I recently noticed Wikipedia starting each page with an appeal from some person or other, asking me to contribute to Wikipedia. Every time I check on the ingredients of garam masala or check the episode listings of  the original Dr. Who (two of the few things for which Wikipedia is useful), I am forced to look at a pleading photo from "a contributor of 18,000 edits" or "a Wikipedia programmer" or , worst of all, the smug and self-important face of Wikipedia's "founder" (who neither originated the concept nor wrote the original code). Besides bringing back horrible memories of my pre-cable days when I had to watch hours of PBS pleading¹ in order to watch my favorite programs, it also seems kind of pathetic. After all, it isn't as if turning a profit compromises one's reliability. Google manged to become a multibillion dollar company while still providing unbiased search services². So why can't Wikipedia find a way to become self-supporting, without resorting to this unseemly begging?

But that is not the reason for my title. Much as I find begging unseemly from a group which could easily be self-supporting, I would still consider contributing were I convinced of the worthiness of the cause. Unfortunately, Wikipedia, despite their many claims of importance, is far from worthy. It is a topic I have covered many times (as can be seen in the Postscript, where I list all my previous essays on Wikipedia), but I feel that, if I am going to oppose their fundraising efforts so openly, they deserve a concise statement of my objections.

The first objection to Wikipedia is a very simple one, but an objection that Wikipedia boosters seem to fail to recognize. Or, should it be brought to their attention, one which they shrug off as "transient" or "irrelevant", which demonstrates quite well their failure to understand the basic concepts underlying an encyclopedia.

And that problem is nothing more than the fact that Wikipedia can be edited by anyone at any time³. And that means that at a given moment, there is no way to tell if what one reads is the thoughtful product of the collaborative efforts of the best minds in the field, or the carefully crafted prank of a snickering twelve year old. Or, if we want to ignore outright vandalism, there is no way to tell if the current article is the product of a thoughtful back and forth between various viewpoints, or if it was just modified by a strong-willed individual pushing a specific position to the exclusion of all others.

Wikipedia boosters tend to dismiss these worries by arguing that such problems are usually remedied quickly, either by corrective edits or by reversion to the previous copy. Vandalism, even when well done, and not immediately obvious as such, tends to stay for a short time before it is noticed by a reader and removed⁴. And, in essays where there are contentious issues, there tends to be a lot of give and take, so any efforts to use an article to push a particular viewpoint are usually caught very quickly and the essay is returned to a more balanced presentation⁵.

Even if such statements are completely accurate, they show that the responder does not understand the purpose of modern encyclopedias⁶. Yes, pranks may be caught and corrected quickly, and flame wars may be resolved my more moderate editors, but for casual users, that is irrelevant. A casual user is consulting an encyclopedia because he lacks some information and needs to find a quick answer. In other words, he is in no position to assess the quality of the response, or at least cannot speak tot he accuracy of some specific statement due to a lack of information. and so, for the casual user, a prank page, or a biased page, is indistinguishable from a valid page⁷.

And therein lies the problem. Even if prank pages or biased pages were corrected within an hour of posting (which is far from the truth), the fact remains that prank and biased pages are visible for a time. Given the casual user's inability to distinguish a valid from an invalid page, this means any page is suspect. And since it is known there are invalid pages, every page is worthless, as it is impossible to know whether a given page is the product of sincere efforts or an opening salvo in a new flame war. Nor does it help to consult the page several times, as a change in contents leaves the question whether the original pages was valid or the new version. In fact, both could be valid, or at least sincerely written, with changes representing simple differences of opinion. All of which means that at any given moment, any given pages of Wikipedia has absolutely no utility in establishing any given fact, as there is no way to tell whether or not that page is accurate or not⁸.

Of course, that is far from the only problem with Wikipedia, though that single problem makes it absolutely useless as a reference should any sort of certainty be required. Yes, it can be sued for trivial issues, such as the two I mentioned above, and a handful of other uses where, if the information later proves false, the impact is negligible, but for any purpose where a definitive answer is needed, Wikipedia is useless. But the problems do not end there.

