Wikipedia talk:Consensus/Archive 9

This is an archive of past discussions. Do not edit the contents of this page. If you wish to start a new discussion or revive an old one, please do so on the current talk page.

Archive 5

←

Archive 7

→

Reverted edits making consensus "one policy among many"

I do believe that wikipedia is run by consensus, not by policy. A system where procedures, processes and policies have primacy is called a bureaucracy. Wikipedia is WP:NOT a bureaucracy. Nor should it be. Bureaucracies have an exceedingly short internet half-life. :-P

--Kim Bruning (talk) 00:57, 3 July 2009 (UTC)

Thoughts: Wikipedia works (idealy perhaps) by consensus, as documented by various policies. Those policies are upheld by consensus, even the policies imposed from outside the community. Policies are maintained and interpreted by consensus. Consensus is informed by the policy text. Consensus and policy don't really compare well. I think this could lead to the prescriptive/descriptive flawed ideology. Policies tend to be prescriptive (perhaps they describe behaviour that will be enforced if you deviate). The consensus policy is not prescriptive. As a descriptive document, it's pretty weak, too. Dunno. In any case, the recent changes sound like philosophising, and do little in the way of meaningfully changing the message of the page to the intended audience. --SmokeyJoe (talk) 02:07, 3 July 2009 (UTC)

According to you, what is the message of the page to the intended audience? And who is the intended audience? --Kim Bruning (talk) 17:35, 3 July 2009 (UTC)

You mean this message "This page documents an official English Wikipedia policy, a widely accepted ..." To me, it means "this is a rule, obey it or there'll be trouble". The intended audience? I think it is the relative newcomer, especially the lead section. Specific points below may be relevant to experienced wikipedians. I worry that leading wikipedians seem to editing as a means of high-level communication with each other. --SmokeyJoe (talk) 23:55, 3 July 2009 (UTC)

@Chillum: NPOV , NOR, and V do NOT supersede consensus. I am especially surprised to see NPOV in the list, because the only way to achive NPOV is *through* consensus. (NPOV and the consensus system are complementary, and they were not just randomly thrown together when wikipedia was founded) --Kim Bruning (talk) 17:37, 3 July 2009 (UTC)

See Chillum's comments on this here. I must admit, I found myself agreeing with much of what he said. –xeno^talk 17:42, 3 July 2009 (UTC)

He's putting the cart before the horse. An NPOV article does not magically come into existence fully formed by a miracle.

Instead, we have to work hard on it. And we need to cooperate. Chillum appears to be using "NPOV" as a battlecry here. That's a lose-lose proposition.

If he wants to win the day, he is going to have to use good negotiation skills to move forward and form a solid consensus on exactly where that Neutral position actually lies. It is no coincidence that the NPOV position often lies close to the consensus of those present. It's similar to the reason why statistical sampling works.

Now; if we wish to have actual Neutral, Verifiable and well Researched articles, we cannot magically make them appear. That's impossible. Instead, we need to follow processes, so that we can reach our objective step by step. To ensure that our objective is obtained, we need to ensure that our processes have primacy.

Never confuse the ends with the means. --Kim Bruning (talk) 18:47, 3 July 2009 (UTC)

I like the way Ward 3000 and yourself are actually going about things there. In fact, you're even figuring ways to ensure that both wikipedia's ethics and psychologists' ethics are met at the same time. This is more likely to result in a neutral and verifiable article.

In contrast, Chillums original position where he tries to exclude psychologists eventually leads to a biased article, because the position of the psychological field cannot be accurately ascertained (because he would chase away all the professionals).

In short, he was claiming to want a neutral, verifiable article, but his actions at the time were preventing the article from reaching that state. I think he started to chill out a bit later on though. ;-)

Incidentally, his position on including or excluding the Rorschach test has little to do with the 3 core Article-writing values, IMHO. It has to do with opposition to censorship. While an admirable goal in-and-of itself, he should go about this in a neutral and considerate manner, taking into account the positions of his peers, as befits a wikipedian.

Once again, if he now continues to do so, I have no doubt that he will be able to ensure both that the article remains uncensored, and that our core values are upheld.

--Kim Bruning (talk) 19:08, 3 July 2009 (UTC)

What on Earth are you talking about? Can you show me exactly what led you to accuse me of attempting to exclude psychologists? I have not obeyed their commands, but that is not an attempt to exclude. Perhaps I have misunderstood you. This is way off topic here, reply on my talk page if you please. Chillum 02:06, 4 July 2009 (UTC)

I think Kim Bruning is talking about how including the Rorschach images violates neutrality by actively engaging in the dispute. If the article weren't so directly involved in the dispute, it would be easier to argue WP:NPOV. As it stands, including the images forces all the editors to take a non-neutral position in the dispute, either for or against. That's not a good situation, is it? It's why the issue is taking so long to resolve. And it's why it's spilling over into this talk page. Danglingdiagnosis (talk) 18:29, 4 July 2009 (UTC)

Consensus cannot override the core elements of WP:NPOV and WP:BLP for example, so it's inaccurate to say "NPOV , NOR, and V do NOT supersede consensus"; certainly superficial or ancillary elements of those policies can be overridden by consensus, but the core elements cannot be superseded by consensus. To say that “the only way to achive NPOV is *through* consensus” is not only false, it is also a red-herring, one still cannot override NPOV by consensus. Dreadstar † 21:26, 3 July 2009 (UTC)

This is sounding like a debate on angels on the head of a needle. Clear and accepted wp:consensus throughout the project is that WP:NPOV and WP:BLP are essential to the continued development of the project. For evidence, look at the lack of substance of controversy at the NPOV recent history and the BLP recent history, as per Wikipedia:Silence and consensus. What is the question that this debate addresses? --SmokeyJoe (talk) 00:23, 4 July 2009 (UTC)

You're just repeating the claim, but are not providing additional reasoning. If we don't write an NPOV article by means of consensus, then what means are used to write an NPOV article (or in fact, any article?) --Kim Bruning (talk) 21:40, 3 July 2009 (UTC)

NPOV is mandated by Jimbo, so I'm not repeating any claims, I'm repeating fact. The means to writing an NPOV article are outlined in our content policies, the core of which are non-negotiable. The real point is that Consensus cannot override core NPOV policy, not this red herring that somehow the ability to write an NPOV article is only due to consensus, that's a false trail. Dreadstar † 22:19, 3 July 2009 (UTC)

It seems to me that we shouldn't try to claim one policy is more important than another. NPOV is inviolable as policy, but frequently can only be interpreted by consensus. Sunray (talk) 22:40, 3 July 2009 (UTC)

I think that an answer can be sought in that direction. We could divide policies into what we want, and how to get there. NPOV is what we want, and consensus is how to get there. That way they don't bite each other. Fair enough?

However, I would like to be very careful: NPOV and the "wiki process", which on this wiki we define on WP:CONSENSUS, are both meta:founding principles, but V and NOR are not.

--Kim Bruning (talk) 23:25, 3 July 2009 (UTC)

Somewhere, did I once read that WP:V and WP:NOR exist as method to ensure WP:NPOV? --SmokeyJoe (talk) 00:26, 4 July 2009 (UTC)

If you did, you had better correct your source. For instance: wikinews has no WP:NOR, but still manages to maintain NPOV. Some other wikipedia's actually don't have WP:V either yet, and those also manage to retain NPOV. Historically, the english wikipedia was able to maintain NPOV without NOR *or* V.

This is understandable: NOR stops kooks from adding crazy things to the wiki, while verifiability improves the reliability of articles; neither of which have much to do with NPOV per-se. --Kim Bruning (talk) 01:25, 4 July 2009 (UTC)

A consensus to violate our neutral point of view, disregard verifiability, or engage in original research is not valid and should be discarded as such. Consensus is not the end all be all of the Wikipedia universe. We can't come to a consensus to say that Jesus is the only path to God, we can't decide to claim the Earth is only 6000 years old without any proof, and we cannot by consensus decide to publish a new variant of the theory of relativity we came up with one day after school. We are an encyclopedia first, consensus is merely a tool to help up create an encyclopedia. Contrary to Kim's belief consensus does not always lead to neutrality, sometimes a strongly bias consensus(particularly in nationalist and religious debates) needs to be overridden in the interests of neutrality.

By the way this is not just my opinion, but a long standing part of policy. From WP:NPOV: "Neutral point of view" is one of Wikipedia's three core content policies. The other two are "Verifiability" and "No original research". Jointly, these policies determine the type and quality of material that is acceptable in Wikipedia articles. Because the policies are complementary, they should not be interpreted in isolation from one another, and editors should familiarize themselves with all three. The principles upon which these policies are based cannot be superseded by other policies or guidelines, or by editors' consensus. Chillum 02:04, 4 July 2009 (UTC)

Consensus is not so much "a tool", but is a method, or a criterion for evaluation. --SmokeyJoe (talk) 02:14, 4 July 2009 (UTC)

Yes, I suppose it is. Chillum 02:15, 4 July 2009 (UTC)

What's wrong with letting consensus be the first step, and letting policy dictate in the event of a dispute? If everyone can agree on an ethical point of view, and I contend that there are some ethical situations that we can all agree on, (See

then consensus should be sufficient, shouldn't it? Or is the goal here to make policies that foresee every possible situation? That's a nice goal, but I don't think it's very realistic. It's a big world out there. Danglingdiagnosis (talk) 18:29, 4 July 2009 (UTC)

@Chillum: To meatball:NameTheConflict, I think you're skipping over a rather significant part of the NPOV quote there: " The principles upon which these policies are based cannot be superseded by other policies or guidelines, or by editors' consensus. "

That's a fairly significant part of the sentence you're ignoring there. Do you know what principles are actually meant? Especially when we take into account the fact that those principles cannot be superseded on-wiki, that's kind of important. --Kim Bruning (talk) 21:35, 4 July 2009 (UTC)

Not to restart this thankfully dead conversation, but, Kim, your assertion that a neutral article cannot be achieved except through the consensus process is false: you have just thoughtlessly asserted that every single article written by any one editor is always, necessarily, and without question non-neutral. This may be likely, but it is not necessarily true, especially for relatively concrete and uncontroversial subjects (e.g., a plant species). WhatamIdoing (talk) 21:37, 1 August 2009 (UTC)

Hmm, can you show me where I seem to say that that? I would think that a page written by a single editor might be just fine, though admittedly I do think everything needs checking (trust, but verify ;-) ). --Kim Bruning (talk) 15:39, 15 August 2009 (UTC)

You said: "the only way to achive NPOV is *through* consensus". WhatamIdoing (talk) 18:21, 15 August 2009 (UTC)

Ut oh

Some content issues – such as copyright violations and certain issues relating to the biographies of living persons – are not normally subject to debate or consensus, primarily because of the risk of real harm inherent in them.

I'm not saying that this isn't current practice (ish) , but ...

The main way we regulate things on wikipedia is through consensus. By withdrawing certain issues from the consensus process, we are essentially removing them from regular, ahhh regulation. That is to say: making things a LOT worse. That can't be good. --Kim Bruning (talk) 16:08, 16 August 2009 (UTC)

The fact is that these things override consensus. If consensus is to violate copyright law, we don't violate copyright law. Consensus is not the only form of regulation. Chillum 16:11, 16 August 2009 (UTC)

Hmm. It's about the way we approach the issue. It's very nice to just say, for instance, "we don't torture", but if there is no mechanism in place to prevent torture, then torture will occasionally take place, no matter how wonderful the ideal may be.

The same is true of more regular things like copyright law, or blp. We only really have one mechanism we use to achieve everything (KISS principle ;-) ), and that's wiki-editing and consensus. If the mechanism doesn't get used, how else are you going to be sure copyright law was applied correctly? And how else can you be sure that some BLP article is doing no harm?

So we need to collectively keep our brains turned on, and keep them working together through the consensus mechanism. Or do you have an alternative mechanism through which to achieve our goals? ;-)

--Kim Bruning (talk) 16:20, 16 August 2009 (UTC)

I suppose we would use WP:Copyright violations for copyright issues, and WP:Biographies of living persons for BLP issues. We have WP:Possibly unfree files if we are not sure about a copyright status, and we have our legal counsel if we are still not sure. If there is consensus that we use a copyright violation, we don't use it. If there is a consensus that something is not a copyright violation, but our legal counsel disagrees then we still don't use it. If there is consensus that something is not a copyright violation, and there is no other indication that it is then we will use it, not because consensus has authority in the areas of copyright issues, but because we have no reason to think it is being violated. Perhaps I am not understanding your objections and I am talking cross purposes? Chillum 16:25, 16 August 2009 (UTC)

Those are rules, not mechanisms.

Now, I'm going to assume that normally, it is consensus to remove a copyright violation. In that case, the end result is that the copyright violation is removed.

Would you agree that that is what would happen, given that my assumption holds? --Kim Bruning (talk) 16:58, 16 August 2009 (UTC)

I am saying that if there is consensus to remove a copyright violation we remove it, and if there is consensus against removing a copyright violation we remove it. BJAODN is a good example such a situation. The mechanism is that such violations are recognized and responded to, in the case of copyright violations usually the "delete" button. While discussion certainly plays a role, it can not override the policies on these matters. Chillum 18:23, 16 August 2009 (UTC)

Well, if there is a consensus against removing a copyrighted image that appears to be a copyvio, there is usually a very good reason. In those cases, chances are good that an arrangement will be made and the image will be kept. For instance, the famous photo of Phan_Thị_Kim_Phúc , which is an iconic photograph, has been kept. --Kim Bruning (talk) 22:14, 16 August 2009 (UTC)

No. If there is (current) consensus against removing a copyvio image it just means that the people who showed up to discuss it don't understand our rules and copyright laws and/or don't care. If enough people show up in a specific discussion to vote in favor of doing something against core principles, we do not do it anyway. To compare it to governments, no government is set up as a direct democracy, simple because a straight vote of uninformed can end up being chaos. In the United States we have constitutional rights and standards that cannot be contradicted by laws, and if enough legislators get together to pass such a law it gets tossed out by the courts. Wikipedia's founding principles are our constitution. DreamGuy (talk) 13:04, 19 August 2009 (UTC)

Fair use is not a copyright violation from the image page: "Permission notwithstanding, we still assert fair use". If there was consensus against removing a copyright violation, the material is removed, period. If there is consensus to go against neutrality, then we stick with neutrality. If there is consensus to use unverifiable negative information against a living person, then we don't do that. Much of what Wikipedia is comes down to consensus, but not everything, consensus is not the end-all be-all of decision making. Sometimes we just have firm rules. Chillum 14:01, 17 August 2009 (UTC)

I think it might be helpful to take a step back and remember that our policies and guidelines are based on consensus. They reflect a community wide consensus about what we can and can not do on Wikipedia. If that community wide consensus changes, the "rules" will change. So really, this discussion is about whether a local consensus at the article level can over-ride community wide consensus at a project level. The answer to that is "no".... community wide consensus takes precidence. Blueboar (talk) 16:17, 18 August 2009 (UTC)

If the community gets together and decides to ignore our founding principles, they should get the heck off this site and form their own project with their own founding principles and leave this one alone. Wikipedia is not an experiment in social networking, it's an encyclopedia, and that means certain rules that cannot be violated. When Jimbo or the Wikimedia board see something out of control they come and lay the smack down, consensus or not, but it shouldn't even get that far most of the time. DreamGuy (talk) 13:04, 19 August 2009 (UTC)

The wording and interpretation of these policies is certainly subject to debate and consensus. How do you think they came to be in the form they are? I agree with Blueboar--the basic meaning of a policy is a wider consensus. I disagree with him about whether the community can make exceptions at an article level. The community can always make exceptions. The community can always decide how to interpret. The community can change the wording or the interpretation either globally or locally. The community, if nothing else serves, can always use IAR. DGG ( talk ) 21:06, 18 August 2009 (UTC)

Kim, I know that you value consensus very highly indeed, but it's simply not the be-all and end-all of Wikipedia's existence. For example, even if every single one of the thousands of Wikipedia editors "fully agreed" to say something plainly libelous in an article, or to turn Wikipedia into a child pornography distribution site, then the Wikimedia Foundation office could override that "consensus decision" -- and could enforce its "anti-consensus" decision technologically, by blocking users, protecting pages, and so forth. Consensus is not an absolute value: it's the currently supported primary process, and it has many advantages, but it is neither the sole process at work nor the only possible process. WhatamIdoing (talk) 05:30, 19 August 2009 (UTC)

Chillum, you state "If there is consensus to go against neutrality, then we stick with neutrality". Who is this we, that sticks with neutrality? The same we in which there is a consensus not to stick to neutrality? Copyvio is different, since the foundation has set clear demands on use of non free content, in contrast the requirement of neutrality, which from the foundation side is much softer worded. Essentially even if we agree to use copyvio's we could, and would, be overruled. Taemyr (talk) 18:25, 19 August 2009 (UTC)

Primacy of consensus for the benefit of continuing education

Primacy of consensus over the letter of any policy is essential to ensure that editors of the day own, and take responsibility for, the project. Requiring consensus means that all editors entitled to opinion need to be know and understand the reason behind the rules. If a local consensus is going off the rails, it needs to be important that a wider collection of editors join the debate to better inform the local group. The alternative, that someone, some “founding fathers” write irrevocable rules, means that continuing education of new editors is no longer a priority. That would be a recipe for decay. --SmokeyJoe (talk) 10:40, 19 August 2009 (UTC)

The existence of "founding fathers" or founding principles that cannot be ignored despite a local consensus in no way means that "continuing education of new editors is no longer a priority". That makes no sense. Any project at all needs an underlying set of rules or principles that need to be followed or else it's just a chaotic mess of nothing thrown together at a whim. All the local consensus in the world to ignore WP:NPOV, for one example, does not make it acceptable. At our core we're an encyclopedia. Anyone who edits for a reason other than that doesn't need to be here. DreamGuy (talk) 12:05, 19 August 2009 (UTC)

If a "local consensus" decides to ignore WP:NPOV, it is important to tell them why WP:NPOV is important to the project, and it is a mistake to try to say that WP:NPOV must be followed because it is a rule written by someone important. --SmokeyJoe (talk) 21:23, 19 August 2009 (UTC)

Board mandated policies include Wikipedia:Non-free content criteria and Wikipedia:Biographies of living persons.

This is not exactly correct, and gives the wrong impression. The board mandates that we have such policies. It doesn't mandate their contents. How we interpret their requirements are almost entirely up to us. DGG ( talk ) 21:00, 18 August 2009 (UTC)

Certainly their contents to a large extent are also mandated. If our non-free content criteria contradicted what the board expressly wanted then it'd like not having it at all, or worse even. DreamGuy (talk) 12:40, 19 August 2009 (UTC)

Yes, some aspects of their general direction is mandated, but that still leaves a enormous amount of flexibility. I am not sure how things would proceed if the board actually did object to something on those policy pages, or considered that something should be there which is not. In practice, I think comments to that effect would be most likely to come from Mike Godwin. I don't think there's ever been an actual official objection, though of course people frequently make well-founded or ill-founded comments about what the board would say. DGG ( talk ) 21:35, 19 August 2009 (UTC)

The "Discuss First" option and Wikipedia Policies

Over a year ago the editors who regularly watch and edit this policy engaged in a long debate over flow charts and what the processes behind WP:CONSESNSUS should be (see Archives 5 and 6). Among the the various models discussed were several that suggested a "Discuss first" approach to changing articles. These were essentially rejected, in favor of "edit first" models.

Meanwhile, the editors who regularly watch and edit many of our other policies and guidelines were more and more insisting on a "discuss first" approach to consensus building, to the point where "discuss first" is the accepted norm for many policy pages (I am sure there are exceptions, but I know it is true at the pages I watch- such as WP:V, WP:NOR, and WP:RS. Today, an increasing number of edits to policy pages are reverted with comments such as "this should be dicussed this first". You can say that this is "in violation" of WP:CONSENSUS... but that is the reality of what is happening in the trenches out there on the policy pages... and the violators are some of our most respected editors.

This discrepancy between theory and practice has come to a head recently at WP:Editing policy, which includes a section on Editing policies and guidelines that explicitly states that major changes to policy pages that would change the substance of the policy (which I take to mean: change "the rules") should be anounced and discussed first.

Therefore, I think it is time to re-open the question of "edit first" vs. "discuss first"... at least as these two models apply to Policy pages. Judging by what is actually happening at other policy pages, the consensus seems to have changed on what the best way to achieve a consensus on policy pages is. "Discuss first" is, at least, an acceptable variation on on the BRD theme. This policy should reflect that change in consensus. Blueboar (talk) 19:46, 19 August 2009 (UTC)

This is actually not about a difference between theory and practice, but a difference in breadth and depth of experience. It looks like we're actually seeing people who have never actually wiki-edited before.

I recently asked Blueboar if (s)he had ever (managed to) simply edit a page along with others in a collaborative fashion, with no reverts happening, and perhaps not even much discussion - aka. a short, to the point description of the outcome of normal wiki-editing as I understand it; (& given fundament by Founding principle #3).

I think (s)he said that no they had not. Blueboar, is that correct?

--Kim Bruning (talk) 20:04, 19 August 2009 (UTC) Note: this is not an accusation, not a trap, and it is VERY important to me to get a good answer. Remember the thread where I said I'd been horrified and needed to think?

Actually, the whole bullshit of discuss-first was what enticed me to get WP:DEVCOM going before it was thoroughly discussed to death — that there is no way to demonstrate the viability of such a committee without actually trying it first, on the principle of being bold. @harej 20:15, 19 August 2009 (UTC)

If Blueboar's answer boils down to 3 letters (Y,E,S) , I think that we're going to have slightly bigger fish to fry. :-) --Kim Bruning (talk) 20:22, 19 August 2009 (UTC)

WP:Consensus makes no explicit reference to editing policy, and I think there's a strong argument for handling that separately. I've tried to skim the WT:Consensus archives 5 and 6 as Kim suggested, but it's making my head hurt a bit and I'm certainly not getting any clarity on distinctions being made between consensus in articles and consensus in policy. Or, for that matter - the issue we were talking about recently - whether on policy pages it's better to discuss first or to edit first. The archives 5 and 6 seem to focus on the relation between policy and practice and local v global consensus, which is a different issue. Rd232 ^talk 20:43, 19 August 2009 (UTC)

No it doesn't make distinctions between editing articles and editing policy. Actually, It makes reference to editing *everything* without distinction.

I believe Kevin Murray was a proponent of "discuss first" at the time.

--Kim Bruning (talk) 21:44, 19 August 2009 (UTC)

OK. So do you have an opinion on whether there is a difference between editing articles and policy? Rd232 ^talk 22:16, 19 August 2009 (UTC)

I do. In some people's theory, there's a big difference. In practice, I don't think there really is (for some suitably large scale and inclusive value of in-practice, perhaps, which includes pages subject to mediation or arbcom.) --Kim Bruning (talk) 22:31, 19 August 2009 (UTC)

Actually, I have had one or two wonderful experiences of the kind you talk about... they are extremely rare (perhaps because I tend to edit articles on highly controvercial topics) but I have experienced a few. That said, in over five years of editing Wikipedia, I have never experienced wiki-editing as you understand it on a policy page. In fact, I would say that "normal" on a policy page is for people to demand extensive discussion on just about any change, especailly one that might alter the meaning and intent of the policy. (hell, at times you could argue that normal on a policy page is an edit war, until some admin step in and forces everyone to discuss by locking the page for a few days)... sometimes the consensus results in the change (or an ammended version) being added... most of the time the consensus is to leave the policy page as it was (people tend to be very conservative when it comes to changing policy pages).

Don't get me wrong... I am not saying that the process currently outlined on this page does not work. I am saying it does not reflect the reality of how consensus building actually takes place on most of our policy pages. I think it should at least acknowledge that reality. Blueboar (talk) 20:45, 19 August 2009 (UTC)

Oh... and just so you can use the correct pronoun... I'm a he. Blueboar (talk) 20:56, 19 August 2009 (UTC)

By the way... you folks might want to see what is stated at Wikipedia:Policies and guidelines#Content changes... this seems to back up what I have been talking about. Blueboar (talk) 21:06, 19 August 2009 (UTC)

That was recently refactored, edited, and messed with too. And messed up.

So you're saying you've experienced wiki-editing a couple of times? Well, that's something, phew. I was about to call in Jimbo and foundation and the Board and the horsemen of the apocalypse (ok, maybe not the horsemen, but I would have pulled weight to get the rest).

