Wikipedia talk:Consensus/Archive 3

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

Contents

Ouch

Wikipedia works by building consensus. This is done through polite discussion and negotiation, in an attempt to develop a consensus. If we find that a particular consensus happens often, we write it down as a guideline, to save people the time having to discuss the same principles over and over. Normally consensus is reached via discussion on talk pages. In the rare situations where this doesn't work, it is also possible to use the Wikipedia:Dispute resolution processes, which are designed to assist consensus-building...

And so it goes. What does it all mean? If we're trying to be all "descriptive, not prescriptive" and all that stuff, then let's face it, "consensus" is an abused word around here. What is consensus, as it relates to the wiki process, can we explain these without trying to define consensus in terms of itself. I wish I could provide more focused criticism, but my head hurts. 192.75.48.150 20:38, 18 December 2006 (UTC)

Does Consensus decision making help any? Note that wikis use consensus per definition. If there is consensus on the state of a page, the page is not changed, if there is not consensus on the state of the page, the page gets changed. It might get several changes before a new consensus is settled upon.
of course, with all the um, things going on all over the place on wikipedia these days (including talk pages, project namespace pages, wikiprojects and goodness knows what else), this basic, simple concept of consensus gets pushed to the background. But it's still there, and it's still the main mechanism, simply because it's there on every page :-) Kim Bruning 21:33, 18 December 2006 (UTC)
Would it make your head feel better if we define it in terms of related concepts? I.e. "consensus is the result of thoughtful dialogue, which takes place both through the editing process itself, and through outside discussion." Actually that's basically how I read the page as it stands, but perhaps it could be clarified to appear less circular. -- Visviva 15:43, 19 December 2006 (UTC)
See also: m:foundation issues #3
Kim Bruning 18:21, 19 December 2006 (UTC)

This anon has a good point. This page needs to describe the overall consensus process in much more concrete terms. The page treats consensus as the way to deal with conflict, rather than the foundation of how a wiki works. We are missing some of the most basic concepts. I'll give examples of what I think is needed:

  • Silence equals consent. This is the ultimate measure of consensus for every page: somebody makes an edit and nobody objects and nobody changes it.
  • Experience matters. People need to put some effort into understanding how things work before they try to change things.
  • Change happens slowly in small steps. Radical changes to the entire project are near impossible to enforce except by decree by Jimmy Wales or by RFA. Otherwise, the change has to happen slowly by experimentation and dialogue. People with good ideas for change should be encouraged to experiment.
  • Multiple ways of doing things can coexist. When there is no clear consensus, people can do things several different ways until one becomes common practice and the other fades into the background. This is often a good way to deal with disagreements.
  • Technological changes have an effect on policy. I'll cite as examples how {{CategoryTOC}} changed categorization policy.
  • Collaboration is essential. Editors must be open-minded and willing to work with others. Experts have to be able to work with novices. People with opposing political views must be able to work together.
  • Quality trumps quantity This is the essential reason why we don't "vote". One good well-thought-out argument is more important than dozens of comments from people who only say "keep" or "delete". Ultimately a consensus is not measured by counting "votes" but by whether conficts get resolved, and they get resolved through creative solutions and force of argument.
--06:59, 20 December 2006 (UTC)
Amen to that. Although I'm not 100% sure all of those would belong here --i.e., really relate directly to consensus -- they do help to elaborate on what m:Foundation issues means by the "wiki process." More and more people are being drawn to Wikipedia because of its prominence as an information source, and often become heavily involved without taking the time to understand how the underlying process works. We need more materials to help educate/acculturate this growing throng. -- Visviva 08:41, 20 December 2006 (UTC)
I'm starting to add some of this. -- Samuel Wantman 22:17, 20 December 2006 (UTC)
Your edit was a step in the right direction. 192.75.48.150 16:13, 21 December 2006 (UTC)

While we're on the subject

It just struck me as weird that "consensus" is a guideline, whereas "consensus can change" is policy... :) (Radiant) 15:51, 19 December 2006 (UTC)

Oddly, that seems about right to me. :-) -- Visviva 16:58, 19 December 2006 (UTC)
If you walk this path all the way, well, that would be interesting. :-) Kim Bruning 18:26, 19 December 2006 (UTC) Have you noticed any other odd taggings, lately?
  • Yeah, several. I'm working on it, though. Any in particular you had in mind? (Radiant) 11:29, 20 December 2006 (UTC)
Not really, it'd just be interesting if you continue looking at the situation on a general scale. I'm curious if you'll draw the same conclusions I did. :-) Kim Bruning 18:14, 20 December 2006 (UTC)
  • Yeah, I am. The two main problems seem to be (1) too much bureaucracy, and (2) many new(ish) editors make incorrect assumptions about how Wikipedia works, based on misinterpretation, partial observation or just not based on anything much, and strongly resist having these assumptions challenged. This is both caused by and causing the utter mess that is the Wikipedia namespace, and is difficult to work with. For a more concrete issue, it seems that the 3RR has become a net detriment, because the relevant process is downright byzantine, and because it tends to give people the impression that edit warring is acceptable if one sticks to the boundaries, and that it's okay to recruit people for the fourth. (Radiant) 10:04, 21 December 2006 (UTC)
This is odd. I don't see how consensus can be anything but policy, since it is an integral part of the wiki process. As I have mentioned above, I think this page needs to describe that process better, and should be less about resolving disputes and polling. Re-written this way, it would be policy as it is a foundation of Wikipedia. A page about how to solve disputes using consensus would be a guideline. -- Samuel Wantman 18:09, 23 December 2006 (UTC)
Yep. As mentioned above, all of our policies and guidelines depend on this thing called "consensus", but no one really knows what that means. — Omegatron 06:25, 25 December 2006 (UTC)
  • That's a good idea. >Radiant< 10:24, 24 December 2006 (UTC)

Policy

This seems like it should be a policy rather than a guideline. I'll change it in a few days if nobody has a problem, but that's a big change so it's best to wait awhile and discuss here. Just H 02:11, 28 December 2006 (UTC)

  • There's related discussion on the talk page of WP:CCC. Arguably consensus is such a basic principle that it could be policy. >Radiant< 10:09, 28 December 2006 (UTC)
    • Arguably, using a Policy to say that a Guideline is not binding is ridiculous.Circeus 19:11, 29 December 2006 (UTC)
      • I don't think WP:CCC is saying that Wikipedia:Consensus is not binding; it's saying that consensus on Wikipedia does not constrain itself. So I don't really see the problem with this remaining at guideline status, although I have no particular objection to the change. -- Visviva 15:35, 30 December 2006 (UTC)
        • It's just that it's irrelevant to state here as a policy, rather than there, that wikipedia decisions are taken by consensus.Circeus 16:54, 30 December 2006 (UTC)

I'm looking at the talk page for WP:CCC. What discussion are you referring to? Where is the difference between a policy and a guideline explained anyway?

  • According to WP:POLICY - none really, but policies "more official and less likely to have exceptions". It's not really made clear what's official and what's not in this context, let alone what would make something more or less official than something else.
  • The ticky boxes for both policy and guideline are identical, except:
    • Guidelines have blue tickies, policies have green tickies.
    • Guidelines are "not set in stone and should be treated with common sense and the occasional exception" in the guideline box, with the last two words linking to WP:IAR. This suggests that policies are set in stone and not amenable to exceptions and common sense.

Missing something? 72.137.20.109 03:58, 30 December 2006 (UTC)

I'm not sure that it's accurate to say that policies are not amenable to exceptions and common sense, but it is accurate to say that making exceptions to policies requires a higher burden of proof than making exceptions to guidelines. There are obvious exceptions to the principle of consensus on Wikipedia: for example, Office actions and decisions of the Foundation supercede any opinions expressed by consensus; similarly, if there were an apparent consensus to include unsourced information in violation of WP:BLP, that wouldn't fly. These exceptions should probably be included explicitly on the page, but if they are I would support this shifting from "guideline" to "policy". —Josiah Rowe (talkcontribs) 03:47, 18 January 2007 (UTC)

merge proposal

I don't think it's pertinent to have these pages separate when one is basically discussing a specific aspect of the other. While keeping them separate was pertinent when their status was different, it seems irrelevant to keep them separate now, and seems closer to POV-forking (where users have been using WP:CCC to assert that consensus is invalid because it can change at a later date.Circeus 17:18, 20 January 2007 (UTC)

Agreed. Since "no binding decisions" has had its name change to "consensus can change" it is more obvious that it actually describes a part of the concensus process, and could well be merged here. // habj 23:20, 25 January 2007 (UTC)
Agreed. CCC should redirect to the appropriate section, though. — Omegatron 00:13, 26 January 2007 (UTC)

Exception

Whether a fact satisifies WP:V and WP:OR is, and always has been, decided by consensus. If this policy will claim that WP:V is not subject to consensus, then in order to avoid trolling wikilawyer editors it is necessary to point out that just because a troll doesn't think that WP:V or WP:OR is met doesn't mean that it actually is not met.

I agree with the spirit of the exception, which is the unverifiable material cannot be included even if a consensus agrees to do so. But if a consensus agrees that the material is verifiable, then it is not OK for a troll to remove it claiming it is not. CMummert · talk 18:03, 20 January 2007 (UTC)

Note that you're talking about a local consensus to include unverifiable changes there. The global consensus (global to the entire project) is that it's not ok to add unverifiable material. I'm not sure why this particular case of global vs local has been chosen, specifically? --Kim Bruning 22:48, 23 January 2007 (UTC)

I don't know - it was added a few days ago, and I was just trying to clarify what was here before it got "stuck" the way it was. CMummert · talk 23:40, 23 January 2007 (UTC)

What? You still don't get it? Do you want me to draw a picture or something?!

Consensus plain.svg

Well, here you go then! ;-)

For review, I applied my -admittedly rusty- vector art skills to actually drawing a flowchart of how consensus works on-wiki. Note that in normal editing on most pages, you generally don't get to the talk page much at all, really.

--Kim Bruning 05:59, 23 January 2007 (UTC)

Wow, nice picture. Throwing that into the main article certainly wouldn't be a bad idea.--Wizardman 06:11, 23 January 2007 (UTC)
One quibble. Being unchanged for a long time creates a presumption of consensus but that's only ever an untested presumption. Lack of further edits might only mean that the edit has so far been overlooked. But I'm not sure how to integrate that quibble into your flowchart in a way that makes sense. This will be helpful. Thanks. Rossami (talk)

Shanel was complaining about not knowing where to start, and verily, a proper flowchart has a Start. :-) (Added) --Kim Bruning 06:48, 23 January 2007 (UTC)

Titoxd came up with an idea to get somewhere in the right direction wrt what Rossami has been saying. Is this an improvement? --Kim Bruning 07:03, 23 January 2007 (UTC)

The next step is, of course, to get a good descriptive text to go along with the picture. Years of confuzzlement and confusion have sort of eroded away any original description, if it even ever really existed. I hope some folks will help. I'll put up the diagram for now at least. --Kim Bruning 15:50, 23 January 2007 (UTC)

Yes, I've run into what I think is an interesting case of confusing (or conflating) consensus with status-quo. I noted how the flow-chart does return to the beginning (i.e. status quo) and that is significant as well. I had the following thoughts regarding an explanation. Not sure if it goes with or just near the diagram.
Lack of change, or status-quo, is necessary to establish consensus, but defending the status-quo is the opposite of building consensus. Consensus building is an inclusive process that seeks to involve more people. Defending the status-quo alienates people, and stunts progress. Consensus cannot be actively maintained, it can only be interactively broken and re-established.
I'm off to work, and this is just my first pass. So please kick it around if you likeDhaluza 11:10, 24 January 2007 (UTC)
By some coincidence, on a wiki, the status quo on a page happens to coincide with consensus on that page. If you spot the page, and you happen to not agree with the status quo, then there is no longer a consensus :-)
You are quite right about defending the status quo too... there's a suggested method bold reverse discuss for dealing with that kind of thing. But all of this needs more writing to be done, IMHO. --Kim Bruning 11:50, 24 January 2007 (UTC)

I really like the idea of a flowchart. — Omegatron 00:17, 26 January 2007 (UTC)

I like it. Makes consensus a lot simpler to work with when there's a flowchart :). Yuser31415 (Editor review two!) 03:02, 26 January 2007 (UTC)
With things a little tighter. Also changed to blue and was a perfectionist about lining up the arrows.

I tried to squeeze things a little tighter so that it could be readable at a smaller size, and ended up completely redoing it, so I uploaded as a different file name. — Omegatron 04:51, 26 January 2007 (UTC)

  • Cute. Can you do a jumping arrow thingy so that there's one "consensus" box at the bottom, rather than two of them? >Radiant< 10:00, 26 January 2007 (UTC)
Actually, I was thinking that it would be better if the arrows both went back up to the original Consensus box to show more clearly that it's a cycle, but it would need to be laid out a little differently, and I didn't know how you draw "unsoldered wires" in a flowchart. Is that what you mean by "jumping arrow"? — Omegatron 15:04, 26 January 2007 (UTC)
With just one consensus oval, to show it's a cycle. A little skinnier, so the words are legible at thumb size.
I tried to implement this and make it skinnier in the process. (The skinnier it is, the more legible the text will be without taking up the entire width of the policy page.) The word "implement" fitting in between those two boxes makes it wide.  :-) Please try to think of a skinnier alternative layout. — Omegatron 03:35, 27 January 2007 (UTC)


Well, it's skinnier, but the crossing lines and placing of the words make it less readable than your last version (the 2nd version on this page is the most readable of the 3 :-) ).

The best way to try and prevent crossing lines is to redesign/ refactor the process to be structured, of course. But that might not be descriptive. Hmph. :-p


Here's some rules that I seem to remember for flowcharts, maybe they're handy: "ovals" should typically only be labled start or stop (I cheated with consensus, because it is the ultimate "stop" point for wikipedia). "Start" may not have any arrows towards it, and may only have one arrow departing. "Stop" may not have any arrows departing. Decisions are associated with the departure point from a diamond. Additional process names are associated with the box from which they start. Crossing lines are evil. --Kim Bruning 05:49, 27 January 2007 (UTC)

Hmm, on the basis of flowchart rules above... maybe putting consensus in a box in the center, or a slight variant where the flowchart is all creative and curvy and circular looking. (Tricky to implement, I know :-P)... hmm... --Kim Bruning 06:06, 27 January 2007 (UTC)

I thought it was a bit funny to have consensus at the top. If you begin with consensus, then there's nothing much to do. I see consensus as belonging at the bottom. SmokeyJoe 06:16, 27 January 2007 (UTC)

You begin with the current consensus. You disagree with it (so now there's no full consensus), so you set about to change things to bring it back to consensus. So consensus is both at the beginning and at the end, forming a cycle. That's what Omegatron has been trying to illustrate better, and how he came up with his third attempt. --Kim Bruning 07:22, 27 January 2007 (UTC)

I do prefer the new diagram's graphic layout, but there should be two ovals, one at the top that represents "Previous Consensus" and one at the bottom that represents "New Consensus". The revert path should return to the top, the edit path should lead to the bottom. This illustrates the point that consensus can change, forming an new starting point, as explained in the note at the bottom.. It also better illustrates the subtle but important corollary that consensus does not exist unless the person proposing the change agrees, not whether the person making the revert disagrees. With one consensus box at the top, the consensus change is not as obvious, and the distinction between consensus and status-quo is blurred. Dhaluza 11:39, 27 January 2007 (UTC)

Great ideas.

I think this will do wonders for people who think their revert warring is justified or who misuse "consensus" to mean whatever they feel like that particular day. Simple graphics are much harder to wikilawyer than text. We should create flowcharts for all of our policies!  :-)

What do you think of the color and linewidths, etc? What about the sideways "Implement"? — Omegatron 06:38, 28 January 2007 (UTC)

I've found that typically, you know you are in a doomed bureaucratic organisation when people have pretty flowcharts for all their policies. Ut oh. ^^;; --Kim Bruning 06:44, 28 January 2007 (UTC)
You started it.  :-) But I think it's actually a good thing in a wiki-based organization where policies can be subtly changed over time to suit the desires of small numbers of people. A graphical flowchart is much less likely to be messed around with and misinterpreted.
I agree with the idea of having a "Previous consensus" and "New consensus", but I don't see any way to have a single "New consensus" box without crossing paths. — Omegatron 01:24, 29 January 2007 (UTC)
Yet another version. Has "Previous consensus" and "New consensus". There's no way to avoid the crossing path with the single outcome, so I made it a "jumping arrow". Start is bold so it's more obvious. Even skinnier.

Yet another version. — Omegatron 02:16, 29 January 2007 (UTC)

Nice! -- Ned Scott 03:01, 29 January 2007 (UTC)
OK, I guess a revision/revert cycle does create a new consensus in a sense, so it is not necessary to return to the top. This would make the diagram cross-over again anyway. The one touch-up I think we should also incorporate is to change the "as you like" in the note on CCC to "as necessary". As you like implies that people can do this on a whim, and there really should be a more substantial purpose for making a change.Dhaluza 10:49, 29 January 2007 (UTC)
Done. — Omegatron 14:28, 29 January 2007 (UTC)

I like the diagram, but, should this policy really say that you should go straight from previous consensus to editing? What happend to "discuss changes on the talk page first"? WP:Bold is important, but sometimes its better to talk first. I suggest two routes from “previous consensus” to “Make an edit”. One direct, the other via “Discuss on the discussion page”. SmokeyJoe 23:48, 29 January 2007 (UTC)

Well, there are other paths to consensus, so we should probably just note that this is the path for many good faith but not all. -- Ned Scott 05:43, 1 February 2007 (UTC)
Consensus-flowchart.png

There's a more general image here, by the way. — Omegatron 00:55, 25 April 2007 (UTC)

Interesting! This is for general purpose real world use? I actually do see this process used on talk pages too at times. (though it is much slower than wiki-editing... ^^;;) --Kim Bruning 01:55, 25 April 2007 (UTC)


The quote in the "Consensus vs. supermajority" section

I can not see that this quote has anything to do with supermajority at all! It does not describe supermajority. Either it is a good description of consensus; then it should be included in the first section. If it is not a good description... then it should be removed. // habj 23:47, 26 January 2007 (UTC)

I feel the section is a bit problematic, and have just had a try at improving it. I don't think the quote is sufficiently authoritative. Perhaps its content should be intregrated into the text, as assuming its content is important, it should be reworded, and reduced to a citation. SmokeyJoe 06:28, 27 January 2007 (UTC)

I'm unhappy with the edit SmokeyJoe made here, as it seems to denigrate the quote (which I feel to be a decent description of consensus as it works — when it works — on Wikipedia). Also, I'm not certain that the use of the word "supermajority" in the text was a reference to the proposal Wikipedia:Supermajority, which I had never even come across; at least, I had always assumed that the use of "supermajority" was just a descriptive term.
The current version of the section seems to suggest that whenever an unreconciled minority exists, consensus has not been reached. I don't think that's accurate, and this recent ArbCom case would seem to agree. I'll see if I can think up a better wording myself — if you don't like whatever I come up with, feel free to revert me. —Josiah Rowe (talkcontribs) 07:54, 27 January 2007 (UTC)
Well, I gave it a shot — the bit I'm fondest of is splitting the quotation out from "Consensus vs. supermajority" into a section on "Consensus in practice", since that's really what it's about. Also, the quote is more about the difference between consensus and unanimity than the difference between consensus and supermajority.
I'm less certain about my wording for the "consensus vs. supermajority" section — I hope that some wordsmith may be able to condense what I was trying to say into something clearer. The point, I think, is that although supermajority is different from consensus, it is possible to have a consensus to abide by the will of the supermajority, and in areas such as RfA that's pretty much what happens. Improvements to my awkward wording are invited. —Josiah Rowe (talkcontribs) 08:15, 27 January 2007 (UTC)

Hi Josiah. Sorry to have made you unhappy. The quote is given a lot of prominance for an except from an email to a mailing list, don't you think? To be used properly, shouldn't it be attributed to the individual? If it contains some valuable wisdom, but the source was not notable, then it should be rephrased with credit given.

Wikipedia:Supermajority has a long history intertwinned with this page. It originated here, was split off, and then rejected.

Afterthought: I don't object to ignoring the rejected policy. Rejected (perhaps like consensus at wikipedia) is a misused word. There was not a consensus to reject it. People gave up before reaching any agreement. SmokeyJoe 10:38, 27 January 2007 (UTC)

"Consensus does not mean that everyone agrees with the outcome; instead, it means that everyone agrees to abide by the outcome." This is not true in the world I live in. People agree to abide to majority decisions, but they don't claim that this makes them consensus decisions. People can be bullied into abiding a decision, which does not make a not consensus.

It seems to be that "consensus at wikipedia" is a perversion of the word consensus. Consider the definitions from the merriam webster and from the oxford english dictionary:

oed.com

1. Phys. General agreement or concord of different parts or organs of the body in effecting a given purpose; sympathy. Hence transf. of the members or parts of any system of things.


1854 G. BRIMLEY Ess., Comte 320 In the universe..he resolves to see only a vast consensus of forces. 1861 GOLDW. SMITH Lect. Mod. Hist. 24 There is a general connexion between the different parts of a nation's civilization; call it, if you will, a consensus, provided that the notion of a set of physical organs does not slip in with that term. 1870 H. SPENCER Princ. Psychol. I. II. ix. 278 A mutually-dependent set of organs having a consensus of functions.

2. a. Agreement in opinion; the collective unanimous opinion of a number of persons. 1861 Sat. Rev. 21 Dec. 637 Bishop Colenso is..decidedly against what seems to be the consensus of the Protestant missionaries. 1880 Athenæum 10 Apr. 474/3 A consensus had actually been arrived at on the main features involved.

m-w.com

Main Entry: con•sen•sus Pronunciation: k&n-'sen(t)-s&s Function: noun Usage: often attributive Etymology: Latin, from consentire 1 a : general agreement : UNANIMITY <the consensus of their opinion, based on reports...from the border -- John Hersey> b : the judgment arrived at by most of those concerned <the consensus was to go ahead>

2 : group solidarity in sentiment and belief usage The phrase consensus of opinion, which is not actually redundant (see sense 1a; the sense that takes the phrase is slightly older), has been so often claimed to be a redundancy that many writers avoid it. You are safe in using consensus alone when it is clear you mean consensus of opinion, and most writers in fact do so.

In both, the reference to unanimity is strong. In my real world experiences, consensus (when not immediately achieved) may only be achieved after a process of considering a multitude of possible positions, and often only a very weak consensus can be acheived. This may mean that the final resolution includes an acknowledgement of differing opinions, as which point the dissentors formally yield. SmokeyJoe 10:32, 27 January 2007 (UTC)

You are correct that the meaning of "consensus" as used on Wikipedia is different (or at least, often different) from its meaning in the English language. I have found that this results in a great deal of confusion and I think it would be better to find a different word. On a related note, the mantra that "decisions on Wikipedia are made by consensus" ignores the fact that, often, decisions aren't made at all. All the fancy flowcharts in the world can't change reality. 6SJ7 11:55, 27 January 2007 (UTC)
While I am usually very deferential to the authority of the writers of published dictionaries, in this case they are wrong. The dictionaries are still using an outdated and simplistic definition of the word "consensus". We use it in the sense commonly used by Change Management and Organizational Decision-Making experts - that is, one mode on a spectrum of decision-making modes. I tried to describe that spectrum here but would appreciate a reference if anyone still has the original article where this model was discussed. Rossami (talk) 14:19, 27 January 2007 (UTC)
Even with the strict dictionary definition, the concept of using consensus on a wiki still makes sense. If you come across a page and disagree with it, then there is no consensus anymore, and therefore you may edit it. Note that consensus in this case is "consensus among all those who have seen the page recently" as opposed to "consensus among all 3 million registered users".
If it helps any, note that there can be many reasons for people to agree with something, even if they are not happy with it. You might agree to a page version because it's perfect. You might agree because you're too lazy. You might agree because it's good enough, and changing it is not worth the effort. You might hate the new version, but want or need to be diplomatic for some reason. You might dislike it, but can live with the changes for now, etc...
Canonically, the decisions of a tyrannical dictator always have consensus, because everyone who disagrees with them has been shot dead, and those wiser souls who are still alive tremble in fear and dare not oppose. For some reason, I do not actually recommend that particular method of consensus formation for use on wikipedia, but it does serve as an (extreme) illustration.
--Kim Bruning 16:31, 27 January 2007 (UTC)
Wow, that's very well put. I really like the pat about noting the difference of consensus between "among those who have seen the page recently" and "the entire community/everybody". -- Ned Scott 16:42, 27 January 2007 (UTC)
The second paragraph of what Kim has written above, is in essence the same as that quote from the mailing list. We don't need it in the form of a quote, we need it described in the actual text of the page. // habj 20:18, 27 January 2007 (UTC)

Since the word concensus is used in several different meanings, also in this text, maybe we need a section that defines different usages of the word? Currently, there is a piped link in the first sentence to consensus decision-making. Unless we really define consensus the way it is used and explained in that article, it should not be there. There are several other articles you could link to, for instance consensus. For now, I just remove the link. // habj 22:35, 27 January 2007 (UTC)

Per AGF

I changed the wording slightly. It used to say "Editors must always assume good faith". I think, per AGF, that users are allowed to cease doing this when there is overwhelming evidence that a user is not acting in good faith, although they must strive forit whenever possible. IronDuke 19:02, 27 January 2007 (UTC)

Sounds very reasonable to me, good idea. -- Ned Scott 19:05, 27 January 2007 (UTC)
A very important point. — Omegatron 06:33, 28 January 2007 (UTC)

Rough consensus

I've made a first mention of Rough consensus again, but we have a bad definition of it. Note that in that same line, I also state that certain processes may have been somewhat misdesigned if they wish to use consensus (because they do not scale well, and have too many people participating (ie, more than Dunbar's number predicts to be a good idea.) Philip reverted it saying there is no consensus, well... note that this is an objective statement based on predictions made by peer reviewed scientific research (see the links). I don't care how broad a consensus there is for the sky being green. Get proof! :-P --Kim Bruning 11:18, 28 January 2007 (UTC)

Exceptions

What was incorrect about the section? I agree that it should be reworded, but not completely removed. Trebor 16:31, 29 January 2007 (UTC)

I have added it back pending discussion on specific failings or rewording. Rossami (talk) 16:58, 29 January 2007 (UTC)
Apologies, I'd removed it already, but ended up scratching my head a little too long.
At any rate, local consensus can, may, and should override wiki-wide policy, (I'd almost say duh, it's a consequence of ignore all rules). Note that things like verifiability will still apply anyway, as long as more people come along and look. Office actions can be overridden by sufficiently large numbers of people forming a consensus, though typically at the meta level. Same goes for all the top brass type people. Finally, foundation issues can be modified by consensus at the meta level, but it's a mite tricky to pull off. So while in practice they're very *hard* to override, they still can be. If it needs to be pointed out here that they're *hard* to override, so be it... but at the same time, I want to continue to stress the impression that everything is consensus based. <still scratching head> --Kim Bruning 18:45, 29 January 2007 (UTC)
"local consensus can, may, and should override wiki-wide policy," I'm sorry, but I can't swallow that. In very cases I can think, it's either policy interpretation that is involved, or policy was changed. Policy is Policy, there's little to add.
I reordered them for now. While the two last are admittedly debatable, the first two, I think, are hardly so.Circeus 18:58, 29 January 2007 (UTC)
On a consensus based wiki, policy is a kind of illusion. What you see written down on pages in the project namespace is a description of what compromises, ideas, and systems people have discovered to work best, anytime they tried to edit as per the chart above. Even when people start out with a vote (like has been experimented with once or twice in the past), the end result still gets smoothed by consensus in wiki-editing. So not only isn't policy policy, in fact it doesn't even exist. (And if you don't believe me, we have a policy that says so explicitly, called Ignore all rules). If this sounds confusing, I might still know one or two better ways to explain it.
The reason some people started calling certain things policy is that that way they could save a lot of time trying to get newbies to do the right thing. I have always been rather opposed to that move, since it makes things rather confusing later on (as we can see :-P ) .
--Kim Bruning 19:02, 29 January 2007 (UTC)
Note that the ordering you have now runs from most debatable to least debatable... Verifiability is the smallest scope concept, and only applies on (en.)wikipedia. Office actions are typically done by arbcom, or certain foundation employees. Jimbo and Board can override office (but not in reverse). Finally, even the board must obey the foundation issues in practice, because people typically join on the basis of the foundation issues, and violating them would cause a large number of those people and in fact entire projects to either defect or fork (or both) --Kim Bruning 19:15, 29 January 2007 (UTC)
"policy is a kind of illusion" But a necessary one if we want to avoid descending into chaos, basically. Consensus comes into play when debate is involved, not when policy appears to contradict itself. If you prefer, Consensus can't override the spirit expressed by policy. Wikipedia:Five pillars makes points that can hardly be overrun by any consensus,if only because they are extensions of the Meta basic principles.
"Note that the ordering you have now runs from most debatable to least debatable..." Sorry 'bout that. My edit apparently didn't register the first time around.Circeus 19:24, 29 January 2007 (UTC)
Because, for the most part, content policies at least are pretty much immutable, I rewrote the entry to emphasize that aspect.Circeus 02:00, 30 January 2007 (UTC)
The thrust of it seems OK to me. I softened the language to should, rather than must; these really mean the same thing in the context of a consensus based system. I think the word "cases" might be better as "articles" but it's not that significant. CMummert · talk 02:11, 30 January 2007 (UTC)
Policy is not an illusion, it is consensus. Policy is the current consensus opinion of how to apply the basic principles in practice. WP:IAR is just the corollary to WP:CCC -- since building a better encyclopedia is a basic principle, with a very wide consensus, any policy that stands in the way has a narrower consensus, and that consensus must change. So it is not contradictory at all. Dhaluza 02:34, 30 January 2007 (UTC)
That's a much better way to explain what I'm trying to say. <bows to the superior explainer> --Kim Bruning 10:26, 30 January 2007 (UTC)

Consensus or Agreement ?

I believe the primitive notion that underpins Wikipedia is agreement. Agreement is defined as "a meeting of minds" in contract (common) law. It requires an offer and its acceptance. In the language of common law, one may 'accept' or 'reject' an 'offer'. A counter-offer is just another 'offer'.
It is equally important to recognise that the flip side of the 'agreement' coin is 'conflict'. ie. People may 'agree' to 'disagree', so disagreement alone is inadequate. Indeed, unless opposing parties choose to agree to 'fight' there is no fight! Without conflict (arising from the diversity of access to Wikipedia editors), the articles and even the rules of Wikipedia would never grow / adapt. Remember Wale's 'Ignore all rules' policy? ie. Every policy, guideline or any other rule may be ignored if it hinders improving Wikipedia.
Do people think some policy needs to be written on agreement and or conflict? Scholzj2006 01:46, 30 January 2007 (UTC)

Failures?

