Cascadia Wikimedians/2024 Wikimedia Summit report

I (Jmabel (talk)) represented the Cascadia Wikimedians at the 2024 Wikimedia Summit April 19-21, 2024 in Berlin. This is primarily my report back to Cascadia Wikimedians, but is also intended for a more general audience.

Nutshell: if you care at all about Wikimedia budget and governance, this is big. Barring tremendously unforeseen events, we will have major and presumably democratizing changes in how the Wikimedia movement determines strategy, and how the discretionary budget is determined.

Summary

The 2024 Wikimedia Summit was generally a success. Although the Movement Charter was not nearly as close to a final document as it should have been coming into the Summit, it felt to me like like literally everyone there came to the Summit with good will and a willingness to work very hard for three days. What emerged at the end, if still a bit murky, is much more coherent than what we started with, and could very well form a basis for movement governance moving forward. Further, it seems that (in marked contrast to a decade ago) the WMF Board and the upper reaches of the Foundation staff are reasonably well aligned with the broader movement and quite willing to hand over primary strategic decisions and a quite substantial portion of the budget to a more democratically chosen body in a relatively short time frame, limited mainly by how rapidly that more democratically chosen body can organize itself into an effective governance mechanism.

In short, the primary current holders of power wish to hand off many, probably most, of their current powers (including the bulk of the budget) and would largely let the (prospective) elected Global Council (GC) set the pace at which a structure driven by the latter takes those on. There was a short list (oddly I don't readily see it online anywhere, if you have it, please add a link) of things the Foundation as such needs to retain for the foreseeable future, such as (1) branding, (2) primary responsibility for the technical infrastructure that keeps the sites running, (3) primary control of the "banner" for the major fundraising campaign. Other than that, though, I was widely assured that even if the GC is not voted in, everyone understands that would be a vote not against democratization but against a particular structure, and the Foundation would still be looking for a way to offload many of their responsibilities to a more democratically based entity or entities.

Context: my mandate

Because Cascadia Wikimedians did not come in with any major issues of our own (we are supportive of a North American Hub, but as it turned out, individual hubs were barely discussed at the Summit, just the hub concept in general), I had negotiated in advance with the chapter that I would have two major foci for the meeting:

Advocate that people whose involvement in the Wikimedia movement consists entirely, or almost entirely, of on-wiki activity—rather than (for example) participating actively in an affiliate or (even more) traveling to international events—need to be better represented and better served.
Represent Wikimedia Commons, which did not have a delegate of its own.

In practice, a third focus quickly emerged: I seemed to have had more experience with governance structures and more familiarity with scholarship on the subject than any but a handful of people present, so I was also able often to contribute usefully about what has and has not historically proven workable.

Context: my situation

I nearly didn't get to attend the Summit. I was on the original list; I was dropped for reasons that were variously explained (budget; overrepresentation of North America), but I will say publicly that I really believe I was dropped mainly because the people making the decision to drop a few of the attendees strongly favored precisely the same people who have been in these meetings over and over. I was reinstated only after some rather extensive lobbying, which I don't feel free to describe publicly. I could lobby successfully only because of the extent to which some of my activity has not been strictly on-wiki. That is to say, the only reason I could be there to advocate for the disenfranchised group (on-wiki users) was that I was merely on the border of that group: unless you count events in Seattle that happen to involve Canadians (I don't), this was only my third national or international in-person Wikimedia community event in over 20 years of robust online involvement in en-wiki (since 2003), Commons (2004, the year of its founding), and Wikidata (2015).

In short, I could only overcome what some would see as corruption of one sort (a lack of power-sharing) with what some would see as corruption of another sort (networking through friends and acquaintances rather than official channels). This is not a sign of enormous health in the movement.

Context: The Summit

Wikimedia Summit 2024 was, deliberately, the last of its kind. Since 2009, Wikimedia Deutschland has repeatedly hosted this conference (originally the "Wikimedia Conference") in Berlin (see Wikimedia Summit#History of the Conferences for details and some minor exceptions). They have announced that they now wish to pass it on to someone else.

There were exactly two tracks to this years Summit, though Track 1 involved very numerous breakout sessions:

Global Charter (GC)
the future of worldwide meetings of affiliates.

