How to prevent edit warring

[Larry Sanger]

There are at least two other general other explanations of edit page wars that I can see--or, rather, two contributing factors: the desire to be right, and plain old rudeness.

First, there's the desire to (appear to) be right, to (appear to) win a debate.  Whatever might explain this or be associated with it, it is evidently a powerful human motivation, and a hard one to turn off.

Second, there's the outraged reaction to offended sensibilities and rudeness.  Often, it isn't so much the rudeness itself that causes the trouble, because that is put to one side quickly.  It is the fact that the rudeness establishes very clearly that you're on opposite sides, and you're fighting.

Put these together with the logistical (and logical) difficulties of staying focused on the issue at hand, and you have a fairly comprehensive explanation of why we like to fight.

[Larry Sanger]

Now, if those are the explanations, how can we prevent edit page wars?  We do that by rendering the factors contributing to verbal strife either nonexistent or inoperative.

I am toying with the general concept of making three rules, namely, (1) always to state a clear topic of discussion, (2) the topic should concern the wording of the article, and (3) only argue about that topic--never the topic itself.

The successful operation of these rules would probably only render strife inoperative--not nonexistent.

Of course, this represents a most direct attack on the first problem I mentioned above, that people don't stick to the topic.  The solution is to tell them to stick to the topic. 

Generally speaking, if we are always only arguing about pieces of text in the article, and how they should read, then while we will still want to win our debates, they will be relevant debates.  But they probably won't be quite so acrimonious, because they will not be about "the big questions" underlying an article.  They'll be about relatively boring and trivial questions, about precise wording.

Again, if we aren't engaging "the big questions," this undercuts much of the occasion for rudeness.  People get rude and hostile toward the other side when they feel their worldview is under attack.  Small pieces of text will not encapsulate worldviews (I should hope--more below).

Still, I need to say much more.  I can't simply say that we will adopt rules (1)-(3) above, period.  I need to address several other issues.  (a) What does it mean to pick a clear topic?  (b) What does it mean to say the topic should concern the wording of the article?  (c) What does it mean to argue only about that topic?  (Is there always such a clear distinction?  And is this really a reasonable requirement?)  And, of course, (d) how do I intend to enforce all this?

(a) A clear topic should be a proposition, statable in a fairly brief sentence, that you want to affirm; it is a "resolution," so to speak.  To say that your topic is clear is to say that a reader can form a reasonable interpretation of it, and not mistake your intention for something else in the near vicinity, conceptually speaking.  It is hard to "mistake your meaning."

(b) When I say the topic should concern the wording of the article, I mean that it should either propose to add some piece of text, or it should identify some part of the text, and then make a definite proposal about how it should be changed: either deleted, edited, or replaced.

Those are perhaps relatively easy questions.  (c) and (d) are harder.

(c) If I am arguing only about a proposition as described in (b), I am arguing about how the text should read.  The trouble is that some argument might get very far afield of the text and yet still be construed as being "about how the text should read."  For instance, in Exhibit 1, Citizen 1 might propose to change the wording so that Jesus' historicity is not simply asserted.  But then Citizen 3 might want to say that the actual historicity of Jesus is, of course, about that wording.  So, in what way was Citizen 3 actually off-topic?

Well, it's subtle, but I hope clear enough.  Citizen 3 was off-topic because the arguments adduced tended to establish Jesus' historicity, not that the article should report Jesus' historicity.  Suppose Citizen 3 established beyond any reasonable doubt that Jesus lived.  That would not actually support in any way at all the claim that CZ should simply assert that Jesus lived; to establish the latter, one must also establish that no serious scholar or any significant portion of the population denies or doubts it.  But in fact some serious scholars and a significant minority of the population does deny or doubt it.  So Citizen 3's argument does not bear on the question at issue.

This is perhaps a little too subtle to expect our poor beleaguered Citizens to draw out of rule (2) alone.  So we need another rule, I think, that makes it perfectly clear.  To wit:

(2a) If the proposition concerns a topic controversial in itself, it is not on-topic to argue in favor of one side of the controversy.

