concerns
  • Call for consensus
  • Ways to resolve blocked consensus
  • Moving on

Each step would have links to pages to help people through the process. A template could be created whenever the process needs to be invoked. It would outline the process, and have links to more detail. That way, everyone would get familiar with the method quickly.

I propose that we start by creating the process for making policy. AND I propose that we use that very same process for creating the process. In other words, I think we should modify what we write based on the experiences we have among ourselves trying to write it.

I volunteer to help facilitate this along, but pretty soon I hope that it is self-facilitating.

So the first step, which I just took, is to float an idea and Solicit opinions about the idea. What I am looking for is what things trigger the process. In most case I think it is an individual recognizing that there is a conflict and trying to solve it. That person writes a proposal in broad terms and asks for comments from others. At this stage, people should be having discussions. The important thing is not to say "I'm in favor" or I'm opposed" but to define the problem well. This means exploring all the manifestations of the problem and creating criteria for any possible solution.

So what do you think? I'd like your opinions.

OK. I've written enough for now --Samuel Wantman 08:50, 27 Feb 2005 (UTC)

By "creating the process for making policy" you talking about changing Wikipedia:Policy and Wikipedia:How to create policy as we write Wikipedia:Consensus or just writing Wikipedia:consensus using consensus?
Wikipedia:How to create policy is incomplete. It just outlines the BEGINNING of the process. If you just write a policy and post it, that doesn't mean it will be accepted and implemented. I think that page was probably all that was needed in past years, but it needs more now. More often than not the problem now is about how to CHANGE policy. If you have a great idea for a change and just post it, it will be reverted almost immediately. If you post the change in a talk page, you might get a few comments. If it is a big change, it might be dismissed as "this is not the way things work here". The status quo rules. I'm constantly reading about users who are frustrated by this.
Wikipedia is supposed to run by consensus, but how should that be played out? I know I am not the only person who has thought of a change for Wikipedia that I think will make it much better, and then been at a loss for how to proceed. If something is posted on the Village pump, how does it become policy? I think this should be clear. What I think we should be doing is formalizing how consensus works within Wikipedia. Not just what consensus is, but how it works in specific cases.
So while we are thinking about how consensus should work, we should also be thinking about how it applies to different areas of Wikipedia. Perhaps we can use certain pages as examples. The process for deciding on a major overhaul of an article is probably not as involved as the process for deciding on a major change of policy. But they should both be clearly spelled out.
By "self-facilitating" are you proposing an automated built-in (mandatory) facilitation process, or a spectrum of users with various levels of involvement in facilitation and variously contributing to the process? Hyacinth 01:02, 3 Mar 2005 (UTC)
I don't think it would require any additional programming. Facilitators have several responsibilities:
  • They help direct the process: Templates could be used to help the process. Anyone creating a proposal could include a template that references discussion on how the process works, adds the proposal to a wikipedia proposal category, etc... For example, each item in the bulleted list above could have a template. So when the decision making process gets to that point, a corresponding template could be added to the discussion.
  • They educate everyone about how the process works: The templates would have links to pages that could help instruct people on how the process should work, and suggestions of techniques for moving the group toward consensus. Anyone could add the templates, so it would be, in a sense, self-facilitating. As these templates begin to show up on more and more pages, people would become much more familiar with how the consensus process works.
  • They keep things moving: Typically, the big problem in facilitating large groups is having enough time for everyone to be heard. We probably need time guidelines for every step of the process.
  • They remain impartial and try to look for middle ground: Anyone can take on this role. This is very much like the current mediation process. The community can maintain a list of people to call on to help sort out any controversial issues.
  • They keep things civil and fair: Everyone can help with this.
--Samuel Wantman 07:36, 3 Mar 2005 (UTC)

Consensus and polling

Consensus should be reached by polite discussion and negotiation, but should never trump policy. I might add that at Wikipedia it should be based on reasons that support our project's goals. A reminder is clearly in order:

Wikipedia's only purpose is to create a free encyclopedia which is accurate, useful and unbiased. NPOV addresses the "unbiased" part of that, although many users are POV pushers and actively disagree with that policy. What is useful has never been pinned down, so there's a lot of trivia; more serious-minded folk simply ignore the trivia, as long as it stays out of the way (which it usually does). Accuracy is fairly well defined: we prefer reference to primary sources, and web links generally suffice for that.

Wikipedia is not an experiment in anarchy or democracy. It is not a social action movement (like "no nukes"). It has a purpose which is mandated from above. That is, the founders created an unalterable manifesto. (Jimbo controls the board he set up and has consistently refused to budge from his initial vision: accuracy & NPOV combined with civility.)

(Sorry if I'm meandering... this is a first draft.)

As a means to the end of assembling and organizing the articles, the wiki format was chosen. But it quickly became apparent that with more than a few dozen users, AssumeGoodFaith was not enough. A hierarchy of policy-enforcers came about despite the best efforts of anarchy proponents to keep everything level. I won't go into the whole history, but we now have an elected arbitration committee empowered to decide crucial issues (subject only to Jimbo's veto).

Civility is the other neglected value I want to talk about. Many people try to push the bounds, inserting nonsenses (I call it "graffiti") into articles; relentlessly reverting others' changes instead of discussing matters on the talk pages; choosing usernames and placing calculated to irritate others; and so on. All of this incivility is disruptive to the project.

Now as to what consensus is: first let us say what it is not. A quick vote by a dozen or two people in favor of letting someone violate a rule (or bend it) is not consensus. It's merely a poll. Uncle Ed 12:18, August 30, 2005 (UTC)

I presume Ed is inspired here by this RFC: Wikipedia:Requests for comment/User names/Trollderella. The story: Ed, as a bureaucrat, renamed a user named Trollderella. An RfC was already underway to discuss whether the name was inappropriate. At the point of Ed's intervention, a count showed that 11 users thought the name was OK, versus 2 who opposed the username.
As regards Ed's larger point: polls are an imperfect measure of consensus, but it's surely better than the opinion of a single bureaucrat.
In the specific case of Trollderella, the policy invoked by Ed cantains the proviso that sysops should only take action if their judgement is that a "rough consensus" has arisen that the username is inappropriate.. Certainly, no such consensus was evident in this case. — Matt Crypto 12:33, 30 August 2005 (UTC)

User:Iit_bpd1962 comments: 29 April 2006

I quote, — Matt Crypto 12:33, 30 August 2005 (UTC)

"Consensus should be reached by polite discussion and negotiation, but should never trump policy. I might add that at Wikipedia it should be based on reasons that support our project's goals. ...." quote ends prematuredly since it is already available above in some sort of "complete in itself" form.

The reasons that support our project's goals can be replaced by a Logic sytem acompanied by the basis ; viz., axioms which are atomic that support the Logic being employed). Logic is more appealing (by virtue of it's nature, as the physical sensation it leaves behind in our mind a sense of semantically completion and thus allows Logic to be usable a tool for the basis for everthing else left to be justified), as a method to justify in toto any reason; and reason is in this sence is to be viewed as the compact mundane word for justfying any result it seeks to justify, which otherwise would need the complete Logical arguments to support the uttered reason which is used in an attempt to justify an observed phenomenon, mental, physical, or of other domains, and must include the very axioms / premises upon which the Logic itself rests.

If any establised, or otherwise (Eureka!, in the public domain), logic, is suspect in eyes of the corncerned police, then it's axioms need to re-examined. On completion of such an exercise, the semantics embedded in the final conclusion of employing such complete logic systems, should then, until proven to be "not in keeping with reality" (physics, chemistry, physical phenomenon), must serve as the sole justification for providing a "reason" to explain anything.

Until proved otherwise, the statement "Logic can prove everything" is valid especially when its axioms are based upon experienced reality, (and here and here alone consensus as to ascertaining that others too have consciousness), may seem abhorent a notion in the very spirit in which science / Mathematics has been entertained. "Question Everthing" is a serious suggestion.

Consensus (of other's views) may lead to circularity in thinking; maintaining potential ignorance by resorting to "established" "athorities". A single individual, starting from experience and employing honesty, fortified by sheer "excersices involving complete / consistent content" can indeed come up with statements whose sole judge shall be, NOT consensus, but the Logic and Logic alone insofar it does not conflict with reality (consensus with physically manifested laws, and one own experience.

If my statements above are accepted, without consensus of anykind, except in the form of querying Nature of NATURE, or seeking answers within one's thoughts, then I shall, penultimately, conclude here that consesnsus seeking, of the usual sort, is the sole cause of beaurocratic delays, dogmatism, and consulting "authorities" for permission to opine, when the Logical sanctity of such an authority may itself be in question.

Consensus is a form of compromise. An argument for non-consensus approach was written by, Ayn Rand's, Fountainhead, where the character Roark addressed the issue of a society harbouring ingratiation [1], "politnesss", etc., as means to get their way: see, The Fountainhead, by Ayn Rand, pages 606 & 677 till 685, First signet printing, April 1952, copyright 1943, the Bobbs-Merril Company, Last copyright, Leonard Peikoff, 1993. This is a reference and cannot be categorized as " resorting to any consensus ".

Thank you.

Rough consensus

Recently there has been a move to change "Votes for Deletion" to "Articles for deletion" under what seems to be the ideas expressed in meta:Don't vote on everything. Both the Wikipedia:Articles for deletion and Wikipedia:Requested moves have been edited to remove the word vote and encourage consensus.

In the Westminster system of government, Cabinet decisions are consensual collective and inclusive. If as a member of the government does not agree with a decision (s)he can resign from the government; (as did several British ministers over the invasion of Iraq). This means that in the Westminster system of government the cabinet always collectively decides all decisions and all ministers are responsible for arguing in favour of any decision made by the cabinet. It is a major cornerstone of the Westminster system.

Wikipedia is very different, for all but uncontroversial trivial propositions, it is unusual for decisions on Wikipedia talk pages to operate on a true consensus. Instead they operate on a rough consensus where it is recognised that a minority are in opposition. The question then arises is how large must the majority be to ignore the opinions of a minority? I think that this page has it about right in describing how the community tries to come to a consensus and how it resolves disputes when a true consensus can not be achived. So I hope that the page does not come under attack from those who would like to idealise the process and not describe how conflict between reasonable Wikipedia editors is resolved in practice. To do that would be unfair on people new to Wikipedia (those most likely to read this page) who do not yet know how a Wikipedia rough consensus is arrived at in practice Philip Baird Shearer 11:45, 4 September 2005 (UTC)

I have changed the wording of the paragraph mentioing numerical standards to indicate that a certain parge is not actually consensus but is used to help determine consensus; this is more in line with both the meaning of the term and policy. Mindspillage (spill yours?) 19:16, 4 September 2005 (UTC)

Minimum threshold?

If a vote has to be held to attempt a consensus, is there a minimum to the number of people that must participate before it becomes reasonable to be taken as consensus? Erath 12:57, September 5, 2005 (UTC)

4? --James S. 23:15, 13 January 2006 (UTC)

Dispute resolution

The final paragraph has been used to justify non-collegial behavior by several Revert Warriors for Truth. After all, if they're fighting a bad thing, they can ignore all rules. I would like to add something like:

The recommended way to deal with this is to draw the attention of more editors to the issue by one of the methods of dispute resolution, such as consulting a third party, filing a request for comment (on the article in question), and requesting mediation. Enlarging the pool will prevent consensus being enforced by a small group of willful editors.

