On Wed, Jun 29, 2011 at 1:35 PM, Philippe Beaudette philippe@wikimedia.org wrote:

*Call for referendum*:  The Wikimedia Foundation, at the direction of the Board of Trustees, will be holding a vote to determine whether members of the community support the creation and usage of an opt-in personal image filter, which would allow readers to voluntarily screen particular types of images strictly for their own account.

Yay on several levels! A referendum--- good thing. Referendums are good things. The last I checked in on the filtration project, they appeared to be acting very thoughtfully and going in a direction carefully chosen to be consistent with our values. We're all a little jittery from last time, so everyone was reading the fine print very very closely and wringing their hands over tiny bits of the wording that could trigger some sort of negative experience--- but if my impression of what I think the plan will be is accurate, I suspect this discussion is likely to find support, or at the very least, little strenuous 'mission/values-based" objections.

-- Putting aside this referendum and this actual ballot item and thinking merely abstractly about referendums and process-- ;too long to read? 1. Pick the right referendum format for the right issue. 2. Let the board members speak up. 3. Public debates also have PR bonuses.

==Formats== One of the things to clarify in any given discussion, whether this is a non-binding discussion, a poll, or whether it's a secret vote intended to be binding (with the obvious understanding that regardless of prior intentions, the board does have a "emergency veto" option). Each of these options is completely valid-- you just need to communicate which of them we're going with on any particular occasion.

A secret vote alone, devoid of rationales, gives us a very tiny bit of information-- yes or no, black or white. We know what the overall outcome is, but we have no clue why it was the outcome. A third-party poll or survey asks specific questions rather than seeking an up-down vote-- questions like "Would you personally use this feature?" vs merely "Should WMF do this?". Answers to surveys can give us more information, but of course that extra information will 'gray' the outcome, making it sometimes harder to interpret the results. An on-wiki 'poll'/discussion-hybrid, of these sort used at RFA/AFD, is our most 'default' decision making process, but this process does not guarantee any clear outcome-- this style is the most prone to "no consensus", and thus it's not as appropriate in cases where an actually black-white decision must be made (e.g. choosing board members). And lastly, there's always unstructured "talk page style" discussion alone.

One of the challenges with referendums is matching the issues to the method. When people vote, they'll have certain expectations about how the outcome will affect things. When people survey, poll or discuss, they bring different expectations. Two nearly equally supported options are 'no consensus' in a on-wiki poll-discussion, but those same sentiments seen through the lens of a vote can result in a 'winner and a runner up'.

Some of the other interesting effects-- you can always go from a discussion to a discussion + survey. You could transform a planned survey into a vote. But I suspect it will be more controversial to go the other way-- to take a vote in progress and turn it back into a mere discussion, discussion-poll, or survey. In whichever process you use, the more participation you get, the 'stronger' the result. The more discussion, the more nuanced the result. Voting gives you black or white, discussions give you all the shades of gray. Finding the 'optimum method' for any given occasion is always probably going to be tricky, especially when you factor in the language issues too.

==Show us Debates==

Another issue, not necessarily relevant to this referendum but more to referendums in general-- when there's a disagreement on the board, we really want to know the reasoning and the question at issue. The board can and often should advise the community. Certainly, this could be just via on-wiki participation in the discussion by members of the board in their role as community members-- but there are other ways of doing it too.

We don't necessarily need active, named participation-- the board could vote to release a statement written collaboratively that describes the diversity of opinion within the board. Where multiple opinions exist, multiple members could collaborate in making sure their advice gets to the community. Adding names to statements is nice because some people trust individual board members they know, but not the board as a whole-- and seeing names they know will help reassure them. I know no one on the board will want to public sign on to a statement that could wind up being controversial-- there's always the risk for blowback. I'd certainly prefer "across the board" participation, even if some of it has to be anonymous/pseudonymous/collective participation.

The point is--- in a normal election, the ballot issue is who will sit on the board, so there's a certain logic in the non-candidate board preserving its neutrality in that issue. But for referendums in general, the community doesn't have to operate in a vacuum.

