User:Piotrus/Morsels of wikiwisdom

From Wikipedia, the free encyclopedia
Jump to navigation Jump to search
When I am relatively happy with a section, it will be "locked" with the slight green background. All other sections are a work in progress and may not represent all or even most of what I intended to say in those matters. Each section has a corresponding discussion section; feel free to suggest improvements there.

Over the years, I've learned several interesting things about Wikipedia. Let me share them with you:

Why edit warriors can win[edit]

On the most dangerous of mindsets[edit]

On radicalization of users[edit]

On the perils of anonymity[edit]

Why good users leave the project, or why civility is the key policy[edit]

On the spirit and the letter of Wikipedia[edit]

Mud sticks, or on activity of editors[edit]

On cabals, canvassing and cooperation[edit]

Model of mass radicalization and conflict generation[edit]

On why good admins don't exist[edit]

On why so many admin heads are seen sticking in the sand when push comes to shove[edit]

On the importance of wikipolitics[edit]

Advocates needed[edit]

On why the little editor doesn't matter (but should)[edit]

Puck: "What are you going to do about it?"

This cannot be over-emphasized: mud sticks[edit]

When to use the banhammer - and when not to: a simple math[edit]

On copyright paranoia[edit]

Please comment on this section here.

Copyright paranoia is evil.

Having said that, I do not mean we should accept copyvio images. In fact, I appreciate Wikipedia's attention to enforcing the copyright; the ensuing inconvenience educates people about the impracticality of the current regime. However, if the copyright is unclear (usually because of an obscure law being confusing, or owner being very difficult to trace) I support erring on the site "copyright expired". What annoys me the most is that we delete images where there is virtually no chance of us getting sued. Images of anonymous authors, dating to pre-1923, are deleted. Consider the following examples:

  • Template:PD-EU-no author disclosure and Template:PD-Poland. In Poland, it seems that unpublished photos of anonymous authors have eternal copyright, as one has to assume that the author is still alive. So it doesn't matter that a photo was taken in late 19th century, the wording of the law "protects" the author, and prevents us from using those images till kingdom come. Perpetual copyright? It already exists.

I often hear the argument that WMF cannot afford to get sued. I do not buy it, not without seeing an argument backed up by numbers saying that an average suit costs X, and we would not X from the enraged worldwide community. IF we ever get sued over copyright, I expect that it will be a very beneficial suit - both for Wikipedia (WMF) and the free culture movement. Such a suit is likely to be covered not only from WMF budget but from other sources (EFF, CC, and similar organizations), should increase the public understanding of free culture, and should generate good publicity for Wikipedia/WMF/free culture (and bad for the whoever is suing us). We should not push for such a suit, but if it happens, I think we should embrace it as a chance to change the world.

For similar reasons, I don't see why we should be so unfriendly towards fair use.

I find it sad that our policies support being very defensive and scared of the copyright laws. They are wrong, and we should be in the forefront of challenging them - just like we did already, by adopting free licenses and showing how great of a project we can make. Time for a little more civil disobedience, I say...

More may follow :)

On conflicts of interests and paid editing[edit]

On what I learned editing Eastern European content area for over five years[edit]

Please comment on this section here.

I have had a lot of time to think about recent and not so recent events, and I've distilled my thoughts into the following analysis. I tried to compress my 5+ years of EE experience and several DR proceedings into it, I hope you find it useful.

WP:EEML was I think the 4th major EE arbcom that I recall, but I know there were some smaller ones as well. The big ones seem to reoccur regularly around fall. There is an English proverb that seems quite applicable here: "Once is happenstance. Twice is coincidence. Three times is enemy action." (see also the "1 2 (3) Infinity" principle)

The enemy here is not "the other side" (and generalization into two sides is a major oversimplification anyway). The enemy here is "us" - all the editors who became involved in EE areas and became radicalized over time. Sure, some are worse than others, but nobody here is a paragon of shining virtue.

What needs to be done to end this vicious cycle of EE arbcoms every year? I for one have enough of them. Yet by myself I cannot end them. Even if I leave the project, nothing will change.

However many sides are out there, they have all proven over they years that they can replenish their ranks. Both leaders and man-at-arms can be banned or leave - but others will step in their shoes and the "battleground" will continue. Those who left only serve the new generations as martyrs - "remember X and Y who were chased off by the others!".