Still, only a couple of times seems pretty dismal, for something that we're actually here to do (it ain't called wikipedia for nothing, you know! ;-) ). I take it no one managed to get the regular wiki-editing going at the other locations you were at?

Well, maybe I won't drag in the big guns, but just Sam and Sj or so <hmmm, scratches head>.

They're working on a bag of tricks, and I wonder if they'd be able to open it in some of the places you've been having trouble in recent history. Let me ponder for a bit. --Kim Bruning (talk) 21:44, 19 August 2009 (UTC)

To refocus: Let me repeat, I am not trying to say that the system outlined on this page does not work, ... what I am saying is that this policy does not reflect actual practice on other policy pages, and I think it should at least acknowledge that actual practice. It needs at least a mention that the "discuss first" concept is being used on policy pages. Shall I come up with proposed language, or is this a non-starter here? Blueboar (talk) 00:11, 20 August 2009 (UTC)

I see it as a starter. I was involved in the previous discussions, but I don't think I engaged on "discuss first for policy pages" issue. However, it was, and it is my view that this page needs to be primarily directed to newcomers editing mainspace pages. WP:Consensus is a prominent reference page for newcomers. The flowchart, which is a nice and simple one, works best for new editors working on an unrefined article. It applies to a case where significant edits can move the page towards consensus, where talking about the edits can be far more confusing than looking at the result of the edits. Policy pages are definitely different. They substantially reflect a long established consensus. There were many other, more complicated flowcharts that might be better applied to established policy. The flowchart remaining at WP:BRD comes to mind. Kim was going to make a subpage study of the many proposed charts (each reflecting a subtly different approach). Maybe this idea should be revisited, perhaps from a section near the bottom, titled, “Refinement of consensus in contentious areas”. --SmokeyJoe (talk) 01:27, 20 August 2009 (UTC)

OK... below is an initial draft for us to ponder. Blueboar (talk) 14:25, 20 August 2009 (UTC)

Consensus on Policy and Guidelines (drafts)

Initial Draft:

The principals outlined in Wikipedia's Policy and Guideline pages form the basis for all editing on Wikipedia. These principals, and in many cases even the specific wording of the policies and guidelines, reflect the long established consensus of the entire Wikipedia community. The community expects and relies on our policies and guidelines to remain stable and consistent. While consensus can change, when it comes to established policy and guideline pages it does so slowly, and only with a lot of prior discussion. For this reason, being overly bold, and making significant changes to established policies and guidelines without some initial discussion on the talk page is often seen as being disruptive.

-------

(post revised Drafts here):

Discussion on Drafts

I am sure that my initial draft needs a lot of work... so please discuss and post alternatives. Blueboar (talk) 14:25, 20 August 2009 (UTC)

So, just to be sure, let me repeat that in my own words: it's being said here that we should not use the wiki-process (founding principle #3) as described at wikipedia:consensus for policy pages, and in fact it is disruptive. This is because we want those pages to be stable and consistent.

Correct?

--Kim Bruning (talk) 19:54, 20 August 2009 (U

Don't be such a drama queen. Where is it written that wiki process consists of edit first, discuss later? Also WP:Consensus doesn't say anything about editing policy. Also, presuming that you mean the fourth pillar (WP:5P), let me quote it: "Wikipedia has a code of conduct: Respect your fellow Wikipedians even when you may not agree with them. Be civil. Avoid conflicts of interest, personal attacks, and sweeping generalizations. Find consensus, avoid edit wars, follow the three-revert rule, and remember that there are 3,005,683 articles on the English Wikipedia to work on and discuss. Act in good faith, never disrupt Wikipedia to illustrate a point, and assume good faith on the part of others. Be open and welcoming." Can't see any contradiction with that. Finally, your concluding sentence is correct: we want policy pages to be stable and consistent, and we do not want policy to be constantly blown about by every passing breeze of an editor with a Bright Idea. Instead it should evolve more slowly. PS Unlike article content, building policy is not actually an objective - quite the opposite (WP:CREEP). Rd232 ^talk 20:12, 20 August 2009 (UTC)

Atually Rd, I think Kim was referring to Founding principle #3... which I note does not spell out exactly what the "wiki-process" actually is (it certainly does not say it is "edit first").

Kim, my intent was not to say we should never use the wiki-process (as you see it). The BRD collaborative cycle works extremely well when creating a new policy/guideline page... and for making minor tweeks to existing policy pages. It even works for major changes... after initial discussion. But yes, being bold and making a major change to established policy without at least some discussion to show that the change is needed and wanted by the broader community is considered disruptive. Blueboar (talk) 21:09, 20 August 2009 (UTC)

Rd232, I don't think I was being dramatic or anything, unless you were reading something into it. I think the wiki-process has something to do with editing pages ;-). I'm also pretty sure "discussion" is not really an essential component of the typical wiki-process at all, simply because the first wikis (and actually most wikis except for mediawiki) didn't/don't actually have talk pages. (it does get slightly more complicated when discussing meatball:ThreadMode vs meatball:DocumentMode, etc..., but the basic premise does hold)

So if you state that some people feel that editing a particular page (Which is the one thing that makes a wiki a wiki) is actually disruptive, I think the logical conclusion to draw is that those people are opposed to using the wiki aspects of mediawiki, in that case.

Fair enough? --Kim Bruning (talk) 22:01, 20 August 2009 (UTC)

Blueboar: The BRD approach is used to deal with issues of filibustering, or where conversation has broken down. If you find the need to use BRD to create a new policy/guideline page, or a *new* page of any kind, you are doomed, I'm afraid, and you had best give up.

The intent of BRD is to start making changes to pages that were recently protected (and now unprotected), "established policy", "the consensus version", etc; for instance as part of a mediation (more typically medcab or fully informal mediation, and less used at medcom) . The intent is to try to find out which people are holding positions that are holding up consensus, and to get the page editing process flowing again. Sometimes they're covering up for some sort of uncertainty, or they do not trust the wiki in some way, and are unwilling to believe that the page can actually be improved. This happens a lot on policy pages, of course, so BRD is a very useful tool there, I've found over the years.

I'm somewhat amused by the fact that the same process that is said to be suitable for representing consensus built up and documented over the course of centuries among leading scientists worldwide; that same process is somehow assumed not be suitable for representing the consensus built up over the last couple of years among this rag-tag band of internet-addicts on a cutting-edge, competitive, and constantly evolving website. ;-)

Obviously, ultimately, the wiki-process does deliver stable, reliable pages. (see: WP:FA).

There are some very good reasons to assume that the wiki-process does work (not even counting empirical evidence).

--Kim Bruning (talk) 22:23, 20 August 2009 (UTC)

In other news, note that I'm going to be at Wikimania the coming week, so might not be able to post regularly (ironically). Is anyone from this discussion going to be there? :-)

--Kim Bruning (talk) 22:25, 20 August 2009 (UTC)

Kim, you definitly are coming across as being overly dramatic, ... no one is opposed to using the wiki aspects of mediawiki... we are simply saying that, on policy pages, there needs to be discussion before using them. No one is saying good faith editing, even on a policy page is disruptive... they are only saying that making a significant edit to an established policy page without discussion first is disruptive.

You have pointed us to Founding principle #3... so let's look at what that actually says: The founding principles include... "The "wiki process" and discussion with other editors as the final decision-making mechanism for all content (bolding mine). All we are really saying is that when it comes to Policy pages, the order is and should be reversed... to: discussion and then the "wiki-process". Blueboar (talk) 22:38, 20 August 2009 (UTC)

Sigh... reading your last, Kim... perhaps it is you who should give up. You are still stuck in a never never land of how things are supposed to work on a wiki. Please break from that, and realize I am talking about what is actually happening on Wikipedia policy pages. Perhaps I don't use or undertand BRD correctly... that's because I'm a simple guy and don't care a whole lot about terminology, flow charts and how things are supposed to work, I look at how they actually work... in this I think I am a fairly typical Wikipedia editor. I think my proposal accurately refelcts what is actually happening on the policy pages. But the more I read from you, the more I get the idea that you are out of touch with the realities of editing on Wikipedia. If something like it is not included in the policy... well, to be honest, it does not really matter. The editors in the trenches will simply "ignore this rule" and continue to demand discussion before considering changes to policy. Blueboar (talk) 23:20, 20 August 2009 (UTC)

That's the point, they are simply following the R-D, while the bold editor follows B. I think Kim has this mostly right. Do you think that "I reverted because this change was not discussed first" is ever a legitimate reason to revert, all other things being equal? Because the changes you propose seem to make it so. M 23:43, 20 August 2009 (UTC)

(ec) Somehow, you're still assuming that I'm a theorist. The irony is that I'm actually very much not. Notoriously so; infamously so. People complain that my "descriptive, not prescriptive" mantra is beginning to grate ;-)

So I'm quite happy that you want to keep things realistic. My problem is to figure out how to convince you that my practical experience is quite as real as yours. Perhaps I've been neglecting that aspect a little. If so, I apologize.

A second aspect is the following: Not everything you encounter on the wiki is a good thing. Some behaviours are not productive at all, such as vandalism. So we do tell people about vandalizing, but we tell them that vandalizing is a bad idea. Of course, sometimes it is difficult to decide whether something is good or bad, and then we need to discuss that. So that's the second aspect we need to discuss.

Once we get those two aspects covered, we can probably reach a consensus on what to actually put up on the page fairly quickly.

--Kim Bruning (talk) 00:01, 21 August 2009 (UTC)

Ok, I've pondered a bit, here's some questions I'm currently wondering about.

What would be sufficient to convince you that I'm not spinning fairy tales?
Can you show that the discuss first approach actually helps keep policy pages tidy, reliable, and in line with community consensus? Since you get to make my life difficult now, it's only fair that I ask difficult questions back. ;-)
Of the pages you edited that were actually fun to work on, do you think those reflected consensus better after you were done with them, or do you think that they failed to reflect consensus better after you were done with them?
Finally, (trick-ish question ;-) ) Is it your position that for each edit to a wiki-page you first discuss, and then afterwards you follow the wiki-process?

--Kim Bruning (talk) 00:16, 21 August 2009 (UTC)

M... you asked: Do you think that "I reverted because this change was not discussed first" is ever a legitimate reason to revert? My answer is YES, on policy pages I do. And based on actual practice on our various policy pages, so do most of the editors who are involved. This is what I am saying needs to be reflected in this policy. Blueboar (talk) 00:34, 21 August 2009 (UTC)

Kim: I am not at all sure how you can convince me that you are not spinning fairy tales... but one thing that would help is if we could agree that I am not talking about the wiki-process in general, but am only talking about how to best discuss the actual current practice on policy pages.

As for your second question... to be honest that is irrelevant. This is exactly why I am saying you are building fairy castles and talking theory. My point all along has been that this is happening, right now, on actual policy pages... not that it is necessarily a good thing. I suppose I am saying that the community consensus on how to edit policy pages has already shifted to discuss first, whether that is a good thing or not, and because it has already shifted, this policy should reflect that shift.

As to your final question... no. small minor changes that don't really change the meaning of the policy certainly don't need to be discussed. And for major changes, once it is clear that the community consensus on a policy issue has changed... and that the policy in question needs to be changed to reflect that new community consenssu, I think being bold in figuring out the language can actually help. That's why I have been saying that the BRD process happens after discussion. Blueboar (talk) 01:12, 21 August 2009 (UTC)

Big question for you. That revert is an edit. All edits must reflect consensus. Does that revert reflect consensus? Why?

How do you know for certain that it reflects consensus? If someone asks you for an exact and in depth analysis of the current consensus as you see it, would you be able to provide it at the point in time when you do that revert? --Kim Bruning (talk) 00:38, 21 August 2009 (UTC)

Yes... I would point to the policy page as it was before the BRD cycle took place. While it is understood that consensus can change... when it comes to policy pages we need some evidence that it actually has changed before accepting major edits. The default is to return the page to what it was before. If I were to edit the WP:Consensus policy so that it was substantially different, I would expect you to revert my change... unless I could demonstrate that the community consensus has changed. That is what I am attempting to do now. Blueboar (talk) 01:12, 21 August 2009 (UTC)

Oh, don't bother. Just add your text to the page. I'll try to NPOVize it as much as possible, and let's see where we end up? --Kim Bruning (talk) 01:39, 21 August 2009 (UTC)

WP:NPOV?? You really don't distinguish between policy and content, do you? Rd232 ^talk 11:50, 21 August 2009 (UTC)

Well, NPOVize-ish, since it's not exactly the same, but close. And yeah, you thought I was kidding? ;-)

It's pretty much the same process, except we're documenting things internal to wikipedia, rather than things in the outside world. The only thing we really can't seem to avoid is original research, as that appears to be inherent to the practice of naval-gazing we call the wikipedia: namespace :-) --Kim Bruning (talk) 19:23, 21 August 2009 (UTC)

Point by point reply:

Ah, how does the process vary between policies, and say some other page you mean? Well, in my experience, it's a continuum. You've got regular wiki-pages at one end, and pages where the editors are currently having an arbcom case on the other end. Policy pages are somewhere in-between.
I see where you're at there. As a counter example: Vandalism is also a very prevalent "procedure". We should certainly mention it, but it is not a best practice, so we can tell people to Not Do That. ;-) In the case of enforced discuss-first, I currently believe that practice is actually harmful. So it helps us both to explain why it is or isn't, according to you. :-)
I think you're conflating wiki-process and BRD process here? I can't quite make out what you're saying. Would you care to clarify?

In the mean time, let's see where we end up on the page? :) --Kim Bruning (talk) 01:47, 21 August 2009 (UTC)

Point by point re-reply:

OK... but I am talking about Policy pages and only policy pages... so could you elaborate on how you see policy page editing and what you mean by in-between?
Yes, that is the firm consensus on dealing with vandalism. The community agrees that vandalism is bad and should be discoraged. But, judging by what is occuring on the policy pages, the comunity does not agree that "Discuss first" is bad. In fact, it's looking increasingly like the community consensus is that "discuss first" is good. For me, the main argument in favor is that "discuss first" gives everyone a quick reality check on whether the community consensus has or has not changed regarding the section of the policy that is under discussion. If lots of people respond negatively, then there is at least an indication that the change does not reflect community consensus. If lots of people repond positively then the oposite is true.
this is why I hate jargon... I probably am conflaiting wiki-process and BRD process. (I tend to not care about process and focus on end results)... Can you explain the difference in simple terms.

And for placing my draft on the page... ok, If you think that is best. will do. Blueboar (talk) 03:21, 21 August 2009 (UTC)

Blueboar, I think you err on the side of being too tentative. Your addition to the policy was not at all disruptive. I have edited the section you introduced with the aim of making it clear and to the point. Do you disagree with my edits? --SmokeyJoe (talk) 06:34, 21 August 2009 (UTC)

I think it's unfortunate that principle #3, which is specifically limited to content, is being invoked in a discussion about what the "proper" behavior is on a non-content page.

In my experience, the level of prior discussion required for a substantive change to be accepted on guidelines varies by page, and even by the section of the page. For example, if you merely re-order WP:ELNO (an act that involves no content changes), you can not only expect it to be very promptly reverted, but also to have many editors angry with you. On the other hand, if you rewrote WP:EL#ADV, there's a chance that no one will care.

I think it would be fair -- even an act of kindness -- to less experienced editors to warn them that well-intentioned changes to non-article pages may irritate other editors, especially if the page has developed a discuss-first culture, and that caution (in the form of a talk page proposal) might be a good first step. We don't need to send newbies off to the dragon just because Kim wants to lead the charge. WhatamIdoing (talk) 06:55, 21 August 2009 (UTC)

"For this reason, being overly bold, and making significant changes to established policies and guidelines without some initial discussion on the talk page is often seen as being disruptive." I am against the use of the word disruptive in this sentence as it has a specific meaning among Wikipedia editors. It is not "disruptive" to use the BOLD cycle, but in the case of established Wikiedia Policy pages using of the BOLD cycle to edit first is probably counter productive to the interests of the editor wanting to make a change. I do not think the proposed wording should contain a club that stops all BOLD editing of policy pages, but rather advises that discussion takes place first.

This really comes down to the content of the major policy pages that by and large have a consensus, and usually changes to the wording even if not meant does change a nuance or two of the meaning (like the use of the word "disruptive" in the sentence I quoted above), which might not be familiar to a person who has not been involved in the discussions over the paragraph/section they are changing. So it is better to assume that the current wording reflects consensus unless under discussion it is shown that it does not. This is very different from editing an article where a section or paragraph that is unsourced, or poorly sourced, is replaced with a sentences that are sourced to reliable sources. Such BOLD editing is quite acceptable and supported by WP:PROVEIT. --PBS (talk) 10:03, 21 August 2009 (UTC)

Well said, Philip. Blueboar (talk) 15:16, 21 August 2009 (UTC)

I've heard the reasoning before, but doesn't this contradict WP:SILENCE and WP:CCC?

I tend to think the content of an article often documents years, decades, or even centuries of consensus built up in the real world, while policy pages document consensus on something as ephemeral as a wiki. --Kim Bruning (talk) 19:20, 21 August 2009 (UTC)

I add to PBS's comment that the statement that such changes are "often seen as being disruptive" is no more than the literal, obvious truth. Perception is not reality: editors may perceive undiscussed changes as being disruptive even if they are not actually disruptive. WhatamIdoing (talk) 20:58, 21 August 2009 (UTC)

Kim, I do not think that discussing changes to the major policy pages before making changes contradicts either WP:SILENCE or WP:CCC. If one suggest a change on the talk page and no one objects then WP:SILENCE is reinforced, as is the argument that consensus can change. --PBS (talk) 22:54, 21 August 2009 (UTC)

I think we're having two discussions. I have no problem whatsoever with discussing things first, as long as it is facultative. I doubt anyone would disagree.

My only issue is with people who want to make a discuss-first stratagem mandatory. Enforcement (through reverts, page protection, peer pressure etc) exposes a corner-case in the wiki system, where it becomes possible that consensus can never be reached (in theory), or at best that the reaching of consensus becomes very difficult (in practice). See archives 5 and 6 for the gory details. So let's not do that. ;-)

--Kim Bruning (talk) 14:55, 22 August 2009 (UTC)

New draft

Policies and guidelines are supposed to state what most Wikipedians agree upon, and should be phrased to reflect the present consensus on a subject. In general, more caution should be exercised in editing policies and guidelines than in editing articles. Minor edits to existing pages, such as formatting changes, grammatical improvement and uncontentious clarification, may be made by any editor at any time. However, changes that would alter the substance of policy or guidelines should normally be announced on the appropriate talk page first. The change may then be implemented if no objection is made to it or if discussion shows that there is consensus for the change. Major changes should also be publicized to the community in general, as should proposals for new policy pages.

This is pretty descriptive. It is not absolute - neither approach is enforced. It distinguishes minor from major changes, which is important. And it is focussed on policy, which has stability requirements content does not. BTW, anyone recognise this text? Yes ladies and gents, it's from WP:EP#Editing policies and guidelines. Because at the time I wrote it, I reflected what I thought was current policy and practice. And I don't see anything's changed in the 3 months since. Rd232 ^talk 11:50, 21 August 2009 (UTC)

This is acceptable to me... I will be happy as long as we inform editors that, when it comes to many policy and guideline pages, at least some prior discussion is expected. Blueboar (talk) 12:46, 21 August 2009 (UTC)

"expected" is exactly the right verb. Kim sometimes seems to be talking as if we want to say it's "required". It's "expected" not "required" in practice and there are good reasons for that, some of which emerge from discussion above. This should be reflected in policy. Rd232 ^talk 17:48, 21 August 2009 (UTC)

Ahhhh! Well, sometimes I've had to deal with people who do try to require it, and will go as far as gaming WP:3RR to do so. It is really hard to reach consensus in that kind of situation, as you can imagine.

Perhaps this is partially what sets my teeth on edge when you tell people to go to talk first.

It is of course FINE to actually discuss things first, but my worry is that people will conflate "it's a good idea to discuss things first, at times, especially if bold editing makes you nervous or depressed", with, "revert, revert, revert like you don't have a care".

Yes, that's silly, but I have seen people manage to somehow twist things that way a little too often ;-)

Would there be some way to say "of course, discussing first is not the only way to do things, there are many approaches to reaching and determining consensus. Please allow people to use the methods they prefer and are accustomed to", or some such.

--Kim Bruning (talk) 19:37, 21 August 2009 (UTC)

I'm confused - doesn't WP:BRD require reverting? Sounds like what you're really complaining about is that reverters won't discuss the changes (!). Rd232 ^talk 20:28, 21 August 2009 (UTC)

Often the reverter won't edit, and indeed they won't discuss much or discuss at all either.

I'm not opposed to discussion, you know!

Just because I don't happen to think that it is the only way to reach consensus; and I'm pretty much opposed to enforcing it at the expense of all the other methods; that doesn't mean that I don't think that discussion isn't important!

BRD distinguishes between 3 kinds of people: The "fisher" looking for Most Important People, the Most Important People themselves, and the innocent bystanders.

In a perfect world, the person doing the "fishing" only makes 1 edit themselves, discusses, and never reverts. In a less perfect world, the person doing the "fishing" might still be able to get a perfect 0RR, but they'll likely end up doing more edits. In the current version of BRD, it is rare for a fisher to do reverts. In the one scenario where they do need to revert, it is at most once per 24 hours.

The intent of the fisher is to get *other* people (especially the innocent bystanders) to start making edits, and to stop people from reverting at all.

The procedure is considered successful when everyone has stopped reverting. :-)

--Kim Bruning (talk) 23:31, 21 August 2009 (UTC)

I have no problems with that... as long as it is clear that "discuss first" is one of those methods, and if people ask you to discuss first, please do. Blueboar (talk) 20:17, 21 August 2009 (UTC)

I see no purpose to Kim's proposed additions in this section, since it detracts from what I think is the main purpose (namely, "Fair warning: if you make changes to policy pages without discussion, other editors might get mad at you").

Kim, when you have people fuss at you about following procedures solely for the sake of following procedures, then you should point them at WP:BURO and be done with it. I, too, have rarely encountered a person that reverts something that they strongly agree with because of a mistaken impression that Wikipedia is a bureaucracy; I have found that correcting this misapprehension is a simple and straightforward task. Since these are rare, I don't see any need whatsoever to repeat "Do not mindlessly follow rules" all over this page. Also, based on my experience of dealing with you, I expect that a fairly high proportion of people that revert your changes outside the main namespace nominally for "not following the procedure" are actually reverting your changes because they disagree with their substance. An edit summary of "please discuss first" might mean "I don't understand WP:BURO", but it more often means, "I can't believe you expect us to agree to this obviously dreadful idea, but I'll give you a chance to make your case on the talk page." WhatamIdoing (talk) 21:10, 21 August 2009 (UTC)

BRD should apply to policies as well as to articles. Nobody actually seems to have come up with a reason why it shouldn't, apart from that it won't always work. The fact that it isn;t allowed can sometimes actually be really harmful. A dispute with WP:PLOT has been allowed to fester for far too long because every time people attempt to edit the policy to reach a consensus the page gets protected and people are told to discuss any changes first. The discussion becomes incredibly lengthy and heated, and nothing much ever changes except people's frustrations. We should be allowed to try and amend policy through the wiki process. That is, after all, how the pages were written in the first place. There is no reason to mandate a discussion first approach, because each situation which requires discussion will be discussed because of the nature of wiki editing. However, in instances where discussion is not needed, allowing admins to curtail the wiki-editing process is harmful to wikipedia. The drafts above are not suitable, since they do not guide people to boldly edit a policy or a guideline page. Hiding T 21:29, 21 August 2009 (UTC)
- "The fact that it isn;t allowed"? Nowhere does policy (or proposals) say it (BRD) isn't! Nobody's "mandating" anything in policy (though practice has a tendency to work out that way).
- Your issues about discussion getting bogged down seem to have more to do with the discussion process than anything else. People can perfectly well go BRD-nuts on draft pages. And if we could find ways to improve discussions, that would help. (Some of it is just having too many people saying similar things, so it goes round in circles rather than progressing.)
- " in instances where discussion is not needed," - huh? That can only be true of uncontroversial edits, which isn't the situation you're alluding to (WP:NOTPLOT)!
- AFAIK a lot of the problem with PLOT is that it was introduced without sufficient discussion to demonstrate consensus,(!) which made it a lot harder later on for people opposing it to accept that subsequent discussion should require a consensus to remove, rather than consensus to retain. Rd232 ^talk 22:15, 21 August 2009 (UTC)

I must be missing the argument here if you are saying policies are open to editing. Although why your draft fails to mention that I find strange, if it is the case that one does not need to post to talk first. Having checked our policy on policies, Wikipedia:Policies and guidelines, I find that it appears to be correct that bold editing of policies is allowed, therefore I fail to understand why your draft fails to make mention of that if it is to adequately summarise existing policies.

