User:Teratornis/Outplacement

Proposal: start a Wikipedia outplacement service for deleted articles

The deletion problem

Lolcatspeak may not convey the serious nature of Wikipedia editing, but it's pretty darned funny.

Wikipedia deletes up to several pages per minute. Some fraction of these pages were created by new users who were probably unaware of Wikipedia's complex policies, as well as unaware of how many pages Wikipedia deletes.

From my experience advising disgruntled editors on the Help desk, I'm becoming more convinced that Wikipedia isn't doing enough to account for its honey trap nature:

Wikipedia attracts millions of visitors, many who have never heard of wikis before.
Many people actually seem to think "wiki" is a synonym for "Wikipedia."
Wikipedia makes it so easy for anyone to edit that for many people, this is the first wiki they try editing on.
Many users who are new to Wikipedia have previous experience with other kinds of user-editable Web sites that work very differently than Wikipedia. On sites such as MySpace, Google Groups, blogs, and Internet forum sites, a user's postings are more or less permanent. That is, there is a concept of ownership of the material we contribute to such sites, and other people generally cannot mess with our stuff. People who have been conditioned for years by using such sites may initially assume Wikipedia works similarly.
However, what isn't initially apparent is just how demanding Wikipedia's editorial standards are. The result is that Wikipedia lures a steady stream of new editors, some fraction of whom may have very little chance of ever agreeing with Wikipedia's goal to build an encyclopedia. Even those who do contribute some encyclopedic content may still want other outlets for their non-encyclopedic expressive urges.

Article outplacement: a possible solution

Now, Wikipedia is not a Web directory nor a mere promoter of other people's sites. This principle is valid in general, but if applied in such a way as to make it harder for new editors to become aware of alternative outlets for their non-encyclopedic edits, may work against Wikipedia's primary goal of building an encyclopedia. Specifically, I'm thinking edit wars and deletion debates may become unnecessarily heated when authors see someone threaten to delete their work altogether. I think it is often much less threatening merely to see one's work moved to another wiki.

By systematically identifying and cataloguing enough of these alternative outlets, such that we build up a comprehensive outplacement directory (that is, a directory of alternative outlets that collectively would accept most of the material Wikipedia currently deletes), I believe we can minimize the waste of editor time that currently goes into unnecessarily heated deletion debates. By freeing up everyone's energy to focus on the primary goal (which is to build an encyclopedia, rather than have fights about it), such a directory would more than justify its seeming violation of the "mere" Web directory principle.

Most people are not naturally NPOV in their thinking. They may find it easier to restrict themselves to NPOV on Wikipedia if they concurrently indulge their POV urges on other wikis that share their biases.

In any case, such an outplacement directory would not need to be an exhaustive list of wikis such as the one in WikiIndex. Instead, we would only need to list the wikis which had actually been found useful as outlets for Wikipedia's article rejects, arranged by topic.

In a few cases, groups of Wikipedia editors who shared a topical interest set up their own wikis to publish the non-encyclopedic content they had a tendency to write on Wikipedia, alongside the encyclopedic content. Wookiepedia is an example; many of its editors contribute to Wikipedia, and their edits on Wikipedia are better because they have an outlet for their non-encyclopedic work. That's an excellent model for calming nerves and making Wikipedia more productive.

I would go so far as to suggest that many if not most WikiProjects should identify, and promote, alternative outlets for non-encyclopedic content in their topical areas, as a way to help their members have a more productive and enjoyable Wikipedia editing experience. For example, in Wikipedia:WikiProject Cycling, I suggested using Bicycling Wiki as such an outlet, because most of the people who edit cycling-related articles on Wikipedia also want to share lots of information about cycling which is unlikely to be encyclopedic (such as information about the local cycling they actually do, most of which is non-notable, but ironically is far more relevant than reading about the elite Lance Armstrongs of the world). That soon led to a productive resolution of one article deletion debate. It was easy to build consensus when we had a way to divide the encyclopedic from non-encyclopedic content, and send each to its appropriate place.

Simply deleting people's hard work seems to contradict Wikipedia's long-standing policy of allowing anyone to edit. If we invite everyone to edit, then I think we owe it to them to make a good-faith effort to find welcoming homes for the articles we encouraged them to start, if we decide their articles aren't encyclopedic enough. With so many newbies having little or no idea that other wikis exist, we should at least make them aware of that fact, so they choose to edit on Wikipedia because they share Wikipedia's goal, not merely because it's the only wiki they know about.