I attended Track 1 throughout, so virtually everything I know about Track 2 is what is available in public sources. My only observation that may not be in those public materials is that the structure of the Summit was to participate in Track 2 at all you had to make that the bulk of your involvement in the second half of the Summit. As a result, virtually no one who had significant focus on the GC could participate in that discussion.

Because of that, I will not attempt to address Track 2 (future worldwide meetings of affiliates) in this document.

As at any conference of this sort, the informal discussions outside of the sessions were every bit as important as what happened in the sessions, both in terms of moving forward the main agenda and of conducting various other business (some of it resolved on the spot, some of it leading to scheduling future meetings either in person or virtually, etc.). There is really no substitute for having at least occasional opportunities for a broad group of people to meet face-to-face so that these spontaneous interactions can happen.

Preparation did not go well

I thought that, in several respects, preparation for the Summit did not go well. Before you get further into this section: yes, most of what concerned me here in terms of GC content was well addressed at the Summit; very little of it was at all addressed beforehand. I think it deserves recapitulation, but don't dwell on it, except maybe to learn some lessons about process.

In particular (as I wrote in an email report to the Cascadia Wikimedians on August 29, 2023 about the session I attended), the August/September 2023 "engagement sessions" did not seem to me at all on the mark. One thing I didn't say at that time, but will say now, is that the online session I participated in took about 60 minutes to deliver about 7 minutes worth of actual information, and had almost no opportunity for feedback or serious discussion. At one point, three panelists spent almost 10 minutes discussing how a particular section came to be added, referring to it only by its section number, before one of them thought to mention what the section was about. I found this lack of a coherent communication strategy ominous to say the least.

Summarizing that August 29 email:

As I wrote at the time on Wikimedia Commons' village pump, "Because the 'community' (vs. Foundation) involvement for the conference is entirely through affiliates and user groups -- not through 'projects' such as Commons -- there is no overt representation for Commons [or any other on-line group - JM 2024-04-25] at the Summit. When I raised this question, someone pointed out that there will be a representative of the Commons Photographers User Group (I'm not sure whom, and there is no indication on that page [I believe that is the role in which Frank Schulenburg attended - JM 2024-04-25]), but of course the focus (so to speak) of that group is much narrower than that of Commons as a whole. … [I]t concerns me that the basis for representation at the Summit seems to completely ignore the many users, probably the majority of users, whose involvement is strictly on-wiki. I raised that issue […and] the response from those running the session was somewhere between 'the train already left the station' and a massive shrug. I do not think the organizers are going to do anything proactive to address this concern," as indeed proved to be the case, hence my negotiated mandate from Cascadia Wikimedians for my involvement, as discussed above.
I saw the intent of creating a Global Council to represent the broader "movement" community as "clearly well-intentioned", but was worried about "exactly the same flaw in terms of omission that I mentioned in the section above" and expressed concern of getting "something like the UN General Assembly: a 'talking shop' with little or no actual power, thick with bloviators and boondogglers."
I noted that there seemed to me to be "a bit too much focus on structures that correspond to overt money flows (hence that failure to recognize on-wiki activity)" and that I do not necessarily think WMF has tended to spend money efficiently.
This is verbatim: "Combining points 2 and 3: WMF has its own, effective, fundraising. The only other entity I'm aware of in the Wikimedia world that has a comparable budget is Wikimedia Deutschland (it's no accident that the Summit is in Berlin). Money is power. And I have a lot of doubt about the power of any entity that is set up that does not have its own source of money. (Cue Billie Holiday's recording of 'God Bless the Child'.)"

The "engagement sessions" were not the end of poor preparation. For example:

