Earlier on, I pledged to blog with a bit more clarity. And how can I reach clarity? By taking a bunch of old topics of mine and talking about them in a new light.

Therefore, it’s a high time that I post a good summary of what I feel about Wikipedia policies and how they’re applied, as they stand right now.

…OK, it turned out to be over 4500 words. I think I failed. Sorry.

This post is a vast expansion of a comment I originally posted on Laurel L. Russwurm’s blog post on Wikipedia’s deletionism.

Honestly, it’s sort of painful to read the blog post in question. The themes are so familiar, but things have changed. I find myself agreeing on points that I would have strongly disagreed on years ago. However, I also realise one thing: My actual view on how Wikipedia processes should operate hasn’t changed one bit over the years. It is the Wikipedia community and the Wikipedia’s general spirit that has changed – and not in the good direction.

Hence, I have to agree with the critics here. I’ve probably implied or even said it before, and now I say it out loud for once and for all, dammit: Deletionists are insane, mergists and inclusionists got it right.

Notability

The primary thing that governs deletions in Wikipedia is the subject’s notability.

Notability should be the other side of the coin to Neutrality. Neutrality, as far as Wikipedia policies go, means that researched facts stand on their own. Notability means that only researched facts are admitted to Wikipedia.

As you may have guessed, while both of these stem from simple principles that have been around almost as long as the whole project, Wikipedia’s actual notability criteria have had a colourful history to say the least.

Long ago, people tried to “quantify” notability. People tried to create intricate policies that determine exactly how many news sources or Google hits there should be on any subject on a specific field.

For example, on software world, people tried to argue if an open source package has been included on Linux distributions, it’s probably notable. People actually tried to base this on something like Debian’s Popularity Contest; if a package has a high popcon rank, then it has probably made a significant dent. But people started arguing about this: Debian has thousands of packages – what is the exact popcon rank cutoff? Similarly, there was huge debates on just how many films an actor should appear to get an article of their own. Since *ahem* some actors appear in more films than others due to the fact that *ahem* certain films are easier to make than others, an alternate proposal regarding pornographic actors appeared, setting the required numbers of appearances a little bit higher. And the debate raged on.

The problem is, once you start creating this sort of notability guidelines that cover every imaginable subject area, the rules become very complex and often debated. “Often debated” means that the rules are inconsistently enforced; an article that passes AfD today may not pass it next week, because the policy kept living on. And the problem is, there are always corner cases where things just don’t fall into nicely defined buckets. Simply put, rules were what you argue them to be.

Which is why it’s refreshing to see that nowadays, we have a “General notability guideline”:

“If a topic has received significant coverage in reliable sources that are independent of the subject, it is presumed to satisfy the inclusion criteria for a stand-alone article or stand-alone list.”

Notability? Sources. Sources? Notability. Can’t get simpler than this.

And nowadays, the subject-specific notability criteria have basically been rendered obsolete by this criterion – they’ve been reduced into rules of thumb. For example, the web page notability criteria is basically “General notability applies. In case of web pages, this means that we’re not a web directory; let’s write about web sites that have actually been written about in news and/or that have been subject of scientific research.”

The deletionists are essentially yearning for the bad old days when the notability criteria were numerous, complex, incomprehensible and changing every week — when there were numbers and limits and you could tell when topics were significant based on Google hits.

Deletionism kills

As I said, Deletionists are insane. And they killed the Wikipedia community spirit.

Deletionists have killed Wikipedia community because they have an overly strict idea of what notability should mean. We should keep in mind that significance of topics has to be always evaluated case by case. Deletionists just want to jump on the delete button if there’s any indication that people are not, right now, fixing the article to the strictest standard imaginable. Deletionists want to jump on the delete button if the article’s subject isn’t worth an article right now.

They want to punish the laziness of other people. Of course, this conveniently always ignores the fact that they’re also being lazy.

Why it is easier to destroy than to care

It reminds me of a discussion, years ago. I know it’s highly silly that I can’t point to the actual conversation when I usually demand sources for this kind of stuff, but I just honestly can’t remember the article what this debate was about.