Is this a valid statement of policy?

Is this the right order?

Is it worth adding?

Septentrionalis 16:17, 19 September 2005 (UTC)

Sounds sane! :-) You could even mention third opinion, article RFCs (which I've been told need fixing) and mediation cabal. Kim Bruning 23:59, 27 November 2005 (UTC)

The numbers

I've removed this:

The numbers are a relatively recent addition (wasn't there in April early March), and as far as I'm aware they are not correct. There is no numerical guideline for AfD, for instance; indeed the name has recently been changes to emphasize that there isn't a numerical vote, and the voting mentality is decried.

However I've found one recent instance where someone appealed to the figure in this document in his claim that a consensus had been reached with a particular numerical vote. This is not a policy document and we should take care to ensure that it isn't mistaken for one. Perhaps the numbers can be restored with suitable wording to ensure that they are not taken as definitive. --Tony SidawayTalk 05:39, 14 September 2005 (UTC)

Well, they give useful starting points for understanding the process. There's nothing more magic about the existence of these numbers than their removal. If they stay removed, people will still use them. In fact, I recall our newest bureaucrats, guardians of consensus, citing the 80%ish thing for when to promote on RfA. -Splashtalk 01:54, 15 September 2005 (UTC)

So how about simply prefacing the list with a "Note: these numbers are not considered binding, are at the very most guidelines, and may be applied with wide variation as the situation demands". Or something. Meantime, it seems to drastic to simply pretend they don't exist. -Splashtalk 01:56, 15 September 2005 (UTC)

I'm not really sure that there is any value in putting down numerical values on which, inevitably, there is no consensus. 60% is a consensus on WP:RM? Says who? The numbers seems to me to be unhelpful because their presence implies more than it is safe to imply. However if you can find a form of words that makes it plain that the numbers are completely without value and yet still manages to make them appear to have some function, go for it. --Tony SidawayTalk 19:35, 19 September 2005 (UTC)
To WP:RM in particular, I have no idea as I don't go there unless I need something. I'll not replace the numbers since I doubt I can find a formulation that will be acceptable to you, so I'll let your opinion prevail instead of mine. However, I'm not at all convinced that 'closing your eyes' to the numbers will make them magically dissolve away, and nor will pointing here and saying "look the numbers have gone". I could be wrong, though, I suppose. -Splashtalk 20:26, 19 September 2005 (UTC)
Well at least we'll have less likelihood of people misguidedly pointing here and saying "look, a consensus is X%". --Tony SidawayTalk 21:16, 19 September 2005 (UTC)

I originally added the numbers to the page. I believe I did so after witnessing some controversy in which the organized vote drives of one side managed to get (barely) the necessary numbers for a supermajority—thus creating a "consensus decision" which was then used to club the other side's valid and unresolved objections into silence. I was, as I recall, appalled; I added the numbers as part of a series of edits meant to point out how bizarre the definitions of consensus used on Wikipedia really were. I am doubly appalled to find out that people are actually citing these numbers as The One True Official Meaning of consensus; I am triply appalled when I realize that this is probably because my description of the weird definitions was dead on. —Charles P. (Mirv) 09:27, 29 September 2005 (UTC)

So, is it hard to breathe with those beans up your nose? I've certainly choked a few times myself. Sadly if you hadn't put the numbers there someone else probably would've.
While in practice there are numerical guidelines for determining consensus I think it is better not to have them here; there are too many people who misinterpret and cite them as if they are consensus itself, and set in stone. And on some—AfD in particular—there is no consensus as to what the numerical guideline should be. Mindspillage (spill yours?) 13:44, 29 September 2005 (UTC)

In the case of , Cabinet government (see above) a true consensus is reached for all decisions because one resigns from the Cabinet rather than support a policy one disagrees with. That is not possible in Wikipedia decision making the best that can usually be achived is a #Rough consensus.

Let me give an example:

  • A person asks for a WP:RM no one expresses an opinion for or against. That is taken to be a consensus to move.
  • A person asks for a page to be moved one other person expresses an opinion that it should not be moved, then after five days it is taken to be no consensus for a move.
  • A person asks for a page to be moved one person expresses an opinion that it should be moved another is against. After five days is it a consensus to move or not?

Assume that in the above example that the move is widely advertised and no one is willing to change their position. In the last case should the page be moved or not? If so why? If not why not?

Unless numbers are added to define what makes a rough consensus, then the whole process leaves far to much arbitrary power in the hands of the administrator who makes the decision. As most administrators do not want to get bogged down in arguments defending their decision, putting numbers up front on what is a rough consensus helps everyone. Putting numbers on this page helps people new to Wikipedia to have a chance in the rough and tumble of Wikipedia consensus building. Because in the real Wikipedia world there is rarely a decision with 100%. Some yardstick is used to decide what is a rough consensus and publishing them helps everyone, reduces residual conflict and protects the administrators who have better things to do than argue contested decisions.

As a real world example look at the straw poll on Wikipedia talk:Naming conventions (use English)#Proposal and straw poll regarding place names with diacritical marks now there are two contrary positions which despite paragraphs of discussion I do not think one person has changed their vote. So how likely is it that a real consensus will ever be possible. In that case it is unlikely that even a rough consensus can be reached but pages have to have names....

The fact is that in reality different thresholds are taken to be a rough consensus depending on the severity of the process. A page move is not as critical as a page delete and a page delete is not as critical as granting adminship, this should be mentioned here and probably quantified with numbers as well. Philip Baird Shearer 23:38, 14 October 2005 (UTC)

If no one has an opinion on what I have said above, it is my intention to reinsert the text at the start of this section back into the page. Philip Baird Shearer 13:04, 16 October 2005 (UTC)

OI! When did we have numeric values of consensus? zOMG! Isn't wikipedia NOT a democracy? I thought consensus != majority. I guess I must have been mistaken?

(see the "often consensus is taken as 75%" , on the page. Err, seriously, WTF?)

Kim Bruning 22:22, 26 November 2005 (UTC)

Seeing the discussion above, I'll just remove the numbers. Kim Bruning 22:35, 26 November 2005 (UTC)

Please discuss changes before making them. You will notice that I left the question open for more well over 24 hours before I made a change. What does "OI!" and "zOMG!" mean? What do you thing a consensus is if it is not the sort used in Cabinet Government? -- Philip Baird Shearer 00:19, 27 November 2005 (UTC)

I have no idea what you think consensus is, but it's definately not a majority. I just said that though. *sigh* See also WP:NOT. Kim Bruning 01:20, 27 November 2005 (UTC)
I think a true consensus is the sort arrived at in Cabinet government as I said earlier in the section. I think what is arrived at in most cases on Wikipedia is a rough consensus. But I have already said these things higher up the page and I persume you have read them so you ought to know what I think a consensus is. So what do you think a wikipedia consensus is apart from "definately not a majority"? How for example should decisions be made over page moves when only three people express an opinion on the matter? --Philip Baird Shearer 13:11, 27 November 2005 (UTC)

Consensus means everyone comes to agree on something, or at least agrees to disagree. It doesn't mean you vote on it and that a certain percentage wins. It has never meant that. RFA votes and other votes taken on Wikipedia do not matter. Regardless of how a vote goes, it is simply to show what the community thinks. Bureaucrats can always give adminship to someone whose vote did not go well. These numbers must be removed. Their addition to this page should have been speedy reverted. — BRIAN0918 • 2005-11-27 01:26

See also this quote by JW. (on the other hand I could provide the results of some polls, that were judged by the percentage) +MATIA 01:48, 27 November 2005 (UTC)

The current page expresses JW position the the sentence before and after the numbers:

Note that these numbers, being mere statistics, are not binding on the editor who is interpreting the debate, and are often varied widely to suit the needs of a particular situation:
However, judgment and discretion are applied to determine the correct action. The discussion itself is more important than the statistics.

Consensus does not mean "agrees to disagree". In a system like cabinet government everyone has to agree or resign. What we are talking about here is a rough consensus. The only thing open to debate is how large that rought consensus ought to be for it to count as an agreement to carry out an action. As I said earlier in this section

Putting numbers on this page helps people new to Wikipedia to have a chance in the rough and tumble of Wikipedia consensus building. Because in the real Wikipedia world there is rarely a decision with 100%. Some yardstick is used to decide what is a rough consensus and publishing them helps everyone, reduces residual conflict and protects the administrators who have better things to do than argue contested decisions.

--Philip Baird Shearer 13:11, 27 November 2005 (UTC)

Read some of the old discussion on AFD closings (this is as old as the wiki). You'd be more accurate, and equally realistic to state that the "number" for consensus is the square root of -1. Kim Bruning 18:07, 27 November 2005 (UTC)
Maybe an example is in order: On AFD, if 99.9% of the people are of the opinion "KEEP, very encyclopedic, and I like the british spelling too", but one single person goes "Oh, no wonder, it's a copyvio from britannica!", then the consensus is read as delete. There is no numeric threshold. Kim Bruning 18:11, 27 November 2005 (UTC)
That's an unreasonable example. The consensus was to retain an article on the topic, but copyright law demands that we remove this particular one. If a new article were written on it, which was original, it would be soundly kept without an AfD. -Splashtalk 18:29, 27 November 2005 (UTC)
Another, more reasonable example. On RFA, if a user has several supports and several opposes, but thoroughly discusses the issues with the opposition, the bureaucrat would not be unreasonable to ignore the arbitrary "supermajority" and promote the user. — BRIAN0918 • 2005-11-27 18:33
It's a perfect example. Everyone agrees that the article must be removed. You will note that no one will take it to deletion review in this situation. (The acid test of consensus is not when everyone agrees, it's when no one opposes.). Kim Bruning 20:46, 27 November 2005 (UTC)
No, that's just not right. If AfD decided to delete an article that actually turned out to be a copyvio, recreations would be speedied until DRV overturned the AfD. If AfD decided to keep an article that actually turned out to be a copyvio, a non-copyvio recreation would be fine. Consensus has nothing to do with abiding by the law. -Splashtalk 21:09, 27 November 2005 (UTC)
If agreement cannot be made, but those who disagree can at least accept the fact that they will always disagree, then that is almost as good as consensus. Picking an arbitrary percentage and ignoring the minority who disagree is never a good alternative, and should never be used as the final determination. In any case, these numbers should not be on this page, nor stated in the way they currently are "consensus is a supermajority". It has never meant that, and whoever added that to the page should've been reverted immediately. It's a complete redefinition of how Wikipedia should work. It shouldn't be taken so lightly. — BRIAN0918 • 2005-11-27 18:13
I concur with Brian. Mindspillage (spill yours?) 18:38, 27 November 2005 (UTC)
<aol>me 2</aol> (But you already knew that) Kim Bruning 21:11, 27 November 2005 (UTC)
Alright, I've replaced the supermajority numbers with the links to the respective articles which discuss these numbers further. They never belonged on this page to begin with. If anywhere, they would belong at something like Wikipedia:Supermajority, since this page is strictly for discussion of Consensus. — BRIAN0918 • 2005-11-27 21:25

Read some of the old discussion on AFD closings There is no need to read old discussions look at the current discussions on Wikipedia:Deletion_review#Decisions to be reviewed. Numbers are mentioned frequently for example they are used in both of the first two nomination:

  • List of Jewish Fellows of the Royal Society
    • Endorsre close. It's 63% for deletion, which makes a close of "no consensus"
    • Overturn either by re-opening afd or putting article on some type of lockdown - unfortunately, I don't know how to proceed in this case. If 75% if max for delete I suppose reopening the article is the only way to proceed - again, I'm not sure.
  • Interchanges on Ontario provincial highway 401
    • Valid AfD (30 deletes/11 keep/9 merge/1 anon)

Brian0918 part of the reason for keeping the numbers on this page is because they show how as the subject becomes more important, what is taken as a consensus rises. You may not be aware of the split in the community over the use of diacritics in article names. If you are not then please have a look at Wikipedia talk:Naming conventions (Norse mythology). Here is a case where deciding on what represents a consensus is necessary. There is no way that many editors are going to change their views on this issue and few seem willing to build a compromise somewhere in the middle. Believe me I've been trying both on that page and on WP:UE to find a compromise, but few have actively taken to the idea as yet. The debate after the attempt to build a consensus on Norse mythology guideline talk page shows why such a paragraph on this page is useful.