If something's non-controversial within the board, the resulting process will be less controversial if you tell us it's not controversial in a way we can verify. Right now, we expect the board will keep the traditional "non-profit boardmember firewall of silence"-- and thus every referendum or issue could, potentially, be very controversial or very boring-- I think in general this ambiguity will tend to make things more controversial, as people out-of-communication with the board on a specific issue may imagine the board at each others throats, or may imagine their favorite board members secretly trying to shout an opinion but being gagged by traditions of silence and neutrality for board members. If "I" know only one board member and I see something ambiguous happening, I might 'project' my own imagined conflicts onto the board in unforeseeable ways.

If there really is a controversy within the board that's ever really so heated members really couldn't discuss it civilly, then perhaps it would actually be wise to remain silent. But my impression is that these kinds of cases are 'imagined' far more frequently than they actually occur-- if indeed they ever occur. Our board members are our best and our brightest. They can disagree without being incivil-- indeed, they can be a model civil disagreement for us. Our whole system is built upon transitioning from civil disagreement to consensus. I don't expect things to fall apart when the board admits that they, like every group of humans, have a disagreement over some details.

On the contrary, I expect civil presentation of such debates would tend to be a moderating influence on subsequent discussion. Usually, it seems like the involved-community tends to imaging far worse conflicts within the board than actually exist, while the uninvolved community tends to imagine board debates as non-existent or utterly trivial. Getting every individual board member on record, freely but civilly stating their own opinion-- on nearly all matters, this practice will tend to minimize conflict, not inflame it.

In those issues where a consensus of the board does not yet exist, we can already expect there to be community controversy, regardless of what the board members say. In these cases, board members can serve a valuable role in helping the community shape its discussion in more productive ways, focusing on what matters without getting hung up on the things that don't matter.

I also recognize that as humans there may, in fact, be issues the board can't discuss in public. That's okay too. The board is a team, and sometimes it may have to stay silent in order to ensure good teamwork. We don't want the team to splinter-- we don't want anyone beaten to death with a baseball bat sitting around the board room table, we don't want the community to hear that anyone's been nearly beaten to death with a cane on the floor of the chamber because debate got too heated. lol. But I don't think that metaphorically & facetious disaster is ever going to occur.

In all but the most extreme of imaginable circumstances, I have complete confidence that every single individual board member's opinion can be a productive part of the discussion, even if their words are as simple as "I personally have no stance or opinion on this question" or "I'm not sure what the answer should be". Just saying that little bit will help people see that it's not a big deal, the house isn't on fire, everything's okay, there's a debate, and like most of our debates, it's kinda boring once you get to the details and probably not something to freak out over. :)

==PR bonuses to public disagreement== There's a saying in at least the US non-profit world, although since I'm connected to that world it may be further spread than just that. The saying is "what happens in the board room stays in the board room". The saying gives words to a very, very old concept, that a team needs a certain confidentiality to be able to form optimum decision making-- that limited confidentiality, though a little contrary to our values, buys us a lot of pragmatic results we wouldn't want to sacrifice. Honest, frank, unquotable discussion is a good way to communicate, and board confidentiality rules are valuable in helping foster that communication.

But at the opposite end of the spectrum we have an extreme that could be called "Whoever enters the board room stays in the board room... and never comes back out."-- namely, that it's sometimes difficult to get our board members, in total, to express _any_ public disagreement, even when such disagreement plainly exists and can be inferred.

Never showing public disagreement is every bit as dangerous as never allowing any conversations to be ever been confidential.

A visible diversity of opinion on the board could have a wide range of positive effects beyond just the editor community. Off the top of my head, a visibly diverse board is far hard to attack and far less prone to being abandoned in the wake of controversy. A monolithic board perpetuates the false impression that a single board member can exert more far direct control over our project than they actually can, and this has invited all manner of attacks. Many of the worst outside PR attacks we've been subjected to have been of the form "Person X works for WMF, Person X is evil, and Person X can directly influence WM policy, thus WM is evil". I won't rehash the list of specific cases, but if you go back and look, most fit that patten in some way.

How do we kill this narrative dead in its tracks?

We can't possibly ensure every person we every talk to is non-controversial to everyone, obviously. We can never claim everyone we associate with is a saint-- we've been incredibly lucky so far, if we've had bad apples they haven't risen to the level I've heard about them and remember them now-- but every organization will, sooner or later, have truly bad apples who are publicly discovered to be truly bad apples.