Just throwing out ideas here, so feel free to shoot me down, but shouldn't there be a section that explains that consensus - god, it cheeses me off when people misspell that - is not perfect as a way of deciding what should be done and may occasionally not work? And that despite this, however, due to its general success and the fact that there is no evidence to support the contention that any other system would be any better, we still use consensus to decide things?

I believe that there have been several ArbCom findings of fact to this effect. Moreschi Deletion! 13:03, 30 January 2007 (UTC)

"It's the worst system, except for all the others?" --Kim Bruning 13:11, 30 January 2007 (UTC) may Mark Twain forgive me
Something like that. So, is this a good idea, or not? Moreschi Deletion! 13:14, 30 January 2007 (UTC)
It's something that we should certainly think about. Maybe try on a separate page first? --Kim Bruning 07:24, 31 January 2007 (UTC)

First draft is at User:Moreschi/Consensus. Cheers, Moreschi Deletion! 22:06, 3 February 2007 (UTC)

2nd person

Does anyone else think the 2nd person "you" is inappropriate in the WP:CCC text that was merged in? Dhaluza 00:42, 31 January 2007 (UTC)

Yes. CMummert · talk 00:58, 31 January 2007 (UTC)
It probably wasn't as obvious when it was a stand-alone article, but it really sticks out now that it is merged in. Dhaluza 01:00, 31 January 2007 (UTC)
I reworded it and also rephrased this sentence: "You cannot declare a new consensus all by yourself without the participants in the discussion agreeing that the previous consensus does not apply anymore." If consensus has changed, you still may not get others to explicitly say it has changed. What matters is whether other editors act as if it has changed, and whether they revert you when you write that it has changed. So maybe someone can expand that sentence a little.
Also, the third para in that section is pretty bad. CMummert · talk 01:25, 31 January 2007 (UTC)
Nice job! Dhaluza 02:24, 31 January 2007 (UTC)

P.S. I also noticed 1st person "we" in the second paragraph as well. Dhaluza 16:08, 3 February 2007 (UTC)

Two Consensus flow-charts?

Hey, I love the flow-chart for consensus-building, but I feel that there really should be an extra level of process for edits to policies, guidelines, and other documents which have gained consensus of a large section of the community. The flow-chart as it now stands works beautifully for articles, but I feel that the first action when building a new consensus for "massive-consensus" articles like policies and guidelines should be at least a post on the article talk page, if not a notice at the Village Pump as well.

I personally feel that essays should also be included in this flow-chart, but I don't think that's as crucial as emphasizing that policies and guidelines need an "extra mile" approach to consensus.--Aervanath 04:31, 1 February 2007 (UTC)

You make 2 errors. First, all policies and guidelines are descriptive, not prescriptive. Second policies and guidelines (including this one), can be edited in the same way, and no harm will come of it. Try it! :-) Worst that can happen is that you simply get reverted, and have to take it to talk anyway. --Kim Bruning 07:56, 1 February 2007 (UTC)

Oh. Right. Be bold! Thanks for clarifying that!--Aervanath 08:04, 1 February 2007 (UTC)

How long to wait for a consensus?

I can't find anything in the policy that tells us how LONG to wait for a consensus to form. This came up recently in a dispute on the WP:COI talk page, with one user giving a 12-hour deadline to respond to his comments before they reverted the article. If this has not been addressed elsewhere, I propose that following be included in the policy:

How much time should I wait for a consensus to form?

Consensus can take 5 minutes to 5 years. Generally, if there is still discussion occurring about an issue, then consensus has not been reached. However, if the discussion has stalled, or there have been no additions to the discussion after the initial proposal, then it is appropriate to assume consensus after the corresponding length of time has passed:
1) For a lightly-edited article, 24 hours;
2) For a heavily-edited article, 3-5 days;
3) For Wikipedia policy and guidelines, 1 week.

Important Note: If you do wait the requisite length of time before making your desired edits, and someone then reverts the edits that you have made, then DO NOT REVERT THOSE CHANGES; it is a sign that the discussion needs to be re-opened, and a new consensus formed.

  • Thoughts? --Aervanath 06:05, 1 February 2007 (UTC)
I tend to give lightly-edited articles more time, not less, on the hypothesis that fewer people are watching it and more people need more time to find and decide to join the conversation. Rossami (talk) 06:37, 1 February 2007 (UTC)
I had been working on the assumption that the heavier the editing, the more editors would be involved, therefore it would take longer to come to a consensus. However, what you say makes sense, too. How about:
1) For a heavily-edited article, 3-5 days;
2) For a lightly-edited article, 1 week;
3) For Wikipedia policy and guidelines, 2 weeks.
  • More thoughts? --Aervanath 07:09, 1 February 2007 (UTC)

I'm not fond of this thread. If we start talking about how long to wait, people will start wikilawyering when someone objects one day after anything posted. This kind of language is not in the spirit of consensus. Consensus forms when it forms, it changes when people give inspired reasons that are convincing. Quantifying the amount of time involved will just make things worse. -- Samuel Wantman 07:21, 1 February 2007 (UTC)

I realize that. I guess I wasn't clear in my proposal. I'm not looking to provoke wikilawyering (you're right, that is a danger), but to provide a guideline for how long to wait before editing/reverting when there has been NO RESPONSE to your proposal. You can't form a consensus without discussion, so if there's no discussion at all, how long should you wait before just being bold and going for it?

Hmm. I think this is actually moving away from something that belongs on the concensus page. I'll post a query over at the Village Pump (Policy). Thanks for your input!--Aervanath 07:32, 1 February 2007 (UTC)
How long should you wait if there's no discussion? You should probably just give up. I've posted several proposals that got absolutely zero response. It means that nobody liked it, or nobody bothered to read it, or both. If it is something you can undertake on your own, you can be bold and try it, and see if it remains. I started Wikipedia:Classification that way. On the other hand, if after lots of heated discussion, nobody responds to a proposed solution, it often means that everyone can live with it, and you should probably be bold and see if it survives once it is posted. -- Samuel Wantman 08:15, 1 February 2007 (UTC)
Actually, I was more sort of thinking along the lines of proposing an edit change on a policy talk page and having no one respond. However, in that case I agree with you that probably the best thing to do is just do the edit and see if someone reverts it. Thanks! --Aervanath 08:27, 1 February 2007 (UTC)

This is already covered. No further wording is needed. --Kim Bruning 08:01, 1 February 2007 (UTC)

True, but it's not really explicit. However, I'm hesitant to make the policy any longer. That would WP:CREEP everyone out, I think. :-P--Aervanath 08:17, 1 February 2007 (UTC)
Well, you're right about one thing. Anything you don't explicitly nail down... someone will come along and abuse it. (see this page as the ultimate example of what can go wrong). I don't feel good about writing tiny tiny bureaucratic rules though. :-/ --Kim Bruning 08:27, 1 February 2007 (UTC)
I think you're right not to feel good about it. We shouldn't have to write down all these rules. Basically, most of them come down to "Don't be a jerk" anyway, so if someone starts abusing the process, we can shut them down then.--Aervanath 08:34, 1 February 2007 (UTC)
<Snicker> You been introduced to The trifecta yet? :-) --Kim Bruning 08:46, 1 February 2007 (UTC)
It's already on my talk page. :) Found it last night. --Aervanath 08:57, 1 February 2007 (UTC)
 :-D --Kim Bruning 08:58, 1 February 2007 (UTC)
It's rarely a good idea to set firm numbers like time limits, but I agree with the general idea. If a proposal has been made, a long time passes with no response, it's probably fine to be bold and try implementing the proposal. The flipside is that it's a bad idea to make a proposal with a time limit/ultimatum along the lines of "If I don't get a response in X...", particularly if it's a relatively short timeline. It's wrongheaded to assume that because a few hours are passed, nobody cares, especially in an ongoing discussion where people have already given opinions. COI was a funny situation in that it wasn't even a case where a proposal received no responses. There was quite a bit of discussion and objection to proposed edits, but one editor seemed to think that if the others had an obligation to continue the conversation even though they had given their opinion, and that a lull in the discussion was an indication that everyone else stopped caring and it was OK to ignore the objections and edit away. I'm always leery of editors who insist "nobody is responding to me" or "not discussed on talk page" when it has been discussed...but they just don't like the responses.
I don't know if Consensus is the best place for an addition along those lines, or if it would be more appropriate somewhere else (assuming it doesn't exist already). --Milo H Minderbinder 14:22, 1 February 2007 (UTC)
I agree with that - the test of consensus is whether your changes are removed, not whether you can wear down your opposition until they stop responding. This comes up all the time in articles, when determined editors will spend months pushing for non-consensus edits of certain articles. Leaving a comment and waiting for a while, depending on the severity of the changes you plan to implement, is a polite thing to do, but a lack of responses doesn't necessarily mean others agree with you. CMummert · talk 14:46, 1 February 2007 (UTC)
"Leave a comment first" is overused perhaps. If you do leave a comment, also edit. If you see an edit you disagree with, check back in 5-10 minutes to see if a comment shows up. --Kim Bruning 15:03, 1 February 2007 (UTC)

Thinking more about it, I am very uncomfortable with this concept. For one thing, the time you wait and the default action you can take can depend very much on the wording you use in your proposal. "I think X is a good idea and will implement unless someone objects" is very different from "I think Not X is a bad idea" or "I've ranted about X for the 10th time and people stopped bothering to respond to me". For another, the appropriate time has more to do with the nature and scope of the change than with whether or not the page is heavily edited. If the edit seems non-controversial, I'm going to be bold. If I expect it to be highly controversial, I will allow months for consensus to emerge. I don't think we can define even loose guidelines here. Sorry. Rossami (talk) 14:51, 1 February 2007 (UTC)

You should be bold in both cases. Worst case you will end up in a bold revert discuss cycle, otherwise you can just continue with normal consensus editing. --Kim Bruning 15:03, 1 February 2007 (UTC)
  • Strict time limits would be pretty much instruction creep. Not such a good idea. >Radiant< 15:23, 12 February 2007 (UTC)

Making page strictly en.wikipedia

This [1] recent edit by CentrX ties consensus specifically to the en.wikipedia definition . I wonder if we're actually creating these pages at the correct location? (Perhaps we should be working on meta?) --Kim Bruning 21:43, 11 February 2007 (UTC)

Dunbar's number, and predictions about RFA, AFD

When discussing RFA, AFD, etc...

" , therefore these processes may have been somewhat misdesigned. As a first heuristic in these decision-making processes, people might first see if the criterion of supermajority is achieved, and on that basis make a first order assumption on how close one is to rough consensus"

This is not correct?

--Kim Bruning 12:34, 14 February 2007 (UTC)

No I do not think it is not correct because I think it is a minority point of view "that the processes may have been somewhat misdesigned." If it were not, then they would not be as they are. That they exist and are in very heavily used shows that there is a broad Wikipedia community consensus on the use of these processes. As for the second sentence please see the British Plain English Campaign. --Philip Baird Shearer 14:23, 14 February 2007 (UTC)
See also Parliament Act which allows for a guillotine on fillibustering. I do not think that such an act means that Parliamentary democracy is somewhat misdesigned. --Philip Baird Shearer 14:29, 14 February 2007 (UTC)
The first section is to do with the fact that they do not take Dunbar's number into account, and may therefore ultimately be self-defeating. Whether or not they have wide community consensus is irrelevant, as communities are quite capable of becoming pathological and self destructive if misdesigned.
The wording here implies that literature suggests/predicts that these processes might be flawed, and therefore these processes should be closely watched. Seeing the discussions ongoing at for instance WP:RFA over the past months, people seem to state that there are known problems, so perhaps the literature-based prediction is turning out to be correct. AFD has long been known to be a problem area.
As for the second sentence. Please {{sofixit}} rather than deleting. That's what wikis are for. :-) --Kim Bruning 15:05, 14 February 2007 (UTC) Perhaps the whole block needs to be rewritten in Plain English. I was initially merely concerned with expressing the situation in as exact terms as possible.
As Kim has asked (on my talk page) for more. There is no point in the follow up sentence if the previous phrase is removed. I think I have said all that I want to say on this subject and others should contribute their POV as to whether the first phrase stays and if the sentence "As a first heuristic ..." is worth keeping. --Philip Baird Shearer 17:14, 14 February 2007 (UTC)
No, you cannot enforce your own personal process over consensus. Please follow regular process. I am reverting you for that reason and that reason alone. There's no hard feelings, and I am assuming that you are acting in good faith. Feel free to come back and discuss at any time. If possible, please provide solid sources or reasoning for your position at that time as well. (see also: Bold revert discuss as a means to effectively implement your reasoning.) --Kim Bruning 18:04, 14 February 2007 (UTC)

Let me expand a bit on my reasoning for that last revert, just in case it isn't fully described elsewhere yet.

We shouldn't typically accept "It's community consensus" as an argument, since it is impossible for a single person to know the thoughts of the entire community. At best, a single person can only put forward their own thoughts and reasoning when challenged, even if those thoughts already constitute a sytnthesis of the concepts of the entire community. Consensus is mostly formed during wiki-editing of a page, so there is also no need for the "it's a consensus" argument.

Note that reverting good faith edits without providing reasoning on the talk page holds other people hostage to your whim, which is probably not what you intend. Let's agree on maintaining the process as described in the flowchart here. Discuss until we have some modicum of (temporary) agreement, and then and only then taking it back to the page itself. (though I have no trouble with "so how does this look! :-)" type edits to the page, even during discussion.)

Something like this has also been written at WP:BRD, but perhaps not to that length.

Philip: Could you put forward your personal argumentation please?

--Kim Bruning 18:13, 14 February 2007 (UTC)

I do not agree with "therefore these processes may have been somewhat misdesigned." (which leads to deleting the second sentence as well), and nothing you have written here has convinced me othewise. As I have said there is a clear community consensus to use these processes and your presumption that "wide community consensus is irrelevant, as communities are quite capable of becoming pathological and self destructive if misdesigned" does not convince me that you are correct on this one. --Philip Baird Shearer 00:19, 15 February 2007 (UTC)
I stated (for context): "...consensus-building can be unwieldy due to the sheer number of contributors/discussions involved, therefore..."
Why do you accept the reference Dunbar's number, but then reject the conclusion? That does not compute... Wait... that wording wasn't really very clear on that either, now was it. How's this? --Kim Bruning 00:37, 15 February 2007 (UTC)

I still do not agree with your wording. The reason for this is that one can only reach a consensus when the group who hold differing points of view strive to reach a consensus and are willing to compromise. In the case of a Westminster cabinet there is a vested interest compromising for two reasons. One, there is a danger of collective loss of decision making (they loose the next election and the other lot get in), if they are not seen to agree, or secondly an individual who does not accept the decision of the others must resign from the cabinet and the loss of a government job is quite a large stick. But in a Wikipedia debate there are no carrots and sticks available to other members of a group to insist that their views prevail (and few to make a majority consider a compromise with a minority) ... But you know all this because it has already been discussed at great length in these talk pages. So to say that some processes are missaligned makes an assumption that there is an correct method within Wikipedia to reach a consensus which is not true. Now we could go off on a tangent and discuss gift cultures, the Cathedral and the Bazaar etc, but I do not think that is necessary nor do not think that the introduction of the article "Dunbar's number" at all helps in this context, because I do not think it is relevant as to why a guillotine on what amounts to filibustering is desirable on some wikipedia administrative processes. --Philip Baird Shearer 17:55, 15 February 2007 (UTC)

Ok, Let's get back to Dunbar's number in a minute. But first, why do you bring up Filibustering? --Kim Bruning 18:30, 15 February 2007 (UTC)

Consensus of few overruling consensus of many

An issue I keep noticing with our guideline pages is the disparity between what the guideline says, and what editors actually do. A consensus of 10s of people can be enough to solidify a guideline, even when it is at odds with the editing habits of 1000s of people. Then, once the guideline is written, it only takes a few committed people to protect that guideline from change. Do we have any way of dealing with this proglem? - Peregrine Fisher 01:14, 17 February 2007 (UTC)

The typical way is to simply define consensus as the 1000 people who are doing it their way, and/or ignoring whatever games this particular set of guideline writers are playing.
Not exactly ideal, I know. I do keep pointing out that whatever is in the project namespace sometimes has little relationship to what people are actually doing, due to this tactic.
A better way is to demand that people write descriptive (not prescriptive) guidelines. Maybe we should finally write that down in a policy. :-)
Finally, the best way is to simply patrol guidelines, and make sure that isn't happening. It's a big world out there though. Could you tell me which guideline(s) you've encountered it on? --Kim Bruning 01:35, 17 February 2007 (UTC)
We really ought to find a better word than "guidelines", because that isn't what they are. They are, at best, common practices, and in some cases they are merely what some people think should be common practices. Perhaps the most egregiously mis-categorized "guideline" that I am aware of is WP:PNSD which is really just an essay, and a "common opinion," since there is no "common practice" or anything really actionable involved. As the discussion on that page indicates, the reason that the main proponent wanted it to be a "guideline" was because people wouldn't listen if it was merely an essay. Someone else on that page (who favored it) said the purpose of making it a guideline was to be able to shut down polls. In other words, it was an opinion that was turned into a "guideline" for purposes of intimidating editors who dared to have a different opinion. Or, looked at another way, it was a supposedly "descriptive" statement that is called a "guideline" to discourage any change in what is being described -- and it thereby becomes prescriptive. 6SJ7 06:06, 19 February 2007 (UTC)

I would not call the "default" behavior of many Wikipedians to be a consensus. A consensus suggest that pros and cons were considered, that there was discussion, that people were aware there were even options. In many cases people just repeat what they saw from another place, and not for any major reason. Common practice can represent a consensus, but not always. -- Ned Scott 08:38, 19 February 2007 (UTC)

  • There was a big discussion about this at a recent naming convention, where members of a wikiproject claimed that since the project had a consensus, the matter was decided even though people who were not members of that project disagreed with the outcome. It took several months of debate, but eventually the wikiwide consensus overrode the project-only consensus. >Radiant< 10:14, 19 February 2007 (UTC)

Asking the other parent

Hmmm, interesting addition. Could our friendly IP editor give some more background information or examples of situations where this applies? (though I might be able to think of one or two myself, perhaps :-) )

--Kim Bruning 01:57, 25 February 2007 (UTC)

A Miracle Happened Here

This article gives a flowchart. It says "Find a reasonable if temporary compromise" in one box. Good plan. But perhaps that should be labeled "A Miracle Happens Here".--Blue Tie 11:53, 30 March 2007 (UTC)

This page is nearly useless. --Blue Tie 14:56, 5 April 2007 (UTC)

  • What do you think is missing? >Radiant< 15:39, 5 April 2007 (UTC)
Clarity and reality. The problem starts with the heart of wikipedia. "The (internet) encyclopedia that ANYONE can edit". It then moves into the sense that wikipedia is NOT a democracy (but really it mostly is). The denial of the democracy here leads to a sense of "no voting". This is enhanced by a desire for "collegiality" on articles. When everyone is super pure and super good this will work. For normal people it will not.
Example: Two people disagree on a paragraph. Where is the consensus? None? Does the paragraph get struck? No consensus for that. So we ask for a third opinion. The third opinion gives a real third opinion not a "tie breaking vote". What is the consensus now?
Example: One lone expert holds out against and then gets overridden by 10 other ignorant editors. What is the consensus?
Example: Five editors guard an article and revert changes instantly. Between them they have at least 12 reverts a day and maybe 15. They chase off other 20 or 30 other editors, one at a time. What is the consensus?


A more realistic guideline would say "Consensus is what someone can get a majority of people who are editing this week or this day to agree to include or exclude in the article. The majority will decide issues through the use of discussions as possible or as they desire and decide the issue through reverts and volume at other times." This may sound like the words of a bitter person. I am not bitter. I feel fine. I am trying to honestly say what I think is wrong with the consensus article. It does not define consensus very clearly. Without a clear definition, it is useless. As evidence, just look at how frequently it gets cited in debates and the nature of those cites. It is almost never cited until someone announces the achievement of consensus. Upon which point someone else is likely to disagree and then the article gets cited as a touchstone for both sides of the issue. On the other hand, if a cite for Attribution is brought up, it typically results in a recognition by all of the issue. That is because WP:V or WP:ATT are clear. WP:CON is not. --Blue Tie 16:01, 5 April 2007 (UTC)
  • I think you are overlooking the important point that, barring a small but vocal minority, Wikipedia editors tend to be reasonable, well-meaning folks who can compromise on issues and can change their mind over discussion. If you make that assumption, consensus is not problematic; if you deny that assumption, Wikipedia is doomed. Forming consensus only becomes problematic when sufficient unreasonable people become involved to drown out the reasonable debate. This does happen on articles on controversial issues, but there is no real way to avoid that in a wiki. >Radiant< 16:05, 5 April 2007 (UTC)
Well, it only comes to play on controversial articles. And as you say, there is nothing that can be done about that.
I have seen consensus work on a controversial topic only once. It was beautiful. It was rare. I know it happens. It requires a very strong assumption of good faith as a minimum and often it requires a level above good faith -- active goodness of heart. I believe I have personally witnessed at least one case where the "sufficient reasonable people" have both: 1) come to their own consensus and value it now above new contributions -- and thus become unreasonable, and 2) they have been attacked for their perspective until they no longer believe that anyone but them has good faith. But let's suppose that this policy is not for those contentious times but rather it is for collaborative times. In collaborative times, a fuzzy guide will work fine, but this guideline also tries to cover consensus when people are disputing and fighting. That also muddies the water. Perhaps the WP:CON page should say: Consensus rarely emerges from a fight and the presence of strong and on-going disputes is an indication that there is no consensus. No one can claim consensus under those conditions -- they can only claim majority opinion. Saying that would be anathema though. --Blue Tie 16:20, 5 April 2007 (UTC)
Current best estimate is ~3K controversial articles where consensus is unlikely to work, and ~3M articles where consensus is very likely to work.
Slogging through working on guidelines for "just" the 3M articles is already way too much work for me, especially since the system is not really designed to help people describe process. (actually make that: it was never designed, full stop. Perhaps sharpening the axe we work with would be a good idea, hmm... )
Sometime in my copious free time, I'll try to figure out what works on the remaining 3K. AFAIK, nothing has been written on the latter topic, yet. --Kim Bruning 17:31, 5 April 2007 (UTC)
Kim, that is an interesting response. On one hand, the guidelines that reflect the majority of the instances (I am accepting your numbers as correct) is clearly descriptive. But in those cases, it is not needed. When it is needed is when there is a problem. The 3k need the guidelines more than the 3m do.
Incidentally, I do not know if it comes thru but I respect almost every word you type. Even though I am sure we do not generally agree, I think you are really fantastic. I am a fan. --Blue Tie 17:51, 5 April 2007 (UTC)
  • It is true enough that consensus rarely emerges from a fight. However, the presence of a dispute does not necessarily mean there's no consensus. In particular, a few controversial articles have a reasonably stable state but get "attacked" by outsiders every week. This is rather unfortunate, and one of the things Stable Versions (WP:STABLE?) was supposed to remedy. The problem with "disputing and fighting" cases is that there really isn't much we can state in policy that would actually work against that, because of the wiki model. >Radiant< 08:47, 6 April 2007 (UTC)

Flowchart: Wait .... how long?

Very nice flowchart, but unless it states for how long the wait is, then it makes the whole thing rather un-followable! --Rebroad 11:01, 10 April 2007 (UTC)

It depends on the number of people participating on the talk page, the size of the article, etc. We need to seriously cut back on detailed rules like "wait 2 days" and emphasize more common sense rules like "wait and see what other people say". — Omegatron 13:25, 10 April 2007 (UTC)

The wait can be for three reasons

  1. First, your edit is too new. No one has seen it yet.
  2. People have seen it, but the person who will object hasn't seen it yet.
  3. Your change has been consensus for years, but consensus can change, and now someone disagrees.

For sanities sake, in all three situations, we just say that your edit "has consensus" until someone finally changes it. (though arguably, in the first case, you don't know for sure). --Kim Bruning 02:52, 12 April 2007 (UTC)

Not policy? Overridden by other guidelines?

Errr, that's not right afaict. CentrX, care to explain?

(And if you reply with something like "there's no consensus for this to be policy" ... well... that would be funny ;-) )

--Kim Bruning 02:48, 12 April 2007 (UTC)

I would go farther than policy, even. The "wiki process", which I take to be essentially the same as consensus decision making, is a foundation principle, along with NPOV. Of course everyone knows this, I just want to point out the irony of a foundation principle that is not a "policy". CMummert · talk 02:57, 12 April 2007 (UTC)
The "wiki process" is not consensus as defined here, and consensus does not not over-ride those other policies. —Centrxtalk • 03:11, 12 April 2007 (UTC)
We should have a foundation-issue tag, perhaps? --Kim Bruning 03:05, 12 April 2007 (UTC)
Consensus is always in reference to the encyclopedia. There cannot be a consensus to over-ride neutrality, or freedom, or even simple reasonable deductions about what is an encyclopedia. Consensus as the basis of everything is meaningless, because consensus requires absolutely a goal. There is no consensus without a goal. In any case, a notion of "consensus" being policy is wholly separate from whether this page itself is policy on the order of any policy. This page precisely fits into what a guideline is, and is nothing like neutrality or verifiability in its firmness of what is acceptable on Wikipedia. —Centrxtalk • 03:11, 12 April 2007 (UTC)
Also, see above, at section #Policy which is the entire extent of the discussion preceding the tag change. —Centrxtalk • 03:14, 12 April 2007 (UTC)
If the content of this page doesn't reflect the accepted policy on consensus, I think the appropriate thing to do is to change the content to reflect the policy, rather than leave the content alone and remove the policy tag. I think that there should be some page, somewhere, that explains the policy on consensus decision making, and this is the natural location for it. CMummert · talk 03:18, 12 April 2007 (UTC)
The problem is there has never been any firm or good definition of consensus. It is appropriate to say that "Decisions on Wikipedia are made by consensus" and tag that statement alone as a policy (i.e. on some "Wikipedia:Decisions" page), but there are problems with making an entire page about "consensus" be a "policy" because any definition always runs into infirmities and holes of the kind that belong with a guideline. We can say "Be bold" is policy too (and that is part of the "wiki process" cited above), but there are specifics and exceptions that make any page about it a guideline. We can say the same for Wikipedia:Etiquette; being polite is a good policy, but that's just an extrapolation of Wikipedia:Civility, with advice and exceptions. The idea of Wikipedia:Do not disrupt Wikipedia to illustrate a point can be said to be "policy" (could it ever be policy that it is okay to disrupt Wikipedia?) but the specifics of it fall under guideline. —Centrxtalk • 03:27, 12 April 2007 (UTC)
So basically, what you are saying is that a lot of pages describing fundamental aspects of wikipedia are/should be marked "guideline", while less important pages are "policy". Back to square one. *sigh* . Well, maybe not quite. You're not going to revert WP:5P back to essay status as well, are you? --Kim Bruning 03:32, 12 April 2007 (UTC) perhaps we should abandon the policy/guideline/essay system, since it doesn't prioritize very well.
Not so much importance, the pages are categorized by primacy and firmness. While ethereal "consensus" has higher primacy, this page itself is simply filled with some good suggestions and is somewhat essay-like; it is no where near canonical or comprehensive or firm. WP:5P, on the other hand, is the most prime of all and because of its simplicity also has the luxury of being firm (just as the sentence "Decisions are made by consensus", formulated in the pillars in "find consensus", has the luxury of being relatively firmer policy). —Centrxtalk • 03:51, 12 April 2007 (UTC)
By that criterion, if a particular principle is divided among multiple pages, would the "tag level" of all the pages drop? --Kim Bruning 04:09, 12 April 2007 (UTC)
It depends on what you mean. An important principle still warrants a super-page that generally encompasses the others. —Centrxtalk • 00:21, 14 April 2007 (UTC)
As far as I'm aware, this is just about the only page that actually takes the time to explain any part of the wiki-process at all. Consensus is a key component of that process, of course.
I agree that consensus by itself doesn't do much, but in many ways it is inherent to the fact that wikipedia is a wiki. A wiki by itself also doesn't do much of course, hence the decision to use this particular wiki to write an encyclopedia. With those basics down, we have the concept of "wikipedia" mostly covered. Throw in a GFDL, and off we go. :-)
Note that verifiability is currently under contention, and people are using the consensus decision making process to decide its ultimate fate.
NPOV is not so much under contention, but that's because it's a foundation issue. Even so, note that even the foundation issues are subject to debate and gradually changing views. (And, in fact, an update to them is required right now)
--Kim Bruning 03:23, 12 April 2007 (UTC)
Wikipedia:Be bold in updating pages, Wikipedia:Ignore all rules, and (much of) Wikipedia:Resolving disputes are all about the wiki process. The concept of verifiability is under no contention whatsoever; the only contention is about replacing or merging the page. Again, the issue here and for Verifiability is the page and where it belongs; there isn't a guideline on Wikipedia whose principle is not sound policy. Neither verifiability and neutrality are under contention because they are fundamental to a wiki encyclopedia, sine qua non to Wikipedia. Consensus cannot over-ride these things; consensus cannot annihilate Wikipedia, it can only fork it. —Centrxtalk • 03:44, 12 April 2007 (UTC)
On wikipedia, policy is just a word that means that something has general consensus (see the policy tag "a wide acceptance among editors"). This doesn't mean that something with a strong general consensus isn't very important. Basically the strong general consensus is what makes them important, and that's why we respect those concepts.
Incidentally, did you notice the discussion about "verifiability, not truth?"
--Kim Bruning 03:47, 12 April 2007 (UTC)
Guidelines also have general consensus and wide acceptance, but they are nevertheless different sorts of pages. There is no guideline on Wikipedia that we should not follow in principle, but all of them, such as this one, are not so strong or even good in their particulars. —Centrxtalk • 00:24, 14 April 2007 (UTC)

how can this page be improved

How can this page be improved to better reflect the true policy? I read through it again, and some parts seem very appropriate. For example, the lede section is very good. It looks to me that sections 1 and 5 are the most problematic. CMummert · talk 11:21, 12 April 2007 (UTC)

That's odd. 1 is supposed to summarize Consensus decision making, although the current version might be rewritten more constructively. 5 is supposed to summarize WP:BOLD (though may need to be better integrated into the text). (The concept there is simply that wikipedia is a wiki, not a discussion site. Use the fine wiki! )
I would think 3 and 4 would cause the most issues.
But that's all just me :-)
Could you explain issues you see with the sections you indicate?
--Kim Bruning 11:46, 12 April 2007 (UTC)
For section 1 (Reasonable consensus-building), it's mostly the tone that I dislike. The second para is not so great.
For section 5 (Note on use of discussion page), there are several issues. The title needs to change (is it a footnote, or a section?) and it shouldn't start with a bullet point. Those are very minor. The prose is choppy, which is less minor. The one para mixes two issues: being bold, and using the talk page instead of just edit summaries. This could be split into two pieces.
I agree that the other sections could also use some work. CMummert · talk 11:58, 12 April 2007 (UTC)
That's probably fair criticism. Have at it! :-) --Kim Bruning 12:20, 12 April 2007 (UTC)

How much makes a consenus?