In other words, if you are arguing about how the article should answer a question that people get hot and bothered about, then you may not argue for one particular answer to the question.  Period!  You can argue that some people say such-and-such; you can argue about details about what people actually do say; but you can't say those things yourselves.

Now, are these rules, (2) and (2a), really a reasonable requirement?  We might agree with the neutrality policy, but still say that it is very germane to argue for one side or another, because this will shed important light on how the article should read.  We need to get the dialectic out in the open, and then we'll be able to write the article better.

Maybe this is true, but I very much doubt it.  I doubt it because I've observed many an edit war over the years, which essentially go to the issue of how the article should be biased, when it shouldn't be biased in one direction at all.  I don't think the articles are ever really improved by such debates.  But I do think that there's an excellent case to be made that they are made worse by those debates, as in the case of Wikipedia.  This could be why Wikipedia articles are so much about the topical controversies and not so much about the uncontroversial aspects of a topic.

I think we ought at least to try to focus exclusively on the wording of articles.

Still, someone might say that there is some useful background information that might need to be imparted on an article's talk page, which concerns the topic and not the wording.  Well, that's all right, if it's not part of a controversy.  But if you are giving the background information in order to argue for a point, where there is another side who will want to argue the opposite, that again would be off-topic.

This brings me to (d), or how I propose to enforce all this.

Ultimately, as a constraint, I would like to say that creating new roles is to be avoided if at all possible, and the employment of the contributors themselves will make for the most scalable solution.

I'll write another whole post about this question.

[Larry Sanger]

I think we need to have some standardized way of articulating what the topic of a talk page section is, such as a template, and/or a protocol much trumpeted and, e.g., integrated into the text of talk pages.  (In fact, I can imagine a bot or a database command appending some text to the bottom of every existing talk page on CZ, explaining the new rule.)  So, if you want to start a new talk page topic, you open a new section, and you use a template that identifies a certain proposition as the proposition you want to debate.  Then the rule is that everything in that section of the talk page must directly bear on that proposition.

What do we do, however, if someone doesn't know about the template, or thinks it's too much trouble to add it, or just ignores the rule to stick to the topic?  I can imagine two things.  First, we just assume that the general announcement, as well as the explanation on every talk page, is enough to spread the word; then, if someone breaks the rules, the other side calls a constable.  Clear enough.  Second, however, we might actually make an exception to our usual rule against complaining about other people's work, in this case, and actually encourage people to teach each other these rules and to enforce them by peer pressure.  Right now, it's against the rules to complain about someone else's behavior on the wiki; it will bring a {{nocomplaints}} template.  {{nocomplaints}} is generally a good idea, I think, because it has helped us to avoid a lot of conflict.  A lot of conflict does involve people complaining about each other.

I think I'm going to plug for the second option.  I think actually that the enforcement of this regime has to involve rank-and-file contributors talking to each other, because otherwise it might be difficult to determine what is and is not on-topic.  The contributors are the ones who will know.  Besides, if we disallow X to tell Y, "That's off-topic," this puts X at the mercy of Y's bloviations, and would very likely start the whole cycle of pointless dispute.  But if we give X an argument-stopper--"Y, that is off-topic, we have to stick to the proposition, see the rules"--then, no prob.  Then, if Y persists, X can call a constable, and moreover, there is a very specific thing that X can tell the constables: "Y is persisting in making off-topic arguments.  Please remove them."  Then the Constabulary does so.

Well, this is what I propose we pilot.

[Joe Quick]

I think I'm going to plug for the second option.  I think actually that the enforcement of this regime has to involve rank-and-file contributors talking to each other, because otherwise it might be difficult to determine what is and is not on-topic.  The contributors are the ones who will know.  Besides, if we disallow X to tell Y, "That's off-topic," this puts X at the mercy of Y's bloviations, and would very likely start the whole cycle of pointless dispute.  But if we give X an argument-stopper--"Y, that is off-topic, we have to stick to the proposition, see the rules"--then, no prob.  Then, if Y persists, X can call a constable, and moreover, there is a very specific thing that X can tell the constables: "Y is persisting in making off-topic arguments.  Please remove them."  Then the Constabulary does so.  [Ed - inserted corrected quote]

I'd love to see this work but I doubt that it will.  I would expect to see a lot of Xs abusing the rules, saying "Y, that is off-topic" in order to push a personal point of view.  And I would expect a lot of Ys to start their arguments with "It is NOT off-topic!" and continue on with even more fervor.