"People can perfectly well go BRD-nuts on draft pages." You are aware that this is a wiki, aren't you? That means every page is actually a draft, policies included. We don't need to make another draft, we have one right in front of us.

You need to revise your understanding of the problems with PLOT. Although I'd be curious how an editor would know when there had been sufficient discussion. ;) At the time it was added, it seemed there had been sufficient discussion, and it seemed like the consensus was too add, which was perhaps born out by the fact that no-one immediately reverted it. PLOT actually demonstrates the flaw in your reasoning. Had Plot been added as a bold move rather than proposed to talk first, it is likely it never would have got added. And perversely, there has been a large discussion which showed a consensus to remove PLOT, but it appears impossible to act upon that consensus because those wishing it to stay do not accept that there was sufficient discussion, even after over 200 people commented. Were we allowed to be bold and edit the policy without the page being protected because "changes to policy pages need discussion", we would not be in the current situation. That editors are not allowed to edit policy pages, and that editing policy pages is discouraged has actually, demonstrably, harmed wikipedia. I'm curious, are you aware of how policies are created? Hiding T 22:55, 21 August 2009 (UTC)

OK, I don't know enough about PLOT, mostly because I don't care. (Can't care about everything.)
"We don't need to make another draft, we have one right in front of us". Sorry, but that's somewhere between anarchy and naivete. Hey, let's change WP:NPOV, I think it should read Northern Point Of View! What the hell, it's only a draft, right? Seriously, you can call it a draft in one sense, because it's always subject to revision, but in another it is a published operational document in current use, and therefore not to be messed with as if it were merely a pre-publication draft. Rd232 ^talk 23:37, 21 August 2009 (UTC)
The draft says " changes that would alter the substance of policy or guidelines should normally be announced on the appropriate talk page first. " That seems substantially equivalent to Wikipedia:Policies and guidelines' "Talk page discussion typically, but not necessarily, precedes substantive changes to policy." it can be tweaked, or the identical wording adopted - fine by me.
The draft doesn't mention Wikipedia:Policies and guidelines' injunction to bold editors to stick to 0RR or 1RR, which is the only thing it has to say about bold editing (it certainly doesn't explicitly permit or encourage bold editing). Maybe the draft should mention that, but it doesn't seem necessary to me.
You say "Were we allowed to be bold and edit the policy" - who's we? Who's not allowing it? Inability to achieve consensus with large numbers of people with diametrically opposed views seems to be of very little relevance to the question of whether individual editors should be encouraged to discuss substantial edits first (i.e. get some indication that it may be acceptable to the rest of the community before changing policy that describes how the community should act). Bold editing as part of a substantial ongoing discussion to help move things forward is different, maybe worth stating explicitly. But that would be new - current policy doesn't say that. Rd232 ^talk 23:37, 21 August 2009 (UTC)

Considering that we know this text is controversial, the people hoping to add a section need to get consensus for those changes first. This is how things have always worked. Some editor going to a new section and rewording it to add new claims that are not supported by consensus and then claim consensus is needed to change his/her preferred wording is just obnoxious wikilawyering. We cannot add it until we have a consensus about what it should say, and we especially cannot add stuff making claims that fly in the face of longstanding practices (like claims that you do not need to discuss before you change policies).

This is the Consensus POLICY, so you'd think the people editing it would try to, I don't know, follow consensus policy in editing it? Instead we seem to have the same old individuals well known for deciding they can just up and change policies whenever they want. DreamGuy (talk) 19:37, 22 August 2009 (UTC)

cleanup/ update

Dang, this page needs cleanup again... Done a little adding of information. Am I doubling up information, or has the information eroded out again somehow? <puzzeled> --Kim Bruning (talk) 01:57, 21 August 2009 (UTC) ... or was it never in there? If the latter... I may have a book to write. ;-) --Kim Bruning (talk) 02:03, 21 August 2009 (UTC)

Hmm... I added "naturally, you do need to know what that consensus is". But this page doesn't really do a good job of telling people how to research and figure out what the current consensus is on a matter, nor how to test it. Sounds like I have a new project for the near future. ^^;; --Kim Bruning (talk) 18:14, 21 August 2009 (UTC)

Three good ways... first, ask on the talk page. Second, ask on the various related noticeboards. Third, ask at places like the Village Pump. OOPS... that means discussing first, doesn't it? Blueboar (talk) 18:35, 21 August 2009 (UTC)

Sure, no problem!

My main issue is with some people who seem to think that it is ok to enforce just their one particular way of doing things. Among other things, it distorts certain tests for consensus.

If someone feels somewhat nervous, and/or would like to build up a better picture of consensus before proceeding, they should certainly discuss things first. Alright? :-)

At the same time, we also have experienced editors who are simply current on consensus on $practice because they're doing $practice every day; we have people who research back archives; we have people who do numeric analysis; people who conduct polls; people who first establish consensus at foundation level; people who lead by example and then document when successful; people who discover that the description of a particular procedure doesn't work in practice and would like to make a note of that; and etc. There's more than one way to do it. (I do have preferences for some of these above others though. Some of those preferences can be objectively motivated

--Kim Bruning (talk) 18:54, 21 August 2009 (UTC)

The larger the community gets, and the more established policy pages are, the less likely the average editor can substantially improve policy without any prior discussion. I think some of your way of thinking may be due to having been around when WP was much, much smaller and much more changeable, policywise. Rd232 ^talk 18:59, 21 August 2009 (UTC)

You can mitigate community size effectively by limiting scope. The remaining factors are number of experienced editors present, and level of assumption of good faith. If you can slowly grow scope, while maintaining that all present act in good faith, then you can still be effective, no matter the community size.

Are you saying the average editor doesn't know how to do this? (am I an average editor? ^^;;) Maybe we ought to write some more documentation on how to do this kind of thing. --Kim Bruning (talk) 19:16, 21 August 2009 (UTC)

Kim, you are far from an average editor. And I mean that both positively and negatively. Blueboar (talk) 19:58, 21 August 2009 (UTC)

"limit scope"? I don't know what you mean and the link leaves me none the wiser. Let's put this another way: policy is (a) descriptive, following practice, in a Brownian motion sort of way; and (b) prescriptive, telling people who don't know the practice what it is; and (c) transformative - trying to change practice. The larger and more established the community, the slower practice is to change, and the less rapidly policy needs to change to accommodate A (i.e. match changing practice). Under heading B it also needs to change less often, as over time policy gets written up pretty well reflecting existing practice and instructing how to follow it. Finally, C becomes ever harder, due to coordination problems, which create a more complex dynamic between policy and practice (eg if policy is changed reflecting local consensus but not global). Combine that with the ability for editors to introduce instability, confusion and error (policy deviating from consensus and/or practice), and the larger and more established the community, the more reluctant editors should be to make substantive changes to policy without prior discussion. Rd232 ^talk 20:22, 21 August 2009 (UTC)

If you don't understand how to limit, expand, or control scope, you may be vulnerable certain political tactics. I certainly recommend taking the time to read meatball on those topics sometime! In the mean time, I'll drop that line of discussion, as a discussion of those topics is too long to fit in this margin. ;-)

Let's try to see if I can show you the way I'm thinking somehow...

Do you know experimental statistics? Given the task to sample noisy data, and given the task to fit a function to that data; should you: A. Take as many samples as possible, B. Take as few samples as possible? What are the upsides and downsides to each? What yields the most accurate fit?

--Kim Bruning (talk) 21:29, 21 August 2009 (UTC)

Yes (albeit a bit rusty). What's the population you're sampling? Edits? Editors? And what's the point you're getting at? Rd232 ^talk 22:01, 21 August 2009 (UTC)

I was going for an abstract data-set to start with, so we can clarify some theory, before heading back to the messy real world. Is that ok with you?

If you prefer something solid to think about by analogy instead, how about sound-sampling from a piece of music, for instance? In that case, is more samples a good thing, or a bad thing?

Either way, I'm thinking about how sample size or sample rate (=same thing as sample size, but now in time) affect perception of the underlying data. --Kim Bruning (talk) 22:35, 21 August 2009 (UTC)

I'm more used to statistics, like population or GDP. Obviously a larger sample is better, though there are diminishing returns. And if the underlying population is changing, more frequent samples are better, though again with diminishing returns. Now what's your point already? Rd232 ^talk 22:39, 21 August 2009 (UTC)

What happens if we consider edits to a policy page to be samples of community consensus? If you look at it that way, would it be better to have more edits or less edits to policy pages per unit of time. (all else being equal)?

Do you think this is an adequate way of looking at policy pages? Why? Why not? --Kim Bruning (talk) 23:41, 21 August 2009 (UTC)

Not, because, obviously, edits are not the only or even the best way to test consensus. WP:TALKPAGE. Rd232 ^talk 10:14, 22 August 2009 (UTC)

Ok, then that is a clear difference of opinion between us. For me, edits are the final, ultimate, buck stops here test of consensus. All else is gravy. The reason for this opinion is because this is a wiki, and the content of the wiki is defined by the edits to it.

If a group of people states that they want to be in Washington, but I'm meeting them in New York, then my conclusion at the end of the day is that they actually ended up in New York at that point in time.

If the "community consensus" on a talk page says that a page should say X, but the page says Y, then my conclusion is that the consensus is actually Y at this point in time. --Kim Bruning (talk) 15:17, 22 August 2009 (UTC)

difference between quality and strength

What's the difference between the quality and strength of an argument? SlimVirgin ^{talk|contribs} 00:32, 23 August 2009 (UTC)

strong argument: "DO THIS NOW OR I WILL BLOCK YOU!" ... that's a pretty strong argument, I'm going to listen to that! ;-)
good quality argument: "I think that if we do it this way, we'll have the following advantages <insert advantages>, because of <insert logical reasons>" ... ok, that's logical, I think we should probably do that.
strong, good quality argument: "I found a copy of this picture at location x,y,z, and apparently it is copyright. we need to adhere to copyright law, so this picture needs to be deleted" Ok, that's logical, and I don't want to break the law, so I'd best listen to this.

--Kim Bruning (talk) 00:37, 23 August 2009 (UTC)

That's not what it means in English though. There is no difference between the strength and quality of an argument. The thing about this policy is that it's incredibly wordy: words are just being thrown in whether they add to the meaning or not, to the point where it's hard to get through it, but when I try to tighten, I'm reverted. That was just one example by the way. There are dozens of others. SlimVirgin ^{talk|contribs} 00:41, 23 August 2009 (UTC)

I didn't revert you, I'm sure. If I have to choose, I'd say drop "strength" and use just "quality". Will that do? --Kim Bruning (talk) 00:43, 23 August 2009 (UTC)

You did revert. [1] I don't mind which, but strength is better. My point is that you have to allow copy edits to be made. It's not just that one example. SlimVirgin ^{talk|contribs} 00:45, 23 August 2009 (UTC)

Nah! Compare it with what was there before you tightend. Some of your tightening changed the emphasis a bit, so I loosened those somewhat again, and tried to rearrange to get it as short as possible.

I've used some of your suggestions to tighten back up a bit again now, but imho, more accurate :-) . See if you can do some more? :-)

--Kim Bruning (talk) 00:51, 23 August 2009 (UTC)

The OWNership comment

(+ [diff for reference) --Kim Bruning (talk) 19:39, 21 August 2009 (UTC)

Kim, I take great exception to your recent addition (now removed) implying that anyone and everyone who prefers to discuss changes before they are added to policy pages is doing so because of OWNership. What happened to WP:Assume Good Faith? Most of the editors who prefer a discuss first approach are not doing so because of ownership, but because thats how they think policy pages should opperate. I also have to say that justifying that addition by claiming that you are adding NPOV shows that you don't actually understand that policy. Blueboar (talk) 18:32, 21 August 2009 (UTC)

I don't tend to accuse people, and I do assume good faith, but I don't shy away from naming the conflict, as part of an effort to resolve it.

To get back on topic: as long as people would like to simply discuss a page first, that's fine. It's when they start reverting people on those grounds that they start getting into ownership territory. MIPs and Primary Editors are strongly correlated. :-P

My edit was just a first pass at trying to improve the text, feel free to edit further, and let's see where we end up!

Finally, please don't use ad hominem in discussions on wikipedia, it is a form of logical fallacy, and it's not very productive. --Kim Bruning (talk) 19:08, 21 August 2009 (UTC)

where was the ad hominem?

I am not sure where you draw the line here: "as long as people would like to simply discuss a page first, that's fine. It's when they start reverting people on those grounds that they start getting into ownership territory"... Someone makes an edit to a policy page that you don't think reflects consensus... what is wrong with reverting and asking for discussion? Blueboar (talk) 19:22, 21 August 2009 (UTC)

That's an Excellent question, let's discuss that!

Maybe you already know (part of) the answer yourself. Here's one question back for starters: How do you know that the edit does not reflect consensus? --Kim Bruning (talk) 19:26, 21 August 2009 (UTC)

Well, that depends on the change... sometimes you know from experience... if someone changes WP:V to say that there is no need for information to be verifiable to reliable sources, I know from years of editing on Wikipedia that this is changing a fundamental principle that has wide acceptance. It is highly unlikely that the consensus has changed on this.

But often I don't know for sure ... but I don't think it does. That's WHY I want to discuss it first... to get a sense of the community... to get input from all the people who follow the policy talk page (and others if needed), to find out whether the community consensus has change or not.

In either case, what is wrong with asking for discussion? Blueboar (talk) 19:49, 21 August 2009 (UTC)

Ok, so next, checking to see if we're in the same universe ;-) : A revert is an edit, right? (it certainly seems to show up in the page history as an edit)

--Kim Bruning (talk) 20:00, 21 August 2009 (UTC)

Yes, technically it is. In practical terms, I would say it is a specific type of edit and many people don't consider a revert to be a "real" edit, they see it more as an "unedit"... but... yes. Blueboar (talk) 20:03, 21 August 2009 (UTC)

It's good thing I checked then: I consider a revert to be a normal edit, for all practical purposes that count; both in terms of how it is handled by mediawiki internally (how it is stored in the database), as well as in terms of how it should be treated with regards to Consensus.

So I guess I'd better ask this next question too:

An edit must reflect consensus, right? Or at least, an edit must strive to reflect consensus, or at the very least be an attempt to achieve it. Is that correct? Or does your position differ subtly here too?

--Kim Bruning (talk) 20:09, 21 August 2009 (UTC)

No, I fully agree with that (is this leading to answering my question?). Blueboar (talk) 20:12, 21 August 2009 (UTC)

Ok, based on the premisses you agreed on (+ some trivial extra):

(1). If someone makes an edit to a page, consensus will lie somewhere between the previous version of the page (or at least the previous good faith aka non-vandalized version), and the new version. This means that a straight revert is not likely to reflect consensus. Since all edits must reflect consensus, a revert is probably the worst edit you can still make in good faith. (and sometimes not even that, if you do it 3 times, we assume bad faith and block you for 24 hours ;-) ). This is also why older wikipedians consider reverts to be rude.

(2) You need to be able to explain all edits you make (this is a minimal requirement of course. You can't reach consensus if you are unwilling to ultimately defend your actions)

(3) Explaining a revert by "has no consensus" is a trivial statement. We already knew it didn't have consensus. But *why* doesn't it have consensus. You still haven't explained that. Therefore...

(4) You cannot meet the criterion at (2) by reverting with as reason "does not have consensus"

Is there a flaw in the logic so far?

Incidentally, while I don't really like reverting, I do agree that it is sometimes ok to do so. In those cases you need to clearly state "why" something doesn't meet consensus. This reasoning cannot be that you didn't see the consensus or whatever (ignorance is no excuse ;-) ). You can't read other people's minds, and you don't even know who has a page watchlisted. Maybe there are 100 people lurking who all think the edit was a great idea, and you just went and slowed down their work! Sure, maybe not, but there's no way to know.

What you can do is know what is in your own mind, and you can know your own opinion. If you disagree with an edit, then you may certainly edit it and correct it, or you may revert it, if you see no option to do the former.

Does this make sense so far? Or is there some portion you disagree with or that seems incorrect? --Kim Bruning (talk) 20:43, 21 August 2009 (UTC)

(undent) Nope... you start off with a faulty assumption... you say: "consensus will lie somewhere between the previous version of the page (or at least the previous good faith aka non-vandalized version), and the new version" but that is not always the case... consensus might lie somewhere between the two edits... but it also might lie solidly with the previous version. Or (less likely on policy pages) solidly with the changed version. Remember that Policy is not determined by consensus between two people... it is determined by the consensus of an entire community... that's thousands of users. Granted, it isn't always easy to determine what that consensus is, but the only way to determine what a large group of people think is to discuss it with them (or at least as many as you can). Blueboar (talk) 21:06, 21 August 2009 (UTC)

In what cases is it true that consensus lies with the previous version, and in what cases is it true that it lies with the changed version. Is this a binary proposition, or are those two extremes of some continuous function? How can you determine what the consensus is unambiguously, in the absence of further discussion? Is that possible at all? --Kim Bruning (talk) 21:18, 21 August 2009 (UTC)

How can you determine what the consensus is unambiguously, in the absence of further discussion? Is that possible at all? It isn't possible... that's the entire point here. Major changes to policy pages need lots and lots of discussion to determine what the community consensus IS.

I suppose you can argue that, both the editor who made the change and the editor who reverted acted without demonstrating that either edit has consensus... which is why it is best to discuss the issue before any edit is made... you determine (as best you can) what the consensus is through discussion and then you edit accordingly. We forgive the error on the part of the reverter, since he is reverting to a version that at least has the presumption of consensus. Blueboar (talk) 21:43, 21 August 2009 (UTC)

(ec) I would not simply forgive either party.

They both have some explaining to do.

Is that a reasonable demand to make? --Kim Bruning (talk) 22:16, 21 August 2009 (UTC)

"How can you determine what the consensus is unambiguously, in the absence of further discussion? Is that possible at all? It isn't possible... that's the entire point here." Point of order, you appear to be incorrect, since it is possible, and that is the entire point here. You can determine the consensus unambiguously as being the point at which the page stabilises. A lot of people seem to miss the trick that ultimately, even in edit war scenarios, as long as people respect WP:3RR, and there are processes in place to ensure they do, editing will solve the issue of where consensus lies, because consensus will lie with the last standing edit. If 49 people feel one line of text should be added, and 50 people feel that line should be removed, and assuming for the sake of argument we ban discussion and allow each person one edit to the page per day, how do you think this situation will play out? Which way will the page look at the end of the editing? There really is actually very little need for discussion if parties are prepared to engage with each other in the spirit of our policies. And we should not frame our policies to reflect anything other than best practise, for fear that best practise becomes impossible.

"Major changes to policy pages need lots and lots of discussion to determine what the community consensus IS." No they do not. Major changes to policy need to be floated to determine consensus. They don't have to be discussed. Have you heard of the concept of "kite flying"? What damages the wiki process on occasion is reversion based on "you can't do that". Such reversion is based on a misapprehension to start with, namely that we can ignore rules. We don't need a rule telling us that we cannot employ a method that may actually work, and when it does, is best practise. The correct thing to say is that "Major changes to policy need to reflect consensus." Determining consensus does not have to involve discussion. I don;t see why we need to mandate that it does. Hiding T 22:12, 21 August 2009 (UTC)

"Determining consensus does not have to involve discussion". If you place zero weight on the value of policy stability, that's fine. Do you? (Because at the limit, policy then ends up a random walk, buffeted by a thousand bold editors changing something every five minutes...) If not - if you place some value on policy stability, great or small - then the less likely (in the editor's own judgement) that the edit matches existing consensus, the more likely they should be to discuss first, to test and possibly change consensus. That's the proposition being made. Rd232 ^talk 10:20, 22 August 2009 (UTC)

I think you are making the assumption that discussion is the only way to determine a stable consensus.

I think you also assume that bold edits are the same as random edits. As per policy,a good faith edit must have the intent of moving the page towards some (new or old) consensus, that is a requirement. It can certainly fall short or overshoot that objective, but the intent is there.

If the editor does not believe that either

their edit either matches consensus and/or
Their edit will reach or obtain a consensus over time,

then they are editing in bad faith. In that case, I will provide personal blessing and assistence in administering meatball:ParkingLotTreatment or some other adequate meatball:SignificantEmotionalEvent of your choosing.

If people edit in good faith, their edits will tend to converge on consensus over time. The more edits are made, the sooner consensus is reached.

The more edits per unit of time, the better the page will stabilize around consensus. (Think normal distribution. If you increase the sample size, the standard deviation tends to become smaller.)

--Kim Bruning (talk) 14:25, 22 August 2009 (UTC)

breaking

(from previous)

I can't see how you can argue that a revert to the status quo ante (where this exists, i.e. where there is a reasonably stable version) is the same as deviating from the status quo in one of many possible directions. Rd232 ^talk 20:25, 21 August 2009 (UTC)

Hmmm! :-) Define "Status quo ante", in a manner that is replicable in such a way that any two random people will always agree on what that is on any given page. --Kim Bruning (talk) 20:45, 21 August 2009 (UTC)

the version that existed on the page prior to the change (from the latin... Status quo anti = situation which was before). Blueboar (talk) 21:17, 21 August 2009 (UTC)

Ok, so in practice, say I am now checking a history of a page, so I click on the revision that is one below topmost, and then click edit, and then click save? And this is a revert to status quo ante?

Is that always permissible? Are there any traps to watch out for? --Kim Bruning (talk) 21:32, 21 August 2009 (UTC)

It is always permissible to revert good-faith non-improvements so long as (1) you personally think that the previous version is better than the newest version, (2) there's no indication that any editor other than the person making the most recent change disagrees with your personal assessment of the two versions, and (3) you stop there, instead of endlessly WP:Edit warring to keep your preferred version in place. It is polite and helpful to explain why you like yesterday's version better than today's version. WhatamIdoing (talk) 21:41, 21 August 2009 (UTC)

Due to reasons, I occasionally encounter the situation where (1) the person has no personal opinion as to the suitability of the edit, (2) there is 1 other editor who agrees that the most recent change is an improvement. (3) the person edit wars.

In that situation, is such a person in the right? What should be done about the situation? --Kim Bruning (talk) 22:24, 21 August 2009 (UTC)

How exactly does a single editor conduct an edit war? If you stop with a single reversion, then what's the other editor doing? Repeatedly adding null edits to really, really, really make the page exactly the same as it was before your change? WhatamIdoing (talk) 20:35, 24 August 2009 (UTC)

Kim's "logical" argument is clearly flawed. For example:

(1) It makes weird and unsupportable assumptions: If you have a strange idea, and dozens of editors reject it outright, then the consensus is for "no strange idea, thanks", not for "let's have 4% of this strange idea, since one editor likes it, and two dozen hate it". It also doesn't allow for the possibility that the editor proposing the change will ultimately reject the idea (e.g., after becoming more educated on the subject). For example: An editor doesn't see a relevant statement on a policy page, and adds one. Someone else reverts it as "already addressed in #This section". This revert improves the page, and the first editor is unlikely to insist on keeping duplicate statements.

Furthermore, there are times when I know, without a doubt, that a change has no consensus. For example: re-numbering WP:ELNO has no consensus. It's not actually necessary to "discuss" whether it's a bad idea to screw up thousands of references to that page in several years of previous edit summaries and talk page discussions.

(2) Kim wrongly conflates "being able to do something" and "having already done it". I suspect Kim makes this error because of confusion between "determining WP:Consensus" and "WP:Dispute resolution". If I revert a change to a policy or guideline, I'm capable of explaining why, and happy to do so at length on request, but I'm not behaving badly if I remove good-faith errors before explaining why they're errors, nor am I behaving badly if I don't instantly start a discussion about why it was an error, especially for something that seems patently obvious to me. WP:IAR and WP:BURO apply to wasting editors' time with unnecessary discussions, you know.

