User:SMcCandlish/Wikipedia's self-management and future

Wikipedia and the organizational life-cycle

The trade-off between proceduralism and efficiency is an organization/project management issue, generally speaking. For example, the new hire that screws the pooch gets canned easily, but someone who's been part of the team for a decade does not, even if the issues they're thought to be causing are greater than those of the newbie. Losing the experience, the institutional memory, can cost more than that problem's resolution is worth in many cases. This effect has Wikipedia treat long-term and generally productive editors with a great deal more leniency and presumption of good-faith than is afforded to more recent editors, despite our closely held Wikipedia:Please do not bite the newcomers guideline. This effect also pertains to retention of, and resistance to any change in, processes and procedures even when they are failing us.

There's a cost–benefit analysis line beyond which this effect no longer holds (or rationally cannot). Project management is a [social] science. Wikipedia has not really been approaching it as one, but the functional needs of an organization eventually force a project's hand, or it fails. Part of the organizational life-cycle is the evolution of processes, and the departure or adjustment of participants, from the wild-and-wooly "visionary founding" era, in response to the changed operational needs of a matured organization. There is often more than one such line the crossing of which can trigger the organization to make a decision to either change or eject someone or something.

I've been through a lot of that (on both sides of it) at various organizations. I thus observe (maybe more easily than average) that WP has had and still has various growing pains in this regard. Just a few examples:

• A still-unresolved need to shift the organization's model of thinking toward behaving like a nonprofit with a mission and a constituency instead of operating just like a software company with a product/service and a userbase.
• Wikipedia co-founder User:Jimbo largely stepping away from an active "über-admin" role, under community pressure to do so.
• A culling of admins from the earliest days who were not really subject to any scrutiny before getting the bit and who have not used it appropriately.
• Imposition of a pseudo-legal body to bindingly adjudicate behavioral disputes, Wikipedia:Arbitration Committee (ArbCom).
• Closure of processes that failed because they lacked enforcement "teeth", including Wikipedia:Requests for comment/User conduct (RfC/U), and more recently Wikipedia:Mediation Committee (MedCom).
• Broad community takeover of a functional but formerly admin process, Wikipedia:Administrators' noticeboard/Incidents (ANI), to compensate for closure of those other processes.
• ArbCom slowly redefining its role and approach over the last few years, not always with sufficient community input (and some community steps to re-steer it).
• WP:Requests for adminship (RfA) becoming more difficult, while at the same time various "bits" that need lower trust levels (PageMover, FileMover, TemplateEditor, etc.) have been spun out from adminship to compensate (and despite all kinds of "the sky will fall!" nonsense about it from old-time admins).
• Creation of the discretionary sanctions system, which gave more power to admins to restrain other editors without community input (i.e., turning an informal volunteer militia into a standing police-and-military force, by way of analogy; another might be the evolution of white blood cells and other immunity-specialized biological processes). This was the proximal cause of greatly increased community concerns about the ease of obtaining administrator permissions – a major turning point.
• Changes to the WMF officers and board, and the fields from which they are drawn (some for the better, but not all of it, especially the increased disconnection between the editorial pool and the employees and board, most of whom are not regular editors at this or any other WMF project).
• WP shifting strongly toward a voting system that we pretend is not one: WP:RFCs are used constantly today, and subject to a lot of formal closure (increasingly based on vote-counting rather than careful policy analysis), when RfCs were originally introduced simply to attract occasional additional attention to discussions, usually without written closures.
• Opening of a large number of topical Wikipedia:Noticeboards, as the number of disputes to resolve has increased.
• Many others; this is just a smattering.

Because of our "Wikipedia is not a bureaucracy" policy, the project is more resistant to creeping officialdom than average, but is clearly not immune. And it cannot be, because increased organizational complexity overall requires increased and more specialized internal structure; this is a biological not just organizational principle.