Hi all. I have a question on the application of consensus. In a feud by a user I'm curious at knowing how many contributors are needed to make a consensus. I understand that most times arguments engage several contributors, however, in some cases, arguments include small numbers of contributors. In these cases, should the argumentors ask help from others or discuss it amongst themselves and have the minor party conceed? Cheers. Captain Scarlet and the Mysterons 12:25, 14 April 2007 (UTC)

Everyone needs to agree (or at least needs to agree to disagree). Else try asking for help at Wikipedia:Editor assistance, they can help you sort out consensus issues. --Kim Bruning 15:15, 14 April 2007 (UTC)
See Wikipedia:Resolving disputes. Also, consensus is about creating an encyclopedia, not about the numbers on opposing sides. —Centrxtalk • 17:05, 14 April 2007 (UTC)

Removing concept that Consensus subject to other guidelines

I think that that is a "suicide clause". Consensus is very fragile, more fragile than even democracy. If anything else stomps on it, it ceases to function. --Kim Bruning 23:28, 16 April 2007 (UTC)

The not formally "verifiable" link

The not formally "verifiable" link to the mailing list for a "definition" is problematic, as is the tone that seems to imply that on Wikipedia, Consensus == Majority/Supermajority

First, we should use a "verifiable" definition of consensus, not one from a mailing list, especially as a primary definition. If we need the link to the posting from the mailing list, it should be posed as a contrasting opinion. The formal definition of consensus includes not only absolutely no voting (the Society of Friends - i.e. the non-religious arm of the Quakers - have been doing consensus-based decision-making for hundreds of years without ever once voting) of any kind, ever, under any circumstances, as well as making sure that every opinion, no matter how contrasting, is heard. When an organization shifts to or uses majority/supermajority voting mechanics in place of consensus, an organization has then stopped using consensus.

I strongly feel that the implication of this policy that consensus and majority/supermajority are equivalent is dishonest and disingenuous. It potentially leads to Deletion Review closers and other admins thinking that nose-counting is consensus-making when it's clearly not, and it leads to us drifting away from the stated point of consensus.

I don't really mind whether policy is changed to reflect the reality of procedure (i.e. some processes use voting instead of consensus) or whether we start using actual consensus again, but the situation as it stands where we at Wikipedia say that we do one thing (consensus) but in fact do another entirely different thing (voting, in some cases), is definitely not okay.

I'll try to furnish some actual edits soon, boldly, and then we can revert and discuss them as needed. --MalcolmGin Talk / Conts 21:46, 17 April 2007 (UTC)

There's already been some attempts to clarify the difference. Tread lightly :-) --Kim Bruning 22:01, 17 April 2007 (UTC)

If no one cares enough to comment, is there consensus?

See Wikipedia talk:Requested moves#How much discussion is needed?; a move request was closed as no consensus when I was the only one commenting, and I supported the move. --NE2 10:07, 20 April 2007 (UTC)

In some cases, like requested moves and {{editprotected}} requests, the admins are looking for evidence of consensus, and may refuse to take action if there isn't any. This is an acceptable, conservative approach to administrative actions. In this case, I think the admin's remark on the article talk page means that he or she is looking to see more comments at the wikiproject talk page. Try restarting the conversation there. CMummert · talk 12:56, 20 April 2007 (UTC)
If no one commented when I posted a link there, why would anyone comment this time? Note how many threads without response are on Wikipedia talk:WikiProject New York City Public Transportation, and the lack of comments at Wikipedia talk:WikiProject New York City Public Transportation/station names. What you're suggesting is compulsory voting, an impossibility on a volunteer project. --NE2 13:02, 20 April 2007 (UTC)
If you feel the request by the admin in question was unreasonable, why not raise the issue on his talk page to see if he will reconsider or at least explain his thinking? If that doesn't lead to a good resolution, you are always free to request the move again with an explanation. CMummert · talk 13:40, 20 April 2007 (UTC)
I already discussed that. I think my move request was pretty clear in showing what the common name is; why would requesting the move again lead to a different result? --NE2 14:01, 20 April 2007 (UTC)
I wish you had indicated from the start that the closing admin in question has already explained their reasoning [2]. The advice given in that comment seems sound. CMummert · talk 14:12, 20 April 2007 (UTC)
I don't understand what I am to do. I've already tried to get others to comment, to no avail. Any more would probably be canvassing. --NE2 14:28, 20 April 2007 (UTC)
Specifically, someone has said that he supports using the common name, but is not sure exactly how one would find that. Would I be allowed to specifically ask him to comment? --NE2 14:53, 20 April 2007 (UTC)
It is all right to do so, particularly if you won't do it too often. Just do it once, and hopefully he will go and support your idea. No mass mailing to everyone, though.--Kylohk 15:30, 2 May 2007 (UTC)

Local or global consensus

I am trying to understand how consensus interacts with guidelines, and in particular whether there is such a concept as "local consensus". Can the editors of an article agree that the article ignore particular guidelines? Or can other editors, citing the guidline, breeze through and enforce it anyway? (Note: I am talking about "guidelines" rather than "policies"). I am as guilty as anyone of editing articles I've never seen before to enforce guidlines to meet WP:EL, Manual of Style, etc. But what if the other editors of the page disagree in consensus? Is there some "Wikipedia wide consensus" that applies over and above any particular article's dedicated editors? Notinasnaid 07:12, 3 May 2007 (UTC)

There is absolutely at least for WP:NPOV, WP:V and WP:NOR, but for guidelines and less frequented pages like the Manual of Style subpages, it is not necessarily so clear. Ultimately, though, if someone really wants to remove an external link from an article or conform to the manual of style, you are going to need to convince them of why that should not be done, in the interests of having a good encyclopedia article, and if there is a good reason in the particular case it might be good to make a small change to the guideline page about it. Guidelines are best practice written down. —Centrxtalk • 15:27, 3 May 2007 (UTC)

I do mean written down, well defined, guidelines. (But not policy). For example WP:EL is a guideline. What if a consensus of editors on a page agree they will ignore the guideline (not policy). Do visiting editors have any right to ignore their consensus and do what the guideline says? Notinasnaid 15:31, 3 May 2007 (UTC)

What is the particular issue? If the consensus of the article editors is to add some advertisement, that is absolutely not appropriate for the encyclopedia. If, on the other hand, WP:EL has some obscure provision about Usenet links not being appropriate on Sundays--which few people have probably read anyway--despite the eminent relevance of the link to the article, then the link should be included and the guideline should be softened. —Centrxtalk • 15:36, 3 May 2007 (UTC)

There isn't a specific issue. My question is an overspill from another discussion. I don't want to complicate this with a specific issue, because I'm trying to understand how Wikipedia processes work, to solve future issues. Do guidelines (such as WP:EL or the manual of style) trump consensus, which is a policy? If guidelines do trump consensus, then how do I deal with this consensus of editors who have made up their mind? Can I warn each of them who reverts me, immune from 3RR warnings myself? If not, how does this work? Notinasnaid 16:18, 3 May 2007 (UTC)

No, guidelines don't trump consensus, but I think you're asking the wrong question here. Guidelines should have rough global consensus to begin with, or they shouldn't be guidelines. They should also not conflict with policy, which should absolutely have consensus (or be mandated from on high, in the case of a few key provisions). But locally, there is wiggle room if there's a significant reason to deviate from the written guideline. If this happens often enough, then the guideline should be modified to describe the exception (they should be descriptive, not proscriptive). But I can't think of any guideline that allows you to violate the 3 revert "rule", no, as long as the edits you are disagreeing with are not vandalism or violating core policies. -- nae'blis 17:25, 3 May 2007 (UTC)
What he said. The general answer is that guidelines can and do have exceptions. >Radiant< 09:42, 4 May 2007 (UTC)
What if there's a proposal based on a guideline, and the only comments saying why the guideline based proposal shouldn't go through are the gist of "I don't the agree with the guideline" and not anything about why it doesn't fit in a particular scenario. Then does the guideline beat local consensus (or lack thereof)? Miss Mondegreen talk  07:10, May 8 2007
  • Quite possibly. Could you please be more specific? >Radiant< 08:16, 8 May 2007 (UTC)
Here's the case that got me interested in consensus. Miss Mondegreen talk  09:06, May 9 2007
*pulling out a 10-foot bargepole* Eeeuw! That's a requested moves spam :-/ . Nothing to do with consensus at all! The amazing thing is that it actually almost worked here, and people point out that the page shouldn't be moved, and provide solid reasoning. Of course, requested moves only entrenches people in one option or another, so you've basically just sunk any chance of moving it. That's the risk you take and the price you pay for using RM. :-/ --Kim Bruning 14:10, 9 May 2007 (UTC)
Please AGF---once a RM failed, there's no way I'd try to move over that, and that's not what I was trying to do. In hindsight, I should have just been bold and moved it and sighted the relevent guideline, but that's not what I'm asking about.
I see this issue of local v. global consensus everywhere, and this was one of the first cases that I was involved in that brought it to my attention, and it's a great one to link to, because it was made clear that the opponents of the move didn't agree with the guideline--one flat out said that my case would work except that he didn't agree with the guideline.
So I'm asking what I or someone else can do if a move or Afd or something splits like this in the future, and it's clear, as clear as a user saying that they disagree with the guideline or policy... Then what happens? Does local consensus overrule global consensus? Can someone close in favor of global consensus if no one raises a legitimate issue with the guideline or policy and how it applies in the case in question? The theory of global consensus is all well and good, but I keep seeing again and again 3 or 5 people showing up and saying "I don't like ____" and overruling Wikipedia's guidelines and policies, and asking what can actually be done about this. Pretend you're closing the RM and those are the only editors who have commented and those are the comments you have to go by. What happens? Miss Mondegreen talk  06:34, May 11 2007
The guidelines form the starting point for a discussion. If people then point out that the guideline is silly "in this particular case", then perhaps it may just be so. At that point, take a look at the guideline, and perhaps modify that instead. :-) . In the example case you point out, it seems like both titles have only ever been released in japan. Insisting on using an English translation does seem just a little odd then, and one person points out that the translation does not appear to come from a reliable source , and that's a guideline too... Hmmm... --Kim Bruning 11:26, 11 May 2007 (UTC) In general this class of problem is mostly "political" in nature, since we do have redirects. If we could have pages under two or more titles with equal priority, this kind of discussion would be entirely unnecessary :-P
Unreliable? The translation came from a Nintendo press release years ago when they were planning to release the game in English. The translated name is copyrighted and trademarked by them. And I wouldn't have put it up for a RM, because the page has a ton of redirects, except that there's a nav box. When the page is in the nav box (which is on and off and that is for political reasons), the page was in their by the Japanese name, which didn't help people in terms of navigation, especially since it was the only page to go by the name. I was just confused by the whole thing. Everyone insisted on it being an official name, even though the guideline says that it only had to be a common name, and it wasn't just a common name, we had a Nintendo press release as a source, and given that that's a Nintendo game, that seems reliable to me--and none of the other articles had issue with it. I just guess I have an issue with two or three people being able to say "eating ice cream is against policy" or "I don't like policy" and if not enough people are there to stop them, policy or guideline is thrown by the wayside. And since wikipedia moves so quickly, I do see this happening and if it's subtle enough, or not too bad, it's often just left to set. Miss Mondegreen talk  04:58, May 12 2007

Continued appeals to "Consensus has been reached"

I'm not quite sure where to ask this, but on a talk page I've been involved in, there are a group of editors who consistently argue against any change based almost exclusively on the claim that "consensus has been reached", despite the fact that there are currently several editors who are trying to discuss changes to this, and despite the fact that for several months, there have been continued debate over the issue.

How do you deal with this? Cogswobbletalk 16:30, 5 May 2007 (UTC)

  • You can also try moving to a more inclusive stance. Instead of trying to decide whether A or B is the correct position, try to come up with a proposal that includes both A and B. For instance by saying "X says A, but this is refuted by Y who says B." or something similar. If you can come up with a wording that both groups agree is accurate, you have created something that can be truly enlightening. Often arguments get bogged down because both sides believe they are protecting the truth. In cases like that, both sides need to move to a broader understanding, which is recognizing that people can have different interpretations of the same facts and events. The issue, instead of determining the truth, is determining how much weight to give each perspective. This can be settled by looking at the citations presented by each side. This moves the discussion from "truth" to "scholarship". When the discussion moves from truth to scholarship it separates the zealot from the scholar. Zealots don't seem to last long here. They end up blocked, or leave on their own accord. -- Samuel Wantman 07:36, 8 May 2007 (UTC)

You disagree with the situation, so possibly there might not be consensus, and those people may just be confused. --Kim Bruning 11:33, 11 May 2007 (UTC)

Hard to know. After all, "What is consensus?" Never defined, it is not even necessary, but it is constantly referred to. Probably the least useful policy on wikipedia because of the vague nature of its definitions and content. Second is WP:MEAT. If we have a bunch of people who agree with one another, editing the same articles... are they meat puppets or not? I saw two guys who are friends and who edit in agreement, get blocked for a week for WP:MEAT violations, even though each was acting to his own personal direction. So... What makes some meat puppets and others not meat puppets? And, in particularly, when we are looking at "Consensus", are all the people who agree with each other Meat puppets? Again, this is another large area of vague interpretations. --Blue Tie 14:06, 11 May 2007 (UTC)
  • Hm, that's interesting. The meatpuppetry policy is intended to prevent a deletion debate on e.g. a website being flooded by users of that website, registering Wikipedia accounts solely to prevent that article from being deleted. Such accounts should generally be ignored rather than blocked. I would like to know which meat case you're referring to, because it appears to have been an improper block. >Radiant< 14:13, 11 May 2007 (UTC)
Yes, it is interesting. Here is the block notice. Here is the blocking admin's explanation -- which seemed pretty thin to me. Here and here I requested further information because it appeared to me that the standards enacted are regularly violated by other people who "guard" a particular article. It looked like a double standard. In reply, I was essentially told to go away. So I did.
Interestingly enough, the article under discussion there was up for a review for featured article. There was a poll taken. I think that 50% said it should not be featured article and 50% said that it should. (In fact, it might have been 17 for and 18 against -- I would have to go back). Many of the problems of the article that would have technically denied it featured status were not actually addressed. Despite the vote, and despite the pesky technical issues there was apparently a consensus for the article to be featured because that is how it was closed -- although just a couple of weeks earlier, Kim, who contributes here, had closed it as not featured (and then re-opened it when someone complained that the close was improper). Isn't consensus weird? But at least its not a democracy... you would have been so proud about how the poll was ignored!
And speaking of weird, the individuals who were blocked for one week, had brought a complaint against an Admin who was the chief editor and contributor on that article. This same admin had been reported for 3rr violations in the past a few times and always got off by the narrowest of constructions of the policy, while to my eye, these two non-admin editors who were questioning some of his actions got hit with a week for rather flimsy reasons. Interestingly, not long after that incident, Raul, who is a strong supporter of that Admin -- to the point of egregiously insulting other editors, was himself blocked for actively and unapologetically edit warring on the article. 6rr and personal insults to other editors in less than 24 hours so he got a 12 hour block overnight while he slept. Pretty rough, I know, but fair is fair. I am sure that had he been a regular editor instead of a member of arbcom, he would have just been told to be careful, but I know they had to set an example, so they were fearless and stern. Later, when a question about the process and the decision to close that article as featured was brought up on the FARC talk page, the Closing Admin closed off all discussion and told everyone to appeal the decision to the boss of "featured article status". Who is ... Raul! Of course, there is clearly no evidence here of a cabal, but you know how kooks get... they might think all of this looks a bit suspicious. Thankfully though we all know that all of this was an unbiased process of consensus, just like the policy says. --Blue Tie 18:22, 11 May 2007 (UTC)
Insofar as I was involved, I discovered that particular actions I took did not have consensus. Therefore I undid those actions, as is correct on wikipedia. After that I stayed away from the Global Warming article, and from the discussions, simply because I lack the time to deal with it (argh!) ^^;; --Kim Bruning 18:12, 12 May 2007 (UTC)
(In case of misunderstanding -- I think you are blameless. That is your habit -- you are always doing things well.)--Blue Tie 18:44, 12 May 2007 (UTC)
  • This seems to be a rather complicated issue. On the one hand making sockpuppets is not usually a problem, on the other hand using them to astroturf an issue is. It's a bit late to do anything about it; in the future I would strongly recommend (if you cannot reach an agreement with the admin in question) to bring up the matter on the admin noticeboard. I would also recommend being less verbose in doing so, brevity is wit and all that. Durova's reaction is rather curt but this may be caused by the verbosity. >Radiant< 15:02, 14 May 2007 (UTC)
All excellent recommendations. I was not upset so I did not pursue it. But it seemed a bit wrong. Thanks for the notice on my talk page. Also... question: What is meant by "astroturf"? In this case it is not sockpuppets, it is two guys who work together editing. I think one is a radio talk show guy and the other is his producer or something like that. I am not sure of the details. They agree with each other though. --Blue Tie 01:00, 15 May 2007 (UTC)

And yet existence of (temporary) consensus can be detected unambiguously. --Kim Bruning 15:17, 11 May 2007 (UTC)

  • Astroturfing. Basically using sock- and/or meatpuppets to make an issue seem more supported than it actually is. >Radiant< 08:23, 15 May 2007 (UTC)
  • Kind of like this. Not Radiant, honestly 08:23, 15 May 2007 (UTC)

To the original question, WP:CCC addresses this quite directly, "A small group of editors can reach a consensual decision about an article, but when the article gains wider attention, members of the larger community of interest may then disagree, thus changing the consensus. The original group should not block further change on grounds that they already have made a decision." Why not just go there? Dhaluza 10:39, 15 May 2007 (UTC)

Well the original question is misleading at best, an outright lie at worst. What the editor (because there was only one editor who used the phrase "Consensus has been reached") at the page was telling Cogswobble, was that one of his points was addressed previously and that he should read the archives, because no one wanted to keep repeating themself. Cogswobble made a few suggestions, some of which were incorporated into the article. At no point in time, was discussion ever cut off. "Consensus can change" does not require editors to constantly repeat all arguments once every "Johnnie-come-lately" comes to an article. Or does it? Ramsquire (throw me a line) 17:02, 3 June 2007 (UTC)
If people don't want to repeat arguments, they can give links or precise directions to exactly where the arguments are. I don't think it's fair to just say that the arguments are in the archives "somewhere". The other person might read the entire archives and not find them because they don't see them as convincing arguments. If people get tired of pointing to the same arguments over and over again, they can make some sort of FAQ or pointer to them. "Convincing arguments are in the archives somewhere" is not a valid argument because it is not refutable. --Coppertwig 17:26, 3 June 2007 (UTC)
Your FAQ suggestion is very interesting because one was created for the page, and Cogswobble deleted it. He felt that it was being used to suppress discussion. In addition, said links were given over and over, and over again. This just seems to be a sour grapes discussion thread. Also of note is that all of these so called editors who felt that consensus had changed have now moved on to other aspects of the article, and the original text reads exactly the same as it did before they got there. A success for consensus actually being reached I would say. Ramsquire (throw me a line) 17:45, 3 June 2007 (UTC)
Undelete the FAQ... And /FAQ pages seem like a good idea in general to me. :-) --Kim Bruning 18:01, 3 June 2007 (UTC) Sounds like a good "alternate rule for heavily edited pages"
The problem with the FAQ is that there were a group of editors who were pretty much using the FAQ, and appeals to consensus, as the primary argument against any change or even discussion about an issue. Their arguments typically went like this "There's already consensus, and I'm sick of repeating this over and over to different editors for the past eight months". Note the hostile attitude of this editor in his response above - "Well the original question is misleading at best, an outright lie at worst". The FAQ was and is being used to discourage discussion. Cogswobbletalk 18:03, 4 June 2007 (UTC)
Well, if you take issue with certain entreis in the FAQ, you can edit them (and go bold revert discuss worst case), or you could challenge the assertions in the FAQ. Would either of those be a good start here, or is there more to this? --Kim Bruning 18:57, 4 June 2007 (UTC)
To Kim Bruning...Yeah, there is a lot more to this, like 4 archives (including 2 RfC's, an ArbCom request, and confirmed and suspected sockpuppetry) of discussion on the Fox News Channel talk page (which would seem to contradict this assertion that discussion was discouraged). Again, discussion wasn't discouraged when we adopted some of Cogswobble ideas, but when we disagreed with his ONE point, because it was addressed previously in detail (and in two RfC's), then we are discouraging discussion. To Cogsobble-- You can't have it both ways. I'm sorry you came in at the end of the discussion, but you have been provided links to the archives, and the FAQ. And I will give you credit you have been one of the more reasonable Johnie-come-lately's (not intended to offend, it's just that its kind of innaccurate to call you a latecomer). Whenever you raise a new point I have and will continue to address your concerns but there is no burden on me or other editors to constantly repeat the same points over and over. Finally, your post here is misleading because we've been through this before and you've admitted that I did not discourage discussion. [3] The problem you have is with one editor in particular. But up there you make it seem as though there was some sort of cabal at that page of which I would be a member. And if you are saying that I have discouraged discussion using the argument that "consensus has been reached" that would be lie, wouldn't it!? There is nothing hostile about my attitude (trying to discern an attitude from a post is an often useless endeavor since in effect, you are simply reading text on a computer screen. It's better to just simply ask the editor instead of assuming hostility--outside a blatant PA of course). I just figured that both sides should be presented. Ramsquire (throw me a line) 19:59, 4 June 2007 (UTC)

Talk pages

I am now in a controversy, with an editor who insists that article talk pages do not operate by consensus. (The point at issue is a project tag he insists on adding to a talk page which only he thinks in relevant to the project.)

Has anyone met this aberration before? If so, should we clarify the page to say explicitly that consensus governs all pages at Wikipedia? If not, will someone click on the link above and tell him so? Septentrionalis PMAnderson 16:29, 12 May 2007 (UTC)

I guess a talk page is somewhat of a different animal. Individual comments are not governed by consensus. They are a means to reach consensus. But what about the common items such as the tags at the top of the talk page? It seems that those should be consensus driven, though most of them are far from controversial. What project tag and page are of concern? maybe the specifics make a difference here. --Blue Tie 19:56, 12 May 2007 (UTC)
I think the account at User:TonyTheTiger/DR bot is intelligible. It links, eventually, to two other discussions and the talk page concerned (Talk:Jon Corzine. Septentrionalis PMAnderson 22:05, 12 May 2007 (UTC)

Ode to Consensus

A yakkity-yak,
A hackity-hack,
A talk page archive or two.
Put 'em together and what have you got?
Something that's scary and blue.

Splash - tk 22:21, 18 May 2007 (UTC)

In serenity
Everyone stops arguing
That is consensus.


Oh, some see through smudgy lenses,
And some with extra senses
And many don't even know their proboscis from their tail.
But I’ll go by consensus
For that is what defends us
From the trolling and the polling that otherwise would prevail.
...with apologies to Silly Wizard. -- Visviva 12:53, 21 May 2007 (UTC)

"Clear consensus"

At Wikipedia:Deletion review/Log/2007 May 23, you can read:

List of people by name – Overturned and deleted. AFD showed a clear consensus to delete which is apparent here too. – Srikeit 03:57, 28 May 2007 (UTC)

Following Srikeit's suggestion, I have posted the issue at Wikipedia:Administrators'_noticeboard/Incidents#List_of_people_by_name, so a discussion may develop there, but my basic question, after carefully rereading Wikipedia:Consensus, is how anyone, even an admin, can boldly claim that there is "clear concensus" in the face of dozens of objections by various users. In the case cited above, this has been going on for months, if not years (Wikipedia:Articles for deletion/List of people by name, Wikipedia:Articles for deletion/List of people by name (2nd nomination), Wikipedia:Miscellany for deletion/List of people by name (May 2007), Wikipedia:Deletion review/Log/2007 May 23, etc.).

User:Ceyockey has commented that "this page set is going the way of other perennially nominated pages and it will, eventually, be deleted simply as a matter of time. Radiant's bringing it here is pretty much a death knell for the page set as this is a highly respected contributor and admin. Therefore [...] there is simply no way it can be kept for the long term owing to continued attempts to delete by persons who vigorously oppose its existence. It is not original research, it is not useless, and it is not unmaintainable but it is unpopular - and that is the reason why it is ultimately doomed to deletion".

I'm willing to learn, so who can explain to me what is "clear" about the consensus above? <KF> 12:55, 28 May 2007 (UTC)

No one, it seems. <KF> 21:35, 5 June 2007 (UTC)

NPOV

I think this policy should state that it doesn't override WP:NPOV. What is the use of having a WP:NPOV if majority rules overrides it? There may be an instance where someone does not want to follow NPOV at all. Ideally, we should reach consensus with the goal of NPOV, but where in this policy does it state that requirement? Some editors could say they have arrived at a consensus in which they desire to only show one POV where other notable POVs exist, overriding the NPOV policy. This policy should state that it doesn't override the WP:NPOV.--09:27, 30 May 2007 (UTC)

While it's not made explicit, both the "consensus can not quickly override existing policies" and "consensus does not trump Foundation Issues" include NPOV, with the latter link mentioning it specifically. It would be foolhardy in my view to try to enumerate all the things that consensus can and cannot override on this page. -- nae'blis 15:33, 30 May 2007 (UTC)
Local consensus on one wikiproject cannot override Foundation issues by themselves. However, Consensus Can Change, and even the Foundation issues evolve ever so slowly over time, as conditions and our understanding change. --Kim Bruning 18:06, 3 June 2007 (UTC)

How to Avoid Consensus

It seems to me that the idea of consensus has one huge loophole. If an editor is involved in a content dispute or just simply wants to be a pest, all he has to do is be a persistent three revert violator. Once reported to WP:3RR an admin may simply lock the page and request further discussion, instead of blocking the disruptive editor. This encourages holdout editors to avoid ever reaching consensus, in the hopes that they could frustate other editors off of the disputed article. Perhaps, we can ask admins to dig a little deeper, before locking an article brought to their attention from WP:3RR. Ramsquire (throw me a line) 17:12, 3 June 2007 (UTC)

Admins should block the 3RR perpetrator, not punish other editors by protecting the page. --Kim Bruning 18:07, 3 June 2007 (UTC)
Unfortunately though, I've seen more than one admin do precisely that. Of course, you don't want 3RR to become a race where the first one there becomes the content dispute winner by getting the competition blocked, but you don't want the situation I described above either. Ramsquire (throw me a line) 18:22, 3 June 2007 (UTC)
Hmm, typically I refer people to User:Dmcdevit/Thoughts. Is that useful? --Kim Bruning 19:04, 3 June 2007 (UTC)
  • Well, there are a number of people that argue that if they never agree to any compromise or suggestion, then the "present version" must stay since there is obviously no consensus for moving away from that version. Since this attitude can be disruptive, such people have a tendency of ending up blocked. In general this approach by definition causes a fuss which attracts attention, and attention tends to show that the approach is invalid. >Radiant< 15:51, 4 June 2007 (UTC)

Moved from Wikipedia talk:Consensus can change

this needs a tag - is it a policy, guideline, edict or something else? Thryduulf 20:39, 31 August 2005 (UTC)

I think a great many of the old grandees of Wikipedia (and no, I'm not including myself in that category - I'm far too new around here - though I would, also) would say that it has, indeed, been policy right from the start. "Edict" sounds like it comes from on-high, which is the wrong idea; it's more of a backbone of the social contract of Wikipedia, I would say.
James F. (talk) 22:04, 31 August 2005 (UTC)
Disagreed. The picture I'm getting is that Wikipedia is evolving into an environment where the community votes on everything procedural.  Denelson83  23:19, 31 August 2005 (UTC)
I would use the term "mutating" rather than "evolving". It is not a welcome change, and it is not a change for the better, and it is a change.
James F. (talk) 00:22, 1 September 2005 (UTC)

This can't be a policy since it's simply not accurate. If binding polls were not permitted, neither RFA nor any of the deletion pages would exist. If a complete absence of voting is desirable (on which the community is far from unanimous), we've got a lot of reform to do. — Dan | Talk 23:26, 31 August 2005 (UTC)

They're not binding polls. The closing admin (or bureaucrat) evaluates using their intelligence to gauge consensus. The distinction may appear slight, but it's important.
James F. (talk) 00:22, 1 September 2005 (UTC)
Fair enough; however, this distinction should be made very clear, and I've attempted to make it so. — Dan | Talk 01:02, 1 September 2005 (UTC)
Just wanted to say thanks for putting the clarification in, even though I've already thanked you in IRC. :-)
James F. (talk) 01:17, 1 September 2005 (UTC)

Voting is not banned

Voting is not banned. What do you think happened during the Elections? How do you think arbitration cases are determined? The policy has always been Don't vote on everything, not don't vote at all. Angela. 03:06, September 1, 2005 (UTC)

I've moved the page from Wikipedia:Voting is banned to Wikipedia:Don't vote on everything to reflect this. Angela.
Elections are not part of the Wikipedia process. Neither are Arbitration cases. Feel free to bring up an actual example of something to do with Wikipedia, as opposed to Wikimedia in general. Otherwise, this should be moved back.
James F. (talk) 09:27, 1 September 2005 (UTC)
Of course arbitration cases are part of the Wikipedia process. After all, their outcome is actively enforced.--Eloquence* 12:15, September 1, 2005 (UTC)
How Arbitration cases are handled internally is irrelevent to the community's use of our decisions, however. The Committee's internal workings are not a community process, and we're talking about that here.
James F. (talk) 14:36, 1 September 2005 (UTC)

Consensus on voting :-)

It appears to me that some of those who are so heavily advocating consensus are at the same time trying to steamroll over those who feel that voting[*] is legitimate under certain circumstances. I hope that I'm mistaken. Titles such as "Voting is banned" are certainly not likely to engender productive discourse, so I'd like to thank Angela for moving the page to a more appropriate title. I have attempted to rewrite the page to reflect the actual policies and practices on the English Wikipedia and the beliefs of its inhabitants.