Brian0918 What I fail to understand is why you object to these numbers appearing here as they are widely used on the pages named and putting them here severs several useful purposes which I have mentioned above and which you have not directly addressed.

I am going to revert the edits because you have not shown that you have a consensus to remove the figures which appear on this page. BTW what would consider constitutes such a consensus? --Philip Baird Shearer 23:24, 27 November 2005 (UTC)

  • I very much doubt that there was ever a consensus to add the numbers to the page in the first place. This page is about consensus, NOT supermajority. Because of the recent changes, we are actually redefining consensus as supermajority. It is clearly stated that supermajority is not the preferred method, but merely a method people have been known to use without being contested. So those same people stick the numbers on this page and give themselves something to follow. It's a completely circular reasoning. "I say consesus is 75%." "Sounds good to me, let's add it to the guidelines." "Consensus is 75%, see WP:Consensus". If you want to elaborate supermajority at Wikipedia:Supermajority, that's fine, but people here seem to be in favor of removing these numbers, and unless you can provide some reasoning for why they should be in the article, we should just agree to disagree and leave them out. After all, the respective articles on RFA, AFD, etc, detail these supposed "supermajority" numbers, so why do they need to be detailed here? Most people will just come here, skip all the other information, and pick out the numbers. That's how it got started, and that's how it remains popular. It's much easier to say "consensus is 75%" than to actually read Wikipedia:Consensus and think about it. — BRIAN0918 • 2005-11-27 23:34
  • Also, you have just reverted out a number of changes unrelated to what we are debating. This proves my point exactly. It's much easier to find your name in the history and rollback to that than to actually read what I changed and revert what you disagree with. — BRIAN0918 • 2005-11-27 23:37
  • These numbers weren't even discussed before being added into the article originally. They were added in to make a point, and were not meant to be taken seriously, and yet, now they are being taken seriously. It's ridiculous. They were just added in, and nobody bothered to revert. They should not be there by default until being discussed. — BRIAN0918 • 2005-11-27 23:44
    • I don't think it's appropriate for the numbers to be here. Having them here seems to give them more weight than they have, as per Brian; better to mention that sometimes numerical guidelines are used to determine rough consensus and leave mention of the standards for each page to those pages. This page should be a higher-level page, anyhow; a listing of what goes on at each individual page on which consensus is called for shouldn't necessarily be here anyhow. Mindspillage (spill yours?) 23:54, 27 November 2005 (UTC)
      • Agreed they should not be mentioned here. This is a different policy. We don't debate here possible supermajority numbers used on RFA, AFD, etc, so why should they be listed here? They were added as a joke/point in the first place, without being discussed at all, so they should be removed as such. — BRIAN0918 • 2005-11-28 00:08
No, I read you changes and considered rolling back to the first version you wrote. It would have been my prefered stratagy to show my willingness to work with you on this issue, but I did not, because it included the links to ADF etc twice which I am sure we would both agree is unnecessary.
"but people here seem to be in favor of removing these numbers, and unless you can provide some reasoning for why they should be in the article, we should just agree to disagree and leave them out." I posted the question here for several days before I re-inserted the text which had been removed, so when I posted it no one objected to me doing so. I have however objected to it being removed, so you do not have a consensus to remove it and I can just a legitimatly claim "we should just agree to disagree and leave them in".
You say "After all, the respective articles on RFA, AFD, etc, detail these supposed "supermajority" numbers". I am not very familiar with WP:AFD please can you point me to a section in the ADF guidelines where there is a detail a number? --Philip Baird Shearer 23:55, 27 November 2005 (UTC)
Please stop skipping what you do not want to reply to. It doesn't work like that. These numbers were added here as a joke/point, and instead of being reverted (maybe because you agreed with their undiscussed addition), they are now embraced as guidelines. You cannot ignore this. They should never have remained in here without discussion in the first place. That is never how it works on Wikipedia. Things are discussed before they are changed. This was never discussed before being added, and at least 5 people here have stated their opposition to its addition. — BRIAN0918 • 2005-11-28 00:01
As for where on AFD: Wikipedia:Deletion_guidelines_for_administrators#Rough_consensus. The number is not detailed specifically, but that is where it should belong if it is to be detailed. Why would people discuss changing the percentage for supermajority on AFD here? — BRIAN0918 • 2005-11-28 00:06
Actually, last time I looked at that (AFD policy), I complained, providing the above example as one of the reasons. The numbers were removed. Perhaps my complaint has since been archived and history may have had sufficient time to repeat itself.
In other news, there's no consensus to keep the numbers in the article either, as you may have noticed. It's under contention. Now we have to negotiate some outcome. Kim Bruning 23:59, 27 November 2005 (UTC)

"Please stop skipping what you do not want to reply to" What have I skipped? I thought I had answered everything. Yes we have to negotiate some outcome. First of all why not make the changes you want which does not include removing the numbers and we will see if we can use that as a basis for further chages. However I have other things to do for the next 18 hours or so. So I'll come back to this later today. --Philip Baird Shearer 00:16, 28 November 2005 (UTC)

Here is the part you are repeatedly skipping: the numbers were added to make a point without any previous discussion. They should not be there by default without being discussed, as it clearly says in {{guideline}}.BRIAN0918 • 2005-11-28 00:20
So there was notice on the talk page for a week before I before I were reinstated the numbers, no one choose to object, so I take that to be a consensus for reinserting them. There was no discussion before they were removed, and I objected to the removal within hours and reverted to the previous version, which means there was not consensus for the removal. --Philip Baird Shearer 20:06, 28 November 2005 (UTC)

Section break

This is interesting. Brian0918 is correct about the numbers being added to make a point. That point was to satirize the Bizarro Consensus that too often supplants the real thing around here. I was trolling, and apparently doing it very well: nobody said a thing about it for six months, and now others are arguing, in all seriousness, in favor of a position that I only meant to mock.

The numbers should stay out. They seem to have opened some eyes to the troubling behavior of operating by majority vote and calling it "consensus", but they've also strengthened the very behavior they were intended to correct. Many people missed the point of A Modest Proposal, but nobody seriously tried to implement it. —Charles P. (Mirv) 06:01, 28 November 2005 (UTC)

  • I think your point would have been better made with more satirical wording. I caught some of it, but still thought it was a serious effort. So, we are up to about 6 people in favor of its removal. Is that enough of a consensus? ;) — BRIAN0918 • 2005-11-28 06:36
    • No, we should definitely edit war over it some more. Seriously though, I don't see why this is a big deal. Whether or not percentages are listed on this page won't influence the way people work at AFD, RM or RFA. Radiant_>|< 09:38, 28 November 2005 (UTC)
      • On the contrary, since Mirv randomly added those numbers here, this page's numbers have been cited ad nauseum as rationale at RFA, AFD, and RM. — BRIAN0918 • 2005-11-28 09:52
        • I'm not really familiar with RM, but to the best of my knowledge both AFD and RFA have used those numbers as rough but non-binding guides for far longer than the half year that they have been here. Consider it a meme. Radiant_>|< 10:15, 28 November 2005 (UTC)
          • The people may have used the numbers, but I don't think they were actually written out as part of the guidelines. — BRIAN0918 • 2005-11-28 10:31
            • But given that guidelines are supposed to be descriptive rather than presciptive, does it not follow that if people generally use a certain criterion then it should be added to the guideline? Radiant_>|< 10:39, 28 November 2005 (UTC)
              • Provided that it is discussed first, and placed in the proper guidelines page, yes. In this case, neither is true. — BRIAN0918 • 2005-11-28 16:16

Brian0918 you write above "but I don't think they were actually written out as part of the guidelines" but they are for both WP and RFA, and you are on record up above as saying: "After all, the respective articles on RFA, AFD, etc, detail these supposed "supermajority" numbers, so why do they need to be detailed here" and "Actually, last time I looked at that (AFD policy), I complained, providing the above example as one of the reasons. The numbers were removed." I wish you would keep to one version of the facts.

What do you consider a consensus to be if it is not the sort of consensus gained in cabinet government? I have somemore questions for you, but I need to understand you position on this specific issue if we are going to agree a compromise.

I am going to reverse the changes made since my last reversal, so that you can do what I suggested before: "Yes we have to negotiate some outcome. First of all why not make the changes you want which does not include removing the numbers and we will see if we can use that as a basis for further chages." --Philip Baird Shearer 20:06, 28 November 2005 (UTC)

  • Please don't quote other people's comments as my own. Consensus is thoroughly discussed in the current article, and the discussion so far here has been in favor of this non-number version, so, rather than go against everyone else, how about actually starting to reply to people's comments individually? That is usually how a discussion occurs. — BRIAN0918 • 2005-11-28 20:11

I am having problems with a router in Holland so the edit below was written before your last posting so if it is not quite in sink with you last one my apologies:

Sorry for attributing a comment to you which was someone else. As you will notice I have struck it out, but not removed it so others can see my mistake and you request for me not to make them. What do you mean "discussion so far here has been in favor of this non-number version". I am not in favour of this non-number version so how do you measure that the discussion has been in favour? Which comment in particular do you think I have not replied to? What do you consider a consensus to be if it is not the sort of consensus gained in cabinet government? --Philip Baird Shearer 20:44, 28 November 2005 (UTC)

"I am going to reverse the changes made since my last reversal" - this isn't how consensus works. Have you considered allowing other editors to contribute and discuss their changes here? - Tεxτurε 20:35, 28 November 2005 (UTC)

Now to answer the last point. The version I am reverting to was the last version for which there was a consensus (or at least no objections at the time it was put onto the page). The version you Brian0918 reverted to is a version for which there are objections. Why do you does Brian0918 not do what I suggested and edit the version I am reverting to so those none controversial changes Brian0918 mentioned are added to the page first so that we can use that as a base to start to build a version we can all live with? --Philip Baird Shearer 20:44, 28 November 2005 (UTC)

You make a good point. But I haven't reverted anything. Are you speaking to someone else? - Tεxτurε 20:50, 28 November 2005 (UTC)

Less hast more speed. The router seems to be behaving its self again now. Yes you are qutit right. I got confused between you comment and Brian0918's revert, and with the router playing up I did not have current version of the article history availible to check the history for who had done the revert. Foolishly I assumed that the revert and the comment were by one and the same person. I tried to patch my comment once but it was not clear. So I have done it again. I hope it is clearer now. --Philip Baird Shearer 21:05, 28 November 2005 (UTC)

  • Lack of discussion for a version is not consensus for a version. That's ridiculous. You are currently the only one objecting to the current version, and you are simply saying that you are against the numbers not being there, without making any attempt at discussing it at all. Meanwhile we have been thoroughly discussing it and you don't bother to reply to any of our points. The changes you are espousing are the controversial ones, everyone except you is currently opposed to them, and they were never discussed in the first place before being added. In fact, they were added as a joke to make a point. — BRIAN0918 • 2005-11-28 20:54

Which point of yours have I not answered? What do you consider a consensus to be if it is not the sort of consensus gained in cabinet government? --Philip Baird Shearer 21:07, 28 November 2005 (UTC)

  • Please reply to everyone's messages individually above before going on with this tangential discussion. — BRIAN0918 • 2005-11-28 21:13

Page Protected, Let's Talk Here

Things seemed to be getting out of control, so let's continue to discuss things here before reverting anymore. karmafist 21:59, 28 November 2005 (UTC)

BRIAN0918 I have looked through the page and given that I have contributed a lot to the discussions I am not sure what else I can say without repeating myself.