Hence any arbcom remedy that is as crude as a block/ban is futile. Few months will be just a delay, indef means a martyr whose role will be taken up by another, or an infestation of socks. Worse, a desire to block an opponent one cannot deal with in terms of content creation leads some to actively engage in wiki-harassment, often combined with a wikilawyering mentality, often contributing to the cycle that ends up at arbcom (see a more detailed analysis of that problem here).

Amnesties, warnings, admonishments and their ilk are futile as well. Those warned will think they got off easily and can resume their actions, probably in a more sikrit and organized way. For every person who is scared off, another will step in, and the scared off (reformed) editor will be seen as semi-matryr/coward, and peer pressure will be put on him to rejoin his brethren.

Is there no hope? Not quite. I do think that a new type of solution needs to be implemented, one more complex than "block them all" or "do nothing". Major pieces of the solution puzzle have already been suggested - and this is not the first time they were. In fact, Irpen Alex suggested something akin to what I am suggesting last time... shame his idea was never taken up (I should've supported it more strongly then... if I did... sigh).

What I suggest this time is a several-pronged approach to deradicalize editors:

First, a mediation.

Mediation and collaborative content creation (which should be the goal of it) should reestablish trust between editors, make them see that the other guy is also an editor who wants to help. Mediation should also be combined with mentorship. There should be enough mediators in MedCabal and MedCommittee to step up to deal with a bunch of editors if the Committee asks them (2-3 mediators should be able handle such a large case), and finding a few mentors/coaches (many who could oversee multiple editors) should also be doable.

Second, an EE noticeboard.

Noticeboard should eliminate the need for "call to arms" in partisan forums.

Thirdlly, collaborative content creation.

Assuming good faith, editors are radicalized because they see others around them radicalized and engaged in wiki-combat. Showing them that those they consider "enemies" can in fact be their partners in buildning the encyclopedia is a major milestone in changing that mindset.

Fourthly, voluntary (or not) restrictions.

Restrictions should penalize editors and limit the chance of further problems. Somebody cannot help but to revert too much? 1RR. Incivil? Civility parole. Loses cool when dealing with editor X? Interaction ban. Loses cool on article Y? Article ban. And so on. Crucially those restrictions should last until mentor/coach/mediator thinks the editors they are overseeing have shown enough good faith and restrain to be allowed to regain their former freedoms. If abused, it should be easy (no red tape) to reinstate restrictions or block them. The editors should know they are watched and will be watched by the community for a long time (there should however be also a watch for those who attempt to bait them, and such editors should be put under severe restrictions as well).

The restrictions should also include measures aimed at preventing forum/block shopping, in particular, bans from reporting others to places like AE (without support of uninvolved admins, perhaps) and in extreme case, interaction and dispute resolution bans. I discuss the logic behind this and a more refined implementation idaes here.

Fifthly, John's community service idea is also good.

Community service should create a way to penalize editors while teaching them about the project (telling them "this is not what you may want to do but it is what project needs") without turning them away from the project ("you are blocked - we don't want you here") and making them felt betrayed by "the system". Of course it means that blocks and community service should not be combined (editors who refuse to do community service should be blocked, but why not give them a chance to do something constructive first?).

And those who refuse to acknowledge they need to join mediation/take up restrictions/mentorship/community service and so on may end up getting banned. Some people cannot be reformed.

Such restrictions should not only target one side, they should affect all editors involved in the EE issues. To determine who should be affected, lists of editors should be submitted by all parties (but they should not affect editors who were on the administrative fringes of the business, just those involved in content disputes).

This approach has multiple benefits, primarily:

  • deradicalized editors can ensure by themselves that the battleground will not reappear, and can later coach and mentor newcomers, turning them not into future warriors, but good editors. Thus the troublemakers are turned into gatekeepers
  • experienced editors will not burn out/be forced to leave, but will continue creating content (and this is the primary goal of Wikipedia, after all)
  • users who fail to improve under that scheme ("extremists") will still end up banned, but the moderates will be reformed (instead of allowed to radicalize into extremists or chased away)
  • obvious but: the EE battleground will finally end

To summarize: blocks/bans and amnesties were tried and failed. If we don't want to see this dramu repeat itself in the future, with slowly changing caste but the same set of issues, wasting everyone's time over and over, we need a new paradigm to break the cycle and to deradicalize editors.

Such approach is indeed our best, last hope.

For a more theoretical and hyperlinked version of this see my mini-essay on "Model of mass radicalization and conflict generation".

See also: #Model of mass radicalization and conflict generation, and Moreschi thoughts (permalink, post has many other good essay links). Also Wikipedia talk:Arbitration/Requests/Archive 2#Would it be within Arbitration Committee scope...