First, there is the standard of proof required by Wikipedia. Past encyclopedias, and traditional encyclopedias of today, usually employ expert editorial staff who determine whether the evidence supports a given article. In many cases, especially in encyclopedias of the past, and in specialist encyclopedias of today, many essays are written by prominent researchers who provide new information as well as individual analysis. And in such cases, the editorial staff general defers to the expert, though not always. Still, in general, it is the expertise of the staff that is used to establish the validity of the evidence, or the value of the analysis, with the assumption being that a consortium of experts can determine with some certainty whether or not a given essay either represents the majority position on many issues, or, if it does not, is at least presented in such a way that any assumptions are made clear and the supporting evidence is clearly provided.

Wikipedia forbids anything similar to the traditional model. With its absurd egalitarian model, there are no experts. Oh, they do break with total egalitarianism to allow for editors, but even there, editors are not those with expertise, but those who post the most often. And even then, editors who post on one topic are free to throw their weight around in areas about which they have no knowledge. Given such a model, Wikipedia falls back on two forms of proof, neither of which fills a reader with confidence.

First, they use the argument made by every high school research paper. "If it's in print, it must be true." That's right, if a writer can site a supporting essay, ANY supporting essay, that is "proof". On the other hand, the opinion of the foremost expert in the field, or facts known by everyone which are not documented, neither of those are acceptable. For example, if you can find an essay which says gravity causes objects to repel one another, you can write it. On the other hand, if I say they attract, but fail to cite a source, I am engaging in "Original Research" and my opinion is not allowed. The standard of proof is not established by expertise, or by empirical evidence, it is determined by the volume of pages written in defense of a position. Which, given how much time most conspiracy nuts spend printing out their beliefs, makes me pretty certain it is "proven" Kennedy was killed by CIA Freemasons working for Martian warlords.

Of course, often the printed word supports more than one position. And in those cases, essays often see saw back and forth, as those with extreme positions try to write an essay favoring their position, and those with strong opinions in the opposite direction try to push it in the other, with a handful of moderates trying to balance the two. Since such cases cannot be decided by reference to cited works, the solution is an even more uncertain means of finding the truth. Majority vote. That's right, when Wikipedia encounters any controversy, the solution is majority rule. And this is fine when it is asking whether two essays should be merged, or how a particular title should appear⁹. But when determining whether a given essay should read one way or another, majority vote is a very poor means of determining the truth. This modern form of compurgation is likely to either end up producing a bland compromise, which rarely matches the truth, or, if the moderates become fed up and sit it out, the vote will end up favoring the extreme position with the most followers (that is the largest minority opinion). Neither is going to get anyone anywhere close to the truth.

There are other issues, but I think there two show more than enough reason to avoid funding Wikipedia. It may sound like a good idea to egalitarians, some fringe libertarians and the extreme fringe of the free software movement. ("Copyright as Politics", "Some Libertarian Analogies", ""Best Practices"", "More Examples From Another Field") But a moment's thought would show how flawed the methodology is, and the incredibly small probability it would produce anything of value. And so, though it may upset many of the true believers, I would suggest that anyone thinking of giving money to Wikipedia instead contribute to some other charity, any other, as prolonging the Wikipedia experiment will do little, except perhaps provide one more lesson in why such egalitarian theories should be avoided.

==========================================================================

1. It is interesting to view those fundraising claims in light of the proliferation of cable channels. I recall many times hearing PBS claim that without voluntary donations (and government subsidies) there would be no broadcasts of cooking shows, travel programs, home improvement shows, foreign programming and so on. However, as The Cooking Channel, Food Network, Travel Network, Home and Garden, BBC America, IFC and a host of other channels now turn a profit specializing in just one of those topics each, it shows how little truth there was to that claim. Of course, in the era of heavily limited broadcast bandwidth, and stringent government control of broadcasting (including PSA requirements, the fairness doctrine, and many other limits on content), it probably would have been difficult to create such channels, but in part that was also because such channels would have been in competition with the subsidized government channels. (Not to mention that PBS channels used up several of those very scarce broadcast channels, making the remaining channels even more expensive and difficult to obtain.) I wrote on this before in my essay "The Debt Ceiling".