[*] Before you bring up a vote/poll distinction: Please look at the dictionary definitions for both words. [4] [5] Neither word automatically implies that the result is "binding", or that a particular process is to be used.--Eloquence* 04:10, September 1, 2005 (UTC)

Before you quote a dictionary at us, remember that we have dozens of terms of art and distinctions of this kind on Wikipedia. "Poll" and "vote" do not mean the same thing on Wikipedia (much as "ban" and "block" have very different meaning).
James F. (talk) 12:15, 1 September 2005 (UTC)
These definitions, to my knowledge, have never been laid out, and we cannot expect any reader to understand an idiosyncratic usage of the terms "vote" and "polls". So let's be clear about what exactly is allowable and what not.--Eloquence* 12:36, September 1, 2005 (UTC)


Community knowledge is very rarely laid out; that's why we're only creating this page now, 4 years after it has been use. But you want definitions; very well, here you go:
Poll
n., system whereby people simplistically but clearly list opinions on each of the sides of a debate in a helpful manner to ease understanding of community consensus.
Vote
n., aka "binding poll", poll (q.v.) system wherein instead of community consensus being allowed to be evaluated, a rigid system which violates the Wikipedia principles of flexibility and open-ness is used where the community is forced to be bound.
James F. (talk) 14:23, 1 September 2005 (UTC)

Voting is not banned, but binding voting might be

Following discussion with James F. about this, it seems the main issue is the binding aspect of a poll or vote rather than the occurrence of such a method itself. Therefore, I've moved the page to Wikipedia:No binding decisions since it isn't only polls/votes that shouldn't be binding, but the result of any decision making process. I feel polls can be a valid decision making process if used correctly, and if they are not binding, they can not be accused of undermining consensus. Angela. 13:17, September 1, 2005 (UTC)

I agree that straw polling can indeed be helpful. I'm happy with the current title, and the current wording too.
Others?
James F. (talk) 14:28, 1 September 2005 (UTC)
Sure; looks OK from here. Mindspillage (spill yours?) 15:36, 1 September 2005 (UTC)
As the one who was initially responsible for the original title (I suggested Kim write it on IRC), I am perfectly happy with this. May I suggest that meta is a better place for this? [[smoddy]] 17:27, 1 September 2005 (UTC)
The problem with putting things on meta is that people immediately say that they don't apply to the English Wikipedia.
*sighs*
James F. (talk) 10:17, 2 September 2005 (UTC)
Which proves our main problem: people are stupid. [[smoddy]] 16:07, 2 September 2005 (UTC)

No binding decisions eh? Does that include the decision for the policy "No binding decisions"? :) --WikiSlasher 09:06, 5 December 2006 (UTC)

Precedent should matter

While I understand what this is trying to say, I think it almost comes across as an endorsement of arbitrary decision-making. I think it's especially important in AfD debates to look at the outcome of recent debates and the presence of similar, long-established articles. Consensus may change, but change will probably take some time, and when it happens, it should be reflected consistently in the treatment of different articles. I've seen the NBD policy used to justify multiple AfD nominations in a short period of time, which I think reflects a misunderstanding of what it's trying to say. --Cheapestcostavoider 20:03, 18 May 2006 (UTC)

If there's no consensus to delete, how do you think re-polling will change that consensus?
Right! It won't.
A poll measures current consensus, it doesn't nescesarily change it. No matter how often you measure, the outcome will always be similar! :-)
(it won't be exactly the same, because consensus is constantly slowly shifting, and a poll only samples that consensus at at a single point in time)
Kim Bruning 11:45, 18 August 2006 (UTC)

Contradiction

Isn't this kind of a contradiction? Wikipedia has no binding decisions. Except the decision that we have no binding decisions. That's binding. :-) --W.marsh 00:40, 18 August 2006 (UTC)

It's a fundamental property of the wiki concept. Any software that implements a wiki will tend to enforce it by default. There is no contradiction with policy and guidelines, as those are made up by users of the sofware, and looked after by hand.
This page happens to also be an example of the kind of knowledge was taken for granted in the first few years of wikipedia. I've been pushing old users to actually write down that knowledge; before they leave, and their knowlege and skills are lost to the communuity.
Kim Bruning 09:15, 18 August 2006 (UTC)

Verification Needed for Assertion of Policy Status

JA: I have requested verification of the claim asserted by the 02 Sept 2005 insertion of the {policy} template that "This page is an official policy on the English Wikipedia. It has wide acceptance among editors". Until somebody provides adequate evidence to support this claim, the template is invalid and unverified and can be removed with impunity at any time by any user. Jon Awbrey 12:12, 18 August 2006 (UTC)

That's not how things work. I've removed the tag, please don't put it back. (note however, that it might be interesting to find verifiability rules for guidelines... that might be a great way to cut down on the mess). Note that this particular rule is a logical consequence of how a wiki works, so there's not much we can do about it. That and it's the only way to ensure that wikipedia will be around for ~100 years. You need to be able to get out if you accidentally paint yourself into a corner. Kim Bruning 15:37, 18 August 2006 (UTC)
Hoping that Wikipedia will be around for an entire century is a bit ambitious. Copies of it will probably still exist, but let's hope that by 2106 there'll be something even better than this. And then I mean, something better than the world wide web. --Thunderhead 16:01, 18 August 2006 (UTC)
You can hope, but in the mean time, this is what we have. It's doable and definately worth going for, even though I know it runs a bit counter to the current-day western psyche. It's still much less ambitious than long now ;-P Kim Bruning 16:57, 18 August 2006 (UTC)

JA: IC, so U do recognize the distinction between a policy and a principle after all. Thanks, I needed that. Jon Awbrey 20:14, 18 August 2006 (UTC)

BRD shortcut

Jon Awbrey, it's great that you're re-inventing bold revert discuss, but I've already done the reverting. Could we get to the next step? Kim Bruning 17:09, 18 August 2006 (UTC)

Binding decisions

We already have a method of making binding decisions, in one small case: the deletion of redirects. WP:CSD G4 permits speedy deletion of recreated pages which are substantially identical to the deleted page; and in general that's a good thing. If someone wants to reopen the discussion on whether to have an article, he need merely write a new article on the topic, and it will be taken to AfD.

But if a redirect is ever deleted, that's permanent. There's only one way to make a redirect from A to B, so anybody, any time, who makes the same redirect will find it speedied. (This is already happening with cross-namespace redirects.) Septentrionalis 18:19, 18 August 2006 (UTC)

History and status of this project

This project page was declared to be a "WikiPedia Policy" by User:Jdforrester on 02 Sept 2005, when that user posted the {{policy|[[WP:NBD]]}} template on the project page, as evidenced by the following history link:

The claims asserted by means of this device, specifically:

  1. "This page is an official policy on the English Wikipedia", and
  2. "It has wide acceptance among editors and is considered a standard that all users should follow",

are in dispute and should be regarded as the personal opinions of the small number of Wikipedia users who actually support them.

JA: Under Extreme Protest, moving the above from main to talk.

JA: Wikipedia broke down for a couple of hours when I tried to post the above. I hope it wasn't something I said. Jon Awbrey 20:04, 18 August 2006 (UTC)

JA: 21 Aug 2006. I tagged this project page as Proposed, and Kim Bruning reverted it:

JA: I do understand the English sentence that asserts of this project page that "It has wide acceptance among editors and is considered a standard that all users should follow". There are three problems with this statement:

  1. There is no evidence given to support it.
  2. There is no reason given to believe that it's true.
  3. There is no source given to say by whom it "is considered a standard that all users should follow".

JA: That sort of claim is inveighed against in WP articles — I cannot imagine why it should be permitted in WP policies. Jon Awbrey 17:24, 21 August 2006 (UTC)

Wikipedia guidelines are odd, relative to the rest of the encyclopedia, since we need to do our own research for them. There's no choice in this. There is no scientific document or engineering procedure that can explain how this particular system can be run at this scale, so we need to design our own, and that's what we've been doing in the past 5 years.
This particular guideline is somewhat stranger, since it's basically a primary source even! A group of old experienced wikipedians was asked to write down one of the design principles of the wiki.
It actually states one of the design principles behind the wiki-engine, especially the way it allows editing (and reverting). So you can say that the actual php code has been written based on this rule, and it sets the framework within which we work.
No matter how you mark this page, or how you edit it, the php code will still enforce the framework. (A rose by any other name ... ) .
Now if you change the php code, then things would change. However, the end result would no longer be called a wiki.
Kim Bruning 17:59, 21 August 2006 (UTC)

JA: 21 Aug 2006. According to my scan of the edit history, subject to correction, there have been a total of 15 editors on the main page, a total of 14 editors on the talk page, and a total of 21 editors all together on this project since its inception, namely:

JA: That does not warrant claims of "wide acceptance among editors", even if all 21 editors agreed about every issue among themselves, which I know for a fact that they do not. Jon Awbrey 18:12, 21 August 2006 (UTC)

With that, you are (partially) answering a point I made a couple of days ago, not the point I made today. Kim Bruning 18:22, 21 August 2006 (UTC)
Right. Would you care to answer the point I'm making now? In a sense you're forgetting to mention Ward Cunningham, I think ;-) Any idea why I think you're forgetting him? Kim Bruning 20:20, 21 August 2006 (UTC)

JA: I just munged the edit histories into a table and sorted — if I missed an edit line please supply. Thanks, Jon Awbrey 20:26, 21 August 2006 (UTC)

It's nice that you're doing this work, but you're not addressing my current point yet. If you'd like to continue on the previous point though, perhaps you'd like to talk with each of these people? Especially James Forrester might be handy. Kim Bruning 20:31, 21 August 2006 (UTC)

JA: As to the point that you are trying to make, I quite frankly hesitate to articulate it, and apparently so do you. So I can but wonder why is that? Jon Awbrey 20:36, 21 August 2006 (UTC)

JA: what you may be missing is that Wikipedia, and wiki communities in general, are not empirical. Wide acceptance may be asserted of anything which is widely practiced, without need to justify this with a poll of the community. It's rather like saying lawns of grass are widely accepted; it's not universal, it may be virulently opposed by a segment of the population, yet any reasonable survey of lawns will find the vast majority have grass as a consituent element. - Amgine 20:32, 21 August 2006 (UTC)

My point was [6]. Kim Bruning 21:07, 21 August 2006 (UTC)

Point, Counterpoint — and Fugue

JA: I was afraid of that. And my point is that no amount of techno-mumbo-jumbo is an excuse for making false or misleading statements on a content, policy, guideline, or project page. Software is made to serve society. Society is not made to serve software. Let's call it "Simple Rule Number 1". Jon Awbrey 21:24, 21 August 2006 (UTC)

Darn, we can't work with those definitions. Please see Social software as a start. But ...there's more to it than that. Kim Bruning 22:16, 21 August 2006 (UTC)

JA: Please try to understand, I did not come here to join some kind of Borg Religious Order (BRO), or fade and dissolve my storm-tossed exhile of selfhood in a mystical etherium of wiki-karmic-goo. I came here to write high quality encyclopedia articles on subjects that I have invested a lifetime getting more or less adequate in. There are all sorts of wikis in space(.com) that have no other purpose but to preserve the cultivation of their lotuses (loti?) by providing their "users" with a Feel God experience, but that's not what WikiPedia claims to be. If the software does not serve the espoused objective well enough, then those of us who really share that objective will find some other tool to that purpose. The software is not a replacement for society. The software is a tool of human purposes that antedate it by many, many evolutionary cycles. Jon Awbrey 02:16, 22 August 2006 (UTC)

Good. That means we're both pragmatic. Now let's discuss how an amalgamation of software and social systems can allow you to write your high quality encyclopedia articles, and also how -without adequate constraint- they can conspire to prevent you from doing so. (see also: Cybernetics, Systems theory, Game theory, Nomic) Kim Bruning 09:09, 22 August 2006 (UTC)

JA: Our Pragma are as Manifold as our Sinns. There are many forms of pragmatism, pragmaticism, mea maxima pragmata. The most pressing problem at the moment is that there is a false or misleading statement at the top of the project page that you have resisted correcting or removing for reasons that you have yet to articulate fully, much less render in the least bit convincing with regard to ordinary standards of acceptability. And that is normally considered a bad thing. When we have dealt with that tiny dust bunny, then we can can think about the next task on the list. Jon Awbrey 13:06, 22 August 2006 (UTC)

I will not be drawn into a philosophical discussion. I have neither the time nor the inclination at the moment, sorry.
This page is policy. I have stated why, you reject my premise. Fine. But since the wiki design leans on that same premise (among several others), you automatically reject the existence of wikis. Oops.
But that's your problem. Resolve it for yourself. Come back when your philosophy and reality are back in the same ballpark. I'm always willing to talk, as you know :-)
Kim Bruning 15:06, 22 August 2006 (UTC)
Perhaps the problem is that we're not quite speaking the same language, even though we both *appear* to be using english. Consider yourself a stranger in a strange land. It would be polite and handy to learn (parts of) our language, especially if your own language lacks words for snow.  ;-) At the same time, I might take some trouble to learn yours? In the mean time, we'll get into big conflicts if we keep up like this. Could we call an editing truce? :-) Kim Bruning 15:19, 22 August 2006 (UTC)

Hello Jon. If I am reading you right, you have three issues:

  1. The coherence of the rule itself: objection based on self-reference
  2. The status of the rule: doubt that this is actually "policy"
  3. The meta-text of the rule: the specific claims of the box with the green ticky mark in it, even if it is "policy".

I hope I have not omitted anything, or mischaracterized you on anything I have included. My responses are:

  1. No comment at this time. Let us get the meta-argument resolved first, and see if there is still an argument left over.
  2. The policy tag adds it to a particular category. I infer either support, or at least lack of objection, from everyone who would have noticed such a thing over the past 11 and a half months. This should not be underestimated.
  3. I think it is a bit misleading to suggest "wide support" among editors, when the overwhelming majority probably rarely visit meta-space at all. But it occurs on many policy pages, and has its roots in the need for people to put ticky boxes on things, which is an issue much broader than this particular page. It may be best discussed for the template in general.

Cheers. -anon 16:04, 22 August 2006 (UTC)

Hypocrisy & Misrepresentation Breed Cynicism & Mistrust

JA: 'Nuff said. Jon Awbrey 16:52, 22 August 2006 (UTC)

Yes they do. :-) Kim Bruning 21:40, 22 August 2006 (UTC)

JA: I'm beginning to see the sand this castle is made on — and I see that you people need some serious help rationalizing your policies. I'm not even sure you want them to be rational. Doesn't matter, the rest of the world — and Oh Yes, Virgilia, the Sandy Claws of Reality will Catch your Act in the Vth — will demand rational justification for what you say and do. I will think on it a while, but right now I desperately need to go work on some genuine articles before I go bats. In the meantime, I would like KB to quit confusing my WikiPedia birthdate with my RealWorld birthdate — I wasn't born yesterday, or late last December. At any rate, back to the grinstone for a now. Jon Awbrey 12:38, 23 August 2006 (UTC)

The empirical justification is that it has worked, and outperformed all other known rulesets for the past 5 years, including at least one ruleset used by the same project under a different name, and one ruleset used by a different project with the same objectives. Kim Bruning 13:41, 23 August 2006 (UTC)

JA: Not Entirely Non Sequitur. I used to think I was the Artful Dodger, but I see now that I've got a lot to learn about the Art of the Non-Sequitur. The best I can do for now is to share the following story, based on true life events, if not exactly ripped from today's headlines.

JA: Every Spring we hang a clear red plastic globe — shaped and colored vaguely reminiscent of a large translucent strawberry, with yellow plastic flower-portals on a subtended green plastic calyx — on a cast-iron scrollwork post in our rose garden, and we fill it full of sugar water on a biweekly basis. Now, the instincts of hummingbirds are clearly plastic enough that they go right to it and sup the refined sugar nectar from where it wells up in the faux-flower ports. But the funniest thing, and it cracks us up all Summer long into Fall as we peek through the bay window that peeps out over the roses, is this — there's more sugar water in a single filling of that globe than the whole gang of hummers in the 'hood could possibly consume in a month, and yet their instincts are not so plastic that they'll ever leave off blustering and dogfighting and just plain enjoy the mix. Moral of the Story? Ay, there's the rub.

JA: Still, I wonder, what your plans are for refactoring that ruleset, or have you even noticed it yet, with all your idees fixed on that rube mechanical bit? Jon Awbrey 05:20, 24 August 2006 (UTC)

It's a continuum. In my career I've found that some some things that looked like social problems could more easily be fixed by software engineering, and some things that looked like software problems could more easily be fixed by "social engineering". I could tell you lots of fun stories about this :-)
Once I was working in a team with one programmer and one social engineer. One day I was sitting next to the programmer, and we'd determined that our problem would take 3 days to solve, if we didn't want to get in everyone elses way. The social engineer went around and talked with several people and asked them if it was ok if we stomped on their areas for a bit. The problem was solved in 3 hours. :-)
On the other side of the coin, there exists a famous program called bittorrent, which uses clever codified social rules to prevent defection (called leeching). By clever enforcement of these social rules on its users, bittorrent allows unprecedented data traffic between machines, and in fact the bittorrent protocol is currently a very significant percentage of all internet traffic. (Use of) Bittorrent has also had several social consequences, most famously leading to the creation and increased popularity of a political party in sweden (!)
So the first thing to understand is not to make arbitrary distinctions between social and software, since they are very much a continuum; in fact I'd consider software to be a subset of social behaviour. If you don't take the entire continuum into account, you will be blindsided. Badly. Mediawiki people have been accused of (deliberately) blindsiding people in the past, for precisely this reason Kim Bruning 13:05, 24 August 2006 (UTC)

JA: Thank you. This has been very liberating. Naturally, I don't believe you folks for a second. But still, even a bit of foma can be freeing. Jon Awbrey 21:34, 24 August 2006 (UTC)

This reminds me of a discussion at the start of a book ... (I think it was Thomas More's Utopia, but don't pin me on that). In one scene, a monk sets out to philosophically prove that the Americas couldn't possibly exist... to a member of Amerigo Vespuccis expedition, who had just returned from there ;-) Kim Bruning 01:04, 25 August 2006 (UTC)

JA: Dear B, I hope you don't think I was being sarcastic. I genuinely experienced a moment of satori and immediately went off and spent a blissflow day in several creative endeavors here and there about WikiParadiso. My sideswipe of suspended belief is merely the escape pause that reflects long experience, and warns against the perils of Icharus. Jon Awbrey 20:00, 25 August 2006 (UTC)

Aware, Bware, Cware

JA: Making this new section to continue a previous discussion in a slightly different light. Jon Awbrey 21:14, 25 August 2006 (UTC)

This needs a rename

...because the title is misleading, and this has been confusing new users. An important purpose of policy/guideline pages is to instruct new users, and because of cognitive laziness they don't always read past the first few lines. The point of this page is not that we make no binding decisions (indeed, that would imply a loner can ignore a consensual discussion entirely). The point of this page, rather, is that consensus can change, and that any consensual decision can later be overturned by consensus (but not by a lone editor who didn't like it). >Radiant< 16:58, 11 September 2006 (UTC)

Got a suggestion? Maybe Wikipedia:Consensus can change, which differentiates top-down decisions and the five pillars? Wikipedia:No stone tablets has appeal too... -- nae'blis 15:45, 12 September 2006 (UTC)
  • Good idea, and done. >Radiant< 20:29, 14 September 2006 (UTC)

Query

"Why on earth would you ask about a specific article content issue on Village Pump before on the talk page? that's what talk pages are for, to discuss articles" -- Derex (talk · contribs)

  • Because this page is about how "Consensus can change". If consensus has been demonstrated a few months ago on an article talk page, asking "is this consensus still valid" on that talk page is very unlikely to get you a meaningful response. Hence, if you wish to demonstrate that consensus has changed, you need to get feedback in a more public spot. >Radiant< 14:58, 6 November 2006 (UTC)

Precedent - in or out

I think it would be nice if this page mentioned how precedent should be used on wikipedia. I, for one, think that precedent can be involved in discussing changes - but changes should not simply be made based solely on precedent. Consensus from discussion is key, and precedent (like statistics) can be made to prove anything has support. Therefore, I think this page should mention something about precedent - whether or not there is consensus for using precedent or not. Did that make sense..? Fresheneesz 09:49, 24 December 2006 (UTC)

  • That didn't really make sense, and the assumption that precedent can be made to prove anything is an obvious fallacy. >Radiant< 10:17, 24 December 2006 (UTC)

Sticky consensus

A popular idea around here is that once a consensus is formed, it becomes "sticky", and you need another opposing consensus to undo it. Like you get a consensus of five like-minded people on a talk page, slap a policy tag on a page and force people to do stupid shit, and you then need to get a lot more than five people to overrule the original five who won't give up their position in order to remove the tag. I think this is quite silly, and not the way things are done. If there's no agreement, there's no agreement.

However, User:Sarenne made a good point about the fact that the people who like the status quo aren't going to be participating vocally on talk pages... What do you think? — Omegatron 00:34, 19 May 2007 (UTC)

  • Yes, that behavior is more common than I'd like, but that doesn't make it right. The sections on "consensus can change" and "local consensus vs. wikiwide consensus" should make that clear. >Radiant< 08:26, 5 June 2007 (UTC)

Precedent for retention after an AfD Keep

What is the status of an article that has had an AfD end as a Keep? Is there any limitation on additional attempts to delete the article and does there have to be any material change in fact to justify initiating a second AfD after a previous AfD established a precedent to Keep the article? The way I read WP:Consensus -- "This does not mean that Wikipedia ignores precedent; for example, editors should not continuously nominate an article to WP:AFD until it reaches their preferred outcome." -- there is an implied "protection" for an article that has a demonstrated consensus for retention. The way I interpret this statement, if AfD1 ends in a Keep, then a strong burden is placed on the prospective nominator of AfD2 (on the same article) to demonstrate that "... there is new information to discuss.", as the policy continues, before starting a new AfD. If a Keep precedent on an initial AfD provides no measure of protection to an article, there is nothing to prevent any other editor from taking a second (or third, fourth or fifth...) stab at deleting the article. And if AfD #37 finally achieves the desired result, there would seem to be nothing in WP:Consensus forbidding the article from being recreated as is, and the AfD cycle starting all over again. Does AfD precedent have any meaning in terms of the future of an article? Alansohn 16:12, 5 June 2007 (UTC)

  • Yes, it does. If you repeatedly recreate a deleted article, it gets locked out. If you nominate an article for deletion for the, say, 37th time, it gets speedily closed. Both cases can get you blocked for edit warring against consensus, and/or WP:POINT. >Radiant< 16:15, 5 June 2007 (UTC)
    • Let's ignore the extreme case. Does the precedent of a Keep in an AfD offer any "protection" from AfD II happening or does it place any burden on the prospective nominator of AfD II to show why the AfD I consensus should be overturned? Alansohn 16:20, 5 June 2007 (UTC)
      • The protection it offers is that the same good reasoning in AfD I applies to AfD II. The people who comment in AfD II and the closing admin in AfD II will look at the discussion in AfD I and the prevailing reasons there, if they were indeed prevailing reasons and if they are still relevant to the article, apply just as much to AfD II. It is not, though, binding precedent in the same sense as precedent is binding in the legal system. —Centrxtalk • 16:27, 5 June 2007 (UTC)
  • Posting AfD#2 immediately after a Keep-close AFD#1 usually results in a quick build up of speedy Keeps before a speedily closed. Since consensus changes, the passage of time is enough to offer up the same article for AfD. Sometimes you'll see an AfD#2 for an article that has not change much in the year since AfD#1 and the article winds up being delete. The prospective nominator of AfD II need only show why the article should be deleted, not why the AfD I consensus should be overturned. If the reason for the AfD I consensus remains valid, most people just repeat the valid Keep reason in AfD#2 rather than use AfD#1 as some sort of precedent. I've seen AfD arguments that articles listed on the main page under DYK have DYK precedent approval to remain on Wikipedia. That precedent argument usually received little to no weight by others. -- Jreferee (Talk) 17:05, 1 July 2007 (UTC)

Policy in a nutshell, potential wording

This may have already come up, but I think this policy should have a nutshell, such as:

Please comment on the proposed wording. -- Jreferee (Talk) 16:47, 1 July 2007 (UTC)

  • That disregards the fact that most content is written by people who feel like writing something, not people who first ask for consensus to support their wording. In other words it contradicts WP:BOLD. >Radiant< 15:46, 4 July 2007 (UTC)

I added the following back in February. It only lasted a few hours and I didn't pursue it further:

-- SamuelWantman 07:07, 5 July 2007 (UTC)
If you added "Consensus is not a vote" or something like that it could be useful. Sophia 07:53, 13 July 2007 (UTC)

"Asking the other parent" section

A good sign that you have not demonstrated a change in consensus, so much as a change in the people showing up, is if few or none of the people involved in the previous discussion show up for the new one.

This appears to be an argument from silence. Discussions move on - new research is linked - how can the previous participants in a discussion be held as some sort of unassailable status quo? Surely if they cared that much they would revisit to explain their take on things again? Sophia 21:22, 12 July 2007 (UTC)

Intro addition

I think this is a crucially important point about consensus that should be in the intro:

"When consensus is referred to in Wikipedia discussion, it always means 'within the framework of established policy and practice'; a common view by a limited group of editors almost never outweighs the consensus of the community as documented by policy."

This is a point that often comes up in dispute, particularly in connection with POV and ownership issues: "It's consensus the article should handle it this way, so your view is marginalized". Unless consensus within a debate operates within the framework of policy, which itself represents the much larger consensus of the community, it's flawed logic.

I've added a brief sentence to this effect. FT2 (Talk | email) 14:05, 16 July 2007 (UTC)

The numbers

Somehow the numbers got added again. I've removed them. --Tony Sidaway 16:11, 17 July 2007 (UTC)

Added section: Consensus on Wikipedia

I noticed that the policy did not in fact desxcribe (other than in the introduction and by allusion) how consensus works on Wikipedia. The first section was, Building consensus, and the second, Changing consensus, but nowhere was there a clear description of what consensus means here. It's not obvious, especially when comparing localized consensus (article debate) with communal consensus (policy).

I've made good on this, adding a section on consensus, with 3 brief examples, in line with other policy pages that explain the meaning for newcomers.

I have also clarified how consensus and established practice are followed even in regard to consensus change, be adding a short paragraph specifically on that point, in the "consensus can change" section that covers consensus change.

FT2 (Talk | email) 21:07, 21 July 2007 (UTC)

Ouwie... I removed it as being somewhat inaccurate, and THEN read the talk page. (I'm human too ^^;;) I'll reproduce the text here and let's see what we can do with it. --Kim Bruning 18:50, 22 July 2007 (UTC)



== Consensus on Wikipedia ==

Consensus on Wikipedia always means, within the framework of communal consensus, as documented by established policies and practice. Consensus never means "whatever a limited group of editors might agree upon", where this contradicts policy and practice. This means:

  • Editing decisions are not based upon pure "vote counts". A "vote" based upon ignoring Wikipedia policies does not usually count as more than a personal view or wish. (See WP:IAR for rare exceptions)
Example: 14 editors vote on an article deletion. 9 users vote for deletion arguing the article is "offensive" or "unnecessary", without realizing that policy deems these non-valid arguments. 3 users vote for deletion stating it is "not notable" (a valid policy-based reason). 2 users vote for keeping and submit evidence that it is notable and verifiable. Even though the "delete" votes are strongly worded and the majority, the consensus of policy related points is that the last 2 editors have shown evidence which backs the view that the subject is notable and verifiable, and other editors have not shown evidence to refute this. Policy tends to support their view that it should be kept.
  • Even strong opinions and strong support expressed in specific polls, almost never change the need to abide by communally-agreed policies, guidelines and practices. Consensus on a small scale is not expected to override consensus on a wider scale very quickly (such as content-related policies/guidelines).
Example: 4 editors who strongly agree on some viewpoint end up dominating discussion on the article's talk page. Even if they all agree, and are all sure they are right, and all sure other editors are wrong, they cannot override the requirement of policy to represent the opposing views neutrally and fairly, because the community has indicated a very high level of consensus that this is non-negotiable.
  • A proposal or edit (eg to a Wikipedia: page) that has the effect of changing policy or practice needs care to be sure it really does have communal approval, and a rationale others will support, or it will be quickly challenged.
Example: The proposal to merge certain key policies under the title Wikipedia:Attribution was a serious and wide-scale discussion but when tested, did not in fact have communal support. It transpired that a significant number of editors had concerns it would replace three well-defined focussed documents by one less direct more general document. It failed.

First paragraph: Actually voting is discouraged at all, if you have an opinion it might override all others, if it is particularly convincing. Slightly different phrasing might be useful here?

Second paragraph lends to much weight to wikipedia-is-a-bureaucracy, which is something we're trying to avoid.

Third paragraph is about proposing things. This is unwise, the correct action is always to be WP:BOLD. See also the provided flowchart. It does help to be bold but stay inside the existing consensus framework, of course... Perhaps some rewording there will make this paragraph work too?

--Kim Bruning 18:57, 22 July 2007 (UTC)

Hiya, and thanks :)
I did think quite carefully about these. They are documenting how we use consensus on Wikipedia, and how it informs our editing decisions. Some rationale for this:
Clearly we do need something that says "this is what consensus means, on Wikipedia". So I've tried to cover what we mean when we use the term. That's what the section is about. WP:CONSENSUS didn't actually have a section on this. It was a bit like having an NPOV policy that explained how to obtain NPOV but didn't actually really ever say clearly what NPOV was, or what we mean by the term, just alluded to it. It's a gap in the policy.
  • Expressing an opinion in a poll is widespread on wikipedia, even if polls are evil. Even if polls aren;t used, people will say "I agree" or "I don;t agree" in prose. So we start from that point, and explain that even though it looks like a vote, it isn't; consensus is never a matter of vote counting but of opinion gathering and consensus. However the starting point is, you have to say somewhere, consensus isn't about votes, even though we often use polls to gather opinions. If this is part of how consensus works on Wikipedia, let's explain it.
  • I've reviewed this and not sure how this "adds weight to bureaucracy". We all share a common view that small, localized consensus doesn't readily outweigh large communal consensus. The wording for this is long standing and in the policy already, by communal agreement to date. Explaining why localized consensus doesn't override wider consensus and hence why discussion and agreement by a few editors can't override core policies, seems useful. It's a regular item in disputes, so a clear statement gives editors something to refer to: "even though you all agree, you still can't ignore other notable viewpoints, WP:CONSENSUS says so". Giving a tangible common example, surely isn't a problem.
  • The point and your comment don't conflict. Editing policy isn't deprecated. People edit policy pages every day. Proposing new policy that way is, but I'm thinking of the normal case where someone edits to say "X is allowed" or "Y isn't okay". The way consensus works is, if you make an edit that will have the effect of changing policy, think about whether you have consensus - will others support it? If not it will probably be quickly reverted. I don't see a problem advising users that's how it goes. There is no contradiction since one should be bold... and consider others' likely views when being bold. (Classic Wikipedia "good practice" - I prepare to boldly edit the article, but I also ask myself is this controversial, does it contradict consensus, should I broach it on the talk page first?)
Reflections? FT2 (Talk | email) 20:18, 22 July 2007 (UTC)


Update: This needs discussing. The additions appear to be a pretty good description of consensus, and useful material, if you disagree we need to discuss it further. Could you comment with insight somewhere in the next few days, if you still disagree? :) Thanks. (Message also on talkk page) FT2 (Talk | email) 23:57, 26 July 2007 (UTC)
Still to be come back to. Or shall I go for it for further consideration, if you are happier with the explanations (having rethunk it). FT2 (Talk | email) 19:12, 3 August 2007 (UTC)
Please clarify whether there are any remaining objections? If none remain, I'd be inclined to re-add and see if others have any. FT2 (Talk | email) 08:06, 23 August 2007 (UTC)

Refactor of ccc explanation

I hope folks like the reworking of "Consensus can change". The new version keeps almost all the original wordage, but trims essay-like stuff and a few other things that aren't so helpful in gaining an understanding of consensus on Wikipedia:

  1. Moved example of WP:ATTRIBUTION and description of ccc as it pertains to policy to a new subsection "consensus and policy change", so this section can focus more directly on consensus change in general.
  2. Described consensus change more, before going into who can and can't declare it is changed.
  3. Trimming a few excess words ("however" etc) and phrases ("when the article gains wider attention, members of the larger community of interest may then disagree" -> "when the article gains wider attention, others may then disagree").
  4. Focus on policy: for example changing the emphasis of consensus seeking, so that it is not so much about seeing whether other editors "agree" (OLD) but more about seeing "what points other editors think are important, and to compare and examine the different viewpoints and reasons." (NEW) The latter is much more in line with wiki-ish collaborative approaches, as well as supporting policy such as AFD where simply stating agreement is far less important than actually raising and considering discussion points.
  5. Improving flow/reorganize sentence order: sentences regrouped into 4 paragraphs, covering 1/consensus is not immutable and can change, 2/ small groups can make decisions but these are not fixed or OWNed, and the wider community can disagree, or they can be questioned, 3/ consensus change distinguishes between good faith concerns for good reason, and disruptive attempts to force a view, 4/ original paragraph on philosophy of consensus change, "Wikipedia is always changing as new editors and information arises", with very slight reword to reduce risk of being taken to imply "Wikipedia is unreliable because it's always changing".