On wikilove, or the importance of positive reinforcement[edit]

Please comment on this section here.

There are two aspects to positive reinforcement: 1) rewarding editors who do good work and 2) rewarding reformed, deradicalized editors.

With regards to 1)

This cannot be over stressed: Wikipedia:Kindness Campaign and similar projects need to be supported; they are a vital glue that holds our community together. Editors who contribute to Wikipedia do it because they find that the project is worth of their time, and that it is easy to do. But over time, they encounter problems, and in order to counter them, they need to be de-stressed, and relax. To that end, showing other editors that we (the community) cares about them, appreciates their hard work and wants them to enjoy their stay is a crucial strategy.

With regards to 2)

From our article on positive reinforcement:

It is in everybody's best interest that the reformed editors stay that way, and positive reinforcement is a good way of achieving this (accompanied, of course, by the threat that relapse will lead to more punishment). This threat is not enough; the positive reinforcement is needed if we don't want to loose those editors; as simply punishing volunteers likely results in them loosing motivation to contribute to the project and leaving (remember: people contribute to Wikipedia not because they have to, but because they want to, and few people are gluttons for repeated punishment).

See also: Wikimedia Strategy Taskforce for Community Health Recommendation #1: Volunteer Recognition, Positive feedback works for editing, say Wikipedia editors, SBU study finds informal awards contribute to higher Wikipedia participation

On wikistress, or the importance of negative reinforcement[edit]

Please comment on this section here.

As I noted in my very first essay here, some cunning editors have mastered the art of negative reinforcement: aiming to chase away the editors they cannot deal with in civil, rational debates, they instead engage in a long-term (months, years) campaign of personal attacks, incivility, stalking and harassment (or, in wiki politically correct speak, wikihounding).

Unfortunately, it works: mud sticks and people have limits on how much abuse they can take. And the result is a net loss of editors to the project. Why stay and contribute if what you get in return is (more and more of) stress, angst, harsh words and a reputation of a troublemaker?

Solution: 1) be nice and encourage niceness and 2) enforce WP:CIV

On ArbCom failings, and how it can be reformed[edit]

Please comment on this section here.

Also, on AE failings ("admin lottery", etc.). See this.

On the distressing trend of editors leaving Wikipedia[edit]

Please comment on this section here.
Note: I am aware of both the cries that the editors are leaving and that all is peachy. The truth, of course, is somewhere in the middle...

(English) Wikipedia's population of active editors has been stable, for the past several years, at around 50,000 (see Wikipedia:Wikipedians for details). You could say: stable is good, right? No.

First of all, we need more editors - just look at the number of inactive WikiProjects and various backlogs, not to mention the systemic bias of the work being done; there are many areas that need more help, and the only way to get it is to have more editors join the project. Sure, you can say - "content is growing, eventually we will do it all, it's just a matter of time". Leaving aside the fact that if the trend of editors growth has now leveled off it can be seen as a warning before the number of active editors starts to diminish, the sheer amount of work needing done (I estimate that our content coverage is barely 1% of the "sum total of [encyclopedic] human knowledge") means that (assuming human knowledge does not increase...) we still need something like 10,000 years to do a good job covering it. Forgive me, I am slightly more impatient than that... :>

Further, the situation is more dire than the word "stable" gives justice to. Slice it or dice it how you want, the bottom line is that since we are still gaining new recruits it means that somebody is leaving. Think about it for a second. Who is leaving? Some newcomers, and some oldtimers. Ideally, we should not be losing anyone, and the loss of oldtimers is doubly worrisome. A significant number of those who leave are experienced editors who contributed a lot to this project - and that are not that many of those (Wikipedia's editors fit nicely into the Pareto principle: (~10% of editors create 90% of content)). While this is somewhat anecdotal, I believe that each experienced user can name at least one respected, quite active colleague who is no longer here... I can name several :( Bottom line is: we are losing experienced editors and gaining n00bies. Now, I have nothing against the newcomers - I welcome them with open arms - but for the project it means that the number of active editors in fact hides a number of editors who are "learning the ropes" instead of building the encyclopedic content (efficiently).

We now have some hard numbers of contributors leaving. Why do the editors leave? 38% of editors who left point to unpleasant atmosphere and "some other editors" as to the reasons they left; that number goes up to 61% for editors with 10+ edits a month. I discussed the theory of that above: #Why good users leave the project, or why civility is the key policy (and linked essays).