  • See my comments in the section #Rough consensus. "So I hope that the page does not come under attack from those who would like to idealise the process and not describe how conflict between reasonable Wikipedia editors is resolved in practice. To do that would be unfair on people new to Wikipedia (those most likely to read this page) who do not yet know how a Wikipedia rough consensus is arrived at in practice"
  • See also my comment: "The fact is that in reality different thresholds are taken to be a rough consensus depending on the severity of the process. A page move is not as critical as a page delete and a page delete is not as critical as granting adminship, this should be mentioned here and probably quantified with numbers as well." (23:38, 14 October 2005)
  • See the comment from lots of other people on the issue of what consensus is for example older≠wiser 13:55, Jan 20, 2005 or -Splashtalk 01:54, 15 September 2005 (UTC)
  • There numbers were in the page for six months and must have been read by a lot of people in that time no-one thought strongly enough that they were incorrect and should have been removed, or even flag direct criticism on the talk page.
  • After they were removed, I added paragraphs to the discussion and waited a week before reinstalling the deleted section no one objected. You say "Lack of discussion for a version is not consensus for a version. That's ridiculous." Why is it ridiculous? If no one had raised an objection when Kim Bruning removed the numbers on the 26 November would you argue that they should be restored as "Lack of discussion for a version is not consensus for a version. That's ridiculous."? Why could he not have said that he wanted to remove the numbers and then waited to see if anyone objected. Which is precisely what I did in reverse.

I think a good place to start re-editing the page would be if you would allow me to revet the page and then you add all the other edits that you think are necessary without deleting the numbers so that we can have a new base line to start to build a consensus from. As I said before I considered doing that with you edits before you removed the numbers [2] but as there were double links in there it did not seem like the best version to keep. but in that version I have some real issues for example I much prefer the words "Precise numbers are hard to establish, and Wikipedia is not a majoritarian democracy, so simple vote-counting is not the key part of the interpretation of a debate." over your new words "Precise numbers for "supermajority" are hard to establish, and Wikipedia is not a majoritarian democracy, so simple vote-counting should never be the key part of the interpretation of a debate." What should and what is are two different things.

Now if there are any other specific point you would like me to address I am more than willing to do so. Just as I would like you to answer my question: What do you consider a consensus to be if it is not the sort of consensus gained in cabinet government? --Philip Baird Shearer 23:19, 28 November 2005 (UTC)

  • Yes, please address this page, for one: Wikipedia:No_binding_decisions . :) Kim Bruning 00:34, 29 November 2005 (UTC)
    • Hello page nice to meet you! Of course they are not binding. But IMHO you should have sought a consensus before making a major change to this page on this talk page as is suggested on the template at the top of the article "but please use the discussion page to propose any major changes". But that is water under the bridge. Lets work together to try to reach a compromise on this issue so that the page can be unlocked and an edit war is not resumed. As you can see from above I have started to addess the new clauses added perhapes we can concentrate on that edit first. Philip Baird Shearer
      • Yes, I did, I reviewed the page, saw only one dissenter in 6 months, so I deleted. Since putting percentages in is a really really really bad idea, it's also sort of common sense to remove them. I'm sorry I didn't do so 6 months ago, but I'm an eventualist, and I was sort of busy trying to fix dispute resolution first. I'm done with that now. So here I am. :-) Kim Bruning 15:59, 29 November 2005 (UTC)
      • The original addition of the numbers on the page was never discussed in the first place. Please, stop ignoring that and trying to act like they should be there by default and be discussed to be removed. They shouldn't be there by default until active discussion leads to the consensus that they should be there. — BRIAN0918 • 2005-11-29 18:09
  • Err... "the sort of consensus gained in cabinet government"? Isn't that where Cabinet members leave if they don't stand whole-heartedly behind the agreed proposal, and the proposals are written and essentially can be pushed through by the Prime Minister, if necessary acting alone? Are you sure you mean that?
    James F. (talk) 00:43, 29 November 2005 (UTC)
    Decisions can not be pushed through by the Prime minster (or someone in a similar position, depending on the Government). All agreements are gained by reaching a consensus, and yes if anyone disagrees with the consensus they have to resign from the Government. Indeed there is an interesting constitutional issue at the moment in Britain. Because the smoking ban tiff in the cabinet you will see from this article how to remain in government Ms Hewitt has to support the Cabinet position. That Sir Liam Donaldson has spoken out against the compromise is very unusual. He justifies this on his use of a medical prerogative, which is a new one under the usual Cabinet rules. Unlike cabinet government there is no carrot or stick for reaching a consensus in a Wikipedia consensus, so the best that can be hoped for in most cases is a "rough consensus". If one accepts that then putting numbers to a rough consensus is not an unreasonable position. As we all know very few people change their positions once they have made them known during the consensus gathering with the RFAs, AFDs, WPs nomination. With a limit on time available to come to a decision during a RFAs, AFDs, WPs nomination, and so few people changing their initial position, the proposition that the majority of opinions are anything but votes is in my opinion disingenuous. Take the recent RFA for Haukurth Wikipedia:Requests for adminship/Haukurth only two(?) people out of 82 people changed their position, over what was a divisive issue. Philip Baird Shearer
That's Haukurths' problem frankly. I've been at several RFAs and AFDs where people turned them around a full 180 degrees. Sometimes even twice. This has become somewhat trickier since some not-so-experienced people started vote-counting, however. What's happening is that more people are entering the system than that we can train fast enough to be able to use it. Kim Bruning 16:13, 29 November 2005 (UTC)
  • I'm still waiting for you to reply to the fact that the content was added by Mirv as a joke to make a point, without any previous discussion, and is now requested by that very person to be removed. Reply to all of these points satisfactorily, and I think we can make some progress. -- BRIAN0918  01:52, 29 November 2005 (UTC)
    • The motive for the original inclusion are not realy pertinent. Whatever the initial motives for placing the numbers on the page, the people who read and edited them in good faith did not know those motives, so they remained there without removal for many months. Further before I reinstated the text I published the fact on the talk page that I was going to do so and no one objected. Now that I have answered that question please answer mine: What do you consider a consensus to be if it is not the sort of consensus gained in cabinet government? --Philip Baird Shearer 11:09, 29 November 2005 (UTC)
    • You have not answered every point. The first time they were put they, they were added without any discussion whatsoever. Then discussion occurred to remove them, and they were removed. Then you wanted them back, so you posted on the talk page, and again, no discussion occurred, so they were added back. On whose user talk pages did you post a notice that you were going to be adding them back, so that you could discuss it first? (since its inclusion had yet to ever be discussed) — BRIAN0918 • 2005-11-29 18:06
      • Why would I need to post it to anyone's page? If they are interested enough in the article it will be in their watch list. Now that I have answered that question please answer mine: What do you consider a consensus to be if it is not the sort of consensus gained in cabinet government? --Philip Baird Shearer 18:27, 29 November 2005 (UTC)
        • I have no idea what the consensus gained in a cabinet government is, I do know what a consensus on a wiki is. Tickle me pink if this isn't a wiki :-) Kim Bruning 18:48, 29 November 2005 (UTC)
        • You have not replied to every point. The first time the numbers were put in, they were added without any discussion. — BRIAN0918 • 2005-11-29 18:55
  • Simple solution. Who added it and who wants it gone is irrelevant.
    1. People do use those percentages on an almost daily basis.
    2. WP guidelines are descriptive.
    3. Therefore, this guideline should describe the percentages that people use.
    4. A caveat that they aren't set in stone would be a good idea.
  • Radiant_>|< 11:58, 29 November 2005 (UTC)
The use of the percentages is erroneous, and prone to DOING VERY BAD THINGS(tm). That's why it's erroneous. So we are sort of trying to stamp them out. If we take them out "there", people point to here. If we take them out here people point to "there". It's catch 22. What are we to do? Kim Bruning 15:59, 29 November 2005 (UTC)
  • I agree that percentages can be problematic and I am wholly averse to tightly sticking to them. Let's see. There's three examples mentioned here (I'm unfamiliar with how e.g. WP:FAC does it).
    • Requested moves is probably easy. It says 60% but that could easily be replaced by, say, "a good majority". Renaming pages isn't that controversial and should generally be done according to people's wishes. I believe RM is a mess in general and have been meaning to check that out.
    • RFA is really up to the 'crats, since it's their judgment call. I am not sure how exactly they make decisions if a nomination is not obvious. Unless we can get a 'crat to comment here, the guideline here should simply point to the 'crats.
    • However, the AFD criterion is part of the general cesspool that is deletion. Lack of an objective criterion would increase the faction wars. This is of course annoying, but several admins that close AFDs are biased, so each side would want to keep the other in check and a simple criterion is the easiest way of doing so.
  • Radiant_>|< 16:46, 29 November 2005 (UTC)
    • For every problem, there is a solution that is simple, logical, elegant, and wrong. ;-) Kim Bruning 17:36, 29 November 2005 (UTC)
    • Why would we list in this guideline the supermajority numbers for other wikipedia processes? If they are supposed to be listed somewhere, they should remain on their individual process guideline pages, not on a completely different guideline page. If they are to be listed together on a related page, Wikipedia:Supermajority would be a more accurate title for that page. This prevents the redefinition of consensus as supermajority, the main problem that was plaguing this page. This page is about consensus, not about supermajority. That is why we have a section on Consensus vs. Supermajority. — BRIAN0918 • 2005-11-29 17:59
BRIAN0918 what do you think are the objective criteria used for deciding on a Wikipedia:consensus? --Philip Baird Shearer 18:32, 29 November 2005 (UTC)
A good objective criterium is "When it doesn't get edited or reverted". Note that consensus on a wiki can be very fluid. It can hang at one point for 6 months, and then suddenly shift. <innocent look> Kim Bruning 18:48, 29 November 2005 (UTC)
I think the quote at Wikipedia:Consensus says it best, and makes it clear that consensus is based on discussion to lead to conclusions, not votes, which only lead to people having to back up their vote (rather than considering to revise it) and other people corralling others to vote with them. — BRIAN0918 • 2005-12-1 20:37
And what happens when after a discussion stakeholders can not reach a true consensus, particularly for things such as AFD? --Philip Baird Shearer 01:22, 2 December 2005 (UTC)
I am not here to discuss AFD policies. This page is about defining and describing consensus. Personally, I would still leave it up to the person closing the AFD to decide either way (as long as that person didn't also vote in the AFD), since that person doesn't have a vested interest in backing up his initial vote, whereas all other voters do. But let's not get sidetracked. — BRIAN0918 • 2005-12-2 01:29
This is a discussion about a Wikipedia consensus not a dictionary definition of consensus. In Wikipedia consensus building there comes a point were not further discussion will bring a true consensus any closer. In these cases the participants need to be able to judge if a rough consensus has been reached. What quantifiable methods do you think they can use to decide if a consensus has been reached? --Philip Baird Shearer 02:01, 2 December 2005 (UTC)
What if consensus is binary? Kim Bruning 02:50, 2 December 2005 (UTC)
I think using that quote will solve nearly all problems:
In fact WP's standard way of operating is a rather good illustration of what it does mean: a mixture across the community of those who are largely agreed, some who disagree but 'agree to disagree' without disaffection, those who don't agree but give low priority to the given issue, those who disagree strongly but concede that there is a community view and respect it on that level, some vocal and unreconciled folk, some who operate 'outside the law'. You find out whether you have consensus, if not unanimity, when you try to build on it.
Of course voting can be used to let people know where everyone else stands, to act as another form of input, but the policies and guidelines regarding voting should be laid out somewhere else, not here. I see voting on here as being more like that in 12 Angry Men. They would take quick votes to see who else doesn't agree, allowing the discussion to be focused on those people. — BRIAN0918 • 2005-12-2 03:22