[Larry Sanger]

Here's a further suggestion.

I think it might be useful to categorize talk page comments into at least two categories: arguments, and explanations.  Arguments require the apparatus described above.  Explanations merely explain why somebody made an edit, without pretending to argue for it.  If someone disagrees with an explanation, he must do so using an argument section.

We might also identify such things as questions and praise.

[Larry Sanger]

I think I'm going to plug for the second option.  I think actually that the enforcement of this regime has to involve rank-and-file contributors talking to each other, because otherwise it might be difficult to determine what is and is not on-topic.  The contributors are the ones who will know.  Besides, if we disallow X to tell Y, "That's off-topic," this puts X at the mercy of Y's bloviations, and would very likely start the whole cycle of pointless dispute.  But if we give X an argument-stopper--"Y, that is off-topic, we have to stick to the proposition, see the rules"--then, no prob.  Then, if Y persists, X can call a constable, and moreover, there is a very specific thing that X can tell the constables: "Y is persisting in making off-topic arguments.  Please remove them."  Then the Constabulary does so.  [Ed - inserted corrected quote]

I'd love to see this work but I doubt that it will.  I would expect to see a lot of Xs abusing the rules, saying "Y, that is off-topic" in order to push a personal point of view.  And I would expect a lot of Ys to start their arguments with "It is NOT off-topic!" and continue on with even more fervor.

You could be right; that's one reason we need a pilot.  If there's a disagreement about topicality, an uninvolved editor or constable (perhaps either) might be called to make a definite ruling.  If someone is repeatedly wrong about topicality, and refuses to learn, he might be banned.

[Greg Woodhouse]

You could be right; that's one reason we need a pilot.  If there's a disagreement about topicality, an uninvolved editor or constable (perhaps either) might be called to make a definite ruling.  If someone is repeatedly wrong about topicality, and refuses to learn, he might be banned.

That's the problem today. What we've been calling edit wars end up being wars of attrition, and those advocating a position for strong ideological reasons will be the last to relent, and the first to cry foul if anyone disagrees with them, or (more to the point) with the approach being taken in the article.

[Joe Quick]

I think I'm going to plug for the second option.  I think actually that the enforcement of this regime has to involve rank-and-file contributors talking to each other, because otherwise it might be difficult to determine what is and is not on-topic.  The contributors are the ones who will know.  Besides, if we disallow X to tell Y, "That's off-topic," this puts X at the mercy of Y's bloviations, and would very likely start the whole cycle of pointless dispute.  But if we give X an argument-stopper--"Y, that is off-topic, we have to stick to the proposition, see the rules"--then, no prob.  Then, if Y persists, X can call a constable, and moreover, there is a very specific thing that X can tell the constables: "Y is persisting in making off-topic arguments.  Please remove them."  Then the Constabulary does so.  [Ed - inserted corrected quote]

I'd love to see this work but I doubt that it will.  I would expect to see a lot of Xs abusing the rules, saying "Y, that is off-topic" in order to push a personal point of view.  And I would expect a lot of Ys to start their arguments with "It is NOT off-topic!" and continue on with even more fervor.

You could be right; that's one reason we need a pilot.  If there's a disagreement about topicality, an uninvolved editor or constable (perhaps either) might be called to make a definite ruling.  If someone is repeatedly wrong about topicality, and refuses to learn, he might be banned.

Well let's give it a shot then.  We've been talking and talking but we haven't done anything about the issue yet.

I still think it would be worthwhile to create articles devoted specifically to public debates over scientific/religious/other contentious ideas.  I think that people who engage in edit wars will often be happy to back down if they can see that their objections (which represent one side or another of an issue) are given full consideration and discussed with care.  See my posts in the other thread for how I think this would work.