Diff: [7]

FT2 (Talk | email) 23:14, 21 July 2007 (UTC)

Consensus vs. Policy

I have been involved in a few discussions regarding a page move and the overall consensus to move the page has been overwhelmingly against. However, according to Wikipedia's official guidelines regarding that topic, the page should be moved. I am curous to know if consensus is superior to official policy. Reginmund 21:08, 25 July 2007 (UTC)

What policy is this? Wikipedia:Naming conventions (films) is not a policy, which is what the title of the film is based on. According to Wikipedia:Naming conventions (films)#Rationale, you use the most common form, which is the one that is used more commonly in primarily english speaking countries. This does not mean you count the residence of each country and pit population size against population size, but that you say "Title X is used in 5 countries", "Title Y is used in 1 country"...."Title X is the more commonly used title".  BIGNOLE  (Contact me) 21:31, 25 July 2007 (UTC)

Ave Maria! The most common form is by quantity of people! Not by quantitly of countries! Now where on the policy does it say that the amount of countries determines over the amount of people? Reginmund 22:31, 25 July 2007 (UTC)

It says, Names of articles should be the most commonly used title for the following reasons - that "used" means you count the number of places it is used in, not the number of peole that "know" it by that name. The people are not "using" the name, the country is using the name when it allows the release of the film.  BIGNOLE  (Contact me) 22:50, 25 July 2007 (UTC)
It says, Names of articles should be the most commonly used title for the following reasons - that "used" means you count the number of people that use it, not the number of countries that it is used in. The countries are not "using" the name, the people are using the name when they refer to it.

It doesn't even say on the guideline page that we should count the number of countries and not places. Stop making up rules. I'm am not going to repeat myself for the fourth time why keeping it is violating at least three rules. Reginmund 00:17, 26 July 2007 (UTC)

You said it violates policy. I'm curious as to the policy it is violating. It clearly states on the page, the title should be that which is used the most in the primarily english speaking countries. If you want to know "how" that is supposed to be measured then maybe you should be going over to that page and asking what it means.  BIGNOLE  (Contact me) 00:20, 26 July 2007 (UTC)

Use the title more commonly recognized by English readers - It doesn't say anything about countries, it says about speakers... 40 million more of them. That's one rule broken. Stop making up your own rules again.

Sometimes different English-speaking countries use different titles, in which case use the native title instead - The native title is of the country in which the film was made. (i.e. United States [WB produced it])

Warner Bros. bought the film rights to the book in 1999. (Taken from the film's page) Well it looks like you broke your own rule since you said that it matters who owns the films rights. Reginmund 21:17, 26 July 2007 (UTC)

(1) You keep saying "it doesn't say anything about countries"...newsflash, it doesn't say anything about counting the population either. You are just as guilty of assuming as anyone else.
(2) Unfortunately, the film was "made" in Britain. Both British AND American countries produced the film. So, again, you are doing some assuming.
(3) Obviously you need some reading glasses, as I have stated (i think this is either the third or fourth time now...collected on multiple pages), that I see the ownwer of the film as the one who owns the film rights. Now lets get something clear, this is where you need to read carefully because you have obviously missed it so many times before when I said it, I have in no way stated that what I believe to be the true identifier of the owner of film is what Wikipedia believes. Please stop assuming, you just make yourself look silly. Also, why are you citing a Wikipedia article as a source for anything?
On an end note, go find out the clarification of whether one should count heads, or count the number of countries that use the title. Then, only then, will you maybe persuade the massive group of editors at Harry Potter and the Philosopher's Stone to change the title.  BIGNOLE  (Contact me) 21:24, 26 July 2007 (UTC)

(1) Newsflash? Fox could do better than you. Since when do we count by countries. We count by people unless this is the U.S. Presidential Election. It says "go by the better known name". 40 million people know it better... that makes it better known. (2) The primary production company as cited by two different sources has shown to be Warner Bros. and the country of production as the U.S.. (3) Stop making up banter that never happened. I never said "you believe the owner of the film is what Wikipedia believes". (4) As for the film rights... It turns out that you were wrong (again) that I was citing from Wikipedia as Wikipedia has cited itself[8] (see under distributor notes) NOW TRY CALLING IT A BRITISH FILM... you are looking dafter by the glyph. Reginmund 00:48, 27 July 2007 (UTC)

Again, where does it say "you count the population"? You're the one making up the rules as you go along. It doesn't say that anywhere. As you are so fond of saying to everyone else, "do your research", and go request clarification, because right now all you are doing is using your own symantics to define what it means. Warner Brothers is not the only production company of that film. It has a British production company, it's made from a British book, the cast is British, it's filmed in Britain....it seems to have more ties to Britain than it does to America. But you can believe what you want, and I'll believe what I want. I find your emotional state rather funny. You make personal attacks against me, while shouting (typing in all caps). You may need to take a Wikibreak my friend, and cool down just a bit.  BIGNOLE  (Contact me) 01:09, 27 July 2007 (UTC)

It doesn't say anywhere but I assume that is how one counts (usually when they attend mathematics class). It says how it is more commonly known. More people know it as "Sorcerer's Stone". It's that simple. It isn't a political issue as you make it out to be. Since when is better known the devil's advocate of better known (by population)... I don't think so. Something like that doesn't require research, just a few mathematics classes. Out of two American companies and two British companies, Warner Bros. is the primary company that produced that film. I have already sourced you on that. They bought the rights, they assembled the cast, they provided the funding, they distributed the film, that makes it theirs, and that makes it as American as NASCAR. It "seems" to have more ties is speculation... therefore... wrong. Just because the film is based on a British novel doesn't make it British. It didn't make The Picture of Dorian Gray British. I find your emotional state rather like one of a hypocritical basket case... touché, but unless you want to compare opinions of each other and not of the film, save it for your message boards. Note that you have also lied. I never made personal attacks against you and this was the only time that I used all capital letters just to EMPHASISE something that I HAVE to KEEP rePEATing since you "seem" to have short-term memory loss or are just too just too arrogant to admit that you are wrong after being proved it several times. Reginmund 06:49, 27 July 2007 (UTC)

"More people know it as 'Sorcerer's Stone'": Source? Matthew 14:24, 30 July 2007 (UTC)
I agree in general with this post. This policy should be defined so it does not conflict with other policies. Many times discussions reach a consensus but use arguments that are not based off of policy. If discussions have a consensus using arguments that are not based off of Wikipedia policies, then WP:CON trumps all other policies, making them useless. There should be a clause in this article that says "The consensus policy does not override other Wikipedia policies. Consensus must be reached using arguments based on Wikipedia policy or else the consensus is invalid."----DarkTea 19:06, 3 August 2007 (UTC)

That is exactly correct. Wikipedia consensus trumps all other policies (except all other rules) making them essentially useless.

Alright alright, before I get accused of being "too zen" again, let me expand a bit:

Actually it's more accurate to say that, "wikipedia policy" is merely a writeup of whatever the consensus is, or at least what it happened to be at the last update. Upside is that policy is organic and fits to what goes on on the ground at all times, downside is that written policy documents are always somewhat out of date!

Another extreme and somewhat odd consequence of this approach is that in certain cases it is actually possible to be reprimanded, blocked or even site-banned for following policy-as-written.[1] Current consensus is that you should always be warned upfront though, so don't be afraid to normally do what you think is fair, or even WP:Ignore All Rules, at least up until you get told otherwise. This gives you a lot of freedom.

This might seem rather odd to people used to a rules based system, as opposed to a consensus based system. Even so, the system has worked on wikipedia this way for half a decade already, and has proved itself to be most effective. It just takes a little getting used to.

Wikipedia is not the only consensus system around. Certain nations apply consensus processes to parts of their rule making, and there are also a small number of (other) organisations that use consensus. Some of those organisations are centuries old and were even involved in the founding of the USA.

Anyway, to summarize, Consensus vs Policy is a false opposition. Wikipedia Policy is basically bits of consensus that folks have written down, somewhat akin to FAQ documents elsehwere

--Kim Bruning 01:27, 26 August 2007 (UTC) [1]The Arbitration Committee is adapted to this environment as well... all Arbcom decisions are supposed to ultimately be "for the good of wikipedia".


Consensus and policy interact at more than one level. It's not as simple as "A trumps B". In roughly descending order of trumpness (!):

  1. At the top level, the topmost basis for all policy is consensus. No policy exists which is not merely a reflection of communal consensus. Even the ability for certain processes to sidestep policy and consensus (eg dictat of Jimbo or office) has previous agreement by consensus, and any processes purporting to "change policy" are in effect merely ways the community accepts as evidence that consensus has changed regarding the sanctioned practices.
  2. At the next level, policy is the codification of certain major themes and principles in consensus, written as clearly and usefully as the community knows how. part of the consensus of the community is that we all agree with each other to be mutually bound by the spirit of these, in order that we're all "working to the same tune".
  3. Next, policy is put to use, as the summarized voice of consensus. At this level, policy "speaks for" communal consensus and carries great authority upon editors. This is in part why WP:IAR exists, because it can never speak perfectly nor cover all contingencies. But to a great extent, policy does reflect communal consensus well, and in other cases we speak of editors being obligated to follow "the spirit of policy" by which we mean, the underlying consensus and intent that the wording was written to capture.
  4. At this point we're down to individual editors and groups of editors. At this level, policy has strong authority. We say that at this level, the agreement of a few editors cannot usually override policy or communal practice, meaning, that consensus by a few editors at this level cannot claim to be "the community changing global consensus" (a higher level consensus).

So we actually have use the term "consensus" in multiple ways (hence the difficulty in wording) to mean both the voice of the community, as well as the voice of a small group of editors. We use "policy" to mean the codification of the community's consensus (as best the community is able to word it), which to individual editors or groups of editors has great authority as it "speaks for" the communal consensus.

This can be summarized fairly briefly:

Policy speaks for communal consensus. At any given time, policies and guidelines represent community views written down as best the community is able. Corollaries are that 1/ Policies as written do not override the consensus intention of the community; and 2/ The agreement of a small group of editors does not (except in rare and exceptional scenarios) have the authority to set aside the communal consensus as outlined in communally written policies and guidelines.

FT2 (Talk | email) 04:00, 26 August 2007 (UTC)

What I think "Consensus vs. supermajority" should say

While the most important part of consensus-building is to thoroughly discuss and consider all issues, it is often difficult for all members in a discussion to come to a single conclusion. In activities such as Requests for Adminship, Articles for Deletion or Requested Moves, consensus-building can be unwieldy due to the fact that more people participate than can effectively cooperate (see: Dunbar's number). These processes may have been somewhat misdesigned, in that they have not scaled cleanly. As a result, it is sometimes hard to determine what the consensus is in such processes.
To compensate for this, people first simply check if the criterion of supermajority is achieved, and on that basis make a first order assumption on how close one is to rough consensus.
Formal decision making based on vote counting is not how wikipedia works (see Wikipedia is not a majoritarian democracy) and simple vote-counting should never be the key part of the interpretation of a debate. When polling is used, it should be seen as a process of 'testing' for consensus, rather than reaching consensus.

Anything else is superfluous. I strongly object to the placing of numeric margins into that section, or any other attempt to introduce an element of vote-counting into our official policy. --Tony Sidaway 09:39, 9 August 2007 (UTC)

  • I believe the numerical margins are less important than they used to be, and I also believe that this is a healthy tendency. We have way too many threads at e.g. WP:DRV that "this deletion had 65% support but by policy it must be 68% and therefore it was invalid". So yes, please get rid of the numbers. >Radiant< 11:52, 9 August 2007 (UTC)


Agree with Tony on numerical levels, and with Radiant that % level reference points are not usually to be encouraged, disagree that policy should be a commentary on other processes. Reword:
While the most important part of consensus-building is to carefully discuss and consider different views, it is often difficult for all members in a discussion to come to a single conclusion.
In processes such as Articles for Deletion, arguments that contradict policy or community views, are not considered valid; consensus is only used to determine the quality and scale of support for the views that are based upon policy, evidence and good practice. In some other processes, notably Requests for Adminship, an informal standard has developed that suggests a candidate with less than a given percentage approval has insufficient communal support.
Formal decision making based on vote counting is not how wikipedia works (see Wikipedia is not a majoritarian democracy) and simple vote-counting should never be the key part of the interpretation of a debate. When polling is used, it should be seen as a process of 'testing' for consensus and for editors' rationale, rather than reaching consensus. Supermajority can be a helpful (limited) indicator of confidence for a process aimed at identifying confidence levels or matters of opinion, but is less useful for deciding whether an article or other matter is factually evidenced and in line with policy.
Try something like that. FT2 (Talk | email) 13:39, 9 August 2007 (UTC)

Without numbers people have no way of knowing what a wikipedia consensus means in practice. I think that over the last year or so this page has degenerated. It is becoming a wish list of how people think a consensus ought to be arrived at not how in practice one is reached. Writing a wish page does not do any favours for anyone, and particularly new editors who are the ones most likely to take this page seriously. For example:

Consensus decisions in specific cases are not expected to override consensus on a wider scale very quickly - for instance, a local debate on a Wikiproject does not override the larger consensus behind a policy or guideline. The project cannot decide that for "their" articles, said policy does not apply.

This is just not true. If the people who edit a page and agree on something, then even if it goes against a policy it makes no difference because 3Rs and the person looses even if thy are on the side of the Angels. And saying well put in an RfC or whatever in most cases if two people turn up for an RfC one is doing well.

Further arguing that there is a community wide consensus on things like Wikipedia policies and guidelines is nothing like that. See for example the WP:ATT saga. At the moment about three or four people are deciding which template should be on the top of that page. Or see the ongoing dispute about where to put ref tags on WP:FOOT at the moment less than a dozen people are discussing the issue. See for example debates over things like Talk:Voßstraße#Page name as the old programming maxim goes "the wonderful things about standards is there are so many to choose from". The guidelines and standards grow ever bigger and ever more complex that there are more and more discrepancies in them.

Pages have to have names. In the case of page names like Voßstraße how can it possibly be decided what the consensus is without deciding on what a rough consensus is? How in practice does one decide such issues because if "simple vote-counting should never be the key part of the interpretation of a debate." then what does decide such an issue? People come to Wikipedia to contribute because they enjoy it. If in such cases as the name of Voßstraße or some other similar issues there is contradictory guideline advice, and an admin makes an arbitrary decision on the page name ignoring what most editors think, then they have the right to feel metaphorically mugged. It is much better, for the Wikipedia project, even if occasionally muppets make the wrong decision that the consensus is made by the editors interested in an issue. Usually most people will support the guidelines when they see that there is a good case for doing so but there are many many cases where this is not clear and I do not think it healthy for the project or for the admin that this page takes the stance that "aunty knows best". For processes such as WP, ADF and RFAs should have clear numerical guidelines because it brings transparency to the processes, and besides in practice if the decision is close the closing admin can usually cast the deciding vote, so everyone goes home happy. --Philip Baird Shearer 19:24, 9 August 2007 (UTC)

Correct me if I am ignorant, should the section simply aim to explain the meaning of consensus? Describing the various methods that communities achieve a consensus exceeds the scope of explaining what consensus means. In a nutshell I think it merely means that everyone abides by a specific outcome. Abiding, unfortunately, can exist for several reasons, including social agreement, disinterest, distraction, or by overt blocks/bans. My point is not to make consensus look dirty, but it seems that the page is trying to deliver far more knowledge than it must. Communities have favorite methods and develop their own tools, but we should not expect a policy statement to iterate every road to consensus. Of course in some cases, a Wikicommunity may add on additional standards, such as a percentage approval in addition to consensus, so that one could agree that supermajority is not consensus, but people can use them side-by-side. —Kanodin 22:55, 9 August 2007 (UTC)


I'm not sure entirely that I agree, though it's certainly material to think over. For example, "Without numbers people have no way of knowing what a wikipedia consensus means in practice" - most consensus decisions don't use a "number" - in practice, the decision process often followed seems to be "of those involved, the general sense seems to be X". This is then taken as (non-binding) evidence of the community's view, by sampling those who are involved in that debate. Whether Wikipedia editors genuinely use a consensus model in all cases is questionable; but that a consensus model is the communal view of how it should work, is probably not.
The comment you make ("This is just not true. If the people who edit a page and agree on something, then even if it goes against a policy it makes no difference because 3Rs and the person looses even if thy are on the side of the Angels. And saying well put in an RfC or whatever in most cases if two people turn up for an RfC one is doing well") is more to the effect that the consensus model is often poorly followed, not that it is not the agreed model. Consensus is still the model that in principle the community feels should be backed, even if in practice this is poorly followed in many cases. I won't disagree with that analysis. But the answer isn't to ditch the policy people back; it's to look at ways to strengthen its application in practice. One way to do this is to make clear what it means in this policy; other ways may be worth discussing too. FT2 (Talk | email) 23:19, 9 August 2007 (UTC)
Kanodin in the case of Cabinet Government in the Westminster model a consensus is always reached because there is a carrot and stick. The carrot is the retention of power the stick is forced resignation (and loss of power) and the rules say you have to publicly support Cabinet decisions or resign. In the case of Wikipedia because there is no stick, it is very rare that there is a consensus is reached. The best we can hope for is a rough consensus. I believe that explaining to people that the more critical the decision (for example deleting a page is more far reaching than renaming it) the closer the rough consensus should be to unanimity, to place numbers on these which are used every day in making decisions helps people accept a decision, particularly if they are new to Wikipedia and do not fully understand how consensus is interpreted day to day on Wikipedia. Without these types of constraints it is far to easy for administrators to act in a capricious manner.
FT2, it is much better for this guideline to report on how a rough consensus is reached in practice on Wikipedia than to write an essay on how a consensus ought to be reached. The only time that people look at this guideline is when a consensus is difficult to reach and so it ought to explain what a rough consensus is not what a consensus ought to be. --Philip Baird Shearer 10:53, 11 August 2007 (UTC)
To show you an example of what I mean. There is clearly not a consensus to remove the numerical sentence in this guideline (I for one am against it) but it has been removed several times recently without any attempt to reach a consensus to do so. If people can not even use the suggested rules for this page that they wish to impose then what is the point of this guideline? --Philip Baird Shearer 10:55, 11 August 2007 (UTC)
It is true that an essay is probably unhelpful. But a 3 paragraph brief summary of the same is not likely to be unhelpful.
Points that need mentioning about how decision-making works on Wikipedia include that 1/ policy related matters trump non-policy, 2/ evidenced stances usually trumps non-evidenced stances, 3/ arguments that would require contradicting policy are usually not relevant/invalid, 4/ within those, the scale and quality of support (and ?sometimes strength of feelings) given to different stances are also considered. I think these 4 points have wide buy-in on most Wikipedia decisions.
The role of consensus in practice is to help render a "decision" from these, and decide when one has been reached. It helps us choose decide between equally valid remaining options which one the community prefers and how strong that preference is. It's not used to decide between all options (valid and invalid) and doesn't have any real relevance until at least the first three of these matters are considered.
In some cases that is all that counts - if there are equally valid views at AFD, then that is how the decision's made usually. If it is a poll for a matter where a high level of communal support is needed (RfA, RfB etc), then specific rule-of-thumb percentages have emerged over time, suggesting the degree of majority needed to allow "communal agreement" to be claimed. That is where and how "supermajority" comes into consensus. But in all cases, policy trumps non-policy, evidenced usually trumps non-evidenced, contradictions of policy are uually not valid argument, and a sense of both scale and quality is sized up by the debate closer. The quality and size of support (and ?sometimes apparent strength of feeling) is then assessed to see if there appears to be something that can be deemed a consensus. These are points I think anyone can agree with, and quite often, it's how it works in practice. FT2 (Talk | email) 14:18, 11 August 2007 (UTC)
  • While you make a number of good points, it does sound like instruction creep if you put it in a bulleted list like that. Also, the above implies that IF all opinions given are not evidenced (or, conversely, all are evidenced) THEN the vote count decides - which is incorrect. I think the only place where the percentages still come into play is RFA/RFB (where they are a rather convoluted issue). AFD used to have a strict "vote cutoff point" in the past, but that has not been true for a long time. >Radiant< 12:12, 13 August 2007 (UTC)
We agree on avoiding CREEP; I wasn't thinking of this as being a recipe or list. More as clarifying the issues on consensus, that this is where and how consensus + percentages fit into Wiki decision making, and that other factors (quality of support, validity of suggestions, evidence, etc) are an essential part of good consensus decision making. Consensus is far from being "just about the numbers" and encouraging "decision by percentage" wouldn't usually respect the intent and spirit of the policy -- to respect and work with the different perspectives (within policy) rather than just find a sufficiently popular one.
Consensus - in the AFD sense - is more than a simple proportion (although lack of a proportion probably means lack of consensus), it is an experienced person's evaluation of the arguments, their quality, the nature of any backing support (policy, evidence), and a sense of how well they would include the views of as many as possible to be heard in a balanced manner.
The place for consensus is to form a view as to which if any valid suggestion can claim communal support.
By contrast, supermajority and percentages are rules of thumb that have evolved over time (as opposed to being imposed by policy), and seem to broadly mark the point of "communal comfort" on certain sensitive matters such as elevated access rights. They are not specific levels to be imposed. They are helpful, but usually not decisional (if that's a word). Too many other factors (policy, evidence, etc) are able to override them in all but a very few debates.
What I do notice is that the very few exceptions where supermajorities have clearly emerged and do generally get considered, are in fact precisely those few processes like RfA and RfB, intended to poll for popularity (of a decision to grant enhanced access, ie community trust levels) rather than poll for best practice or policy compliance. That's likely to be why commonly acknowledged supermajority levels have clearly emerged just for these, but have not clearly emerged for (say) AFD and content issues -- because on almost all other polls, popularity and numeric percentage concerns are second to policy and evidence-related concerns.
For this reason, I suspect there's a rule of thumb about successful use of supermajorities -- supermajority guidelines are more likely to arise and be useful (if anywhere) on decisions which are mainly about community likes, dislikes and comfort levels (eg, enhanced access such as RfA/B, policy change, etc) and less useful on decisions that are mainly about policy/guideline compliance and evidence oriented polls (eg, AFD).
FT2 (Talk | email) 17:07, 13 August 2007 (UTC)

In effort to forestall an incipient edit war, I have returned the section to its wording of 07:03, 9 August 2007 (last edit by User:Kanodin) immediately prior to my edit removing the following two paragraphs:

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.
See the pages for RM, AFD and RFA 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. Judgment and discretion are essential to determine the correct action, and in all cases, the discussion itself is more important than the statistics.

This has the net effect of undoing all of my removals, and also removing the words added by Philip Baird Shearer, which I think should not be there:

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.

Whatever the outcome of the debate on consensus versus supermajority I really do think it's wrong to pre-empt the development of thinking on consensus by, as it were, chiselling cold, hard numbers into this policy document. Omitting the numbers doesn't mean that numerical thresholds cannot develop--rather the reverse. Without hard numbers in the document those involved in developing different aspects of policy can continue to develop whatever informal metrics ("rules of thumb") they wish. --Tony Sidaway 08:15, 16 August 2007 (UTC)

I did not add the words they have been there for at least a year. The numbers should be here because this document is primarily for those new to Wikipedia who need guidence, and as I said above I do not like the move over the last year or so that "Auntie knows best" with administrators acting as judges rather than facilitators. I think it is important to get over the idea that the larger the effect of a decision on Wikipedia the closer to unanimity the decision should be. --Philip Baird Shearer 09:19, 16 August 2007 (UTC)

That idea would not need percentages specified to convey. FT2 (Talk | email) 10:01, 16 August 2007 (UTC)
The reason for the "move over the last year" is that deciding debates "by the numbers" has proven to be ineffective in a number of situations. In particular, it allows the resultant votes to be swayed by the ignorant, by people missing the point of the debate (which is alarmingly common), by people unfamiliar with or dissenting from important policies, by people unfamiliar with the larger picture, and by people being canvassed for a particular POV. >Radiant< 10:41, 16 August 2007 (UTC)

So you are of the school that "Auntie knows best". Does that mean that in cases like Vossstrasse One should simply impose the rules of common English and ignore the opinions of those that think Voßstraße should be used? --Philip Baird Shearer 09:48, 23 August 2007 (UTC)

I oppose this edit by User:Radiant! for two unrelated reasons. The first is a matter of procedure: I oppose editwarring, especially on a policy page, and especially without concurrent use of the talk page. Instead of repeatedly reverting without using the talk page as Radiant! has done recently, I urge everyone involved to use the talk page without repeatedly reverting.

The second reason is a matter of content. I think it's probably better to include the sentence with the numbers. Radiant! stated in an edit summary: "Those words you mention contradict the previous paragraph as well as existing practice; they are misleading." (Please state your arguments on the talk page, not only in edit summaries.) The words do not seem to me to contradict in any way the preceding paragraph. Please explain where you think the contradiction is. I don't see how they're misleading. Please explain. I don't see how they contradict existing practice. Please explain. Some decisions, such as featured picture, AfD etc. do involve some consideration of numbers by some people (as well as consideration of validity of arguments, etc.) so I don't think it's factually incorrect. The role of policy is not just to document what is being done but to state what should be done -- the ideal to strive for. However, as someone pointed out above, documenting current practice for the benefit of orienting newcomers is also of value. The sentence with the numbers is stated very softly: no hard numbers are asserted; the words "about" and "over" indicate that the numbers mentioned are not to be applied as hard limits.

Re allowing votes to be swayed by the ignorant etc.: perhaps some people closing a poll count the number of votes that are associated with apparently valid arguments, that do not appear to be sockpuppets etc., while ignoring the rest. The existence of problems with a system does not necessarily mean we don't want to use that system, since there are probably also other problems with alternative systems.

As a matter of precedent, a quick examination of the page history suggests that the sentence with the numbers was present steadily (was there on all the pages I spot-checked) from Feb 2006 to about a month ago. --Coppertwig 13:59, 23 August 2007 (UTC)

Is there any way to split the page into a policy page and a guideline page? The numbers should perhaps be on a guideline, not a policy. --Coppertwig 14:08, 23 August 2007 (UTC)

After re-reading some of the above and straightening out in my mind some confusion as to whether this is a guideline or a policy, I now think that if it's policy, then the numbers sentence should not be included, but if it's a guideline, it should be. (However, I don't support any edit to remove the numbers sentence at this point in time as I don't think consensus has been reached on removing it, and apparently it's been here as policy pretty steadily for some time.) I suggest moving the numbers sentence to a guideline page such as Wikipedia:Polling is not a substitute for discussion, which has a section about how to carry out a straw poll ("Straw poll guidelines"), or Wikipedia:Straw polls.

Reasons for not including the numbers sentence in policy: Policy should be prescriptive, not descriptive. (It wouldn't make sense to repeal the law against stealing just because stealing regularly occurs.) It could be dangerous to write voting methods into policy, given the possibility of clever use of sockpuppets to take over Wikipedia (evil laugh Wooooooo- -hoo-hoo-hoo-hoo-hoo.) --Coppertwig 16:30, 23 August 2007 (UTC)

Meaning of consensus within the context of policy

Re this edit [9] by Omegatron: I prefer the previous version. Omegatron's is simply factually incorrect: when people talk about "consensus" during a discussion, they rarely mean the consensus which supports the policies (though that is assumed); what they usually mean is the consensus for a specific decision about a specific article. "Will almost never" is good in the sense of WP:IAR. --Coppertwig 19:15, 22 August 2007 (UTC)

I don't understand your objection. — Omegatron 01:59, 23 August 2007 (UTC)

Old version:

When consensus is referred to in Wikipedia discussion, it always means 'within the framework of established policy and practice'. Even a majority of a limited group of editors will almost never outweigh community consensus on a wider scale, as documented within policies.

Omegatron's version:

When consensus is referred to in Wikipedia discussion, it refers to the framework of established policy and practice. Even a strong majority amongst a limited group of editors cannot outweigh community consensus on a wider scale.