If we don't improve the atmosphere around this project, and make it more welcoming and conductive to editors sticking around, this project, at best, will never reach its goal ("sum total of human knowledge"), and at worst, will eventually stagnate and start deteriorating.


On the old and new generation of editors[edit]

Please comment on this section here.

I do wonder if it is a good title. What I mean is - there are two types of editors: those who respect WP:IAR and those who don't. For the first, improving Wikipedia matters more then other rules; for the others - it doesn't matter what the goal is, if a rule was not followed, the editor who broke it must be sanctioned.

I have a gut feeling that in the old days, the first few years of Wikipedia's activity, a higher percentage of editors respected IAR than today. It is just a gut feeling, though, based on some hard-to-express observations of how the community has been chancing, and what others wrote on those subject.

A study of WP:IAR references over the years could be quite enlightening...

On why the most active and dedicated editors are seen as a danger to the community[edit]

Please comment on this section here.

There is an interesting study that shows that most dedicated volunteers are a common target of attacks from within their own community. There are two major reasons for such attacks: 1) some see the very active and selfless volunteers as raising standards too high and/or making regular volunteers look bad; 2) others see them as social rule-breakers; deviants, weirdos whose high and selfless activity is "just not right", they are threatening the community with their very existence (they are so different, there must be something wrong with them, they are not like regulars, and thus must be chased away).

In our wiki environment, this is compounded by the mud-sticking phenomena I described earlier.

Further, the project was built by the uber-active editors (10% of editors created 90% of content) ([2]). But that 10% (or perhaps, the top 10% of that 10%...) is being the target of negative reinforcement...

On deradicalization of editors[edit]

Please comment on this section here.

On citing every sentence[edit]

This time I took a rare digression into an issue dealing less with behavior of editors, and more with content quality. For that reason, I've decided to split this into its own essay here.

On the importance of under-appreciated WikiProjects[edit]

WikiProjects are an ancient Wikipedia institution, dating to the very beginnings of Wikipedia (the idea was first put forward in September 2001). It is my belief that WikiProjects are of vital importance to our community; at the same time, they have never been a focus of much attention from the WMF (with regards to research, or general support), or from the community. This seems to be a major blindspot in our vision; we are constantly looking for how to improve our content, and editors' number and retention - yet we are ignoring one of the best tools at our disposal (yes, the WikiProjects).

Let me pose several hypotheses that I believe should be studied:

  1. WikiProjects are conductive to developing friendly and motivating atmosphere, thus enforcing editor's retention ratio;
  2. WikiProjects are conductive to developing content in their area (there should be a correlation between activity of a WikiProject and quality/quantity of related Wikipedia content);
  3. general participation in active WikiProjects is relatively low, with only few significantly active editors making most WikiProjects alive. It is likely that most active WikiProjects are run by only several dedicated individuals;
  4. most WikiProjects are inactive, as the few editors who kept them alive became inactive, and where not replaced by a new generation. As of 2 April 2012, Category:Active WikiProjects had 687 entries; semi-active, 115, defunct and inactive, 53 and about 350, respectively, but in my personal experience, most WikiProjects classified as active are semi-active or worse.

If the above hypothesis are true, there are several quick conclusions. First, much of Wikipedia content is here because of WikiProjets. Second, WikiProjects exist thanks to only a small number of editors. Therefore, the project would benefit immensely if WikiProjects were even more active. Finally, even a small increase in WikiProject membership (one - two active editors for a project) would have a major impact on the content.

The following research questions should be asked:

  1. ) why are you / aren't you involved with a WikiProject
  2. ) what makes your WikiProject active / useful? (camaderie? tools?)
  3. ) what made you stop your involvement in a WikiProject? (I predict that the answer is most likely: there were not enough active editors to make it interesting)
  4. ) how to the most active WikiProject function? What are they, and what makes them successful?