(3) is trivial: If you disagree with a change enough to revert it, then that change does not have consensus by definition. Explaining why you disagree is critical to resolving disputes, but entirely unnecessary to determining whether consensus for a change exists.

(4) The "conclusion" is invalid, because the assumptions are invalid and/or far too limited, but again, this is a conflation of "figuring out whether this change is agreed upon" (usually a trivial issue) and "figuring out what to do next" (sometimes a very complicated task). WhatamIdoing (talk) 21:41, 21 August 2009 (UTC)

I'm glad your reply edit-conflicted me, because it was way more helpful :) Rd232 ^talk 21:44, 21 August 2009 (UTC)

I disagree with WhatamIdoing's point (3) If it has been shown through discussion that dozens of editors do approve of a change... it does has consensus (or at least a strong indication of it), as you point out in your first point your lone opposition is not enough to challenge this consensus. Otherwise you are spot on. Blueboar (talk) 21:52, 21 August 2009 (UTC)

True that. Consensus != unanimity. Rd232 ^talk 21:58, 21 August 2009 (UTC)

Would you like to deprecate WP:CCC? This allows a single editor to challenge current consensus.

Without WP:CCC, say 5 people form a local consensus on an issue. How can the rest of the community enforce their position instead? Any single editor from the community could be reverted by the gang-of-5. ;-) --Kim Bruning (talk) 22:02, 21 August 2009 (UTC)

How do you get from "Consensus != unanimity" to deprecating "consensus can change"? The only way I can see that making sense is if there were no such thing as a talk page. The issue of local v global consensus .... no I don't see how you can get there. Rd232 ^talk 22:21, 21 August 2009 (UTC)

Let me expand the example scenario:

There are 5 people who have formed a local consensus, and are feeling a tad WP:OWN-ish.
An editor (1 person) comes along and tries to discuss. They ignore him.
The editor makes an edit. They revert him
"lone opposition is not enough to challenge this consensus", so the editor has no recourse, and leaves. This is Blueboars statement from earlier in this thread (Kim, please do not take comments out of context... what I said was: If it has been shown through discussion that dozens of editors do approve of a change... it does has consensus (or at least a strong indication of it), as you point out in your first point your lone opposition is not enough to challenge this consensus.)
Another 1 editor comes along and tries to discuss. They ignore him and/or revert him. wash, rinse, repeat from step 3.

Many editors will have come by, but none can actually prevent the 5 editors from owning the page, provided we accept Blueboar's statement.

The initial situation is not atypical. As it stands, you could normally apply WP:CCC+WP:BRD to break the hold of the gang-of-5. --Kim Bruning (talk) 22:57, 21 August 2009 (UTC)

Yes, if WP:CCC and WP:BRD were magic wands instead of three-letter shortcuts, that analysis would make sense. In practice, if the 5 aren't willing to discuss when asked nicely, they're also going to ignore CCC when quoted at them and BRD behaviour when used and just revert with little or no discussion. That's the dynamic you've postulated, and neither CCC nor BRD do anything to change that. Now in the real world, the key to breaking such OWNership is discussion, possibly by persuasion, probably by bringing in other editors by whatever means - RFC, whatever. You're confusing an issue of discuss first v edit first with local v global consensus (again). To build on your gang of 5 example though, here's a way discussion (with or without editing) is better than editing alone: that way at least every time a lone editor comes by and comments on talk (but there's no substantial discussion and bold edits are reverted), there's a record of it that others may see and be encouraged by. If they merely boldly edit and then sod off without discussion when they're reverted, that disappears much more easily in the edit history than in the talk archive. Rd232 ^talk 23:50, 21 August 2009 (UTC)

(ec)

Re blueboar inline: Ah ok. I took an example with 5 people, partially because I'd actually encountered that case listed on WP:AN once, and managed to solve it at the time. If you have dozens of people, it starts to get really really hard to change the page indeed. I sometimes worry about what will happen once real-world astroturfers figure out that vulnerability. It would probably take some heavy IAR-ing to fix. --Kim Bruning (talk) 23:52, 21 August 2009 (UTC)

Re Rd232: At the time, WP:CCC protected me from being blocked by admins all too easily, so I had time to apply WP:BRD to ascertain whether the problem was real (it was), and to get some idea as to who the 'ringleaders' were.

I could then immediately talk with them, and IIRC they were somewhat dismissive, but I can be fairly persuasive, and having WP:CCC to refer to was better than not having it at all, of course ;-).

Several of the things you suggested were also done, including one or two (other) admins helping out. I hope I've said it before, but you are not wrong in the things that you DO do.

It's just that there are more things that you can also do. More than you thought you could! :-)

--Kim Bruning (talk) 00:05, 22 August 2009 (UTC)

(ec), @WhatamIdoing I think we're actually not to far from agreement.

(1) You're right, I forgot to qualify this properly. It only holds if the edit is reasonable and made in good faith. I'm not sure if that qualification rules out all exceptions yet. Am I missing any? If not, are we in agreement? If I did miss some, I'll probably agree with anything reasonable, and then we'll be in agreement, afaict.

(2) only requires that you be capable, and that you respond if challenged. We're fully in agreement here, afaict.

(3) we agree this is trivial

(4) If (3) is trivially true, and you are challenged as per (2), then if you answer the challenge with "the edit did not have consensus", you have answered with a trivial truth. Is that logically solid enough?

Now the next question is,

if I'm a mediator, or admin, and I actually issue the challenge as per (2), should I accept a trivial truth as an answer? Will this allow me to resolve the dispute in any way?

If I am an ordinary user, and I issue the challenge as per 2, sould I accept a trivial truth as an answer? Will this allow me to resolve the dispute in any way?

--Kim Bruning (talk) 22:02, 21 August 2009 (UTC)

Any chance you could discuss this like a normal person instead of a failed Turing test? This numbered conversation about logic needs to be corralled somewhere it can't hurt innocent bystanders. Rd232 ^talk 22:27, 21 August 2009 (UTC)

Oh no! I've been made! I can no longer hide the fact that I dream of electric sheep!

--Kim Bruning (talk) 23:04, 21 August 2009 (UTC)

Kim... you never answered my question (or if you did, it is burried where I could not find it) so let me ask again... If someone makes an edit to a policy page that you don't think reflects consensus... do you think it is wrong to revert and ask for discussion? (please do not answer by asking another question) Blueboar (talk) 22:37, 21 August 2009 (UTC)

I'm not Kim, but my answer is that it rather depends on why the edit was reverted, doesn't it. I've seen edits reverted and then a message posted to talk along the lines of "while I agree with the change, I'm not sure such an edit should be made without discussion." Ive always found that sort of reversion incorrect. Hiding T 22:59, 21 August 2009 (UTC)

Please read the question again... I say why the edit was reverted. Blueboar (talk) 23:04, 21 August 2009 (UTC)

Actually, part of what I'm discussing in this section is that maybe you think you're saying why the edit was reverted, but in reality you're saying something that has no real meaning. :-/

Here's a straight answer for you, in 2 bullet points:

* While it is not always a nice thing to do, it is generally accepted that you can revert people and ask them to discuss things, provided you meet the 3 criteria mentioned by WhatAmIDoing: (1) you personally think that the previous version is better than the newest version, (2) there's no indication that any editor other than the person making the most recent change disagrees with your personal assessment of the two versions, and (3) you stop there, instead of endlessly WP:Edit warring to keep your preferred version in place.

* I think it is wrong to revert someone merely for the sake of asking them to discuss on talk.

--Kim Bruning (talk) 23:14, 21 August 2009 (UTC)

"I think it is wrong to revert someone merely for the sake of asking them to discuss on talk." Logic (and practice, as I'm sure you know) disagree with you. Substantive edits made without discussion on talk can slip by without people noticing, especially if followed rapidly by minor edits. So reverting to discuss ensures wider exposure to what may be a bad idea. Again, in an article it might be better to just leave it unreverted and start a discussion. But policy is always a working document, and so reversion may be appropriate for substantive undiscussed changes - even if the rverter agrees with the changes (but recognises that he might change his mind or others might disagree). To give an analogy: do web developer teams in large companies develop on (a) a testing server or (b) the live production server? Yes, it's b, because they may have a great idea, but they may also have a terrible idea, or destroy things, or whatever. Hence the testing server - the equivalent, in this analogy, of the talk page. Rd232 ^talk 00:01, 22 August 2009 (UTC)

But wiki-editors in large companies edit on the production server, and they tend to not use talk pages at all. Wiki's are not the same as regular websites, somehow. --Kim Bruning (talk) 00:10, 22 August 2009 (UTC)

Hmmm, I suspect that this is due to the presence of revision control (aka. "the history tab"), which allows you to have an infinite number of "testing servers".

--Kim Bruning (talk) 00:30, 22 August 2009 (UTC)

OK, now I'm sure I've wandered into a Turing test. Pff, I've got better things to do. Unwatching. Rd232 ^talk 00:36, 22 August 2009 (UTC)

<undent> Let me step back and review the background. Policy pages are highly visible and highly influential pages, so edits cannot be done lightly. Therefore it is not only the reverter's obligation to explain their revert; it is the initial editor who also must clearly explain their intentions, preferably in the talk page. At the same time, the reverter my clearly state the true reason of the objection. The reason "no consensus" is just as bad as "IDONTLIKEIT". How the first editor can defend their edit if they don't know the factual objection? In other words, you cannot put the whole blame on one side in the "edit-revert" pair. There is endless number of scenarios. My guess is that we revert mostly occasional editors who peek into page want to improve something, but don't know the underlying history and a long road to the trade-off. Therefore I am inclined to think that a revert is not a big drama. If the change is non-critical, it may wait. If someone is fixing a showstopper problem, say so in edit summary and start the discussion. If you are right indeed, people may come up with better remedies. - Altenmann >t 23:32, 21 August 2009 (UTC)

convenience break

I have realized that the various "sides" involved in this discussion are focussed on different things... For example, several people seem focused on the revert and what happens after it, while others have been focused on the bold change and what happens before it. I think we need to sync up here, so we are all discussing the same thing.

To me, the "discuss first" concept is all about what happens before a change is made. So let's discuss the pros and cons of discussing an edit before you make it. First pro... you can determine if the desired change is totally unacceptable (ie clearly and obviously does not reflect consensus), is worth discussing (either because we are not clear as to what consensus is, or because the person made a good argument that might change consensus), or should be implimented (ie clearly does reflect consensus). Thoughts? Blueboar (talk) 00:12, 22 August 2009 (UTC)

I'd like to re-iterate that I have no trouble with people who would like discuss things before they edit the wiki, as long as they don't force me to follow their approach (which I consider to be inefficient, and at times impossible). Live and let live, as it were. Is that ok? --Kim Bruning (talk) 00:16, 22 August 2009 (UTC) "some of my best friends discuss first!"

OK, but can you explain why do you think it is inefficient and at times impossible? Blueboar (talk) 00:29, 22 August 2009 (UTC)

Because it can take a very long time to get to an edit, and sometimes no edit happens at all. --Kim Bruning (talk) 12:44, 22 August 2009 (UTC)

OK... why is that a bad thing (and remember we are talking about Policy pages here, not articles)? Blueboar (talk) 12:55, 22 August 2009 (UTC)

I don't make the distinction, remember? All wiki-pages must be as accurate as possible. The one way to achieve anything on a wiki-page is to make an edit. No edit - no page. So in general, more edits is a good thing, less edits is a bad thing. The first rule of making a page more accurate, is it needs more editors making more edits.

Some people say that stability is important. Sure. If every edit moves a page somewhat closer to consensus, then edits will tend to converge on consensus and the page will stabilize at consensus.

Some people say that a (policy) page should be static and not change much. I think we ought to be indifferent towards static-ness on a particular page:

If the underlying consensus described by a policy page is currently static and not moving, then more edits to a page won't substantially change the semantic content of the page, except maybe to make it clearer.
If the underlying consensus is currently on the move, or uncertain, then the page should also be on the move, and/or represent the underlying uncertainty.

The intent of the policy page is to document consensus, after all. It must describe the current practical experience that people have, both on what things are wise to do, and on what things are not wise to do. (and remember, just because a lot of people do something, doesn't immediately make it a wise thing to do! -think eg. smoking- ;-) --Kim Bruning (talk) 14:07, 22 August 2009 (UTC)

Re: If the underlying consensus described by a policy page is currently static and not moving, then more edits to a page won't substantially change the semantic content of the page, except maybe to make it clearer. - this is the most rediculous comment I have seen in this entire disucssion, and seems to indicate that you no idea of what actually happening on most of our policy and guideline pages. Our core policies, such as WP:V and WP:NOR, enjoy a consensus that is very static and unmoving... and yet they are constently being edited in ways that change the semantic content of the page. Sometimes the change to the semantic content is intentional, sometimes it isn't. Blueboar (talk) 14:56, 22 August 2009 (UTC)

If the semantic content of the pages is constantly changing, one would expect the page to stop tracking community consensus rather quickly, correct? Do these pages still represent community consensus today? Why or why not?

If the pages still represent consensus today as they did in the past, then apparently the sum total of those semantic changes hasn't added up to much, has it?

Unless community consensus has changed.

Well, all of this is easily checked. How about looking at the page history for WP:NPOV, and see how much the page has really changed over the years?

If you are right, then due to the accumulation of all those semantically significant changes, what is understood to be NPOV today will be substantially different from what we understood it to be in -say- 2007.
If I'm right, the page may use different words, but the message will still be essentially the same as in 2007.

I've already checked the page for myself, and I stand by my position. Once you've checked for yourself, do you disagree? --Kim Bruning (talk) 15:11, 22 August 2009 (UTC)

Is there a consensus?

It's time for a reality check... would everyone please take a look at the current language in the Consensus on Policies and Guidelines section. I think this accurately reflects both practice and community consensus... Does anyone disagree? Blueboar (talk) 13:25, 22 August 2009 (UTC)

Still muddling with the wording a bit. [2] Will this do? "Expected" is a fairly strong word, and it can be confused with "required", imho. --Kim Bruning (talk) 14:39, 22 August 2009 (UTC)

fair enough... personally I think "expected" (and yes... that does have implications of "required") is the most accurate word as far as discribing what is actually happening on the policy/guideline pages. I realize that you disagree with the trend, but the simple fact is that editors are expected to discuss significant edits to policy and guideline pages before the edit is made. Whether this is a good thing or not is quite frankly irrelevant at this point. So is there a better word than "expected"? Blueboar (talk) 15:24, 22 August 2009 (UTC)

Policy describes current best practice, not current most common practice. It's a subtle difference, but somewhat important, because among other things (per reductio ad absurdum):

vandals could take over the wiki "new ip users are expected to vandalize the wiki"
new improvements to practice would never be documented "only pete and his 3 friends are using this new fangled way to fix 1000 pages per minute and make unicorns and rainbows. No one else is doing it, so it's not policy; let's revert it!"

--Kim Bruning (talk) 15:31, 22 August 2009 (UTC)

I see... I think this is coming down to... who determines what best practice actually is? Blueboar (talk) 15:45, 22 August 2009 (UTC)

That was never explicitly codified. It comes down to finding a consensus about the underlying logic and wisdom of each approach. Hence why I'm discussing at all; else I'd have added your exact text myself, you see. --Kim Bruning (talk) 16:12, 22 August 2009 (UTC)

I didn't ask how... I asked who? ... Jimbo? Admins? Arbcom? The community at large? Blueboar (talk) 16:26, 22 August 2009 (UTC)

It's a consensus system, until and unless you finally somehow manage to figure out how to bring it down on top of us dude.

So until then, the answer is you, me, and everyone else who is here or decides to join in! :-) --Kim Bruning (talk) 16:31, 22 August 2009 (UTC) Shall I invite Jimbo? <innocent look>

Am I correct in assuming that you mean this in a broad context... implying the entire community... and not limited to the contributors at WP:CONSENSUS?Feel free to invite Jimbo, or anyone else... I think we are getting into some fundamental issues here, and I would welcome his input. Blueboar (talk) 17:03, 22 August 2009 (UTC)

Errr, we're discussing WP:CONSENSUS, what did you think we were doing this whole time, if not discussing issues fundamental to wikipedia? ;-)

And lots of people have this page on their watchlist, if they want to fix it, they will. Anyone doing policy patrol will come across the page sooner or later. I'm pretty sure the entire community will input.

Just to be sure (at the risk of being WP:BEANSy): I'd be somewhat opposed to an ATT-like sudden expansion of scope, for some strange reason. --Kim Bruning (talk) 17:27, 22 August 2009 (UTC)

And there is part of the problem ... we are only partially discussing WP:CONSENSUS... we are also talking about every other policy and guideline on Wikipedia.

Do you think it is possible that a new best practice can be determined organically, through countless small discussions that do not take place on this page?... I do. I think that is what occured in this instance. I think the community by-passed a discussion on this page, reached a community wide consensus of what best practice on editing policies and guidelines should be, and now this page is having to catch up. Blueboar (talk) 18:00, 22 August 2009 (UTC)

I agree up to the point where you said "best practice". I assume you mean "common practice"? --Kim Bruning (talk) 18:10, 22 August 2009 (UTC)

No, I mean best. The community at large decided organically that the best practice is to discuss changes first, so they have implimented that practice... they didn't discuss it at WP:CONSENSUS, they just did it. Blueboar (talk) 18:20, 22 August 2009 (UTC)

A best practice is the most effective method, regardless of how many proponents it has. I'm pretty sure that "discuss first" is not the most effective method known to the community, even though it is commonly known by the community.

As a common method, I agree that it must be mentioned on the page. However, many people use other methods, those methods are also effective, and they are certainly not deprecated. We should not remove mention of other methods from the page.

Finally, just like your POV is not necessarily NPOV, your practice is not necessarily the best practice. Though of course it might be, which is why it is always important for people to listen to each other and learn from each other.

--Kim Bruning (talk) 18:31, 22 August 2009 (UTC)

You say: I'm pretty sure that "discuss first" is not the most effective method known to the community... well, based on practice the community at large thinks you are wrong. The community at large is pretty sure that "discuss first" is the most effective method known to them... that's why they they have implimented it.

You know, I think in many ways this is the reverse of what happened with ATT... with ATT, the policy wonks surprised the community by implementing something that the community was not prepared for. With discuss first, the community has surprised the the policy wonks with something the policy wonks were not prepared for. Well... guess what... the community will simply ignor the policy wonks if they don't get with the program and follow community consensus. Blueboar (talk) 19:10, 22 August 2009 (UTC)

Um. I'm not a policy wonk, I'm a huge advocate of describing best practice. Eg. Check out the recent history of WP:5P, to see how that went down not so long ago. You remember that "could never happen" fairytale situation where bold edits don't get reverted on policy? ;-) Well... see for yourself.

*AND* we've seen bold editing of several other policy pages, even as we were discussing them. Many of those are still there. Including bold edits to THIS page, may I remind you?

*AND* we see bold editing of policy pages every day. Just review page history. In fact, you admitted as much in a different thread.

So who is actually discussing objective, measurable, confirm-able reality here? ;-)

--Kim Bruning (talk) 19:56, 22 August 2009 (UTC)

If you are editing 5P... you definitely are a policy wonk... that's ok, I am too. Seriouly Kim, own up... when was the last time that you spent a day on wikipedia without working on a single policy page

As for the fact that there are bold edits here and on other pages... irrelevant... I have never said that being bold doesn't work (I have freely admited from the beginning that it does, just as you have admitted that the discuss first model does work)... all I have been saying is that the consensus out there in the real wiki-world (if that isn't an oxymoron) is that "discuss first" is indeed considered best practice (or at least the better practice) and because of this it has become expected. Blueboar (talk) 20:56, 22 August 2009 (UTC)

The problem with making it "expected" or "required" is that suddenly it becomes hard to edit or maintain policy.

I understand that people are doing that. I understand why people are doing it. But THAT practice is wrong.

Discussing is fine. Reverting and requiring discussion is wrong. People should not force methodologies on each other.

And ok, I have been editing policy a lot lately, and not getting around to other stuff. It started happening when I was mediating, and badly documented policies started making it harder and harder to mediate.

Then I started fixing them. Then I got to the point where I started losing sleep... http://xkcd.com/386/

--Kim Bruning (talk) 21:08, 22 August 2009 (UTC)

re: ...becomes hard to edit or maintain policy. I'll take this in two parts...

re: becomes hard to edit policy... I don't think that is a bad thing. Making significant changes to policy shouldn't be easy... as I think we agree, policy pages need to be somewhat stable and consistent. It is important to remember that hard to edit does not mean impossible to edit. Discuss first does not mean ONLY discuss... it means discuss then edit. It can even mean discuss, edit, discuss some more, edit some more, etc.

re: becomes hard to maintain... I disagree completely. It is far easier to maintain the policy if things go slow and there are not a lot of edits all at once. Blueboar (talk) 22:01, 22 August 2009 (UTC)

I think that deliberately slowing down editing makes a page unstable. I've written a couple of analogies and models on this talk page today to support that. See the analogy of the aeroplane pilot keeping the vehicle straight and level that I used to explain what I mean to Dreadstar. --Kim Bruning (talk) 22:07, 22 August 2009 (UTC)

Wow... our differences really are fundamental... I think deliberately slowing down editing makes a page more stable. And Wikipedia isn't an airplane. Blueboar (talk) 23:24, 22 August 2009 (UTC)

Indeed it is not. Though I wouldn't actually mind flying in a wikiplane (where all the passengers try to keep the plane in the air through consensus). People once did an experiment on that (simulator, of course). Everyone was given a controller, and the system averaged out the control inputs and sent them to the aircraft. After a couple of tries and getting the hang of things, the players actually kept the plane airborne! In that case, I think consensus was determined once every 1/25th of a second, or once every 1/50th, as those are typical numbers used by controller designers.

Could you imagine telling those people to discuss first, while the aircraft was hurtling towards a mountain top? ;-) --Kim Bruning (talk) 23:31, 22 August 2009 (UTC)

Can you imagine causing it to hurtle toward a mountain top by fiddling with the buttons while everyone was asking you to stop? SlimVirgin ^{talk|contribs} 23:39, 22 August 2009 (UTC)

(ec) Controller inputs get averaged out. I'd actually be surprised if some people didn't try that! They'd end up disappointed though. Poor terrorists! ;-)

So summarising: one of the differences between you and I is that I would trust a wiki with my life, and you wouldn't :-P --Kim Bruning (talk) 23:51, 22 August 2009 (UTC)

And how stable would the flight be if someone kept changing the speed of the engines, tweeking the angle of the flaps, and was constantly lowering and raising the wheels while you were trying to fly? Blueboar (talk) 23:50, 22 August 2009 (UTC)

Well, the control inputs get averaged out, so assuming that most of the people want a stable flight, the flight will actually end up being pretty smooth. Possibly smoother than if it were controlled by just a single pilot.