I think both versions are wrong. When someone talks about consensus they mean the consensus of the editors taking part in a debate. That is the whole point of involving more editors in a debate through RFCs etc, it is assumed that the more editors that are involved the more likely it is that the consensus will reflect the community consensus on a wider scale. Trouble is of course that this is often not true in which case the 3RR rule means that a determined mico majority can impose their views. --Philip Baird Shearer 09:45, 23 August 2007 (UTC)

The second is wrong for sure - consensus in everyday discussion does not simply mean "the framework of established policy and practice". What is important here is that consensus is trying to get a sense of "if the community had to pick a way of saying/doing it, how would it be said/done". The community is thousands of individuals with widely differing views, but even so, it has a number of established principles and practices with high 'buy-in' (NPOV, OR, CITE, NPA, etc). Even when policy itself changes, it changes in a manner (and via a process) that the community has tacitly approved.
To underpin this, I'm not aware of an exception where change (including policy/practice change) has been imposed on the community without following some method that has wide buy-in as a communal practice; even if it is imposed by the Foundation or Jimbo, that's still for exceptional cases a process with wide acceptance. Lesser changes are by discussion, BRD, or "silence=consent", all of which have buy-in. Changes failing to gain agreement that they are done with communal approval and some form of communal agreement, following an explicit or implicit communally agreed practice, are almost inevitably reverted, I think.
This is the starting point for the above sentence. It's to note that when we discuss, claim or seek consensus, we always mean "within the framework of established policy and practice". Editors may agree or disagree on content points, have disputes, reach mediated agreement, be arbitrated, edit, revert... and all these are fine, but any claim of consensus by a group of editors (whether that there is consensus in a dispute, in a sanction, in a content issue) is always within the framework of established communal standards and practices. if it's not, then it isn't a viable consensus.
The purpose of the sentence is to remind editors that it isn't enough to have agreement on a point, and then one can say "there is consensus on this". The agreement is always within the broader framework that thousands of editors have agreed NPOV is mandatory, OR is a core policy, personal attack is never acceptable, unverified information needs support, text must meet GFDL, etc. Any matter described as "consensus" must be said or done within the framework of communal practices, whether it be a dictat by Jimbo, a content issue, a use of a DR process, or a core policy change. Unilateral declaration that a small group are happy and therefore communal practices are ignored beyond that are never permissible.
It is in that sense that we say When consensus is referred to in Wikipedia discussion, it always means 'within the framework of established policy and practice'. That is a core principle of consensus on the Wiki, and also explains how a consensus of a few editors, and consensus of the community, relate to each other. FT2 (Talk | email) 11:02, 23 August 2007 (UTC)

I have problems with the bold text above. For example look at the dispute about reference tags on WP:FOOT and WP:CITE those claiming there is a consensus for prescription after punctuation are not claiming it is "within the framework of established policy and practice", they are claiming it because no one edit warred by changing the guidline, but only discussed discussed the issue on the talk page. The change from WP:V to WP:ATT dispute was not about "within the framework of established policy and practice" it was about changing established policy. --Philip Baird Shearer 11:20, 23 August 2007 (UTC)

  • It may not be the best wording, but the point is important. An individual wikiproject, or group of editors working on an article, cannot meaningfully choose to override important policies or guidelines just because they want to, because their "small consensus" is assumedly overridden by the "Wikiwide consensus" in support of the policy/guideline. >Radiant< 11:22, 23 August 2007 (UTC)

To give an example of where policy is ignored why are lots of European names given in forms which are clearly not WP:NC "Generally, article naming should prefer what the majority of English speakers would most easily recognize, with a reasonable minimum of ambiguity, while at the same time making linking to those articles easy and second nature." eg Lech Wałęsa. It happens with Polish names because the majority of editors who take an interest in Polish articles tend to be Polish and they are less discomforted by the use of Polish squiggles than not to use them. One can argue the issue on the talk page but it will not be changed because the local consensus is not to change them to English spellings (see Talk:Lech Wałęsa/Archives/2012/April#Requested move). --Philip Baird Shearer 11:30, 23 August 2007 (UTC)

Isn't that similar to the Gdanzkig debate? >Radiant< 11:38, 23 August 2007 (UTC)
No it is not similar to the Gdansk/Danzig debate, but is is the same as the Gdańsk Gdansk debate. The vote on which name to use at different times was with the use of the name Gdansk yet despite that the page has been moved to Gdańsk. --Philip Baird Shearer 12:29, 23 August 2007 (UTC)
Why single out the Poles, many people who's mother tongue is not English press for similar changes eg voss-strasse or the recent debate to depreciate the use of Kosovar and Kosovan on the Kosovo page. The point I was making is that a local consensus often overrides the Wikipedia wide consensus and these things often do not go away, but remain on the talk pages of articles for years whatever the policies say. --Philip Baird Shearer 12:46, 23 August 2007 (UTC)
  • That does not, however, mean we should change the policy to accomodate that. >Radiant< 12:49, 23 August 2007 (UTC)
No but it does mean that the policy should reflect what people mean when they say consensus and it is not always "When consensus is referred to in Wikipedia discussion, it always means 'within the framework of established policy and practice'." --Philip Baird Shearer 12:52, 23 August 2007 (UTC)
(edit conflict) I reverted to the earlier version of the paragraph pending discussion, as there doesn't seem to be consensus at this time for Omegatron's change.
Omegatron, I'll try to explain. Here's one way of looking at it: there are two different activities: changing policy, and changing articles. Changing policy typically involves a consensus process among a large number of editors. Changing articles typically involves a consensus process among a small number of editors. When people are editing an article and use the word "consensus", they are usually talking about the latter, i.e. the consensus process for changing the article; they're not usually talking about the consensus process that developed whatever policies they're invoking during the article editing process. Your version makes it sound as if whenever people use the word "consensus", they are always talking about the former, i.e. the broad consensus by which policies are developed.
The second objection is related to WP:IAR. It's dangerous to put in the word "never" because cases will be found where an exception should be made. Exceptions should not be made lightly, but can be made in consideration of the apparent reasons for the rule being overridden and the likelihood that cases like the one under consideration were intended to be covered by the rule. The wording "almost never" is sufficient to warn people not to override the rule lightly. --Coppertwig 12:55, 23 August 2007 (UTC)
Ah, I see your point, Philip Baird Shearer. How about "When consensus is referred to in Wikipedia discussion, it should always mean 'within the framework of established policy and practice'.".
If the discussion is about changing established policy, I'm not sure that's quite accurate. But I can't think of a better wording at the moment and I think it's OK to leave it like that, since even when changing an established policy or practice, there is still a framework (or should be) of other established policy and practice as to how to go about changing that policy or practice; or to put it another way, we would generally only change one policy or practice at a time, so the framework in general is still there the whole time. --Coppertwig 13:04, 23 August 2007 (UTC)
(Comment - It's an accurate statement even for policy related changes. There's very clear established policy, practice and consensus about how established policy can change. For example:- Jimbo dictat yes, legal OFFICE requirement yes, community discussion yes, minor changes and improved wording on policy talk pages with general agreement yes, BRD for wording improvements if not disputed yes, proposed out of the blue yes if agreed but unlikely to be successful, group of 4 editors on some article talk page deciding its consensus no, arbcom member unilateral proclamation no, random IP editor wandering in and saying IDONTLIKEIT, no. So it's clear that even when it's a policy change, the change is only effective if it accords with communally-established practices. It's almost tautological in fact. So I wouldn't worry about that aspect either.) FT2 (Talk | email) 14:08, 23 August 2007 (UTC)

Nutshell

In general, I think that "nutshell" statements are pointless and often counter-productive. While convenient, they encourage readers to assume that they understand the full policy just by reading (or quoting) the nutshell and miss all the nuance that's described on the rest of the page.

In this case, the "nutshell" has been moved four paragraphs down the page, defeating what little purpose it ever had. I think the template should be removed from this page. Removing it will clean up the flow of the page but not sacrifice any content. Rossami (talk) 14:37, 23 August 2007 (UTC)

Quis custodiet ipsos custodes?

I think there is a misunderstanding of the paragraph

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.

It is not saying that a poll is the way to decide what is a "rough consensus" (I prefer that phrase to "supermajority"). It is saying when there is not unaminity this is how Wikipeida editors usually decide if the is a "rough consensus" for change.

For people who are new to Wikipedia, it is important that they understand what Wikipedia consensus is (so that they can contribute positively to discussion and not feel that they have been mugged -- after all people only contribute because they like to do so -- and many decisions, particularly those with a guillotine time limit, explicitly in RM, AFD and RFA and implicitly in most forum where a Wikipedia decision need to be made without unanimity (which is what anyone with knowledge of Cabinet government would expect it to mean -- most educated people from the Commonwealth of Nations).

The problem with "Administrators who close the discussion decide", argument is first not all decisions made by consensus in involve an administrator, and secondly it is by no means agreed that "auntie know best" (see Wikipedia:Admin accountability poll/RFA should be...). The major reason "auntie know best" is not agreed to is Quis custodiet ipsos custodes?. Administrators make mistakes for two principle reasons, either they support one side of a discussion and intentionally or unintentionally are persuaded by the arguments presented by that side, or they make a mistake because they are unaware of some of the complications in the policies -- For example when deciding on cases like Liancourt Rocks which takes priority WP:NC or WP:NPOV and is Dokdo the most common name and even if it is non NPOV? (I'm glad I did not do that one) -- to be seen fair one would have to put in an appeal process, and few of us want to see more "instruction creep".

So explaining that when a decision is made it is guided by a Wikipedia consensus and that a Wikipedia consensus varies between 60% to over 80% depending upon the decision, with the more critical processes tending to have higher thresholds, is an accurate description of how a Wikipedia consensus is reached in most cases and is likely to lead to more editors accepting such decisions. --Philip Baird Shearer 09:38, 24 August 2007 (UTC)

Wrong. Most consensus in most situations is not even decided by admins, and ~50% is likely decided by anons and non-regulars. In over 90% of situations there aren't even enough editors present to hold a viable majority type vote in the first place. I'm basing this on publicly published and discussed hard numbers (including talks at wikimania 2006 and 2007). --Kim Bruning 00:51, 26 August 2007 (UTC)

I agree with you that consensus building does not always involve administrators and it has nothing to do with votes! As I wrote above "argument is first not all decisions made by consensus in involve an administrator," and "It is not saying that a poll is the way to decide what is a "rough consensus" (I prefer that phrase to "supermajority"). It is saying when there is not unaminity this is how Wikipeida editors usually decide if the is a "rough consensus" for change.". I do not know what percentage of the consensus building exercises result in unanimity but for those decisions that are made that do not end in unanimity, numbers indicating what a rough consensus are useful. --Philip Baird Shearer 12:04, 26 August 2007 (UTC)

Not unless you want to murder the definitions of both consensus and rough consensus. I think we've been over this several times over the years? Just repeating your position over and over won't get you consensus here :-P --Kim Bruning 12:23, 26 August 2007 (UTC)

Consensus is easy it is an "[a]greement in opinion; the collective unanimous opinion of a number of persons"(OED) and I think we can agree that that rarely happens on Wikipedia. So what do you think a Wikipedia "rough consensus" is? --Philip Baird Shearer 12:29, 26 August 2007 (UTC)

A wikipedia rough consensus is where people in general agree to a course of action, several agree to compromise, and outliers are ignored. In certain situations, it is possible that a single person ultimately determines the outcome of a discussion (because their arguments are very convincing indeed), at other times very large groups of people together still can't reach a decision. (and in fact the downside of consensus is that it is more difficult to work with in larger groups). --Kim Bruning 12:49, 26 August 2007 (UTC)

Unless the people involved in a consensus building exercise agree that one person's "arguments are very convincing indeed" and so agree a consensus how does one judge that that person's "arguments are very convincing indeed"? --Philip Baird Shearer 12:53, 26 August 2007 (UTC)

Straw man. One doesn't, of course. The argument is convincing only if people are actually convinced. (It does occasionally happen, believe it or not. :-) --Coppertwig 13:47, 26 August 2007 (UTC)
why do you say straw man, Kim put the argument forward that "it is possible that a single person ultimately determines the outcome of a discussion (because their arguments are very convincing indeed" but that is not true unless the others in the discussion change their minds and unless everyone does, one is still going to have to work with a rough consensus. --Philip Baird Shearer
Didn't we cover this before sometime? An example from a while ago: on an AFD (before we had a separate copyvio system) "Copyvio" was a pretty darn convincing reason to delete. --Kim Bruning 14:51, 26 August 2007 (UTC)

The numbers came back again

So I removed them. --Tony Sidaway 19:44, 25 August 2007 (UTC)

Did you not read the section above? There is not consensus to remove them. --Philip Baird Shearer 12:05, 26 August 2007 (UTC)

I have read the section above. There is clearly no consensus to include them. Stop adding them. --Tony Sidaway 12:10, 26 August 2007 (UTC)

The numbers have been in the article long time (looking at the history with 500 edits set returns the edit on 5 April 2005 which has the numbers in it, and the initial number of 80% was added on 5 December 2004). It is up to you to show that there is a consensus to remove them. --Philip Baird Shearer 12:30, 26 August 2007 (UTC)
Sure, I've just checked all the referenced pages, and can't find a useful citation to support the position of that paragraph. (I looked at each page one by one, and ended up removing each of them from the list). Hmph, I think use of {[citation needed}} in the project namespace might be way useful. ;-) --Kim Bruning 12:50, 26 August 2007 (UTC)

Just because some people have been running around deleting numbers does not mean that they are not still in use. See for example the recent debate on Liancourt Rocks‎, and I come back to my initial point on these processes, particularly those where an administartor is involved I come back to Quis custodiet ipsos custodes? --Philip Baird Shearer 12:58, 26 August 2007 (UTC)

The way forward there is obviously not going to be found by using numbers though. For instances like these, we should introduce some way to put different names on equal footing. Let me ponder on that for a bit. --Kim Bruning 15:03, 26 August 2007 (UTC)

Consensus vs. supermajority

I suggest that we remove the word supermajority and instead use the words "rough consensus" as I think the term supermajority implies a poll and is not helpful. It also divides opinion on this talk page in what I think is an artificial divide. --Philip Baird Shearer 12:49, 26 August 2007 (UTC)

Interesting thought! :-) --Kim Bruning 12:51, 26 August 2007 (UTC)

How about something like this:


Start the section "Consensus in practice" with:

The OED defines consensus as an "[a]greement in opinion; the collective unanimous opinion of a number of persons" but in the Wikipedia project it is widely agreed that consensus does not mean that everyone agrees with the outcome; instead, it means that everyone agrees to abide by the outcome. The following description of Wikipedia consensus ...

=== Consensus vs. supermajority ===

While the most important part of consensus-building is to thoroughly discuss and consider all issues, it is often difficult for all members in a discussion to come to a single conclusion. So that decisions can be reached and development of Wikipedia can continue there is wide spread agreement among Wikipedia editors that a "rough consensus" is sufficient for a decision to be implemented. Deciding what a "rough consensus" is has developed gradually over the length of the project. The numbers mentioned as being sufficient to reach rough consensus vary from about 60% to over 80% depending upon the decision, with the more critical processes tending to have higher thresholds.

For example there are three processes which have a self imposed guillotine time limit of five days (although if no consensus has been reached, and there is a consensus to do, this may be extended) these are Requested Moves, Articles for Deletion, and Requests for Adminship, and although occasionally there is unanimity, it is common that agreement is reached to implement the decision based on the "rough consensus" of those who have expressed an informed opinion.

In other cases where there is a large community interest it is unusual for a unanimous consensus to coalesce around one agreed outcome, due to the fact that more people participate than can effectively cooperate (see: Dunbar's number). In such situations often a decision to change Wikipedia will be implemented if a "rough consensus" to make the change can be agreed upon.

Need something here about the use of polls and a rough consensus.


--Philip Baird Shearer 14:04, 26 August 2007 (UTC)

Consensus on Wikipedia does not mean that everyone agrees to abide by the outcome. People who have been blocked for repeatedly reverting are not agreeing to abide by the outcome; they are being forced to accept the outcome. People who have given up or run out of energy or are busy with other things are not "agreeing" to abide by the outcome; they are merely abiding by the outcome. One could say that they are "accepting" the outcome.
I don't accept that on Wikipedia the term "consensus" means something different from its dictionary definition, and I'm not sure that this is widely accepted. I think it's more productive to use a different term than to try to use words in ways that go against the dictionary, which is confusing since many (particularly newcomers) will assume the dictionary meaning at least as a connotation.
I agree that "rough consensus" is a better term than "supermajority". "Rough consensus" brings an image into my mind of a few editors on an article page deciding to run with an edit even though one or two are not very happy with it, while "supermajority" brings up an image of the listing and counting of votes. The description of consensus (or rough consensus) should focus mainly on the former and then only briefly mention the latter. Also, "rough consensus" suggests to me that it's based partly on how strongly the dissenters feel, while "supermajority" sounds as if it's just going to count how many of them there are and then (if numbers warrant) overrule them regardless of how they feel about it.
It's important to emphasize that people should not just ignore the opinion of a small number of editors, but should engage in discussion and address the points that those editors bring up, and try to reach consensus (in its true dictionary meaning). --Coppertwig 14:19, 26 August 2007 (UTC)


Policy isn't an essay and should not be apologetic for poliies, nor speak to "you" (the reader); I've reworded the first two paragraphs of this section mainly for style and wordiness issues (including over complication).
The only addition or change of any significance is to rewrite the sentence that states if supermajority is used one must review the discussion, and noted that this includes checking the quality and strength of consensus, the nature of other concerns, and any overriding policy-based matters. I don't think that can be very controversial, it's what any debate closer has to do anyhow. But should be said. FT2 (Talk | email) 14:47, 26 August 2007 (UTC)

Rough consensus is not determined by numbers. What Coppertwig and FT2 point out that the quality of the arguments of the pro and dissenting sides are very important to the decision. It is quite fun to shift a consensus in a sane direction by providing good arguments, and discussing with people one by one. --Kim Bruning 15:02, 26 August 2007 (UTC)

Of course a rough consensus is determined by numbers. If it is not then how one determine the quality of the argument? If the the other people in the debate do not change their opinion it would suggest that the quality of the argument is a subjective point of view. --Philip Baird Shearer 15:35, 26 August 2007 (UTC)

I am suffering from deja vu over this let us not progress this argument in this thread as it is already covered in the section Quis custodiet ipsos custodes? above. Instead lets see if we can agree to remove the subheading Consensus vs. supermajority. --Philip Baird Shearer 15:43, 26 August 2007 (UTC)

Well there's one thing I haven't seen before...
My current understanding of the universe:
Quality == Quality; Quantity == Quantity; Quantity != Quality.
Are you trying to convince me that? :
Quality == Quantity
--Kim Bruning 15:57, 26 August 2007 (UTC)
In other news, perhaps we could reword and not use majority, or we could recognise that people do use majority as a kind of early approximation. You have me convinced me that we may want to mention something about that. --Kim Bruning 16:01, 26 August 2007 (UTC)
  • In light of the above, I don't think that this wording is practical: "As a quick rule of thumb people will often check whether a supermajority exists". This implies to novice people that a simple supermajority count may be used instead or consensus, and similar wordings have in the past given rise to arguments of the kind of "we're in the majority, so you must shut up now". >Radiant< 14:01, 27 August 2007 (UTC)


"If people encounter a majority in favor their ideas, people will often decide to spend their time on other matters. If there is no clear majority, people might decide to participate in the discussion proper, and try to negotiate a more favorable compromise." -- what's this trying to say, before it gets reworded into better English? "If its obvious, people may not choose to waste time, if it's not obvious they discuss more"? FT2 (Talk | email) 15:10, 27 August 2007 (UTC)
Interesting question. On the one hand, it alludes that people who share the majority opinion might not voice their opinion because it's not necessary; based on, say, AFD, this does not appear to be the case in general. On the other hand, it points to the problem that frequently, majorities are unwilling to compromise, because they "hold the vote" already. >Radiant< 15:20, 27 August 2007 (UTC)

(Edit conflict) Quite frankly, I find the entire "consensus vs supermajority" section to be garbled, confusing and lacking in cohesion. What, for example, is meant by If people encounter a majority in favor their ideas, people will often decide to spend their time on other matters? Is this supposed to mean that if people are confronted with a (super)majority, it will often be enough to resolve the issue without further debate? If so, then I suggest it should say as much, because otherwise the sentence appears to have no discernable point.

Then there is More in general formal decision making based on vote counting is not how wikipedia works. "More in general formal decision making"? Can someone translate this into English for me?

Then there is As a quick rule of thumb people will often check whether a supermajority exists, which is simpler to identify, and which can often indicate whether there is a rough consensus [but] The actual determination can only be done by carefully reading the entire discussion to find out what the actual consensus is. Isn't that self-contradictory? If the "actual determination can only be done by carefully reading the entire discussion", what is the point of holding a vote to begin with?

BTW, I find it hard to understand the wrangle over whether numbers should appear on this page or not, what's the difference when a user only needs to click on the "supermajority" link to find it approximates to 60%+ anyway? Gatoclass 15:27, 27 August 2007 (UTC)

Update: I've now made some edits designed to address these issues, which I think make the meaning more clear. Gatoclass 16:39, 27 August 2007 (UTC)
It should be "More in general comma formal decision ..." - and note that Wikipedia:Supermajority is a rejected proposal, not some kind of rule. >Radiant< 15:40, 27 August 2007 (UTC)
Thanks for that, I've rephrased the sentence slightly in line with your explanation :) Gatoclass 16:39, 27 August 2007 (UTC)
After "will often decide to spend their time on other matters" or "will often be enough to resolve the issue without further debate", I suggest adding something like "on the other hand, sometimes even a single user will continue to put forward arguments, which other users should consider." This is needed, otherwise people will say "consensus is against you", citing this part of the policy, and refuse to listen to any arguments. Also finding some way to change the wording so it suggests (sometimes, often, usually) informally estimating the numbers on each side of an argument without any straw poll; i.e. someone just looks at a discussion and decides what to do based on it, i.e. whether to make an edit, participate in the discussion or go away. The current wording seems to imply a formalized straw poll. --Coppertwig 16:43, 27 August 2007 (UTC)
I see what you mean. I'll see if I can find a way to rephrase that. Gatoclass 16:57, 27 August 2007 (UTC)
Okay, I've re-edited that sentence, does it read better now? Gatoclass 17:14, 27 August 2007 (UTC)
Thank you very much for listening to my concerns. Yes. I like this version: "Sometimes however, users who find themselves in the minority may choose not to concede, in which case discussion should continue in an effort to try and negotiate a more favorable compromise." However, I don't like the words "genuine concerns" introduced by FT2, since it makes it too easy for people to dismiss concerns by saying they're not "genuine" -- as they tend to seem not to be to those who disagree with them. The policy should encourage discussion when someone still doesn't agree, not only "sometimes". It need not explicitly say "always", though. --Coppertwig 19:17, 27 August 2007 (UTC)
Yes, FT's version does appear to indicate that he thinks a supermajority should be a means of silencing a minority. That's the impression I get when I read it anyhow. But then, it's late and maybe my judgement is askew. I'll come back and have another look of it tomorrow. Gatoclass 19:33, 27 August 2007 (UTC)


That's not my view at all. Given how often I've emphasized that consensus need to respect all views (majority and minority), that numbers alone don't necessarily reflect (policy based) consensus, and that one should consider the quality of different viewpoints and the policy-based reasons touching upon by them, when considering the importance of numbers from a poll.... you'd think that this would be rather evident :) I'll buy that it's late and one of us may have written or read poorly though. Discuss tomorrow? FT2 (Talk | email) 21:45, 27 August 2007 (UTC)
Well let me put it this way. You removed the statement that a supermajority is "not a binding vote". You then added an additional caveat implying that a minority position should only be recognized when it constistutes a "genuine concern". The net effect of these changes is to considerably weaken the position of the minority, to the point where users could not unreasonably assume the majority view can indeed override the minority whenever it decides the latter's arguments lack validity.
I'm not saying you're wrong to imply such a thing. Maybe the policy is meant to be deliberately vague to leave both minorities and majorities unsure of their position, which might arguably encourage them to engage more rather than use policy to take an intransigent position. Maybe the section should hint that the majority can prevail where it concludes that the minority's objections are not "genuine" or otherwise lacking in validity (in practice this is what often seems to occur anyway). It's just that I'm not sure that's how the policy is intended to operate, that's all. Gatoclass 02:00, 28 August 2007 (UTC)
I'll try to explain once more, if it'll help. If it doesn't, please let me know. Consensus is a complex and subtle issue. We want to respect all views that have a good foundation in policy, whether majority or minority, yet not bind ourselves up in stalemate when it dead-ends, and not encourage a state where either is simply ignored. We need to respect both a majority, that many people think the same; and also respect a minority, that some have strong concerns. These are complex problems to balance.
By way of respect of minorities and consensus vs. supermajority, here's a personal example. I had about 80% approval midway through my first RfA. RfA "rule of thumb" is roughly, <70% doubtful, 75% likely agreement, 80% almost certain agreement. And yet, I (unilaterally and voluntarily) withdrew. A significant minority felt I hadn't used edit summaries sufficiently, and I wanted to respect their concerns by ensuring a demonstrable track record. There was no good reason to override concerns that could be validly addressed. I use that as an example of where supermajority and consensus differ: I had a supermajority. But RfA requires consensus, and consensus implies that minority concerns are addressed and taken seriously, not ignored, where reasonable. My 80% was not (in my view) 'consensus', and out of respect for that, and for no other reason, I withdrew from RfA.
I removed the statement that supermajority is not a binding vote because it seems to make a point that is better made elsewhere, and has been made elsewhere (if you state that minority concerns must be considered, that implies that majority is not an automatic binding vote, surely?) Also, if you allow that any minority position must be given equal weight to any other, you fail to distinguish those positions which concur with communal views (eg that there is a policy-related issue) and those which do not (eg trolling). That's a pretty important distinction to make, but hard to codify since "personal views" may be good faith concerns or otherwise. In practice we solve this by leaving the interpretation of whether a "genuine [good faith] concern" exists and what weight should be given, to other discussion parties. But the broad principle, that good faith genuine concerns should not be ignored, but should be considered at the minimum, is an important principle.
With regards to your 2nd point, policy usually tries to briefly document the main communal "best practices". They are part explaination and tutorial, and part rulebook, in a way. So lets try to see what we know about consensus and supermajority on the wiki:
  1. We know that supermajority is usually respected, but their view is not automatically binding in any debate, especially in the face of a good policy based objection.
  2. We know that even if there is a supermajority, the core process is still consensus seeking, we still seek to reassure and include minorities and take account of their concerns if possible. Or at the least we do not treat them as ignorable just for being a minority.
  3. We try to distinguish good faith vs. bad faith, or at least encyclopedic vs personal interests. So in most debates, whether at AFD and RFA or on article debates and WP:DR, we tend to give more weight to some kinds of concerns and views, less weight to others.
  4. We do not use consensus or respect of minorities as a license for unhelpful activity (trolling) or puppeteering.
  5. We know that closing a debate often takes experience to balance, because it requires a fine sense of what policy says, what the spirit of policy says, and the weight and quality of points made, as well as sheer numbers.
  6. We know that on processes to do with "communal trust" broadly respected supermajority rules of thumb have emerged (RfA, RfB), but on debates more related to policy points (AFD, PP) they have not. (Probably since in the former we are testing for supermajority support vs the latter where we're testing for compliance.)
  7. We know we want to encourage people to work collaboratively, respect all views, and this requires ready compromise in some cases, but standing ones ground in others. We want people to judge how others see things, and consider their own stance rather than merely react to push their own view.
  8. If there is still disagreement, and its not a debate specifically ascertaining popular support (RfA, RfB, new policy, etc), then the preferred recourse is to to seek more (or more experienced) eyeballs rather than quoting numbers.
This is broadly core stuff about consensus on the wiki, I don't think any of the above are much disputed. The question is, how do we summarize supermajority vs. consensus without accidentally giving either side a "weapon" to sidestep consensus.
The wording is delicate since we want to guide people towards a communally sanctioned consensus approach, but we do not want to either give the majority a justification to unreasonably ignore other significant views, nor give the minority a justification to unreasonably hold articles hostage. Edits which push too much on either side (as opposed to pushing for reasonable collaborative consensus seeking whatever the discussion) are likely to miss the point.
FT2 (Talk | email) 10:05, 28 August 2007 (UTC)
After all that, I can only assume you *do* want to keep the text nice and woolly :b
but we do not want to either give the majority a justification to unreasonably ignore other significant views, nor give the minority a justification to unreasonably hold articles hostage
That's why I said I thought the "genuine concerns" bit could arguably be retained. However, as both your edit and mine have since been supplanted by those of others, this discussion appears to have become somewhat moot. Gatoclass 12:19, 28 August 2007 (UTC)
Seems so. I may step back a little a bit from this discussion, to see what others make of it and where others take it for a while. My main concern will be that we don't misrepresent consensus in any proposed wording. Above all other policies, consensus may be one of the most tricky to pin down, because it reflects a human question of judgement and balance based upon experience, rather than a factual observation on objective matters. Every attitude needed, whether compromise or tenacity, can be used in a consensus-building way or a consensus-destructive way. It's important that we codify the attitudes of mind and checks and balances, almost more than the actual judgement calls, since we can always obtain judgement calls from an understanding of what consensus is intended to do. FT2 (Talk | email) 18:15, 28 August 2007 (UTC)

Consensus v. supermajority: Arbitrary break #1

In reply to Gatoglass: the difference is whether the numbers appear in policy or not, therefore whether people are going to say things like "the policy says 80%, and we have 80%!" and refuse to listen to other editors' arguments. --Coppertwig 16:45, 27 August 2007 (UTC)
Yes, but if the policy page itself refers to the concept of supermajority, what is the point of pretending that "supermajority" means something other than what the included wikilink says it means? Gatoclass 16:57, 27 August 2007 (UTC)


Not bad, I've tried to build on your version to have yet another go at this thorny wording. (DIFF). Changes:

  1. Changed wording of "thoroughly discuss". Emphasis switched to "all issues are considered and listened to".
  2. Removed redundant extra wordage sentence "not a binding vote", anyone who reads the rest (especially as rewritten) will see this.
  3. Added the need to consider "the basis of objection of those who disagree" to the comment about "if there is a supermajority". Rather important.
  4. Rewrote 2 sentence third paragraph to clarify the "softer" but crucial side of consensus seeking: 1/ that compromise is part of consensus building, and 2/ if it stalls outside views are often sought, which is in practice what happens and what we do want people doing. Not just arguing between "sides".
  5. last, rewrote the sentence "Except in standard processes like WP:AFD and WP:RM, polling is discouraged, and should not be seen as a means of reaching consensus", to read, "For this reason, the use of polling alone as a means of decision-making (as opposed to decision-exploring) is widely discouraged". More to the point, less specific.

FT2 (Talk | email) 18:17, 27 August 2007 (UTC)

I disagree with most of your changes. I think they just make the text by and large more repetitive and woollier and harder to understand. I think "not a binding vote" should remain because it is crystal clear - why assume the reader will glean this from the rest of the discussion? Experience strongly indicates this will not be the case.
Maybe the bit about "genuine concerns" should remain, because it leaves the door open to just ignoring WP:DICK's I guess, but then I'm not sure if that's the actual intention of the policy. Gatoclass 18:49, 27 August 2007 (UTC)
The problem is that consensus is inherently a fairly woolly subject. The whole "what percentage" debate arises as a response to the fact that that the community by and large rejects bright line decision-rules, preferring to rely instead on discussion, compromise and willingness to work with others.
So for example, there is no requirement to "thoroughly discuss" (although its often a Good Thing) - in practice viewpoints are often ignored if they do not make an impact. The statements that one must consider XYZ other things first makes clear that percentage alone is not binding; it doesn't need restating. It is important to make clear users (on both sides) are expected to think about when to compromise and when not, as an integral part of consensus, not just rely on majority of numbers. These are specifics that matter. Hope this clarifies. FT2 (Talk | email) 21:45, 27 August 2007 (UTC)

I've edited again, making some essential changes, and some less essential ones. I can't quite write clearly in FT2's favored style, so I hope FT2 or someone else can edit it into that particular style if that's their preference.