To repeat myself: I hope that the page does not come under attack from those who would like to idealise the process and not describe how conflict between reasonable Wikipedia editors is resolved in practice. To do that would be unfair on people new to Wikipedia (those most likely to read this page) who do not yet know how a Wikipedia rough consensus is arrived at in practice. The fact is that in reality different thresholds are taken to be a rough consensus depending on the severity of the process. A page move is not as critical as a page delete and a page delete is not as critical as granting adminship, this should be mentioned here and quantified with numbers as well. --Philip Baird Shearer 17:05, 3 December 2005 (UTC)

Now that the page is unprotected I am going to revert it. As I have suggested several times before please make whatever changes you think are not controversial first (ie do not immediately remove the numbers) and let us see if we can build a consensus from a new base. --Philip Baird Shearer 17:13, 3 December 2005 (UTC)

The numbers do not belong on that page. I have restored an earlier version without the numbers. --Tony Sidaway|Talk 07:38, 6 December 2005 (UTC)

I have read your previous contributions to this talk page and I disagree with you. I think the numbers do belong on the page. How for example does one decide what the consensus is if there are a hundred contributors to a consensus building process?
In the day to day running of WP:RM how does one decide if a page should be moved or not moved, if only three people express an opinion and they all do not agree. If two are in favour of the move and one is against should he page be move or not? If as there are at the moment there is a backlog of 75 outstanding requested moves, if each and every decision for a WP:RM has to be made by agreeing to agree a consensus, then the system would become unmanageable. Few administrators are going to want to move pages in the backlog if they are going to be involved in dozens and dozens of exchanges explaining why they consider the consensus on a particular move has or has not been reached.
Without numerical guidelines it comes down to the whimsical opinion of the administrator who has to make the decision. Now that administrator may be honest and always apply the same criteria when making a decision on what constitutes a consensus. But is a consensus to move a page the same as a consensus to delete a page or should the consensus for the latter be closer to a true consensus? If so how much closer? --Philip Baird Shearer 10:15, 6 December 2005 (UTC)
If your right honorable opponents were to believe that numerical guidelines were useful, they'd have pushed for a poll. In fact, if you think numerical guidelines are it, why aren't you pushing for a poll? <innocent look> Kim Bruning 21:54, 8 December 2005 (UTC)
Because I am working towards the true consenus over this issue :-) --Philip Baird Shearer 14:53, 10 December 2005 (UTC)
<grin> Hallelujah! Kim Bruning 16:07, 10 December 2005 (UTC)

How to handle a situation of majority overruling NPOV?

What recourse is there for a situation where a majority (not even a consensus) is allegedly violating presentation neutrality? Isn't the {npov} template added to an article when there is an in good faith neutrality dispute? What if this majority then claims there isn't even a dispute and removes the tag? zen master T 01:58, 30 November 2005 (UTC)

Wikipedia is not (supposed to be) run by majority. If there's a majority trying to harm NPOV, you can especially ask for one of the older users to come in and knock heads together. (I'm a bit worried about some of the less experienced users and admins these days, they seem to think majority is right or something) Kim Bruning 04:57, 30 November 2005 (UTC)
I liked the "knock heads" part :) +MATIA 13:43, 30 November 2005 (UTC)
If you have time, I hope you'll stop by the conspiracy theory talk page. Your input would be welcome. Read the archives, familiarize yourself with the discusssion, and knock heads as you think appropriate. Tom Harrison (talk) 13:54, 30 November 2005 (UTC)
Yes, can some real outside people look at the Conspiracy theory article? I feel like I am being toyed with over there. I believe the {npov} template should be added to the article given the neutrality dispute(s), but your interpretation of the situation is your own. zen master T 22:28, 30 November 2005 (UTC)

What is the deletion threshold for a template?

It doesn't seem to be documented anywhere. Should it be 50%? 60%? 67%? 75%? 80%? (Those choices taken from both this article and the apparent current practice.)

  • I vote 67% if at least two people are using it, 60% if one person is using it, and 50% if it isn't being used by anyone. James S. 21:15, 13 January 2006 (UTC)
LOL! And you ask this on the page for *consensus*? :-) Wikipedia is NOT a democracy. Have a nice day. Kim Bruning 23:56, 13 January 2006 (UTC)
I am hoping that people will share their honest opinions. All the other namespace deletions have documented threshold values. Please be kind. --James S. 00:09, 14 January 2006 (UTC)
  • Depending on the current position of the moon, it is either 65% or 43%. As the average age of the involved individuals decreases, the percentage should increase. — 0918BRIAN • 2006-01-14 00:04
Ha, ha. Assuming that the moon is below the horizon and everyone is about the same age? --James S. 00:09, 14 January 2006 (UTC)
Then try for consensus. (erm, just in case, I'm not being obtuse. Just, um, could I maybe nudge you to read the rest of the talk page, right above your question? :-/ ) Kim Bruning 00:14, 14 January 2006 (UTC)
I've read it; I have been trying to shepherd a compromise for a long-running and apparently nasty dispute involving a deleted article -- which shares some of the problems with Conspiracy theory. But what does that have to do with the threshold for template deletion? If you don't like majority (50%) then what do you like? --James S. 00:21, 14 January 2006 (UTC)
Wikipedia:Consensus (see also: Consensus) .You may have heard of it. It's described around here. (If you're truely lost, even after reading those two links, please drop a note on my talk page, and I'll scratch my head and figure out how to explain... somehow.) Kim Bruning 02:33, 14 January 2006 (UTC)
I am trying to ask a civil question. To quote from this article:
* 75-80% or larger majority support for a Wikipedia:Request for adminship
* two-thirds or larger majority support for Wikipedia:Articles for deletion
* 60% or larger majority support for Wikipedia:Requested moves
If there is a reason that template deletion should not be subject to similar guidelines, please be civil and let me know. --James S. 02:57, 14 January 2006 (UTC)
These numbers are the opinion of one individual who keeps redefining consensus and readding the content to the article. — 0918BRIAN • 2006-01-14 04:13
Really? Are you serious? Without numeric standards, everything becomes completely subjective and consensus devolves into whomever has the most influence, time to argue or mold the debate to their position. Even the IETF counts by show of hands. This is completely nuts. Those deletion thresholds aren't just here, they are in the instructions presented to new users such as myself who read the introductory material about the deletion process. I'm astonished, and disappointed. I had hoped that there was a quantatative support for decision making. How sad. --James S. 04:44, 14 January 2006 (UTC)
  • These processes are not votes. They are discussions. Discussions don't have hard numbers in them, because one well argued position can be significantly more important than multiple poorly argued ones. We are not the IETF, and they are not a beacon of consensual decision making (which is a fine art not easily learnt). -Splashtalk 04:47, 14 January 2006 (UTC)
  • Wikipedia is not a democracy. Consensus was never about having the most random current voters to agree with you. Wikipedia is always changing, so the "current editors" are of little value as numbers for votes. Their opinions and rationale are what is of value. — 0918BRIAN • 2006-01-14 04:48
Huh? What about the numeric methods described in detail on Wikipedia:How to hold a consensus vote or the "60%" clearly indicated on the lead section of WP:RM? Are those also the work of a lone rebel numericist? Is this some kind of a joke? --James S. 04:50, 14 January 2006 (UTC)
Please do spend some time reading through some controversial AfDs and TfDs. RM works on consensus just the same as everywhere else. This is no joke at all. -Splashtalk 04:52, 14 January 2006 (UTC)
I have been reading a lot of recent TfDs carefully, and I thought I saw the 2/3rds AfD standard, which is mentioned elsewhere, by the way. What do you think brought me here. I had such high hopes. --James S. 04:56, 14 January 2006 (UTC)
That page you cite, "How to hold a consensus vote" was a proposed policy that was ditched. It should be clear why. — 0918BRIAN • 2006-01-14 05:08

Look, I fully understand the need for flexible standards in votecounting to give the tabulator some discretion, but I thought for sure this would be.... Oh, nevermind. This is just silly. --James S. 05:00, 14 January 2006 (UTC)

It is not primarily votecounting, that's the key barrier to cross. Ok, look at it this way. Around Wikipedia, the ruling determinant is 'rough consensus' (or in some cases a 'general consensus'). Consensus is best determined through an experienced editor reading a debate, and reaching an informed, neutral decision. There are fuzzy levels outside of which the community is generally likely to question a decision which is not given a thorough, convincing reason. For most deletion debates, the lowest threshold most people will accept on pure numerics is, approximately, two-thirds. Many admins will not see that as a decent baseline, and some prefer towards 80% agreement. But those are just numbers. They can and are overridden by the contents of the debate, and an admin is by no means bound by magic numbers in reaching their decisions. If some editors are clearly being silly, for example, they can expect to be given the appropriate treatment in light of others not being silly. The numerics are the merest of guidelines, and are, very approximately, those that will, probably meet with general "you are within discretion" comments from the community. Admins may move outside mere numerics if they can see reason to do so. The debates are not votes and the numbers are not the prime determinant. -Splashtalk 05:04, 14 January 2006 (UTC)

As James S said above "Really? Are you serious? Without numeric standards, everything becomes completely subjective and consensus devolves into whomever has the most influence, time to argue or mold the debate to their position" --Philip Baird Shearer 00:36, 15 January 2006 (UTC)

Get rid of percentages

All the discussion about numbers leads me to wonder if Wikipedia should move in a very different direction about consensus, perhaps trying to emulate tradition formal consensus decision making in a modified form. In formal consensus, ONE PERSON can block consensus. Because of this, there is the impetus for the majority to listen to the views of the minority and hear their concerns and act on them. When we start attaching numbers to the process, the discussion devolves into "I'm right, you are wrong".