(IIRC in the case of the simulator, I think they only simulated joystick input. But making an expanded program and demonstrating it might be a really cool thing to do for wikimania 2010 :-) )

--Kim Bruning (talk) 23:55, 22 August 2009 (UTC)

I sure as hell am not getting on a plane with you in charge. SMASH! Blueboar (talk) 00:01, 23 August 2009 (UTC)

Oh, that's too bad. It's using the same theory as was used to generate our first 2M articles or so, and those methods are still used to maintain most of them. Are you arguing that wikipedia is impossible? ;-) --Kim Bruning (talk) 00:25, 23 August 2009 (UTC)

Oh come now Kim. I don't see a claim that Wikipedia is impossible on this page. I understand the benefit of being a slippery debater but there are limits. Chillum 00:32, 23 August 2009 (UTC)

Nope, somehow, despite the fact that the plane is bouncing all over the place and people are getting air sick, the plane is still flying and has not hit any mountains... but it sure as hell has been a bumpy ride, and the passengers are starting to agree that maybe they should stop pushing all those buttons, or at least ask what they do before they push them (well... to be honest, most passengers want everyone else to stop pushing buttons, but think it is probably OK if they continue to do so). Blueboar (talk) 01:29, 23 August 2009 (UTC)not to belabor the analogy or anything

Ah... did I mention some folks actually DID the experiment once? Apparently after a little practice, it was quite stable. Do not assume that the behaviour of the pilots is random! I've had some trouble finding the reference, and I'm out of time right now. I'll try to remember to find it later. Feel free to remind me! --Kim Bruning (talk) 04:24, 23 August 2009 (UTC)

Kim, you're making the rather dramatic and thoroughly false assumption that the "Wikipedians for pornography" and the "Wikipedians to protect the innocence of children" are trying to aim for the same place. I'm sure we could find half a dozen similar divergences without trying too hard -- even if we limited our search to what we could document in ArbCom cases. (Let's see: Homeopathy, Scientology, date de-linking...) WhatamIdoing (talk) 20:50, 24 August 2009 (UTC)

Frequent bold editing of policy

I don't think we want to make a statement that editors should be "frequently and boldly" editing policy; policy does need a certain amount of stability which "frequent bold editing" doesn't really allow for. Significant policy changes need clear consensus, and generally this should be sought before making the actual changes. Minor tweaks can be bold, but not major changes, at least imho. Dreadstar ☥ 20:31, 22 August 2009 (UTC)

Why does policy need a certain amount of stability? And for what it is worth, I have to say I've never really seen policy change too much, even when bold and frequent editing has been allowed. There isn't too much drift in the scope of any policy, the devil is usually in the detail. I'm always amused by people who object to allowing people to boldly edit policies because it will mean someone can amend NPOV to read "Northern Point Of View", as if telling such a person they can;t edit policy will prevent them from doing such a thing. Anyone who wants to vandalise a policy will do so regardless of the rules. Anyone who thinks they will be able to generate a consensus to amend the NPOV to read "Northern Point Of View" without reversion is, well, maybe saying one flew over the cuckoo's nest is too much, but I think everyone will catch my drift. We don't need rules to catch extreme examples, because extreme examples will happen regardless of the rules. We need the rules to document best practise, and best practise is everyone editing in a good faith manner, assuming that everyone else is editing in a good faith manner and editing and discussing where necessary to reach a consensus. If the goal was to have policy stay stable, I would imagine the first people to write the policies would have protected them, surely? Which seems to be anti-wiki, doesn't it? Hiding T 20:56, 22 August 2009 (UTC)

(ec)

See: #convenience_break.

Consensus can be found through editing, and often is.

Perhaps somewhat counter-intuitively to some; Rapid, continuous, bold editing can, will, and does attain stability. This is a form of stability known as dynamic stability.

I've sort of got poor Blueboar cornered in the mentioned section, since we've noted that there are a lot of edits to NPOV (bold or otherwise), and NPOV has actually remained pretty much stable; exactly like I claim it would.

As an analogy, take an aeroplane pilot. To keep the aircraft on a steady course, you cannot simply jam a log into the controls and hope that the machine will fly straight. There will always be slight gusts of wind, or unexpected vibrations, or turbulence to send the machine off course.

To keep the plane flying straight and level, the pilot needs to continuously make small course corrections back onto the desired flight path. In the case of turbulence, the pilot occasionally needs to make *large* corrections to stay on course.

A similar situation exists on-wiki. People need to be able to make changes large and small, as appropriate.

We cannot *force* people to discuss first, as the dynamics are then similar to jamming a log in the controls and praying. (see archives 5 and 6 for details on that)

--Kim Bruning (talk) 21:02, 22 August 2009 (UTC)

<ec>So you're saying that policies should be unstable? Some elements of policy are actually non-negotiable (e.g. WP:NPOV), so yes, a certain amount of stability is not only needed, it is expected. And just to clarify, "stable" does not mean locked in or unchanging as would be the case if policy pages were protected. I've seen "bold and frequent editing" turn into edit wars, blocks and policy pages being protected....so I don't think we need to be saying such a thing - especially regarding policy pages. Dreadstar ☥ 21:11, 22 August 2009 (UTC)

I'm saying that policies should be pretty stable. I just disagree on the procedure. Most bold editing does not lead to edit wars, blocks, or protections, fortunately, nor do discussions on the talk page. --Kim Bruning (talk) 21:39, 22 August 2009 (UTC)

That was @Hiding, not you Kim, a combo of those dratted edit conflicts and my lack of desire to top post. :) Dreadstar ☥ 21:42, 22 August 2009 (UTC)

Kim, this page documents an English Wikipedia policy, a widely accepted standard that all editors should normally follow. Changes made to it should reflect consensus. Same goes for other policies. While occasionally simply changing policy to reflect consensus can be done, it is more often the case the consensus needs to be tested on the policy talk page first. Chillum 21:12, 22 August 2009 (UTC)

Three letters: Yes.

I'd only want to add that there are in fact several ways to test and look for consensus, including but not limited to WP:SILENCE, WP:BRD, WP:IAR, etc...

My only concern is if people wish to make "discuss first" mandatory, as that would make some of the other methods more difficult. If discussing first is facultative, that's fine, I do it myself, after all. :-) --Kim Bruning (talk) 23:22, 22 August 2009 (UTC)

Throughout this discussion, which Kim has started on many pages over the years, Kim is the only person I've seen advocating bold editing on policies. Given that he says the policies are descriptive, not prescriptive, he should take that seriously and stop trying to misdescribe the way editors actually behave toward the policies. Copy edits and clarifications are fine; indeed, welcomed if they're good ones. But substantive changes are not welcomed without discussion. Policies that are edited so much they become unstable end up not saying anything that people can trust, so they fall into misuse. If you're relying on policy while you're editing, you have to know that it says more or less what it said when you last looked at it. SlimVirgin ^{talk|contribs}

I'm usually a first responder because I do policy patrol. Didn't we both agree to do that, way back when? :-)

We both show up, but normally you just leave after just a few back-and-forths. After that, other people tend to show up and the documented policy settles on allowing both "bold" editing as well as discussion.

One thing I'll admit to though: On or near my RFA in early 2004, I believe I told people that I would eventually be banned for trolling, on the basis that I would be pushing the traditional wiki-approach and consensus, and that this might fall out of favor as wikipedia slowly aged and finally died (like all online communities eventually do). I think they laughed at the concept at the time. :-P

What has surprised me most is that I've never really been accused of trolling so far. Wikipedia has lasted longer than I expected. I wonder if your rumbelings about my edits this year are just you, or whether they are the beginning of the end that I predicted over half a decade ago? :-) --Kim Bruning (talk) 22:34, 22 August 2009 (UTC)

You've been accused of trolling (rightly or wrongly) many times, Kim. And it's not just your edits this year; you've been doing this to policy pages off and on for years, and I've never see the efforts gain support. As I said earlier, if you believe the "descriptive not prescriptive" meme, please take seriously that, as a matter of fact, whether they're right or wrong, people do not want to see the policies become unstable because of BOLD editing. SlimVirgin ^{talk|contribs} 23:01, 22 August 2009 (UTC)

Ok, diffs for those accusations please! I know just one case where you'll find a diff, but that was retracted.

MzMcbride points out below that we may all have different views of what "bold editing" actually means. As a consequence of my particular definition, the only kinds of edits you are allowed to do all lead towards consensus (I'm very old fashioned, and not all wikis use talk pages, so then that's the only way to reach consensus) . Making more edits thus will tend to get you to consensus quicker.

I've always assumed that everyone just took that approach as a given (it's kind of fundamental to making a wiki work, after all) . Possibly that was a foolish assumption?

--Kim Bruning (talk) 23:10, 22 August 2009 (UTC)

This conversation seems to be supporting the current state of Wikipedia:Consensus#Policies_and_guidelines? --SmokeyJoe (talk) 21:40, 22 August 2009 (UTC)

The policies and guidelines reflect established consensus. Stability and consistency is important, and editors are therefore expected to discuss proposed changes on the talk page before making them.

@chillum, Slimvirgin, Smokeyjoe. *sigh* I just walked through this with a couple of other editors today. It's also been walked through in the archives. While I'd normally be happy for this opportunity to take each of you by the hand and walk you through it, would you mind taking a look at previous discussion on this matter first? (On this page, and also see archives, especially #5 and #6) :-) I'm supposed to sort of be packing for wikimania at the moment ^^;; --Kim Bruning (talk) 21:44, 22 August 2009 (UTC)

I consider myself familiar with those archives. A previously made point remains: Most of this discussion distracts from the primary purpose of this page, which is the explanation of WP:CONSENSUS to newcomers, as it applies to mainspace. Most of the discussion would be more appropriate to Wikipedia:Consensus/Further study --SmokeyJoe (talk) 21:53, 22 August 2009 (UTC)

Cool! I thought the point of this page was to explain consensus for all purposes. Are you proposing we split the page into beginner and advanced or so? You know, that might actually work. <scratches head> --Kim Bruning (talk) 22:09, 22 August 2009 (UTC)

I think it could work. I also think that wikipedia-consensus has not yet been sufficiently well defined for a robust explanation beyond the beginner level. Enjoy wikimania. --SmokeyJoe (talk)

It *STILL* hasn't been adequately defined? If you really think so, what have we been doing wrong all these years? I do think that it is defined, but the definition is fragmented across the entire project namespace, and is subject to erosion.

I was thinking of maybe writing a book or so, so that the information is in one place. <scratches head some more> --Kim Bruning (talk) 22:25, 22 August 2009 (UTC)

Definition of "bold"

I think the primary problem in this discussion is the lack of a clear definition for what a "bold" edit is. Some people are interpreting "bold" to mean changing the underlying ideas in a page. Others are interpreting "bold" to mean only changing the presentation of those ideas, but keeping the ideas the same. I think it might be best to move forward not using the word "bold" at all in these discussions and instead focusing on what you actually mean.... --MZMcBride (talk) 22:07, 22 August 2009 (UTC)

By "WP:BOLD" I just mean editing without necessarily discussing beforehand. (one might still read talk pages, or consult polls, or do anything else, the defining property of bold editing for me is simply that you don't post to talk before making the edit) --Kim Bruning (talk) 22:16, 22 August 2009 (UTC)

I'd be happier to see "Bold" point to WP:BRD, which says:

Try to make the edit and its explanation simultaneous: Many people will first make an edit, and then explain it on the talk page. Somehow there will always be some fast-off-the-hip reverter who manages to revert you right in the middle, before you have time to complete your explanation. To try to prevent this, reverse the order, first edit the talk page, and then make your edit immediately afterwards. This way your explanation will already be there at the moment of the expected revert. Don't hesitate between the two actions though, since for some reason people tend to be accused of bad faith if they do that. Best of all, if the page has little activity right now, you might be able to prepare edits to page and talk page summary, and save them simultaneously.

I read this as saying that if there isn't an existing discussion supporting the change, then you should prepare and post a decent explanation. This should dampen the problem of incessant fiddling. --SmokeyJoe (talk) 22:29, 22 August 2009 (UTC)

<blush> --Kim Bruning (talk) 22:48, 22 August 2009 (UTC)

SmokeyJoe, I'd like to see one weasel word added to the statement above -- either "editors are typically expected to discuss" or "to propose substantial changes on the talk page", or something like that. We don't want people to waste time seeking permission to fix a typo or to wikilink a word. WhatamIdoing (talk) 21:09, 24 August 2009 (UTC)

"essay opposing use of consensus to violate the editing policy"

That's a problem, I hope someone will fix it. WP:EGG, and WP:BOLD isn't a policy, ~~and policy can contradict a guideline but it can't "violate" a guideline; if there's any violating going on, it would be the other way around.~~ on second thought, just fixing the EGG would work for me - Dank (push to talk) 02:26, 1 August 2009 (UTC)

I'm a bit confused: What's the problem with Wikipedia:Linking#Piped links? WhatamIdoing (talk) 21:39, 1 August 2009 (UTC)

Hi, WhatamIdoing. The EGG problem is that the link seems to advertise WP:Editing policy but instead gets you to WP:Be bold, which isn't policy. - Dank (push to talk) 01:47, 2 August 2009 (UTC)

That was needlessly sloppy of me; thanks for spotting it. Is this better? WhatamIdoing (talk) 07:16, 12 August 2009 (UTC)

I really like the section linked, thanks. - Dank (push to talk) 11:58, 25 August 2009 (UTC)

BRD

@chillum: we already reverted, discussed, and went back to bold (in this case with Slimvirgin leading out in BOLD mode, strictly as per WP:BRD). So no need to blanket revert :-) --Kim Bruning (talk) 00:58, 23 August 2009 (UTC)

Adding "this is not mandatory" seems a little over the top. "Expected" is sufficiently flexible in nature not to mean "mandatory". Dreadstar ☥ 01:20, 23 August 2009 (UTC)

Oh, phew, I thought you were reverting because you thought it was not not mandatory (aka, required).

The reason I keep going back to this is that the "expected" makes me feel all nervous and depressed somehow. :-/ I would like to emphasize the non-mandatoriness, lest someone reverts based on the word "expected". Any suggestions? --Kim Bruning (talk) 01:24, 23 August 2009 (UTC)

Prozac? Blueboar (talk) 01:31, 23 August 2009 (UTC)

See, I don't think we want to emphasize "non-mandatory", and actually, I'm of the opinion that major, substantive changes to policy should be discussed before implementing. Sure, go ahead and be-bold-with-no-discussion to fix typos, misspellings or even changing the wording or formatting slightly to make it clearer or easier to read and understand, but I'm thinking that major changes should always be discussed first. Dreadstar ☥ 01:40, 23 August 2009 (UTC)

Well, if you read the archives, you'll see that if you mandate that substantive changes should be discussed first, you can end up in a situation where substantive changes become impossible or at least very hard, even if such a change is necessary. So I have this notion that that might not be such a great idea. --Kim Bruning (talk) 04:28, 23 August 2009 (UTC)

You seem to be alone, or almost alone, in thinking that BOLD in policy editing is a good thing. If you believe in consensus, I hope you'll act accordingly. SlimVirgin ^{talk|contribs} 04:31, 23 August 2009 (UTC)

How do you come to that conclusion? Hiding T 13:06, 23 August 2009 (UTC)

Yes, there are a number of editors that agree with Kim. M 00:13, 24 August 2009 (UTC)

I get the impression Kim is equating "discuss first" with "don't edit until the discussion is over". I certainly don't. For me, what "discuss first" is really about is alerting others to a potential change, explaining why you think it is needed, determining the degree of support and/or opposition by other editors, and giving others a chance to explain their view. Once discussion is under way, I see nothing wrong with bold editing. In fact, doing so can move the discussion along and help everyone to reach a consensus. Blueboar (talk) 18:29, 23 August 2009 (UTC)

As someone currently involved with a couple of people who have decided to "boldly" edit substantive guidance, I would say this is not a good idea. In poorly-watched pages a couple of people can decide to alter long-established guidance with wide-ranging implications with little or no community input. Guidance and policy need to be fairly stable or no one knows where they are. Editing articles is hard enough when policy is relatively stable. If you have to keep reading every policy page weekly in case someone has boldly reversed it, that makes things impossible. People need to be discouraged from altering policy to suit their immediate needs and acting in defiance of the built up experience and usefulness that lies in a guideline that has developed gradually and functioned well over the years. The warnings against non-consensus editing are very useful in dissuading editors from over-eager policy meddling. I would support greater limitations on guideline and policy changes so that significamt changes had to be widely advertised and approved by the community before implementation. Xandar 00:35, 24 August 2009 (UTC)

Xandar is referring to WP:Naming conflict, where a small group of editors involved with the Catholic Church article are reverting edits to a part of that guideline, reasons being "it's consensus", and "you need more consensus to change guidelines". The part in question says "an article is titled with the name which its subject refers to itself as", which contradicts the policy that common (and neutral) names are preferred. This is a very good case against the "don't be bold, get consensus" position. I'm now involved in that discussion, but I've seen plenty of cases where obvious corrections to misleading or improperly added wording have been fought on the basis of only 'don't change what's written'. M 02:22, 24 August 2009 (UTC)

I don't get why this is difficult. If you think you are doing the right thing, then do it. If someone reverts, or otherwise actively disagrees, then get conensus first, or at least make a clear argument before trying again. --SmokeyJoe (talk) 03:39, 24 August 2009 (UTC)

M, can you give some examples? SlimVirgin ^{talk|contribs} 05:08, 24 August 2009 (UTC)

M is totally and inexcusably misrepresenting the situation from a position of partiality. Two editors have suddenly decided they want to totally reverse the long-standing naming convention WP:Naming conflicts without gaining any sort of widespread consensus. Major edits were made on the basis of one or two people's ideas, taking out huge sections of the convention that had worked well for years, and when these were reverted, they started forum-shopping and then edit-warring rather than trying to find if there was any consensus for such radical changes. The page has had to be protected for some days to stop the constant reversions to un-agreed policy. His contention that the guidelines conflict is simply nonsense. This guidance has worked well for the past 4 years and is deferred to by both WP:Naming conventions, WP:NPOV and the manual of style. One of the chief proponents of the changes is also currently active in trying to change the name of the Catholic Church article. However the idea that opposition is "all from Catholics" is utterly misleading. Xandar 13:57, 24 August 2009 (UTC)

Actually, the real problem is that editors on both sides are editing the guideline with an agenda (ensuring that the guideline supports their position in the Catholic Church/Roman Catholic Church naming dispute). In this case, neither "bold editing first", nor "discuss first" would have made a bit of difference. Blueboar (talk) 14:10, 24 August 2009 (UTC)

This is an interesting example of a misunderstanding and these differences probably tell a tail. Here is the last edit I made to the guideline back on February 1 2009. Here is the difference between Feb. 1 and 26 April. Talk page over the same period. As one can see the changes were discussed on the talk page with no objections. Here is the revert that Xandar made on 10 June. And the comments on the talk page for that day: Wikipedia talk:Naming_conflict/Archive 1#About self-identifying names. So I disagree that both sides are editing with an agenda, as the changes where initially made in April to make the guideline more in line with the naming conventions policy. As can be seen from the talk page this major change to the wording was proposed 31 March 2009 and four weeks later after no one said anything the change was made. It only became a two sided issue on 10 June, when an article dispute split over into a guideline and perhaps if Xandar had followed Kotniski behaviour (and talked through his/her proposed reversal before doing and waited a few days, to see if (s)he had consensus for the revert, an edit war could have been avoided. --PBS (talk) 15:03, 24 August 2009 (UTC)

A number of the editors are entirely uninvolved with the Catholic Church dispute. My original edits actually made Xandar's position clearer within the guideline (I removed rambling examples, and pseudo-philosophical talk about rocks not having names for themselves). Xandar agreed with the changes, but then reverted back to "the original" (since obviously it has consensus...). I said several times that Xandar's position seemed sensible, since it was unforgivably neutral. After several great points were brought up against this, I changed my mind. It's entirely incorrect to label this as a dispute between equally-biased sides. M 21:33, 24 August 2009 (UTC)

PBS. The original change was made on 26th April after a note on the empty talk page. I saw the major change, made just on Kotniski's say-so on 10th June and immediately reverted to the original, giving my reasons on the talk page. Since there had clearly been no discussion, let alone community consensus to such a major change, this was the right thing to do. After some inconclusive debate on the page, certainly no consensus for change Kotniski again removed large sections of the guidance on 22nd July, which I again reverted. After a further period of discussion, we seemed to be reaching some sort of middle way, by shortening the guidance - which was one of the declared aims of Kotniski and PMAnderson. However PMAnderson then broke the developing attempts at consensus change by completely ripping apart the page and inserting, a completely new policy at absolute odds with the existing and long-established policy. Kontiski then joined him in this and discussion was moved to the Name conventions forum without informing me or any of the other objectors to their scheme. Gaining no consensus on that forum either, they came back to the naming conflicts forum and PMA started edit-warring until the page was protected. Xandar 01:30, 25 August 2009 (UTC)

M. What exactly were these "great points" which suddenly made you decide that the existing policy, which has never, as far as I have seen caused a single problem, and which has solved many, should now be reversed, even at the cost of mayhem across Wikipedia?? Xandar 01:30, 25 August 2009 (UTC)

Buenos dias. ;-) Would this be a job for a mediator to sort out? --Kim Bruning (talk) 13:22, 26 August 2009 (UTC)

It's beginning to look like it. Some of the discussion is getting heated. Xandar 21:10, 26 August 2009 (UTC)

Does this help?

Moving us back to the issue of "discuss first" and "bold first"... I have added the following (shown in bold) to the guideline...

The policies and guidelines reflect established consensus. Stability and consistency is important, and editors are therefore expected to discuss any substantial proposed changes on the talk page before making them. This does not mean that a final consensus on a proposed change must be agreed to before any edit can be made. Bold editing can often help move a stalled discussion forward.

This clarifies my position on the "discuss first" method, and, I hope, addresses the conscerns that some have expressed about endless discussions. Blueboar (talk) 16:34, 24 August 2009 (UTC)

Blueboar, I removed this because you're trying to find a compromise between most editors on the one hand, and two or three on the other. As I see it, there is strong consensus that BOLD is not helpful on policy pages. If we want to change that, or test it, I think a wiki-wide RfC is in order, because it affects every policy and guideline. SlimVirgin ^{talk|contribs} 19:50, 24 August 2009 (UTC)

Hmmmm... not sure if there is a consensus that you should never make a bold edit... Looking at both comments and practice, the issue seems to be about whether to initiate a change with a bold edit or with discussion. With the consensus being: initiate with a discussion.

Take the very section you and I are editing now... I see it as a perfect example of what I am talking about with my addition. I started things off a few days ago with a discussion, in which I expressed why I thought this policy needed changing. Others then expressed their opinions on the matter. I floated some proposed language, and got (at the time, hesitant) agreement to add it to the page. It was then boldly edited by several editors until it reached its current state. Meanwhile, discussion continued. I think it worked fairly well. No? Blueboar (talk) 21:47, 24 August 2009 (UTC)

I'm curious as to how it is that SlimVirgin gets to have an opinion that actually defines consensus, while the rest of us are merely whistling in the wind. Did I miss a cabal meeting or what? ;) I'm always interested in claims as to where the consensus lies, the people making the claims, and the positions they themselves hold. Has anyone actually delineated a reason why bold edits to policies and guidance should be prohibited? As far as I can see the consensus seems to be that the consensus making process seems to be the same with either discuss first or edit first. I can't therefore work out why we would guide against one. Hiding T 22:21, 24 August 2009 (UTC)

Blueboar, the issue is not whether people should make one edit, then talk, or never make an edit before talking yada yada. The issue is whether BOLD as a concept ought to apply to policies, and I see a strong consensus saying no, not just here but on multiple pages. Hiding, I see this stuff because I edit a lot, I see people refer to policies a lot, I see them pissed off when the part they referred to suddenly disappears, I see people reverting policy changes a lot. SlimVirgin ^{talk|contribs} 23:58, 24 August 2009 (UTC)

Ah, you're the one doing all the editing. ;) Okay, so if I understand you, some sections of policy are untouchable because it helps some users to win arguments? I'm trying to work out how we relate that to the idea that consensus can change. I don't really get the rest of your argument, because no-one is arguing that a bold edit to policy should stay simply because it is a bold edit to policy. We do all agree that policy should be subject to challenge, don't we? If we agree that, then I'm not sure why it matters what form the challenge makes, after all, Wikipedia is not a bureaucracy. Now, would it be possible to focus on the potential harms and gains to the community if we move from a dynamic mode to a static one? That would actually be of some benefit, and open up areas of debate that are probably more conducive to reaching consensus. If we treat our policies as static pages, we are effectively locking people out of a very important part of the wiki-contract; ultimately, we are dis-enfranchising them. What will that lead to, that dis-enfranchisement? Well, it's plausible it will lead to frustration, and a lack of participation, which is actually bad for Wikipedia in the long run. Nobody here is arguing that we do away with discussion, but I think it's a bit disingenuous to state that consensus is behind the idea that consensus can't change. If consensus can change, it follows that policies are dynamic, and the pages they are written upon are open to editing. At no point does this idea that pages are open to editing mean they have to be edited, or that discussion should not ensue where consensus is unclear. What it means is that discussion is not the only method for reaching consensus. If you look at how most of our policies formed, you will find they evolved more from editing the actual policy than from talking about it on the talk page. It is therefore quite clear that best practise is to edit. All that said, I know you agree with the bold editing of policies approach, because you do the same yourself, as can be seen at Wikipedia:Administrators and Wikipedia talk:Administrators. Comparing the history of the two pages, we find you've edited the policy without discussing on the talk page. So either you've acted against the community consensus you claim to know so well, or we all agree that bold editing to policies isn't a problem, it's the edits themselves which can be problematic. Hiding T 10:19, 25 August 2009 (UTC)

But I didn't change the policy. I tightened the writing. We're talking only about substantive changes, edits that change the meaning of a policy.