“I want to delete this article. It doesn’t have any sources. In the deletion discussion years ago, I said there were no sources. You defended the article back then. Sources. Please.”

*sigh* Here you go. Very good sources, as you can see. Took two minutes of Googling. Why the heck didn’t you do this?”

“Burden of the proof is on the people who want the articles to stay. I’m not the one who wants to keep the article. It’s not my responsibility.”

Yes. Burden of the proof is on the people who want the articles to stay. And now that I think of it, it’s a shame we don’t require people to think of constructing an encyclopaedia. It’s a shame we don’t want people to want the articles to stay.

This is the problem of deletionism as it exists now: if you want to be a deletionist, you don’t have to be concerned of the primary activity of building an encyclopaedia: Adding interesting facts and finding sources for them. You don’t even have to be interested of sane sort of fact-checking or janitorial work: removing things that are demonstrably completely unsupported, or even wrong, and don’t agree with any established sources in the field. You don’t need to entertain the idea that some fact might be supported or unsupported by a source somewhere. All you need is a desire – or at very least cold-bloodedness – to destroy the work of others, and the will to look for passages or even whole articles that just don’t happen to have sources right now.

If we demanded people to be constructive, they might actually go looking for the sources. You know, if you genuinely cared about the subject – say, some obscure now-defunct website from 1996 that none of the kids these days have heard of – you’d be aware of the literature and press. Oh, goody, an obscure website! You’d be exhilarated to see what, if anything, the magazines said about the website when it still existed. What did people write about the site in the still-fairly-nascent Web?

In this light, it’s more reasonable to just leave the [citation needed] tags there until the Hell freezes over, or until someone who actually cares about sourcing facts either proves or disproves the fact. Or, goddess willing, an expert on the subject happens to stumble upon the blatant falsehood that a lot of people have previously been rightfully sceptical about, and removes it.

A cautionary example from among the opposition

If we allow people to delete things just because they might be unsourced, we get this (captured for archival in case folks get antsy). This example is from Conservapedia, a place which encourages thugs to delete things they disagree with on a whim, with zero accountability or room for complaint. Oh, the article has one error in it? Delete the stuff that follows the error. Reduce the article to “Internet Explorer is a web browser.”

Certainly an uncontroversial article in this form. In Conservapedia, you cannot contradict administrators – and this was an administrator’s decision. (As of writing, the article redirects to “web browser”, because the admin in question says the article isn’t maintained. Of course it isn’t – that would contradict the orders from the administration…)

You hear a lot of babble about how the administrators in Wikipedia are evil, but they’re still normal users. In Wikipedia, the decisionmaking is up to the regular users; because the policy is formed by the user base as a whole, everyone can shape its development.

And because everyone is responsible for how the policy shapes, we have to remember that there’s a small threat looming in the horizon: Everyone can become a deletionist.

I’m not really concerned about threats from the Wikipedia administration, as admins actually have to display some accountability and there are actual rules governing them. I’m concerned about the threat from below: misguided regular users who think things should work in some bizarre way.

Rational people should take over

Which is why I’m asking the people to be informed. The people who currently shape the site’s development have this ridiculous deletionism in their heads. Take the stuff over. Start bringing sources to articles that don’t have them. Contest pointless deletions. Kill the time limits.

Why are people deleting articles? “Because the policy says so” isn’t a valid answer. “Because this article is demonstrably irredeemable” is a better answer.

And this is what I think people should basically do: If any given subject is, in any way, “redeemable” — as in there somehow exists a way for people to somehow see the facts are straight somehow — then people should just leave the articles alone. The only situation where people really, really need to get sources to the article as soon as possible is in biographical articles of living people; as it has been demonstrated, people get surprisingly annoyed if the facts are wrong in biographies. But for other topics, there just isn’t a burning need to get the sources to the articles, as long as there’s a reasonable expectation the sources will come eventually.

Why deletionism is futile

It’s a shame that the deletionists won momentarily. But it’s also a good thing to keep in mind that they don’t know that they’re fighting a futile war. Deletionism cannot be codified into policy without everyone smelling a rat.

Ultimately, we must realise that many of the Wikipedia processes have outlived their usefulness.