If there is a one vote block, then people are very reluctant to block. You must be willing to stand up and say, "Everyone is about to make a very big mistake, and for the sake of everyone I must block it." Blocking something is considered a very rare event. People who block often loose all their credibility. People who block must be willing to argue their case. All these things are lost when we try to put numbers and percentages on the process.

So, here is what I propose: Get rid of ALL percentages. Change the process so people are to state "Approve" and "Have concerns". If someone says, "Have concerns" they must state what they are, and hopefully, everyone else will try to address them. If the concerns are not stated or the person does not engage in dialogue, they can be discounted. An admin will read through the concerns and decide if they need to be seriously considered. The results would be "Consensus" and "No Consensus". If there is no consensus, there can be an additional week of dialogue to try to work on consensus. At the end of the second week (with extensions if the discussion seems to warrant additional time), someone would "Call for consensus". The response would be "Approve", "Approve with reservations", "Stand aside", and "Block". The Admin facilitating this would have to read through it and decide what to do. They might have to talk with the person saying block. They would have to use their judgement to see if blocks are serious and if the concerns are serious, and not count votes. They would also write a decision explaining the result and this could be challenged, in which case there would be more discussion. I think this would change the dynamic of the discussion and the contentiousness of the debates. The point of consensus is to try and get people to work together. -- Samuel Wantman 07:31, 14 January 2006 (UTC)

At the risk of seeming numericist, I would point out that, as many people above and in similar discussions have recognized, the workability of such a proposal decreases in proportion to the number of participants, which in this case is going nowhere but up. I think failing to provide a clear range of percentages within which the tabulator has discretion is simply asking for trouble. The amount of trouble involved, strictly increasing over time. If you all want to deal with the extra work it involves, knock yourselves out. Just don't ask for my help, and I won't say I told you so. --James S. 21:24, 14 January 2006 (UTC)
I have been following this tact at Wikipedia:Categorization. A new set of guidelines have been worked on for months, and is about to "go live". The process was slow, took quite a bit of effort, but was never contentious, and the issues and opinions of everyone were made clear. Now that we are at the end of the process, things are very quiet. I suspect that many people just don't want to put in the time to work through issues to the point where there is consensus, and leave the issue to those that do. The current voting procedure encourages knee-jerk reactions. Often the same issues come up over and over in multiple CfDs or votes on articles, and they remain unresolved until someone takes the effort to start a discussion on the issue. Then the process I describe above or something like it does happen. The big problem that I see, is that most people have no idea what consensus means, and just think of it as some type of supermajority vote. Unless we make a concerted effort to change that view, and incorporate a true consensus process things risk getting worse and more contentious.
As for dealing with large numbers, I would say if you ask for concerns instead of having a vote, you determine if there already is a clear consensus, or if there are serious issues that need discussion. I see this all the time at CfD. Just as the issues are being made clear, the time runs out and there is a decision that leaves the issues unresolved and many contributors dissatisfied. Take away the word 'vote' and say that the issue will be discussed offline until a consensus is reached. I suspect that large number of people will voluntarily remove themselve from the discussion with the expectations that those most concerned will figure it out. --Samuel Wantman 21:57, 14 January 2006 (UTC)
The easy way to deal with large numbers would be to use the style at WP:RFC, where different "views" are written, and people sign their names in agreement with the view. This way, the number of differing opinions is kept as small as possible, making the weighing of opinions much easier. — 0918BRIAN • 2006-01-14 22:00

Brian0918: How does one "weigh opinions" are some opinions heavier than others? how many people have to hold contrary views on a subject before no consensus is judged to have formed? What happens if two or more people disagree that there or is not a consensus on a previous survey? Seems to me we end up with a recursive problem if there are not ways of deciding in advance what constitutes a consensus for a particular type of Wikipedia survey.

Wantman: without numerical guidelines giving a lot of discretionary powers to the administrator, (What happens if the administrator involved is not disinterested?) and an unfair advantage to people who are used to gaming the system. Also what happens over issues where the community is split over an issue like The WP:UE diacritics dispute (see Ubeda, Zurich etc) or the failed move of the Ivory Coast page. or the Gasoline/Petrol debate or Tram/Street car. ... --Philip Baird Shearer 00:29, 15 January 2006 (UTC)

  • My reply to you is simple. I don't know how simple/difficult consensus will be to determine in every situation, but I know that it won't and should never be as simple as a numerical count. You are arguing to the contrary. — 0918BRIAN • 2006-01-15 00:40
  • My replay is that "good faith" is the currency that drives consensus. If people object to a decision, they can object and resurrect the discussion. Consensus only happens when everyone agrees that there is consensus. It doesn't mean that everyone is perfectly happy with the decision, or that everyone's concerns have been met. It does mean that everyone agrees that the result is the consensus of the group. If people do not operate in good faith, they spend their currency. Their views will be discounted, and their credibility diminished. -- Samuel Wantman 01:18, 15 January 2006 (UTC)

Let's discuss a practical example. The Wikipedia community is split on the use of diacritics in article names, but pages still have to be named. How does one build a consensus on such an issue as the number of people involved in the debate runs to more than 100 and the number who have changed their opinion on the issue could be counted on one hand? The problem is not so much new pages, but the moving of pages that already exist to fit in with one guideline or another because no wording which would unify the guidelines can be agreed upon. So we end up with debate after debate, see the Wikipedia:Naming conventions (Norse mythology) for a good example of the consequences of this. If you can suggest how such a contentious issue can be sorted out without discussion of the number of participants which go to make up a rough consensus, I would be most interested to read it --Philip Baird Shearer 13:40, 15 January 2006 (UTC)

Have the Nordic diactriticists and unicoders considered mediation for their deadlock? --James S. 23:57, 15 January 2006 (UTC)
That's usually the next step to consider when things get that serious. — 0918BRIAN • 2006-01-16 00:16
This type of debate is exactly the type of situation that shows the shortcomings of voting. If there are strong opinions and the numbers are split fairly evenly, it means that there needs to be more discussion, analysis and collaborative effort to come to a creative solution that both sides can accept. A rough outline for this process might be something like this:
  • Solicit opinions
  • Collect information
  • Analyze the problem
  • Agree on criteria
  • Brainstorm possible solutions
  • Create a rough proposal
  • Hear people's concerns
  • Modify proposal to address concerns
  • Call for consensus
  • Resolve blocked consensus
  • Implement the solution
  • Evaluate the outcome
This takes time, but at the end you have a solution that everyone can agree to implement even if they are not 100% in favor. And I should point out that often the result of this is that there is no consensus, but a way to accommodation both sides. -- Samuel Wantman 01:38, 16 January 2006 (UTC)
  • Agreed. This is not a quick solution, but there is no point to quick solutions except for the "instant gratification" factor. The current version of Wikipedia matters little, so it is better to have a long, involved discussion than a quick "shut up and vote". — 0918BRIAN • 2006-01-16 02:19

IMHO neither of you have addressed the example of diacritics. This debate has been ongoing for well over a year. All the steps down to "Call for a consensus" have been done time and again. How do I know there is no rough consensus over this issue? Because whenever it is debated a substantial minority disagree with the majority, which ever way the majority happens to be for any one specific debate. How do I know this? Because people count the number of people who have expressed an opinion and all agree that there is no consensus over diacritics. Yet articles have to have page names, and editors put in requests to move them. I presume that neither of you would suggest that all articles which can have diacritics in their names should be deleted until a true consensus on this issue is reached. So a working rule of thumb has to be agreed for how to address the wordings of guidelines and for individual requested moves, because without theses not guidlines could be editied or pages moved.

Further to date, I am not convinced by your arguments that somehow a rough consensus can be agreed upon, unless there is someway of quantifying what a rough consensus is, and giving examples of how different procedures use different numerical thresholds depending on how critical the issue is helps people decide these issues. --Philip Baird Shearer 12:35, 16 January 2006 (UTC)

Let's do it here

There seems to be a small revert war on this page. So clearly, there is not a consensus about consensus. I believe that the page should state that there is costumary numerical percentages, but I think it should be de-emphasized. To start the article with talk about percentages, misses the important points that need to be stated about consensus. We can continue to discuss the use of numbers, but I think it would be wrong to state any percentages as being firm guidelines. -- Samuel Wantman 01:26, 15 January 2006 (UTC)

  • When people see percentages, they assume that those are the final say, and don't bother to read any of the actual policies/guidelines on the matter. That's how newcomers like James S. get so confused on the matter. The version of the guideline without the arbitrary percentages (percentages which were originally added to this guideline as a joke, mind you), does a perfectly good job of explaining where polling should fit into the picture, and how it should be handled. If people want simple numbers, they can find them on the appropriate policy pages (RFA, AFD). From WP:NOT:
    "Wikipedia is not an experiment in democracy. Its primary method of finding consensus is discussion, not voting. In difficult cases, straw polls may be conducted to help determine consensus, but are to be used with caution and not to be treated as binding votes."0918BRIAN • 2006-01-15 01:37

I have found that when there isn't a consensus it works to state both sides and make it clear that there isn't a consensus. While Brian0918 was writing his comments, I tried to do this. Let us agree to disagree. -- Samuel Wantman 01:42, 15 January 2006 (UTC)

  • As long as the percentages are in there, people will ignore all the rest. This article is about consensus, not about voting. We've been dealing with PBS for the last several months. There is no consensus with him. — 0918BRIAN • 2006-01-15 01:47
  • I've implemented my version of the compromise, which mixes in what you added. Rather than completely ignore the percentages, it points to where people can find information on them, since this page is about consensus, not voting. — 0918BRIAN • 2006-01-15 01:50
I am ok with this, but since I don't like percentages and agree with Brian0918 on this issue, I am not the person who needs to agree. -- Samuel Wantman 02:01, 15 January 2006 (UTC)

As I have told PBS multiple times before, if you want to talk about supermajority, write Wikipedia:Supermajority. There, you can add whatever arbitrary numbers you want, because the article will actually be about supermajority voting. This article is about consensus, not supermajority voting. — 0918BRIAN • 2006-01-15 02:32

I think that saying "That's how newcomers like James S. get so confused on the matter." is very patronising. We are not talking about a consensus on this page. As I have said several times an example of a consensus being used to govern a country appearers in the Westminster System of Government. There are carrots and sticks involved in a Cabinet reaching a consensus which do not apply in a Wikipedia discussion. If for example the loosing side of a discussion were not allowed to edit a page for six months, or if they expressed a dissent from the consensus after a consensus had been reached, a consensus would be achieved and adhered to in many more cases than they do at the moment. But as the rewards and penalties are not going to change in Wikipedia consensus building, (and I would not suggest that they are,) the best that can be hoped for is a "rough consensus", and for that to work in a fair and efficient way one needs to agree what a rough consensus is before embarking on a rough consensus building exercise. I think it is better that the rough consensus is defined in a none ambiguous way as that reduces conflict and is something that most reasonable editors can agree on. Further it helps people new to the talk pages clear guidelines and does not leave them feeling that they have been mugged by the regular contributes to a Wikipedia process. That mugged feeling is something that people have expressed, particularly over the WP:VFD WP:AFD which are not at all explicit. --Philip Baird Shearer 13:26, 15 January 2006 (UTC)

Wikipedia:Supermajority created

As requested, I took a stab at starting out Wikipedia:Supermajority. Please have a look. It's really short at the moment because I'm trying to not rush, and, I don't really know what else should be in it, except for a threshold for deleting templates. With all the userbox controversy, that should most certainly be there. My proposal from the discussion above is copied to Wikipedia talk:Supermajority.