Another take on the problem

This is another take on the problem, which I will integrate with the above presentation:

Wikipedia's deletion procedures may seem to the bewildered novice like an expression of bureaucratic indifference, because we cite all the rules and regulations only after a new user wasted hours editing an article we don't want. Wikipedia by design makes it very easy to create new articles, without first requiring the novice to know anything about all the rules and regulations we will hit him over the head with after the fact. If Wikipedia were some sort of an exclusive private club, or perhaps an elite unit of the armed forces, this sort of quasi-hazing ritual we put many novices through might be desirable. But Wikipedia is not like a private club; Wikipedia is a tax-exempt charity, and I'd like to think charities should operate differently than, say, fraternities or the Navy SEALs. (If an organization gets a tax break, I think it should try to serve the public rather than punking people.) We inadvertently mislead large numbers of new users, wasting their time, and subjecting them to unnecessary stress, because thousands of novices look at Wikipedia and get the wrong idea. Why not try to present ourselves to novices in a way that increases the percentage who get the right idea? After all, we are trying to build an encyclopedia, and if the encyclopedia is being systematically misunderstood, that would seem logically inconsistent with the purpose of an encyclopedia, which is to explain things. If we can't explain Wikipedia to new users before they create new articles, then what can we explain? Wikipedia doesn't need to list thousands of public wikis, because WikiIndex already does. I think we would fulfill our charitable obligations (justifying our tax break) merely by insuring that before a new user creates a new article here, he or she has a reasonable opportunity to become aware of the following facts:

Many new users form incorrect impressions about the kinds of articles suitable for Wikipedia, with the result that Wikipedia deletes up to several articles per minute.
Wikipedia is not the only wiki - there are thousands, covering a wide range of interests, accepting many articles that Wikipedia would reject.
Before creating a new article on Wikipedia, an editor should first determine whether the article is suitable for Wikipedia, or whether another wiki would be more suitable for it.

The above three sentences are simple, but many new users might take months to learn this information the hard way. It's evidently far easier for novices to learn how to create a new article - because we purposely made it easier.

Stanton Grant

I have some questions about the Stanton Grant described here:

Press release: Wikipedia to become more user-friendly for new volunteer writers (December 3, 2008)
foundation:Stanton Grant Q&A

For example:

Who are the new contributors you are hoping to attract?

We are hoping to attract new contributors who are just as smart and knowledgeable as the people who have always written for Wikipedia and its sister projects, but who -to date- have been unable or reluctant to participate because of the barriers posed by the interface. There are countless individuals who read Wikipedia and would be great writers/editors, but are daunted by complex wiki syntax. They may not even realize that they can edit Wikipedia. They are the people we are targeting with this project.

My critique: I have serious issues with some of the assumptions that appear to underlie the above statement. These assumptions appear to be:

There are people who are just as smart as the people who wrote Wikipedia, but they are unable (or unwilling) to learn wikitext markup.
If we could relieve these people of the cognitive burden of having to learn wikitext markup, they would then feel motivated to contribute to Wikipedia.
Their contributions would be of sufficiently high quality to justify an expensive project aimed at making Wikipedia's software interface palatable to them.

Breaking the assumptions down further:

Learning wikitext markup is so difficult for some "smart" people that they are "unable" to learn it.

This seems difficult to believe, but I suppose it is possible. After all, some cognitive skills (such as musical ability) can be mysteriously absent in some otherwise bright people. However, consider that wikitext is simple enough for middle school children to learn (I heard that Wikipedia had one administrator who was 13 years old). These are people who the government classifies as having brains that are insufficiently developed to operate motor vehicles, to vote for political candidates, and to do many other adult activities. And yet large numbers of them can learn wikitext editing with no help beyond coming to Wikipedia of their own initiative and reading the friendly manuals. I find it very difficult to think of an adult as being "smart" if he or she cannot do what a bright 12-year-old can do. Certainly, the burden of proof is on whoever makes the counterintuitive claim that bright adults cannot keep up with unaided schoolchildren.

Wikitext is the simplest markup language I have ever seen. Of course on Wikipedia the markup gets complex because of all the templates and so on that people use to extend the syntax, but the basic markup is stunningly simple, compared to other powerful markup languages like LaTeX and DocBook. Learning wikitext markup is straightforward for anyone who can read and follow instructions. It's no more difficult than preparing an elaborate dinner by following recipes from a cookbook.

What definition of "smart" are the project leaders using here? Is this a version of "smart" that is in some sense quantifiable? The best available quantitative measure of intelligence is IQ. IQ score correlates with a person's ability to manage complexity, to learn, and to make sense of new and unfamiliar situations. The project leaders should study a representative group of accomplished Wikipedia editors, to determine whether success on Wikipedia correlates with IQ (I strongly suspect it does). If it does, then the project leaders should clarify exactly what they are trying to accomplish, which would have to be one or both of the following:

Make Wikipedia easier to use by people who have IQs comparable to Wikipedia's successful users, but really cannot learn wikitext markup.

Obviously, before attempting to do this, the project leaders should first identify a sample group of such people. I strongly suspect such people are so rare as to hardly be worth the trouble of trying to accommodate - indeed, we don't know in advance whether Wikipedia can accommodate them with any available technology.

Make Wikipedia easier to use by people who have IQs substantially lower on average than Wikipedia's successful users.

If it turns out that having a high IQ is the principle determinant of successful editing on Wikipedia, this goal would be like trying to squeeze blood from a stone. It would join the long list of failures in the history of educational reform.

