Usability Yes. Social Features, Uh-Oh.

The thing that makes a category different from a group: there is no central talk page or noticeboard for someone who labels themselves with a category. But if you have a group, you can post a notice that says "this article has been edited by Bodnotbod and is clearly anti-(our point of view). Let's go in there and restore balance (by pushing our point of view as a conglomerate of 20 people)!"

I can definitely get your idea to delete groups, not unlike MfD. But we should aim for removing certain privileges first, at soft solutions where it makes sense. We might want to list a few things that might make a group disruptive: a non-neutral goal, or active membership who aren't neutral, or a group that has become just plain incivil.

I think it's too much to ask that people in their hearts are neutral. What's important is that the articles themselves are neutral. I think neutrality in articles can often arise from (let's say) two editors who each have opposing views, know about their side of the argument and present that argument, and the other guy knows his side of the argument and presents that. So neither party is in themselves neutral but the article ends up being so.

I'm just worried about cross article activity. It only takes 12 editors to push a new policy under stealth... or for one of those 12 editors to find another consensus that has emerged among 30 editors, and invite the rest of their group to disrupt that policy. Or to bring that group to an AFD. Or to bring that group to a discussion for merger.

I agree with you that we can't expect every individual in the group to be neutral. But the hope is that there is enough diversity of viewpoints among the active group members that it doesn't become a think tank for a single POV. If we have a group that is assembled of handpicked editors who all believe that minorities have a scientifically lower IQ than whites, then that group can use their strength in numbers to do a lot of damage. That's different from having a group where some people are card carrying communists, but they are balanced out by several active members who are libertarians.

I understand your position and I'm sure you understand mine.

They're irreconcilable, so I think we'll have to agree to disagree.

To my mind, without some feature to support people gathering in groups there is no social networking recommendation. Groups are at the heart of all social networking sites. Even Twitter, which doesn't explicitly group people, will have tacit/loose groups simply by virtue of the fact that people share friends and interests with each other.

So really the question now becomes whether you are willing to allow this recommendation to be submitted. If you're firmly against it being submitted then that leaves us at stalemate, since obviously I'm equally passionate about the opposite view.

I guess that means we'd have to draw on input from others and see what the consensus is.

As I understand it, Philippe has been calling for more input for this wiki and inviting people to look at the recommendations. So, if we do get new eyes looking at it, we can hope to see which way the wind blows.

Apologies, this post sounds like a bit of a huffy, almost ultimatum like post. It's not meant to be. I definitely understand your concerns and I enjoy your contributions and our discussions. But, as I say, I feel that it is impossible to provide networking without enabling groups (both good and bad). I hope I'm on the winning side of this argument as it's something I think could be quite spectacular. But I respect your opposition.

Truth is we know there will be "Social Features" in the future, the Foundation want it and put enough incentives already to make sure Task Forces get that point.

The debate is more around potential damages & safety break system.

We know we have MfD as safety system and also local Arbitration Committee ruling if necessary. Not sure that could deter motivated PoV pusher groups.

Yeah, for better or for worse, KrebMarkt is right. The foundation decided before they began this process that they want social networking features. Pushing back would just create a lot of controversy and anger, without any guarantee that it would stop it. And I don't want to do that. I want to make it work.

Imagine if somehow I created a group about "social networking" that seems neutral on the surface. (Pardon the paradox.) Then I look for Wikipedians with the category "Wikipedians against social networking", and recruit all 50 of them to the group. Then I post a message to the group and say "hey, me and Bodnotbod are having a discussion about social networking, please offer your opinion". Then you are debating me, and 50 people who are against social networking. Do you think that's going to represent a reasonable range of viewpoints? Do you think that's going to result in a consensus?

Nothing I've said is against fundamentally against groups. We just need a way to review the conduct of groups and sanction them if they get out of hand. Part of that is if a group takes on a fundamentally non-neutral or incivil purpose. But part of that is if people are recruited *because* they share a point of view, essentially creating an opinion-cartel. The idea is to get groups that share a common interest, but have some diversity of viewpoints. Not even total equality. Just something to prevent them from becoming a vehicle for partisanship and divisiveness. We have enough trouble building consensus as is.

Another approach is to have an approval process for groups. We don't have to make this an argument about whether to do it or not. We can brainstorm ways to do it well.