Wikipedia has its own article deletion process. A casual observer might notice a weird disparity: despite of the deletion process and hundreds of articles that get put on Articles for Deletion every day (and, hey, many articles still survive that ordeal just fine), there’s 3.7 million articles in Wikipedia. I’m betting here that more than enough completely and utterly offensive articles (in deletionist’s minds) are seeping through the process – and if deletionists are saying “don’t worry, we’ll get around to dealing with offensive articles one day”, well, it just may take some time.

That is the mental image I’ve had a long time. AfD is a tiny bureau that pretends it manages a huge country.

I’m sure everyone is well aware of what happens when tiny bureaus manage huge countries. We find people who have deluded themselves into thinking they’re somehow important. It leads to boundless absurdity, which is only hilarious to people who have decided to stay firmly outside of the borders of that country. It’s always heartbreaking when I remember the fact that to the unfortunate souls who still reside in the country, it’s anything but hilarious.

It’s an inefficient process. Why people keep referring to it, or pretend it’s constructive, I just don’t know.

I don’t have the data to back this up, but my gut feeling tells that majority of Wikipedia’s deletions do not come from Articles for Deletion. Wikipedia has two other deletion processes: the proposed deletion process for things that should be mostly uncontroversial, and speedy deletion process. I’ve come to the conclusion that the proposed deletion process was a good idea at the time, but nowadays, there’s just too many unwatched and neglected pages for it to work properly any more: All it does is that it allows people to go to an obscure topic, say “I’ve never heard of this thing, not that I’d be expert on the subject or anything”, and off it goes, because no one’s watching. It relies too much on honesty and watchfulness.

Speedy deletion is a necessary process, however, because it tackles things that are more or less uncontroversial: Spam, silly vandalism from bored schoolkids, copyright violations, unused fair-use images and utterly broken things should get removed. There are also technical hurdles: edit histories of articles cannot be merged without temporary deletions, so that’s a job for admins. You could start questioning some of the finer points: CSD A7 requires that new biography articles should have some references on it. It’s a more difficult call if the persons in question seem to have accomplished something, but really, if there were no such deletion criteria, we’d have about five billion articles on the variation of “Billy Joe Random is on the second grade of Backwater School, and also the king of the world”. We spend a lot of time thinking of the controversial cases, but we must always remember that there are cases that fall far beneath the fine line.

There’s also the Oversight process, which they’re apparently renaming to a far more appropriate name, “Suppression”. That has a good ring to it; they’re just calling it like it is, so that people who wield that power will remember that this stuff is not done lightly. Certain information, like private personal data, obviously shouldn’t be on Wikipedia and it has to be removed permanently. To use it to remove other things would be completely unethical.

A technical and legal concern: what the deletions actually do?

…Which leads to a few inconvenient conclusions.

A big problem with deletions is that it’s used for a lot of things. Technically speaking, deletion hides revisions from non-sysops.

Wikimedia staff, however, has stated that this “hiding” is only a convenience for administrators, and there’s no guarantee that deleted revisions will be retained in perpetuity. (Apologies if this has changed since then; I don’t know if there’s an official Foundation position on this matter, all I know that this was a statement from the Wikimedia technical staff years ago, and finding a link to that is challenging enough…)

This wouldn’t be a problem if people truly wanted the deleted revisions gone.

The thing is, we’re increasingly turning toward the situation where potentially useful information is chucked on the wayside. Deletion processes don’t actually seek to destroy the pages; they just want to reject things that aren’t satisfactory. This is my primary technical opposition to deletions: it’s not always used to mark things that are deleted in the sense that no one wants them back. It’s used to signal that these pages need more work (sometimes, admittedly, a lot of work) before they can be admitted on Wikipedia or any other Wikimedia project. And they unfortunately do that by making the pages harder to access.

There’s a whole lot of transwikiing going on that does not export the entire histories of articles. People on other wikis may be copying Wikipedia articles without histories, referring to articles on Wikipedia, and Wikipedia community suddenly decides “OK, this is a specialist topic, and it looks like the other wiki has a copy of the article – let’s delete it here.”