The following are my suggestions to WMF/community at large. We should:

  1. ) recognize the positive role WikiProjects play in our community (support "WikiLove" towards WikiProjects); the goal here is to increase the editors willingness to be involved with WikiProject by making them feel proud/good about their involvement;
  2. ) develop tools to make them more efficient. A major function of WikiProjects is likely the fostering of camaraderie (ties) between editors of common interests, but there are also some tools that make projects useful beyond mere discussion boards. For example, Wikipedia:WikiProject Poland has lists of most popular articles, a matrix of quality vs. importance, article news feed, new article feed, cleanup lists, notes on Poland-related article naming conventions, dedicated awards, and other tools. Wikipedia:WikiProject Sociology promises support for sociology-instructors and professionals. There is surely more than can be done.
  3. ) aid WikiProject recruitment
  • editors active in topic areas should be encouraged to join a WikiProject, and post to the relevant (WikiProject) discussion page;
  • similar WikiProjects would benefit from merging, at least as far as having one centralized discussion place, so that they can reach the active level and look active to their members and newcomers; this is particularly relevant with regards to semi-active/inactive WikiProjects, which periodically attract an editor who looks around, sees no activity, and abandons them (critical mass of active editors is likely the key);
  • there is room to boost WikiProject membership through recruiting outside Wikipedia (professional organizations); thus synergy with outreach program.
  • the welcome and check out a related WikiProject of interest to you new editor welcome template I developed at User:Piotrus/w should be popularized; it is also an example that a useful tool can be easily developed (within minutes);

On Request for Administration[edit]

Thoughts on inefficiencies in RfA placeholder.

See also: Guide to RfAs...

What Wikipedia governance is[edit]

How decisions are made?

Despite what WP:CONSENSUS would have newbies (and some old hands who should know better) believe, Wikipedia is not a consensus decision-making system. Instead, it is a deliberative democracy, which is a system that adopts elements of both consensus decision making (clearly a popular community norm endorsed by WP:CONSENSUS), but also of a majority rule. The point being, WP:NOTAVOTE moaning notwithstanding, Wikipedia has a lot of voting (see Category:Times that large groups of Wikipedians supported something and Category:Wikipedia_surveys_and_polls, which nearly always ends up with support being number of votes). Votes, however, are not equal: people who don't explain their votes may have them not counted, hence the stress on deliberation.

Of course, Wikipedia is also mostly a direct democracy, as most of the times we can make our own decisions. However, there are elements of representative democracy, as there are elected officials (administrators and arbitrators in particular) who are vested with the power to make certain decisions, and do not have to be concerned with the opinions of the people from outside their groups (see how much a non-admin opinion counts on WP:AE, or a non-arb at an arbitration case).

There are small but noticeable examples of charismatic authority (status of Jimbo), and benevolent dictatorship (Wikimedia Foundation can overrule community, and has done so, although this is exceedingly rare and reasonably limited to justified self-preservation cases such as "community cannot chose to disobey American law because this will result in Wikipedia servers being seized and the project closed").

How work is organized?

Wikipedia claims it is not a bureaucracy... what a joke. Wikipedia has tens of thousands of policy pages, from a manual of style, to policies on copyright and civil behavior.

McKeon, Viégas, and Wattenberg (2007) note that Wikipedia has “myriad guidelines, policies and rules” and “complex and bureaucratic processes [that run] counter to naïve depictions of Wikipedia as an anarchic space.” Viégas Wattenberg, Kriss, & van Ham (2007) also found that policy pages are growing on Wikipedia almost as quickly as content pages.

This is a problem: Clay Shirky notes: “Process is an embedded reaction to prior stupidity,” meaning “an organization slowly forms around avoiding the dumbest behaviors of its mediocre employees, resulting in layers of gunk that keep its best employees from doing interesting work.” Some editors (through only a small number, it appears) leave because they find the Wikipedia environment to bureaucratic.

The applicability of WP:IGNORE is an interesting question.

Spek, Postma, and Herik (2006) conclude that Wikipedia may be seen as an ultimate self-managing team. Benkler (2006, p. 104), in his discussion of peer-based commons-production model, notes that Wikipedia is the strongest example of a discourse-centric model of cooperation based on social norms. Sanger (2007) mentions Wikipedia's self-selecting membership. Viégas, Wattenberg, and McKeon (2007) use Ostrom’s theory of collective self-governance. Bruns (2008) discusses Wikipedia as a prime example of his ad-hoc meritocracy where produsers (content producers and end users) participate at will, limited only by their skills and interests. In my Konieczny (2010) I argue that Wikipedia has many qualities of an adhocracy, characterized by:

  • work organization rests on specialized teams
  • few barriers to enter or leave a team
  • decentralization / freedom from hierarchy
  • little formalization of behavior
  • roles not clearly defined
  • culture based on non-bureaucratic work ("getting the job done as quickly as efficiently as possible")

For decision making, that system appears to be a deliberative, direct democracy (with a few elements of a representative democracy)

For organization of work, the system appears to be a mix of bureaucracy and adhocracy

See also[edit]