I have often written that Wikipedia proves that a million monkeys typing for a million years still produce gibberish. Of course, that is a bit of a harsh assessment, but it is not that far off the mark. Wikipedia, despite all the high sounding ideals behind it, shows that consensus is not the best route to truth.

Before I wrote about how the administrative model adopted by Wikipedia leads to problems, but that will not be my topic today. Instead, I am going to look at Wikipedia at its best. I am going to grant Wikipedia all of its assumptions. I will grant it everything it alleges, and show that it still falls far short of a traditional encyclopedia. After that I may also look at why a few of these assumptions are unfounded, but my main goal is to show that, even when everything works right, Wikipedia is just a bad idea.

Now the basic assumption behind Wikipedia is that the majority is engaged in an effort to produce an accurate report. The theory does allow for the occasional stray malcontent or vandal, but the overall theory requires that the majority are sincere. Given these assumptions, the model works as follows. First, an interested individual posts an article. Another individual reads the article and either spots an error and corrects it, or else decides it needs additional information and expands it. This continues in an ongoing process and the article slowly gets closer and closer to the truth.

However, even assuming we have nothing but sincere truth seekers working on the article, there is one very big problem with this model, the assumption that all involved can agree on what is true. Or agree on what is settled and what is unsettled. This is obviously most prevalent in contentious areas, but it happens everywhere. Even in subjects as dry as mathematics or philology. Allow em to give an example.

Let us suppose we have an article on the theories of John Maynard Keynes. Let us further suppose that someone posts the statement "Henry Hazlitt proved these theories false." To the poster, this is a fact. However, to a dedicated Keynesian, this is not. So he changes it to read "Henry Hazlitt claimed to have proved these theories false." The original poster returns, and arguing that the proofs have not been successfully challenged,  reverts to his original sentence. And thus begins one of those "reversion wars" that plague Wikipedia. Both sides claiming that their position is true, and both marshaling seemingly valid arguments.

Another problem comes in the form of questions of relevance. For example, is it relevant to an article on the World Trade Center to state "some allege that George W Bush had them blown up." Some would argue that these allegations are important, as they provide an explanation for the collapse. Others would argue that, as they are fringe beliefs not supported by evidence,t hey are clearly not relevant. The second group could also argue, though they rarely do, if we allow every fringe belief to have a single sentence, the storage requirements would exceed all available storage on the planet, as the conspiracy theories surrounding the twin towers alone run into the thousands of pages.

And both of those show a problem inherent in Wikipedia. It wants to be, in Wikipedia jargon, NPOV. That is "neutral point of view", adopting no position but providing information on all possible perspectives. It sounds noble and the Wikipedia supporters argue that it is essential, as they don't want to exclude any aspect of the topic, but in reality it simply leads to chaos. Without an overarching set of beliefs, it is impossible to establish what is truth and what is opinion, as well as establish criteria for relevance.

Let us take an extreme example. Suppose we have a historian who takes one specific reading of Hegel very seriously, believing that history is simply the materialization of the various national spirits, and that individuals have no relevance, being nothing but tools for expressing these national spirits. In his mind, mention of the role of any individual in events is irrelevant, and he would edit in accord with his philosophy. The problem is, how can we argue against him while maintaining NPOV? The only possible counter argument is that his philosophy is incorrect, but to say that we must propose an alternate philosophy on which to base that claim.

And that is the truth of the NPOV claim. There is not really a neutral point of view, but a smuggled point of view, a Wikipedia consensus which informs all the posts. But where a traditional encyclopedia would ahve an explicit editorial philosophy, Wikipedia instead has a silent, unacknowledged perspective. And, so long as readers and posters accept this view, that is fine. However, should one try to post in opposition tot his consensus, he will find himself suddenly accused of forcing his point of view into the proceedings. It is, in essence, a very hypocritical position, this smuggled POV. The editors and posters are just as opinionated as those they chastise, but they do not admit it.

In addition, this smuggled point of view raises one other issue, is an unspoken consensus the best way to establish truth? As I said, posts are generally evaluated against this unrecognized consensus and judged accordingly, but is that how truth is established? Is the consensus the best measure of accuracy? I would argue that it is not. Conventional wisdom can be wrong as often as it is right. For example, a vast majority since the 1970's have believed that oil companies earn huge profits. The truth is that earnings in the big oil companies are actually average or slightly lower in comparison to capital investment. And if conventional wisdom on one matter can be wrong for almost 40 years, why would the silent consensus of Wikipedia be any more correct?

And it is not just a consensus, but a clearly political one. As with most trendy internet phenomena, Wikipedia seems to follow the political perspective I first identified on Slashdot as "left-libertarian". It is a contradictory view that distrusts all things governmental yet still adopts a number of views on the political left. It is largely anti-military, anti-war, anti-big business, yet anti-government. It believes in AGW, believes extinction is a massive problem, and is open to many conspiracy theories. It has a vague belief in government helping the "poor" though it opposes most specific applications. It thinks most people are "sheep", racist, falsely patriotic and easily led. In other words, it is your traditional arrogant liberal viewpoint trying to subscribe to the GPL-type of pseudo-libertarianism. Not all Wikipedia editors fall into this category, but enough do that this political position informs many Wikipedia articles.

And those are just the problems that occur when we grant all of Wikipedia's assumptions. In reality, many are not as ready to allow corrections as Wikipedia's theory would assume. Many epople have a strong viewpoint they will restore over and over again, not matter what arguments are raised against them. The result of this is that those with the most time, and who feel strongly enough to dedicate themselves to controlling their Wikipedia article, will end up winning. The article will not be determined by the best argument, but by the person most committed to seeing their viewpoint in print.

Now, I am not saying Wikipedia is worthless. It is fine for finding bits of trivia, birth dates, and some very basic information, but that is it. I would say it is valuable for researching something which is not in dispute, but the problem is that many of us are not aware of what constitutes a controversial issue in mathematics or Slavic history, so we could think we are reading an unbiased article only to discover we inadvertently just heard one side of an ongoing debate. Given that, I have to say, the utility of Wikipedia is very, very limited, and for the foreseeable future, I am sticking to information that has been vetted by an editorial board with an explicit policy. Even a highly politicized journal is more useful, as at least the bias is obvious for all to see, and I can correct the information for bias. Wikipedia provides no such check, pretending to a neutrality it fails to achieve in practice.

POSTSCRIPT

There are two matters that did not fit in the article above, but I feel the need to mention.

First, there is an additional problem with Wikipedia. Even if we assume Wikipedia were perfect and corrected for errors, the truth is that anyone can edit it. So, when you do your research, who is to say a vandal has not just changed something, and the information you are receiving is not false? Even fi the editors were perfect and the end result was a fully accurate, unbiased encyclopedia, at any specific moment the information is completely unreliable, and thus worthless.

Second, I know some of what I wrote sounds like a bitter poster, but that is not the case. I have made a few trivial corrections in Wikipedia. I corrected a typo in the article on porphyry, removed a badly worded sentence in the article on Maimonides, removed a strange and apparently irrelevant paragraph from the article on sticky toffee pudding, and removed an irrelevant sentence in the article on dachas. I may have corrected one or two other minor issues. As far as I know, my changes are still there. I am not a Wikipedia contributor, embittered or otherwise. My criticism is based solely on my observations about the shortcomings of the model upon which Wikipedia is based.

NOTE

I have written all these things elsewhere in the past. However, most were either in comments about Townhall columnists, on GunnyG's blog, or on Slashdot several years ago. As I had never put it together in a single location, I decided to write this essay.