It's not quite fair to say they're untouchable to help people win arguments. They need to be stable to help people edit. I can either explain to other editors a million times that, when you cite a work, you must cite the work that you read, even if it's not the original, plus the original that you haven't seen, or I can quickly link to WP:SAYWHEREYOUGOTIT. When I do the latter, I want to know that it's there, and not that someone in a fit of BOLDness has done away with it. There's nothing wrong with that need. No one wants to get bogged down reinventing the wheel several times a day, year in year out. SlimVirgin ^{talk|contribs} 10:29, 25 August 2009 (UTC) \

Well, when a simple heading tweak can break a redirect, I think it's very hard to know what can be proscribed to ensure that everything works smoothly. I don't think anyone wants to get bogged down reinventing the wheel, but it is interesting that the tyres we use today have very little in common with a slice of a tree trunk besides the general shape. I mean, even the propulsion system has changed beyond all imagination, from simply pushing and pulling to vehicles that can fly! Sort of implies there is value in reinventing the wheel, wouldn't you say? So I guess we kind of agree that we don't mind people being bold, except when we do, and when we do we should work out what the consensus is, because we don't know whether we're going to like someone being bold until after the event, so there's no mileage lost in being bold after all. Unless, I haven't misread you, have I, and you're suggesting no-one do anything that might inconvenience you, are you? ;) Hiding T 10:54, 25 August 2009 (UTC)

The problem we've been seeing is editors asserting their right to be BOLD by reverting against multiple editors, waiting a short time, then starting up again with the same edit on another page. BOLD's proponents have been giving it a bad name perhaps. But really, what would be the sense of me going now to the MoS and removing that inline citations are required? You do need to know what you're doing if you're going to be BOLD, but if an editor does know what they're doing, does have a good idea of how the edit will be received, it isn't really BOLD, is it? Evolution, not revolution, is how policies change, and from time to time we'll have paradigm shifts too. But the time has to be right for a paradigm shift, and whoever initiates it usually has to know that. SlimVirgin ^{talk|contribs} 11:07, 25 August 2009 (UTC)

Could you give me some examples of these "problem editors" or disputed edits, please, and also, can you clarify which "we" you are using, either the Royal "we" or, um, actually, there isn't another one, is there? After all, each person only speaks for themselves on Wikipedia, don't they? Or can I start saying that actually, we haven't seen this problem you (or would that be "the other we"?) describe at all, and then where does that get us? Let's just assume you mean you've encountered a problem. Personally, I've encountered different ones, so perhaps you and I together can accept our experiences are of equal value, and work out a mutually acceptable solution. Although I have to say, I can't quite follow you when you say BOLD's proponents have been giving it a bad name, because you then go on to say that any edit which isn't reverted isn't a BOLD edit. Correct me if I'm wrong, but if we changed the definition of vandalism to apply only to all the good edits in Wikipedia, we could safely say Wikipedia doesn't have a vandalism problem, couldn't we? Hiding T 11:44, 25 August 2009 (UTC)

Hiding, one more point. Each policy and guideline has a core of editors who look after it. Those editors know what the consensus is regarding "their" policies, or at least they do if they're active content contributors. I help to maintain V and NOR, and I've been helping to write them and watch the talk pages for around four years. Plus I edit articles a lot, so I know what the consensus seems to be regarding those policies. I know which changes are likely to "take" and which not. I know what the community's concerns are, so if an edit addressing those concerns arrives, I know to try to keep it. And if one arrives that's very much opposed to best practice, I know people will reject it. Most policies and guidelines have a number of editors like that who keep an eye on them. You can call it "ownership" if you want to be negative, but you can also see it more positively as guardianship. What I've been concerned about with the changes we've been discussing is that they're not coming from people who have been tremendously active editors, so I have to wonder how they can know what the consensus is about this or that. It's hard for anyone to know, but if you're not active, it's obviously very much harder. SlimVirgin ^{talk|contribs} 10:38, 25 August 2009 (UTC)

Hmmm. I'm going to have to think about this and what implications it has. Because, as you say, it would be very easy to characterise your statement as akin to ownership, and even to ponder ideas like local consensus, cabals and elitism. I'm sure that's not what you mean, so I'll have to think hard to try and work out what you do mean. After all, I'm sure we agree that no-one on Wikipedia speaks for anyone but themselves, so it would be wrong to revert something simply because it might not "take". After all, the whole thrust of the argument against allowing bold editing of policies is that that never happens. Maybe, as you suggest above, we need to ponder having an RFC over whether the consensus really is that certain editors are granted stewardship, no guardianship, wasn't it, over policies and guidelines. Hiding T 11:06, 25 August 2009 (UTC)

The policies have a strong descriptive element, as well as a prescriptive one. That is, they have to describe best practice. If you see an edit to a policy you're familiar with, in an area you're familiar with, and you know it doesn't describe best practice, and doesn't reflect consensus, you know it won't "take." The right thing to do in that situation is remove it. Are you disagreeing with that? SlimVirgin ^{talk|contribs} 02:39, 26 August 2009 (UTC)

I disagree with your assertion that no editor speaks for anyone but himself or herself: I fairly often answer questions with statements that amount to "The community consensus is X, which I personally dislike." (See, for example, WP:REFPUNC, which says that putting a footnote before punctuation is acceptable, and related to which I have corrected several erroneous assertions to the contrary in GA-related pages, even though I hope that every editor will freely choose to put the footnote immediately after the punctuation, because this style suits my sense of aesthetics better.)

I sincerely hope that most of our editors are capable of identifying the existence of other opinions other than their own, and even accurately explaining the other side's views. If they aren't, then disputes will never be resolved. WhatamIdoing (talk) 18:08, 25 August 2009 (UTC)

WP:OWN doesn't magically reduce the reputation of people who have shown good judgment on a policy page over a long period of time, and WP:BOLD doesn't magically enhance the reputation of people who like to edit first and think later. Even though Wikipedia is more like a never-ending masquerade ball than a cocktail party, people are still people, and reputation is still something people try for, especially because there's no hierarchy to substitute for reputation. - Dank (push to talk) 12:23, 25 August 2009 (UTC)

I don't follow you. Are you suggesting that all people who utilise bold editing "like to edit first and think later"? I sincerely hope not, since I like to assume good faith, so let us assume that you are not. So if you aren't saying that, then you are saying that because some people may get it wrong, we shouldn't do it. Would that be right? And if that is right, then it follows that we shouldn't actually do anything, should we, because some people may (even do) get it wrong. Which isn't actually what we do at all, is it? We don't strip all admins of power because of the odd bad egg, do we? Although now you come to mention it... But that's a topic for another day. So basically, no, we should not prevent bold editing to policies, where that bold editing is preceded or followed by an explanation, because sometimes it is a perfectly good way to get things done, and after all, getting things done is more important than how they get done, because, well, we all know why. Hiding T 12:48, 25 August 2009 (UTC)

SV... I think we are both responding to the same shift in community consensus. But we clearly have a different different ideas as to details. Still, I do suspect that the consensus is at least closer to your way of seeing things than mine. Blueboar (talk) 00:24, 25 August 2009 (UTC)

Some respected editors are bold-ish and others aren't. I would characterize consensus as "Changes made to [a policy page] should reflect consensus"; that's at the top of every policy page. If we discard that statement in favor of BRD (which is only an essay), then we enable the guys who defend every edit by saying "I was only following BRD". - Dank (push to talk) 02:49, 25 August 2009 (UTC)

Out of curiosity, how do you ensure you reflect consensus? I mean, changes to articles have to reflect consensus, but we don't require people to discuss changes to an article on talk first, do we? Hiding T 12:51, 25 August 2009 (UTC)

Dank: BRD is defined as a specific special case of the consensus system. The procedures defined on the BRD page are methods to ensure that all changes reflect consensus. So by definition, if you are doing BRD properly , you are well within the normal bounds of consensus.

If you are making a choice "between BRD and consensus" , it's similar to asking me to -if talking with you- make a choice between your heart or your chest. Aka... HUH? --Kim Bruning (talk) 13:31, 26 August 2009 (UTC)

Some editors (particularly the ones who have done their homework) get good results from bold edits, most don't, and implying that bold edits are always welcome is misleading and a little bit mean to the less-experienced editors who really didn't know better and were expecting to get good advice from our BOLD guideline. - Dank (push to talk) 14:49, 27 August 2009 (UTC)

If you don't do your homework at all, either before or after, then I don't think you can expect to ever have results, no matter what you do, right? :-P Perhaps we need a WP:DOYOURHOMEWORK ;-) --Kim Bruning (talk) 17:43, 6 September 2009 (UTC)

It's an essay that's had only 111 editors and 282 edits in four years. [3] Even WP:BOLD is just a guideline. By trying to force BRD into policy pages, it is, in effect, being promoted by stealth. SlimVirgin ^{talk|contribs} 07:36, 25 August 2009 (UTC)

All that said, bold editing has been policy since near enough year dot. Attempting to color the practise as only a guideline seems to be an attempt to demote it through stealth. Hiding T 12:17, 25 August 2009 (UTC)

BRD is a descriptive text about things that do work and don't. Slimvirgin, I'm not sure what the relevance is of the other points you are making, in that context. --Kim Bruning (talk) 13:35, 26 August 2009 (UTC)

What do we mean by "discuss"?

OK, I think most of us are agreed about: "Stability and consistency is important, and editors are therefore expected to discuss any proposed changes on the talk page before making them, if they would substantially alter the policy". The still open question is... what do we mean by "discuss"? Blueboar (talk) 12:58, 25 August 2009 (UTC)

I don't get the impression most of us agree with that at all. I think most of us agree with something like "Policies and guidelines are standards of content and conduct that have widespread community support. Amendments to them require thought, therefore talk page discussion may precede substantive changes to policy. Bold editors of policy and guideline pages are strongly encouraged to follow WP:1RR or WP:0RR standards. Minor edits to improve formatting, grammar, and clarity may be made at any time." That doesn't really leave any questions open, after all. Hiding T 13:16, 25 August 2009 (UTC)

If tens of thousands of people feel that they have given assent to a particular section of policy (by discussing, editing, or simply following the policy without complaint), then a lone editor who changes the policy without any previous discussion about the change may find that they get quickly reverted, and people might (or might not) be less likely to support their positions in future edits and discussions. On your questions above, Hiding, my position is that you're looking for a one-size-fits-all answer. Different editors feel comfortable with different levels of boldness, and some can get away with it better than others can. It's not possible to solve this problem by making a policy statement that "bold is good" or "bold is bad". If I'm forced to make a snappy statement that covers all cases, the best I can come up with is "Changes made to [a policy page] should reflect consensus." (And that's just what the community seems to have settled on, otherwise it wouldn't be on every policy page.) That's a wonderfully flexible statement, when you take the individual editor into account. If Kim makes a change without discussion to a policy page, I know that in part it's because he honestly believes that that's the best way to proceed, he knows that that action is going to create resistance all by itself, he's aware of the consequences and is prepared to deal with flak over it. I might disagree, I might push back, but I wouldn't think he's being sneaky or sloppy. If another editor does the same thing, I might come to a different conclusion about what they're up to (and this opinion would sometimes affect how I respond to them, but I wouldn't share this opinion except at a voluntary editor review or RFA, and even then, only if it seemed really necessary and relevant to an issue raised there). It would be misleading and even mean to give editors the mistaken impression that WP:BOLD gives them some kind of blanket pardon for frequent, undiscussed changes to policy, because it doesn't provide that protection at all, not in the court of public opinion and not (in hard-core cases) at ANI or Arbcom. - Dank (push to talk) 13:56, 25 August 2009 (UTC)

The entire reason why we are discussing this is that community consensus on how to edit policies has already changed to "discuss first". The simple fact is that "Discuss before you edit" has, to all intents and purposes, already become Wikipedia policy. The community consensus isn't saying discussion is an option... it's saying disussion is needed, and expected (in fact it is comming close to being required). All else goes from that basic fact.

So the question isn't should we discuss before we edit or not... THAT has been decided out there, beyond this page. What this page needs to figure out is what the community means by "discuss before you edit". Blueboar (talk) 14:13, 25 August 2009 (UTC)

Agreed, and my answer to that last question is just above. Everyone is different, and a certain prior discussion might seem like overkill from a particular editor on a particular subject, and not enough coming from a different editor on a different subject. Whether a person chooses to revert and how they discuss will depend on that subjective opinion. Still, it would probably be kind to create an essay giving editors at least a list of examples of bold changes that quickly got reverted because of lack of prior discussion, and let people draw their own conclusions about what will fly. - Dank (push to talk) 14:22, 25 August 2009 (UTC)

We need to agree on a bare minimum acceptable discussion... otherwise the idea of "Discuss before you edit" is meaningless (as it gives people the option to, subjectively, decide that no discussion is needed). Blueboar (talk) 14:45, 25 August 2009 (UTC)

Thinking a bit more about what Dank is saying... Can we at least agree that some amount of discussion is expected (boardering on required), but how much discussion will be needed depends on the specific situation. Blueboar (talk) 15:00, 25 August 2009 (UTC)

Btw, I'm only talking about policy, and maybe those guideline pages that get edited as much as policy pages do (and maybe we should say at the top of the talk pages of the heavy-traffic guidelines, paraphrasing, "This guideline is discussed and cited a lot. Be aware that many editors are expecting the same serious attitude here towards discussion and edits that applies on policy pages.") It would probably be useful to give editors a heads-up that even a good edit is likely to get reverted on a policy page unless they make the case for the change first; at a minimum, it would be smart to ask first if anyone knows if the desired change has been discussed before, on that page or elsewhere. - Dank (push to talk) 16:58, 25 August 2009 (UTC)

I think each page has its own "culture", but in general, yes, I agree that the average policy page takes a dimmer view of well-intentioned but undiscussed changes than the average guideline. Within guidelines, there's a lot more variation. For example:

Changes to WP:Reliable sources's substance are strongly resisted, and I'd expect even a small change to meaning to be promptly reverted, and would not be surprised to have substantial irritation directed at the "bold" editor.
Changes to WP:MEDMOS sometimes are done with an edit summary like "Here's what I think we should do, please revert or improve", or with a change followed by an explanation at the talk page. Nobody's angry at bold editors, although undiscussed changes tend to have a short lifespan.
I recently re-wrote half of WP:SISTER without a single word of discussion. (Of course, I knew that I was on solid ground, as no one can legitimately object to providing examples of how to format these links and describing the usual locations for different types of links.)

In general, I think that it's fair to warn editors that bold editing of most policies and some guidelines may not help them win friends and influence people. Discussion is generally "expected"; only on a few pages is it -- in actual practice -- "required". WhatamIdoing (talk) 18:23, 25 August 2009 (UTC)

I'm confused. You guys seem to be saying that what I wrote is not current consensus. Trouble is, it's cut and pasted from an arbitration committee endorsed policy. So something is obviously wrong. I'd like to understand why you continue to claim your position has consensus though, so perhaps you can walk me through it slowly. Cheers. Hiding T 19:08, 25 August 2009 (UTC)
Link please? - Dank (push to talk) 20:34, 25 August 2009 (UTC)

Hiding... read back over the last few days of discussion. Blueboar (talk) 20:37, 25 August 2009 (UTC)
This is interesting, and I'm sort of confused myself. The two of you seem to be arguing that changes to policy pages shouldn't be allowed without a discussion first, but policy itself seems to contradict that stance, and is further supported by the fact that policy pages aren't generally protected. To be blunt, the above thread of posts appears to be an attempt to foist some sort of fait acompli on... something.
— $V = I * R$ (talk) 00:13, 26 August 2009 (UTC)

Ohms... That is exactly the situation. In many ways we do have a fait acompli... This policy page no longer matches the community consensus... the consensus as to what the policy should be has changed, and that change now needs to be reflected on this Policy page. The problem is that, in this case, the consensus shift happened organically, out there on all the other policy/guideline pages instead of here on the relevant policy talk page. Essentially the editors who watch and work on our other policy pages have agreed that you need to discuss changes first. If you want to put a label on it, what occured was an unspoken mass invocation of IAR ... saying, we don't care what WP:CONSENSUS says, on this page you have to discuss first. It has reached the point where this is the norm at most of the other policy pages, which tells me that this practice now enjoys the consensus of the community. Thus, the need for this page to reflect that consensus.

What makes this ironic is that we are talking about needing to change the policy about Consensus... because what that policy no longer has consensus (as it applies to Policy pages). Blueboar (talk) 03:04, 26 August 2009 (UTC)

In many ways, the shift to "Discuss First" on policy pages is similar to what is going on with the "flagged revisions" experiment on articles... we are seeing a fundamental shift in the concept of how Wikipedia needs to work in the future. The entire BOLD concept is being re-evaluated (not necessarily abandoned completely, but certainly changed). For articles, "flagged revisions" means that there is going to be a (needed) check on overly-bold editing. The "Discuss First" shift on policy pages is part of that same trend. It too is a check on overly-bold editing. If the flagged revisions experiment works... it may well end up being applied to Policy pages... and that may cause yet another a change in consensus. But until that happens, we need to bring this page up to speed. Now... the current language (with the addition of the short sentence mentioning that discussion is expected on policies and guidelines) does this. Thus I come back to my question... do we need to tweek it to explain what we mean, or is what we have enough? It sounds like people think it is enough. That works for me. Blueboar (talk) 03:25, 26 August 2009 (UTC)

Agreed. SlimVirgin ^{talk|contribs} 08:37, 26 August 2009 (UTC)

eh... this whole idea of "we need to bring this page up to speed" just makes me uncomfortable. I know that I've seen some significant debate over proposals to protect all policy pages, for example. To be straightforward about it, I think that a couple of you are seeing what you want to see based on your own viewpoints. There probably are specific, individual, policy pages that have some entrenched positions, but it seems to me to be a real stretch to extend that to the point of thinking that changes shouldn't ever be made without getting some sort of "approval". I don't think that you're wrong exactly, but I think that you're probably over extending the conclusions that you're drawing from personal observations.
—

V = I * R

(talk) 08:58, 26 August 2009 (UTC)

Hang on. So the consensus page and the policy page contradict, and some of us think the policy page has best practise and some of us think the consensus page has best practise? Is that what the issue is? And the change to this policy is quite recent, isn't it? So actually, when you think about it, that's the change that needs to be reverted and discussed, isn't it? I mean, we're fairly evenly split here, so that's probably an indication that consensus hasn;t been reached on the issue yet, so whichever practise we're using, either BRD or discuss first, that would mean the disputed text shouldn;t be in the page while we discuss it, doesn't it? Hiding T 09:16, 26 August 2009 (UTC)
There, I added an edit which takes a bit more flexible route. Essentially, some engagement on talk pages should be seen to be required. "Drive by" changes, and certainly WP:POINTy changes, should be actively discouraged. Engaging in discussion is essential to the consensus building process, however, and locking out that initial change can really have chilling effect, espeically on true BRD style editing.
— $V = I * R$ (talk) 09:44, 26 August 2009 (UTC)

Stability

Caveat: I think it's understood that we need wider discussion before we do anything significant (and it was noted in the Signpost that similar discussions are happening in various places, as usual), otherwise whatever we decide on is likely to be rejected even if it's great. Another caveat: I'm fine with the idea of a bias in favor of the version that tens of thousands have assented to in one way or another and against the lone edit that doesn't seem to reflect the history of the page ... the problem is, which version is the "status quo"? Is it the version the last time I looked at the page, or a week or a month ago? If we're not careful, all we're doing is adding another layer of stuff to argue about ... which is the "stable" version ... before we can start arguing about substance, which will probably mean a lot of wasted time. The answer I like best is one I can't personally argue in favor of because of potential conflict of interest: the best arbitrary "stable" point would be once a month, meaning the version that shows up in the WP:Update ... once a day would mean you potentially get a daily argument on which edits just before midnight represented consensus, and once every 3 months would be rejected since we all know consensus can and does change more often than that. (The conflict of interest is that I've done almost all the work at WP:Update so far.) If we do this, then whoever's doing the Update should follow the edits and discussions on each policy page for a day or two after the end of the month and use good judgment on picking a cut-off if necessary to represent a more stable version. Obviously, this wouldn't work for guidelines that don't get updated. - Dank (push to talk) 13:27, 26 August 2009 (UTC) got a better idea, see below 14:52, 27 August 2009 (UTC)

I think people are missing the point here... you can hedge this all you want ... it won't change what is happening out there. So, the question is very simply... do you want this page to reflect what the consensus actually is (you are expected to discuss major changes to Policy/Guidelines before you edit), or do you want this page to reflect what a few policy wonks think it should be? Blueboar (talk) 14:05, 26 August 2009 (UTC)

Major change from which version, Blueboar? (Btw, I've gotten some good feedback on this, I'm going to start in this month's Update fudging a little on the midnight Aug 31 cutoff if it looks like no one has had a chance to respond to a recent edit. It's going to be a huge amount of work, and as always, anyone is welcome to do any of it.) - Dank (push to talk) 14:10, 26 August 2009 (UTC)

From the version that was on the page prior to any non-discussed major edits, as that version reflects the most recent consensus (as determined by the last discussion that took place). Actually, if no one makes major edits without prior discussion, you don't have any question as to which version... what is currently on the page should reflect the most recent consensus (because no one has changed it since the last discussion took place). You only get arguments over "which version" to use when changes are made without discussion. Blueboar (talk) 14:40, 26 August 2009 (UTC)

Okay, then let's see if we can all come to an agreement on what the established version of this page is, the version against which future edits should be compared to see if they're "major changes". If we can't do it for this one page in a reasonable period of time, then odds are slim that we can do it for all policy pages, let alone guidelines. Does anyone want to suggest a date/time stamp as the "established" version? - Dank (push to talk) 15:01, 26 August 2009 (UTC)

This looks like an interesting idea, because if this can't be done, wouldn't it sort of imply that we don't actually expect policies to be stable and consistent? Hiding T 15:10, 26 August 2009 (UTC)

Look, I'm really confused now. If Blueboar is right and the page should reflect what is happening now, well, then people are using BRD at policy and guideline pages. So I'm not sure we'd say they aren't, when they are. And if there are discussions about it all over Wikipedia, that kind of means there isn;t a consensus against doing it either. SO I am completely unclear as to how we get from those two points to the point of saying that "the consensus actually is (you are expected to discuss major changes to Policy/Guidelines before you edit)", when that's not current practise or current consensus. Unless I'm missing something. Are there any policies or guidelines which people don't boldly edit? Hiding T 15:05, 26 August 2009 (UTC)

At the moment some people are attempting to boldly edit policies and guidelines... and they are being told in no uncertanin terms, "no, discuss first". Blueboar (talk) 15:31, 26 August 2009 (UTC)

Blueboar, my position is that this idea that we can all agree for every policy page at every point in time what the "established" version is, the position that gets the benefit of the doubt against new edits, is a non-starter; I doubt that we can even get consensus on this one question for this one page at this one point in time. Hiding, the problem is that people read BOLD, and see that changes get made to policy pages all the time, and they come to the conclusion that they should go ahead and make an edit that looks good to them without doing the hard work of reading the page history and talk page, and almost always, their change doesn't "stick" if it's at all significant. People who feel like their idea is shot down without a fair hearing come to the conclusion that they're being disrespected and/or that a surly cabal OWNs the policy page; very rarely is either of these conclusions correct, but we can't stop people from thinking these things, and this is a needless source of hurtful conflict. We need to come up with a system that stops setting people up and knocking them down. The only solution I can think of is, someone picks a date/time stamp somewhere around the end of the month, and for edits made during the month, editors are encouraged to revert edits and discuss on the talk page if they feel the case hasn't been made that the new edit is an improvement. Picking a monthly, somewhat arbitrary standard to judge new edits against will present several difficulties, but I haven't heard any alternatives yet that overcome both of the problems I mentioned in this paragraph. (I won't argue in favor of using WP:Update per se since that's kind of my baby, but until someone suggests a better idea, I'll argue in favor of some sort of semi-arbitrary, regular set points to compare against future edits.) - Dank (push to talk) 16:02, 26 August 2009 (UTC)