I hope I'm not overstating, but I did not get the sense that anyone on the Movement Charter Drafting Committee (MCDC) had read any significant works on what does and does not work in creating governance structures. To my amazement, no one had read (or even heard of) C. Northcote Parkinson and his work Parkinson's Law, an easy read (I read it in high school) that, as our article about him says, "led [Parkinson] to be regarded as an important scholar in public administration and management." And this isn't just a case of "well, they'd read different books": they seem me to have approached this subject as if no one had ever done this before.
I gather that Richard Haslam (en:User:Nosebagbear) emerged early as the leader of the MCDC and that his untimely death was a severe blow to the committee and to the process, both in terms of loss of leadership and of mourning a colleague. I've never been on a committee that had to deal with something like that, and my sincere sympathies go out to all who were involved, and this alone may be an explanation for a fair amount of what did not go well.
I fully understand why the WMF wanted to be very "hands off" with the committee. They apparently provided a fair amount of material resources (I'm not privy to the details) and otherwise stayed in a reactive mode, because they did not want to overly influence what the community might come up with, or to be seen to be throwing their weight around. Still, some of the higher-ups must have seen that the committee was floundering a bit, and apparently didn't think to say, "Here's a budget to bring in some outside advisors of your own choosing to help move this along."
Very little effort seems to have been made to create any robust online engagement in the months and weeks leading up to the Summit. The one online meeting I was asked to attend in the month before the Summit was one where I apparently had been accidentally included in a group defined by certain parameters that did not apply to me (I'll spare you the details), and days before the meeting someone realized that and disinvited me. The mistake was entirely understandable, but it did head me off from setting up any such online meeting of my own while there was still time to do so. There was one other meeting I was invited to, on short notice about a week before the conference, but it was at 02:00 my time, and I was too tired that night to participate in a meeting at that ungodly hour.
As of this writing, there is still no clear statement of what in the GC represents a continuation of current policy/organization/etc. vs. what is new.

I could go on, but I believe I've made my point.

The Summit

Thankfully, the Summit itself went a lot better than that. After the formality of a welcome from Tobias Lindner, Minister of State at the German Federal Foreign Office, and a much more on-point keynote from Amitabh Behar, Executive Director of Oxfam (discussing how his organization transitioned leadership from representatives of donor countries to those of recipient countries), we settled down to business, alternating "gallery walk" sessions (where anyone could add a sticky note to a large copy of a draft document) with breakout sessions on various topics. Much credit to the facilitators from Wikimedia Deutschland, who did an amazing job of keeping things generally on schedule and on topic. My literally only criticism of their otherwise excellent facilitation was despite an initial statement in the plenary to the effect that "Whoever is in the room, those are the right people; whatever is decided that is the right decision" etc. (if anyone has that verbatim I would love to quote it correctly, please let me know), three separate times I saw a facilitator fail to accept that approach, and try to steer a productive discussion back to what had been pre-planned. In at one case, that was an absolutely wrong decision: all they did was derail a productive discussion. In the other two cases, the effects were basically neutral in terms of results. But to give credit where it is due: in all other respects, the facilitation was excellent.

One more process-related remark and then I will try to get to substance: while we got a lot done, and the group actually came remarkably close to consensus, the last significant piece of business of the Summit was not terribly well handled. There was a vote via Slido on 46 propositions. A lot of these propositions were so incoherent that I can't really believe the vote meant a lot. Three examples:

"30. Full transparency in resource distribution could include: (a) Policy; (b) Criteria; (c) Evaluation; and (d) Safety & Vetting." Who can possibly be against "transparency"? Well, when it comes to "Safety & Vetting", I can. That necessarily involves some degree of confidentiality, and I hope we don't end up putting something in the Charter that says otherwise.
"45. There must be mechanisms to protect staff from internal / external lobbying." Yeah, maybe. Sounds good in principle, but another way to say the same thing would be "Staff should be inaccessible to ordinary movement members," a much less positive-sounding way to say the same thing."
"28. The Charter should specify that the Global Council technology advancement committee should solicit input from Hubs and Affiliates in order to provide advice to WMF and other tech providers." I agree that they should solicit such advice. I doubt this is really a Charter-level matter.

Consensus on key matters

That said, I believe we reached genuine consensus on most key matters. Some of the following was formally agreed, some was not, but here were my most important takeaways (with a very few opinionated side-comments from me):