The obvious way to kill this narrative is to use the clear fact-- a board member is very, very limited in what they can do, and the foundation itself is very limited in what it can do. If you have a problem with us, you cannot attack or blackmail our board members in order to make us change our minds-- and the easiest way to demonstrate that is for us to show to the whole world that our board does not agree, our board members cannot dictate our articles' content or our projects' policies.

In the US, while the content of public school libraries is occasionally controversial, but the content of a university library is virtually never controversial, even if the university is a public university. The dispute is not always about child-access-- talk to protesters and they will often tell you that the objected content is universally wrong for adults and children alike. It's NOT just age difference-- people accept an 18 year old can read absolutely anything at a university, but someone people will adamantly protest that same 18 year old reading the same content at their high school. So why then do they protest public k-12 but only rarely protest libraries at public universities?

I've been fascinated by this effect for a long time, and my conclusion is this. Protesters who object to content will equally object to it everywhere, but they only protest it in places they reasonably expect someone actually has the power to change things.

At a university library, they know the librarian cannot remove a book-- they know the the university president can't remove a book-- they don't expect _anyone_ has the power to change the free nature of the university library-- often they truly expect that nothing short of a US constitutional amendment can affect the contents of a university library. In contrast, a k-12 school is an institution that runs on a hierarchical model, people usually 'expect' there to be 'some authority' that decides what is allowed and what is not.

In yet another illustration of the counter-intuitive conclusion from game theory-- sometime the player with more power is actually at a disadvantage because of that power. A university president doesn't have much control over things, so he's free to say "I wish there was something that could be done about that horrible author, but there's absolutely nothing I can do about it. If you're still upset, call your congressperson." But since an elementary school teacher is often expected to have that kind of 'censorship power', an elementary school official is a huge target for pressure.

The further we can get away from the model of elementary schools and towards the model of the global universities, the better.

The public still seems to think that if someone can just convince Jimmy (or some other prominent leader), that leader can then just change things accordingly. There are stories upon stories of that follow this patten. This myth, laughable to us insiders, keeps cropping up when the broader society discusses us. Maybe someone in our organization has defrauded us, seduced us, corrupted us, bribed us, or infiltrated us. The common misunderstanding that keeps resurfacing is this myth that if you can just convince some person to change their mind, you would instantly convince WM to change its actions. This confusion invites bad-faith people to actually attack us to try to change things, and this confusion invites good-faith people to speculate about conflicts of interest.

One of the biggest weapons we could have against these sorts of criticisms and attacks is a wider understanding that our leaders don't have that kind of power-- which insiders already know. This narrative is fueled by confusion, it can be solved with truth. Oour leaders don't want that kind of editorial control, they don't ask for it, they don't need it, they don't have it, and they can't get it no matter how badly they want it. So-- if you have a problem with us, you can't attack the board members or the foundation staff in order to solve it.

The clearest signals we could send to kill that myth once and for all is to let people see that our board is, in fact, extremely diverse in nature and extremely limited in authority. Regular referendums with public contribution from the board will help all parties understand the reality that is our status quo-- people disagree all the time, and we like it that way.

Alec

On Wed, Jun 29, 2011 at 01:35:00PM -0700, Philippe Beaudette wrote:

*Please distribute widely*

*Call for referendum*: The Wikimedia Foundation, at the direction of the Board of Trustees, will be holding a vote to determine whether members of the community support the creation and usage of an opt-in personal image filter, which would allow readers to voluntarily screen particular types of images strictly for their own account.

This is false. Having participated: there are no questions asking whether members of the community "support the creation of an opt-in personal image filter".

Even if the question had been asked as stated, it has been rendered moot by further discussion and collaboration.

==Summary of discussion so far==

(please help if you think I'm forgetting major points!)

A personal image filter -in itself- does not cause as much trouble as you might think...

... however, finding and/or generating the meta-data required to make such a filter work has some flaws. * The data may be open to attack * Marking images may entail a POV value judgement * Marking images may require an impractically large amount of volunteer work

( For completeness: In the discussion, A small number of people have also taken the opportunity to advocate actual censorship of wikipedia and wikimedia commons; or have advocated the use of the (new) metadata in parental filters or other censorship tools. )

== Next steps ==

I suggest that we now concentrate on constructively discussing the meta-data ecology; including viability, security, and practicality.

sincerely, Kim Bruning