Randomran:

I want to address one point in your post, where you said "The foundation decided before they began this process that they want social networking features". That's not correct.

It's true that as research began, that emerged as a very strong recommendation from a number of experts and users, but when the process started it wasn't really something that was even on my mind. As I read more research and looked more closely, I've come to believe that it's a smart thing for us to do (even with the caveats that you and others have raised). "The Foundation" is a nebulous concept, and I can't really speak for it, but I can say this... there were no preconceived notions starting this process. Everything went on the table. Sue was very clear as we started that there were no "sacred cows" beyond the very core beliefs of the movement. Social features certainly aren't one of those. :-)

I'm going to fade back into the background now and continue reading (I enjoy very much reading the LiquidThreads posts that this task force and so many others are generating - it makes me very very happy) but I don't want there to be any misunderstanding of our motives coming into this.

Thanks!

@Philippe

I think that "Social Features" making it into the recommendations short list was close to 100% from the very beginning. You simply can't ignore the Enormous rock disturbing the flow of the river.

Just permit some of us to express theirs resentment that "Social Features" recommendation emerged clearly from a movement from Top to Bottom rather than Bottom to Top as expected from the premises.

If "Social Features" screw up, Experts and People at the top should take the blame.

That said we are moving over and prepare for the worst while hopping for the best.

KrebMarkt gives a cookie to Philippe because Philippe has a hard job ;)

Thanks Philippe. I remain skeptical. I doubt that a dozen well-reasoned and well-researched criticisms could stop the organizational inertia that is behind social networking. But I still found your post somewhat reassuring, and I know this job is hard.

That said, I don't want to stand in the way of social networking. Maybe if resources are finite. But if other priorities are taken care of, I'm okay with social networking. My concern is making sure that the foundation does it right.

I think a big part of that really is making sure that we're very selective about how group features are used, since group misconduct is actually a big part of the problem that our research identified. The other part is focusing on other aspects of social networking (e.g.: profile pages, cleaner one-to-one communication, cleaner ways to follow your friends...).

Random, I completely agree that the situation you outline were groups supported could come about and would be deeply undesirable.

I'm not sure if this will work, but I'll throw an idea out: so, going on the assumption that groups were linked to categories, I outlined a potential issue giving the example of our categories "pro-life" and "pro-choice" as relates to abortion (obviously there will be many such oppositional categories on WP).

Both of those categories are under the broader category of "abortion". So perhaps if people were unable to join either of the subcategories but could only join the broader one it would effectively mean that both sides are on the same ship and would be interacting with (and opposed to) each other. The fact that both warring tribes are in the same place means that they may tend to police each other.

Anyone see any merit in this idea? Could that work?

KrebMarkt, I actually wouldn't have so grave a perception that social features has been a top down imposition. I've been putting in a lot of the work related to it on this wiki and I'm just a regular volunteer (I'm not even an admin on en:wp which is pretty much the only project I contributed to before this Strategy wiki began).

I honestly can't remember how I first came to think about social features. It may have been something I got from the collection of interviews (some of which were with the WMF board). It may, though, have been something I picked out as an emerging trend in the hundreds of proposals that were submitted (at the beginning of the process I spent hours and hours trying to review as many of them as I could).

Either way, what inspired me to create the page social features and then draft the recommendation was emphatically not that I felt "it's what the board wants" or "it's going to happen, so somebody may as well write it up". It came from a genuine interest at the possibilities and from my sense that such an innovation would be a pleasure to use AND genuinely improve community numbers and contributions.

So, although the WMF are very keen on social features, much of the work done on this Strat wiki to that end was taken up by me; an average and unremarkable volunteer. Obviously the fact that I was able to draft the recommendation headed by quotes from those higher up or more well-known was gratifying because it gave me a rather powerful way to start the recommendation but it was coincidental; I made my draft out of a personal passion for the concept, not because I felt pressured or obliged in any way.

No problem. We can't know who influenced who and when. It's simply mutual influence between community & the Foundation. In this case the interviews acted as both a catalyst & amplifier. I consider the "Social Features" issue settled.

Personal opinion wise i'm for "Social Features" however i'm clearly against the excess of enthusiasm around it. It may be detrimental to others less glamorous recommendations.

Now all we have to do is make sure it won't screw up on "unacceptable" magnitude. Damages control duty for everyone.