TLDR summary: pick an apparently stable point in the edit history of each policy page on the last two or first two days of each month, and treat this version for the rest of the month as if it were "consensus"; that is, to get a change to stick, the arguments will have to be strong enough to overcome the presumed consensus. If we do a bad job of picking the "stable" version, we can fix that problem by rational discourse and hopefully get it right next month. Besides, most policy pages are relatively stable; the damage done by a slightly-out-of-line choice will be minimal. The claim is that this system is better than what we've got now and the alternative Blueboar proposes, for the reasons given above. - Dank (push to talk) 16:47, 26 August 2009 (UTC) got a better idea, see below 14:56, 27 August 2009 (UTC)

I think that we're over-thinking this. Editors are perfectly capable of discussing, on the talk page, which version they think is the previous stable version. It may even vary by section, if changes are being made and discussed in multiple sections: the immediately prior version may be the canonical version for the relevant section, without being accepted for some other section. WhatamIdoing (talk) 16:55, 26 August 2009 (UTC)

Okay, I'll keep an open mind, since some people assert this is doable. What would you say is the "previous stable version" for this page (WP:CONSENSUS), as of the last edit? - Dank (push to talk) 18:00, 26 August 2009 (UTC)

I know that this is going to sound somewhat harsh, and want to make clear that I understand where the "we need to keep it stable" thinking is coming from and respect it. This whole idea behind some artificial stability, identifying a stable version, the "correct way to behave"... all of it is just a wrong-headed attitude to have. Wikipedia has, by now, an institutional culture, and none of us is going to be able to change that, especially not here on a relatively backwater project talk page. Blueboar, I understand completely what you mean about getting this page to reflect reality, I simply dispute your perception of that reality. There are definitely some policy pages that have become locked down, but that's more of a problem that is caused by clumps of "cabals" who have taken over in some areas then it is reflective of the real wider reality. As Xandar (talk · contribs) points out briefly below, I think that we're risking marginalizing policy completely, here. I think that this is exactly what lead me to start Wikipedia:Village pump (proposals)/Archive 51#Policy and Guideline improvement drive, and then to join WP:PROJPOL.
—

V = I * R

(talk) 23:48, 26 August 2009 (UTC)

I know that there are cabals of all shapes and sizes, but every time I've seen someone claim that their edit was reverted because the cabal wouldn't permit it on a policy page (no comment on guidelines), that's not what it looked like to me. It looked to me like someone thought they understood the issues, and saw others boldly editing policy pages, and thought their edit would be welcome too, and then they came to the conclusion when it got reverted that they were being disrespected and/or being ganged up on by a tight-knit cabal, and frankly, I see very little evidence of either flagrant disrespect or tight-knit cabals operating on policy pages. Policy is just hard; there's a lot to read before you edit, and others who have done their homework often independently come to the same conclusion, that a new edit doesn't sufficiently respect the previous discussions and long-standing consensus involving a whole lot of people. - Dank (push to talk) 00:16, 27 August 2009 (UTC)

I can agree with pretty much all of what you're saying here, it's just that... that's a common criticism of one aspect the entire WikiMedia "universe". The openness of all of our processes leads to problems, and I don't think anyone will deny that, but in balance the benefits of openness far outweigh the costs of being closed.
—

V = I * R

(talk) 07:09, 27 August 2009 (UTC)

Experiments

P.S. I understand from the "fait accompli" comment that, as always, some are anxious that some people are getting together and trying to force a change on everyone else. I don't think that's happening, and I'm not in a hurry. It would make me happy if we can agree on some of the problems and experiment with some proposals. I'll pick a few policy pages and float the idea of seeing if we can get everyone to pretend that the beginning of the month version is a working "consensus", and see if that actually fixes any of the problems mentioned above. - Dank (push to talk) 18:33, 26 August 2009 (UTC) got a better idea, below 13:06, 27 August 2009 (UTC)

Given the fact that WP:CONSENSUS isn't opperating under "discuss first" at the moment, it is very difficult (if not impossible) to determine a "previous stable version". If we had been opperating under that system, I would argue that the last change that had a clearly discussed indication of "yeah, go ahead and make the change" was back when I originally added my proposed language for a "Consensus on Policies and Guidelines" section (this dif). I definitly don't think we should revert back to my edit... I happen to agree with some of the subsequent changes that took place afterwards (and I think they would have recieved consensus had they been discussed)... but there wasn't any real discussion about them (well... SV did open discussion on some of her changes, but I don't see much in the way of a reply... I'm not sure if that counts as discussion or not). Blueboar (talk) 19:05, 26 August 2009 (UTC)

"big" (e.g. WP:BLP or WP:N) or "contentious" (e.g. WP:NOT or WP:CSD) policy pages are watchlisted by a fair number of editors, so I should imagine that a discussion among folks who watch one of those pages could be sufficient to make a small change. The problem stems from rapid or wide ranging changes. Most editors and admins don't actually check a process page or policy page immediately prior to performing an action. when was the last time you read WP:BLOCK? or WP:ADMIN? People operate on tribal knowledge. Their knowledge of policy stems from discussions they have heard about it, memory of the most salient experience with it and a good deal of fudging (where gaps in knowledge are filled by assumptions of what a reasonable policy would do). Ideally, we would like editor and admin action to correspond perfectly with policy. There are a few routes to this: 1) Make editors follow policies step by step. 2) Continuously adjust policy until it meets editor/admin actions and 3) attempt to write policy so that a reasonable assumption of what the policy ought to say ends up being what the policy says. Each of these solutions has fatal problems, so we try to go with another way. Policies should be a reasonable codification of stuff we already do. They should be stable, because we don't want too many people operating under too many incorrect memories. They should follow practice, because many impressions of a policy will be generated by experience with practice, and they should be reasonable, if not for the prima facia requirement, then because we would want the policy to approximate an assumption of its contours. In policy space, this means we should favor inertia for inertia's sake. Consequently, discussion should involve interested editors and should at least include some affirmative case for the proposed change. negative consequences of this stance are outlined well by dan above, but I'm not convinced that the broad positives (policy matches roughly with practice) are swamped by the frustration felt by editors attempting to change policies. Protonk (talk) 20:01, 26 August 2009 (UTC)

As someone who spends most of his time editing articles, I have to say that the "frustration" felt by editors who don't get their bright policy change wheezes through as quickly as they would like, is of minor concern. Policy and stability go together. Not rigidly, but so that article writing is possible. If policy is changing every week on the whim of "bold" editors, with none of the affected being informed, then in the end policy will either start being ignored altogether, or there will be uneccesary mayhem in the articles affected. If editors can't refer to a policy in the knowledge it is likely to be there next week, what is the point? At the moment there seem to be too many people seeking change for change's sake. I would feel that in addition to the "watchers" on a policy page being consulted, major changes should be advertised Wikipedia-wide, or the editors of articles likely to be affected by a proposed substantial change informed that such a change is in the offing. Xandar 21:25, 26 August 2009 (UTC)

Affected by what? A community-wide consensus? How is that in any way a problem? If you're not the sort of person who has certain policies watchlisted, or Dank's very informative monthly digest of policy changes - I forgot the link, or you don't drop by WP:CENT sometimes, then you don't need to be spammed. M 21:51, 26 August 2009 (UTC)

I don't think that Xandar is proposing spamming anyone; I think s/he proposes that major discussions get listed at WP:CENT, or WP:VP, or WP:RFCpol. These are perfectly reasonable and very normal ways to both get more input (more datapoints = more reliable measurement of actual consensus) and to advertise changes (providing practical information to busy editors). WhatamIdoing (talk) 23:41, 26 August 2009 (UTC)

I think that he makes a good point. I've had a long standing grievance with Wikipedia's structural failings in the area of "community". The problem is, that discussion is far beyond the scope of anything which could possibly be resolved here.
—

V = I * R

(talk) 23:51, 26 August 2009 (UTC)

That's fair. We are way north of Dunbar's number here, making the sort of collaboration envisioned by folks in 2001 rather hard. Protonk (talk) 01:11, 27 August 2009 (UTC)

Well, my response to that is, 'good luck'. What's a substantive change? When I merged Wheel War into Admin, was that substantive? I didn't really change policy, but, um, I pretty much removed an entire policy page with me' own bare hands. Should I have notified the community, first? When people support an action, they don't say a thing, only the opposition does, so perhaps it wouldn't have gone through. It's all "who cares about one wee redundant sentence, one extra policy page" until it's "wait, you cut the size of a policy page in half?" (+how dare you, did you get consensus first!?) Anyway, it works well enough now, attempts to make it 'work better' have some serious drawbacks. M 02:27, 27 August 2009 (UTC)

I think that this is a particularly salient point. Quite often, bringing an issue up on the talk page only leads to those who want to stop whatever it is, saying something. There's a natural selection bias that occurs there. It's also the primary issue that WP:BRD is intended to address, because making an edit to the page (either article or project page) is discussion of a sort. The problem is getting people to stick around to participate in the continuing discussion, really. I assure you that there are plenty of mainpace articles that have been, are, and will continue to be much more controversial then policy and guideline pages will ever be (one of which brought one of the participants in this thread here, by the way...) There is already, and there will always be, additional stability to policy and guidelines simply do to their nature as policy. I don't see that we need to throw up additional barriers to that... although, I might be open to a Flagged Revisions implementation, eventually (I need more exposure to it, before I'm comfortable with recommending that).
—

V = I * R

(talk) 07:03, 27 August 2009 (UTC)

Quite often, bringing an issue up on the talk page only leads to those who want to stop whatever it is, saying something... and why is saying something a bad thing? Blueboar (talk) 11:49, 27 August 2009 (UTC)

If you want nothing to change, then forcing people to the talk page is certainly a means to do that. What is being pointed out is the fact that people often don't say anything unless they perceive a problem, which leads to selection bias. This is hardly an issue confined to Wikipedia, as for example Nixon famously used the phrase "Silent Majority" when talking about the Vietnam War.
—

V = I * R

(talk) 12:15, 27 August 2009 (UTC)

How is "forcing" people to the talk page a means to achieving no change? Blueboar (talk) 13:19, 27 August 2009 (UTC)

It seems to me that you just want to add an (unenforcable, anyway) extra hoop for people to go through in order to edit any project pages that happen to be policies (or guidelines?). It's obviously not just you, by the way, but it's easier to just say "you". To be upfront about it, it seems to me that you're saying "I know what's going on better then you, so you have to get my approval first!" I don't object to getting people to discuss what they think should be changed on policy and guideline pages. What I do object to is "before". That's just going to keep people from even trying to make changes to policy or guidelines, weather that's the intent here or not.
—

V = I * R

(talk) 22:52, 27 August 2009 (UTC)

My point is that a change to POLICY and some major guidelines can often cause widespread troubles that the few people making the change don't consider or foresee. If someone decides to change the guidance "Use Pinyin as standard for Chinese", and doesn't involve the Chinese article community in that change, it will cause hundreds of problems down the line. In guidelines and policies we do not need a presumption in favour of change such as Ohms Law and others seem to advocate. If it aint broke, dont fix it. Change should require a real need for change, so great that the majority of comment for an advertised change from users would be supportive. Xandar 01:33, 28 August 2009 (UTC)

I think that part of the problem here may be a bit of a misunderstanding of just what policy and guidelines actually represent. They are not, have never been, will not be, and cannot be actual "rules". WP:OFFICE makes real "rules", and bureaucrats are the only ones actually capable of enforcing any actual rules regardless. Policy and guidelines really just describe "this is why and how we do what we do". That can be used as rules by all of us, to a certain extent, but that is done based on our own (usually implied) consensus in saying "yea, I agree with the citation of <whatever> in this case". There's more in depth and understandable coverage of this topic on WP:POLICY, if you're interested.
—

V = I * R

(talk to Ω) 05:24, 29 August 2009 (UTC)

I'll respond to this on your talk page, Ohms. - Dank (push to talk) 12:36, 29 August 2009 (UTC)

Plan

An essay and a plan: User:Dank/Essays#Consensus on policy pages. - Dank (push to talk) 14:28, 27 August 2009 (UTC)

Groan. WP:CONSENSUS is the second page I come to in the monthly update ... feel free to yell at me (preferably on the relevant talk page) if you would have picked a different version than the one I pick (I'm going to have to give this some thought). I have to start early this month because this update is going to take a lot longer than the previous ones. - Dank (push to talk) 18:36, 27 August 2009 (UTC)

Okay, I'm trying to figure this out as I go along. It's safe to say that if I habitually say "I don't like the current version, can you change it before I do the Update?", people are going to yell at me. OTOH, I can't make a decision if I actually don't understand the language, and I'm not clear on what "Editors are therefore expected" in the last edit means ... expected by God? By everyone? By many policy editors? - Dank (push to talk) 19:03, 27 August 2009 (UTC)

Expected by the community. Blueboar (talk) 20:29, 27 August 2009 (UTC)

Okay. I don't think you're using the word "community" in the sense that it's used on pages such as WP:BAN, but I should probably back away from making suggestions and concentrate on trying to pick a version that appears more stable than the other choices to include in the WP:Update. The current argument seems to be between requiring a talk page comment simultaneously vs. ahead of time. - Dank (push to talk) 21:15, 27 August 2009 (UTC)

I think that over-simplifies the argument slightly... a discussion implies more than just posting a talk page comment ahead of time... it implies some degree of back and forth (such as waiting to see if others have any questions about your proposed change, and then answering those questions). Blueboar (talk) 22:22, 27 August 2009 (UTC)

Ironically enough though, there is obviously not consensus for this change. My suggestion is to simply mention that discussion about this specific area of the policy is under debate and is therefore not stable. That way you're not making an arbitrary choice, and you're faithfully reporting what's actualyl occurring.
—

V = I * R

(talk) 23:02, 27 August 2009 (UTC)

I agree with Ohms as to how to report what is happening on this page. As for "not consensus for this change"... well, that depends on how you measure consensus. I think there is a solid community consensus behind what the change ... but I would agree that there is not yet a local consensus on this page (which is what is so ironic... the very policy that tells us that consensus can change, is having difficulty with editors acknowledging that it has changed). Blueboar (talk) 00:07, 28 August 2009 (UTC)

I'm having trouble with meshing a requirement to make one kind of edit before you can make another kind of edit with WP:BURO. I've always taken the last sentence in BURO to be pretty important (the one about a mistake in procedure not invalidating your post), but I just searched through the talk archives at NOT and find only 8 hits on BURO, none of which is on point. I've seen it cited a lot on other talk pages, though. Don't we have to modify that sentence, which has been policy since at least 2005, before we can write in a requirement here to require talk before certain edits? If so, maybe we should be making the argument over there first. And I don't know how that argument will turn out, or how I want it to turn out. - Dank (push to talk) 04:48, 28 August 2009 (UTC)

This is why I like the word "expected", and don't like the word "required"... "Expected" is descriptive, while "required" is proscriptive. The current language does not invalidate major editing without discussion... but it does let editors know that doing so is strongly discouraged. Blueboar (talk) 12:45, 28 August 2009 (UTC)

←Good point. On the previous question: WP:Update has never reported on any of the discussions or any kind of meta-information, only on what's actually on the page. It would be nice if I could keep using that distinction to set appropriate boundaries on my role. I'd prefer to give you some idea what would seem like "stability" to me on this page, and you guys can work together (or not) to head in that direction (or not), and I'll report on what shows up on the page. This gives you an incentive to make an edit that attempts to respect and/or mention all significant points of view, since that version would be more likely to be selected. I've made a tweak to WP:Update to describe what I'm looking for: a late-month version that looks stable, or if there isn't one, then whichever version looks like the best starting point for the next month's discussions.

So: I understand that a lot of editors don't agree with the language that Blueboar and others are asking for, but that doesn't invalidate his point that our current policy and the WP:Be bold guideline are giving bad advice that needs to be fixed. One approach to getting the word out and gathering consensus might be to add something like this to the nutshell at the top of every policy page: "Policy edits that don't take other points of view into account often get reverted; ~~with a request to read the prior discussions~~ discussing before editing may help avoid misunderstandings." I don't know if this is maybe less forceful than you're looking for, Blueboar, but I think this plus the effects of the new Update might be enough to push things in the right direction. Plus, I'm not sure that I'm on board with the idea that it's best to always discuss significant edits first; I mean, a lot of people don't, and it works out fine. I'm going to be spending a lot of my time reading policy discussions for the next few months, and I'll keep that in mind. - Dank (push to talk) 16:44, 28 August 2009 (UTC)

I have no problems with the idea of adding a nutshell statement like that. Advising editors to read past discussions, and discuss major revisions before editing is a very good idea in my book. Blueboar (talk) 17:40, 28 August 2009 (UTC)

Thanks. Does anyone want to balance that with something like "Everyone is welcome to contribute, but ..."? What we should shoot for is not to get in the way of people's own good sense about whether their edit will be welcome or not. Most people have a natural aversion to speaking up when they're not sure what they're talking about, and WP:BOLD attempts to undercut that hesitation, which is fine on some pages but usually a disservice to the contributor on policy pages. So I'm aiming for something neutral ... not encouraging, not discouraging. If we can settle on some language, let's float the idea somewhere else or have an RFC, then put it at the top of every policy page. If the language survives there, then we've got something stable. - Dank (push to talk) 19:21, 28 August 2009 (UTC)

I think that the nutshell idea is a good idea, and the original language seems good to me. Adding a bunch of platitudes or passivity to it isn't really going to help anything. My main concern all along is what you just expressed well: "I'm not sure that I'm on board with the idea that it's best to always discuss significant edits first; I mean, a lot of people don't, and it works out fine." The way things were looking there for a while, it looked like an attempt was being made to turn this document into some sort of weapon. I know from discussion that wasn't really the intent, but you know what they say about the road to Hell...
—

V = I * R

(talk to Ω) 01:03, 29 August 2009 (UTC)

I need some guidance here on what to do when the end of the month is 3 days away, and the talk page clearly shows a significant, unresolved conflict, but only one of those positions is currently represented on the project page, as is the case right now: "Editors are therefore expected to discuss substantive changes on the talk page before making them." Would it make sense to aim for a version I could put in the WP:Update that tries not to pre-judge the outcome, on the theory that pretending that we're already done here would only inflame the opposition and make progress harder? How about this: "Many though not all editors are expecting discussion before significant edits and may be quick to revert policy edits that don't take other points of view into account." It's unlikely that would be the language that anyone would settle on when we're done, but does it work as a quick way to suggest that changes are on the way and we're not done yet? - Dank (push to talk) 03:14, 29 August 2009 (UTC)

Dan, you really don't like simple discriptive statements, do you? What is it about "Editors are therefore expected..." that bothers you so much? It's accurate, and it gets directly to the point. And, if I am correct and the change reflects community consensus, then it is unlikely to "inflame the opposition". Blueboar (talk) 04:19, 29 August 2009 (UTC)

I know you don't want to, but the only way that you can summarize what the policy states at any specific point in time, while attempting to stay neutral, is to use what has been stable and refer to ongoing discussions about any portions that are under dispute/discussion. You can't force these things, you know?
—

V = I * R

(talk to Ω) 05:16, 29 August 2009 (UTC)

Everyone is making excellent points, and I hope I have time to do justice to them, but I've got a lot of pages to cover. Blueboar, I'm making an assumption that most people haven't changed their position, which means that there is still a significant contingent that would support your word "expect" and a significant contingent that would oppose it. I could be wrong; we could find out with an RFC. You are making the case that those people are simply wrong, that on most policy pages, discussion is expected before major changes. Personally I get what you're saying and think it's important, although I can point you to a bunch of pages this month where significant changes were made without discussion and it worked out just fine. But how I would choose to put it isn't the point, at the moment ... the point is that we have and always have had significant opposition along the lines that Ohms is expressing. (tweaked:) If I adopt your version for WP:Update over the last version of the month (which will mean "no comment" when there's a clear ongoing debate)), I'd be saying, "It looks to me like the debate has stopped for now in favor of Blueboar's edit". That's not accurate; debate hasn't stopped, and I expect there's a lot more ahead, and if I screw up more than a few times and represent open discussions as closed at WP:Update, this experiment will be a failure and I'll have to go back to always using the last version of the month. Your suggested language has a subtext: "The issue is closed, we won". That's inconsistent with what I'm trying to do with WP:Update; since I'm pretty sure any version we can get in the next 3 days won't be stable, I want the version I select to be the one most likely to foster productive conversation over the next month, not the one that says "In your face". I am totally aware that I may be dreaming here that WP:Update can have this kind of positive effect, but I'd like to give it a shot.

Ohms, thanks, your feedback has clarified things. What I need to do is to say at WP:Update that if I select the last version of the month, it means one of two things: either that version is stable so there was no need to reject it in favor of an earlier version, or none of the late-month versions appears stable so there was no reason to prefer one over the other. No one is going to accept me or anyone else in the role of making the call on every policy page as to whether debate has ended yet, nor making the call on summarizing the opposing positions. I'm not going to wear that hat; it's too big for me, you could fit 100 Wikipedians comfortably under that hat. - Dank (push to talk) 12:19, 29 August 2009 (UTC)

Yup, that's pretty much what I was getting at. Anyway, I wanted to ask, have you seen WP:PROJPOL? I'm sure that we could use your broad outlook there, and it could even be a means to "advertise" WP:UPDATE some more.
— V = I * R (talk to Ω) 13:09, 29 August 2009 (UTC)

I'll have a look sometime soon. - Dank (push to talk) 16:52, 29 August 2009 (UTC)

Dan, could you explain how WP:UPDATE works and what your goal is (you can do so at my user talk page if you think it will detract from this discussion)... surely this isn't the first time that a policy page change was still under discussion at the end of the month? How did you deal with such issues in the past?

As for the subtext of my comments... in many ways you are correct. There is an element of "The issue is closed, we won" in what I am saying (although I don't see it as a contest to be won or lost. I see it as simply needing to have policy reflect a shift in community consensus). The reason why there is this subtext is that the issue wasn't "won" or "lost" on any specific policy page or in a discussion... the change in consensus occured organically... essentially it occured dispite what was stated on policy pages. Now the policy pages that are affected need to catch up. There is some local resistance from a few outspoken editors to doing so. At a local level, there appears to be a lack of consensus... but at the community level I think there is a consensus, and a fairly strong one. If you want to frame this as a contest to be won or lost... I think the community consensus out weighs the local consensus, hence the "we won" subtext. Blueboar (talk) 17:40, 29 August 2009 (UTC)

My apologies, I got confused above (and struck before I saw this comment) ... it's not that I think that your edit that includes "editors are expected" was equivalent to "in your face", you chose the language you think reflects reality, which is of course fine. I'll continue on my talk page in response to your comment there. - Dank (push to talk) 18:32, 29 August 2009 (UTC)

The thing about actually writing this down is that it "sets it in stone", so to speak. I know that your perception is correct for some policy pages, and for those policy pages where the "you must discuss any change before making it" model really are in effect there is (rrently) a good reason for that. The real problem is extending that to include all policy (and, by extension, guideline) pages.
— V = I * R (talk to Ω) 23:40, 29 August 2009 (UTC)

Another concern

I just realized that there may be a larger concern, here. We seem to be duplicating, in part, what is already at WP:PGLIFE. I think that we may acutally be guilty of the oft overused WP:CREEP citation here, with the addition of a Policies and guidelines section at all to this policy.
— $V = I * R$ (talk to Ω) 02:23, 29 August 2009 (UTC)

There's clearly some overlap in the territory covered; another thing to work on. Hopefully all the things we've discussed will get sorted out by the end of September. - Dank (push to talk) 03:14, 29 August 2009 (UTC)

Yes, there is, and the more I think about it (and re-read what's here and at WP:PGLIFE), the more I start thinking that we should remove the additions, end the debate here, and go there instead. This really is the wrong forum for this entire discussion. The idea behind consensus itself doesn't change just because we're in the Wikipedia namespace, after all.
—

V = I * R

(talk to Ω) 05:03, 29 August 2009 (UTC)

I'm confused... I raised the issue here because Kim, and a few others, more or less insisted that we do so (seeWT:Editing policy), and now people are saying that this is the wrong venue.