Considering IQ may not be politically correct, but if success on Wikipedia does in fact correlate strongly with IQ, the project leaders will ignore this at their peril.

Imagine if someone attempted to turn every person into a professional athlete in one of the major sports. We all know that would be silly, because most people simply lack the ability to excel in sports, regardless of how hard they might train, and furthermore many people simply do not like to exercise. It's abundantly clear that some people are made for sports, and some people aren't. If you want your team to win, you do it by recruiting the best athletes. You don't win by trying to make your team accessible to the scrubs. There is not much difficulty in identifying the athletes. You just set up competitions, and the athletically inclined youngsters will gravitate to them on their own initiative.

Wikipedia may already be an efficient mechanism for recruiting the kinds of people who can succeed on Wikipedia. More on this below.

There are some people who are unwilling to learn wikitext markup, but they would contribute to Wikipedia if the mechanics of editing were simpler for them.

(Stopping here for now. The main points to make are whether the mechanics of putting text onto Wikipedia represents the main hurdle. I think the cognitive burden of learning Wikipedia's culture, policies, and guidelines is at least an order of magnitude greater, and the fact that all this extra stuff exists is even less apparent to new users than how start editing a page. As evidence, consider the staggering death toll of new articles by new users, or even by relatively experienced users. There are already thousands of Wikipedia users who can figure out how to create new articles and mark them up, but have no inkling of the stupefyingly complex rules for allowable content. I think that is by far a more serious problem than worrying about how to recruit still more victims for the deletion-fest.)

Also point out how Wikipedia is a top-five Web site. The four sites ahead of us: Google, Microsoft, Yahoo!, and AOL. Are any of the other top sites easily editable by people who are not willing to make a major life commitment? Before you can put your stamp on any of the top four sites, you have to apply for a job, get interviewed, get hired, probably relocate, go through some kind of training and orientation, and learn about the corporate culture. There is little or no opportunity for incompetent or unmotivated people on these other top sites. More than likely, for a site to get into the top five, it must have some mechanism for weeding out people who are incompetent or unmotivated.

The project leaders should identify some collaborative Web sites that are usable by people who cannot use Wikipedia, and see how successful those sites are. If no one else knows how to give such people a free hand and get content from them that creates a top-five site, maybe it just isn't possible with current technology. At the very least, the project leaders should know whether they are trying to do something that no one has done before.

Random scribbles: also mention Wikipedia's size and growth. Wikipedia is almost unimaginably huge, and continuing to grow fast. How many more articles and users do we need? Has the current growth trend shown any sign of leveling off? Is the goal to reach, say, 10 million articles on the English Wikipedia? If nothing changes, is there any doubt that we are going to get there soon enough? Do we want to try to go faster, and if so, why? Is being the fifth most popular Web site not good enough? Is Wikipedia having any noticeable difficulty in attracting quality contributors in any topic area?

Can we identify some contributions these "smart" but "unwilling" or "unable" people would make which cannot be made by anyone who can learn wikitext markup? Even if there are some smart people who can't get past Wikipedia's interface, is there anything to suggest we will miss them? What is there that Wikipedia's current and growing population of editors cannot write? If we have, say, 500 people who can write competently about a given topic, do we really need another 500 people to write on the same topic? Anything the 1000 can do, the 500 can presumably do in twice the time. Since there is no deadline on Wikipedia, why do we care whether it takes N years, or 2*N?

Of course there are some project goals I like, such as tools to make footnotes and table editing easier. For example, we could use a database of reliable sources, which would allow one person to enter a source, and thereafter everyone else could cite it in any other article without having to slog through editing those tedious citation templates again. As a side benefit, we could easily see how Wikipedia cites a given source (and the primary authors could also see this). It might be nice to optionally break out the citation markup from the rest of the markup. A WYSIWYG table editor would also be nice, and not just for the people who don't want to learn wikitext. There are many types of edits which are easier to make in a visual format. It's nice to be able to edit from both the visual side and the markup side, since different tasks work better in one than the other.

Of course everything depends on the skill of implementation. A WYSIWYG table editor might be great, or it might suck, depending on who builds it and how good they are at coding.

I would like to see references moved into their own database or namespace, with automatic archiving of the original Web page if one exists (to prevent link rot). That way, every time someone adds a reference to Wikipedia, it would become easily reusable in other articles without everybody having to edit those clumsy citation templates over and over. It would be nice if Wikipedia was smart enough to automatically extract the essential reference field entries from, say, pages on at least the 500 most-cited Web sites (author, date, title, URL, etc.), so you could generate a one-click reference for whatever Web page you're looking at. Currently it takes at least half a dozen copies and pastes to get all the citation template field data from the source document browser tab into the Wikipedia brower tab.

To-do

Some things to do with this page:

Look up all my Help desk contributions relating to deletion, and coherently summarize the various ideas I have written.

Partial list (add to this):

Wikipedia:Help desk/Archives/2007 September 4#Page deleted without Consultation (permanent link)