The key thing I removed is the reference to policies and guidelines. I know that it's automatic to reference other policies and guidelines, and typically that's good ... but in the case of consensus: since all policies and guidelines are based on consensus, if you refer back to them, you get HORRIFIC circularities. So don't do that ;-)

Other than that, I'm not entirely satisfied with this wording, and I'm not sure if I should expand further into "safe polling" territory, or maybe leave more out, or maybe make that some totally new guideline page. <scratches head>

--Kim Bruning 22:34, 27 August 2007 (UTC)

I still think my version was more clear, although, as I suggested to FT above, maybe it's not meant to be clear, I really don't know.
But one thing I will say is that I'm kind of suprised to see how easy it is to edit Wiki policies. This doesn't seem right to me, because it potentially creates instability. I can't help but think that policy pages should be locked most of the time and only unlocked to change the wording when consensus about something is reached on this page. Gatoclass 02:09, 28 August 2007 (UTC)
No instability is caused by your editing of this page. As you can see, the sky has not fallen, and all is still right with the world. Conversely, your edits may well have helped enlighten future generations. <grin>
Note that wikipedia is a wiki. We're using the wiki editing process to write an actual encyclopedia. Why should we not use the same process for the writing of some mere guidelines? O:-) --Kim Bruning 15:07, 28 August 2007 (UTC)

Hmmm, the last paragraph now also somehow ignores RFA, AFD, etc... which isn't too good. My fault, I'm sure <scratches head> --Kim Bruning 22:41, 27 August 2007 (UTC) any ideas how to put that better?

No it's not your fault, FT removed those references. I agree they should be restored. Gatoclass 02:12, 28 August 2007 (UTC)
I don't see benerfit in mentioning them. A policy on consensus should provide general and specific guidance, but it doesn't need to specify here, which processes work which ways. These aren't actually useful as examples; what difference does it make specifically to WP:CONSENSUS to know that the WP:RFA process has a rough (informal) sense of supermajority? It doesn't seem to add anything to this policy, doesn't help people reach consensus generally, and if true and communally agreed for RFA/AFD it will be mentioned on those policies where it's more directly relevant. FT2 (Talk | email) 09:14, 28 August 2007 (UTC)

My silence is not a my approval of the recent changes, so there is no consensus for them :-) I'm away in the real world for a day or two however the numbers should be restored and so should much of the text that has recently been deleted. --Philip Baird Shearer 10:07, 28 August 2007 (UTC)

  • On the contrary, I believe the present page is way too long as it stands. >Radiant< 10:43, 28 August 2007 (UTC)
Oh that's annoying, I was sort of editing assuming that you were still present - trying to make sure that your opinions were taken into account. And I was assuming you'd correct us if we'd missed something. So now I have no idea where we stand. :-/ But thanks for warning, else the departure from potential consensus would have grown larger. --Kim Bruning 15:11, 28 August 2007 (UTC)

Thanks for taking out the example of WP:ATT. Not only is Radiant! right that it's a poor example as it was essentially vetoed by Jimbo, but also it says it was "duly promoted to policy", a statement for which there is not consensus. --Coppertwig 16:37, 28 August 2007 (UTC)

Consensus is not (necessarily) about talk page discussion

Over 90% of the time the talk page doesn't factor into consensus formation on a (mediawiki type) wiki at all. In fact, there exist wikis that have no talk pages, and still work by way of consensus. :-) So I don't want to revert a revert, though. Gatoclass, can you explain what's up? --Kim Bruning 16:49, 28 August 2007 (UTC)

Yes, but your heading gives the impression that the prior discussion was not about talk pages but something else, and that what follows is specifically to do with talk page discussions, and neither is true. So I just don't think it's an appropriate header, that's all. Gatoclass 17:01, 28 August 2007 (UTC)
Hmm, okay, fair enough... I'm just thinking that most of the consensus page is about the <10% minority of cases, whilst it might be handy to describe the vast majority too. So how to go about that? <scratches head some more> --Kim Bruning 20:14, 28 August 2007 (UTC)

Quite frankly, I think the latter half of the page has become a bit of a mess what with all the recent changes, it just seems to ramble on with no particular point in mind, so trying to come up with an accurate heading for any of it would be a bit of a challenge IMO.

Also, like FT I'm reluctant to continue trying to edit this page right now, since it seems the people currently trying to edit it have divergent views and styles resulting in a dog's breakfast, and trying to straighten it out from this point looks like a headache I'd prefer to avoid - especially given the unlikelihood that any edits I make would survive for five minutes anyhow.

Personally, I think it would probably be better at this point just to wind the page back to an earlier version and then perhaps to discuss proposed changes on this talk page first, rather than continuing with the edit free-for-all which has been going on lately. Gatoclass 08:19, 29 August 2007 (UTC)

Oh come on. :-) Try your changes out on the page, and see what sticks and what doesn't. If style worries you, edit for style. It will work out. It has done so over a million times before. :-) --Kim Bruning 21:20, 29 August 2007 (UTC)
Quite honestly Kim, policy pages are really not my thing :) I'm not sure how I got into this in the first place, I just jumped in and made a couple of spontaneous edits that I hoped would end the existing wrangle.
Generally speaking, I'm not terribly passionate about Wikipolitics/policies and I have other Wikipriorities right now. So I think I'll cede the field to those who are more passionate about this topic. Perhaps I'll drop in again in a week or two to see how things are shaping up. Thanks for the gracious invitation though :) Gatoclass 01:02, 30 August 2007 (UTC)
  • I think what Kim is referring to is the article editing process, whereas Gatoclass is referring to the discussion-based processes such as AFD/FAC/RFA and so forth. The latter obviously wouldn't work on WP without "talk page" discussion. >Radiant< 14:23, 30 August 2007 (UTC)
*Nod* those are pages that are in the absolute minority though. I do agree that they exist, of course ;-) --Kim Bruning 20:56, 30 August 2007 (UTC)
Yes, but people seem to refer to this CON page primarily from, or with respect to, those structured debate areas. Hence the confusion. >Radiant< 07:49, 3 September 2007 (UTC)
There's actually some debate about if we even still need any of those areas, interestingly enough. Mediawiki has gotten a lot tougher and can stand up to more abuse, but there hasn't been a corresponding lightening of rules ^^;; --Kim Bruning 01:09, 5 September 2007 (UTC)
I'm not sure there's a defined answer. How about, enough people to convince everyone else that there's no consensus, in a way that doesn't get them blocked for being a pain in the butt? —Preceding unsigned comment added by Wikidemo (talkcontribs) 04:08, 16 November 2007 (UTC)

Silence equals consent

(pasted from discussion at the Pump) WP:CON suggests that "silence equals consent" in the context of editing an article. Does that translate to silence equals consent in terms of approving proposals for new policies, guidelines, etc.? --Kevin Murray 23:37, 30 August 2007 (UTC)

That depends very strongly on the circumstances. Do you have a specific issue in mind? If you're not sure about whether consensus exists on a policy, wider consultation and comment will always trump 'silence' or an assumed consent. TenOfAllTrades(talk) 23:49, 30 August 2007 (UTC)
Let's be blunt. Sometimes an interested party does not notice a change to a particular section of WP until it is brought to their attention. Silence may mean many things. dr.ef.tymac 01:46, 31 August 2007 (UTC)
I don't think this is reasonable given how we treat guidelines and policies. The premise that silence is consent needs to be complemented by the idea that if someone raises an objection the old consensus is overturned and a new one (possibly the same of course) has to be constructed. Generally, however, once something is a guideline more than a single voice is needed to remove that status. Christopher Parham (talk) 01:50, 31 August 2007 (UTC)
I'd agree it's very difficult to provide an answer without a specific example. In terms of proposing entirely new policies/guidelines, a strong and clear consensus is needed to establish the guideline or policy as such (obviously moreso for policy than for guidelines). If it's standard tweaking and revision that takes place constantly in a wiki, than in the absence of objection, consensus may be assumed. If it's a significant revision, without outside solicitation (here for example) and clear discussion, such changes can validly challenged as inappropriate or invalid. At least, that's my understanding of how it's intended to work. You're welcome to some grains of salt with those thoughts. :) Vassyana 02:51, 31 August 2007 (UTC)
Silence equals consent when it comes to common practice, I think—and common practice may eventually become a policy or a guideline. When practice varies too much, however, or if practice is criticized, a more formal type of consensus is required. (Commonly in the form of supermajority, and preferably in the form of unanimity.) GracenotesT § 03:20, 31 August 2007 (UTC)
I should think a wording of "silence defaults to consent" would be better applied when it comes to policy, guidelines and the like. As mentioned by dr.ef.tymac silence is sometimes a result of non-awareness, but also it is the case that proponents of change argue they have implied consensus since the supporters of the status quo have not demonstrated their support previously - although the proponents have never demonstrated how this might be proven. It is difficult to define the degree of consent by silence (even by counting traffic against non-consent comment), but it should be recognised that it forms part of the process of consensus. LessHeard vanU 09:37, 31 August 2007 (UTC)
I certainly do not think that one person raising an objection means there is no longer consensus - that is not feasibl in a Wiki environment, with thousands of editors varying their attention. I think this is especially so when it comes to policies. "Airplane" pretty much means the same thing around the world and anyone who knows anything bout airplanes, even a newbie, can ad significant important content to the article on airplanes. But Wikipedia policies are unique to Wikipedia and there is no reason why a newbie would understand their value or implications. When thinking about consensus and decision-making and editing processes in gneral, we need to consider articles and policies separately: the former ideally require expertise in a subject matter; the latter ideally require expertise in Wikipedia. For that reason I think the threshold for change of policies must be much much higher than for articles - one person's objection surely is far below that threshold. I would even say fifty editors who have been around only a few weeks or months would still be below the threshold. Moreover, I think we need to distinguish between changing the text, and changing the policy - again this is because policies are fundamentally different from articles. An article about airplanes is about real things out there in the world about airplanes. A policy is itself the actual policy. A policy page is about itself, not about something out there outside of Wikipedia. Now, when it comes simply to improving the wording of a policy page - editing the text for a more elegant style, or clarity - newbies may have very valuable suggestions, and if say ten editors are active on the page and seven or eight agree to an edit, fine, why not. But if an editor - even someone who has been here 5 years ith 100,000 edits - proposes an edit that in effect changes the policy itself, I do not think that a consensus of "whichever editors happen to be active on the talk page that day" is enough. Ideally, I think such a proposal should involve creating a new page that presents the proposal, with its own talk page dedicated solely to debaing the proposal to change the policy (the talk page of the policy should be reserved for improving the text). And I feel very strongly that any editor or group of editors who want to change a policy need to publicize the proposal as widely as possibl: an announcement on the list-serve, on the Admin's bulletin board, and on the talk pages for other policies and guidelines, minimally. Slrubenstein | Talk 09:51, 31 August 2007 (UTC)
Sometimes improving the "clarity" of policy has the effect of subsequently altering the perceived meaning of policy. This both is usually a consequence of good faith editing, and unremarked by those who check only the diff and not its overall effect. Also, proposed changes to policy is often worked upon outside Wikipedia space, for very good reasons, and then presented as a proposed change. The authors will then have a their considered arguments marshalled, while other interested parties only become aware of it at that time and whose responses are fragmented and hastily assembled. Having a workshop space linked to relevant policy pages may be useful in reviewing the gestation of proposed policy change. LessHeard vanU 10:19, 31 August 2007 (UTC)
"Proposed policy change" is very much deprecated, since the raw numbers say it simply doesn't work on-wiki (at least not on en.wikipedia). --Kim Bruning 00:37, 5 September 2007 (UTC)
  • Silence equals consent only if sufficient people have actually read the page. That goes for articles as well as policies, though. For instance, if there haven't been any edits to an article for over a month, it does not follow that the present version is therefore consensual. >Radiant< 12:34, 3 September 2007 (UTC)
  • (sorry, I lost sight of this thread so am a bit late here...)Um... how does 30 days after the last edit deprecate the implied consensus as when it was 1 day after. "If it ain't broke, don' fix it!" and all that. LessHeard vanU 20:36, 5 September 2007 (UTC)
  • Based on the responses above, I would say that the phrase "silence equals consent" as it it used at WP:CON leads to potential confusion. I think that it should be removed or more clearly explained. --Kevin Murray 17:17, 3 September 2007 (UTC)
Please note recent changes to WP:CON, which attempt to reflect the concerns above. But we may need to fine tune or go further. --Kevin Murray 22:26, 4 September 2007 (UTC)
"Silence is consent" is the very cornerstone of Consensus in practice as used on a wiki, as far as I'm aware. And wiki-based consensus is the only method known to mankind to practically manage so many pages with so many collaborators, that I'm aware of. I am always willing to listen to improvements of course. But pulling out the bottom brick of wikipedia sounds a little tricky, unless you happen to have some replacement on-hand? --Kim Bruning 00:36, 5 September 2007 (UTC)
I believe that "Silence is consent" entered this policy around November '06 with little discussion and/or notice. --unsigned
A common complaint was that certain things were not yet codified, pretty late in the history of wikipedia. Some things still aren't. Of course, then writing things down itself gets criticized. Damned if you do, damned if you don't :-P --Kim Bruning 03:03, 5 September 2007 (UTC)


I would like to correct Slrubenstein above. Most wiki-pages are edited by up to maybe 30 people total throughout their lifetime, with maybe only 3-5 people working on the page at any one time. I think we should concentrate first and foremost at ensuring that this vast majority of pages continues to be edited efficiently. Perhaps we should split the guidelines for "high traffic" (minority pages) and "low traffic" pages somehow? --Kim Bruning 00:36, 5 September 2007 (UTC)

I could see some distinction between the content pages and the infrastructure pages. --Kevin Murray 01:56, 5 September 2007 (UTC)
Define "infrastructure pages". --Kim Bruning 03:03, 5 September 2007 (UTC)

I would never say that "Silence equals consent" but you can say that "Silence implies consent". Consensus is demonstrated through the edits and comments of the community of informed editors. But we are also encouraged not to comment unless we have something important to say. Those who choose not to comment are still presumed to have read the discussion - and to have made a conscious decision not to add a comment. Rossami (talk) 01:42, 5 September 2007 (UTC)

I agree as long as a broad base of editors have the opportunity to be informed on policy issues. --Kevin Murray 01:56, 5 September 2007 (UTC)
Define "broad base"? How broad is broad? How do you propose to convert the time and energy these people spend into useful work? I never want to see another GNAA AFD, or see things like the ATT ever again, if at all possible, thanks :-) Those were utterly unmanageable. --Kim Bruning 03:03, 5 September 2007 (UTC)
I think that the statement below by Wikidemo (17:59, 5 September 2007 (UTC)) very ably identifies the problem. Broad based is a realtive term to the effect of the page in question. As we are a society of writers, I think our writers should determine the course of the project, not a small band of policy gurus. --Kevin Murray 20:44, 5 September 2007 (UTC)

removed section, several issues

Removed: " However this process is only valid if a significant number of editors are made aware of the changes; unilateral or minority changes which go unnoticed do not specifically represent consensus."

only if a significant number of editors are made aware of the changes
  • Problems with WP:POLL, as this practically forces superfluous polling
  • This is asking for scaling issues, wrt use of the wiki. Do NOT invite large numbers of editors to a page, unless you are in fact trying to find the actual value of Dunbar's number for wikis :-P
  • It is assumed that people with an actual interest in changes will have watchlisted a page. (see Bold Revert Discuss for details.
  • People who have no actual interest may be forced to make decisions at the drop of the hat. This is not actually consensus. (see the Attribution blowup.)
  • Mediation and negotiation is best done with only one or two people at once, never encourage pileups. Wiki based consensus is difficult or impossible when there are large pileups. So Don't Do That.
  • Normally most pages are left alone by most people. It might be wise to leave it that way.
unilateral
  • First use of the word "unilateral" in guidelines, afaik. So called "unilateralism" is specifically permitted through Be Bold, and IS permitted, because that's the way you edit on a wiki, remember? : Edit the wiki, others will correct and improve.
minority changes
  • Are exactly what we are looking for!
which go unnoticed
  • By whom, by which measure? This is not enforceable.
Do not specifically represent consensus
  • Once again, this is untestable in practice. Don't go making one up either. There's a test in the wording just preceding this section, just use that already!

--Kim Bruning 00:27, 5 September 2007 (UTC)

  • This list of objections is idealistic but not practical. Bringing notice to a broad base of editors does not constitute polling. Limiting the discussion or analysis to those with special interests (e.g., watchlisted) encourages special interest groups to gain control of segments of WP. Yes, we edit unilaterally, but we build consensus as a group. --Kevin Murray 00:39, 5 September 2007 (UTC)
For precedents look at the recent reforms to the notability infrastructure where broad discussions brought excellent results. I am less concerned about the articles and perhaps we don't want to solicit unproductive attention to many areas, but some areas have benefited tremendously by attracting broad attention to dislodge special interest control. --Kevin Murray 00:46, 5 September 2007 (UTC)
Since articles are our bread and butter, it might be a good idea to use this working implementation[1] us wide eyed idealists came up with ;-)
--Kim Bruning 01:02, 5 September 2007 (UTC)
[1] 2 million pages and counting! \o/
*The "discussions" about our "notability infrastructure" started out sane, but devolved to pure power-politics. Many were disenfranchised. I was not pleased.


Whoops, maybe I'm a bit quick at reverting though. I should really be asking Kevin where he's headed? --Kim Bruning 02:56, 5 September 2007 (UTC)

I agree that the passage seems at tension with BRD, but a milder form of this statement might have a place in a policy. There are a lot of backwater articles on wikipedia. The number of unwatched pages is enormous, and users simply can't rest on the longevity of their edits as "consensus" if there is no true examination of the edits. I think the gold nugget in this sentence is "changes which go unnoticed"; this isn't a hard measure, but a rule of thumb like the entire topic of consensus. The other language about "unilateral or minority changes" is a a timebomb. Cool Hand Luke 03:14, 5 September 2007 (UTC)

Changes which go unnoticed might suddenly change quite considerably, once they *are* noticed, indeed. Hmmm! :-) --Kim Bruning 03:27, 5 September 2007 (UTC)
We ought to put that statement back in, and more, with respect to policy and guideline pages. Unilateralism is a huge problem there. Small bands of people are pronouncing new rules, sometimes on obscure pages, to the surprise of everyone else on Wikipedia, then going out and enforcing them. The assumption that people who care about an issue will know about and participate in the policy debate is not a good one. There are far too many policy and guideline pages, and they're too poorly organized, for even the most enthusiastic policy lover to get a handle on everything. It creates a bureaucratic class that operates without oversight, often through aggressiveness and stubbornness. Moreover, policy page edits are often done in an unpleasant way that approaches incivility. This is particularly daunting to those who don't spend their entire lives hanging out on policy talk pages. For that reason, major policy changes that affect other areas of policy, wikiprojects, or large numbers of articles, should be advertised and any change considered tentative until it's clear it's truly accepted. If twelve editors agree to prohibit something in main space, and a hundred thousand editors writing a hundred thousand articles have that something, have those twelve editors truly created a consensus on Wikipedia? I think not.
I also don't think "you snooze you lose" (which is what it really means) or you claim the sandbox, you make the rules for the sandbox, are valid rule-making principles. It's a description de-facto of how rules are made, but that doesn't make it right. We might as well introduce other descriptions, like "contentious editing that scares everyone off the page = consensus".Wikidemo 17:59, 5 September 2007 (UTC)
This would be assuming that the "proposed" process actually works, which it doesn't. If you see people doing this, just feel free to mark such a page {{rejected}}, to revert their edits if the edits were to an established page, or if they are bullying you, to tell them you don't think they have consensus and would they please go away. (if they indeed do not have consensus, your co-editors will take a similar stance, and the bullies won't have a leg to stand on).
Be careful to listen if people point out that they are documenting common practice though! In the latter case, listen first, and go out and check to see if they are indeed correct (if you didn't already intuitively know that upfront, which might well be possible!).
Even so, perhaps people haven't done that enough lately, can you give some examples where we might need to check people's behavior? --Kim Bruning 05:50, 9 September 2007 (UTC) That and we should have more people on policy patrol :-)
I was making a general observation. Although I've seen plenty of these things, it would be unfair of me to call attention here to people I think have not acted reasonably. One flare-up I haven't been deeply involved in yet is WP:OR. Both sides accuse the other of acting without consensus and there seems to be much aggressive editing going on (e.g. fighting over protecting and unprotecting a page, and which is the "right version"). It makes me wonder whether one should have to deal with bare-knuckled policy fighting just to try to contribute to Wikipedia policy. Wikidemo 06:41, 9 September 2007 (UTC)
I suspect that Original Research, Verifiability and Reliable Sources are indeed being WP:OWNed by a number of people who enjoy playing (what I call) wikinomic. I'm not pointing to any one person (mainly because I haven't checked who's busy there at the moment), and I'm not saying all (or even many) play the game consciously. But still.
One reason I'm so careful around the Ignore All Rules and consensus policies is that I want to be careful. If you are out to defang the other guy, it's way too easy to also end up defanging yourself (oops!)
I am always very very interested in names by the way. If you have issues with anyone's behavior, you should talk with them. This does help, possibly they didn't even realize they were causing you difficulties. If you don't feel up to that right now, please e-mail me, and let's see what we can do. --Kim Bruning 03:06, 11 September 2007 (UTC)

Across pages

Hello, I was wondering if there is a general guideline or principle regarding statements like:

if it's added/removed in this article, it should be added/removed in the other article.

For example, the case currently is that in the United Kingdom article they are including some information about the the personal union between countries specifically via the Queen, and in the Canada article, there is a discussion if that info is better served in a daughter article. I'm not sure, but I think I remember that consensus usually happens per page, and if the editors on one page don't think it needs to be in one page, the effects of it being in another page don't really matter that much. Any thoughts would be much appreciated. Regards, -- Jeff3000 18:53, 9 September 2007 (UTC)

I would post the issue at both pages and try to get a consensus among all of the editors. WP is one project, not a confederacy of individual page clubs (theoretically). --Kevin Murray 21:36, 9 September 2007 (UTC)
A mixed bag, I think. Sometimes the individual article pages are laboratories for deciding when something applies and when it doesn't. Sometimes the result isn't consistent or unanimous. As long as no great damage is done it's useful to allow different outcomes, and see where things are headed. The outcome of one consensus process is instructive but not binding (as a lawsuit outcome might be binding) for future cases. If it's truly a broad issue, or something important where the two cases are truly comparable, I agree that it's best to alert people on the other page and reach consensus in tandem. If it's a very widespread issue it is best discussed on policy pages and notice boards to avoid needless duplication of efforts, problems having to do with notice, and inconsistent results. For example, there were recent deletion debates about whether there should be national origin-based list articles (list of German-Americans, list of Lebanese-americans, etc). In individual debates some lists were kept and some deleted, seemingly without a consistent pattern. That kind of thing needs to be centralized and discussed in one place.Wikidemo 00:55, 21 September 2007 (UTC)

The flow chart says I should take "it" here

Of course, the flowchart doesn't represent consensus so I don't have to follow it but....

For whatever good it does, the flowchart gives the false impression that it's the only way to achieve consensus. It should be noted as a representative or possible consensus process, not the only one. There are others.

For example, it starts out suggesting that the way to reach consensus starts with making an edit. However, it's a lot more complex than that. Often it starts on the talk page or by reviewing past consensus discussions. It's often a bad idea to make a change that one knows is contentious, a major shift, re-opens an old debate, or alters a recently-reached consensus, without first offering to talk about it. When you do talk about it, the article talk page is normally the place but not always. Issues that are particularly troublesome, or of broad applicability, get talked about on policy and guideline talk pages, and notice boards.

It's not terribly reflective of the consensus process for policy and guideline pages, where consensus is less a matter of reaching agreement on what the page should say, but an effort to gauge what represents the community norm.

There are plenty of other counterexamples.

So, I have nothing against the flow chart. It simply should not claim to be the single universal process for consensus. We should note that somewhere, best in the image caption.

-- Wikidemo 00:49, 21 September 2007 (UTC)

I think the flow chart is great and that you are making things too difficult. I don’t think a reasonable interpretation is that it claims to be “the single universal process”. Your concerns are largely covered by the section “Note on use of discussion page”. --SmokeyJoe 06:09, 22 September 2007 (UTC)


Actually, it is a very good idea to edit directly and WP:BOLDly on a page. That's what wikis are for. It is an especially good idea for contentious changes, major shifts, reopening-s of old debates and exceptionally good for showing that their recently reached "consensus" really wasn't.
The huge temptation of talk pages is to go there first, thereby losing huge amounts of tempo and wasting massive amounts of valuable volunteer time. Neither of which would be lost if you simply assume everyone is editing in good faith in the first place.
Only *after* normal wiki-editing breaks down should you actually go to the talk page.
Some people enjoy holding running conversations on talk pages or IRC, I have no problem with that either, as long as wikiediting is the norm.
Policy and guideline process is heavily abused. In general, typical wiki-editing is actually pretty decent to maintain best practices documents, but there is a stronger temptation to subvert the process. :-/
See also: WP:BRD.
--Kim Bruning 23:26, 23 September 2007 (UTC) A wise and careful person may also wish to check previous history and discussion about a page, to prevent history repeating itself, but this is not currently seen as a necessity.

Note on use of the discussion page

This needs working on, and maybe integrating into the main consensus text.

Hmm, people seem to think that sapping tempo from wiki-editing is a good idea. I'm not entirely sure why, because every time it has happened in my experience, a page gets mired and people get sick and tired and walk off. Tempo is what makes wiki-editing fun. :-) And if it isn't fun, what other reason is there to edit? --Kim Bruning 23:43, 23 September 2007 (UTC)

I'm not sure what "sapping tempo" is all about. The thought behind this section and Kim Bruning's added text is good but I found it overly chatty. This page already has a lot of exhortations. So I shortened it to be more to the point, avoiding SHOUTING ABOUT BEING BOLD (which is not despite the prior text the primary rule of editing, but one among quite a few others), etc. Wikidemo 22:40, 26 September 2007 (UTC)
\o/ Thank you for tidying up the note a bit. :-)
Oh and ah...
Rules of editing specifically?
Didn't they go something like this?
  1. You are editing the encyclopedia
  2. So you should go actually edit the encyclopedia!
  3. So go edit already!
  4. What are you waiting for? it's a Wiki!
  5. You're still here? Seriously, don't worry, you can't break anything.
  6. Alright, you want me to tell you to play nice with others? Go play nice with others.
  7. And yes, you should probably actually use decent sources, and not make a mess, and all that... but isn't that obvious?
  8. Oh god, you're not going to hang around and quote that crap back at me all day are you? I only just made that all up 'cause you wanted some more guidelines.
  9. No really, lalalalalalala I can't hear you!
  10. So like, go have fun already! :-)
It's an as decent a summary of what we're doing as any other, if you ask me O:-)
Wikiediting is typically only fun when people actually do wikiediting. When folks first have to sit down and fight out tiny details and so forth, things quickly get mired (you lose tempo) and you often end up needing a mediator to pull things back out of the swamp. :-/ --Kim Bruning 14:04, 27 September 2007 (UTC)

Worth reading

The New York Times has today an article on how scientists came to a consensus that low fat diets were healthy, while getting the science completely wrong. It is worthwhile for Wikipedians to consider the thinking in the article Points it makes are that 1) whatever position is taken first is more likely to be the outcome than if a different position was taken first, 2) social pressures on dissenters don't help get the right answer, and 3) if a group is given data leading to the right/wrong outcome in a 60/40 ratio, there is still a 1/3 chance of the group coming to the wrong decision. All of us should take care with regard to these issues. GRBerry 14:28, 9 October 2007 (UTC)

How do we assess if consensus is achieved?

I'm not being facetious here - when reading edit disputes, I often see these mandarin-like opaque admonitions that we must "seek consensus" but "wikipedia is not a democracy" and we should not be "voting". What metric do we use to assess consensus if we are not supposed to count the frequency of expressed opinions which agree with each other? --feline1 15:28, 22 October 2007 (UTC)

This is a really good question. One problem that I see with "voting" is what I call drive-by votes where people unfamiliar with the issue stumble in to express opinions without an indepth understanding or to promote a narrow agenda. Unfortunately when you remove the quantitative assessement, evaluating consensus becomes subjective and unlcear. Sometime it becomes who can shout the loudest or the longest. For some reason though the system seems to work fairly well over time, although not at every moment. --Kevin Murray 16:53, 22 October 2007 (UTC)
Yes - so basically, "voting" is "bad" because: people have to be presented with a finite set of "either/or" choices, when the reality may be more complex; also, a pure Yes or No vote says nothing about the competance of the voter to make the decision...... but I do not see how recognising these flaws in "democracy" helps us assess objectively when "consensus" has been reached. Clearly, in effect, expressed views ARE being counted and "weighted" in some way so that those deemed more "valid" count for more...?--feline1 16:58, 22 October 2007 (UTC)
Clearly. It seems to be a combination of weight of numbers, strength of argument, and consistency with other trends at WP. An abstract evaluation of consensus is to try to assess the intent of those who have not expressed specific opinions, but indicate their preferences through their actions particularly in the policy areas. --Kevin Murray 17:11, 22 October 2007 (UTC)
I think voting is a good "last resort" to break deadlock and edit wars, when everyone's been over each others' positions and no-one has anything new to say. Indeed I've sometimes had votes go against my position and found it actually reduced my "wikistress", as it meant the issue could be closed and everyone could move on. —Ashley Y 09:28, 12 November 2007 (UTC)

RTFS Read The Fine Schematic. You know consensus is achieved (at least for now) when no one wants to make further changes. :-) --Kim Bruning (talk) 09:42, 13 December 2007 (UTC)

Primacy

Purely as an observation, I note that this project page including its "Exceptions" section has a sort of "constitutional" or "Basic Law" status, in that all other policy (yes, even WP:IAR) defers to it. —Ashley Y 02:27, 11 November 2007 (UTC)

Indeed a successful application of Ignore all rules depends upon the existence of consensus that the application is appropriate. --Tony Sidaway 06:17, 13 November 2007 (UTC)
People often construe policy poorly based on the overly simplistic (and generally fallacious, if it comes to the point of using it in argument) maxim that "policy trumps guideline." Expressing one's position with reference to primary policy does not create a sword by which all other arguments may be stricken. Guidelines and non-primary policies clarify, interpret, give rules for applying and enforcing, and otherwise refine primary policy so in that sense they modify policy. Some useful cannons of interpretation:
  • The specific overrules the general
  • Last in time prevails
  • If there are two possible reasonable interpretations of a rule, one interpretation by which it or another rule is invalid, and the other interpretation by which both rules are valid, the proper interpretation is the one by which both rules are valid.
  • New rules should be invalidated as being in conflict with older rules only if there is no interpretation by which they are in conformity, the older rules have primacy, and the enactment of the new rule is not deemed an intentional, valid repeal of the old rule.Wikidemo 07:00, 13 November 2007 (UTC)

Opening sentence

I got a chuckle when I saw the edits on the opening sentence- a problem with consensus over the article Consensus! My take is that the sentence is very strong, and it reflects the overall nature of Wikipedia. I fully agree that it should stay, and I welcome others to make their comments on it. Monsieurdl 14:32, 12 November 2007 (UTC)

"Wikipedia works fundamentally by building consensus."