They want to bring this all to a vote very soon: in June, as I understand it.
- This is the one point on which I am in dissent. As they say, no one remembers when the PCjr shipped [because it was a legendary failure]. Yes, this has dragged on too long, but I find it hard to believe that a committee that has had this much difficulty getting its job done will suddenly become wonderfully efficient.
To ratify the Charter, it must be approved by (1) a majority vote of the affiliates, (2) a majority vote of "movement-based voters" (exact suffrage requirements TBD, I believe), and (3) a majority vote of the Board.
While this is still in flux, there will almost certainly be about a 100-person Global Council Assembly (GCA); that, in turn, will choose somewhere between 5 and 15 people as a Global Council Board (GCB) that will effectively function as a steering committee. There may be any number of other committees.
The GCA will probably have one annual face-to-face meeting and a few virtual meetings; the GCB will need something more intensive than that.
The GC needs a staff of its own, of people not beholden to the WMF Board, C-level staff, etc.
We are leaning heavily toward a policy under which no one can be both a paid employee of any movement organization and a member of the GCA. If you are in a paid position, you already have your own base of power in the movement; also, you have a built-in conflict of interest in driving discretionary budgets.
The Charter itself should deal mainly with the "evergreen" matters that we can reasonably expect will not change much over the foreseeable future. We will need supplementary documents with official status, but we want a charter more like the famously stable Constitution of the United States (even if that may now be under some threat) than the Constitution of California ("one of the longest collections of laws in the world").
We need a structure that, by design, will give us a geographically diverse and gender-diverse GCA. Grant-making will also have similar requirements for equity/fairness.
It is not as clear how we will get diversity of skills.
People generally agreed with the term limits built into the current draft Charter.
The GCA should become, over time, the main organ of movement governance, with primary responsibility for movement strategy and for at least most aspects of the budget. While WMF remains the main fundraising body, the board and executives appear to be ready to hand over much of this power and responsibility almost immediately, and to develop a plan to hand over more over time.
It looks like once the GC is established, affiliates as such may not have a formal role in selecting GCA members. More likely, members will be selected geographically (with a few "thematic" seats, such as LGBT, independent of that). Individuals will vote for representatives of a geographic of thematic constituency. We'll allow a lot of flexibility about this (e.g. a Syrian in exile in South America could choose either geographic constituency, or could opt for a thematic constituency), but each person will only vote once. I like this, because it give a lot more potential involvement to people who are "strictly on-line".

Representation and access

Besides the formal work of the meeting, I want to talk a little about representation and access. I know how valuable it has been for me to have had a chance to attend two Wikimedia Summits (2018, 2024) and a WikiConference North America (November 2023 in Toronto). I also am acutely aware that a lot of people who "should be in the room" don't have that kind of access. I had the good fortune to set up an "Open Space" meeting Sunday, April 21 that led to a pretty good discussion of this matter. Here are the Etherpad notes from that, a combination of live notes by User:Slaporte and a pass-through by me after the meeting.

Many Wikimedians, including many excellent contributors, are not inclined to go to physical (or even virtual) meetings. Many have no formal affiliation, either because of lack of information about such groups, because of geography, because of lack of inclination to join such a group, or because they have one or another compunction about the relevant geographic affiliate in their area. Examples:

Lack of information: you can easily edit on Wikipedia or Commons or whatever for years and have no idea of anything that happens off-wiki, especially if you are working almost entirely in "article space" and article talk pages.
Geography: for our own group Cascadia Wikimedians, we represent an area running roughly from the South Oregon coast to somewhere north of Vancouver, BC, but our leadership and meetings tend to be mostly in Seattle, occasionally in Portland, and rarely elsewhere. That does not leave someone living in, say, Corvallis, Oregon or Victoria, BC a lot of motivation to join.
Lack of inclination: let's face it, not everyone likes meetings. It should be possible for introverted people (for example) to stay informed and to be represented, even if face-to-face meetings are not generally how they'd do it.
Compunction: I'll leave out the country name, but one particular Balkan country has had its national Wikimedia group taken over by arch-nationalists. No one else there feels welcome to participate in the group any more.

Even as someone active in an affiliate and at the Summit on behalf of that affiliate, I think it is good that apparently in the future affiliates as such will not be the means of voting on more global matters, because while affiliates are valuable, most are not effectively particularly democratic. At best, they represent those who are inclined toward active participation.

There is a similar set of issues for representation of communities defined around activity on a particular wiki (in my case, if Wikimedia Commons as such were structurally represented in governance, I'd probably be more aligned with that constituency than with a geographic one). Someone from the Foundation remarked to me at the Summit that when (for example) Commons produces a "wish list" for tech, it remains an open question to the WMF tech team how representative it is of the wishes of Commons at large, even when some points get near-unanimity among those who participate. (Me: if you've got 10 admins and 10 other users and they pretty much all agree, in a reasonably publicized discussion anyone else could have joined, really what is the chance they were unrepresentative of that wiki community?)

Here are a few highlights, but for more detail see the Etherpad.