The shrinking editorial pool

Overall, Wikipedia has a shrinking editorial pool (we do get new editors every day, but they do not make up for the attrition rate, though the rate has slowed). This is pretty much inevitable, up to a point. The "sky is falling" reasoning about it could be factors, but I think at most they are minor ones. Some of these are:

• Our internal culture has become toxic – too much in-fighting over bureaucratic trivia.
• There are too many rules.
• The systemic complexity of MediaWiki's markup language and our templating system, and category structure, and etc., are just too much for most people.
• Wikipedians are mostly cut from the same cloth (American, young, single, left-leaning, white, male computer nerds – by far our largest single bloc of editors, though not outnumbering all the others combined).

Some of these rationales are outright specious. For example, contributing at a high level has gotten easier in most respects, as templates provide more utility, and as other tools like WP:Visual Editor have come online (and their kinks have mostly been worked out). One does not actually need to be a template editor at all to be a good Wikipedia editor. And we know from long experience that a recurrent dispute can most easily be forestalled by creating a rule about it (even an arbitrary one); if a question comes up over and over again, the efficient thing to do is write down and stick to the most consistent answer arrived at by previous consensus discussions. There is also less dispute, on average, between editors today than there ever was before, less dispute in total, and less raucous dispute, because of the tightening of our dispute-resolution systems and, yes, all our additional rules. The most affecting of the above concerns is surely editorial group-think, as it both skews our public content (WP:Systemic bias) and our internal policy interpretation and problem-solving.

The real, primary reason for editorial decline is that all the "sexy" work has already been done. Meta:Eventualism stopped being a viable wikiphilosophy by about 2005, because the world was already using WP as an actual encyclopedia, not a draft of one and not a project to someday make one. The project was "positively invaded", if you will, by a large influx of excited "I can hardly believe this idea is real" people, mostly from SlashDot in the early to mid-2000s, who became a massive workforce and got the job done, over a surprisingly short period of time.

Today, we already have basic articles on most topics about which we need one. The work that is left is far more difficult and generally less appealing:

• Making poor articles good, and pretty good ones great.
• Adding additional articles to cover:
  • new events and newly notable people;
  • gender and other minority coverage gaps, and many cross-cultural ones
• Defending the work from vandalism and (more and more) from organized pushing of points of view by off-site socio-political and commercial interests.

Thus, people leave, and newer editors are harder to recruit, except for coverage of topics with substantial fan bases (new TV shows, etc.). Those being the main draw for new editorship is problematic in and of itself for quality reasons (notability problems, non-neutral writing, immature writing by kids and by amateur bloggers with no training in the craft, etc.).

However, the diehards who are committed to the project, not just temporarily thrilled by it, mostly stick around. They have a much slower drop-off rate than those with temporary and usually topically focused enthusiasm. This is why the project will continue and isn't in any long-term danger that we can discern; it's simply less of an activity hive than it was ten years ago. You don't need an army of construction workers on-site after the building has been erected; you need a few maintenance people. And we have them. We've done amazing things (with more people, back when), and we continue to do so with a smaller crew because we love the project; for many of us, it is our primary hobby.

A risk of ideological shift

The actual danger is in so many old-timers leaving that institutional memory erodes enough for the purposes/goals of the project and the mores of its editorial community to radically change direction. This actually happens to a lot of organizations. A high-profile example is the ACLU shifting from a constitutional rights and actual civil liberties organization into a fairly-far-left civil rights and entitlements organization, in the space of about a generation or two. It's quite effective in this newer role, but the change alienated a large portion of its original constituency and made outright political enemies from many of them. A more recent example is what's become of both of the major US political parties; they would be nearly unrecognizable to members of these parties from our grandparents' time (except inasmuch as some people from that generation are still living and were "boiled slowly" as the changes gradually happened).

The institutional memory erosion isn't very likely here, because of the depth to which we've documented what Wikipedia is and how it works and must work. Our internal policies have barely shifted in any substantive way since the late 2000s. The last attempt I can think of was the 2007 merger of WP:V, WP:NOR, and WP:RS policies and guideline into a single "WP:Attribution" policy, by a small but influential "friends of Jimbo" cadre. The community rejected the idea, and the WP:ATT page is only retained today as a disused overview of our sourcing policies. (If you care to examine that dispute, you can save much time by skimming the pro and con essays, rather than wading through all the votes and the multiple pages of discussions that led up to the poll, especially since the poll was manipulated to present a misleading proposal in the first place.)