Unfortunately, it is not true. It is too strong as a statement of fact. The fundamentals of wikipedia are really the technical aspects, and why it works fundamentally has to do with human psychology and why editors choose to devote time contributing. There are plenty of things about it that work but are not "consensus". --SmokeyJoe 21:06, 12 November 2007 (UTC)
Consensus is fundamental to the social process of creating the encyclopaedia, though. —Ashley Y 22:41, 12 November 2007 (UTC)
Creating an encyclopedia does not require a "social process". The vast majority of articles are created by people who have no interaction with anyone else. Even those situations which do have interaction, do not require "building" any consensus, nor do they require the entity defined as "consensus" on this page. For most articles its editors have the same interests in mind for writing a factual article. Wikipedia works fundamentally because it is openly editable; Wikipedia works fundamentally because its articles are supposed to be neutral; Wikipedia does not work fundamentally by reason of the "consensus" on this page that this sentence introduces. "Consensus is an inherent part of the wiki process." is a much less problematic sentence, which is redundant with the first sentence at its most reasonable interpretation. —Centrxtalk • 03:18, 13 November 2007 (UTC)
Well, you could create an encyclopaedia all by yourself if you like. But creating this encyclopaedia very much is a social process, and consensus is fundamental to it. —Ashley Y 04:14, 13 November 2007 (UTC)
Consensus is not fundamental, it is a goal, or at least most contributors seem to agree that it should be. The building of wikipedia involves many situations of consensus building. I think that more energy for the building comes from the drama of conflict rather than any pre-existing, fundamental consensus. Whether or not a social process is required (I think it is important but not required, single authored pages exist) is irrelevant to the point that the opening statement is false. --SmokeyJoe 06:13, 13 November 2007 (UTC)
The opening statement is correct in context. At one level, Wikipedia works fundamentally by responding to HTTP requests. At another level, Wikipedia works fundamentally by allowing anyone to edit any page, and keeping track of the changes. And at this level, Wikipedia works fundamentally by consensus. —Ashley Y 06:35, 13 November 2007 (UTC)
¡Ay, caramba! The word "fundamentally" neither strengthens nor weakens the sentence. It just makes it a word longer. It's a filler word (in its absence the phrase is "works by" which impliedly claims primacy anyway) and if there is any meaning beyond filler that meaning is a little ambiguous and not terribly enlightening. Policy pages are worth getting right of course, but if anyone is actually getting wikistressed by this, a deep breath and maybe a nap. Wikidemo 06:41, 13 November 2007 (UTC)
Without HTTP requests, there would be no Wikipedia website. Without a wiki, Wikipedia would never have been created. Without "consensus", things work just fine. "Free", "Wiki", and "Encyclopedia" are fundamental; consensus is just a mechanism for producing parts of the encyclopedia. —Centrxtalk • 04:49, 15 November 2007 (UTC)
Without consensus, Wikipedia would be an endless stream of edit wars, and no progress would be made. We would, for instance, not have usable policies, since they also depend on consensus. Some articles would surely exist, but we would not have anything that could be called an encyclopedia.
I'm fine without the word "fundamentally", but the first sentence says something very central to the project. —Ashley Y 05:12, 15 November 2007 (UTC)

Contradiction... again

Somebody in an earlier post said that this rule conflicted w/ others. And now, I agree with him.

Heres the idea. A consensus is a majority opinion, right? Opinions arent allowed in articles. right? Im debating an anonymous user over at the arena rock article, and he says that consensus (majority opinion) can be used when sources arent present. But that conflicts directly with WP:OR, which disallows opinions to be inserted.

So, simply, this rule conflicts with WP:OR. Prepare to be Mezmerized! :D 19:40, 15 November 2007 (UTC)

Consensus is something more subtle than simple majority opinion. It's the prevailing considered opinion agreed to among the participants, or something like that. The existence of a consensus process may be in tension to rules, necessarily so. Rules shape consensus and conensus shapes rules. But the consensus policy itself does not conflict with the other policies. There are subjects where consensus does not apply, certain "non-negotiable" things like OR. Where people try to apply consensus when it doesn't belong, that doesn't mean anything is wrong with the consensus policy, it means those people in that instance may be failing to consider the OR policy. Content may be inserted with or without a source. However, if a specific piece of content is challenged as possibly inaccurate the burden is on the person inserting (or proposing to keep) the content to provide sources. If they cannot it is fair to delete the content.
I think all of that misses the point with the Arena rock article. The whole thing isn't properly sourced. Most of the citations are done (in the wrong format, I should add) to a single blog. A blog is not an acceptable source for this kind of thing. If the fact is worth mentioning it's worth citing to a proper source. The terms "Arena rock" and "Stadium rock" are widely used, more so in the 80s and 90s than today, but you should be able to find thousands of sources for this. Millions of google hits. Try Rolling Stone or some of the other music-oriented magazines, newspaper music critics, etc. I wouldn't start deleting content over it, but simply insisting on better sources (because it is sourceable) and possibly modifying the article as you learn what these sources say. One thing you might find is that "anthem rock" is not synonymous because it refers to a type of power ballad - things are done in arena rock that are not anthems, and not all anthems are done in arenas. You'll also find there are three different subjects, and have to decide which to cover, or perhaps cover all three: 1) the word as a neologism that exists apart from the rock styles involved; 2) a style of music; and 3) a phenomenon of performance, promotion, and consumption of music that is different than the underlying musical style (big hair heavy metal works just fine in a small theater, it does not have to be done in a stadium). Anyway, you might find that all in the sources.
Hope this helps. Wikidemo 21:33, 15 November 2007 (UTC)
A quick follow up. Here is one source, a BBC program on stadium rock [10]. The "director's notes" section on the page is only semi-reliable. He is just writing his summary opinion. But the content of the proram itself is probably more reliable. Anyway, that's the kind of source you need to look for, not a blog. And note, they are clearly saying Led Zeppelin is stadium rock, so you can use the BBC rather than consensus or a blog as a source to verify that claim. Wikidemo 21:38, 15 November 2007 (UTC)

"A consensus is a majority opinion, right?"
Wrong.
Consensus is a group acquiesced position. It can be like a compromise, but not necessarily. On a well defined question, there may be no possible consensus. Consensus finding typically requires a refinement of the question. In the end, a consensus position may require acknowledging multiple divergent opinions. It’s all a little fuzzy, but in the end, a majority does not define consensus. See the failed Wikipedia:Supermajority and its discussion. --SmokeyJoe 00:28, 16 November 2007 (UTC)

That's a good point. A consensus means most everybody agrees - not necessarily because they approve or think that's the very best outcome, but they all reach an understanding. It's like if two people love apples and hate oranges, and one person hates apples and loves oranges, but they can all agree on a banana, the banana is the consensus. Orange would be majority rule. And if oranges and bananas are both illegal, apple would be the only choice and consensus would not matter. Wikidemo 03:04, 16 November 2007 (UTC)

I generally agree, but it does leave a question: how many people does it take to prevent a consensus? —Ashley Y 04:00, 16 November 2007 (UTC)

Saying consensus is the way you make decisions, but saying it is "fuzzy" and "subtle" might just be a way of justifying it's usage. Also, it provides a convenient excuse for not giving a proportion for failed consensus. There really should be a category for decisions that failed to reach a consensus, so people can further work on them. Otherwise, as I mention below, apathy by the opposition can be confused with consensus. I myself have succumbed to apathy. The great horde of the Wikipedia believers prevents me from engaging in any continuing opposition, because I have better things to do in life. --Fandyllic (talk) 1:22 PM PST 6 Dec 2007

Tracking consensus decisions on Wikipedia

Does Wikipedia track situations where consensus was or was not achieved? I've seen quite a few decisions made by admins that do not seem to honor consensus (trying to negotiate agreement with parties of minority viewpoints to abide by the outcome), but rather snap decision making with the apparent hope that apathy will allow the decision to appear as a consensus decision. --Fandyllic (talk) 1:14 PM PST 6 Dec 2007 —Preceding comment was added at 21:15, 6 December 2007 (UTC)

Gotta love SineBot. —Preceding unsigned comment added by Fandyllic (talkcontribs) 21:25, 6 December 2007 (UTC)
Actually the latter is correct, for wikipedia-style consensus. Basically you determine whether or not you have consensus by whether anyone disagrees with you. Since everything can be reverted, the fastest way to discover this is to simply make the change, and see if anyone reverts you. If they do, you know you should talk with them first.
The reason you've seen admins do this more often than other people is because they are more experienced with the system, and are therefore more likely to apply it in a quick and efficient manner.
If you disagree with someone, just tell them, and if all is well they'll then gladly talk with you and discuss the changes they made. --Kim Bruning (talk) 09:47, 13 December 2007 (UTC)
Interesting comments, but I don't see how they answer my question. From what I can glean, the answer to my question is: No, Wikipedia does not track situations where consensus was or was not achieved. The fact that people talk about lack of consensus means nothing in terms of achieving consensus or even caring if consensus is achieved on a consistent basis. The stance of many folks on Wikipedia appears to be more about preserving the appearance of consensus than actually achieving it. --Fandyllic (talk) 12:46 AM PST 23 Dec 2007

re: sudden edit war

For the record, I definitely believe COGDEN's edit were appropriate. Although I can see where "steadfast policy" advocates can dispute the "BOLD edit" bit (the rest of the edit being very much a mere wording adjustments that affects neither the letter, and certainly not the spirit of this or WP:BOLD), it happens all the time, despite what they would prefer. This basically reminds me of a dictionary review by Michael Quinion: "Many people say they want their dictionaries to be authoritative, when what they really mean is that they want them to be conservative, most commonly reflecting the language that was taught to them in their youth, the language that they feel comfortable with." It seems some user are not uncomfortable with these specific changes to policy, but rather with any, even perceived alteration to policy as it was understood at the time they internalized it. Of course, this is edit warring on a policy page, and is still unacceptable regardless of what is going on philosophically. Circeus (talk) 06:14, 15 December 2007 (UTC)

I disagree about COGDEN's edits. A bold edit is by definition a singular event by a single user that does not have anything at all to do with consensus, especially on a policy. The wording COGDEN added to attempt to address this issue crosses the line into formal Wikilawyering, and is totally unacceptable. Dreadstar 07:18, 15 December 2007 (UTC)
Actually, you've just completely ignored what I just said said, i.e. it happens all the time. I'm fairly sure if I look at any policy or guideline page over the last year, I can dig edits that were never really discussed, but never contested either. Circeus (talk) 07:25, 15 December 2007 (UTC)
No, I didn't ignore it - sorry if it looked that way - I read what you said, but I just think it's irrelevant to "consensus" that someone makes a bold edit that stays. That's not the definition of "consensus". Sure, if other editors read that edit and agree "silently" then it can be an unspoken sort of consensus, but I think it is something that is not necessary to codify in this policy, it would be confusing and easily misused. Editors used to add unsourced critical content to biographies of living persons, but just because it was 'happening all the time' doesn't mean it's the right thing to do. That's an extreme example, but it proves the point that just because something is being done, that it should included in a policy or guideline. Dreadstar 07:48, 15 December 2007 (UTC)
Have you looked at the page? I'll quote from the first paragraph: "[...] generally someone makes a change or addition to a page, and then everyone who reads the page has an opportunity to either leave the page as it is or change it. In essence silence implies consent if there is adequate exposure to the community." While changing a point of policy should involve discussion, especially where practice is unclear, an undiscussed changes that is never challenged (usually a minor one, and especially one that comes to be actually quoted) is effectively consensus until someone demonstrate that it isn't (cf. the first section). As such, my assertion is that the edit most certainly reflected current practice as described on the very page. Circeus (talk) 08:08, 15 December 2007 (UTC)
Yes, I've "looked" at the page. The quote you provide is different than the one being disputed, they are two similiar but different statements. The statement that you partially quoted has a larger context to it, and includes some important detail "In essence silence implies consent if there is adequate exposure to the community. In the case of policy pages a higher standard of participation and consensus is expected." The context and detail of the disputed edit is narrower and does not supply adequate detail or context..and adding that detail and context to what is essentally unnecessary to begin with, seems to be wikilawyering to present a certain POV. I dunno, perhaps I'm being a little too nitpicky, but it's understandable in light of some recent events. I would like to see the opinons of others, including SlimVirgin, who also disputed the change in wording. Dreadstar 08:34, 15 December 2007 (UTC)

Eh? Didn't we already cover this? --Kim Bruning (talk) 09:06, 15 December 2007 (UTC)

Did we? What did we decide? Dreadstar 09:17, 15 December 2007 (UTC)

The real issue

For those who wonder what is going on, a dispute at WT:NOR has boiled over and splashed here. Dreadstar is one of a group of editors there who assert that "longstanding" policy cannot be changed without demonstrating a broad consensus for change. Since only a small number of editors are actively involved in a talk page discussion at any given time, by this logic the small number of objectors can easily be dismissed as inconsequential compared to the larger but silent community. So in effect, this false premise is used to subvert WP:CCC and maintain the status quo. The reason this is a false premise is the number of defenders of the status quo is also small compared to the community, but they rely on a fundamental misapplication of the principle of silence=consensus. Once a non-trivial number of editors break the silence with cogent arguments given in good-faith, consensus is no longer implied.

I have challenged this position several times there, most recently in this thread. There is strong consensus for using extra caution in editing policy pages, and making extensive use of the talk page to reach consensus, because policy pages do need some degree of stability. But this extreme application of those ideals just creates an impossible threshold for change, causing stagnation, which is contrary to the letter and spirit of this policy. Dhaluza (talk) 17:13, 15 December 2007 (UTC)

That's a distortion of what I've stated, Dhaluza. What I have maintained all along, is that in cases where long-standing policy content has recieved prior consensus, it requires new consensus to change that content (or some act by those in authority, such as Jimbo). A few editors have come along and disputed the long-standing consensus for WP:PSTS, and have been unable to find consensus for their changes. Consensus and inertia are important to maintaining strong policy, else chaos will reign with policy undergoing significant changes from day to day, if not more frequently. That is my position. The issue has only "spilled" here because COGDEN made edits that he subsequently appears to use to prevail in disputes he was already involved in other places. Dreadstar 17:28, 15 December 2007 (UTC)
It wasn't intended as a distortion, and I still think it accurately represents the whole of what you are describing. I agree that editing policy, and then quoting your recently edited version (without disclosing your involvement) is not good practice. But regardless of one editor's alleged behavior, consensus can change, and one group in a dispute cannot claim the silence of the masses represents consensus for their position, regardless of the status quo. Stability in policy is desirable, but stagnation is not. First, it is not clear there is a "long-standing" consensus for the case you cite, because there have been continuous changes over time, some substantial, and not all those changes were discussed while they were happening. So if the full effect of an edit to a page (of any type including policies) is not noticed until a later time, questions raised at any time are equally valid. It is completely against many policies and principles, including this one, to suggest that time trumps substance, and the legitimate concerns of the community can be ignored once the subjective threshold of "long standing" has been reached. And hysterical predictions of the demise of Wikipedia if "long standing" elements of policy descriptions are changed is just nonsense. Quoting from this policy:
A small group of editors can reach a consensual decision, but when the article gains wider attention, others may then disagree. The original group should not block further change on grounds that they already have made a decision.
Now, by way of full disclosure, I should point out that I edited that sentence, but that was almost a year ago, so now it's "long standing" ;-) Dhaluza (talk) 22:19, 15 December 2007 (UTC)
I can agree, the original group should not block changes that are made with new consensus. I do disagree that a few editors can dispute and change long-standing content that was originally put into a policy by that prior consensus (without having reached consensus for the changes they propose). The distortion I mention comes from leaving out the bits about 'prior conensus'. Without prior consensus, then yes, disputed content should removed until consensus is found. It may be a simple, small difference in wording, but it has a big impact when left out. Dreadstar 09:08, 16 December 2007 (UTC)
A few editors can and do. Policy is descriptive, not prescriptive. Our personal preferences are somewhat irrelevant. --Kim Bruning (talk) 11:12, 16 December 2007 (UTC)
Sorry, I left out the number of editors opposing those "few editors" who are trying to make what are non-consensus changes to content that was put into place by prior consensus. Dreadstar 10:18, 17 December 2007 (UTC)
And that is where your approach runs afoul of this policy. It is not necessary to demonstrate a new consensus for change; demanding this as a treshold is blocking further change. A few editors disputing policy in good faith, whether long standing or not, indicates that upon gaining wider attention any prior assumption of consensus was either not valid, or too narrow. If the group of objectors is indeed small, then as the issue gains even wider attention, this will become obvious. But if the number of objectors grows with the exposure, then what may appear to be long standing consensus is not really a consensus at all, and all parties must work to form a new consensus.
Now surely you can appreciate this irony: Since you obviously disagree with this "long standing" policy, you may want to change it. But since at least one other editor and myself disagree with you, then by your own method we could demand that you demonstrate a consensus to change this policy, then refuse to join in such a consensus, thus blocking any change. Dhaluza (talk) 15:19, 16 December 2007 (UTC)
(crossover) Hey, that's interesting. If several people are actually following that flawed procedure, would that be what's causing Vassyana so many headaches, as described here: Wikipedia_talk:Disruptive_editing#Blocking_consensus ? --Kim Bruning (talk) 06:27, 17 December 2007 (UTC) Perhaps the behavior that is perceived as stonewalling is actually unintentional and in good faith?
Kim, I don't think it's flawed, I just didn't describe the situation completely - this very same discussion has been going on for months and months and months, so I've shorthanded some of the details. Let me know if the below is flawed...I tried to add in all the details to try and clarify what I'm trying to get across....;) Dreadstar 10:52, 17 December 2007 (UTC)
Dhaluza, again, distortion. In the case we are talking about, (WP:NOR),
  • It is not: "It is necessary to demonstrate a new consensus for change";
  • it is instead: "It is necessary to demonstrate a new consensus to make a disputed change".
A single editor or even a small group of editors cannot just perform a disputed override to content on a policy page that has long-standing, previous consensus and make whatever changes they want. If there is a dispute, then new consensus must be found...if this isn't followed in contentious circumstances like this, then all will be chaos...
And no, there's no irony here. Your last statement is almost correct. I say "almost" mainly because I'm not sure exactly what change you're referring to. Dreadstar 07:02, 17 December 2007 (UTC)

Ok, let me try to make my postion clearer:

"The original group should not block further change on grounds that they already have made a decision."

This statement is true, the prior decision all by itself is meaningless in the face of newly established consensus. But my statement is that the original group as a part of the current group that's trying to achieve consensus can disagree with the proposed changes, and if that disagreement causes there to be no consensus for the changes, then the content that is currently in place by the prior consensus stays in place until a new consensus is found to change it. The new changes being proposed by the 'small group of editors" is being disputed by a significant number of editors, so there is no conensus for the changes. That is the situtation that I've been trying to convey, and the one my comments above are all based on.

This is regarding disputed content on Policy pages, but the proposed changes are also disputed. The current content had prior conensus and has been long-standing policy. The new, non-consensus content that is also being disputed can't just replace the current content...there is no consensus to do so.

Certainly, a small group of unopposed editors can come in and change content they dispute, even if it's long-standing policy put into place by consensus...but that's not the situation I'm talking about. I have never believed that there can't be changes because there was prior consensus, or that a small group of editors can't change prior-consensus policy; and I hope I've clarified that with the above. Dreadstar 10:28, 17 December 2007 (UTC)

First, let's be careful to stay on-topic here--we are only discussing the general principles of consensus, not the specific case at WT:NOR (which should stay there). The problem with the case you have laid out is the faulty assumption that the status-quo at any point in time necessarily represents an actual consensus at some prior time. This is further compounded by the corrolary assumption that when consensus is lost, the status quo must stand until new consensus is formed. The problem with these assumptions is it creates a disinsentive for the defenders of the status quo to work to form a new consensus. This policy wisely balances the two extremes by saying we do not ignore precedent, but we are not bound by it either. Your position is that assumed prior consensus is binding, unti a new consensus is formed. But since there would be no reason for people who have internalized the old version to take a risk, their loss aversion will naturally make them unwilling to accept that a new consensus could be better for all involved. So if we followed this path, policies would become stagnent and progress would be effectively thwarted. Dhaluza (talk) 13:17, 17 December 2007 (UTC)
We are fully on-topic here. I only mention NOR becuase you created this subsection to those "who wonder" what is going on, and gave history on the NOR dispute. The specifics of my statements are directly related to the dispute over WP:NOR, but can apply as general principles for all policies. Again you misquote me. You continue to distort what I have clearly stated, this is not correct: "you have laid out is the faulty assumption that the status-quo at any point in time necessarily represents an actual consensus at some prior time"; nor is this statement correct "Your position is that assumed prior consensus is binding, unti a new consensus is formed.". In the NOR dispute there was clear and undeniable prior consensus for the long-standing PSTS policy content. I did not say that prior consensus is binding until a new consensus is formed, my statement included additional issues such as when the proposed changes are disputed and no consensus is found for those changes, non-consensus changes that are meant to replace content that has been long-standing policy that was put into place by prior consensus. I have stated this clearly, over and over again. It is impossible for anyone to have a discussion with a party that is continually distorting the comments of others, the history of the issue, etc. So I'll just have to leave off replying to you for now, Dhaluza. Dreadstar 19:57, 17 December 2007 (UTC)
Dreadstar, you continue to assert that I am engaging in distortion, a loaded word ill suited for a civil discussion, even after I have assured you that that is not my intention. I ignored this when used as a noun, because under WP:AGF it was possible that you were only referring to the result, not the intent. But when used as a verb, especially in the context above, I can not see any non-judgmental way to read this. I can accept that my interpretation differs from your intended point, but that does not mean I am deliberately misrepresenting you. It would help if you tried to be more open to other perspectives, and mindful that any ambiguity in interpretation is the responsibility of the writer, not the reader.
Notwithstanding the form of your augments, I also take issue with the substance. I deliberately distanced my comments from the WT:NOR discussion, because I don't want to bring that dispute here; but you reconnected it anyway, putting my comments out of context. Status quo does not necessarily equal consensus, and it can be difficult to determine what the prior consensus was on a constantly changing wiki. Consensus can be lost gradually, and absent a detailed process such as a formal draft that was widely commented on and approved in a specific form, there is no way to show "clear and undeniable prior consensus" for a particular version of a document that has changed continuously over time. The wiki model works on assumed consensus, but that assumption has limits.
Also, I see that you are trying to draw a distinction between "consensus for change" and "consensus to make a disputed change", but I fail to see the difference. Either way, blocking changes based on prior consensus is specifically disallowed by the WP:CCC passage quoted above. You are correct that a new consensus must be formed when there is a dispute, but this requires all parties to work toward that consensus. Is is not permissible for one party to the discussion to demand the other complete a snipe hunt by demanding a new consensus they will never join. That is the fundamental flaw in your approach I am trying to point out, without arguing the merits of any particular case. Dhaluza (talk) 03:05, 18 December 2007 (UTC)
The distinction is in wording left out of the quote you posted above. The missing key word is "demonstrate", which is an active act. A bold edit that is not disputed nor discussed is passive consensus. When one is asked to "demonstrate" consensus, that is an 'active' act. That's the distinction between them, especially with disputed changes...which is clearly seen when comparing the differences on how consensus works on a passive bold edit change with an active disputed change discussion that's trying to find and demonstrate consensus. Sure, you can try to demonstrate consensus on a passive bold edit change, by saying "look, nobody objected", but is that really consensus or just the simple fact that no one noticed...?
  • It is not: "It is necessary to demonstrate a new consensus for change";
  • it is instead: "It is necessary to demonstrate a new consensus to make a disputed change".
It may be a subtle difference, but it is a difference. I apologize if I've seemed to accuse you of purposefully distorting, I'm not really suggesting that...heck, I leave things out all the time. If there's a specific comment you think that truly accuses you, I'll strike it out. My apologies.
And again, no one is blocking changes based solely on prior consensus, the disputed changes are being blocked due to lack of consensus for the proposed changes. Prior consensus is only being used to keep the current content from being removed by the few editors that are disputing it. If there were no prior consensus, and no current support for the policy content..then it should just simply be removed until consensus is found for a new version..or even to keep it out all together. Dreadstar 19:27, 18 December 2007 (UTC)
The whole controversy about the last paragraph of "Consensus can change" is being blown way out of proportion. CONS is clear, and has been clear for a very long time. The important point, that nobody disputes, is that
"A small group of editors can reach a consensual decision, but when the article gains wider attention, others may then disagree. The original group should not block further change on grounds that they already have made a decision."
This is the only point that I ever cited in the NOR dispute, and nobody is disputing it here. The changes to the last paragraph are simply to clarify what the above quote means, and to correct an addition from about July 2007 that could be taken out of context.
Nobody has ever established a consensus that for Wikipedia policy pages, a different set of editing rules apply, or that policy pages are a kind of legislative code that requires enactment to change. From the beginning, policy pages have just described what consensus currently is. They were never proscriptive. That would be against the whole spirit of Ignore All Rules—the first Wikipedia policy. COGDEN 20:16, 17 December 2007 (UTC)


Dreadstar: you are reversing the way consensus works on wikipedia. This is a wiki: You must edit to gain consensus (not gain consensus to edit). To explain: Many wikis do not even have talk pages. In such a situation, editing the article or page is the only way to communicate, if you do not do so you can never gain consensus in the first place. When introduced, talk pages were never meant to prevent editing, but merely as a means to declutter the same process.

Have people been using the process you describe during the NOR debate? (Perhaps artificially enforcing it with reverts?) --Kim Bruning (talk) 00:54, 18 December 2007 (UTC)

  • Kim, that is an over simplification. It is appropriate to be bold, but if reverted discuss. However, some pages need more stability and it is courteous to discuss changes first when practical. --Kevin Murray (talk) 00:58, 18 December 2007 (UTC)
    • Kevin, I think she has it right. Policy pages involve more eyes (and more assholes), so additional effort is needed to build consensus, and discussion is strongly encouraged. But stability is achieved through consensus, it is not an alternative to consensus. Dhaluza (talk) 01:57, 18 December 2007 (UTC)
That's exactly what happened, Kim. COGDEN's non-consensus changes to NOR policy were artificially supported by his and his few supporters reversions. Over and over. Basically, COGDEN tried to introduce major changes to WP:NOR by bold editing. Fine. He was reverted, but while the issue was being discussed in detail, ad nauseam on the talk page, COGDEN and his supporters kept adding and reverting back to non-consensus major policy changes, while many other editors who opposed the changes kept reverting them over and over again. It became a mess and several of us felt an RfC on COGDEN was warranted...and if you look at the RfC, you'll see a strong agreement that COGDEN was disrupting the policy page with his aggressive, unilateral attempts to change the policy. The situation is well summarized, imo, by this statement. Naturally there are other views, but that one has the most support.
The first bold edit to change policy is fine, but if it's reverted, then discuss on the talk page and find consensus..but that's not what happened, the non-consensus changes kept getting reverted back into the policy; and the reasoning was because he and a few other editors were disputing the current policy, and so saying that it should be removed or changed immediately because it no longer had consensus due the fact that they were now disputing it. Now, the old content that COGDEN was disputing did have prior consensus, and it was long-standing policy that is still supported by a large number of editors. COGDEN needed to find consensus for his changes and not continue trying to change by reversion and so-called "bold editing", which is not bold editing under the circumstances I've just described. Consensus and discussion when edits are reverted and disputed - that's the way to do it. And this is the situation I've talking about with all my comments about consensus on this talk page. Dreadstar 19:08, 18 December 2007 (UTC)
That's generally what happened. In my case, at least, I made a large number of "proposal edits" designed for compromise, or to move toward consensus. If one edit was reverted (which it almost always was), I tried another edit, sometimes compromising more or trying some other approach that I thought might gain acceptance based on the previous reversions or what was happening on the talk page (which was also usually reverted without comment too). The reason I did this for a while is that due to the volume and confusion on the talk page, nobody's proposals were ever looked at or discussed. Vassyana, for example, had some great proposals that nobody commented on until the edit was attempted, then the angry reversions came like a landslide. The edit comments, unlike anything anybody said on the talk page, got noticed. This was a bit of an experiment on my part, and it had some success (the compromise tag and some language remained on the page for nearly a month without reversion), but in the end it was unsuccessful and I gave up. Often, after a new compromise or proposal was reverted, I would discuss it on the talk page, although such discussions were easily forgotten given the volume of talk page edits. Sometimes, I just gave up on a particular compromise and tried some other approach. COGDEN 20:05, 18 December 2007 (UTC)
I can't believe I read that whole thread. ...and yet, apathy wins again. --Fandyllic (talk) 12:58 AM PST 23 Dec 2007

Dreadstar seems almost correct, except I do not recognize the concept of "prior consensus" which seems to be contrary to my understanding of how consensus works on wikipedia. COGDEN also seems to be mostly correct. The problem lies with the people doing reverts in favor of either version. These people are basically edit warring, as per the somewhat older policies to that effect. Note that the 3RR is an electric fence. You will almost certainly be blocked if you edit war beyond that point. But there is no policy that says you can't be blocked for edit warring and essentially gaming the system before that point.

Based only on the comments by COGDEN and Dreadstar here (the situation might be more complex) the problem seems to be caused mostly by reverting parties on both sides.

One solution might be institute a "0 revert rule" on the page perhaps, where any (non-vandalism) reverter is instantly blocked from editing for 24 hours and doubling for each consequent revert. Reverting "back" would also be a violation. A non-revert-edit (ie an edit with substantially new material) would be the only type of edit permitted. This would probably solve the situation, as the page could be edited essentially normally per the consensus process, and would eliminate those who are not quite current on the process. (and/or give them a time-out to study the documentation ;-) )

Let me stress that "0-revert" is not "0-edit". Some admins have some slight difficultly with the difference. An admin who accidentally blocks on an edit would have to be reprimanded and his actions undone. (If they wheel-warred, they may even need to be de-adminned). Hmph.

--Kim Bruning (talk)

Yes, indeed, there are a lot of different solutions to revert warring, Kim, but fortunately, the war seems to be over, with productive discussions taking place on the talk page. One minor thing, it is incorrect to say "The problem lies with the people doing reverts in favor of either version.", no - the problem (as I see it) is editors revert warring to keep non-consensus, hotly disputed changes in place - but that's been adequately addressed and is no longer an issue on NOR. If you want to discuss what I mean by "prior consensus", we can continue on our talk pages. Kudos to COGDEN for his statement above, I feel we're on - or at least almost on, the same page, and working towards consensus! Dreadstar 01:16, 25 December 2007 (UTC)