[Stephen Ewen]

There are at least two other general other explanations of edit page wars that I can see--or, rather, two contributing factors: the desire to be right, and plain old rudeness.

First, there's the desire to (appear to) be right, to (appear to) win a debate.  Whatever might explain this or be associated with it, it is evidently a powerful human motivation, and a hard one to turn off.

Second, there's the outraged reaction to offended sensibilities and rudeness.  Often, it isn't so much the rudeness itself that causes the trouble, because that is put to one side quickly.  It is the fact that the rudeness establishes very clearly that you're on opposite sides, and you're fighting.

Put these together with the logistical (and logical) difficulties of staying focused on the issue at hand, and you have a fairly comprehensive explanation of why we like to fight.

There are others.

Take lack of trust by writers in a writer who has evidenced hostility toward a subject.  Especially when a power differentiation exists between an editor and author, this can be a problem. 

So, for example, if an editor is present on the Jesus article who has arrogantly expressed serious hostility to all the Christian dupes of the world for believing in a person who he has declared never existed (I am summarizing the hostile view by using that language), as is the case with the example and editor you gave, there is going to very naturally be, and really should be, a reluctance to simply engage the person as an honest broker of the topic, as you suggest be done.

In my view, edit wars and talk page controversies on controversial article topics are an inherent and serious weakness of the strong collaboration model, more than it is anything else such as a lack of rules of discourse.  Controversial topics draw people of extreme viewpoints like moths to flames, because it is there they can hope to exercise power, control over public content so as to have an authoritative wrapping to their views.  This is why Wikipedia's "Assume Good Faith" policy is simply idiotic, in my view--as dictum, it just does not correspond to reality.  Meanwhile, resident extremists on a page, playing just inside the rules, drive away actual honest brokers.  Yes, on Citizendium. 

Publications who hand select authors can neatly avoid all this by soliciting articles from writers who are known to be honest and sympathetic brokers on a controversial topic--they don't select an avowed anarchist to write an article on neoliberalism, for example.  Honestly, I think moving closer to that traditional model for controversial topics is the best solution, although I am wide open to contrary arguments and am certainly for a pilot, as you are proposing.

[Greg Woodhouse]

Publications who hand select authors can neatly avoid all this by soliciting articles from writers who are known to be honest and sympathetic brokers on a controversial topic.  Honestly, I think moving closer to that model for controversial topics is the best solution, although I am wide open to contrary arguments and am certainly for a pilot, as you are proposing.

That's half of the equation. The hard part is persuading skeptical authors that the editors (or others?) will be fair and honest. The trouble is that if an author is determined that the earth moves around the sun in a giant square, then the most sympathetic editor in the world will still disagree, and this will only serve to confirm the author's suspicion that the editor is not treating him/her fairly. I try to go into "teaching" mode, providing explanations and asking questions on the talk page. Sometimes it works, and sometimes it doesn't.

[Stephen Ewen]

The hard part is persuading skeptical authors that the editors (or others?) will be fair and honest.

Not at all, for editors.  The persuasion is contained in the list of publications they bring to the table; or, perhaps in some cases, the absence of them.  "Assume Honest Brokerage" should not really have to be done when evidence of it (or not) is possible.

For authors pushing their determination that the earth moves around the sun in a giant square, and the like, dishonest brokers simply should not last long on the project, of their own doing.

[Larry Sanger]

Greg and Steve, your point is well taken indeed.  There's another whole way to think about the cause of edit wars--it's that they are caused for the most part by partisans.  Steve, you say that fairness is achieved in the traditional model when a good editor can find an "honest broker" to represent a debate.  That's a good point, but I wouldn't give up on the open collaborative model quite yet.

I consider whether the system I propose might not in fact solve the partisanship problem.  Even if partisans can be caused to fight only about the wording of articles, and stick to that, people will still fight acrimoniously and as partisans, not as honest brokers, at least as much as they can.

The more general problem is that Partisanship Will Find a Way.  Partisans view articles as opportunities for as it were propaganda, and they will "game the system," whatever the system is, to advance their view.  They need not do so consciously, much less as a part of a big organized partisan conspiracy (dictated e.g. by the Central Committee).  No, they merely have to take the view that they're in the a struggle with the forces of darkness and they need to make sure their position, which is the fair and balanced one, is stated clearly.  And if the other side isn't going to play fair (i.e., agree with them) then they don't have to, either.