As long as the community resists any charismatic "sweeping change" stuff proposed by a small group of editors next year or in 2022 or in 2030, things really should be fine. We simply cannot afford to let systems like RfA and ANI and AfD and ArbCom and RM and RfPP become so ossified they no longer function properly. There are many ways an organization can fail (entire books have been written about this).

Resistance to change – cutting both ways

So often, the community is entirely aware that some subsystem of Wikipedia is broken, the ways in which it is, and even which solutions would probably work, but is paralyzed by unreasonable fear that any change will have catastrophic unintended consequences. This leads to ossified processes worsening over time, as they slip further away from what the project actually needs at its current stage, and to gradual but uncontrolled compensatory shifts that compound the problems.

For example, the change in tenor at RfA over the last decade was not engineered by anyone or decided by explicit consensus. It just gradually degraded into an awful process hardly anyone wants to subject themselves to, because and only because the community's concerns about abuse of admin power went unaddressed any other way. Our collective refusal to make conscious changes has necessarily forced unconscious ones.

A similar and even more controversial example is us still permitting anonymous IP editors, despite the proportion of their edits today that are unconstructive, and the drain this puts on legitimate-editor productivity. Given that we require no ID or any form of credentials, and editors can be as pseudonymous as they like, and creating an account is easy and fast, we actually have no incentive to permit IP edits any longer and have not had one since ca. 2010. The community clings to permitting them for ideological reasons (an ideology that isn't even shared by many current active editors and is primarily found in old-timers). "The encyclopedia anyone can edit" doesn't translate to "the encyclopedic anyone can edit without logging in", but we (collectively, not all individually) irrationally treat these concepts as synonymous.

Something has to give at some point (lots of somethings do). It's simply inevitable. Various people will quit over it (or them, there being more than one community issue to resolve about which people feel very strongly). Editors wring their hands and pull their hair about this predictable outcome, as if it's like proposing limb amputation, but it's actually more like shedding old hair. It's a natural and necessary process of the organizational lifecycle for tired people to leave a project when it's no longer comfortable for them, and for new project members to come on board who are really into the current system and have no ingrained memory of "the good old days" (which often were not very good – the tinted spectacles of nostalgia).

WP is – mostly – naturally resistant to ideological takeover because of the depth of the editorial pool and their involvement in policy formation and stability. This is a blade that cuts its wielder when that same resistance to untoward influence also gets in the way of making necessary procedural changes, or when the editorial pool becomes too like-minded (when ideological influence of a particular stripe is taken for proper, for Truth, instead of untoward influence). I've already witnessed first-hand what happens to wikis that end up controlled by a "good ol' boy's club". The FIFO sets in hard, and new volunteers become increasingly repelled (in both senses). We can directly observe this effect on-site in a number of wikiprojects and their slide into "Me and Jim-Bob here, we own this topic." There's a legitimate risk of this problem spreading; we saw early warning signs of it in the WP:ATT dispute of 2007, a decade and a half ago.

A symptom of institutional change-resistance being counterproductive is the tendency of long-term editors to react at noticeboards as if any criticism of another long-term editor is some kind of injustice. Too often, this devolves straight into "any ends justify a means I agree with" sentiments and even verbal attacks against those reporting problem behavior (especially if they have less tenure and less skill at policy manipulation to get their way than the good ol' boys). It's clearly a combination of overprotectiveness of people with slivers of institutional memory, unhappiness that WP has become a complex project with structures of rules and management instead of the crazy, free-wheeling experiment founded in 2001, and reactionary response to anything challenging the status quo – even though the status quo is often something accidental and undesirable, a build-up of unconscious compensatory measures to work around elephant-in-the-room failing structures we collectively refuse to fix almost until it's too late. Wikipedia's growth along the organizational life-cycle requires it to force the ejection or adaptation of ill-fitting systems and participants that cannot keep up with the changing needs of the project, even if sentimentality and familiarity causes us to resist this fact.