Please help expand it and/or comment on the talk page.

Thanks to all of you who had the patience with my template question here. I respect your idealism, and am glad that there is a clear consensus that in situations where consensus is unlikely to be achieved, such strict idealism should be set aside and a pragmatic numeric range is used for deletion decision making. --James S. 11:03, 15 January 2006 (UTC)


Wikipedia:Supermajority was created specifically to remove the non-consensus related content from this page, now you want to merge it back? AFD, RFA, etc all use supermajority. Ask any bureaucrat. Why else would you want the numbers in this article so desperately, unless they were being used regularly? — 0918BRIAN • 2006-01-16 23:39

  • Removed. --James S. 09:34, 17 January 2006 (UTC)
  • Brian, you're wrong. Both AFD and RFA plainly state that they work on Consensus, not Supermajority, right at the top. And those numbers are being used on a daily basis. We should not tightly stick to them, of course, but since guideline pages are DEScriptive they should contain a description of the status quo, to prevent confusion among newbies. Radiant_>|< 11:37, 17 January 2006 (UTC)
    • While those pages might state that they work on consensus and it may still be held as an ideal by some (however unrealistic), the fact of the matter is that those pages operate on a principle of supermajority voting -- to call that "consensus" is Orwellian double-speak. olderwiser 13:37, 17 January 2006 (UTC)
    • I am not wrong. If you look at the last few bureaucrats elected, they all said that their only rationale for promotion (except in some rare exceptions) was whether some percentage (such as 80%) of the votes were Support. In any case, the understanding of how RFA/AFD work should be left on WP:RFA and WP:AFD, where they can easily be discussed/changed/updated as required, rather than left on an unrelated page, where they may not be as frequently updated to the current standards. This is not the holding ground for other processes' policies. — 0918BRIAN • 2006-01-17 13:49
      • So basically you're saying that the notion of consensus, in most Wikiprocesses, has been, or has to be, deprecated and replaced by a simple percentual limit known as supermajority. Whether that's the case for RFA is really the crat business, but on AFD it is not a good idea. Radiant_>|< 14:49, 17 January 2006 (UTC)
        • No I am not saying that's how it should be. I'm against simple vote counting; it redefines "consensus" as "supermajority". That's why I was against those percentages being on the page, because readers simply skipped reading any of the policy and went straight to the percentages, assuming that they represented what "consensus" means. I specifically asked one of the bureaucrats to define "consensus" on RFA, and he said "80%". — 0918BRIAN • 2006-01-17 14:52
          • People shouldn't be lazy like that, but the simple reality is that closing any AFD as "keep" with 80% or more delete-votes, or as "delete" with 60% or less keep-votes, is likely to be problematic. This needs to be part of this descriptive guideline page, and it needs to be worded to make it clear it's not a strict cut-off point. Also note that RFAs have been known to pass with as little as 70% support. Radiant_>|< 15:03, 17 January 2006 (UTC)
            • Yes, I agree that it should be clearly explained.. on the processes' policy/guideline page, not on the page of a different policy. The current version of this page clearly explains the situation, and points people in the right direction for finding out more specific information. — 0918BRIAN • 2006-01-17 15:07
  • A difference between "rough consensus" and "supermajority" is that a "rough consensus" has criteria beyond a simple count of votes. On AFD at least, I do not impose a fixed limit when closing debates. I usually look for about two thirds delete majority, but my discretion extends past just deciding who's votes are valid. If a concern over verifiability was never answered I'm willing to delete at a bit below two-thirds, if all the "delete" votes say "not notable" and nothing else, and someone provides real evidence of notability I'll put the threshhold a bit higher. Sjakkalle (Check!) 14:59, 17 January 2006 (UTC)
  • For the most part wikipedia moved over to super majority for everything except article content.Geni 15:01, 17 January 2006 (UTC)
    • Er, what? Please provide evidence of that claim? Radiant_>|< 15:03, 17 January 2006 (UTC)
      • WP:RFA,WP:RFAr, WP:AFD,WP:MFD,WP:Sfd, WP:RDV ,WP:IFD The currently arbcom elections in thoery could be decided on a striaght majority (they wont becuase some people have been able to rack up very high levels of support). The orignal votes for the 3 revert rule and the expansion of speedy to cover obvious copyvios were supermajorities.Geni 15:09, 17 January 2006 (UTC)
        • That's not evidence, that's simply a list of processes. Note Sjakkalle's remarks just above regarding AFD; I tend to agree with him. Same applies to MFD/SFD/IFD. You are most definitely wrong about RFAr, and regarding RFA I can now cite Linuxbeak that "RFA certainly seems to be heading towards supermajority, but that is not how we want it to end up. Indeed, I have talked with three other bureaucrats regarding this very manner and we're not exactly liking how it's heading." Radiant_>|< 15:35, 17 January 2006 (UTC)
        • While I may agree with you to more or less a degree about some of those processes, it is not the case for the arbcom elections. — 0918BRIAN • 2006-01-17 15:36
          • It is in fact unknown how the outcome of the so-called ArbCom election will influence Jimbo's decision to appoint people. It is not a given that the people with most support will in fact get appointed; hence, ArbCom doesn't work by supermajority (nor consensus for that matter). Radiant_>|< 15:48, 17 January 2006 (UTC)
            • That's what I said (in less detail). — 0918BRIAN • 2006-01-17 15:51
              • AfD consensus (and other kinds of *fD) should not, IMHO be determined by supermajority rules. The closers I admire most remember "When in doubt, don't delete", and take into account the content of the comments, not just do a nose count. Sjakkalle has it exactly right. I admit that while I think Snowspinner's current proposal on AfDs being influenced by experts is a bit radical in mechanics the idea is sound... someone with bonafides in the topic area saying "this is notable and here's why" should carry more weight than randoms who come by to vote in AfD for fun (I wonder if I'm in that camp???)... the "and here's why" is important, but I'd allow some shorthanding on the why if the person's bonafides were already established. WP:RFAr doesn't fit, the arbitrators have their own process for various steps and it is more like real voting. WP:RFA seems to follow supermajority at first glance but I'd rather think the 'crats exercised their judgement in close calls, which they seem to. Wikipedia is not a democracy nor should it be. IMHO. ++Lar: t/c 16:48, 17 January 2006 (UTC)
Denial isn't going to help the problem. The things I listed are for the most part decided by supermajority. Arbcom elections could in thoery be decided by striaght majorities under jimbo's rules if you consider the theoretical case where only a few people get 50%.Geni 16:52, 17 January 2006 (UTC)
I'm sorry, was that directed specifically at my comment? I would like to think I'm talking about how things ought to be more than how they are. But I think others gave some counterexamples too. Apologies if I'm misunderstanding you. ++Lar: t/c 17:02, 17 January 2006 (UTC)
  • Defining a supermajority as what's necessary for deletion is IMHO dangerous. In AfD, sock/meatpuppet voting is a constant occurrence and having a policy/guideline that states exactly that just gives them ammunition. Supermajority could be used for other processes as far as I'm concerned (policy votes, RFA, etc) but definitely not deletion-related discussions. howcheng {chat} 17:04, 17 January 2006 (UTC)
  • Ditto Sjakalle. The concept of voting is dangerous because it implies that there's an objective way of looking at these things. A "community decision" (I do not believe pure consensus is applied on AfD or RFA; a rough consensus, yes, but not pure) allows for some leeway. I've deleted articles that would be kept under a pure voting system because they were obvious hoaxes or copyvios. I find it unfortunate most people try to apply concepts from real world governance (the right to a fair trial, innocent before proven guilty, freedom of speech, the right to vote, etc.) to Wikipedia, because without some modification, it ain't going to float. I think consensus isn't scaling/working because of Extreme Unction's second law, but being the idealist that I am, I try to believe I am applying consensus to AfD. Certainly consensus still applies on FAC, although I fear it isn't going to scale for much longer. And as for RFA? Well, it's really hard to say. I guess it's a cross between supermajority and consensus. I'm just not sure how it's balanced -- whether consensus overrules supermajority or vice-versa. (I hope nobody got confused by what I'm saying.) Johnleemk | Talk 17:20, 17 January 2006 (UTC)

There was no consesus for Brian0918 to removed the text from this article which is now in wikipedia:supermajority it is my intention to reinsert the figures (yet again) into this article, irespective of whether the new supermajority article is merged into this one. --Philip Baird Shearer 19:14, 17 January 2006 (UTC)

  • What is the basis for your reinsertion of this material--material that was originally added as a joke without any prior discussion on its addition? You have been ignoring the discussion on this page, and only replying when you're about to reinsert your numbers, and even then, your reply simply says "I'm going to reinsert the figures".... This is not how productive discussion works. — 0918BRIAN • 2006-01-17 19:48
Philip: does this mean you'll accept a supermajority, if they ask to remove the numbers? Kim Bruning 21:00, 17 January 2006 (UTC)

Time to make this a loop as we have been here already. GOTO #Page Protected, Let's Talk Here --Philip Baird Shearer 00:05, 18 January 2006 (UTC)

History doesn't repeat itself, it rhymes. At any point in time you have numbers on the page, I get to hold a supermajority vote to take them back off ;-) Therein lies the technical difficulty of wikinomic (I'm considering ways of ending wikinomic in the wikipedia: namespace, but I'm not sure how yet). Kim Bruning 02:13, 18 January 2006 (UTC)

How to reach consensus here

This discussion has become an example of how consensus works. There are two very different sides, and now we have to resolve this issue. Let me state what I think the current situation is. If we can agree that this is the current situation, perhaps we can also agree on how to state it on the project page.

  1. Consensus is one of the founding principals of Wikipedia.
  2. The exact process of reaching consensus has not been clearly defined or agreed to.
  3. In many situations decisions are, in effect, being made using supermajority voting to expedite the process.
  4. Supermajority voting does not have fixed numerical requirements for approval. The numbers are set more by custom than by agreement.
  5. Stating a clear policy about supermajority voting is seen by many as a slippery slope, by which supermajority voting will supplant consensus.
  6. Our effort should be put into strengthening consensus processes.

and I'd like to add one that I think is very important:

  • Even when supermajority voting is used, it should be seen as a process of 'testing' for consensus, rather than reaching consensus. The stated outcome is the best judgement of the facilitator (often and admin). If there is strong disagreement with the outcome from the Wikipedia community it is clear that consensus has not been reached.