@Bodnotbod I trust you in both neutrality & integrity and appreciate your efforts writing the recommendations. :)

Bodnotbod, I really like that idea. The overall idea makes sense. We can't stop people from having a POV. But we can make sure that groups aren't organized around a POV. That people with multiple POVs end up in the same group.

How do we generalize it into a statement of principle? I'm not sure. But let me try a few "assertions":

• Groups are organized around areas of content, to allow editors with diverse viewpoints to collaborate.
• Groups should otherwise be neutral and civil in their purpose.
• Groups must not take on a partisan purpose, whether through the scope of its mission or its recruitment practices.

Not sure if that's too aggressive, or not aggressive enough. I hope it's close to the mark, though. I think it describes why putting the abortion and anti-abortion people into the same group could ensure neutrality overall.

Thanks a lot for your work so far. The CH task force couldn't have pulled it off without you.

Cheers Random, the feeling's entirely mutual.

Hmm, so it sounds like we're concluding that the ability to join a "category/group" should not be switched on by default. And that, instead, the formation of such a group should be applied for and then approved. Would it please folk if I make an effort to build this into the recommendation?

I don't know that it has to be "invite only". Just that we want to keep track of who is doing the inviting, and if there is a pattern to the editors they are recruiting. A wide recruitment of active editors is good. But a systematic recruitment of editors who all share a specific POV would give the group undue weight.

But the creation of a group in the first place should definitely have some kind of approval process. And a recall process, too, for extreme cases.

@Randomran

I don't see why anyone could not join any groups as long he/she is an autoconfirmed user. There should be groups log showing who joined or left and when.

The less strict for someone to join a group and broader range of opinions it will permit within this group. So i'm against the idea of "invite only".

Definitely agree. We want groups to be very open. That said, there are people who notify users to join groups already, and we want to make sure they're not canvassing a particular POV to dominate a group.

• Groups creation should require "rationals" to be filled (Groups must have a purpose)
• Groups created should all be listed in a centralized place (no hidden group)
• Groups specific to a project should be listed to the project page (no hidden group & no concurrent groups)

There should be one group per subject/scope to avoid concurrent groups with same scope but different PoV.

Anyone is free to join any groups unless the groups scope require special privileges (example: templates admins).

Removing an editor from a group can by requested by others group members however they must not have the power to decide the removal. (that to avoid "Happy few cabal club").

There is a natural tendency for human beings to be aware of each other, keep track of each other and form groups. This natural tendency manifests itself in Wikipedia/media, and it can have a wide range of effects, both positive and negative, with a strong correlation with topics.

Introducing "social features" will strengthen such effects. The basic question thus becomes: what measures should be put in place to curb the negative effects (without affecting the positive effects); this should be itemized per "social feature" that is considered.

I have not yet seen this happening on this page. - Brya 05:27, 23 January 2010 (UTC)

It's definitely something we're discussing, and your concern is something that is raised by a lot of people.

Do you have any ideas to curb the negative effects, while protecting/promoting the positive?

Well, as I said, it looks to me that this should be discussed per proposed "social feature", in relationship to the characters of that particular feature. The logical thing to do would be to deal with this social feature by social feature. Actually, in-depth experience with that particular social feature in another environment would be very useful in identifying potential disavantages. - Brya 05:14, 24 January 2010 (UTC)

Like I said, we've been discussing the disadvantages of groups on social networking. You only have to look at some of the groups in the "beliefs and causes" column to understand what it would mean for Wikimedia projects:

• "Official Petition To Remove 'F**k the Troops' Groups"
• "FACEBOOK: STOP ALLOWING GROUPS THAT MOCK SPECIAL NEEDS AND DISABILITIES"
• "Groups called 'FUCK BULGARIA' to be REMOVED from facebook"

It's not just the vulgarity or incivility. It's the advocacy -- there groups are organized around identity, politics, and ethnicity. Even if the groups were to use more civil language, it's not likely that it would lead to the improvement of content or community. You'd have subcommunities that are organized to "fight back" against other groups, and use their strength in numbers to overwhelm normal discussion and "win" the argument. Equally as bad would be if Wikipedia divide into "pro" and "anti" groups, and the groups turn every article into a battleground.

We don't want groups to become rally points so that editors can organize a cabal with a common point of view. We want groups to be more like Wikiprojects, where editors of many points of view discuss and arrive at a consensus.