I do agree that there is an overlap with WP:PG, and we should address this issue there as well ... As for this policy, I think it should be mentioned. While the "idea behind consensus itself doesn't change" between article space and policy space, the method by which one is expected to achieve consensus does change (as bold editing still enjoys community consensus in article space, while discussing first has consensus on Policy/Guideline pages). Blueboar (talk) 14:05, 29 August 2009 (UTC)

I obviously missed the earlier conversation, so my apologies for rehashing an old subject. I think that it's a good topic to continually discuss, however. I see what you're saying, and it is a point, but I'm still left wondering if this is the best venue to address the issue. Working in some mention of WP:PG might be a better solution...
— V = I * R (talk to Ω) 14:11, 29 August 2009 (UTC)

Actually, I just read Wikipedia talk:Editing policy#rm contradiction of multiple policies/ redundant. It's entirely possible that I'm seeing what I want to see here, but isn't User:Kim Bruning saying the same thing? "To centralise discussion, and simplify policy pages in general, the topic of modification of policy/guideline/essay pages should be merged to Wikipedia:Policies and guidelines." is what ~~she~~ he started out with, in that thread.
— V = I * R (talk to Ω) 14:17, 29 August 2009 (UTC)

Well, as I stated in the earlier discussion... I don't mind a centralized discussion... nor do I really care where that discussion takes place. The goal is to make policy match community consensus... and that means all our policies.

Since the issue seems to impact several policy/guideline pages, we probably should at least leave a note on the talk pages of these policy/guidelines, informing editors that a centralized discussion about this issue is taking place, and pointing them to that centralized discussion (where ever it is). I have left a note at WT:PG, linking to this discussion (and proposing a suggested edit to that page). But if people want to shift the centralized discussion over to that page, I have no problem doing so. Are there any other policy and guideline pages that need to be notified of the discussion (where ever it is)? Blueboar (talk) 14:45, 29 August 2009 (UTC)

I dont' think so. I can do some digging around to look though... later.
— V = I * R (talk to Ω) 15:07, 29 August 2009 (UTC)

Scope

WP:PG's scope is over policy in general. It's the most appropriate place to have this discussion, but near the start, there were objections to moving this conversation there (I have no idea why, it's the most correct place). The other issue here is that policy related to editing and maintaining policies and guidelines belongs at PG - so 'summaries' of PG don't belong here, nor over at our other policies. This causes a lot of confusion. As you can see, someone added that original paragraph just a couple of months ago, and now it's already conflicting with the actual policy. There are very good reasons for the content rules described at WP:PG#Content. M 21:58, 29 August 2009 (UTC)

Stability (2)

Slightly unrelated, but there has been some talk about stability - I think it's very important to point out that policy is stable, but policy pages need not be. The reason policies are stable is because it's very hard to sway the general opinion of thousands of our editors, it takes a long, long time. Policy pages, however, may have mistakes in them that have been in there for a few months - serious mistakes and contradictions. These can be removed right away, and "policies are stable" is absolutely no excuse to keep them. M 22:05, 29 August 2009 (UTC)

I completely agree that errors and contradictions need to be fixed ... and I see the discuss first model as the best way to achieve this. All it takes is a bit of explaining about what the problem is, and where the contradiction is... then editors are all on the same page... they all understand why the policy page needs fixing and can discuss the best way to fix it. Blueboar (talk) 22:58, 29 August 2009 (UTC)

I think that is way too cautious for errors and contradiction fixing. --SmokeyJoe (talk) 23:18, 29 August 2009 (UTC)

(ec)Of course. The message and meaning of the policy page should ~~remain stable~~ have a good degree of stability. If the words themselves are a problem, and if you know what you are doing, be bold in improving accuracy and clarity. Out of courtesy, the layout should have some degree of stability. --SmokeyJoe (talk) 23:13, 29 August 2009 (UTC)

Yes! I absolutely agree about he ideas behind policies remaining stable despite what we try to do. I was just talking about this with Dank on my talk page, as a matter of fact. Even substantive changes, despite best efforts to garnish community support and participation, are only going to have a muted effect. If it takes all sorts of debate to push through a change, it's very likely that whatever the change is, is likely to simply be ignored anyway. Unless and until Admins start blocking people based on some (ill thought out) change, people largely just aren't going to notice. The expression "Like herding cat's" comes immediately to mind, here.
— V = I * R (talk to Ω) 23:34, 29 August 2009 (UTC)

There are also shortcuts. I plan on cleaning up WP:NAME, and I'll remove a whole bunch of headings, and I won't be leaving shortcuts. I should not be prevented from doing so. As for policy pages, they describe what the community prescribes, using exhortative language. There's a crucial distinction here, too. Policy is relatively stable (note: not "should be" stable, but "is"). Policy pages reflect that, and themselves have no need for additional stability where that stability gets in the way of clarity or correctness. M 03:03, 30 August 2009 (UTC)

Well... now you've got me slightly uncomfortable, here. There may be good reasons to leave some headings and shortcuts, and good reason to remove some of them. I know that my own ideas of what is best don't always turn out to be perfect... this seems like the sort of thing we're talking about below, where people start becoming rigid in their views.
— V = I * R (talk to Ω) 02:57, 31 August 2009 (UTC)

Discussion

M, a discussion will not prevent you from making your planned edits. It sounds like you are anticipating that your edits will face resistance. That may be, but a discussion should actually help minimize that resistance. It will let others know what that you are trying to do and, more importantly, why you are trying to doing it. Also, a discussion will allow you to quickly find out why there is resistance, and address it (either by further explanation on the talk page, or by accounting for the concerns expressed in your proposed edit.) Blueboar (talk) 12:32, 30 August 2009 (UTC)

I think that a level of angst over fiddling with important things is less a problem than important things becoming out-of-date. If something can be fixed, then do it. If there is resistance, evidenced by a revert or talk page complaint, then you should ease up on your boldness. M has recently been acting boldly, generating just some little complaints. Where there's complaint, he should explain and discuss. Where there is not complaint, he must be doing OK (see WP:SILENCE). It's is possible that no one will notice some bad edits for a little while; but after a week or two it's far easier to revert everything he did than it was for him to do it in the first place. --SmokeyJoe (talk) 13:02, 30 August 2009 (UTC)

Indeed, M is now finding his work being wholesale reverted. Perhaps is would be better to decribe this observation, that if you make too much change without support then you risk seeing your efforts reverted. --SmokeyJoe (talk) 05:31, 31 August 2009 (UTC)

Replying on your talk page, Joe. - Dank (push to talk) 12:13, 31 August 2009 (UTC)

(edit conflict)I agree with what you're saying, but there's a reality to acknowledge here as well. There are cases where people become rigid in their thought, even after (or because of!) extended discussion. The fact is, Wikipedia is simply huge, and the viewpoints of people are just too varied to have actual "legislative" policy. There seems to have been a move to more prescriptive policy recently, and I admit that the pull towards that viewpoint is compelling, but I don't think that it will stand for long... anyway, yea, I agree but with caveats.
— V = I * R (talk to Ω) 13:12, 30 August 2009 (UTC)

I agree that some people can be rigid in their thought... but that happens no matter what method we use to edit... in the "edit boldly" method, rigid thougt is expressed in revert wars. In the "discuss first" method it is expressed in people talking at each other instead of with eachother. I think we would agree that both are wrong. The key is to somehow break the rigidity, and the only way I can think to do that is through discussion.

The point of "discuss first" is that you can often avoid rigidity in the first place... people understand what you are trying to do (and why) and you understand their concerns (and why they have them). Blueboar (talk) 13:54, 30 August 2009 (UTC)

A rule that suggests "discuss first" does not avoid rigidity, it condones and offers support for it. There have been numerous examples presented, too, where 'discuss first' would yield a selection bias for objections. People who support are Silent, yet anyone can think of an objection to doing something sensible, if you ask them for one. M 21:08, 30 August 2009 (UTC)

What M said here is basically where I'm coming from s well. Saying "You must discuss first" is just... an attempt to control the outcome, basically. Really, the worst part of it is that it ignores the fact that edits are discussion. Their not a particularly good form of discussion, but they are a form of communication nonetheless. If you look at it in that light. then the arguments to prevent changes before discussion end up being turned around against the whole proposal to require discussion first.
— V = I * R (talk to Ω) 03:03, 31 August 2009 (UTC)

Concerns

I absolutely agree that if there are concerns brought up, they should be addressed before trying to force the changes through. But how exactly am I supposed to address the concern that, and I quote, "changes [were] made without discussion"? Do policy maintainers need to work in bands now? When this reason is given, it is nearly always the only reason given. This indicates that people actually believe that "you should have discussed first" is on its own a good reason to revert, even when they apparently have no other objections. It needs to be made very clear that this is simply not the case. M 21:05, 30 August 2009 (UTC)

Of course if everyone does discuss first, then no one will get reverted with "you should have discussed first".

Actually, I do think "please discuss this first" is a valid reason to revert, even on its own. I look at an edit summery like that as a request for information. Sometimes I need to know what the intent behind an edit is before I know whether I support it or not. By saying "discuss this first" I am essentially asking the editor who made the edit to explain WHY the edit was made. Once I understand that, I often find that I agree with the intent of the the edit, but not the wording of the edit itself. But if I understand the intent, I can then suggest alternative wording. Blueboar (talk) 22:07, 30 August 2009 (UTC)

That's circular. You're saying that if people followed the 'discuss first' rule, they wouldn't be breaking the 'discuss first' rule. The issue is whether we need a discuss first rule. We don't. When I get reverted with the reason "discuss first", I don't see it as a request for more information, I see this as a legitimate objection to the new text. "I don't know your intent" is not legitimate, "this[link] specific section has become unclear, and I've undone it while leaving the changes that I have no objection to intact. Do you mean A, or B?" is a great objection. If a reverter can't articulate an objection, why should other editors need to provide long explanations for simple actions, and play guessing games as to what the objection might be? You know what I see as a request for more information? A request for more information, preferably on the talk page. M 23:50, 30 August 2009 (UTC)

No I am saying that if people followed the discuss first model (it isn't a rule... yet), there would be significantly less edits reverted (because they would have been discussed) and no edits reverted with "please discuss this first". As for "I don't know your intent"... I not only think this is legitimate reason to revert, I think "what is your intent?" it is a vital question and an excellent reason to revert. I would suspect that half of the disagreements that take place on policy pages could be avoided if the "changing" editor fully explained their intent, and if other editors bothred to ask. Blueboar (talk) 00:12, 31 August 2009 (UTC)

I know that there are hundreds and maybe thousands of Wikipedians who feel strongly about one aspect or another of how policy gets made who might join in if their comments were likely to be read, so I really want to be careful not to talk more than my share; please don't hesitate to tell me, anyone, if you'd like for me to talk less and pay more attention to what you're saying.

Having said that: Blueboar, wouldn't page protection be better than encouraging people to revert whenever no prior discussion had taken place? Either way would require either lack of opposition or a substantive discussion before the edit could stick, but without page protection, wouldn't we be inflaming tempers to make it appear the page can be edited by anyone (which is what "not being protected" has meant so far) and then slap their wrist with a reversion when they try it? Maybe we should be talking about page protection ... which I don't think the community would accept for all policy pages at all times, but it's possible we could come to an agreement that, because policy pages represent a lot of work by a lot of people, it wouldn't be out of line with our principles to protect policy pages more often, including in cases of "slow-burning" conflicts. Would this position be a compromise that might work for you (and others)? - Dank (push to talk) 00:40, 31 August 2009 (UTC)

First, despite obviously good intentions, I think that Dank is correct that the effect of trying to force a "discuss first" policy is simply additional conflict. The worst part of that is the fact that the conflict generated wouldn't be about the subject actually in dispute, but would likely be about process, which doesn't get anyone anywhere. The only way to practically put a "discuss first" change into effect is through some sort of technical means anyway, sude as page protection or (probably more practical) Flagged Revisions.
— V = I * R (talk to Ω) 03:46, 31 August 2009 (UTC)

Agreed, not only are the discussions here good-faith, both sides make solid points. Look at the Updates for July and August, and it's not hard to find 50 changes where a requirement [although note that the main argument is about "expectations" not "requirements"] for prior discussions would probably have had a net positive effect, and 50 entries where the requirement wouldn't have helped even for significant changes (because there's no evidence anyone had a problem or was likely to have a problem with the changes) ... and if it doesn't help, then it probably hurts by impeding change. This discussion has been very helpful for me, not because someone made some argument that settles all the questions in a snappy way, but because the back-and-forth has been useful at pruning out some suggestions that might not work. An experiment: I don't think much would be harmed by protecting just the See Also sections of one or more policy pages ... we could easily justify the protection under our current understanding of WP:PROTECT because links keep moving in and out in unproductive ways. Maybe we could start there, as a way to launch a general discussion about the advantages of discussing first, in some cases? I think there's a good chance that advantages of prior discussion outweigh the disadvantages in some of these cases. - Dank (push to talk) 12:44, 31 August 2009 (UTC)

I see this as being similar to the WP:BURDEN "rule" at WP:V, Just as it is up to editors who want to add information to an article to support it with a citation, so is it up to an editor who wants to make a substantive change to a policy to explain it.

As for enforcement... With WP:BURDEN, the only enforcement is that unsoursed info might be removed, there is no penalty other than removal. With "Discuss first" the enforcement is that undiscussed edits might be reverted. There is no penalty other than being reverted.

At WP:V we do not say "you may not add unsourced info", we are saying (essentially) "you are expected to provide a source". Here, we do not say "you may not boldly edit" we are saying "you are expected to discuss first". Blueboar (talk) 13:10, 31 August 2009 (UTC)

An impressive argument. - Dank (push to talk) 13:31, 31 August 2009 (UTC)

The problem I see in saying "you are expected to discuss first" (and, I'm assuming that we're talking about using something like that in an edit summary while reverting, since we're talking about "enforcement") is that it's just dismissive. While we're here having this conversation the intent is clear. "Out there" in the day to day workings of things, where the context of this discussion is absent, it just doesn't come across. I see revert edit summaries all the time where people will use similar non-responsive replies such as "There's no consensus for this" or something equally vague. Every time I see that I'll invariably check the talk page and there's usually nothing there (about what's directly being disputed anyway). It is actually better here on policy and guideline pages (generally speaking), but that problem still exists. At the very least, start a discussion on the talk page, then revert, and provide a link to the section in the edit summary.
— V = I * R (talk to Ω) 13:53, 31 August 2009 (UTC)

But that is why I raise the analogy to WP:BURDEN... I feel the burden to explain why a substantive edit is needed rests with the person making the edit. At minimum that means leaving a comment on the talk page (ie at least starting a discussion). In both cases, if you don't meet this burden, other editors may challenge your addition/change and remove it.

Now, meeting the burden does not always end the discussion. With sourcing someone may still have a problem with the edit and may remove the material, but the burden slightly shifts as they have to explain why they are removing it (for example they may point out that the source provided is not reliable, or that does not actually support the statement.) With policy the same thing happens, if the burden to explain is met by the editor who desires a change, then the burden shifts to those who don't agree to explain why they don't agree (ie they continue the discussion).

Even that may not resolve the issue... with sourcing there may be debate as to whether the source is reliable or not... so you bring in third parties or go to RSN for advice and opinion. With policy editing there may be debate as to whether the change is a good one, so you bring in third parties and file RfCs.

But, it all starts off with the burden of citation or explanation being with the person who initiates the change. Blueboar (talk) 14:26, 31 August 2009 (UTC)

This is excellent, both as a way to "sell" some version of this proposal to the skeptical, and as a way to model how its done ... you're saying, we should use what we know about how BURDEN applies to article discussions and apply that to policy discussions, with the talk page and archives taking the place of references. Nice work, except ... with an online reference, you click on the reference, and boom ... either it supports what you're saying or it doesn't. OTOH, he talk page archives are so long that people are going to need help to find the arguments that support the current text ... if they're even in that page's talk archives and not somewhere else ... and even if they could find them, people don't in general summarize the main points when they're finished discussing, so it would be hard to understand the outcome if you weren't "there". TLDR summary: there are similarities to BURDEN, but since something unreadable is taking the place of references, the results will be less than we hoped for. I'm left in the same place: this kind of thinking will clearly make some discussions better and others worse. Can we please select specific pages or sections of pages where we think pushing prior discussion is very likely to make things better rather than worse, run the experiment for a month, and see what happens? - Dank (push to talk) 15:06, 31 August 2009 (UTC)

I admit that my analogy is not perfect. Essentially, all I am saying is that I think the person who wishes to make a change has an initial burden of explanation, and that reverting an unexplained policy edit is analogous to removing an unsourced statement. Beyond that, the analogy is not apt... while there are similarities, I agree that there are also significant differences. Blueboar (talk) 15:22, 31 August 2009 (UTC)

Okay, how would this work: we pick pages and sections of pages where we expect people will be happy with the results, and put a notice at the top and bottom of the talk page, along the lines of (and we should get this as tight as possible after we agree on the basic idea): "This month, per (link section here at WP:CON that supports this), we're experimenting with this (page or section). Any editor may post here a challenge per (linked section in WP:CON) to any addition or change, and if there's still no evidence of at least a good-faith attempt at discussion after a week, the edit may be reverted. The point here is that the nutshell at the top of every policy page, "Changes to (this page) should reflect consensus", has up to now always applied equally to an edit or to a reversion of that edit; this is an experiment that shifts the burden of proof in the direction of the person making an addition or change, not the person making the reversion, as long as they allow sufficient time for discussion." - Dank (push to talk) 15:54, 31 August 2009 (UTC)

P.S. We might want to add "This is only an experiment, currently under discussion at WT:CON" and/or "Reverters should provide a link to at least one relevant previous discussion if they're aware of any such discussion." - Dank (push to talk) 16:14, 31 August 2009 (UTC)

Another possibility: "If you plan to revert per (linked section at WP:CON), you must leave a note on the editor's talk page giving a link to the contested edit and also giving the link to (WP:CON) so that they understand that this is only an experiment designed to gather data on how to make policy discussions more productive."

Um... I am a little confused by what the "experiment" is? Blueboar (talk) 16:35, 31 August 2009 (UTC)

The experiment is to add some text to WP:CON, something along the lines you're suggesting, but add "on selected pages where lack of discussion has led to difficulties in the past". Also add what Ohms and many others are saying about what to look for as evidence of when this new approach is not working and what to do about it. Then pick policy pages where the new approach has the greatest chance of making a good impression; each policy page has a "culture" that's a little different. ~~We might or might not add text to the nutshell on all policy pages along the lines we seemed to agree to earlier.~~ Then we can come back to this at the end of September, gather data, and decide whether to expand or shrink the number of pages affected. - Dank (push to talk) 17:01, 31 August 2009 (UTC)

Concerns (2)

What are the effects of the change? We often consider such 'effects' when writing policies, taking into account WP:BEANS and wording things in ways that influence behavior. Let's get a bit more information about what we want to encourage and discourage:

Discourage people from editing policy pages as if these changes could then be used to settle their disputes.
Encourage cleanup, even substantial cleanup.
Prevent editors from upsetting careful wording.

I've been working on policy cleanup, and my view of the last point is very negative - it amounts to "revert editors when they remove wording that managed to sneak into unrelated policy when someone added a rambling summary a year or so ago, and that you've found helpful in shutting up people who disagree with you". What I personally would like is this:

Prevent editors from ever being reverted based on procedural reasons ("you didn't discuss first", "the wording has been here a long time").

I would be fine with adding the cautionary wording if the above was very, very firmly stated. People need to supply reasons and objections - if they can't articulate one, or refuse to, how re we going to reach consensus on anything? M 21:15, 31 August 2009 (UTC)

RE #2, substantial cleanup. I recommend that you don't do such extensive cleanup in one edit, or even in day. Stuff gets wrongly thrown out in sporadic clean-ups. People can take insult ("we've all maintained those words for a long time, you're telling us we've been pretty stupid, and wow do you know better"). Surely you had given the clean up some consideration, why couldn't you have shared some of you thinking along the way? --SmokeyJoe (talk) 21:35, 3 September 2009 (UTC)

Blueboar, correct me if I'm wrong, but you would like to go all the way to "On all policy pages, always, the burden is on the person who makes a change, not on the person who reverts that change [after a week or so, probably?] for failure to make any argument or point to any discussion." [There's a problem with knowing in all cases which edit is the "change" and which is the "status quo", but leave that aside for the moment.] Correct me if I'm wrong, M, but from what you just said, it sounds like you would prefer an equal burden on the person making a change and the person reverting that change on policy pages, in all cases, which is how "changes should reflect consensus" has usually been interpreted. (Btw, some version of "When editing this page, please ensure that your revision reflects consensus" has been in the {{policy}} template consistently since at least January 2007, and it's often been quite helpful in settling disputes. From Sept 2008 to the end of July, the sentence before that one has linked some word ("normally" or "try" or "attempt") to Use common sense, which was a popular addition; it's too early to say if the removal of that link is going to stick.) - Dank (push to talk) 15:19, 1 September 2009 (UTC)

It's interesting that you mention {{Policy}}, as I was thinking about that myself earlier...
— V = I * R (talk to Ω) 15:53, 1 September 2009 (UTC)

I would say there is an equal burden to explain... On both the person making the initial change and those who may disagree with it... however, the person making the initial change has the initial burden. They need to open the discussion. That said, in situations where a policy has sat in a stable condition for a reasonable amount of time, and someone then comes along and reverts to a prior version... I would call that revert an "initial" change. What that reasonable amount of time may be is an open question, but I think we would all agree that there comes a point where a stable version is assumed to have consensus (per WP:SILENCE). A revert after that point would be indicating a need to re-evaluate that consensus, and thus a "change". Blueboar (talk) 16:26, 1 September 2009 (UTC)

It's not up to the person making the change to counter every single possible objection. The very point here is that they may not know what these are. Often, just making the change is enough explanation - "oh, the editor thinks that this word makes it more accurate". Their burden is satisfied through making the edit. The objector, on the other hand, is expected to have a clear and definite objection. Can you find an edit where a person made a change to policy, and did not provide in your view an adequate explanation? Because it seems that the only valid changes to policy are the ones with "as per talk" in the summary. M 22:07, 2 September 2009 (UTC)

Well said, that expresses my own general view nicely. "As per talk", and similar, should only be used during or after some sort of dispute, anyway.
— V = I * R (talk to Ω) 01:16, 3 September 2009 (UTC)

re: M's last: I don't disagree... for one thing, it is not possible for anyone to know "every single possible objection" that might get raised. But, I do think it is up to the person making the change to explain why he/she thinks the change is needed before he/she makes it. If they do that, then it is then up to those who may object to explain why they object. The burden of discussion is shared... but it must be initated by the person wanting the change. Blueboar (talk) 22:15, 3 September 2009 (UTC)

One experiment we could try, Blueboar, would be an editnotice (or WP:FILTER if there's some technical problem with the editnotice). It might say something like: "This is a policy page. If one of your edits is reverted, and you start a discussion on the talk page and don't get a prompt and respectful response, please post a message at (WT:POLICY, WP:VPP maybe?)". If we go this route, I think we should avoid making promises or telling people how we expect it to work ... just give them enough so that they can infer that maybe they're supposed to be leaving a message on the talk page if they get reverted, and let them know that the community wants to help if things get stuck. - Dank (push to talk) 01:08, 4 September 2009 (UTC)

Hmmm... several of our policies and guidelines have Noticeboards, places where one can ask a question about the policy as it relates to a specific situation... and if needed get assistance with conflicts from a neutral third party... perhaps we need a WP:Policy/Noticeboard as a centralized place to ask questions and seek neutral assistance in resolving edit disputes on policy pages. Blueboar (talk) 02:16, 4 September 2009 (UTC)

I'd help cover it if we had one. - Dank (push to talk) 02:30, 4 September 2009 (UTC)

WP:VPP is supposed to serve that role, right now. I like the idea of adding an edit notice to policy pages, though. That sounds like an excellent idea, to me.
— V = I * R (talk to Ω) 03:49, 4 September 2009 (UTC)

It would be nice, but VPP isn't serving that purpose; none of the current threads at WP:VPP involves a conflict that I recognized from a policy talk page. FWIW, my preferred place to do this would be the policy talk pages themselves, as long as there's a page we can watchlist somewhere where people can drop a note saying that someone got reverted and isn't getting any response to their talk page message. VPP would currently be the logical place for this, although it gets a lot of edits and that might turn some people off who would otherwise be willing to help with policy-mediation-type issues. - Dank (push to talk) 11:49, 4 September 2009 (UTC)

That people don't use VPP as it really should be used is no excuse, though. The reasons for that board are written right at the top...
— V = I * R (talk to Ω) 14:25, 4 September 2009 (UTC)

I'm not opposed. Let's start with WP:VPP as the default, then, and if people either aren't willing to ask or aren't willing to answer there because of heavy unrelated traffic, then we'll pick another page. - Dank (push to talk) 15:12, 4 September 2009 (UTC)