This creates a copyright problem. Sure, you can blame the badly transwikiing people for creating the problem, but they had every expectation that the article would be retained in Wikipedia’s publicly accessible article histories in some form.

There’s also the problem with people who use deleted articles as sources for quotes. Yes, yes, you shouldn’t cite Wikipedia, but this doesn’t stop people from doing that – journalism just isn’t what it used to be.

This is why, in my opinion, Wikipedia has to reach either of these decisions: Either the deleted revisions are given the same precedence as the extant revisions what comes to retaining them for posterity, or we quit deleting things that people refer to when building articles in Wikipedia or other places. The first option would guarantee that if people request seeing full histories of articles, even if they are deleted, they will always be available. The second option would guarantee that such revisions are always available anyway, without administrator intervention.

Because we always have the Suppression feature for stuff that absolutely must be gone.

Or perhaps we need a whole new host of technical features to allow us to rethink the whole problem of unsatisfactory material. If only we weren’t so clung to this whole hack of deleting pages that may have potentially useful material to build pages upon, we might start considering that sort of processes.

What is news? Where should the material go?

Another topic that gets frequently tossed around are the “news” articles. Wikipedia already has good policies for that: We acknowledge articles shouldn’t cover news for news’ sake, and we acknowledge that not every topic needs a distinct article. Conversely, things that are in news can and should be mentioned, if that benefits the articles as whole. Looking at the news sources and documenting the history is very important – but it should be just that, documenting history, not drawing original conclusions or politicising.

The blog post I’m commenting on referred to the deletion debate of the article on death of Mark Duggan; it was subject of news articles, and as such the facts should be verifiable and we can say the subject is obviously notable enough for inclusion. The only question we then have is whether we should mention it on the article on 2011 England riots (or whether the riots themselves should be covered elsewhere…?) or does it warrant an article of its own.

You may think the deletion debate on Mark Duggan was absurd. It was.

But I ask you to reflect one sobering thought (har!). Wikipedia used to have an article titled “Mel Gibson DUI incident”.

It was banal.

People went through three frigging AfD discussions to get rid of it – once with no consensus, twice with Keep. It was nominated twice as a Good Article, and I just don’t want to even think of looking up how the GA discussions went. Someone, somewhere though “Sure, an actor doing dumb shit is just what warrants all possible efforts for the article to reach the vaunted Featured Article status.” But since it didn’t pass, I’m going to guess that GA process probably still has people who think things through.

AfD discussions were overwhelming. The article was kept.

Nowadays, it’s a redirect to a rather modest-sized section in the Mel Gibson article.

Is Mel Gibson’s drunken behaviour worth articles of their own? Let’s be accommodating and say that a promient brainfart like this requires a mention in an article. It’s in the news, apparently, so why not? But is it worth an article of its own? I’m going to suggest “probably not”, but that’s just me.

But it went through the Process three frigging times and came through scott free.

And afterwards, rational people just went to the talk page, did an improptu RFC on whether to redirect the thing, and went ahead with it when there was ample evidence that many people support the idea.

…You know, got a consensus on the matter.

Outside of the AfD.

It’s fairly obvious why people should stay out of AfD to reach any kinds of sane consensuses.

Deletion culture

When I rambled about Wikipedia’s deletions in the blog post comments, another user chimed in about the deletion discussions.

This reminded me of just how absurd and inefficient the Articles for Deletion process is.

The user commented on Wikipedia cliques who come to torpedo the discussions. In the past, I’ve been a little bit apprehensive of the subject because Wikipedia is actually quite transparent; people always speak of these shadowy cliques, but the fact that people rarely name the names and provide any evidence is somewhat puzzling.

But now that it was mentioned, I realised one thing: It doesn’t matter if there are cliques or not. It’s the environment that potentially fosters these cliques that is the problem. I personally couldn’t care less who bullying who; I’m more concerned about the possibilities for bullying that these processes generate. People always abuse the processes.

I’m not pointing fingers at anyone. I’m saying AfD is the maker of monsters.

We need to get rid of the bureaucracy and bureaucratic processes. All of the abuse against constructive editors I’ve seen in AfD is always tied to the processes. Without the protection from the policies, these cliques would have no power.