One very good idea that was floated was for some events—maybe not something as business-oriented as the Summit, but things like Wikimania or even a less specifically focused regional conference—we need to set aside a certain number of scholarships specifically for people who have never attended such an event before, or who have not attended one in the last five years.
When people think Wikimedia/Wikipedia, they think about our online projects. Wikipedia is much more well known than our movement. Even Commons might be more well known than our movement. This fact often gets lost in this parallel universe of conversations among offline Wikimedians and full-time Foundation employees, which are so different from the conversations among online contributors. No one said the following in these terms in the meeting, but I'll say it now: we need to remember en:Wikipedia:What Wikipedia is not, and in particular the "Community" section. This is not intended to be primarily an experiment in anarchy, democracy, bureaucracy, etc. This is intended primarily to diffuse knowledge, drawing on a basically egalitarian model. If something runs counter to that end, then it is probably the wrong way to go. And the bulk of that happens online. The rest is mainly to facilitate that.
It should be easier for online contribution to be a route to being hired by the Foundation, but one reality is that almost always that once someone is hired by the Foundation their perspective is liable to change rapidly, and they may quickly cease to be representative of whatever constituency they came from originally. Still, it would at least make for a WMF staff who were more conversant with the various projects.
Shortage of newsletters or other ways to track what is happening without reading what amount to voluminous primary sources (the various Village pumps, chats, DRs, noticeboards, etc.)
- Is there anything trying to fill this gap other than en:Wikipedia:Wikipedia Signpost? Anything in any other language?
- How to get good news out without being to PR-ish?
- How can online contributors, in particular, have more effective points of contact when they want information or involvement?
What can we do so that "insiders" more readily hear from online contributors (and, for that matter, from end users)?
How can we better incorporate online contributors into the Charter? (A question I personally feel has been far too little asked this last year or so, and it may be too late for this round of Charter work.)
How do we prevent on-wiki efforts to influence things from going into a black hole? It sometimes seems that the only ways this ever succeeds right now are through back channels. How can we make it about what you know rather than who you know?
Do we need more of a system of ambassadors?
Meta Wiki needs an equivalent of "No Stupid Questions." Ask anything, respond kindly.

One topic that kept coming up was the difficulty of navigating the content already on meta-wiki, what should be added to that, and what other means might make such information more accessible. The TL;DR is that English [or whatever language is the origin of the particular document] writing needs to be decent in the first place, and then translation needs to be done by someone with the right capacity. Professional payment shouldn't be off the table. There is a lot more in the Etherpad, but here are a few highlights:

Things like annual plan need to use much simpler and clearer English.
When you write deliberately for translation, you're forced to write clearly.
A committee should have someone in a secretarial role who writes the target language (usually English) well, and can help craft a good document.
Translation problems amplify any problems in the original.
In terms of broader understanding: for really crucial areas, we should have documents that talk about the rationale of the decisions that went into a plan/rule/etc., what ideas were considered and rejected, what prior work (from inside or outside the movement) influenced this particular document, etc.
Frank Schulenburg, in particular, remarked that the German translations of English texts were usually poor enough that even though German is his native language he sticks to the English original. His suspicion is that many of these translations were not done by native German speakers.
Generally low estimate of machine translation. Certainly it should not be used to produce documents on its own, but it can be useful to:
- Produce a draft translation to be edited
- Assist in dialogue. We do this a lot on Commons: someone asks a question in, say, Korean; we do a Google Translate to English and mark it as such, reply in English followed by a Google Translate of that to Korean and mark it as such. Ideally an indication as to whether it was just straight machine translation, or if the person using it knew the target language well enough to tweak it.
- Machine translation may not be great, but it can be a lot better than someone trying to carry on a conversation in a language they barely can use."You can run their well-written French through a translation tool and get something a lot better than that person's bad English."

There is a lot more on the Etherpad, including Stephen LaPorte's summary, which takes a different angle from mine. Quoting part of his summary here:

All participants at these events:

Need to represent the contributors' perspective, and maximize it
Need to bring info back to online contributors
We'll never be perfect, so keep in mind our bias and whose not in the room

Other Open Space sessions

Of course I could only attend one Open Space session, but I see that several others left notes in Etherpad. I've gathered those at Wikimedia Summit 2024/Open Space, and I hope someone will write some notes about the other sessions.