I concede all this.  I state it explicitly merely to make it clear that I understand it.

But I do consider that worst of the edit wars, the ones that make us say that there is a serious problem to solve (like the ones noted above), might be prevented by the rules I've put forward.  Yes, people will still try to work within the system to advance their causes.  But they won't be able to do so very effectively under the rules I've proposed.  Consider again that partisans will have to state a very clear topic of discussion, which concerns only the wording of the article, and will have to stick to that topic.  This in itself will prevent some of the worst edit wars, which proliferate (as my exhibits demonstrated) because people quickly lose the thread of the discussion and wander into more controversial topics.  Also, consider my additional rule, (2a), which says, "If the proposition concerns a topic controversial in itself, it is off-topic to argue in favor of one side of the controversy."  But this is precisely what partisans so love to do, which causes edit wars to explode and go on and on.  When people start favoring one side of a controversy, that's the flashpoint of the edit war.

Privately, Mike Johnson made one of the most incisive criticisms of the proposal yet, which is that it might make the learning curve too steep.  As I design a pilot, I will bear this in mind very carefully.

[Jitse Niesen]

I think we need to have some standardized way of articulating what the topic of a talk page section is, such as a template, and/or a protocol much trumpeted and, e.g., integrated into the text of talk pages.

That may prove to be quite a hassle. If it prevents a debate which generates a lot of heat and bad feeling without improving the article, then it is of course worth it. However, the vast majority of discussions on talk pages are good-natured and productive. In such cases, adding another template (or whatever the mechanism is) and thinking about how to formulate the topic are additional efforts with no or little gain. Perhaps the template or protocol could be restricted to articles that are likely to be generate unproductive discussions? Or perhaps people should initially be free to skip the protocol and only be obliged to follow it if somebody asks for it?

Additionally, there is a danger that people are less willing to use the talk page and have edit wars on the article, while communicating via edit summaries or not at all. But a pilot will give us an idea on whether that will happen or not.

[Stephen Ewen]

Also, they may resort to email.

But here is some interesting encouragement from David Gerard, who I respect and enjoy reading, to work at the problem, some way.

The main problem with Wikipedia is the obvious problem with a massively collaborative enterprise: how to work with idiots you can’t get rid of, who consider you an idiot they can’t get rid of. Collaboration is hard.

Academia has evolved mechanisms to deal with antisocial idiots (throw them out) and antisocial experts (put them to work in a locked room and keep them away from humans); wikis are still working on the problem. Perhaps Citizendium will come up with something Wikipedia hasn’t as yet.

From: http://thearticlewriter.com/blog/2007/07/12/online-editor-extraordinaire/

[Larry Sanger]

I think we need to have some standardized way of articulating what the topic of a talk page section is, such as a template, and/or a protocol much trumpeted and, e.g., integrated into the text of talk pages.

That may prove to be quite a hassle. If it prevents a debate which generates a lot of heat and bad feeling without improving the article, then it is of course worth it. However, the vast majority of discussions on talk pages are good-natured and productive. In such cases, adding another template (or whatever the mechanism is) and thinking about how to formulate the topic are additional efforts with no or little gain. Perhaps the template or protocol could be restricted to articles that are likely to be generate unproductive discussions? Or perhaps people should initially be free to skip the protocol and only be obliged to follow it if somebody asks for it?

Additionally, there is a danger that people are less willing to use the talk page and have edit wars on the article, while communicating via edit summaries or not at all. But a pilot will give us an idea on whether that will happen or not.

Jitse, both of your suggestions are very reasonable and well motivated, and I am kicking myself for not thinking of them myself!  Sure, on totally uncontroversial topics, requiring the protocol might get in the way.  So we might say that this becomes a requirement for specified topics, when triggered by a contributor or a constable.

Well, I'm ready to start.  I think we should start with [[Bill Clinton]], [[global warming]], [[oriental (word)]], and maybe a few others.