2. Some may object that Google's handling of China shows this is false. I would contend that instead it shows that Google's management made a bad decision. Google could have easily remained competitive without giving in to Chinese demands for content restrictions. Perhaps not as competitive, but it is questionable how much advertising revenue is generated by having approved Chinese access, versus the covert Chinese use which would have existed had Google refused. There is also the question of goodwill from consumers, and the loss for some time due to the handling of the Chinese question. Regardless, the point is, it is possible to remain profitable, and certainly possible to be simply self-funding, without "selling out" or compromising one's objectivity. (To the degree that is possible.)

3. It does not change my argument that certain known vandals, or those who insist on posting contentious content, can be blocked by their IP, or that often vandalized articles can be locked for a time. Users cannot easily tell when an article is locked, nor can they tell at all when a vandal is excluded. And neither makes the remaining articles any more dependable.

4. Though this does not always work out in real life. As we saw with the hoax article which was picked up and incorporated into an obituary. ("Vindication") Nor is that the only possible problem. As I discussed in a number of essays, there are a number of sites which reproduce content directly from Wikipedia. Should such sites harvest their content while a fraudulent article is being displayed, that article may remain in circulation for some time. In fact, given Wikipedia's "citation as proof" policy ("Wikipedia Absurdity, Or How To Create Your Own Citation"), which we will discuss later in this essay, that fraudulent citation may eventually be reincorporated into Wikipedia itself, bringing back an error which had been purged earlier.

5. As I will discuss later, and as I wrote in several essays, the more common result is that a partisan for a competing theory will post an equally contentious essay in a different direction, and the entry will tip back and forth between extremes with only brief respites when more moderate editors notice the ongoing "flame wars" and merge the competing versions, or, in extreme cases, lock the essay for a time.

6. Historically, encyclopedias were not always general reference works. many older encyclopedias were intended to provide relatively scholarly essays for specialists. Even today many special purpose encyclopedias provide much more scholarly essays. To the degree an essay goes beyond providing basic factual statements the amount of bias increases, by which I mean that more scholarly essays contain many more assumptions about how things work than basic factual statements. However, such scholarly essays normally make most such assumptions explicit (unless a given assumption is common enough in the field to be effectively "understood" by all). Thus, general purpose encyclopedias tend to contain less of a viewpoint than a more scholarly work, but suffers more from what bias it does contain, as it tends to not make such bias explicit.

7. Obviously, very bad pranks and some explicit bias is noticeable by even those with little knowledge. And, similarly, if a biased essay or prank contradicts those facts a user knows to be true, the user will recognize the page as invalid. The problem exists, mostly, when the prank or bias is not so overt, or the user knows very little about the topic, which makes it much more difficult to spot such problems. Of course, even when a user notices a bias, or recognizes certain sections as vandalized, it is impossible to tell if the part in which he is interested is also invalid, as many pages contain bias only in some sections, and vandals often change only certain parts of the page.

8. We are not discussing here factual errors, or improvements in our understanding. I will grant that printed encyclopedias often contained mistakes, as do the best intentioned Wikipedia pages. By "true", in this context, I mean the product of well-intentioned collaboration, using the best available information, with as little bias as possible. In printed encyclopedias, for the most part, every essay fit that description, and remained constantly true through the life of the book. Wikipedia pages cannot say the same.

9. Actually, the splitting or merging of essays and the choice of titles is more important than I suggest here, and probably also should not be left to majority rule. The choice of what topics to cover independently and which to lump together can have a serious impact on the final result. But, in terms of consequences, deciding such things by majority opinion is a lot less damaging than letting the majority decide what is and is not true.