Wikipedia’s history has many projects that that had to be struck down because they were too bureaucratic. Esperanza was a good-natured effort to help people, but its downfall was the bureaucratic organisation. Another famous effort was Association of Members’ Advocates, which I don’t remember much about, except it was apparently a place for people to wikilawyer around. Sounds harmless – except it was shut down, so it may not have been harmless. Hmm…

The AfD, at least when I was active on it years ago, was plagued with confusion and contradictory ideas. It mutated into a process for determining whether or not articles are notable enough or not. The process used to be called “votes for deletion”, and people tried to use the term “democratic” to describe the process.

Of course, we had to explain people that the process isn’t democratic. It’s not about votes. Wikipedia has special templates to remind people that the deletion process isn’t a ballot. The deletion process retained its voting process: Someone proposes something, there’s a period of gathering people’s opinions, a decision is made. It retained it’s outward appearance of voting process: People list their general opinions.

Yes, it’s supposed to be about consensus, not headcounts. But isn’t it time we made this whole thing a little bit less confusing? Isn’t it time we just started acting on the agreed upon rules, and quit questioning, time after time, whether those agreed-upon rules actually apply to the articles?

In the Mark Duggan case, people babbled endlessly about the meaning of notability… instead of working on actual solutions to the problem that would work in their image. People love to discuss, but they don’t want to work on solutions.

We shouldn’t require people to come out and say their opinion on the matter, because an opinion alone doesn’t help construct an encyclopedia. We just need the consensus to fix the articles; we don’t need a consensus to be formed in an overly bureaucratic and newbie-crushing process like AfD.

Without AfD, life would continue just as it did, just unhampered by some of the common sources of grief: If you have an opinion that a full-blown article is not helping people, go ahead and put it on a talk page and argue about it there. If you want to change anything, we should start requiring that people are actually willing to implement the solutions they propose, and demonstrate that their realised vision is much better than the solutions others are merely speculating about.

I’m betting people’s willingness to destroy would be significantly hampered if destruction required an effort to construct.

Do people care about the deletion process any more?

And here’s another question: Do people actually want to use the AfD process any more?

It’s obvious Wikipedia is still growing. It’s obvious the policies still stand. But do people really want the AfD process?

The deletionists want a more complicated process. They want the AfD to stay as it is. They want to drive away the clueful people from AfD process so they can get consensuses they want. They want the complex notability criteria, criteria that rely on quantifiable significance.

But the people don’t want a complex notability criteria. They just want to create articles.

The complex, wikilawyerable notability criteria don’t exist anymore. Our policies were horrible enough as they were, and we don’t want things to be quantifiable. We want sources. If we have the sources, all we can do is nod and keep going. Can’t stand in the way of the progress.

The deletionists want hot air. They want an environment where people don’t fix articles, but spend time debating about how to fix the articles. They want the bureaucracy to stay.

The deletionists want the ability to end discussions on technicalities. They want to be able to say “you may have the required news sources, but we don’t care about that, because arbitrary criterion X isn’t fulfilled”.

When they notice that the community has long since rejected the notion of quantifiable significance, they resort to a rhetorical assault. Result? Long boring deletion debates. Lots of time wasted.

Sane people would just admit that while they have some good reliable sources for their facts, the incident might not be worth covering in a separate article. Just mention it somewhere. Make articles accessible. Wikipedia doesn’t need a gigantic article on Mel Gibson’s drunk driving. But if Wikipedians absolutely have to mention that Mel Gibson has problems with drunk driving, do mention it – but don’t make it the big point. Despite his faults, I’m pretty sure Mel Gibson has accomplished a bit more in his life than a bit of weird driving.

Final words

I’m not having as much fun blogging about Wikipedia as I once did – especially since what I said isn’t exactly that new. I’ve spent far too much time thinking of Wikipedia’s policies than anyone could possibly cope with. There’s a lot of sound, rational philosophy in Wikipedia’s policies, but the concepts must be kept simple and unambiguous, and not turned into gigantic monsters like what AfD became.

Here we are, here we stay – I’d rather have spent 4500 words on some other topics. But I’m quite happy that I got to say what I was going to say. Now to the other topics…