So let us try to first work together to see what we can all agree with, and then we can talk about how we want to say it on the page. -- Samuel Wantman 21:12, 17 January 2006 (UTC)

  • What you say above sounds like a plausible reason for merging the supermajority page, since it's used to test for consensus, and not as a stand-alone process. My point is mainly this: we have convention that certain percentages are (and are not) acceptable signs of consensus. New users frequently ask after them. This is a descriptive guideline page, and must therefore explain what those percentages are, as well as why you should not rely upon them blindly. Radiant_>|< 21:34, 17 January 2006 (UTC)
    • It is plausible to explain where voting stands, but specific percentages should not be given in this article. They should be in their own RFA/AFD policy pages (as they are). This is a guideline about consensus on Wikipedia, not a voting policy about RFA and AFD. — 0918BRIAN • 2006-01-17 21:42
      • It's an example of how a theoretical concept is used in practice. Radiant_>|< 23:16, 17 January 2006 (UTC)
  • Are you both saying that you both agree with what I have written? I'd like to start with reaching consensus where we agree. An important part of the consensus process is finding common ground. -- Samuel Wantman 22:53, 17 January 2006 (UTC)
    • Yes, I agree with what you said. Oh yes and please add that consensus does not equate to unanimity. Radiant_>|< 23:16, 17 January 2006 (UTC)
      • Consensus does mean unanimity, that is how Cabinet Government under the Westminster System works. People can argue against a proposition but once it is agreed to then they either have to resign from the Cabinet or support it. Because of the repercussions for all in the Cabinet, resignations are not usually in the interests of the stake holders particularly the person or persons who resign, so usually a compromise is reached. There are exceptions which prove the point, the resignation speech of Geoffrey Howe from the Thatcher Cabinet was a catalyst in her downfall. As a Wikipedia dispute is nothing like this, and the repercussions of dissent are less sever for all involved, we are not arguing over a "true consensus" but a "rough consensus". Once that is agreed then all one has to do is define what a rough consensus is. The numbers should be displayed here because their correlation helps to explain what a rough consensus is. That the larger the effect on the Wikipedia project of a decision, the closer the rough consensus should be to a true consensus and this page should reflect this point with numbers. --Philip Baird Shearer 23:45, 17 January 2006 (UTC)
        • Okay, let me paraphrase that, consensus does equate to unanimity in cabinet government, but not on Wikipedia (I think this point is important because wikilawyers sometimes claim that since they disagree with something, there's no consensus). Radiant_>|< 00:00, 18 January 2006 (UTC)
        • I'm going to disagree with the technical definitions of "unanimity" and "consensus" offered just above this comment. (My definitions are based on the commonly-accepted models of group decision-making used in Change Management studies.) Unanimity in its strict application means that everyone agrees with the decision. Consensus in its strict sense means that everyone has been heard (which is a higher standard than just "had the chance to speak" because it requires listening) and that everyone can live with the decision. It means a lack of disagreement - which is not the same as unanimous agreement. By the way, unanimity and even consensus are known to be unattainable decision-making standards in groups over a certain size (though the size can be affected strongly by cultural and other influences).
          But, as both of you have already said, the Wikipedia definition is "rough consensus" - a different animal altogether. My understanding has always been that our "rough consensus" meant that everyone gets heard and that most everyone can live with the result. As a community, we recognize that the very openness of a wiki makes our process vulnerable to abuse by trolls and fringe minorities if we try to live up to purist definitions.
          By design, we've avoided strict numerical formulations because, as Samuel Wantman said above, any "voting" or polling we do is merely a way to measure how close we already are to rough consensus. I like his synopsis. Rossami (talk) 00:51, 18 January 2006 (UTC)

If we are in agreement with the principles above, is there an issue BESIDES whether to put the percentages on this page or leave them off? I think there is an argument to not being specific about percentages because that helps move us down the slippery slope towards fixed percentages. But what is the problem with this edit?:

Precise numbers for "supermajority" are hard to establish, and Wikipedia is not a majoritarian democracy, so simple vote-counting should never be the key part of the interpretation of a debate. When supermajority voting is used, it should be seen as a process of 'testing' for consensus, rather than reaching consensus. The stated outcome is the best judgment of the facilitator, often an admin. If there is strong disagreement with the outcome from the Wikipedia community, it is clear that consensus has not been reached. Nevertheless, some mediators of often-used Wikipedia-space processes have placed importance on the proportion of concurring editors reaching a particular level. This issue is controversial, and there is no consensus about having numerical guidelines. That said, the numbers mentioned as being sufficient to reach supermajority vary from about 60% to over 80% depending upon the decision, with the more critical processes tending to have higher thresholds. See the pages for RFA, AFD, and RM for further discussion of such figures. The numbers are by no means fixed, but are merely statistics reflecting past decisions. Note that the numbers are not binding on the editor who is interpreting the debate, and should never be the only consideration in making a final decision.
However, judgment and discretion are applied to determine the correct action. The discussion itself is more important than the statistics. In disputes, the term consensus is often used as if it means anything from genuine consensus to my position; it is possible to see both sides in a revert war claiming a consensus for its version of the article.

Is this a possibility for a compromise? If not, what changes would make it ok WITHOUT puting in specific numbers, and WITHOUT removing the numbers? -- Samuel Wantman 00:56, 18 January 2006 (UTC)

  • Sounds good, well done. Radiant_>|< 01:02, 18 January 2006 (UTC)
  • It seems like the page also ought to have a link to the related discussion(s) on Meta: such as m:Polls are evil if those are still current. Rossami (talk) 05:41, 18 January 2006 (UTC)

Anyone else want to weigh in on this? Brian, can you live with this? -- Samuel Wantman 20:49, 20 January 2006 (UTC)

  • I've made some grammar/spelling fixes, and switched your use of "consensus" in one sentence to "supermajority", which is the proper term. I'm alright with the new version. I would also suggested linking to WP:VIE either in the section, or under See Also. — 0918BRIAN • 2006-01-21 22:59

The next step in consensus is for the facilitator to implement the solution that in his/her best judgement is the rough consensus of the group. I just did that by copying over the proposed text. This is the true test of consensus. If I'm wrong about my judgement, I'm sure to hear about it! -- Samuel Wantman 23:28, 21 January 2006 (UTC)

  • I'm sure you'll hear about it regardless of your judgment. — 0918BRIAN • 2006-01-21 23:35

It has to work in all areas

If we're going to use supermajority for AfD, then we have to do it for DRV, as well. And why not RFA, too? User:Zoe|(talk) 00:08, 18 January 2006 (UTC)

  • I don't think we'll be doing that :) However, generally DRV does use supermajority (as far as I know that's the sole exception) because otherwise we'd possibly need a review of the interpretation of deletion review (e.g. WP:DRVRV, WP:DRVRVRV etc ad nauseam). Radiant_>|< 00:10, 18 January 2006 (UTC)
    • DRV only takes a simple majority to overturn. User:Zoe|(talk) 00:39, 18 January 2006 (UTC)
      • The use of majority/supermajority on DRV was one that was arrived out after long discussion and with a strong tip of the head to pragmatism. In those talks a general distaste for vote counting was the norm, whether the level was set at >50%, >75%, or any other number. End the end a hard and fast rule was created to close off any loopholes and to avoid recursion as mentioned above. (Although I'd have said WP:DRV2,WP:DRV3, etc.) It bears commenting that DRV has a fairly small participation base who tend to take longer to consider than on other XfD forums. An item than does not receive near-unanimousness there is the exception. - brenneman(t)(c) 06:42, 18 January 2006 (UTC)

Gimme the numbers!

<this space reserved>

Before we start rewriting wikipedia policy from scratch, let's get some numbers and discover if we need to. I've requested some graphs and statistics from gmaxwell, and I'm waiting until he gets home and can provide them. Kim Bruning 02:28, 18 January 2006 (UTC)

Polling is not necessarily the end of the process

"Rough" consensus means that everyone has the opportunity to be heard, and most everyone can live with the decision. Clearly rough consensus is something other than super-majority. For those processes where rough consensus is the goal for which we are striving, a poll is not necessarily the end of the process. Further discussion may or may not be disruptive, but it should not be considered disruptive merely because a super-majority has expressed its opinion. I think it is important that the policy make these points clear. Otherwise, there may be a tendency to believe that polls settle matters, period. --BostonMA 03:39, 18 January 2006 (UTC)

m:voting is evil does not appear as a Wikipedia guideline. That is why a statement ought to be made in the consensus guideline that in cases where rough consensus is the objective, rather than super-majority, then polls do not necessarily end the process. --BostonMA 23:45, 19 January 2006 (UTC)
  • It is now. WP:VIE. What was your point again? :)
  • Seriously though. Polls are generally used in two ways. (1) someone makes a statement and immediately calls a vote. WRONG. Generally, people will either close the poll, or create a new section called "Polls are evil" to vote against the vote. The main problem is that it reduces the issue to a yes/no question (or at least a pick between a limited number of options) and that the best option may not have been found yet and would be found after discussing.
  • (2) after a lengthy discussion, all arguments are laid out. Ideally, this is already consensus. Sometimes, people will still disagree, and it is unclear whether either side is consensual or simply a vocal minority (and that does not imply that either side is in fact consensual). In such cases, a poll may be appropriate to affirm a statement (which was produced through discussion) as consensual. Oe of the points is to identify a vocal minority as such. Radiant_>|< 01:52, 20 January 2006 (UTC)
As in, a poll does not decide whether to accept or deny something. A poll determines if it's worth it to negotiate further :-) Kim Bruning 11:05, 30 January 2006 (UTC)
No, as in, a poll may be appropriate to affirm a statement (which was produced through discussion) as consensual, and to identify a vocal minority as such. A good poll ends the issue, but a good issue generally doesn't end in a poll. Radiant_>|< 11:56, 30 January 2006 (UTC)

Guideline, quote

As far as I know, this has always been a guideline, though I've often seen people claim it is a policy, and accuse others of "violating the policy of consensus". As for Jimbo's quote, it would be a personal attack and an ad hominem to call someone an idiot in a discussion, and I'm sure Jimbo understands that fully. This quote simply shows why consensus is more than just a vote, and why voting doesn't always work. People may have idiotic rationale, so assuming that their rationale are sound (by putting it to a simple vote) would be idiotic, as would attacking the people, rather than their rationale. — 0918BRIAN • 2006-01-21 22:46

  • Policy or guideline is really not all that important a distinction. However, P or G pages do not have quotes on them. That's because quotes are easy to misinterpret, as is this one. Radiant_>|< 23:10, 21 January 2006 (UTC)

Now outmoded

This whole concept is entirely outmoded. Is there a template for policies we used to believe in but have now been trashed? Grace Note 04:46, 22 February 2006 (UTC)

Without examples you're not going to get very far. And I still see consensus is hard at work on FAC. Johnleemk | Talk 13:45, 22 February 2006 (UTC)

A comment on the concept of consensus as a governing principle

Robert was surely aware of the early evolutionary development of parliamentary procedure in the English House of Lords resulting in a movement from "consensus," in its original sense of unanimous agreement, toward a decision by majority vote as we know it today. This evolution came about from a recognition that a requirement of unanimity or near unanimity can become a form of tyranny in itself. In an assembly that tries to make such a requirement the norm, a variety of misguided feelings--reluctance to be seen as opposing the leadership, a notion that causing controversy will be frowned upon, fear of seeming an obstacle to unity--can easily lead to decisions being taken with a psuedoconsensus which in reality implies elements of default, which satisfies no one, and for which no one really assumes responsibility.

Robert's Rules of Order Newly Revised, 10th edition, pp. xliv-xlv.

For your contemplation. Kelly Martin (talk) 17:00, 25 February 2006 (UTC)

Okay, Kelly, but the question then is: is pretending that we do things by consensus the best way of resolving that problem? Or is allowing the empowered to make unilateral choices, so long as they are willing to be "responsible" for them better? What form would "responsibility" take? I'm guessing you don't believe your "constituency" should be allowed to disapprove your actions, so how do you see it?Grace Note 00:12, 26 February 2006 (UTC)

THIS IS AN ARCHIVE SEE Wikipedia talk:Consensus