==========================================================================

POSTSCRIPT

As mentioned in the essay, I have written many essays about my objections to Wikipedia. The following list should contain every previous reference:

The Failure of Wikipedia
Final Comment on Wikipedia (For Now, Anyway)
Wikipedia?
Now I know Why
One More Wikipedia Problem
Very Short Digression On Wikipedia
Wikipedia Syndrome
 Wikipedia Absurdities
Stop Confusing Me With The Facts!
Mystery Quotes
Opinion Masquerading as Fact
Funny Numbers
Endangered Species
Have to Love Wikipedia
Sterility of Formal Economics
Wikipedia Absurdity, Or How To Create Your Own Citation
Some Libertarian Analogies
Proof Positive
Why People Don't Take Academics Seriously
Deceiving Themselves?
A Question About Language
Roman Legions, Hopscotch, Killer Gays, "Got AIDS Yet", WMDs and a "Damn Piece of Paper"
Very Short Digression On Wikipedia
One More Wikipedia Problem
The Power of Myth on the Internet
Vindication
Life is Strange
The Tragedy of the Creative Commons
The Taxonomy of Trivia
All Conspiracies Great and Small [not specifically on Wikipedia, but many references to problems]
The Inherent Weakness of Evil [mostly in the footnotes]

It is possible there are some omissions, as it is a topic I have revisited many times. However, the links within the articles themselves should help to correct any such errors.

I would also like to include a few other essays:

Some Thoughts on the Media
The Press Versus The Nation
The Death of Impartial Media
An Adversarial Press?
Reporters' Opinions
The Impossibility of Unbiased Reporting
The Rebirth of Skepticism
Media Double Standards and a Proposed Solution
Contradictory Beliefs and Practices
Are They Admitting It?

These essays discuss the concept of an "impartial media", but have a clear applicability to Wikipedia. Just as all media has some degree of bias, so too the contents of any encyclopedia display the intellectual biases of the contributors. Provided those premises are made clear, at least to the degree of presenting underlying assumptions behind any analysis, it is harmless. But Wikipedia, like the media, tries to remain "impartial" and as such smuggles in various biases without admitting as much, which can be more harmful than an open bias. (The preceding essay discussed this as well.)

POSTSCRIPT II

I discussed the fact that there is some bias in every essay. And I stand by that position. There is no such thing as an unbiased work, at least provided the work goes beyond simple statements of fact such as "My hand is hot" or "That shirt is red". Anything more involved includes within the statements certain assumptions. Many times they are simple assumptions, and assumptions that are shared by almost everyone. For example, the belief in a mechanistic cause and effect. However, encyclopedias usually go beyond simple cause and effect, and discuss issues such as politics, economics, physics and other areas where one's assumptions play a significant role.

In the past, most essayists writing encyclopedias had a clear intellectual bias. Voltaire, Diderot, Erasmus and others had already made clear the foundation of their thinking, so it was not difficult to understand what was at the root of their essays. Likewise, in many specialist encyclopedias of today, especially those focusing on religion or history, the writers are usually well known and their assumptions documented, and so it is easy to compensate for bias.

The only problematic encyclopedias are the modern, more anonymous encyclopedias, where the authors are either not named, or not well know. And encyclopedia publishers understand this. In order to avoid charges of bias, and to appeal to the largest audience, they usually approach a given topic in one of two ways. Either they accept the majority opinion in the field, or, if there is no single majority opinion, they present a survey of the most common positions. In this way they can present what is seen as an unbiased essay, even though it still contains countless unstated assumptions.

I only mention this as Wikipedia struggles constantly to remain "neutral", while failing to recognize that true neutrality is impossible, there will always be some assumptions with which someone will disagree, and the best alternative is simply to make one's assumptions clear, so that the reader can accept or reject them.

POSTSCRIPT III

I found a great quote in my post "Wikipedia Syndrome", explaining precisely what I have been trying to say about the inevitability of bias:

I will probably be shunned by academia for saying so, but knowledge is, at its root, discriminatory. Discriminating between the essential and the inconsequential is the basis of all understanding, so "discrimination" is essential to true knowledge, and impartiality is anathema to real learning.

As that statement should make clear, all knowledge involves some assumptions, some bias, and some discrimination between what one considers important and unimportant. Trying to understand a topic while having NPOV (A "Neutral Point of View") is simply impossible. Until you make assumptions about how things work, you can understand nothing, and so any understanding, of necessity, implies preconceptions and bias.